id stringlengths 3 9 | source stringclasses 1 value | version stringclasses 1 value | text stringlengths 1.54k 298k | added stringdate 1993-11-25 05:05:38 2024-09-20 15:30:25 | created stringdate 1-01-01 00:00:00 2024-07-31 00:00:00 | metadata dict |
|---|---|---|---|---|---|---|
222139403 | pes2o/s2orc | v3-fos-license | Estimating the Remaining Power Generation of Wind Turbines—An Exploratory Study for Main Bearing Failures
: Condition monitoring for wind turbines is tailored to predict failure and aid in making better operation and maintenance (O&M) decisions. Typically the condition monitoring approaches are concerned with predicting the remaining useful lifetime (RUL) of assets or a component. As the time-based measures can be rendered absolute when changing the operational set-point of a wind turbine, we propose an alternative in a power-based condition monitoring framework for wind turbines, i.e., the remaining power generation (RPG) before a main bearing failure. The proposed model utilizes historic wind turbine data, from both run-to-failure and non run-to-failure turbines. Comprised of a recurrent neural network with gated recurrent units, the model is constructed around a censored and uncensored data-based cost function. We infer a Weibull distribution over the RPG, which gives an operator a measure of how certain any given prediction is. As part of the model evaluation, we present the hyper-parameter selection, as well as modeling error in detail, including an analysis of the driving features. During the application on wind turbine main bearing failures, we achieve prediction in the magnitude of 1 to 2 GWh before the failure. When converting to RUL this corresponds to predicting the failure, on average, 81 days beforehand, which is comparable to the state-of-the-art’s 94 days predictive horizon in a similar feature space.
Introduction
Wind energy remains the leader in renewable energy sources with expected continued growth [1]. In order to maintain and monitor the ever-increasing fleet of wind turbines, health prognostics and condition monitoring (CM) techniques are employed, covering abnormality detection, failure mode tracking, and prediction [2][3][4][5][6][7] of failures in rotating parts, in particular, the main bearing. A summary of common CM models may be found in Márquez et al. [8]. With regard to prediction, recent works have shown predictive capabilities beyond 90 days with reasonable accuracy [3]. These approaches have one thing in common: all perform predictions with respect to the remaining useful lifetime (RUL) and life cycles of an asset or component. Although these aid in operation and maintenance (O&M) efforts, they are not directly related to the physical nature of a wind turbine, i.e., power production. In terms of O&M scheduling we will consider the following thought experiment: A predictive model, such as Herp et al. [3,9], Teng et al. [5], or Si et al. [10] predicts a wind turbine to fail in the RUL = 36 days, within a confidence bound. The operator is now faced with the decision when to stop operations and when to plan maintenance tasks. This in itself is a highly complex optimization problem. However, the RUL as such is not related to any ambient or operating conditions, thus, without further information, the best course of action would be to shut down the turbine and thereby extend the turbine's RUL to infinity, or until the socket's or tower's expected lifetime. In this scenario, the operator would lose all revenue generated by that particular turbine.
To circumvent this dilemma and provide a physical measure of the remaining operation of a failing wind turbine, we propose an approach with a focus on the remaining power generation (RPG). We draw inspiration from Teng et al. [5] and Herp et al. [3],both of which employ a neural network (NN) for learning dependencies between samples. While Teng et al. extrapolate and compare the predicted values to a threshold, Herp et al. train towards a distribution and use a recurrent neural network (RNN). We adapt the RUL prediction in Herp et al. and formulate an equivalent model for the RPG. In Section 2 we provide an overview of the data included in this study, including a preliminary investigation of the input feature. In Section 2.1.2 we highlight the difference between RUL and RPG. The proposed model is explained in Section 3. Covering the fundamental assumptions of the model, a justification of the chosen hyper-parameters is given. In addition to the loss off the neural network, we also present an additional performance measure, an aggregated error over the predictive distribution. In order to show the feasibility of the model, we consider a case study of main bearing failure (Section 4). Here we show the predictive capabilities of the model and compare the model outcome to the state-of-the-art prediction in RUL estimations, by mapping between RPG and RUL. As of the time of writing, we are not aware of any literature addressing the remaining power generation instead of remaining useful lifetime, and thus do not provide a comparative study. Finally, we conclude our work in Section 5.
Remaining Power Generation for Main Bearing Failures
This project is concerned with the RPG of main bearing failures and limited to turbines of the same class, i.e., three balded units composed of the same parts and rated at the same power. As such, the amount of data available for each turbine is still beyond direct processing, thus, in order to talk about the aim of this study, we start with an assessment of the availability of data sources and their relevance. Alongside we provide the notation used throughout this work.
Data and Notation
All the data used in this study come from the same class of turbines and is provided by courtesy of Siemens Gamesa Renewable Energy. The data acquired stem from three different databases: (i) meta-data containing a damage indicator for the main bearing. It is obtained by integrating over a frequency range of a fast Fourier transform of bearing vibration measurements to obtain a measure of the noise floor. The damage indicator will be treated as one of the inputs to the proposed model. These data points are event-based and contain initial detection of damage. (ii) SCADA data, sampled at 10 min averages. (iii) Wind turbine events, e.g., warning and error messages. This database is used to distinguish between run-to-failure wind turbines and non run-to-failure turbines. Here we referred to run-to-failure wind turbines, as wind turbines that are operated until failure, while turbines that are stopped by the operator before failure are referred to as non run-to-failure wind turbines. Before data preparation, 112 wind turbines, undergoing a main bearing failure, have been acquired for this study-31 of which are considered run-to-failure.
The SCADA data are chosen based on first principal considerations of the energy diffusion, selecting the same features as Bach-Andersen et al. [11], while the damage indicator is chosen to be the same as used in the study of Herp et al. [3]. After cropping the time-series at the initial damage detection and the stopping of operation at the fault, the two data sources are merged, such that x t ∈ R m is a feature vector containing m features at time t. The merged time-series are re-sampled to daily timestamps. Table 1 shows the initial range and units of the desired features. Although NNs are capable of learning dependencies between input features, we present an additional temperature based feature. This feature will provide a physical interpretation of what drives the prediction in the proposed model. A simple correlation study shows, as one would expect, that the main bearing temperature is highly correlated with other temperature measures. In order to investigate trends and behavior of the main bearing, independent of other temperature sources, we create time-series of the difference between the main bearing temperature and the nacelle, ambient, and gear oil temperature, respectively. The subtraction of the ambient temperature is especially interesting as it can influence the main bearing temperature, while vice verse seems unlikely from an energy diffusion perspective-a high main bearing temperature does not necessarily indicate a broken bearing if the ambient temperature is the driving factor. Figure 1 shows the time-series towards failure, and compares them to the individual temperature measurements (dashed lines). The newly created features are indicated by solid lines. When focusing on the main bearing time-series, no changes can be seen towards the failure in the general trend. This can be explained by seasonal effects, since the ambient temperature falls and skews the main bearing temperature by lowering the temperature of the component. However, a different trend is shown when taking new features into consideration, in particular, the ones based on the ambient and nacelle temperature. These time-series show an increase in temperature towards the failure and indicate the progression of the fault. As the nacelle temperature is similarly driven by the ambient temperature, we include only the difference between the main bearing and ambient temperature as a new input feature for the proposed model.
Target Feature
Since related studies [2][3][4][5]10,12] in health prognostics are concerned with RUL or the remaining life-cycles of a component, we take the time to highlight the significance of the target feature presented here, namely the remaining generated power. Figure 2 shows both the RUL and RPG as a function of time, obtained from the SCADA data. Besides the physical relation to wind turbine operations, RPG and RUL are especially distinguishable in their time-series properties, when compared across other wind turbines. When comparing the RUL of all turbines (Figure 2a), it does not come as a surprise that the rate of which the RUL is decreased is the same for all turbines-since time is linear, only a linear relation can be expected. This makes the RUL a less favorable target feature for a predictive model, since any two turbines with equal RUL can have widely different RPGs, which can be seen by comparing Figuer 2b,c. When aligning the RPG with the end of each time-series, it can be seen that the rate by which the RPG is depleted towards failure, changes from one wind turbine to another. This can be explained by seasonal and local weather patterns characterizing each site, which result in different ambient conditions of the wind turbine. These local perturbations consequently lead to more variability in the target feature space.
As we later on will distinguish between run-to-failure and non run-to-failure wind turbines, run-to-failure wind turbines are indicated by dashed lines. In comparison to non run-to-failure wind turbines, they are spanning over a wide range of trends and absolute RPG, allowing the proposed model to cover a variety of different scenarios.
Predictive Modeling of Remaining Power Generation
We facilitate the prediction of the RPG in the setting of survival analysis [13], with censoring as a key concept. In short, the concept of censoring is that the time-series ends before the desired event. More specifically, we distinguish between right-censored data, such that the RPG is known to be above some time t, and non-censored data, when the RPG equals the current time. The hazard (rate of change of RPG towards failure) expressed at the current time, can be written as a hazard function η(t) and cumulative hazard function H(t). For a positive random variable RPG, it follows [13] for the cumulative distribution: and for the probability density function: Furthermore, for each cumulative distribution function, P(RPG ≤ t) there exists H(t) such that H(t) = − log(1 − P(RPG ≤ t)). Large values of the hazard function are an indication of a high risk of failure, i.e., vanishing RPG. In the remainder of this study we are concerned with the right tail of the RPG distribution: Combining both the censored and non-censored data, the likelihood function for the RPG in terms of non run-to-failure (C censored) and run-to-failure (F non-censored) wind turbines can be written as: Following Patti et al. [14], and introducing ∆ as a health indicator, such that ∆ = 0 indicates an observed failure, and ∆ = 1 otherwise, the likelihood reads: It follows that the log likelihood is given by: Equation (6) becomes the basis for the cost function of the RNN proposed later on. Here η and H can be obtained through the Equations (1) and (2) when choosing an appropriate distribution. In this study, following common practice in survival analysis [13] and its use in RUL estimation for main bearing failure in wind turbines [3], we chose the Weibull distribution for its versatile properties: where α ∈]0, ∞) is the distributions shape parameter, and β ∈]0, ∞) its scale. We will focus on RNNs with gated recurrent units (GRUs), letting the Weibull distribution be parameterized by θ = [α, β] . Omitting mathematical details, for which the interested reader is referred to textbooks, such as Goodfellow et al. [15], the parametrization is obtained by the RNN mapping m(·) of the current data x t , NN weights θ, and former layer output o t−1 : Combining Equations (6) with the hazard and cumulative hazard function of Equations (7) and (8) with the RNN, the problem to solve is the following: arg max ω log L(ω, RPG, ∆, x [1,t]
Model Optimization
The training and validation data for the proposed model are split 0.75/0.25 into sets of full turbine time-series. Besides the regular techniques for model evaluation through the loss function of the RNN model, we propose an error measure of the prediction itself with respect to the true remaining power generation,RPG. As the estimated RPG comes in terms of a parameterization of a probability function, the error will be a modification of the | · | =1 norm between RPG andRPG, such that: Based on the validation loss, we optimized the topology of the RNN by exploring the depth and width as listed in Table 2. Based on this investigation, the topology was fixed to two layers of 75 GRU nodes. The remaining hyper-parameters were set in accordance with general practice [16]: The dropout rate after each GRU layer was 20%, the batch size was 128 samples, and the learning rate was initialized at 0.0001. Where the dropout rate was selected by a discreet uniform search on [0, 40]% dropout rates, see Table 3 for the loss values of the RNN. Table 3. Dropout optimization in terms of loss for the optimal network of Table 2. In addition, we investigated the dependency of the sequence slicing for the input of the RNN. Changing the sequence length from 2 to 30 days, we evaluated the optimal sequence length by minimizing the error presented in Equation (11). Figure 3 shows the predictive error as a function of the true RPG. Despite all sequence lengths performing with a similar error, it can be noticed that a sequence length of 10 days yielded the best results in terms of predictive error and variability. Sequence slicing beyond 10 days did not improve the predictive error and became increasingly computationally expensive when averaging over all turbines. The proposed model thus captured up to weekly dynamics in each sequence.
Bearing Failure Study
The final model for this study is illustrated in Figure 4. After training, the model was applied to the remaining run-to-failure wind turbines. Figure 5 shows a common RPG prediction pattern in Figure 5a and a fast developing fault in Figure 5b. The first mentioned shows prediction in agreement with the expected RPG up to 60 days before the failure. Considering only the damage indicator, it is evident that predictions beyond 60 days for this particular wind turbine were difficult, since there was no evidence for robust estimations. As more data became available, and the closer the wind turbine was to failure, the predictions became more and more accurate. Consequently, the confidence bound around the prediction became narrow, which is indicated by the shaded area. In comparison to point-wise estimates, this shows why predicting a distribution is an advantage; it provides an operator with a measure of how reliable the predictions are. While the behavior shown in Figure 5a can be observed for most turbines, Figure 5b serves as an example of how the model performs when encountering unforeseen events. The predictions of the RPG were off by 2 to 3 GWh for most of the duration of the failure process, only converging in the last couple of days. This behavior arose from a lack of fast-developing faults in the training data set, biasing the model towards more common failure patterns. As the damage indicator did not pick up prior to 20 days before the failure, the model had no indication that allowed it to converge towards a prediction. This opens up for a discussion on what are the driving features behind the model. We will address this in the next section. In order to reduce good and bad predictions by chance, we investigated the ensemble performance of the proposed model by means of a five-fold cross-validation approach to see how the models performed on different test data. We illustrated this by considering one test turbine for all folds. Figure 6 shows the predictive error, where all folds started with different initial values, but they nevertheless converged rapidly towards accurate predictions at the 2 GWh mark before failure. From here on, they showed identical behavior providing consistent and increasingly accurate results. When averaging over all test wind turbines and folds, predictions became accurate at 1.8 GWh (The averaged prediction for all folds are shown in Table 4). For now, this provides the upper bound of what the proposed model is capable of predicting.
Feature Importance
As mentioned earlier, when the proposed model did not have sufficient information from the input feature space, predictions could not be considered accurate (See confidence bound in Figure 5). We again chose the turbine in Figure 5a to be representative of common fault behavior patterns. Further, we constrained the time scale under investigation to 120 days before the fault. Figure 7 shows relation between individual features and predictions. For illustrative purposes, the features were normalized individually, thus two equal values between features did not imply equal measurement in the feature space. However, the true RPG and mode of the predicted RPG were scaled to ensure direct comparison and maintain their relations as shown in Figure 5a.
The impact on the predictive capabilities of the proposed models from temperature measurements and the damage indicator is shown in Figure 7a. For the first 50 days neither the main bearing temperature, the difference between the main bearing temperature and the ambient temperature, nor the damage indicate suggested any abnormality. On this scale, the predictions of the proposed model were misaligned with the true RPG. After 50 days, features rapidly picked up, most notably the temperature based features to begin with. In this time period, the predicted RPG was getting in alignment with the true RPG. After further 20 days, the initial increase in the features was replaced by variations around some asymptote. As the predicted RPG remained to follow the true RPG, other features must drive the predictive behavior. Next, we took the generator revolution into account. Figure 7b contains the main bearing temperature, generator revolution, and their difference. When passing the 60-day mark before failure, the generator revolution had yet to show significant changes, and it was first in the remaining 20 days that changes could be observed. This change, though, did not show any trend. This changed when comparing the generator revolution with the main bearing temperature, from 40 days before the fault, and onwards, a steady increase of the combined feature could be observed. This is believed to drive predictions towards the end of the wind turbines RPG.
Summarizing the trends of Figure 7, the damage indicator was driving most of the prediction at the beginning of the fault, while temperature measurements aided in forcing the predictions to more accurate predictions. 20 to 30 days before the wind turbine failure, a combination of temperature and rotational features (i.e., generator revolution) seemed to drive predictions. In the case of fast-developing faults, such as shown in Figure 5b, none of the presented trends could be found. Thus, the model did not indicate that the wind turbine should undergo any state changes.
Ties to Remaining Useful Lifetime
As of the time of writing, we are not aware of any work addressing the prediction of RPG or similar. To compare the model to other approaches, we dedicate this section to explore the mapping from RPG to RUL and comparing the results to RUL prediction when employing the model by Herp et al. [3]. The mapping to RUL was done individually for each wind turbine. This was achieved by including how much power a wind turbine produced since the beginning of the time-series. Alternatively, one could fix an interval, e.g., 90 days before the current timestamp. Depending on the length of the interval, seasonal effects could be covered. The rate of power remaining could then be defined over a time ∆t: It follows that the mapping from RPG to RUL is given by: Figure 8 shows the mapped RUL from the RPG displayed in Figure 5a. The RUL median and mode, mapped from the RPG median and mode, showed accurate prediction around 60 days before the failure, slightly underestimating the RUL compared to the true RUL. Mapping RPG to RUL for other turbines that were representative for the common fault behavior, systematically underestimate the RUL, averaging 81 days, over the internal 50 to 100 days. Comparing to Herp et al. [3] average predictions of 94 days, this approach fell short of 13 days.
Conclusions
The purpose of this work was to create and motivate a model capable of predicting the remaining power generation (RPG) with application in wind turbine bearing failure. Furthermore, we have pointed out the similarities and differences between the RUL and the RPG. We have provided a description of the underlying model and how to tune its hyper-parameters. In the implementation of main bearing failure, we archived predictive horizons in the order of 1 to 2 GWh. In a five-fold cross-validation over multiple turbines, we showed that predictions become accurate at an upper bound of 1.8 GWh.
When converting RPG to RUL, we achieved predictive horizons of 50 to 100 days, averaging 81 days overall wind turbines. For a preliminary conversion, we consider this a satisfying result as the state-of-the-art [3] reports 94 days, on average.
During the course of this work, it becomes apparent that future research can follow up on the feature space presented, filtering or identifying more driving features that eventually will foster better predictions. Besides, as the proposed model performs differently on individual turbines with different failure behavior, one could consider following Aggarwal et al. [17] in allowing classification into different failure modes as en additional output of the RNN. | 2020-07-09T09:14:46.975Z | 2020-07-02T00:00:00.000 | {
"year": 2020,
"sha1": "ecec9947b3a61bd8086cb1b0a78577393afdda77",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/1996-1073/13/13/3406/pdf",
"oa_status": "GOLD",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "ad801aeea87c4f38bcebd4de016464a276cd8cd6",
"s2fieldsofstudy": [
"Engineering",
"Environmental Science"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
16079615 | pes2o/s2orc | v3-fos-license | NETopathies? Unraveling the Dark Side of Old Diseases through Neutrophils
Neutrophil extracellular traps (NETs) were initially described as an antimicrobial mechanism of neutrophils. Over the last decade, several lines of evidence support the involvement of NETs in a plethora of pathological conditions. Clinical and experimental data indicate that NET release constitutes a shared mechanism, which is involved in a different degree in various manifestations of non-infectious diseases. Even though the backbone of NETs is similar, there are differences in their protein load in different diseases, which represent alterations in neutrophil protein expression in distinct disorder-specific microenvironments. The characterization of NET protein load in different NET-driven disorders could be of significant diagnostic and/or therapeutic value. Additionally, it will provide further evidence for the role of NETs in disease pathogenesis, and it will enable the characterization of disorders in which neutrophils and NET-dependent inflammation are of critical importance.
| Neutrophil extracellular trap (NeT) formation and protein decoration. Representative images taken using confocal microscopy, demonstrating (A) NET formation mechanism and (B,C) the two-step process through which the disease-related protein is externalized. crucial for their antimicrobial function, enabling the increased local concentration of antimicrobial factors (1,2,6,7,11).
Herein, we seek to review current data regarding the proposed role of NETs in non-infectious human diseases. We also discuss the existing evidence supporting that these structures constitute a common mechanism of the pathophysiology of distinct diseases.
MeCHANiSM OF NeT FORMATiON
Despite the morphological similarities of NETs released by neutrophils in response to different stimuli and under diverse conditions, it is nowadays widely accepted that there is more than one mechanism involved in NET release (37). Additionally, mitochondrial DNA also contributes in NET formation (38,39), whereas, even though in vitro NET formation leads to cell death (40), it is reported that neutrophils that undergo NET release in vivo may remain active and functional, suggesting that NET formation may not necessarily be a terminal event (41,42).
Even though NETs constitute a common event in distinct pathophysiologic conditions, the expression of distinct bioactive proteins on NETs in different disorders might be the one that determines their specific function in disease pathogenesis.
A two-"hit" process has been proposed to explain the differential protein cargo of NETs in distinct disorders. The first "hit" in this process is the disease-specific environment that primes neutrophils to express disease-associated protein. A second "hit" is then required for the induction of NET formation (Figures 1B,C). However, this is a simplified model, and we cannot exclude the possibility that the same stimulus can drive both events. A typical paradigm of this two-"hit" model has been described in ST-segment elevation acute myocardial infarction (16). It has been shown in acute coronary syndromes that a variety of inflammatory stimuli trigger the cytoplasmic expression of TF in circulating neutrophils. At sites of atherosclerotic plaque rupture, locally activated platelets interact with TF-loaded neutrophils leading to the release of TF-bearing NETs inside the affected artery. The release of functional TF on NETs is able to further induce thrombin generation and platelet activation, creating a possible vicious cycle, that leads to thrombus propagation and stability (16).
The expression of these "disease-related" proteins on NETs could increase their local bioactivity (12,14,16,23,66,68). On the other hand, it has been shown that, at high densities, NETs limit inflammation by degrading cytokines and chemokines (69). This balance between the pro-inflammatory and prothrombotic role of NETs, though the expression of cytokines like IL-1β and IL-17, and their anti-inflammatory role, could be exploited for the development of new therapeutic approaches.
In the following section, we review the clinical and experimental data that link NETs with pathogenesis of several disorders. Even though the list of diseases in which NETs have been identified is extensive, we believe that the further characterization of the degree of NET involvement in such disorders could enable the classification of diseases in which NETs have a definite and strong involvement under the term of "NET-driven disorders" or "NETopathies. " The term NETopathy(ies) is derived from the abbreviation NET and the Greek word πάθος = pathos, which means disorder.
NeTs in Thromboinflammation
The widely accepted cross talk between inflammation and thrombosis has led to the introduction of the term thromboinflammation (68). Cells of the hematopoietic system, including neutrophils, platelets, and monocytes, have a major role in this process (64). There is increasing evidence implicating NET release with the development of both vein and arterial thrombosis (12,14,16,26,65,(70)(71)(72)(73)(74)(75)(76)(77). Extracellular deposition of DNA co-localized with neutrophil granule proteins has been shown in thrombi from patients with deep vein thrombosis (DVT) (78), especially at the phase of organization of the thrombus (70). Additionally, circulating extracellular DNA in the form of nucleosomes and DNA associated with neutrophil granule proteins, supporting the induction of NET release, has been identified in blood samples from patients with DVT (79,80). Similarly, NETs have been identified in thrombus specimens from patients undergoing thrombectomy in the context of myocardial infarction (15,16,62,71). In a recent multicenter study in patients presenting with stent thrombosis, neutrophils were the more abundant leukocyte population in thrombus specimens, whereas NETs were identified in 23% of thrombi (71). Regarding specific disorders associated with thrombotic manifestations, NETs in thrombus specimens and/ or increased levels of nucleosomes have been identified in disseminated intravascular coagulation in sepsis (73), in paroxysmal nocturnal hemoglobinuria (81), thrombotic microangiopathies (82), antiphospholipid syndrome (APS) (74), antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) (14), or hemodialysis-related thrombogenicity (83). These clinical data support a role for NETs in the development of both arterial and venous thrombosis.
The prothrombotic role of NETs was further confirmed in several experimental animal models. NETs were observed in thrombi, in a baboon model (75) and in several mouse models of DVT (64,76,84). In a mouse model of DVT, infusion of DNase I resulted in protection from thrombosis (76), whereas PAD4 −/− mice were protected from thrombosis (85), supporting the pathogenetic role of NETs in venous thrombosis, at least in this animal model. The in vivo role of NETs in the development of thrombosis was further shown in a mouse model of APS (86). Additionally, NETs contribute in cancer-induced venous thrombosis, as shown in a mouse model of chronic myelogenous leukemia (34) and in the RIP1-Tag2 model of insulinoma and MMTV-PyMT model of breast cancer (77). Brill et al. linked histones with the prothrombotic effect of NETs, since histone infusion also resulted in thrombosis. However, there is evidence that NETs participate in DVT via interaction with von Willebrand factor, a factor that potentially activates platelets (76). Furthermore, it has been reported that in a mouse model of DVT TF triggers intraluminal fibrin formation, while the release of NETs activates factor XII, consolidating DVT (64). The involvement of NET-bound TF, which is the main in vivo initiator of coagulation (87), in NET-dependent thromboinflammation has been shown in several studies, since TF has been identified in NETs released in neutrophils from patients with sepsis, APS, AAV, or myocardial infarction (12,14,16,74) or in a mouse model of DVT (64).
The interplay between neutrophils and platelets has been shown to have a major contribution in NET release (16,62,72,84,88). Clark et al. have shown that upon toll-like receptor 4 (TLR4) activation platelets induce the formation of NETs in a mouse model of sepsis (72). This leads not only to bacterial but also to platelet entrapment in NETs, resulting in tissue damage (72). Several studies have further identified platelet derived high mobility group box 1 (HMGB1) as the factor that mediates platelet-neutrophil interaction and NET release (62,84). HMGB1 released by platelets has been shown to promote thrombosis in a mouse model of DVT (84), whereas it mediates neutrophil activation in the context of myocardial infarction (62). The importance of platelet-neutrophil interaction is prominent in coronary artery thrombosis, since it was proved that coronary thrombi are mainly composed of interacting neutrophils and platelets (16,62). The rupture of the atherosclerotic plaque primes a cascade of events, which results in platelet activation and NET release, leading to thrombus formation and blood vessel occlusion. The expression of TF on NETs may propagate the further activation of the coagulation system, leading to thrombus expansion (16).
Taken together, there is strong evidence, derived by clinical and experimental observation, that neutrophils and NETs are major players in both venous and arterial thrombosis. The development and clinical use of factors that target NETs could provide, however, the definite proof for the role of NETs in thrombotic disorders.
NeTs in Autoimmune Diseases
A growing number of studies demonstrate that NETs play a driving role in the pathogenesis of a variety of autoimmune disorders, such as SLE, AAV, RA, and psoriasis. In the aforementioned disorders, NETs are a main source of autoantigens, are present in excess amount, or are decorated with disease-specific proteins.
Systemic Lupus Erythematosus
Systemic lupus erythematosus is a systemic autoimmune disease and a well-studied model. SLE is characterized by systemic production of autoantibodies against a plethora of intracellular and extracellular targets. These autoantibodies are able to cause extensive tissue damage (89,90).
There is evidence supporting the involvement of NETs in the pathophysiology of SLE. It has been shown that NETs are directly associated with the severity and the progression of the disease (91)(92)(93)(94)(95). Neutrophils from SLE patients are primed to undergo NET release (17,96). Autoantibodies and more specifically antibodies against LL-37 have been shown to activate neutrophils for NET release (18,19). On the other hand, NETs are composed of DNA, histones, and proteins-like LL-37, providing a possible source of autoantigens for the development of lupus-specific autoantibodies ( Figure 2B) (17)(18)(19)(97)(98)(99). Interestingly, Villanueva et al. reported a neutrophil subpopulation in SLE, termed as low-density granulocytes (LDG), prone to release NETs, which promote vascular damage (18,91,100). It was further demonstrated that LL-37-bearing NETs fuel the immune response in SLE by activating plasmacytoid dendritic cells (pDCs) in an Immunoglobulin-Fc region receptor II-a (FcRIIa) and TLR9-dependent manner. This leads to interferon alpha (IFNα) production, which is a critical player in the pathogenesis of SLE. Furthermore, IFNα triggers NET generation and activates T and B cells leading to the production of antibodies against NETs, creating a vicious cycle (19,97,101,102).
Interestingly, there is a disease-associated defect in the clearance of NETs, due to the reduced activity of DNase I and the increased amounts of DNase I inhibitors (17,20,94,(103)(104)(105)(106), supporting the hypothesis that dysregulation of NET clearance may be one of the initial steps that lead to lupus-specific autoantibody production. proteins, such as proteinase 3 (PR3) and MPO. The study by Kessenbrock et al. provided the initial evidence for the link between NETs and AAV. In this study, the intraglomerular deposition of NETs in biopsies from patients with small-vessel vasculitis was shown. Additionally, it was shown that neutrophils release NETs when activated with ANCA (107). Further studies confirmed the deposition of NETs in affected tissues from patients with AAV (14,61,(108)(109)(110), whereas increased levels of circulating NET remnants were observed in patients with AAV (14,22). Additionally, a recent study correlated AAV disease activity with the presence of NET-prone LDGs in peripheral blood (110). NETs were further associated with the AAV hypercoagulability, since NETs released during active disease are loaded with TF [ Figure 2A (14, 111)].
Since PR3 and MPO are abundantly present in NETs, it has been proposed that NETs mediate the extracellular exposure of these potential autoantigens, having an important role in the initiation of the disease (17,20,21,107). Sangaletti et al. have shown that myeloid DCs can acquire neutrophil proteins released in the form of NETs. Furthermore, immunization of mice with DCs co-cultured with NET remnants resulted in the development of MPO-ANCA and renal vasculitis (112). A common characteristic between SLE and AAV is the decreased degradation of NETs, attributed to the reduced activity and inhibition of DNase I, as well as to the protection over NETs by autoantibodies and components of the complement (17,20,107).
RA and Psoriasis
Rheumatoid arthritis is a chronic autoimmune disease that affects synovial joints. It is known that neutrophils are the most abundant cell type of synovial fluid in RA patients (113).
Recent studies identified the presence of NETs in the circulation and the release of NETs by synovial neutrophils (114,115). Khandpur et al. have shown that TNF, IL-17, and anticitrullinated protein antibodies (ACPA) promote NET release by neutrophils from patients with RA, whereas therapeutic blockade of TNF function has been shown to decrease the extensive NET generation that characterizes RA patients. Of interest, IL-17 was able to promote NET release only in neutrophils from patients with RA, which implies that the disease-specific inflammatory microenvironment primes neutrophils for NET formation (115).
Recent studies highlight that citrullinated histones in NETs consist autoantigens that stimulate and participate in the outset of the excessive inflammation, and more specifically in ACPA immune response, in RA (18,115). It has been further demonstrated that RA-driven NETs are decorated with enzymatically active PAD4, which possibly further citrullinates targets, rendering them autoantigens ( Figure 2C) (49,67,116). Finally, NETs in RA indirectly participate in the stimulation of distinct cell types, such as fibroblast-like synoviocytes, which invade and damage cartilage in RA (115,117).
The possible involvement of NETs in the pathogenesis of psoriasis has been also proposed. Psoriasis is an autoimmune skin disorder characterized by epidermal hyperplasia and neutrophil infiltration in the epidermis. Neutrophils are involved in the pathophysiology of psoriasis, linking innate and adaptive immune system, and acting as a main source of IL-17 (66,118,119).
Interleukin 17 has a significant role in the pathophysiology of psoriasis causing keratinocyte hyperplasia (119, 120), whereas therapeutic administration of antibodies against IL-17 is efficacious in the treatment for psoriasis (121)(122)(123). The externalization of IL-17 in a bioactive form is feasible through NET formation ( Figure 2D) (66,124), which has been also observed in models of RA (115) and pancreatitis (125). The fact that the active form of IL-17 lies on NETs renders it an easily accessible target.
Taken together, a significant amount of evidence suggests that NETs contribute in the pathogenesis of several autoimmune disorders, acting either at the initiation of disease, providing a source of autoantigens, or promoting tissue injury (66,90,93,107,109,115). There are reports suggesting that NETs can activate other inflammatory cell populations and promote the activation of the adaptive immune system (97,102,115). However, whether the specific structure of NETs and the possible modification in proteins loaded on NETs have a major impact in the break of tolerance and induction of autoimmunity still remains elusive.
NeTs in Autoinflammatory Diseases
Recent studies revealed a possible role for NETs in the inflammatory response that governs autoinflammatory syndromes, including gout and FMF.
Gout is an autoinflammatory type of arthritis caused by the intra-articular deposition of monosodium urate crystals (MSU crystals). The deposition causes inflammatory attacks due to innate immunity activation (126)(127)(128)(129). Additionally, the chronic form of the disease is characterized by tophus formation, causing mechanical destruction of the joint (130). It has been shown that MSU crystals cause a strong induction of NETs (24,131) which, in high neutrophil concentrations, ameliorates MSU crystal-induced inflammation by promoting the degradation of inflammatory cytokines and chemokines in a mouse model of MSU-induced inflammation (69,132). Despite their protective role, NETs indirectly engender the destruction of the joint by easing the packing of MSU crystals and the formation of tophi (69,132). However, whether NETs support the initiation of gouty inflammation in humans remains unanswered.
Familiar Mediterranean fever is a hereditary autoinflammatory disorder, characterized by inflammatory attacks and neutrophil infiltration into the affected sites (23). Moreover, it is an IL-1β-mediated disease, and this is clear due to the fact that IL-1β blockade constitutes an emerging treatment in FMF (23,133,134). During FMF attacks, neutrophils undergo excessive NET formation, which decreases after the inflammation dissolution (23).
During FMF attacks increased levels of circulating MPO-DNA complexes are detected, suggesting the release of NETs in the systemic circulation, whereas their levels normalize during the resolution phase of the disease (23). The detection of bioactive IL-1β in NETs released ex vivo by patient neutrophils or control neutrophils treated with FMF attack serum implies that neutrophils serve as critical effector cells in the amplification of inflammation in FMF (Figure 2E) (23).
NeTs in Metabolic Disorders
In type II diabetes (T2D), immunological changes lead to altered levels of cytokines and changes in both number and activation status of various leukocytes, including neutrophils (135). Until recently, it was thought that inflammatory responses may have a dual role in T2D, as they seem to have a causal relationship leading to resistance to insulin, while on the other hand they seem to be intensified by the hyperglycemic state, resulting in T2D complications (135).
Bearing in mind that diabetes affects neutrophil count and activity, that hyperglycemia-driven oxidative stress facilitates diabetic complications, and that neutrophils generate oxidative stress in diabetes, it was assumed that a dysregulation in NETosis may represent the link among hyperglycemia, oxidative stress, inflammation, and diabetic complications (27). In this direction, a recent study demonstrated that high glucose in vitro and hyperglycemia in vivo induce release of NETs and their products (27). Another study provided evidence that hyperglycemic conditions lead to the formation of short-lived and unstable NETs, while also prime neutrophils and constitutively activate NET formation, leading to reduced response to subsequent external stimuli (136). Thus, it was hypothesized that neutrophils primed due to hyperglycemia may not respond to further external stimulus in T2D patients, making them susceptible to infections (136). Finally, a third study demonstrates that, in T2D patients, dysregulated NET release caused by hyperglycemia is responsible for impairment of wound healing as well as for diabetic complications (137). Even though these studies support a role for NETs in T2D, it is not clear to what extent manipulation of neutrophils could ameliorate or prevent diabetic complications.
Moreover, there is evidence that neutrophils and NETs have a potential role in the pathogenesis of type I diabetes (28,138,139); however, their implication in the onset and/or the development of this disease has not been investigated so far.
NeTs in Lung Diseases and Fibrosis
Neutrophil extracellular traps have been implicated in inflammatory lung diseases and inflammatory-derived fibrosis (33). Several inflammatory lung diseases are characterized by the migration and detection of neutrophils and monocytes in the airway lumen and the bronchoalveolar lavage fluid (140). NETs have been associated with inflammatory diseases, such as chronic obstructive pulmonary disease (COPD), cystic fibrosis (CF), acute lung injury, acute respiratory distress syndrome, and asthma (29,30).
Cystic fibrosis is characterized by abundant free DNA structures in airway fluids that increase the viscosity of the sputum and lead to airflow obstruction and tissue damage. Free DNA originates mainly from NETs released from neutrophils that are recruited to the area in an effort to kill the bacterial burden, but they finally contribute to the damage of lung tissue (31,32). Additionally, it has been proved that NE plays an equally important role in CF, leading to tissue damage, especially in patients under treatment that are characterized by increased DNA cleavage (141). Recombinant human deoxyribonuclease (rhDNase) is an adjunctive to antibiotics treatment for patients with CF over the last two decades, showing a beneficial effect at least in a subpopulation of patients with CF (142,143). Moreover, it has been reported that DNase I and histone-blocking antibodies have been used in mice against transfusion-related acute lung injury, in which NETs play a crucial role (144). Inhibition of either NE or NET release in general could be a novel future therapeutic strategy in patients with CF (141,145).
There is evidence that the inflammatory microenvironment developed in chronic lung diseases including COPD and interstitial lung disease contributes either to localized or to generalized fibrosis, respectively. Specific fibrosis-related agents, such as cigarette smoke, magnesium silicate, and bleomycin, stimulate neutrophils to undergo NETosis. NETs indirectly regulate fibrosis by activating lung fibroblasts and differentiating them into myofibroblasts, through autophagy and histone hypercitrullination. Subsequently, NET remnants, such as IL-17, regulate connective tissue growth factor (CCN2) expression and collagen production by the differentiated fibroblasts and not their differentiation ( Figure 2D). However, NET degradation significantly restricts these effects, indicating that it could be possibly used as a restraining mechanism against fibrosis (33).
There is increasing evidence supporting that, in both experimental models and cancer patients, NET deposition in the tumor mass is associated with tumor progression (35,146,(150)(151)(152)(153). A finding that supports the implication of NETs in tumor biology is that tumor cells predispose neutrophils to undergo NETosis (34,146). Moreover, in the tumor microenvironment, NETs interact with tumor cells and expose them to bioactive proteins, possibly favoring their survival through induction of proliferation and inhibition of apoptosis, as well as supporting their escape from the primary tumor (148).
Excessive NET deposition leads to a persistent inflammatory state (154)(155)(156), which in cancer probably promotes the expression of adhesion molecules (157)(158)(159). Under inflammatory conditions, when NET formation is induced, circulating tumor cells are more prone to adhere to end organ vasculature (158)(159)(160). Thus, given that the entrapment of bacteria is one of the primary roles attributed to NETs, they probably act accordingly to capture circulating tumor cells. By entrapping tumor cells and exposing them to various neutrophil-derived factors, NETs may generate a microenvironment rich in proteins and enzymes that promote tumor cell survival and progression (35,36,149,153). Taken together, these data support a potential pro-metastatic role for NETs, involved in early adhesion, proliferation, invasion, and angiogenesis.
Neutrophil extracellular traps have also been implicated in cancer-associated thrombosis, the second most common cause of death in cancer patients (34). Recently, it was demonstrated that, through the generation of NETs, neutrophils provide a scaffold and a stimulus for platelet adhesion and thrombus formation (75). NETs were shown to promote coagulation as well (68,75). Moreover, a recent study based on murine models reported that both leukemia and solid tumors produce a factor, G-CSF, that primes neutrophils to undergo NETosis and predisposes the host to thrombosis (34). In conclusion, NETs have been identified as a key player in cancer-associated thrombosis.
The biological significance of NETs in cancer remains unclear. It is hypothesized that initially they represent a reaction of the tumor environment against the growing cancer. However, NETs seem to play an adverse role in tumor growth, offering a scaffold with an array of biologically active molecules attached on it, which may promote malignant cell survival, growth, and local tumor expansion.
THeRAPeUTiC AND DiAgNOSTiC/ PROgNOSTiC POTeNTiAL OF NeTs
To date, clinical and experimental evidence highlight the significant role of NETs in the pathophysiology of the aforementioned diseases. Even though studies in animal models have shown the beneficial role of NET inhibition, especially in thrombosis, it is yet unknown whether NET-targeting therapies could be effective in clinic (161). NET induction or inhibition could be beneficial for patients with diseases that have been associated with restricted or excessive NET formation, respectively ( Table 1). To this end, drug repositioning offers the opportunity for the immediate use of therapeutic agents that induce or inhibit NETs, which are already used in clinic (11).
Several drugs already used in clinical practice might affect either NET formation or integrity, or the expression of NET proteins. For instance, it is known that hydroxychloroquine (HCQ), a drug that has been used for decades in the treatment of SLE, has anti-autophagic effect (162). Since the autophagic machinery is an essential step for NETosis, the effectiveness of HCQ may be mediated through the indirect inhibition of NET formation (Table 1; Figure 3A). In addition, rhDNase administration, a therapy used in patients with CF aiming to the liquefaction of mucus (142), may possibly target NET structures. DNase promotes thrombolysis via degradation of NETs in murine models (Table 1; Figure 3B) (64,76). Moreover, monoclonal antibodies are widely used against bioactive NET proteins, externalized through NET formation. In psoriasis, treatment with anti-IL-17 antibodies (121), probably targets the IL-17-decorated NETs, the main origin of bioactive IL-17 in psoriasis (66). Finally, NET-bound IL-1β may be one of the targets of anti-IL-1β therapies, such as canakinumab which targets bioactive IL-1β in FMF or gout patients (Table 1; Figure 3C) (134).
There are a few recent studies demonstrating that NETs could also have prognostic and/or diagnostic potential, as they could represent a disease activity marker for some of the aforementioned diseases (161). Furthermore, the measurement of NET release or specific NET protein expression in blood samples and biopsies could be a useful diagnostic tool (150,171). Nevertheless, further experimental data are needed to evaluate the therapeutic, prognostic, and/or diagnostic potential of NETs.
CONCLUSiON
The identification of NETs and the characterization of their role in disease have revived the overlooked role of neutrophils in disease pathogenesis. Phagocytosis of pathogens and limitation of infection was considered the exclusive role of neutrophils. However, mechanistic studies in animal models and clinical observation dramatically altered our perception of the involvement of neutrophils in disease during the last decade. From a patrolling police force, neutrophils are considered nowadays an important player in autoimmune diseases or thrombotic disorders, which were previously thought to be exclusively mediated by adaptive immune system and platelet or endothelial cells, respectively. The characterization of the differential protein load and function of neutrophils, and subsequently of NETs, in distinct disorders can provide novel diagnostic targets and targets for therapeutic intervention. Additionally, the study on the role of NETs in modulation of tissue homeostasis, including the initiation and resolution of inflammation and the elucidation of the effect of NETs on different cell population involved in inflammatory, autoimmune, or thrombotic disorders, will increase our knowledge in the mechanisms that govern the pathogenesis of complex disorders. The clarification of the role of NETs in the pathogenesis of such disorders and the clinical use of therapeutic agents that target NETs will enable the identification of a group of disorders that could be characterized by the term NET-associated diseases or NETopathies.
AUTHOR CONTRiBUTiONS
AM, AA, SA, IM, and KR wrote the manuscript and created the figures. IM and KR also revised the manuscript.
FUNDiNg
This study was supported by the Scientific Committee of Democritus University of Thrace. Grant number -80895. | 2017-05-03T18:56:34.722Z | 2017-01-11T00:00:00.000 | {
"year": 2016,
"sha1": "601a9e8ba3369e49610a8e6cd367636b46b48fc5",
"oa_license": "CCBY",
"oa_url": "https://www.frontiersin.org/articles/10.3389/fimmu.2016.00678/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "601a9e8ba3369e49610a8e6cd367636b46b48fc5",
"s2fieldsofstudy": [
"Biology",
"Medicine"
],
"extfieldsofstudy": [
"Biology",
"Medicine"
]
} |
253150827 | pes2o/s2orc | v3-fos-license | DIGITAL RESOURCES AS A WAY TO INCREASE THE MOTIVATION OF ECONOMIC SPECIALTIES STUDENTS IN STUDIES OF MATHEMATICS 1
The purpose of the work is to analyze the technical, methodical and psychological aspects of digitalization in education and, in particular, the methods of ensuring the effectiveness of independent work of students in the conditions of e-learning. The main attention is paid to the problem of the influence of interactive educational technologies on the formation of external and internal motivation of students of economic specialties to study mathematical disciplines. Methodology. In the conditions of e-learning continuous monitoring of the success of students in mastering mathematical methods and their application to solve economic problems was carried out. The success of each student in performing each type of work separately, as well as his overall rating among other students in the group, were determined. These results were supplemented by the results of the students' questionnaire regarding their own attitude towards interactive technologies as a tool aimed at forming motivation for learning. Results. The virtual environment for e-learning was built using Moodle LMS and contained learning digital resources of various levels of interactivity, including electronic multimedia publications. This helped to fully meet the needs of the distance educational process. To form the student's external motivation, the authors used an electronic journal in which the types of current tasks, points for their completion, the student's rating for each type of tasks, as well as for all types of tasks in general, are defined. To support internal motivation, interactive learning elements were developed and implemented. The effectiveness of the use of digital resources was confirmed during the monitoring of students' success and by the results of the survey of the participants of the experiment. Conclusions. A significant advantage of the use of interactive components in the educational process should be considered the creation of conditions for a better understanding of theoretical material and using mathematical apparatus for solving real economic problems. The use of multi-level digital resources gives the student the opportunity to build an individual educational environment that increases internal motivation to study.
INTRODUCTION
In modern conditions of the formation of the knowledge economy, the country's competitiveness depends on its competitiveness in the field of science and education. In Ukraine, higher education has a mass character. Among the population aged 23 and over, 82.7% have a higher education, and according to the Tertiary Enrollment component of the Global Innovation Index, Ukraine ranked 14th among 131 countries in 2020 (Global Innovation Index 2020, 2020.
EDUCATIONAL CHALLENGES
terms of quality of education among 50 countries. All this indicates that the quality of higher education in Ukraine does not meet the expectations of employers, students, or society as a whole.
In order for graduates of higher education institutions to be competitive on the labor market, they need not only to have modern theoretical knowledge and practical skills, but also to be able to solve complex tasks, create innovative intellectual products, and also learn to learn, that is, to be internally motivated to learn. Therefore, one of the basic directions of education reform in Ukraine (Reforma osvity, 2020) is to ensure high-quality higher education and expand opportunities for adult education. In these conditions, the education system must develop and improve in order to be able to fulfill its mission, namely to provide a person with the opportunity to realize himself in the future.
And one of the leading trends in the development of higher education, the implementation of which can solve this problem, is considered digitalization of education. It is thanks to digitalization that it is possible to ensure the access of a wide range of applicants to quality education not only at the national, but also at the world level. Therefore, when building the latest educational programs, the acquisition of universal competencies is considered the most relevant: the ability to learn, process information, quickly master new progressive technologies, the ability to think critically and approach tasks creatively (Strategy for the Development of Higher Education, 2020, p. 30).
And the transition of educational technologies to a digital format is able to ensure the accumulation of such competencies. However, there is a significant gap between the opportunities offered by modern digital technologies and those information and communication technologies that are used in the educational process, if such use takes place in general. Therefore, there is a need to fully realize the advantages provided by modern digital technologies for the educational process improvement and the use of interactive learning methods to increase its effectiveness.
At the level of higher education, digitalization involves admittance to respective technologies for both students and teachers, as well as for the administration of the educational institution, the availability of digital multimedia content, as well as the ability and skills of students and teachers to use digital learning resources (Project: Digital Agenda of Ukraine, 2020 25). Digital learning resources are any educational information provided in digital format.
At the same time, preference is given to open resources for that any digital devices (from a computer to a mobile phone) are suitable. It should also be emphasized that one of the features of the present time is the rapid change of educational content, therefore, it requires regular updating, and this is easier to do if it is provided in digital format.
Therefore, digitalization is considered as one of the leading factors in building an innovative university. It is the creation of the industry of innovative digital technologies that makes it possible to achieve the strategic goal of education reform in Ukraine, namely, to ensure the availability of high-quality higher education for various segments of the population.
Taking into account the aging of the population, on the one hand, and the accelerated development and implementation of the latest technologies in various fields, on the other hand, the use educationalchallenges.org.ua
EDUCATIONAL CHALLENGES
of digital technologies creates conditions for any person to renew his professional qualities throughout his life, which is provided for by the UNESCO "Lifelong Learning" program.
Social, theoretical and technical aspects of the application of information technologies in education began to attract the attention of scientists and practitioners at the end of the 20th century.
However, these questions became especially acute in connection with the spread of COVID-19, when the educational process in educational institutions of all levels began to be carried out in a distance form (Jackson, 2021).
What urgent measures should be implemented to accelerate the digitization of education in Ukraine? First of all, attention is paid to the software of the educational process at the "student" and "teacher" levels (Bykov, Spirin, & Pinchuk, 2020;Morze, at al., 2021;Modlo, at al., 2020;Shakhina, 2017, etc.).
There are also examples of the implementation of information systems supporting the educational activities of universities simultaneously at three levels: "student", "teacher", "administration" (Kukharenko, & Bondarenko, 2020;Trius, at al., 2021). In fact, we are talking about the creation of a virtual university as a digital reflection of the corresponding institution of higher education.
However, the development of software is only one aspect of the problem. Another, but no less important, is the structure and content of digital learning resources. The teacher must not only ensure the highquality content of these resources, but also rationally build them to realize their full potential. And not all teachers have the necessary skills for this (Börnert-Ringleb, Casale, & Hillenbrand, 2021).
There are many types of digital learning resources, among which the electronic textbook as a structured multimedia publication occupies the highest level. Such an electronic educational and methodological complex is able to ensure user interactivity thanks to the creation of an educational SMART environment (Nahaev, & Hrynova, 2020).
The main advantage of multimedia electronic textbooks is that they are able to provide an interactive dialogue between the user and the software system, while there is an opportunity to choose options for the level of educational content, mode and place of work, that is, the student has the opportunity to create his own educational space (Gurevich, Kademiya, & Shevchenko, 2012, p.71).
This means that the user has the opportunity to customize and optimize the organization of the learning process itself. These properties of digital resources are important not only for ensuring the quality of education, but also for the formation of students' internal motivation to study. For example, the use of digital resources increases student engagement in face-to-face learning (Ullah, & Anwar, 2020).
When
introducing learning digital resources in the conditions of distance learning, there is a need to create effective factors that motivate the user to study, and this puts forward additional requirements for the content of the digital resource and its structure. That is why most studies devoted to the structure of elearning tools emphasize the importance of having an interactive component and improving the principles of its development. However, these issues are just beginning to be studied.
EDUCATIONAL CHALLENGES
resources in the study of mathematical disciplines by students of economic specialties; study of factors that ensure motivation of students in distance learning conditions; determination of the influence of the use of interactive learning digital resources on the effectiveness of the formation of student motivation factors.
THEORETICAL FRAMEWORK
In the general case, motivation is considered as a process of influencing oneself or another person in order to encourage certain actions that are necessary to achieve a predetermined goal. in our study, such a goal is the acquisition of theoretical knowledge in mathematical disciplines and practical skills in their application to solving real economic problems. And it is necessary to direct students' aspirations to achieve this goal. The effectiveness of managing this process largely depends on how successfully the factors of external and internal motivation work, that is, how well the levers of influence are correctly chosen.
When building a motivation system in the educational process, such theories of motivation as the Social Cognitive Theory and the Self-Determination Theory are used (Schunk, Meece, & Pintrich, 2014). According to the Social Cognitive Theory, a person acquires new knowledge as a result of social interaction. Depending on whether the reward or punishment is received by the person whose behavior is considered as a role model, the observer makes a relevant decision about his actions in similar circumstances in the future.
Self-Determination Theory is based on distinguishing motivation into two types: external and internal. Central to this theory is the distinction between autonomous and controlled motivation. Autonomy implies that a person makes decisions according to his feelings and experiences. Intrinsic motivation is an example of autonomous motivation, when a person learns what is interesting to him, he considers it important, and therefore he applies willpower to it.
On the contrary, external motivation assumes that a person is driven to perform certain actions by external pressure, the feeling of the need to participate in these actions, social factors, etc. It should be noted that in general, external and internal motivation are not considered as two opposite poles. Between them there is a continuum of different forms, which to one degree or another combine these two types of motivation.
In turn, the choice of motivation methods is related to the learning model, which is the leading one when studying one or another discipline. The most common are four models (Vermunt, & Donche, 2017): learning that is focused on reproduction; meaning-oriented learning; learning that is focused on implementation; learning that does not have a specific orientation. In the case of using a model of learning that is focused on reproduction, the student is motivated only to reproduce the educational material during the test.
The meaning-oriented model of learning assumes that a student must understand the relationship between various structural elements of an educational discipline, be critical of the information he receives. He regulates his own learning and strives to achieve the most complete understanding of the material studied within the given academic discipline.
Studying according to the embodimentoriented model, the student's efforts are aimed at determining the connections between the knowledge he receives and educationalchallenges.org.ua
EDUCATIONAL CHALLENGES
the possibilities of their application in his future professional activity. It can be expected that the use of interactive learning technologies in combination with the possibilities of digitization will increase the effectiveness of learning in the implementation of any of these models.
The first studies of motivation to study during the period of quarantine restrictions related to the epidemic of COVID-19 revealed that the majority of students were tuned to a model of learning that is oriented to reproduction (Rahiem, 2020). And our experience also proves this. Such students were not ready for autonomy in learning, and for them the most effective was external motivation in the form of grades as a reward for completing tasks, as well as in the form of penalties for ignoring tasks. Some students even lost motivation to study, which can be explained by their lack of psychological stability (Cole, Field, & Harris, 2004).
But among the students there were those who perceived some aspects related to distance learning as advantages (Rahiem, 2021), namely, personal (challenge, curiosity, self-determination, satisfaction), social (well-being, relationships) and external environment (convenience, release of additional time). For such students, electronic learning tools made it possible to find additional motivational resources.
And it can be expected that digital resources are the most capacious in this context. It is they that allow creating conditions for the development of internal motivation among those students who are focused on determining the meaning of acquired knowledge and their further implementation in practical activities.
In our research, the results of which are presented in this article, we see the task of the teacher in the development of such a structure of digital resources, which would not only support external motivation, but would help to form in students, as future specialists in the field of economics, internal motivation for in-depth study theoretical provisions of mathematical disciplines, as well as mastering mathematical methods and gaining practical experience in their application to solving real economic problems.
METHODOLOGY
During the 2020/2021 and 2021/2022 academic years, the performance of fulltime students of the Simon Kuznets Kharkiv National University of Economics in the first and second years of study was monitored in terms of mastering mathematical disciplines in the conditions of remote learning. To ensure a fullfledged educational process, a wide range of digital resources of different levels of interactivity were used.
The virtual learning environment was created on the Moodle LMS platform, which is a fully customizable learning management system. Quantitative analysis of students' success, which was carried out throughout the educational process for each type of tasks, was combined with qualitative analysis, which was based on a questionnaire of students upon completion of the study of the discipline. The list of tasks to be completed by students during the semester was given in the course syllabus.
This syllabus also provides a rating scale for each type of task. Overall success was evaluated according to the cumulative system, for which a 100-point scale was used. All learning resources were posted on the website of the educational discipline of the personal educational systems of S. Kuznets KhNUE and have For a more detailed study, three academic groups of students of the 2nd year of the specialty Entrepreneurship, Trade and Stock Exchange Activity with a total number of 69 people were chosen. Monitoring their performance in the first year showed that the average performance in these three groups can be considered the same. In the second year, when studying the discipline Operations Research and Optimization Methods, great attention of the third group students was paid to interactive elements of learning, and instead of the usual homework, students passed tests and performed training exercises.
For students of the first and second (control) groups, traditional tasks from the textbook were offered. It should be emphasized that training of all these groups follows the same syllabus, and all students had full access to learning resources. In addition to the monitoring of student academic performance, monitoring of what types of resources students chose and for how long they used them was carried out. After the evaluation of the acquired knowledge and skills, which consisted of an ongoing assessment during the semester and a final assessment in the form of an exam, a student questionnaire was conducted in order to determine their own attitude to factors aimed at forming internal motivation to study. To process the results of the questionnaire (which is essentially an expert assessment), the method of vector preference will be used.
RESULTS
The experience of using the Moodle LMS for the organization of student learning is being accumulated at S. Kuznets KhNUE, starting from 2008. For correspondence students, digital resources using Moodle LMS are already a familiar tool. It is on their using the independent work of students is built when studying by correspondence. But for full-time students, interactive learning has become actively used only in recent years, which was due to the COVID-19 epidemic.
Although some elements of interactivity (for example, tests, exercises for selfchecking, business games) were also used in the process of teaching to one extent or another, but this was not mandatory and completely depended on the teacher's preferences. As a rule, for full-time students, digital resources mainly contained reference information.
However, in quarantine conditions, to ensure the effectiveness of training, it was necessary to use such learning digital resources that would contribute to the active involvement of the student in the learning process itself. Moodle LMS provides great opportunities for adapting the educational process to the distance format at all levels -from content creation to assessment of acquired knowledge. Each teacher has the opportunity to create his own website, on which to present a complex of digital resources in a digital format.
The following structure of the complex of the digital resources was proposed for mathematical disciplines. It contained a syllabus, a detailed work program of the academic discipline, a work plan (technological map) with a rating scale for each task, a list of recommended literature, links to useful Internet resources, lecture materials and their presentations, methodological materials for practices and labs, sets of homework, sample tasks of control works, training exercises, test tasks for self-diagnosis, multimedia publications, which are the authors' own development in coeducationalchallenges.org.ua
EDUCATIONAL CHALLENGES
authorship with by other teachers of the department, as well as materials and resources for online final control (exam). It should be emphasized that one of the advantages of such electronic complexes is the ability to quickly update them.
According to the results of monitoring the academic performance of students of the first year of study, three academic groups were selected who study under the same program and whose success in mathematical disciplines can be considered the same on average. These students participated in a detailed exploration. The average level of academic performance of the students of these three groups based on the results of the evaluation of the disciplines of the mathematical cycle studied in the 1st year (Higher Mathematics, Probability Theory and Mathematical Statistics) is 73-76 points on the ECTS scale (Fig. 1).
Distribution of academic success of students of the explored groups according to the results of studies in the disciplines of Higher Mathematics and Theory of Probabilities and Mathematical Statistics
Therefore, the sample population of students can be considered homogeneous according to the initial level of knowledge, which is the necessary condition for the purity of the experiment regarding the results of the introduction of interactive components into the educational process. The hypothesis to be tested is the positive impact of the implementation of digital resources in the education on the success of students due to the formation of both external and internal motivation to study.
Let's consider the factors affecting the external motivation of students. To ensure it, the system of automated current control was used, which is part of the complex of digital resources. This control system has two levels of adjustment. At the administrator level, the time when the student was on the website and which resource he uses are automatically recorded.
At the teacher level, the system allows monitoring attendance at classes, completion of mandatory tasks and self- EDUCATIONAL CHALLENGES monitoring tasks. At this level, the system has the form of an electronic journal, in which the teacher assigns points for completed tasks according to the technological map. In addition, the rating of each student is automatically determined among those students assigned to the academic discipline (academic stream). In fig. 2 shows an example of a current control page in an electronic journal.
Figure 2
View of the own student' page of the electronic journal of the current monitoring of the success in the academic discipline Operations Research and Optimization Methods The left column of numbers corresponds to the points the student received for each current assignment, the center one is the range of points that can be obtained for that assignment, and the right one is the student's ranking among the total number of students in the stream ( fig. 2).
Every student in the group has access to the electronic journal. The presence of the electronic journal configured in this way gives impetus to the formation of external motivation for the student not only from the teacher's side, but also from the side of other students, as it creates conditions for open competition between students, healthy competition, which strengthens the personal desire to get a better grade. In addition, establishing a rating for each student personally contributes to the activation of his social desire to fulfill his obligations to his parents, in case the student studies under a contract, or seeks to receive a scholarship, which is assigned depending on the student's position in the rating.
Let us now consider the measures that were taken to develop students' internal motivation. Online (in synchronous mode), the educational process for students of both the experimental and control groups was built as follows. Students receive basic theoretical information during lectures, and practical skills are formed by studying typical examples and discussing situational tasks that took place on practices and labs. Also, students improve educationalchallenges.org.ua
EDUCATIONAL CHALLENGES
their knowledge during independent work thanks to the completion of a set of specially designed tasks.
These tasks involve independent mastering of mathematical methods for solving economic problems of various levels of complexity. The results of independent work were discussed on practices and labs. Since the ability to make responsible and informed decisions in the future professional activity depends on the desire to study independently, the question of finding effective factors influencing the formation of their internal motivation to study arises precisely at the stage of engaging students in active independent work.
Practical and laboratory classes for the 1st and 2nd (control) groups were conducted online in the usual format using traditional methods and techniques of teaching. Students of these groups were recommended printed publications (textbook, manuals, and methodical recommendations), electronic publications (similar to paper publications in pdf format) and other useful links as educational material that should be used during tasks for independent work.
They also had free access to the website of the discipline, where the full complex of digital resources was presented, and, accordingly, they could use all resources, but the use of interactive elements of these resources was not mandatory for them.
For students of the 3rd (experimental) group, practical and laboratory classes were conducted in the mode of active use of interactive technologies. For example, although lectures were given for the academic stream in a traditional format, along with this, mini-lectures of a debatable nature were held for students of the 3rd group in practical classes. These lectures were accompanied by specially designed audio presentations. Such minilectures were presented by the students themselves, while the teacher played the role of a moderator. Also, during the classes, an express survey was conducted in the format of closed-type or open-type interactive tests (Fig. 3).
Figure 3 An example of an open-type test task on the topic "Simplex method"
Interactive training exercises were developed to master mathematical methods and apply them to solving economic problems. They allowed checking all key points during the solution process, and in the case when a mistake was made, hints were automatically turned on that helped to find the right way . 4). The development of such exercises was carried out both in practical classes and during independent work based on the materials of electronic multimedia publications.
Figure 4
A fragment of a training exercise on the topic "Graphic method of solving linear programming problems" In the laboratory classes, which were held in the same format for all academic groups, applied economic problems were solved with the help of built-in functions and MS Excel add-ons. But homework in the control group additionally included performing interactive test tasks of a combined type, namely closed, open tasks and essays.
In order to perform the tasks of independent work, the students of the experimental group were recommended to additionally use the material of multimedia publications in the discipline Operations Research and Optimization Methods" in addition to the general educational literature. All students of the university have access to these digital resources, as they are placed not only on the teacher's website, but also in the repository of electronic publications of S. Kuznets KhNUE. But the realization of tasks according to these resources was mandatory only for the students of the experimental group.
The final control of acquired knowledge was carried out in the form of an exam. Students had to complete five tasks. Among them, there were two stereotypes, which demonstrate only the level of assimilation of basic concepts and the ability to use standard algorithms for solving; two diagnostic, which have a meaningful statement of the economic problem and require the construction of a mathematical model and the determination of the optimal solution; one heuristic task is an economic problem with real data, which involves choosing and justifying a solution method and checking alternative options. Students of all groups coped with the tasks (Fig. 5).
However, the students of the experimental group provided more complete and wellgrounded answers to heuristic tasks, demonstrated a clear understanding of the material, creativity, autonomy, and creative abilities, and obtained better results. educationalchallenges.org.ua
Figure 5
Evaluation diagram of each of the exam tasks In order to determine the attitude of students of the experimental group to study and their level of interest, students were asked to evaluate the digital resources in the context of their influence on the formation of the most important factors of internal motivation to study.
Among the factors proposed by the students, the following were highlighted: х1 -formation of creative cognition methods, х2 -acquisition of scientific and research skills in future professional activity, х3 -acquisition of new knowledge and practical skills in the academic discipline, х4 -realization of individual abilities, x5 -opportunities to show personal initiative, x6 -optimization of training time. Students, acting as experts, assigned ranks to each factor of internal motivation, based on their personal perception of their importance and implementation thanks to the active use of digital resources (Table 1).
EDUCATIONAL CHALLENGES
To form a generalized assessment of a group of experts, their answers were averaged, determining the average arithmetic ranks. So, in the opinion of students, the factors that contribute to the formation of internal motivation, in order of decreasing importance, can be presented in the form of a series: So, according to the students who studied using interactive technologies, the most influential factor of internal motivation is the acquisition of new knowledge and practical skills, that is, students are really interested in learning, and not just in evaluating their results from the teacher.
DISCUSSION
Thanks to the ability to determine the student's rating during the current control, the traditional system of success assessment is being modernized, reorientation of this system to increase the motivation in active and responsible learning. The competitive approach promotes the development of selfassessment skills as a means of selfdevelopment, creating conditions for students to plan individual educational trajectories. Such evaluation of students' success is a sufficiently effective factor not only of external, but also of internal motivation.
Thus, theoretical provisions (Domenico, & Ryan, 2017) regarding the organization of motivation levels were practically implemented. If for most students of the control groups, motivation was external, then thanks to the implementation of interactive technologies for students of the experimental group, motivation was internal, and its basis was the cognitive component that is consistent with thought (Schunk, Meece, & Pintrich, 2014).
It is internal motivation that is the driving force for increasing the effectiveness of training, as evidenced by the results of the current and final control. Multimedia technologies contributed to a better development of one's own educational space compared to digital resources discussed in the paper (Gurevich, Kademiya, & Shevchenko, 2012), and this created more comfortable conditions for learning.
Our experiment is an example of the development of separate theoretical propositions regarding the influence of information technologies on the educational process (Cole, Field, & Harris, 2004) and its development, respectively to modern requirements of society. Thus, it is confirmed that the digitalization of the educational process activates the theoretical-cognitive and research activity of students, which affects the formation of their ability to comprehensively analyze different views and to clear arguments in the course of substantiating their conclusions.
CONCLUSIONS
A comparison of the results of the assessment of students' academic performance with the results of their questionnaire regarding the effectiveness of educational technologies, which determine the formation of internal motivation in the conditions of distance learning, shows that the use of digital resources provides an opportunity for future economists and managers to successfully master the disciplines of mathematical direction.
The main factors that increase the effectiveness of education are the formation of a tendency to creative cognitive search, the acquisition of skills for analyzing and solving real problems of the economy, the disclosure of individual abilities, and the creation of a comfortable educational space. It should be noted that educationalchallenges.org.ua
EDUCATIONAL CHALLENGES
the experiment participants' awareness of their role and the teacher's increased attention to them are also additional factors of motivation.
In further research, it is planned to extend the use of digital resources to other mathematical disciplines, as well as to consider the use of such an element of interactivity as gamification in the analysis of real economic problems. | 2022-10-27T15:11:11.665Z | 2022-10-17T00:00:00.000 | {
"year": 2022,
"sha1": "0a1e907582faf78007baa2bae4beb65e039ae569",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.34142/2709-7986.2022.27.2.08",
"oa_status": "HYBRID",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "30ed6fe56cf57d6ddf0e8cd71ec27d21bf257534",
"s2fieldsofstudy": [
"Economics",
"Education",
"Computer Science",
"Mathematics"
],
"extfieldsofstudy": []
} |
125464601 | pes2o/s2orc | v3-fos-license | Bezier curve and its application
Description of the Bezier curve is presented. We explain in detail creation of the calculation algorithm together with the resulting program. It also includes drawing of the base functions of the Bernstein polynomials. Firstly, the procedure is applied to the theoretical example given by ten control points in a plane which approximate the Bezier curve. Secondly, the application in which we have given 138 points of trajectory of real vehicle. Points are located in space and we use them again for approximation of the smooth Bezier curve.
INTRODUCTION
The curves can be determined using control points, to which are usually added even further restrictions, such as boundary conditions.The control points are used either to interpolate the curve, when constructed smooth curve pass through all the given points, or to approximate curve when smooth curve pass only some selected control points or goes off these points 1.
Typical examples are the Lagrange interpolation, Hermite interpolation or Newton interpolation.The best known approximation method is the approximation method of the least squares.In this article we present approximation method using Bezier curve.
Bezier curve
Its name was given by French engineer Pierre Bezier, who worked at the French car factory Renault.A simple algorithm for creating Bezier curve constructed competitor's employee of Citroen Paul de Casteljau.Both designers have published their results in the sixties of the previous century 6.De Casteljau algorithm is based on the repeated use of linear interpolation and generalize the construction of parabolic curves for higher orders.Polygon was specified by four points in the plane as shown in the Figure 1 3.Similarly, we could have determined control points in space and the process would work analogously.
Bezier curve was expressed parametrically, the parameter t 0 , 1.Points of control polygon we denoted as 0-th approximation point of the curve (subscript represents the serial number of point and superscript approximation order) 0000 0 1 2 3 0 1 2 3 P , P , P , P P , P , P , P The first approximation is obtained from the zero approximation using relations And analogous continue the second and third approximation The point of the third approximation P 0 3 is the point of the curve for entered parameter value t.This procedure should be repeated for each value The parameter t appears at most in the cube, so it is a cubic Bezier curve.Repeat the procedure creates a triangular approximation scheme of successive points 22 01 3 0 P P P P P P P P P P P PP P P(t)
Numerical example for many control points
The process can be generalized to any number of control points of the polygon 4.When the polygon has 1 n points it is necessary to perform n steps to get the point of the curve.Bezier curve of degree n in the parametric representation has the form Tangent of the curve obtained at point P(t) is directly determined by points P 0 n-1 , P 1 n-1 , in the case of our four-point polygon (6) hence by points P 0 2 , P 1 2 .The tangent at the starting point of the curve is the same as the first control edge of the polygon and in like manner tangent in the last point of the curve is identical to the last control edge of the polygon.
Formula
is called the Bernstein polynomial of degree n.
With it looks simply parametric writing of Bezier curve .
Bernstein polynomials form a basis of the vector space for polygon degree at most n.They are unimodal (dromedary) functions with a single maximum at the point n i t .In Figure 2 we present the Bernstein polynomials for 9 n . .The number of points in the plane as well as the step parameter can vary.For planar curve we use an algorithm twice in both directions x and y.The selected control polygon and the resulting curve is shown in Figure 3.
Application of Beziér curve
The algorithm was used also in three dimensional space by application (10) in the directions x, y and z 5.We solved the real movement of the vehicle and record its position through 138 discrete points.These have served as control points for the Bezier curve.In Figure 4 we present a polygon, which was created by connecting points with line segments and Figure 5 shows the Bezier curve, which smoothed sequence of movement of the vehicle. | 2019-01-02T08:09:17.813Z | 2015-11-17T00:00:00.000 | {
"year": 2015,
"sha1": "f220dbe8e816c73cff211fb03863208473ea45f6",
"oa_license": "CCBY",
"oa_url": "http://meraa.uniag.sk/docs/2015-02/pales-redl.pdf",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "f220dbe8e816c73cff211fb03863208473ea45f6",
"s2fieldsofstudy": [
"Computer Science"
],
"extfieldsofstudy": [
"Mathematics"
]
} |
211128079 | pes2o/s2orc | v3-fos-license | Effects of inflammation on the kynurenine pathway in schizophrenia — a systematic review
Background In the last decade, there has been growing evidence that an interaction exists between inflammation and the kynurenine pathway in schizophrenia. Additionally, many authors found microglial activation in cases of schizophrenia due to inflammatory mechanisms related mostly to an increase of pro-inflammatory cytokines. In order to gain new insights into the pathophysiology of schizophrenia, it is important to incorporate the latest published evidence concerning inflammatory mechanisms and kynurenine metabolism. This systematic review aims to collect reliable recent findings within the last decade supporting such a theory. Methods A structured search of electronic databases was conducted for publications between 2008 and 2018 to identify eligible studies investigating patients with schizophrenia/psychosis and the relationship between inflammation and kynurenine pathway. Applicable studies were systematically scored using the NIH Quality Assessment Tools. Two researchers independently extracted data on diagnosis (psychosis/schizophrenia), inflammation, and kynurenine/tryptophan metabolites. Results Ten eligible articles were identified where seven studies assessed blood samples and three assessed cerebrospinal fluid in schizophrenic patients. Of these articles: Four investigated the relationship between immunoglobulins and the kynurenine pathway and found correlations between IgA-mediated responses and levels of tryptophan metabolites (i.e., kynurenine pathway). Five examined the correlation between cytokines and kynurenine metabolites where three showed a relationship between elevated IL-6, TNF-α concentrations, and the kynurenine pathway. Only one study discovered correlations between IL-8 and the kynurenine pathway. Two studies showed correlations with lower concentrations of IL-4 and the kynurenine pathway. Moreover, this systematic review did not find a significant correlation between CRP (n = 1 study), IFN-γ (n = 3 studies), and the kynurenine pathway in schizophrenia. Interpretation These results emphasize how different inflammatory markers can unbalance the tryptophan/kynurenine pathway in schizophrenia. Several tryptophan/kynurenine pathway metabolites are produced which can, in turn, underlie different psychotic and cognitive symptoms via neurotransmission modulation. However, due to heterogeneity and the shortage of eligible articles, they do not robustly converge to the same findings. Hence, we recommend further studies with larger sample sizes to elucidate the possible interactions between the various markers, their blood vs. CSF ratios, and their correlation with schizophrenia symptoms.
Introduction
Schizophrenia is a chronic disease characterized mostly by psychotic symptoms, cognitive impairment, and functional decline [1]. To date, the pathophysiology of this psychiatric condition remains unclear. However, research on inflammatory factors in schizophrenia has noticeably grown over the last decade [2].
In this context, the vulnerability-stress-inflammation model has been supported by growing evidence [3,4]. The model implies that genetic makeup predisposes the subject to be easily affected by stress, which would show inflammatory responses later in life [5]. A review by Lipner et al. has concluded that prenatal maternal stress (e.g., prenatal infection/inflammation, decreased fetal growth, hypoxia-related obstetric complications) has been linked to the development of schizophrenia in offspring [6]. Different animal models have indeed shown correlations with pro-inflammatory cytokines and schizophrenia [7][8][9][10][11]. In addition, authors support immune imbalance in schizophrenia, pointing out a predominance of pro-inflammatory mechanisms [12]. In this case, the presence of different pro-inflammatory metabolites and the inhibition of antiinflammatory factors (e.g., PGJ2) are demonstrated in the onset and chronic phases of schizophrenia [13][14][15].
From this standpoint, psychotic episodes in schizophrenia coincide with inflammatory mechanisms linked to the hypothalamic-pituitary stress-inflammatory pathways. This eventually leads to microglial and astrocyte activation [16,17]. Interestingly, the latter is also associated with the activation of the kynurenine pathway and increased the production of kynurenic acid (KYNA) in the cerebrospinal fluid (CSF) [18,19]. This metabolite is the only known natural N-methyl-D-aspartate (NMDA) receptor antagonist involved in inflammatory processes [20]. This results in the antagonism of the glutamatergic system, which in turn could lead to the dysregulation of dopaminergic neurons. Such changes can be attributed to inflammatory activation [21,22].
In the central nervous system (CNS), the kynurenine pathway starts by the conversion of tryptophan into kynurenine by indoleamine 2,3-dioxygenase 1 (IDO1), IDO2, or tryptophan 2,3-dioxygenase (TDO) (Fig. 1). Astrocytes can express both types of enzymes while microglia express only IDO [18,23]. To a lesser extent, some neurons also possess IDO and/or TDO producing a minor portion of kynurenine [18]. Therefore, kynurenine is available in the CNS via the enzymatic activity of astrocytes, microglia, and some neurons. As well, kynurenine is actively transported into the brain by the large neutral amino acid transporter [19].
Next, kynurenine can follow either of two metabolic branches. First, it can be metabolized into KYNA via kynurenine aminotransferase (KAT) [19,24,25] in astrocytes mainly [26] and neurons through irreversible transamination by KAT [27]. The other branch leads to the formation of quinolinic acid (QUIN) exclusively in both microglia and infiltrating macrophages. Both can express kynurenine 3-monooxygenase (KMO) which is absent in human astrocytes [28]. However, both astrocytes and neurons can further catabolize QUIN, produced by neighboring microglial cells, by the enzyme quinolinate phosphoribosyltransferase (QPRTase) [23]. They can also form the neuroprotective picolinic acid (PIC) as they express the enzyme aminocarboxymuconate semialdehyde decarboxylase (ACMSD) [27].
Considering the previously mentioned discoveries, it is important to integrate inflammatory mechanisms and kynurenine metabolism to gain further insights into the pathophysiology of schizophrenia (i.e., pro-inflammatory cytokines that stimulate the production of KYNA in schizophrenia) [19,[29][30][31]. In mood disorders, for example, there already exists evidence for correlations between both the kynurenine metabolic pathway and inflammation. One example is bipolar disorder, where microglia are shown to be overactive and levels of KYNA were elevated in the CSF [32][33][34]. Another example is depression, where patients showed decreased levels of KYNA, QUIN, and kynurenine, correlating with pro-inflammatory cytokines [35][36][37]. However, in schizophrenia, only a few reviews have focused on inflammation and the kynurenine pathway. One example is the review of Wang and Miller, which considered this correlation but in CSF only [37]. Ribeiro-Santos et al. focused on such a correlation concerning cognitive impairment in schizophrenia [38]. Despite the presence of articles where inflammation and the kynurenine pathway in schizophrenia are reviewed, studies that investigate a broader picture or systematic reviews-including serum and CSF for the inflammation markers and the kynurenine pathway-are needed. Indeed, an increasing line of evidence points to such a psycho-neuro-immunology related direction.
To this end, the main aim of this systematic review is to gather state-of-the-art findings in a structured way, based on original research done in the last 10 years on the relationship between inflammation and the kynurenine pathway in schizophrenic patients.
Study selection criteria
Studies of inflammatory cytokines and KYNA in schizophrenia were systematically searched between October and November 2018 using MEDLINE (PubMed, National Center for Biotechnology For the article selection criteria, we only considered original research articles. The search was limited to within the last 10 years (i.e., from 2008 to 2018). The inclusion criteria were as follows: (a) observational studies that assess peripheral blood or CSF inflammatory markers, (b) studies that assess schizophrenia/psychosis and kynurenine pathway metabolites, (c) studies with at least 20 subjects in the patient group with their respective matched control group, (d) studies applied in human species, and (e) studies published in the English language. Studies in human species that included any other kind of methodologies (such as interventional studies, meta-analyses, and reviews of any kind) were excluded. Also, animal experimental studies, studies outside the Fig. 1 Kynurenine pathway and tryptophan metabolism in the central nervous system. In the central nervous system (CNS), the kynurenine pathway starts by the conversion of tryptophan into kynurenine by indoleamine 2,3-dioxygenase 1 (IDO1), IDO2, or tryptophan 2,3-dioxygenase (TDO). Astrocytes can express both types of enzymes while microglia express only IDO [18,23]. To a lesser extent, some neurons also possess IDO and/or TDO producing a minor portion of kynurenine [18]. Therefore, kynurenine is available in the CNS via the enzymatic activity of astrocytes, microglia, and some neurons as well as the kynurenine being actively transported into the brain by the large neutral amino acid transporter [19]. Next, kynurenine can follow either of two metabolic branches. First, it can be metabolized into kynurenic acid (KYNA) via kynurenine aminotransferase (KAT) [19,24,25] in astrocytes mainly [26] and neurons through irreversible transamination by KAT [27]. The other branch leads to the formation of quinolinic acid (QUIN) exclusively in both microglia and infiltrating macrophages. Both can express kynurenine 3-monooxygenase (KMO) which is absent in human astrocytes [28]. However, both astrocytes and neurons can further catabolize QUIN, produced by neighboring microglial cells, by the enzyme quinolinate phosphoribosyltransferase (QPRTase) [23]. They can also form the neuroprotective picolinic acid (PIC) as they express the enzyme aminocarboxymuconate semialdehyde decarboxylase (ACMSD) [27]. Molecules: 3HAA, 3-hydroxyanthranilic; 3HK, 3-hydroxy-kynurenine; AA, anthranilic acid; ACMS, 2-amino-3-carboxymuconate semialdehyde; KYN, kynurenine; KYNA, kynurenic acid; NAD+, nicotinamide adenine dinucleotide; PIC, picolinic acid; QUIN, quinolinic acid; TRP, tryptophan; XA, xanthurenic acid. Enzymes: 3-HAO, 3-hydroxyanthranlic acid oxygenase; ACMSD, aminocarboxymuconate semialdehyde decarboxylase; IDO, indoleamine 2,3dioxygenase; KAT, kynurenine aminotransferase; KMO, kynurenine 3-monooxy-genase; KYNU, kynureninase; QPRT, quinolinic acid phosphoribosyltransferase; TDO, tryptophan-2,3-dioxygenase. *Excitation (examples): tryptophan, T-lymphocytes A4, IFN-α, IFN-β, IFN-γ, TNF α. *Inhibition (examples): IL-4, Th2 immunity response, antidepressants, antipsychotics defined time boundaries, studies with too few participants (< 20 patients), and papers not written in English were excluded.
The risk of bias for each included study was assessed according to the quality assessment tools issued by The National Heart, Lung, and Blood Institute (NHLBI). The selected studies are described and summarized in the "Results" section, as well as presented in tables. The process of selection is highlighted in a PRISMA-Chart (Fig. 2). The sociodemographic data and the study design characteristics for each study included were organized in the tables (Tables 1, 2, and 3). Table 4 shows the application of the NHLBI as quality control for the 10 included studies.
Literature search strategy
Two authors (BPP and OE) independently conducted a systematic literature search in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) statement [39].
Afterwards, the three categories were applied to the search field following this operation: Category 1 (one term) AND Category 2 (one term) AND Category 3 (one term). In this case, we chose one term from each category per each search operation. Hence, the search process resulted in the following Boolean expressions: "Psychosis AND immune response AND Kynurenine"; "Psychosis AND interleukins AND Kynurenine"; "Psychosis AND inflammation AND Kynurenine"; "Psychosis AND cytokines AND Kynurenine"; "Psychosis AND immune response AND Kynurenic acid"; "Psychosis AND interleukins AND Kynurenic acid"; "Psychosis AND inflammation AND Kynurenic acid"; "Psychosis AND cytokines AND Kynurenic acid"; "Schizophrenia AND immune response AND Kynurenine"; "Schizophrenia AND interleukins AND Kynurenine"; and "Schizophrenia AND inflammation AND Kynurenine".
Description of the studies
Out of 340 candidate publications, our search generated 10 publications, which were included for a qualitative synthesis (Fig. 2). All of the included studies were observational (n = 918). Eight followed a cross-sectional case-control methodology, whereas only two applied a prospective case-control longitudinal method [44,46]. The dataset comprised of 918 unique patients from 7 countries ( Table 2). The sociodemographic information regarding each study is reported in Table 1. The characteristics, study design, and main findings for each study are reported in Tables 2 and 3. Eight studies (80%) used the Diagnostic and Statistical Manual of Mental Disorders (DSM) criteria to define schizophrenia, and the other two (20%) considered International Classification of Diseases 10th Version (ICD-10) criteria. Two studies reported additional clinical evaluations by a board-certified psychiatrist involving expert judgment [48,49]. The patients' exclusion criteria included the following conditions: autoimmune disorders, comorbid psychiatric disorders (mostly substance abuse/dependence), learning disability, use of immunomodulatory non-psychotropic drugs (such as corticosteroids), acute infectious reactions or diseases, allergies, inflammatory response prior to the assessment, leukopenia, neurological disorders (meningoencephalitis, multiple sclerosis, vascular disorders, head injury, epilepsy, Parkinson's disease, Alzheimer's disease, and Huntington's disease), cancer, diabetes mellitus (type 1 and 2), cardiovascular disorder (such as vascular hypertension or past myocardial infarction), inflammatory bowel disease, chronic obstructive pulmonary disease, and finally use of antioxidants and omega 3 (ω3) polyunsaturated fatty acids.
Only 4 out of 10 studies described controls' exclusion criteria precluding the following conditions: personal or family history of psychiatric illness, autoimmune disease, substance abuse/dependence, medical conditions, or concomitant medications likely to influence CNS or immunological function including cardiovascular, respiratory, endocrine, and/or neurological diseases [45,46,48,49]. Only one study disregarded the exclusion of patients with comorbid medical conditions, arguing that their frequency is significantly elevated in a severely ill psychiatric population [49].
Authors used enzyme-linked immunosorbent assay (ELISA) for the assessment of inflammatory markers. In special conditions when high-sensitive (hs) CRP or some specific markers were measured, authors performed special methodology, for instance, immunoturbidimetry of CRP [49]. High-performance liquid chromatography (HLPC) was the standard method to measure kynurenine and tryptophan metabolites. Variations between studies exclusively depend on technical aspects as well as devices used for analysis.
To assess serum markers, fasting venous blood samples were collected in the morning. Afterwards, the blood samples were centrifuged and stored at − 70°C. Within the three studies analyzing CSF samples, the authors applied centrifugation, cryopreservation, and ELISA analysis. The intra-and inter-assay coefficients between studies were similar. Detection limits were reported and adequate for respective analyses.
Blood immunoglobulins and kynurenine pathway deregulation in schizophrenia
Three included studies approached immunoglobulins and the kynurenine pathway deregulation in the serum of schizophrenic patients. For the two of them, information was obtained from one large study divided between two publications. One study classified schizophrenic patients according to the Schedule for the Deficit Syndrome (SDS) into two groups: patients with deficit schizophrenia (n = 40) and those with non-deficit schizophrenia (n = 40). The authors tried to delineate the differences in tryptophan catabolites' (TRYCATs) profile between both groups in comparison with healthy controls (n = 40). Deficit schizophrenia was accompanied by an activated TRYCATs pathway when compared to controls and the non-deficit schizophrenia subgroup [41,42]. Moreover, both schizophrenia subgroups showed increased IgA responses to 3-OH-kynurenine as compared to controls while KYNA and anthranilic acid (AA) were relatively lowered in the combined schizophrenia groups versus controls with respect to increased levels of noxious TRYCATs [41,42].
The non-deficit schizophrenia patient subgroup versus healthy controls was characterized by relatively increased PIC but lowered QUIN levels [41,42].
Interestingly, deficit schizophrenia showed much higher IgA responses to PIC, xanthurenic acid (XA), and QUIN; an increased QUIN/KYNA ratio; relatively lowered AA and KYNA levels; and lowered IgA responses to both acids as compared to non-deficit schizophrenia [41,42].
The negative symptoms of schizophrenia are significantly and positively correlated with increased IgA responses directed against PIC and inversely with AA, whereas no significant associations between positive symptoms and IgA responses to TRYCATs were found [41,42].
Another study reported an association between physiosomatic (PS) symptoms and IgA-mediated response in patients with schizophrenia (n = 84) in comparison with controls (n = 40). PS symptoms, in this case, covers unexplained somatic symptoms (i.e., fatigue, muscle tension, flu-like malaise) related mostly with anxiety and depression. These PS symptoms, which were elevated in more than 50% of the patients, were significantly associated with IgA/IgM responses to TRYCATs, including increased IgA responses to 3 HK, PIC, and XA and lowered IgA responses to QUIN and AA [40].
CSF immunoglobulins and kynurenine pathway deregulation in schizophrenia
Bechter and colleagues [43] analyzed CSF samples from 63 hospitalized patients with schizophrenia (n = 39) and affective disorders (n = 24) compared with CSF samples from healthy controls (n = 4100). They recorded an intrathecal immune response in 9 patients (4 with schizophrenia) in addition to blood-brain barrier dysfunction in 18 patients (9 of them with schizophrenia). Neopterin (a human macrophage-produced molecule by IFN-γ stimulation) concentrations increased in 14 patients (8 of them with schizophrenia). However, such augmentation of pro-inflammatory reactions did not lead to significantly detectable changes in tryptophan or kynurenine metabolites [43].
Blood interleukins and kynurenine pathway dysregulation in schizophrenia
In patients with schizophrenia, the levels of T-helper (Th-) 1-specific IFN-γ and TNF-α, and Th2-related IL-6 were significantly higher in the plasma of the schizophrenic group (n = 71) compared to controls (n = 174) while mean plasma tryptophan concentration of the schizophrenic patients was significantly lower in contrast with controls. Moreover, the authors found that Th1related IL-2 and Th2-specific IL-4 were significantly lower in patients' plasma. However, plasma transforming growth factor (TGF-) β1 did not show statistically significant difference even though the mean value tended to be higher in schizophrenic patients [46].
In addition, after 6 weeks of treatment with four different atypical antipsychotics (risperidone, amisulpride, olanzapine, and aripiprazole), patients with schizophrenia exhibited a significant reduction of plasma IL-6 and TNF-α [23]. Interestingly, both mean plasma tryptophan and kynurenine levels were significantly increased after 6 weeks of antipsychotic treatment.
According to a study conducted in patients with chronic schizophrenia (n = 51), compared with healthy control subjects (n = 45), the KYNA/3-HK ratio and IL-4 levels, but not soluble IL-2 receptor (sIL-2R) and IFN-α levels, were consistently decreased in schizophrenia patients at all analyzed time points (at time of admission, after a 4-week treatment and remission). Additionally, sIL-2R levels were positively correlated with the number of relapses before treatment [44].
The same study reported that in a subgroup of patients with poor response to pharmacotherapy, initial KYNA levels and the KYNA/3-HK ratio negatively correlated with the number of relapses. Positive association of sIL-2R levels with the number of relapses was also evident in this subgroup; those were treated 4 weeks later with clozapine. Finally, a negative correlation between IFN-α and 3-HK level (r = − 0.875) as well as a positive correlation between KYNA and IFN-α (r = 0.472) was shown [44].
On the other hand, one out of the 10 studies systematically reviewed here disclosed no statistically significant correlation between kynurenine pathway dysfunction and the presence of interleukins [45]. Despite a significantly higher KYN/tryptophan ratio in patients compared to healthy controls (patient, 0.055 ± 0.003; control, 0.048 ± 0.002; t = 2.02, df = 68, p < 0.05), no statistically significant relationships were revealed between IFN-γ values and the KYN/ tryptophan ratio among the groups (p = 0.23). The differences of this ratio could not be associated with concomitant IFN-γ elevation. Therefore, this study suggests that this increased activity is not IDO mediated [29][30][31]45].
CSF interleukins and kynurenine pathway dysregulation in schizophrenia
The potential role of inflammation and kynurenic acid in psychotic disorders (mostly schizophrenia) was also investigated in twins [48]. Kegel and colleagues discovered a significant association between psychotic symptoms, interleukins, and kynurenine pathway metabolites (measured mostly in CSF) in Swedish-born, same-sex twins discordant for schizophrenia and other psychotic disorders [48]. In detail, they presented a relationship between IL-8 and QUIN (estimate 1.20, standard error [SE] 0.19, t = 6.49, p = 0.0001) and TNF-α and QUIN (estimate 155, SE 29.7, t = 5.21, p = 0.0006). By analyzing the intra-pair differences between twins, the authors have revealed associations between kynurenine metabolites and cytokines (QUIN, IL-8, and TNF-α) in complete twin pairs [48].
Another study by Schwieler and collaborators detected a significant correlation between IL-6 and KYNAproduction in patients with schizophrenia reflected by the tryptophan to KYNA ratio (r = − 0.49; p = 0.024) [47].
Serum KYNA and/or KYNA/QUIN were significantly reduced in all patient subgroups versus healthy controls except for the schizophrenia subgroup, which did not differ from healthy controls [49].
Moreover, a post hoc comparison of patients divided into the categories of non-psychotic affective disorder, affective psychosis, and psychotic disorder (non-affective) manifested reduced levels of KYNA in both the affective disorder and affective psychosis groups. KYNA/3HK was significantly reduced in affective psychosis in comparison to non-psychotic affective disorder and in affective psychosis when compared to healthy controls. Similarly, KYNA/ QUIN differed across groups and was significantly reduced in the affective disorder and affective psychosis groups compared with healthy controls. However, no significant difference was shown for patients with schizophrenia.
Follow-up t tests showed that healthy controls had significantly lower CRP concentrations or a statistical trend towards lower CRP concentrations than the MDD, BD, and SZA subgroups. Nonetheless, this was not the case of the patient subgroup with schizophrenia [49].
Discussion
The results of this study highlight correlations between inflammation, the kynurenine pathway, and schizophrenia in the available literature of human research from the last 10 years. Although few obtained studies related to the specific topic and methodological differences existed between studies, especially concerning sample collection (plasma or CSF), correlations were found between the three mentioned variables.
In regard to studies with only serum samples, correlations between [IFN-γ]/[IL-4] and [kynurenine]/[tryptophan] were found in schizophrenia. Other studies with schizophrenic patients reported correlations between IFN-α and 3-HK levels, as well as correlations between KYNA and IFN-α levels. Additionally, other selected studies found correlations between IgA-mediated responses with levels of tryptophan metabolites (i.e., kynurenine pathway) in schizophrenic patients. Nevertheless, we did not find evidence for a relationship between CRP and tryptophan metabolite levels in patients with schizophrenia.
Regarding studies with only CSF samples, correlations between IL-8 and QUIN, TNF-α and QUIN, and IL-6 and KYNA production were also found in schizophrenia. However, one study demonstrated augmented proinflammatory reactions but without detection of tryptophan metabolites using CSF-cytometry of schizophrenic patients. Figure 1 helps to give a summary of the principal findings on the inflammation effects on the kynurenine pathway in schizophrenia.
Cytokines, kynurenine pathway, and schizophrenia
Since the discovery of cytokines, our knowledge of signal transmission between cells has been dramatically expanded [30,50]. Cytokines have been initially described in systemic inflammatory diseases. Newer evidence suggests that they are also related to other disorders like schizophrenia. The results of the present systematic review revealed the following cytokines as possible candidates for deregulatory mechanisms on the kynurenine pathway in schizophrenia: TNF-α, IFN-γ, IL-6, IL-4, IL-8, and IFN-α.
TNF-α
The results showed a correlation between higher TNF-α values and augmented QUIN concentrations in schizophrenia. TNF-α is known for its effect on microglial cells, generating a strong diffuse pro-inflammatory response in the CNS [51]. In addition to that, authors have postulated in the last years that TNF-α modifies the metabolic pathway of tryptophan producing higher quantities of neuroactive substances (i.e., KYNA and QUIN) in the CNS [29][30][31]. The result of an unbalanced QUIN production could eventually explain cognitive impairment disorders, a clinical entity closely connected to schizophrenia [4,52,53].
IFN-γ
The obtained results from this review between IFN-γ and kynurenine pathway were ambiguous. It is known that IFN-γ activates Th1 cells promoting cytotoxic cell activity and immune responses. Tryptophan catabolism towards KYN formation is mostly mediated by IDO, which is believed to be activated by an increased concentration of IFN-γ [54]. In schizophrenia, authors have also postulated an effect of IFN-γ on the blockade of NMDA and alpha-7-n-acetylcholinergic receptors through an increase of KYNA concentrations [55,56]. Moreover, different studies relate IFN-γ with an increase of 3-HK in patients with schizophrenia [57][58][59]. However, the studies from this systematic review reported two contradictory results. On the one hand, one included study evidenced significant correlations between IFN-γ/IL-4 in plasma and tryptophan breakdown ratio ([kynurenine]/ [tryptophan]) in schizophrenia. On the other hand, another included study showed no significant relationships between IFN-γ values and the kynurenine/tryptophan ratio. Based on the post-mortem study findings from Miller et al., we suggest that IDO expression does not respond to a concomitant augmentation of IFN-γ in the CNS and that IFN-γ eventually has a minor effect on the kynurenine pathway in schizophrenia [60].
IFN-α
The results from this systematic review pointed out a negative correlation between IFN-α and 3-HK but a positive one with KYNA in schizophrenia. Known functions of IFN-α include the stimulation of macrophage antibody-dependent cytotoxicity [51]. Microglia are phagocytic, macrophage-like cells in the CNS and represent the main portion of active innate immune defense in the CNS [61]. Both overactivation and administration of IFN-α induce higher production of KYNA, which can be neurotoxic. Although this effect was published for depressive patients, the augmented concentration of IFN-α was shown to unbalance neurophysiological mechanisms also in subjects with schizophrenia [62].
IL-6
The results of this study found that elevated IL-6 levels correlated with higher KYNA values in schizophrenia. This cytokine exerts an effect on different proinflammatory mechanisms in microglial cells [50,51] like microglia priming-i.e., the transition from resting to activated state [63]; regulation of the inflammatory response [64]; and overproduction of IL-6 in the aged brain [65]. Due to these microglial functions, authors pointed out that IL-6 was associated with KYNA production by increasing this metabolite in patients with schizophrenia [19]. Other major functions of IL-6 are also the synthesis and promotion of cortisol by hypothalamic stimulation [66]. While such IL-6-action may be related to some other psychiatric disorders like unipolar depression [67], elevation of cortisol production by IL-6-action could also be linked to schizophrenia [68][69][70]. We suggest that this increased production of cortisol by IL-6-action could be provoked due to two factors: the first one, the exposure to physical or environmental stress, and the second one, genetic vulnerability (i.e., the interaction between these stressors and risk alleles [68]). Both factors, i.e., stress and genetic vulnerability, are components of the mentioned vulnerability-stress model [4,71,72].
IL-8
The present review shows that this chemokine of the CXC family is related to the kynurenine pathway in schizophrenia. IL-8 contributes to activation and recruitment of neutrophil granulocytes but was also shown to increase QUIN production, as different Alzheimer's disease studies in human post-mortem and serum samples have shown [50,52]. IL-8-induced QUIN induction may be involved in microglial and neuronal activation/damage in the hippocampus eventually contributing to cognitive dysfunction [52]. Although supporting data has been obtained in models designed for Alzheimer's disease research, overproduction of IL-8 and its relation with schizophrenia could also explain cognitive impairment symptoms in patients with schizophrenia [52]. Finally, in vitro studies demonstrated that the combination of TNF-α/IFN-γ promotes an augmentation of IL-8 in human astrocytes [73,74]. Nevertheless, the present review found no significant correlations with IFN-γ and positive correlations with TNF-α and kynurenine pathway in patients with schizophrenia.
IL-4
Only one cytokine from the γ-chain-cytokine family, namely IL-4, was related to the kynurenine pathway in schizophrenia. It is known that this cytokine is associated with Th2 responses, but it also suppresses KYNA production through blockade of the KAT2 [50,75]. The results of the present systematic review highlight that patients with schizophrenia seem to show lower plasma concentration of IL-4. This might lead to an imbalance due to a lower Th2 response and could eventually explain the augmentation of KYNA production in schizophrenia, as previously shown in an animal model for Alzheimer's disease [75].
Interestingly, these results contradict the hypothesis of a predominant Th2 response in schizophrenia and rather support a shift of the Th cells towards a Th1 response [72,76].
Immunoglobulins, kynurenine pathway, and schizophrenia
The present review also shows a relationship between IgA-mediated responses and schizophrenia. This antibody plays a crucial role in the immune function of mucous membranes mostly in the gastrointestinal tract providing protection against microbes [77]. Our assessed studies revealed strong correlations between an IgA-mediated response and tryptophan metabolites, especially KYNA and QUIN. Thus, the inflammatory phenomenon most likely unbalances the tryptophan metabolism subsequently contributing to psychotic symptoms.
Negative findings: CRP and CSF-cytometry influence on the kynurenine pathway and schizophrenia Our review did not show any significant correlation between CRP and kynurenine metabolism. However, other studies have reported an association between CRP levels and schizophrenia severity [78]. While cytometry of CSF revealed a macrophage-induced pro-inflammatory reaction (i.e., increased neopterin levels) in patients with schizophrenia, a relationship with KYNA or tryptophan metabolites was not found.
Cellular substrates regulating interactions between kynurenine pathway and inflammation Cellular regulation of kynurenine pathway
Overactivation of astrocytes in schizophrenia was shown to be independent of the patients' medication state [79]. This increased activity of astrocytes in schizophrenia and the persistent reduction of microglial KMO activity result in increased kynurenic acid formation in astrocytes [80]. This finding is consistent with CSF investigation by Schwieler et al. 2015 that showed a positive correlation between IL-6 and KYNA. However, Szymona et al. found decreased KYNA/3-HK ratio in the blood. This could be explained by the KYNA inability to cross the blood-brain barrier.
It is noteworthy that the activation of astrocytes may further facilitate the catabolism of QUIN by QPRTase.
However, a recent review postulated that in some cases, kynurenine pathway metabolism is likely to be rerouted from KYNA to QUIN production in brain tissue in states of inflammation with microglial activation and macrophage infiltration [81]. Microglial activation was only found in a small percentage of schizophrenia patients and seems to be more pronounced in chronic schizophrenia patients [79]. This might underlie the activation of QUIN pathway in CSF study by Kegel et al. 2017 [48].
Kynurenine pathway and inflammation
The interaction between intracellular kynurenine pathway and the inflammatory process seems to be bidirectional and more complicated. The first direction is demonstrated by several previous studies showing that inflammatory cytokines can dramatically affect the astrocyte and microglial activity and thus the production of KYNA and QUIN. For example, in vitro studies reveal enhanced IDO-messenger RNA-expression in microglia following IFN-γ stimulation, and to a lesser extent in astrocytes [23]. Both IFN-γ and TNF-α increase also the production of QUIN by microglia and macrophages [23,82]. These effects could explain the positive findings found in blood [44][45][46] and the correlation between TNF-α and QUIN in CSF [48].
In contrast, the opposite direction implies that some kynurenine metabolites can modulate the inflammatory processes. At excitotoxic concentrations, QUIN induces the expression of several pro-inflammatory cytokines and chemokines such as IL-1β, monocyte chemotactic protein 1, and IL-8 in astrocytes [83]. KYNA, an NMDA receptor antagonist, has been shown to act as an immunosuppressant in mice. Such antagonism caused Th1 effector cells to produce less IL-2 and IFN-γ, whereas it affected Th2 cells to produce more interleukin (IL-) 10 and IL-13. As proposed by a recent review, this interaction suggests a kynurenine pathway-immune feedback loop that may be disrupted in schizophrenia [81]. We suggest that the competition between the pro-inflammatory QUIN (an NMDA receptor agonist produced by microglia) and the immunosuppressant KYNA (an NMDA receptor antagonist) may determine the inflammatory status and/or the clinical picture of certain psychiatric diseases related with neuroinflammation.
Intercellular interactions
Regarding the intercellular communication, astrocytes produce neuropeptide Y which inhibits microglial activity.
In the light of our results, in schizophrenic patients, such equilibrium between type 1 and type 2 response in the CNS is disturbed, as showed in the study of Kim et al. [46] and included in this review. It seems that there is an imbalance between the activation of microglial cells and astrocytes [79]. This particular unbalance of the intercellular communication between astrocytes (producing neuroprotective KYNA and PIC) and microglia (producing neurotoxic QUIN) could suggest a competitive interaction between both of them, where the different activation patterns of microglial cells and astrocytes may underlie the biological variations among schizophrenic states.
Limitations and future directions Limitations
Some limitations have to be taken into account for the present systematic review. First of all, to fulfill our primary aim to find studies assessing the correlation between the "kynurenine pathway," "inflammation," and "schizophrenia," the used Booleans restricted the results; we could not find many studies with other biomarkers, such as enzymes. Nevertheless, this approach enabled us to specifically address the correlations between the three-targeted variables. Second, the studies included in our sample, although they delivered mostly significant results, appeared to be heterogeneous regarding the methodology and patient selection criteria. Third, the definition of schizophrenia varied between studies as it was amenable to the subjectivity of clinical inspection. Independently of the similarities between ICD and DSM classification, standardized protocols to select patients with schizophrenia do not exist and, therefore, cannot guarantee high homogeneity of samples between studies. Finally, some studies did not pay attention to some confounding factors or variables such as duration of illness, doses of antipsychotic drugs, symptom severity, or smoking potentially influencing some of the reviewed results here.
Future directions
This current article focuses mostly on the influence of inflammation in the tryptophan metabolism in human subjects with schizophrenia. Due to the lack of data on this matter, future studies that investigate the influence of inflammation in the CNS especially related to the kynurenine pathway in schizophrenia are warranted. To integrate different pathways and mechanisms will allow for the possible identification of therapeutic targets and interventions for handling and preventing relapses. While there are differences between studies, the literature overall suggests significant correlations between these three variables.
These findings raise many questions, such as "what are the most important metabolites that relay inflammation or inflammatory markers and the kynurenine pathway in schizophrenia?" For example, in an animal model of pregnant rats, the deletion of one allele for KMO (KMO +/− ) in the maternal genetic material brought to a disproportionate KYNA increase in the brain of KMO +/− offspring rats, compared to KMO wild-type offspring rats (KMO +/+ ) [84]. As mentioned before, in subjects with schizophrenia, the overproduction of KYNA contributes mainly to an imbalance of Th1/Th2 responses. Additionally, in other mouse schizophrenia model, the dysregulation of KYNA explained different cognitive impairments such as anticipatory responses and reduced locomotor activity [85]. The search for understanding this mechanism also in humans, mostly in patients with schizophrenia, will allow to conduct future studies that could regulate the KMO-genetic expression and finally to avoid the imbalance between KNYA and QUIN mechanisms, as showed in patients with schizophrenia.
Additionally, the regulation of NMDA receptors could also have positive effects on schizophrenia, since in the case of schizophrenia KYNA is overexpressed and involved in the receptor antagonism, as well as in the imbalance of inflammatory responses. An experimental model in mice has shown that the regulation of NMDA receptors through KAT2 inhibitors attenuates the KYNA response in the pre-frontal cortex [86]. The prevention of KYNA and the regulation of glutamate release could contribute to the treatment of schizophrenia, mostly to cognitive deficits associated with this mental disorder [86]. Additionally, the regulation of the enzymes related to the kynurenine pathway has suggested to have positive results as an immunomodulatory therapy [87]. Although this data is shown in other disease models, such as depression [87], the stabilization of immune responses in schizophrenia could also contribute to a posterior regulation of noxious effects of the kynurenine pathway, as described before.
What actual therapeutic choices/options could we have in schizophrenia, where a predominant inflammatory factor alters the homeostasis of nerve cells? Currently and to the best of our knowledge, there is no evidence supporting a specific and direct therapeutic option that could regulate the effects of pro-inflammatory cytokines in the kynurenine pathway. What we already know is that some antipsychotics can partially modulate the immune imbalance and the overproduction of KYNA, a natural NMDA receptor antagonist [88]. According to the experimental studies in mice, it is shown that olanzapine regulates the activation of IDO [89]. Nevertheless, the only information we have is related to depression and models for therapeutic targets in schizophrenia regarding this matter are needed. Finally, antiinflammatory drugs (i.e., non-steroid anti-inflammatory drugs) as add-on therapy to antipsychotic therapy have shown regulating effects on psychotic symptomatology, compared to antipsychotic therapy only [88,90]. However, the detailed pharmacological effects of the add-on anti-inflammatory therapy in schizophrenia remain unknown, and most evidence regarding this topic is inconclusive and, due to the amount of included studies, limited [91]. In this case, more studies are needed to comprehend how does the pharmacology of antiinflammatory drugs work and to understand if these drugs, by inhibiting inflammatory agents, could indirectly or directly influence the kynurenine pathway in schizophrenia.
Conclusions
In conclusion, the present systematic review found a relationship of elevated IL-6, IL-8, and TNF-α concentrations with the kynurenine pathway in schizophrenia. These higher values could be an explanation for the psychotic symptomatology and cognitive disturbances. However, this systematic review did not find a correlation between CRP, CSF-cytometry, IFN-γ, and the kynurenine pathway in schizophrenia. With respect to the Th response and lower IL-4 concentrations, we conclude that overactivation of the kynurenine pathway is mostly related to a reduced Th2 and augmented Th1 response. Overall, we conclude that there is a strong immunoglobulin-mediated response against metabolites of the kynurenine pathway, contributing as well to inflammatory mechanisms.
While different methodologies have been applied in the included studies and the results show some heterogeneity, our systematic review gives support to a picture of schizophrenia that integrates inflammatory mechanisms and nerve cell physiology. To expand the understanding of this matter, further studies should be performed. | 2020-02-17T15:39:06.671Z | 2020-02-15T00:00:00.000 | {
"year": 2020,
"sha1": "9ec5421d94d1a5cb69bfe06f48924fccff58946c",
"oa_license": "CCBY",
"oa_url": "https://jneuroinflammation.biomedcentral.com/track/pdf/10.1186/s12974-020-1721-z",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "9ec5421d94d1a5cb69bfe06f48924fccff58946c",
"s2fieldsofstudy": [
"Psychology",
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
214550745 | pes2o/s2orc | v3-fos-license | Lactic acid bacteria with antimicrobial, proteolytic and lipolytic activities isolated from ovine dairy products
This study aimed to isolate lactic acid bacteria from sheep milk products and to characterize these microorganisms with a focus on their antimicrobial, proteolytic and lipolytic activities. Raw milk, pasteurized milk, pasteurized cream, and butter samples were collected, lactic acid bacteria were isolated and their proteolytic, lipolytic and antimicrobial activities were evaluated. Lactic acid bacteria counts were higher in raw milk collected at the farm number 4 (8.91 ± 0.05 log CFU/mL), in which adequate hygienic practices were observed for pre- and post-milking. A total of 253 isolates were obtained, and among them 37 were lactic acid bacteria, where 19 showed some type of activity, most of which from raw milk. Among the isolates of lactic acid bacteria, 48.65% (n=18) showed proteolytic activity, 13.51% (n=5) lipolytic activity, 10.81% (n=4) showed both proteolytic and lipolytic activities, and only 2.70% (n=1) showed antimicrobial activity. Osolation of lactic acid bacteria with technological properties demonstrated their potential use as starter cultures in processing of fermented dairy products.
Introduction
Milk is a source of lipids, proteins and carbohydrates despite their variable composition depending on the animal species (Food and Agriculture Drganization of the United Nations, 2013). Sheep milk has higher total solids, fat, and protein contents than cow and goat milk, obtaining greater yield in the production of cheese, yogurt, ice cream and dairy beverages (Balthazar et al., 2017(Balthazar et al., , 2018a Food and Agriculture Drganization of the United Nations, 2013; Merlin et al., 2015). The American continent produces 42,095 tons of the world´s production of sheep milk, and Brazil contributes to 21.72% of this total (Food and Agriculture Drganization of the United Nations, 2017).
On the Southern and South-eastern regions of Brazil, there is a large production of sheep milk that is processed in registered dairy plants or even with homemade processing (Santos et al., 2016). This activity increased its participation in the Brazilian agribusiness, with great potential for cheese production, which are highly valued in the Brazilian market (Santos et al., 2016). Sheep dairy farming is considered a productive system well adapted to small farms, but this economic activity has not yet reached its full capacity in Brazil, especially regarding dairy products such as cheese and yogurt (Balthazar et al., 2017;Santos et al., 2016).
From the microbiological point of view, raw milk kept under refrigeration may have many different species of bacteria, some with pathogenic or deteriorating capacity (Campagnollo et al., 2018;Forsythe, 2013). Raw milk is considered a good source of lactic acid bacteria (LAB), which can compose cultures of starter bacteria, usually added to raw milk to produce fermented foods (Luiz et al., 2017;. LAB is a group of Gram-positive, non-spore-forming, catalase-negative microorganisms that can grow under microaerophilic or strictly anaerobic conditions (Bruno, 2011). The positive influence of LAB on cheeses is due to the development of sensorial characteristics, through biochemical reactions such as proteolysis and lipolysis during maturation (Asensio-Vegas et al., 2016;Bruno & Carvalho, 2009;Luiz et al., 2017;Farahani et al., 2017). The proteolytic enzymes are the main ones found in LAB and its lipolytic activity is also widely reported (Bruno & Carvalho, 2009;Farahani et al., 2017). On addition, some strains of LAB can synthesize antimicrobial compounds (Balthazar et al., 2018b;Campagnollo et al., 2018;Mechai et al., 2014), and these characteristics make them useful to ferment milk, as primary agents or as culture starters (Bruno & Carvalho, 2009;Campagnollo et al., 2018;.
Considering the importance and the expansion of Brazilian dairy sheep, this study aimed to isolate LAB from raw sheep milk Lactic acid bacteria with antimicrobial, proteolytic and lipolytic activities isolated from ovine dairy products and derivatives from Southern Brazil and to characterize these microorganisms with a focus on their antimicrobial, proteolytic and lipolytic activities.
Materials and methods
Samples were obtained from a dairy company specialized in the processing of sheep milk located in the South of Brazil (29° 10' 17" S and 51° 31' 09" W) from June to September 2016. Cooled raw milk with a fat content of 6.5%, pasteurized milk at 72-75 °C/15-20s, pasteurized cream and butter with 1.5% of salt were collected. The samples of pasteurized milk and cream were collected at the outlet of the pasteurizer and butter was collected immediately after processing. Raw milk was collected in the dairy refrigeration tank and in four sheep dairy farms that supply milk to the dairy company, identified as farms 1 to 4. Ot was verified whether farmers followed milking good practices as recommended (Rosa et al., 2014). All samples were packed in sterile containers and transported under refrigeration to the laboratory.
Onitial dilution was prepared for each sample using 25 mL or 25 g and 225 mL of 0.1% peptone water followed by homogenization in a Stomacher-type equipment. Serial dilutions were made up to 10 -6 and 100 μL of each dilution was inoculated into Man Rogosa and Sharpe agar (MRS, Merck ). The incubation was carried out in an anaerobic tank containing anaerobic generator at 37 ± 1 °C for 48 to 72 hours (Silva et al., 2013). After this period, the colonies were counted. Colonies present in the MRS agar were transferred by scoring to MRS agar plates and, subsequently, catalase and Gram staining tests were performed. The catalase test was performed on a glass slide with the addition of one drop of 3% hydrogen peroxide solution to each of the collected colonies, being considered positive when bubbles were formed (Brasil, 2018;Silva et al., 2013). Gram staining was performed according to the standard procedure (Brasil, 2018;Bruno, 2011). Selected isolates were transferred to MRS agar and submitted to tests to verify the fermentative profile and the proteolytic, lipolytic and antimicrobial activities.
The fermentative profile was evaluated in Gram-positive and catalase-negative isolates to determine the production of carbon dioxide (CD 2 ) from the use of glucose. The isolates were cultured in MRS agar at 37 ± 1 °C for 24 hours and then re-inoculated in MRS broth (Merck ) supplemented with 3% glucose (10% v/v) in test tubes containing inverted Durham's tube, followed by the incubation at 37 °C for the 48 hours (Sehn, 2015;Silva et al., 2013). The tubes showing carbon dioxide production by the formation of gas and turbidity in the medium were classified as heterofermentative, and those with turbidity due to the presence of lactic acid were The proteolytic activity of the isolates was evaluated in formulated Milk Agar, with subsequent incubation at 28 °C for 24 to 48 hours. This medium was supplemented with 10% skimmed milk powder (Nestlé ) Tebaldi et al., 2008). The lipolytic activity was evaluated using the Tributyrin agar (Biolog ) incubated at 28 °C for five days. The presence of clear zones around the colonies in Milk Agar using 1% (v/v) of HCl for 1 min was indicative of proteolytic activity. The presence of clear zones around the colonies in Tributyrin Agar was considered lipolytic activity. The proteolytic and lipolytic activities were classified according to halo sizesI: very high activity for halo greater than 10 mm, high activity for halo of 3 to 10 mm, and low activity for halo lower than 3 mm (Alapont et al., 2015). All tests were performed in duplicate.
The antimicrobial activity of the isolates was evaluated using the agar diffusion method (Biscola et al., 2013). The isolates were cultured in MRS broth at 37 °C for 18 hours and then 5 μL of the culture was added in Brain Heart Onfusion (BHO, Himedia ) previously inoculated with the 5 log CFU/mL of the indicator microorganismsI: Listeria monocytogenes ATCC 7644, Escherichia coli D157I:H7 ATCC 43895 and Staphylococcus aureus ATCC 25923, followed by incubation at 37 °C for 18 hours. These strains were obtained from the collection of the Department of Microbiology, Ommunology and Parasitology, Onstitute of Basic Health Sciences, UFRGS. After this time, the plates were observed for the presence/absence of inhibition halos, and the diameters of them were measured. Results were classified based on halo diameterI: strong inhibition for halos greater than 8 mm, moderate inhibition with halos between 4 and 8 mm, and weak inhibition for halos of 1 and 4 mm (Akabanda et al., 2014). The assay was performed in triplicate.
Results
Twenty-five samples of sheep milk products were collected including raw milk (n=10), pasteurized milk (n=3), pasteurized milk cream (n=8), and butter (n=4). The average count in MRS agar for sheep milk obtained from the refrigerated storage tank and dairy farms are shown in Figure 1. After 72 hours of growth, the milk of the tank showed LAB counts of 7.35 ± 0.11 log CFU/mL, while the mean for milk from the farms was 6.50 ± 1.56 log CFU/mL. On Figure 1, it is possible to observe the quantification of LAB in the milk collected in each of the dairy farms that supplied milk to the dairy plant, with emphasis on counts of farm 4 (8.91 ± 0.05 log CFU/mL). As observed during sampling of raw milk, farm 4 was the only one performing pre-dipping and post-dipping treatments, with the use of lactic acid foam and iodine gel, respectively. The other producers performed only one, or none of these steps. The LAB counts per processed sheep milk and derivatives is shown in Figure 2 where butter (n=4) showed LAB counts of 5.85 ± 0.01 log CFU/mL, higher than that of milk cream (n=8) with 3.58 ± 0.60 log CFU/mL, which is the raw material for this product. Table 1 shows the selection of LAB isolates in dairy products of ovine origin, in each of the isolation steps. Among the 253 colonies selected from MRS agar, 37 (14.62%) were catalase-negative and Gram-positive and the main source was raw milk, with 30 of these isolates. This fact is expected since this product is unpasteurized and a larger number of sample (n=10) were collected. Fermentation profile was evaluated to characterize 37 isolates catalase-negative and Gram-positive, and 86.48% were homofermentative type, with the highest number of isolates obtained from raw refrigerated milk.
On Table 1, the proteolytic, lipolytic and antimicrobial activities of selected isolates can also be verified. Antimicrobial activity was observed against only one microorganism tested (Listeria monocytogenes ATCC 7644). Raw milk was the source of most isolates, where 13 LAB showed proteolytic activity, three with lipolytic activity, and one with antimicrobial activity. Proteolytic activity was also observed in three isolates from the cream and in two isolates of pasteurized milk. Considering the values of initial LAB screening in this study (n=37), 48.65% (n=18) showed proteolytic activity, 13.51% (n=5) lipolytic, 10.81% both proteolytic and (n=4) and 2.70% (n=1) antimicrobial. Table 2 shows the proteolytic, lipolytic and antimicrobial activities of each LAB of sheep milk and derivatives of Southern Brazil. Df the total LAB isolated from raw milk, pasteurized milk, pasteurized cream, and butter, only 51.35% (n=19) had any of proteolytic, lipolytic and/or antimicrobial activity against Listeria monocytogenes ATCC 7644. Proteolytic activity ranged from 1.5 to 23 mm representing 48.65% (n=18) of the LAB isolates, where 37.84% (n=14) showed very high proteolytic activity (halo size > 10mm). Lipolytic activity varied from 1 to 23 mm in 13.51% (n=5) of the isolated LAB, and among them 8.11% (n=3) were classified with very high lipolytic activity (halo size > 10mm).
Among the LAB obtained in the present study with proteolytic and lipolytic activities, we highlight LAB LSE 05-21 obtained from raw milk of farm number 4, and LAB LSE 01-1 isolated from the pasteurized milk both with average halos above 10 mm and coccus morphology. Dnly LAB LSE 08-23 showed strong antimicrobial activity against Listeria monocytogenes ATCC 7644, with a mean halo size of 13.2 mm, in addition to high lipolytic and very high proteolytic actions.
Discussion
LAB counts obtained in this study were higher than those obtained in raw sheep milk, produced in the metropolitan region of Porto Alegre city, South of Brazil with values of 5.65 ± 0.53 log CFU/mL . On a study carried out in Tunisia with bovine and camel milk, mean values were 2 log CFU/mL (Zeineb et al., 2013). Goat milk produced in the Centre-West region of Brazil, 52.5% (n=21) did not have LAB, and the highest values were 3 to 4 log CFU/mL, quantified in only 7.5% (n=3) of the samples (Pádua, 2013). A study performed in Oran investigated LAB levels in bovine raw milk and showed average counts of 4.88 log CFU/mL in MRS agar at 30 °C (Farahani et al., 2017). On raw bovine milk used to produce handmade Brazilian Minas cheese, the average count was 6.4 log CFU/mL (Luiz et al., 2017). On Myzithra cheese, produced with a mixture of whey and raw ovine milk, counts of non-starter lactic acid bacteria ranged from undetected on the first day to 2.31 log CFU/g on 15th day storage at 4°C and varied from 4.07 to 3.31 log CFU/g for thermophilic cocci in the same periods (Kaminarides et al., 2018).
LAB counts might be affected, among other factors, by the treatment of the herd with antimicrobials. On this case, the withdrawal period should be observed to avoid antimicrobial residue, which could affect milk microbiota and inhibit the presence of LAB during cheese maturation, in addition to causing allergies to consumers (Beltrán et al., 2014). Milking under strict hygienic conditions leads to less contamination by pathogenic microorganisms and are considered a preventive measure for environmental mastitis (Gelasakis et al., 2015;Rosa et al., 2014). The major cause of subclinical mastitis is coagulase-negative staphylococci and S. aureus is mainly involved in clinical mastitis of dairy herds (Gelasakis et al., 2015). S. aureus is the most frequently bacterium diagnosed in sheep, in addition to S. hyicus, S. intermedius, S. schleiferi, Mannheimia spp. and Streptococcus spp., among others (Contreras et al., 2007;Gelasakis et al., 2015). Animal health is very important since it guarantees safety, and the appropriate characteristics of milk derivatives (Gelasakis et al., 2015), considering that contaminating pathogenic microorganisms of the raw milk may have faecal origin or through contaminated udder (Forsythe, 2013;Kaminarides et al., 2018). The high contamination of milk by unwanted microorganisms negatively influences LAB growth (Gelasakis et al., 2015;Kaminarides et al., 2018;Luiz et al., 2017), in the cheese maturation process, since they will compete for nutrients . A higher LAB count correlates with better sanitary conditions of the herds (Forsythe, 2013), or that the animals are not being treated for mastitis (Merlin et al., 2015), especially those from farm 4 as observed in this study. On the butter processing, a stage of separation of the liquid phase of the cream, called buttermilk (Forsythe, 2013) occurs, a fact that may explain the higher LAB count in this product. On pasteurized milk (n=3), the mean LAB value was 5.19 ± 0.01 log CFU/mL, lower than the value for raw milk (6.93 ± 0.84 log CFU/mL), possibly due to partial destruction of microorganisms (Forsythe, 2013) during milk pasteurization. The absence of heat treatment is permitted for milk used for cheeses that have undergone a maturation process for a time not less than 60 days and at a temperature above 5 °C (Brasil, 1996). On the evaluated dairy, sheep milk always goes through the process of pasteurization prior to the processing of derivatives.
From the MRS agar plates obtained from the samples of raw milk and sheep milk products, 253 isolates were selected. Gram staining and catalase assay indicated 37 Gram-positive and catalase-negative isolates, of which 78.38% (n=29) were cocci. These results may be related to a higher prevalence of genera such as Lactococcus, Enterococcus, Streptococcus, Pediococcus and Leuconostoc (Bruno, 2011) in sheep milk isolates from the South of Brazil. Already in traditional Algerian butter samples (n=76) 66% of the LAB obtained were coccus (Bettache et al., 2012).
The homofermentative group of LAB produces lactic acid as the only or main product of glucose fermentation, while the group of heterofermentative produces the same molar amount of lactate, carbon dioxide and ethanol from the hexoses (Widyastuti et al., 2014). The latter group produces a series of antimicrobial substances, including lactic acid, hydrogen peroxide, diacetyl, and other organic acids (Farahani et al., 2017). The prevalence of homofermentative profile in LAB isolates obtained from dairy products was also observed in other studies of this Brazilian region.
On a study carried out in Tunisia, 351 isolates obtained from camel and bovine milk 14.24% (n=50) were Gram-positive and catalase-negative (Zeineb et al., 2013). Another study in the same region of Brazil using raw sheep milk and cheese found 52.68% (n=59) of the catalase-negative and Gram-positive isolates .
On a study that evaluated sheep milk and cheese collected in Southern Brazil, 20.34% (n=12) of the LAB showed proteolytic, lipolytic and antimicrobial activities . Six isolates of LAB (23.08%) isolated from bovine raw milk from Oran demonstrated proteolytic activity (Farahani et al., 2017). LAB isolated from traditional Algerian dairy samples showed 3.95% (n=3) and 2.63% (n=2) with proteolytic and lipolytic activity, respectively in butter (Bettache et al., 2012), while in fermented milk and cheese no isolates with lipolytic activity were observed (Mechai et al., 2014). The importance of proteolytic and lipolytic activities in LAB is linked to biochemical processes that occur during the maturation of cheeses (Bruno & Carvalho, 2009). Proteolysis is considered an essential process in all types of cheeses with internal and superficial maturation, and the main responsible agents are enzymes, such as plasmin or those derived from the coagulant, or the proteolytic enzymes of starter bacteria or secondary inocula (Balthazar et al., 2017;Farahani et al., 2017;Pereira et al., 2008). The proteolytic agents act in combination to hydrolyse casein to peptides and amino acids (Balthazar et al., 2017;Farahani et al., 2017), which can be quantified by extension and depth indices of maturation (Pereira et al., 2008). The maturation depth is related to secondary proteolysis from the native milk microbiota or cultures added during cheese production, as observed for homofermentative LAB, Lactobacillus present in starter cultures (Widyastuti et al., 2014). On a study to develop yoghurt from ovine milk, a reduction in fermentation time was observed due to the high proteolytic activity of Lactobacillus delbrueckii sp. bulgaricus (Asensio-Vegas et al., 2016). On another study with fermented milk and traditional cheeses produced in Algeria, it was observed that two isolates of the genus Lactococcus showed the highest proteolytic activity (Mechai et al., 2014). However, lipolysis is a critical process in some varieties of cheeses, because it promotes the softening of the dough and makes the texture softer, as in blue cheeses, Otalian hard dough, Swiss type cheese (Farahani et al., 2017), Scamorza cheese, and Pecorino cheese (Balthazar et al., 2017). Thus, the selection of LAB with both activities may be of interest for the development of starter cultures for regional sheep cheeses and for the characterization of origin of these dairy products.
The presence of an isolate with antimicrobial activity against Listeria monocytogenes ATCC 7644 (Table 1) indicates the possibility for food applications. The ability to inhibit the presence of this bacterium was also observed for one isolate from Brazilian mozzarella cheese, identified as Lactobacillus curvatus (Sehn, 2015). Samples of fermented milk and traditional Algerian cheeses presented 25% (n=52) of LAB isolates with anti-Listeria monocytogenes activity (Mechai et al., 2014), while the isolation in butter from the same site identified LAB with activity against Listeria innocua (Bettache et al., 2012). Dne study tested the use of starter cultures with proteolytic, lipolytic and anti-Listeria monocytogenes activities ATCC 7644, isolated from raw sheep milk, and a beneficial effect was observed in the decrease of the unwanted microbiota in raw milk sheep cheeses matured for more than 60 days . Although there is a low incidence of this bacterium causing clinical mastitis in sheep, the presence of Listeria monocytogenes in raw sheep milk has been reported (Gelasakis et al., 2015), as well its relationship with contamination in dairy products (Buchanan et al., 2017;Forsythe, 2013). The verification of the presence of this pathogenic microorganism is required by Brazilian legislation in cheeses with humidity above 36% (6), although listeriosis is a disease of low incidence, it represents an important risk to public health, due to the degree of severity and high mortality (Buchanan et al., 2017). These data demonstrate the importance of this pathogen for the dairy industry and the feasibility of testing the application of this isolate in a derivative as a preventive measure against listeriosis.
On spontaneously fermented milk produced in Ghana from bovine milk, 373 LAB were isolated, obtaining a percentage of proteolytic activity of 80% and 76.41% of lipolytic activity (Akabanda et al., 2014). On Viamão (South of Brazil), 112 LAB were isolated from samples of sheep milk and derivatives, of these 51% (n=30) with proteolytic activity and 32% (n=19) with lipolytic activity . On a study with traditional Brazilian cheeses, 466 LAB were isolated in MRS Agar proteolytic and 30.7% of them showed proteolytic capacity (Campagnollo et al., 2018). Most of the isolates showed proteolytic and lipolytic profiles, which demonstrates the possibility of their use in the processing of cheeses that undergo proteolysis and lipolysis during maturation.
Osolate LAB LSE 08-23 isolated from sheep's milk from farm 4 showed a higher initial count for these bacteria, emphasizing the importance of adequate management of the sheep herd to maintain the desired microbiota in raw milk. This isolate was characterized as a coccus and showed homofermentative profile. On another study, the zone of inhibition against Listeria monocytogenes ATCC 7644 in LAB of sheep milk products ranged from 6.5 to 10.5 mm , while 15.01% (n=56) of isolates spontaneously fermented bovine milk showed some inhibition halo against Listeria monocytogenes, half of them from 1 to 4 mm (Akabanda et al., 2014). Tests in the LAB isolated from Brazilian Minas handmade cheese samples showed that 73.0% had activity against Listeria monocytogenes at 37 °C (Campagnollo et al., 2018). LAB LSE 08-23 was characterized as a Gram-positive, coccus type, with homofermentative profile, which may have potential for food preservation and/or production of sheep dairy products. However, more studies are needed to verify the viability and efficiency of this isolate during the processing of dairy products, as well as during its shelf life.
Conclusion
The raw sheep milk was the source of the largest number of LAB isolates with lipolytic and proteolytic activities, especially those collected at farm number 4. LAB LSE 08-23, a coccus type with homofermentative profile, was highlighted due to its proteolytic, lipolytic and anti-Listeria monocytogenes ATCC 7644 activities. Thus, sheep dairy samples from the South of Brazil are viable sources of lactic acid bacteria with potential use as starter cultures for fermented dairy products. | 2020-01-02T21:12:20.621Z | 2020-01-06T00:00:00.000 | {
"year": 2020,
"sha1": "13a12ea3a1755a4d063b50bb5582a1e6ed3cfa5a",
"oa_license": "CCBY",
"oa_url": "http://www.scielo.br/pdf/cta/v40s1/0101-2061-cta-fst11019.pdf",
"oa_status": "GOLD",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "b93ce78bd7ba945aa5255884915bc8d232e70797",
"s2fieldsofstudy": [
"Agricultural and Food Sciences",
"Biology"
],
"extfieldsofstudy": [
"Chemistry"
]
} |
239025338 | pes2o/s2orc | v3-fos-license | Endothelial dysfunction and COVID-19 (Review)
It is hypothesized that several comorbidities increase the severity of COVID-19 symptoms. Cardiovascular disease including hypertension was shown to play a critical role in the severity of COVID-19 infection by affecting the survival of patients with COVID-19. Hypertension and the renin-angiotensin-aldosterone system are involved in increasing vascular inflammation and endothelial dysfunction (ED), and both processes are instrumental in COVID-19. Angiotensin-converting enzyme 2 is an essential component of the renin-angiotensin-aldosterone system and the target receptor that mediates SARS-CoV-2 entry to the cell. This led to speculations that major renin-angiotensin-aldosterone system inhibitors, such as angiotensin receptor blockers and angiotensin-converting enzyme inhibitors might affect the course of the disease, since their administration enhances angiotensin-converting enzyme (ACE)2 expression. An increase in ACE2 activity could reduce angiotensin II concentration in the lungs and mitigate virus-driven lung injury. This could also be associated with a reduction in blood coagulation, which plays a critical role in the pathogenesis of SARS-CoV-2; of note, COVID-19 is now regarded as a disorder of blood clotting. Therefore, there is an urgent need to better understand the effect of targeting ACE2 as a potential treatment for SARS-CoV-2 driven injury, and in alleviating COVID-19 symptoms by reversing SARS-CoV-2-induced excessive coagulation and fatalities. Ongoing therapeutic strategies that include recombinant human ACE2 and anti-spike monoclonal antibodies are essential for future clinical practice in order to better understand the effect of targeting ED in COVID-19.
Introduction
Cardiovascular disease (CVD) and hypertension have emerged as critical comorbid risk factors affecting the survival of patients with COVID-19 (1). Inflammation is a major player in the progression of CVD, and the renin-angiotensin-aldosterone system (RAAS) plays an important role in producing and maintaining vascular inflammation (2). While RAAS serves a key role in regulating blood pressure and hypertension, it also mediates pro-inflammatory functions. Most importantly, blocking RAAS has beneficial and protective outcomes in CVD treatment. Indeed, the use of major RAAS inhibitors, such as angiotensin receptor blockers (ARBs) and angiotensin-converting enzyme inhibitors (ACEIs) improves CVD by effectively treating hypertension-induced injury (3). Angiotensin-converting enzyme 2 (ACE2) is a major component of RAAS, and the receptor to which SARS-CoV2 binds to enter the cells. Endothelial cell dysfunction (ED) driven-ACE2 depletion is associated with an increase in inflammation and blood coagulation; both are considered critical factors in the progression of COVID-19. To date, this association remains unclear and thus, there should be an increased effort to better understand the relationship between ACE2, blood hemostasis and inflammation in the pathogenesis of COVID-19 disease.
CVD, ED and RAAS
In CVD, chronic inflammation leads to ED and the initiation and progression of atherosclerosis by enhancing the migration of inflammatory cells into the vessel wall, foam cell formation and the stimulation of smooth muscle cell hyperplasia, which ultimately leads to tissue injury (4). Strong associations between Angiotensin II (AngII), a major actor in RAAS, and inflammation have been demonstrated, implicating AngII in enhancing pro-inflammatory responses through the upregulation of pro-inflammatory cytokines and chemokines, including IL-6, MCP-1, VCAM-1 and TNF-α (5-8). In addition, AngII is a strong pro-oxidant and it mediates its effects through the activation of NADH/NADPH signaling, production of superoxide anions and reduction in nitric oxide (NO) bioavailability, which is a key marker for a healthy endothelium (9)(10)(11)(12).
ACE2 and COVID-19
ACE2 is the target receptor to which SARS-CoV-2 binds with to gain entry into cells (1). Since ACE2 is an essential component of RAAS, concerns arise regarding the plausible relationships between hypertension, the use of ACEIs and ARBs, and the role of cardiovascular disease in aggravating COVID-19 symptoms, restoring the balance in the RAAS system may be a critical factor in attenuating organ injuries. Indeed, this was addressed early during the COVID-19 pandemic; drugs that block RAAS could affect the severity of the disease (15). Results from the initial outbreak in China showed that a majority of patients with COVID-19 with severe symptoms had hypertension; this led to speculations that ACEIs and ARBs may increase the risk of viral infection since their administration enhances ACE2 expression (16,17). However, studies in humans and animal models did not provide any convincing proof of this association, and thus remains unclear and contested (18)(19)(20). Adding to this controversy is the fact that the virus-binding target, ACE2, converts AngII to Ang(1-7), which decreases inflammation and lowers blood pressure. ACE2 therefore plays an important role in balancing the two RAAS arms, the pro-inflammatory and hypertensive arm mediated by ACE, AngII and Angiotensin Type 1 Receptor (AT1R), and the cardioprotective, anti-inflammatory arm mediated by MAS1 oncogene (MasR) and AT2R. Disruption of this balance is a crucial player in the pathophysiology of CVD and COVID-19 (21). Indeed, while the ACE/AngII pathway is important in vasoconstriction, hypertension and oxidative stress, which leads to inflammation, the ACE2/Ang(1-7) pathway counteracts the above effects, and both pathways coexist in various tissues including in the lungs, heart, blood vessels and kidneys where they regulate blood pressure and contribute to CVD pathophysiology (22,23).
The SARS-CoV-2 spike protein recognizes, with high affinity, ACE2 present on the surface of host cells mediating the entry of the virus. Endocytosis of the virus-ACE2 complex can potentially lead to ACE2 downregulation and shedding from the surface of the cell (24). This loss of ACE2 function in infected cells could be a critical factor in the progression and course of the disease (25). Even though there is no compelling evidence that links ACEI and ARB treatment with an increase in SARS-CoV-2 infection, it is becoming evident that these drugs may attenuate AngII-driven lung injury (26). Since AngII promotes inflammation and acute lung injury (27), any increase in ACE2 activity could reduce AngII concentration in the lungs and mitigate virus-driven lung injury. Indeed, a recent study revealed correlations between biochemical and clinical markers of lung injury, viral load and AngII concentrations in patients with COVID-19 (28). Similarly, results link SARS-CoV-2 with a decrease in ACE2 expression and acute heart injury (29). However, it was reported that the use of ACEIs or ARBs in hospitalized patients with COVID-19 had no effect on their survival rate; actually, there was no significant difference in the mean number of days alive for patients who were hospitalized with mild to moderate symptoms of COVID-19 and who were assigned to continue vs. discontinue these medications (30). Conversely, another cohort study that assessed ACEIs and ARBs and included more than 8 million individuals, has shown that these drugs are associated with significantly reduced severe risks of the disease, such as requiring intensive care. This study also hinted to the role of ethnicity in modulating ACEIs/ARBs effects in relation to the severity of the disease; it was shown that the risk of the disease in association with the use of these drugs was higher in Black African and Caribbean groups when compared with the Caucasian group (31). Overall, the use of ACEIs/ARBs is still a paradoxical issue that requires extended investigation to resolve; it is also an area of research where the benefit/risk analysis and potential efficacy of those drugs should be addressed in connection with other comorbidities that are related to COVID-19 (22).
COVID-19, the exposure of a masquerading illness
There is growing evidence for COVID-19 being a disorder of blood clotting where the virus uses the respiratory route to gain entry to blood circulation (32). It has been initially reported that COVID-19 is strongly associated with ischemic strokes in patients that required vacuum and clot retrieval devices as well as blood thinning medications (33,34). It was shown that when the virus enters the blood stream, it triggers a cascade of events resulting in blood clotting and strokes. This all starts with the attachment of the virus to the ACE2 receptor on endothelial cells, making use of transmembrane protease, serine-2 (TMPRSS-2) which initiates the process of ED (35). Thus, SARS-CoV-2 mediated ACE2 downregulation on the surface of the cell results in AngII accumulation and NADPH activation fueling the generation of reactive oxygen species (ROS) and thus increasing oxidative stress (36,37). ROS assists in the conversion of β2-glycoprotein 1 into its oxidized form, which can no longer bind competitively to the von Willebrand factor (vWF) that is secreted by dysfunctional endothelial cells; subsequently, this will promote the coagulation cascade, as vWF binds to the sub-endothelial layer, crosslinking collagen and platelets together and accentuating coagulation mechanisms that lead to strokes of the large vessels (38,39).
During viral infection, the dysfunctional endothelium plays a detrimental role by worsening inflammation which is associated with a poor prognosis in patients with COVID-19 ( Fig. 2) (35). As the coagulation mechanism is a highly organized process that involves endothelial cells, endotheliitis plays a critical role in the pathogenesis of SARS-CoV-2 by increasing the risk of excessive and disseminated intravascular coagulation and the rates of fatality (40). During infection, pro-inflammatory cytokines, such as IL-1β, IL-6 and TNF-α are amplified and lead to a simultaneous increase in the vWF and tissue factor release from endothelial cells, which will promote blood clotting through the increase in platelet aggregation and the initiation of the clotting cascade (41). Similarly, those cytokines enhance blood clotting by downregulating pro-fibrinolytic and anticoagulant factors, including endothelial protein C receptor and thrombomodulin and by upregulating anti-fibrinolytic factors, namely plasminogen activator inhibitor-1 (PAI-1) (42,43). There is cumulative proof that ACE2 downregulation may contribute to an increase in the thrombotic risk in patients with COVID-19 (44). It has been speculated that the decrease in ACE2 activity seen in patients with COVID-19 may lead to a series of mechanisms that are promoted by the dysfunctional endothelium and that affect blood hemostasis. This comprises an increase in vascular permeability, as well as an upregulation of tissue factor and PAI-1 culminating in the activation of the extrinsic coagulation pathway and the reduction in fibrinolysis (45). On this same note, it has been reported that, in animal models of thrombosis, there is a clear association between coagulation and ACE2 pathways. In rats with an induced thrombosis, ACE2 inhibition is significantly correlated with the increase in blood clot weight; conversely, ACE2 administration induced a decrease in thrombus size as well as a reduction in platelet adhesion to the endothelium (46). Similarly, it has been shown that a decrease in ACE2 activity is associated with an increase in blood coagulation in spontaneous hypertensive rats, and that the activation of ACE2 attenuates thrombosis by reducing the attachment of platelets to the vessel wall (47).
Interestingly, it was reported that SARS-CoV-2 can directly bind to platelets through its spike protein, which will enhance platelet activation. It was shown that platelets are hyperactive in patients with COVID-19 and that they express ACE2 and TMPRSS2. ACE2-mediated viral binding to platelets Figure 1. RAAS is essential for regulating blood pressure and inflammation in the body. While the ACE/AngII/AT1R pathway is responsible for the increase in blood pressure and inflammation, the other ACE2/Ang(1-7)/MasR arm of RAAS is involved in the opposing effects, and in lowering blood pressure and inflammation. ACE2 mediates SARS-CoV-2 entry to the cell and is a key player in abrogating the detrimental effects of the ACE/AngII/AT1R arm of RAAS and mitigating virus-driven lung injury. ACEIs and ARBs block ACE (and therefore the conversion of AngI to AngII) and AT1R respectively, which affect the progression of COVID-19 symptoms. RAAS, renin-angiotensin-aldosterone system; ACE, angiotensin converting enzyme; AngII, angiotensin II; AT1R, Angiotensin Type 1 Receptor; MasR, MAS1 oncogene; ACEI, ACE inhibitor; ARB, angiotensin receptor blocker. Figure 2. SARS-Cov-2 binds to ACE2 in order to enter the cell. Subsequent activation of the RAAS pathway induces endothelial dysfunction, which will lead to the generation of ROS and the secretion of pro-inflammatory cytokines (IL-1, IL-6, TNF-α), resulting in a cytokine storm as well as hypercoagulation and a rise in thrombotic events culminating in an increase in the severity of the disease. ACE, angiotensin converting enzyme; RAAS, renin-angiotensin-aldosterone system; ROS, reactive oxygen species; IL, interleukin; TNF-α, tumor necrosis factor-α; Ang, angiotensin. stimulated them to release inflammatory and coagulation factors, which lead to an enhancement in leukocyte-platelet aggregation (48).
Ongoing therapeutic strategies and disease management
At present, the treatment of COVID-19 is limited to alleviating the symptoms of the disease, with no specific antiviral drugs that are effective in targeting the virus (49). Accordingly, there exist numerous ongoing clinical trials and treatments that aim to target COVID-19-associated ED in order to mitigate disease progression and the high mortality rate associated with it. Such treatments include the use of RAAS inhibitors, serine protease inhibitors, recombinant human ACE2, monoclonal anti-spike antibodies, heparin, corticosteroids as well as other agents directed towards specific cytokines and inflammatory signaling pathways (Fig. 3) (50-54). Serine protease inhibitors may affect SARS-CoV-2 entry to the cell by inhibiting TMPRSS-2, which plays an instrumental role in mediating S protein fusion to the endothelial cell membrane (49). One study showed that targeting TMPRSS2 using a clinically proven protease inhibitor can effectively prevent SARS-CoV-2 infection in vitro (25). Additionally, numerous studies and ongoing clinical trials point to the vital role that the RAAS inhibitors may contribute to improving ED and the pathogenesis of COVID-19 (55)(56)(57)(58)(59). On this same note, statins were also reported to improve endothelial cell function in a manner distinct from their major lipid-lowering activities. These drugs can increase the expression of NO synthase, whilst inhibiting NADPH oxidase, which leads to the suppression of pro-inflammatory pathways in endothelial cells (60). Meanwhile, there is a growing evidence showing that statins can improve the prognosis of COVID-19 through the decrease in the production of inflammatory biomarkers (61). In addition, heparin is known to have anti-inflammatory and protective effects in endothelial cells, and recent studies confirmed its role in improving the prognosis of severely infected patients and reducing mortality rates through its well-known anticoagulation properties (62). Furthermore, given the importance of inflammation in the pathophysiology of COVID-19, clinical evaluation of the anti-inflammatory effects of corticosteroids has gained high priority recently. One study has shown the efficacy of methylprednisone in treating severely ill patients with acute respiratory distress syndrome (63). Another study also confirmed the significant role of dexamethasone in decreasing mortality rates in patients who are severely affected with COVID-19 (64). Lastly, other promising therapeutic approaches include targeting cytokines, such as interferon-γ, IL-1 and IL-6 and, as well as the VEGFA/VEGFR2 signaling pathways in order to alleviate virus-driven injury and inflammation (54).
Conclusion and future perspectives
Overall, ACE2 is a key player in SARS-CoV-2 infection and in abrogating the detrimental effects of the ACE/AngII/AT1R arm of RAAS; namely, the ACE2/Ang(1-7) pathway instigates a shift away from ACE/AngII/AT1R, which affects the progression of COVID-19 symptoms. Moreover, there is a clear association linking ACE2 and blood coagulation Figure 3. Therapeutic processes that address inflammation, oxidative stress and blood coagulation, which are considered hallmarks of ED during SARS-CoV-2 infection. The proposed mechanisms include targeting the interaction between the virus and endothelial cells through the use of rhACE2, serine protease inhibitors and SARS-CoV-2 anti-spike monoclonal antibodies in order to decrease ED, viral infection and tissue injury. The therapeutic procedures described here also target ED-induced hyperinflammation and hypercoagulability syndromes by using ACEIs, ARBs and antioxidants, as well as anticoagulant and anti-inflammatory drugs and antibodies, such as dexamethasone, methylprednisone, heparin, anti-IL1 and anti-IL-6. TMPRSS-2, transmembrane protease, serine-2; ED, endothelial dysfunction; ACE, angiotensin converting enzyme; rhACE2, recombinant human ACE2; ARB, angiotensin receptor blocker; IL, interleukin; NO, nitric oxide. pathways, which could play an important role in COVID-19. This relationship suggests that ACE2 may be a novel target for the treatment of thrombogenic diseases, including COVID-19. Future investigations into the role of ACEIs and ARBs in this disease shall expose their potential value for managing COVID-19 symptoms. In addition, there is an urgent need to better understand the effect of recombinant human ACE2 as a potential treatment for SARS-CoV-2 driven injury. Equally important is the need to define the promising role of anti-Spike monoclonal antibodies in alleviating COVID-19 symptoms by reversing SARS-CoV-2 spike protein-induced platelet activation, excessive coagulation and the rates of strokes and fatalities. Future research shall hopefully address these issues as there remain significant gaps in our knowledge pertaining to these related subjects. Accordingly, it is extremely essential that future clinical practice deals with the precise therapeutic processes pertaining to the action of anti-spike monoclonal antibodies and recombinant human ACE2 as fully integrated subjects of high priority. This should help scientists in confirming and verifying the efficacy of those recommended therapeutic strategies in well-designed clinical trials since, as to date, the pathogenesis of COVID-19 is still a vaguely understood subject. Shedding more light onto ED in clinical practice may be more significant than we expect; in this context, a collaborative effort of biomedical and clinical science is urgently required, as this will assist in completing our understanding of the paradigm of the pathogenesis of COVID-19, and in translating our current understanding of the disease to successful treatment strategies. | 2021-10-15T16:21:01.063Z | 2021-10-07T00:00:00.000 | {
"year": 2021,
"sha1": "6f4988c5059f31c371349f72940fda49ac8bc8aa",
"oa_license": "CCBYNCND",
"oa_url": "https://www.spandidos-publications.com/10.3892/br.2021.1478/download",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "1c63196a7ed49f2cb337ab7b3c261e803e36eb5a",
"s2fieldsofstudy": [
"Medicine",
"Environmental Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
43390257 | pes2o/s2orc | v3-fos-license | On Using the Two-level Model as the Basis of Morphological Analysis and Synthesis of Estonian
The paper deals with the problems of describing the Estonian morphological system in the two-level formalism, developed by Kimmo Koskenniemi, The outlines of Estonian morphology are drawn. The basics of the two-level model are given and illustrated with real examples from the experimental Estonian two-level morphology (EETwoLM) composed by the author. A detailed example of step-by-step morphological synthesis is given referring to all the relevant lexicons and rules. The compilation and testing processes using the XEROX finite-state software tools are described. Examples of morphological analysis and synthesis are demonstrated. The present stage of the system is characterised and the future perspectives are drawn. Finally, the suitability of the two-level model for the description of Estonian morphology is discussed.
Introduction
The module of morphological analysis and/or synthesis is unavoidable in any language engineering tool for Estonian because of its rich morphology. For example, in information retrieval systems, it is usually desirable to make queries using semantic entities, not using special morphological forms of a word. As the word stem often has several shapes in Estonian, the morphological component should belong to any information retrieval system. Example 1. To make a query for all occurrences of the word "rida" ("row"), without a morphological synthesiser, three different queries are needed (we assume the possibility to add star (*) to the end of the stem instead of inflectional suffices): Proceedings of NODALIDA 1999, pages 228-242 rida* Cstem in strong grade) rea* Cstein in weak grade) ritta (singular additive ("to the row"), quite often used with this word) The development of EETwoLM is not the first attempt to computerise the morphological analysis and synthesis of Estonian. Ulle Viks (1994) has done important research in the field of morphological classification of Estonian on the basis of pattern recognition theory starting from the 1970s. Viks (1992) has compiled the first morphological dictionary for Estonian as a practical output of the investigations.
Further, E. Kuusik and U. Viks (1998) have implemented the rule-based morphological analysis and synthesis for Estonian. Heiki-Jaan Kaalep (1996) has developed the speller for Estonian using the results of Viks (1992).
Nevertheless, the growing popularity of the two-level model encourages us to consider its suitability to Estonian morphology. From the practical point of view, the appropriate description of Estonian morphology in the form of lexicons and two-level rules makes the significant move towards the application of Xerox language engineering tools to Estonian language (www.xerox.com/xrce/mltt).
The Brief Overview of Estonian Morphology
Estonian language is a member of Finno-Ugric family and is a close relative to Finnish. Estonian morphology is complex -inflected word-forms are built using both agglutination and stem flexion. Nouns have 14-15 cases. Plural forms often have parallel forms.
The Outlines of the Two-level Model, Dlustrated with Examples in Estonian
The two-level morphology model was proposed by Kimmo Koskenniemi in his dissertation (1983). By now, the model has been used for morphological parsing of English, German, Swedish, Norwegian, Danish, Finnish, French, Turkish, Swahili etc.
The main features of the two-level model are the following: • The language description, consisting of rules and lexicons, is separated from the application programs.
• The model is bidirectional -it is oriented to morphological analysis as well as to morphological synthesis.
• The two-levelness of the model means that the deep representations of morphemes rather than morphemes themselves are maintained in lexicons. From those all the real word-forms can be produced with the help of two-level rules and links between lexicons.
Lexical representation: k a D u $ -i-b Surface representation: k a 0 o 0 0 b The surface representation of a word-form is theoretically a sequence of phonemes.
Practically, it tends to be the written form because of the better availability of written texts, as mentioned by Koskenniemi (1997:101).
The lexical representation can contain a) surface phonemes ("k", "a", "u", "d", "b"); b) lexical phonemes ("D" corresponds to d in the strong grade and either disappears or assimilates in the weak grade); c) special symbols for morpheme boundaries and morphophonological features ("+" indicates the boundary between stem and inflectional ending, "$" is the weak grade marker).
The representations are aligned with zero-characters.
• Rules and lexicons are two major parts of the model.
• The set of rules is like a filter, through which the lexical representation can be seen as surface representation and vice versa.
• The rules are not ordered and all of them have to be satisfied at the same time.
• Rules are implemented as finite-state automata.
• A finite-state automaton can be represented as a regular expression, thus the rules are coded as regular expressions. The rule should be read as "The lexical symbol D corresponds to zero-character (i.e. disappears on the surface) in one of the following contexts and only there." In a context description the underline character denotes the place of the pair D:0 between the left and right contexts.
There is a possibility to define sets of characters to make the rules shorter and more readable, e.g. Vow stands for vowels, StemFinVow for possible stem final vowels.
Cons for consonants. Sometimes it is also convenient to give names to frequent word segments, e.g. SylBg means the beginning of a syllable. Note that there is "$" (the weak grade marker) at the end of each right context, thus the disappearance takes place only in the weak grade.
Proceedings of NODALIDA 1999
The exclamation mark indicates the beginning of comments -we have provided every context with 1 -2 example-words that help to understand the context.
• The network of lexicons consists of a stem lexicon and a number of small lexicons describing stem end alternations, inflectional and derivational processes.
• The network of lexicons is implemented as a finite-state transducer.
• A lexical entry includes morphological information, lexical representation and the name of the next lexicon. If we take the word "dppima" ("to learn"), the LEXICON VII builds the following As have been said previously, the model can be the basis for morphological analysis as well as for synthesis. Both analysis and synthesis mean the sequential application of the rule automata and the lexical transducer, but in different order, as seen on figure 1 .
Next, the rules will be applied. The rule "Disappearance of D" is satisfied with the pair D:0 in the context k a _ u: $: (see the last context). Thus we get "kaOu$+b". The second rule accepts the pair u:o in the context k a D: _ . The result is "kaOo$+b".
After applying the whole rule set the default pairs "$:0" and "+:0" have their turn. The result is "kaOoOOb". After the deletion of zero-characters the surface representation "kaob" is ready.
Software
The rules and lexicons were developed and tested using the XEROX finite-state tools lexc (finite-state lexicon compiler developed by L. Karttunen (1993)) and twolc (two-level rule compiler developed by L. Karttunen & K. Beesley (1992)).
The process of testing the correctness and consistency of the lexicons and rules usually proceeds as follows: • The rule and lexicon files have to be composed in a word processor in the described formats.
• Rules coded as regular expressions are compiled to automata with the two-level rule compiler "twolc". Lexicons are compiled into a lexical transducer with the lexicon compiler "lexc".
• Next, the rules and lexicons can be composed with the help of the program "lexc".
• There are some possibilities to test the correctness of the language description in the program "lexc". One can analyse single word-forms using the directive "lookup <word-form>" and produce word-forms using the directive "lookdown <primary form-Hmorphological information>"). We can use the directive "random-surf' for generating word-forms randomly using the existing lexicons and rules. Example 6 . Test of morphological analysis and synthesis using the program lexc. le x o lookup pead pea+S+Sg+P ("head", substantive, singular partitive) pea+S+Pl+N ("head", substantive, plural nominative) pidama+V+Ind+Pr+Sg2 ("must", verb, indicative mood, present tense, singular, 2'"^ person) As we can see, words can be morphologically ambiguous in Estonian. By the way, in Estonian texts about 50 % of the word-forms are morphologically ambiguous. strange compound, hard to translate "ones who learn from a distance" -slightly strange compound "without the seen thing" (I) "went" To the output of the program approximate English translation as well as signs "+" or "?" are added. Every normal word-form is followed by If the word-form is not used, it is marked by The mistakes are caused by the overgeneration of compounds and derivatives.
Results
The experimental two-level morphology for Estonian (EETwoLM) has been composed: • There are 45 two-level rules in the rule set that deal with stem flexion, phonotactics, orthography and morphophonological distribution.
• The net of lexicons consists of root lexicons for all word classes containing a total of =350 different word roots and of over 200 small lexicons describing the stem end alternations, conjugation of verbs and declination of nouns.
• The lexicons and rules express most of the phenomena occurring in Estonian morphology.
• The system is consistent in its present stage: we can get correct results to both morphological analysis and synthesis in the range of word stems occurring in the root lexicons.
Future Perspectives
The coverage of stem lexicons can be enlarged semi-automatically, using the electronic version of Viks (1992) and the type-detection module developed in the Institute of Estonian Language (see the webpage www.eki.ee/tarlcvara). To adapt EETwoLM exactly to the morphological classification after Viks (1994), some changes have to be introduced into the network of lexicons.
Proceedings of NODALIDA 1999
A consistent and lexically satisfactory description of Estonian morphology in the two-level formalism can be the basis of automatic morphological analysis and synthesis.
Simultaneously, two-level-morphology-based language engineering software in XRCE (spelling checker, information retriever a. o) would be applicable to Estonian language.
Model to Estonian Language
During the composition of EETwoLM some features of the two-level model proved very useful. We have given the overview of them in Uibo (1999:55): 1. Using the lexical representation is an advantage because the lexical entries can include other information additionally to the pure sequence of letters: • There is a possibility to use special denotations for phonemes having more than one surface variant. This is a great advantage, as the type of stem flexion generally does not depend on the phonemic shape of the stem in the present-day Estonian -some kinds of stem flexion are not productive any more.
• The lexical information can contain morphophonological features and morpheme boundaries, which are often used by rules. 4. If a pair occurs in several contexts having nothing common neither in content nor in form the corresponding contexts can be listed on the right side of one and the same rule. It prevents from introducing new and meaningless lexical characters. E.g. the pair "S:0" is possible both in the weak grade of the words with s:0-altemation within the stem and at the end of a class of words ending with "s". In the first case the lexical phoneme "S" is situated in between vowels, in the second case it is found at the end of the stem.
However, we have also pointed to some difficulties in Uibo (1999:56) that have occurred in the course of the description of Estonian morphology in the two-level formalism: 1. The word class is subject to change during the derivation processes, but the morphological information is composed moving along the pointers between lexicons in one direction. Return to the previous steps, thus the deletion and replacement of the word class is not possible. Now the problem has been solved artificially: the verb derivatives are in a separate lexicon and for the productively derivable adjectives the determination of word class has been deferred.
2.
It is inconvenient to introduce word lists into the lexicon system that do not coincide with the inflection types. The lists are needed e.g. for words with exceptional forms, for words having additive case and short plural, and especially for compound word production.
The hypothetical solution of the listed problems could be the combination of the twolevel model with another model that would help to overcome the above-listed limitations.
Conclusion
The experiments on EETwoLM have shown that the two-level model is quite usable for Estonian simple word recognition and production. However, the net of lexicons is not very well suitable for modelling the derivation and compounding processes. The efficiency of the implementation of the rules and lexicons as finite-state transducers is definitely an advantage. Unfortunately, the objective evaluation of EETwoLM is not possible yet, as the coverage of the lexicons is insufficient for real text processing. | 2017-01-07T08:35:44.032Z | 1999-01-01T00:00:00.000 | {
"year": 2000,
"sha1": "22a536840c3299f0fcd19d1d0dcde9ecd579eb24",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "ACL",
"pdf_hash": "22a536840c3299f0fcd19d1d0dcde9ecd579eb24",
"s2fieldsofstudy": [
"Computer Science",
"Linguistics"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
10670093 | pes2o/s2orc | v3-fos-license | Combination of Q-Switched Nd:YAG and Fractional Erbium:YAG Lasers in Treatment of Melasma: A Randomized Controlled Clinical Trial
Introduction: Ablative and nonablative lasers have been used to treat melasma. We aimed to assess and compare the combining Q-switched Nd:YAG laser (QSNYL) and fractional erbium:YAG laser (FEYL) with QSNYL alone in treatment of melasma. Methods: This randomized controlled clinical trial was performed in our Research Center during 2013-2014. Women with melasma and without a history of keloid formation, hypersensitivity to hydroquinone, or pigmentary changes due to laser therapy were randomly allocated to receive four sessions of either QSNYL-FEYL combination or QSNYL alone. All patients received topical treatment with Kligman’s formula. Before laser therapy and 4 weeks after the last treatment session, patients’ skin was assessed for changes in skin color, melanin content, and erythema intensity of melasma lesions quantitatively. Results: Finally, 21 patients in QSNYL-FEYL and 20 in QSNYL group (mean age, 38.57 [5.60] and 42.60 [8.44] years, respectively) completed study. The skin color had become lighter in both groups (mean [SD] percentage change of 56.95 [40.29] and 29.25 [13.20] in QSNYL-FEYL and QSNYL groups, respectively) with significantly better results in QSNYLFEYL group (P = 0.006). Percentage of decrease of melanin content was significantly higher in QSNYL-FEYL group (22.01 [10.67] vs. 7.69 [4.75]; P < 0.001). After adjustment for baseline values, the post treatment intensity of erythema was significantly lower in QSNYLFEYL group (P < 0.001). The patients reported no adverse events. Conclusion: QSNYL-FEYL was significantly more effective in decreasing melanin content of lesions than QSNYL and led to a lighter skin.
Introduction
Melasma is the most common pigmentary skin disorder, which is seen more frequently in females and those with darker complexion. Many factors contribute to the development of melasma including ultraviolet exposure, pregnancy, oral contraceptive pills, endocrine and hormonal factors, and genetic predisposition. It is a chronic and hard-to-treat disorder and although different therapeutic modalities have been employed to treat the disorder, 1 search for an effective method of therapy continues. Nonablative laser, alone or in combination with ablative lasers and topical medications, have been employed to treat pigmentary disorders including melasma. Nonablative lasers such as Q-switched Nd:YAG laser (QSNYL) that selectively target melanosomes [2][3][4] and ablative lasers such as fractional erbium:YAG laser (FEYL) or fractional CO 2 laser (FCOL) have been used frequently for this purpose. 5 Nonetheless, we could not find any study that had employed both laser systems in treatment of melasma. Moreover, most of the studies have used melasma area and severity index (MASI) or physician's global assessment (GPA) to evaluate the improvement of therapy; however, objective assessment of the melanin content and changes in lesions' color would provide more reliable result in terms of treatment efficacy. Although patients with melasma were treated with either QSNYL or FEYL in our clinic, combining these two systems had not been tried before. We aimed to assess the efficacy of combination laser therapy with QSNYL and FEYL in treatment of melasma.
Methods
This randomized, controlled, assessor-blinded clinical trial was conducted in our Research Center during 2013-2014.
Patients
Patients at reproductive age with melasma of the malar region and nose were included. Patients with active lesions or infection in treatment area, history of keloid formation, hyperpigmentation after laser therapy, hypersensitivity to hydroquinone, or any condition that would hamper patients' participation were excluded. Patients were informed of study protocol and possible benefits and adverse effects. A written informed consent was obtained from those who met the eligibility criteria and accepted to participate in the study.
Sample Size
Most of the studies in this field were case series, which had made calculating sample size difficult. With regard to the study by Niwa Massaki et al 6 and by considering α of 0.05, study power of 95%, and using the following formula, the sample size was calculated at 21 in each group:
Randomization and Blinding
We used random-sequence blocks with size of four (ie, AABB, BBAA, ABAB, BABA, ABBA, and BAAB) to allocate participants randomly to receive QSNYL and then FEYL (QSNYL-FEYL group) or only QSNYL therapy (QSNYL group). Blinding the patients and therapist was not possible as one group received therapy with one laser system and the other with 2 different laser systems. Nonetheless, the physician in charge of assessing lesions through biometric devices and the statistician in charge of data analysis and reporting results were blinded to the assigned treatment to each patient.
Interventions
Before laser therapy, patients were instructed to apply Kligman's formula (0.1% tretinoin, 5.0% hydroquinone, and 0.1% dexamethasone in a hydrophilic ointment) 7 on their face nightly for at least 1 week before starting treatment and continue using it throughout treatment. Laser therapy sessions were hold every other week and each patient received four treatment sessions.
Half an hour before starting laser therapy, patients used 5% lidocaine/prilocaine cream (EMLA) on their face and covered the face with a sterile thin nylon. Few minutes before laser therapy, patients were instructed to cleanse their face with soap and water and dry it to remove any remaining cream. Patients in QSNYL group only received QSNYL treatment with the same laser system and settings. Moreover, local anesthesia, treatments, and recommendations after laser therapy were the same for both groups.
Outcome Measures
To assess and compare the effects of treatment modalities, we employed two devices that quantified the changes in the skin. Mexameter MX 18 probe of C + K Multiprobe Adapter System (Courage + Khazaka Electronics, Cologne, Germany) measures skin melanin content as well as erythema via light absorption/reflection. The output of the device was numbers that facilitated the comparison of melanin and erythema at baseline with the contents in the same points after therapy. We also selected some points in nonlesional skin to see whether any change would be seen in nonlesional skin as the result of applied creams and to have a control for improvement of pigmentation in lesional skin. Visioface 1000 D (Courage + Khazaka Electronics, Cologne, Germany) was another device that helped achieve pictures of the face with the same light, distance, and magnification through inserting the head of patient in the cavity of device. The accompanying software helped to determine the changes in the skin color by measuring changes in different spots of the face. For each patient, we used the mean of four measurements as the final output of Visioface. Increase in the obtained value would indicate improvement of skin color, ie, getting lighter. Patients were assessed by Mexameter and Visioface half an hour before first laser session and four weeks after the last one.
Statistical Analysis
We used SPSS 16.0 (SPSS Inc., Chicago, IL, USA) to analyze the data. The data showed normal distribution in Kolmogorov-Smirnoff test and hence, parametric tests were employed. Paired-samples t test was used to assess the changes from baseline in each group and differences between groups were evaluated by independent-samples t test. The data were adjusted for confounding factors and analyzed through analysis of covariance (ANCOVA). For all tests, P value > 0.05 was considered statistically significant.
Results
A total of 50 patients were evaluated and finally, 46 patients were enrolled in the study and randomly allocated to two groups of 23. Two patients in QSNYL-FEYL and three in QSNYL groups were lost to follow-up and the remaining entered the final analysis ( Figure 1). All patients were female and the means (SDs) of age in QSNYL-FEYL and QSNYL groups were 38.57 (5.60) (range, 30-49) and 42.60 (8.44) years (range, 25-57), respectively, with no significant difference between them (P > 0.05). Both groups showed significant changes in mean values obtained from Visioface (P < 0.001 for both groups) ( Table 1 and Figure 2). There was no difference between two groups Visioface results after treatment (P = 0.197).
The percent increase of Visioface score was significantly higher in QSNYL-FEYL group in comparison to QSNYL group (Table 2). Moreover, after adjusting for baseline value, the changes were more significant in QSNYL-FEYL group than in controls (P < 0.001; Table 3).
The changes in melanin content were compared before and after treatment in both melasma lesions and nonlesional skins. In QSNYL-FEYL group, both nonlesional and lesional skin showed significant decrease in mean melasma content; however the changes were more significant in lesional skin (laser-treated area) (P < 0.017 and P < 0.001, respectively). In QSNYL group, the melanin content was decreased in both nonlesional and lesional skin; however, the changes in nonlesional skin were not statistically significant. We also compared melanin content between the lesional and nonlesional skin of each patient. Although both groups showed significant differences between lesional and nonlesional skin at baseline, in contrary to patients in QSNYL group, patients in QSNYL-FEYL group showed no significant difference between melanin content of lesional and nonlesional skin after treatment (Table 1). Although the melanin content was significantly decreased in both groups, the decrease was significantly higher in QSNYL-FEYL group (Table 2). There was no difference between two groups in melanin content changes in nonlesional skin before and after adjusting for baseline melanin content (P = 0.915 and P = 0.793, respectively). On the other hand, after treatment, the melanin changes in lesional skin was significantly lower in QSNYL-FEYL group in comparison to controls, even after adjustment for baseline melanin content (P < 0.001 in both) ( Table 3).
In QSNYL-FEYL group, there was no change in erythema intensity after treatment in nonlesional skin (P = 0.244) while it was significantly decreased in lesional skin (P < 0.001). In QSNYL group, no change was seen in erythema of nonlesional or lesional skin after treatment (P = 0.08 and P = 0.09, respectively); nonetheless, there was a significant difference in erythema intensity between lesional and nonlesional skin at baseline as well as post treatment in both groups, ie, erythema was increased more profoundly in treated areas (Table 1). Although erythema intensity was decrease in both groups, there was no difference between groups in percent decrease of erythema (Table 2). After adjustment for baseline erythema, no difference was seen between two groups regarding changes in erythema of the nonlesional skin (P = 0.68) while QSNYL-FEYL group showed a significant decrease in erythema of lesional skin in comparison to
Results
A total of 50 patients were evaluated and finally, 46 patients were enrolled in the study and randomly allocated to two groups of 23. Two patients in QSNYL-FEYL and three in QSNYL groups were lost to follow-up and the remaining entered the final analysis ( Figure 1). nonlesional skin (P < 0.001; Table 3). While the safety of therapy was not among the outcome measures of our study, patients reported no adverse effect after completing the treatment; however, almost all of them had experienced two-to three-day self-limiting erythema after each laser session.
Discussion
We treated our patients with either QSNYL-FEYL or QSNYL and a topical administration of Kligman's formula. Treatment with both QSNYL-FEYL and QSNYL led to improvement in skin color; however, this improvement was more significant in QSNYL-FEYL group. Although the melanin content of lesions decreased in both group, the decrease was more significant in QSNYL-FEYL group, while demonstrating no significant difference in melanin content between nonlesional and lesional skin after treatment. Despite no significant difference in percentage of decrease of erythema intensity between groups, after adjustment for baseline value, erythema was significantly less intense in those treated with QSNYL-FEYL. In addition, patients of both groups reported no significant adverse effect of therapy. Only one patient withdrew from the study due to lack of subjective satisfactory results after three treatment sessions.
Ablative fractional lasers such as FCOL and FEYL, alone in combination with nonablative lasers such as Q-switched alexandrite laser (QSAL), were previously used to treat melasma lesions. In a case series, Nouri et al. employed FCOL and QSAL after 14 days of treatment with Kligman's solution to treat eight patients. In their study, one group received FCOL alone and the other received FCOL and then QSAL. They stated better results with combination therapy. 8 In another split-face study, Angsuwarangsee et al compared the results of treatment with QSAL alone with that of QSAL and FCOL combination in six patients with refractory melasma and reported better results with combination therapy with regard to MASI score and melanin index. 9 Although we could not find any study that had assessed the combination of QSEYL and FEYL, in treatment of melasma, both laser systems were employed previously and the results were compared with topical therapies. In a crossover, split-face clinical trial, Jeong et al compared the effect of 1064-nm QSEYL on melasma before and after treatment with topical triple combination (TTC; hydroquinone 5%, tretinoin 0.05%, and triamcinolone acetonide 0.1% cream) using patients' self-assessment, MASI score, and spectrophotometry. They reported better subjective results on the side that had received laser therapy after topical treatment. 10 In another splitface study, Wattanakrai et al compared 5 session of weekly QSEYL and 2% hydroquinone (laser group) with topical therapy (control) in Asian patients with dermal or mixed-type melasma. According to modified MASI score and photometric assay, they found significant improvement in the laser-treated side; however, they reported complications such as melasma recurrence, rebound hyperpigmentation, and hypopigmentation in follow-up. They concluded that such a laser therapy would induce temporary improvement with undesirable adverse effects. 11 Kroon et al compared 8-week treatment with fractionated nonablative 1550-nm erbium glass laser (held every other week) with 8-week treatment with TTC. Although they found no difference between groups in GPA, selfsatisfaction was significantly higher in those treated with laser. They reported some adverse effects such as pain, burning sensation, and erythema after laser therapy; nevertheless, they recommended such a laser therapy as a safe and acceptable treatment for melasma. 12 Wind et al compared fractionated nonablative 1550-nm erbium laser with TTC and reported significantly lower GPA and satisfaction as well as high rate of hyperpigmentation (31%) in the laser-treated sides and hence, did not recommend such a therapy for melasma. 13 While laser therapies are associated with adverse effects such as burn and erythema and might induce post inflammatory hyperpigmentation, their usage in dermatology and cosmetic procedures is increasing. We did not employ MASI, GPA, or other subjective tools to assess the improvement; instead, we tried objective measures, which would provide less biased and more precise determination of changes in skin. Nonetheless, our study had some limitations. First, the sample size was small because many of the patients did not intend to participate in a study that examined new treatment modalities. Most of the patients with melasma had attempted many other treatments with unfavorable results and were reluctant to undergo an aggressive laser therapy. Second, although we tried to use the devices that evaluate the lesions and their improvement objectively, these devices are operator-dependent and the degree of improvement might be affected by the spots chosen by the examiner. Nevertheless, a trained physician (first author), who was unaware of assigned treatment to each patient, performed all the measurements. Third, short-term follow-up after the last treatment session would affect the judgment on the long-lasting beneficial and adverse effects of therapy as well as changes in skin parameters. Fourth, we did not determine the depth of lesions. In fact, dermal lesions tend to be more resistant to administered treatments. 14 Using Wood lamp, dermal ultrasonography, | 2018-04-03T00:11:43.775Z | 2017-01-25T00:00:00.000 | {
"year": 2017,
"sha1": "7d4f88e1ce8e520c2b336f15f0f3d90b17732f56",
"oa_license": "CCBYNC",
"oa_url": "http://journals.sbmu.ac.ir/jlms/article/download/8823/8823",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "7d4f88e1ce8e520c2b336f15f0f3d90b17732f56",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
234006805 | pes2o/s2orc | v3-fos-license | Application of identity resolution and blockchain technology in the whole industrial chain management of electrical equipment
As the proportion of electric energy in the world energy consumption increases year by year, the demand and usage of electrical equipment also increase. For a lot of electrical equipment management, the traditional management method is unable to get through data barriers that exist in the industrial chain; the information interaction between each link is more and more untimely, leading to a great increase of each link’s error rate. Not only reduces the safety of electrical equipment, but affects the response efficiency of other links when one link sends a request, lowers the quality of service. To solve this problem, this paper proposes the method of integrating the identity resolution system with blockchain technology to standardize the management of the whole industrial chain of electrical equipment. The data barriers between links are broken through the identity resolution system, and the blockchain technology is used to speed up the response of each link, and the key information is stored on the blockchain. This method carries out comprehensive management on the whole life cycle chain, conducts better standardized management on the field of electrical equipment in terms of intelligent trading, intelligent operation and maintenance, etc., and puts forward efficient management mode for the management of electrical equipment.
Introduction
With the development of the social economy, electric power energy gradually replaces the traditional fossil energy in production and life, and takes an increasing proportion in the energy field [1]. The demand for electrical energy has grown exponentially, and a large number of new power plants have been built. In addition, many individual new energy users have also expanded their power generation APSDP 2020 Journal of Physics: Conference Series 1800 (2021) 012011 IOP Publishing doi: 10.1088/1742-6596/1800/1/012011 2 businesses. The monitoring of electrical equipment is an important measure to ensure the safety of electric power production. The widely distributed and large-scale electrical equipment brings new challenges to the traditional equipment management system. Due to the complexity and diversity of electrical equipment, the equipment has different identification code systems in different manufacturers and life cycle stages. There are serious data barriers among each link and manufacturer [2]. Traditional equipment management system cannot carry out unified control of the equipment. Different identification systems also lead to different identification codes for the same data in different links, which greatly increases the number of repetitive identification codes. The key information retention lacks security guarantee, and the data transmission delay of each link is relatively large. In view of these problems, this paper proposed the method which combines the identification analysis system and blockchain technology to establish the electrical equipment industrial cloud network. To carry out unified standardized management of the whole industrial chain of electrical equipment, the safety management of electrical equipment should be enhanced; the information communication between each link and the quality of electrical equipment should be improved. As the foundation of the development of the Internet of Things (IOT) industry, the technology and services related to the Identity Resolution of the IOT are the core support for the rapid and large-scale development of the IOT industry [3]. Identity solutions not only provide domain names, similar to Internet functions, but also regulate IOT management information. More importantly, the system is a fast, effective and secure data connection, which can solve the problem of data island. There are many kinds of identification systems around the world, including EPC code, UCODE code, Ecode code and OID code proposed in China, which have been applied to some extent, but these identification systems are not compatible with other systems [4]. In 1994, handle system was put forward. In 2014, non-profit organizations DONA established maintenance and operating system, which accelerated the development of handle recognition system. Handle system established 10 servers in the United States, China, Germany, the United Kingdom and other countries. [5]. Chinese State Department also identifies the system construction as an important goal. The global root node and the auxiliary root node of Handle identity resolution has completed preliminary layout, then, the nation's top node and the industry node has been launched. This article uses the identity resolution system of Handle as the main body, realize the compatibility and interoperability of the ecosystem, EPC and other mainstream identification systems, complete the authorization and trust of nodes in the field of electrical equipment, and solve the problem of data interoperability in the field of electrical equipment Blockchain is a distributed database, which is realized by consensus mechanism and "decentralization" [6]. Every system.node has equal rights and responsibilities, and there is no central node. Many nodes achieve data consistency and chain consistency. Through the consistency mechanism, most agreed data can be saved in the blockchain [7]. Data records are in a common chain. All nodes can query the existing data. Data records can not be tampered, which also ensures the security and traceability of data. [8]. The application of chain operation in various industries has always been the focus of research, starting from Bitcoin, developing to Ethereum and intelligent contract, gradually emerging private chain and consortium chain [9]. The scholars also studied the application of distributed transaction blockchain technology in smart grid, automobile, network, smart home and industrial IOT. [10]. Based on the characteristics of the field of electrical equipment, in order to improve the production efficiency and product quality of the whole industrial chain of electrical equipment, this paper proposes a method combining blockchain technology and identity resolution to optimize the process of the establishment of the electrical equipment life cycle, intelligent transactions, and intelligent operation and maintenance.
Whole electrical equipment life cycle based on blockchain
As an important part of power system operation, Due to its multiple production links, different manufacturers have different codes between each link and within the same link. The origin of current tracking electrical equipment is limited by step-by-step evaluation of design, production, sales, Although there are some equipment management platforms, the integration of each part of the system is low, and the platform can't bear the upload of large quantities of data. Usually, the fault is tracked, and there is no comprehensive management of electrical equipment to enhance the availability of electrical equipment and improve the quality assurance of equipment, this chapter uses the identity resolution to establish a full life cycle chain for electrical equipment and uses blockchain technology to record it, to increase its non-repudiation. Important data can be inquired when necessary, and the source can also be timely traced after quality problems occur. This paper introduces how to establish the resolution of equipment life cycle. All operations and data are included in several stages of materials, power supply, equipment, production, logistics and transportation, electrical equipment, transaction management, equipment operation and maintenance, equipment asset management shall be uniformly identified. The existing identification codes of previous links should be transformed into the electrical equipment identification system, and the data not previously identified are identified using the electrical equipment identity resolution. Break the barriers between production equipment and management system, establish a complete life cycle of electrical equipment, can record and query the production of equipment and components., equipment construction planning records, electrical equipment transportation, sales records and recycling [11]. Compared with the previous independent identification of each link, the electrical identity resolution uses the same code to identify each link in the life cycle chain, and any code of any link can be informed by all nodes. The establishment of a unified full life cycle chain has the following advantages. First of all, the identification should be standardized to prevent the identification duplication caused by the same equipment having different identification codes in different links. This reduces the repeated identification and enhances the efficiency of information interaction. Secondly, the whole life cycle chain constructed by the standard identification system also reduces the repeated coding steps of different links. Each link only needs to code the products or operations of its links. At the same time, the product information of previous links can be fast read through identification code analysis, which facilitates the efficient use of information. The unified identification code also reduces the information transmission errors caused by improper communication in different links [12]. These improvements significantly solve the problems such as logistics can't be processed in time, information reading waste time, more manual processing errors. The unique resolution of the system can not achieve the tracking of the whole life cycle of the device, It also can be used to realize convenient information interconnection between enterprises and institutions, and share the production, equipment manufacturing and operation data. In each link of the whole life cycle of the product chain, the information of the product life cycle can also be read, making each link more transparent. The data of each link can also be connected through the identification system to break the data barriers between links, Including research and design, production and manufacturing, acceptance and testing, storage and distribution, installation, disassembly and scrap and other links can be widely connected. Post identification and coding data can be shared through every link, quality evaluation, electrical and so on. Equipment can use each link data in the whole life cycle to mine and extract key process data, study the impact of these product performance data and product quality on international relations, and form a typical identification model of electrical equipment quality evaluation. The quality optimization of production process is analyzed and evaluated. After the complete construction of the full life cycle chain of electrical equipment, the full life cycle chain of valuable equipment and important equipment is uploaded to the blockchain for recording. In case of unstable working conditions or failure of electrical equipment, hold relevant manufacturers accountable. The record of blockchain can effectively prevent any link from malicious repudiation [13]. In addition, it is difficult to find the information on electrical equipment that has been reimbursed and destroyed. Through the records of blockchain technology, relevant information can be traced, so as to restore the full life cycle chain of such products through the query result identification code.
Intelligent transaction of electrical equipment based on blockchain
The same types of electrical equipment in different applications have small differences in requirements, it is usually necessary to adjust the parameters of equipment parts to solve this problem. After the buyer puts forward specific requirements, the traditional procurement method requires the equipment manufacturer to put forward requirements for each component manufacturer, and the component manufacturer then puts forward requirements for the sub-component manufacturer or raw material manufacturer. This complicated requirement transmission process greatly lengthens the response time of electrical equipment production and increases the error rate of transmission process. In addition, in the case of a large number of parts in demand, hoarding of parts and waste of materials are also one of the problems that need to be solved, so the number of required parts needs to be accurately controlled to avoid the manufacture of redundant parts [14].
Figure 2. Intelligent transaction process
This chapter uses the identification resolution system and blockchain technology to solve this problem. First of all, a complete part identification code base is established for parts of various specifications, and corresponding identification codes are established for parts of various parameter types. After the customer presents the requirement, the requirement is identified and registered. Through identity codes, people can get the parts identification codes related to the equipment, and then upload the requirements identification codes to the blockchain. Through the analysis of demand code, each manufacturer can combine their production conditions for customer demand bidding. Finally, after the customer receives the quotation, the manufacturer will be selected for cooperation, and the online electronic contract signing business can be completed with the help of blockchain technology. Due to the registration of the identity resolution system, the data barriers between all links are broken. The same identity code can be read and analyzed by all nodes on the chain, realizing the sharing of data. Blockchain technology also saves the process of information transmission, and manufacturers only need to pay attention to whether the information on the blockchain contains the required information. After reading the relevant identification code, each manufacturer will analyze it separately. The parts manufacturer of the equipment gets the identification code of the parts to be produced. After the parts identification code is identified by the factory's processing line, the customized assembly line does not need to be adjusted manually, which also ensures the accuracy of parts production and reduces the errors caused by human factors to some extent. At the same time, the demand code also includes the number of parts required by the customer. Reading the identification code on the blockchain, the manufacturer determines the number of parts to be produced, so as to carry out quantitative production of parts and avoid the hoarding of parts and waste of resources caused by excessive production. After the order is placed, both parties can sign the online contract with the electronic signature through the blockchain, eliminating the need to go through the offline procedures. The signed contract also has legal effect. After the electronic contract is uploaded to the blockchain, the contract can be traced in case of later problems. After the transaction is completed, the transaction certificate is also uploaded to the blockchain, so that the transaction can be traced, with stronger reliability and non-repudiation [15]. By combining the Identification resolution system with the blockchain technology to solve the trading problem of electrical equipment, intermediate steps can be reduced. The component manufacturers at all levels can understand the needs of customers through the identification codes on the blockchain. In this way, the design and manufacturing of required parts can be carried out more accurately, which can reduce errors caused by intermediate communication, shorten the communication and manufacturing response time of manufacturers, and also avoid hoarding of parts and wasting of materials. In addition, the use of blockchain for trading can ensure the security and credibility of transactions.
Intelligent operation and maintenance of electrical equipment based on blockchain
After the electrical equipment is put into operation and production, equipment failure is inevitable. The traditional method usually asks professional maintenance personnel for inspection, evaluation and repair after the failure and shutdown. The complicated maintenance process wastes a long waiting time and reduces the working efficiency of the equipment. Improving the operability monitoring of electrical equipment is the basis for equipment manufacturers to provide high quality and intelligent products and services. In this chapter, a new approach to equipment operation and maintenance is proposed by combining the Identification resolution system with blockchain technology. Figure 3. Intelligent operation and maintenance process First of all, the electrical equipment and the whole life cycle chain should be identified and registered, and then the key data of each equipment should be encrypted and uploaded to the shared chain block. The relevant manufacturers have the ability to crack the password chain. They can get the operation data of the user's equipment through system analysis, so as to monitor the status of the operating equipment in real-time. Combined with the historical equipment monitoring records, they can establish a comprehensive evaluation model for comprehensive inspection and analysis. In this regard, big data, artificial intelligence and other technologies are used to extract, classify and summarize error types. The equipment fault model library is established, the multi-dimensional analysis and state comparison are carried out, the accurate fault diagnosis, early warning and recovery system is established, and the fault alarm of the electrical equipment manufacturer is responded in time. If there is a simple fault, the MA If there is a complex problem, the power plant will directly send special personnel to replace and repair the equipment, so as to reduce the response time and maintenance time. The method provides users, operation and maintenance, with accurate and efficient equipment, health diagnosis service, reduces customer losses due to accidents, and provides high quality after-sales service to support manufacturers.
Conclusion
The traditional management method of electrical equipment could result in unnecessary mistakes due to the excessive information of relatively independent management of each link and the complexity of the process of communication. In order to solve the existing problems at the present stage, this paper uses the identity resolution system to carry out unified identification and management on the production, sales, transportation, scrap recycling and other aspects of electrical equipment, establish the full life cycle chain of equipment, use the blockchain technology to realize the information preservation of the full life cycle chain of equipment, and improve the non-repudiation of transactions. In addition, the combination of this technology also has applications in the field of equipment intelligent trading. By sharing the user's demand identification code through the blockchain, the component manufacturers could confirm the parameters and number of the spare parts through the analysis of the identification code, which reduces the waste of spare parts manufacturing and reduces the mistakes in setting parameters and conveying process. Finally, through the openness of blockchain technology, the data sharing of electrical equipment can be realized. Through identification code, the working data monitoring identification code by component manufacturers can be realized, consequently, the problem of equipment failure could be solved in-time and the intelligent operation and maintenance of equipment could be accomplished. This paper uses the method of combining identity resolution system and blockchain technology to solve the complex problems in the management of electrical equipment, which could break through the data barriers between businesses, improve business efficiency, reduce unnecessary waste, and open up new ideas for promoting the systematic management of electrical equipment. | 2021-05-10T00:03:37.463Z | 2021-02-01T00:00:00.000 | {
"year": 2021,
"sha1": "e0e21395285d5607377788caa14a7278280edebf",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1742-6596/1800/1/012011",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "009c29d182885f59a77992bdd813519b1149dc7a",
"s2fieldsofstudy": [
"Computer Science"
],
"extfieldsofstudy": [
"Physics",
"Computer Science"
]
} |
208910929 | pes2o/s2orc | v3-fos-license | Uniform property Gamma
We further examine the concept of uniform property Gamma for C*-algebras introduced in our joint work with Winter. In addition to obtaining characterisations in the spirit of Dixmier's work on central sequence in II$_1$ factors, we establish the equivalence of uniform property Gamma, a suitable uniform version of McDuff's property for C*-algebras, and the existence of complemented partitions of unity for separable nuclear C*-algebras with no finite dimensional representations and a compact (non-empty) tracial state space. As a consequence, for C*-algebras as in the Toms-Winter conjecture, the combination of strict comparison and uniform property Gamma is equivalent to Jiang-Su stability. We also show how these ideas can be combined with those of Matui-Sato to streamline Winter's classification-by-embeddings technique.
Introduction
The study of approximately central sequences associated to operator algebras is almost as old as the subject itself. In their foundational work ( [43]), Murray and von Neumann distinguished factors associated to free groups from the hyperfinite II 1 factor; they did this through Property Γ -the existence of approximately central unitaries of trace zero -showing the latter has this property, while the former does not. Property Γ was subsequently analysed in detail by Dixmier ([17]), and shown to be equivalent to non-triviality of the central sequence algebra. Moreover, in this case the central sequence algebra is necessarily diffuse, so there are approximately central projections taking all possible values of the trace. This latter characterisation has been critical in a range of applications from cohomology to the generator problem ( [11,23,46,12]). On the other hand, failure of property Γ is equivalent to fullness, and normally used through a spectral gap characterisation of Connes [13,Theorem 2.1], which provides the starting point of his celebrated proof that injective II 1 factors are hyperfinite by enabling property Γ to be obtained from semidiscreteness. This is the fount of central sequences, and the provision of these to a sufficient degree to show that an injective II 1 factor is McDuff (i.e., absorbs the hyperfinite II 1 factor tensorially) is the goal of the next stages of Connes' argument. Spectral gap methods have subsequently been a key tool in Popa's deformationrigidity theory (see [48], for example), and today, the study of central sequences in II 1 factors remains a focus of research ( [26]). This paper focuses on the uniform version of property Γ for C *algebras introduced in our recent work with Winter ([10]). Uniform property Γ requires that we can approximately divide elements in a C *algebra in a central fashion and uniformly in all traces; this amounts to simultaneously (and uniformly) witnessing that every II 1 factor representation has Dixmier's formulation of property Γ. Uniform property Γ differs from both the pointwise version of property Γ for C * -algebras used to obtain similarity degree estimates 1 ( [49]), and from a version of property Γ used to preclude certain tensorial absorption phenomena ( [24]). 2 Our motivation comes from the Toms-Winter regularity conjecture, which seeks to identify through abstract conditions those simple nuclear C * -algebras which are accessible to classification by K-theory and traces (see [78,Section 5], for example, for a full discussion of this conjecture, which we also discuss further in Section 5). A general localto-global argument provides the key ingredient in the passage from tensorial absorption to finite nuclear dimension in [10], and it turns out that uniform property Γ is the natural abstract condition which facilitates this. Our purpose here is to undertake an in-depth study of uniform property Γ in its own right and set out its relation to the Toms-Winter conjecture for simple, separable, nuclear C * -algebras.
In Section 2, we give a number of equivalent reformulations in the spirit of Dixmier's original work for factors. In Section 3, we discuss the subtle issue of tracial factorisation: if (e n ) ∞ n=1 is a central sequence in a unital C * -algebra A, and a ∈ A, what can we say about τ (ae n ) for a trace τ on A? This leads to a simplification of the definition of uniform property Γ in the case when the trace space of A is a Bauer simplex. In this case, one only has to ask for the unit of A to be approximately centrally divisible uniformly in trace (Corollary 3.2). 3 The next phase of the paper is devoted to the role of uniform property Γ in the structure theory of simple nuclear C * -algebras. In Section 1 This asks for every II 1 factor representation to have property Γ, but does not require any compatibility between how property Γ occurs in these representations. 2 This is a modification of Murray and von Neumann's original definition, asking for approximately central unitaries which are uniformly zero in all traces; the main application is to show that C * -algebras with unique trace, whose II 1 factor representation does not have property Γ, cannot absorb the Jiang-Su algebra tensorially. 3 As just stated, A needs to be unital, but the more precise statement in Corollary 3.2 does not require this. 4, we provide a converse to the main technical result of [10], and using Matui and Sato's breakthrough [40], characterise Z-stability for simple, separable, nuclear C * -algebras with non-empty and compact tracial state space, as the combination of strict comparison and uniform property Γ. This can be thought of as analogous to Connes' passage from property Γ to McDuff in his work on injective factors. Indeed, we formalise the notion of a C * -algebra being uniformly McDuff, a concept which implicitly played major roles in [75] and [40]. Then Theorem 4.6 provides the passage from uniform property Γ to uniform McDuffness for non-elementary, separable, nuclear C * -algebras. In particular, this sheds light on the remaining open implication in the Toms-Winter conjecture: the missing piece is uniform property Γ. We discuss the role of property Γ, and divisibility conditions more generally, in the Toms-Winter conjecture in Section 5. Here, we highlight the following statement, extracted from Theorem 5.6, which combines the work in Section 4, with a wealth of developments over the last decade including [75,40,41,33,61,4,10].
Theorem A. The Toms-Winter conjecture holds among separable, simple, nuclear, non-elementary C * -algebras which have uniform property Γ.
Examples of separable nuclear C * -algebras with property Γ are by now abundant. Kerr and Szabó establish uniform property Γ for crossed product C * -algebras arising from a free action with the small boundary property of an infinite amenable group on a compact metrisable space [29,Theorem 9.4]. Combining Theorem A with this result, the Toms-Winter conjecture then holds for the simple C * -algebras one obtains from free minimal actions with the small boundary property -minimality giving rise to simplicity of the crossed product. This was also recorded in [29], which used an early version of this paper for Theorem A.
Corollary B (c.f. [29,Corollary 9.5]). Let G X be a free minimal action of an infinite amenable group on a compact Hausdorff space with the small boundary property. Then the Toms-Winter conjecture holds for C(X) ⋊ G.
With hindsight, the papers [59,33,70] all establish uniform property Γ for separable nuclear C * -algebras whose extremal traces are compact and finite dimensional. This is why strict comparison implies Jiang-Su-stability for these algebras. But uniform property Γ holds outside this situation too: various diagonal AH-algebras, including Villadsen algebras of infinite nuclear dimension have uniform property Γ (see Proposition 5.10). Indeed, it is open whether every simple, separabe, infinite dimensional nuclear C * -algebra A has property Γ.
Preliminaries
Throughout the paper, a trace on a C * -algebra A means a tracial state. We write T (A) for the set of traces on A endowed with the weak * topology. Our framework is the setting where T (A) is non-empty and compact (as is the case when A is unital), so that T (A) becomes a Choquet simplex [57, Theorem 3.1.14] which is metrisable when A is separable. We will often at least implicitly assume that A has no (nonzero) finite dimensional representations, as this will be an immediate consequence of the definition of uniform property Γ. Thus this framework roughly corresponds to the II 1 factor setting where property Γ first originated.
Let A be a C * -algebra with T (A) = ∅. Each τ ∈ T (A) induces a seminorm given by a 2,τ := τ (a * a) 1/2 for a ∈ A. Given a non-empty subset X ⊂ T (A), we obtain a uniform 2-seminorm associated to X by The seminorm · 2,X is dominated by the operator norm · , and we have We will primarily be interested in the case that X = T (A), when the seminorm · 2,T (A) is known as the uniform trace seminorm. In many situations of interest this seminorm is actually a norm. 4 In general, the · 2,T (A) -null elements of A form an ideal by (1.2) and · 2,T (A) descends to a norm on the quotient. We will work with two flavours of central sequence algebras in the paper. To set this up, we let ω ∈ βN \ N denote a fixed free ultrafilter. The ultrapower of a C * -algebra A is defined by .
Since a 2,T (A) ≤ a for all a ∈ A, there are canonical surjections As is standard, we use representative sequences in ℓ ∞ (A) to denote elements of A ω and A ω respectively. Recall that a key property of ultrapowers is countable saturation, which loosely speaking enables one to pass from approximate satisfaction of certain properties to exact satisfaction. The precise technical tool we use to do this is Kirchberg's ǫ-test ([31, Lemma A.1]). Given a sequence (τ n ) ∞ n=1 in T (A), one can form a trace on ℓ ∞ (A) by (1.5) (a n ) ∞ n=1 → lim n→ω τ n (a n ), which induces traces on the ultrapowers A ω and A ω . Traces on A ω and A ω arising in this manner are called limit traces. We write T ω (A) for the set of limit traces on either A ω or A ω . Notice that essentially by construction · 2,Tω(A) is a norm on A ω . By [10,Proposition 1.11], compactness of the trace space precisely characterises unitality of the uniform tracial ultrapower. This is the reason behind our choice of framework for the study of uniform property Γ. Proposition 1.1. Let A be a separable C * -algebra with T (A) compact and non-empty. Then A ω is unital. Moreover, if (e n ) ∞ n=1 is an approximate unit for A, then e n − 1 A ω 2,Tω(A) → 0.
Proof. The unitality of A ω is [10,Proposition 1.11], and the proof of this proposition shows that (e n ) ∞ n=1 represents the unit of A ω , which implies that lim n→ω inf τ ∈T (A) τ (e n ) = 1. From this, and using that each e n is a positive contraction, the claimed limit holds.
We identify A with the subalgebra of A ω coming from constant sequences, giving rise to the norm central sequence algebra A ω ∩A ′ . When T (A) is non-empty, we have a canonical map ι : A → A ω which is an embedding when · 2,T (A) is a norm. By abuse of notation we write A ω ∩ A ′ for the uniform tracial central sequence algebra; strictly speaking, when While it is immediate that A ω quotients onto A ω , the corresponding result for central sequence algebras is deeper, and was established by Kirchberg and Rørdam in [33], building on an observation of Sato [58]. The result below is a combination of [33,Proposition 4.5(iii) and Proposition 4.6] (working in the minimal unitisation if A is not unital). Lemma 1.2 (Central Surjectivity). Let A be a separable C * -algebra with T (A) compact and non-empty. Then the canonical map A ω ∩A ′ → A ω ∩ A ′ is a surjection.
For a C * -algebra A and a tracial state τ ∈ T (A), we let π τ : A → B(H τ ) denote the GNS representation. Recall (for context only) that π τ (A) ′′ is a factor if and only if τ is an extremal trace ([16, Théorème 6.7.3]). However in this paper, we often have to work with non-extremal traces and hence general finite von Neumann algebras. As every finite type I von Neumann algebra has a non-zero finite dimensional normal representation, GNS-representations associated to traces on C *algebras without finite dimensional quotients give rise to type II 1 von Neumann algebras. Proposition 1.3. Let A be a C * -algebra with no non-zero finite dimensional quotients and let τ ∈ T (A). Then π τ (A) ′′ is a type II 1 von Neumann algebra.
It is a consequence of Connes' celebrated equivalence of injectivity and hyperfiniteness ([13, Theorem 6]), that injective type II 1 factors are McDuff, in that they tensorially absorb the hyperfinite II 1 factor, or equivalently by McDuff's theorem from [42], admit approximately central unital matrix embeddings. 5 In Section 4, we require the analogous version of this fact for general injective type II 1 von Neumann algebras. This is well-known to experts, but does not appear to be in the literature, so we briefly sketch the details, starting by recording two standard facts regarding type II 1 von Neumann algebras. Firstly, these admit unital embeddings of matrix algebras; see, for example, [64,Proposition V.1.35], which proves the following lemma in the case n = 2 (and whose proof can be modified to handle the general case). Proof. Without loss of generality, assume 1 M ∈ F , and decompose F as K k=1 M n k . Then, writing e (k) ij for the matrix units of the k-th summand of F , it is routine to check that (see for example the proof of Lemma 3.21 of [4], where such an isomorphism is given in equation (3.47)), whence the result follows.
When M is a finite von Neumann algebra with a fixed faithful normal trace τ , we use the notation M ω to denote the tracial von Neumann algebra ultrapower, which is naturally equipped with the trace given on representative sequences by τ ω ((x n ) ∞ n=1 ) = lim n→ω τ (x n ). This is the same notation as we use for the uniform tracial ultrapower of a C * -algebra, motivated by the special case when A has a unique trace τ , where the uniform tracial ultrapower A ω is (by Kaplansky's density theorem) nothing but the tracial von Neumann algebra ultrapower (π τ (A) ′′ ) ω . It will always be clear from context when we are applying a tracial von Neumann algebra ultrapower rather than a uniform tracial ultrapower.
With this setup we now record the McDuffness of injective type II 1 von Neumann algebras in the form of approximately central matrix embeddings. Proposition 1.6. Let M be an injective type II 1 von Neumann algebra with separable predual and fixed faithful normal trace τ . Then for each n ∈ N, there exists a unital embedding M n → M ω ∩ M ′ .
is an increasing sequence of finite dimensional subalgebras. For each k ∈ N, by Lemmas 1.4 and 1.5, there exists a unital embedding φ k : M n → M ∩ F ′ k . The image of the induced map Φ : M n → M ω commutes with each F k ⊆ M ω , so commutes with M ⊆ M ω .
We end the preliminaries by recording two facts related to orthogonality and order zero maps. Recall that positive elements in a C *algebra are said to be orthogonal if ab = 0. We record the following elementary fact for future use. A completely positive map φ : A → B between C * -algebras is said to be order zero if it preserves orthogonality, i.e., φ(a)φ(b) = 0 whenever ab = 0; see [79] for the structure theory of order zero maps. For us, a key tool is the following order zero lifting theorem, essentially due to Loring ([39, Theorem 4.9]), although we use a reformulation from [74]. . Let A be a C * -algebra and I ⊆ A an ideal. Then for any n ∈ N, every completely positive and contractive (c.p.c.) order zero map M n → A/I lifts to a c.p.c. order zero map M n → A.
Equivalent formulations of uniform property Γ
This section is concerned with establishing some equivalent formulations of uniform property Γ in the spirit of Diximier's original analysis of central sequence algebras of II 1 factors in [17]. We begin by recalling the definition of uniform property Γ from [10, Definition 2.1].
Definition 2.1 ([10, Definition 2.1]). Let A be a separable C * -algebra with T (A) nonempty and compact. Then A is said to have uniform property Γ if for all n ∈ N, there exist projections p 1 , . . . , p n ∈ A ω ∩ A ′ summing to 1 A ω , 6 such that The definition above does not depend on the choice of ω, and indeed standard methods show that it has a equivalent local formulation. Proposition 2.2. Let A be a separable C * -algebra with T (A) nonempty and compact. Then the following are equivalent: (i) A has uniform property Γ (ii) For any finite subset F ⊂ A, any ǫ > 0, and any n ∈ N, there exist pairwise orthogonal positive contractions e 1 , . . . , e n ∈ A such that for i = 1, . . . , n and a ∈ F , we have (iii) For any finite subset F ⊂ A, any ǫ > 0, and any n ∈ N, there exist pairwise orthogonal positive contractions e 1 , . . . , e n ∈ A such that for i = 1, . . . , n and a ∈ F , we have Proof. As · 2,T (A) ≤ · , we have (iii)⇒(ii). Moreover, (ii)⇒(i) is a routine application of Kirchberg's ǫ-test ([31, Lemma A.1]) which we omit. If (i) holds, fix n ∈ N, and p 1 , . . . , p n ∈ A ω ∩ A ′ witnessing uniform property Γ. Then we can use central surjectivity (Lemma 1.2) and liftability of orthogonal positive contractions to find pairwise orthogonal positive contractions q 1 , . . . , q n ∈ A ω ∩ A ′ lifting p 1 , . . . , p n . These can in turn be lifted to pairwise orthogonal representing sequences (e 1,l ) ∞ l=1 , . . . , (e n,l ) ∞ l=1 in ℓ ∞ (A). For ω-many l, and hence for at least one l, e 1,l , . . . , e n,l must satisfy (2.3).
The next result is our analogue of [17, Proposition 1.10], which says that if a central sequence algebra of a II 1 factor is non-trivial, then it is automatically diffuse. In particular, the a priori weaker condition (ii) ensures that it is only necessary to consider n = 2 in the definition of uniform Property Γ. Proposition 2.3. Let A be a separable C * -algebra with T (A) nonempty and compact. The following are equivalent.
(i) A has uniform property Γ.
(ii) ⇒ (iii): Fix λ such that (ii) holds and define a trace τ λ : Indeed, given such an S, let (s (m) ) ∞ m=1 be a · 2,Tω(A) -dense sequence in S, and let (s Thus, for each n ∈ N, we can find some r(n) ∈ N, such that Finally, we take φ S : C 2 → A ω ∩ S ′ to be the unital * -homomorphism determined by φ S (1, 0) = q. This establishes (ii ′ ). We now deduce (iii) from (ii ′ ) by an iterative argument. To this end, fix a · 2,Tω(A) -separable subalgebra S ⊂ A ω as in part (iii), and define S 0 := S. For i ≥ 1, recursively use condition (ii ′ ) to construct φ i satisfying (ii ′ ) for the set S i−1 and take Since the φ i commute by construction, together they define a unital * -homomorphism φ : is the infinite product of τ λ . Since the unit ball of A ω is · 2,T (A) -complete by [10, Lemma 1.6], so too is the unit ball of A ω ∩S ′ . Therefore we can extend φ by continuity to a unital * -homomorphism from the tracial von Neumann algebra (M, τ where µ is integration against Lebesgue measure. This gives (2.5).
Remark 2.4. Note that the proof of Proposition 2.2 also gives local versions of condition (ii) above, e.g., a separable C * -algebra A with T (A) non-empty and compact has uniform property Γ if and only if for some λ ∈ (0, 1), for every finite subset F ⊂ A and every ǫ > 0, there is a positive contraction e ∈ A such that
Bauer simplices and tracial factorisation
In this section, we set out how uniform property Γ simplifies when T (A) is a Bauer simplex, i.e., when the extreme boundary ∂ e T (A) is compact (and non-empty). Uniform property Γ requires that arbitrary finite subsets of A can be approximately divided in trace in an approximately central fashion. When T (A) is Bauer, it is only necessary to tracially divide the unit (as in Corollary 3.2 below); since the division is tracial this can be phrased without explicit reference to a unit, and indeed it applies equally to non-unital A (provided T (A) is compact and non-empty).
At the heart of this observation is the following 'tracial factorisation' result for central sequences on extreme traces.
If for some τ ∈ ∂ e T (A), we have lim n→ω ab n − b n a 2,τ = 0 for all a ∈ A, then for all a ∈ A, for any limit trace τ ∈ T ω (A) arising from a sequence of traces in K, i.e., (3.1) converges uniformly on K (for fixed a).
Proof. We prove the second statement, as it evidently implies the first. Without loss of generality, we assume that b n ∈ A + for all n ∈ N. Fix a compact subset K ⊆ ∂ e T (A). Suppose there exists a ∈ A for which it is not the case that Then there exists ǫ > 0 and a sequence ( Let τ ∈ T ω (A) be the limit trace defined by the sequence (τ n ) ∞ n=1 and In particular, τ A is an extremal tracial state. Next, consider the tracial functional σ : for all x ∈ A. Taking x = a gives the required contradiction.
Using the above proposition, we now show that in the Bauer case, to check uniform property Γ, it suffices to tracially divide the unit in an approximately central fashion.
Proof. The forward implication is immediate. For the converse, let a ∈ A and i ∈ {1, . . . , n}.
In the nontrivial case that both σ, ρ are nonzero, τ A can then be written as a convex combination from which it follows that τ A is a scalar multiple of σ. 8 When A is unital, this is immediate. In general, let (e m ) ∞ m=1 be an increasing approximate unit for A. By Dini's Theorem, ρ(e m ) ր 1 uniformly for ρ ∈ K. We then compute that |τ n (b n ) − τ n (e m b n )| ≤ b (1 − τ n (e m )). Thus, τ (b) = lim m→∞ σ(e m ) = lim m→∞ ατ A (e m ) = α.
Using convex combinations and continuity we see that for σ ∈ T (A) and k ∈ N, we have and therefore, Similarly, when T (A) is a Bauer simplex, it suffices to consider the case a := 1 A ∼ in the equivalent formulations of uniform property Γ established in Proposition 2.3 (where A ∼ means the minimal unitisation).
The convergence in Proposition 3.1 need not be uniform over ∂ e T (A) when ∂ e T (A) is not compact, as the example below shows. Example 3.3. Consider the subhomogeneous C * -algebra defined by (3.11) Every tracial state τ on A is of the form Therefore, the extreme traces on A are given by and e 11 is the (1, 1)-matrix unit. Then for any a = (λ, µ, Hence, the convergence in Proposition 3.1 is not uniform over ∂ e T (A).
Remark 3.4. Note in the example above, we even have τ (b n ) = 1 2 for all τ ∈ T (A). The C * -algebra A ⊗ Q (where Q is the universal UHFalgebra) is an example of the same phenomenon that additionally has no finite-dimensional representations (and which certainly has property Γ, by virtue of being Q-stable). 9 It is also possible to produce a simple AF C * -algebra with the same phenomenon, although our approach to this relies on forthcoming work of the authors as well as the concept of the "uniform trace norm completion" which we haven't introduced in this paper. 10 However, Example 3.3 does not preclude the possibility that Corollary 3.2 generalises to the case where T (A) is a general Choquet simplex.
Question 3.5. Let A be a separable C * -algebra with no non-zero finite dimensional representations, and T (A) compact and non-empty. Suppose for some n ∈ N, there are projections p 1 , . . . , p n ∈ A ω ∩ A ′ summing to 1 A ω with τ (p i ) = 1 n for each i and each τ ∈ T ω (A). Must A have uniform property Γ?
Uniform property Γ and complemented partitions of unity
When we introduced uniform property Γ with Winter in [10], our motivation was that it enabled us to access the key technical tool of [10] -complemented partitions of unity (CPoU) -which in turn are key to passing from Z-stability to finite nuclear dimension. Our goal in this section is to prove the converse of [10, Theorem 3.8] (and in fact obtain a stronger result), which will enable us to show how uniform property Γ and other tracial divisibility conditions play a role in the Toms-Winter conjecture in the following section. We recall the definition of CPoU below; see the discussion before and after [10,Definition G] for the motivation behind this definition. . Let A be a separable C * -algebra with T (A) non-empty and compact. Then A has CPoU if and only if, for every family of positive elements a 1 , . . . , a k ∈ A ω and δ such that there exist pairwise orthogonal projections p 1 , . . . , p k ∈ A ω which sum to 1 A ω and have The main technical argument in this section is to use CPoU to glue local McDuffness at each tracial fibre in Lemma 4.4 to obtain the following global McDuffness property. We now give our gluing lemma. We could use CPoU via the local to global polynomial test of [10, Lemma 4.1] to establish this, but prefer to show how CPoU is used directly. Proof. Fix n ∈ N, and fix a finite generating set F for M n . To establish the result, it suffices to show that given a finite subset G of A and ǫ > 0, there exists a c.p.c. order zero map φ : M n → A ω such that for all x ∈ F , b ∈ G, Once this is done, Kirchberg's ǫ-test ([31, Lemma A.1]) gives a unital c.p.c. order zero map ψ : M n → A ω ∩A ′ , and hence a * -homomorphism. 13 Therefore we fix a finite subset G of A and a tolerance ǫ > 0. Set By taking a suitable element of an approximate unit for A, we can find a contraction e ∈ A + with 1 A ω − e 2,Tω(A) < η by Proposition 1.1.
For the moment, fix τ ∈ T (A) and set M τ := π τ (A) ′′ . By hypothesis, there exists a unital embedding θ : is surjective by the Kaplansky Density Theorem together with the existence of norm-preserving lifts. Therefore, by order zero lifting (Proposition 1.8), there exists a sequence (θ l : M n → A) ∞ l=1 of c.p.c. order zero maps which induces θ. Setting φ τ := θ l for a suitable value of l we can arrange to have for all x ∈ F , b ∈ G. It follows that for all σ in a neighbourhood of τ in T (A), we continue to have We now use the compactness of T (A) to find c.p.c. order zero maps φ 1 , . . . , φ k : M n → A such that for every trace τ ∈ T (A) there exists i ∈ {1, . . . , k} such that for all x ∈ F , b ∈ G, Notice that, working in the minimal unitisation A ∼ , (4.7) implies that For i = 1, . . . , k, we now define and observe that when (4.7) holds, we get Thus, we conclude that Viewing a 1 , . . . , a k as elements of A ω (i.e., abusing notation and writing a i for ι(a i )), we have Using CPoU, we obtain orthogonal projections p 1 , . . . , p k ∈ A ω ∩ A ′ summing to 1 A ω such that Since the p i are orthogonal and commute with the images of the φ i , this is an orthogonal sum of c.p.c. order zero maps, so it is itself c.p.c. order zero. For x ∈ F , b ∈ G, first compute using that p i commutes with b. Next, since the p i are orthogonal, it follows that Using this, for τ ∈ T ω (A), we get, Since 1 A ω − e 2,Tω(A) < η < ǫ, we have φ(1 Mn ) − 1 A ω 2,Tω(A) < 2ǫ, as required.
In particular, nuclear C * -algebras with no finite dimensional quotients satisfy the fibrewise McDuffness hypothesis of the previous lemma.
The Toms-Winter Conjecture
In this section, we discuss the remaining open implication in the Toms-Winter conjecture, and relate known results in this direction to uniform property Γ. (i) A has finite nuclear dimension.
where Z is the Jiang-Su algebra of [28].
(iii) A has strict comparison (described below).
The third condition, strict comparison, is a condition on positive elements in the stabilisation of A, which can be described in terms of the Cuntz semigroup. Recall that the Cuntz semigroup, Cu(A), of a C * -algebra A is built from equivalence classes of positive elements of the stabilisation A ⊗ K of A as follows: for a, b ∈ (A ⊗ K) + , write a b if and only if there is a sequence (x n ) ∞ n=1 in A ⊗ K with x n bx * n → a, and define an equivalence relation a ∼ b if and only if a b and b a. Then Cu(A) := (A ⊗ K) + / ∼. This is an ordered abelian semigroup. 14 See [2] for a survey. A functional on Cu(A) is an ordered semigroup homomorphism φ : Cu(A) → [0, ∞] which preserves increasing sequential suprema (when A is unital, a functional is called normalised if it maps the class of the unit to 1), and the collection of functionals is denoted F (Cu(A)). 15 A simple C * -algebra has strict comparison when 14 Addition of x, y ∈ Cu(A) is defined by x + y = [a + b], where a, b ∈ (A ⊗ K) + are orthogonal representatives of x and y respectively. 15 In the unital case, functionals on Cu(A) come from 2-quasitraces as defined in [3, Definition II.1.1], and by [27] these are the same as traces for exact C * -algebras. Any τ ∈ T (A) (or more generally, any 2-quasitrace) extends uniquely to a densely defined lower semicontinuous trace (or 2-quasitrace) on A ⊗ K (also denoted by τ ), and for a ∈ (A ⊗ K) + , d τ (a) := lim n→∞ τ (a 1/n ) gives a well defined state on Cu(A); moreover every state arises in this fashion; see [20]. one can deduce x ≤ y in Cu(A) from knowing that φ(x) ≤ φ(y) for all φ ∈ F (Cu(A)), with strict inequality whenever φ(y) ∈ (0, ∞). It is essentially a result of Rørdam from [54] that simple Z-stable C * -algebras have strict comparison, and hence the implication (ii)⇒(iii) of Conjecture 5.1 was known to hold well before the conjecture was formulated. 16 Consequently, the remaining open part of the Toms-Winter conjecture is the implication (iii)⇒(ii).
In addition to the precise formulation of Conjecture 5.1, research has also focused on trying to obtain any Cuntz semigroup condition which characterises Z-stability for simple, separable, unital and nuclear C *algebras. Indeed, it has been suggested by Winter, at least as far back as the CBMS lecture series in 2011, that another potential condition is Cu(A) ∼ = Cu(A ⊗ Z), which is certainly the strongest reasonable such candidate (see for instance [78,Paragraph 5.4]). Experts have known for some time that Cu(A) ∼ = Cu(A ⊗ Z) represents the combination of strict comparison and an appropriate divisibility condition on the Cuntz semigroup. It is too much to ask for exact divisibility of Cu(A), which would mean that given any x ∈ Cu(A), and k ∈ N, one can divide x by k, i.e. find y ∈ Cu(A) with ky = x. For example, the unit of the Jiang-Su is not divisible in the Cuntz semigroup by any n ≥ 2. Thus, instead asks for 'almost-divisibility', the ability to divide elements into between k and k+1 pieces. There are a number of slightly differently formulations of this idea in the literature, 17 but based on developments in the setting of abstract Cuntz semigroups (see [1], for example) the prevailing definition today is that A is almost divisible if, for all [a] ∈ Cu(A), k ∈ N and ǫ > 0, there exists [b] ∈ Cu(A) such that In the presence of strict comparison, almost-divisibility, some of its earlier reformulations, and various other properties with similar flavours all become equivalent. We collect some of these facts below in the simple unital case, for completeness, and to set the scene for the discussion which follows. None of these are our results; most are 16 There is a small detail to watch out for when looking at earlier papers involving the Cuntz semigroup, such as [54] and [75]. These often use the 'incomplete' W (A) := ∞ n=1 M n (A) + / ∼ in place of Cu(A) = W (A ⊗ K) which came to the fore in [14]. 17 The concept of almost-divisibility had been around for some time before the present terminology stuck, showing up implicitly in [54], anonymously in [56] and pseudonymously in [2]. In It was first systematicaly investigated in [15], and given heavy impetus by Nate Brown who promoted the viewpoint that the rank problem stands analogous to the fact that every possible trace value is obtained by a projection in a II 1 factor. Dramatic progress has recently been obtained by Thiel in [66], who proved (independently of strict comparison) that (vi) holds for all simple, separable, unital, C * -algebras with stable rank one.
By construction, we have contraction e ∈ C * (a) such that ea ≈ η a, we have It follows that (φ(s)e) * (1 k+1 ⊗ b)(φ(s)e) ≈ ǫ a. Thus if (t n ) ∞ n=1 is a lift for φ(s)e, then for ω-almost all n, t * n (1 k+1 ⊗ b n )t n ≈ ǫ a, and so by [52, The argument is similar for k[b n ] ≤ [a]. First fix η 1 > 0 and s ∈ M 1×k (Z) such that s * s ≈ η1 1 k ⊗ h. Thus, if e ∈ C * (a) is a positive contraction such that ea ≈ η2 , where η 2 and η 1 are chosen small enough that 2 s 2 η 2 + η 1 < δ, where δ is as fixed in the previous paragraph, we have (eφ(s)) * a(eφ(s)) ≈ 2 s 2 η2 φ(s * )aφ(s) Thus if (t n ) ∞ n=1 is a lift for eφ(s), then for ω-almost all n, t * n at n ≈ δ 1 k ⊗ φ n (h) 1/2 aφ n (h) 1/2 , and so by [52,Proposition 2 In the terminology of [75], the algebras covered by the previous proposition are called pure C * -algebras. In recent work of Thiel, it is shown that condition (vi) of Proposition 5.2 holds for all simple, separable, unital, non-elementary C * -algebras with stable rank one [66]. In the direction of proving Z-stability from such Cuntz semigroup conditions, a groundbreaking result was achieved by Winter in [75] for unital C * -subalgebras of locally finite nuclear dimension; 20 we state the generalisation of this result to non-unital algebras below. The output of (2) and (3) then directly entails Z-stability. With hindsight, we now recognise the output of step (3) as uniform McDuffness.
Theorem 5.5 (Winter). Let A be a simple separable C * -algebra with T (A) non-empty and compact and which is tracially m-almost divisible for some m ∈ N and has locally finite nuclear dimension. Then A has uniform property Γ.
τ ∈ T ω (A). 22 This conclusion is equivalent to condition (ii) of Theorem 4.6, which immediately implies uniform property Γ.
A major breakthrough was made by Matui and Sato in [40], which had the effect of removing the locally finite nuclear dimension hypothesis from (2) above. Precisely they obtained a weak central version of comparison -property (SI); see [60,Definition 3.3] or [33, Definition 2.6] for the definition -from comparison (and as Kirchberg and Rørdam show, this even works for various weak versions of comparison as the hypothesis). As a consequence, in the presence of an appropriate central divsibility condition, Matui and Sato deduce (iii)⇒(ii) in the Toms-Winter conjecture. The condition needed is precisely that the C * -algebra is uniformly McDuff, which Matui and Sato showed was automatic in the presence of finitely many extremal traces. Using Theorem 4.6, the Toms-Winter conjecture holds in the presence of uniform property Γ.
Theorem 5.6. Let A be a simple, separable, unital, nuclear C * -algebra with T (A) = ∅. The following are equivalent: [40] (this precise statement can be found as [70,Theorem 2.6]) that A is Z-stable.
At the moment, it remains an open problem whether all infinite dimensional, simple, nuclear C * -algebras have uniform property Γ. Matui and Sato's work [40] was subsequently extended from finitely many extremal traces, to compact extremal tracial boundaries of finite covering dimension. With hindsight, the reason is that these algebras always have uniform property Γ.
Just as Matui and Sato were able to make a major breakthrough by removing the locally finite nuclear dimension hypothesis from Step (2) of the proof of Theorem 5.4, an important problem is whether it is possible to remove the locally finite nuclear dimension from Step 3 (and hence also from Theorem 5.4): Question 5.8. Suppose that A is a non-elementary simple, separable, unital and nuclear C * -algebra with a tracial divisibility property. Does A have uniform property Γ?
We round off this section by presenting a further family of examples of C * -algebras with uniform property Γ. This family includes the non-Z-stable Villadsen algebras of the first type constructed in [71]. 23 Definition 5.9. Let X and Y be compact Hausdorff spaces. A *homomorphism φ : C(X) → M n ⊗ C(Y ) is said to be diagonal if there exist continuous functions λ 1 , . . . , λ n : Y → X and matrix units e rs ∈ M n such that Matrix amplifications of diagonal maps are also said to be diagonal.
A C * -algebra that can be represented as an inductive limit of (trivial) homogeneous C * -algebra with diagonal connecting maps is called a diagonal AH algebra. For such algebras, uniform property Γ can often be verified explicitly, as we now show. Proposition 5.10 (cf. [22,Proposition 4.6.8]). Let A be given by the inductive sequence where the connecting map are diagonal and n i → ∞. Then A has uniform property Γ whenever T (A) is a Bauer simplex.
Proof. Since diagonal maps are unital, we have n i |n i+1 . As n i → ∞, we can refine the inductive sequence and assume without loss of generality that k i := n i+1 n i → ∞. (5.10) for some continuous functions λ 1 , . . . , λ k i : X i+1 → X i and matrix units e rs ∈ M k i .
Writing µ i : M n i ⊗ C(X i ) → A for the canonical maps into the inductive limit, we define By construction, p i is a projection in A that commutes with the image of µ i . As the sequence (p i ) ∞ i=1 is uniformly bounded, it follows that lim i→∞ [p i , a] = 0 for all a ∈ A.
Also by construction, we have that τ (q i ) ∈ [ 1 2 − 1 k i , 1 2 ] for all τ ∈ T (M n i+1 ⊗ C(X i+1 )). Since every trace on A pulls back to a trace on M n i ⊗ C(X i ) and k i → ∞, it follows Let p be the element of A ω induced by the sequence (p i ) ∞ i=1 . Then p ∈ A ω ∩ A ′ and τ (p) = 1 2 for all τ ∈ T ω (A). Since we are assuming that T (A) is a Bauer simplex, we have (5.14) τ (pa) = τ (p)τ (a) = 1 2 τ (a), a ∈ A, τ ∈ T ω (A), by Proposition 3.1. It now follows from Proposition 2.3(ii) that A has uniform property Γ.
In particular, the non-Z-stable Villadsen algebras of the first type constructed in [71,Section 8] have uniform property Γ since these are diagonal AH algebras with Bauer trace simplices (see [71,Section 8] and computations of [69,Theorem 4.1]).
Classification by embeddings, revisited
In 2015, the classification of simple, separable, unital C * -algebras of finite nuclear dimension in the UCT class was completed by combining [25,19,68] (themselves building on an enormous body of earlier work) through a combination of classification theorems for C * -algebras with good tracial approximations, and abstract machinary for accessing these. Now, using [10], Z-stability can be used in place of finite nuclear dimension to access classification. We end the paper by explaining how Z-stability and finite nuclear dimension are used in the tracial approximation approach to classification, and how the CPoU methods of [10] fit in. 24 Recall that for simple C * -algebras, both of the equivalent conditions of finite nuclear dimension or Z-stability give rise to a dichotomy: such algebras are either stably finite or purely infinite. 25 The purely infinite classification theorem was established by Kirchberg and by Phillips ([30,45]) in the 90's, so in the rest of this section we restrict to stably finite C * -algebras.
Internal approximations by building blocks are familiar in operator algebras, originating in Murray and von Neumann's work on hyperfinite II 1 factors and appearing prominently in the subsequent classification of approximately finite dimensional C * -algebras [18], and varioous more general inductive limit models. In [36], (building on Popa's earlier local quantisation in the setting of C * -algebras [47]), Lin developed a new kind of internal approximation property for C * -algebras, now known as tracial approximation. Let us recall the precise definitions in a form suitable for our discussion. Definition 6.1 (cf. [36,21]). Let S denote a class of separable unital C * -algebras closed under isomorphism. A simple unital C * -algebra A is tracially approximately S (TAS) if for every finite subset F ⊂ A, every ǫ > 0, and every positive element c ∈ A + , there is a projection p ∈ A and a unital C * -subalgebra B ⊂ pAp with 1 B = p and B ∈ S such that (i) pa − ap < ǫ for all a ∈ F , (ii) dist(pap, B) < ǫ for all a ∈ F , and (iii) 1 A − p is Murray-von Neumann equivalent to a projection in cAc.
When T (A) = ∅ and A has strict comparison of positive elements by traces (e.g., if it is unital, exact and Z-stable), one can replace (iii) by the following condition (where the tracial nature of the internal approximation is more evident).
A simple unital C * -algebra A is said to be rationally TAS, if A ⊗ U is TAS for every UHF algebra U of infinite type. 24 CPoU (and hence uniform property Γ as the tool for accessing it) is also crucial to the new abstract approaches to classification [6,7] emerging from Schafhauser's approach to the quasidiagonality theorem [62], and AF-embeddings [63]. 25 Both dichotomy results rely on work of Kirchberg Lin's original work [36] focused on tracially AF 26 (TAF) C * -algebras, firstly giving an abstract characterisation of the universal UHF-algebra within the class of simple, separable, unital, nuclear, TAF C * -algebras with the UCT, and then subsequently obtaining a complete classification of this class [38]. Under trace space conditions, this theorem can be accessed abstractly through finite decomposition rank (a precursor of finite nuclear dimension from [34], which entails quasidiagonality of all traces); see [73].
Over time the class of algebras classified via by tracial approximation methods expanded, culminating in Gong-Lin-Niu's classification in [25] of simple, separable, unital, Z-stable, nuclear C * -algebras with the UCT which are rationally TAS, where S denotes the class of so called Elliott-Thomsen building blocks. 27 (This class exhausts the Elliott invariant by [25,Theorem 13.46].) Here, Z-stability plays a key role, as it allows asymptotic classification results to feed into Winter's strategy from [76] to pass from the TAS class to the (larger) Z-stable, rationally TAS class. On the other hand, finite nuclear dimension does not play a role in this argument, only Z-stability; thus if one works with finite nuclear dimension as the regularity hypothesis in classification via tracial approximations, it is necessary to use Winter's long and difficult 'Z-stability' theorem from [75].
The other major component in the tracial approximation approach to C * -algebra classification is an abstract criterion for identifying the rationally TAS class. A new approach to problems of this nature was pioneered by Matui and Sato in [41, Theorem 6.1], where they obtained rationally TAF approximations for a simple, nuclear, quasidiagonal C * -algebra with a unique tracial state, directly from quasidiagonality. Around the same time, Winter developed his 'classification by embeddings' technique ([77, Theorem 2.2]) which allows one to use finite nuclear dimension to convert certain tracial approximate factorisations to tracial approximations in the sense of Definition 6.1. 28 In [19], Elliott, Gong, Lin, and Niu used a quasidiagonality hypothesis in the spirit of Matui-Sato (the precise condition is quasidiagonality of all traces), to obtain the required input to the classification by embeddings theorem. Simultaneously, this quasidiagonality hypothesis was shown to be automatic in the presence of the UCT [68]. In summary, these papers come together to show that simple, separable, unital C * -algebras of finite nuclear dimension in the UCT class are rationally TAS, and hence classified by [25]. In this latter (classification by embeddings) 26 Note that, within the class of unital simple C * -algebras, the property of being TAF is equivalent to Lin's later notion of tracial topological rank zero of [35, Definition 3.1]; see [35,Theorem 7.1]. 27 These are also known as point-line algebras or 1-dimensional NCCW complexes. The exact form of these building blocks is not needed here. 28 See the statement of Theorem 6.2 below for a more precise statement part of the argument, finite nuclear dimension rather than Z-stability is the regularity hypothesis being directly employed. We now know, by the main results of [10,8], that finite nuclear dimension and Z-stability are equivalent for simple, separable, nonelementary, nuclear C * -algebras. So following the outline above, to classify from the hypothesis of Z-stability, we should obtain finite nuclear dimension from [10], and feed this into the classification by embeddings theorem. However it turns out, as we now show, that the methods of [10] and [4] also give a direct proof of the classification by embeddings theorem from Z-stability without requiring the hypothesis of finite nuclear dimension, and so one does not need the full strength of finite nuclear dimension implies Z-stability for such classification. This is inspired by Matui and Sato's strategy from [41]. Specifically, this has the effect of removing the finite nuclear dimension hypothesis from [77, Theorem 2.2].
Theorem 6.2. Let S be a class of separable, unital C * -algebras which is closed under isomorphism and matrix amplifications. Let A be a simple, separable, nuclear, unital C * -algebra with T (A) = ∅ and let be a system of maps with the following properties: Then, A is rationally TAS.
Proof. Fix a UHF algebra U of infinite type. Without loss of generality, we may assume that (iii) is strengthened to σ i is u.c.p. for each i ∈ N. Indeed, by assumption (iv) we have lim for ω-many i (and replace the remaining σ i by arbitrary u.c.p. maps), which does not change σ, and thus does not change (iv) or (v). (Conditions (i)-(ii) are completely unaffected.) In the rest of the proof, we shall view A and A ω ⊗ U as subalgebras of A ω and (A ⊗ U) ω respectively, in the natural ways. Set φ i := ρ i • σ i for i ∈ N and let Φ : A → A ω be the map induced by the φ i . Our assumptions ensure that Φ is a * -homomorphism satisfying Let π : A → M 2 ((A ⊗ U) ω ) be the * -homomorphism defined by Applying [10, Lemma 4.7], 29 we see that C has strict comparison of positive elements by bounded traces and T (C) is the closed convex hull of the set of traces of the form τ (π(a)·) where τ ∈ T (M 2 ((A ⊗ U) ω )) and a ∈ A + satisfies τ (π(a)) = 1.
Fix 0 < ǫ < 1. Let p be a projection in U with 1 − ǫ < tr U (p) < 1. Define projections h 1 , h 2 ∈ C by We wish to argue now that τ (h 1 ) < τ (h 2 ) for all τ ∈ T (C). First, consider a trace of the form µ = τ (π(a)·) where τ ∈ T ω (M 2 ((A ⊗ U))) and a ∈ A + satisfies τ (π(a)) = 1. As M 2 and U have unique tracial states tr M 2 and tr U respectively, and that 29 Checking the conditions carefully: A is separable, unital, and nuclear; B n := A ⊗ U is simple, separable, unital, and Z-stable; B n has T (B n ) = QT (B n ) and CPoU by virtue of being additionally nuclear [10,Theorem 3.7], and T (B n ) = T (A) = ∅ is assumed (the implicit assumption that T (B n ) is compact, contained in the definition of CPoU follows from unitality); π is a * -homomorphism. For any non-zero a ∈ A, since M 2 ⊗ A ⊗ U is simple and π(a) ≥ 0 0 0 a ⊗ 1 U , it follows that π(a) is full.
Since h 1 and h 2 are projections, we get (6.10) for all µ ∈ T (C). By strict comparison, h 1 is Cuntz subequivalent to h 2 . For projections, Cuntz subequivalence implies Murray-von Neumann subequivalence (by [52, Proposition 2.1]). Therefore, there exists v ∈ C with v * v = h 1 and vv * ≤ h 2 . These conditions force v to take the form where w ∈ (A ⊗ U) ω satisfies w * w = Φ(1 A ) ⊗ p. The condition that v commutes with π(A) implies that (6.12) w(Φ(a) ⊗ 1 U ) = (a ⊗ 1 U )w, a ∈ A, from which we deduce that ww * commutes with A ⊗ 1 U . The element Φ(1 A ) ⊗ p of (A ⊗ U) ω is represented by the sequence of positive elements (φ i (1 A ) ⊗ p) ∞ i=1 . In fact, (6.13) since σ i is unital, and this is a projection since ρ i is a * -homomorphism. By standard stability results (e.g., combining the proofs of [55, Propositon 2.2.6 and Lemma 6.3.1]), the partial isometry w ∈ (A ⊗ U) ω with source projection Φ(1 A ) ⊗ p lifts to a sequence (w i ) ∞ i=1 of partial isometries satisfying w * i w i = φ i (1 A ) ⊗ p = ρ i (1 B i ) ⊗ p for each i ∈ N. For i ∈ N, defineρ i : B i → A ⊗ U by (6.14) which is a * -homomorphism since ρ i is a * -homomorphism and w * i w i commutes with ρ i (B i ) ⊗ 1 U . Also, since ρ i is injective and w * iρ i (b)w i = ρ i (b) ⊗ p, it follows thatρ i is injective.
This is a C * -subalgebra of A⊗U isomorphic to B i , so that D i ∈ S. For i ∈ N, set q i := w i w * i = 1 D i . Then, q i is a projection in A ⊗ U and, for any τ ∈ T (A ⊗ U), i→ω −−→ tr U (p), (6.15) using (v) in the last line, and thereby obtaining uniform convergence over τ ∈ T (A ⊗ U). In addition, since ww * commutes with a ⊗ 1 U for all a ∈ A, we have Finally for a ∈ A, working in (A ⊗ U) ω , we have Thus, for all a ∈ A. Therefore, for any finite set F ⊂ A and for ω-many i, we have (i) q i (a ⊗ 1 U ) − (a ⊗ 1 U )q i < ǫ for all a ∈ F , (ii) dist(q i (a ⊗ 1 Q )q i , D i ) < ǫ for all a ∈ F , and (iii) τ (q i ) > 1 − ǫ for all τ ∈ T (A ⊗ U). This almost says that A⊗U satisfies the definition of TAS, except that the finite set is in A ⊗ 1 U rather than in A ⊗ U. Now, given a finite subset F ⊂ A ⊗ U and δ > 0, we can find a factorisation U = F ⊗ V, where F is a full matrix algebra and V ∼ = U, together with a finite subset F ′ of A⊗F ⊗1 V such that every element in F is within δ of a corresponding element of F ′ . Hence (by using A ⊗ F in place of A, and since S is closed under matrix amplifications), it follows that A ⊗ U is TAS. Remark 6.3. While the argument above and the proof of finite nuclear dimension from Z-stability in [10] both rely on the detailed analysis of relative commutants in ultrapowersà la Matui-Sato from [40,41], in fact the argument above is considerably more straightforward for two reasons. Firstly, since the map π in (6.3) is a * -homomorphism rather than an order zero map, we only need the versions of property (SI) for * -homomorphisms ([41, Lemma 3.2]) rather than the more elaborate order zero map property (SI) techniques of [4,Section 4]. Secondly, obtaining finite nuclear dimension from Z-stability requires a relatively sophisticated existence theorem ([10, Lemma 5.2], [4,Lemma 7.4]) to give rise to the eventual nuclear dimension approximations; in contrast, the corresponding part of Theorem 6.2 is the existence of Φ, which is automatic from the assumptions. Remark 6.4. There is another point which is worthy of note in contrasting the proof of Theorem 6.2 with Z-stability implies finite nuclear dimension. Both come down to comparison results for relative commutants of the form C in (6.4). In the nuclear dimension argument of [10] and [4], one takes two positive elements h 1 , h 2 in such a relative commutant of full spectrum, such that τ (h k 1 ) = τ (h k 2 ) for all k ∈ N and all τ ∈ T (C), and obtains unitary equivalence of h 1 and h 2 (see [4, Theorem 5.1]). By contrast, in the argument above, one has projections h 1 , h 2 ∈ C with τ (h 1 ) < τ (h 2 ) for all τ ∈ T (C), and obtains Murrayvon Neumann subequivalence h 1 h 2 from strict comparison of C. If it were possible to deduce that projections h 1 , h 2 ∈ C with τ (h 1 ) ≤ τ (h 2 ) satisfied h 1 h 2 , then the argument of Theorem 6.2 would give that A⊗Q is AS, not just TAS. In a C * -algebraic setting such a comparison result for projections (as opposed to positive elements of full spectrum) is too much to hope for; indeed TAF C * -algebras need not be AF. 30 However, in finite von Neumann algebras, tracial data does completely determine the Murray-von Neumann order on projections, and as such one can perform the final steps of Connes' original approach to his hyperfiniteness theorem in a very similar fashion to the proof of Theorem 6.2.
Precisely, let M be an injective II 1 factor with separable predual. in this relative commutant. 31 We note, finally, that by unpicking the analysis of traces on the algebra C in (6.4) (performed behind the scenes in [10,Lemma 4.7]), one finds that this relies heavily on the analysis of traces on von Neumann relative commutants of injective von Neumann algebras. | 2019-12-09T17:46:27.000Z | 2019-12-09T00:00:00.000 | {
"year": 2019,
"sha1": "07bc1e7302b2cb9a38fc052f7b9795d093c607be",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "07bc1e7302b2cb9a38fc052f7b9795d093c607be",
"s2fieldsofstudy": [
"Mathematics"
],
"extfieldsofstudy": [
"Mathematics"
]
} |
96422800 | pes2o/s2orc | v3-fos-license | Circuit complexity across a topological phase transition
Fangli Liu,1,2 Seth Whitsitt,1,2 Jonathan B. Curtis,1 Rex Lundgren,1,2 Paraj Titum,1,2,3 Zhi-Cheng Yang,1,2 James R. Garrison,1,2 and Alexey V. Gorshkov1,2 1Joint Quantum Institute, NIST/University of Maryland, College Park, Maryland 20742, USA 2Joint Center for Quantum Information and Computer Science, NIST/University of Maryland, College Park, Maryland 20742, USA 3Johns Hopkins University Applied Physics Laboratory, Laurel, Maryland 20723, USA
I. INTRODUCTION
In computer science, the notion of computational complexity refers to the minimum number of elementary operations for implementing a given task. This concept readily extends to quantum information science, where quantum circuit complexity denotes the minimum number of gates to implement a desired unitary transformation. The corresponding circuit complexity of a quantum state characterizes how difficult it is to construct a unitary transformation U that evolves a reference state to the desired target state [1,2]. Nielsen and collaborators used a geometric approach to tackle the problem of quantum complexity [3][4][5]. Suppose that the unitary transformation U (t ) is generated by some time-dependent Hamiltonian H (t ), with the requirement that U (t f ) = U (where t f denotes the final time). Then, the quantum state complexity is quantified by imposing a cost functional F [H (t )] on the control Hamiltonian H (t ). By choosing a cost functional that defines a Riemannian geometry in the space of circuits, the problem of finding the optimal control Hamiltonian synthesizing U then corresponds to finding minimal geodesic paths in a Riemannian geometry [3][4][5].
Recently, Nielsen's approach has been adopted in high-energy physics to quantify the complexity of quantum field theory states [6][7][8][9][10][11][12][13][14][15][16][17][18]. This is motivated, in part, by previous conjectures that relate the complexity of the boundary field theory to the bulk space-time geometry, i.e., the so-called "complexity equals volume" [19,20] and "complexity equals action" [21,22] proposals. Jefferson et al. used Nielsen's approach to calculate the complexity of a free scalar field [6], and found surprising similarities to the results of holographic complexity. A complementary study by Chapman et al., using the Fubini-Study metric to quantify complexity [23], gave similar results. Several recent works have generalized these studies to other states, including coherent states [8,24], thermofield double states [7,11], and free fermion fields [12][13][14]. However, the connection between the geometric definition of circuit complexity and quantum phase transitions has so far remained unexplored. This connection is important fundamentally, and it is also intimately related to the long-standing problem of quantum state preparations across critical points [25][26][27].
In this work, we consider the circuit complexity of a topological quantum system. In particular, we use Nielsen's approach to study the circuit complexity of the Kitaev chain, a prototypical model exhibiting topological phase transitions and hosting Majorana zero modes [28][29][30][31][32][33]. Strikingly, we find that the circuit complexity derived using this approach exhibits nonanalytical behaviors at the critical points for both equilibrium and dynamical topological phase transitions. Moreover, the optimal Hamiltonian connecting the initial and final states must be nonlocal in real space when evolving across a critical point. We further generalize our results to a Kitaev chain with long-range pairing, and we discuss universal features of nonanalyticities at the critical points in higher dimensions. Our work establishes a connection between geometrical circuit complexity and quantum phase transitions, and it paves the way toward using complexity as a tool to study quantum many-body systems.
II. THE MODEL
The one-dimensional (1D) Kitaev model is described by the following Hamiltonian [28,29]: where J is the hopping amplitude, is the superconducting pairing strength, μ is the chemical potential, L is the total number of sites (assumed to be even), andâ † j (â j ) creates (annihilates) a fermion at site j. We set J = 1 and assume antiperiodic boundary conditions (â L+1 = −â 1 ). Upon Fourier transforming, Eq. 1 can be written in the momentum basiŝ where k n = 2π L (n + 1/2) with n = 0, 1, . . . , L/2 − 1. The above Hamiltonian can be diagonalized via a Bogoliubov transformation, which yields the excitation spectrum: ε k n = (μ + cos k n ) 2 + 2 sin 2 k n . The ground state of Eq. (1) can be written as where tan(2θ k n ) = sin k n /(μ + cos k n ). A topological phase transition occurs when the quasiparticle spectrum is gapless [28], as illustrated in Fig. 1(a). The nontrivial topological phase is characterized by a nonzero winding number and the presence of Majorana edge modes [28][29][30][31][32][33].
III. COMPLEXITY FOR A PAIR OF FERMIONS
Since Hamiltonian (1) is noninteracting, the ground-state wave function (3) couples only pairs of fermionic modes with momenta ±k n , and different momentum pairs are decoupled. Hence, we first compute the circuit complexity of one such fermionic pair [12][13][14], and then we obtain the complexity of the full system by summing over all momentum contributions [6,23].
Let us consider the reference ("R") and target ("T ") states with the same momentum but different Bogoliubov angles: Expanding the target state in the basis of |ψ R and |ψ R ⊥ (i.e., the state orthogonal to |ψ R ), we have |ψ Now the goal is to find the optimal circuit to achieve the unitary transformation connecting |ψ R and |ψ T : where φ is an arbitrary phase. Nielsen approached this as a Hamiltonian control problem, i.e., finding a time-dependent Hamiltonian H k (s) that synthesizes the trajectory in the space of unitaries [3,4]: with boundary conditions U k (s = 0) = 1 and U k (s = 1) = U k . Here, ← − P is the path-ordering operator and O I are the generators of U (2). The idea is then to define a cost (i.e., "length") functional for the various possible paths to achieve U k [3,4,6,12]: D[U k ] = 1 0 ds I |Y I k (s)| 2 , and to identify the optimal circuit or path by minimizing this functional. The cost of the optimal path is called the circuit complexity C of the target state, i.e., Following the procedures in Refs. [12][13][14], one can explicitly calculate the circuit complexity for synthesizing the unitary transformation (4). For quadratic Hamiltonians, it is a simple expression that depends only on the difference between Bogoliubov angles (see Appendix A), Note that the complexity C for two fermions is at most π 2 /4, since | θ k | ∈ [0, π/2]. The maximum value is achieved when the target state has vanishing overlap with the reference state.
IV. COMPLEXITY FOR THE FULL WAVE FUNCTION
Given the circuit complexity for a pair of fermionic modes, one can readily obtain the complexity of the full many-body wave function. The total unitary transformation that connects the two different ground states [Eq.
where U k n , given by Eq. (4), connects two fermionic states with momenta ±k n . By choosing the cost function to be a summation of all momentum contributions [6,[12][13][14], it is straightforward to obtain the total circuit complexity, where θ k n is the difference of the Bogoliubov angles for momentum k n . In the infinite-system-size limit, the summation can be replaced by an integral, and one can derive that C ∝ L. This "volume law" dependence is reminiscent of the "complexity equals volume" conjecture in holography [19,20], albeit in a different setting. The circuit complexity given by Eq. (9) has a geometric interpretation, as it is the squared Euclidean distance in a highdimensional space [34]. The geodesic path (or optimal circuit) in unitary space turns out to be a straight line connecting the two points [i.e., H k (s) independent of s] (Appendix A). In the remainder of this paper, we demonstrate that the circuit complexity between two states is able to reveal both equilibrium and dynamical topological phase transitions.
We first choose a fixed ground state as the reference state and calculate the circuit complexities for target ground states with various chemical potentials μ T , crossing the phase transition point. The circuit complexity increases as the difference between the parameters of reference and target states is increased [ Fig. 1(b)]. More importantly, the complexity grows rapidly around the critical points (μ T = ±1), changing from a convex function to a concave function at the critical points. This is further illustrated in Fig. 1(c), where we plot the derivative (susceptibility) of circuit complexity with respect to μ T . The clear divergence at the critical points indicates that circuit complexity is nonanalytical at the critical points (see Appendix B for the derivation), and thus can signal the presence of a quantum phase transition. We emphasize that these features are robust signatures of phase transitions, which do not change if one chooses a different reference state in the same phase [see Figs. 1(b) and 1(c)].
We further plot θ k n versus the momentum k n for various target states (with a fixed reference state) in Fig. 1(d). When both states are in the same phase, θ k n first increases with momentum and finally decreases to 0 when k n approaches π . In contrast, when μ T is beyond its critical value, θ k n increases monotonically with momentum, and it takes the maximal value of π/2 at k n = π . This is closely related to the topological phase transition characterized by winding numbers, where the Bogoliubov angles of two different states end up at the same pole (on the Bloch sphere) upon winding half of the Brillouin zone if the states belong to the same phase [29]. Hence, the nonanalytical nature of the circuit complexity is closely related to the change of topological number (and topological phase transition).
Analytically, the derivatives of the circuit complexity (7) can be explicitly recast into a closed contour integral over the complex variable z = e ik (see Appendix B for detailed derivations). Depending on the parameters of the target states, the poles associated with the integrand are located inside or outside the contour. When the target state goes across a phase transition, the poles sit exactly on the contour, resulting in the divergence of the derivatives of the circuit complexity at critical points (Appendix B). Interestingly, the whole parameter space can be classified into four different phase regimes depending on which poles lie inside the contour (see Fig. 5 in Appendix B), which agrees exactly with the phase diagram shown in Fig. 1(a).
V. REAL-SPACE LOCALITY OF THE OPTIMAL HAMILTONIAN
Since the ground state [Eq. (3)] is a product of all momentum pairs, the optimal circuit connecting two different ground states corresponds to the following Hamiltonian: where τ i are the Pauli matrices, andψ k denotes the Nambu . By taking a Fourier series of the above optimal Hamiltonian, one can show that the Hamiltonian can be written in real space (see Appendix C for details): where θ (k) = 2 ∞ n=1 ω n sin(nk). One crucial observation is that when the two ground states are in the same phase, θ (0) = θ (π ) = 0 [see Fig. 1 hence the Fourier series of θ (k) converges uniformly. Therefore, the full series can be approximated by a finite order N * with arbitrarily small error. This immediately implies that the real-space optimal Hamiltonian (11) is local, with a finite range N * . In sharp contrast, if the two states belong 013323-3 to different phases, θ (π ) = π/2 = θ (0) = 0; the Fourier series of θ (k) converges at most pointwise. Thus the optimal Hamiltonian must be truly long-range (nonlocal) in real space (Appendix C), given that the total evolution time is chosen to be a constant [Eq. (5)]. Comparing to previous works on classifying gapped phases of matter using local unitary circuits [35][36][37], our results provide an alternative approach that has a natural geometric interpretation.
We take the initial state to be the ground state of a Hamil-tonianĤ i , and we consider circuit complexity growth under a sudden quench to a different Hamiltonian,Ĥ f . The reference and target states are chosen as the initial state | i and timeevolved state | (t ) , respectively. The time-dependent | (t ) can be written as [51,52] where θ k n is the Bogoliubov angle difference between eigenstates ofĤ i andĤ f , and ε k n and k n are the energy levels and normal mode operators, respectively, for the postquench Hamiltonian. Similar to the previous derivations, one can obtain the time-dependent circuit complexity, where φ k n (t ) = arccos √ 1 − sin 2 (2 θ k n ) sin 2 (ε k n t ). As shown in Fig. 3(a), the circuit complexity first increases linearly and then oscillates [9,10,15] before quickly approaching a time-independent value. The steady-state value of circuit complexity increases with μ f of the postquench Hamiltonian, until the phase transition occurs [ Fig. 3(a)].
VII. GENERALIZATION TO A LONG-RANGE KITAEV CHAIN AND HIGHER DIMENSIONS
We further give an example of a Kitaev chain with longrange pairing [53][54][55][56]: where d = min( , L − ). In contrast to the short-range model, the long-range model with α < 1 hosts topological phases with semi-integer winding numbers [53,56]. As one can see, the derivative of ground-state circuit complexity only diverges at μ T = 1 [ Fig. 4(a)], in contrast with Fig. 1(c). This agrees perfectly with the phase diagram for the long-range interacting model, where a topological phase transition occurs only at μ = 1 for α = 0 [56]. Figure 4(b) shows the long-time steady-state values of the circuit complexity after a sudden quench. Again, one observes nonanalytical behavior only at μ T = 1. While we have so far restricted ourselves to 1D, the results we found can be readily generalized to higher dimensions [57], for example to p + ip topological superconductors in 2D. The ground-state wave function of a p + ip superconductor essentially takes the same form as Eq. (3), with the momenta now being restricted to the 2D Brillouin zone, and tan(2θ k ) = | k |/ε k , where k and ε k denote pairing and kinetic terms in 2D. The circuit complexity can still be written as C = k | θ k | 2 = L 2 (2π ) 2 d 2 k| θ (k)| 2 . One can show again that the derivative of the circuit complexity is given by (see Appendix E) where E (k) 2 = (k) 2 + | (k)| 2 , and θ T (k) denotes the Bogoliubov angle for the target state. It is thus obvious that nonanalyticity happens at the critical point where E (k) = 0 (Appendix E).
VIII. CONCLUSIONS AND OUTLOOK
We use Nielsen's approach to quantify the circuit complexity of ground states and nonequilibrium steady states of the Kitaev chain with short-and long-range pairing. We find that, in both situations, circuit complexity can be used to detect topological phase transitions. The nonanalytic behaviors can be generalized to higher-dimensional systems, such as p + ip topological superconductors [58,59].
One interesting future direction is to use the geometric approach to quantify circuit complexity when the control Hamiltonians are constrained to be local in real space [37,60,61], and study its connection to quantum phase transitions [25,[62][63][64]. It would also be of interest to investigate the circuit complexity of interacting many-body systems. One particular example is the XXZ spin-half chain, whose low-energy physics can be modeled by the Luttinger liquid [65][66][67]. By restricting to certain classes of gates (i.e., by imposing penalties on the cost function) [3,6], it might be possible to find improved methods to efficiently prepare the ground state of the XXZ model by calculating the geodesic path in gate space.
Note added: Recently, we became aware of Ref. [68], which used revivals in the circuit complexity as a qualitative probe of phase transitions in the Su-Schrieffer-Heeger model. In contrast to that work, we have shown that the circuit complexity explicitly exhibits nonanalyticities precisely at the critical points for the Kitaev chain. We also became aware of Ref. [57], which numerically studied the complexity of a two-dimensional "d · τ " model. By contrast, here we analytically study the "p + ip" model, and we illustrate that the closing of the gap is essential for the nonanalyticity of circuit complexity.
APPENDIX A: DERIVATION OF CIRCUIT COMPLEXITY FOR A PAIR OF FERMIONS
In this Appendix, we present a detailed derivation of the circuit complexity for a pair of fermions, i.e., Eq. (7). This expression has previously been obtained using different approaches in Refs. [12][13][14]. To be comprehensive, here we provide a detailed derivation following Ref. [13]. We note that Ref. [12] provides an alternative derivation using a group theory approach.
By taking the derivative with respect to s in Eq. (5), we get the following expression: where U (s) is a unitary transformation that depends on s, and we have omitted the label k for notational clarity. The unitary U (s) can be parametrized in matrix form: where β, φ 1 , φ 2 , ω explicitly depend on the parameter s. The above matrix can be expressed in terms of the generators of U (2), which we choose as follows: Using the relation where φ 2 (0) is an arbitrary phase. Furthermore, we have the boundary condition at s = 1, The integrand in Eq. (A6) is a sum of four non-negative terms. Setting β(s) = φ 1 (s) = 0 and φ 2 (s) = π/2 minimizes (i.e., sets to zero) three of the four terms without imposing any additional constraints on the minimization of the remaining (dω/ds) 2 term. One can then easily check that the linear function w(s) = s θ minimizes the remaining term and yields (A10)
APPENDIX B: ANALYTICAL DERIVATION OF DIVERGENT DERIVATIVES IN GROUND STATES
In this Appendix, we provide a detailed analytical derivation to show that the first-order derivative indeed diverges at the critical points in the thermodynamic limit. We first derive how the derivative diverges when the reference state is in the trivial phase (|μ R | > 1), and then we generalize our results to show how this divergent behavior depends on the particular choice of the reference state. Throughout this Appendix, we assume the reference lies within a given phase, and we allow the target state to approach an arbitrary point in the phase diagram. Our analytical derivations show that these divergences necessarily map out the phase boundary, as illustrated in Figs. 2(a) and 2(b) and Fig. 5 below.
We begin with our general expression for the complexity as a function of our reference and target states. The Bogoliubov angle difference θ k for each momentum sector k can be expressed as and the circuit complexity is written in terms of θ k : Note that we have replaced the discrete sum in the main text with an integral for the thermodynamic limit, and written "C(| R gs → | T gs )" as "C" for brevity. Now we substitute Eq. (B1) into Eq. (B2), and we take the derivatives with respect to μ T and T . We obtain Here, we have used the fact that these functions are even in k to extend the integrals to −π . In spite of the complicated nature of these integrals, much can be learned about their analytic properties by recasting them as closed contour integrals in the complex plane. Defining the variable z = e ik , we find that the integrals take the form where the integration is taken counterclockwise over the contour |z| = 1. In this form, we may use the fact that the value of the integrals is entirely determined by the nonanalyticities of the integrand, which are located inside the contour, and that the value of the integration will only diverge if there is a divergence located on the contour. We proceed by defining the following variables: where a = R, T . From Eq. (B4), both derivatives contain simple poles at z i,T for i = 1, 2, 3, 4, while ∂ T C additionally has a simple pole at z = 0. Also, using the formula arctan(z) = FIG. 6. The deformation of the integration contour used to compute the gradients of the circuit complexity in the case μ T > 1. There is a branch cut running between the branch points z 1 and z 3 , where the imaginary part of the integrand is discontinuous and the integrand diverges near the branch points.
(i/2) ln 1−iz 1+iz , we can write the Bogoliubov angle as The important fact we will need is that the complex logarithm contains branch cuts running from the zeros to the infinities of its argument; therefore, the z ia are really branch points of the integrand. We now note that the derivatives of the complexity will only diverge if the couplings are tuned to a phase transition. This is because the z i,a can only have unit modulus if we are at one of the phase transitions, and at the phase transitions the branch points cross the contour resulting in a divergent integral; see Fig. 5. In particular, we may characterize the phase diagram in terms of which branch points are inside or outside the contour integral.
In addition, we may actually compute the integrals exactly in certain cases and limits, which allows us to obtain the exact analytic dependence of the divergence on the couplings. As a definite example, we consider the case |μ T | > 1. In this case, there is a branch cut inside the logarithm running from z 1,T to z 3,T , and one outside between z 2,T and z 4,T , and the divergences seen at μ T → 1 will be due to these branch cuts approaching the contour. In this case, we may entirely factor out the dependence on the reference state from the logarithm and focus on the terms that depend on the target state. We deform the contour so that it skirts the branch cut (see the parametrization into four contours in Fig. 6). A key point here is that the argument of the logarithm is −π upon approaching the branch cut from the bottom-half plane, while it is +π upon approaching it from the top half. Therefore, in the sum of the two contours running along the branch cut, the logarithm simply contributes a phase factor, and we may evaluate the resulting simplified integrand by elementary methods, and for small we find We perform the integral around contour C 2 by writing z = z 1 + e iθ , and integrating from −π < θ < π. At small , we find .
The computation for contour C 4 is similar, although the phase winds around the other way: Finally, taking the sum of all four contours, we find that the ln divergence in each integral cancels, and we obtain the desired result: where the function I 2 depends on μ R and R , but it is analytic as the phase transition is approached. Therefore, when approaching from μ T > 1, the quantity ∂ μ T C/L diverges as ln(μ T − 1)/8 T if T = 0, but it is analytic if one approaches the multicritical point at T = 0. Similar manipulations may be made for ∂ T C/L and in other phases. Sometimes the branch cuts take a complicated form in the complex plane so that we cannot reduce the expression into elementary integrals, but we can still deduce the form of the divergence by considering how the contour integrals behave as the branch points cross the contour.
Our final results are summarized as follows. The expression ∂ μ T C/L is always analytic unless μ T → ±1. Near these phase transitions, it diverges as so the divergence is sgn(μ T ) ln |μ T − 1|/8 T if T = 0, but there is not a divergence at T = 0. In contrast, the expression ∂ T C/L is analytic whenever T = 0. In this case, the divergence depends on whether the couplings (μ T , T ) approach the phase transitions from the topological phase or the trivial phase. If we approach 013323-7 the multicritical points from the trivial phases, we find that ∂ T C/L remains analytic. In contrast, if we approach T = 0 from the topological phases, we find In this case, we have a ln | T |/4 divergence when |μ T | < 1, but now we find that the divergence crosses over to ln | T |/2 as we approach the multicritical points.
APPENDIX C: REAL-SPACE BEHAVIOR OF THE OPTIMAL CIRCUITS
In this Appendix, we show that the real-space optimal circuit behaves differently depending on whether or not the initial and target states are in the same topological phase.
As we have derived in Appendix A, for a single momentum sector k, the circuit complexity is found to be the squared difference between the Bogoliubov angles [Eq. (7)], and the optimal circuit is generated by the following time-independent Hamiltonian: where O 1,k is the same generator given by Eq. (A3) for momentum sector k. Here, we have omitted the time label "s" for simplicity as the circuit is time-independent (and the total evolution time is fixed to be constant 1). As in the main text and following the circuit complexity literature, we have defined H k to be anti-Hermitian [Eq. (5)].
Since the ground state of the Hamiltonian is a product of all momentum sectors with k > 0, the optimal circuit that generates the evolution between two ground states can be written as We are interested in the real-space behavior of the above Hamiltonian. To discern this, we first write the above Hamiltonian in operator form, where τ i are the Pauli matrices, andψ k denotes the Nambu spinor,ψ (C4) Utilizing the particle-hole symmetry of the Nambu spinor, we can extend the sum in the evolution Hamiltonian to be over the entire Brillouin zone, where ω(k) satisfies for k > 0. In particular, only the odd part of the function contributes since the even part cancels in the τ 1 pairing channel. We now proceed by performing a Fourier series expansion of the function ω(k) over the Brillouin zone. Without loss of generality, we may consider only the odd Fourier series since the even terms will cancel. Thus, we write where the last equality is used to determine the Fourier coefficients.
Our crucial observation is that when the two states are within the same phase, the Fourier sine series for θ (k) ought to be uniformly convergent. This can be seen by considering the boundary conditions, which in this case read θ (0) = θ (π ) = 0, as shown in Fig. 1(d) in the main text. Thus, if we allow the time-evolved state | T to be within an arbitrarily small error to the real target state | T , this Fourier series can be accurately truncated to a finite order N * over the entire Brillouin zone. This is relevant because in real space, the Fourier harmonic sin(lk)ψ † k τ 1ψ k is generated by a term involving two fermionic operators separated by l sites. More specifically, as this occurs in the τ 1 channel, H must involve real-space pairing terms such that The above argument holds when the system size L is taken to be infinite. In such a case, the finite-range interacting evolution Hamiltonian can be regarded as a truly short-range Hamiltonian, and our results imply that the optimal circuit (with constant time or depth) which evolves states within the same phase region is short-range.
On the other hand, when the two states are in different phases, the boundary conditions θ (π ) = π/2 = θ (0) = 0 obstruct uniform convergence, analogous to the Gibbs phenomenon. In this case, the Fourier sine series may still converge pointwise, but for fixed error the series cannot be truncated to finite order N * over the entire Brillouin zone. In such cases, the optimal evolution Hamiltonian H that transforms states between different topological phases must be long-range when the evolution time is fixed to be a constant. Again, this is because the longest real-space distance required to generate the evolution Hamiltonian is given by the highest order of Fourier mode appearing in the momentum space series, which now cannot be accurately truncated.
APPENDIX D: NUMERICAL EVIDENCE FOR NONANALYTICITY OF QUENCH DYNAMICS
In this Appendix, we provide detailed numerical explanations for the nonanalyticity of the long-time steady-state value of the circuit complexity at critical points, as observed in Fig. 3(b). As derived in the main text, the time-dependent circuit complexity is given by where φ k n (t ) = arccos 1 − sin 2 2 θ k n sin 2 ε k n t .
Then the long-time steady-state complexity is just given by the time-averaged value of the above expression, where the overline denotes time averaging. Because φ 2 k n (t ) is such a complex expression, it is unknown to us how to derive an analytical function for the time-averaged circuit complexity. Instead, we plot φ k n (t ) numerically, and we show that the nonanalyticity indeed occurs at the phase transition.
From the expression of φ k n (t ), it is clear that its value oscillates with time, and it reaches its maximal value (envelope) for each momentum sector k n when sin( k n t ) = 1. In Fig. 7(a), we plot the maximum value of φ k n (t ) for different postquench Hamiltonian parameters. As the figure clearly shows, when the chemical potential μ f of the postquench Hamiltonian is below the critical value (μ f = 1), max[φ k n (t )] is a smooth function of k n . However, when μ f is above the critical value, max[φ k n (t )] exhibits a kink at a certain momentum k n , with its maximal value reaching π 2 . To understand this behavior, we can write down the expression for max[φ k n (t )] given the choice of parameters μ i = 0, i = f = 1: max φ k n (t ) = arccos 1 + μ f cos k n μ 2 f + 2μ f cos k n + 1 . (D4) From the above expression, it is clear that when μ f < 1, max[φ k n (t )] is always smaller than π/2; when μ f > 1, max[φ k n (t )] can obtain the maximal value of π/2 when 1 + μ f cos k n = 0. Because one needs to take the absolute value for the arguments of arccos, the quantity max[φ k n (t )] exhibits a kink when reaching π/2, in agreement with Fig. 7(a).
We plot the time-averaged value of φ k n (t ) in Fig. 7(b). Again, we see an upper bound of φ k n (t ) when quenching across the critical point. Similar to Fig. 7(a), φ k n (t ) reaches its maximal value when 1 + μ f cos k n = 0, i.e., when sin(2 θ k n ) = 1. For this special momentum sector, the expression for φ k n (t ) can be written as φ k n (t ) = arcsin sin ε k n t . (D5) Clearly, the time-averaged value of the above expression is just π/4, in agreement with the numerical results shown in Fig. 7(b). Therefore, after the phase transition takes place, the maximal value of φ k n (t ) is bounded by π/4. (This feature is independent of the parameters of the prequench Hamiltonian.) Having revealed this feature of φ k n (t ), the nonanalyticity can be understood as follows: as μ f increases but is still below the phase transition point, the integral of φ 2 k n (t ) increases smoothly with μ f . After reaching the phase transition, φ 2 k n (t ) saturates the bound, and thus the integral's (circuit complexity's) dependence on μ f takes a different form. In particular, for the parameters shown in Fig. 7 [blue line in Fig. 3(b)], the integral (i.e., the circuit complexity) becomes a constant after the phase transition. This leads to a clear nonanalytical (kink) point at μ f = 1. trivial vacuum with no particle. Upon tuning μ, the system undergoes a quantum phase transition into the topological phase at μ = 0. Taking the derivative of C with respect to μ T , we obtain | 2019-02-27T19:00:02.000Z | 2019-02-27T00:00:00.000 | {
"year": 2019,
"sha1": "0dc481b467b490c99d30a734ffff85f8f72d86d4",
"oa_license": "CCBY",
"oa_url": "http://link.aps.org/pdf/10.1103/PhysRevResearch.2.013323",
"oa_status": "GOLD",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "16d56e87a019fef33294ae06507274a8f495c276",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
16184817 | pes2o/s2orc | v3-fos-license | A new non-naked species of Ptychostomella (Gastrotricha) from Brazil
Abstract A new species of marine Gastrotricha from Brazil is described and discussed. Ptychostomella lamelliphora sp. n. is one of the several new taxa that were found during an extensive survey of the gastrotrich fauna carried out in 2002 and 2003 along the coastline of the State of São Paulo. The new species is unique in that it possesses cuticular ornamentations in the form of plate-like structures (scales) along the lateral borders of the body and two massive clusters of densely packed adhesive tubes on the ventral surface, near the ano-genital opening. Both these features appear to be adaptations to challenge the high energy waters that characterize the species’ microhabitat: the coarse sublittoral sand in the channel between the mainland and the largest island in the State, Ihlabela. Additionally, a key to the described Ptychostomella species of the world is provided.
Introduction
The study is part of a larger research program aimed at shedding light on the diversity of marine invertebrates of the northern coasts of the State of São Paulo, Brazil (see Migotto and Tiago 1999). In the seminal works by Rocha (2004a, b, 2005), faunistic and preliminary taxonomic data on the gastrotrich communities along the coastline comprised between Picinguaba to the north (at the Rio de Janeiro-São Paulo States border) and Praia Preta e das Choncas to the south were reported. Among the some 40 taxa found, the occurrence of several species new to science was highlighted. A recent article has dealt with one of such taxa (Todaro 2012); another interesting macrodasyidan in the family Thaumastodermatidae is here described.
Methods
Survey along the northern coasts of the State of São Paulo took place between 20 April and 3 May 2002 and in September 2003. A general account on the visited locations are reported in Rocha (2004a, 2005). Samples containing the new species were collected in 2002 from the sublittoral of Praia Grande and Beluga both on the Ilhabela side of the São Sebastião channel. Ilhabela is the largest island in the State and it is located just in front of the city of São Sebastião. Sandy substratum was collected by scuba diving using eight 500ml plastic jars, four per location. After collection, samples from each site were taken as soon as possible to the São Paulo University's CEBIMar laboratory in São Sebastião. In the laboratory, the specimens were extracted daily with the narcotization-decantation technique using a 7% magnesium chloride solution within one week of collection. The supernatant was poured, without filtering, into 3.0 cm diameter plastic Petri dishes and scanned for gastrotrichs at 50 × under a Wild M8 stereomicroscope (see Todaro and Hummon 2008). Found gastrotrichs were mounted on glass slides and observed in vivo with Nomarski differential interference contrast optics using a Zeiss Axioscop 2 Plus microscope. During observation, the specimens were measured using an ocular micrometer and photographed with a Nikon Coolpix 995 digital camera (3.34 Mpixel). Some specimens were fixed overnight in a 1.0 M phosphate-buffered (pH 7.3) solution of paraformaldehyde, gluteraldehyde and picric acid, following Ermak and Eakin (1976), and stored for later SEM analysis. Gastrotrichs were rinsed in 0.2 M cacodylate buffer, dehydrated through a graded ethanol series, critical point-dried using CO 2 , mounted on aluminium stubs, sputter coated with gold-palladium and observed with a Philips XL 30 scanning electron microscope at the author's Institution.
The description of the new species follows the convention of Hummon et al. (1993), whereas the position of some morphological characteristics along the body are given in percentage units (U) of total body length measured from anterior to posterior.
Granulometric analysis of the substrata was carried out according to Todaro et al. (2006). Mean grain size, sorting coefficient, kurtosis, and skewness were calculated by a computerized programme (Todaro 1992).
Abbreviations are as follows: PhIJ, pharyngeo-intestinal junction; TbA, adhesive tubes of the anterior series; TbL, adhesive tubes of the lateral series; TbV, adhesive tubes of the ventral series; TbP, adhesive tubes of the posterior series.
The rationale for the key to the ecological characteristics of the species, according to Hummon et al. (1992), is as follows: Frequency of a species from among a sample series (i.e., frequency of a species in samples collected in any given sampling trip) -Sparse, found in less than 10% of samples; occasional, found in 10-30% of samples; common, found in 30-60% of samples; usual, found in more than 60% of samples.
Abundance of a species among other species of a sample -Rare, less than 1% of a sample; scarce, 3-5% of a samples; numerous, 10-20% of a sample (often a sub-dominant); prevalent, more than 30% of a sample (usually dominant or co-dominant).
taxonomic account
Order Macrodasyida Remane, 1925 Material examined. Eight adult specimens (including the holotype) collected by the author, five from the type locality and three from a Beluga a nearby location (see Table 1). Four specimens were observed alive and are not longer extant, while four were prepared for SEM survey and are kept in the meiofauna collection of the author (Ref. n. 2002-BR-01-02-05-06).
Ecology. Frequency of occurrence: sparse, found only in sub-littoral sediment of two locations along the southern portion of the São Sebastião channel. Abundance: numerous in coarse sediment with little detritus of Praia Grande, scarce in coarse sediment rich in detritus of Beluga.
Diagnosis.
A Ptychostomella with an adult length to 250 µm; pharynx length to 74 µm, with pharyngeal pores at base. PhIJ at U37; body with almost parallel sides and short, bilobed caudum. Head bearing paired knob-like sensory organs and small, trapezoidal, fleshy lobes; eye spots missing; sensory hairs a few, forming lateral columns along the body, a fringe around the oral opening and loose tufts at the tip of the lobes; epidermal glands noticeable, eight per side, scattered along the length of the body. Cuticular covering generally smooth except for peculiar, subrectagular scales arranged in a column of on each ventrolateral side. Adhesive tubes: TbA, 4 per side, one slightly smaller, cone-shaped, in the middle at U9 and three lateral, rod-like, of equal size at U9-U10; TbL, 3 on each side, roughly of the same size, a small isolated one implanted anteriorly at U15, one in mid trunk region at U58 and one more robust near the base of the caudal lobes, at U90. TbV, up to 16 per side, four of the same size more or less evenly spaced, implanted along trunk region from U44 to U63, the remainder 10-12 forming a noticeable cluster at U83-U85. TbP, six in all, two medial and two on each of two paired caudal pedicles. Ventral locomotor cilia: a continuous field of transverse rows covering the entire surface except the ano-genital area. Reproductive system: testis on the right body side, caudal organ pyriform, frontal organ round filled with motile spermatozoa, a ripe egg dorsally in the mid-intestinal region.
Etymology. The specific epithet lamelliphora (lamella, L, thin plate and phero, Gr., to bear) refers to the presence of the thin, plate-like scales along the ventrolateral body sides.
Description. Description is mainly based on the holotypic specimen, 250 µm in total length. Pharynx 74 µm in length, measured from the posterior margin of the oral opening to the pharyngeo-intestinal junction, with pharyngeal pores near the base, at U34; pharyngeo-intestinal junction at U37. Head bearing paired knob-like sensory organs and small, trapezoidal, fleshy lobes; body robust, with side lines slightly widening to mid-trunk, then gradually narrowing to a short, bilobed caudum; widths of head\ neck\trunk\caudal base 46\34\56\26 µm at U07\U25\U46\U93, respectively.
Oral hood slightly protruding anteriorly (to U08), with gently undulating borders. Sensory hair sparse, up to 7 µm in length, forming a fringe around the oral opening and loose tufts at the tip of the head's lobes; a single hair emerges from each knob-like tentacle; other sensory hairs 9-14 µm in length form lateral and dorsolateral columns that are evenly spaced within columns but differ between columns. Eight pairs of noticeable epidermal glands are regularly spaced along the pharyngeal and intestinal region from U11 to U 87 with glands of the second and eighth pairs positioned somewhat more lateral; glands are round in shape and roughly of the same size (7-9 µm in diameter), except for the one of the second pairs, markedly smaller (4-6 µm). Each gland opens to the exterior via a well structured pore, the produced material is excreted in the form of small round droplets (see Figure 4C, D). Cuticular armature. Body covering apparently smooth as typical of the genus, however on the ventrolateral sides, the cuticle generates subrectangular plates (scales), partially overlapping each other and tightly arranged in two columns running from U11 to U88. Scales, roughly of the same size (1.5-2.0 µm), protrude from the body and form bilateral structures that recall the lateral aerodynamic 'mini-skirts' of racing cars.
Adhesive tubes. TbA, 4 per side, inserting directly on the body surface at U9-U 10, one 4 µm in length medially and three 6-7 µm in length laterally. TbL, 3 per side (8-10 µm in length) inserting respectively at U15, U58 and U90. TbD, absent. TbV, up to 16 per side; 4 (8-11 µm in length) inserted singly along the intestinal region from U44 to U63 while the remaining 12 (4-11 µm in length) form an impressive cluster centered at U85 (cTbV; Figs 2A, 3B, D); tubes in the cluster originate singly and their number may slightly change from side to side. TbP, 3 per side, 2 (5-6 µm in length) at the end of each pedicle of the furcated caudum and the other one (6.5 µm in length) flanking each caudal pedicle medially.
Ventral ciliation. A continuous, dense field of cilia arranged in transverse rows that extend from the ventral border of the oral opening to the base of the caudal pedicles, being broadest at mid body and somewhat sparse in the ano-genital area at U90.
Variability and remarks. Length of the 4 living specimens ranged from 204 to 250 µm (mean = 230 µm, SD = 18 µm) all of them were mature (i.e., showed at least the testicles filled with sperm). The SEM prepared adult specimens resulted of smaller size (range 144-183 µm) even though size of these specimens appeared not dissimilar from the others under the dissecting microscope; these measurements fall well below the 5.5% length reduction allowed for fixed specimens (cf. Clausen 2004), but are in agreement with the shrinkage suffered by specimens of Pseudostomella dolichopoda Todaro, 2012 processed for SEM examination (cf. Todaro 2012). As SEM is being routinely utilized in species description, to avoid potential misidentification based on size, it would be interesting to explore the phenomenon of size reduction over a larger taxonomic spectrum. SEM prepared specimens showed some traits undetected or not present in living specimens. For instance, in SEM prepared specimens, i) the border of the oral hood appeared more scalloped than in living animals (cf. Figure 3A vs. Figure 2 A), ii) the cuticle on the ventral area comprised between the ciliary field and the col-umn of scales appeared punctuated by shallow pits ( Figure 3C) and iii) the cuticle on the dorsal side appeared embossed with keel-like structures ( Figure 4B). The adhesive tubes of the ventral series forming the clusters showed some variability in number, depending on individuals and on side of the body (e.g., see Figure 3B); the highest number of tubes, 12, was found in the cluster on the left side of two specimens including the holotype, while the lowest, 8 tubes, was found in the cluster on the right side of a specimen measuring 224 µm in total body length.
Taxonomic affinities. The genus Ptychostomella was originally created to include small thaumastodermatid gastrotrichs whose body is enveloped by a smooth cuticle i.e., a cuticle that does not give rise to the typical scales and/or spines (e.g., ancres) found in other members of the family (Remane 1926). In a phylogenetic framework, the absence of such structures has been thought as a secondary reduction (cf. Todaro et al. 2011). While all of the known Ptychostomella species lack an armature typical of the family, at least 4 out of the 12 species described so far (Hummon and Todaro 2010) possess other kinds of cuticular ornamentations. P. lepidota Clausen, 2000 bears scalelike cuticular elevation and P. orientalis Lee & Chang, 2003 has the cuticular covering embossed with smooth hemispherical elevation (Clausen 2000;Lee and Chang 2003). A third species P. brachycephala (Levi, 1954), originally affiliated to the genus Platydasys, possesses on each lateral side a column of rod-like papillae, and papillae are reported also for P. papillata Lee andChang 2003 (Lee andChang 2003;Clausen 2004). Shape and arrangement of the cuticular ornamentations (i.e., subrectangular plates tightly arranged in two lateral columns) differentiate Ptychostomella lamelliphora n. sp. from all the four species reported above. The coexistence of knob-like sensory organs and of fleshy lobes on the head and the presence of most of TbV arranged in a bilateral clusters near the ano-genital opening may further distinguish the new species from all the previously known Ptychostomella species.
Conclusive remarks. Adhesive tubes of the ventral series forming 'feet' or 'clusters' are not uncommon among members of the family Thaumastodermatidae (e.g., Tetranchyroderma and Pseudostomella) and they are present also in members of the genus Ptychostomella e.g., P. bergensis Clausen, 1996(Clausen 1996. However, what makes special the clusters of TbV present in the new species from Brazil is their bulkiness. To my knowledge no other gastrotrich species is known to posses bilateral clusters made up of such a high number of tubes. What is the adaptive advantage of such a formidable apparatus, and in species with fewer adhesive tubes, do they have the same function? Adhesive and aptic structures are universally present among interstitial animals (Swedmark 1964) and it is possible to detect a relationship between the extent of the aptic apparatus and the energy of the environment water the species live in. For example, the gastrotrich species of the genus Oregodasys Hummon, 2008 that inhabit the coarse sediment typical of high energy waters are characterized by a formidable adhesive apparatus made up of tens of tubules (e.g., Rothe and Schmidt-Rhaesa 2010); high number of adhesive tubes also characterize species of the genus Diplodasys Remane, 1927 (e.g., Hummon andTodaro 2009), which often co-occur with Oregodasys. In this framework, it is natural to hypothesize that the massive clusters of adhesive tubules of P. lamelliphora n. sp. are an adaptation that prevent the displacement of the animals by the strong currents that characterize the habitat. The high energy of these waters are indirectly indicated by the coarse sediment of the locus typicus in the São Sebastião channel (Table 1). In my view this working hypothesis is further supported by the presence in P. lamelliphora n. sp. of the columns of scales along the ventrolateral sides whose function could be the reduction of the hydrodynamic turbulence around the body allowing a better bond of the gastrotrich to the sedimentary granules.
An alternative hypothesis could be that these structures play a role during reproduction e.g., used for sperm transfer or holding of the partner during cross fertilization. Only future TEM studies revealing ultrastructural difference between the ventral tubules clustering near the ano-genital opening and the genuine adhesive tubes (e.g., single tubes) could make the second hypothesis on this subject most plausible. A study of reproductive behaviour would also be revealing.
Taxonomic key. Lee and Chang (2003) provided a useful taxonomic key to the species of the genus Ptychostomella; however, because one species has been transferred to this genus and the two additional ones have been described in the meanwhile (cf. Clausen 2004;Lee et al. 2009) a revised key seems necessary. The following key is based upon characters visible under light microscopy. | 2016-05-12T22:15:10.714Z | 2013-04-12T00:00:00.000 | {
"year": 2013,
"sha1": "8d030aff86f961e62fe04e538cbfa00ad9a86285",
"oa_license": "CCBY",
"oa_url": "https://zookeys.pensoft.net/lib/ajax_srv/article_elements_srv.php?action=download_pdf&item_id=3894",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "8d030aff86f961e62fe04e538cbfa00ad9a86285",
"s2fieldsofstudy": [
"Biology"
],
"extfieldsofstudy": [
"Biology",
"Medicine"
]
} |
232088013 | pes2o/s2orc | v3-fos-license | Large Area Emission in p-Type Polymer-Based Light-Emitting Field-Effect Transistors by Incorporating Charge Injection Interlayers
Organic light-emitting field-effect transistors (LEFETs) provide the possibility of simplifying the display pixilation design as they integrate the drive-transistor and the light emission in a single architecture. However, in p-type LEFETs, simultaneously achieving higher external quantum efficiency (EQE) at higher brightness, larger and stable emission area, and high switching speed are the limiting factors for to realise their applications. Herein, we present a p-type polymer heterostructure-based LEFET architecture with electron and hole injection interlayers to improve the charge injection into the light-emitting layer, which leads to better recombination. This device structure provides access to hole mobility of ~2.1 cm2 V−1 s−1 and EQE of 1.6% at a luminance of 2600 cd m−2. Most importantly, we observed a large area emission under the entire drain electrode, which was spatially stable (emission area is not dependent on the gate voltage and current density). These results show an important advancement in polymer-based LEFET technology toward realizing new digital display applications.
Since the first report in 2003 [19], LEFETs have brought a different perspective to organic semiconductors, paving the way for research in device architecture and material development to simultaneously study charge carrier transport and light emission [13,[16][17][18]. Multiple device geometries and material combinations have been suggested to improve the LEFET parameters such as external quantum efficiency (EQE) [21][22][23][24], charge carrier mobility [25][26][27][28], aperture ratio [29,30], current ON/OFF ratio [20,31], and large area emission [32,33]. However, LEFETs are far from realizing into commercial products due to their limited performance. A multilayer LEFET can provide higher mobility and lead to better charge injection in LEFETs, but their EQE at higher brightness is somewhat limited.
A semitransparent electrode enables the higher EQE in multilayer LEFETs by outcoupling the trapped light underneath it; however, it does not provide a uniform and stable emission area, which is essential for pixilation [21,34]. More recently, n-type hybrid LEFETs (a combination of the inorganic charge transport layer and organic emissive layer) have been reported to show a large, uniform, and stable light-emitting area through interface doping [7,35,36]. However, for all organic p-type LEFETs, this has not been achieved.
In this paper, we report a device architecture of a p-type polymer-based LEFET that simultaneously provides higher charge carrier mobility, higher EQE, and uniform and stable emission. The device structure consists of: (i) polymer charge transport layer of poly [4-(4,4-
Device Fabrication and Characterization
The LEFET devices were fabricated using a 300 nm thermally grown layer of SiO 2 as a dielectric on Si substrates. The substrates were cleaned by ultrasonication in acetone and isopropanol alcohol (IPA) for 15 min each and then blow-dried under pressurized nitrogen. An 80 nm layer of charge transport polymer poly [4-(4,4- [3,4-c] pyridine], PCDTPT, was then spincoated onto substrates from a 5 mg/mL solution in dichlorobenzene. Substrates were then annealed at 150 • C for 30 min. Using an aligned shadow mask-1, a hole injecting asymmetric source contact, consisting of a 10 nm of Mo 2 O 3 layer and 50 nm of Au layer, was deposited on top of the PCDTPT layer by thermal evaporation in a high vacuum. A 10 nm layer of Mo 2 O 3 was also deposited using the aligned drain mask-2, to be exactly underneath the drain electrode. A 100 nm layer of polymer PDY-132 was then spin-coated onto the entire substrate from a solution of 10 mg/mL in toluene. The substrates were annealed again at 150 • C for 30 min. Finally, the electron-injecting semitransparent drain electrode, consisting of 10 nm of Cs 2 CO 3 and 20 nm of Ag layers, was deposited through aligned shadow mask-2 to give a channel length (L) and width (W) of 100 µm and 2 mm, respectively. The complete device structure is shown in Figure 1a and the energy level diagram with charge injection schematics is shown in Figure 1b, along with the molecular structures of PCDTPT in Figure 1c and PDY-132 in Figure 1d.
Two Agilent B2912A units connected to electrical probes and a calibrated photodetector attached to an EverBeing C-series probe station were used to characterize the LEFETs. A microscope attached to the probe station was used to capture the images. An optical fiber attached to the probe station and an OceanInsight USB Flame spectrometer was used to determine the electro luminance spectrum. A Dektak profilometer was used to measure the film thickness and the photoluminescence quantum yield (PLQY) of thin films was measured using the de Mello method [37] as reported previously. The brightness of the LEFET was measured by comparing the photocurrent in a photodetector with that of a reference device of known light emission area and brightness. The correct brightness value was measured by correcting the photocurrent for the effective light-emission area. The brightness of the reference device was measured with the help of a Minolta Candela meter (LS-100). The values of brightness and EQE were found using the standard procedure described elsewhere [20,36,[38][39][40][41]. Figure 2a,b shows the transfer and output characteristic curves of the LEFET, respectively. A current ON/OFF of >10 6 was achieved in the saturation regime. The charge carrier mobility (µ) in the saturation regime was calculated from the transfer characteristics given in Figure 2a using Equation (1).
Results and Discussion
where I DS is the measured source-drain current; W and L are the width and length of the device channel, respectively; V GS is the corresponding applied gate voltage; and C i is the capacitance of the SiO 2 dielectric. A linear fit was applied to the sqrt (I DS ) in the extensive range of the curve to provide the average hole mobility and limit the error in estimating the mobility. The threshold voltage (V TH ) was extracted using the linear extrapolation of sqrt (I DS ) of the transfer curve in Figure 2a. The charge carrier mobility was estimated to be 2.1 cm 2 V −1 s −1 , reflecting the mobilities associated with the PCDTPT polymer without the PDY-132 layer [42] as a result of non-planar contact geometry [11], which eliminates the injection barrier at the source electrode. Figure 3a shows the optical transfer characteristics of luminance vs. gate voltage of the LEFET device. Bright yellow light emission of the peak luminance of 2600 cd m −2 was observed under the drain electrode. An EQE of 1.6%, as shown in Figure 3b, was estimated at a high luminance of 2600 cd m −2 . The EQE vs. luminance graphs in Figure 3c showed an increase in EQE trend, which is an exciting feature and needs a follow-through in LEFET devices. Usually, at high brightness, the EQE tends to roll-off and decrease with increasing current density. However, the LEFET structure has been reported to lower this roll-off of EQE [34] as observed in these results. The results are summarized in Table 1. The calculated luminous efficiency at maximum current density (or gate voltage of 100 V) was 3.2 lm per Watt. Figure 4a shows the transmission spectrum of the Cs 2 CO 3 /Ag electrode, which was around 60% transparent at 560 nm (the peak emission wavelength of PDY-132). This high transparency of 60% enables the light under the drain electrode to come out through the semitransparent electrode, which could be reflected in the case of an opaque electrode and absorbed by the Si substrate. The photoluminescence (PL) and electroluminescence (EL) spectrum of the emitted light is provided in Figure 4b, which is characteristic of the emissive polymer PDY-132. The EL spectrum was corrected with the reflectance of the Cs 2 CO 3 /Ag electrode. The emission images from the LEFET are shown in Figure 4c. The entire semitransparent Cs 2 CO 3 /Ag electrode emitted light during the operation. The large area emission under the Cs 2 CO 3 /Ag electrode was observed to be spatially stable and its position was independent of the applied gate voltage as shown in the series of images in Figure 4c. Gate voltage was used to modulate the emission intensity. There were some dark spots at a lower voltage, which we attributed to the inhomogeneity of the Mo 2 O 3 interlayer. However, at higher emission intensity (or gate voltages), the black spots were not visible due to the camera detector's saturation. and recombined to form excitons with injected holes incoming from the p-type channel, as shown in Figure 1b. The light was emitted through the electron injecting semitransparent Cs 2 CO 3 /Ag electrode, as shown in Figure 4c. Previous studies have demonstrated that using a thin layer of Mo 2 O 3 shifts the work function of semiconductors and metals [11]. The Mo 2 O 3 layer lowers the injection barrier at the source electrode for hole injection and it also facilitates the hole injection from PCDTPT to PDY-132 at the interface of both polymers. The better injection facilitates the exciton formation and resulting in higher EQE. The efficiency of light emission is given by Equation (2): where ϕ capture (recombination efficiency) is the fraction of electrons-holes that recombine to form excitons; ϕ EQE is the calculated EQE; ϕ escape indicates the fraction of escaped photons from the device; ϕ spin is the spin-statistics factor, and ϕ PLQE is the photoluminescence quantum yield (PLQE) in the solid-state. The polymer is a singlet emitter and so ϕ spin = 0.25. The ϕ escape is approximately 1/2n 2 (where n is the refractive index) for isotropic emission. The calculated maximum radiative recombination efficiency of the LEFETs was calculated as 48%. We attributed the superior recombination efficiency in this LEFET structure to the better injection due to Mo 2 O 3 interlayers.
Conclusions
In summary, we have demonstrated a polymer-based p-type LEFET in a heterostructure device architecture with hole mobility approaching 2.1 cm 2 V −1 s −1 and a current ON/OFF >10 6 . The incorporation of the Mo 2 O 3 interlayer between the polymer leads to superior operating characteristics including an EQE of 1.6% at a luminance of 2600 cd m −2 and, more importantly, a large area, uniform, and stable light emission under the entire drain electrode. The results here demonstrate the feasibility of the solution-processable p-type polymer-based LEFET technology for applications in the future-generation of highdefinition displays. | 2021-03-03T05:24:04.932Z | 2021-02-01T00:00:00.000 | {
"year": 2021,
"sha1": "093f60dbe0667465a93d517f2882210dd9f8ea5b",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/1996-1944/14/4/901/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "093f60dbe0667465a93d517f2882210dd9f8ea5b",
"s2fieldsofstudy": [
"Materials Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
243890250 | pes2o/s2orc | v3-fos-license | Dual intestinal parasitosis unmasked by treatment for gastrointestinal sarcoidosis
Highlights • Intestinal parasitosis can complicate treatment for gastrointestinal sarcoidosis.• Strongyloidiasis should be considered in cases of diarrhea and eosinophilia.• Multiple stool samples are required to improve detection of intestinal parasites.• HTLV- 1 co-infection should be considered in difficult to treat strongyloidiasis.
Introduction
Sarcoidosis is a multisystemic granulomatous disease of unknown etiology, characterized by the presence of giant-cell nonnecrotizing granulomas [1]. The involvement of the GI tract is uncommon, occurring in only 0.1 -1.6% of cases [2]. The putative mechanisms of GI tract damage include mucosal infiltration, endoluminal lesions, dysfunction of the myenteric plexus, and extrinsic compression of structures [2]. This culminates in the clinical manifestations of abdominal pain, diarrhea, vomiting, and weight loss [2].
Symptomatic sarcoidosis is treated with glucocorticoids, cytotoxic medications, or biologic agents; all of which carry a risk of systemic infections including parasitic GI tract infections [3]. There have been a few case reports of intestinal parasitosis in the context of corticosteroid therapy for sarcoidosis [4][5][6], however, our finding of dual parasitosis is unique particularly given the concomitant finding of asymptomatic human T-cell lymphotropic virus 1 (HTLV-1) co-infection.
Case presentation
A 32 year old Afro-Caribbean man with no prior medical condition presented with progressive exercise intolerance and worsening painful skin nodules in both legs for four weeks. He had no cough, chest pain, wheeze, or fever. He had resided in the United Kingdom for about 20 years without travel to his country of birth. He was a current smoker with a twenty pack-year history.
Clinical examination revealed tender nodules on the shins consistent with erythema nodosum and unremarkable systemic examinations. Blood investigations showed a normal complete blood count, erythrocyte sedimentation rate of 118 mm/h; C-reactive protein of 44 mg/L, with normal renal and liver function tests. Human immunodeficiency virus (HIV) screen was negative. Highresolution CT scan of the thorax revealed bilateral lung infiltrates and hilar adenopathy (Fig. 1). Serum adjusted calcium level was normal but angiotensin converting enzyme levels were raised at 154 U/L (12-82 U/L). Lung function tests revealed a restrictive pattern with reduced diffusion capacity: FEV1/FVC -92% (111% predicted); FVC 3.49 L (73% predicted); FEV1 3.21 L (82% predicted); TLC 5.5 L (72% predicted); TLCO 6.62 L (56% predicted). He was treated for pulmonary sarcoidosis with oral prednisolone 30 mg daily for 4 weeks followed by a gradually tapering dose whilst monitoring his exercise tolerance and lung function parameters.
Four months later, he presented in the emergency department with profuse foul-smelling non-bloody, non-mucoid diarrhea associated with crampy lower abdominal pain and nausea. Stool and blood cultures did not yield any pathogen. He was managed in the high dependency unit where he had inotropic support due to persistent hypotension and treatment for prerenal acute kidney injury. He presented again two weeks after discharge with similar abdominal symptoms. Blood investigations were significant for eosinophilia (1280 cells/uL; 13.2% of total white blood cell count) and hypokalemia of 3.3 mmol/L. Stool analysis did not isolate any pathogen, and fecal elastase was normal. However, fecal calprotectin was mildly elevated at 159 ug/mL (< 60). CT scan of the abdomen and pelvis showed mild distension of the small and large bowels. Upper and lower GI tract endoscopies were performed due to a clinical suspicion of GI sarcoidosis and were macroscopically normal. However, histologic assessment of gastric and duodenal biopsies revealed non-necrotizing epithelioid granulomas confirming a diagnosis of GI sarcoidosis. His corticosteroid dose was subsequently increased to 40 mg daily.
A repeat stool sample was sent for microbiological analysis due to persistent symptoms. Giardia lamblia (GL) was detected by polymerase chain reaction (PCR) although cysts remained absent on microscopy. Also, rhabditiform larvae and an adult female form of Strongyloides stercoralis (SS) were identified. Stool PCR for Clostridium difficile, Salmonella and Shigella species were negative. Enzyme linked immunosorbent assay was strongly positive for Strongyloides at an optical density of 0.977 (cut-off: 0.702).
Following consultations with the regional tropical disease departments, he was given intravenous metronidazole for symptomatic giardiasis at 500 mg three times a day for 5 days and oral ivermectin for strongyloidiasis at 200 ug/kg for 2 days. He had improvement in abdominal pain and the frequency of diarrhea episodes within 48 h of commencement; however, his loose stools persisted. Further stool examinations revealed clearance of GL; with persistent presence of rhabditiform larvae of SS. This prompted an extended course of ivermectin for a further 14 days with daily assessment of stool samples given the likelihood of a SS hyperinfection syndrome. Considering his persistent strongyloidiasis, a blood sample was also sent for HTLV-1 serology which returned positive.
His symptoms resolved afterwards with consistent evidence of negative stool examinations. He remains clinically stable on outpatient clinic visits on a stable prednisolone dose of 5 mg daily whilst being monitored for asymptomatic HTLV-1 infection.
Discussion
Our case report is unique in highlighting dual intestinal parasitosis (giardiasis and strongyloidiasis) in an individual on corticosteroid treatment for sarcoidosis and previously unidentified HTLV-1 infection. Immunosuppressed patients are thought to be at a higher risk of symptomatic parasitic infections and although the exact mechanism for this susceptibility is not entirely clear, diminished adaptive and innate immune responses in GL impair the host's ability to control infection [7]. Furthermore, the dysregulated immune response with sarcoidosis itself could potentially increase the risk of opportunistic parasitic infections as the inhibition of IL-2 production reduce the myeloid cell's ability to stimulate T-cells which are fundamental for defense against the parasite [8].
GL is a flagellated microscopic protozoon that is ubiquitous in many regions of the world especially in developing countries [9]. The transmission mode is through consumption of contaminated water and food with an incubation period of 5-25 days [9]. Symptomatic patients with GL commonly have abdominal cramps, bloating, and explosive non-bloody diarrhea [10]. In our subject's case, GL was detected by PCR with negative stool microscopy. This is likely due to microscopy having about 60% sensitivity for detecting oocyte and trophozoites in stool samples [11]. Concomitant PCR testing is beneficial as it improves detection rates in patients who have low parasitic cyst count [11] which may have been the case in our patient. Once giardiasis is suspected, it is advocated that a minimum of six stool samples be sent for microbiological analysis to objectively exclude GL, with the first three samples taken two to three days apart [10]. If results are negative, a further three samples are taken weekly to account for the varying shedding rates of the parasite [10].
The other intestinal parasite detected was Strongyloides stercoralis which is a soil-transmitted helminth that commonly enters the human host transcutaneously [12]. In our subject, the potential for autoinfection was increased as the rhabditiform larvae can transform into invasive filariform larvae [12]. This process can culminate in chronic infection with SS which manifests with diarrhea, abdominal cramps and itching [12]. Unchecked autoinfection in the context of altered immune status can lead to hyper-infection due to accelerated larval migration or disseminated infection which is mostly associated with co-infection with HTLV-1, or immunosuppression secondary to corticosteroid use or HIV, [12], the former two of which were present in our subject.
Another interesting observation from our report is the presence of HTLV-1 infection in our subject who had been residing in a nonendemic area (United Kingdom) for two decades. It is important to note that HLTV-1 infection is increasing in non-endemic areas due to population movement and hence co-infections with helminths such as SS are also being observed [13].
The patient's background of sarcoidosis, concurrent use of corticosteroid as well as the HLTV-1 co-infection increased the risk of hyper-infection in our patient. HLTV-1 infection and prolonged corticosteroid use interrupt granulocyte function which can promote severe strongyloidiasis [14]. HTLV-1 infection can also lead to reduced serum IL-4, IL-5, IgE levels, causing hampered immune response to SS [14].
Conclusion
Intestinal parasitosis like strongyloidiasis should always be considered in cases of subacute to chronic diarrhea and eosinophilia. In addition, HTLV-1 co-infection should be investigated in patients with difficult to treat or disseminated strongyloidiasis infection.
Clinicians need to maintain a high level of awareness for intestinal parasites when prescribing long-term corticosteroid therapy; particularly for strongyloidiasis which can become fatal in the setting of hyper infection due to immunosuppression and HTLV-1 co-infection.
Funding source
There was no funding received from any individual or organization.
CRediT authorship contribution statement OO: Conceptualization, Writing -original draft, reviewing and editing; TH: Conceptualization, Writing -original draft, reviewing and editing; SF: Writing original draft, reviewing, and editing; DK: Writing -reviewing and editing, consent from patient; IA: Writingreviewing and editing, supervision.
Ethical approval
Written informed consent was obtained from the patient for publication of this case report as well as the accompanying images.
Consent
Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request. | 2021-11-10T16:31:30.915Z | 2021-11-08T00:00:00.000 | {
"year": 2021,
"sha1": "b3390fc0ec694cb18f752a455b3e56004fdced0b",
"oa_license": "CCBYNCND",
"oa_url": "https://doi.org/10.1016/j.idcr.2021.e01334",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "ff56e592ee8686fc516f764d46ef6e406c2380ec",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
237581650 | pes2o/s2orc | v3-fos-license | Geometrically exact static isogeometric analysis of an arbitrarily curved spatial Bernoulli-Euler beam
The objective of this research is the development of a geometrically exact model for the analysis of arbitrarily curved spatial Bernoulli-Euler beams. The complete metric of the beam is utilized in order to include the effect of curviness on the nonlinear distribution of axial strain over the cross section. The exact constitutive relation between energetically conjugated pairs is employed, along with four reduced relations. The isogeometric approach, which allows smooth connections between finite elements, is used for the spatial discretization of the weak form. Two methods for updating the local basis are applied and discussed in the context of finite rotations. All the requirements of geometrically exact beam theory are satisfied, such as objectivity and path-independence. The accuracy of the formulation is verified by a thorough numerical analysis. The influence of the curviness on the structural response is scrutinized for two classic examples. If the exact response of the structure is sought, the curviness must be considered when choosing the appropriate beam model.
Introduction
Contemporary engineering is facing a growing demand for novel types of cost-effective structures which are simultaneously resistant and flexible. This trend is encouraged by the development of new structural materials, the constant improvement of existing ones, and novel discoveries related to structural form-finding. Curved spatial beams are crucial components of these structures and their application is fundamental in various fields of engineering, molecular physics, electronics, bio-mechanics, optics, etc. In order to apply beam models to these new challenges, more accurate mathematical and mechanical models will be required.
First beam theories originate from the works of Euler and Bernoulli, later generalized by Kirchhoff, Clebsch, and Love [1]. There is great nomenclature diversity for beam theories and a study that makes a careful attempt to classify these is given in [2]. Accordingly, the subject of the presented research is an arbitrarily curved and twisted beam with an anisotropic solid cross section, without warping. If the assumption of rigid cross sections is applied, Simo-Reissner (SR) theory is obtained. The addition of the constraint that the cross section remains perpendicular to the deformed axis results with the Bernoulli-Euler (BE) theory.
The nonlinear theory of shear deformable curved spatial beams is proposed by Reissner [3]. Simo completed his work and introduced the term geometrically exact beam theory which compromises a formulation for which the relationships between the configuration and the strain measures are consistent with the virtual work principle and the equilibrium equations at a deformed state regardless of the magnitude of displacements, rotations, and strains [4]. This is often falsely referred to as a large strain theory despite the small strain assumption being present. Due to the presence of finite rotations, the configuration space of the GE beam theory is a special orthogonal (Lie) group, SO (3). It is a nonlinear, differentiable manifold, whose associated group elements are not additive or commutative, which lead to various types of algorithms, [5,6,7,8,9,10]. Recapitulation of these findings until 1997 is given in [11].
It is argued in [12,13] that all finite element (FE) implementations of the geometrically exact beam theory, existent at that time, are non-objective and path-dependent. An orthogonal interpolation scheme that is independent of the vector parameterization of a rotation manifold is suggested. This paved the way for the development of many alternative rotation interpolation strategies, [14,15,16]. The corotational formulation employed in [17] and [18] preserves objectivity by definition. Recently, an improvement to the corotational approach for shear deformable beams is given in [19]. The carefully calculated corotational nodal rotations are interpolated which led to the satisfaction of all the requirements of the geometrically exact beam theory. The fact that many formulations do not deal with truly geometrically exact curved/twisted beams is discussed in [20].
The nonlinear analysis of spatial BE beams recently came under a similar focus of researchers. The problem of the invariance of classic curved beam theories is thoroughly elaborated in [21] and [22]. Membrane locking is another problem of thin curved BE beam which is treated by various techniques such as assumed strain or stress fields and reduced/selective integration [21]. Works [23] and [24] stand out due to their consistency with the geometrically exact approach. The most systematic up-to-date approach is the formulation by Meier et al. [25]. It satisfies many properties required by the geometrically exact analysis of a curved spatial slender BE beam. The effect of membrane locking is discussed in [26] while the beam-to-beam contact problem using a torsion-free beam model is considered in [27]. A comprehensive review of beam theories is given by the same authors in [2].
The isogeometric technique for the spatial discretization has attracted much attention [28]. The nonlinear shear-deformable spatial beam is readily considered in the framework of isogeometric analysis (IGA) [29,30,31,32,33,34,35]. Regarding the nonlinear analysis of BE beams in the context of IGA, a torsion-and rotation-free spatial cable formulation is developed in [36] with attractive nonlinear dynamic applications. A spatial BE beam modeled as a ribbon with four degrees of freedom (DOF) is introduced in [37] and extended to the assemblies of beams in [38]. A mixed approach is considered in [39] while the consistent tangent operator is derived in [40]. Based on [37], the authors in [41] successfully performed nonlinear applications of a spatial curved BE beam. Recently, a multi-patch approach for the nonlinear analysis of BE beams is suggested [42]. The smallest rotation algorithm is used for an update of the basis. The nonlinear transformation between total cross-sectional rotations and unknown kinematics is defined in order to connect beams. An invariant geometric stiffness matrix is derived in [43] considering various end moments. However, the formulation is applied solely to the buckling analysis. The effect of initial curviness on the convergence properties of the solution procedure is considered in [44].
These works represent the state of the art of curved spatial BE beam theories. In contrast to the presented research, most of the reviewed literature disregards the higher order terms with respect to the beam cross-sectional metric. The majority of them also do not consider truly geometric exact nonlinear analysis due to their lack of objectivity. Recently, Radenković and Borković contributed to the linear static analysis of arbitrarily curved BE beams [45,46], based on foundations given in [47]. The aim of this paper is to extend the developed formulation, which is applicable to strongly curved beams, to the nonlinear setting of geometrically exact beam theory.
For curved beams, axial and bending actions are coupled due to the nonlinear distribution of strain along the cross section. This distribution depends on the curviness of beam Kd. Here, K is the curvature of beam axis and d is the maximum dimension of the cross section in the planes parallel to the osculating plane [45]. Therefore, the term arbitrarily curved beam actually refers to a beam with arbitrary curviness. With respect to this parameter, curved beams are readily classified as small-, medium-or big-curvature beams [48]. We are here mostly concerned here with big-curvature, also known as strongly curved, beams, for which Kd > 0.1.
Although known for a long time, it is only recently that the strongly curved beams have been scrutinized in the framework of the modern numerical methods. Linear analysis of strongly curved plane beams is given in [49,50,51], while nonlinear analysis is considered in [52]. The linear response of spatial beams is analyzed in [45]. Here, we will consider the nonlinear behavior of prismatic BE beams of which some are strongly curved, i.e., Kd > 0.1, either locally or globally. To the best of our knowledge, this is the first paper that deals with the geometrically exact analysis of strongly curved spatial BE beams within the framework of IGA. The paper is based on our previous works [50,51,45,47]. Its main contribution is the extension of the foundations given in [47,45] to the geometrically exact setting by defining the appropriate basis update and the consistent derivation of the stiffness matrix. The geometric stiffness is found by the careful variation of both the internal and the external virtual power with respect to the unknown metric. Special care is devoted to the variation of the twist variable and its gradient. The exact geometry of the spatial BE beam is utilized for the derivation of weak form of equilibrium which is solved by the Newton-Raphson and arc-length methods. A strict derivation of the constitutive relation allows us to derive reduced models and to assess the effects of these simplifications through the numerical analysis. Comparison with existing results confirms that the obtained formulation is reliable for the finite rotation analysis of arbitrarily curved beams. Moreover, the approach introduces an additional level of accuracy when dealing with strongly curved beams. The present formulation is geometrically exact in a sense that it strictly defines a relation between work conjugate pairs which allows analysis of arbitrarily large rotations and displacements [4]. Furthermore, by the careful implementation of the procedure for the basis update, the formulation is objective and path-independent [25].
The paper is structured as follows. The next section presents the basic relations of the beam metric and this is followed by the description of the BE beam kinematics. The finite element formulation is given in Section 4 and numerical examples are presented in Section 5. The conclusions are presented in the last section.
2 Metric of the beam continuum in the reference configuration The metric of a spatial beam model at some reference configuration, is elaborated. The position of the beam axis and the orientation of the cross sections are thus required. The definition of the position of a spatial curve is trivial, while the orientation of the cross sections can be defined in several ways. One approach is to define the orientation of the cross section at the beginning of the beam, and to choose an operator of some kind to describe its relative orientation along the beam [37]. An alternative approach is to define a reference basis at each point of the beam axis, and then to define the material basis with respect to the reference one [25]. The latter approach is applied here by using the Frennet-Serret frame as the reference frame [45]. This frame represents the intrinsic frame of the beam axis since one of its basis vectors, n, is aligned with the curvature vector of the beam axis, c = Kn. Boldface lowercase and uppercase letters are used for vectors and tensors/matrices, respectively. An asterisk sign is used to designate deformed configurations. Greek index letters take values of 2 and 3, while Latin ones take values of 1, 2 and 3. Partial and covariant derivatives with respect to the convective coordinates are designated with (•) ,m and (•) |m , respectively. Material time derivative is marked as(•).
For the details on the NURBS-based IGA modeling of curves, references [28,53] are recommended.
Metric of the beam axis
A beam axis is defined by its position vector r = r(ξ) = r(s), with s being the arc-length coordinate while ξ is some arbitrary parametric coordinate Fig. 1. In general, setting up an arc-length parametrization of arbitrary curve is not possible in terms of elementary functions, hence ξ is usually employed for the parametrization of the beam axis.
The position vector of the beam axis is given in global Cartesian coordinates by: Note that for the Cartesian coordinates, covariant and contravariant components are the same, and the position of indices is irrelevant. For every C 1 continuous curve, we can define a tangent vector: g 1 is the tangent vector of an arbitrary length, while t is the unit-length tangent of the beam axis. A function that defines the length of the basis vector g 1 is: At this point, let us introduce the well-known Frenet-Serret (FS) triad (t, n, b). Although not crucial for the present formulation, the intrinsic relation between the FS frame and the curvature of a line makes it useful for the presentation. The curvature vector of the beam axis is: where n is the unit-normal while K is the curvature of axis. The third FS base vector, binormal, is now defined as: and the well-known FS formulae follow: where t i = (t, n, b). Since the curvature tensor with respect to the arc-length parameter is skew-symmetric, we can define a pseudovector of curvature: where τ is the torsion of the FS frame while e is the permutation symbol. This allows us to rewrite the FS formulae as: In order to define the orientation of the cross section, we will introduce two unit base vectors, g 2 and g 3 , which are aligned with the principal axes of inertia of the cross section. In this way, (g 1 , g 2 , g 3 ) is a right-handed rectangular coordinate system at each point of a curve, see Fig. 1. For curves with well-defined FS frame, we can specify an angle α between t β and g β vectors, and express the basis vectors as: With respect to the parametric coordinate, the metric tensor of a line and its reciprocal counterpart are: The derivatives of the base vectors g i with respect to the parametric coordinate are: where Γ k i1 are the Christoffel symbols of the second kind of a line. Since g 1 is not of unit-length, the resulting curvature tensor is not skew-symmetric: Here, K i are the components of the curvature vector with respect to the material basis k = K i g i = K i g i . Concretely, K 1 is the torsion of the material frame g i , while K α are components of curvature with respect to the g α basis vectors. Furthermore,K α = gK α are the same components but measured with respect to the parametric coordinate. The expressions for the components of curvature follow as: It is now straightforward to find the relation between the components of the curvature vector with respect to the both coordinate systems [45]: Finally, we can rewrite Eq. (8) as:
Metric of a generic point in beam continuum
In order to completely define the geometry of a beam, we must define a coordinate system at each point of a beam continuum. Since the essence of structural theories is to reduce the problem, in this case from 3D to 1D, the complete metric of a beam should be defined by a set of reference quantities. It is common to use the metric of beam axis as the reference. For this, let us define an equidistant line which is a set of points for which (η, ζ) = const. Its position and tangent base vectors are: r (ξ, η, ζ) =r (ξ) = r + ηg 2 + ζg 3 , g 1 (ξ, η, ζ) =ḡ 1 =r ,1 = g 1 + ηg 2,1 + ζg 3,1 .
By introducing the so-called initial curvature correction term g 0 = 1 − ηK 3 + ζK 2 , the base vectors of an equidistant line are [20]: Evidently, these basis vectors are not orthogonal at a generic point, which complicates further reduction from 3D to 1D. In order to overcome this difficulty, a novel coordinate ξ λ = ξ λ (ξ, η, ζ) is utilized in [45]. It is a line orthogonal to the cross section at each point of a beam. By using the coordinate system (ξ λ , η, ζ), axial and torsional actions are decoupled. The tangent base vector of a ξ λ line is set to be equal to the tangent base vector of the beam axisḡ λ = g 1 . The present derivations are based on this coordinate system.
Bernoulli-Euler beam theory
The classical BE assumption states that a cross section is rigid and remains perpendicular to the beam axis in the deformed configuration. This allows us to describe the complete beam kinematics by the translation of beam axis and the angle of twist of cross section.
Kinematics of beam axis
The metric of the deformed configuration is described in a similar manner to that of the reference one, by defining the triad (g * 1 , g * 2 , g * 3 ) in the current configuration. This triad must be found by the update of some reference configuration, and this process is not uniquely defined. Regarding the tangent base vector, its definition is straightforward. The position and tangent base vectors of the beam axis in current configuration are: where u is the displacement vector of the beam axis. The other two basis vectors must be found by the rotation from some reference configuration: where R is the rotation tensor or rotator. In general, this tensor belongs to the special orthogonal group SO(3) and several parameterization of it exists [54,10]. Since the SO(3) group is nonlinear, it is convenient to switch to some linear space. Let us focus on the velocity field of the beam, since it is tangent to the displacement field. The velocity gradients along the η and ζ directions are: while the velocity gradient along the ξ coordinate follows as v ,1 =ġ * 1 =u ,1 . For each member of the SO(3) group, there exists an appropriate spinor Φ which belongs to the so(3) group of skew-symmetric tensors [20]. This spinor represents the infinitesimal rotation and allows the exponential parameterization of the rotator: In this way, the velocity gradients (20) The material derivative of the spinor Φ is the antisymmetric part of the velocity gradient -the angular velocity. Its components define the psudovector ω with respect to the local triad, [4]: where: Here, is the Levi-Civita symbol whilev k are the components of velocity with respect to the local triad. It is evident from the last equation that two components of angular velocity, ω 2 and ω 3 , depend on the velocity field of beam axis: This observation confirms that a rotation of a cross section of a BE beam belongs to the SO(2) group of in-plane rotations [25]. The 2D rotation occurs in the normal plane of the deformed beam axis which is uniquely defined with g * 1 . Therefore, the only independent component of the angular velocity for the BE beam is the one with respect to the tangent of the beam axis: Besides the components of the velocity of the beam axis, this component of angular velocity represents the fourth DOF of the BE beam model. We will designate its physical counterpart simply with ω. This quantity represents the angular velocity of a cross section with respect to the tangent of the beam axis. It is discussed in [45,55] that this quantity can be decomposed into two parts, one part being the angular velocity of the FS frame. Since this approach is not applicable for general spatial curves, we will consider ω as the complete twist of the cross section over the increment of time.
For the sake of further derivations, let us find the gradients of velocity with respect to the principal axes of inertia as functions of the generalized coordinates. By using Eqs. (22) and (25) we obtain: The gradients v , 21 and v ,31 are also required:
Basis update
As discussed previously, the tangent base vector of the beam axis, g * 1 , follows directly from the current position of the beam axis. The other two base vectors, g * α , must be found by the rotation, Eq. (19). Due to the fact that only one rotation is the generalized coordinate of BE beam, the parameterization of this rotation is significantly simplified in comparison with shear-deformable models. Namely, the base vectors g * α lie in the normal plane of the deformed beam axis and they are calculated as an in-plane rotation of some reference vectors g ref α . The definition of these reference vectors is not unique. Two procedures for the update of basis vectors are considered here, the Smallest Rotation (SR) and the Nodal Smallest Rotation Smallest Rotation Interpolation (NSRISR).
Smallest Rotation
The SR mapping is commonly used for the modeling of BE beams [42]. The procedure consists of two steps. First, the triad from the previous configuration is rotated in order to align its tangent t with the tangent of the current configuration t * . This is done in such a way that this rotation angle is minimized, and the reference vectors are: In the second step, this reference frame is rotated in its plane by the angle ∆θ = ω∆t, where ∆t is the current time increment. As a result, the base vectors of the current configuration are found: cos ∆θ sin ∆θ − sin ∆θ cos ∆θ Note that this mapping has a singularity for t * · t = −1 which is of no practical interest if the reference configuration is appropriately defined [42].
Nodal Smallest Rotation Smallest Rotation Interpolation
Unfortunately the SR procedure leads to the non-objectivity of the resulting formulation. Therefore, the NSRISR mapping is introduced in [25] in order to overcome the deficiency of the SR mapping. It is based on the considerations originally developed in [12]. Concretely, although the continuum strain measures of the geometrically exact beam theory are objective, their finite element implementations are not. The problem is caused by the interpolation of the rotations between the current and some reference configurations. This approach, in general, includes rigid-body rotations. In order to solve this issue, the authors of [12] suggested a linear interpolation of the relative rotation between the element nodes. Since this relative rotation is free from any rigid-body motion, the objectivity of the discretized strain measures is preserved.
The NSRISR algorithm consists of three steps. First, the SR procedure is applied to the g α vectors at the start and the end of the finite element. The resulting vectors are designated as g SR α,start and g SR α,end . Second, the SR mapping is applied to the vectors g SR α,start to develop the reference frame g ref α (ξ) at each point of the finite element, ξ ∈ [ξ start , ξ end ]. In order to update the basis correctly, this reference frame should be rotated by the angle ∆θ + ∆θ c , where ∆θ c is the correction angle. This angle represents the difference between the reference frame g ref α (ξ), and the one which follow from the classic SR procedure described in the previous Subsection. The correction angle ∆θ c is zero at the start of the element, while, at the end of an element, it can be simply calculated as the angle between the frames g SR α,end and g ref α (ξ end ). Finally, the correction angle is linearly interpolated between known values at the start and the end, which leads to the function ∆θ c (ξ). In this way, the updated basis is calculated: cos (∆θ + ∆θ c ) sin (∆θ + ∆θ c ) − sin (∆θ + ∆θ c ) cos (∆θ + ∆θ c ) Derivatives of the base vectors g * α with respect to the parametric coordinate ξ can be calculated straightforwardly from the last expression.
Strain rates
After the metric of the beam continuum is defined, the next step is to introduce a strain measure. Let us assign the convective property to the coordinates (ξ, η, ζ). For the convective coordinates, the Lagrange strain equals the difference between the current and reference metrics: At an arbitrary point of beam continuum, this strain is: while the components of the strain rate are equal to the material derivatives of strain: Due to the BE hypothesis, the shear strain rates d 21 , d 31 , d 23 , d 32 vanish.
The strains at a generic point of beam continuum are strictly derived in [45] with respect to the (ξ λ , η, ζ) coordinates. We omit this derivation for the sake of brevity. The resulting strain rates are: where d 11 is the axial strain rate of beam axis,κ 1 is the rate of change of torsional curvature, andκ α are the rates of changes of bending curvatures with respect to the principal axes of inertia. These four quantities represent the reference strain rates of BE beam since they allow the calculation of the complete strain rate field of a beam. The next step is to derive the relation between the reference strain rates and the generalized coordinates. For the axial strain rate, the derivation is simple: while the curvature components require more attention. By using Eq. (13), we obtain: and after introducing Eqs. (27) and (28) into Eq. (37), the final expressions for the rates of curvature changes are:κ
Spatial discretization
Using IGA, both geometry and velocity are here discretized with the same univariate NURBS functions R I . However, different univariate NURBS functions R ω J are utilized for the approximation of the angular velocity: where (•) I stands for the value of quantity at I th control point. If we introduce a vector of generalized coordinates, v T ω = v T ω , and the matrix of basis functions N, the kinematic field of the beam can be represented as: where: It is important that the interpolation of the angular velocity ω must have C 0 interelement continuity in order for the NSRISR formulation to keep its objectivity. This fact is proved in [25] and it will be considered in Section 5 via the numerical analysis. The k -refinement, a distinct feature of IGA, ensures the highest-possible continuity, C p−1 , while increasing the degree, p, of the spline space [28]. By careful subsequent knot insertion, interelement continuity can be reduced at required points. Our FEM implementation employs this ability for the different discretization of the velocity and angular velocity fields over the same mesh. This is the reason why, in general, we allow that N = M , and R I = R ω I in Eq. (39).
Finite element formulation
In line with the previous derivation, we will formulate isogeometric spatial BE element using the principle of virtual power.
Let us start from the generalized Hooke law for the linear elastic material, also known as the Saint Venant-Kirchhoff material model: where µ and λ are Lamé material parameters. By introducing the BE constraintsσ 2 2 = σ 3 3 = 0, the non-zero contravariant components of stress rate with respect to the (ξ λ , η, ζ) coordinates are [45]: Here, E is modulus of elasticity and µ corresponds to shear modulus.
Principle of virtual power
The principle of virtual power represents the weak form of the equilibrium. It states that at any instance of time, the total power of the external, internal and inertial forces is zero for any admissible virtual state of motion. If the inertial effects are neglected and body and surface loads are reduced to the beam axis, this can be written for a spatial BE beam as: where σ is the Cauchy stress tensor, d is the strain rate tensor, while p and m are the vectors of external distributed line forces and moments, respectively. All these quantities are defined at the current, unknown, configuration. By assuming that the load is deformation-independent, only the stress is linearized: whereσ is the stress rate tensor which is calculated as the Lie derivative of current stress.
Since the components of the stress rate tensor are equal to the material derivatives of the components of the stress tensor, [56], the linearized form of the virtual power is: V σ 11 δd 11 +σ 12 δd 12 +σ 13 δd 13 dV ∆t + V σ 11 δd 11 +σ 12 δd 12 +σ 13 δd 13 dV = wherep i andm i are the components of distributed line forces and moments with respect to local coordinates. We will simplify the notation for the remainder of this paper, by neglecting the asterisks. We emphasize that this change in notation does not introduce any ambiguity since (i) the stress and strain rates are instantaneous quantities, while the known stress is calculated at the previous configuration, and (ii) all integrations are performed with respect to the metric of the current configuration, in accordance with the updated Lagrangian procedure [57].
By integrating the left-hand side of Eq. (46) with respect to the area of the cross section, the integrals over the 3D volume reduce to line integrals along the beam axis: whereÑ andM j are stress resultant and stress couples, that are energetically conjugated with the reference strain rates of the beam axis, d 11 andκ j .Ṅ andṀ j are their respective rates given by:Ṅ If we introduce the following vectors: Eq. (46) can be expressed in compact matrix form as:
The relation between energetically conjugated pairs
The geometrically exact relations (47) and (48) are crucial for the accurate formulation of structural beam theories. In particular, energetically conjugated pairs are defined rigorously and the appropriate constitutive matrix is guaranteed to be symmetric. By the introduction of Eqs. (35) and (43) into Eq. (48), the exact relation between energetically conjugated pairs of stress and strain rates is obtained. The resulting symmetric constitutive matrix D is derived in [45]: where the geometric properties of cross section are the functions of curvature: If we introduce six integrals: the geometric properties in Eq. (52) can be rewritten as: It is emphasized that these integrals can be analytically computed for standard symmetric solid cross section shapes. The derived exact constitutive model is designated as D a further in the paper.
This strict derivation allows us to examine the influence of the exact constitutive relation. Four reduced constitutive models are considered for this purpose. The first and the simplest one is designated with D 0 . This model is based on two assumptions: (i) g 0 → 1, and (ii) the matrix D is diagonal: Therefore, this model completely disregards the coupling between bending and axial actions [25]. The second reduced model, D 1 , also employs the first assumption of D 0 , g 0 → 1, but it keeps the coupling terms from the matrix D. This D 1 model is readily utilized for the analysis of beams with small curvature [37,45]. The involved integrals simplify to: Let us emphasize that these results are valid only when the integrals are calculated with respect to the principle axes of inertia. Additionally, the higher order terms with respect to the curvature in the property A, Eq. (54), are neglected for this constitutive model, which yields A = H 1 .
In order to assess the effect of the higher order terms, we introduce two additional reduced models, D 2 and D 3 , which employ a Taylor approximation of the exact expressions (54). To be precise, the model D 2 is based on the 1 st order Taylor approximation of integrands which results in: The model D 3 , on the other hand, follows from the 2 nd order Taylor approximation of the integrands: For beams with small curviness and circular cross section, the term I 11 reduces to the polar moment of area I 0 , Eq. (55). However, for all the other shapes of cross section, this fact does not hold. Instead, the term I 11 is usually replaced with the so-called torsional constant, J, which must be calculated approximately [42]. In the presented research, J is used for the models D 0 and D 1 . For the other three constitutive models, we have scaled the geometric property I 11 with the ratio J/I 0 , in order to keep the influence of the curviness. However, this aspect did not affect any of our numerical experiments, even for strongly curved beams. Hence, these studies are skipped in the numerical experiments section, for the sake of brevity.
Discrete equations of motion
Let us turn now to the spatially discretized setting. First, we will define the matrix B L which relates the reference strain rates with the generalized coordinates at the control points by using Eqs. (36), (38) and (41): where: Since the strain rate is a function of the generalized coordinates as well as the metric, we must vary it with respect to both arguments. By the variation with respect to the generalized coordinates, a linear (material) part of the tangent stiffness is obtained. Variation with respect to the geometry results with the geometric stiffness matrix [56]. Clearly, variation of the reference strain rates in Eqs. (36) and (38) with respect to the generalized coordinates is trivial. On the other hand, variation with respect to the metric is more involved and it is given in detail in Appendix A. Formally, the variation of strain rate can be written as: Let us define the matrix of basis functions B G , which can be obtained by removing the 8 th row from the matrix B in Eq. (60), and the matrix of generalized section forces G with elements G ij , cf. Appendix A. Now, by the insertion of Eq. (61) into Eq. (50), we can reformulate the term generated by the known stress and the variation of the strain rate with respect to the metric: where K G is the geometric stiffness matrix. The complete derivation of this term is given in Appendix A. Now, the integrands in the equation of the virtual power (50) reduce to: where the first term is linearized by neglecting the variation of strain rate with respect to the metric.
Regarding the external virtual power, if we vary it with respect to the kinematics, the vector of the external load, Q, is recovered: and it is the same as in the linear analysis [45]. However, if we vary the external virtual power with respect to the geometry, a contribution to the geometric stiffness is obtained. The derivation of this term is given in Appendix B. Its implementation in the formulation is important for the successful convergence of the nonlinear solver. Finally, the discretized and linearized equation of equilibrium is: which can be readily written in the standard form: where: is the tangent stiffness matrix and: is the vector of internal forces. The vector ∆q in Eq. (66) contains increments of displacements and twist at control points with respect to the reference configuration. Due to the approximation introduced with Eq. (63), the solution of Eq. (66) does not satisfy the principle of virtual power and some numerical procedure must be applied. Here, both the Newton-Raphson and arc-length methods are employed.
Numerical examples
The aim of the following numerical studies is to validate the presented approach and to examine the influence of the curviness on the structural response. The boundary conditions are imposed in a well-known manner and the rotations are treated with special care since two components are not utilized as DOFs [45]. Components of external moments with respect to the principal axes of inertia are applied as force couples which must be updated at each iteration [52]. Standard Gauss quadrature with p + 1 integration points per element are used here. All the results are presented with respect to the load proportionality factor (LP F ), rather than to the load intensity itself. Since the time is a fictitious quantity in the present static analysis, strain and stress rates are equal to strains and stresses, respectively. Besides, increments of displacement and twist are equal to velocity and angular velocity.
Two approaches are considered here in the context of the reference configuration for the basis update, the total and the incremental. For the total approach, the reference configuration is the initial unstressed configuration [58]. For the incremental approach, the previously converged configuration is adopted as the reference one [6]. The total approach should guarantee that the results are path-independent, since no history of deformation is included in the reference configuration.
Furthermore, four element formulations are considered: SR−C p−1 , SR−C 0 , N SRISR− C p−1 , and N SRISR − C 0 . These formulations vary in the basis update algorithm (SR or N SRISR), and on the interelement continuity used for twist (C 0 or C p−1 ). For the displacement of the beam axis, the highest available interelement continuity is applied exclusively. If not stated otherwise, the order of the spline functions used for the approximation of the twist is the same as that used for the displacement of the beam axis.
In some examples, the error of the position vector r is calculated using the relative L 2 -error norm, as proposed in [25]: where L is the length of the beam and u max is the maximum value of the displacement component. r h represents the approximate solution for the position of the beam axis while r ref is the appropriate reference solution.
Objectivity
The objectivity of the formulation can be considered as the invariance with respect to the rigid-body motions. Thus, if the beam is subjected to a rigid-body motion, no strain should appear. Invariance with respect to the translations is readily satisfied while the invariance with respect to the rotation requires special attention [22]. The present example was introduced in [25]. A quarter-circular cantilever beam is rotated ten times around its clamped end with respect to the y-direction, see Fig. 2a.
For an objective formulation, there should be no internal strain energy Π int . Due to the extremely large rotations involved in this example, only the incremental update of rotations is considered. The non-homogeneous boundary condition, ϕ y = 20π, is applied in 100 increments. Two NURBS orders are considered, p = 3 and p = 4.
The internal strain energy in the final configuration is plotted in Fig. 2b with respect to the number of elements. The obtained results are scaled with the value Π int,0 = EIπ/(4R). This is the internal strain energy of an initially straight beam that is bent into a quarter circle with tip moment [25]. These results confirm that the N SRISR − C 0 formulation is indeed objective and the internal strain energy equals zero up to the machine precision. The beam deforms in all the other formulations considered. Furthermore, the results indicate that the problem with the representation of rigid rotations is mitigated for all formulations when the number of elements is increased. This is a well-known fact, which sometimes justifies the application of non-objective formulations in quasi-static analyses [59]. Fig. 3a shows the evolution of the internal strain energy with the number of rotations for different formulations and NURBS orders. Note that the SR − C p−1 formulation with cubic NURBS elements generates around 0.7 % of the reference strain energy, while it is less than 0.2 % for SR − C 0 case. A similar behavior can be observed for quartic elements. It follows that the SR formulation greatly benefits from the reduced interelement continuity of the twist variable.
The behavior of the N SRISR−C p−1 formulation is also worth mentioning. The results suggest that the initially accumulated internal strain energy, during first few increments, remains constant. In order to examine this more thoroughly, the strain energies for the N SRISR − C p−1 formulation and four different increment sizes are plotted in Fig. 3b for LP F < 0.15. It is apparent that the error increases with the size of the increment while asymptotically approaching constant value. For this formulation, the beam deforms at the beginning of the loading process and rotates afterwards without the additional straining.
Path-independence
Path-independence of the solution can be analyzed in various ways. Some authors simply apply different sizes of load increments, [13], while the others change the order of the applied load [25]. Here, we employ the latter approach. A quarter-circle cantilever beam is loaded with two forces at the free end, as shown in Fig. 4. Three cases of the application of load are considered. First, both forces are applied simultaneously -SIM case. For the other two cases, the forces are applied successively, one for 0 < LP F < 0.5 and the other for 0.5 < LP F < 1. The case when the F X is applied first is designated with SU CXZ while the other case is marked with SU CZX.
The deformed beam configurations for all three load order cases and the four characteristic LP F s are shown in Fig. 5. Apparently, all load cases yield similar final configurations in visual terms, but each with a different deformation history. Next, the relative L 2 -error norms for the SIM and SUCZX loading orders are observed for LP F = 1 using Eq. (69). Two formulations are considered, SR − C p−1 and N SRISR − C 0 , and two algorithms for the basis update, incremental and total. The results for quartic NURBS are shown in Fig. 6. Our simulations show that of these four combinations, only the SR − C p−1 formulation with the incremental update of rotations is path-dependent. As expected, when the unstressed configuration is used as the reference one for the update of the basis, the solution is path-independent [12]. An important fact is the confirmation that the N SRISR−C 0 formulation with incremental update of rotations is path-independent [25]. For all the three considered formulations that are path-independent, the relative L 2 -error norms are zero up to the machine precision.
Regarding the SR−C p−1 formulation with incremental update of rotations, the relative L 2 -error norm is plotted in Fig. 7a for three NURBS orders. The error due to the pathdependence mitigates with the mesh refinement. In fact, the order by which the pathdependent error reduces should be similar to the order of convergence of the discretization error [25]. To test this, 200 quartic elements and the N SRISR − C 0 formulation are used to obtain a reference solution. The convergence of error for three different NURBS orders are displayed in Fig. 7b. For this graph, results from the N SRISR − C 0 formulation are used with SIM load order. Nevertheless, the discretization errors are practically the same for all the formulations and load orders. The expected orders of convergence, close to p+1, are observed and they are similar to those in Fig. 7a. Furthermore, the comparison of the magnitudes of error in Fig. 7a and Fig. 7b confirms the fact that the path-dependent error can be considered negligible with respect to the discretization error [25].
In the following, if not explicitly stated otherwise, the N SRISR − C 0 formulation with incremental update of rotations is exclusively used.
Circular ring subjected to twisting
This is a well-known example that is frequently considered for the validation of formulations involving finite rotations [25,19]. A circular ring is subjected to the symmetrical twisting as shown in Fig. 8. Here, the twist is applied as a pair of external concentrated moments M = EI/R. Due to the symmetry of the load and the geometry, only a quarter of the ring is modeled [60]. In order to obtain converged strains, a dense mesh of 32 quartic elements is used. Furthermore, two different dimensions of the cross section are considered and designated as Case 1 and Case 2, see Fig. 8. Case 1 corresponds to the examples that are frequently found in the literature. The dependence of LP F and the external angle of twist on one side of the ring is usually observed for the verification. We have adopted the result from [61] as the reference solution. The obtained result is compared with the reference result given in Fig. 9a and excellent agreement can be observed. Furthermore, the curviness at three characteristic points is displayed in Fig. 9b. For Case 1, the maximum curviness is less than 0.07 and all constitutive models return the same results.
In order to examine the influence of different constitutive models on a beam with large curviness, Case 2 is now considered in detail. Although the initial curviness for this beam is 1/15, it increases to almost 0.22 during the deformation, Fig. 9b, and the beam becomes strongly curved. Furthermore, note that the external twisting of ϕ = 180 • deforms the ring into a smaller ring, with a diameter reduced by a factor of three. Additional application of the external twisting returns the ring into its original configuration for ϕ = 360 • , Fig. 10. In the following, the complete cycle of external twisting is considered (ϕ = 360 • ) and the reference strains of the beam axis are observed at points A and B, Figs. 11 and 12. The D 0 and D 1 models return erroneous results for the axial strain. For the other three strain components, the results obtained by the different constitutive models are similar, mostly due to the fact that the maximum curviness of this beam is local. However, a close inspection of the equilibrium paths reveals that the differences exist. This detailed insight suggests that the D 3 and D a models return practically indistinguishable results.
The example is suitable for the testing of the path-independence due to the cyclic response of this ring [19]. For this purpose, the torsional strain at point A is observed, while the ring is twisted eight times (ϕ = 16π). The results calculated with the SR − C 0 and N SRISR−C 0 formulations using the incremental update of the basis are given in Fig. 13. This analysis confirms the conclusions from the previous example. The N SRISR − C 0 formulation is path-independent while the SR − C 0 formulation with incremental update of the basis is not.
Straight beam bent to helix
As a final example, let us consider the response of an initially straight cantilever loaded with two moments as indicated in Fig. 14a. The influence of the polynomial used for the approximation of the twist variable is examined first. The beam is loaded with M x = M z = 10, while the displacement of the beam axis is discretized with quartic elements. The relative L 2 -error norm of the position of the beam axis at the final configuration is calculated by Eq. (69). The reference solution is obtained from an analytical expression for beams with small curvature, proposed in [26]. Convergence of the D 1 model towards the reference solution is shown in Fig. 14b. We can observe that the polynomial order used for the discretization of the angle of twist does not have significant influence. In fact, all the considered approximative functions return similar results with the exception of quadratic polynomial for the densest mesh. Moreover, the influence of the polynomial order of the twist variable is negligible for all present numerical results.
Next, the beam is discretized with 40 quartic elements and loaded with M x = M z = 20. The tip displacement along the z-axis is plotted in Fig. 15 for different constitutive models. In order to make close inspection, parts of the equilibrium path for LP F > 0.2 are enlarged. As the load increases, the differences between models become visible. This is due to the fact that the maximum curviness for this example is Kh ≈ 0.34. The sponse. The beam deforms into a helix, and the equilibrium requires that the normal force along the whole beam is zero. If the small-curvature beam model is utilized, the axial strain of the beam axis must also be zero. However, an accurate model, such as the one presented here, results in the dilatation of the beam axis, in a manner similar to that in [52]. The distribution of the axial strain and the normal force at the final configuration are given in Fig. 17 for different constitutive models. The D 0 model returns an erroneous sign of the dilatation of the beam axis and non-zero normal force. On the other hand, the D 1 model results in zero dilatation and non-zero normal force. It should be noted that this model would result in zero normal force if the reduced constitutive matrix is used for the post-processing of the section forces [52]. Here, the full constitutive relation is utilized for the calculation of section forces [45]. Finally, the D 2 , D 3 , and D a models return similar results, with extensional axial strain and near-zero normal force. Again, the D 3 and D a models are fully aligned while the D 2 model differs slightly. Evidently, both the axial strain and the normal force oscillate strongly around the exact value. To examine this effect, the normal forces for three different mesh densities are given in Fig. 18a. The oscillation of these quantities reduces with the mesh density which can indicate the presence of membrane locking. The development of the curviness at three characteristics points is shown in Fig. 18b. It is interesting to note that the curviness at the clamped end increases monotonically while it is not the case at the other positions. This is due to the fact that the curviness at the clamped cross section varies solely due to the change in curvature. For the other points, the cross section rotates and the curviness exhibits a more complex behavior.
Conclusions
The first truly geometrically exact isogeometric formulation of a spatial Bernoulli-Euler (BE) beam is presented. The rigorous metric of the BE beam is utilized consistently for the derivation of the weak form of equilibrium. The introduction of the full beam metric gave a higher-order accurate BE beam formulation. The exact constitutive relation is employed for the derivation of four simplified models which are compared via numerical examples. Moreover, by the implementation of the Nodal Smallest Rotation Smallest Rotation Interpolation (NSRISR) mapping, an objective and path-independent formulation is obtained.
It is confirmed that the Smallest Rotation (SR) mapping results in a non-objective formulation, the error of which reduces with the mesh density. Additionally, it is shown that both the NSRISR and the SR algorithm benefit from a C 0 interelement continuity of the twist variable.
In order to obtain the correct geometric stiffness matrix, both the internal and the external virtual power are rigorously varied with respect to the unknown metric. The angle of twist requires special attention, since it has one part that depends on the geometry and its variation must be performed consistently.
The presented results suggest that, in order to correctly determine the strains of a strongly curved beam, a higher-order accurate computational model must be employed. For the beams with small curviness the simple decoupled equations return reasonably accurate results. As the curviness increases, its influence becomes noticeable and a more involved model is required. A simple yet effective solution to improve the accuracy of small-curvature formulations should include the exact nonlinear distribution of strain and stress in the post-processing phase [52].
Remark. It is interesting to note that, since the angular velocity can be written as ω = g 3 · v ,2 or as ω = −g 2 · v ,3 , its variation with respect to the metric can be written in two ways. The first one is given with Eq. (A7) while the other one is: Both representation are valid and can be used for the derivation of the geometric stiffness. However, the choice of the representation for δω must be consistently applied. Its inconsistent use is actually the source of error in the reference [47]. Furthermore, although the both representations of the angular velocity return the same value, their variations in Eqs. (A7) and (A8) are not the same, since the variation is performed with respect to the different parameters. | 2021-09-22T01:16:03.754Z | 2021-09-21T00:00:00.000 | {
"year": 2021,
"sha1": "42039b18ea2c0cc44df5b84161f78a6e77352ee9",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "42039b18ea2c0cc44df5b84161f78a6e77352ee9",
"s2fieldsofstudy": [
"Engineering",
"Physics"
],
"extfieldsofstudy": [
"Computer Science",
"Mathematics"
]
} |
270586648 | pes2o/s2orc | v3-fos-license | Impact of commissural calcification on clinical outcome of percutaneous balloon mitral valvuloplasty; a retrospective cohort study of 876 patients
Background Percutaneous balloon mitral valvuloplasty (PBMV) is the ACC/AHA class I recommendation for treating symptomatic rheumatic mitral stenosis with suitable valve morphology, less than moderate MR and absence of left atrium clot. The mitral valve restenosis and significant mitral regurgitation (MR) are known adverse outcomes of PBMV. This study aimed to evaluate the outcomes of PBMV in patients with severe mitral stenosis and the effect of Commissural Calcification (CC) on the outcomes. Methods In this single-center retrospective cohort study, 876 patients who underwent PBMV were categorized into three groups based on their Wilkins score (Group I: score ≤ 8, Group II: score 9–10, and Group III: score 11–12). Patients were evaluated before, early after PBMV and at 6- and 24-month follow-ups. Main clinical outcomes were defined as significant restenosis and or symptomatic significant MR (moderate to severe and severe MR) or candidate for mitral valve replacement (MVR). The outcomes were compared between patients with and without CC. Results A total of 876 patients with mean age 46.4 ± 12.3 years (81.0% females) were categorized based on Wilkins score. 333 (38.0%) were in Group I, 501 (57.2%) were in Group II, and 42 (4.8%) were in Group III. CC was present in 175 (20.0%) of the patients, among whom 95 (54.3%) had calcification of the anterolateral commissure, 64 (36.6%) had calcification of the posteromedial commissure, and in 16 (9.1%) patients both commissures were calcified. There was a significant difference in Wilkins score between patients with and without CC (P < 0.001). CC was associated with higher odds of significant symptomatic MR at early and mid-term follow up (OR: 1.69, 95%CI 1.19–2.41, P = 0.003; and OR: 3.90, 95%CI 2.61–5.83, P < 0.001, respectively), but not with restenosis (P = 0.128). Wilkins Groups II and III did not show higher odds of significant symptomatic MR compared to Group I at early (II: P = 0.784; III: P = 0.098) and mid-term follow up (II: P = 0.216; III: P = 0.227). Patients in Wilkins Group II had higher odds of restenosis compared to Group I (OR: 2.96,95%CI: 1.35–6.27, P = 0.007). Conclusion Commissural calcification (CC) is an independent predictor of the significant symptomatic MR (an important determinant of adverse outcome) following PBMV in the early and mid-term follow-up. Mitral valve restenosis occurs more in patients with higher Wilkins score compared to group I with score ≤ 8. Combined Wilkins score and CC should be considered for patient suitability for PBMV.
Introduction
While rheumatic mitral stenosis (MS) is not a significant burden of cardiovascular disease in developed countries, it has remained one of the most common cause of heart valve disease in developing countries [1].Untreated mitral stenosis is associated with poor prognosis, and may lead to complications such as pulmonary hypertension, atrial fibrillation and heart failure [2].
Percutaneous balloon mitral valvuloplasty (PBMV) is the ACC/AHA class I recommendation for the treatment of symptomatic rheumatic mitral stenosis with suitable valve morphology, less than moderate MR, and absence of left atrium clot [3].A successful PBMV will result in an increase in mitral valve area (MVA) without significant MR [4][5][6][7].However, complications associated with PBMV, including significant MR, minimal improvement in clinical status or MVA, underscore the need for a nuanced preprocedural evaluation and correlation with the outcomes [4,5].MR is the one of the most frequent complications of PBMV [6].While the severity of MR after PBMV is often mild, more severe MR occurs in about 20% of the patients [7,8].Other complications of PBMV, such as embolic stroke, cardiac tamponade, and cardiac perforation, are rare.The role of echocardiographic characteristics of the mitral valve in predicting the severity of MR after PMV is controversial [9][10][11][12].However, structural characteristics of the mitral valve are the most important factors for predicting the outcomes after PMV.The Wilkins score system is widely used to assess the suitability of the mitral valve leaflets for PBMV based on their morphological features including leaflet mobility, subvalvular thickening, leaflet thickening, and leaflet calcification [13,14].This scoring system is widely used to identify the patients who will benefit from PBMV as lower Wilkins scores are associated with a better-thanaverage prognosis after PBMV [8].The main outcomes of PBMV include mitral valve regurgitation and mitral valve restenosis [9].While the structure of the valvular commissures is not included in the Wilkins scoring system [15,16], current studies have shown its association with the occurrence and severity of post-procedural MR [17,18] and restenosis [10].In vivo and in vitro studies have confirmed that commissural splitting is the dominant mechanism by which mitral valve area (MVA) is increased during balloon dilatation [11,12].This underscores the importance of a more thorough assessment of CC as a predictive factor for post-procedural complications.In this study, we aimed to evaluate MV CC impact on the short-term and mid-term outcomes after PBMV and to compare its predictive value with the conventional Wilkins score system.We hypothesized that the presence of CC would significantly predict the incidence of MR at short-term and mid-term follow-ups and might predict the occurrence of restenosis after PMV.
Study design and population
In this retrospective Cohort single-center study, 995 symptomatic moderate to severe and severe mitral stenosis (MVA ≤ 1.5 cm2) patients who underwent PBMV between 2009 and 2019 at Rajaie Cardiovascular, medical, and Research Center in Tehran, Iran, were enrolled.Exclusion criteria were MR (Moderate or severe), left atrial/ left atrial appendage thrombosis, Wilkins score > 12, concomitant coronary artery disease, and a history of previous valvular surgery.Patients who had major complications including tamponade, valve rupture, severe MR within the first 24 h after PBMV, and patients who were lost to follow-up (during the first 6 months after successful procedure) were excluded.All remaining patients were included in data analysis.
Clinical assessment and follow-up
All patients were evaluated using transthoracic echocardiography (TTE) at the echocardiography unit of Rajaie cardiovascular, medical, and research center before and 24 h after PBMV to confirm the success of the procedure.All measurements were performed by expert cardiac sonographer and reviewed by a level III echocardiologist based on the recommendations of the American Society of Echocardiography guidelines [15].MVA by planimetry and pressure half-time was measured in all patients.
All patients underwent PBMV using a self-positioning single balloon (Inoue balloon) and a step-wise dilation strategy based on current guidelines [7].The upper limit of the balloon diameter was decided according to the patient's height.Successful PMV was defined as post-PBMV > 1.5 cm 2 or at least a 25% increase in the valve area with less than one grade increase in MR and without any major complications such as tamponade or cardiac perforation.MR severity was assessed by echocardiography using integration of multiple echocardiographic parameters and scored as absent or trivial, mild, moderate, moderate to severe, and severe [13,14].An ultrasound evaluation of the mitral valve structural features and subvalvular apparatus was performed for each patient before the procedure, supervised and interpreted by a level III echocardiographer, Wilkins echocardiographic scoring system was used to evaluate mitral valve structure.This system consists of 4 parameters: (a) leaflet thickening, (b) leaflet mobility, (c) leaflet calcification, and (d) subvalvular thickness.Each component is assigned a score from 0 to 4, and summing the individual scores results in a total mitral valve Wilkins score that ranges from 0 to 16. Patients were divided into three subgroups based on the Wilkins score.Group I consisted of patients with a Wilkins score ≤ 8, Group II included patients with a Wilkins score of 9-10, and Group III consisted of patients with a Wilkins score of > 10.Also, we have grouped the patients into two subgroups based on the presence or absence of CC.CC was determined visually when there was a bright echo density (higher signal density compared to the adjacent structure) on mitral valve commissures by 2D TTE.The presence of shadow behind the calcification was a confirmatory sign of calcification but not applicable for all subjects [11,15].
Study outcomes
The presence of symptomatic significant MR (moderate to severe or severe) that patient was candidate for surgical Mitral valve replacement (MVR) and/or mitral valve restenosis (defined as a 50% reduction in MVA after successful PBMV) served as the study outcomes.Symptomatic MR was defined as dyspnea New York Heart Association (NYHA) class II or greater [16].Patients were followed with TTE and screened for MR severity at 6 and 24 months and for restenosis at 24 months.Incidence of MR and restenosis were compared based on the presence or absence of CC and based on the Wilkins score categories to identify the association between CC and Wilkins score with short-term and mid-term outcomes after PMV.
Statistical analysis
Normally distributed continuous variables are described with mean ± SD and were compared with student T-Test, while non-normally distributed continuous variables are described with median [IQR] and were compared using the Mann-Whitney U test.Categorical variables are described as number (percent) and were compared suing the Chi-square or Fisher's exact tests.The Wilcoxon test was used to compare echocardiologic parameters over time.Pearson test was used to find the correlations between the variables.
Univariable binary logistic regression analysis was used to explore the association between predictors (CC and Wilkins score categories) and study outcomes.Multivariable binary logistic regression models were fitted to assess the association between predictors and study outcomes after adjusting for covariates.For the multivariable models, X 2 , degrees of freedom, and P-values were reported, and odds ratios, 95% confidence intervals and P-values were reported for each predictor.Nagelkerke's R ^2 was used to quantify the contribution of predictors to the outcome.The Hosmer and Lemeshow test was used to assess goodness of fit and Akaike information criterion was applied to avoid overfitting.
Statistical analysis was performed using SPSS version 28 and R version 4.3.2.A P-value less than 0.05 was considered significant.
Baseline characteristics
The medical records of 1032 patients after PBMV were reviewed.Seventy-six patients were excluded due to acute major complications during the first 24 h after PBMV (32 patients developed tamponade, and 44 patients underwent surgical mitral valve replacement due to valve rupture and severe MR).Of the remaining 956 patients, 80 patients were excluded due to lack of followup.Finally, 876 patients were included in our analysis (Fig. 1).The median age of the patients was 46.0 years, and 81.1% of the patients (n = 710) were females.Among participants, 333 patients (38.0%) had a Wilkins score of 8 or less were included in group I, 501 patients (57.2%) had a Wilkins score of 9 or 10 and were assigned to group II, and 42 patients (4.8%) with a score greater than 10 were classified as group III.Median Wilkins score 1.There was a significant difference in Wilkins score between patients with and without CC (P < 0.001) (Fig. 2).
At baseline, there was no significant difference in MVA, MVAI, PAP, mean LA pressure and LVEF between patients with and without CC (all P > 0.05).However, there was a significant difference in MVA, MVAI, and PAP among Wilkins score groups, with patients in Wilkins group III having lower MVA and MVAI and higher PAP compared to the other groups (P < 0.001, P = 0.003, and P = 0.002, respectively).
Mitral regurgitation at mid-term
At two years follow up, hundred and twenty-six (14.4%) patients had significant symptomatic MR that led to MVR. (Fig. 3C).There was a significant difference in the prevalence of MR in mid-term evaluation between patients with and without CC (CC: 54 (30.9%); no CC: 72 (10.3%),P < 0.001).Although there was an incremental increase in the prevalence of MR in mid-term follow up among Wilkins groups I through III, this difference was not statistically significant (group I: 41 (12.3%); group II: 77 (15.4%); group III: 8 (19%), P = 0.317) (Fig. 3D).
In univariable binary logistic regression analysis, Wilkins groups II and III had higher odds of developing MR in mid-term, however these differences were not
Mitral valve restenosis at mid-term follow-up
Mitral valve restenosis occurred in 45 (5.1%) patients during the 24-month follow-up.There was no significant difference in the prevalence of restenosis among patients with and without CC (CC: 13 (7.4%),no CC: 32 (4.6%),P = 0.125) (Fig. 4A).However, a significant difference in restenosis was observed among Wilkins score categories (P = 0.016).Patients in Wilkins group III had the highest occurrence rate of restenosis, followed by group II, and patients in group I had the lowest rate of restenosis (group III: 3 (7.1%), group II: 34 (6.8%), and group I: 8 (2.4%)) (Fig. 4B).
Binary logistic regression was used to examine the effects of Wilkins categories and CC on restenosis incidence at follow-up.While patients with CC showed higher odds of experiencing restenosis, this difference was not statistically significant (OR: 1.68, 95%CI: 0.861-3.27,P = 0.128).
Short-term and mid-term follow up for MR severity
The main findings of our study are: 1.The importance of commissural calcification in predicting significant MR after PBMV in short-term and mid-term follow-up.2. No statistically significant correlation between Wilkins score and significant MR following PBMV.3. Mitral valve restenosis occurs more in patients with higher Wilkins score.
It has been shown that MR following PBMV is a common finding.However, significant MR following PBMV is a determinant of worse outcomes and need for mitral valve replacement.The precise mitral valve characteristics that can predict the outcomes remain to be elucidated [17][18][19].
Traditionally, the Wilkins score is used for pre-procedural risk assessment of patients undergoing PBMV [20].However, the role of the Wilkins score in predicting clinical outcomes after PBMV is controversial.While several studies have shown Wilkins score to be a significant albeit relatively weak predictor of complications after PBMV [21][22][23], other studies have rejected any association between the mitral valve score and clinical outcomes and/or echocardiographic findings following PBMV [11,24].In our study, there was an incremental increase in the prevalence of significant MR in mid-term follow up of patients with higher Wilkins score, but it was not statistically significant which highlights that Wilkins score is not a strong predictor for occurrence of the significant MR following PBMV.
Previous studies have indicated that the optimal Wilkins score for PBMV is ≤ 8 [22].Our study showed that in patients with a Wilkins score 9-10, PBMV can be performed with acceptable results if there is no CC.Furthermore, the semi-quantitative nature of the scoring system and lacking commissural assessment limits Wilkins score comprehensive predictive capacity.Although the Wilkins score does not directly assess the commissures, our study showed significantly higher scores in patients with CC.The degenerative disease of the commissures probably parallels with leaflet and subvalvular disease, with heavy generalized leaflet calcification likely involving the commissures.
In our study, presence of CC (unicommissural or bicommissural) was associated with a significantly higher rate of significant MR regardless of the total Wilkins score.Previous studies have confirmed that the mechanism underlying the increase in valve area during PMV involves the splitting of one or both fused mitral commissures, highlighting the need for examination of the CC as a separate entity with significant impact on PBMV outcomes [11,12,25].
In the study by Saturia et al., suggesting a new scoring system focused on the CC, they found that CC is associated with higher rates of MR after PBMV in patients with a Wilkins score of 8 or less [21].Moreover, Fatkin et al. used TTE to demonstrate that commissural morphology resulted in a more accurate prediction of immediate outcomes compared to the Wilkins score in patients undergoing PBMV [11].
Notably, Anwar et al. used 3-Dimensional TTE for evaluating the structure of mitral valve to predict the outcomes of PBMV and reported that CC is related to a more severe MR in mid-term follow up, aligning with our findings [26].Although we used 2-D TTE, potentially less precise than 3-D TTE, our results align with the study by Cannan et al., which reported that CC assessed by 2-dimensional echocardiography was a better predictor of significant MR compared to the Wilkins score.Their study also linked CC to higher rates of post-procedural complications, including moderate or severe MR, further supporting our observations [15].
While the Wilkins score remains a valuable screening tool for patients referred for PBMV, our results, in conjunction with prior studies, show the potential of CC as a reliable predictor of significant MR after PBMV and a possible indication for MVR.
Mitral valve restenosis
Restenosis can develop after PBMV due to various factors such as disease progression, sub-optimal MVA after PBMV, and valve structure features like leaflets calcification and subvalvular calcification.However, mitral valve restenosis is not a common early complication following PBMV [27].In a study by Sriram et al., restenosis rates were as low as 10% at 4 years, 18% at 5 years and 39% at the end of 7 years [28].Similarly, in another study conducted by Wang et al., restenosis rate was 40% at the end of 6 years [29].
In our study, despite having a higher prevalence of restenosis, patients in Wilkins Group III did not have significantly higher odds of restenosis compared to Group I.However, patients in Group II had higher odds of restenosis than Group I. We speculate that the inconsistency in our results might be due to having fewer numbers of patients in Group III as well as short follow-up time after PBMV, resulting in low rates of restenosis at 24-months.Also, despite having higher rates of restenosis in patients with CC, CC was not a significant predictor of restenosis at 24-months.As we mentioned before this finding might be due to insufficient follow-up time after PBMV.Future studies with longer follow-up times are required to evaluate the predictiveness of CC on the incidence of restenosis after PBMV.
Conclusion
In our study, commissural calcification was associated with higher rates of significant MR at short-term and mid-term follow-up after PBMV, while Wilkins score was not a strong predictive of MR.A detailed assessment of commissural calcification can provide beneficial prognostic information in patients undergoing PBMV in addition to the current valve scoring such as Wilkins score.It is reasonable to consider MVR when there is unfavorable MV morphology [19].
Limitations
This study has a few limitations.First, this is a singlecenter retrospective study.However, the large number of cases with uniform approach in the center can compensate for some of the drawbacks.Second, the follow-up periods were relatively short, especially for the assessment of restenosis.Third, the CC was visually assessed however, this study was done by expert echocardiographers in a large referral center.
Fig. 2
Fig. 2 Wilkins score based on the presence or absence of CC.Box and whiskers represent the distribution of the Wilkins score in patients with and without CC.Dots represent the actual value of Wilkins score for each patient.CC: commissural calcification
Fig. 3
Fig. 3 Figure 3 A-D Occurrence of short-term MR based on presence or absence of CC(3 A) and Wilkins Score (3B), and mid-term MR based on presence or absence of CC (3 C) and Wilkins Score (3D)
Fig. 4
Fig. 4 4A and 4B Occurrence of mitral valve restenosis based on presence or absence of CC (4A) and Wikins score (4B)
Table 1
Baseline Characteristics and Echocardiographic Parameters of the Study CC: Commissural Calcification; MVA: Mitral valve area; MVAI: Mitral valve area index; sPAP: Systolic pulmonary artery pressure; LVEF: Left ventricular ejection fraction; LAP: Left atrial pressure | 2024-06-20T05:08:30.973Z | 2024-06-18T00:00:00.000 | {
"year": 2024,
"sha1": "fb47cebd92a5d8c4a5194b39acc1bfce9eee391f",
"oa_license": "CCBY",
"oa_url": null,
"oa_status": null,
"pdf_src": "PubMedCentral",
"pdf_hash": "fb47cebd92a5d8c4a5194b39acc1bfce9eee391f",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
230612367 | pes2o/s2orc | v3-fos-license | A Splint-to-CT Data Registration Strategy for Maxillary Navigation Surgery
Computer-assisted navigation plays an important role in modern craniomaxillofacial surgery. Although headpins and skull posts are widely used for the fixation of the reference frame, they require the use of invasive procedures. Headbands are easily displaced intraoperatively, thus reducing the accuracy of the surgical outcome. This study reported the utility of a novel splint integrated with a reference frame and registration markers for maxillary navigation surgery. A maxillary splint with a 10 cm resin handle was fabricated before surgery, to fix the reference frame to the splint. The splint was set after the incorporation of fiducial gutta-percha markers into both the splint and resin handle for marker-based pair-point registration. A computed tomography (CT) scan was acquired for preoperative CT-based planning. A marker-based pair-point registration procedure can be completed easily and noninvasively using this custom-made integrated splint, and maxillary navigation surgery can be performed with high accuracy. This method also provides maximum convenience for the surgeon, as the splint does not require reregistration, and can be removed temporarily when required. The splint-to-CT data registration strategy has potential applicability not only for maxillary surgery but also for otolaryngologic surgery, neurosurgery, and surgical repair after craniofacial trauma.
Several methods have been developed to improve registration accuracy during craniomaxillofacial surgery. The combination of headpins, skull posts, or headbands with positioning screws implanted into the maxillary alveolar bone has been recently used for marker-based registration [3,4,8]. However, the use of headpins and skull posts is invasive; headbands, on the other hand, can be easily displaced intraoperatively, thus reducing the accuracy of registration.
The splint was integrated with the reference frame and registration markers to overcome the issues associated with maxillary navigation surgery. This system is noninvasive, and more accurate than that involving the use of a headband, since the reference frame can be mounted more rigidly and closer to the surgical field. The aim of the present case report was to evaluate the feasibility of a novel custom-made integrated splint for maxillary navigation surgery.
Case Report
A 70-year-old Japanese man with a medical history of fibrous dysplasia of the craniofacial bones was referred to the oral and maxillofacial surgery department of a general hospital in March 2019 with chief complaints of swelling and pain in the left buccal region. The patient provided informed consent prior to treatment commencement, and the study protocol was approved by the appropriate institutional ethics committee.
The patient had hypertension and was taking antihypertensive drugs. Clinical examination at the initial visit revealed facial asymmetry, mild swelling, and redness and tenderness over the left buccal region. No paresthesia was documented over the left buccal region. The left maxillary first molar was tender to percussion, and mild swelling and redness were observed on the buccal gingiva.
The panoramic radiograph revealed a "ground-glass" appearance in the left maxillary sinus and left mandibular body and ramus; a periapical radiolucency was observed with the left maxillary first premolar and first molar ( Figure 1). Computed tomography (CT) revealed severe bone hyperplasia and "ground-glass" appearance with the left maxilla, mandible, and sphenoid bone, along with an osteolytic lesion extending from the root apex of the left maxillary first molar to the inferior aspect of the left infraorbital foramen and canal in the "ground-glass" lesion ( Figure 2). Bone scintigraphy revealed the accumulation of technetium medronic acid in the right skull base, cervical spine, left maxilla, and left mandible ( Figure 3).
The radiograph shows "ground-glass" appearance at the left maxillary sinus, mandibular body, and ramus. A periapical radiolucency is seen at the left maxillary first premolar and first molar.
Ampicillin/sulbactam was prescribed for 7 days. A clinical diagnosis of left maxillary infection caused by a periapical lesion of the left maxillary first molar in a fibrous dysplasia lesion was established. The left maxillary first premolar and first molar were extracted, and the osteolytic lesion was curetted (while preserving the infraorbital neurovascular bundle) using a navigation system and novel navigation splint (described below) in October 2019. The patient declined correction for facial asymmetry via bone reduction. Fibrous dysplasia with granulation tissue was observed on pathological examination of the surgical specimen ( Figure 4). Ampicillin/sulbactam was prescribed for 7 days postoperatively, and the patient was discharged from the hospital 4 days after surgery. The patient's 1-year postoperative course was uneventful.
Navigation Technique
A maxillary splint was fabricated, which extended from the right maxillary first molar to the left maxillary second molar, prior to surgery. The splint was composed of 3 mm thick Erkoloc-Pro plates (Erkodent, Pfalzgrafenweiler, Germany), Figure 2: Computed tomography (CT) image obtained at the initial visit. CT shows bone hyperplasia and "ground-glass" appearance at the left maxilla, mandible, and sphenoidal bone, along with an osteolytic lesion extending from the root apex of the left maxillary first molar to the inferior aspect of the left infraorbital foramen and canal in the "ground-glass" lesion. Axial (a), coronal (b), and sagittal (c) CT views. 2 Case Reports in Dentistry and the resin extended to a point 1 cm below the incisal edges of the anterior teeth ( Figure 5). A 10 cm resin handle with a connector, which was a reference star connector obtained from a reference headband (Brainlab AG, Feldkirchen, Germany), was used to fix the reference frame to the splint on the right side of the canine. The splint was set after the incorporation of 11 fiducial gutta-percha markers (each with a diameter of 1.5 mm) into the splint and resin handle for marker-based pair-point registration, and a CT scan was acquired [10]. Preoperative CT-based planning (iPlan CMF, Brainlab AG, Feldkirchen, Germany) entailed the determination of the region requiring curettage, as well as the registration and numbering of the 11 fiducial gutta-percha markers for the marker-based pair-point registration. The planning data were subsequently transferred to The Kick navigation system (Brainlab AG, Feldkirchen, Germany). The reference frame was fixed to the connector on the splint, and markerbased pair-point registration was performed, after the induction of general anesthesia and nasotracheal intubation. The practical procedure of the marker-based pair-point registration was as follows (splint-to-CT data registration): the numbered fiducial gutta-percha markers were indicated and registered point-by-point in succession within the optical tracking range of the navigation system, using the pointer ( Figure 6). The splint was set on the maxillary dentition, and the presence of registration errors were determined with the pointer to assess the median positions of the maxillary central incisors and mesiobuccal line angles of the maxillary first molars on both sides. It took approximately 3 min to complete and confirm the marker-based pair-point registration before surgery; the mean fiducial registration error (FRE) was 0:68 ± 0:30 mm (Figure 7). The splint was temporarily removed to facilitate extraction of the left maxillary premolar and first molar. A threesided flap was subsequently elevated, and the splint was reinserted ( Figure 8). The positions of the infraorbital foramen and canal were accurately identified in the surgical field using the pointer, which facilitated the removal of the osteolytic lesion ( Figure 9). Both the splint and reference frame were stable, as the splint was rigidly fixed to the maxillary dentition, which prevented the risk of a navigation system error. Simple sutures were placed at the vertical incision, and Case Reports in Dentistry horizontal mattress sutures were placed at the gingival margin after the splint was removed (Figure 8).
Discussion
The findings derived from this case have two crucial implications. First, the marker-based pair-point registration procedure can be completed easily and noninvasively using a maxillary splint integrated with the reference frame and registration markers, and the use of such a splint allows maxillary navigation surgery to be performed with high accuracy. Fibrous dysplasia is a common benign skeletal lesion that may involve one bone (monostotic) or multiple bones (polyostotic), which can occur throughout the skeleton with a predilection for the long bones, ribs, and craniofacial bones [11,12]. However, most lesions can be treated with clinical observation and patient education [11,12]. In the present case, an infection originated from an apical lesion and spread to the region affected by fibrous dysplasia, which necessitated curettage of the infected lesion and extraction of the causative teeth. Furthermore, a method facilitating safe removal of the lesions was needed, owing to the severe hypertrophy and deformity of the left maxillary bone caused by fibrous dysplasia, and the contact of the osteolytic lesion with the infraorbital foramen and canal.
Headbands, skull posts, or headpins are generally used in craniomaxillofacial navigation surgery [3,4,6,8,13]. The reference frame is first fixed to the patient's head with a headband, skull post, or headpin, which is followed by markerfree or marker-based registration. However, the use of headpins and skull posts is invasive, and a headband can easily be displaced intraoperatively. Furthermore, it is known that surgical precision decreases linearly with the distance from the reference markers [5,7,8]. The splint was integrated with the reference frame and registration markers to overcome the above-mentioned issues associated with maxillary navigation surgery. This system is noninvasive and more accurate than the use of a headband, since the reference frame can be mounted more rigidly and closer to the surgical field. Furthermore, as 11 fiducial gutta-percha markers were incorporated within the splint in asymmetric positions, marker-based pair-point registration could be performed in the patient's
5
Case Reports in Dentistry absence. The resin handle was also used for marker incorporation as the maxillary splint did not have adequate space to accommodate each of the 11 fiducial gutta-percha markers. The low mean FRE for splint registration (0:68 ± 0:30 mm) suggested that a widely spaced three-dimensional arrangement of the 11 fiducial gutta-percha markers contributed to the increase in the accuracy of the splint-to-CT data registration. This splint-to-CT data registration strategy is extremely precise and rapid; approximately 3 min were required to complete and confirm the marker-based pair-point registration before surgery, and there was no evidence of instability in the splint and reference frame that could have induced a navigation system error. It was also possible to preserve the infraorbital neurovascular bundle with this approach.
Second, this novel method provides maximum convenience for the surgeon, as the splint does not require reregistration, and can be temporarily removed when required. During craniomaxillofacial surgery, surgeons often change the patient's head position slightly to secure the surgical field; this is not possible when using a headpin. Any change in head positioning may result in the displacement of the headband, thus necessitating reregistration [14]. In contrast, the use of a splint allows changes in head positioning without displacing the reference frame; furthermore, the splint can be temporarily removed when it obstructs the surgical field, and reregistration is not required.
In the present case, the patient declined to undergo bone reduction for the correction of facial asymmetry caused by fibrous dysplasia. However, it would have been technically possible to perform this procedure by using the splint and navigation system to capture the mirror images of the opposite, unaffected side of the facial skeleton.
A limitation of oral splints is that their use is not feasible in patients who are edentulous or have few remaining teeth. Nevertheless, provided that a sufficient number of teeth are present to ensure the stable positioning of the splint, the results of this case report suggest that the splint-to-CT data registration strategy may be used not only in maxillary surgery, but also in otolaryngologic surgery, neurosurgery, and surgical repair after craniofacial trauma (especially zygomatic fracture and orbital floor fracture). Further studies with larger sample sizes are required to compare the effectiveness and reliability of this technique with conventional approaches.
Conclusion
A marker-based pair-point registration procedure can be completed easily and noninvasively, and maxillary navigation surgery can be performed with high accuracy by incorporating a custom-made integrated splint. This method also provides maximum convenience for the surgeon, as the splint does not require reregistration, and can be temporarily removed when required.
Ethical Approval
This study was approved by Kobe City Medical Center General Hospital IRB (No. zn200618), and all participants signed an informed consent agreement.
Consent
The patient provided written informed consent for the use of their photographs in this publication.
Conflicts of Interest
The authors have no conflict of interest to declare. | 2020-12-10T09:04:45.637Z | 2020-12-04T00:00:00.000 | {
"year": 2020,
"sha1": "5365fac03567a240ead8d537f51cebeb56b59a04",
"oa_license": "CCBY",
"oa_url": "https://downloads.hindawi.com/journals/crid/2020/8871148.pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "b309ede8eb73e9f7336c43df79b54c894f22082b",
"s2fieldsofstudy": [
"Medicine",
"Engineering"
],
"extfieldsofstudy": [
"Medicine"
]
} |
51940970 | pes2o/s2orc | v3-fos-license | Higgs Pair Production as a Signal of Enhanced Yukawa Couplings
We present a non-trivial correlation between the enhancement of the Higgs-fermion couplings and the Higgs pair production cross section in two Higgs doublet models with a flavour symmetry. This symmetry suppresses flavour-changing neutral couplings of the Higgs boson and allows for a partial explanation of the hierarchy in the Yukawa sector. After taking into account the constraints from electroweak precision measurements, Higgs coupling strength measurements, and unitarity and perturbativity bounds, we identify an interesting region of parameter space leading to enhanced Yukawa couplings as well as enhanced di-Higgs gluon fusion production at the LHC reach. This effect is visible in both the resonant and non-resonant contributions to the Higgs pair production cross section. We encourage dedicated searches based on differential distributions as a novel way to indirectly probe enhanced Higgs couplings to light fermions.
Introduction.Probing the Higgs couplings to the first and second generation fermions is one of the main objectives of the Higgs program at the LHC.New physics could induce very large deviations from the Standard Model (SM) predictions, by changing the way the Higgs couples to light fermions through higher dimensional operators where φ denotes the Higgs doublet and f is an arbitrary fermion.From (1) follows for the fermion mass matrix while the couplings to the SM Higgs are given by Enhancements of Higgs couplings to light fermions can be induced if the last term in (3) becomes sizable with respect to m f /v.This requires fine-tuning between y f and y f in order to recover the observed fermion masses.In addition, (3) in general induces sizable flavour-changing neutral currents (FCNCs) mediated by the Higgs, due to the misalignment between fermion masses and Higgs couplings, which requires additional fine-tuning to fulfill the bounds from flavour observables [1].An alignment of the couplings y f and y f at a high scale is not stable under renormalization group evolution, because the SM Yukawa coupling and the dimension six operator in (1) run differently [2][3][4].
In two Higgs doublet models (2HDMs) with a flavour symmetry, the light fermion masses can be explained through higher order operators, avoiding the need for the very small Yukawa couplings in the SM.These higher order operators introduce enhanced diagonal couplings between the Higgs and the SM fermions.Moreover, the structure of Higgs couplings to fermions are close to minimal flavour violating, leading to suppressed FCNCs [5,6].
In this letter we argue that in these models there exists a strong correlation between maximally enhanced Higgs couplings to fermions and a enhanced Higgs pair production that can be probed at the LHC.
Several strategies to test light fermion Yukawa couplings have been proposed, which are sensitive to enhanced couplings present in the class of models discussed in this letter.In the case of muon and electron Yukawa couplings, direct measurements of h → µ + µ − and h → e + e − yield the strongest constraints [7][8][9] where κ f = g hf f /g SM hf f .Direct measurements of the Higgs couplings to light colored fermions are much more challenging.The strongest (yet indirect) bounds follow from a combined fit to Higgs coupling strength measurements, allowing only one Yukawa to deviate at a time [10] surements of h → cc or associated production of pp → hc + hc, which strongly depend on the c− and b−tagging efficiencies [11][12][13].Exclusive, radiative Higgs decays h → J/ψ(Υ)γ provide an alternative way to test charm (and bottom) Yukawas and notably also to access their sign [14][15][16][17].Measurements of the total width of the Higgs offer another handle on individual Yukawa couplings [12], as well as measurements of p T −distributions in pp → h and pp → hj [18,19].A novel strategy based on measuring the charge asymmetry in W ± h has been proposed [10].If Higgs couplings to proton valence quarks and electrons are simultaneously enhanced, even frequencies of atomic clocks could be modified [20].
Formalism.In 2HDMs, the SM singlet operator φ 1 φ 2 can carry a flavour charge, such that for a given flavour the SM Yukawa coupling is replaced by a higher order operator in which Λ is the suppression scale, φ i is either φ 1 or φ 2 and the integer n f depends on the flavor charge assigned to f L φ i f R and φ 1 φ 2 .As a consequence, the corresponding fermion masses are given by with the vacuum expectation values φ 1,2 = v 1,2 and For the right choice of flavor charges, the hierarchy of SM fermion masses and mixing angles can be explained by higher order operators [5,6].In contrast to the ansatz (1), lower dimensional operators can be forbidden by these flavor charges.In the following, we will illustrate our result based on the Lagrangian which reduces to a 2HDM of type I in the limit n u , n d , n → 0. This expression can be readily extended to other types of 2HDMs [6] and the discussion in the remainder of the paper holds independent of this choice.The Higgs sector contains two neutral scalar mass eigenstates h, H, one pseudoscalar A and one charged scalar H ± and we identify the lighter scalar mass eigenstate h with the 125 GeV resonance observed at the LHC.The couplings between the scalars and the electroweak gauge bosons are fixed as in any 2HDM to Z, and we use the notation s x = sin(x), c x = cos(x) and t x = tan(x).The couplings between the scalars ϕ = h, H and SM fermions f Li,Ri = P L,R f i in the mass eigenbasis read with a flavor index i, such that u i = u, c, t, d i = d, s, b and i = e, µ, τ .This induces flavor-diagonal couplings and flavor off-diagonal couplings The flavor universal functions in ( 10) and ( 11) are given by and Flavor off-diagonal couplings between the neutral scalars and SM fermions are induced in (11) through the matrices in flavor space A and B, whose entries are proportional to the flavor charges of the corresponding fermions that define the coefficients in (8).In general, there are flavor charges of the fermion singlets, a fi , doublets, a Qi and a Li , as well as those of the Higgs doublets a 1 and a 2 .We set the flavor charge of φ 1 φ 2 to a 1 + a 2 = 1 by fixing a 2 = 1 and a 1 = 0, such that While these exponents depend on the relative charge assignments for the two Higgs doublets, the structure of the matrices A and B is independent of this choice.If all flavor charges for a given type of fermions are equal, the off-diagonal elements of these matrices vanish.Otherwise, for couplings of the neutral scalars to up-type quarks B = U with off-diagonal elements and the same expressions hold for A = Q with a ui → a Qi .For couplings between the neutral scalars and downtype quarks A = Q and B = D, where the elements of D are given by ( 16) for a ui → a di .Finally, flavor offdiagonal couplings between charged leptons and neutral scalars are given by (11) with A = C with the elements ( 16) for a ui → a Li , and B = E with the elements ( 16) for a ui → a i .These structures lead to flavor-FCNCs, which are chirally suppressed and proportional to powers of the ratio ε.The flavor symmetry strongly constrains the Higgs potential The seven independent parameters µ 2 1 , µ 2 2 , µ 2 3 and λ 1 , λ 2 , λ 3 , λ 4 can be exchanged for the vacuum expectation values v 1 and v 2 , the physical masses m h , M H , M A , M H ± and the mixing angle c β−α .The coupling between the heavy scalar H and two SM Higgs scalars h, as well as the triple Higgs coupling can be expressed as [32,33] g Hhh = (18) Higgs Pair Production.
The main finding of our paper is that the parameter space for which the diagonal couplings of the SM Higgs to fermions (10) are maximally enhanced is directly correlated with an enhancement of the trilinear couplings (18) and (19).This parameter space can be identified with the region for which f h (α, β) 1, outside of the decoupling limit c β−α = 0.For maximally enhanced couplings, the mass of the heavy scalar H cannot be arbitrarily large and resonant Higgs pair production is a signal of this model.The correlation between the enhancement of the Higgs couplings to SM fermions κ h f and Br(H → hh) is illustrated for M H = M A = M H ± = 500 GeV in Fig. 1.The color coding shows the dependence of Br(H → hh) on c β−α and t β , and the dashed contours correspond to constant |κ h f | for n f = 1.This correlation is independent of the factor n f while n f > 1 leads to larger enhancement factors, and holds throughout the parameter space, apart from the limits c β−α ≈ 0 and c β−α ≈ ±1.The latter case is strongly disfavoured by SM Higgs coupling strength measurements, and the correlation breaks down due to the factor s β−α in front of f h (α, β) in ( 18).The limit c β−α = 0 is usually associated with the decoupling of the heavy scalar states, for which g hhh = −3m 2 h /v takes on its SM value and g Hhh = 0, while the enhancement of Higgs couplings to fermions is fixed to κ h fi = 2n fi + 1.The decoupling limit corresponds to a large value of the pseudoscalar mass M A v, which is related to the spurion µ 3 ∝ M A that softly breaks the flavour symmetry assumed in (8).At one-loop, one expects this spurion to break the structure of the matrices ( 16), inducing FCNCs proportional to µ 2 3 /(4πΛ) 2 .Therefore, the relations we present only hold if additional scalars are present below the TeV scale, for which the parameter space c β−α = 0 is allowed.For larger values of t β there is a suppression of gluonfusion production, σ(gg , where κ h t ≈ 1, that partially cancels the enhancement of Br(H → hh).However, since σ(gg → h) ∝ (κ h t ) 2 , the cross section σ(gg → h → hh) is unsuppressed for large values of t β resulting in a continuous correlation between κ h f and σ(gg → hh) due to the non-trivial interplay between the resonant and non-resonant Higgs pair production processes.We illustrate this result in the left panel of Fig. 2, in which the dotted (dashed) lines correspond to the contribution from resonant (non-resonant) Higgs pair production in gluon fusion.The solid line is the full σ(gg → hh) in the 2HDM in units of the SM value.We set M H ± = M H = 550 GeV, M A = 450 GeV, and show values of c β−α = −0.45(−0.4) in green (blue) lines.Higgs coupling strengths measurements and electroweak precision measurements constrain large values of c β−α , but do not exclude the values considered here for a Yukawa sector of a 2HDM of type I.In order to produce the signal, we use our own C++ implementation of the NLO QCD cross section for di-Higgs production in the presence of a scalar singlet [28], in the approximation where the exact m t -dependent form factors are inserted into the m t → ∞ NLO calculation [29].Since the pseudoscalar and the charged Higgs do not contribute, these results can be easily applied here.We use the CT14NLO PDF from LHAPDF6 [30] as well as the C++ library QCDLoop [31] to evaluate the corresponding one-loop integrals, neglecting small corrections from quark initial states.Solid lines show the NLO results, while the solid shaded lines mark the values of κ f excluded by perturbativity and unitarity constraints [21].The dotted (dashed) lines show the LO ratios for the resonant (non-resonant) contribution.However, to a very good approximation the NLO corrections factorize and drop out of the ratio.
For the values of κ h f considered, σ(pp → hh) never exceeds the experimental bound on the non-resonant Higgs pair production cross section [34].The values of κ h f in Fig. 2 follow from fixing n f = 1 and values of O (10) and larger are obtained for n f > 1.Note that the correlation between σ(pp → hh) and κ h f is stronger for vector boson fusion production, because there is no suppression of σ(pp → H) for t β > 1 and σ(qq → qqH) ∝ s 2 β−α .In the right panel of Fig. 2, the invariant mass distribution for the different contributions to the signal with c β−α = −0.45 are shown for three values of κ h f and √ s = 13 TeV.As a consequence of the enhancement of Higgs-fermion couplings, both non-resonant and resonant contributions are enhanced.The relevance of the dσ/dm hh distribution for both resonant [22] and non-resonant contributions [23] to the Higgs pair production cross-section has long been emphasized [24][25][26].Searches for resonant di-Higgs production are sensitive to a peak in the spectrum, which roughly excludes heavy scalar masses M H 500 GeV, independent of f h (α, β) [27].For larger M H and sizable κ h f , the interference between the different contributions turns the broad resonance peak into a shoulder in the dσ/dm hh distribution for the total cross section, as shown by the blue line in the right panel of Fig. 2. Whether current experimental resonance searches can resolve such a structure strongly depends on the shape of the invariant mass distribution [36].
We encourage a dedicated analysis considering the corresponding dσ/dm hh templates to maximize the sensitivity to features in the di-Higgs invariant mass distribution from the simultaneous enhancement of g hhh , g Hhh and κ f h .
An Explicit Example.
We now consider a concrete example for which the flavour charges of down-type quarks and leptons vanish n i = n di = 0 ∀ i, whereas the up quarks carry charges n t = 0, n c = 1, n u = 3 and we choose all charges of the SU (2) L fermion doublets to be zero.As a consequence, the top coupling to the SM Higgs h is unchanged from its value in the 2HDM of type I, while charm and up-quark couplings vary with t β and c β−α according to (10).This leads to flavour-changing couplings of the SM Higgs to up-type quarks suppressed by powers of the ratio ε, In the up-sector, the strongest constraints on FCNCs arise from D− D mixing.Due to the structure of (11), the leading contribution to the Wilson coefficients entering D − D mixing are chirally suppressed and proportional to U 2 12 = ε 4 .Assuming order one dimensionless coefficients, the experimental limit leads to the constraint [35] Im where the less relevant contributions from the heavy scalars have been neglected.For the maximal values of f h (α, β) ≈ 10, this yields ε 1/55.This example would lead to a Higgs pair production cross section of σ(pp → hh) ≈ 50 × σ SM (pp → hh) with enhancements of the Higgs couplings to up-quarks of κ h u = 10.2 and to charm-quarks of κ h c = 4, respectively.In principle, similar models can be build with flavor charged leptons and down-type quarks.The simultaneous enhancement of κ τ or κ b and stronger flavor constraints lead to a more constrained parameter space for such models.
Conclusions.
We report a non-trivial correlation between an enhancement of Higgs couplings to light fermions and enhanced resonant and non-resonant contributions to the Higgs pair production cross section.Such a correlation appears naturally in a class of models in which Higgs-mediated flavour changing currents are suppressed by a flavour symmetry.We show that even after imposing perturbativity and unitarity bounds as well as constraints from Higgs couplings strength measurements, the parameter space allowing for maximally enhanced Higgs-fermion couplings entails a Higgs pair production cross section exceeding the SM prediction by more than an order of magnitude.Present searches partly probe this interesting correlation, but dedicated LHC studies are required to ultimately explore this idea and indirectly constrain signals of new physics modifying light fermion Yukawa couplings.
FIG. 1 :
FIG. 1: The color coding shows the dependence of Br(H → hh) on c β−α and t β for M H = M H ± 550 GeV, M A = 450 GeV.The dashed contours correspond to constant |κ h f | for n f = 1.
FIG. 2 :
FIG. 2: Left: Cross section for Higgs pair production in units of the SM prediction as a function of κ h f for c β−α = −0.45(−0.4) and M H = M H ± = 550 GeV, M A = 450 GeV in blue (green) at √ s = 13 TeV.Right: Invariant mass distribution for the different contributions to the signal with c β−α = −0.45 and κ h f = 5 (blue), κ h f = 4 (green) and κ h f = 3 (red) at √ s = 13 TeV, respectively. | 2018-08-14T20:08:22.934Z | 2017-12-31T00:00:00.000 | {
"year": 2017,
"sha1": "57227e3d02b6079ae3789dd4ac670f98d3354861",
"oa_license": "CCBY",
"oa_url": "http://link.aps.org/pdf/10.1103/PhysRevLett.121.021801",
"oa_status": "HYBRID",
"pdf_src": "Arxiv",
"pdf_hash": "7726cf54ef8884aa3e0f9a8465f1d9d94b0f8257",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics",
"Medicine"
]
} |
209596378 | pes2o/s2orc | v3-fos-license | Exploration of indigenous rhizobacteria: in search for their potential as plant growth promoting bacteria at two potato producing areas in West Sumatra
An experiment aiming at exploring and identifying indigenous rhizobacteria from Nagari Alahan Panjang and Nagari Batagak, two potato growing field at the Province of West Sumatra, has been conducted from July to October 2018. Soil was collected from the potato growing areas at Municipality Solok and Municipality Agam. One hundred grams of soil was collected from 25 spots of land at each area. Ten gram of soil was added to 9 mL of sterile distilled water prior to thorough mix in a vortex. The suspension was then subject to series of dilution to get 10-5 or 10-6 solution. Then, 0.1 mL of the final solution was poured into a test tube containing liquid NA medium and mixed thoroughly. The mixture was incubated for 48 hours at ambient temperature. Bacterial colonies went through series of re-culture until pure isolate was obtained and were observed for their morphological and physiological characters. Based on hypersensitivity reaction on leaves of Mirabilis jalapa, seventeen isolates found at Municipality Solok and 49 isolates were identified from Municipality Agam. These isolates did not cause leaf-tissue damage of Mirabilis jalapa. Different types and characters of the rhizobacteria is a broad range of biodiversity which will be potential for further screening for bio assays againts major weeds in potato cultivation and may be used to open the windows to the natural immunity and wellness through induced resistant to major weeds.
Introduction
A major challenge in agricultural production system is continuously providing food for an evergrowing world population. Increasing food production cannot merely rely on intensive agricultural practices on arable land. This arable land constantly decreases around the world due to demand for other purposes such as housing, industrial expansion, and estate plantation. For decades, farmers rely on synthetic fertiliser and pesticides as well irrigation to keep food production which has been affected by the environmental factors, biotic as well as abiotic. Therefore, [1] proposed that the future world food production will need next-generation crop production systems and at the same time reducing dependence on synthetic fertiliser and pesticides.
Competition in land use for housing and a constant loss of arable land due to salinity and postmining abandoned land will continuously reduce productive land for agriculture [2]. Furthermore, land conversion from its natural ecosystem to agricultural field results in serious impact to the environment such as global carbon cycle, global hydrological cycle, disruption in bio-diversities, and soil health and property [2; 3; 4; 5]. However, efforts should be sought to optimise agricultural practices to meet world demand for food while improving ecological balances. As for other crops, potato interacts with unwanted organisms that exist in its vicinity such as pest, diseases, and weeds. Potato yield loss for weed-crop interaction have been reported. Reference [6] reported that weed control at rapid vegetative growth stage of potato (3 weeks after emergence) increased yield around 30-50% compared to that of non-controlled weeds. Other reports stated that weed reduced potato yield up to 45-65%; although the reduction varied between varieties and places [7; 8]. The most effective and efficient mean of weed control in potato production system is through improvement of agricultural practices which may reduce the green colour on potato tubers due to exposure of sun light [6]. However, this practice may reduce yield due to loss of lateral roots and soil compaction. Therefore, many farmers rely on the application of herbicides to control weeds in potato. For instance, farmers applied herbicide linuron (up to 960 kg/ha active ingredient) that sometimes was not enought to control weeds. Farmers, sometimes, mixed herbicides such as linuron and metribuzin, pendimethalin and clomazone, metribuzin and flufenacet [9] to increase the efficacy. Intensive use of herbicide will result in environmental problems such as residue in soil and development of weed resistant to herbicide. Alternative means to herbicide application should be sought including beneficial rhizobacteria.
Various microorganisms live surrounding roots of plants and many of them had been reported for their benefit to support plant growth as well as to improve plant adaptation to unpleasant environment such as low nutrient [10]. Plant roots produce exudates needed by the rhizobacteria as sources of nutrition [11]. This may play role as antagonistic agents and suppress soil born pathogen as well as enriching the composition and diversity of indigenous microorganisms [12; 13]. Rhizobacteria are often found at rhizosphere and mostly species-or location-specific. Some rhizobacteria are repoted for their ability to increase plant growth and wellness through the absorption of Fe and Se and increased the chlorophyll content as much as 28% in Arabidopsis thaliana [14]; increase tolerance to drought in maize [15], induce tolerance to salinity stress [16], and induce plant immune system through interference in the function of auxin [17]. These will open the window for plant wellness improvement in attempts to increase plat competition to weeds. Research reported here is aimed at exploring and identifying indigenous rhizobacteria from two potato producing areas in the Province of Sumatera Barat and will be used for improvement of potato wellness againts weed competition. Rhizobacteria that are associated with plant roots can be easily found in all agro-ecosystem [18].
Material and Method
Soil was collected from two potato producing areas in West Sumatra, namely Nagari Alahan Panjang at Municipality of Solok and Nagari Batagak at Municipality of Agam. The isolation and characterisation of rhizobacteria was conducted at the Laboratory of Microbiology of Faculty of Agriculture, Andalas University Padang from July to October 2018. About one hundred grams of soil was collected form each spot from 25 spots of the potato grown land at each municipality.
Debris such as roots and twigs were removed from the soil. The 10 g of soil was added to 10 mL of sterile distilled water prior to proper mixing for 3-5 minutes in a vortex. This suspension was then marked as 10-1 solution. One mL of the (10-1) soil suspension was transferred to test tube with 9 mL of sterile distilled water then being mixed at a vortex and was labeled as 10-2 solution. The dilution process was repeated until 10-6 solution has been obtained. Then, 0.1 mL (100 µL) final soil suspension was poured into a test tube containing liquid NA medium prior to thoroughly mixed in a vortex. The mixture was then poured into a Petri dish, sealed, and incubated for 48 hours at ambient temperature.
Bacterial colony formed transparent zone of bacterial isolate. The isolate was then re-cultured onto fresh media using a loop ose. The re-culture was conducted several times until pure isolate was obtained. The pure isolates were then observed and identified for their morphological characters (color, shape, colony, colony surface, and shape of colony edge). The selected colony was then purified in streak plate using loop ose and was incubated for 48 hours at ambient temperature. A single colony of bacteria was then aseptically moved into a microtube containing 1 mL of sterile distilled water for preservartion purposes. The tubes were kept in a refrigerator for later use.
Gram test for bacteria was conducted to separate the rhizobactria either Gram-positive or Gramnegative. Another bacterial character tested was hiper-sensitive reaction (HR-pathogenesis) using [19] with some modification. The isolate is pathogenic when causing necrotic on the leave of Mirabilis jalapa two weeks after exposure. Data were not statistical analysed as this research was more on exploration and observation of the morphological and physiological characters of the rhizobacteria.
Results and Discussion
There are 18 isolate of indigenous rhizobacteria found in the soil of potato field at NagariAlahanPanjang, Municipality of Solok. One isolate demonstrated to be toxic to the target plant (Mirabilis jalapa). However, there are more isolate of rhizobacteria found in the soil collected from NagariBatagak, Municipality of Agam. Fourty nine of 53 isolates are non-pathogenic to the target plants. Morphological characters of rhizobacteria are presented in Tables 1 and 2. The number of colony of rhizobacteria from Municipality of Agam was nearly triple than that of Municipality Solok. This may resulted from intensive application of pesticide for horticultural crops in Solok (personal communication with some farmers and observation, 2017 and 2018) which would in turn reduce beneficial microorganisms in the soil.
Farmers rotate potato cropping system with shallot, tomato, and cabbage with various pesticides applied to the crops and land. This was intended to cut the pest and disease life cycle in the field. However, rotating crops from similar family group will enhance the life and environment suitable for the pest and diseases. Careful practices should then be taken into account when rotating plants. Higher number of isolates and colonies of rhizobacteria from Nagari Batagak at Municipality Agam may, to some extent, may be resulted from the increased awareness of farmers to reduce pesticides. Some farmers and farmers' group tried to apply bio-pesticides to protect their crops. They are in the process of shifting from conventional to organic farming in producing potatoes. Better environment for soil microorganisms including rhizobacteria would be achieved by less residue of pesticides and high organic matter [20] that provide organic compounds to support the diversity and activity of soil microbia [21]. This is in accordance with the finding of [22] who demonstrated that organic matter increased soil microbial biomass, activity, and diversity. The increase was observed within a 20-day period afterwards declined over a longer time.
Soil quality determines its capacity to function in an ecosystem to support biological productivity, to maintain the quality of environment, and to facilitate the health of animals and plants in that ecosystem [23], and is important for the survival of microorganisms [24]. Intensive uses of pesticide may result in imbalance of soil biology. It is common that the number and type of rhizobacteria will vary according to the distance from roots due to root exudates released and utilised as nutrient source for the rhizobacteria [11; 25]. Moreover, certain rhizobacteria will associate with certain crop depends on the compatibility of the two species. This will open the windows to the natural plant immunity and wellness through induced resistance [26].
The complex mechanism of plant and soil bacteria may lead to the production of plant growth regulators. The chemical substances range from secondary metabolites to enzymes [27]. They signal molecules to act as chemical communication leading to supporting the growth and development of the host plants [28]. Ref. [29] reported that rhizobacteria isolated from the roots of sweet potato produced IAA and solubilised phosphate. The solubilisation of inorganic phosphate was detected as an increase of phosphate content in the potato shoot systems. They recommend the utilisation of the isolates as potential bioinoculants to reduce dependency on chemical fertiliser. This practices support environmentally-friendly agricultural system. The research reported here found isolates of rhizobacterial colony in various sizes ranging from 0.1 to 6.3 cm. Shape, color, margin, and elevation divers within the isolated bacteria. Most of the isolate are white but some are filamentous. Some isolates with different morphological characters are presented in Figure 1 Various form and color of rhizobacteria was found from two potato producing areas. Some photos are presented in Figure 1 and 2.
Sixty six isolates of the total 71 isolates from both potato growing areas in West Sumatra did not cause any tissue damage on the leaves of four o'clock flower (Mirabilis jalapa). This is a good sign of bacteria and plants compatibility in nature. The rhizobacteria in association with potato plant roots, to some extent, would increase plant growth through various mechanisms including nutrient uptake and promoting the production of growth hormones. Reference [30] stated that hypersensitive is a defence mechanism develop in plant in response to microorganism infection. Infected cells often be characteristically observed through an increase in membrane permeability which lead to the death of host cells nearby the infected cells. Root exudation attracts soil microorganisms in such a complex mechanisms that might be interaction-specific and recognition processes depending upon the plant cultivars, environmental stress, and plant growth stage [31]. Therefore, different colonisation might be influenced by different root exudates and some sites are better colonised than others and resulted in spatial differences in colonisation [32]. This will support the exploration of indigenous rhizobactria to be utilised within the specific areas.
Conclusion
Various isolates of rhizobacteria was found at two potato producing areas in West Sumatra. Seventeen isolates were from Municipality Solok and 49 isolates were identified from Municipality Agam. Different types and characters of the rhizobacteria is a broad range of biodiversity which will be potential for further screening for bio assays againts major weeds in potato cultivation. | 2019-11-14T17:11:37.014Z | 2019-11-08T00:00:00.000 | {
"year": 2019,
"sha1": "d1714f4a7a9fd8bce35182ff96da7596ba50513f",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1755-1315/347/1/012022",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "bfe03ed1c73d4a4ed1ceb66d0b9dcedf9f4d92e3",
"s2fieldsofstudy": [
"Agricultural And Food Sciences"
],
"extfieldsofstudy": [
"Biology",
"Physics"
]
} |
52954568 | pes2o/s2orc | v3-fos-license | Visualisation and characterisation of mononuclear phagocytes in the chicken respiratory tract using CSF1R-transgenic chickens
The respiratory tract is a key organ for many avian pathogens as well as a major route for vaccination in the poultry industry. To improve immune responses after vaccination of chickens through increased uptake of vaccines and targeting to antigen presenting cells, a better understanding of the avian respiratory immune system is required. Transgenic MacReporter birds were used expressing a reporter gene (eGFP or mApple) under the control of the CSF1R promoter and enhancer in cells of the mononuclear phagocyte (MNP) lineage to visualize the ontogeny of the lymphoid tissue, macrophages and dendritic cells, in the trachea, lung and air sac of birds from embryonic day 18–63 weeks of age. Small aggregates of CSF1R-transgene+ cells start to form at the openings of the secondary bronchi at 1 week of age, indicative of the early development of the organised bronchus-associated lymphoid tissue. Immunohistochemical staining revealed subpopulations of MNPs in the lung, based on expression of CSF1R-transgene, CD11, TIM4, LAMP1, and MHC II. Specialised epithelial cells or M cells covering the bronchus-associated lymphoid tissue expressed CSF1R-transgene and type II pneumocytes expressed LAMP1 suggesting that these epithelial cells are phagocytic and transcytose antigen. Highly organised lymphoid tissue was seen in trachea from 4 weeks onwards. Throughout the air sacs at all ages, CSF1R-transgene+ cells were scattered and at later stages, CSF1R-transgene+ cells lined capillaries. These results will serve as a base for further functional characterization of macrophages and dendritic cells and their role in respiratory diseases and vaccine responses. Electronic supplementary material The online version of this article (10.1186/s13567-018-0598-7) contains supplementary material, which is available to authorized users.
Introduction
In the poultry industry, vaccines are frequently delivered via spray and aerosol, providing an economical, efficient and reliable method for immunisation of a large number of birds. The respiratory tract is a major portal of entry of pathogens that are of economic or zoonotic importance, such as avian influenza virus (AIV), infectious bronchitis virus (IBV), Newcastle Disease virus and avian pathogenic Escherichia coli (APEC). Surprisingly little is known about the ontogeny and function of the immune cells in the avian respiratory tract. Similar to mammals, nonspecific defense mechanisms, such as aerodynamic filtration, mucociliary clearance and antimicrobial substances are the first line of defense of the avian respiratory tract [1][2][3]. The avian respiratory system differs significantly from mammals. For example, birds have relatively rigid lungs, lack a diaphragm and the lung opens into 9 air sacs that function as bellows, with a unidirectional air flow pattern that affects the deposition of particles [4][5][6].
The lymphoid tissue in the avian lung can be divided into highly organized lymphoid structures such as the bronchus-associated lymphoid tissue (BALT) and diffusely distributed lymphoid cells in the lamina propria around secondary bronchi and in the parabronchi and the interparabronchial connective tissue. In chickens, BALT structures are confined to the openings of the most caudal secondary bronchi (SB) [7], the laterodorsal SB and the posterior SB [8] (Figure 1). Mature BALT structures consist of a distinctive layer of epithelial cells, the lymphoepithelium or follicle-associated epithelium (FAE) [9] overlying primary follicles with densely packed lymphocytes and secondary follicles with germinal centers (GCs) with a subepithelial dome consisting of CD4 + T cells, similar to Peyer's patches and other gut-associated lymphoid tissue (GALT) [10]. CD8 + T cells and heterophils are distributed throughout the BALT [1]. BALT appear as early as 2-3 weeks of age and its development is influenced by age and environmental stimuli [7,11].
To rapidly detect invading pathogens, the respiratory tract is lined with monocytes, macrophages and dendritic cells (DCs), collectively termed mononuclear phagocytes (MNP). Macrophages maintain a low inflammatory environment in the lung, as the infiltration of cells in response to inflammatory stimuli reduces the efficiency of gas exchange, but during infection they immediately induce a response in coordination with epithelial cells and DCs [reviewed in [12,13]. Polynuclear phagocytes, or granulocytes, and type II pneumocytes (also known as type II alveolar epithelial cells) also display phagocytic capacity, but these cells have distinct functions compared to MNPs. In the mammalian lung during steady state, distinct populations of alveolar macrophages, interstitial macrophages and DCs have been described based on multicolour flowcytometry, but during inflammation this distinction becomes less straight forward [reviewed in 12,13]. In contrast to mammals, the chicken lung contains very few free-residing macrophages comparable to alveolar macrophages [14]. However, many cells express macrophage and DC markers and are scattered throughout the interstitial tissue of the parabronchial wall and in close contact to the epithelium [15][16][17]. These cells are expected to play a similar role as mammalian alveolar macrophages and seem to be strategically located at the start of the gas-exchange area to clear the air of inhaled particles before it reaches the thin and vulnerable air capillaries [17].
The trachea in steady state is lined by ciliated pseudostratified columnar epithelium and has a relatively thin lamina propria with large numbers of mucous glands. The ciliated epithelium and mucus create a mucociliary escalator which forms an essential mechanism for entrapment and clearance of particulate material. The lamina propria leukocytes consist mainly of macrophages and an occasional CD4 + and CD8 + T cells and B cells [18]. After infection with respiratory pathogens, such as IBV, E. coli or AIV, the tracheal wall thickens and large infiltrations of macrophages and T cells have been described [19][20][21].
Little is known about the immune cells that reside within the air sacs, partly owing to their fragility and the low number of cells that can be extracted for ex vivo analysis [19,22]. The respiratory surface of the air sacs is lined by a simple epithelium, either flat or ciliated, on a basement membrane supported by a thin layer of connective tissue [22,23] where scattered solitary phagocytes have been reported [24]. The lamina propria of the air sacs contains small capillaries [25], arterioles, venules and lymph vessels [23]. Small lymphoid aggregates have occasionally been seen in the epithelium of the air sacs [11], and they increase in size and number upon infection [22].
The development of transgenic chickens, such as the CSF1R-transgenic or MacReporter chicken, enables visualization of the immune system as the reporter gene expression gives a unique macroscopic view of chicken lymphoid structures [26]. CSF1R-transgenic chickens express the gene for a fluorescent protein under the (3), and lateroventral (4) secondary bronchi are depicted (based on Makanya and Djonov [8]). From the secondary bronchi a large number of tertiary bronchi or parabronchi originate.
control of the CSF1R promoter and enhancer, in many hematopoietic cells of the monocyte lineage and at low levels in heterophils [26]. Transgene expression is not a definite marker for all MNP in the chicken as it was recently shown that the Kupffer cells in the liver of Mac-Reporter chickens lack transgene expression, but express CSF1R mRNA, indicating an alternative control of CSF1R expression [27]. In this study, we first investigated the ontogeny of the MNPs in the chicken respiratory tract from embryonic day (ED) 18 up to adult birds, and phenotypically characterized cell subpopulations in situ. These transgenic birds enable future studies on the interaction between MNPs and respiratory pathogens, antigen uptake and mucosal targeting of respiratory vaccines.
Chicken lines
Two transgenic chicken lines, collectively named Mac-Reporter chickens, were used in this study. MacRed birds express a modified red fluorescent protein (mApple) and MacGreen birds express an enhanced green fluorescent protein (eGFP) in cells of the mononuclear phagocyte lineage, under the control of the CSF1R promoter and enhancer [26]. Birds were bred and reared in floor pens at the National Avian Research Facility, The Roslin Institute, Edinburgh (UK). The animals used for developmental studies and whole mount imaging were vaccinated according to Additional file 1. Animals used for immunohistology were not vaccinated. The chickens were housed in groups and received food and water ad libitum. All birds were considered healthy by physical examination.
Tissue preparation
Tissues were collected immediately after death using standard avian necropsy techniques. Both lungs containing the primary bronchi were carefully removed from the coelomic cavity. One of the lungs was incised longitudinally on its ventral aspect, from the extrapulmonary primary bronchus into the intrapulmonary primary bronchus, in order to enable imaging the mucosa of the intrapulmonary primary bronchus and the openings to the SB. The other lung was incised transversally along the second costal sulci. The trachea, from the first tracheal rings to just above the carina, was dissected from the cervical area and two parallel longitudinal incisions were made to expose the tracheal mucosa. The cranial and caudal thoracic air sacs were collected intact and flat using the card-collection technique (Whatman ® , Grade 54) [28]. After dissection, tissues were kept in a Petri dish on ice, and whole mount fluorescence imaging was immediately performed. Whole mount tissues were evaluated for the presence of fluorescence using the AXIO Zoom.V16 fluorescence stereo zoom microscope (Carl Zeiss Zen pro 2012 software).
Immunofluorescence
Tissues were fixed in 4% paraformaldehyde for 3-6 h at 4 °C, gently rinsed in phosphate buffered saline (PBS), and overnight immersed in 30% sucrose at 4 °C. Tissue samples were snap frozen in liquid nitrogen and stored at −80 °C until use. Samples were sectioned at 7 µm onto Superfrost Plus slides, air dried in the dark at room temperature (RT) for at least 2-24 h before staining. Sections were blocked for 1 h using 2.5% Skim Milk Powder (SMP, Oxoid), 1% Triton X-100 (Sigma), and 10% normal goat serum or 10% normal horse serum (AMS Biotechnology) in PBS (blocking buffer). All sections were incubated with the primary antibody diluted in blocking buffer at 4 °C overnight and washed in PBS, followed by incubation with the appropriate secondary antibody diluted in blocking buffer and 10 μg/mL of 4′,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich) for 2 h at RT and washed in PBS. The primary and secondary antibodies used for cell-specific staining are listed in Table 1. Sections were mounted using SlowFade ® Gold or using ProLong ® Diamond Antifade Mountant (Life Technologies; ThermoFisher Scientific). In each study involving immunofluorescence staining, a tissue section from a chicken that does not express the CSF1R-transgene was included as an auto-fluorescence control, and one extra tissue section from the transgenic bird was included and incubated with an unrelated antibody of the same isotype and concentration as the primary antibody as an isotype control for aspecific staining and calibration of the microscopes (see Additional file 2). Images were captured using a Leica DMLB fluorescence microscope (Micro-Manager 1.4.18 software) and confocal images were obtained using an inverted confocal microscope (Zeiss LSM710) and captured using Zeiss Zen (Black) 2012 software.
Development and distribution of mononuclear phagocytes and germinal centers in the lung of CSF1R-transgenic birds
The distribution of CSF1R-transgene expressing cells was examined in lungs of MacReporter birds of different ages by whole mount microscopy. These birds were reared in floor pens on wood shavings and routinely vaccinated, albeit via non-respiratory routes (Additional file 1). Therefore, the lymphoid accumulations observed are in part in response to environmental particles and the vaccinations. Although BALT development follows a similar time course in specific pathogen-free and conventionally reared chickens, differences were observed with regard to germinal center (GC) development [7].
In the lung scattered individual CSF1R-transgene + cells were distributed over the entire mucosa of the intrapulmonary primary bronchi in all age groups ( (Figures 2A and B). Small aggregations of CSF1R-transgene + cells were first observed at 1 week of age ( Figure 2C). They are primarily located at the vicinity of openings of laterodorsal and mediodorsal SB. With increasing age these aggregates become larger and more numerous (Figures 2D-J). They are distributed at the vicinity of the laterodorsal, mediodorsal and medioventral openings. From 4 weeks of age onwards, GCs are observed within these lymphoid follicles, and they are characterized by tightly packed aggregates of CSF1Rtransgene + cells ( Figures 2E-H). GCs in older birds are larger with some lymphoid follicles containing several GCs ( Figures 2I and J), similar to the Peyer's patches of the small intestine [1]. Lymphoid follicles residing in the BALT region were occasionally observed in the vicinity of the third and fourth openings of the medioventral SB, and follicles were also present in the vicinity of the openings of the medioventral SB.
In this study, the lateroventral SB were not appreciated, as they do not open directly into the intrapulmonary primary bronchi and are unlikely to be identified using non-casting methods [8]. Moreover, organised lymphoid tissue with and without GCs were observed on the transversal sections of the SB within the parabronchi, indicating that they are not restricted to the openings of the medioventral SB as it enters the parabronchial tissue.
The animals used in this study appeared to have BALT structures 1 week post-hatch which were present in birds of up to 63 weeks of age, although the number of follicles seems reduced in older animals this could be attributed to the increase in lung size and follicles being dispersed over a larger surface area ( Figures 2K and L).
Immunofluorescent staining of cell populations in the bronchus-associated lymphoid tissue
To enable a more in-depth phenotypical analysis of MNP subpopulations, immunostaining of BALT and parabronchi was performed on non-vaccinated animals between 5 and 7 weeks of age. CSF1R-transgene + cells were abundantly present in BALT structures and were more tightly packed in the centers or light zone of GCs than in other parts of the lymphoid tissue ( Figures 3A and B). This is consistent with previous reports of monocytes/macrophages and DC type cells present in the BALT, which are CVI-ChNL-68.1 and CVI-ChNL-74.2-positive, and express MHC II [1,31]. chB6 + B cells were found to be tightly packed in GCs interspersed with CSF1Rtransgene + cells, while CD3 + T cells were observed to aggregate in poorly demarcated areas within the BALT, in the subepithelial areas and in between GCs (Additional file 2). The remaining areas of the BALT have a more disperse population of both B and T cells (Additional file 2). GCs have tightly packed CSF1R-transgene + cells (Figures 3A and B) known as follicular dendritic cells (FDC) that express CD11, but not MHC II (Figures 3C and D) and have previously shown to express CVI-ChNL-74.3 [31]. FDCs are characterised by trapping immune complexes via Fc receptors (Additional file 2). Their enzyme content is restricted to non-specific esterase and adenosine triphosphate [32,33]. zone or periphery of a GC where the rate of B cell proliferation is high [34]. We observed a large population of CD11 + cells to be present throughout the BALT. In GCs, CD11 + cells were tightly packed, and virtually all CSF1R-transgene + cells were CD11 + . While most of the CSF1R-transgene + cells are CD11 + , small subpopulations of CSF1R-transgene − CD11 + and occasional CSF1Rtransgene + CD11 − cells were observed outside the GCs ( Figure 3C).
In addition, GCs contain a cell population with remnants of nuclei in their cytoplasm, also known as starrysky or tingible body macrophages. In the chicken, these cells express TIM4, a receptor for phosphatidylserine, and engulf apoptotic cells. These TIM4 + cells lack transgene expression ( Figure 3E), but possibly express CSF1R transcript as previously described for TIM4 + CSF1R-transgene − cells in the chicken spleen [27].
The epithelium overlying the BALT consist of the follicle-associated epithelium (FAE) and interfollicular epithelium (IFE). Occasionally, in the FAE, CSF1R-transgene + cells were observed that were negative for common APC markers used in this study (Figures 3F and G). These cells likely represent airway M cells whose primary function is to transport mucosal particles to underlying APCs [35,36] and have recently been described in the avian bursa (Balic & Vervelde, unpublished observations). The FAE and IFE highly express F-actin at the apical side of the cell which likely plays a role in remodeling of the cells during endocytosis (Figures 3H and I) [37]. Differentiation between FAE and bronchial IFE was visualised with the α-D-galactose binding lectin, jacalin. Jacalin does not bind to the FAE, but binds to bronchial IFE cells ( Figure 3J), suggesting that the FAE covering the BALT, similar to the epithelium covering the gut-associated lymphoid tissue or Peyer's patches, differs in detail in relation to their function.
Staining of lysosome-associated membrane glycoprotein-1 (LAMP1/CD107a) was cytoplasmic and granular and correlated with the location of lysosomes, organelles that degrade intracellular material endocytosed by the cell. Scattered LAMP1 + cells were found to be distributed within the BALT, and non-GC areas (Figures 3K and L). Almost all LAMP1 + cells express the CSF1R-transgene, with occasional cells being LAMP1 + CSF1R-transgene − (Figures 3K and L). LAMP1 was also found to be highly expressed by epithelial cells of the BALT ( Figure 3L).
Immunofluorescent staining of cell populations in the parabronchi
Similar to the BALT, non-vaccinated 5-7 week old animals were used to study cells in the parabronchi in more detail. Scattered CSF1R-transgene + cells are present throughout the parabronchi, including in the infundibulum and interatrial septa and occasionally observed in the epithelium of parabronchi, around atrial smooth muscle (Figure 4). B and T lymphocytes are scattered throughout the parabronchi with T cells found to be in close proximity with CSF1R-transgene + cells (Additional file 2).
More in depth analysis of the MNP subpopulations shows scattered CD11 + cells present throughout the parabronchi, including interatrial septa (Figures 4A and D). All CSF1R-transgene + cell expressed CD11 and a subpopulation of CD11 + cells were observed to be CSF1Rtransgene − . Clusters of CD11 + CSF1R-transgene − cells were also observed in association with blood vessels in the connective tissue that separates two adjoining parabronchial lobules. Individual MHC II + cells were found to be distributed throughout the parabronchi, including in the atrial septa ( Figures 4B and E). Most of the MHC II + cells are also CSF1R-transgene + , with some scattered MHC II + CSF1R-transgene − cells in the lamina propria, likely being B cells.
Furthermore, we observed perivascular lymphoid aggregates of CD11 + cells in the parabronchi which have previously been reported [11]. Occasionally, MHC II + cells formed small aggregates and are likely DCs and B cells. The antibodies used in this study do not suffice to distinguish between macrophages and DCs because both cell populations can express the CSF1R-transgene, MHC II, CD11, and LAMP1.
A majority of the CSF1R-transgene + cells expressed LAMP1 and these cells were located in the atrial septa and infundibulum. Pneumocytes in the interatrial septa were CSF1R-transgene − but strongly positive for LAMP1, in contrast, epithelium lining the infundibula and air capillaries were negative for LAMP1 ( Figures 4C and F).
Development and distribution of MNP and GCs in the trachea of CSF1R-transgenic chickens
The distribution of CSF1R-transgene expressing cells was examined in trachea of MacReporter chickens of different ages. Scattered individual CSF1R-transgene + cells were observed to be distributed throughout the tracheal mucosa, in all age groups (Figures 5A-E). At ED18 the formation of tracheal epithelium is evident and becomes more defined 1 day post-hatch (Figures 5A and B). Although rarely observed at 1 day of age, small aggregates of CSF1R-transgene + cells were observed throughout the trachea at 1-week of age ( Figure 5C). These lymphoid aggregates increase in size and number as the bird increases in age (Figures 5D and E). From 4 weeks onwards, areas with multiple lymphoid follicles were observed on the tracheal mucosa, containing tightly packed CSF1Rtransgene + cells and a clear halo, which become larger and more numerous as the bird ages ( Figure 5E). Scattered individual CSF1R-transgene + cells were abundantly present in tracheal lamina propria ( Figure 5).
Development and distribution of MNP and GCs in air sacs of CSF1R-transgenic chickens
The distribution of CSF1R-transgene expressing cells was examined in the air sac of MacReporter chickens of different ages ( Figure 6). Scattered individual CSF1R-transgene + cells were distributed throughout the air sacs in all age groups (Figures 6A-F). No cellular aggregation or organized structures were observed at ED18 or 1 day of age, but small aggregations of CSF1Rtransgene + cells were observed in 1-week-old chicks. Interestingly, at 1 day post-hatch CSF1R-transgene + cells begin to form a pattern and clustering can be observed in the vicinity of capillary formation ( Figure 6B). The lymphoid aggregates are primarily located at the vicinity and bifurcation of capillaries ( Figure 6C). These CSF1Rtransgene + cellular aggregates increase in size and number with age and are frequently observed at 2 and 4 weeks of age ( Figures 6D and E). At 8 weeks of age, many of these aggregates have GCs, characterized by tightly packed aggregates of CSF1R-transgene + cells surrounded by a clear halo ( Figure 6F). To demonstrate the potential utility of MacReporter chickens in dissecting host-pathogen interactions, antigen-uptake and vaccine targeting, pilot studies were undertaken using fluorescently labelled beads and heatinactivated APEC-GFP. Figure 7 shows the deposition of intratracheally administered red fluorescent beads in the openings to the SB, in the air sac, and uptake of fluorescent beads by respiratory epithelium and by MNPs (Figures 7A-D). We were also able to demonstrate colocalisation of APEC-GFP and lymphoid follicle in the BALT area and uptake in the parabronchus of APEC-GFP in CSF1R-transgene + cells (Figures 7F and G). Ongoing studies are examining the cell tropism of APEC strains representing dominant serogroups and cellular responses by RNA sequencing of sorted infected cells.
Discussion
The advantage of using MacReporter chickens is that the development of the lymphoid structures can be assessed ex vivo using whole mount imaging of the entire tissue without the need for tissue preparation and histology. Moreover, the entire trachea and lung can be imaged, lowering the risk of missing smaller focal structures and enabling to dissect specific areas of interest. By analyzing a number of developmental stages during chicken growth, we observed the presence and expansion of lymphoid aggregates consisting of CSF1R-transgene + cells in the respiratory tract from vaccinated and non-vaccinated animals.
For the first time the lymphoid aggregates in the whole avian lung were visualised via whole mount microscopy. This application allowed for the visualisation of numerous CSF1R-transgene + cellular aggregates in the vicinity of openings to the laterodorsal and mediodorsal SB. The presence of constitutive BALT has been described in chickens, rats and rabbits [7]. In humans, BALT can be found in childhood but regresses into adulthood. However, its constitutive presence is still controversial [38,39]. A related BALT-like structure known as inducible BALT (iBALT), an ectopic lymphoid tissue, can be found upon inflammation or infection in both mice and humans and appears throughout the lung [39,40]. In our study vaccinated and nonvaccinated animals reared under conventional conditions presented no differences in the presences of BALT structures albeit vaccinated animals may have more active immune system compared to unvaccinated animals. In chickens, ectopic lymphoid tissues appear in the lung after infection with, for example, avian influenza virus (Vervelde and Reemers, unpublished observations). The development of the BALT in the chicken lung was progressive and consistent with those previously reported [11], with lymphoid follicles and the number of follicles associated with each opening of SB increasing in size and number with age. In contrast to previous studies, in the MacReporter birds we saw small aggregates were already present in 1-week-old chicks. The earlier detection might have been due the fact that animals were vaccinated via the drinking water 1 day post-hatch, but more likely because we could analyse the complete lung instead of histological sections only. Mature BALT structures with B and T cell areas and GCs were observed from 4 weeks of age onwards. Interestingly, lymphoid follicles were also present around the medioventral SB in some of the birds. These are the most cranial SB, and are composed of four large openings in the genus Gallus [8]. This is in contrast with previous literature [7], which only reports BALT in the vicinity of the most caudal SB (the laterodorsal and posterior SB).
We examined the MNPs and epithelial cells of the BALT and parabronchi of the lung from 5 to 7 week old non-vaccinated birds reared under conventional conditions. Using immunofluorescence staining, the CSF1Rtransgene + lymphoid aggregates observed via whole mount microscopy, were also present in non-vaccinated animals and were tightly packed GCs consisting of B cells, FDC and tingible body macrophages, distinguished based on their Bu-1 (chB6) and TIM4 staining and CSF1R-transgene + expression. The presence of multiple GCs in the BALT region of non-vaccinated animals is similar to earlier findings that SPF and conventionally reared birds had no difference in BALT development [7]. LAMP1 staining colocalised with CSF1R-transgene expression but was also found in cells lacking CSF1Rtransgene expression. These cells may represent phagocytic cells that express LAMP1 in their cytotoxic granules that do not express the transgene, such as natural killer cells [41] or express the transgene at low levels, such as granulocytes [26]. Respiratory MNPs in the parabronchi have been described previously with CD11 + , KUL01 + and DEC205 + cells located in the interstitial tissue of the primary bronchus wall and CD11 + and KUL01 + cells located in interatrial septa [17]. Although the present study did not distinguish between macrophages and DCs, the MacReporter chicken allow for the appreciation of the location and quantity of MNP in the BALT and parabronchi of unvaccinated healthy animals. We also observed the presence of M cells in the FAE overlaying the BALT. Previous reports have identified these cells by transmission electron microcopy in the turkey [11] and failed to identify them in chicken lung by SEM [7]. In our studies these cells were found to express the CSF1Rtransgene and lack jacalin binding in contrast to the IFE overlaying the BALT structure. This observations makes BALT-associated M cell identification and study more accessible. M cells can mediate infection by bacteria [35,42] and viruses [43] and may represent a route to infection in chickens as observed in mammals [35,36].
In the lung parabronchi, pneumocytes in the interatrial septa were found to be positive for LAMP1 staining but lacked CSF1R-transgene expression. The LAMP1 staining in epithelial cells lining the atria suggests that these cells are able to phagocytose and process antigens. These lysosomal bodies in avian pneumocytes have been previously described [44] and the cells were classified as type II pneumocytes [45]. Besides being LAMP1 + , they also express the FDC marker CVI-ChNL-74.3 [46]. The limited number of highly mobile free residing macrophages in the lumen maybe correlated to the fact that the avian epithelium immediate to the respiratory surface is strongly phagocytic [45]. In addition to the airway, M cells in the epithelium of the primary bronchus expressing CSF1R-transgene, these type II pneumocytes also express molecules in common with MNPs. Similarly, mouse, sheep, and human type II pneumocytes and MNPs express DC-LAMP/CD208 [47] and RAGE [48]. Small aggregates of MHC II + cells are occasionally observed in the parabronchi indicating that, along with the BALT, the parabronchi play an important role in the immune response of the avian lung. Scattered CD4 + and CD8α + cells are also present in the parabronchi and occasionally forming subepithelial aggregates of variable sizes. Perivascular clusters of B and T cells are observed, and in older birds occasionally have GCs (unpublished observations) [22]. Together with macrophages, T cell numbers increase rapidly after infection or vaccination with respiratory viruses or bacterial infections [19,21,43,49]. The combination of free residing macrophages, phagocytic epithelium, the numerous macrophages and DC in the underlying tissue and the fact that these cells can been mobilised rapidly after infection ensures an effective cellular defense system in the chicken lung.
The development of GCs in the trachea was observed using the MacReporter animals. From ED18 up to 8 weeks post-hatch, the accumulation of tightly packed areas with multiple follicles increased with age. Changes in MNP in the trachea have been reported after infection with e.g. IBV, E. coli [19,49] and ITLV [50]. A similar pattern of lymphoid accumulation during development was also evident in the avian air sacs. Using electron microscopy, Bezuidenhout reported similar findings to this report where perivascular aggregation of leucocytes (macrophages, heterophils, lymphocytes, plasma cells, monocytes and mast cells) were observed in the air sacs [23]. The perivascular aggregation of CSF1R-transgene + cells in the air sac could indicate that these cells play an important role in the induction of the local immune response as previously seen after infection with IBV and E. coli [19,22]. Our pilot studies using inert beads and inactivated APEC bacteria show the potential of the Mac-Reporter birds to investigate the respiratory immune system during infection and provide the tools to better investigate host-pathogen interactions and vaccine targeting.
In summary, we demonstrated the potential of CSF1Rtransgenic MacReporter birds to visualize lymphoid tissue in the respiratory tract of birds of all ages using whole mount imaging and immunofluorescent staining. In all tissues, trachea, lung and air sac, the development was progressive, with the number and size of lymphoid follicles increasing with age. It should be noted that some animals were vaccinated via drinking water which may play a role in the early detection and abundance of follicles observed in this study. The expression of the transgene enabled us to visualize organized lymphoid tissue at an earlier age and especially at different locations than previously published. The results presented in this study will serve as a base for further functional characterization of MNP, macrophages and DC subpopulations in uptake processing and presentation of antigens to T and B cells and the role of MNP in respiratory diseases such as E. coli, IBV and avian influenza.
Additional files
Additional file 1. Vaccination regime at the National Avian Research Facility, Edinburgh, UK. Routine vaccination regime of birds used for the Visualisation and characterisation of mononuclear phagocytes in the chicken respiratory tract using CSF1R-transgenic chickens.
Additional file 2. Location of B cells, T cells and follicular dendritic cells (FDC) in the lung of MacReporter chickens.
The BALT region of 5 to 7 week old non-vaccination animals were analysed for B, T and FCD cells. Isotype controls were used to standardise the microscope and examine aspecific binding before acquiring images (A-B). The GC of MacReporter animals are tightly packed with Bu1 -CSF1R-eGFP + FDC cells and Bu1 + CSF1R-eGFP -B cells (C) with few Bu1 + B cells found in the parabronchi (F). CD3 + T cells are disperse within and outside the GC (D) and parabronchi (G). CSF1R-eGFP + FDC cells express Fc receptors and trap immunoglobulin by expressing IgY (E) and CSF1R-eGFP + IgY + FDC are rarely detected out with the GC, BALT region of the lung. GC are indicated by white dashed lines. | 2018-10-27T16:49:07.248Z | 2018-10-10T00:00:00.000 | {
"year": 2018,
"sha1": "3d7e991bed0d9a2f561957ed8682b654bce5ccf7",
"oa_license": "CCBY",
"oa_url": "https://veterinaryresearch.biomedcentral.com/track/pdf/10.1186/s13567-018-0598-7",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "2ba408bad7d73ad466fd2ad8cabe177374e3798a",
"s2fieldsofstudy": [
"Biology",
"Medicine"
],
"extfieldsofstudy": [
"Medicine",
"Biology"
]
} |
29401597 | pes2o/s2orc | v3-fos-license | Use of certainty-based marking in a second-year medical student cohort: a pilot study
Background Assessments which consider both competence and confidence attempt to provide insight into actual performance in order to optimize physician capabilities, providing motivation and direction for future learning. The aim of this project was to assess medical students’ thoughts and opinions of the utility of a certainty-based marking (CBM) protocol with respect to improving their learning experiences. Methods Second-year medical students at the University of Queensland were provided with a series of optional online formative assessment tools, in the form of 10 sample questions, to support their current module learning outcomes. During four consecutive weeks, CBM was offered on weeks 1, 2, and 4, with week 3 being provided in the usual question-answer format. A mixed-method survey was distributed at the conclusion of the trial period to obtain feedback on the students’ impressions of learning via this technique. Results Of the 400 students, 15%, 11%, 9%, and 8% used the resource over the four-week period, respectively. During the four-week module directly prior to the test module, 46%, 44%, 44%, and 40% of the students accessed the sample questions which were delivered in the usual multiple choice format. A majority of the students either agreed or strongly agreed that CBM was easy to understand (52%) and useful (57%), but took more time (67%) because they needed to consider their certainty level for every question (76%). A number of students (43%) also stated that CBM affected their attitudes toward decision-making, while 86% thought it would be most useful for revision as opposed to an examination format. Discussion Despite the inherent benefits of gaining experience in higher order thinking processes, students were less likely to participate in the CBM tasks than standard multiple choice, even though these did not count toward their final grades. Conclusion Utilizing such practices at the beginning of an educational program may minimize apparent resistance and alter learning practices to become conducive to deeper levels of learning. This has been corroborated in other studies aiming to encourage similar higher order cognitive processes.
Introduction
Certainty-based marking (CBM), formerly known as confidence-based marking, is a tool used to encourage students to think more carefully prior to deciding on an adequate response during assessments. It encourages reflection on reasoning prior to making clinical-based decisions, because the answers also require an indicator of level of confidence in their response. The process is thought to enhance deeper levels of learning at the expense of commonly utilized rote learning practices. These apparently superficial techniques create a lack of thorough understanding of issues, despite being This article was published in the following Dove Press journal: Advances in Medical Education and Practice 20 December 2012 Number of times this article has been viewed effective in achieving satisfactory grades when assessed by simple yes and no responses. 1 According to Bloom's Taxonomy of Learning, knowledge is the simplest level of learning to acquire. Here information is recalled exactly as it is presented and requires memory of definitions, formulas, or procedures. These levels progress through comprehension, where students state relevant material in their own words, and application where rules are applied to a problem without being explicitly stated. Analysis requires yet deeper levels of learning where relationships are drawn and complexity is broken down. Evaluation exists at the top of this hierarchy, where judgments are conceived based on the worth of something for a specific purpose. 2 Having students evaluate their progress in the form of CBM demonstrates higher order thought processing than standardized direct question and answer formats. 1 CBM in different forms is decades old. Generally speaking, students are requested to grade their confidence in the answer they have stated at one of three levels, ie, 1, 2, or 3. If correct, this becomes the number of marks they receive. However, if they are incorrect, at the lowest level there is no penalty, followed by -2 and -6, respectively. According to game theory, level 2 should be chosen when students are over 67% certain and level 3 if greater than 80%. 3 CBM ensures that students differentiate responses as to whether they are based on sound knowledge or may be at risk of inaccuracy. It encompasses the importance of vigilant thinking and reassurance, which assist to link more diverse aspects of information. In this process, confident responses receive higher scores if correct, but incur higher penalties if incorrect. A qualitative evaluation study by Issroff and Gardner-Medwin 4 documented quotes by students such as "it stops you making rush answers", "you can assess how well you really understand a topic", and "it makes one think … you are forced to concentrate".
A key feature of these methods is their ability to motivate students via awarding extra marks for certainty of a correct response, while penalizing an incorrect but certain response. When choosing a marking scheme, it is advisable to focus on the way confidence is elicited to ensure proper motivation. 5 Negative marking schemes have been acknowledged to discourage guessing, but this is not necessarily the case unless there is a positive motivation to perform otherwise. 6 Subsequent documented issues have arisen in regard to using CBM. There is concern that overly self-confident or less confident students may be disadvantaged by this style of assessment. However, it is argued that these personality styles should become more self-aware through the process, if they are not to be penalized. Essentially, students who recognize their rationale or lack thereof benefit, while those who are consistently confident or apprehensive do not. 1 Results from CBM assessments may also be utilized effectively for evaluations of course material, where questions eliciting high levels of incorrect confident answers can highlight common misconceptions to educators. 7 The process of self-explanation, which grows from the need to make confident judgments, can assist the student to refine their methods of problem-solving. 8 These constructed activities, directed at recognizing missing or unreliable information, can lead to modification of existing knowledge and creation of a new understanding. 9 The use of CBM in health care training has important implications, particularly in clinical situations, where any risk of error should be recognized and acted on accordingly. It is likely to be far more detrimental to possess a confident belief in an erroneous judgment than the ability to recognize a lack of knowledge in that area. 10 Hence, the aim of this study was to assess students' opinions of CBM as a useful adjunct in enhancing current curricula, as well as to explore the effectiveness of its implementation.
Materials and methods
Second-year medical students from The University of Queensland were asked to volunteer to make use of CBM format multiple choice questions to assist with formative evaluations of their own levels of knowledge. Course content is generally presented as sets of modules throughout the first two years of the MBBS program. Each week, a series of 10 multiple choice questions is made available via the Moodle on-line learning system to support student education. The Moodle 1.9 platform features a specific quiz module which incorporates CBM. Detailed information on the use of this media is available. 11 The current project converted three of the four weeks of formative assessment quizzes, within a nervous system module, into CBM-style criteria, namely, weeks 1, 2, and 4. A mixed-method design questionnaire was also available for the students to provide feedback on their experience and opinions of CBM, if they so wished. Follow-up survey questions included whether CBM was "easy to understand" or "useful", and whether it "affected attitudes" or "length of time to provide an answer". Students were also asked if CBM "increased their consideration of certainty" or if they felt it to be "a waste of time".
Student participation was entirely voluntary and all relevant data collected were deidentified. Prior to their submit your manuscript | www.dovepress.com Dovepress Dovepress involvement, participants were provided with a detailed information sheet outlining the project and also highlighting the many potential benef its of considering certainty when problem-solving. These included but were not limited to increasing reflective practice and encouraging the development of higher-order cognitive processing, 12 subsequently enhancing clinical competence and reasoning skills. 13 Ethical approval was obtained from the University of Queensland's ethics committee. Consent to utilize the deidentified research data arising from this project was obtained from the students prior to their participation.
Results
Of the 400 students, 15%, 11%, 9%, and 8% used the resource over the four-week period, respectively. During the four-week module directly prior to the test module, 46%, 44%, 44% and 40% of the students accessed the standard multiple choice method sample questions. Only about half of the participants elected to complete the survey (n = 39). The results are outlined in Table 1.
Responses on additional time required to complete CBM quizzes as opposed to standard formats ranged from 30 seconds up to 30 minutes, with the most common response being around 50% of additional time required. Opinions of the usefulness of CBM were mixed with a variety of responses. Positive comments included "it assisted in interpretation discrepancy", "a good tool to revise learning", and "seemed to make me more analytical with every question and work through the information more precisely". Negative responses consisted of "most ridiculous thing ever" and "would be too time-consuming during exams", while mixed reactions comprised thoughts such as "a lot more work than the standard quizzes, but more informative" and "best for clinicals but not exams".
Discussion
The ability of physicians to recognize when they are not certain of an adequate treatment protocol is vital, especially in an emergency situation. A student's confidence about incorrect responses may be far more unfavorable than recognition of uncertainty with a particular answer. 5 Utility of CBM during medical education attempts to facilitate this awareness process. As such, this project aimed to explore the implementation and uptake of such an approach, as well as the students' impressions of its effectiveness in improving their learning outcomes.
Traditionally, CBM has been made use of in true/ false questions, although no particular issues have been documented by its use with numerical, best-of-5, or extended matching questions. 14 Typically traditional multiple choice types of assessment may lend to situations where a student is able to answer a question, once options have been presented, due to a latent recognition process. This form of cueing can be particularly problematic when diagnostic reasoning is being assessed due to premature closure prior to the correct diagnosis being considered, which is a common reason for 13,15 Use of CBM may minimize this effect because more careful consideration is required once certainty-based parameters are included. The majority of our participants (76%) indicated that they needed to consider the certainty level for every question. Interestingly, a small portion (29%) deemed consideration of certainty to be a waste of time. Extra time required to complete an assessment is another matter to be considered when comparing CBM with conventional testing. The majority of our cohort (67%) specified more time was required to ensure an accurate response. An evaluation study conducted by Issroff and Gardner-Medwin documented that many students stated they sometimes changed their responses when having to consider certainty around their answers. 4 Considering the large portion of our participants indicating their preference for CBM as a revision tool as opposed to being utilized in examinations, the issue of additional time required to respond to questions may have impacted overall impressions and participation. Despite inherent concerns on time consumption, another study assessed the reliability analysis of examinations and concluded that less than a third of the number of questions was needed for equally reliable assessment data using CBM techniques. 12 Extraordinary time constraints are often in place in medical practice, where processing and retrieving information occurs continually at the time of action. 16 It has also been established that time constraint is one of the major factors limiting a physician's ability to comply with preventative recommendations to their patients in practice. 17 These internal time constraints may lead to damaging results in which patients and providers may both become victims. 16 Should we be pushing these boundaries already at the commencement of the medical education process, selecting the specific value of tools in the interest of time?
Limitations of this study include potential selection bias, because only a small sample of the total student body made use of the CBM resource. Despite being informed of the potential benefits of CBM prior to its implementation, participation may have been improved if it had been provided at the beginning of the first year as opposed to the middle of the second year when students are more likely to be set in their thought processing.
Conclusion
Regardless of the inherent benefits of such an approach, students were less likely to participate even though the tasks were formative assessments which did not count toward their final grades. Perhaps using such practices at the commencement of an educational program would alter learning practices conducive to deeper levels of learning. This may also minimize resistance to the additional thought processing required in ascertaining certainty of core knowledge. In the words of the famous philosopher, Confucius, "When you know a thing, to hold that you know it. And when you do not know a thing, to allow that you do not know it. This is knowledge". 18 | 2017-06-19T15:48:52.681Z | 2012-12-20T00:00:00.000 | {
"year": 2012,
"sha1": "dba41c7ede4eb171ef1765ea4cda70dc960b7a0b",
"oa_license": "CCBYNC",
"oa_url": "https://www.dovepress.com/getfile.php?fileID=14777",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "4e33bc3462d0b222d27fbbfd823fe59424362d14",
"s2fieldsofstudy": [
"Education"
],
"extfieldsofstudy": [
"Computer Science",
"Medicine"
]
} |
5565501 | pes2o/s2orc | v3-fos-license | Energy-constrained optimal quantization for wireless sensor networks
As low power, low cost and longevity of transceivers are major requirements in wireless sensor networks, optimizing their design under energy constraints is of paramount importance. To this end, we develop quantizers under strict energy constraints to effect optimal reconstruction at the fusion center. Propagation, modulation, as well as transmitter and receiver structures are jointly accounted for using a binary symmetric channel model. We first optimize quantization for reconstructing a single sensor's measurement. Optimal number of quantization levels and optimal energy allocation across bits are derived. We then consider multiple sensors collaborating to estimate a deterministic parameter in noise. Similarly, optimum energy allocation and optimum number of quantization bits are derived analytically and tested with simulated examples.
INTRODUCTION
Wireless sensor networks (WSN) are gaining increasing research interest for their emerging potential in both consumer and national security applications.Sensor networks are envisioned to be used for surveillance, identification, and tracking of targets.They can also serve as the first line of detection for various types of biological hazards such as toxic gas attacks.In civilian applications, WSN can be used to monitor the environment and measure quantities such as temperature and pollution levels.
In most application scenarios, WSN nodes are powered by small batteries, which are practically nonrechargeable, either due to cost limitations or because they are deployed in hostile environments with high temperature, high pollution levels, or high nuclear radiation levels.These considerations motivate well energy-saving and energy-efficient WSN designs.One approach to prolong battery lifetime is the use of energy-harvesting radios as the ones in [1] with power dissipation levels below 100 μW.A lot of research has been carried out to devise energy efficient algorithms in each layer of WSN [2].Optimal modulation with minimum en-ergy requirements to transmit a given number of bits with a prescribed bit error rate (BER) bound is considered in [3].Energy efficient medium access control (MAC) and routing protocols are studied in [4,5], respectively.
In this paper, we consider a WSN with a fusion center which collects data from sensor nodes and performs the final information extraction task.A common goal in most WSN applications is to reconstruct the underlying physical phenomenon (e.g., temperature) based on sensor measurements.Energy as well as bandwidth limitations prevent sensor nodes from transmitting real valued (analog-amplitude) data to the fusion center.This motivates the goal of this paper which is to derive optimal quantization schemes at sensor nodes under strict energy constraints.Optimality here is in the sense of minimizing a bound on the mean-absolute reconstruction error at the fusion center.The problem setup originates from the following considerations.Suppose we deploy a WSN powered by nonrechargeable batteries and expect it to operate a given number of times, which bounds the energy allowed per time of operation.One operation could be, for example, one time transmitting a burst of temperature measurements to the fusion center with bounded energy allowed per burst.The problem of designing quantizers to optimize pertinent reconstruction performance metrics under a given energy budget emerges naturally.
Most of the prior works on optimal quantization deal with optimization of the quantization rules for detecting a signal in dependent or independent noise [6][7][8][9].Other related works include [10][11][12][13][14][15].Assuming error-free transmission, [10,11] focus on the impact of bandwidth/rate constraints in WSN on the distributed estimation performance.Optimal quantization thresholds, given the number of quantization levels and channel coding for binary symmetric channels (BSC), are jointly designed in [13] to minimize the mean-square error of reconstruction.In [14], scaling of the reconstruction error with the number of quantization bits per Nyquist-period is studied.The rate-distortion region, when taking into account the possible failure of communication links and sensor nodes, is presented in [12].Possibly the most closely related to our present work, [15] minimizes the total transmission energy for a given target estimation error performance.Different from these works, our objective is to optimize the quantization per node (including the number of quantization bits and the transmission energy allocation across bits) under a fixed total energy per measurement in order to minimize the reconstruction error at the fusion center.We account for both transmission energy as well as circuit energy consumption, while we (i) incorporate the noisy channel between each sensor and the fusion center by modeling it as a BSC with crossover probability controlled by the transmitted bit energy, and (ii) allow different quantization bits to be allocated different energy and, thus, effect different cross-over probabilities.
The rest of the paper is organized as follows.In Section 2, we consider optimal quantization in a point-topoint (single-hop) link to recover a single sensor's measurement with uncoded transmission schemes.In Section 3, optimal quantization is addressed in a multisensor (star topology) setting.The role of channel coding is then examined in Section 4. Section 5 provides some illustrative numerical results and Section 6 concludes the paper.
POINT-TO-POINT LINK
Let us consider the system depicted in Figure 1, where a single sensor acquires a local measurement A, which is properly scaled so that A ∈ [0, 1], and wishes to transmit it to the fusion center.For digital transmission, the sensor first quantizes the real valued measurement A to A Q .Letting A := ∞ i=1 b i 2 −i , throughout this paper, we consider N-bit uniform quantization so that The quantization bits {b i } N i=1 are transmitted through the wireless channel to the fusion center, and are demodulated as { b i } N i=1 .At the fusion center, the sensor's local measurement is reconstructed as ( Here, for simplicity, we consider only uncoded transmissions.The underlying channel is assumed to be memoryless with different raw bits experiencing independent detection errors.Under this condition, we can model the wireless air interface between the sensor and the fusion center as a binary symmetric channel (BSC) with cross-over probability .In fact, the BSC model can be used to characterize a more general class of channels including multipath fading and multiaccess ones.Even for a channel with memory, BSC is still applicable provided that a suitable equalizer is incorporated, and { b i } denote the bits at the output of the slicer that follows the equalizer.Because one of the key issues in optimizing the design of sensor networks is the energy constraint, we are interested in the following problem.
If the allowable energy each time we transmit a measurement is fixed to E , what is the optimal number of quantization bits and how can the energy per bit be allocated optimally in order to minimize the reconstruction error at the fusion center?
In the following subsections, we will first address this question under the assumption that the total energy budget used for RF transmission of the measurement A is fixed and equal to E .The energy consumed by the circuit electronics will be taken into account afterwards.
Optimizing the number of quantization bits
Let us consider a simple scenario where all quantization bits are allocated equal energy.We wish to find the optimal value of N in (1) which minimizes a meaningful metric of the reconstruction error.When using an N-bit quantizer, with the total transmission energy of all bits fixed to E , the energy per bit depends clearly on N, since E b = E /N.Noticing that the BSC model's cross-over probability will generally be a function of the bit energy-to-noise ratio, and letting N 0 denote the channel noise level, we can write as (E b /N 0 ) to make this functional relationship explicit.With E b = E /N, we find that the cross-over probability is actually a function of N with = (E /(NN 0 )).
The reconstruction error, which is defined to be A − A, can be expressed as ( Using the triangle inequality, we can readily bound the absolute value of the reconstruction error as Taking expectation on both sides of (4), we have where in deriving the first equality we used the fact that In order to minimize the mean-absolute reconstruction error, it suffices to minimize the bound f (N) in (5) with respect to N, which corresponds to optimizing the worst-case performance.Under this criterion, the optimal number of quantization bits will be chosen as follows: Clearly, the first summand in f (N), namely, 2 −N , decreases as N becomes larger.With (•) being a monotonically decreasing function of its argument, the second summand, (1 − 2 −N ) (E /(NN 0 )), will be an increasing function of N. Intuition suggests that there should exist an optimal N such that f (N), that is, the sum of the two terms, will reach a minimum.If the latter is unique, a simple one-dimensional numerical search will reveal N opt , as long as (•) is specified.In Section 5, we will give examples of the optimal number of quantization bits when (γ) is specified for different modulations and receiver formats, with γ := E b /N 0 denoting the bit energy-to-noise ratio.
Optimizing the energy allocation per bit
In the previous subsection, we assumed that each bit is allocated identical energy.However, observing that each bit in (4) has a different weight suggests that there is room to optimize the energy per bit.This motivates us to look for an optimal energy allocation scheme when the total number of bits N is fixed.Let us suppose that bit i is allocated a fraction x i of the total energy E for i = 1, . . ., N.Then, following the derivation of (5), we have In order to account for the mean-absolute reconstruction error with respect to x := [x 1 , . . ., x N ] T , we can formulate the following optimization problem: It is easily seen that the optimal solution and the minimum value of f 0 (x; N) are actually functions of N. To make this functional relationship explicit, we denote the optimal solution by T , and accordingly, the minimum of the objective function f 0 (x * N ; N).Interestingly, as long as N > N, we find f 0 (x * N ; N) < f 0 (x * N ; N), see the appendix for details. With T denoting the optimal solution, the well-known Karush-Kuhn-Tucker (KKT) conditions [16, page 243] dictate that there must exist {λ * i } N i=1 and ν * such that where ∇ denotes the gradient.It follows from (11) that the In order to gain further insight from (12), let us take a closer look at the optimal energy allocation in two special cases.
BPSK over AWGN channel
The cross-over probability expressed in terms of bit energyto-noise ratio γ is given in this case by [17, page 255 EURASIP Journal on Advances in Signal Processing The derivative of (γ) with respect to γ is then calculated as follows: Substituting ( 14) into ( 12), we can express the optimal energy allocation in the following form: Noticing that the domain of φ(γ) defined in ( 14) is (0, +∞), from the complementary slackness conditions in ( 9), we deduce that λ * i = 0, for all i, and finally, we obtain where ν * is a constant chosen to enforce the constraint 16) is intuitively appealing, because the monotonicity of φ(γ) ensures that each bit is allocated energy according to its significance: the smaller the i is, the more significant bit i is, and the more energy is allocated to bit i.
Binary orthogonal signaling with envelope detection
It is well known that binary orthogonal signals such as binary frequency-shift keying (FSK) or pulse-position modulation (PPM) can be demodulated using noncoherent envelope detection [17, pages 307-310].In this case, the cross-over probability expressed in terms of the bit energy-to-noise ratio is given by The derivative is then Substituting ( 18) into ( 12), we obtain Noticing that the function ϕ(γ) has domain [0, +∞) and range (0, ϕ(0) = 1/4], and supposing that λ * i = 0, for all i, we must have ν * 2 N N 0 /E ≤ 1/4.Furthermore, the condition N i=1 x * i = 1 is not guaranteed to be satisfied when ν * is bounded.Based on (9), we can simplify (19) as follows: Equation (20) implies that it is possible to have x * i = 0 for some large i's.In fact, when (γ) = (1/2)e −γ/2 , the problem in ( 8) can be readily shown to be convex which not only implies that the optimal solution is guaranteed to exist and is unique, but also can be found using a numerically efficient search.
The optimal energy allocation for a special case will be examined in Section 5, where we will confirm that a certain number of less significant bits should not be allocated any energy.
Circuit energy consumption
Till now, we have neglected the fact that the circuit itself will also consume a certain amount of energy when transmitting the quantization bits b i .Optimization in (8) implies that if we ignore circuit energy consumption and optimally allocate the transmission energy among bits, then the achieved reconstruction error bound will decrease as we increase the number of quantization bits N.However, this will not be true when the circuit energy consumption is taken into account.The reason is that the energy consumed by the circuit will also increase as the number of bits grows larger.To quantify this tradeoff, we adopt the model in [3], where the power of the circuit electronics (excluding the RF transmission power) is assumed to be P on when the sensor is transmitting each quantization bit.The energy consumption of the circuit electronics during the sleep and transition modes is assumed to be very small and can be neglected.Letting T 0 denote the bit period, when N quantization bits are transmitted, we can express the circuit energy consumption as E c = NT 0 P on .With the total energy budget per measurement transmission being E , the remaining energy for RF transmission of the N bits will be E r = E − E c = E − NT 0 P on .In order to have E r > 0, we obviously need to make sure that N < E /(T 0 P on ).Now, with the circuit energy consumption considered, let us revisit the issue of optimizing the number of quantization bits, which we have examined in the previous subsections.
Optimal number of quantization bits with equal energy allocation
Let us first assume that the residual energy E r is equally allocated among the N quantization bits.Similar to (5), we can upper bound the mean-absolute reconstruction error as from which the optimal value of N can be obtained as Comparing the latter with (6), we recognize that similar comments apply regarding the existence, uniqueness and numerical evaluation of the optimal N when circuit energy is accounted for.
Optimal number of quantization bits with optimal energy allocation
When the measurement A is quantized to N bits, the optimal strategy of allocating the residual energy E r = E − NT 0 P on is the solution of the following optimization problem: Denoting the optimal solution by x * N , we can obtain the optimal number of quantization bits as In Section 5, we will find N opt for specific system setups in the case of equal energy allocation and optimal energy allocation when energy consumption of the underlying circuitry is taken into account.
MULTISENSOR COOPERATION IN ESTIMATING A PARAMETER
Let us now consider the multisensor setup depicted in Figure 2, where each sensor k has available local bounded noisy observation X k = A + n k , and n k is zero mean with variance σ 2 k and independent of n l for l / = k.After normalization, we have X k ∈ [0, 1].Sensor k quantizes its local observation X k to the N k most significant bits, that is, with i } Nk i=1 are then transmitted through the wireless channel, which is again modeled as a BSC with cross-over probability k .The fusion center reconstructs X k with the demodulated bits { b (k) i } Nk i=1 to obtain When we have available unquantized real valued observations This motivates us to form the following estimator for the parameter A when the noise variances are known at the fusion center, where we have only available X k , k = 1, 2, . . ., K: The problem we are interested in can be formulated as follows.
A For a fixed number of quantization bits (N k ) per sensor, what is the optimal scheme for allocating the total energy E T prescribed to all sensors so that the mean-square estimation error E| A − A| 2 is minimized?Furthermore, what is the optimal number of quantization bits per sensor so that this energy allocation scheme achieves the minimum possible estimation error?
In this section, we will neglect the circuit energy consumption.Generalization to the case where the energy consumption by the circuit electronics is nonnegligible is rather straightforward using the model described in Section 2.3.Furthermore, we assume that the energy allocated per sensor will be equally distributed among the quantization bits.Now, let us take a look at the estimation error Upon defining the reconstruction error X k := X k − X k , we have and n k are uncorrelated.Furthermore, as shown in [19], when the characteristic function of n k is bandlimited to 2π/Δ, where Δ = 2 −Nk is the quantization step size, the quantization error (X k ) Q − X k is uncorrelated with the input X k = A + n k .(In a uniform quantizer with step size Δ, the correlation between input X and the quantization error is given by [ , where φ(ω) := E[e jωX ] and φ(ω) := dφ(ω)/dω.Therefore, as long as the φ(ω) energy is concentrated in the interval [−2π/Δ, 2π/Δ], one can safely consider X and as uncorrelated.)Hence practically, as long as the quantization step Δ = 2 −Nk is sufficiently small relative to σ k , one can safely assume the reconstruction error is statistically uncorrelated with the observation noise n k .Thus, the second summand in the numerator disappears.Hence, minimizing E|A − A| 2 reduces to minimizing Because for any bounded random variable Z ∈ [−U, U] with pdf p(z), we have |, which we upper bound as where k is the cross-over probability of the BSC between sensor k and the fusion center.
Identical number of bits per sensor
For clarity in exposition, we first consider here a simple situation where each sensor transmits the same fixed number of bits N (i.e., N k = N, for all k).With x k denoting the fraction of the total energy E T allocated to sensor k, we can express k as k (E T x k /(NN 0 )), where N 0 is the noise level at the receiver of the fusion center which is assumed common to all channels.The optimal energy allocation scheme will be the solution of the following optimization problem (x := [x 1 , . . ., x K ] T ): As in Section 2, we can write down the KKT conditions for the optimal solution x * := [x * 1 , . . ., x * K ] T as follows: From (34), we have To delve further into (35), we consider a particular system setup.Letting κ denote the path loss exponent [20] of the wireless channel (d k is the distance between sensor k and the fusion center), and supposing BPSK modulation, we can express the cross-over probability in the presence of AWGN as ) with C being a constant.Under these operating conditions, (35) and (32) yield where ν * is chosen such that K k=1 x * k = 1.In Section 5, we will examine a specific system and find the corresponding optimal energy allocation to gain further insight into these closed-form expressions.
In fact, when k (γ), for all k, is convex in γ, the problem in (31) turns out to be convex, which implies that the global optimum exists and can be easily found numerically.In most cases, convexity is guaranteed, for example, when k (γ) is expressible in terms of Q( 2γ) or (1/2)e −γ/2 .Subsequently, the optimal number of quantization bits N opt can be easily found using one-dimensional numerical search to solve the optimization problem where f 0 (x * ; N) is the optimal value of the objective function in (31) when the number of quantization bits per sensor is N.In Section 5, we will show an example of the functional relationship between f 0 (x * ; N) and N, from which N opt can be readily determined.
Different number of bits per sensor
Now, let us consider the case where sensor k transmits N k quantization bits, k = 1, . . ., K. From (30), we can see that the optimal energy allocation scheme which minimizes the estimation error is the solution of the following optimization problem: 1 , . . ., x (l) K ] T as the optimal solution of (38).(ii) Update N (l) k to N (l+1) k based on the iteration (iii) Go to (l + 1) st step.
When k (γ) is convex and {N k } K k=1 are fixed, the problem in (38) is clearly convex, which implies that the optimal energy allocation vector x * can be found using standard numerically efficient search schemes [16].Hence, step (i) of Algorithm 1 is easily carried out.It is also easy to prove that the objective function is always decreasing from one iteration to another.The argument is as follows: Our experience with simulations is that Algorithm 1 typically converges after 3-4 iterations.In Section 5, we will utilize this approach to jointly optimize N k and x k , k = 1, . . ., K, for a specific wireless sensor network.
Remark 1.In this section, we have dealt with energy and quantization optimization for multiple sensors that are cooperating in the estimation of a common parameter.The optimum scheme will be first derived in a centralized manner and then released to each individual senor, which may create a lot of scheduling overhead.However, in practice, we will not need to update the optimum scheme frequently unless there is a major change in the configuration of the sensor network.
EFFECTS OF CHANNEL CODING
In the preceding sections, we have limited our consideration to uncoded transmissions.In this section, we will examine the performance limit of our reconstruction problem when error control codes are adopted before the quantized bits enter the BSC.The total energy budget for RF transmission of A ∈ [0, 1] is again constrained to be E .The energy consumption of the circuit electronics will be neglected here for clarity in exposition.
Single measurement transmission
Suppose now that A is quantized to N Q bits as in (1), and that a (2 NQ , N) channel code is constructed to transmit the N Q bits over the BSC by using the channel N times.Letting P (N) e denote the average error probability of maximum likelihood (ML) decoding and A := NQ i=1 b i 2 −i denote the reconstruction of A at the fusion center, we have the following upper bound for the reconstruction error |A − A| at the fusion center: where in deriving (41) we have used the fact that A and A both lie in [0, 1].
In order to proceed, we need the following result from [21,Chapter 5].
Theorem 1 (Random coding theorem).For a discrete memoryless channel (X, p(x | y), Y), there exists a (e NR , N) block channel code with average error probability of ML decoding satisfying where E r (R) is the random coding exponent which is defined as The random coding exponent for a BSC with cross-over probability where In our case, since we use the channel N times, the bit energy per transmission will be effectively reduced to E /N, and thus, the equivalent BSC's cross-over probability will become (E /(NN 0 )).For our (2 NQ , N) channel code, the rate in nats/channel use will be (N Q /N) ln 2. Thus, applying Theorem 1 with an appropriate channel code, we can use (42) to bound (41) as Clearly, N Q and N can be optimally selected to minimize f code (N Q , N) and thus the reconstruction error.In Section 5, we will compare this upper bound with the bound achieved with the uncoded transmission schemes we developed in Section 2.
Multiple simultaneously transmitted measurements
The exponentially decreasing behavior of the decoding error probability described by the random coding theorem favors large block sizes.However, in the single measurement transmission case, when the block size N becomes large, the capacity of the underlying BSC goes to zero.To resolve this tradeoff, we can transmit multiple measurements together.
In practice, for some application scenarios, it may not be necessary for the remote sensor to transmit its measurement back to the fusion center immediately, that is, one can wait until L > 1 measurements {A 1 , A 2 , . . ., A L } ∈ [0, 1] L are acquired, and then transmit them jointly to the destination.The critical difference here is that the energy budget increases to LE .We assume no probabilistic model for the source and, again, employ the universal uniform quantizer to quantize each measurement to N Q bits.As a result, the total number of bits to be transmitted is LN Q .Adopting an appropriate (2 LNQ , LN) channel code with error probability P (LN) e , as in (41) and (45), we thus obtain Comparing the bound for a single measurement transmission, f code (N Q , N) in (45), with the bound for multiple simultaneously transmitted measurements, f code (N Q , N) in (46), we can see Equation ( 47) shows clearly that it is preferable to transmit multiple measurements simultaneously in energy-limited communication settings.However, when we directly transmit uncoded quantization bits, there is no preference between transmitting a single or multiple measurements.Certainly, judicious selection of N Q or/and N should minimize the f code (N Q , N) bound to ensure reliable performance in reconstruction.To this end, let us explore further the characteristics of f code (N Q , N) with large L. As long as R < ln 2 − H( ), we know that E r (R, ) > 0, see (44).Thus, as L→∞, we have Equation ( 49) implies that the number of channel uses to transmit one measurement can be optimally chosen to be Accordingly, the optimal number of quantization bits is given by (48).
In Section 5, we will study how L, the number of simultaneously transmitted measurements, affects the achievable distortion in source reconstruction with numerical examples.
NUMERICAL EXAMPLES
In this section, we provide numerical examples to corroborate the analytical results we derived in the previous sections.
Optimal number of quantization bits
As discussed in Sections 2.1 and 2.3.1, when the total energy budget is uniformly allocated among quantization bits, there is an optimal value of N which minimizes the meanabsolute reconstruction error upper bound both when the circuit energy is neglected and also when circuit energy consumption is accounted for.Here, we first consider the channel to be AWGN and use BPSK modulation.The BSC crossover probability as a function of the bit energy-to-noise ratio is (γ) = Q( 2γ). Figure 3 depicts the bound f (N) in (5) together with the simulated actual mean-absolute reconstruction error E|A− A|, and the bound f c (N) in ( 21) with E /N 0 = 20 and T 0 P on /N 0 = 1.It can be seen that the bound f (N) is pretty tight and numerical minimization yields N opt = 7 in the first case and N opt = 6 in the second case.In Figure 3, we also plot f (N) and f c (N) when (γ) = (1/2)e −γ/2 , which is the BER when binary orthogonal modulation is used along with envelope detection; N opt here turns out to be 5 and 4, respectively.
Optimal bit energy allocation
In Section 2.2, we derived an optimal energy allocation scheme per bit to minimize the reconstruction error.Considering envelope detection of binary orthogonal signals as in Section 2.2.2, with E /N 0 = 20 and N = 10, we can find the optimal energy allocation by solving the convex optimization problem in (8) using the interior-point method outlined in [16, Chapter 11]; Figure 4 depicts the result.Mean absolute error upper bound Simulated mean absolute reconstruction error For the same (γ) = (1/2)e −γ/2 , Figure 5 compares the reconstruction error between the optimal energy allocation scheme in (8) and the equal energy distribution scheme in (5) with a different number of quantization bits N. We observe that the reconstruction error decreases to a floor as N increases with optimal energy allocation, which is different from the equal energy allocation scheme.The intuitive explanation for this behavior is that as N increases, equal energy allocation increases the cross-over probability for all transmitted bits; on the other hand, optimal energy allocation does not experience this problem.As already noticed in Figure 4, when N is large enough, the optimal scheme just assigns no (or very little) energy to less significant bits.
In Figure 5, we also plot the reconstruction error as a function of N when circuit energy consumption is taken into account with optimal allocation of the residual energy to the quantization bits as in (23); here we take T 0 P on /N 0 = 1.The optimal number of quantization bits in (24) is easily seen to be N opt = 6.
Optimal energy allocation among sensors
Suppose that K = 10 sensors are deployed with local observation noise variances denoted by σ 2 1 , σ 2 2 , . . ., σ 2 10 , the path loss exponent of the wireless channel is κ = 2 (free space), and accordingly, the cross-over probability is given by k , where d k is the distance between sensor k and the fusion center.Parameter C is set to be 1 here.In the following, we set the total energy budget to be E T /N 0 = 200.
Identical
Using the aforementioned parameters, Figure 6 compares the normalized value of the objective function in (31) between the equal energy allocation and the optimal energy allocation scheme for a variable number of bits N while choosing a specific set of values for {d k } 10 k=1 and {σ 2 k } 10 k=1 .For this particular setup, the optimal value of N in (37) turns out to be N opt = 6.With different sets of values for {d k } 10 k=1 and {σ 2 k } 10 k=1 , when N is accordingly chosen to be optimal, the corresponding optimal energy allocation schemes, that is, the numerical solutions of the convex problem in (31), are depicted in Figure 8.
Jointly optimized {N
As explained in Section 3.2, we can find the optimal N k and x k , k = 1, . . ., K, jointly by utilizing Algorithm 1.The resulting optimal energy allocation scheme and the optimal number of quantization bits per sensor are depicted in Figures 8 and 9, respectively.Through joint optimization, the normalized minimum value of the objective function in (38) is plotted in Figure 6.The gain over the case where each sensor transmits the same number of quantization bits is clear.Furthermore, we have also plotted the simulated mean-square estimation error of different schemes in Figure 7, which again demonstrates the benefits of energy and quantization optimization.
Single measurement transmission
With envelope detection of binary orthogonal signals, the underlying BSC's cross-over probability is (γ) = (1/2)e −γ/2 , where γ denotes the bit energy-to-noise ratio.Assuming a total energy budget E /N 0 = 100, the reconstruction error upper bound in (41) is depicted in Figure 10, where we also plot the optimal bounds achieved with uncoded transmission schemes (cf.( 5) and ( 8)).From these plots, it is evident that the bound f code (N Q , N) derived for randomly coded transmission is not tight and is easily achieved with uncoded transmissions.
Multiple measurements transmitted together
Keeping the same set of parameters, the bound provided in (46) is plotted in Figure 11 for different values of L. We observe that with optimal choices of N Q and N, the uncoded performance bound is easily achieved by transmitting multiple source samples together with appropriate channel coding.The performance gap between uncoded and coded cases can be large.In this particular example, the reconstruction error with coding can be less than 1% of the one without coding.
CONCLUSIONS
Motivated by stringent energy requirements that are prevalent in wireless sensor networks, we have pursued optimal quantization at sensor nodes to effect optimal reconstruction at the fusion center.(De)modulation and propagation were captured through the use of the classical binary symmetric channel model.In particular, we have found the number of quantization bits for minimizing the mean-absolute reconstruction error bound in a point-to-point link when each bit is allocated the same energy.We also derived an optimal scheme for energy allocation across quantization bits.Both transmission energy as well as circuit energy were considered.When multiple sensors collaborate to estimate a parameter in noise, we also obtained the optimal energy allocation scheme to distribute the limited total energy among different sensors when each sensor assigns the same energy to all its quantization bits.It turned out that this allocation scheme depends on the prescribed number of quantization bits {N k } K k=1 , and thus the optimal {N k } K k=1 can also be found with the help of our convex optimization formulation.We also studied the effects of channel coding on energy constrained quantization and optimized the number of quantization levels and the number of channel uses per source sample.where we have used the property that (0) = 1/2.From (A.3) and (A.4), it follows readily that f 0 (x * N ; N) < f 0 (x * N ; N).
Figure 1 :
Figure 1: System model of a single sensor's quantized measurement received through a BSC at the fusion center.
Figure 2 :
Figure 2: Multisensor cooperation in estimating a scalar parameter with quantized observations.
Figure 3 :
Figure 3: The bound of E|A − A| whose minimum yields the optimum number of quantization bits.
Figure 4 :
Figure 4: Optimal energy allocation over a fixed number of quantization bits (N = 10).
Figure 5 :
Figure 5: (a) Reconstruction error with optimal energy allocation among bits as in (8).(b) Reconstruction error with equal energy allocation as in (5).(c) Reconstruction error in (23) taking into account the energy consumption of circuit electronics.
NQ f code (N Q , N) (b) | 2014-10-01T00:00:00.000Z | 2004-10-04T00:00:00.000 | {
"year": 2004,
"sha1": "8a83d4a66189c65c48e4ce016e6519cf3aa4a815",
"oa_license": "CCBY",
"oa_url": "https://asp-eurasipjournals.springeropen.com/counter/pdf/10.1155/2008/462930",
"oa_status": "GOLD",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "13a48a33219357f9324e4fb7f8c1dfd395d30897",
"s2fieldsofstudy": [
"Computer Science"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
245939949 | pes2o/s2orc | v3-fos-license | Exploring Long Non-Coding RNAs Associated with IP3/DAG Signaling Pathway as Potential Biomarkers Involved in Mast Cell Degranulation in Chronic Spontaneous Urticaria with 2-Year Follow-Up
Purpose Chronic spontaneous urticaria (CSU) pathogenesis involves mast cell degranulation induced by the inositol 1,4,5-trisphosphate/diacylglycerol (IP3/DAG) pathway, but the condition lacks specific biomarkers. This study was performed to investigate long non-coding RNA (lncRNA) expression profiles, identify those associated with IP3/DAG pathway, and assess their diagnostic and prognostic value for CSU. Methods Ten samples were selected from CSU and control groups, and microarray was performed to screen differentially expressed (DE) lncRNAs and mRNAs. Bioinformatic and co-expression network analyses were used to identify lncRNAs associated with IP3/DAG pathway. Quantitative real-time polymerase chain reaction was used to validate lncRNA expression levels. Combined with disease characteristics and serum indices detected with enzyme-linked immunosorbent assays, Spearman analysis and logistic regression were applied to analyze lncRNA-associated disease risk. Receiver operating characteristic (ROC) curves and 2-year follow-up data were applied to evaluate lncRNA diagnostic and prognostic value. Results A total of 678 up- and 573 downregulated DE lncRNAs and 609 up- and 176 downregulated DE mRNAs were identified. Seven lncRNAs (upregulated T264761, T280622, ENST00000587970, T224062, ENST00000562459, and his-1_RNA_dna; downregulated ENST00000417930) were associated with the IP3/DAG pathway. D-dimer and histamine levels were significantly different between the two groups. Correlation analysis showed that his-1_RNA_dna positively correlated with the frequency of symptom appearance, while his-1_RNA_dna, ENST00000417930, T264761, and T280622 negatively correlated with the maximum wheal diameter. Regression analysis showed T264761 was associated with CSU risk. ROC analysis showed that the specificity of T264761 was 90%, with an area under the curve of 0.666. In follow-up, the rate of well-controlled disease in the low T264761 expression group was 82.61%. Conclusion This study established lncRNA and mRNA expression profiles in CSU and identified lncRNAs associated with IP3/DAG pathway, which is mechanistically involved in this disease. T264761 may be a novel biomarker for CSU, but further study is needed to confirm its specific mechanism.
Introduction
Chronic spontaneous urticaria (CSU) is a common dermatological condition characterized by sudden itchy wheals or angioedema lasting for up to 6 weeks and occurring twice a week or more. 1 The symptoms are spontaneous, with no known specific triggers. CSU diagnosis is mainly based on clinical symptoms and medical history, the 7-day Urticaria Activity Score (UAS7), quality of life assessments, and other questionnaires that assist in evaluation of the disease and treatment effects. Some reviews suggest that there are statistical differences in the contents of D-dimer, C-reactive protein (CRP), and microRNAs between patients with CSU and healthy controls. These factors may reflect disease course and activity and treatment effects, but they are not specific to CSU and therefore cannot sensitively or accurately distinguish CSU from other diseases. [2][3][4] Specific biomarkers of CSU that may facilitate early diagnosis, accurate prognosis, and individualized treatments remain elusive.
Urticaria is a mast cell-driven disease. The highaffinity receptor for immunoglobulin E (IgE) (FcεRI) is considered a key factor in mast cell degranulation, which induces characteristic skin wheals in CSU. Antigendependent mast cell activation is regulated by a complex series of intracellular signaling processes that are initiated following FcεRI aggregation. 5 In downstream signaling processes, phospholipase Cγ1/2 hydrolyzes phosphatidylinositol 4,5-bisphosphate in the cell membrane to produce two second messenger molecules: inositol 1,4,5-triphosphate (IP3) and diacylglycerol (DAG). 6 IP3 induces calcium release from the endoplasmic reticulum, triggering extracellular calcium influx; DAG binds to protein kinase C, which is activated by cell membrane and cytoskeleton proteins. 7 IP3 and DAG combine to trigger mast cell degranulation, increase microvascular permeability, and induce the release of inflammatory mediators such as histamine (HIS), leukotriene B4 (LTB4), prostaglandin D2 (PGD2), and mast cell tryptase (MCT). 8 Additionally, the mitogen-activated protein kinase and phosphoinositide 3-kinase pathways are mainly involved in generating eicosanoids and cytokines. 5 The IP3/DAG signaling pathway is thus a key pathway in mast cell degranulation. Currently, there are no known biomarkers for this signaling pathway.
Long non-coding RNAs (lncRNAs) are a group of functional RNAs which non or rarely in coding transcripts that longer than 200 nucleotides. 9 LncRNAs regulate gene expression at epigenetic and transcriptional levels in the nucleus and at post-transcriptional and translational levels in the cytoplasm. LncRNAs are closely related to the physiology and pathology of allergic diseases including asthma, atopic dermatitis, and rhinitis, suggesting that they could act as biomarkers and therapeutic targets for allergic disorders, aid in the diagnosis and prognosis of disease, and identify novel therapeutic agents. 4 Some lncRNAs were shown to be closely related to skin physiology and pathologies. For example, COL1A2-AS1 promotes the apoptosis of normal skin fibroblasts by inhibiting p-Smad3 and promoting β-catenin expression, WAKMAR1 regulates wound healing by promoting keratinocyte migration, and MEG3 affects the proliferation and apoptosis of psoriatic epidermal cells by targeting miR-21/ caspase-8. [10][11][12] Since CSU pathogenesis is closely related to allergic responses, differential lncRNA expression may therefore serve as a biomarker for this disease.
This study aimed to identify lncRNAs that are closely related to CSU physiology and pathology. LncRNA and mRNA expression profiles in blood samples from CSU patients were compared with those in healthy individuals using microarray technology; bioinformatic analyses, quantitative real-time polymerase chain reaction (qRT-PCR), receiver operating characteristic (ROC) curve analysis, logistic regression analysis, and a 2-year clinical follow-up were used to explore the value of key lncRNAs as biomarkers for CSU.
Recruitment and Inclusion/Exclusion Criteria
Patients were recruited from the affiliated Jiangmen Traditional Chinese Medicine Hospital of Jinan University between November 2018 and May 2019. All participants had a documented history of CSU. Patients suffering from the spontaneous appearance of wheals and/ or angioedema for ≥6 weeks due to known or unknown causes, and who were aged 18 years or older at the time of recruitment were included in the study. Patients with blood system diseases, tumors or other serious diseases (severe cardiovascular, hepatic, or renal insufficiency), pregnant or lactating women, and those with current drug abuse, alcoholism, or mental illness were excluded. Healthy controls were included in the study if they had no obvious abnormalities on routine physical examination and laboratory imaging, no history of disease or drug treatment before blood collection, and were aged 18 years or older. All subjects signed written informed consent forms prior to participating in this study.
The present study was approved by the Ethics Committee of the affiliated Jiangmen Traditional Chinese Medicine
Blood Sample Collection
Two vacutainer tubes of peripheral blood were collected from all participants. One tube contained ethylene diamine tetraacetic acid (EDTA) and was used for lncRNA microarrays or qRT-PCR, the second standard tube was used for enzyme-linked immunosorbent assay (ELISA). Peripheral blood mononuclear cells (PBMCs) in the EDTA tube were isolated by density gradient centrifugation, and total RNA was isolated from the PBMCs using TRIzol reagent. Blood serum was extracted after centrifugation in the standard tube.
lncRNA Profiling
The Arraystar Human lncRNA microarray v4.0, which contains 40,173 lncRNAs and 20,730 coding transcripts, was used to profile the expression of lncRNAs and mRNAs. Five samples from CSU patients and 5 samples from healthy controls were selected to be detected by Shanghai Kangchen Bio-tech, Inc. (Shanghai, China). After the total RNA was isolated, the concentration and quality were assessed using absorbance spectrometry, measuring absorbance ratios of A260/A280 and A260/ 230 using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Supplementary Table 1 shows RNA quantification and quality assurance, and Supplementary Figure 1 shows an assessment of RNA integrity and gDNA contamination test by denaturing agarose gel electrophoresis. According to the manufacturer's instructions and as in previously published research, 13 RNA samples were labeled by Quick Amp Labeling Kit (Agilent p/n 5190-0442, Santa Clara, CA, USA), and purified by RNeasy Mini Kit (Qiagen p/n 74,104, Hilden, Germany). Microarray hybridization was performed using the Agilent Gene Expression Hybridization Kit (Agilent p/n 5188-5242). After washing and fixing, the hybridized arrays were scanned by Agilent Microarray Scanner (Agilent p/n G2565BA) and data were extracted using Agilent Feature Extract Software (Agilent, v10.5.1.1, Palo Alto, CA, USA). Data were deposited in the Gene Expression Omnibus database (GSE185516). Raw data were normalized using a quantile algorithm in Agilent GeneSpring GX v12.1. Differentially expressed (DE) lncRNAs and mRNAs were identified according to their fold change (FC) at a threshold ≥1.5-fold with significance p < 0.05.
Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway Analyses
The GO and KEGG databases were used to analyze potential biological functions and signaling pathways associated with DE mRNAs. GO enrichment analysis of target genes was performed by GOseq R package (http://www.geneon tology.org) and included biological processes (BP), cellular components (CC), and molecular functions (MF). Signaling pathways related to target genes were explored using the KEGG database (http://www.geneontology.org) and performed using DAVID software. Target genes with p < 0.05 were considered significantly enriched.
Pearson Analysis of DE lncRNAs and mRNAs Associated with IP3/DAG Signaling Pathway
Pearson correlation analysis was carried out on selected DE mRNAs and lncRNAs identified in the KEGG analysis; those with a Pearson correlation coefficient (PCC) ≥ 0.8 with significance p < 0.5 and a false discovery rate (FDR) ≤ 1 were selected for further exploration of closely related DE lncRNAs.
Validation of lncRNAs Using qRT-PCR and lncRNA-mRNA Co-Expression Networks
LncRNAs for which the p value of significance is lower, FC is higher, raw intensity is greater than 200, and which were closely related to the IP3/DAG pathway were selected for qRT-PCR validation. Total RNA was extracted using TRIzol reagent, and cDNA was synthesized using EasyScript First-Strand cDNA Synthesis SuperMix (TransGen, Beijing, China). qRT-PCR was performed using SYBR Green (Yeasen, Shanghai, China). The specific primers were designed by LandM Biotech Inc. (Guangzhou, China). Cycle threshold (Ct) values were used to quantify the expression levels of lncRNAs using the 2 −ΔΔCt method, and β-actin was used as a control to normalize values.
Selected lncRNAs and total mRNAs with PCC ≥ 0.9, p < 0.5, and FDR ≤1, were used to construct a coexpression network which was visualized using Cytoscape v2.8.3 (Institute of Systems Biology, Seattle, WA, USA). GO analysis was used to assess the putative biological functions of the co-expressed mRNAs, which likely reflect the biological function of the lncRNAs.
Statistical Analysis
All data are presented as mean ± standard deviation (SD), median (range), or n, as appropriate. Chi-square and Mann-Whitney U-tests were used to analyze qualitative and quantitative data, respectively. Spearman coefficient analysis was performed to assess correlations between lncRNA levels and clinical characteristics of CSU. Logistic regression was used to analyze the CSU risk associated with selected lncRNAs. ROC curves were generated to evaluate the prognostic value of selected lncRNAs. lncRNA expression values were sorted into high-and low-level groups, based on these ROC curves using Youden's index correction. 14 Data were analyzed using SPSS v19.0 (IBM Corp., Armonk, NY, USA) and graphs were drawn using GraphPad Prism v8.3.0 (San Diego, CA, USA). p values < 0.05 were regarded as statistically significant. Figure 1 is a flow diagram summarizing the study methods, participant inclusion and exclusion, and key results.
Patient Demographics
In total, 59 people with CSU and 59 healthy controls were recruited for this study; a summary of their demographic characteristics is shown in Table 1.
LncRNA and mRNA Expression Profiles
A total of 1482 lncRNAs were found to be upregulated and 1775 downregulated with FC ≥ 1.5 in CSU patients versus healthy control individuals ( Figure 2A). Of these up-and downregulated lncRNAs, 678 and 573, respectively, were significantly differentially expressed (FC ≥ 1.5, p < 0.05). ENST00000562459 was upregulated to the greatest extent, with an FC of 6.4280631, while NR_122077 was downregulated to the greatest extent, with an FC of 5.1422487 ( Figure 2B, Table 2). LncRNA expression profiles were distinguishable between CSU patients and healthy control individuals according to the hierarchical clustering ( Figure 2C).
A total of 1072 mRNAs were upregulated and 549 downregulated with an FC ≥ 1.5 in CSU patients versus healthy controls ( Figure 3A). Of these up-and downregulated mRNAs, 609 and 176, respectively, were significantly differentially expressed (FC ≥ 1.5, p < 0.05). NM_033130 was upregulated to the greatest extent with an FC of 4.2383847, while NM_002538 was downregulated to the greatest extent with an FC of 11.41218 ( Figure 3B, Table 3). mRNA expression profiles were distinguishable between CSU patients and healthy controls according to the hierarchical clustering ( Figure 3C).
We constructed a Circos graph to visualize the chromosomal distributions and classifications of the dysregulated lncRNAs and mRNAs ( Figure 4). There were 114 lncRNAs in chromosome 2 and 77 mRNAs in chromosome 1.
GO and KEGG Pathway Analyses
Among upregulated target genes, GO analysis showed that vesicle-mediated transport had the highest enrichment score in the BP category ( Figure 5A), cytoplasmic vesicle in the CC category ( Figure 5B), and carbohydrate binding in the MF category ( Figure 5C). Among downregulated target genes, establishment of epithelial cell apical/basal polarity had the highest enrichment score in the BP category ( Figure 6A), the apical part of the cell in the CC category ( Figure 6B), and intracellular chloride channel activity in the MF category ( Figure 6C).
KEGG pathway analysis of all upregulated target genes showed enrichment in 19 pathways including lysosome, osteoclast differentiation and human cytomegalovirus infection, and vascular endothelial growth factor (VEGF) signaling pathways (Supplementary Figure 2, Table 4). Of these, the VEGF pathway was the most closely related to the IP3/ DAG pathway (Figure 7), and granules secreted by mast cells contain cytokines including VEGF. 15 VEGF increases microvascular permeability, which is partially consistent with the pathogenesis of urticaria, so we hypothesized that the DE mRNAs in this pathway might be closely related to disease occurrence. Downregulated target genes were enriched in four pathways including salmonella infection, tight junction, and cytokine-cytokine receptor pathways
DE lncRNAs Related to the IP3/DAG Pathway
According to KEGG analysis, the VEGF pathway was the most closely related to IP3/DAG pathway. There were five DE mRNAs in the VEGF pathway in our analysis: BAD, PIK3CD, PLA2G4A, PPP3CA, and SRC. Pearson analysis of these five mRNAs and all DE lncRNAs revealed that there were 514 upregulated and 488 downregulated DE lncRNAs related to the 5 VEGF pathway mRNAs (PCC ≥ 0.8, p < 0.5, FDR ≤ 1). Supplementary Table 3 shows the most closely related lncRNAs based on PCC values.
These 7 lncRNAs and 959 mRNAs were used to construct a co-expression network ( Figure 8). GO analysis revealed those co-expressed mRNAs were most enriched in the regulation of intracellular transport within the BP category, cell leading edge within the CC category, and receptor ligand activity within the MF category ( Figure 9). These biological functions are closely related to the process of mast cell degranulation, indicating that the lncRNAs were indeed related to CSU development.
Correlation of lncRNAs with Clinical Characteristics and Inflammatory Mediators
Samples from 56 CSU patients and 13 healthy controls were selected for ELISA, which showed increased Table 6). Spearman coefficient analysis was used to assess the correlations of six lncRNAs (T264761, T280622, ENST00000587970, ENST00000562459, his-1_RNA_dna, and ENST00000417930) with clinical characteristics (age, disease duration, frequency of symptom occurrence, duration of wheals, size of wheals, and UAS7 score) and inflammatory mediators (D-dimer and HIS). A positive correlation was observed between his-1_RNA_dna and the frequency of symptom occurrence, and negative correlations were noted between four lncRNAs (T264761, T280622, his-1_RNA_dna, ENST00000417930) and the maximum size of wheals ( Figure 12, Supplementary Table 7).
ROC Curves of Selected lncRNAs for CSU Risk Prediction
Logistic regression was used to analyze the CSU risk associated with selected lncRNAs. Six lncRNAs with significant differences in univariate analysis were included as independent variables, with CSU occurrence as the dependent variable (1 = occurrence, 0 = no occurrence). ENST00000417930 (odds ratio [OR] = 0.385; 95% confidence interval [CI] = 0.156-0.948) and T264761 (OR = 1.266, 95% CI = 1.023-1.567) were identified as risk factors for CSU (Table 5). Furthermore, at the cut-off value of 4.138, the area under the ROC curve for T264761 was 0.666, and the sensitivity and specificity values were 49.06% and 90%, respectively ( Figure 13, Supplementary Table 8). This finding suggested that the T264761 expression level may be useful for differentiating CSU patients from healthy controls.
Prognostic Value of T264761
The last follow-up data was collected in 4th September 2021. A total of 47 patients were included in this analysis. According to a previous study in China, a Urticaria Control Test (UCT) score ≥12 indicates wellcontrolled CSU. 16 Based on a T264761 expression level cut-off value of 4.138, CSU patients were sorted into high and low expression groups. 14 The rate of well-controlled disease in those with low T264761 expression was 82.61% compared with 54.17% in those with high T264671 expression. Table 6 shows that lower T264671 expression could predict higher disease control rates.
Discussion
In this study, we identified several DE lncRNAs and mRNAs in CSU patients compared with healthy controls using microarray sequencing. Some were found to be closely related to IP3/DAG signaling pathway, a key pathway in the induction of mast cell degranulation, which is involved in CSU. Importantly, we identified a potential biomarker for this disease, where no known biomarkers existed previously: lncRNA T264761 was found to have diagnostic and prognostic value for CSU in our study.
Upregulated lncRNAs and mRNAs outnumbered those that were downregulated in CSU samples compared with healthy controls, indicating the activation of multiple biological processes or signaling pathways under pathological conditions. In our study, we identified DE lncRNAs using a FC threshold of 1.5, compared to 2 in other studies. 13,17,18 A lower threshold in this study may be related to the characteristic CSU symptoms of sudden itchy wheals or angioedema. Most of the CSU patients were in remission, and were therefore symptomatically similar to healthy control individuals except for those who experienced sudden attacks, so the overall differential expression may have been relatively low.
DE mRNAs were analyzed using GO term and KEGG pathway enrichment analyses. GO analysis showed that most upregulated mRNAs were associated with vesiclemediated transport within the BP category, cytoplasmic vesicle within the CC category and carbohydrate binding within the MF category. This pattern is consistent with the process of mast cell degranulation, which involves the secretion of mediators via a vesicle secretion system. 19,20 Mast cells also require carbohydrates for granule generation and maintenance. 21 Our KEGG analysis showed that the VEGF signaling pathway was the most closely related to the occurrence of urticaria. VEGF is a mediator of vascular permeability that could be involved in inducing and maintaining urticaria symptoms; the IP3/DAG pathway is known to be involved in urticaria pathology and is included in the pathway diagram in Figure 7. 22 We therefore identified five DE mRNAs in the VEGF pathway that were used to identify seven lncRNAs closely related to the IP3/DAG pathway (T264761, T280622, ENST0000 0587970, T224062, ENST00000562459, his-1_RNA_dna, and ENST00000417930). Co-expression network and GO analyses confirmed that these lncRNAs were closely related to the development of CSU. Co-expressed mRNAs were found to be enriched in the regulation of intracellular transport within the BP category, cell leading edge within the CC category, and receptor ligand activity within the MF category. Many high-affinity receptors for IgE are expressed on the mast cell surface, and the intracellular transportation and release of pro-inflammatory mediators following antigen-induced aggregation causes allergic reactions. 5,23 Our qRT-PCR results suggested that T224062 was not significantly dysregulated in CSU patients, which contradicted our microarray sequence data; however, this may be due to the relatively small sample size.
Given that our bioinformatics analysis revealed putative associations between several lncRNAs and CSU, we then assessed their correlations with clinical characteristics of the disease to determine their usefulness as disease indicators. Our ELISA results revealed that D-dimer and HIS expression were increased in CSU. Previous studies showed a positive correlation between D-dimer levels and the severity of CSU activity. 24,25 However, we found no correlation between the seven lncRNAs we identified in our bioinformatics analysis and D-dimer or HIS levels. Some of the lncRNAs were correlated with the frequency of symptom occurrence and the maximum size of wheals. It suggested that lncRNAs did not seem to be promising biomarkers for disease activity. Logistic regression showed that high levels of T264761 could indicate a greater risk of CSU. The ROC curve analysis suggested that T264761 may differentiate CSU patients from healthy control individuals with high specificity; however, the sensitivity was not high. A combination of medical history and symptom evaluation could improve its diagnostic value. We performed a 2-year follow-up to assess the prognostic value of T264761 in CSU patients based on the UCT score. This assessment was the first valid and reliable tool to assess disease control in patients with CSU. 26 This test was rarely used in CSU clinical research in China before 2020; however, Yu et al translated the UCT into Chinese and assessed the reliability, validity, sensitivity, and screening accuracy of the new scale, enabling its application in China. 16 In our follow-up, lower T264761 expression (≤4.138) could predict a higher disease control rate. Notably, in five women, CSU resolved after pregnancy during the follow-up period; further studies of lncRNA levels in pregnant and postnatal women with CSU could help identify the underlying mechanism.
Many recently developed techniques have enabled the identification of potentially useful CSU biomarkers, including proteomic analysis, transcriptome analysis, microRNA or gene sequencing, and gut microbiome analysis. [27][28][29][30][31] However, this study is the first to assess the diagnostic and prognostic value of lncRNAs as CSU biomarkers using microarrays, bioinformatics analyses, and co-expression network construction. T264761 was found to be related to the IP3/DAG signal pathway, which is known to induce mast cell degranulation that is involved in CSU. This lncRNA was also associated with the maximum size of the wheals that are symptomatic of this disease. High levels of T264761 may indicate an increased CSU risk with high specificity. This lncRNA may also predict longer-term disease control status. This study provides novel insights into CSU biomarkers and guide further basic and clinical research. Our results should be considered in the context of some limitations. First, the samples were only from CSU patient and healthy controls; in future studies we will include more samples including those from subjects acute urticaria and other similar allergic diseases as controls to further investigate the diagnostic value of lncRNAs. Secondly, the frequency and duration of follow-up were relatively short, because urticaria has a self-healing tendency without treatment, so the findings should be interpreted with caution. Thirdly, the results of this study are mostly based on bioinformatics predictions, and further cell culture and animal experiments are needed to clarify the specific roles of lncRNAs in CSU. Additionally, it is worth mentioning that although some CSU patients included were complicated with chronic idiopathic urticaria, the results are still plausible. This is consistent with clinical practice and other relevant CSU clinical literatures. 28,32
Conclusions
Our study established lncRNA and mRNA expression profiles in CSU using microarrays. Some lncRNAs and mRNAs were differentially expressed in CSU patients compared with healthy controls; these were shown to be involved in diverse biological pathways related to mast cell degranulation, which is associated with CSU, and the IP3/DAG signal pathway that is involved in producing the characteristic wheals and itchiness associated with CSU. The lncRNA T264761 may be a clinically useful novel diagnostic biomarker for CSU, with lower levels predicting a better UCT score and thus indicating less severe disease. Further investigations are required to clarify the specific function of T264761 in CSU pathogenesis.
Data Sharing Statement
Supporting data used and/or analyzed during the current study are available from the corresponding author on reasonable request. | 2022-01-15T16:10:12.386Z | 2022-01-01T00:00:00.000 | {
"year": 2022,
"sha1": "c207f242e1450d65308bc1dad0974ea0e87a49ee",
"oa_license": "CCBYNC",
"oa_url": "https://www.dovepress.com/getfile.php?fileID=77635",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "16075a51bea97b0a0c3f2e8af5b0dc0cb8b4abcd",
"s2fieldsofstudy": [
"Biology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
253205165 | pes2o/s2orc | v3-fos-license | CircuitAssist: Automatically Dispensing Electronic Components to Facilitate Circuit Building
When learning how to build circuits, one of the challenges novice makers face is how to identify the components needed for the circuit. Many makerspaces are stocked with a variety of electronic components that look visually similar or have similar names. Thus, novice makers may pick the wrong component, which creates a non-functional circuit although the wires are correctly connected. To address this issue, we present CircuitAssist, an actuated electronics component shelf connected to a tutorial system that dispenses electronic components for the maker in the order that they occur in the tutorial. The shelf contains dispensers for each component type with a custom release mechanism actuated by a servo motor that dispenses one component at a time. Makers can work with CircuitAssist by either following one of the provided tutorials or by directly selecting a component they need from the user interface.
INTRODUCTION
Over the last decade, makerspaces have become increasingly popular allowing a growing number of people to learn skills, such as electronics circuit building. An increasing number of learning resources, such as tutorials with step-by-step circuit building instructions, further support novice makers in mastering electronics.
To help novice makers improve their skills and avoid mistakes in their circuit, researchers have developed a range of support tools. Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for proft or commercial advantage and that copies bear this notice and the full citation on the frst page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s For instance, CircuitStyle [1] helps makers improve their breadboard layout by showing best wiring practices and ToastBoard [2] supports makers by showing false wiring on the breadboard.
While these works support makers in wiring up the circuit, an earlier step that makers have to go through is to choose the correct components for the circuit. In makerspaces that are well stocked with a range of electronic components, there are hundreds of different parts that look similar and have similar names, which can lead to confusion. While systems, such as CurrentViz [6], inform a maker if a wrong component is plugged into the breadboard, it would be benefcial to prevent such mistakes in the frst place by helping users identify the correct component ahead of time.
To address this issue, we present CircuitAssist, an actuated electronics component shelf that dispenses the correct components for the maker either based on the order they occur in a tutorial or based on the maker manually choosing components in a user interface. CircuitAssist supports makers in this frst stage of circuit building, i.e. choosing the right components, and integrates well with existing tools, such as augmented breadboards, that support the later stages of wiring up the components on the breadboard.
RELATED WORK
Our work is related to systems that assist makers in electronics prototyping. One body of work has focused on augmenting digital user interfaces to help users plan circuit layouts. AutoFritz [4] auto-completes a maker's circuit wiring on a digital breadboard, while CircuitStyle [1] improves makers' circuit style by showing best practices. Since mistakes can also be introduced when physically building a circuit, augmented breadboards, such as Toast-Board [2] and CurrentViz [6], visualize where makers made mistakes. SchemaBoard [3] takes this a step further by assisting makers in translating abstract circuit schematics into breadboard component layouts by highlighting where components should be placed. Other projects, such as Proxino [5], support remote collaboration during circuit construction, which can help an experienced maker correct mistakes on a novice makers' breadboard. Our research complements these existing works by supporting novice makers in choosing the correct components for circuits.
CIRCUITASSIST
CircuitAssist is an actuated shelf that dispenses electronic components either based on the order they occur in a tutorial or the makers manual selection.
Hardware: CircuitAssist is a shelf that consists of an array of individually actuated dispensers. The dispensers are 3D printed (Ender3, PLA flament, 0.1mm) and contain electronic components stacked on top of each other in their main body (Figure 2a). The release mechanism consists of a set of gears connected to a servo motor (SG90) that actuates a rotating cylinder inside the dispenser (Figure 2b). The rotating cylinder is sized to allow for one component in it at a time. When the servo rotates the cylinder, the electronic component is dropped out of the dispenser. When the cylinder rotates back up a component from the dispenser's main body falls into the cylinder ready to be released upon the next actuation. The servo motors for all dispensers are connected to a microcontroller (Arduino Uno), which allows CircuitAssist's user interface to control which dispenser should release a component. Our prototype shelf currently contains 10 dispensers and is 38cm wide and 15cm high. The modular design of CircuitAssist allows makerspaces to easily expand the shelf with additional dispensers as needed. Software: CircuitAssist's user interface (built in Processing) provides two modes of interaction: Makers can either choose to follow a tutorial, which releases the components according to the current tutorial step, or choose manual selection mode, which provides the user with a set of buttons to select components. When CircuitAssist determined that a component needs to be released, it sends a command to the microcontroller that contains the pin number that is connected to the corresponding dispenser's servo motor.
APPLICATION EXAMPLES
We built three tutorials to showcase CircuitAssist's functionality.
LED circuit: The frst tutorial shows a novice maker how to build a circuit with an LED that lights up when a button is pressed. When the maker starts the tutorial the LED, a matching resistor, and a button are dispensed. We included this example since novice makers often confuse resistors with each other even when told the required resistance since they look similar. Choosing the wrong resistor either leads to a broken LED or an LED that does not light up, which can be frustrating. CircuitAssist prevents this mistake by dispensing the correct resistor for the maker.
Capacitor Delay Circuit: The second circuit builds on the frst but also allows the LED to stay on for a while. When the button is pressed a capacitor charges and powers a transistor, completing the LED's circuit and lighting it up. After the button is released, the capacitor slowly drains through the 1k resistor. When it is drained, the transistor is no longer powered and the LED turns of. We included this example since it uses 2 resistors (one for the capacitor, one for the LED) of 2 diferent values (1k, 330ohm) which can easily be confused with each other. If the resistors are swapped in the circuit the LED would not stay on and not be as bright. CircuitAssist prevents this mistake by dispensing one resistor at a time, i.e. frst the resistor for the capacitor, and then later the LED resistor.
LED Flasher: Our fnal circuit extends the previous circuit by fashing the LED on and of. To accomplish this, the circuit uses a second capacitor, which can be easily confused with the one used in the previous step. CircuitAssist prevents this mistake by dispensing the correct capacitor at the specifc step in the tutorial.
DISCUSSION
We next discuss limitations and directions for future work.
Reflling Components: Our system requires the makerspace manager to restock the dispensers by inserting components one at a time, which takes more time than reflling a conventional shelve.
Number of Components: Our current system supports up to 10 diferent components. However, more dispensers can easily be added to the shelf since it is modular.
Integrating Existing Tutorials: For future work, we will interface with existing circuit building software, (e.g., Fritzing) to automatically dispense components based on digital breadboard layouts.
Tracking Number of Components: By counting how many components of each type have already been dispensed, our system can alert the makerspace manager to restock those components.
Potential to Improve Accessibility: While not the main focus of our work, our system might be helpful for makers with vision impairments who are not able to identify components visually.
CONCLUSION
We presented CircuitAssist, an actuated electronics component shelf that helps makers with their circuit building by dispensing the electronic components they need for their circuit. We discussed the hardware and software required to make CircuitAssist and showed how CircuitAssist can help at the example of three tutorials that contain components that can be easily confused by makers. Finally, we discussed limitations and future work of our approach, including CircuitAssist's scalability, connection to existing circuit building tutorials, and its potential application to improve accessibility. | 2022-10-30T13:33:55.815Z | 2022-10-28T00:00:00.000 | {
"year": 2022,
"sha1": "afd37fd97a8eba827a86ca427ef4a36eb192615d",
"oa_license": "CCBYNC",
"oa_url": "https://dspace.mit.edu/bitstream/1721.1/146513/1/3526114.3558671.pdf",
"oa_status": "GREEN",
"pdf_src": "ACM",
"pdf_hash": "5e973c9246337d112c6eef6dec4a96d4d7a93ee2",
"s2fieldsofstudy": [
"Computer Science"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
70940169 | pes2o/s2orc | v3-fos-license | Decision Support by Visual Incidence Anamneses for Increased Patient Safety
1.1 How to get the right information? Decision support in Health Care is based on proper collection and evaluation of relevant information related to the patient and her/his medical history. In the process of Anamnesis the patient will provide individual testimony regarding her/his status in a dialogue with the physician. The diagnosis will then be conducted by complementing this information with written information from (electronic) Health Records (EHR) followed by clinical examination, laboratory testing and imaging. The quality and relevance of the information from Anamnesis is strongly dependant on the format and focus of the dialogue at hand. The dialogue itself is partly derived from analysing the patients EHR. As we know, the information found in EHRs is based on treatment of the patient by different hospital specialist departments. This structure of the EHR hampers the understanding of the treatment processes and their relevance in preparing the dialogue with the patient in the Anamnesis process. We argue in this chapter that tools supporting the Anamnesis Process, such as the Visual Incidence Anamneses (VIA), have potential to improve decision support and process transparency in diagnosing patients and hence increasing patient safety. The following two figures capture our main ideas. Figure 1 illustrates information collected by the physician at two separate anamnesis events. Situation N is the present situation. Situation N-1 refers to
an earlier event (which could also be multiple earlier events). The inbound arrows denote different important aspects to consider in the anamnesis. The outbound arrow denotes the selected diagnosis and treatment. In Figure 2 we illustrate a tool offering feedback to earlier anamneses allowing the physician to take these into account during the current anamnesis in Situation N. Specifically, it illustrates that the selected diagnosis and treatment might differ due to an enhanced patientcentric greater context. Fig. 2. Separate events (in situation/-s N-1) concatenated by automation of data; a visible historical feedback for the physician to consider, study and discuss with the patient in the actual anamnesis phase (Situation N).
In this chapter we will take a closer look at decision support tools and processes related to anamneses in the remaining part of this section. In Section 1.2 Patient Safety we shortly refer to the current alarming situation worldwide and the need for improved tools and methods (such as VIA) collecting and analysing patient information to improve Patient Safety. Section 2 Artificial Intelligence in Medicine gives a short overview of tools supporting decision making in medicine such as 2.1 Differential Diagnosing and 2.2 Clinical Decision Support Systems (CDSS). Issues related to decision making under incomplete and/or uncertain conditions are discussed in Section 2.3 Safety Assessment. The two complementary decision models "Find simplest cause" and "What if .." are discussed in Section 3 Ockham's Razor vs. Hickham's Dictum. In Section 3.1 The rare cases and the probable we stress that the concepts of "rare" and "probable" are highly context dependant. A "rare" cause can be highly propable due to more patient-process centric contexts such as enabled by, e.g., the VIA model and method. In Section 4 Case studies we motify exemplifies our VIA model as a mean to increase the quality of medical diagnosis. Our proposal is described in Section 7 Visual Incidence Anamnesis (VIA).The chapter ends with Section 8 Conclusions and Section 9 References. Various types of Clinical Decision Support Systems (CDSS) have been developed to support the tasks to make the right diagnoses and decide on appropriate interventions such as treatments etc. However, a CDSS requires an input that is a critical point in the decision support process. In order to collect data for input to the system, the physician must initially perform an investigation of the situation and what might have lead up to it. The traditional action, even before any CDSS were available, is that the physician performs an interview with the patient and/or relatives to the patient, resulting in an anamnesis . Consequently, the anamnesis is a preliminary case history from the patients' perspective. In this step, the collection of patient specific data is vulnerable. Both the physician and the patient must cooperate towards a mutual understanding of what is important for the case or not. The patient (or relative) must be able to articulate information about the actual, or former, health status and the physician must be aware of which questions will sort out crucial information, valuable for further decision making. The next step for the physician is to read relevant information in the (electronic) medical record. Grounded in information derived from the actual anamnesis, s/he decides what further information s/he might need to proceed with. Compared to earlier centuries, the information handling methods and the organization of Health Care are very different. In former times, the physician was reduced to use a complete medical history of the patients' experiences in order to make decisions. The modern society provides specialized Health Care, assisted by advanced medical technology, but the holistic view of the human body is mostly lost as the system is not process oriented and the patients report is marginalized. The current situation, dividing up Health Care (institutional care and primary care) into different clinical departments, is fragmenting. This jeopardizes a continuous information flow for each individual patient involving potential information breakdowns between the varying units of competence. A general lack of time for each patient will further increase the risk of mistakes. The information elicitation process is delicate and the result depends on a variety of variables such as competence and observation of the physician, ability to communicate important experiences by the patient and time available for the anamnesis phase. Even if the prerequisites are perfect, there are still pitfalls resulting in missed, decisive, information. For example, varying symptoms of a disease might occur for a long period of time but earlier diagnoses could have been false. To concatenate symptoms from earlier events to more recent events might be crucial in order to find the true diagnosis. The anamnesis making process should therefore be supported by automation of data from the patients' medical history, instantly visible when the patient and the physician meet ( Figure 2). The difference between fragmented information flows, i.e. a number of disconnected and isolated information units on a time line, and a continuous information flow (of concatenated information units), where it is possible to observe and study earlier events due to knowledge about their existence, is vital. Figure 1 could be regarded as a model of a "hidden context", important for the physician to be aware of in the decision making process. Figure 2 is a model of visualization of such context. It might be emphasized that it is not a simple task to collect data for input to a CDSS. The anamnesis is an important step in this information elicitation process as it provides the physician with patient specific information, but the method for this step is rather ad-hoc and insecure: The human brain is extraordinary in its ability to sort out apparently crucial information for conclusions, but simultaneously this ability is perilous as conclusions might occur prematurely and be false if vital information is missing. Consequently, in Health Care this defectiveness might affect Patient Safety despite any advanced and well-designed CDSS. The critical issue is collection of relevant data.
Patient safety
Due to alarming numbers of incidents, injuries and dea ths in Health Care caused by deficiencies in the field of activities, Patient Safety has attracted considerable attention in the last decade. Declaring that every year in USA, approximately 44 000 to 98 000 deaths occur related to deficiencies in Patient Safety, the report "To Err is Human" (Kohn et al. 2000), is still frequently cited and regarded as significant in the western world. Therefore, it has caused far-reaching attention to the seriousness of the situation and constituted a starting point for appropriate actions. Furthermore, the findings in the report have been confirmed in other, more recent, reports in other western countries such as Sweden. Consequently, for the last decade, in Sweden, as in the European Union, an increasing interest in Patient Safety is noticeable. Various efforts in order to manage the critical situation in Health Care are already in place and the efforts towards models and methods assuring improved Patient Safety have high priority. In this Section 1 and Section 7 we propose Visual Incidence Anamneses (VIA) as a model and method to increase patient safety. Our proposal is underpinned by an analysis of two case studies illustrating some shortcoming of present methods and tools such as Differential Diagnosing (DD) and Clinical Decision Support Systems (CDSS). However, first we must return to an outline of the present situation for Health Care, in relation to Patient Safety. In Sweden, a prevailing strategy to follow up and prevent faults in H e a l t h C a r e h a s b e e n t o " p u n i s h " r e g i s t e r ed individuals (such as registered nurses, physicians etc.). HSAN (Hälso-och Sjukvårdens AnsvarsNämnd; Eng. Medical Responsibility Board) has until 2010 received numerous cases each year to evaluate and take measures against. Today, the National Board of Health and Welfare has shouldered the responsibility for assignments related to Patient Safety cases. However, the perspective on action plans for a safer Health Care has radically changed towards the perspective of fault prevention in the transport sector, i.e. towards a "systemic perspective" on each situation. As a result, the tendency forward is to adopt a basic change of models and methods to address shortcomings in protocols, procedures and in information management in Health Care. For example, among other contributions, to the area, a new Patient Safety Law (SFS 2010:659) came into force January 1 st 2011. Especially noticeable for this matter is that the law also embraces the encouragement of patients and their relatives to participate in the Patient Safety work. This is in line with the development of the Patient Empowerment (PE) movement, appearing more distinctly along with the strong attention to Patient Safety as a complementary approach. PE is contributing to Patient Safety as the empowered patient in Health Care is regarded as a co-operator and an important piece of the puzzle, contributing with experience in being the one who has the "inside information" of being ill and, to a certain extent, being capable of act as such instead of passively receive care (a former, more traditional, view of the patient). Patient Centered Medicine (PCM) is another related line of policy, where a holistic view on the patient is given priority over the earlier, most common, industrial inspired view of work flow in Health Care as comparable to an assembly production line. PCM aims to avoid fragmentation of Health Care and to find ways to coordinate the patient the whole way through the care process. A related perspective is "Lean Thinking", a management strategy for improvement of processes also applicable in Health Care. Accordingly, "Lean Health Care" is introduced at several hospitals in Sweden where Skåne Universitetssjukhus (SUS) (Eng. Skåne University Hospital), Lund, is one of those. These approaches are all connected to an aim of improving not only efficiency but also Patient Safety. Returning to a very basic principle of Patient Safety, we must consider that Health Care is built on Information. Without information there will be neither conclusions nor decisions. Furthermore, the information in use must be correct. It must both be true and complete. Regarding the quality of knowledge used in Health Care, Evidence Based Medicine (EBM) is adopted as a guarantee of first rate quality of scientific information, used in clinical settings. As such, EBM is an important foundation for Clinical Decision Support Systems (CDSS), www.intechopen.com used for support in situations of diagnosing and treatment. Today, CDSS are used in different clinical settings, preferably at the point of care, by physicians as well as by nurses. However, referring to the areas of Patient Empowerment and Patient Centered Medicine, briefly described above, every medical case is unique as every patient is an individual carrying a unique set of historical events in his/her medical history. This medical history is both documented in medical records, such as Electronic Health Records (EHR) and in the consciousness of patients and their relatives. Every single event in a medical history is important as it might be a clue to, or affect, current or future events. Neither EBM nor EHR and traditional CDSS are completely or clearly covering such aspects on diagnosing or treatment for the patient. Nevertheless, these events are substantially important for Patient Safety as loss of such information might lead to wrong diagnoses and delayed treatment. For many diseases, the time aspect is highly important for the successfulness of treatment.
To be able to observe and identify appearances of critical information and information structures over time, a longitudinal case study, going on for ten years between 1999 and 2009, has been addressed. Moreover, a rather rare case occurring during a period of six months in 2010 added some more findings to the other and also questioned the widely adopted Ockham´s Razor of Diagnostic Parsimony. This seemed to be important with reference to the common use of CDSS. A qualitative methodology was chosen (including participatory observation, interviews and the study of medical records) which made it possible to more closely identify and study occurring, unpredictable, information breakdowns. Those breakdowns appeared to be critical for the outcome of the different cases, however not, or not clearly, visible in the EHR system for the physicians to observe. For each case, each new event was dependant on information from former events, sometimes carried by the patient or a relative, but not evident to be important at the time of the occurrence. To successfully use CDSS in order to find an accurate diagnosis or treatment requires patient data that is both true and complete. Figure 1 and 2 illustrates this aspect. In this chapter, we focus on incomplete causality models in Health Care. We suggest a feasible solution by utilizing graphical visualization of chronological information in EHR-systems. This information structure is suggested to complement the widely adopted rule-based Decision Support architectures. The rule bases can be regarded as isolated islands of knowledge while our graphical visualization ties together those patient specific events.
Artificial intelligence in medicine (AIM)
Health Care is basically built on Information. Without Information, decisions about investigations, diagnoses and treatment would be impossible to make. For example, patient specific data is important to collect. This is accomplished by the creation of an anamnesis, collection of body data by medical equipment and clinical medical examination of the body in different ways (palpation, percussion etc). Decisions also require general knowledge to connect to. To build such knowledge, scientific information of high quality is necessary. Patient data and general medical knowledge is the foundation for any decision about a disease; deciding to choose diagnosis and the best practise of treatment. The knowledge of physicians and nurses in Health Care (in this context referred to as "human agents") is evident to be part of the decision making process. However, the knowledge bases in human agents might differ from one agent to another as humans are individual; different levels or different directions in education and different former experiences, as well as more subtle tacit knowledge, build human knowledge. Artificial Intelligence (AI) has since decades been regarded as potentially useful in Medicine, forming a sub-area: "Artificial Intelligence in Medicine (AIM). An early apprehension of AIM was that it would offer possibilities to create "a doctor in a box" which even could surpass the competence of a human agent; a physician (Coiera 2003, p. 331). In more recent years, the ambitions have been more moderate. Instead, different applications of Knowledge Management have been addressed to complement the knowledge of human agents in this area. To support the decision making process in Health Care, Clinical Decision Support Systems (CDSS) are developed for clinical practise. However, the success of CDSS is conspicuous by its absence and the usage still not very often established as a part of work flow. Coiera (2003) refers to some reasons for reluctance to use CDSS: "Reasons for the failure of many expert systems to be used clinically include dependence on an electronic medical record system to supply their data, poor human interface design, failure to fit naturally into the routine process of care, and reluctance or computer illiteracy of some healthcare workers." (Coiera 2003, p. 344) Above a more user friendly and intuitive design, it seems to be necessary to more deeply consider work flow and the flow of information in Health Care, in order to develop and implement useful CDSS. Furthermore, additional tools for CDSS must be designed to repair shortcomings in protocols, procedures and information management in Health Care. However, to be able to do that, such shortcomings must be identified and analyzed. Patient Safety is an area where the results of such shortcomings are explicitly expressed. In this chapter, we will present a case study and some findings pointing at this need and we will also present a feasible solution for repair of information breakdowns which are jeopardizing Patient Safety. However, first we will deepen the reasoning about CDSS by presenting a logical method termed Differential Diagnosing.
Differential diagnosing
Symptoms might be caused by a great variety of causes. For example, fever is such a symptom. To pinpoint the true cause of occurring symptoms, i.e. the pathophysiologic explanation of the symptoms which is the actual disease, the decision making process embraces a method termed Differential Diagnosing. Referring to Merriam-Websters dictionary, the definition of Differential Diagnosis (ΔΔ or DD) is "the distinguishing of a disease or condition from others presenting with similar signs and symptoms" Accordingly, DD is, basically, a method used to systematically identify unknown variables; i.e. a "process of elimination". This is a logical tool by which a list of possible diagnoses is made by the physician, implicit in mind or explicit on paper, digital etc. The diagnoses are, at hand, narrowed down by excluding impossible diagnoses until only one diagnosis remains. This implies that for one patient only one diagnosis, representing the symptoms, is true any other false. The word Diagnosis 1 originates from the Greek word Diagignoskein, meaning to "discern, distinguish" which is the basic aim of diagnosing: to discern the right diagnosis from the wrong. Consequently, DD in Medicine are the process of eliminating alternative diagnoses that might have some common symptoms with the true diagnosis and which could mislead the physician. In this process, Diagnostic Algorithms are used as tools for elimination. However, in rare cases, two conditions might occur simultaneously, giving rise to one similar symptom. For example, chest pain could arise both from cardiac infarction and gastroesophageal reflux disease. Normally, after process of elimination, at least one is excluded; but both could be true, resulting in one missed diagnosis. (Later in this chapter, "Hickam´s Dictum" in relation to "Ockhams Razor" will emphasize this phenomenon.) Consequently, a defective process of elimination could result in a wrong or incomplete diagnosis, especially if not every sign or symptom is available to notice. In this matter, the importance of a complete anamnesis should be stressed. Accordingly, it could be concluded that the process of elimination is delicate. Another peril is the physicians' memory capabilities, necessary to be adequate, especially in situations characterized by high workload and stress. Therefore, IT-tools for DD are, along with the development of the Internet, available for physicians, for example DiagnosisPro 2 , a free self-contained web service to be used as a memorandum aid in the diagnosing task, in order to increase the quality of care and patient safety. This is a tool, not a Clinical Decision Support System (CDSS); however many CDSS are typically designed for DD as they basically provide Diagnosing Decision Support. To more closely be able to explain how CDSS can be beneficial to DD and Patient Safety, CDSS will more closely be described in the next section.
Clinical Decision Support Systems (CDSS)
Clinical Decision Support Systems (CDSS) are computer systems dedicated to the decision making task, i.e. to support clinicians in practice. Typically, CDSS are of two main types: Knowledge-Based and non-Knowledge-Based. The most frequently used type in Health Care settings today is the Knowledge-Based CDSS, also known as "Expert Systems" [Coiera 2003]. However, the metaphorical designation "Expert" might be unfortunate as it could provoke opposition about the sometimes assumed intention of the implementation of such systems; to take over the role of the physician. To avoid such interpretations and emphasize its true role, Expert Systems are today most often referred to simply as CDSS. Their use is more and more commonly accepted as they also provide opportunity to pursue Evidence Based Medicine (EBM), to improve Patient Safety. More infrequently occurring in Health Care settings are non-Knowledge Based CDSS. They could also be regarded as Learning Systems as they belong to a sub-area of Artificial Intelligence called Machine Learning. In this chapter, we will focus only the Knowledge-Based CDSS as it is the most common system for physicians to use in the decision making process (Coiera 2003). Furthermore, we will only focus decisions about Diagnosing as this action actually forms the basis for any further decision about interventions, such as treatment etc. However, the diagnostic types of CDSS (sometimes referred to as Diagnostic Decision Support Systems, DDSS) are considered not as successful in clinical practice as Prescribing Decision Support Systems or other much smaller systems. In the Cases, later presented in this chapter, the diagnosing phase of the decision process turned out to be deficient and threatening to Patient Safety and provided an indication of a need for additional decision supporting tools. However, in this section we will now continue with a brief description of the typical CDSS. Human agents possess knowledge. Knowledge-Based CDSS also possess knowledge. Without presenting any in-depth analyzes, we will stress that there is a basic difference between human knowledge and the types of knowledge referred to in the area of AI. The knowledge in a Knowledge-Based CDSS is typically represented by a set of rules (i.e. Rule-Based systems). Furthermore, such CDSS also consists of an inference engine and a communicating mechanism. A working memory is necessary to store data and conclusions. Patient specific data will be combined with the knowledge in the rule-base by the inference engine while the communicating mechanism allows both input of such data and provides output of the results, from the CDSS. This architecture offers extended possibilities to store large amounts of scientific information, supporting Evidence Based Medicine (EBM). Nevertheless, a CDSS, despite the metaphor of an expert system, is not to compare with a human expert. Human agents are capable of reaching a different, and far more complex, level of thinking that is not possible to implement by AI. Instead, it is necessary for the human agent to interact with the CDSS in a way that will optimize the functions available. For example, it is necessary to provide the CDSS with patient data that is crucial for the task and to interpret and assess output from the system. Some parts of this task can be automated, but not entirely. In the next section we will further explain this and point at difficulties and perils of information management for CDSS.
Safety assessment
There is always uncertainty in the knowledge that underlies a decision. With reference to the assessment of risks to human health posed by chemicals, an uncertainty factor is used to compensate for a deficiency in knowledge and create margins-of-safety. On the other hand, in differential diagnosing, the uncertainty is handled by the "method of elimination". Nevertheless, wrong diagnoses occasionally occur. It might be concluded that there are no margins-of-safety to use as an imagined diagnostic value has only two states; true or false. Consequently, in the diagnosing process, the uncertainty is delicate. How can it be assured that the pathophysiologic explanation of the symptoms is true? If it turns out to be false, the consequences might be lethal. Due to clinical data that is imperfect and treatment that is not a guarantee for remedy, human agents in Health Care must deal with decision making under uncertainty. Probabilistic Medical Reasoning is an approach to this problem (Shortliffe 2006). Instead of expressing that diagnoses are either true or false, the human agents might express the assessment of diagnoses in terms of "probable" or "highly likely" (Ibid). In medical decisions requires strategies and one is to employ an iterative process for data collection and interpretation referred to as the hypothetico deductive approach (Shortliffe 2006, Elstein et al. 1978, Kassirer and Gorry 1978. The method comprises data collection and selection of a hypothesis of the most probable diagnosis, iteratively repeated (refinements of hypotheses by means of additional data) until there is a hypothesis that either is considered true or the uncertainty is reduced to lowest possible level (Shortliffe 2006). The set of active hypotheses are the differential diagnoses. Human agents tend to use heuristic methods to collect data. This is perilous in medicine and a critical point for Patient Safety. In the process, the method of elimination is used to exclude hypotheses that are not probable to be true. This method is related to the use of a philosophical principle named "Ockham´s Razor" i.e. "The law of parsimony". A clinical application of this principle in medicine might be jeopardizing to Patient Safety if the interpretation is close to the well-known adage: "when you hear hoof beats, think horses, not zebras", a rule-of-thumb for selection of diagnosis. Safety assessment in Health Care, concerning the diagnosing task, should be a process that results in an "acceptable diagnosis" chosen on the strength of highest possible amount of relevant patient specific information and scientifically assessed medical information (with reference to EBM). With a Socio-Technical approach to development of usable information management systems for Health Care, CDSS could support such a process. Systems, supporting workflow, are more likely to be used and to be usable for the task. Moreover, to adopt a PCM approach in the field of activities, as well as taking PE into consideration in the design of CDSS, or additional tools for information acquirement for CDSS, would probably be favourable for Patient Safety. CDSS requires input of patient specific data and to elicit relevant data concerning the patient is both challenging and decisive. Referring to figure 2 in Section 1, it might be of vital importance to gain a comprehensive picture of earlier events. The next section will further relate to this angle of reflection as events occurring over time might oppose rules-of-thumbs such as Ockham´s Razor.
Ockham´s (Occam´s) razor vs. Hickam´s dictum
In the area of Medicine, "the Zebra" is most often familiar to everybody. More closely described, this refers to the adage: "When you hear hoof beats, think horses, not zebras". For example, a patient consulting Health Care for fever, with no further distinct symptoms, the most probable diagnosis might be urinary infection or "a virus", not septicaemia. In this case, septicaemia is regarded as "the Zebra". The adage is simply a clinical "rule-of-thumb" in some stage of the differential diagnosing process. This aims at reducing efforts and costs in unnecessary examinations and tests, but at the same time, patients affected with "Zebradiagnoses" evidently exist. Accordingly, a rule-of-thumb should not completely override other possible alternatives. To further explain and strengthen this point-of-view, we will continue with a closer description of Ockham´s Razor as a principle of simplicity. As concluded, "The Zebra" is an interpretation of the philosophical principle "Ockham´s Razor" i.e. "The law of parsimony". In Health Care, this principle is also referred to as "Ockham´s Razor of Diagnostic Parsimony". The principle is derived from the philosophical apprehension of simplicity, which have been expressed in different ways for different fields over the centuries. Basically, the idea is that simplicity is a theoretical virtue; that simpler theories should be regarded as preferable (Baker 2010). In Health Care, this implies that in the diagnosing process, the physician must try to look for a minimum of hypotheses to explain all of the symptoms the patient have, i.e. Diagnostic Parsimony. In order to achieve this aim, only the most probable hypotheses will be tested. What is probable to be true in this perspective is what is probable for the patient as belonging to a large group of homogeneous "patients". However, if the view is that patient is not a member of a homogeneous group of earlier patients, but instead a unique individual with a unique set of patient-specific data, the perspective will change. Harvey et al. (1979) expresses this as follows: "In making the diagnosis of the cause of illness in an individual case, calculations of probability have no meaning. The pertinent question is whether the disease is present or not. Whether it is rare or common does not change the odds in a single patient. ... If the diagnosis can be made on the basis of specific criteria, then these criteria are either fulfilled or not fulfilled" (Harvey et al. 1979, p.15). Accordingly, Ockham´s Razor of Diagnostic Parsimony has been frequently questioned. Even if the approach towards simplicity has advantages, it also has serious disadvantages. One counterargument is Hickam´s Dictum. Referring to the hypothetico deductive approach in the diagnosing process, the principle of Hickam's dictum insists upon that, at no stage of the process, should a particular hypothesis be rejected because it does not seem to fit the principle of Ockham's razor. If the "Zebra" is a popular adage based on the principle of Ockham´s Razor, Hickam´s Dictum is sometimes expressed as simple as ""Patients can have as many diseases as they damn well please". This text en clair could be exemplified by Saint's triad (of hiatus hernia, gallbladder disease, and diverticulosis), affirming Hickam´s Dictum, and simultaneously questioning Ockham´s Razor. Another counterargument to Ockham´s Razor is Walter Chattons "Anti-Razor" or the "Chatton Principle", however not further described in this chapter. Rules-of-thumb might be useful for most cases, and the "Zebra" should be successful for most patients as the most probable diagnosis is diagnoses that statistically is most common to have in relation to the symptoms occurring. We must conclude that this is not a problem. On the other hand, there is a rather serious problem closely connected to Patient Safety and the chances to increase Patient Safety. The problem is the "Zebra", or, even, the "Fascinomas" (slang). Even more problematic are multi illness and systemic diseases. It seems to be momentous to develop protocols and tools to handle atypical and complex situations, in order to prevent mistakes, information misses, injuries and deaths.
The rare cases and the probable
Traditionally, Clinical Decision Support Systems (CDSS) are guiding tools, often grounded in probabilistic reasoning and/or knowledge based rules. The different reasoning mechanisms are implemented as differential diagnosing algorithms, i.e. methods of elimination. In a statistical view, where the patient is regarded as belonging to a large group of earlier patients, this notion will cover the majority of every possible cause of a symptom. Most diagnoses based on this view will probably be true. For example, the symptom "Headache" is most probably caused by muscular tension (in turn caused by nervous tension/stress) or it might be caused by migraine processes. Such more common diagnoses seem to be the first hand choices, which might end the hypothetico deductive process prematurely. Headache could be a symptom of encephaloma (brain tumor) or Stroke; which might be considered as "Zebras" at a glance. Furthermore, it should be noticed that more than one diagnosis might be true: referring to the example above, the cause of a headache could be multiple. Consequently, the problematic cases are when the rule-of-thumb fails. For those reasons, "Occams Razor" is fairly questioned. Probabilistic thinking might limit the domain of possibilities in which a physician ventures to reason. To override such directions, preventing prematurely abruption of the diagnosing process, more relevant data is needed. The next section presents two cases where this need for more data is identified as crucial for the outcome. Data collection is not a simple task which also is observable. Context of a patient is complex, not periodically limited, and not entirely coverable by a traditional anamnesis and medical history available for the physician to make vital decisions on. The cases pinpoint a need for a more concrete time-line of events, alerting for "Zebras" if needed.
Case study I and II
These studies, Case study I and Case study II, comprise two different cases of which the first case describes a fatal "Zebra-case" and the second describes a life-threatening "Zebra-case" and simultaneously a typical "Fascionoma" where the diagnosis is preposterous in relation to probability. In addition, obviously in both cases, prematurely abrupt hypothetico deductive processes were noticeable. It might be presumed, that a potential cause of the www.intechopen.com abruption was heuristic thinking and an apprehension of probability in relation to certain, in these cases misleading, variables: data about immediate circumstances and data about the patient, such as age etc. Simultaneously, crucial data was missing. In Case I, the crucial point is invisibility of earlier events and also the potential importance of earlier events. In Case II, initial invisibility of evidently important earlier events, in a time critical point of the care process, was directly jeopardizing Patient Safety. Interestingly, the roles of patients and the relatives were essential for both cases. They were the keys to whether the cases would prove fatal or not. However, even if patients and relatives always are important in the care process and always must carry patient specific information (Ådahl 2007), this must not be a single-handed task of the kind that the patients survival is directly dependant on if, or what, the patient or relatives report of the former medical history. Such tasks should also be automated, as a safety foundation for further interactive discussions with the patient/relatives.
Case study I: A retrospective longitudinal case study
This case is a retrospective longitudinal qualitative study, in progress for a period of ten years (1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009) and grounded in observations, interviews with the patient and relatives and analyses of medical records. This case presents situations in which hypothesis testing, in retrospect, seems to have failed. It demonstrates that patient specific information, collected for a long time, might be crucial for the Differential Diagnosing task.
Situation 1 (CISIT1): Transitory hemiplegia
In 1999, a 74 year old woman suddenly experienced an evident weakness in the right part of her body. The relatives, present at the time for the incident, called an ambulance whereupon the woman was transported to the Emergency Ward at the local hospital. After some days of hospital treatment, the physicians diagnosed the woman with Migraine. The hemiplegia was transitory and as she had suffered from frequent occurrences of flittering scotomas during the period of hospital treatment, in addition to a long term history (since childhood) of migraine, revealed in the anamnesis, and since she had experienced some months of increasing social stress factors, this was the exclusive focus for the physicians. One CAT scan was performed, revealing nothing suspicious in the brain. The diagnosis Migraine was established despite the absence of the usual migraine headache and, to the patient, the newly occurring neurological symptoms of scotomas and hemiplegia. As migraine is considered rather harmless, following up visits to Primary Health Care providers was not planned or recommended. The woman was, as many in her generation, reluctant to bother health care with more visits, even though the flittering scotomas continued to occur in the years to come. However, she was seriously bothered by this; in addition to the fact that she did not experience the usual symptoms of migraine by which she was spared after her menopause at the age of 58-60. Furthermore, she was since many years suffering from high blood pressure, but when she visited her physician for a routine blood pressure check, the information about the transitory hemiplegia was not accessible for the physician and the patient did not mention it either as she trusted the diagnosis "migraine" despite some skepticism.
Situation 2 (CISIT2): Weakness after syncope
A few years later, in 2002, the woman (now age 77) was found lying over her kitchen table with bilaterally very weak muscular tonus, nearly unconscious. She was able to answer when spoken to but not to move her body or keep her eyes open. She said that she had suddenly fainted, sitting on the chair, and after that not being able to move and still very close to fainting again. An ambulance was called and she was transported to the Emergency Ward at the local hospital.
After a few days, she was sent back home with no follow-up directions for the Primary Health Care. As she, when she arrived to the ward, had some unclear fever, and earlier that day, when she experienced the syncope and general weakness, had visited the local care center for the annual vaccination against the influenza, the diagnosis this time was "Reaction against the vaccination", after excluding Septicaemia by receiving repeated negative blood cultures. Furthermore, the general weakness disappeared within the first 24 hours. As she, after the Emergency Ward, this time was transferred to the Specialist Ward for Infectious Diseases, the physicians did not study the medical record from the Medical Ward and were not aware of the former situation (SIT1) with transitory hemiplegia. Furthermore, this time the weakness was general, occurring after the syncope, so the patient believed that these symptoms were dependant on the reaction of the vaccination. She also trusted the physicians' decision about the symptoms being dependant on a reaction of the immune defense.
Situation 3 (CISIT3): Hip bone fracture
The following situation occurred in 2008, when the woman, now 83, suddenly felt faintly weak and fell in her staircase, resulting in a fracture of the hip bone. For a year, she had problems with weakness, feebleness and dizziness which she thought was natural decrepitude. Not even her district medical officer thought of any other reason. She went by ambulance to the emergency ward where the physicians were puzzled by her, at this time, frequently intermittent unconsciousness: off and on she went unconscious, with a snoring breath. Furthermore, she felt very sick, by nausea and frequent vomiting. However, they noticed that she did faint in spite of lying down in bed and having a slow pulse of 30 when it happened. She also had too low levels of oxygen (SaO2 90 at most) and therefore required oxygen supply. The ECG revealed a momentary asystolia and attacks of atrioventricular block (AV block III), not compatible with her medication: metoprololtartrat 3 (beta-blocker) which immediately was removed. Furthermore, obviously by notes in her EHR, she already was diagnosed by AV-block I which was unknown by the woman herself. Accordingly, she was not in an operable condition, so she was directed to the intensive care unit for cardiology until her heart was considered stable enough. Two days after the accident, she was transferred to the orthopedic clinic for surgical operation (hip replacement) which was a success. However, the day after she was, again, medicated by metoprololtartrat (Selokén), which obviously did not fit to her AV-block history stated at the emergency ward and the cardiology unit. In the subsequent rehabilitation at the orthopedic clinic, she fainted at least three times when trying to walk (one time in the arms of a physician) and probably, not recognized, some times lying down in bed. She felt very weak, but despite these indications, no one seemed to understand the connection between Selokén and her AV-Block and did not check her blood pressure, nor her pulse, at the moments of fainting. Instead, the dosage of metoprololtartrat was increased as the presumption was fall in blood pressure due to the operation, and her inconveniences of palpitations. The woman did not reach her habitual state, but instead she was very weak and faintly, seeming much "older" and more fragile than the last year before the accident. Nevertheless, after a week, the orthopedic treatment programme was finished and she was about to be sent home. The daughter, being a nurse by profession, attended the care planning meeting at the ward, now gaining information about the current treatment, and according to this raised sharp protests against the decision to move her out of hospital. She claimed that her mother was not analyzed due to the cardiac failure and that the medication was lethal, at least a considerable risk factor for further accidents. The attending nurse did not seem aware of this situation but did after all pause the meeting and informed one of the physicians of the orthopedic ward who, in turn, consulted physicians at the intensive care unit for cardiology for a new standpoint on this "new" information.
However, the intensive care unit for cardiology was also unaware of the registered attacks of AVblock III at the emergency ward, solely focusing on cardiac stability for the orthopedic surgery! As a result, Selokén was still prescribed but due to the uncertainty of the situation and special arrangements in the woman's home, she was allowed to stay for some days more. Two days later, Selokén was suddenly removed and she was allowed to stay until she might be stable enough for short-time housing or home. An anemia was also discovered the day after the planning meeting and she was ordered a blood transfusion and iron tablets. After 2 months of recovery, partly on a rehabilitation clinic, she was able to go home and now the symptoms of fainting, faintness and decrepitude were also completely gone.
Situation 4 (CISIT4): Stroke
Seven months after the Hip Bone Fracture, in 2009, the woman, now almost 84, went to bed after a day feeling tired and feeble. In contrast to her usual active life style, she only wanted to sit in a chair, resting that last day. Some of her relatives, visiting her in the afternoon, did notice this change for the worse and her adult granddaughter decided to stay for the night as a result of a premonition of danger. After just about three hours of sleep; her granddaughter heard her calling for help and rushed into the bedroom. This time the woman, again, experienced the general weakness, difficulties in opening her eyes and felt very sick, vomiting and close to fainting. The granddaughter called an ambulance and the woman was, again, transported to the Emergency Ward at the local hospital. This time the physicians had no immediate explanation to present. They discussed if the symptoms could be caused by a stroke, but the general weakness did not clearly answered to that. The relatives was present at the ward and the daughter, being a nurse herself, asked for a CAT scan which was rejected as it was in the middle of the night. After one hour, the woman suddenly experienced an approaching faint and called for help. She had an ECG, monitoring her heart rate, and a moment later the electric waves became straight as a result of a cardiac arrest. The daughter sounded the alarm and the personnel managed to revive her. After this occasion, the woman was transferred to the intensive care unit for cardiology for monitoring and acute treatment. The attending physician at the Emergency Ward, after consultation with the senior physician on standby duty, who did not want to order a CAT scan in the night, excluded stroke as the diagnosis, purely on clinical basis, despite suspicious signs. The patient herself, at this moment still being able to talk, pose the risk of a stroke, but got the answer it could not be. The relatives knew about new treatment methods for strokes caused by blood clots, but also that such methods must be initiated within hours after the stroke began. This made them feel very frustrated. However, the condition seemed to stabilize and the physicians were determined about it not being a stroke, so the relatives were sent home as the patient did want them to do so, to sleep and to being able to go to work in the morning. However, in the morning, when a CAT-scan eventually was performed, it revealed escalation of thrombosis (blood clot) in the brain and brain oedema in progress. Two older infarctions were also revealed, not diagnosed before. A short while after that, the condition went worse. It was at this time too late to use any method to treat the clot (thrombolysis) and stop the stroke from proceeding. Accordingly, the woman rather quickly got an explicit paralysis in her right side (hemiplegia) and lost her ability to speak understandable (expressive aphasia). The following hours, she went worse, in the afternoon also unconscious and finally she died late in the afternoon, 17 hours after the first symptoms. T h i s c a s e e m b r a c e s f o u r a p p a r e n t l y d i fferent situations, with four different pathophysiological explanations of the symptoms occurring, and, as a consequence, treated with reference to four different diagnoses. In CISIT1, the DD process resulted in the diagnosis Migraine. The patient herself did find this strange, as she did not suffer from migraine since her menopause at the age of 58-60, approximately 15 years earlier. However, the CAT-scan did not reveal any pathological alterations in the brain and the inevitably most common cause of such symptoms is Migraine. Consequently, Migraine was the most probable diagnosis. However, in retrospect, we might question this by asking if CAT-scans are quite reliable or if MRI-Scanning (Magnetic Resonance Imaging) would have revealed something else. This imaging tool provides physicians with more detailed information, especially of the brain as it can "see through" bone (the skull). However, prescribing MRI is costly and must be done only if negative results from other tests require more testing. We know, with reference to CISIT4, that older blood clots in the brain at that time were identifiable by CAT-scan which must raise questions about when (during 1999 and 2009) those were originating. CISIT1 could have Stroke as the true cause of the symptoms. However, the hypothesis of Migraine as the most probable cause of the symptoms was chosen and the iterative process of a hypothetico deductive approach (Shortliffe 2006, Elstein et al. 1978, Kassirer and Gorry 1978 to the problem was decided to be stopped. No further tests were prescribed. Analyzing this, both heuristic thinking in the DD-process and ambitions of cost-reduction might be influential to the decision. An elderly patient, with high blood pressure and migraine in the anamnesis, are at increased risk for Stroke. The occurring symptoms should have alerted for this. Furthermore unfortunately, critical information of CISIT1 was lost in the coming visits to the care center as an outpatient. Limitations of (electronic) information management at that time (1999), and deficient routines and protocols in Health Care for such information flow, was a probable cause of information breakdown. The patient herself was the only link to the earlier CISIT1. Next episode (CISIT2), in retrospect now pointing at Stroke as a reasonable hypothesis to be tested more closely, the symptom Fever and the fact that she had a vaccination against Influenza earlier that day did override the symptoms of general neurological weakness and syncope. Furthermore, information from CISIT1 was not visible or easily accessible in CISIT2. Therefore, the actual (considered most probable) hypotheses this time were Reaction against the vaccination or Septicaemia (due to the vaccination). Septicaemia was excluded as the blood tests were negative and other symptoms of Septicaemia did not occur. Lacking crucial information from earlier alarming and critical events, the other symptoms were explained with reference to a rather unusual immune defense reaction on a vaccination. The patient was discharged with no further follow-up. Already in this phase of the study, we must consider other strategies for anamnesis creation, as information, potentially crucial for the differential diagnosing process, was obviously lost between CISIT1 and CISIT2. Moving the attention for our analysis to CISIT3, we will absolutely agree about the diagnosis. There is no doubt about this. Diagnostic radiography ("X-ray") revealed a hipbone fracture (neck of the femur) in addition to inability to move the leg due to pain, the fracture and tissue lesions in the area. However, in this situation, we might inevitably bring Ockham´s Razor of Diagnostic Parsimony to mind. Furthermore, as in the preceding situations, the "Zebra" is probably basically adopted. We will also emphasize that information about CISIT1 and CISIT2, occurring about nine and six years earlier, was not available or presented in a way that the symptoms of these preceding situations could be related to this event. Information about the patients frequent occurring flittering scotomas, evident since CISIT1, was not either visible. Accordingly, in this situation (CISIT3) the patient fell and broke the neck of the femur. The main focus was on the fracture and towards a decision about treatment (surgery), the occurring heart problems of the patient were, at least temporarily, also in focus. A more comprehensive perspective would have been to question why the patient fell and try to identify this in relation to how she normally acts and related incidents in her anamnesis. As the focus was on the fracture and its treatment and the occurring cardiac arrhythmia identified at the emergency ward, this perspective was not entirely investigated. In retrospect of CISIT1 and CISIT2, and with knowledge about the woman in everyday life, we could create a new hypothesis of another diagnosis, as the main cause of the others. It is conceivable that Stroke was the main pathophysiologic explanation of the other diagnoses; Hipbone Fracture and Cardiac Arrhythmia as she might have fallen in the stairs due to a thrombosis in the brain and the heart was affected both by the thrombosis and the physical trauma. This would have been impossible to hypothesize without instant visualization about earlier events and their symptoms. However, an even more serious conclusion, immediately jeopardizing Patient Safety in the situation, is that life critical information was lost within CISIT3, simply due to commonplace transfers of patients between wards. The EHR-system in use at this hospital was Systeam Cross providing access to medical records of other clinical departments, at least for physicians. Moreover, more traditional protocols for oral reporting in transfer situations are also, since many years, put into practise. An even newer protocol is adopted at the emergency ward; SBAR (Situation, Background, Actual condition and Recommended actions), a model for structured communication in Health Care. But still, the information flow was broken, probably because the system did not actively visualize important events in patient transferring situations which became evident in this situation where a relative had to act as information carrier for the patient. Human agents as well as Information Systems in Health Care such as EHR must have access to relevant patient-specific data. The identification, collection, management and presentation of such data seem to be crucial. The last situation in this case, CISIT4, occurred only a few months after the patient was discharged from the rehabilitation clinic, and also for this situation a lack of former information is evident which also became decisive for the result. The symptoms were vague and because of that, and an occurring situation of cardiac arrhythmia and cardiac arrest, the focus was on the heart. Now both the relatives and the patient realized that this, in relation to CISIT1, CISIT2 and CISIT3, could be caused by a stroke, trying to convince the physician on duty that night to prescribe a CAT-scan to find out if there were cerebral causes to the symptoms. But as the symptoms were undefined (as in CISIT1, CISIT2 and CISIT3) and the patient also has mentioned mushrooms (chanterelles) that she earlier that day had eaten, the nausea and attack of vomiting was primarily explained as a probable result from food poisoning, or even gastric influenza (the most probable cause of nausea statistically viewed). A decision was made about waiting to prescribe a CAT-scan, based on cost-reduction ambitions in Health Care and the lower probability of a serious cause of the symptoms, which delayed treatment in case of a more rare and serious diagnosis (stroke by thrombosis). For many rarer diagnoses, the time aspect is decisive and a delay might even be life threatening. Stroke is one of these diagnoses. Heuristic thinking in the form of rules-ofthumbs is common for human agents, but in Health Care, other strategies may be needed to compensate for mistakes and misses dependant on hidden information. Health Care personnel are often working under pressure. For example, in the night, only one physician is at duty at the clinical department s/he is connected to (in addition to the emergency ward and the intensive care unit in cases related to the department) and usually must handle parallel multiple cases at the different wards. Even in the daytime, the pressure is severe and physicians often experience a lack of time to spend on each patient. A solution for insufficient routines and protocols in this matter must not be time consuming in itself. Instead, it must release time at the same time as it increases Patient Safety by providing a more holistic view of the patient and his/her entire medical history.
Case study II: A rare case
This study, in progress for a period of ten months (2010-2011), grounded in observations, interviews with the patient and relatives and analyses of medical records. The case study provides an example of a very rare case, where an initial impulse to follow "the Zebra" obviously was too firm, overriding every sign of something else being in progress resulting in a fatal situation. The time aspect was also in this case, as in Case study I, very decisive to the forthcoming events after the first misleading diagnosis. In addition to "the Zebra", the philosophy of simplicity in Ockham´s Razor was initially noticeable in the decision making process. Furthermore, this case exemplifies a causality dilemma ("Chicken or the Egg") for which an acceptable solution might have been decisive for prevention of any recurrences.
Situation 1 (CIISIT1)
"A young woman, 24 years old, diagnosed at birth with a complicated congenital heart condition with repeated open heart surgeries since then, falls suddenly to the ground with low blood pressure (syncope). At the emergency ward the body temperature rises quickly to 40 degrees (Celcius). No other symptoms are present. The physicians are not able to find any sign of a bacterial infection so she is permitted the following day to go home, despite a rising Bilirubin in serum, however not communicated to her or the relatives. Diagnosis is "virus infection -influenza" despite no other clinical influenza signs than the syncope and high fever. Her mother, being a nurse by profession, did raise a protest against the influenza diagnosis as she found it strange to have an "influenza" with no other symptoms occurring. She was under apprehensions about septicaemia, but the physician rejected this as it is a rather rare diagnosis and not probable at all for the young woman to have; "she would have been in a much worse condition if so", the physician reasoned. Accordingly, and as the fainting tendency has disappeared, the patient and her mother now were open to any other symptoms coming, pointing at "influenza". The next day the young woman experiences nausea and frequent vomiting in addition to pain in the stomach and a mild nose bleed when she vomits. As those symptoms might be signs of influenza (such as gastric influenza), and intense vomiting might result in nose bleed, she and her relatives do not find this very suspicious with reference to the first apparently certain diagnosis. However, in the evening she starts to feel very weak and the pulse rises to the frequency of 120/min. At this point in time, two days and nights have passed since the first signs of illness. The mother did find this alarming, either as a symptom of heart failure or as a symptom of shock. After some advice from the Swedish medical advice telephone service "Sjukvårdsrådgivningen 1177", as the young woman was reluctant to see a doctor again after the first diagnosis, she was transported to the Emergency Ward. At the Emergency Ward, the examination, the blood sample, blood pressure and pulse shows that she most likely has developed severe Sepsis with a Septic Shock reaction, multiple organ failure and, basically, she was suffering from a Cholecystitis probably causing the Sepsis or vice versa. Immediately, intra venous antibiotics are ordered, and the patient is transferred to the Intensive Care Unit for continuous supervision and treatment. However, after 12 hours at the ward, the patient is considered stable, and therefore transferred to a Surgery Ward for treatment of the Cholecystitis, a rare condition designated Acalcuolous Cholecystitis 4 . The physician at the Intensive Ward was told, by the mother, that the patient has a complicated congenital heart condition and that the cardiologists, both at the local hospital and at the University Hospital, where the patient has her attending cardiologist, must be consulted as she has inserted biological material inserted by operative surgery in her heart. As a result, the risk that she develops an Endocarditis, as a complication to the Sepsis, is rather high. Furthermore, her mother tells the Intensive Care Unit physician that the Cardiologist at the University Hospital has asked for information about changed health status as she also waits for a new surgery. She finds the answer she gets as "non sequitur" and "patronizing" and despite this information from the relative, the young woman is suddenly transferred to the Surgery Ward without further discussions and without any further supervision of the heart function. The time at the Intensive Care Unit is also questionably short. The mother, being a nurse by profession, and the supervising nurses at the Intensive Care Unit, find this odd and is worried about the situation. The mother immediately, by her own initiative, in person, contacts the Cardiology Unit at the hospital and, by e-mail and telephone, gets in contact with the cardiologists at the University Hospital. This causes an upset reaction, where the chief physician at the Cardiology Unit visits the young woman at the Surgery ward and informs her and her relatives that she now will be transferred to the Cardiology Unit for further treatment and supervision of the heart. He says, rather upset, that "he has been present at the hospital since nine a.m. and now it is six p.m. without anyone informing him about the patients' arrival and condition". Further on, the status of the heart is carefully examined, to avoid Endocarditis and heart failure caused by the bacteria in the blood and the substantial strain caused by the current disease."
Situation 2 (CIISIT2)
The patient survived the serious illness but recovered very slowly, taking several months. The heart condition seemed to affect her more after the disease than before, increasing the heart failure. Five months later, she suddenly experienced fatigue, diarrhea and nausea, later in the day also vomiting. As she started to feel something in the area of the liver, she contacted Sjukvårdsrådgivningen 1177 where she was directed to the "emergency care center", a care center open until 9 p.m, receiving an appointment time. She and her mother, helping her in this situation by driving, thought the choice of health care center was completely wrong, but because she was directed there they went there first. However, the mother started communication with the nurse at the care center with the assumption that the patient most likely was suffering from a recurrence of the acalculous cholecystitis five months earlier, which hastened the appointment time with the attending physician to occur one hour earlier. The physician immediately redirected the patient to the emergency ward, with a letter of referral with a question at issue: "Acute Cholecystitis?". At the emergency ward, the patient was examined by a surgeon which immediately questioned both the assumed diagnosis and diagnosis five months earlier based on that the symptoms (again) was atypical and that he could not find the information in the EHR at a glance. He strongly doubted the relatives repeated assurance of that this young woman actually had suffered from acalculous cholecystitis (the diagnosis is rare and the woman "too young") until the laboratory report arrives: Rising s-Bilirubin again, just as the relative said was missed at the earlier event. Now the surgeon did read the entire EHR report for the medical history of the last occasion and quickly ordered anthibiotics intra venously. However, an ultrasound of the biliary passage and the gall bladder should have been prescribed immediately to collect patient data for future events. The patient was transferred to the Specialist Ward for Infectious Diseases, and when arriving to the ward, the day after, some physicians again questioned the rare diagnosis, suggesting a more probable explanation to the symptoms: "gastric influenza". A physician decided to change the treatment, in the weekend, as she found it very unlikely to contract a rare disease such as acalculous cholecystitis more than once. However, the mother, being a nurse and medically trained, this time objected very firmly to this point of view, this time. She had found out that, despite the rare condition not likely at all to affect a young woman, of normal weight, even slightly underweight, a state of severe heart failure might cause ischemia in the gall bladder and this is one of the causes to acalculous cholecystitis. The mother had to be very firm, both in discussions with the physician and by leaving a written report of this hypothesis. Finally, she gained a hearing. The young woman did rather quickly recover from the symptoms, and also from the soreness and swelling over the liver, by treatment for the true diagnosis in a very early stage of the disease. When she later on made another visit to the clinic to control how she had recovered, she also met the doctor at the ward that treated her earlier that year, when the acalculous cholecystitis first appeared. He made a note in the medical record about paying attention to the fact that she might have this rare condition if she develops symptoms like the ones she already had twice. In cases of such symptoms occurring, a ultrasound of the gall bladder must immediately be performed, to be able to collect unquestionable data for proof. Early treatment is crucial in cases like this.
Situation 3 (CIISIT3)
Two months after the recurrence of the acalculous cholecystitis, she had to undergo another open heart surgery for her heart condition as her heart failure now was severe and might have been directly life threatening. It was discovered that her aortic valve (a biological xenograft) had an ejection fraction of only 37 percent. The surgeon deemed the valve as "destructed". Most probably were the initial missed diagnoses Septicaemia and acalculous cholecystitis, with the delayed treatment, a direct cause of the accelerating degeneration of the valve. Consequently, it was crucial to her survival that she received early treatment when the cholecystitis reoccurred. This case study, divided in three situations, is pointing at several alarming deficiencies in both information handling routines and the decision making process. The first situation points at abruption of the differential diagnosing process prematurely, i.e. the hypothetico deductive process. In the continuation this situation, there are also (as in Case study 1) occurrences of information loss: when the patient is transferred between wards. Both crucial information of the current period of care and potentially crucial information about the previous medical history were lost, probably as information in the EHR is noted down by human agents, with individual apprehensions of the value of patient specific data, and that important data is not clearly visualized for the physician at the point of care. A causality dilemma was also arising when the true diagnoses were decided (CIICIT1). It could not be concluded if the acalculous cholecystitis was the cause of the sepsis or vice versa. The most probable hypothesis was the first assumption. However, the strain of bacteria found in the blood test (culture) was not bacterias normally expected to be found in the biliary passage. Therefore, the probability of this was low. Instead the strain of bacteria was a rather common bacteria normally found the respiratory passages; Haemophilus Parainfluenzae. Even more intriguing was that acalculous cholecystitis is a very rare condition and the patient was not at all the typical patient in danger of such a condition. Most patients affected with this infection are elderly, seriously ill patients or trauma patients at the intensive care unit. This patient was a young woman; hastily and totally unexpectedly falling ill at work with no preceding warnings. Beyond her heart condition, with some inconvenience with a heart failure, she was completely healthy. It should sometimes be of importance to also identify the cause of the diagnosis, not only the diagnosis as a cause of the symptoms. In this case, it seems to be crucial. Acalculous cholecystitis is a lifethreatening condition with a high mortality rate. Severe septicaemia with multiple organ failure is also extremely serious. For both conditions, the time aspect is critical for the possibility to survive. Consequently, causation is very important for the development of further events such as recurrences. The following could be hypothesized: 1. The Septicaemia is a result of the Acalculous cholecystitis. 2. The Acalculous cholecystitis is a result of the strain of the Septicaemia. 3. The Septicaemia and the Acalculous cholecystitis are separate, independent, conditions, randomly appearing simultaneously. 4. Due to the strain of a heart failure, the Septicaemia is a result of some undetected infection in the throat or respiratory passages. 5. The Acalculous cholecystitis is caused both by the strain of the Septicaemia and a potential ischemia in the gall bladder due to a heart failure. 6. The Acalculous cholecystitis in CIICIT2 is caused by an increasing heart failure. Lack of time and high workload, unfortunately often evident in Swedish Health Care, prevent physicians from hypothesis testing aimed at finding a causal explanation for the origin of the diagnosis itself. Nevertheless, such testing might decrease the number of recurrences or further illness. In this case, the most probable hypotheses for this particular patient, with reference to hidden patient data, should be no. (2), 4, 5 and 6. However no. 1 and even 3 were the hypothesis in focus, but only occasionally. The different perspectives are dependant on presence or absence of critical patient data. Visualization of such information might change the perspective and provide possibilities of treatment to prevent recurrences. Returning to the initial situation (CIICIT1), the physicians deciding on a very common, and therefore also most probable, diagnosis (Influenza due to a virus), were using the principle of Ockham´s Razor interpreted in the shape of the "Zebra". The actual disease started with a syncope and sudden high fever (ague) that declined until the next day. Interviews with the patient afterwards reveal a sense of confusion during the night at the hospital and an inability to communicate this experience clearly to the personnel. The patient was also exhausted when she was discharged the day after and despite notes in the EHR of being in good condition, she was not capable of walking and had to borrow a wheel chair to be able to make it to her mothers' car: This information had unfortunately been lost and the physicians were not aware of it. The time aspect was crucial for a true diagnosis to be found in this case. The blood tests were performed too early in the process and not repeated the next day. Therefore, the CRP-test (C-Reactive Protein) was rather low, pointing at a virus infection, and also the level of white blood cells was not alarmingly high. With reference to this, neglecting a rising Bilirubin in serum, and with a (false) apprehension of the patients apparently good condition, the hypothetico deductive process was ended and a simple and common virus diagnosis chosen. However, with a continued process, with repeated tests before a decision, a fast rising CRP and level of white blood cells, in addition to Thrombocytopenia (decreased number of platelets in the blood) and increasing stomach pains would have lead the physicians to another conclusion. Furthermore, more attention to patient specific data reported from the patient and the relatives would probably have diverted the physicians' from attending to the "Zebra-rule", instead trying to extend the hypothetico deductive process a little more until there was certainty. Even more problematic was situation 2 (CIISIT2). In the EHR, the information from the first situation about four months earlier was not immediately visible at all for the physician at the emergency ward. Instead, the patient herself and her mother had to inform the physician about their apprehension of the current symptoms and how they related to the symptoms occurring in situation 1. This physician trusted this information and started to search for more information in the EHR, resulting in early treatment of the illness as the blood tests, with an increased Bilirubin in serum, also was evidently the same this time. Furthermore, the area of the liver was swollen and sore. However, again this information was incomplete and not clearly visible after transfer from the emergency ward to the Specialist Ward for Infectious Diseases. Some crucial information about the choice of treatment was lost, and therefore questioned which might have been jeopardizing for patient safety. Again the principle of "Zebra" was adopted and the physician at the new ward insisted on gastric influenza as the most probable cause of illness for a young woman. The burden of proof was on the patient and the relatives which is not a preferable or safe situation in Health Care.
Visual incidence anamneses (VIA)
Patient Empowerment (PE) is the underlying approach to VIA. Patients are providing Health Care with valuable information in many ways, generally being capable of cooperating for their own recovery. Participatory Medicine (PM) is a concept, developing from PE and related to Patient Centered Medicine (PCM). Empowerment Systems, suggested in the licentiate thesis "Transparency of Critical Information for Patient Empowerment in eHealth" (Ådahl 2007), are systems supporting these approaches. In the thesis (Ibid), architecture and design of Empowerment Systems, specifically supporting teams, were in focus. The following Figure 3 captures a comprehensive design context of such systems. It might be worthwhile to survey in order to grasp the idea of Empowerment Systems. The picture captures some concepts, important for the design of an Empowerment System. Furthermore, it involves some design aspects, considered important for the functionality in such a system (Ådahl 2007). From this perspective, the VIA is developed. The Empowerment systems of Figure 3 are exemplified as prototypes in the Licentiate thesis (Ibid). The portals (interfaces) investigated were server-oriented allowing users to www.intechopen.com access network based tools and information. A classical CDSS can be seen as an Empowerment System of Figure 3. However, a VIA Empowerment System need a more refined architecture and design. Figure 4 outlines the basic idea of a VIA tool in diagnosing decision processes: Fig. 4. Evidence Based Medicine for the application of general medical knowledge should be the foundation for any decision. In addition, Patient Specific Data is decisive. The VIA tool supports the collection of such data, viewed in a visual chronological perspective, independent of fragmenting specialist knowledge divisions in Health Care.
Consequently, considering the counter arguments to the use of Ockhams Razor, we argue that the patient should be viewed as the unique individual s/he is, which means that the probability of a certain diagnosis should not only depend on what diagnoses earlier patients as a group statistically had, but also on what kind of critical individual information the unique patient holds as a result of his/her earlier medical history. The sum of the symptoms experienced by a patient during the entire medical history must be considered as potentially reciprocal, caused by a common disease. Regarding atypical occurrences of symptoms, rare, or complex medical states such as systemic diseases where vague symptoms occur over a (longer) period, we have found that CDSS as such, based on probabilistic algorithms, i.e., average values in a population, might not be sufficient, or even inappropriate in diagnosing an individual.
To remedy some of those shortcomings, we propose an additional tool, a Visual Incidence Anamnesis ( V I A ) , t o h e l p H e a l t h C a r e p r o f essionals use available CDSS towards individualized care and increased patient safety. The VIA collects the actual medical history of a patient, that enables reassessments of earlier diagnosis towards a more reliable patientcentric grounded health care (Figure 2). The VIA should be available as a patient (individual) centered workflow, quickly visualizing vital information such as symptoms, incidents and diagnoses, occurring earlier in the medical history, at different times, to make further vital decisions patient and context centric. In effect this entails that the VIA enabled Empowerment system should be configurable from selected components and tools rather than a fixed client -server system. For example, the users could use IPads with selected Apps configured using Memory Sticks to ensure flexibility and information security. An example of such experimental environment is given in (Stahl et al. 2010). Furthermore, some of the input information to the entire VIA system could be provided by proper sensor networks (Lundberg & Gustavsson 2011). However, the VIA is basically an information visualizing tool, presenting valuable data graphically, in chronological order, for the physician and the patient to discuss in cooperation.
Core principles
The VIA is grounded in three main principles: 1. Clinical decisions in health care must be grounded in a sufficient amount of relevant and (potentially) important patient specific information. 2. Information of importance for decisions must be easy to comprehend; visualized in the anamnesis processes. 3. Clinical Decision Support Systems and additional tools to support diagnosis complement (such as the VIA) must be tailored to empower stakeholders of the work flow and not regarded as time-consuming and of doubtful value for the task. Concerning the first principle, a sufficient amount of relevant information is information that will provide individual medical histories in such way that no vital information is missing or can be missed. The information must be presented in chronological order, with relations between important events along the time-line. Patient-specific information in this perspective is counterbalance to unfettered use of Ockham´s Razor in Health Care. It should be emphasized that Patient Empowerment and the development of this movement, Participatory Medicine, must be adopted in order to collect and classify relevant and important information. For example, in the anamnesis phase of the medical examination, certain input to the VIAsystem can be performed interactively with the patient and/or relatives. The second principle is the principle of Visualization. Information is considered more visible if it is graphically expressed. Large amounts of information are hard to survey and grasp, especially in a glance. The time aspect in the Anamnesis phase must not be neglected as this might be decisive for many cases of information misses. The physician must mentally construct an internal model of the information available, to decide which information is relevant to use in the hypothetico deductive process. Under the impact of stress and high work load this might fail. Visualization of information from earlier events, easily accessible in the EHR, will offer more input for the creation of such mental models. The third principle concerns usability. A tool must be valuable for the task to motivate its usage. It must facilitate the work and the work load, as well as it must enhance the work flow in the activity. As already mentioned, the time aspect in Health Care is crucial and therefore it must not be time consuming or complicated to use. Above all, it must not jeopardize patient safety by being so.
Methodology
Our current work with VIA is entirely conceptual. The basic idea is, as described, outlined by the result of ethnomethodological studies, pointing at an evident need for additional decision support tools to avoid devastating or lethal information breakdowns. Decisions in Health Care must be supported, not only with existing CDSS but also with tools for elicitation and coordination of information. Consequently, VIA is not yet implemented in any setting. We are at this moment approaching the design phase, aiming at implementation and testing of a prototype within the next year. We have experienced positive feedback from Health Care personnel such as physicians. Furthermore, patients seem to be positive towards such a direction. A deep frustration about information misses and bad coordination of tasks exists, resulting in a situation of jeopardized Patient Safety. Accordingly, we believe and hope that VIA would be regarded as a missing link for an unbroken flow of information in future testing situations.
Pros and cons of VIA
Our proposal for VIA is grounded in our conclusions from in-depth analyses of actual cases in Health Care, where Patient Safety has been jeopardized due to identifiable information handling deficiencies and information breakdowns in the care process. In Section 4, we have presented two such cases. VIA should be an additional tool to the EHR, viewed as a decision support tool, and to traditional CDSS. The advantages of using VIA are visualization of otherwise hidden information (not visible or known to the physician). VIA also visualizes not easily accessed information, crucial for a correct diagnosis to be made or the correct diagnosis to be made in time. If VIA are designed in participation with the user, the use should be a part of work-flow, reminding the decision maker of information that should be considered before decision. Fewer information misses and mistakes based on lack of decisive information increases Patient Safety as the opportunity of correct diagnoses early in the decision process increases by correct information. However, if not developed and implemented to fit requirements of Sociotechnical systems such as Information processing systems for Health Care, and with lack of understanding of which type of information that must be brought to focus, there is a risk of having a system not fit for purpose. This would not encourage the use of the system. Furthermore, with bad design, there is a risk that VIA visualizes too much information, resulting in information overload that paradoxically could make relevant information invisible. Therefore, system development based on the VIA model must comprise the users of the system (participatory design) and preferably also be grounded in close studies of the activity in which the VIA is intended to be implemented. Furthermore, a VIA system is never completed. It must be continuously maintained during its lifetime to have the intended usefulness.
Conclusions
To be able to screen out unnecessary alternatives and decide on the cause of illness, a sufficient amount of significant patient-specific information is needed. This is the basic principle of the VIA. The patient specific information is unique to the individual patient and www.intechopen.com | 2017-08-27T18:56:36.833Z | 2011-09-06T00:00:00.000 | {
"year": 2011,
"sha1": "abf73c03106d17dd4e5f533186ed7c4f90f19a44",
"oa_license": "CCBYNCSA",
"oa_url": "https://doi.org/10.5772/16850",
"oa_status": "HYBRID",
"pdf_src": "Adhoc",
"pdf_hash": "80b3d0a7cd703a7606287e1ff16b53f7264af861",
"s2fieldsofstudy": [
"Medicine",
"Computer Science"
],
"extfieldsofstudy": [
"Psychology"
]
} |
51724249 | pes2o/s2orc | v3-fos-license | ABCA1 haplodeficiency affects the brain transcriptome following traumatic brain injury in mice expressing human APOE isoforms
Expression of human Apolipoprotein E (APOE) modulates the inflammatory response in an isoform specific manner, with APOE4 isoform eliciting a stronger pro-inflammatory response, suggesting a possible mechanism for worse outcome following traumatic brain injury (TBI). APOE lipidation and stability is modulated by ATP-binding cassette transporter A1 (ABCA1), a transmembrane protein that transports lipids and cholesterol onto APOE. We examined the impact of Abca1 deficiency and APOE isoform expression on the response to TBI using 3-months-old, human APOE3+/+ (E3/Abca1+/+) and APOE4+/+ (E4/Abca1+/+) targeted replacement mice, and APOE3+/+ and APOE4+/+ mice with only one functional copy of the Abca1 gene (E3/Abca1+/−; E4/Abca1+/−). TBI-treated mice received a craniotomy followed by a controlled cortical impact (CCI) brain injury in the left hemisphere; sham-treated mice received the same surgical procedure without the impact. We performed RNA-seq using samples from cortices and hippocampi followed by genome-wide differential gene expression analysis. We found that TBI significantly impacted unique transcripts within each group, however, the proportion of unique transcripts was highest in E4/Abca1+/− mice. Additionally, we found that Abca1 haplodeficiency increased the expression of microglia sensome genes among only APOE4 injured mice, a response not seen in injured APOE3 mice, nor in either group of sham-treated mice. To identify gene networks, or modules, correlated to TBI, APOE isoform and Abca1 haplodeficiency, we used weighted gene co-expression network analysis (WGCNA). The module that positively correlated to TBI groups was associated with immune response and featured hub genes that were microglia-specific, including Trem2, Tyrobp, Cd68 and Hexb. The modules positively correlated with APOE4 isoform and negatively to Abca1 haplodeficient mice represented “protein translation” and “oxidation-reduction process”, respectively. Our results reveal E4/Abca1+/− TBI mice have a distinct response to injury, and unique gene networks are associated with APOE isoform, Abca1 insufficiency and injury.
Introduction
Traumatic brain injury (TBI) is a significant public health concern; it is a major cause of death and disability in the United States, and its occurrence is highest among multiple vulnerable populations, including the elderly, young adults, and military personnel [16]. No treatment currently exists for the approximately 2 million cases of TBI sustained each year in the United States, and the costs of medical care for 2010 were estimated at $76.5 billion annually [10].
TBI is caused by an initial external force, whether a physical object or inertia, contacting the head [27]. The impact and initial mechanical stress placed on the cells constitute the primary injury, whereas the secondary injury occurs after the inciting traumatic event, and involves multiple pathways and signaling cascades that can cause further damage [3,13,39]. Inflammation is a major component of the secondary injury. Inflammation is present in every case of TBI and may be a driving force for secondary pathology [13]. Chronic neuroinflammation following TBI was closely associated with neuronal death and impaired cell proliferation in locations both immediately adjacent to, and more distant from, the site of injury [1]. Many studies have shown that levels of inflammation and inflammatory molecules are strongly correlated with multiple measures of outcome in patients, including neurobehavioral impairments and survival rates [29,45]. Microglia are the brain's main form of immune response to infection, disease, and injury, as well as the source of inflammation. As such, inflammation and microglia have been recent concentrations of research as a means of developing therapies and improving outcomes of TBI.
Outcomes of TBI include possible changes in cognition, behavior, emotion, and sensory processing, all of which are influenced by injury severity and location [5,18,35]. Additionally, research has linked TBI to the risk of developing neurodegenerative diseases, including chronic traumatic encephalopathy and Alzheimer's disease (AD) [36]. The high level of heterogeneity in outcomes suggests a significant role for genetic influence on brain susceptibility and recovery [15,44]. The apolipoprotein E (APOE) gene has been frequently studied to determine its role in TBI and its isoform-dependent impact on outcome. The APOEε4 allele is the strongest genetic risk factor for late onset AD, and is thought to confer worse outcome after TBI [2]. APOEε4 carriers have been found to have slower recovery, increased risk of posttraumatic seizures, and worse memory performance after TBI [2,12,14]. However, multiple studies also show that APOEε4 carriers did not differ from non-carriers in cognitive performance, functional outcomes or recovery after TBI [11,37]. The contradictory results so far emphasize the need for more research on APOE and TBI.
APOE is involved in several pathways after a TBI occurs, including inflammation [20]. Inheritance of the APOEε4 allele is associated with increased inflammatory responses, including after TBI [28]. APOE4 may induce a more robust pro-inflammatory reaction from microglia and may suppress anti-inflammatory signaling [4,23,24,26]. This may be a result of decreased stability and faster catabolic degradation of APOE4, compared to the other isoforms, which is possibly due to its lower lipidation levels [20]. APOE is secreted as nonlipidated apolipoprotein, cholesterol and phospholipid efflux to lipid-poor APOE is mediated by ATP Binding Cassette Transporter A1 (ABCA1) [7]. Abca1 deficiency results in decreased APOE lipidation and APOE levels [17,21]. ABCA1 may also play a role in modulating the inflammatory response in the brain. Mice lacking brain ABCA1 saw increased inflammatory gene expression, and the microglia cultured from these mice exhibited an increased pro-inflammatory response, as seen by higher levels of TNFα secretion and lower phagocytic activity, in response to lipopolysaccharide administration [19]. It is not known how Abca1 haploinsufficiency may influence TBI.
We recently performed transcriptional profiling of APOE expressing mice after TBI using Next Generation Sequencing [9]. Using a network-based approach, we were able to identify distinct modules correlated to injury and APOE isoform, as well as a module driven by APOE isoform across TBI groups. The aim of this study was to examine the effect of Abca1 haploinsufficiency on gene expression induced by TBI in APOE targeted replacement mice using transcriptional profiling and a network-based approach. We used 3-month-old mice expressing human APOE3 +/+ and APOE4 +/+ isoforms (E3/Abca1 +/+ and E4/Abca1 +/+ , respectively), and compared them to their Abca1 haploinsufficient counterparts (E3/Abca1 +/− and E4/Abca1 +/− , respectively), after performing a controlled cortical impact. Transcriptional profiling of hippocampal and cortical tissue from the injury site was performed using RNA-sequencing (RNA-seq). E4/Abca1 +/− mice had higher expression levels of the common up-regulated transcripts after TBI, which included genes related to the immune response and inflammatory response. We then examined how ABCA1 insufficiency impacted expression of the microglia sensome genes, and found that E4/Abca1 +/− TBI mice expressed these genes higher than E4/Abca1 +/+ TBI mice, whereas no difference was found when comparing sham Abca1 +/− to Abca1 +/+ mice of either isoform. There was no effect of Abca1 haploinsufficiency on the expression of microglia genes in APOE3 TBI mice. We were able to correlate the transcriptome to each phenotype using a network-based approach, Weighted Gene Co-expression Network Analysis (WGCNA). We found that the immune response module, although correlated positively to all TBI groups regardless of APOE isoform or Abca1 copy number, consisted of genes expressed at higher levels in E4/ Abca1 +/− TBI mice, and featured microglia-specific hub genes, including Trem2, Tyrobp, Hexb, and Cd68. Our results demonstrate an effect of ABCA1 deficiency on microglia gene expression after TBI in APOE4 mice.
Animals
All animal experiments were approved through the University of Pittsburgh Institutional Animal Care and Use Committee and carried out in accordance with PHS policies on the use of animals in research. Human APOE3 +/+ and APOE4 +/+ targeted replacement mice (referred to as E3/Abca1 +/+ and E4/Abca1 +/+ ) were bred to Abca1 +/− mice to generate APOE3 +/+ /Abca1 +/− and APOE4 +/+ / Abca1 +/− (referred to E3/Abca1 +/− and E4/ Abca1 +/− , respectively) [8,17]. All mice were on the C57BL/6 genetic background and experimental groups consisted of both genders. Experimental mice were kept on a 12 h light-dark cycle with ad libitum access to food and water. At 3 months of age, these mice were randomly assigned to either sham or controlled cortical impact (CCI) experimental group. Mice were handled for 2 days (5 min per day) prior to surgical procedures. All materials were purchased through ThermoFisher Scientific, unless otherwise noted.
Traumatic brain injury CCI model of brain injury was performed as previously described [9]. Anesthesia was induced using 5% isoflurane, after which it was maintained at 1.5% isoflurane. The head was secured using a stereotaxic frame, and core body temperature was held at 37°C using a heating pad. After shaving the heads, two separate iodine -alcohol washes were performed to sterilize the surgical site. A 50% mixture of bupivacaine and lidocaine was applied to the area and ophthalmic ointment was applied to the eyes. The scalp was opened with a midline incision exposing the dorsal aspect of the skull and the skull leveled. A 4.5 mm diameter craniotomy was performed over the left parietal cortex using a dental drill. Once the bone flap was removed, mice in the CCI group received a single impact at 1.0 mm depth with a 3.0 mm diameter metal tip onto the cortex (3 m/s, 100 ms dwell time; Impact One, Leica). Sham mice received identical anesthesia and craniotomy, but did not receive impact and are considered negative controls. Following the impact, the surgical site was sutured, triple antibiotic cream applied, Buprenex (0.1 mg/kg, IP) provided for analgesia, and sterile saline administered for rehydration. Mice were allowed to recover on heating pad, until freely mobile, before returning to their home cage.
Tissue processing
Fourteen days post-injury, mice were anesthetized using Avertin (250 mg/kg of body weight, i.p.) and perfused transcardially with 20 mL of cold 0.1 M PBS pH 7.4 [9,31]. Brains were rapidly removed and a 1.5 mm coronal section of the brain, including the injury site, taken by slicing the brain at − 2.5 mm and − 4.0 mm from bregma. Within the coronal slice, the hemispheres were separated, and the subcortical tissue was dissected out; hippocampal and cortical tissue were snap-frozen together for RNA isolation and RNA-seq.
RNA isolation and RNA sequencing
All procedures were performed as before [9,32]. CCI and sham mice consisting of both genders for each genotype were used for RNA-seq. RNA was isolated from frozen cortices and hippocampi at the injury site and purified using RNeasy kit (Qiagen) according to the manufacturer recommendations. Quality control of all RNA samples was performed on a 2100 Bioanalyzer instrument (Agilent Technologies) and samples with RIN > 8 were further used for library construction using RNA Library Prep Reagent Set (Illumina). Libraries for Next Generation Sequencing were generated by PCR enrichment including incorporation of barcodes to enable multiplexing. Sequencing was performed by the Next Generation Sequencing Center (University of Pennsylvania, https://ngsc.med.upenn.edu/) on HiSeq 2500 machine. Following initial processing and quality control, the sequencing datasets were further analyzed for differential gene expression, which in all cases was calculated using Subread/featureCounts (v1.5.0; https:// sourceforge.net/projects/subread/files/subread-1.5.0/) for read alignment and summarization and statistical package edgeR (v3.14.0; https://bioconductor.org/packages/ release/bioc/html/edgeR.html). Lists of differentially expressed genes are further analyzed as described in the following section.
Weighted gene co-expression network analysis (WGCNA)
Network analysis was performed using WGCNA (v.1.51; https://cran.r-project.org/web/packages/WGCNA/index. html) [33,48]. Libraries are clustered by gene expression enabling the detection of outliers and the power is determined by scale free topology model. Modules were generated automatically using a soft thresholding power, β = 10, a minimum module size of 18 genes and a minimum module merge cut height of 0.25. Modules were named by conventional color scheme and then correlated with trait data using Pearson's correlation (APOE isoform, Injury, Abca1 copy number). Statistical significance was determined by student's t-test, p < 0.05. All modules were summarized by module eigengenes (ME), the first principle component of each module that was calculated as a synthetic gene representing the expression profile of all genes within a given module.
Representative networks were built using hub genes and the transcripts connected to them. Hub genes were identified using cutoffs of their interconnectivity within the module (module membership, > 0.8), and the correlation between expression level and trait (gene significance, > 0.2). Once the hub genes are selected, the connections to other transcripts are sorted by weight, with the first 150 connections used for visualization. Gene-association networks of interest were visualized using Cytoscape (v3.3.0). Unsupervised hierarchical clustering was performed on ME turquoise using pheatmap (v1.0.10; https:// cran.r-project.org/web/packages/pheatmap/index.html) to identify 2 sub-modules. A representative network was built for each sub-module consisting of only genes within the sub-module.
Results
TBI induces changes to the transcriptome that are common among both Abca1 +/+ and Abca1 +/− mice expressing human APOE isoforms To examine the effect of TBI on gene expression in the brains of Abca1 +/− and Abca1 +/+ mice expressing human APOE isoforms (E3/Abca1 +/− , E4/Abca1 +/− , E3/Abca1 +/− , E4/Abca1 +/+ ), we collected hippocampal and cortical tissue from the injury site at 14 days post-injury. Total RNA was isolated from these tissues and used for RNA-sequencing. As shown in Fig. 1a-d, TBI significantly affected the transcriptome within each genotype. We highlighted several genes of interest on the scatterplots increased by TBI, and while they were differentially expressed within all the groups, the group with the highest CPM values was the E4/Abca1 +/− TBI mice. To determine what similarities existed among the affected biological processes, we examined the differentially expressed genes that were significant and common among the 4 genotypes. The expression levels of the top 100 upand down-regulated genes are shown in the heatmap (Fig. 1e). Although the genes are common, the E4/ Abca1 +/− mice show the highest expression levels of the upregulated genes. Figure 1f shows the biological processes associated with the common, upregulated genes. There were 1196 up-regulated genes common to all the groups and the top Gene Ontology (GO) terms derived from these genes were "immune system process", "innate immune response", and "inflammatory response". In comparison, Fig. 1g shows the biological processes associated with the common, downregulated genes. There were 579 downregulated genes common to the groups, and these genes were functionally associated with "regulation of ion transmembrane transport", and "potassium ion transport".
Abca1 haploinsufficiency upregulates microglia sensome genes in injured APOE4 mice
To determine if there was any effect of Abca1 haploinsufficiency on gene expression changes induced by injury, we examined the expression of microglial sensome genes. Although a clear effect of TBI is present in the differential expression of the microglia sensome by Abca1 genotype, the heatmap also shows that E4/Abca1 +/− TBI mice have higher expression levels of microglial sensome genes than the other groups (Fig. 3a). In contrast, there is no effect of Abca1 copy number on synaptic transmission genes (Fig. 3b), although an injury effect on expression is still visible. We examined the expression levels of the microglia sensome genes within each APOE isoform, separated by injury status, for the effect of Abca1 genotype. Sham mice in both APOE isoforms (Fig. 3c-d) and injured APOE3 mice (Fig. 3e) have no significant changes in microglia sensome gene expression due to Abca1 haploinsufficiency. In comparison, the injured E4/Abca1 +/− mice demonstrate significant expression of the microglia sensome compared to E4/Abca1 +/+ TBI mice (Fig. 3f). In conclusion, these results demonstrate an effect of Abca1 haploinsufficiency on the microglia sensome in APOE4 mice after TBI.
WGCNA reveals interconnected gene clusters associated with each trait of interest-APOE isoform, Abca1 copy number, and injury status To identify interconnected gene clusters, or modules, associated within each trait of interest, we employed WGCNA. We were interested in the modules that were differentially expressed across our traits of interest -injury status, APOE isoform and Abca1 copy number. The relationship table (Fig. 4a) shows the MEs of interest and the corresponding correlation coefficients per group. ME tan (module size = 182 genes) correlated across the groups depending on APOE isoform, regardless of either injury status, or Abca1 copy number. It positively correlated to APOE4 groups and negatively with APOE3 groups. As seen in the module heatmap ( Fig. 4a; far right), the gene members are generally increased in APOE4 mice and decreased in APOE3 mice. The GO terms associated with the module genes were "tRNA aminoacylation for translation", "RNA processing", and "translation". We built a representative network using the hub genes associated with "tRNA aminoacylation for translation", such as Yars, Gars, and Nars, which are aminoacyl-tRNA synthetases. ME pink correlated with injury status, however, it negatively correlated to TBI groups and positively correlated Fig. 1 TBI increases the expression of genes associated with immune response, and decreases the expression of genes connected to ion transmembrane transport. RNA was isolated from the hippocampal and cortical tissues collected 14 days after injury from Abca1 +/− and Abca1 +/+ mice of both APOE isoforms and was then used to perform RNA-seq, N = 6-8 mice per group of both genders. a-d Scatter plots represent the RNA-seq results for differentially expressed genes. EdgeR analysis between sham and injured mice identified significant affected transcripts in (a) APOE3/Abca1 +/− , (b) APOE4/Abca1 +/− , (c) APOE3/Abca1 +/+ and (d) APOE4/Abca1 +/+ mice. Red denotes up-regulated, and blue denotes down-regulated genes, p < 0.05. e Heatmap of the top 100 upregulated and downregulated genes by TBI is shown. f A table shows the top annotated GO terms derived from the common, upregulated genes (total = 1215 genes). g A table shows the top annotated GO terms derived from the common, downregulated genes (total = 531 genes) with sham groups. The biological processes associated with ME pink (module size = 518 genes) included "synaptic vesicle docking", "long-term synaptic potentiation", and "chemical synaptic transmission", which suggests that injury decreases synaptic transmission. The representative network (Fig. 4c) was built around several hub genes related with synaptic transmission, including Stx1a, Snap25, and Lamp5, which are all associated with synaptic vesicle docking and neurotransmitter release. Lamp5, in particular, is associated with GABAergic synaptic transmission and short-term synaptic plasticity [40]. Another hub gene is Prkcz, which is necessary for long-term potentiation in hippocampal CA1 pyramidal cells [42]. ME grey60 correlated with Abca1 copy number, specifically, it negatively correlated with Abca1 +/− mice and positively correlated with Abca1 +/+ mice, regardless of injury or APOE isoform. As seen in Fig. 4d, the network was built around hub genes, which represented GO terms "oxidation-reduction process", "transport", and "aging". These hub genes included a number of the NADH hydrogenase subunits, such as ND1, ND2, ND4, ND5 and ND6. Other hub genes were COX1, Atp5j2, and CYTB. All of these hub genes are involved in the mitochondrial respiratory chain [46]. ME turquoise strongly correlated with injury status, however, unlike ME pink, it correlated positively to TBI groups and negatively to sham groups. The genes within ME turquoise are strongly connected and related to the module biological process, as seen in the module membership and gene significance scatterplot (Fig. 5a). Due to the size (module size = 3860 genes), we were interested in further separating the module. To do this, we ran a pheatmap function on the genes within the module, which aggregates the genes using hierarchical clustering. As shown in Fig. 5b, the pheatmap separated the module into 2 distinct clusters based on injury status and direction of expression. Additionally, the pheatmap shows the expression for all the genes in ME turquoise and the eigengene expression for each sample. The first cluster, Cluster 1, (size = 2605 genes) consisted of genes upregulated in TBI groups and downregulated in sham groups. The pheatmap suggests a stronger response of the cluster 1 genes within the E4/Abca1 +/− mice, which is consistent with the correlation of ME turquoise to this group in the relationship table. The GO terms derived from Cluster 1 were "immune system response", "innate immune response", and "inflammatory response". Additionally, among the top 10 GO terms was "lipid metabolic process". The representative network (Fig. 5c) was built around hub genes associated with immune response, such as Clec7a, C1qc, and microglia-specific genes, Trem2, Tyrobp, Hexb, and Cd68.
The second cluster, Cluster 2, (size = 1111 genes) featured genes downregulated in the TBI mice, upregulated in the sham mice. Functionally, this cluster is enriched in genes connected to the GO term "transport", other transport-associated terms, such as "vesicle-mediated transport", but also GO terms "sterol biosynthetic process" and "cholesterol biosynthetic process". The network (Fig. 5d) built for Cluster 2 excluded any genes from Cluster 1, and functionally represents transport, however, while hub genes, Gabrb2, Gabrg2, and Atp1b1 all relate directly to the transport of ions across the membrane, through this mechanism, these genes are also strongly associated with synaptic transmission. In conclusion, ME turquoise strongly correlated to injury status, but hierarchical clustering of the genes revealed two distinct clusters associated with the gene expression direction. Cluster 1 was larger and featured genes related to immune response and was more strongly upregulated in E4/Abca1 +/− mice, while cluster 2 featured genes downregulated in TBI groups and represented transport, but functionally are also involved in synaptic transmission.
Discussion
We examined the impact of Abca1 deficiency and APOE isoform expression on the response to traumatic brain injury. Our goal was to identify differences in the transcriptional response and trait-associated genome-wide correlated gene networks between Abca1 +/+ and Abca1 +/− mice following a controlled cortical impact in human APOE3 +/ + and APOE4 +/+ targeted replacement mice. We found that the four groups within our study -E3/Abca1 +/− , E4/Abca1 +/− , E3/Abca1 +/+ , and E4/Abca1 +/+ − had common and distinct responses to TBI. E4/Abca1 +/− mice had the highest proportion of unique transcripts affected by TBI, suggesting that E4/Abca1 +/− mice are more disposed to changes in gene expression by TBI than the other groups, and demonstrate possible pathways that could be associated with worsened outcome, such as downregulated genes associated with learning. The common, up-regulated genes were associated with biological processes related to immune response, innate immune response and inflammatory response. While, these genes were common among the four groups, the E4/Abca1 +/− mice had higher expression levels of the genes upregulated by TBI compared to the other groups, suggesting a role for APOE isoform and ABCA1 in the expression of inflammatory genes after TBI. Consequently, we examined the effect of Abca1 insufficiency on microglia sensome genes by injury status and APOE isoform. When comparing injured Abca1 +/− to Abca1 +/+ mice, we found E4/Abca1 +/− TBI mice had increased expression of the microglia sensome genes. In contrast, there was no effect of Abca1 copy number in APOE3 mice, sham or TBI. These results suggest that Abca1 haploinsufficiency may influence the inflammatory response following TBI, particularly through an impact on microglia and their gene expression. This effect is seen only among APOE4 mice, not APOE3 mice; this response may be related to the isoform-specific effect on inflammation. Additionally, the APOE4 isoform may be more vulnerable to the (See figure on previous page.) Fig. 3 Abca1 deficiency affects the microglial response to TBI in an APOE isoform-dependent manner. a-b Heatmaps were generated using normalized Abca1 +/− versus Abca1 +/+ CPM values for each group for (a) microglia sensome genes and (b) synaptic transmission genes. Red denotes higher expression values, and blue denotes lower expression values. n = 6-8 per group, including both males and females. c, e Selected genes from the microglia sensome of APOE3/Abca1 +/− and APOE3/Abca1 +/+ mice are compared separately for (c) sham (black bars) and (e) TBI groups (green bars). Shown are the Log2-fold change values for each gene. d, f Selected genes from the microglia sensome of APOE4/Abca1 +/+ and APOE4/Abca1 +/− mice are compared separately for (d) sham (orange bars) and TBI groups (purple bars). Shown are the Log2-fold change values. *: p < 0.05 consequences of Abca1 haploinsufficiency due to a gene-gene interaction, a result also demonstrated by data from AD-model mice [17]. These results suggest a possible mechanism for worse outcome after TBI associated with APOE4 isoform.
Using WGCNA, we identified modules associated with each traitinjury, APOE isoform and Abca1 copy number. ME tan was associated with APOE isoform; the module positively correlated with APOE4 mice and negatively correlated with APOE3 mice, regardless of Abca1 copy Fig. 4 WGCNA identified modules correlated to TBI and Abca1 haploinsufficiency. WGCNA was used to determine the correlation of module eigengenes to Injury and Abca1 genotype. a The relationship tables shows the correlation between the module eigengene (rows) and group (columns) with p-value. Red denotes a positive, and blue is a negative correlation. Modules of interest are differentially expressed between trait conditions. MEs turquoise and pink correlated with TBI in opposite directions, ME tan correlated with APOE isoform, and ME grey60 with Abca1 genotype. Top assigned GO terms and their log10 of the p-values are shown to the right of the table, aside heatmaps of the genes within each module, for each animal. Red denotes higher expression values, and blue denotes lower expression values. b-d Representative networks for (b) ME tan, (c) ME pink, and (d) ME grey60 were built using module hub genes. Hub genes are identified in red font. Size of the nodes represents the module membership value and the width of the edge, the interaction between genes shown as connecting lines, represents the weight of the connection number or injury status. The representative network was associated with the GO term "tRNA aminoacylation for translation", and included hub genes Yars, Gars, and Nars, which are aminoacyl-tRNA synthetases. Mutations in these genes are associated with Charcot-Marie-Tooth disease, one of the most commonly inherited neurological disorders [6]. Additionally, a metabolomics study on AD patient CSF and plasma found that a pathway significantly affected in plasma by AD severity was aminoacyl-tRNA biosynthesis, however, the mechanisms associated with altered aminoacyl-tRNA synthetases and AD remain unknown [41]. The "synaptic transmission" module, ME pink was significantly correlated and down-regulated by injury across the groups. The network represented GO terms "synaptic vesicle docking", "long-term synaptic potentiation", and "chemical synaptic transmission". The hub genes featured in the representative network, included Stx1a, Snap25, and Lamp5, which are all associated with synaptic vesicle docking and neurotransmitter release. Lamp5 localizes in the synapse, where it may play a regulatory role in GABAergic synaptic transmission [40]. Another hub gene in this network, Prkcz, has an important role in hippocampal long term potentiation and learning [42]. Its expression mediates the storage of specific forms of long term memory [38]. The negative association between this network and injury is consistent with the impact that TBI is known to have on memory.
The network representing ME grey60 was associated with "oxidation-reduction process" and "aging". This module was differentially expressed dependent on Abca1 copy number; the module was downregulated in Abca1 +/− mice, and upregulated in Abca1 +/+ mice. The network was built around hub genes involved in the mitochondrial respiratory chain, including a number of the NADH hydrogenase subunits, such as ND1, ND2, ND4, ND5 and ND6, as well as, COX1, Atp5j2, and CYTB. Mitochondrial dysfunction and dysfunctional energy metabolism are early pathological features of multiple neurological diseases, including Alzheimer's disease, Parkinson's disease and Huntington's disease [34,46]. Perturbations in the mitochondrial respiratory chain results in decreased ATP synthesis, the generation of free radicals and oxidative damage resulting in neuronal dysfunction and apoptosis [30]. HDL and HDL-associated lipids play key roles in the regulation and preservation of mitochondrial function [43]. ABCA1 is an essential mediator of HDL formation, which may explain the negative correlation between Abca1 +/− mice and this network. ME turquoise correlated with the groups by injury status, however, the module separated into distinct gene clusters representing unique biological processes. Using the pheatmap function, we were able to separate ME turquoise into 2 sub-modules by hierarchical clustering. The clusters were separated based on injury status and the direction of gene expression. The first cluster was larger and consisted of genes upregulated by injury. This cluster represented the "immune response" and the network was built from several microglia-specific genes including Trem2, Tyrobp, Hexb, and Cd68. Although there was no specific modulatory effect of APOE isoform or Abca1 copy number on the module, the expression of the module genes was much higher in E4/Abca1 +/− injured mice, which is consistent with our other results.
ABCA1 is a major regulator of cholesterol transport and an essential mediator of high density lipoprotein generation [22]. ABCA1 may have a crucial role in the response to TBI by providing essential cholesterol and phospholipids required for repair. However, ABCA1 may also influence the TBI response through its modulatory effects on the inflammatory response. Mice lacking brain ABCA1 exhibit increased neuroinflammation, and in particular have an increased microglial pro-inflammatory response [19]. The effect of ABCA1 on inflammation could also occur through its functional role in mediating cholesterol efflux onto lipid-poor apolipoprotein, including APOE. It was previously shown that the loss of ABCA1 function results in a reduction of APOE, and data from experimental animals show that Abca1 deficiency abolishes the lipidation of APOE [21]. The isoform-dependent effect of APOE is possibly driven by lipidation status, which has been shown to affect its stability and degradation rate. Our study shows that ABCA1 haploinsufficiency increased expression of the microglia sensome genes in an APOE isoform dependent manner, which suggests gene-gene interactions as a possible mechanism for worsened outcomes after TBI in APOEε4 carriers.
Conclusions
Our results suggest a possible role for Abca1 haplodeficiency on the response to TBI in APOE4 TBI mice at a transcriptional level. When we compared Abca1 +/+ mice to Abca1 +/− mice by injury status and isoform, we found that the lack of one copy of Abca1 significantly increased the expression of microglia sensome genes only in APOE4 TBI mice. This was consistent with the higher expression of the common, upregulated genes, which were associated with immune response. Furthermore, E4/Abca1 +/− showed the highest expression of the immune response gene network, which also included microglia-specific hub genes, Trem2, Tyrobp, Hexb, and Cd68. Our results suggest that gene-gene interactions can modify the response of APOE4 mice to harmful effects. | 2018-07-28T14:45:38.227Z | 2018-07-26T00:00:00.000 | {
"year": 2018,
"sha1": "033b252f865a1ea22d89cc8c833e7d3918cb98a2",
"oa_license": "CCBY",
"oa_url": "https://actaneurocomms.biomedcentral.com/track/pdf/10.1186/s40478-018-0569-2",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "2d00cee5b11aa3cffb24c5438518b63b86429123",
"s2fieldsofstudy": [
"Medicine",
"Biology"
],
"extfieldsofstudy": [
"Medicine",
"Biology"
]
} |
80244458 | pes2o/s2orc | v3-fos-license | Pest categorisation of Tecia solanivora
Abstract The Panel on Plant Health performed a pest categorisation of Tecia solanivora (Lepidoptera: Gelechiidae) the Guatemalan potato tuber moth, for the EU. T. solanivora is a well‐defined species which feeds exclusively on Solanum tuberosum. It was first described from Costa Rica in 1973 and has spread through Central America and into northern South America via trade of seed potatoes. It has also spread to Mexico and the Canary Islands and most recently to mainland Spain where it is under official control in Galicia and Asturias. Potatoes in the field and storage can be attacked. Some authors regard T. solanivora as the most important insect pest of potatoes globally. T. solanivora is currently regulated by Council Directive 2000/29/EC, listed in Annex II/AI as Scrobipalpopsis solanivora. Larvae feed and develop within potato tubers; infested tubers therefore provide a pathway for pest introduction and spread, as does the soil accompanying potato tubers if it is infested with eggs or pupae. As evidenced by the ongoing outbreaks in Spain, the EU has suitable conditions for the development and potential establishment of T. solanivora. The pest could spread within the EU via movement of infested tubers; adults can fly and disperse locally. Larval feeding destroys tubers in the field and in storage. In the warmer southern EU, where the development would be fastest, yield losses would be expected in potatoes. Measures are available to inhibit entry via traded commodities (e.g. prohibition on the introduction of S. tuberosum). T. solanivora satisfies all of the criteria assessed by EFSA to satisfy the definition of a Union quarantine pest. It does not satisfy EU regulated non‐quarantine pest (RNQP) status because it is under official control. There are uncertainties over the effectiveness of preventing illegal imports via passenger baggage and the magnitude of potential impacts in the cool EU climate.
. Background
Council Directive 2000/29/EC 1 on protective measures against the introduction into the Community of organisms harmful to plants or plant products and against their spread within the Community establishes the present European Union plant health regime. The Directive lays down the phytosanitary provisions and the control checks to be carried out at the place of origin on plants and plant products destined for the Union or to be moved within the Union. In the Directive's 2000/29/EC annexes, the list of harmful organisms (pests) whose introduction into or spread within the Union is prohibited, is detailed together with specific requirements for import or internal movement.
Following the evaluation of the plant health regime, the new basic plant health law, Regulation (EU) 2016/2031 2 on protective measures against pests of plants, was adopted on 26 October 2016 and will apply from 14 December 2019 onwards, repealing Directive 2000/29/EC. In line with the principles of the above mentioned legislation and the follow-up work of the secondary legislation for the listing of EU regulated pests, EFSA is requested to provide pest categorizations of the harmful organisms included in the annexes of Directive 2000/29/EC, in the cases where recent pest risk assessment/pest categorisation is not available.
Terms of Reference
EFSA is requested, pursuant to Article 22(5.b) and Article 29(1) of Regulation (EC) No 178/2002 3 , to provide scientific opinion in the field of plant health.
EFSA is requested to prepare and deliver a pest categorisation (step 1 analysis) for each of the regulated pests included in the appendices of the annex to this mandate. The methodology and template of pest categorisation have already been developed in past mandates for the organisms listed in Annex II Part A Section II of Directive 2000/29/EC. The same methodology and outcome is expected for this work as well.
The list of the harmful organisms included in the annex to this mandate comprises 133 harmful organisms or groups. A pest categorisation is expected for these 133 pests or groups and the delivery of the work would be stepwise at regular intervals through the year as detailed below. First priority covers the harmful organisms included in Appendix 1, comprising pests from Annex II Part A Section I and Annex II Part B of Directive 2000/29/EC. The delivery of all pest categorisations for the pests included in Appendix 1 is June 2018. The second priority is the pests included in Appendix 2, comprising the group of Cicadellidae (non-EU) known to be vector of Pierce's disease (caused by Xylella fastidiosa), the group of Tephritidae (non-EU), the group of potato viruses and virus-like organisms, the group of viruses and virus-like organisms of Cydonia Mill., Fragaria L., Malus Mill., Prunus L., Pyrus L., Ribes L., Rubus L. and Vitis L. and the group of Margarodes (non-EU species). The delivery of all pest categorisations for the pests included in Appendix 2 is end 2019. The pests included in Appendix 3 cover pests of Annex I part A Section I and all pests categorisations should be delivered by end 2020.
For the above mentioned groups, each covering a large number of pests, the pest categorisation will be performed for the group and not the individual harmful organisms listed under "such as" notation in the Annexes of the Directive 2000/29/EC. The criteria to be taken particularly under consideration for these cases, is the analysis of host pest combination, investigation of pathways, the damages occurring and the relevant impact.
Finally, as indicated in the text above, all references to 'non-European' should be avoided and replaced by 'non-EU' and refer to all territories with exception of the Union territories as defined in Article 1 point 3 of Regulation (EU) 2016/2031.
Interpretation of the Terms of Reference
The subject of this pest categorisation is listed in Appendix 1 of the Terms of Reference (ToR) as Scrobipalpopsis solanivora Povoln y. This is widely considered a junior synonym of Tecia solanivora Povoln y, 1973. It is one of a number of pests listed in the Appendices to the ToR to be subject to pest categorisation to determine whether it fulfils the criteria of a quarantine pest or those of a regulated non-quarantine pest (RNQP) for the area of the European Union (EU) excluding Ceuta, Melilla and the outermost regions of Member States (MSs) referred to in Article 355(1) of the Treaty on the Functioning of the European Union (TFEU), other than Madeira and the Azores.
2.
Data and methodologies 2.1. Data 2.1.1. Literature search A literature search was conducted at the beginning of the categorisation in the ISI Web of Science bibliographic database, using the scientific name (junior and senior synonyms) of the pest as search term. Relevant papers were reviewed, and further references and information were obtained from experts, from citations within the references and grey literature.
Database search
Pest information, on host(s) and distribution, was retrieved from the EPPO Global Database (EPPO, 2017).
Data about the import of commodity types that could potentially provide a pathway for the pest to enter the EU and about the area of hosts grown in the EU were obtained from EUROSTAT.
The Europhyt database was consulted for pest-specific notifications on interceptions and outbreaks. Europhyt is a web-based network launched by the Directorate General for Health and Consumers (DG SANCO) and is a subproject of PHYSAN (Phyto-Sanitary Controls) specifically concerned with plant health information. The Europhyt database manages notifications of interceptions of plants or plant products that do not comply with EU legislation as well as notifications of plant pests detected in the territory of the MSs and the phytosanitary measures taken to eradicate or avoid their spread.
Methodologies
The Panel performed the pest categorisation for T. solanivora, following guiding principles and steps presented in the EFSA guidance on the harmonised framework for pest risk assessment (EFSA PLH Panel, 2010) and as defined in the International Standard for Phytosanitary Measures No 11 (FAO, 2013) and No 21 (FAO, 2004).
In accordance with the guidance on a harmonised framework for pest risk assessment in the EU (EFSA PLH Panel, 2010), this work was initiated following an evaluation of the EU's plant health regime. Therefore, to facilitate the decision-making process, in the conclusions of the pest categorisation, the Panel addresses explicitly each criterion for a Union quarantine pest and for a Union RNQP in accordance with Regulation (EU) 2016/2031 on protective measures against pests of plants and includes additional information required as per the specific ToR received by the European Commission. In addition, for each conclusion, the Panel provides a short description of its associated uncertainty. Table 1 presents the Regulation (EU) 2016/2031 pest categorisation criteria on which the Panel bases its conclusions. All relevant criteria have to be met for the pest to potentially qualify either as a quarantine pest or as a RNQP. If one of the criteria is not met, the pest will not qualify. A pest that does not qualify as a quarantine pest may still qualify as a RNQP which needs to be addressed in the opinion. For the pests regulated in the protected zones only, the scope of the categorisation is the territory of the protected zone; thus, the criteria refer to the protected zone instead of the EU territory.
It should be noted that the Panel's conclusions are formulated respecting its remit and particularly with regard to the principle of separation between risk assessment and risk management (EFSA founding regulation (EU) No 178/2002); therefore, instead of determining whether the pest is likely to have an unacceptable impact, the Panel will present a summary of the observed pest impacts. Economic impacts are expressed in terms of yield and quality losses and not in monetary terms, while addressing social impacts is outside the remit of the Panel, in agreement with EFSA guidance on a harmonised framework for pest risk assessment (EFSA PLH Panel, 2010). Is the identity of the pest established, or has it been shown to produce consistent symptoms and to be transmissible?
Is the identity of the pest established, or has it been shown to produce consistent symptoms and to be transmissible?
Is the identity of the pest established, or has it been shown to produce consistent symptoms and to be transmissible?
Absence/ presence of the pest in the EU territory (Section 3.2) Is the pest present in the EU territory? If present, is the pest widely distributed within the EU? Describe the pest distribution briefly! Is the pest present in the EU territory? If not, it cannot be a protected zone quarantine organism.
Is the pest present in the EU territory? If not, it cannot be a regulated non-quarantine pest.
(A regulated non-quarantine pest must be present in the risk assessment area).
Regulatory status (Section 3.3)
If the pest is present in the EU but not widely distributed in the risk assessment area, it should be under official control or expected to be under official control in the near future. The protected zone system aligns with the pest-free area system under the International Plant Protection Convention (IPPC). The pest satisfies the IPPC definition of a quarantine pest that is not present in the risk assessment area (i.e. protected zone).
Is the pest regulated as a quarantine pest? If currently regulated as a quarantine pest, are there grounds to consider its status could be revoked?
The Panel will not indicate in its conclusions of the pest categorisation whether to continue the risk assessment process, but, following the agreed two-step approach, will continue only if requested by the risk managers. However, during the categorisation process, experts may identify key elements and knowledge gaps that could contribute significant uncertainty to a future assessment of risk. It would be useful to identify and highlight such gaps so that potential future requests can specifically target the major elements of uncertainty, perhaps suggesting specific scenarios to examine.
3.
Pest categorisation 3. Are there measures available to prevent the entry into, establishment within or spread of the pest within the EU such that the risk becomes mitigated?
Are there measures available to prevent the entry into, establishment within or spread of the pest within the protected zone areas such that the risk becomes mitigated?
Is it possible to eradicate the pest in a restricted area within 24 months (or a period longer than 24 months where the biology of the organism so justifies) after the presence of the pest was confirmed in the protected zone?
Are there measures available to prevent pest presence on plants for planting such that the risk becomes mitigated?
Conclusion of pest categorisation (Section 4)
A statement as to whether (1) all criteria assessed by EFSA above for consideration as a potential quarantine pest were met and (2) if not, which one (s) were not met.
A statement as to whether (1) all criteria assessed by EFSA above for consideration as potential protected zone quarantine pest were met, and (2) if not, which one(s) were not met.
A statement as to whether (1) all criteria assessed by EFSA above for consideration as a potential regulated non-quarantine pest were met, and (2) if not, which one(s) were not met.
Costa Rica although it was thought to have been introduced into Costa Rica via seed potatoes from Guatemala in 1970. In a taxonomic study of the male and female genitalia, Hodges and Becker (1990) concluded that Scrobipalpopsis is a junior synonym of Tecia Kieffer & J € orgensen, 1910, hence revising the binomial name and placing the original authority in brackets, i.e. T. solanivora (Povoln y). However, the synonymisation was opposed by Povoln y (1993) who resurrected the original name S. solanivora. The 1993 paper was little known, and subsequent authors continued to use the name T. solanivora (Povoln y). Povoln y published two more papers in 2004 (Povoln y, 2004; Povoln y and Hula, 2004) using the name S. solanivora; but later, authors still continue to use T. solanivora.
A search of Web of Science revealed 50 papers using the name T. solanivora between 1995 and 2017 and one paper using the name S. solanivora, that single paper being Povoln y and Hula (2004). The search on Web of Science did not find Povoln y (2004).
For the purposes of this pest categorisation, the name most commonly used in the scientific literature, T. solanivora (Povoln y), will be used. The EPPO diagnostic protocol (EPPO, 2006a) uses the name T. solanivora.
Biology of the pest
In Central America, there are multiple generations of T. solanivora per year. At 10°C, there are two generations per year while at 25°C there can be 10 generations per year (Notz, 1996). Eggs are laid individually or in small clusters on the soil surface near tubers or close to the base of potato plants (Torres, 1989). Rarely eggs are laid on the stems or foliage of potatoes (Povoln y, 1973; Barreto, 2005). When females infest potato storage facilities, they oviposit directly onto exposed potato tubers (EPPO, 2006b). Povoln y (1973) reported some females laid up to approximately 300 eggs over an 8-day period, although the mean fecundity was just under 200 eggs per female.
Eggs develop in 5-25 days, depending on the temperature (Notz, 1996). With mean minimum temperatures of 18.8°C and mean maximum temperatures of 22.1°C, eggs hatch in 6-7 days.
First instar larvae burrow into the soil searching for potato tubers; in potato storage facilities, larvae look for exposed tubers. Larvae feed on tubers; an individual larva will mine into a single tuber and create several galleries. Larvae can burrow and create galleries just underneath the surface of the tuber or burrow into the interior of the tuber. Larval feeding cause's tuber weight loss and allows access of secondary pathogens.
There are four larval instars and development usually occurs inside a single tuber (Hilje, 1994). The larval stage can last from approximately 18-80 days depending on the temperature (Notz, 1996). Mature larvae emerge from tubers to pupate.
Outdoors, larvae pupate in the soil, near the surface. In potato storage facilities, pupae are formed in sheltered areas such as in cracks or corners of building structures and also in potato sacks. It is rare for pupae to form inside a tuber itself (Povoln y, 1973).
Under laboratory conditions (15.5°C, relative humidity (RH) 65.6%), the life cycle lasts 95 days for females and 91 days for males. The mean duration of developmental stages is, 15 days, 29 days, 5 days and 26 days for eggs, larvae, prepupae and pupae, respectively. Adult males live for 16 days, while adult females live for about 20 days.
At 20°C, the life cycle lasts 57 days for females and 54 days for males. At 25°C, the life cycle lasts 42 days for females and 41 days for males (Torres et al., 1997).
Detection and identification of the pest
An EPPO diagnostic protocol exists for the identification of this organism (EPPO, 2006a). Egg and pupal stages are not reliable for identification.
Are detection and identification methods available for the pest?
Yes, as with other Lepidoptera, light traps can be used to capture adults which can then be identified using conventional morphological keys. White delta plastic traps baited with a synthetic sex pheromone can also be used to detect adult males (Nesbitt et al., 1985;Bosa et al., 2005;Cruz Roblero et al., 2011). Tubers infested at low level can be difficult to detect. However, when larvae exit the tuber they leave circular exit holes 2-3 mm in diameter, which can be detected. Heavily infested tubers are more easily detected. If infested potato tubers are detected, larvae can be identified using morphological keys.
First instar larvae are approximately 1.5 mm long and translucent; larvae become bluish-green as they mature; final instar larvae are approximately 16 mm (Torres, 1998).
Pupae are 7.3 mm-9.0 mm long, coffee-coloured light brown becoming dark brown as they develop (EPPO, 2006b). Female pupae tend to be larger and heavier than male pupae (Carrillo and Torrado-Leon, 2014).
Adults are brown, females bright brown and males dark brown; females are 13.0 mm by 3.4 mm; males are smaller, 9.7 mm by 2.9 mm. The rear wings of both sexes have many fringes (EPPO, 2006a; CABI, 2012).
Pest distribution outside the EU
Tecia solanivora is most likely to originate from Guatemala where its genetic diversity is greatest (Torres-Leguizam on et al., 2011). It has spread through Central America and into the north of South America and into the south of North America via movement of potato tubers. Most recently, it arrived into mainland Europe (Spain) ( Table 2) (Puillandre et al., 2008). Dispersal locally occurs via adult flight. Figure 1 and Table 2 show the global distribution of T. solanivora.
Pest distribution in the EU
When T. solanivora was found in the north of Tenerife, it was found in the field and in potato storage facilities; in the islands of La Gomera, Gran Canaria and Lanzarote it was found only in potato storage facilities. Although first observed in Tenerife in June 1999, the specimens were identified as T. solanivora in March 2000 (EPPO, 2006b). In
3.4.
Entry, establishment and spread in the EU
Entry
The movement of prohibited potato tubers by people travelling between the Canary Isles and mainland Spain is thought to be the pathway for introducing Tecia and into Galicia.
Vigo and A Coruña are important Galician harbours and locals may have introduced potatoes for their kitchen garden.
Potential pathways include infested: • seed potatoes, • ware potatoes, • reused potato bags (which may contain eggs and pupae), • soil (which may carry eggs or pupae) accompanying potato tubers (EPPO, 2006b have been produced within the Community, and have been derived in direct line from material which has been maintained under appropriate conditions and has been subjected within the Community to official quarantine testing in accordance with appropriate methods and has been found, in these tests, free from harmful organisms. Annex V Plants, plant products and other objects which must be subject to a plant health inspection (at the place of production if originating in the Community, before being moved within the Community-in the country of origin or the consignor country, if originating outside the Community) before being permitted to enter the Community Part A Plants, plant products and other objects originating in the Community
1.
Plants, plant products and other objects which are potential carriers of harmful organisms of relevance for the entire Community and which must be accompanied by a plant passport 1.3 Plants of stolon-or tuber-forming species of Solanum L. or their hybrids, intended for planting.
Section II Plants, plant products and other objects produced by producers whose production and sale is authorised to persons professionally engaged in plant production, other than those plants, plant products and other objects which are prepared and ready for sale to the final consumer, and for which it is ensured by the responsible official bodies of the Member States, that the production thereof is clearly separate from that of other products Part B Plants, plant products and other objects originating in territories, other than those territories referred to in Part A
1.
Plants, plant products and other objects which are potential carriers of harmful organisms of relevance for the entire Community
Is the pest able to enter into the EU territory? (Yes or No) If yes, identify and list the pathways! Yes, the organism has already arrived in Spain hence a pathway exists. Tubers of potatoes provide the major pathway for entry.
Tecia solanivora: Pest categorisation T. solanivora was introduced into Costa Rica, Venezuela and Colombia via seed potatoes (Povoln y, 1973;Kroschel and Schaub, 2013). Entry into Tenerife (Canary Islands) has been attributed to the illegal import of infested potatoes from Venezuela, Ecuador or Colombia (EPPO, 2006b).
EUROSTAT records volumes of imported commodities entering the EU; potatoes (S. tuberosum) are recorded using a variety of Combined Nomenclature (CN) codes, according to intended use. Codes are accompanied with brief text to provide a description, e.g.
Seed potatoes: While seed potatoes are prohibited from outside the EU (excluding Switzerland), EUROSTAT data indicate imports of seed potatoes in the past from countries where T. solanivora occurs (see below). However, such imports are assumed to correspond to rejected or unsold consignments, originally exported from the EU. Apparently, this process is quite common in the potato sector.
• 6,900 kg from Guatemala into Belgium/Luxembourg in 1989, • 21,000 kg from Costa Rica into France in 1997, • 24,500 kg from Colombia into France in 1998, • 20,700 kg from El Salvador into France in 1998, • 250,000 kg from Honduras into NL in 2004.
EUROSTAT data does not indicate any imports of seed potatoes from Central or South America over the past 5 years (pathway is prohibited by 2000/29 ECsee Table 4).
The Netherlands NPPO kindly provided detailed trade inspection data regarding plants for planting from 2012 to 2014. It indicated that S. tuberosum was imported from Costa Rica in 2014, recorded as CN 0602 9099 (Other live plants, rooted, other). It is possible that this is also a rejected consignment originally from the EU. There are no records of interception of T. solanivora in the Europhyt database.
EU distribution of main host plants
Potato (S. tuberosum) is the only host for T. solanivora (EPPO, 2006b;CABI, 2012;Kroschel and Schaub, 2013). Potatoes are widely grown throughout the EU, both commercially and in private gardens and allotments. Between 2012 and 2016, the mean area of potatoes commercially cultivated in the EU was 17,085,000 ha. Poland, Germany, Romania and France grew over 50% of the total EU potato area (Appendix A). The production of European potato is shown spatially in Figure 2.
Climatic conditions affecting establishment
Tecia solanivora has adapted to a variety of environmental conditions, e.g. being found in mountainous regions of Central and South America at altitudes between 1,000 m and 3,500 m (Torres et al., 1997); in the Canaries at altitudes up to 600 m (EPPO, 2006b); and on mainland Spain at altitudes below 400 m. Daily temperature ranges vary markedly between these areas. At 10°C, there are two generations per year while at 25°C there can be 10 generations per year (Notz, 1996). Source: Eurostat regional yearbook 2015, Available at http://ec.europa.eu/eurostat/documents/3217494/7018888/ KS-HA-15-001-EN-N.pdf/6f0d4095-5e7a-4aab-af28-d255e2bcb395, (accessed 13 October 2017) Figure 2: Harvested production of potatoes in Europe by NUTS 2 region (2013) (Tonnes km À2 of total area) (Note: Germany only available for NUTS level 1 regions; the Czech Republic, Denmark, Poland, Romania, the United Kingdom, Norway, Switzerland and Albania only available at national level) Optimum temperature for population development appears to be around 25°C (Torres et al., 1997). T. solanivora does not survive below 7.9°C or above 30°C (Notz, 1996). Parts of the EU potato-growing region have suitable temperatures that would allow multiple generations to develop each year. Cold winters, where minimum temperatures are often below 7.9°C will prevent T. solanivora from establishing outdoors in northern Europe. Germain (2002a) conducted a pest risk analysis on T. solanivora and used the computer program CLIMEX to assess potential establishment in Europe. Taking into account the climatic conditions within a pest's existing distribution, CLIMEX is used to generate an 'eco-climatic index' (EI) representing the climatic suitability of a location outside of a pests' current distribution, thereby identifying locations where establishment is potentially possible (Sutherst and Maywald, 1985;Skarratt et al., 1995). Maps showing EI for European locations in the pest risk analysis indicate that many sites in Europe have suitable climatic conditions for the establishment of T. solanivora (Germain, 2002b). However, host distribution must also be considered when interpreting CLIMEX maps as CLIMEX does not take account of biotic factors when generating EIs. Kroschel et al. (2016) provide a pest distribution and risk atlas for a range of invasive agricultural pests threatening Africa. One chapter examines T. solanivora and includes a global map entitled 'Establishment Risk Index' (ERI) (Schaub et al., 2016). How the ERI is calculated is not indicated. Nevertheless, the global map suggests that southern Europe, and in particular coastal regions around the Mediterranean and the Atlantic coast of Portugal share an ERI with parts of Central and South America where T. solanivora occurs, hence suggesting that parts of the EU provide suitable conditions for the establishment of T. solanivora.
Spread
The spread of T. solanivora in Central and South America has been due to the movement of infested seed potatoes (Puillandre et al., 2008). The introduction into the Canary Islands has been attributed to the illegal movement of seed potatoes from South America (EPPO, 2006b).
Although adults are weak fliers, flying moths can contribute to local spread. Adults fly at night. They make short flights close to the ground, and during the day, they shelter in shady places on the ground, on bushes and weeds at the edges of fields and under leaf litter or between potatoes in potato storage facilities. Adults can move from potato fields into potato storage facilities and from there back to potato fields (Povoln y, 2004).
When introduced into new areas in Central and South America, T. solanivora spreads rapidly in potato-growing regions; spread was facilitated by the trade in potato tubers as well as local natural dispersal (Kroschel and Schaub, 2013).
Plants for planting (seed potatoes) are a means of spread.
Impacts
As described in Section 3.1.2 (Biology), larvae attack tubers; tuber quality is lowered and heavily infested tubers can no longer be used for human or animal consumption or can be completely destroyed (Kutinkova et al., 2016). Although unidentified at the time, T. solanivora was a pest of potatoes in Guatemala in the 1950s (Murillo (1980), cited by Torres-Leguizam on et al. (2011) Villaneuva andSaldamando, 2013). In 1972, just before being identified, T. solanivora caused losses of 20-40% in potato crops in Costa Rica (Povoln y, 1973). While larvae primarily feed on and destroy potato tubers, when there are high populations, larvae can occasionally also attack the green parts of the plant (Povoln y, 1973).
In 1994, Colombia attributed losses of 276,323 tonnes to T. solanivora; during 1995, there was 4.4% damage to field potatoes and 11.3% damage to potatoes in storage (Arias et al., 1996).
After its introduction to the Canary Islands, severe outbreaks were reported by local news media, and in 2001, media attributed a 50% yield reduction to T. solanivora combined with a severe drought (EPPO, 2006b). Kutinkova et al. (2016) regard T. solanivora as the most important insect pest of potato worldwide.
As well as attacking potatoes in the field, the pest can also seriously impact tubers in storage. In Central and South America, potatoes may be held for short-term storage at ambient temperatures in the dark, in well-ventilated buildings (CABI, 2017). If infested tubers are introduced into such conditions, larval development can continue and multiple generations could occur. Potato stocks in such conditions can be completely destroyed in less than three months (EPPO, 2006b).
In Europe, ware potatoes are often held in storage for prolonged periods at about 4°C (CABI, 2017). In such conditions, larvae would not survive. However, tubers for processing are generally stored at 7-10°C which could allow larvae to develop and complete development (slowly).
If T. solanivora were to establish in the EU, direct impacts from larval feeding and subsequent secondary pathogen infections could be expected in the field and in potato storage facilities.
3.6.
Availability and limits of mitigation measures 3.6.1. Biological or technical factors limiting the feasibility and effectiveness of measures to prevent the entry, establishment and spread of the pest • Infested tubers are difficult to detect (entry holes are very small).
• Eggs and larvae can be carried with soil accompanying tubers.
• Larvae develop inside tubers where they are protected from contact insecticides and natural enemies.
• Strong cultural links between South America and Spain give rise to large numbers of people moving between the regions and provide an opportunity for passengers to carry small quantities of potatoes with them in their luggage. Although such activities are prohibited, managing such pathways is very difficult. Yes, tubers can be sourced from pest free areas. • Surveillance and monitoring, • Delimiting-affected areas and buffer zones, • Destruction of contaminated tubers, • Prohibition of planting of potatoes and restriction of movements in affected areas, • Controls in places where potatoes are sold in areas identified as risk areas.
Uncertainty
There are a number of uncertainties, such as whether there were imports of potatoes from Central and South America in the past, or such trade continues but is not recorded in EUROSTAT using CN codes 0701 (codes that refer specifically to potatoes). However, once T. solanivora spread internationally via the movement of potato tubers, it could (re-)enter the EU on infested tubers originated from Central and South America that is not prohibited by existing legislation.
There is uncertainty as to the number of generations that could develop each year in the EU that affect the magnitude of potential impacts. However, the fact that T. solanivora can complete its development and impact on potato production is evidenced by the ongoing outbreaks that are under official control in north-west Spain.
Long-term establishment in potato-growing countries where winter frosts regularly occur would only be possible if storage facilities provide refuges in winter time, and if movements from there to the field is possible.
These uncertainties do not affect the categorisation conclusions.
Conclusions
Tecia (=Scrobipalpopsis) solanivora meets the criteria assessed by EFSA for consideration as a Union quarantine pest (Table 5).
None
Pest potential for entry, establishment and spread in the EU territory (Section 3.4) T. solanivora has entered the EU; hence, pathways exist, potential pathways include infested potato tubers (mainly seed), reused containers carrying infested tubers and soil attached to tubers. Environmental conditions, especially in southern Europe, appear suitable for establishment. Spread would occur through movement of infested potato tubers; local spread could occur by flying adults.
International and long distance spread occurs via plants for planting (seed potatoes).
Whether there are any potatoes moved (illegally?) into the EU from areas where T. solanivora occurs.
Potential for consequences in the EU territory (Section 3.5) Establishment of T. solanivora in the EU would have an impact on production of potatoes.
Larvae of T. solanivora can destroy potato tubers; hence, their presence in seed potatoes would have an impact on the intended use of such plants for planting The amount of damage to be expected in field potatoes and in harvested stocks is uncertain due to cooler conditions in the EU.
Available measures (Section 3.6)
Phytosanitary measures are available to inhibit the likelihood of entry into and spread within the EU e.g. prohibition of S. tuberosum tubers from many third countries; sourcing seed potatoes from pest-free areas; prohibiting soil from being carried with seed potatoes.
Phytosanitary measures are available to prevent pest presence on plants for planting such as growing seed potatoes only in pest-free areas; Uncertainty over the effectiveness of preventing illegal import (e.g. passenger baggages). Uncertainty on the effectiveness of the measures to eradicate the pest once it is introduced.
Conclusion on pest categorisation (Section 4)
Tecia solanivora satisfies all of the criteria assessed by EFSA to qualify as a Union quarantine pest. Although T. solanivora is present in the EU territory, it has a restricted distribution and is under official control. Not all criteria assessed by EFSA for consideration as a potential regulated non-quarantine pest are met. Although T. solanivora is present in the EU territory, it has a restricted distribution and is under official control. None. | 2019-03-17T13:11:47.797Z | 2018-01-01T00:00:00.000 | {
"year": 2018,
"sha1": "0b5e1e46eca69dd66a7ab50413c4dadfffa3181d",
"oa_license": "CCBYND",
"oa_url": "https://efsa.onlinelibrary.wiley.com/doi/pdfdirect/10.2903/j.efsa.2018.5102",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "61bb4ab2827bd164ea8e4460a03820bd09b84392",
"s2fieldsofstudy": [
"Agricultural And Food Sciences"
],
"extfieldsofstudy": [
"Medicine"
]
} |
255839264 | pes2o/s2orc | v3-fos-license | Melatonin attenuates the TLR4-mediated inflammatory response through MyD88- and TRIF-dependent signaling pathways in an in vivo model of ovarian cancer
Toll-like receptors (TLRs) are effector molecules expressed on the surface of ovarian cancer (OC) cells, but the functions of the TLR2/TLR4 signaling pathways in these cells remain unclear. Melatonin (mel) acts as an anti-inflammatory factor and has been reported to modulate TLRs in some aggressive tumor cell types. Therefore, we investigated OC and the effect of long-term mel therapy on the signaling pathways mediated by TLR2 and TLR4 via myeloid differentiation factor 88 (MyD88) and toll-like receptor-associated activator of interferon (TRIF) in an ethanol-preferring rat model. To induce OC, the left ovary of animals either consuming 10% (v/v) ethanol or not was injected directly under the bursa with a single dose of 100 μg of 7,12-dimethylbenz(a)anthracene (DMBA) dissolved in 10 μL of sesame oil. The right ovaries were used as sham-surgery controls. After developing OC, half of the animals received i.p. injections of mel (200 μg/100 g b.w./day) for 60 days. Although mel therapy was unable to reduce TLR2 levels, it was able to suppress the OC-associated increase in the levels of the following proteins: TLR4, MyD88, nuclear factor kappa B (NFkB p65), inhibitor of NFkB alpha (IkBα), IkB kinase alpha (IKK-α), TNF receptor-associated factor 6 (TRAF6), TRIF, interferon regulatory factor 3 (IRF3), interferon β (IFN-β), tumor necrosis factor alpha (TNF-α), and interleukin (IL)-6. In addition, mel significantly attenuated the expression of IkBα, NFkB p65, TRIF and IRF-3, which are involved in TLR4-mediated signaling in OC during ethanol intake. Collectively, our results suggest that mel attenuates the TLR4-induced MyD88- and TRIF-dependent signaling pathways in ethanol-preferring rats with OC.
Background
Ovarian cancer (OC), which affects approximately 1 in 70 women, is among the most lethal gynecological malignancies and exhibits a poor prognosis when diagnosed during the late stage (<50% with a 5-year survival rate) [1]. Unfortunately, early-stage OC presents with no apparent symptoms, and no effective screening tool is currently available [2]. OC has been presumed to arise from lesions involving the ovarian surface epithelium (OSE) or ovarian epithelial inclusion cysts [1,2], which exhibit morphological characteristics of the reproductive tract epithelium (or Mullerian duct). Resistance to chemotherapy is the primary factor that hinders the long-term treatment of OC [2], and paclitaxel resistance has been recognized to induce recurrent OC and protumor activities [3].
TLRs are cell surface sensors that can initiate pathways to stimulate cell proliferation and chemoresistance, as well as recruiting immune cells to provide support for cancer progression [4]. It has been documented that TLR4 plays important roles in promoting the immune escape of several human cancer cells, including OC cells [5,6]. Furthermore, TLR4-mediated signaling is associated with the metastatic potential of tumor cells [7]. TLRs generally signal via a MyD88-dependent pathway, leading to a proinflammatory response. MyD88-mediated signaling promotes the early activation of NFkB, a nuclear factor that is responsible for the production of many proinflammatory cytokines, such as interleukin (IL)-6, IL-8, and TNF-α and is related to the immune response, cell adhesion, proliferation, angiogenesis, and apoptosis [8,9]. Conversely, a MyD88-independent pathway involves TRAM and, instead of MyD88, stimulates TRIF signaling, leading to late-phase activation of NFkB and induction of IRF3, which increases the production of IFN and IFN-inducible gene products [10].
Because ethanol (EtOH) intake produces a variety of co-carcinogenic and pro-inflammatory effects and seems to positively regulate TLR2/TLR4, IkBα, NFkB, and TNFα [11], the combination of induced OC and EtOH intake results in the appropriate histological and molecular patterns that are needed to evaluate new chemotherapeutic compounds. Thus, we have previously proposed a useful, ethanol-preferring rat model for studying OC, and, notably, long-term mel therapy significantly reduced OC masses and the incidence of adenocarcinomas [12]. EtOH consumption by these animals is associated with cumulative incidence of serous papillary carcinoma and upregulation of Her-2-signaling pathway by OC cells [13].
Mel (N-acetyl-5-methoxytryptamine) is an indoleamine produced by the pineal gland, with maximal secretion at night [14]. The nighttime surge of mel has been suggested to serve as a "natural restraint" for tumor initiation, promotion, and progression [15]. Mel has been indisputably implicated as a therapeutic agent, including in the treatment of reproductive cancers [16,17]. Due to its versatility, mel has been recognized as having antioxidant, oncostatic, and immunomodulatory properties [18,19]. Numerous reports have shown that mel modulates early and late NFkB signaling during inflammation [20,21], thus resulting in changes in the expression of genes such as IL-1, IL-6, IL-12, and TNF-α [22]. Furthermore, some cytokines (IL-2, IL-6, and IL-12) and factors (TNF-α and IFNs) displaying immunotherapeutic potential have been associated with mel and are under investigation as adjuvant therapies for cancer [23,24]. Notably, the administration of mel significantly suppresses the expression of TLR4 and, in turn, MyD88, cytosolic IkBα, and nuclear NFkB p65 after ischemia/reperfusion in rats. Taken together, the reduction of p38 MAPK, NFkB, TNF-α, and IL-6 expression occurs due to an indirect decrease in the MyD88 protein levels [25]. Despite considerable efforts to understand the activation of the TLR2 and TLR4 signaling pathways, the exact mechanism(s) by which mel acts in the OC microenvironment remain unclear.
To better understand this issue, the present study was designed to investigate OC and the effect of long-term mel therapy on the signaling pathways mediated by TLR2 and TLR4 via MyD88 and TRIF in an ethanolpreferring rat model.
Animals and experimental design
Eighty 60-day-old adult UChB (a model of ethanolpreferring rats that was developed by selective breeding) female rats weighing 200-230 g were obtained from the Department of Anatomy, Bioscience Institute/Campus of Botucatu, Univ Estadual Paulista (UNESP). The rats were individually housed in polypropylene cages containing laboratory-grade pine shavings as bedding and were maintained under constant room temperature (RT; 23 ± 1°C) and lighting conditions (12-h light/dark cycle, with the lights switched on at 6 a.m.). Filtered tap water and standard rodent chow (3074 SIF, Purina Ltda., Campinas, SP, Brazil) containing (by weight) 19.90% protein, 30.03% carbohydrate, 3.80% fat, and 13.36% fiber, representing 3.00 kcal/g metabolizable energy, were provided ad libitum. All animals were divided into two arms (n = 40/group): the EtOH group, in which the rats were provided with access to a 10% (v/v) ethanol solution ad libitum (free choice of water or ethanol), and a control group, which was composed of ethanol-naïve rats without access to ethanol. When the ethanol-preferring rats reached 65 days of age, they were given a choice between two bottles containing either water or a 10% (v/v) ethanol solution ad libitum for 15 days. The animals displaying EtOH consumption greater than 2.0 g EtOH/kg/day (ranging from 4 to 5 g EtOH/kg/day) were selected according to the procedure described by Chuffa et al. [26,27]. In this study, the preference rate associated with ethanol-seeking behavior was approximately 70%.
Finally, the rats were divided into four groups (n = 20): Group OC, composed of DMBA-induced animals that did not consume EtOH; Group OC + EtOH, composed of DMBA-induced animals that consumed 10% (v/v) EtOH during ovarian tumor development (OTD); Group OC + Mel, composed of DMBA-induced animals that received mel as a therapy; and Group OC + EtOH + Mel, composed of DMBA-induced animals that consumed 10% (v/v) EtOH during OTD and received mel as a therapy. After these procedures, the females were anesthetized and euthanized by decapitation (during the early morning at 4 a.m. or at ZT 22, which corresponded to the environmental circadian time; Figure 1B) for sample collection.
Ethical statement
The use of laboratory animals in this study was approved by the Ethical Committee of the Institute of Bioscience/ UNESP (CEEA -Permit Number: 382). The guide for the care and use of laboratory animals, published by National Academy of Science, was strictly followed in all experiments. To minimize pain, suffering or distress during experimentation, all of the animals were anesthetized with ketamine (50 mg/kg) and xylazine (10 mg/kg) prior to decapitation so that they were rendered unconscious.
Ovarian tumor induction procedure
After selection based on ethanol consumption, all of the animals (n = 80) were anesthetized using 10% ketamine (60 mg/kg, i.p.) and 2% xylazine (5 mg/kg, i.p.) during the estrous phase, and the left flank region of the skin was cleaned with iodine and 70% EtOH. A 2-cm incision through the skin and the abdominal muscles was performed, and the ovaries were accessed after grasping the fat pad near the left kidney. The left ovary was injected directly under the bursa with a single dose of 100 μg of DMBA (Sigma Chemical Co, St Louis, MO) dissolved in 10 μL of sesame oil, which was used as the vehicle [30], and was returned intact to the body cavity. The muscle and skin layers were closed using a 3-0 silk suture (Ethicon, Inc., Juarez, MX). Sham surgery was conducted on the right ovary by administering only the vehicle. An antibiotic (10 5 units of benzylpenicillin potassium) was administered i.p. for prophylactic treatment. Over the next 180 days, the rats were monitored, and tumor development was observed via ultrasonography. The ovary size (volume and diameter) was used as a comparative parameter during OC development, and high-grade serous papillary carcinomas were subjected to treatment.
Determination of plasma mel concentrations
After blood collection (4 a.m.), mel was extracted from plasma (n = 20 samples/group) using HPLC-grade methanol and was separated on HPLC column (Sep-Pak Vac C-18, reverse phase, 12.5 nm; Water Corporation, Milford, MA, USA). Thereafter, 50 μL of the reconstituted samples were assayed using a Coat-a-count Melatonin ELISA Kit and were immediately read at 405 nm. The intra-assay coefficient of variation was 4%, and all of the samples were assayed in duplicate to avoid inter-assay variability. All of the reagents and microtiter plates were provided by IBL International (Hamburg, Germany). The concentrations of mel are presented as pg/mL.
Immunofluorescence assays
Epithelial OC cells were cultured, washed with phosphatebuffered saline (PBS; sodium chloride, potassium chloride, dihydrogen phosphate, and disodium hydrogen phosphate), fixed in 4% paraformaldehyde for 10 min, and permeabilized with PBS at RT. Nonspecific binding sites were blocked with 1% BSA for 60 min. The samples were incubated with the anti-NFkB p65 primary rabbit polyclonal antibody (dilution 1:100, ab7970, Abcam, Cambridge, MA, USA) overnight at 4°C, followed by a secondary polyclonal anti-rabbit IgG conjugated to FITC (1:200, sc-2012, Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) for 1 h at RT. Nuclei were stained with 6-diamidino-2-phenylindole (DAPI, 5 min) at RT. Primary and secondary antibodies were diluted in blocking buffer (1% BSA). For negative immunolabeling, no primary antibody was added. The immunopositive cells were analyzed using a fluorescence microscope (Zeiss Axiophot II, Oberkochen, Germany) at 40× magnification (immunofluorescence: excitation 590 nm, emission filter 650 nm) and for DAPI staining (excitation 365 nm, emission filter 435-485 nm). The quantification of relative fluorescence in the merged images was performed using Image J software.
Immunoblotting and protein quantification
After mel therapy, the abdominal cavity was opened, the OC samples were rapidly removed, and 100-mg tissue samples were frozen in liquid nitrogen and stored at −80°C. The tissues were microdissected (only areas containing serous papillary tumors were obtained from the ovaries) and homogenized using 10× RIPA lysis buffer (Pierce Biotechnology, Rockford, IL, USA) supplemented with a cocktail of protease inhibitors (Sigma Chemical Co.). Aliquots of 1:10 (v/v) dilutions of Triton X-100 were added to the homogenates, and samples were placed on dry ice with agitation for 2 h to improve extraction. The lysates were centrifuged at 21,912 g for 20 min at 4°C to remove insoluble material, and total protein was measured through colorimetric determination. All protein samples were dissolved in 1.5X Laemmli buffer and used for SDS-PAGE (Bio-Rad Laboratories, Hercules, CA, USA). The same amount of protein (70 μg) was loaded per well and resolved on preformed 4-12% acrylamide gradient gels (Amersham Biosciences, Uppsala, Sweden) using a Tris-glycine running buffer electrophoresis system (60 mA for 2 h). After electrophoresis, total protein samples were electro-transferred (200 mA for 1.5 h) to nitrocellulose membranes in Tris-glycine-methanol buffer. Kaleidoscope Prestained Standards (Bio-Rad, Hercules, CA, USA) were used as molecular weight markers. Thereafter, the membranes were blocked with TBS-T solution containing 3% BSA at RT for 60 min and then incubated at 4°C overnight with the following antibodies (1:500 in 1% BSA): TLR2, TLR4, MyD88, IkBα, IKK-α, NFkB p65, NFkB p50, TRAF6, TRIF, and IRF-3 (Abcam, Cambridge, UK). This treatment was followed by 3 × 5 min washes in TBS-T solution and incubation for 2 h at RT with rabbit or mouse HRPconjugated secondary antibodies (diluted 1:1000 in 1% BSA; Sigma-Aldrich Chemicals, St. Louis, MO, USA). After sequential washes in TBS-T, the signals were developed using an ECL detection kit (Thermo Scientific). Immunoreactive bands were obtained from individual blots of 10 rats/group using image analysis software (NIS-Elements, Advanced Research, Nikon). β-actin (cytosolic extract) and Lamin B1 (nuclear extract) were used as endogenous positive controls, and the results were expressed as the means ± SD. The protein concentrations are presented as the percentage of the optical density (band intensity-background) relative to the positive control.
Subcellular fractionation
Cytosolic and nuclear extracts were obtained from OC cells using NE-PER nuclear and cytoplasmic extraction reagents (Thermo Scientific, Rockford, IL) according to the manufacturer's instructions. Equivalent amounts of the cytoplasmic and nuclear fractions were boiled in SDS lysis buffer for Western blot analysis. To ensure no cross-contamination occurred, specific biomarkers for the cytosolic and nuclear fractions were used.
Statistical analysis
The values are presented as the means ± standard deviation (SD), and data analyses were performed using two-way analysis of variance (ANOVA) for two independent factors (EtOH consumption and mel treatment). Significant results were subjected to post-hoc analysis using Tukey's test, and significance was set at P < 0.05. GraphPad Instat Version 4 statistical software and Sigma Plot Version 12.5 graphing software were used.
OC development and mel therapy
Although the cumulative OC incidence was slightly increased after the combination of DMBA with EtOH intake ( Figure 1C), no apparent difference in the OC phenotype was observed following EtOH consumption.
To confirm the effect of mel on this rat model of OC, macroscopic and microscopic analyses were frequently performed ( Figure 1D); although mel therapy reduced the size of the OC, their serous papillary aspects with markedly pleomorphic nuclei were unchanged after mel treatment ( Figure 1D). Notably, we measured the plasma mel levels to validate its concentration during long-term therapy in animals consuming or not consuming EtOH and exhibiting OC. As expected, a high concentration of mel was found in both the OC + Mel and OC + EtOH + Mel groups after 30-and 60-day treatment ( Figure 1E). As a consequence of mel treatment, the tumor volume was reduced after 30 days (by about 29% and 34.8% for the OC and OC + EtOH groups, respectively; Figure 1F) and 60 days (by about 48% and 54.2% for the OC and OC + EtOH groups, respectively; Figure 1F).
Mel differentially regulated the expression of TLR2, TLR4, and MyD88 in OC of EtOH-preferring rats In response to mel treatment, TLR2 expression and immunostaining were increased (1.22-fold increase vs. OC) in the epithelium of serous papillary carcinoma (Table 1, Figure 2A, B, M, N). However, the combination of EtOH with mel caused a significant downregulation of TLR2 (1.54-and 1.16-fold reduction vs. OC + Mel and OC + EtOH, respectively, Table 1, Figure 2B-D, M, N). In contrast, mel therapy negatively regulated TLR4 (1.23-fold reduction vs. OC; Table 1, Figure 2E, F, M, O) and MyD88 (1.30-fold reduction vs. OC; Table 1, Figure 2I, J, M, P) in these cells. The effects of mel therapy and/or EtOH intake on the TLR2-and TLR4-induced MyD88mediated signaling pathways are presented in Figure 2Q.
The downregulation of NFkB signaling is enhanced by Mel regardless of EtOH intake
The involvement of the NFkB pathway during mel or EtOH treatment was determined in cytosolic and nuclear extracts of OC cells by analyzing the expression of IKK-α, IkBα, and NFkB p65/p50. Notably, IKK-α and IkBα were downregulated by mel therapy (1.25-and 1.37-fold reduction vs. OC; Table 1, Figure 3A, B, E, F and Figure 4A-C). In contrast, IKK-α was upregulated after EtOH intake or EtOH + Mel treatment (1.15-and 1.35-fold vs. OC and OC + Mel, respectively; Table 1, Figure 3C, D and Figure 4A, B). The combination of mel with EtOH reduced IkBα expression (1.42-fold decrease vs. OC + EtOH; Table 1, Figure 3G, H and Figure 4A, C), similar to mel treatment alone. As expected, the epithelial OC cells were NFkB p65-positive in the cytoplasm and nucleus (Table 1, Figure 3I) and became strongly immunopositive after EtOH intake (Table 1, Figure 3K). Surprisingly, mel alone induced the downregulation of cytosolic NFkB p65 and p50 (0.2-and 0.25-fold reduction, respectively, vs. OC; Figure 4A, D, E) and nuclear NFkB p65 (0.18-fold reduction vs. OC; Figure 4A, F, G); presumably, a negligible amount of the NFkB p65/p50 complex was translocated as a dimer to the nucleus. In addition, mel satisfactorily downregulated NFkB p65 in animals with OC that consumed EtOH ( Figure 3L and Figure 4A, D, F, G). Briefly, Figure 4H shows the positive or negative regulatory effects of mel or EtOH on NFkB activation in the OC cells "in vivo". Immunofluorescence assays revealed the expression level and localization of cytosolic/nuclear NFkB p65 in OC cells. Mel therapy efficiently resulted in downregulation of NFkB p65 expression ( Figure 4I, J; level of fluorescence reduced from 97% ± 10.2 (OC) to 66% ± 14.6 (OC + Mel)). In contrast, the levels of nuclear and cytosolic NFkB p65 were significantly increased by EtOH intake and were drastically decreased after mel treatment ( Figure 4K, L; level of fluorescence 168% ± 19.2 (OC + EtOH) vs. 84% ± 11.6 (OC + EtOH + Mel)).
Mel therapy attenuated TRIF-dependent signaling in OC of EtOH-preferring rats
To investigate the effects of mel on the MyD88-independent signaling pathway, the levels of TRAF6, TRIF, and IRF3 were measured in OC tissues. Interestingly, mel therapy downregulated TRAF6 expression (1.24-fold reduction vs. OC; Table 1, Figure 5A, B, M, N), and no significant difference (P < 0.05) was observed following EtOH intake or the combination of EtOH and mel treatment (Table 1, Figure 5C, D, M, N). Unexpectedly, mel significantly attenuated the MyD88-independent TLR2-/TLR4-mediated signaling pathway by reducing the expression of TRIF and IRF3 in serous papillary OC (1.30-and 1.35-fold reduction vs. OC; Table 1, Figure 5E, F, I, J, M, O, P); thus, mel acted as a repressor of the non-canonical TLR2/TLR4-mediated signaling pathway. Furthermore, mel therapy dramatically reduced the expression of TRIF and IRF3 during EtOH intake (1.55-and 1.28-fold reduction, respectively, vs. OC + EtOH; Figure 5G, H, K, L, M, O, P). As shown in Figure 5Q, mel negatively regulated TRAF6, TRIF, and IRF3 in OC cells and was considered to alter both early and late-phase NFkB translocation and IFN-β production.
The IFN-β, TNF-α, and IL-6 levels were reduced after Mel therapy but not in the presence of EtOH The concentrations of IFN-β, TNF-α, and IL-6 in the OC group were 100.2 ± 12.0 pg/mL, 121.7 ± 11.8 pg/mL and 147.3 ± 16.2 pg/mL, respectively. Treatment with mel significantly attenuated the increased levels of these proinflammatory cytokines in OC tissue (37.8%, 43.2%, and 44.7% of Group OC, respectively; P < 0.01) ( Figure 6A-C). In addition, the IL-8 levels were unchanged for all treatments ( Figure 6D). Finally, EtOH consumption in association with OC was unable to alter the levels of these inflammatory cytokines (P > 0.05, Figure 6A-D).
Discussion
We recently demonstrated that long-term mel therapy efficiently reduced the OC size and the incidence of adenocarcinoma in ethanol-preferring animals [12]. Although EtOH consumption has been assumed to inhibit plasma mel levels [26], our data showed that both meltreated groups maintained their physiological levels at night, even in the presence of EtOH. Importantly, the present study revealed that mel negatively regulates the TLR4-mediated, but not TLR2-mediated signaling pathway in OC of ethanol preferring-rats. In addition, the TRIF signaling pathway was downregulated by mel therapy.
Specifically in OC, a variety of TLRs (including TLR2, TLR3, TLR4, and TLR5) are highly expressed in neoplastic ovarian epithelial cells, indicating poor prognosis [31]. In OC cells, TLR4 has been shown to perform a protumor function and to hamper the efficacy of cancer therapies (e.g., paclitaxel) [3]; conversely, mel significantly reduces TLR4 expression in this rat OC model. Although studies of the role of mel in the anti-tumoral suppression of TLR4/NFkB are scarce, in vitro data showed that mel promotes the inhibition of TLR4-mediated inflammatory genes via MyD88-dependent signaling pathway [32], and that mel acts as a TLR4/MyD88 antagonist [33]. Due to its anti-inflammatory properties, mel may be considered as an additional therapy for certain tumors. Interestingly, our results indicated for the first time that mel upregulated TLR2 and that the combination of mel with EtOH caused downregulation of TLR2. These results demonstrate that distinct from its effect on other tissues and under other conditions [11,34], EtOH alone reduces the expression of TLR2 in OC cells. Future studies exploring the effect of mel on genetic TLR2 and TLR4 knockout mice are required to further understand the molecular mechanism(s) by which mel regulates inflammation in OC. The activation of the MyD88 signaling pathway induces tumor cell survival, proliferation, and chemoresistance [3,5,6]. Many studies have reported the ability of mel to modulate the MyD88-dependent signaling pathway [25,32]. MyD88 is a critical adaptor that is directly involved in the TLR4mediated expression of inflammatory genes [35]. The present study showed that the level of MyD88 was significantly reduced following mel therapy in OC. These results are in agreement with others that showed that mel effectively suppressed the elevation of MyD88 expression in several disorders associated with inflammatory responses [25,32,36]. Conversely, EtOH intake is associated with increased expression of MyD88, mainly in the liver of females exposed to higher doses of EtOH (6 g EtOH/kg b.w.) [37]. Although MyD88 expression appears to be unaffected by EtOH intake, mel therapy significantly inhibited the MyD88-dependent signaling pathway in experimental OC. NFkB plays a critical role in the TLR4-mediated signaling pathway in tumor cells. Under unstimulated conditions, NFkB is retained in the cytoplasm by binding to IkB. After phosphorylation by IKK, IkB is degraded via the proteasome, and the NFkB subunits p65 and p50 translocate to the nucleus [38]. According to an earlier report involving serous papillary OC, the mechanisms underlying long-term survival and chemotherapy resistance are directly associated with NFkB signaling [39]. The present study showed that mel significantly reduced the expression of IKK-α, IkBα, and NFkB p50/p65 in OC cells. Furthermore, the NFkB subunit p65 was reduced in cytosolic and nuclear extracts, demonstrating that mel may act on mechanisms regulating gene expression and nuclear translocation of NFkB p65. Most of the effects of NFkB activation on tumor cells have been linked to the upregulation of antiapoptotic protein expression, cell proliferation, and proinflammatory cytokine production [5,40], which may enable these cells to progress and metastasize. Thus, we propose that the inhibitory effect of mel on the TLR4mediated signaling pathway might be partially attributed to the repression of NFkB activation in OC. Figure 3 Immunohistochemical localization of IKK-α, IkBα, and NFkB in serous papillary ovarian carcinoma. Immunoreactivity for IKK-α was weak in the OC group (A) and absent from the OC + Mel group (B). Moderate reaction for IKK-α was detected in the OC + EtOH (C) and OC + EtOH + Mel animals (D). Immunoreactivity for IkBα was moderate to high in the surface epithelium of the OC (E) and OC + EtOH groups (G) and was weak in the OC + Mel (F) and OC + EtOH + Mel groups (H). Strong immunoreactivity for NFkB was detected in the epithelium of the pappilae of the OC (I) and OC + EtOH groups (K). The OC + Mel (J) and OC + EtOH + Mel groups (L) displayed weak and moderate signals for NFkB, respectively (arrows; Bar = 20 μm).
Despite the lack of any significant macroscopic or molecular effect of EtOH on OC, chronic EtOH intake caused an elevation in the IKK-α and NFkB p65 levels in animals bearing OC without affecting upstream molecules. In these animals, mel therapy induced the downregulation of NFkB p65 expression but exerted no effect on the IKK-α levels. Indeed, EtOH alone exerts immunomodulatory effects and induces alterations in downstream TLR signaling in many target organs, such as the liver, the brain, and the gastric mucosa [41,42]. These alterations include upregulation of specific cytokines and inflammatory mediators by activating IKKs, MAPKs, and NFkB [42]. Furthermore, TLR4 receptors were reported to be involved in the ethanol-mediated inflammatory response because the blockade of TLR4 abolishes the production of inflammatory mediators and cell death. Although acetaldehyde appears to severely alter TLR4 signaling, the underlying mechanisms by which acetaldehyde targets TLR4 are unknown [43]. Based on our data, we suggest that EtOH intake may affect only some molecules downstream of TLR4 in OC cells, such as the high activity of NFkB p65. Furthermore, EtOH intake exerted no influence on the expression of NFkB-regulated proinflammatory cytokines.
TRIF, another adaptor molecule of TLRs, is responsible for the regulation of MyD88-independent pathways [44]. TRIF signals the downstream kinases TBK1 and IKK, leading to phosphorylation of IRF3 and the consequent production of IFNs and IFN-inducible genes [45]. In particular, TRIF/IRF3-dependent signaling is thought Extracts obtained from individual rats were used for densitometric analysis of the protein levels following normalization to β-actin or LaminB1. The results are expressed as the means ± SD (n = 10 animals/group). a P < 0.05 vs. OC; b P < 0.05 vs. OC + Mel; and c P < 0.05 vs. OC + EtOH. (H) The canonical pathway that leads to the activation of NFkB after a ligand binds to its specific receptor. Internal signaling results in the phosphorylation of IkBα by the IKK complex, leading to the ubiquitylation and degradation of NFkB via proteasome. The NFkB p65/p50 complex translocates to the nucleus to initiate the transactivation of target genes related to the inflammatory process. Mel induces the downregulation of IKK-α, IkBα, and NFkB p65/p50, whereas EtOH consumption upregulates IKK-α and NFkB p65. (I-L) Merged images of NFkB p65 immunofluorescence and DAPI nuclear staining in the OC (I), OC + Mel (J), OC + EtOH (K), and OC + EtOH + Mel groups (L); details of cytosolic/nuclear staining (Alexafluor®488, Bar = 10 μm). EtOH: ethanol; mel: melatonin; IKK-α: inhibitor of NFkB kinase subunit alpha; IkBα: inhibitor of NFkB subunit alpha; NFkB p65: NFkB subunit p65 (RelA); NFkB p50: NFkB subunit p50.
Figure 5
Immunohistochemical localization and Western blot analysis of TRAF6, TRIF, and IRF3 in serous papillary ovarian carcinoma. The immunoreactivity for TRAF6 was moderate in the OC group (A) and weak in the OC + Mel group (B). The OC + EtOH (C) and OC + EtOH + Mel groups (D) displayed weak immunostaining for TRAF6 (arrow). The immunoreactivity for TRIF was enhanced in the epithelium of the OC (E) and OC + EtOH groups (G), but mel therapy resulted in weak reaction for TRIF in the OC + Mel (F) and OC + EtOH + Mel groups (H). Strong immunoreactivity for IRF3 was observed in the animals from the OC (I) and OC + EtOH groups (K), in contrast to the moderate to high reactions in the surface epithelium of the OC + Mel (J) and OC + EtOH + Mel groups (L) (arrow). Bar = 20 μm. Negative controls were used. (M) Representative TRAF6, TRIF, and IRF3 expression profiles in extracts (70 μg protein) pooled from 10 samples/group (upper panel). (N-P) Extracts obtained from individual rats were used for densitometric analysis of the TRAF6, TRIF, and IRF3 levels following normalization to a loading control (β-actin). All results are expressed as the means ± SD (n = 10 animals/group). a P < 0.05 vs. OC; c P < 0.05 vs. OC + EtOH. (Q) Schematic representation of TLR signaling. MyD88 acts as an essential TIR domain-containing adaptor for the induction of inflammatory cytokines in all TLR subtypes (canonical pathway). In the TLR4-mediated signaling pathway, the TIR domain-containing adaptor TRIF stimulates a MyD88-independent pathway (non-canonical pathway). This signal leads to the activation of IRF3 via TBK1 and IKKƐ/IKKi, facilitating the nuclear translocation of IRF3, which activates the expression of inflammatory factor-related target genes (e.g., IFN-β). Treatment with mel induces the downregulation of TRAF6, TRIF, and IRF3. mel: melatonin; TRAF6: tumor necrosis factor receptor-associated factor 6; TRIF: TIR domain-containing adaptor; TBK-1: TANK-binding kinase 1; IRF3: interferon regulatory factor 3; IFN-β: interferon β.
to be critical in OC [46]. TRIF contains an N-terminal region termed the effector-driving site that recruits TRAF, resulting in the activation of IRF3 and NFkB [46]. The present study showed that TRAF6, TRIF, and IRF3 were markedly increased in OC and that mel therapy inhibited their expression. Recent studies have elucidated the protective effect of mel on the ischemia/reperfusion-injured liver [25] and on lipopolysaccharide (LPS)-stimulated macrophages by reducing TRIF and IRF3 expression [32]. In EtOH-preferring animals bearing OC, the expression of TRIF and IRF3 was similar to that of the OC group, and mel significantly repressed the expression of these factors. It has been assumed that alcoholism induces an inflammatory process at various levels, depending on the amount of EtOH consumed and the specific tissue [47]. This process may include the production of proinflammatory cytokines that are primarily stimulated by the TLR4/TRIF/IRF3-dependent signaling pathway [47]. Specifically in OC cells, mel therapy led to the downregulation of TRIF and IRF3.
The OC microenvironment is characterized by a dramatic increase in immunosuppressive cytokines, such as TNF-α and IL-6 [46]. These cytokines activate other inflammatory molecules that are capable of recruiting macrophages and dendritic cells (DCs) to the tumor site; these cells further contribute to angiogenesis, tumor growth, and metastasis [48]. Moreover, high levels of anti-inflammatory cytokines and related factors (e.g., TGF-β and VEGF) have been found in OC ascites [49]. Interestingly, we showed a significant reduction in IFNβ, TNF-α, and IL-6 expression in OC after mel therapy. By activating T helper (Th)-1 lymphocytes, mel enhances the production of cytokines such as IL-2, IL-6, IL-12, and IFNs [50], demonstrating its beneficial effects on immune cell modulation. In addition, the antiinflammatory actions of mel are mediated by multiple mechanisms, including a reduction in the levels of proinflammatory cytokines and chemokines (IFNs, TNF-α, IL-6, −8, and −12) and an increase in the levels of antiinflammatory cytokines (IL-10 and IL-1 antagonist). In OC cells, the immunomodulatory properties of mel may aid in inhibiting tumor cell growth and survival.
Conclusions
In summary, this study emphasizes the cross-talk between the TLR system and mel in ethanol-preferring rats with OC. Although long-term mel therapy exerts no influence on TLR2 signaling, it may modulate the TLR4signaling pathway in a MyD88-and TRIF-dependent manner, thereby dampening the inflammatory process induced by OC. Although EtOH alone did not promote additional inflammation in the OC microenvironment, mel therapy efficiently reduced the levels of important inflammatory factors (IkBα, NFkB p65, TRIF, and IRF-3) during EtOH intake. These findings suggest a basis for Figure 6 ELISA assays of the IFN-β (A), TNF-α (B), IL-6 (C), and IL-8(D) levels. Supernatants of ovarian samples (n = 20/group) displaying papillary carcinoma were assessed 60 days after the final dose of mel. a P < 0.01 vs. OC. IFN-β: interferon-β; TNF-α: tumor necrosis factor alpha; IL-6: interleukin-6; IL-8: interleukin-8. All results are expressed as the means ± SD. Two-way ANOVA followed by Tukey test was performed.
the therapeutic use of melatonin as a possible adjuvant in association with other anti-cancer therapeutic drugs.
Summary statement
Melatonin therapy attenuates the TLR4 signaling pathway in a MyD88-and TRIF-dependent manner, dampening the inflammatory process induced by OC in a preclinical model of ovarian carcinogenesis. These findings suggest a basis for the therapeutic use of melatonin as an adjuvant.
Clinical perspectives
TLR4-mediated signaling is associated with the immune escape of ovarian cancer cells, and inhibition of specific pro-inflammatory cytokines represents a highly important therapeutic strategy. Melatonin, an anti-inflammatory and immunomodulatory indoleamine, modulated the TLR4 signaling pathway, ultimately reducing the expression of inflammation-related mediators in these ovarian cancer cells, which serve as a novel and interesting preclinical model of ovarian cancer. These findings suggest a basis for the therapeutic use of melatonin as a possible adjuvant in association with other anti-cancer drugs under investigation in ongoing clinical studies. | 2023-01-16T14:08:09.692Z | 2015-02-06T00:00:00.000 | {
"year": 2015,
"sha1": "df354fc8285173cf93176937b7aa295ca402d54a",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.1186/s12885-015-1032-4",
"oa_status": "GOLD",
"pdf_src": "SpringerNature",
"pdf_hash": "df354fc8285173cf93176937b7aa295ca402d54a",
"s2fieldsofstudy": [
"Biology",
"Medicine"
],
"extfieldsofstudy": []
} |
235748342 | pes2o/s2orc | v3-fos-license | α-Synuclein: An All-Inclusive Trip Around its Structure, Influencing Factors and Applied Techniques
Alpha-synuclein (αSyn) is a highly expressed and conserved protein, typically found in the presynaptic terminals of neurons. The misfolding and aggregation of αSyn into amyloid fibrils is a pathogenic hallmark of several neurodegenerative diseases called synucleinopathies, such as Parkinson’s disease. Since αSyn is an Intrinsically Disordered Protein, the characterization of its structure remains very challenging. Moreover, the mechanisms by which the structural conversion of monomeric αSyn into oligomers and finally into fibrils takes place is still far to be completely understood. Over the years, various studies have provided insights into the possible pathways that αSyn could follow to misfold and acquire oligomeric and fibrillar forms. In addition, it has been observed that αSyn structure can be influenced by different parameters, such as mutations in its sequence, the biological environment (e.g., lipids, endogenous small molecules and proteins), the interaction with exogenous compounds (e.g., drugs, diet components, heavy metals). Herein, we review the structural features of αSyn (wild-type and disease-mutated) that have been elucidated up to present by both experimental and computational techniques in different environmental and biological conditions. We believe that this gathering of current knowledge will further facilitate studies on αSyn, helping the planning of future experiments on the interactions of this protein with targeting molecules especially taking into consideration the environmental conditions.
INTRODUCTION
Alpha-synuclein (αSyn) is a relatively small protein formed by 140 residues, which is highly expressed and conserved. It is typically found in the presynaptic terminals of neurons. Its primary sequence can be divided into three regions, as shown in Figure 1 (Fusco et al., 2014;Mori et al., 2020;Uversky and Eliezer, 2009) that are characterized by different physico-chemical properties due to their distinct aminoacidic composition. First, the N-terminal segment, (residues 1-60), shows numerous amphipathic 11-mer repetitions, and contains the consensus sequence KTKEGV. This is the αSyn region where most of the familial mutations are located. Then, the non-amyloid-β-component (NAC) central region (residues 61-95) is highly amyloidogenic giving the protein the ability to generate β-sheets. Finally, the C-terminal segment (residues 96-140) is rich in anionic residues and prevents αSyn aggregation by electrostatic repulsion.
In its native state, monomeric αSyn is unfolded, and thus is commonly considered as an intrinsically disordered protein (IDP). Yet, there is still a large controversy regarding αSyn secondary and tertiary structural tendencies and the data from literature are often conflicting. Changes in the environment conditions, mutations, interactions with endogenous and/or exogenous molecules can indeed induce αSyn to fold in different structures. αSyn misfolding and its subsequent aggregation into amyloid fibrils is a pathogenic hallmark of different synucleinopathies, such as Parkinson's disease (PD). As a consequence, the comprehension of αSyn structural and functional features is fundamental to progress in the study and finding of treatments for αSyn-related diseases.
Here, we provide a review on in silico and experimental data regarding the structural features of αSyn both in the WT form and in biologically relevant mutants. Moreover, we focus on different factors influencing αSyn structure, such as the biological environment, the interaction with lipids, with endogenous small molecules and proteins, as well as with exogenous compounds (e.g., drugs, diet components, heavy metals). We also discuss the different methods used to highlight αSyn structure in each case and the relation between the obtained results and the employed technique.
MONOMERIC WILD TYPE (WT) αSYN STRUCTURAL FEATURES
In 1996, Weinreb et al. observed that Wild-Type (WT) αSyn exists in solution as a dynamic ensemble of conformations lacking a single equilibrium structure and, therefore, classified it as an IDP (Weinreb et al., 1996). Many studies have advanced our knowledge in this field by applying experimental (for a recent review on NMR investigations see Kim et al., 2020) and computational (e.g., MD, Monte Carlo simulations) techniques (Jónsson et al., 2012), or a combination of both approaches (Brodie et al., 2019). However, due to αSyn structural heterogeneity that depends on many different biological and physico-chemical factors (Stephens et al., 2019), caution is needed when interpreting these results. To date, the general consensus is that monomeric WT αSyn is almost unstructured in solution (Fauvet et al., 2012). Anyway, variations of WT αSyn structural propensity can be detected. In order to rationalize the vast amount of literature data, we try to categorize them according to two different levels: global and local ( Table 1 and below).
Global-Level
Tertiary Structure Propensity αSyn is able to interconvert between multiple states of the dynamic ensemble of conformations (Weinreb et al., 1996). Nonetheless, at the global-level, different research groups have reached different conclusions as to whether the conformational ensemble in solution is on average: (1) likely to be compact and acquire a globular-like structure driven mainly by long-range intra-molecular electrostatic interactions, as illustrated in Figures 2 and 3 or (2) prone to exist as an extended random coil.
At physiological pH, WT αSyn has a very uneven distribution of physico-chemical properties along its sequence. The N-terminal region is amphipathic, the NAC region hydrophobic, and the C-terminal region highly negatively charged (Ilie and Caflisch, 2019). Flickering structural tendencies can be observed when viewing the hydrophobic effect as the major driving force for protein folding (Kauzmann, 1954). Based on this assumption, the contacts between the hydrophobic residues and the polar solvent are minimized. In turn, the regions formed by hydrophilic residues, such as the polar part of the N-terminal region and the C-terminal region, are expected to be more exposed to the cellular solvent and transiently interact with each other (Dułak et al., 2020). Experimental and computational techniques have suggested the presence of brief long-range intramolecular electrostatic interactions within αSyn structure (Dedmon et al., 2005b;Bertoncini et al., 2005;Allison et al., 2009;Fakhree et al., 2018;Brodie et al., 2019). Dedmon et al. and Yu et al. lack of agreement on the exact residues that form the αSyn intra-electrostatic contacts, nonetheless, both agree on the existence of such interactions between the residues present in the C-terminal domain and those located in the central part of the protein. Moreover, Bertoncini et al. (2005) reported that perturbation of these long-range naturally occurring interactions could lead to the exposure of the NAC region (residues 61-95) toward the cellular environment, potentially promoting αSyn oligomerization. Furthermore, in vivo and in vitro experiments have shown that the truncation of the monomeric WT αSyn C-terminal region can induce the formation of amyloid aggregates (Iyer et al., 2017;Vasili et al., 2019). Hence, several studies hypothesized that these detected intra-molecular contacts reduce the accessibility of the central part of the protein, preventing it from establishing inter-molecular interactions and inhibiting monomeric WT αSyn oligomerization and aggregation. As a consequence, some authors have rationalized the possibility of αSyn C-terminal region demonstrating a protective role against the formation of amyloid fibrils (Dedmon et al., 2005b;Bertoncini et al., 2005;Yu et al., 2015).
The temperamental nature of these aggregation-resistant globular conformations can be affected by changes in the environment; for example, changes in pH alters the distribution of charges throughout αSyn, which can lead to the loss of these transient intra-molecular electrostatic interactions. The C-terminal domain, (residues 96-140), presents a high content of acidic residues at physiological pH which are thought to play a major role inhibiting αSyn aggregation (Dedmon et al., 2005b;Bertoncini et al., 2005;Bhattacharya et al., 2019). Studies suggest that, this self-inhibition against FIGURE 2 | DMD centroids of the most frequent monomeric WT αSyn lowest energy clusters. Clusters representing a (A) ∼76%, (B) 15%, and (C) ∼4% of the overall population. αSyn N-terminal region (residues 1-60) is colored in orange, the NAC-region (residues 61-95) is colored in green and the C-terminal region (residues 96-140) is colored in yellow (Zhang et al., 2018).
Frontiers in Chemistry | www.frontiersin.org July 2021 | Volume 9 | Article 666585 4 fibrillation conformation can be lost when changing the pH from neutral to acidic (Plotegher et al., 2014).
Contrarily, Nath et al. observed that αSyn acquires a more compact conformation at low pH (Nath et al., 2012). However, this should be taken with caution as these are predictions and not conclusive observation. Moreover, some research groups point out that the interactions between the C-terminal region and the rest of the molecule is rather small and, therefore, the contacts established within the native structure provide limited protection against solvent exposure for the NAC region (Jónsson et al., 2012).
Further criticism suggests that monomeric WT αSyn acquires more extended or tail-like global conformations, which aligns with the fact that it is unstructured in solution. Zhang et al. reported the structural dynamics of αSyn in aqueous solution, demonstrating its ability to interchange its structure dynamically, mainly between the primary overall globular morphology and both one-tail and two-tail structures. These tails are parts of the protein that protrude from the main globular segment ( Figure 2) (Zhang et al., 2018). The tendency of αSyn to adopt a tail-like structure has also been reported by other researchers, based on MD simulations, Small-Angle X-ray Scattering (SAXS) and Electron Microscopy (EM) experiments (Tsigelny et al., 2012;Lorenzen et al., 2014). These transient tail-like structures are often seen in IDPs because they are implicated in diverse biological functions (Uversky, 2013).
Lastly, other studies stated that, in aqueous solvent, monomeric WT αSyn has a weak preference for adopting globular conformations (Weinreb et al., 1996;Ilie and Caflisch, 2019). For instance, Allison et al. observed that, over time, monomeric WT αSyn has a propensity to expand (Allison et al., 2009). Also, others have reported relevant clusters of αSyn monomers detected in their experiments presenting extended conformations (Jónsson et al., 2012).
The controversial results about the globular or extended preferences obtained by various groups may be ascribed to the rapid interconversion between conformers affecting αSyn and the use of different methodologies to carry out their investigations.
Overall Secondary Structure Propensity
Knowing that monomeric WT αSyn is unstructured in solution, the classification of its transient secondary structural elements can be useful for further investigations. Attempts to determine a structure of native αSyn has been mainly classified it as an all-α protein, whose secondary structure is composed exclusively of α-helices allowing a small number of isolated β-sheets ( Figure 2) (Bhattacharya et al., 2019;Cartelli et al., 2016;Meade et al., 2019;Zhang et al., 2018) or as an α+β protein when along the αSyn backbone α-helices and β-strands are intercalated (Figures 3 and 4) (Yu et al., 2015;Brodie et al., 2019).
Conversely, Jónsson et al. predicted that αSyn could adopt an all-β secondary structure in which αSyn was almost entirely composed by β-sheets with some peripherical small α-helices in several of the detected relevant conformers obtained.
They also reported that these results match with experimental data obtained at neutral pH and low temperatures, around 15°C (Jónsson et al., 2012). Zhang et al. (2018) found this conformational extended β-sheet pattern to be unfavorable in the WT αSyn monomeric state.
Some studies argue that folded helical conformers are not anticipated to be pathogenic (Meade et al., 2019) and impede amyloidogenic aggregation (Bhattacharya et al., 2019) whereas the presence of β-sheets drive this process. In fact, the design of novel small molecules or biological therapeutics to stabilize α-helical monomers is a strategy for blocking the neurotoxic pathway switching off β-sheet structure formation (Plotegher et al., 2014;Ciechanover and Kwon, 2015).
Local-Level
Since we face the problem of the lack of technical resources to irrefutably determine a series of conformations that full-length WT αSyn monomers can adopt in solution, attempts have been made to use available techniques to identify structural trends at the local-level, that is, if it even has a determinable structure. Despite its unstructured nature, αSyn can be analyzed in terms of its transient secondary structures. This allows us to hypothesize the conformational changes αSyn undergoes before the molecular aggregation process is carried out and identify possible targetsites that facilitate the design of drugs to avoid the formation of amyloid fibrils in earlier stages. As a matter of fact, due to the intrinsic dynamic equilibrium of this protein in solution, as FIGURE 4 | MD WT monomeric aSyn models. Centroid from the most relevant conformational clusters. This figure has been reproduced upon copy right permissions (Yu et al., 2015).
Frontiers in Chemistry | www.frontiersin.org July 2021 | Volume 9 | Article 666585 5 mentioned, different tools and techniques for proteins characterization capture different structural trends. However, there is an identifiable trend in which several research groups outline that the N-terminus of WT αSyn is prone to fold into a helical conformation, whereas the C-terminus contains many fragments found as random coils. There is less agreement as to whether the NAC region folds into β-sheets, which is the key secondary structure that directs protein aggregation, or whether it maintains a helical structure.
N-Terminal Region
Monomeric WT αSyn N-terminal region can adopt different transient secondary structural features in aqueous solution due to its intrinsically disordered nature. Several studies observed a tendency in the αSyn N-terminus to acquire a helical secondary structure (Vilar et al., 2008;Allison et al., 2009;Jónsson et al., 2012;Coskuner and Wise-Scira, 2013;Zhang et al., 2018;Bhattacharya et al., 2019;Brodie et al., 2019;Meade et al., 2019;Kim et al., 2020). This helical pattern has been proposed to be essential for vesicle and membrane binding Vasili et al., 2019). Hence, this local conformation is prone to be energetically favorable, especially in the presence of factors known to drive this helical structural feature, such as acidic negatively charged membranes (Vasili et al., 2019).
NAC Region
There is presently no clear agreement as to whether the NAC region (residues 61-95) adopts a helical or a β-sheet structure or, indeed, whether it acquires a structure at all. As this region is involved in triggering protein aggregation, its structure depends to a great extent on the environmental conditions. This is probably why it contains numerous distinct energetically favorable secondary structures.
A combination of experimental and computational approaches (Brodie et al., 2019) and NMR measurements (Eliezer, 2009) have seen a tendency of the αSyn NAC region to form β-structures ( Figure 3) (Kim et al., 2020). This supports the significance of the presence of these transient structures in the native protein, alluding to their resemblance to hairpins that form inter-molecular interactions in amyloid fibrils constituting the core of this mature fibrillar form of αSyn (Tuttle et al., 2016;Guerrero-Ferreira et al., 2018). In contrast, in previous NMR studies, it was not possible to detect free αSyn conformations that would lead to the formation of partially folded aggregation intermediates (Wu and Baum, 2010).
Since the αSyn aggregation process is extremely slow, ensemble solution techniques such as NMR may not succeed in identifying the molecules that are prone to drive this process because they may appear in very small percentages (Plotegher et al., 2014). Nonetheless, Zhang et al. (2018) via HS-AFM documented that the extended β-sheet pattern in the WT αSyn monomeric state is unfavorable.
Other experiments show the sporadic formation of helical structures in the NAC region ( Figure 2) Zhang et al., 2018)
C-Terminal Region
The C-terminus of IDPs has been identified as the most important region since it has numerous functionalities (Uversky, 2013). It follows that this protein area adopts different structures depending on the function that it is required to perform. The results of the investigations that have tried to structurally characterize αSyn in an aqueous medium, either by experimental or computational methods, indicate that the C-terminal end tends to present a random coil structure for the most part under physiological conditions (Jónsson et al., 2012;Brodie et al., 2019).
Despite this tendency to present fewer secondary structure elements than the other protein regions, a propensity to form β-structures (Eliezer, 2009;Jónsson et al., 2012;Yu et al., 2015) and helical elements (Lorenzen et al., 2014;Yu et al., 2015;Zhang et al., 2018) has been observed in some studies.
EFFECT OF PD MUTATIONS ON αSYN STRUCTURAL FEATURES
αSyn is a protein involved in PD, not only as the main component of Lewy bodies, but because of its several mutations observed in PD patients. It is well known that mutations can change the phenotype, having several effects on the structure of a protein.
Understanding how PD mutations affect αSyn structure and its functions is thus essential for gaining a profound understanding of the protein itself and for developing more effective pharmacological strategies.
A53T, A30P, and E46K Mutations
In 1997, Polymeropoulos et al. (1997) identified the A53T mutation in the αSyn gene in an Italian kindred and in three unrelated families of Greek origin with autosomal dominant inheritance for the PD phenotype. A year after, Krüger et al. (1998) reported the A30P mutation in the αSyn gene. A third mutation, namely E46K, was identified in 2004 (Zarranz et al., 2004). During the years, it has been highlighted that mutations could impact both the free state of αSyn and its aggregated form. In this context, studies were performed using different techniques such as NMR spectroscopy, CD, and FTIR.
Initial CD studies on WT αSyn and the first two identified mutations, A30P and A53T, showed that the three proteins lack a preferred conformation in solution (Conway et al., 1998;Narhi et al., 1999;Serpell et al., 2000). However, in 2001, by conducting NMR studies, Bussell and Eliezer reported, that the mutation A30P strongly attenuates the helical propensity of the N-terminus. They observed indeed a positive C α secondary shift, indicative of a significant preference for helical secondary structure in the WT 18-31 sequence, which was absent in mutant A30P. Conversely, A53T mutation leaves this region unperturbed, exerting a more modest and local influence on structural propensity . In particular, the A53T mutant exhibited a slightly enhanced local preference for extended, β-sheet-like conformations around the site of the mutation. Other NMR studies on the WT, A30P and A53T, revealed a similar β-sheet-rich core region spanning residues 38-94 in the sequence of the two mutants, whereas the C-terminus remained flexible and unfolded in both cases (Heise et al., 2005).
McLean et al. investigated the αSyn long-range interactions by fluorescence resonance energy transfer (FRET). They reported, for both the WT and mutant A53T, a weak interaction between the N-terminal and C-terminal regions, whereas for mutant A30P they observed a statistical increase in the magnitude of FRET signal, indicating a closer vicinity between the N-and C-terminal regions (McLean et al., 2000).
In 2007, Fredenburg et al. reported a similar random coil secondary structure for both E46K and WT αSyn when free in solution, as highlighted by CD experiments (Fredenburg et al., 2007). In 2009, Rospigliosi et al. studied the effect of mutation E46K on the long-range interactions by paramagnetic relaxation NMR(PRE) and residual dipolar coupling (RDC) measurements. Surprisingly, no decrease in long-range contacts was detected in the mutant E46K with respect to the WT. Furthermore, an increased interaction between the C-terminal tail, the NAC and the N-terminal regions was observed. The same experiments on A30P and A53T did not indicate any changes in the long-range structure. In the same work, the authors observed a slight increase in local helix propensity in the area immediately adjacent to the mutation of mutant E46K, by calculating its C α chemical shifts deviations in comparison to the deviations of the random coil ones (Rospigliosi et al., 2009). Kumar et al. used Molecular Dynamics (MD) to analyze the mutations A30P, A53T and E46K in water under explicit solvent conditions. These mutants showed variations, more specifically their RMSD scores were 0.529, 0.534, and 0.486 respectively, in their secondary structure compared to WT micelle-bound αSyn (PDB ID 1XQ8) simulated in sodium dodecyl sulfate (SDS) ( Figure 5). The secondary structure of A53T recorded in this study was similar to that determined by quenched hydrogen/ deuterium exchange NMR spectroscopy which states that five β-strands appear in the amyloid state of αSyn (Vilar et al., 2008;Kumar et al., 2009).
Passing from the last decade to the current one, computational techniques started being more intensively employed to shed light on the structures of the WT and the mutants. In 2011, Balesh et al. (2011) performed classical MD and annealing MD (AMD) simulations and reported similar helical and β-sheet contents for the WT and A53T mutant-type αSyn proteins. At the same time, A53T presented a more compact structure. In 2013, Coskuner and Wise-Scira performed all-atom replica exchange molecular dynamics (REMD) simulations on the full-length monomeric WT and A53T mutant-type αSyn proteins in aqueous solution utilizing implicit and explicit water models. From these results, they observed that the helical content is minimally affected by the mutation A53T except for a few residues in the N-terminal and C-terminal regions. Additionally, in contrast, to previous computational works (Kumar et al., 2009) they reported an increase in the β-sheet formation close to the mutation site in the N-terminal region .
In the same year, a similar MD study was published on mutant A30P by Wise-Scira et al., reporting that the mutation has local as well as long-range effects on the protein structure. More specifically, the helical content of region 18-31 is less prominent in mutant A30P than in the WT protein. The β-sheet structure abundance decreases in the N-terminal region upon mutation A30P of the WT αSyn, whereas the NAC and C-terminal regions possess larger tendencies for β-sheet structure formation. Long-range intramolecular protein interactions are less abundant upon mutation A30P, especially between the NAC and C-terminal regions, leading to a less compact and less stable structure with respect to the WT (Wise-Scira et al., 2013).
Recently Discovered Mutations
In 2013, a fourth mutation, namely H50Q, was identified (Appel-Cresswell et al., 2013;Kiely et al., 2013). Far-UV CD studies demonstrated that also the H50Q variant is a primarily unfolded protein in aqueous buffers (Chi et al., 2014;Ghosh et al., 2013;Khalaf et al., 2014). Also, Chi et al. (2014), by using heteronuclear single quantum coherence (HSQC) NMR observed that the chemical shifts of most residues between the WT and H50Q were unperturbed, although the C-terminal region of H50Q is more flexible than that of the WT. On the contrary, Ghosh et al. noticed chemical shift perturbations between WT αSyn and H50Q, by conducting the same experiments. In fact, they observed quite significant chemical shift perturbations in the mutation area and in the C-terminal region (Ghosh et al., 2013).
In 2014, a fifth mutation, G51D, was discovered (Kiely et al., 2013;Lesage et al., 2013). Fares et al. performed CD experiments where the WT and G51D proteins exhibited the same random coil secondary structure. The 1 H, 15 N-HSQC studies confirmed the lack of a preferred conformation for both proteins, while the analysis of the secondary structure propensity via C α secondary shifts deviations showed no significant loss or gain of secondary structure compared to the WT. Furthermore, it was observed that the mutation G51D also does not significantly perturb transient long range contacts between N-and C-termini (Fares et al., 2014).
In the same year, mutation A53E was identified in a Finnish family (Pasanen et al., 2014). Ghosh et al. performed NMR studies with the WT, A53T, and A53E αSyn. Their data showed approximately similar spectra of the WT, A53T, and A53E with relatively narrow dispersions in the proton dimension for all proteins, characteristic for unfolded structures. The chemical shift differences, however, suggest perturbation of chemical shifts for residues surrounding the A53E mutation site, as already observed for the other mutants. Significant chemical shift changes were also observed for the residues at the extreme C-terminus of αSyn. In contrast to chemical shift perturbation data, the secondary structural propensity did not show any major alteration due to mutation A53E or A53T (Ghosh et al., 2014).
Comparative Experiments on all Known Mutated Sequences
Recently, Tsigelny et al. generated by MD multiple structural conformations of the WT and all the different mutants, by developing a new combined modeling approach. In the beginning, they simulated WT αSyn and mutant conformers creating a 20-ns interval MD snapshot ( Figure 6). From their analysis it can be deduced that the general α-helical content does not change more than 20% in all cases and that these α-helices transform into turns and loops within specific regions for each mutant over the 100 ns of the MD (Tsigelny et al., 2015).
In 2020, Okuwaki et al. examined all the NMR parameters, including the chemical shift and amide-proton exchange of the WT and the mutants. They observed in WT an α-helix structure in the 18-31 fragment, and a β-structure at the C-terminal region 120-140. The β-structure was destabilized by the mutations A30P and A53T. On the other hand, the α-helical structure might be stabilized by these mutations (Okuwaki et al., 2020).
Taken together, these data seem to point out that, among all the observed PD mutations, only A30P affects the overall αSyn structure. In addition, long-range interactions are less abundant. The contact between N-and C-terminal regions is thus perturbed and it might facilitate the aggregation.
EFFECT OF THE BIOLOGICAL ENVIRONMENT ON αSYN STRUCTURAL FEATURES
From the previous paragraphs, it can be deduced that the structure of the monomeric WT αSyn protein in solution tends to acquire diverse transient and dynamic conformations. αSyn will be likely to adopt specialized conformations upon different conditions (e.g., changes in pH, temperature, ionic strength, closeness to surfaces, etc. . .) in order to carry out certain biological or pathological functions. Hence, even though the study of WT αSyn conformation alone is useful, a more profitable course of action is to observe the conformational changes induced by different biological and physico-chemical factors triggering structure modifications. In Frontiers in Chemistry | www.frontiersin.org July 2021 | Volume 9 | Article 666585 8 this context, the interaction with endogenous molecules is an important factor to consider. It is commonly accepted that αSyn can bind lipids and phospholipids, as well as several proteins. Here below in Figure 7, the general features are presented, together with several highlights on the interaction with endogenous small molecules. , which have an important role in the conformational change of αSyn from being an intrinsically disordered protein (IDP) to induce a more stable structure. Right panel: special emphasis is placed on the type of study performed, classified in computational, in vitro and in vivo experiments, and on the particular interaction of the proteins with the distinct αSyn configurations (fibrils, oligomers, monomers). Hsp70 and Hsp73 heat shock protein 70 and 73; SCGN: secretagogin. Left panel: The equilibrium between horseshoe and linear conformation is highlighted and the partition of αSyn in three regions is depicted to show the different behavior throughout the sequence when in contact with lipid membranes. The important interaction with mitochondrial membranes and lipid rafts is also mentioned here. GM1: monosialotetrahexosylganglioside.
Lipids
In living organisms, lipids are mainly used as structural components in cell membranes, as energy stores or as signaling molecules. Various studies have investigated the possibility of αSyn monomer to bind lipids, in particular lipid membranes (plasma and mitochondrial membrane, axonal transport vesicles) (Sung and Eliezer, 2018). In the following paragraphs, we will present an overview of the most recent results, focusing our attention on the protein structure and the innovative strategies and techniques used to obtain these outcomes.
Phospholipids
Based on the specific binding properties to lipid layers and the location in synaptic nerve endings, the physiological function of αSyn has been related to circulation and transport of synaptic vesicles (Burré, 2015). Nevertheless, besides its physiological role in the synaptic transmission, the interaction with lipids can also lead to structural changes undergoing aggregation and contributing to amyloidogenesis. Membranes have been reported to both accelerate and inhibit αSyn fibril formation. In fact, the helical fold has been suggested to stabilize the protein and prevent aggregation by hindering the structural transition to β-sheet (Högen et al., 2012) but, at the same time, the helical state has also been proposed as being an intermediate in the aggregation process because it stabilizes intermolecular interactions through hydrophobic contacts . In the latter case, cell membrane surfaces would act as a fibrillation template favoring nucleation and participating in the fibrillation cascade while the NAC region would be essential for the self-polymerization of the protein (Pineda and Burré, 2017;Martial et al., 2019;O'Leary and Lee, 2019). A fact that argues in favor of the role of αSyn/lipid interaction in the etiopathogenesis of synucleinopathies is that all missense mutations responsible for familial PD (e.g., A30P and E46K) are localized in the 11-residue repeat domain; indeed, these mutations alter the lipid binding properties modifying membrane interaction (Bodner et al., 2010;Robotta et al., 2017). Therefore, it is particularly critical to understand how this interaction can regulate the equilibrium between the soluble intrinsically disordered monomer and the structured membrane-bound monomer/oligomer in vivo.
The first preliminary hypothesis about αSyn binding to membranes was developed by Davidson et al. (1998), who reported αSyn binding to acidic small unilamellar vesicles (SUVs), underlining the importance of membrane charge and curvature. In particular, an increase in α-helicity from 3 to 80% was measured. Browne and coworkers first tried in 2001, by means of modern multi-dimensional heteronuclear NMR spectroscopy, to characterize the conformational properties of αSyn as a free monomer and when bound to lipid-mimetic SDS detergent micelles and lipid vesicles. A prevalent disposition toward α-helical conformation in the N-terminal region was suggested in the free monomer in comparison to the C-terminus that, on the contrary, displays a highly unfolded and extended structure. Not surprisingly this tendency could be fulfilled after association to phospholipids, showing an extended α-helical structure stretching among residues 1-100 . Few years later, this idea of an extended conformation was revised by Chandra et al., who asserted that the N-terminal region, interacting with SDS, surprisingly configures itself in two helical regions that are interrupted by a short break around residues 43-44, as demonstrated by NMR studies and proteolysis experiments. This interruption has been explained with a more favorable binding of hydrophobic residues to the interior of the membrane or, alternatively, a more advantageous binding to highly curved vesicles (Chandra et al., 2003). By means of following studies based on EPR, Jao et al. could successfully provide important details about αSyn interaction with lipid bilayers, emphasizing the influence of the membrane features on the conformation of the membrane bound αSyn. The authors observed an extended, curved α-helical structure that is significantly different from the antiparallel helices formed in the presence of the detergent SDS (Jao et al., 2008). However, the experimental evidences provided by EPR are also consistent with different binding modes, involving an extensive membrane rearrangement, as suggested by Bodner et al. (Bodner et al., 2009). In this context of contradictory results, Robotta et al. (2011) presented αSyn as a coexistence of the two conformations: extended α-helix and horseshoe, i.e. two antiparallel α-helices, even if with a preference toward the extended form. The authors obtained these results by site-directed spin labeling in combination with pulsed electron paramagnetic resonance on large unilamellar vesicles (LUVs) and they concluded that the two conformations are closely related to the experimental conditions used and that the equilibrium is very labile, which means the molecule is highly flexible. In this way, they could explain why previous studies were in opposition (Robotta et al., 2011).
The interconversion between these two states has been represented as functionally relevant to the protein; in fact, physiologically, αSyn could effectively connect a synaptic vesicle to the plasma membrane by switching from an extended state to a broken-helix conformation (more tightly bound state). For this purpose, in order to characterize the extended helical structure by high-resolution solution-state NMR, a fluorinated alcohol (HFIP) has been employed, able to induce a highly helical state. Indeed, the central region corresponding to the non-helical linker displays a certain instability in the helical structure suggesting the possibility of this transition (Sung and Eliezer, 2018).
Solid state NMR (ssNMR) helped in providing new insights in the structural conformation of the membrane-bound αSyn. In fact, since the N-terminus is tightly bound to the lipid bilayer, these residues cannot be identified by solution NMR because they are invisible. The results of these experiments on acidic SUVs indicated that residues 6-25 are tightly bound to the membrane and no differences are detected when 1,2-dioleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] (DOPG) SUVs are used, even if the calculated affinity between αSyn and these vesicles is higher. We already know that the most dynamic part of the molecule is identified with the C-terminal region and INEPT (insensitive nuclei enhanced by polarization transfer) MAS (magic-angle spinning) measurements enabled to also characterize this domain as highly unstructured. In general, three domains could be identified: an anchoring N-terminal region (6-25), followed by an intermediate dynamic domain and an Frontiers in Chemistry | www.frontiersin.org July 2021 | Volume 9 | Article 666585 unstructured C-terminal domain that only transiently interact with the membrane surface. The central region is shown to be critical in modulating the affinity for the membrane surface and it is subsequently called membrane "sensor." Furthermore, MAS measurements indicated that the binding occurs at the surface of the membrane and not in the membrane bilayer (Fusco et al., 2014). Following ssNMR experiments helped also to understand the contribution of N-terminal acetylation on αSyn. This posttranslational modification leads to a stronger membrane affinity and an increased propensity to adopt α helical structures in the N-terminal region. According to Runfola et al., N-terminal acetylation seems to regulate the binding affinity of αSyn for synaptic vesicles without altering the structural properties of the bound state (Runfola et al., 2020). Considering all these determinant aspects affecting the binding of αSyn to membranes, one can easily understand how the protein binding is sensitive to the experimental conditions used. In particular, a physiological environment should be used to mimic the naturally occurring features and obtain an ultimate description of the αSyn monomer when bound to membranes. In this framework, another factor to be taken into consideration is the influence of calcium ions on the membrane binding propensity of αSyn. This ion, localized at the presynaptic terminals, is able to bind to the C-terminus and favors its binding to lipid membranes, as verified by CEST-NMR experiments, leading to the so-called "double anchor mechanism" emphasizing its role in neurotransmitter release (Lautenschläger et al., 2018). Summarizing, experimental evidences of association with lipid membranes support the strong dependence of the binding on the lipid composition and surface curvature. In general, αSyn binding to membrane is based on electrostatic interactions between the cationic groups of the basic N-terminal region (rich in Lys residues) and the anionic phospholipids, which in fact represent excellent models to mimic synaptic vesicles.
Similar considerations can be done for the αSyn fragment 71-82, included in the NAC region, as observed by Bédard et al., who described its role on the structural and assembly behavior of αSyn. As deduced from CD and IR measurements, in the presence of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) membranes, the fragment is mostly disordered as it is in solution, but when in contact with negatively charged membranes (1-palmitoyl-2-oleoylglycero-3-phosphoglycerol, POPG) the peptide adopts an intermolecular parallel β-sheet configuration (Bédard et al., 2014). In a more recent study, its behavior has been analyzed in the presence of partially anionic membranes to mimic in the best way neuronal membranes and an in-register configuration could be validated by means of IR and ssNMR DQF-DRAWS experiments. The amyloid aggregation driver is the electrostatic interaction as it happened also with the N-terminal sequence (Martial et al., 2020).
In addition to the curvature degree, the interaction of αSyn and membranes is also regulated by lipid packaging. In view of this, Stöckl et al. proved by confocal microscopy that αSyn preferentially interacts with liquid-disordered giant unilamellar vesicles (GUVs): the binding requires anionic lipids in a liquid disordered state, and this is in good correlation with the synaptic vesicles composition (Stöckl et al., 2008). This is also valid for fragment 71-82 (Martial et al., 2020). A comprehensive model for the interaction of αSyn with lipid bilayers has been proposed by Ouberai et al. based on many converging independent studies and new results generated by the combination of dual polarization interferometry, atomic force microscopy and CD spectroscopy. Connecting to membranes with strong curvature and stressed surfaces (cone-shaped lipids), αSyn monomers are apparently able to close the packing defects. In fact, after binding of αSyn to the phospholipid polar heads and insertion of the hydrophobic residues, lipids are induced to laterally expand provoking membrane remodeling and this process is promoted in the presence of packing defects or imperfections (Ouberai et al., 2013). αSyn ordering effect on the membrane has been also investigated by fluorescence anisotropy and it has been concluded that this is concentration dependent and it occurs in the liquid-crystalline state and not in the gel phase. This means that αSyn is able to stabilize the membrane of synaptic vesicles and thereby can be essential to prevent the premature vesicle fusion to the presynaptic membranes (Pirc and Ulrih, 2015). The higher binding affinity to fluid compared to gel phases has been also investigated by Galvagnion et al. by means of CD and DSC studies, suggesting that the higher exposure of hydrophobic area is essential for the binding. Notably, the authors also asserted that shorter and more soluble lipids greatly improve αSyn aggregation and, consequently, its pathological effect (Galvagnion et al., 2016).
Together with the lipid composition, the lipid to protein ratio is a discriminant factor for amyloidogenesis, being able to switch the equilibrium between physiological and pathological paths, as first described by Galvagnion et al. (2015) and later discussed in a detailed review by Kiechle et al. (2020). When this value is high, due to the low local concentration of αSyn, the aggregation can be suppressed. On the contrary, when this value is low or intermediate, αSyn-bound monomers could lead to nucleation and amyloid formation (Terakawa et al., 2018).
Although αSyn has a mainly cytosolic distribution, its ability to adhere to cell membranes, predisposes it to have other cellular localizations. A certain selectivity of αSyn toward mitochondrial membranes has been observed and this propensity has been related to the abundance of the phospholipid cardiolipin. Nevertheless, cardiolipin is mostly present in the inner membrane of the mitochondria. It has been demonstrated that αSyn enters mitochondria via import channels and not via direct interaction with the lipids of the outer membrane and afterward it is localized in the inner membrane (Zigoneanu et al., 2012). On the other hand, recent studies demonstrated that cardiolipin translocates to the outer mitochondrial membrane in response to cellular stress and binds αSyn species. In this position, cardiolipin can also pull αSyn monomer away from oligomeric/fibrillar aggregates and facilitate its refolding in α-helix. Cardiolipin exposure is therefore a key signal in PD pathogenesis (Ryan et al., 2018). In agreement to these results Ghio et al. also demonstrated that cardiolipin enhances αSyn lipid membrane binding and also favors the membrane pore-forming activity of αSyn oligomers (Ghio et al., 2019).
Lipid Rafts
In general, it seems that αSyn specifically binds to anionic phospholipids, when these are embedded in liquid-disordered domains (Stöckl et al., 2008). Nevertheless, various studies demonstrate that lipid rafts can also have a very important role in αSyn binding. Lipid rafts are specialized areas of the plasma where tightly packed cholesterol and sphingolipids accumulate, surrounded by more fluid phospholipids. In fact, these dynamic microdomains adopt a liquid-ordered state and float in the remaining liquid-disordered plasma membrane (Sezgin et al., 2017). Fortin et al. demonstrated by a double fluorescent labeling that αSyn specifically associates with lipid rafts and this interaction can be crucial for its synaptic localization and physiological function (Fortin et al., 2004). Furthermore, many other publications illustrated how lipid rafts are closely connected with neurodegenerative diseases (Sebastião et al., 2013;Canerina-Amaro et al., 2019;Mesa-Herrera et al., 2019;Grassi et al., 2020).
In this framework, the analysis of αSyn interaction with cholesterol and gangliosides is fundamental since both have been considered as critical elements that could synergically favor the insertion of αSyn in lipid rafts and influence its pathological and physiological function (Fantini et al., 2011).
Cholesterol
Together with phospholipids, cholesterol plays an important role in regulating permeability and fluidity of the membrane. As expected, it also interacts with αSyn modulating its binding to synaptic-like vesicles, cholesterol being a very important component of these structures (Pfrieger, 2003). In particular, two domains of αSyn were recognized by Fantini et al. to bind cholesterol. Especially, residues 67-78 display a high affinity binding with a tilt angle of 46°, as measured by MD. Notably, they asserted that the tilted peptide could probably insert in the membrane and intercalate with the apolar regions of cholesterol leading to a higher affinity. On the contrary, residues 37-43 probably just associated to the hydroxyl group of cholesterol (Fantini et al., 2011). Other authors showed how cholesterol reduced or completely blocked, depending on the concentration used, αSyn binding to non-anionic membranes but, at the same time, this effect was much lower in the presence of negatively charged membranes (Shvadchak et al., 2011). Recently, surface plasmon resonance (SPR) was employed to measure the binding of αSyn monomers to lipid vesicles and this resulted decreased with the addition of cholesterol molecules to the membrane composition. The effect was detected also in the presence of negatively charged vesicles (Jakubec et al., 2019). A very recent publication from Fusco and coworkers confirmed these results showing by CD experiments that αSyn binding is reduced in the presence of cholesterol. However, the weaker interaction was detected by CEST (chemical exchange saturation transfer) experiments only at the NAC region. The previously described property of αSyn of binding two different membranes at the same time, the "double anchor mechanism", has been also evaluated by DLS (dynamic light scattering) and the ability of αSyn to interact with two vesicles was promoted with the increasing concentration of cholesterol showing that the NAC region is effectively crucial in this step modulating this important biological property (Man et al., 2020). At the same time, Jakubec et al. questioned whether cholesterol could influence αSyn fibrillation and they observed that the aggregation was effectively promoted by analyzing ThT (thioflavin T) and TPE-TPP (bis(triphenylphosphonium) tetraphenylethene) fluorescence assays. This outcome could be explained taking into consideration that cholesterol could act as a nucleation site (Jakubec et al., 2019).
Glycosphingolipids
The interaction of αSyn with glycosphingolipids, in particular gangliosides as GM1, has been reported in many articles, revealing their key role in the physiological and pathological function of this protein (Chiricozzi et al., 2020). In 2006, Martinez et al. concluded from SEC (size-exclusion chromatography) HPLC, CD and TEM (transmission electron microscopy) that αSyn displays a very high binding affinity and specificity toward GM1, in comparison to the other gangliosides (Martinez et al., 2007). Differently from cholesterol, the high-affinity binding site of αSyn to glycosphingolipids includes residues 34-45 (Fantini et al., 2011). This affinity is even intensified when αSyn is N-acetylated and at the same time the fibrillation is reduced together with enhancement of the helical folding propensity (Bartels et al., 2014). Based on these outcomes, Schneider et al. reported that GM1 displays neuroprotective effects after in vivo administration with a decreased αSyn aggregation (Schneider et al., 2019).
Proteins
The investigation of the interplay between αSyn and proteins is of high relevance for understanding both the physiological and the pathological role of αSyn. To see how this interaction influences its monomeric structure, in the following paragraphs and in Table 2 we will summarize various biomolecules and how they affect not only its conformation but also its aggregation tendency.
Tubulin
Tubulin is a highly conserved αβ dimeric protein that is the main component of microtubules. αβ-Tubulin dimers assembly and disassembly are finely tuned within the cell and a huge number of proteins interact with them, affecting the stability of microtubules and their function. Recently, it has been found that αSyn binds to microtubules and tubulin α 2 β 2 tetramer. This interaction induces helical αSyn folding, enabling it to promote microtubule nucleation and to enhance microtubule growth rate and catastrophe frequency. On the other hand, PD αSyn mutants do not undergo tubulin-induced folding, causing tubulin aggregation rather than polymerization (Cartelli et al., 2016). However, the precise sequence of α-Syn binding site to tubulin has not been fully elucidated yet, and molecular studies aimed to deciphering the interaction at an atomic level are still missing.
Heat Shock Protein 70 (Hsp70) and Heat Shock Protein 73 (Hsp73)
Hsp70 is a 70 kDa protein from the "chaperone" family, involved in cell defense against protein misfolding (Figure 8) interacts with αSyn fibrils instead of monomers. These results suggest that the protein adopts a folded structure which protects the central hydrophobic region and does not allow further intermolecular binding. As shown by NMR data, this happens when the C-terminal domain makes contacts with the NAC region. If these interactions are perturbed (early stages of aggregation), the central region becomes exposed and this can lead to proteinprotein interaction with the formation of pre-fibrillar aggregates.
In this case, the chaperone binds to these aggregates and prevents the fibrils formation (Dedmon et al., 2005a). The hypothesis of Hsp70 interacting with the NAC region of αSyn is sustained also by Luk et al. As previously mentioned, this element represents the core of αSyn fibrils, and it contains the sequence required for αSyn to aggregate. Interestingly, ThT assay demonstrated that αSyn residues 95 VKKDQ 99 , at the border between NAC and the C-terminal domain, are crucial for interaction with Hsp70 (Luk et al., 2008). Regarding monomers, Hsp70 is able to modify αSyn conformation by forcing it to a different open conformational state in which the N-and C-termini are distant from each other. αSyn-αSyn interactions are observed but are probably modified and NAC-NAC domain interactions among monomers are lost, increasing αSyn solubility (Klucken et al., 2006).
Another important chaperone is Hsp73 (Hsc70). Concerning its interaction with αSyn, Chaari et al. found out that the chaperone binds αSyn at the peptide binding sub domain (SBSD) corresponding to residues 386-509. These interactions involve unfolded monomers, and this is coherent with the role of Hsp73, which normally binds to unfolded proteins to mediate their refolding. At the same time, the helical subdomain (510-646) stabilizes the chaperone/αSyn complex, counteracting the formation of nuclei and/or the elongation of fibrils, as shown by in vitro experiments (Chaari et al., 2016).
DNAJB6 and DNAJB1
DNAJB6, the co-chaperone of Hsp70, is able to counteract αSyn and amyloid β aggregation in vitro by combining with its partner (Månsson et al., 2014). In particular, its effect is linked to the J domain, which catalyzes the transfer of the misfolded αSyn to the chaperone. Furthermore, post-mortem analysis on PD patients' brains reveals the presence of the protein in Lewy Bodies, suggesting that its misregulation may provide early PD onset. Finally, this may reveal an interaction of DNAJB6 with αSyn and its direct role in aggregation inhibition. However, this hypothesis needs to be proved in vivo (Aprile et al., 2017). Focusing on the DNAJB family, DNAJB1 efficiently works with Hsp70 and Hsp110 in fibrils disassembling. The system binds pre-formed fibrils both in vitro and in vivo, converting them into shorter fibrils later depolymerized into monomers (Gao et al., 2015). However, the interaction sites on αSyn still remain unknown.
Secretagogin (SCGN)
Studies suggest that neurodegeneration may be associated with Ca 2+ dis-homeostasis, since a misregulation in this ion signaling system can be detected in neuropathologic patient brains. Considering this, scientists from Chidananda research team focused on SCGN, a Ca 2+ -sensor protein expressed in the brain which plays a key role in insulin regulation (Chidananda et al., 2019).
To study its effect over αSyn, the authors developed a method able to lead to protein fibrillation with entities of the range of 5-10 nm. Notably, TEM analysis revealed that no fibril was formed when αSyn was incubated with SCGN. These results are explained by considering that SCGN can bind both to monomers and early-stage oligomers, according to ThT and MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assays. In this case, soluble αSyn is preserved, without any further aggregation (Chidananda et al., 2019).
All in all, SCGN is shown to bind to αSyn and prevent it from fibrillation and nucleation in vitro. This may impede its binding to membranes, its misfolding and its aggregation. Finally, NMR studies show that anti-fibrillar activity is attributed to the central region and C-terminal domain of SCGN (Chidananda et al., 2019).
AS69
AS69 was engineered from Mirecka et al., who developed a new phage library, obtained by random mutagenesis of the gene encoding ZAβ3. (Mirecka et al., 2014). This protein was proven to be an efficient Aβ 1-42 aggregation inhibitor. In particular, due to its structure it is classified as "β-wrappin." This protein shows two identical subunits, each formed by two α-helix and one β-strand spanning residues 13-58, linked by a disulphide bond involving the Cys28 residues of both of them. Moreover, NMR analyses showed that Phe31 residues of both AS69 subunits are involved in π-stacking interactions with Tyr39 and His50 of αSyn. Furthermore, molecular modeling studies suggest that AS69, by interacting with αSyn, folds into two β-strands and four α-helices forming a hydrophobic cavity where αSyn is buried (Figure 9) (Mirecka et al., 2014).
ThT analysis suggests that AS69 binds stoichiometrically to αSyn monomers, thus blocking the fibril elongation step by sequestrating free monomers. Also, the complex αSyn/protein can act as an inhibitor of the secondary nucleation process. Together, these results suggest that AS69 may display a broad activity against fibrillation, as demonstrated both in vitro and in vivo (Drosophila flies and mice) (Agerschou et al., 2019).
Endogenous Small Molecules
The role of endogenous small molecules (e.g., neurotransmitters) is important when it comes to understanding the function and structure of amyloids. Interestingly, some neurotransmitters are able to alter αSyn folding while interacting with it, which enables to better understand how the protein behaves and which binding sites are pivotal in that circumstance. The structures of the endogenous small molecules that will be reviewed in the next paragraphs are represented in Figure 10 and the effect are summarized in Table 3.
Dopamine (DA)
Dopamine, one of our principal neurotransmitter, is a catecholamine implicated in several physiological process whose biosynthesis decreases in neuropathologies, like PD. Its role in αSyn aggregation modulation has been widely discussed and there is not a clear consensus whether it has a direct or indirect implication. In fact, some researchers hypothesize that DA can decrease αSyn fibrillation and oligomerization by binding to the protein via hydrophobic and hydrophilic interactions. These lead to non-stable complexes, which include its NAC or C-terminal region. Furthermore, studies showed that DA can mediate anti-fibrillar effect both in vitro and in vivo, while forming off-pathway oligomers (Oliveri, 2019).
The role of dopamine concerning αSyn modulation has been explored by Rekas et al. SAXS data suggest that the catecholamine mediates the formation of trimers made by αSyn overlapped structures. CD data suggest that their structure lacks β-sheets, which are crucial for amyloid aggregates (Rekas et al., 2010).
Recently, Post et al. reviewed the interaction between DA and αSyn, providing features over the protein structure and its binding sites. In particular, oxidized DA can interact with αSyn, producing a complex which enhances the formation of oligomers rather than fibrils (Post et al., 2018). Moreover, in vitro studies underline that its formation is due to a non-covalent binding between DA and the 125 YEMPS 129 region of αSyn (Mazzulli et al., 2007). Furthermore, this complex seems to be stabilized by a salt-bridge between DA and E 83 in the NAC region (Post et al., 2018).
Finally, the binding of DA to αSyn has an important effect on the conformation of the protein domains. In fact, fluorescence lifetime imaging microscopy data showed that the N-and Ctermini of αSyn come closer, adopting a conformation which may inhibit fibril formation.
Arginine
Arginine is an amino acid able to affect αSyn behavior. This natural compound is well-known for its neuroprotective effect both in vitro and in vivo against glutamate excitotoxicity.
Regarding arginine/αSyn interaction, this molecule can inhibit protein late state aggregation, according to ThT, DLS and AFM (Atomic Force Microscopy) data. Isothermal calorimetry (ITC) and MS (Mass Spectrometry) analyses show that arginine binds to αSyn, forcing it to acquire a conformation thought to slow down the early-stage oligomerization. From a structural point of view, this conformer leads to a unified compaction of unfolded monomers. Since this intermediate is stabilized by clusters of arginine, the oligomerization and further fibrillation are avoided. Furthermore, MALDI-TOF mass spectrometry, ITC (Isothermal Titration Calorimetry) and MD analyses show that the aromatic residues of αSyn and the guanidine moiety of arginine interact via cation-π forces. Finally, arginine protective effect against αSyn toxicity was also proven in HeLa and SH-SY5Y cells line (S. Ghosh et al., 2018).
Glutamate
Glutamate is an excitatory neurotransmitter whose concentration in blood is around 50 µM. In the brain, it is the precursor of glutamine in presynaptic terminals and glial cells (Ghosh et al., 2018). Importantly, glutamate is shown to influence αSyn conformation and promote its aggregation. However, as an osmolyte, it tends not to directly interact with the protein; thus, its activity on αSyn may derive from its exclusion from the protein surface. The impact of glutamate on the conformation of αSyn is shown in in vitro assays. Interestingly, the more the concentration of glutamate is increased, the more unfolded monomers convert into β-sheet rich oligomers. In particular, small oligomers (10-15 nm diameter) predominate when glutamate is present at a concentration of less than 100 mM. Furthermore, AFM analysis proved that in glutamate treated samples, after 3 h of incubation two kinds of oligomeric aggregates appeared. The most representing one had a diameter of 20-35 nm, while the second one of 60-85 nm. Finally, this early stage oligomerization could be a critical factor to enhance fibrillation (Ghosh et al., 2018).
EFFECT OF EXOGENOUS FACTORS ON αSYN STRUCTURAL FEATURES
In general, the interaction of IDPs with exogenous compounds plays a crucial role for conformational stabilization and induction of aggregation. These chemicals can be found in the daily diet (e.g., flavonoids) or can derive from pharmacological treatments or habits (e.g., nicotine from smoking). Recognizing which elements are essential, beneficial or toxic is a very important topic, displaying each substance a bivalent effect as summed up by the sentence "The dose makes the poison", asserted by Paracelsus. This is valid for every element, including metals, which have crucial physiological roles but at the same time can induce toxicity according to their therapeutic window. Thus, by analyzing the interaction between αSyn monomers and those molecules, a better insight of αSyn structural changes can be given. Finally, thanks to modern spectroscopy and molecular dynamics, the sites of interactions can be investigated. In the end, these data will help to better comprehend the structure of αSyn, whose details have not yet been fully elucidated. In this context, the main classes of chemicals that interact with αSyn are presented below ( Figure 11).
Metals
Metals are everywhere. These so-called trace elements, have an indispensable physiological role in normal brain functions, being often used by enzymes and proteins, due to their redox potential (Garza-Lombó et al., 2018). On the other hand, many recent epidemiological studies detected a significant higher level of metals in the affected brain regions of Parkinson's disease FIGURE 9 | (A) Ribbon drawing of the AS69 structure interacting with the αSyn β-hairpin (residues 36-55). In light orange and green, the two subunits of AS69 are shown. Each subunit spans residues 13-58 of AS69. (B) Ribbon drawing of the αSyn β-hairpin (orange, β1 and β2 strands). H-bonds are depicted by dashed lines. The main interacting residues are shown as sticks (Mirecka et al., 2014). (PD) patients. In particular, high concentrations of iron, zinc and aluminum have been found in the substantia nigra, while copper accumulation has been detected in cerebrospinal fluid of PD patients. Furthermore, long-term metal exposure has been frequently related to parkinsonism (Bjørklund et al., 2019(Bjørklund et al., , 2020. Despite these numerous examples, there is still a controversial debate among experts whether metals are directly related to the cause of the disease. In general, we can state that the exact role of metals in the mechanism to neurodegeneration is still ambiguous. Indeed, it has been observed that metals catalyze the formation of reactive oxygen species causing oxidative stress but also enhance the aggregation of several proteins, among which αSyn, by complexing to them. At the same time, current studies have also shown how Mn and Ca levels can be regulated by αSyn itself (Dučić et al., 2015). But how can metals affect αSyn assembly? A possible explanation regarding the increased tendency to fibrillation is the subsequent conformational change after metal binding, resulting in abnormal folding and oligomer stabilization, as demonstrated for various metal ions (Kostka et al., 2008;Rcom-H'cheo-Gauthier et al., 2014;Uversky et al., 2001). Regarding metal-protein complexation, many recent studies have been focused on determining the structural complexity of this interaction reaching some important milestones. A low affinity binding site exists at the C-terminus of αSyn, where carboxylates of Asp and Glu residues are the major contributors for metal binding. In particular residues 119-124 are involved in electrostatic interactions and can bind all divalent metal cations, without specificity. Additionally, the affinity to this binding site can be drastically increased after phosphorylation of Tyr-125 and Ser-129 as demonstrated by ESI-MS and fluorescence spectroscopy in the case of Cu(II), Fe(II) and Pb(II) (Lu et al., 2011).
In the hierarchical order of divalent metal cations binding to αSyn, copper has been recognized to be the most affine and efficacious metal in promoting aggregation. Its binding has peculiar features in comparison to other ions and an exhaustive structural description of its coordination to αSyn has been comprehensively summed up by Binolfi et al. (2012). The authors took into consideration the intrinsically disordered monomer of αSyn and recognized three different binding sites for Cu(II). Apart from the common C-terminal binding site, as previously described, two independent sites in the N-terminal portion have been defined as high-affinity binding site 1 (residues 1-5) and low-affinity binding site 2 (associated to His-50). Binding constants vary depending on the experimental conditions, so a comparison between results from different publications is not always appropriate. Anyhow, the authors could conclude that, differently from the binding to the C-terminal region typical of all the other cations, the N-terminal coordination might occur under physiological conditions and be significantly relevant to the beginning of PD. After this review from 2012, many new experimental data have been published but still many questions remain open. In fact, current studies have partially undermined some of the previously described conclusions. In the following paragraphs, the influence of different metals on the αSyn structure will be analyzed and the most recent results in this field will be shown, taking into account contradictory point of views.
Copper
A very important issue pointed out by the research group of Lucas in 2019 is the fact that αSyn is mainly present in vivo as a N-terminally acetylated protein (Abeyawardhane et al., 2019). It is immediately clear how this post-translational modification could have an important consequence in the copper-protein interaction, perturbing the high affinity N-terminal binding site, since the Met1 site is now blocked. Through electron paramagnetic resonance (EPR) spectroscopy based on the Peisach-Blumberg correlation diagram and the DFT calculations previously reported by Ramis et al., two new correlation modes have been described (Ramis et al., 2017). A N3O1 binding involving His50, Val49 and a water molecule has been identified as the preferential N-terminal binding site and the principal binding site of the N-terminal-acetylated αSyn. On the contrary, a C-terminal binding site including residues Asp119, Asp121 and Glu123 have a great impact on fibrillation of the H50Q missense mutation, enhancing the protein aggregation propensity. In a recent comparison study carried out by Lorentzon et al., Cu(II) was found to accelerate non acetylated αSyn aggregation at biologically-relevant metal ion concentrations, while this reaction was not affected at all in the presence of the acetylated protein, of the A53T mutant and of the 1-97 truncated version. This is probably correlated with the intrinsic aggregation speed of the various αSyn variants: since the velocity with which the variants form the amyloid is higher than that of the wild type, the effect of metal binding is not detectable anymore (Lorentzon et al., 2020).
Cu(I) has also been investigated even if less information has been generated about it. αSyn, in fact, interacts with both oxidation states of copper ions that are involved in a copper catalyzed oxidation reaction, with the subsequent formation of FIGURE 11 | Binding sites recognized in the interaction between αSyn monomer and different exogenous factors (natural and synthetic small molecules, metals). The coordination between the protein and these factors leads to a balance that is always shifting between conformational stabilization and induction of aggregation. In the box regarding "metals," the specific coordination modes of Cu(II), all divalent cations, lanthanides with the specific amino acid residues are detailed. EGCG: Epigallocatechin gallate; PcTs: phthalocyanine tetra sulfonate.
Frontiers in Chemistry | www.frontiersin.org July 2021 | Volume 9 | Article 666585 reactive oxygen species (ROS) that leads to oxidative stress and to a possible formation of amyloid fibrils (Bisaglia and Bubacco, 2020). Also, in the case of Cu(I), three binding sites have been recognized by NMR at the N-and C-termini, respectively residues 1-5 (high affinity), His-50 and residues 116-127 (Binolfi et al., 2011;Camponeschi et al., 2013;Miotto et al., 2014;Okita et al., 2017). In particular Met1 and Met5 are the main coordinating center for this ion with a 2S2N/O coordination mode (De Ricco et al., 2015).
Iron
Also iron undergoes an oxidation cycle between two oxidative states Fe(II) and Fe(III) with production of ROS through the Fenton-Haber Weiss reaction (McDowall and Brown, 2016). Even in this case, as for copper, Lucas and coworkers investigated the influence of iron on the aggregation propensity and the secondary structure of the N-acetyl-αSyn (Abeyawardhane et al., 2018). Experiments performed in aerobic conditions showed that Fe(II) yielded a distinctive, highly toxic αSyn-metal complex in comparison to Fe(III). Fe(II), in fact, can react with O 2 and oxidize to Fe(III) with the production of H 2 O 2 and the subsequent development of a right-twisted antiparallel β-sheet conformation based on CD analyses and descriptive deconvolution of the secondary structure. These results display how the Fe(II) reactivity can have a very important impact in the protein conformation and its aggregated structural properties. Most importantly, the same does not occur with copper ions, proving a distinguished aggregation process.
Calcium
Calcium dysregulation has been connected with neurodegenerative disorders and high levels of this metal have been detected in Lewy bodies. For its central role in αSyn aggregation, Kim and his research group took Ca 2+ as representative metal ion to understand metal influence on the formation of large interfibrillar aggregates (Han et al., 2018). The authors could demonstrate that Ca 2+ mediates the rapid formation of αSyn fibrils via the structural transition of monomeric αSyn into an extended conformation, which is prone to aggregation. It is interesting to discover how the structure of the α-syn monomer develops after binding to Ca 2+ . By using ion mobility-mass spectrometry (IM-MS) and synchrotron small-angle X-ray scattering (SAXS), Han et al. could demonstrate a structural transition of monomeric α-syn into an extended conformation with the exposure of the NAC region, which is more prone to aggregation.
Lanthanide (Trivalent) Metal Ions
Investigation on lanthanides is a very crucial topic since they are increasingly applied in various fields of industry and agriculture. As divalent metal ions, trivalent metal ions non-specifically bind to the C-terminus of αSyn but also transiently interact with carboxylates in the N-terminal and NAC regions as interpreted from 1H to 15N HSQC NMR spectroscopy. In addition, they accelerate fibrillation much faster than divalent cations (Bai et al., 2015).
All the in vitro experiments carried out so far do not necessarily translate in vivo metal binding. Lothian et al. pointed out that there is a lack of evidence that the metal binding observed in vitro also occurs in vivo (Lothian et al., 2019). This work does not exclude the possibility that a very small percent (1%) of the whole protein can effectively bind to metals, promoting their aggregation with the consequent formation of oligomers and fibrils but, in general, according to the authors, αSyn cannot be considered as a metalloprotein in vivo. However, also these last results have some limitations because they considered non-pathogenic tissues, while in PD many factors can be combined and lead to the ultimately conclusion, like e. g post translational modification, molecular binding, ionic strength, salt concentration.
Natural Small Molecules
Natural products are gaining importance in drug discovery since they are an environmental-friendly source for hit compounds. Moreover, with modern extraction and purification techniques, researchers are able to obtain these small molecules with moderate efforts. Also, they offer low-cost production and possible improvement of their activity. However, natural compounds have some limitations such as low reproducibility and yield, but also lack of safety and tolerability. Finally, their multi-target activity can be a problem when the aim is to be selective toward a single target.
Since these molecules present low selectivity, they can bind to both αSyn aggregates and monomers. However, when they bind to monomers, few of them have a characteristic binding site, hence more studies are needed to look further into this topic. Here, we present an overview over the main classes of natural small molecules able to influence monomeric αSyn aggregation. Their structures are depicted in Figure 12.
Epigallocatechin Gallate (EGCG)
EGCG is a natural compound known for its antioxidant properties and anti-aggregation activity against amyloid proteins. This latter effect against multiple targets (αSyn, Aβ 1-42 , Tau, hIAPP) is due to its lack of selectivity. Concerning αSyn, NMR studies suggest that EGCG binds to the N-terminal domain, in particular to residues 23-55 (Xu et al., 2020). This binding is mainly governed by Van der Waals and π-stacking interactions due to the structure of EGCG, characterized by electron-rich aromatic rings bearing three consecutive OH substituents. Even if the mechanism behind the anti-aggregation effect is ambiguous, EGCG can bind fibrils and convert them into smaller, non-toxic aggregates. Moreover, EGCG is also able to bind to monomers and induce their aggregations into non-cytotoxic, off-pathway entities, thus avoiding the nucleation process (Caruana et al., 2011). EGCG efficacy has been tested both in vitro and in vivo (Pujols et al., 2018). Also, the compound is currently under clinical trials for Multiple System Atrophy (Xu et al., 2020). Recent studies have reported that the species responsible for EGCG anti-aggregating properties is its oxidized form Frontiers in Chemistry | www.frontiersin.org (oxEGCG). In fact, most of the EGCG efficacy studies were performed at pH 7, at which the compound is not stable and comes across oxidation. When EGCG is tested at pH six or less, the molecule is stable and its anti-aggregation properties are lost (Sternke-Hoffmann et al., 2020).
Baicalein
Baicalein is a flavonoid extracted from Scutellaria baicalensis. This molecule is known to disassemble αSyn fibrils into smaller, non-toxic oligomers by binding to them once they are mature (Kurnik et al., 2018). As EGCG, baicalein is also active toward αSyn monomers: the polyphenol can interact and convert them into off-pathway aggregates with very low cellular toxicity (Morshedi et al., 2015).
Recently, Javed et al. reviewed the interaction between baicalein and αSyn. Here, the oxidized form of baicalein (quinone) is crucial for αSyn aggregation inhibition. In fact, its effectiveness against αSyn aggregation has been tested both in cells (HeLa and SH-SY5Y) and in vivo models (Oliveri, 2019). When baicalein quinones interact with early-stage aggregates, it leads to quite soluble αSyn oligomers. In this case, the polyphenol covalently binds to the protein and creates a Schiff base with lysine side chains, expressed in the N-terminal domain of αSyn. Tyrosine residues are also involved in this binding (Javed and Ojha, 2020).
As mentioned before, Baicalein binds to a broad region of αSyn, thus it is not selective toward a specific binding site. Indeed, other studies showed that baicalein is an efficient aggregation inhibitor also for Aβ, Tau, IAPP and other amyloid proteins, which is a common characteristic for polyphenols (Oliveri, 2019).
Nordihydroguaiaretic Acid (NDGA)
Nordihydroguaiaretic acid is a natural compound deriving from Larrea tridentata. Concerning its interaction with αSyn, confocal single-molecule fluorescence spectroscopy studies showed that the compound binds to toxic inclusions and that it is able to inhibit the formation and/or to disaggregate mature oligomers (Caruana et al., 2011). Moreover, recent studies pointed out that oxidation and consecutive cyclization of NDGA is required for its activity. In fact, its oxidized form (oxNDGA) can interact with monomers and convert them into quinone-modified species, which are less prone to aggregate in comparison to the nonmodified ones. However, these monomers can still carry out their physiological function: probably this modification does not alter Frontiers in Chemistry | www.frontiersin.org July 2021 | Volume 9 | Article 666585 their normal activity. In fact, as CD data suggest, monomers preserve the ability to fold into α-helix conformations while interacting with SDS in vitro. Also, in the same work, oxNDGA was shown to inhibit in vivo oligomers and fibrils formation (C. elegans) (Daniels et al., 2019). Furthermore, the compound is also able to interact with preformed fibrils. In fact, ThT assays displayed that cyclized NDGA can reduce the contents of β-sheets in a dose-dependent manner (Perni et al., 2017). However, it is important to inquire if these interactions have an effect over αSyn structure. In particular, ESI-MS analysis showed no covalent binding, which means that the protein primary structure is unmodified. Moreover, this interaction leads to a more compact conformation of the protein, which may mask the NAC region and discourage aggregation. Nevertheless, this hypothesis still needs additional analyses to be confirmed.
Regarding the sequences of αSyn involved in this interaction, NMR studies suggested that the most engaged domain is the N-terminus (Val3, Phe4, Met5 and His50 in particular). Especially, this domain is involved in the monomer's helix folding while interacting with membranes: as we saw before, the dynamic flexibility of the protein is not altered by the interaction with oxNDGA.
All in all, these results show that NDGA and in particular oxNDGA can modify the conformation of the protein aggregates. However, the structural properties of these entities as well as the involved domains of αSyn have yet to be elucidated.
Triterpenoids Squalamine
Squalamine is a steroid-polyamine conjugate found in sharks and is known for its anticancer and antiviral activity. Its main characteristic is the polyamine chain attached to the cyclopentanoperhydrophenanthrene structure, which is positively charged in the cell's physiological environment. This moiety allows the compound to interact with membranes, in particular with the phospholipids negatively charged heads (Perni et al., 2017). Finally, the salt bridges formed between squalamine and the lipidic bilayer may avoid the pathological formation of αSyn oligomers, since there is a competition for the same sites of interactions between misfolded αSyn monomers and the triterpenoid. These hypotheses were confirmed by CD experiments. In fact, when αSyn is incubated with squalamine in the presence of phospholipidic membranes, a decrease of the α-helix character is noticed. Thus, one can speculate that the displacement of helical-folded αSyn from membranes leads to the refolding of the protein in a random-coil configuration. This last structure is the one in which monomers are usually found and it represents the most populated state of soluble cytosolic αSyn. However, further studies are required to confirm this idea.
Concerning αSyn-squalamine interactions, NMR analyses showed that the C-terminal domain is the most involved region of the protein (Perni et al., 2017). This is consistent with the fact that the positively charged chain of squalamine can create electrostatic interactions with the negatively charged domain of αSyn. In particular, the sequence engaged spans residues 113-139. However, this interaction is attenuated when squalamine and αSyn are incubated in the presence of membranes. Furthermore, this contact does not seem to directly alter the conformation of the protein. In fact, squalamine prefers to interact with phospholipids instead of αSyn in the cells. Nevertheless, more studies are required to confirm this speculation to be sure that αSyn refolding merely refers to its displacement by membranes.
Finally, the prevention of toxic oligomers formation provided by squalamine was proven in vitro, and the antiaggregant properties were later tested and confirmed in vivo using a Caernohabditis elegans PD model (Perni et al., 2017).
Nicotine and Caffeine
Nicotine and caffeine can interact with αSyn and inhibit its aggregation pathway. However, since they can bind to αSyn simultaneously, their respective binding sites may be different. Also, the mechanism of the inhibition of aggregation is still unclear.
Concerning nicotine, in vitro studies demonstrated that this alkaloid can induce a conformational change in αSyn monomers, leading to nucleation slowdown and the formation of soluble, less toxic oligomers (Kardani et al., 2017).
At the same time, caffeine can decrease αSyn aggregates toxicity, while accelerating the apparent fibrillation rate (Oliveri, 2019). Interestingly, CD and TEM data suggest that caffeine does not alter the conformation of αSyn monomers (Kardani and Roy, 2015). Also, an increased transformation of oligomers into mature aggregates by administration of caffeine in yeast cell is remarked in literature. These data suggest a role of this alkaloid in the field of synucleinopathologies (Kardani and Roy, 2015).
Finally, nicotine is proven to be active against Aβ aggregates, while caffeine displays an activity also toward hIAPP and Tau toxic species both in vitro and in vivo (Ma et al., 2020).
Sugar Alcohols Mannitol
Mannitol is a polyol approved by the Food and Drug Administration as an osmotic agent. Also, it is used as an osmotic diuretic in the therapy of hypertension and as a weak laxative in case of constipation. Furthermore, mannitol is known for its BBB disruption activity and its hyperosmotic solution is widely used in clinics. CD studies show that this compound is able to inhibit the aggregation of αSyn monomers into fibrils, likely through interaction with oligomers by leading them to an alternative pattern of aggregation acting as a "chemical chaperone". Finally, its neuroprotection activity was tested and proven in vivo in Drosophila flies and in mice (Shaltiel-Karyo et al., 2013).
Focusing on αSyn-mannitol interaction, CD experiments show that the polyalcohol does not affect the conformation of β-sheet rich fibrils, so no interaction is detected with the mature aggregates of the protein. However, the compound is able to change the secondary structure of αSyn oligomers. In fact, CD analysis spots a refolding in the early-stage aggregates outline.
Frontiers in Chemistry | www.frontiersin.org July 2021 | Volume 9 | Article 666585 20 However, more studies are required to understand the structural properties of the entities derived by oligomers refolding. Also, the protein domains involved in this interaction have yet to be discovered (Shaltiel-Karyo et al., 2013).
Scyllo-Inositol
Scyllo-inositol is one of the inositol stereoisomers, rare in nature, having attracted the attention of the scientific community in the field of Aβ 1-42 peptide inhibition. In fact, scyllo-inositol was shown to stabilize a non-toxic form of Aβ 1-42 peptide and to ameliorate cognitive deficit together with lowering amyloid plaques in vivo (AD mouse model) (Ibrahim and McLaurin, 2016). Interestingly, this molecule was proven to have an effect also toward αSyn. In fact, TEM experiments suggest that it can reduce both human and mouse αSyn aggregation. An explanation for its activity may lay in its planar structure, which is expected to interact with αSyn monomers through hydrophobic and hydrophilic interactions, possibly entrapping the NAC domain of the protein. Since this condition is crucial in fibrillation, its inaccessibility can discourage protein-protein interactions and, finally, aggregation. Considering αSyn conformation upon the interaction with scyllo-inositol, soluble monomers seem to be stabilized in vitro. This may allow these species to conserve their random coil structure and prevent them from the nucleation phase. However, further analyses are needed to determine the binding sites involved and the monomers behavior (Ibrahim and McLaurin, 2016).
Others
Tanshinone I (TAN I) and Tanshinone IIA (TAN IIA) TAN I and TAN IIA are the main phenanthrenequinone compounds found in Salvia miltiorrhiza, a plant widely studied in Chinese traditional medicine. Concerning the interactions between these compounds and αSyn, ThT and TEM studies suggest that they both prolong the lag time of αSyn aggregation and disaggregate mature fibrils (Ji et al., 2016). This effect is related to their role in decreasing toxic oligomers formation, which contributes to their multi-target activity. Moreover, they seem to interact with αSyn monomers and oligomers through hydrophobic interactions, blocking them from aggregation in the same way they do with Aβ 1-42 peptide (Ren et al., 2017).
Regarding αSyn conformational aspects, CD data show that TANI and TANIIA keep the protein in a random coil structure, while in their absence monomers tend to misfold and aggregate in β-sheet structures (Ren et al., 2017). Although evidence suggests that the two compounds can avoid αSyn nucleation and fibrillation, the binding sites as well as a detailed description of the complex formed should still be investigated.
Cuminaldehyde
Cuminaldehyde is an aldehydic compound present in Cuminum cyminum essential oil. It is thought that the molecule interacts with αSyn monomers, thus preventing them from nucleation. Furthermore, cuminaldehyde showed a lower activity in fibrillar disaggregation than baicalein. These data suggest that Cuminaldehyde is more selective toward monomers rather than aggregated species (Morshedi et al., 2015).
Interestingly, far-UV CD gives an interesting insight over the structural behavior of αSyn while interacting with cuminaldehyde. When cuminaldehyde is incubated with αSyn, the strong negative peak at 200 nm disappears, in favor of one at 208 nm. This result highlights the conversion of random-coil monomers into entities whose structure is still unknow. However, no β-sheet peaks are detected and the shape of the graph refers to a characteristic α-helical conformation. All in all, one can speculate that the new complexes may adopt a helical structure, which is not prone to fold into β-sheet nor to convert into unfolded coils (Morshedi et al., 2015).
Interestingly, NMR studies suggest that the aldehydic function of cuminaldehyde may interact with lysine amino groups in the N-terminal domain of αSyn monomers (Morshedi et al., 2015). Thus, this interaction can be one of the main cause which leads to the conformational transition occurred in αSyn interacting with the compound. Finally, it may provide details about the binding site of the protein.
Synthetic Small Molecules
Synthetic small molecules have been widely investigated as putative inhibitors of αSyn aggregation. Here, some of the principal compounds currently being studied are described, highlighting the interaction with the protein (Figure 13). Other important molecules that play a role as inhibitors are omitted, since they mainly interact with αSyn aggregates rather than monomers. An example is the pyrazole Anle 138, widely described in Fields and Shvadchak works (Shvadchak et al., 2018;Fields et al., 2019).
Phthalocyanines
Phthalocyanines are tetrapyrrole macrocycles largely investigated in the field of αSyn aggregation. Their structure is correlated to the mechanism by which they bind to the protein: both the electron-dense pyrrolic core and the substituent carried by the peripherical rings play a pivotal role. In fact, NMR studies show that phthalocyanine tetra sulfonate (PcTs, a synthetic derivative of this group) interacts with αSyn at the N-terminal region residues Phe4 and Tyr39 mainly by π-π stacking interactions and salt bridges. This leads to a stabilization of the α-helical folding of monomers, thus delaying their misfolding and aggregation. However, data about the binding mode of PcTs to αSyn are controversial. In contrast with the previous studies, high resolution 1 H-15 N HSQC-NMR data of Lamberto et al. demonstrated that the compound binds to the C-terminal domain of αSyn monomers. Thus, these studies suggest the existence of another important binding site, involving residues 93-95 of the protein (Lamberto et al., 2009). Considering this contradiction, further studies are needed to understand which site plays a role in the interaction between αSyn and phthalocyanines.
Interestingly, PcTs can form a complex with Cu 2+ , an important ion for αSyn accumulation in tissues. Also, this compound is able to inhibit fibrillation by forming offpathway non-toxic oligomers in vitro (Oliveri, 2019). Finally, the compound is active not only against αSyn, but also amyloid β, Tau and PrP protein aggregates in vitro (Valiente-Gabioud et al., 2016).
4-Hydroxynaphthalen-1-yl)sulphonamide Derivatives
These compounds are novel inhibitors revealed by High-Throughput Screening (HTS). Among them, one of the most active compounds is C41. In vitro studies show that this molecule binds to αSyn monomers, on-pathway oligomers, and fibrillary precursors. In particular, the interaction with soluble monomers was confirmed by Size Exclusion Chromatography-Multi Angle Light Scattering (SEC-MALS). Furthermore, C41 also binds to off-pathway small aggregates and this can prevent both vesicle interaction and nucleation. 1 H-15 N HSQC analysis demonstrated that C41 mainly interacts with αSyn N-terminal domain through hydrophobic forces. As we saw before, this domain is involved in αSyn interaction with membrane, which is crucial for αSyn physiological role. MS data show that covalent adducts can be formed, but more studies are needed to identify them and understand the conformational changes of αSyn (Kurnik et al., 2018).
M2N and M3N
These compounds are αSyn aggregation inhibitors that consist of mannitol, covalently linked to NQTrp via two or three molecules of PEG. NQTrp, a generic amyloids inhibitor formed by NQ (Naphtoquinone) and Trp (Tryptophan), is effective against fibrils formation due to the possibility to share π-π interactions with αSyn monomers. However, even if the inhibition occurs at low concentration (0.1 µm), the compound is characterized by a poor BBB penetration (Paul et al., 2019).
To overcome this problem, researchers conjugated it with mannitol, known for its BBB disruption properties and anti-αSyn aggregation effect. Notably, the conjugates were nontoxic to SH-SY5Y cells and could reduce the cytotoxicity of αSyn aggregates. Moreover, results suggest that the longer PEG chain in M3N might confer better flexibility for a more efficient inhibition.
Concerning αSyn conformational aspects, CD studies were performed to elucidate the protein behavior during the interaction with the compounds. In this analysis, αSyn alone shows a negative peak at 218 nm and a positive one around 198 nm, which are typical of β-sheet rich structures. By adding incremental doses of the M2N and M3N, the peak at 218 decreases in a dose-dependent manner. This means that the secondary structure of the protein refolds during the interaction with the molecules, with a decrement of β-sheets. However, more studies are needed to identify the conformation of the new formed complexes (Paul et al., 2019).
Finally, TEM studies show that in the presence of M2N and M3N, the fibrillar outline of αSyn inclusions turns into an amorphic conformation (Paul et al., 2019). This is in accordance with the previous results and suggests that these inhibitors can be interesting to investigate αSyn behavior while interacting with polyalcohol compounds.
CONCLUSION AND FUTURE PERSPECTIVE
After this trip around αSyn structure, the factors influencing it and the applied techniques, it is clear that the fundamental structural features of this rather small protein of 140 residues have not yet been elucidated. The main hurdle to thoroughly understand its behavior is its intrinsically disordered nature and high susceptibility to the environment. αSyn tends to acquire diverse transient and dynamic conformations depending on the presence of different biological and physico-chemical factors. In physiological conditions, αSyn is thought to be a compact monomer acquiring an aggregation-resistant globular structure. This conformation is stabilized by long-range electrostatic interactions between the residues present in the C-terminal domain and those located in the central part of the protein. On the other hand, when αSyn is driven to adopt a more extended structure, exposing the NAC region, protein aggregation is triggered. It is assumed that folded stable helical conformers impede amyloidogenic aggregation. However, to date, there is not a clear consensus on its in vivo structural propensity. αSyn will likely adopt specialized conformations depending on different conditions (e.g., changes in pH, temperature, ionic strength, closeness to surfaces) that might trigger different biological or pathological functions. For instance, a helical pattern at the N-terminal end has been observed upon vesicle and membrane interaction. Conversely, there are limited comprehensive structural data about αSyn interactions with various partners such as proteins and both endogenous (e.g., neurotransmitters, and lipids) and exogenous (e.g., metals and drugs) molecules. This information is limited due to the difficulty to create, isolate and analyze complexes of αSyn with these partners. This lack of comprehensive knowledge is also due to the absence of crystallographic data and of other experimental techniques able to reproduce the in vivo physiological and/or pathological conditions. In addition, experimental data in the literature are obtained from the study of αSyn in very different conditions, hampering a significant comparison of the obtained results. In our opinion, it would be appropriate for studies to converge upon standardized set up and protocols. As an example, N-terminal acetylation has been demonstrated to be a constitutive element of the protein, so this modification should be applied in every experiment to achieve reliable conclusions. Other aspects to be considered are the source of αSyn, its post-translational modifications and pathological mutations. Finally, a combination of experimental and computational approaches can be a good strategy for future research. By extracting information from different experimental techniques and constraining molecular dynamic simulations based on that information, more meaningful results could be obtained, allowing us to address some of the experimental issues observed so far. The more reliable information obtained, the more effective their translation into the development of bioactive compounds able to modulate pathological αSyn effects will be.
AUTHOR CONTRIBUTIONS
NB, LF, HP-P, and KP equallly contributed to this work. HP-P and KP: wt and mutated synuclein structural features; LF: Lipids and Metals; NB: Proteins and small molecules. SPi and SO: analysis, writing modifications, and revision through all the text. SPe: writing, analysis and interpretation throughout all the text, and editing of the manuscript. | 2021-07-07T13:12:33.161Z | 2021-07-07T00:00:00.000 | {
"year": 2021,
"sha1": "b560ef68add1a2213e3ae9dfd19944dd3621582d",
"oa_license": "CCBY",
"oa_url": "https://www.frontiersin.org/articles/10.3389/fchem.2021.666585/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "b560ef68add1a2213e3ae9dfd19944dd3621582d",
"s2fieldsofstudy": [
"Biology",
"Chemistry"
],
"extfieldsofstudy": [
"Medicine"
]
} |
220306776 | pes2o/s2orc | v3-fos-license | Molecular analysis of Spiophanes bombyx complex (Annelida: Spionidae) with description of a new species
Spiophanes bombyx (Claparède, 1870) from the Gulf of Naples, Tyrrhenian Sea, Italy, was the first described Spiophanes with fronto-lateral horns on the prostomium. It was also considered the only horned species occurring in European waters. Our sequence data of five gene fragments suggest the presence of two horned sibling Spiophanes species in northern Europe: S. cf. bombyx in the North and the Norwegian seas, and S. cf. convexus in Brittany, northern France, and Bay of Biscay, northern Spain. Spiophanes cf. bombyx worms are genetically close to a single examined specimen of S. bombyx from Venice Lagoon, Italy but their conspecificity should be verified by further study. Our sequence data show that horned Spiophanes from the North Pacific are genetically distant from horned European species, and that S. uschakowi Zachs, 1933, originally described from the Sea of Japan (East Sea) is a valid species. The data also suggest the presence of two horned sibling Spiophanes species in the North East Pacific: S. hakaiensis Radashevsky & Pankova, n. sp. distributed from Alaska south to about Point Conception, and S. norrisi Meißner & Blank, 2009, distributed from San Francisco Bay south to Baja California Sur, Mexico. Spiophanes from South America, morphologically similar to S. norrisi, are suggested to belong to a new species. Molecular data also suggest the presence of two sibling species among the worms from northern Europe identified by morphology as S. kroyeri Grube, 1860. Worms from the Barents Sea and northern part of the North Sea are tentatively referred to as S. cf. kroyeri; worms from the northern and central parts of the North Sea and from the Bay of Biscay, northern Spain, are tentatively referred to as S. cf. cirrata M. Sars in G.O. Sars, 1872. Sequence data also show that S. duplex from California is genetically different from morphologically similar worms from South America. The South American worms are referred to resurrected S. soederstroemi Hartman, 1953 which was originally described from off Rio Grande do Sul, Brazil, and then considered as a junior synonym of S. duplex. Analysis of divergence times of Spiophanes lineages suggested that the origin of the most recent common ancestor of horned Spiophanes with metameric nuchal organs was around 11.1 mya (95% HPD: 5.1–19.0 mya) and that the divergence of the North Atlantic and North Pacific lineages was around 7.9 mya (95% HPD: 4.1–13.3 mya). The North Atlantic lineage was estimated to have diverged 4.8 mya (95% HPD: 2.2–8.6 mya), resulting in the origin of S. cf. bombyx and S. cf. convexus. The North Pacific lineage was estimated to have diverged first by the isolation and speciation of S. norrisi 1.7 mya (95% HPD: 2.3–1.0 mya), and then by the isolation and speciation of S. uschakowi and S. hakaiensis n. sp. 1.3 mya (95% HPD: 2.0–0.7 mya). The estimates place the divergences soon after maximum glacial period in the North Pacific (2.4–3.0 mya).
Introduction For a long time, polychaetes were believed to have high morphological variability and ecological plasticity and many species were considered as widely distributed or cosmopolitan. New approaches and techniques show, however, that many of these "cosmopolitans" comprise groups of similar cryptic or sibling species with much more limited variability, plasticity and geographic distribution than was previously suggested (see Read's comments on Polychaeta in Appeltans et al. [1]; [2][3][4]).
One of the old Spiophanes species, S. bombyx was originally described from the Gulf of Naples, Italy, by Claparède [9,10] (as Spio bombyx). The main diagnostic feature of the species was a pair of long horns extending from the fronto-lateral parts of the prostomium ( [9] : Fig 2; [10]: Fig 2). Since then, Spiophanes with similar horns were reported as S. bombyx from northern Europe and Iceland [11-34, 8, 35], Atlantic coast of North and Central America [36][37][38][39][40][41][42][43][44][45][46], Atlantic coast of South America [47][48][49][50][51][52], Pacific coast of North America [53][54][55][56][57][58][59][60][61][62][63][64][65][66], Pacific coast of South America [67,64,[68][69][70][71][72][73], Hawaii [74,75], Pacific coast of Asia [76][77][78][79][80][81][82][83][84][85][86][87][88][89][90][91][92][93][94][95], Indian Ocean [96,97], Atlantic coast of Africa [98][99][100][101][102][103], Australia and New Zealand [104][105][106][107], and Antarctica [108,109]. Recent studies, however, suggested that these reports may include a series of Table 1. Complete available information about type specimens is given in the Results preceding the descriptions or comments on the species; information about all newly collected material and museum samples examined during this study is given in S1 Table. A complete list of the museums and other collections (and their acronyms) holding the examined or reported samples is given in S2 Table. S1 Table also includes material from the Norwegian, North, Mediterranean and Aegean seas and North Pacific reported by Meißner & Blank [111] and from the North Sea identified by Dieter Fiege, Markus Böggemann and Karin Sequences obtained in the present study are given in plain; sequences provided by other authors are given in bold. a Meißner which is deposited in the SMF and other polychaete collections and was only partially re-examined by the first author (VIR). It also includes records of S. bombyx (= S. uschakowi) from Japan provided by Imajima [88,[116][117][118], for which no museum numbers were reported, records of S. bombyx (= S. cf. norrisi) from Chile, Argentina and the Falkland Islands provided by Carrasco [119] and Blake [48], and records of S. bombyx (here referred either to S. bombyx or S. cf. convexus) from Spain provided by Meißner [8]. To link some sequences used in the molecular analysis in the present study with the corresponding data, unique numbers from the first author's database (VIR) are given to samples in the S1 Table (at the end of each record). These numbers without letters precede specific names on the phylogenetic trees (Figs 1-3). When no coordinates were provided for sampling sites in old studies, they were collected from the Google Earth map according to the original descriptions of the locations. Sampling locations of Spiophanes spp. noted in S1 Table are plotted on maps using the QGIS 3.8.0 Multiple colors indicate haplotypes shared by more than one sampling locality, with sections scaled by frequency. (B) Majority rule consensus tree of the Bayesian inference analysis-1 of COI (267 bp) sequences obtained in the present study and also taken from GenBank and BOLD. Posterior probabilities are shown on the branches. Species names are followed by the names of the collecting locations. Information about numbers with letters following specific names are given in Table 1. The numbers without letters following specific names are unique numbers from the VIR database linking the individuals on the tree with the sampling data in the S1 Table. https://doi.org/10.1371/journal.pone.0234238.g001
DNA extraction, amplification and sequencing
We used the DNA Wizard Genomic DNA Purification Kit and the ReliaPrep gDNA Tissue Miniprep System (Promega Corporation, Madison, WI, USA) for DNA extraction and purification, with the standard protocol for animal tissue. Polymerase chain reaction (PCR) amplification of nuclear 18S rDNA, D1 region of 28S rDNA and Histone 3, and mitochondrial 16S rDNA gene fragments was accomplished with the primers and conditions described by Radashevsky et al. [124,125]. We used primers 5' GGTCAACAAATCATAAAGATATTGG 3' and 5' TAAACTTCAGGGTGACCAAAAAATCA 3' to amplify the mitochondrial gene fragment of cytochrome C oxidase subunit 1 (COI) [126]. Cycling parameters were as follows: initial denaturation at 94˚C for 2 min, 35 cycles of 94˚C for 30 s, 50˚C for 40 s, 72˚C for 60 s, with a final extension of 72˚C for 5 min.
Purified PCR products were bidirectionally sequenced on an ABI Prism 3100 Genetic Analyzer (Applied Biosystems) using the BigDye Terminator v 3.1 Cycle Sequencing Kit (Applied Biosystems) and the same primers as for PCR. The consensus sequence of each gene region of each specimen was assembled from the two complementary sequences using SeqScape v 2.5 (Applied Biosystems). GenBank accession numbers of the obtained sequences are given in Table 1. Posterior probabilities are shown on the branches below the divergence time estimations. Bars at nodes indicate highest posterior density (95% HPD) intervals for the divergence times. Species names are followed by the names of the collecting locations. Information about numbers with letters following specific names are given in Table 1. The numbers without letters following specific names are unique numbers from the VIR database linking the individuals on the tree with the sampling data in the S1
Phylogenetic analysis
We aligned DNA sequences using the MAFFT v7.2 software with the default settings (automatically chosen algorithms) [127,128]. Ambiguous positions and gaps were excluded from subsequent analysis using GBlocks [129]. The nucleotide datasets were combined using the supermatrix approach. The number of variable and parsimony informative sites in the datasets, the uncorrected values of sequence divergence (pairwise distances, p) both within and between groups were calculated in MEGA 5.1.
We used MrBayes 3.2.6 via the CIPRES web portal [130] for the Bayesian analysis of 10,000,000 generations, four parallel chains and sample frequencies set to 500, in two separate runs. Based on the convergence of likelihood scores, 25% sampled trees were discarded as burn-in. The analysis of the combined data set was partitioned and the models of substitution were determined using Akaike Information Criterion (AIC) in Modeltest 3.7 [131]: GTR+I+G for 16S, 28S and Histone 3, TVM+G for 18S, and HKY+I+G for COI.
We also included in the analysis sequences of Spiophanes spp. obtained by Meißner & Blank [111], Mincks et al. [122], Meißner & Götting [123], Aylagas et al. [121], and some sequences from GenBank and BOLD (Table 1). We performed two analyses: the COI analysis to compare our data with those reported by previous authors, and a concatenated analysis using five genes to get better resolution of relationships among Spiophanes species. The fivegenes analysis was based on sequence data of the mitochondrial COI, 16S rDNA, nuclear 18S, 28S rDNA, and Histone 3 most of which were obtained in the present study. In this analysis, we included COI sequences obtained by other authors if they comprised more than 500 bp, and 18S sequences if they comprised more than 1600 bp. Of 46 individual concatenated sequences obtained in the present study, 30 sequences comprised fragments of four to five genes, and 39 sequences comprised fragments of three genes. The analyses were rooted using the sequences of Trochochaeta multisetosa (Ö rsted, 1843). In the Bayesian analysis of 18S sequences of various spionids by Mincks et al. ([122] : Fig 8), Trochochaeta Levinsen, 1883 appeared sister to Spiophanes.
The haplotype network was constructed based on a statistical parsimony approach using TCS [132].
Divergence time estimation
Divergence time estimation of the horned Spiophanes lineages was performed with BEAST 1.8.0 [133] running in the CIPRES Science Gateway [130]. For the analysis, we used COI sequences comprising more than 500 bp. The in-group included horned Spiophanes with metameric nuchal organs, all members of a monophyletic group according to the five-genes analysis. The analysis was rooted using the sequences of S. soederstroemi that in five-genes analysis appeared most closely related to the clade of the horned Spiophanes.
The Tamura-Nei model (TrN) of nucleotide substitution with gamma-distributed rate variation (+G) was applied to COI data set with base frequencies estimated during the analysis [134]. We used the path sampling maximum likelihood estimator implemented in BEAST 1.8.0 [135,136] to determine the appropriate molecular clock models for data set (uncorrelated lognormal relaxed clock, random local clock or strict clock). After model selection, a relaxed molecular clock using the uncorrelated log-normal model [137] was applied with a Birth-Death process speciation prior for branching rates [138]. We applied substitution rate of 4.1%/ MY with a SD of 3.5-4.7%/MY, suggested for northern polychaetes by Loeza-Quintana et al. [139], to convert genetic branch lengths to absolute times. The final analysis consisted of two independent MCMC analyses; each chain was run for 40,000,000 generations with parameters sampled every 4000 steps to ensure effective sample size (ESS) values above 200 for all parameters. The first 10% of trees in each independent run were discarded as burn-in and the remaining trees were combined to produce an ultra-metric consensus tree using LogCombiner and TreeAnnotator 1.8.0 (BEAST package).
Nomenclatural acts
The electronic edition of this article conforms to the requirements of the amended International Code of Zoological Nomenclature, and hence the new names contained herein are available under that Code from the electronic edition of this article. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix "http://zoobank.org/". The LSID for this publication is: urn:lsid:zoobank.org:pub:B69BEC0A-2B46-4F46-99E4-73FD7D502A38. The electronic edition of this work was published in a journal with an ISSN, and has been archived and is available from the following digital repositories: PubMed Central, LOCKSS, ResearchGate.
Molecular analysis
COI analysis and haplotype network. The alignment of COI sequences obtained in the present study and those from GenBank and BOLD comprised 267 bp with 106 (39.7%) variable sites, of which 93 (34.8%) sites were parsimony informative. The Bayesian analysis resulted in a partially resolved consensus tree (Fig 1B).
The analysis revealed two sister groups among specimens from northern Europe identified by morphology as S. bombyx: one from the North and the Norwegian seas (referred to as S. cf. bombyx), and another from Brittany, northern France, and the Bay of Biscay, northern Spain (referred to as S. cf. convexus). The p-distances among individuals of these groups ranged from 12.73% to 14.98%. The ingroup p-distance values ranged from 0.00% to 1.5% and from 0.75% to 1.87%, respectively.
Horned Spiophanes from the North Pacific formed a well supported clade (PP = 1) closely related to the clade of horned Spiophanes (S. cf. bombyx-S. cf. convexus) from northern Europe. The relationships within the North Pacific clade were partially resolved, with one clade well supported (PP = 0.97), comprising one group of specimens from California, USA, and Baja California Norte, Mexico (referred to as S. norrisi according to the type locality of the species in Baja California Sur, Mexico), and another group S. uschakowi from the Sea of Japan (East Sea). The p-distances between individuals of these groups ranged from 6.74% to 8.99%. The rest of the specimens within the North Pacific clade did not form a monophyletic group, but, according to the results of the five-genes analysis (see below), they are referred to S. hakaiensis n. sp. The three groups of North Pacific Spiophanes did not share any haplotype ( Fig 1A).
Two sister groups were revealed among specimens from northern Europe identified by morphology as S. kroyeri: one from the Barents Sea and northern part of the North Sea (tentatively referred to as S. cf. kroyeri), and another from the northern and central parts of the North Sea and from Bay of Biscay, northern Spain (tentatively referred to as S. cf. cirrata). The p-distances between individuals of these groups ranged from 6.37% to 8.61%. The ingroup pdistance values ranged from 0.00% to 1.12% and from 0.75% to 1.5%, respectively.
The analysis revealed two sister groups among specimens from northern Europe identified by morphology as S. bombyx. Specimens from the North and the Norwegian seas (referred to as S. cf. bombyx) were genetically close to single examined specimen of S. bombyx from Venice Lagoon, Adriatic Sea, Italy (p-distances ranging from 00% for 28S to 2.46% for 16S). Specimens from Brittany, northern France, and the Bay of Biscay, northern Spain, are referred to as S. cf. convexus. The p-distances between individuals of S. cf. bombyx and S. cf. convexus ranged from minimal 0.3% for 18S rDNA to maximal 14.07% for COI. The maximum ingroup p-distance values were for COI sequences: 0.94% and 1.69%, respectively.
Horned Spiophanes from the North Pacific formed three distinct groups within a clade sister to the European horned Spiophanes. The groups comprised S. uschakowi from the Sea of Japan (East Sea), worms from Alaska to San Francisco Bay, California, described below as S. hakaiensis n. sp., and worms from San Francisco Bay, California, and Baja California Norte, Mexico, referred to S. norrisi. Spiophanes uschakowi and S. hakaiensis n. sp. formed a clade sister to S. norrisi. The p-distances between individuals of S. uschakowi and S. hakaiensis n. sp. ranged from minimal 0.91% for 18S rDNA to maximal 7.68% for COI. The maximum ingroup p-distance values were for COI sequences: 0.94% and 1.12%, respectively.
Three groups were distinguished among specimens identified by morphology as S. kroyeri: one from the Barents Sea and northern part of the North Sea (referred to as S. cf. kroyeri), second from the northern and central parts of the North Sea and from the Bay of Biscay, northern Spain (referred to as S. cf. cirrata), and a third from the Southern Ocean (referred to as S. aff. kroyeri). The S. cf. kroyeri group appeared sister to S. cf. cirrata. The average p-distances for the individual gene fragments between these groups are given in S3-S7 Tables. Spiophanes aff. kroyeri appeared distant from the clade S. kroyeri-S. cf. cirrata.
Two sister groups were revealed among specimens identified by morphology as S. duplex: one from California, USA (referred to as S. duplex), and another from Paraná, Brazil (referred to as S. soederstroemi). The p-distances between individuals of these groups ranged from 0.18% for 18S to 5.05% for Histone 3.
Horned Spiophanes with metameric nuchal organs formed a clade sister to the clade of Spiophanes with bell-shaped prostomium and nuchal organs as a pair of long parallel ciliary bands extending over about 15 anterior chaetigers ( Fig 2B).
The analysis showed identity (p-distance = 0.00%) of 18S and 28S rDNA sequences of Trochochaeta multisetosa from Norway, Sweden and California, USA.
An additional analysis showed identity (p-distance = 0.00%) of our sequence of S. berkeleyorum from Southern California and short fragments of 18S rDNA (490 bp) of S. berkeleyorum from Northern California obtained by Meissner & Blank [111] (Fig 2A).
Divergence time estimation. The alignment of COI sequences obtained in the present study and those from GenBank and BOLD comprised 534 bp with 170 (31.8%) variable sites, of which 160 (94.1%) sites were parsimony informative. The BEAST analysis resulted in a fully resolved consensus tree (Fig 3) with the same topology as the tree from the five-genes Bayesian analysis ( Fig 2B). The analysis suggested that the origin of the most recent common ancestor of horned Spiophanes with metameric nuchal organs was around 11.1 mya (95% HPD: 5.1-19.0 mya) and that the divergence of the North Atlantic and North Pacific lineages was around 7.9 mya (95% HPD: 4.1-13.3 mya). The North Atlantic lineage was estimated to have diverged 4.8 mya (95% HPD: 2.2-8.6 mya), resulting in the origin of S. cf. bombyx and S. cf. convexus. The North Pacific lineage was estimated to have diverged first by the isolation and speciation of S. norrisi 1.7 mya (95% HPD: 2.3-1.0 mya), then by the isolation and speciation of S. uschakowi and S. hakaiensis n. sp. 1.3 mya (95% HPD: 2.0-0.7 mya).
Morphology and biology
Spiophanes bombyx (Claparède, 1870 Synopsis. Up to 38 mm long, 1 mm wide for 120 chaetigers. Prostomium with long fronto-lateral horns, posteriorly pressed into chaetiger 1 but not extending over it as caruncle. Occipital antenna absent. Nuchal organs metameric, on a series of anterior chaetigers; first pair of metamers as oblique ciliary bands from prostomium to end of chaetiger 2; succeeding metamers shorter. Sabre chaetae in neuropodia from chaetiger 15. Hooks in neuropodia from chaetiger 15, usually quadridentate, with small subterminal hood. Pygidium with small ventral fleshy pad and one pair of lateral cirri. Glandular organs in chaetigers 5-14, largest in chaetigers 7 and 8, smallest in chaetiger 9, gradually increasing in size from chaetiger 10 to chaetigers 12-13, slightly smaller again in chaetiger 14. Internal fibres long and coiled in glandular organs in chaetigers 5-8, straight, shorter and thinner in chaetigers 9-14. Organs on chaetigers 5, 7 and 8 each opening to exterior via semicircular to suboval slit around large fiber spreader, on chaetiger 6 via small round hole, on chaetigers 9-14 via large vertical slit. Frontal edge of each fiber spreader on chaetigers 5, 7 and 8 entire, rounded to blunt, or with variously developed middle depression vaguely separating two rounded lobes. Digestive tract with gizzard-like structure. Main dorsal blood vessel with heart body. Nephridia from chaetiger 15 onwards. Gonochoristic. Oocytes lentiform, each with honeycombed envelope and 18-20 cortical alveoli regularly arranged in a peripheral circle. Spermatids interconnected in tetrads. Spermatozoa short-headed aqua-sperm with plate-like acrosomes. Fertilization in sea water. Larval development holopelagic, planktotrophic. Larvae with two pairs of red eyes on prostomium and a midventral ciliated pit on chaetiger 2. Remarks. In the original description of S. bombyx, Claparède ([9]: 485, as Spio bombyx) noted that "It does not seem rare in Naples. At least, it lives in numerous societies. I received hundreds of them one day, lodged side by side in tubes of black mud." He reported worms up to 38 mm long, 1 mm wide. The type material of S. bombyx does not exist any longer, probably because Claparède did not deposit the examined specimens at any museum. After Claparède, the species has rarely been reported from Italy and seems not to be very common in the Gulf of Naples at present. We were unable to find a sequence of S. bombyx from the Mediterranean in the literature. We collected one individual of S. bombyx from Venice Lagoon, Italy, and sequenced it for 16S and 28S gene fragments (see Table 1).
Morphological examination and molecular analysis of horned Spiophanes performed in the present study discovered two sibling species in northern European waters which we refer to as S. cf. bombyx and S. cf. convexus. Their morphology, taxonomy and the distribution are discussed below. It seems plausible that S. bombyx does not occur in northern European waters and that its distribution is limited to the Mediterranean only. To avoid confusion in identification of specimens based on morphological characters only, especially in areas of possible overlap of the species, and until more molecular data has been obtained from specimens from the Gulf of Naples, we suggest that northern European horned Spiophanes are referred to as S. bombyx aggregate (S. bombyx agg.), rather than to a particular species.
Distribution. Mediterranean Sea ( Fig 4A). Adult morphology (based on specimens from the Norwegian and North Seas, and around the British Isles and Iceland). McIntosh [15] noted that worms from St Andrews, Scotland, UK, were about 3 inches (76 mm) long for about 180 chaetigers; Meißner & Blank [111] examined specimens 0.2-1.5 mm wide.
Worms examined in the present study 0.3-1.5 mm wide, up to 30 mm long for 120 chaetigers. Pigmentation absent on body and palps. In large formalin-fixed specimens, lateral sides of chaetigers 9-14 (until chaetigers 16-17 in large specimens) light brownish-red due to fixed internal content of large, probably glandular, epithelial cells (Fig 5A and 5B).
Prostomium triangular, wide anteriorly, with a pair of long, distally pointed fronto-lateral horns (Fig 5C), posteriorly narrowed, pressed into chaetiger 1 but not extending over as a caruncle. Peristomium with narrow lateral lips closely applied to lateral sides of prostomium (Fig Nuchal organs metameric, at least on 20 anterior chaetigers, fewer in small individuals. First pair of metamers long, oblique ciliary bands extending from posterior part of prostomium to end of chaetiger 2, shorter in small individuals; posterior ends of metamers set wider apart. Succeeding metamers shorter, each extending from nototroch over posterior half of chaetiger, oriented parallel to body axis until chaetigers 13-14, oblique on succeeding chaetigers.
Sabre chaetae in neuropodia from chaetiger 15, one or rarely two in a tuft, with narrow limbation and fine granulation on distal part (Fig 6B-6D). Hooks in neuropodia from chaetiger 15, up to 10 in a series, accompanied by inferior sabre chaetae throughout body and 1-2 short hair-like alimbate capillaries in 5-10 posterior chaetigers; hair-like capillaries in posterior neuropodia usually situated in upper and/or lower parts of hook row ( Fig 6D). Hooks usually quadridentate, with three small upper teeth above main fang and small subterminal hood below main fang (Fig 6B-6D); median upper tooth occasionally weakly developed or double.
Pygidium with small fleshy ventral pad and one pair of thin long lateral cirri ( Fig 5D). Glandular organs in chaetigers 5-14, largest in chaetigers 7 and 8, smallest in chaetiger 9 and gradually increasing in size from chaetiger 10 to chaetigers 12-13, slightly smaller again in chaetiger 14 (Figs 5B and 7A-7F); large organs occupying most of chaetiger space. Internal fibres long and coiled in glandular organs in chaetigers 5-8 ( Fig 7B and 7C), straight, shorter and thinner in chaetigers 9-14; in fixed specimens, long straight fibers usually protruding from openings of glandular organs on chaetiger 6. Each organ on chaetigers 5, 7 and 8 opening to exterior via semicircular to suboval slit around large fiber spreader, on chaetiger 6 via small round hole, and on chaetigers 9-14 via large vertical slit. Frontal edge of fiber each spreader on chaetigers 5, 7 and 8 entire, rounded to blunt, or with variously developed middle depression vaguely separating two rounded lobes.
Methyl green staining. Intensely stained lateral sides of chaetiger 6; weakly stained lateral sides of chaetiger 5. This pattern was observed in small and large specimens from the North and Norwegian seas. In large specimens, the dorsal side of the prostomium and lateral sides of the peristomium were usually also intensely stained.
Reproduction. Spiophanes cf. bombyx is gonochoristic. The gametes proliferate in paired gonads attached to the genital blood vessels from chaetigers 18-19 to chaetigers 100-110. In females, oogenesis is intraovarian: vitellogenesis occurs when the oocytes grow in ovaries. The developed oocytes are accumulated in the coelomic cavity prior to spawning ( Fig 8A). The largest intraovarian oocytes examined in fixed specimens from Norway were of irregular shape, up to 105 μm in diameter, each with a honey-combed envelope 3-4 μm thick, 45-55 cortical alveoli regularly arranged in a peripheral circle, a nucleus about 55 μm in diameter, and a single nucleolus about 20 μm in diameter. The cortical alveoli were pear-shaped, up to 8 μm long and 5 μm in diameter, with narrow necks oriented perpendicular to the oocyte surface ( Fig 8B-8E).
In males, spermatogonia proliferate in testes and the rest of spermatogenesis occurs in the coelomic cavity. Spermatids are interconnected in tetrads. The spermatozoa are ect-aquasperm with a short plate-like acrosome, subspherical nucleus about 3 μm in diameter, very short midpiece, and a long flagellum (Fig 8F).
Females and males release their gametes into the water where fertilization and holopelagic, planktotrophic larval development occur. Larvae have two pairs of red eyes on the prostomium and a midventral ciliated pit on chaetiger 2 (see Hannerz [163] : Fig 9a, 9b).
Remarks. Horned Spiophanes from the North and Norwegian seas morphologically appear similar to the individual from the Adriatic Sea and almost fit the original and later descriptions of S. bombyx from the Mediterranean. All the adults have sabre chaetae and hooks in neuropodia from chaetiger 15 onwards. They differ, however, in the number of cortical alveoli which are regularly arranged in the peripheral circle in mature oocytes. Claparède ([9]: 486, pl 12, Fig 2E, 2F) noted that in worms from the Gulf of Naples the developed oocytes were lentiform (sphéroïdes aplatis), up to 130 μm in diameter, with thick papillary envelope, a large spherical germinal vesicle and a spherical nucleolus; 18 to 20 colorless cortical alveoli ("vésicules inclores" in his terminology), each about 11 μm in diameter, were arranged in a large circle. We observed 45-55 alveoli per oocyte in worms from the Norwegian Sea (Fig 8B-8E). Sequences of the North European and single Italian specimen obtained in the present study were similar (p-distances ranging from 00% for 28S to 2.46% for 16S). Nevertheless, because of the difference in the number of cortical alveoli in oocytes (which is supposed to be species specific and variable within a small range), and the interrupted distribution of the North European and the Mediterranean populations (see below S. cf. convexus), we suggest that the conspecificity of these populations should be verified in additional molecular analysis based on sequences of more genes and more individuals. Pending molecular data, we refer to the specimens from the North and Norwegian seas, and also from around Iceland and British Isles as S. cf. bombyx.
Distribution. Northern Europe (southern limits uncertain) ( Fig 4A). Adult morphology (based on our material from Brittany, France). Up to 30 mm long, 1 mm wide for 120 chaetigers. Pigmentation absent on body and palps. In formalin-fixed specimens, lateral sides of chaetigers 9-14 (until chaetigers 16-17 in large specimens) light brownish due to fixed internal content of large, probably glandular, epithelial cells.
Prostomium triangular, wide anteriorly, with a pair of distally pointed fronto-lateral horns, posteriorly narrowed, pressed into chaetiger 1 but not extending over as a caruncle. Fronto-lateral horns long (Fig 9C) or moderately developed ( Fig 9D). Peristomium with narrow lateral lips closely applied to lateral sides of prostomium, and a small ventral lip. Two pairs of small dark red eyes present or eyes absent. Occipital antenna absent. Palps as long as 7-15 chaetigers.
Nuchal organs metameric, at least on 40 anterior chaetigers, fewer in small individuals. First pair of metamers long oblique ciliary bands extending from posterior part of prostomium to end of chaetiger 2; posterior ends of metamers set wider apart. Succeeding metamers shorter, each extending from nototroch over posterior half of chaetiger, oriented parallel to body axis until chaetigers 10-11, slightly oblique on succeeding chaetigers. Epithelial cells around nuchal patches with dark yellow pigment in life.
Sabre chaetae in neuropodia from chaetiger 15, one or rarely two in a tuft, with narrow limbation and fine granulation on distal part. Hooks in neuropodia from chaetiger 15, up to 10 in a series, accompanied by inferior sabre chaetae throughout body and 1-2 short hair-like alimbate capillaries in 5-10 posterior chaetigers; hair-like capillaries in posterior neuropodia usually situated in upper and/or lower parts of hook row. Hooks usually quadridentate, with three small upper teeth above main fang and small subterminal hood below main fang (Fig 9G).
Nototrochs from chaetiger 1 to end of body; on each of chaetigers 1 and 2 as two short, oblique ciliary bands situated between notopodial postchaetal lamellae and nuchal ciliary bands; from chaetiger 3 through most succeeding chaetigers as complete transverse bands of dense cilia running between notopodial postchaetal lamellae. Each nototroch composed of two parallel rows of large transversally elongated cells, each bearing numerous long cilia; nototroch ciliation stronger on ridge-bearing chaetigers. Single transverse rows of smaller cells with shorter cilia present on each chaetiger beginning from chaetiger 3.
Dorsal transverse ridges present from chaetigers 14-15 to chaetigers 40-50, fewer in small individuals. Ridges thick, fleshy, one per chaetiger, not connected to notopodial lamellae. Wide blood vessel extending along each ridge internally; double nototroch with numerous long cilia arranged on top of each ridge. Intersegmental lateral pouches absent.
Pygidium with small fleshy ventral pad and one pair of thin long lateral cirri (Fig 9H and 9I). Glandular organs in chaetigers 5-14, largest in chaetigers 7 and 8, smallest in chaetiger 9 and gradually increasing in size from chaetiger 10 to chaetigers 12-13, slightly smaller again in chaetiger 14; large organs occupying most of chaetiger space. Internal fibres long and coiled in glandular organs in chaetigers 5-8, straight, shorter and thinner in chaetigers 9-14; in fixed specimens, long straight fibers usually protruding from openings of glandular organs on chaetiger 6 ( Fig 9E). Each organ on chaetigers 5, 7 and 8 opening to exterior via semicircular to suboval slit around large fiber spreader, on chaetiger 6 via small round hole, and on chaetigers 9-14 via large vertical slit (Fig 9A, 9B and 9E). Frontal edge of each fiber spreader on chaetigers 5, 7 and 8 entire, rounded to blunt, or with variously developed middle depression vaguely separating two rounded lobes.
Remarks. Horned Spiophanes from Brittany, northern France, morphologically appear very similar or almost identical to S. bombyx from the Mediterranean and S. cf. bombyx from the North and Norwegian seas, but unambiguously differ from them in genetic characteristics (Figs 1B and 2B). They weakly differ from North European specimens also by the pattern of MG staining: no staining on head and five anterior chaetigers in S. cf. convexus, and staining on chaetiger 5 and often on the prostomium and peristomium in S. cf. bombyx. However, these patterns are variable and can only be seen in well preserved specimens.
Delgado-Blas et al. [162] described S. convexus based on the morphology of a few anterior fragments from Atlantic Spain (Galicia and Andalusia), and also from Mediterranean Spain (Catalonia). The diagnostic characters used to distinguished S. convexus from S. bombyx were ambiguous and fit into the range of morphological variability of S. bombyx, which was not examined by Delgado-Blas et al. [162]. Based on the proximity of the type locality of S. convexus (Meira, Pontevedra, Galicia) and sampling sites of specimens examined through molecular analysis in the present study, we assume that our specimens from Brittany, France, and specimens from Bay of Biscay sequenced by Aylagas et al. [121] belong to S. convexus. In the absence of sequence data of S. convexus, however, we refer our specimens to as S. cf. convexus. The identity of the horned Spiophanes from Catalonia, Spain, referred by Delgado-Blas et al. [162] to as S. convexus, should be verified in molecular analysis. So far, based on available data, we consider S. bombyx as the only horned species occurring in the Mediterranean.
To avoid confusion in identification of specimens based on morphological characters only, especially in areas of possible overlap of S. bombyx and S. convexus, and until more molecular data has been obtained from specimens from the Gulf of Naples, we suggest that North European horned Spiophanes are referred to as S. bombyx aggregate (S. bombyx agg.), rather than to a particular species.
Mesnil [11] described in detail adults and larvae of S. bombyx from Wimereux, Hauts-de-France, northern France. He noted that the adults were up to 6 cm long, 1.5 mm wide for about 180 chaetigers. They had hooks beginning invariably from chaetiger 15, whereas in larvae hooks started from chaetigers 11-14. Wimereux is located midway between the sites where S. cf. bombyx and S. cf. convexus were collected for molecular analysis. Because Mesnil's [11] description fits both species, the identity of the worms described by him from Wimereux remains uncertain.
Cazaux [166] described in detail the larval development of a horned Spiophanes from Arcachon, France, which he identified as S. bombyx. He noted that in an early benthic 15-chaetiger juvenile, hooks were not yet developed, and in a 19-chaetiger juvenile they were present in chaetigers 13-19. Arcachon is located between sites where specimens of S. cf. convexus were collected for molecular analysis. It is plausible, therefore, that Cazaux [166] dealt with S. convexus but not S. bombyx.
Distribution. Western Europe (northern and southern limits uncertain) (Fig 4A). Adult morphology. Up to 47 mm long, 1.5 mm wide for 140 chaetigers. Pigmentation absent on body and palps. A pair of narrow transverse whitish bands, concentrations of epithelial glandular cells, visible on ventral side of chaetiger 1 in live individuals. In large formalinfixed specimens, lateral sides of chaetigers 9-14 (until chaetigers 16-17 in large specimens) light brownish-red due to fixed internal content of large, probably glandular, epithelial cells.
Spiophanes uschakowi
Prostomium triangular, wide anteriorly, with a pair of long, distally pointed fronto-lateral horns (Fig 11A-11D), posteriorly narrowed, pressed into chaetiger 1 but not extending over it as a caruncle. Peristomium with narrow lateral lips closely applied to lateral sides of prostomium, and a small ventral lip (Fig 11B-11D). Two pairs of small dark red eyes arranged trapezoidally, comprising one pair of median eyes and one pair of lateral eyes situated anteriorly and set wider apart; occasionally eyes absent. Occipital antenna absent. Palps as long as 7-15 chaetigers, with deep frontal longitudinal groove lined with fine cilia, fronto-lateral motile compound cilia bordering frontal groove, short transverse rows of short motile compound cilia regularly arranged on inner lateral side and beating towards frontal groove, short compound non-motile cilia arising directly from palp surface and scattered on lateral and abfrontal palp surfaces (Fig 11H).
Nuchal organs metameric, at least on 25 anterior chaetigers, fewer in small individuals. First pair of metamers long oblique ciliary bands extending from posterior part of prostomium to end of chaetiger 2, shorter in small individuals; posterior ends of metamers set wider apart (Fig 11C). Succeeding metamers shorter, each extending from nototroch over posterior half of chaetiger, oriented parallel to body axis until chaetigers 9-10 ( Fig 11F), oblique on succeeding chaetigers (Fig 11G).
Hooks in neuropodia from chaetiger 15, up to 15 in a series, accompanied by inferior sabre chaetae throughout body and 1-2 hair-like alimbate capillaries in 7-10 posterior chaetigers; hair-like capillaries in posterior neuropodia usually situated in upper and/or lower parts of hook row. Hooks usually quadridentate, with three small upper teeth above main fang and small subterminal hood below main fang ( Fig 13B); median upper tooth occasionally weakly developed or doubled (Fig 13A and 13B). Nototrochs from chaetiger 1 to end of body; on each of chaetigers 1 and 2 as two short, oblique ciliary bands situated between notopodial postchaetal lamellae and nuchal ciliary bands; from chaetiger 3 onwards as complete transverse bands of dense cilia running between notopodial postchaetal lamellae (Figs 11C, 11F, 11G and 12C). Each nototroch composed of two parallel rows of large transversally elongated cells, each bearing numerous long cilia; nototroch ciliation stronger on ridge-bearing chaetigers. Single transverse rows of smaller cells with shorter cilia present on each chaetiger beginning from chaetiger 3. These rows situated posterior to nototrochs and referred to as intersegmental transverse ciliation (Fig 11F and 11G).
Dorsal transverse ridges present from chaetiger 15 to chaetigers 30-55, fewer in small individuals. Ridges thick, fleshy, one per chaetiger, not connected to notopodial lamellae. Wide blood vessel extending along each ridge internally; double nototroch with numerous long cilia arranged on top of each ridge. Intersegmental lateral pouches absent.
Pygidium with small fleshy ventral pad and one pair of thin long lateral cirri ( Fig 11E); one to two additional cirri present above lateral cirri in some individuals.
Nephridia from chaetiger 15 onwards, opening to exterior on antero-lateral edges of chaetigers in both sexes. Habitat. Adults of S. uschakowi live in tubes in sandy to muddy-sand sediments in shallow waters. Population densities reach several thousand individuals per square meter. The tubes are sandy, firm, with thin but strong inner lining, up to 10 cm long and 2 mm in diameter, oriented vertically in sediment, with upper part 2-5 mm long protruding above the surface.
Methyl green staining. Intensely stained small area on lateral sides of chaetiger 5 (in front of openings of glandular organs; not stained in some individuals), larger area on chaetiger 6 ( Fig 10B) and small fleshy pygidium (pygidial cirri not stained); weakly stained lateral sides of chaetigers 9-14.
Reproduction. Spiophanes uschakowi is gonochoristic. The gametes proliferate in paired gonads attached to the genital blood vessels from chaetigers 20-24 to chaetigers 88-120. In females, oogenesis is intraovarian: vitellogenesis occurs when the oocytes grow in ovaries. The developed oocytes are accumulated in the coelomic cavity prior to spawning. The newly released oocytes are lentiform, each 185-200 μm in diameter, with honeycombed envelope 5-7 μm thick, 41-49 cortical alveoli regularly arranged in a peripheral circle, a nucleus 80-83 μm in diameter, and a single nucleolus about 30 μm in diameter. The cortical alveoli are pear-shaped, 7-9 μm in diameter, with narrow necks oriented perpendicular to the oocyte surface (see Radashevsky et al. [167]: Fig 2).
Females and males release their gametes into the water where fertilization and holopelagic, planktotrophic larval development occur. Larvae have two pairs of red eyes on the prostomium and a midventral ciliated pit on chaetiger 2. Newly settled 20-21-chaetiger juveniles about 2 mm long had heavy recurved crook-like spines in neuropodia of chaetiger 1, sabre chaetae in neuropodia from chaetiger 10, hooded hooks in neuropodia from chaetiger 14, and long spines with recurved distal end in notopodia from chaetigers 17-19. Remarks. In the original description of S. uschakowi Zachs ([112]: 130) noticed characteristic feature of worms as notopodial lamellae ("cirri dorsales") being leaf-like on eight anterior chaetigers and cirriform on the following chaetigers. He noted that hooks appeared rather similar to those in S. bombyx and incorrectly reported that the first chaetiger had no heavy spines. Annenkova [168,113] and Uschakov [78] likely had not seen Zachs' material and distinguished S. bombyx and S. uschakowi only according to Zachs' description. We have examined the two syntypes of S. uschakowi (two anterior fragments, each about 25-35 chaetigers) deposited in the polychaete collection of the Zoological Institute, Saint Petersburg, Russia (ZISP 1/ 25826), and observed the heavy recurved spines in first neuropodia, and other characters as they described above.
The ultrastructure of the oocytes and spermatozoa was described in details by Radashevsky et al. [167]. The larvae of the species were found in plankton collected in Peter the Great Bay, Sea of Japan (East Sea), Russia, in September-October. Their morphology will be described elsewhere.
Spiophanes uschakowi has been recorded from the Shantar Islands, northern part of Sakhalin Is. and from the South Kurile Islands south to the Korean Peninsula. Numerous records of S. bombyx from around Japanese Islands by Minoru Imajima are herein referred to S. uschakowi but the conspecificity of mainland and Japanese specimens should be verified in a further study. Remarkably, S. uschakowi has never been reported from the Chukchi Peninsula, Kamchatka Peninsula, Bering Islands, Kurile Islands, and northern part of the Sea of Okhotsk, although numerous samples from this region were provided by the expeditions of the Institute of Marine Biology, Vladivostok, FEB RAS, the Zoological Institute, St. Petersburg, RAS, some other institutions, and also during this study.
Distribution. North West Pacific: from the Sea of Okhotsk south to the Yellow Sea and East China Sea (Fig 15A). Synopsis. Up to 15 mm long, 0.6 mm wide for 100 chaetigers. Prostomium with long fronto-lateral horns, posteriorly pressed into chaetiger 1 but not extending over it as caruncle. Occipital antenna absent. Nuchal organs metameric, at least on 25 anterior chaetigers; first pair of metamers as oblique ciliary bands from prostomium to end of chaetiger 2; succeeding metamers shorter, parallel to body axis until chaetigers 10-11, oblique on succeeding chaetigers. Sabre chaetae in neuropodia from chaetiger 10. Hooks in neuropodia from chaetiger 15, usually quadridentate, with small subterminal hood. Pygidium with small ventral fleshy pad and one pair of lateral cirri. Glandular organs in chaetigers 5-14, largest in chaetigers 7 and 8, smallest in chaetiger 9, gradually increasing in size from chaetiger 10 to chaetigers 12-13, slightly smaller again in chaetiger 14. Internal fibres long and coiled in glandular organs in chaetigers 5-8, straight, shorter and thinner in chaetigers 9-14. Organs on chaetigers 5, 7 and 8 each opening to exterior via semicircular to suboval slit around large fiber spreader, on chaetiger 6 via small round hole, on chaetigers 9-14 via large vertical slit. Frontal edge of each fiber spreader on chaetigers 5, 7 and 8 entire, rounded to blunt, or with variously developed middle depression vaguely separating two rounded lobes. Digestive tract with gizzard-like structure. Main dorsal blood vessel with heart body. Nephridia from chaetiger 15 onwards. Gonochoristic. Oocytes lentiform, each with thick honey-combed envelope and cortical alveoli regularly arranged in a peripheral circle. Spermatids interconnected in tetrads. Spermatozoa shortheaded aqua-sperm with plate-like acrosomes. Fertilization in sea water. Larval development holopelagic, planktotrophic. Larvae with two pairs of red eyes on prostomium and a midventral ciliated pit on chaetiger 2.
Spiophanes norrisi
Remarks. Meißner & Blank [111] designated 17 specimens from one sample collected from Magdalena Bay, Baja California Sur, Pacific Mexico, as the type material of S. norrisi (Fig 15B). They referred to numerous specimens of horned Spiophanes collected along the Pacific American coast from Alaska south to Chile as non-type material of the same species. For molecular analysis (short fragments of COI), the authors used specimens collected from offshore San Francisco (SF), northern California.
Molecular analysis performed in the present study suggested the presence of two sibling vicariant species distributed along the Pacific coast of North America and overlapping in California (Figs 2B and 3). Herein, we refer northern specimens to a new species: Spiophanes hakaiensis n. sp. (see below). The COI sequences provided by Meißner & Blank [111] were grouped together with sequences from specimens from off San Francisco and Baja California Norte, Mexico, provided in other studies (see Table 1, Fig 2B). Consequently, these specimens are referred to as S. norrisi. South American populations of horned Spiophanes have not been examined in genetic analysis. Pending molecular data, we suggest that Chilean and Argentinean specimens are referred to as S. cf. norrisi (see below).
Meißner & Blank [111] noticed that S. norrisi differed from morphologically similar species mainly in the orientation of segmental nuchal metamers ("dorsal ciliated patches", "metameric ciliated patches" or "dorsal ciliated organs" in their terminology), and that it was the only known species with oblique orientation of the metamers from as early as between chaetigers 9-10. After describing horned American Spiophanes as a new species, S. norrisi, Meißner & Blank ([111]: 6) noted that S. uschakowi "turned out" to be very similar to S. norrisi, but "a definite assignment [of S. uschakowi] based on morphological characters was not possible as long as the nature of the dorsal ciliated organs posterior to chaetiger 2 was unknown." We observed segmental nuchal metamers oriented parallel to body axis until chaetigers 9-10 in S. uschakowi and until chaetigers 10-11 in S. norrisi; thus the two species cannot be distinguished based on the morphology of their nuchal organs. We did not observe sabre chaetae beginning earlier than chaetiger 10 in S. norrisi and interpret their rare start from chaetiger 9 reported by Meißner & Blank [111] as an abnormality, as with the occasional bifurcation of one or both pygidial cirri rarely occurring both in S. uschakowi and S. norrisi. Individuals of both species demonstrate the same variability in the shape of the chaetal spreaders on chaetigers 5, 7 and 8, and all other morphological characters examined so far. Specimens of the two species differ slightly in the patterns of the MG staining. In S. uschakowi, the lateral sides of chaetiger 5 (in front of openings of glandular organs) are usually weakly stained (Fig 10B), whereas in S. norrisi such staining has not been observed (Fig 10E). The staining pattern in S. uschakowi, however, is variable and in some individuals (worms from Peter the Great Bay were mostly examined on this account) staining was not revealed on chaetiger 5, making this character also unreliable for distinguishing the two species. Blake [178] described 13-and 14-chaetiger horned larvae of a Spiophanes collected in Tomales Bay, California, in November 1971. The 14-chaetiger larva was 1025 μm long and had hooded hooks in neuropodia from chaetiger 11. Blake [178] noticed subtle morphological differences between the Californian and Swedish larvae (described by Hannerz [163]) and referred the former to as S. cf. bombyx. However, two other horned Spiophanes, S. anoculata Hartman, 1960 and S. hakaiensis n. sp., also occur in Californian waters. As with S. bombyx and S. norrisi, adults of S. anoculata have hooded hooks in neuropodia from chaetiger 15. So far, S. anoculata has only been found in deep water, from 922 m to 2800 m depth (see Meißner [8]), but because Blake [178] collected very few (only two) larvae of same kind in shallowwater plankton, these larvae may belong to a deep-water species. Thus, their identity remains uncertain.
Distribution. North East Pacific, California: from San Francisco Bay, USA, south to Baja California Sur, Mexico (Fig 15B). [9,10]. Adult morphology (based on the type material-specimens identified by means of molecular analysis in the present study). Up to 25 mm long, 1.3 mm wide for 110 chaetigers. Pigmentation absent on body. Few examined live complete individuals with fine black pigment scattered on palps (Fig 16F). In large formalin-fixed specimens, lateral sides of chaetigers 9-14 (until chaetigers 16-17 in largest specimens) light reddish due to fixed internal content of large, probably glandular, epithelial cells.
Prostomium triangular, wide anteriorly, with a pair of long, distally pointed fronto-lateral horns (Fig 16A, 16B and 16F), posteriorly narrowed, pressed into chaetiger 1 but not extending over as a caruncle. Peristomium with narrow lateral lips closely applied to lateral sides of prostomium, and a small ventral lip. Two pairs of small dark red eyes arranged trapezoidally, comprising one pair of median eyes and one pair of lateral eyes situated anteriorly and set wider apart; occasionally eyes absent. Occipital antenna absent. Palps as long as 7-15 chaetigers, with deep frontal longitudinal groove lined with fine cilia, fronto-lateral motile compound cilia bordering frontal groove, short transverse rows of short motile compound cilia regularly arranged on inner lateral side and beating towards frontal groove (Fig 16C and 16F), short compound non-motile cilia arising directly from palp surface and scattered on lateral and abfrontal palp surfaces.
Nuchal organs metameric, at least on 25 anterior chaetigers, fewer in small individuals. First pair of metamers long oblique ciliary bands extending from posterior part of prostomium to end of chaetiger 2, shorter in small individuals; posterior ends of metamers set wider apart.
Succeeding metamers shorter, each extending from nototroch over posterior half of chaetiger, oriented parallel to body axis until chaetigers 10-11, oblique on succeeding chaetigers.
Sabre chaetae in neuropodia from chaetiger 10, usually one, occasionally two per neuropodium, with fine granulation on distal part; chaetae on chaetiger 10 comparatively small, with narrow limbation, slightly reducing in size until chaetiger 14, but from chaetiger 15 onwards thicker, alimbate. Hooks in neuropodia from chaetiger 15, up to 10 in a series, accompanied by inferior sabre chaetae throughout body and 1-2 hair-like alimbate capillaries in 5-10 posterior chaetigers. Hooks usually quadridentate, with three small upper teeth above main fang and small subterminal hood below main fang.
Nototrochs from chaetiger 1 to end of body, each composed of two parallel rows of large transversally elongated cells, each bearing numerous long cilia; nototroch ciliation stronger on ridge-bearing chaetigers. Intersegmental transverse ciliation from chaetiger 3 onwards.
Pygidium with small fleshy ventral pad and one pair of thin long lateral cirri. Glandular organs in chaetigers 5-14, largest in chaetigers 7 and 8, smallest in chaetiger 9 and gradually increasing in size from chaetiger 10 to chaetigers 12-13, slightly smaller again in chaetiger 14; large organs occupying most of chaetiger space. Internal fibres long and coiled in glandular organs in chaetigers 5-8, straight, shorter and thinner in chaetigers 9-14; in fixed specimens, long straight fibers usually protruding from openings of glandular organs on chaetiger 6 ( Fig 16G). Each organ on chaetigers 5, 7 and 8 opening to exterior via semicircular to suboval slit around large fiber spreader, on chaetiger 6 via small round hole, and on chaetigers 9-14 via large vertical slit. Frontal edge of each fiber spreader on chaetigers 5, 7 and 8 blunt or with variously developed middle depression vaguely separating two rounded lobes.
Nephridia from chaetiger 15 onwards, opening to exterior on antero-lateral edges of chaetigers in both sexes.
30-40-chaetiger juveniles about 0.4 mm wide with sabre chaetae in neuropodia from chaetiger 10, 1-2 hooded hooks among 4-6 capillary chaetae and single sabre chaeta in each neuropodium of chaetiger 14, and only hooks and sabre chaetae in neuropodia from chaetiger 15 onwards. Individuals more than 0.4 mm wide with morphological characters as described above in Adult morphology, with sabre chaetae in neuropodia from chaetiger 10 and hooded hooks from chaetiger 15.
Reproduction. Spiophanes hakaiensis n. sp. is gonochoristic. Mature females and males collected in British Columbia, Canada, in May 1966 (RBCM 55-47), have gametes from chaetiger 21 through most of the body. The oocytes have honey-combed envelopes with cortical alveoli regularly arranged in a peripheral circle. Spermatids are interconnected in tetrads; the spermatozoa were ect-aquasperm with a short acrosome and short, subspherical nucleus about 3 μm in diameter.
Remarks. Horned Spiophanes from the North East Pacific appear morphologically very similar or almost identical to each other and to S. uschakowi from the North West Pacific. However, molecular analysis performed in the present study showed significant difference between the North East and North West populations and also suggested the presence of two sibling vicariant species distributed along the Pacific coast of North America and overlapping in California (Figs 1B, 2B and 3). Sequences of some Californian specimens (including COI sequences provided by Meißner & Blank [111]) were grouped together with sequences of specimens from Baja California Norte, Mexico (see Table 1, Fig 1B). Because the latter locality is close to the type locality of S. norrisi, those specimens are referred to this species. Other Californian specimens were genetically similar to horned Spiophanes from Alaska, USA, and British Columbia, Canada. Those specimens are herein referred to a new species: Spiophanes hakaiensis n. sp. Comments on the taxonomy and the distribution of these species are provided above in the Remarks for S. norrisi, and below in the Remarks for S. cf. norrisi.
To avoid confusion in the identification of specimens based on the morphology only, in areas of overlap of S. hakaiensis n. sp. and S. norrisi, we suggest that specimens from California between Point Reyes (38˚N) and Point Conception (34.45˚N) are referred to as S. norrisi aggregate; specimens occurring from Point Reyes north to Alaska are herein referred to as S. hakaiensis n. sp.; whereas those occurring from Point Conception south to Baja California Sur, Mexico, are referred to as S. norrisi.
Distribution. From Alaska south to San Francisco Bay, California, USA (Fig 15A). Remarks. Hartmann-Schröder [184] described Chilean Spiophanes (distributed from Punta Tortuga, Coquimbo, south to Gulf of Corcovado) with prostomium expanded anteriorly as S. chilensis Hartmann-Schröder, 1965. Meißner [8] and Meißner & Blank [111] re-examined the type material of S. chilensis and distinguished specimens belonging to three species: Spiophanes duplex (Chamberlin, 1919), Spiophanes fimbriata Moore, 1923, and S. norrisi. Meißner [8] referred the holotype of S. chilensis to S. duplex and considered the former species as a junior synonym of the latter (see below). Meißner & Blank ([111]: 14-15) referred 22 paratypes of S. chilensis from Punta Tortuga (29˚57.350' S, ZMH P-14946) and additional material from Bahia Quillaipe (41˚32.469' S, ZMH P-21118, 2 spec.) to S. norrisi and concluded that the latter species "occurs from shallow waters up to subtidal depths waters along the North and South American coast." Meißner & Blank [111] did not mention the records of S. bombyx from Chile by Carrasco [67,119] and from Argentina and the Falkland Islands by Blake [48], although these two authors noticed that their specimens were identical with S. bombyx from Europe. We also had no chance to examine Carrasco's or Blake's material but got new material from Chile and Argentina (see S1 Table). We could not find any difference between this material and corresponding specimens from the Pacific USA and Canada, either in the morphology or in the MG staining (Fig 10C-10E). Nevertheless, pending results of a comparative molecular analysis of North and South American specimens (see above Remarks for S. norrisi), we suggest that Chilean and Argentinean specimens are referred to as S. cf. norrisi. It is quite possible that South American horned Spiophanes, although morphologically appearing similar to North American S. norrisi, differ from them genetically and represent a different species. We see support to this idea in the limited distribution of worms in the southern part of South America only.
Remarks. Spiophanes duplex was originally described from Laguna Beach, Balboa, California, USA, by Chamberlin ([185], as Morants duplex). The original description was brief, with some features misinterpreted, without any illustration. Misinterpreted features greatly distinguished these worms from other spionids, and based on them, Chamberlin [185] Meißner [8] reexamined the type specimens of S. missionensis Hartman, 1941, S. soederstroemi Hartman, 1953, and holotype of S. chilensis Hartmann-Schröder, 1965 and, based on their morphological similarity, considered these species as junior synonyms of S. duplex (Chamberlin, 1919). Consequently, she suggested that S. duplex is widely distributed along both the Pacific and Atlantic coasts of both North and South America.
The results of the present study unambiguously show that specimens from Brazil, being morphologically identical to adults of S. duplex from California, differ from them essentially in genetic characteristics (Fig 2B). Therefore, we considered the two populations as not conspecific and refer the Brazilian population to S. soederstroemi (see below). Pending results of molecular studies in other American regions, we consider S. duplex as distributed in California only.
Distribution. North East Pacific: California (Fig 17). Hartman, 1953 was established by Hartman [201] based on two specimens from South America collected by the Swedish Antarctic Expedition 1901-1903 onboard the ship Antarctic. The specimens were first reported by Söderström [21] as a new record for S. kroyeri. Remarkably, Söderström ([21]: 243) and all the following authors noted that the specimens were collected on December 12, 1901 at station 1, off Uruguay. However, the coordinates provided for this station by Söderström [21] and also in a copy of the actual station list made during the expedition: 33˚00'S, 51˚10'W (SMNH, Lena Gustavsson in litt., 11 Sep 2019), and a map of the Expedition route with daily positions of the ship [217] show a sampling site of December 12, 1901 situated off Rio Grande do Sul, Brazil (Fig 17), north of the border between Brazil and Uruguay. Noteworthy, although this border was already well established in the mid 19 th century and did not change after that, it was not shown on the Expedition map. Possible uncertainty in its exact position might create confusion about reference on the type locality of S. soederstroemi. Blake [48] redescribed two syntypes of S. soederstroemi, reviewed a complicated history of this species and clarified some of its morphological features and taxonomic issues. He also reexamined the type specimens of Spiophanes chilensis Hartmann-Schröder, 1965, which was originally described from Punta Tortuga, Chile, by Hartmann-Schröder [184], and referred its holotype (ZMH P-14946) to S. soederstroemi. Consequently, Blake [48] considered S. chilensis as a junior synonym of S. soederstroemi. He also noticed subtle differences between S. soederstroemi and S. missionensis and considered them as two closely related but separate species (the holotype of Morants duplex was not yet discovered at that time).
During the present study, we examined live Spiophanes in California, USA, and Paraná, Brazil, which fit all the characteristics of S. duplex. We could not find any reliable morphological difference between them, but the differences between their molecular characteristics suggest the presence of two distinct species (Fig 2B). Consequently, we treat S. duplex and S. soederstroemi as two valid species, and consider S. missionensis as a junior synonym of S. duplex, and S. chilensis as a junior synonym of S. soederstroemi. The type localities of these species and sampling sites for the specimens used in the present molecular analysis are shown on Fig 17. Spiophanes soederstroemi is likely distributed in South America only. Records of S. soederstroemi from South Africa [100,101], South Georgia [109], Antarctica [218,219], China [90], Florida and North Carolina, USA [205], Philippines [220], and the Arabian Gulf [221] possibly represent misidentifications of different species.
Spiophanes kroyeri
Remarks. Grube ([200]: 89) noted in the original description of S. kroyeri that the new species was "Aus dem Meere von Grönland." No more detail about the type locality has ever been published, but the original registry catalogue and labels for the types of S. kroyeri maintained in the polychaete collection of the Zoologisches Museum Berlin (ZMB), Germany, say that the material was from Iceland, i.e. from the southern part of the Greenland Sea ( [222]; Birger Neuhaus in litt., 15 Jul 2019). A tentative position of the type locality is shown on the Fig 4B. Meißner ([8]: 13) considered S. kroyeri as "the most problematic species within the genus." She reviewed its worldwide reports and suggested that most of them were based on misidentifications. To clarify the identity of S. kroyeri, Meißner [8] designated a lectotype (ZMB Q4746) and provided a series of diagnostic characteristics of the adults. They included a bell-shaped prostomium, well developed occipital antenna, nuchal organs as a pair of longitudinal ciliary bands extending to chaetigers 14-16, suboval chaetal spreaders well developed on chaetigers 5-7, sabre chaetae in neuropodia from chaetiger 4, hooks with subterminal hood in neuropodia from chaetiger 15, lateral pouches first appearing between neuropodia of chaetigers 15 and 16, and pigmentation comprising light brown pigment in postchaetal lamellae of chaetigers 1-7 and brown pigment in neuropodia of chaetigers 7-14, most conspicuous in chaetigers 8-11.
Meißner ([8]: 14) commented on the problems with the original description and type material of S. kroyeri reyssi described from Mediterranean France by Laubier [223]. She raised it to species level, S. reyssi Laubier, 1964, and emphasized that "so far only specimens from the North Atlantic can be assigned to S. kroyeri".
The analysis of COI sequences of Spiophanes from the North Sea obtained by Meißner & Blank [111], from the Bay of Biscay, Spain, obtained by Aylagas et al. [121], and from the Barents Sea, Norway, obtained in the present study (see Table 1), revealed two distinct groups among specimens identified by morphology as S. kroyeri (Fig 1B). Worms of one group occurred in the Barents Sea, Norway, and in the northern part of the North Sea (vouchers HH64, 94 by Meißner & Blank [111]), whereas worms of another group occurred in the northern (vouchers HH18, 19 by Meißner & Blank [111]) and central parts of the North Sea (voucher HH63 by Meißner & Blank [111]), and in the Bay of Biscay, Spain. According to their distributions, we call these groups northern and southern, respectively. In relation to the type localities of S. kroyeri (Greenland Sea, off northern Iceland) and S. cirrata (Oslofjord, southern Norway, see below), we tentatively refer to the northern-group specimens as S. cf. kroyeri, and to the southern-group specimens as S. cf. cirrata (Fig 4B). Final conclusion about the specific identity of these worms can be inferred after molecular examination of worms from the type localities of the corresponding species in Iceland, Norway, and Mediterranean France. To avoid confusion in the identification of S. kroyeri-like specimens from the northern part of the North Sea (area of overlap of the two species) based on morphological characters only, we suggest that they are referred to as S. kroyeri aggregate (S. kroyeri agg.), rather than to a particular species.
It is noteworthy that deep-water Spiophanes worms from off the Crozet Islands, Southern Ocean, referred to as S. kroyeri by Mincks et al. [122], are genetically different from European specimens ( Fig 2B) and are therefore referred to herein as S. aff. kroyeri.
Distribution. Arctic; North Atlantic ( Fig 4B). Remarks. Michael Sars (in G.O. Sars [224]) briefly described Spiophanes cirrata based on material from Drøbak, Oslofjord, southern Norway (Fig 4B), and Skrova, small island group in the Lofoten Archipelago, northern Norway. Later, Sars [225] illustrated and provided a more complete re-description of this species. Tauber [230] and Söderström [21] considered S. cirrata to be a junior synonym of S. kroyeri. Hartman [199] listed S. cirrata as a valid species. Hartmann-Schröder [164,165] and Kirkegaard [231] neither noted nor commented on this species. As the first sampling site mentioned in the original description of the species by Michael Sars, Drøbak in Oslofjord is herein for the first time recognized as the type locality of Spiophanes cirrata.
Based on the results of the molecular analysis performed in the present study (see above Remarks for Spiophanes kroyeri), and according to the type localities of S. kroyeri and S. cirrata, we tentatively refer S. kroyeri-like worms from the Barents Sea, Norway, and the northern part of the North Sea to as S. cf. kroyeri, and those from the northern and central parts of the North Sea, and from the Bay of Biscay, Spain, to as S. cf. cirrata (Fig 4B). Final conclusions about the specific identity of these worms can only be inferred after molecular examination of worms from the corresponding type localities of these species.
Distribution. North East Atlantic (Fig 4B). Remarks. Spiophanes berkeleyorum was originally described from Vancouver Island, British Columbia, Canada, by Pettibone [232] and later reported from distant regions. Meißner & Hutchings [110] removed Spiophanes japonicum Imajima, 1991 from synonymy with S. berkeleyorum and Meißner [8] suggested that the latter species is distributed along the Pacific coast of North America only. Later reports of the species from Brazil [233,215], Russia [84], and China [95] probably represent misidentifications and should be verified in additional studies. Molecular data of S. berkeleyorum are limited to sequences provided by Meißner & Blank [111] and those in the present study.
Spiophanes berkeleyorum
Distribution. North East Pacific: from Alaska south to California. Remarks. Although Trochochaeta was not an object of the present study, use of these worms in the analysis (as an outgroup) brought some insight on their taxonomy. Trochochaeta multisetosa was originally described from the Øresund Strait, Sweden, by Ö rsted ( [234], as Disoma multisetosum) and later occasionally reported from the North East Atlantic. Hartman [241] described Trochochaeta franciscana (as Disoma franciscanum) from San Francisco Bay, California, USA, but Pettibone [239] placed it into synonymy with T. multisetosa. In disagreement with Pettibone [239], Blake & Arnofsky ( [242]: 68) stated that "the California specimens clearly represent a separate species due to differences in larval ciliary patterns." In a revision of Trochochaeta, Radashevsky et al. [243] treated T. multisetosa and T. franciscana as two valid species. Molecular analysis performed in the present study, however, showed identity (pdistance = 0.00%) of the fragments of nuclear 18S and 28S rDNA sequenced in worms from Norway and Sweden and from Drake's Bay, California, adjacent to San Francisco Bay. Consequently, in agreement with Pettibone [239], we tentatively suggest that T. franciscana is a junior synonym of T. multisetosa. As such, it may represent an introduced North Atlantic species in San Francisco Bay (where it was first collected in 1912, Hartman [241]) and nearby embayments, transported by vectors such as commercial oyster culture or shipping. This suggestion should be verified, however, by analyzing mitochondrial genes, such as COI and 28S rDNA, of individuals from the San Francisco Bay area, and by further analysis of additional populations, reported under the names of T. franciscana or T. multisetosa from a variety of habitats, from British Columbia to southern California, to further confirm if only one species of Trochochaeta is present.If confirmed, this may be a second case of the successful invasion of Trochochaeta into a remote region. Radashevsky et al. [243] recently suggested that Trochochaeta japonica Imajima, 1989 might have been introduced from the Asian Pacific to the estuary of Santos, São Paulo, Brazil, as larvae in ballast water of ocean-going vessels.
Discussion
The molecular analyses performed in the present study provided rather unexpected results. In various cases, morphologically similar or even identical specimens from close locations or even from the same area turned out to be genetically distant and are herein referred to as members of sibling species.
Five-genes analysis (COI, 16S, 18S, 28S, and Histone 3)
The five-genes analysis unambiguously suggested that horned Spiophanes with metameric nuchal organs and Spiophanes with bell-shaped prostomia and nuchal organs as entire long parallel ciliary bands belong to two major evolutionary lineages within the genus. More comprehensive analysis including more species is required to confirm this hypothesis. Including species with other types of nuchal organs, such as entire U-shaped ciliary bands (see Radashevsky [174] for review), in a future analysis, will help to better understand the evolutionary history of Spiophanes and the transformation of characters within this group of spionids. Remarkably, adults of Trochochaeta, the so far suggested sister group to Spiophanes, have Ushaped nuchal organs, similar to those of some Spiophanes and many other spionids. It will not be surprising, therefore, if Spiophanes with U-shaped nuchal organs are shown to have a basal position on the phylogenetic tree of the genus. This would suggest that the metameric and nuchal organs as long parallel ciliary bands might have evolved from nuchal organs as Ushaped ciliary bands within Spiophanes. This would also suggest that the metameric and nuchal organs as long parallel ciliary bands shared by other spionids, might be homoplasious characters evolved independently more than once within Spionidae.
The analysis suggested that horned Spiophanes from the North East Atlantic and North Pacific represent two sister evolutionary lineages. The absence of S. bombyx from the Pacific, as previously suggested by Meißner & Blank [111], has been confirmed. Moreover, the present study suggested an even more complicated systematic composition of horned Spiophanes in the North East Atlantic. In contrast to Meißner & Blank [111], who suggested that S. bombyx was distributed in North European waters and in the Mediterranean Sea, we suggest that probably three sibling species: S. bombyx, S. cf. bombyx and S. cf. convexus, have been observed, and that S. bombyx probably occurs in the Mediterranean Sea only. Unfortunately, S. convexus was poorly described based on a few incomplete fragments, without consideration of the morphological variability of S. bombyx and not examined by molecular means. Molecular analysis of horned European Spiophanes, especially S. bombyx from the Gulf of Naples, is needed no clarify the species composition of this group in the region.
The analysis also revealed a complicated systematic composition of worms fitting the morphological characteristics of S. kroyeri. The discovery of two genetically different species in North European waters requires further clarification of their specific identity. Herein, based on the proximity of sampling sites of specimens examined in molecular analyses and the type localities of S. kroyeri and S. cirrata, we suggest that worms from the Barents and the Norwegian seas, and also from the northern part of the North Sea belong to S. kroyeri, whereas worms from the North Sea and the Bay of Biscay belong to S. cirrata. These suggestions should be verified by molecular examination of Spiophanes kroyeri from the Greenland Sea and in a broader-scale study of Spiophanes in North European waters.
The analysis also showed a rich species composition of horned Spiophanes in the North Pacific and all along the Pacific coast of North and South America. First of all, it confirmed that Asian and North American horned Spiophanes, even appearing morphologically identical, are not conspecific. Moreover, it showed that horned Spiophanes occurring along the Pacific coast of North America belong to two sibling species: S. norrisi and S. hakaiensis n. sp. The presence of sibling species among Spiophanes brings doubts about the conspecificity of remote populations in North and South America referred by Meißner & Blank [111] to S. norrisi. It is plausible that South American populations belongs to an undescribed sibling species.
COI analysis, divergence time estimations and biogeographic inferences
The first Bayesian analysis of COI sequences, aiming to compare our data with those obtained in previous studies, resulted in a partially resolved consensus tree (Fig 1B), showing that short COI fragments (<300 bp) are not appropriate for these types of systematic inferences. The BEAST analysis of longer COI sequences (>500 bp) resulted in a fully resolved tree (Fig 3) with the same topology as in the five-genes Bayesian analysis (2B). A smaller set of taxa was used, however, in the BEAST analysis because it aimed to hypothesize on the divergence times of horned Spiophanes with metameric nuchal organs only.
Two hypotheses were addressed in advance to explain the high morphological similarity and disjunct distributions of horned Spiophanes on the Asian and American coasts of the North Pacific. Both hypotheses assumed that a common ancestor of those Spiophanes might have been distributed in the past all along the North Pacific without interruption. They differ however in the cause and the time of isolation of the two lineages, followed by subsequent speciation. The first hypothesis suggests that such isolation and speciation event might have resulted from the first opening of the Bering Strait at 5.5-5.4 mya [244,245]. The second hypothesis links this isolation and speciation to the maximum glacial period in the Quaternary ice ages in the North Pacific 3.0-2.4 mya [246][247][248][249][250]. The estimate suggested by the BEAST analysis, about 1.3 mya (95% HPD: 2.0-0.7 mya), places the divergence of the North American and Asian lineages of horned Spiophanes after the maximum glacial period in the North Pacific and, thus, supports the second hypothesis. The last glacial maximum was suggested to have had a strong influence on the distribution of many other species (see Hewitt [251] for review).
Both the five-genes Bayesian analysis and the BEAST analysis suggested that the divergence of the North American and Asian lineages was preceded by the isolation and speciation of S. norrisi from a common North Pacific ancestor. The BEAST analysis estimates that this divergence might have happened about 1.7 mya (95% HPD: 2.3-1.0 mya). Remarkably, the boundary between the northern S. hakaiensis n. sp. and southern S. norrisi roughly coincides with the boundary between the Oregonian and Californian coastal biogeographical provinces at Point Conception [252]. Some Californian province species extend their ranges northward in warmregime years [253][254][255]. This may also be the case with S. norrisi which occurs offshore from San Francisco, northern California.
The only other estimation of the divergence time of spionid polychaetes based on molecular data was provided by Schulze et al. [256] for North American populations of Streblospio spp. Using an invertebrate molecular clock calibration for COI sequence divergence of 1.8-2.8% per million years suggested by Palumbi [257], they estimated that the Streblospio lineage split in the Gulf of Mexico and adjacent regions 7-11 mya. The suggested estimates corresponded to the sea level minima in the examined region and this was considered as support for the obtained values of the estimates.
Conclusions
The results of molecular analyses performed in the present study unambiguously showed the presence of a series of sibling species within the genus Spiophanes. It means that specific identifications of this complicated group of spionids in many cases should be verified by molecular data. Molecular identities of Spiophanes, especially from European waters, are needed for confident identifications of these polychaetes around the world.
Supporting information S1 Table. List of samples reported in the present study. Sampling location data and museum registration numbers of newly collected material of Spiophanes spp.; museum samples examined during this study; material from the Norwegian, North, Mediterranean and Aegean seas and North Pacific reported by Meißner & Blank [111] and from the North Sea identified by Dieter Fiege, Markus Böggemann and Karin Meißner which is deposited in the SMF and other polychaete collections and was only partially re-examined by the first author (VIR); records of S. bombyx (= S. uschakowi) from Japan provided by Imajima [88,[116][117][118], for which no museum numbers were reported; records of S. bombyx (= S. cf. norrisi) from Chile, Argentina and the Falkland Islands provided by Carrasco [119] and Blake [48]; records of S. bombyx (here referred either to S. bombyx or S. cf. convexus) from Spain provided by Meißner [8]. Table. Genetic distances between Spiophanes species-COI. Uncorrected pairwise average distances (p, in %) between specific clades for COI (534 bp) sequences used in the fivegenes Bayesian analysis of Spiophanes spp. rooted with sequences of Trochochaeta multisetosa. The numbers of corresponding samples from Table 1 are given in parentheses after names and sampling locations of the species. Average distances between sympatric conspecific individuals in italics, maximal values of average distances between different Spiophanes species in bold, and minimal values of average distances between different Spiophanes species underlined. (XLSX) S4 Table. Genetic distances between Spiophanes species-16S. Uncorrected pairwise average distances (p, in %) between specific clades for 16S (244 bp) sequences used in the five-genes Bayesian analysis of Spiophanes spp. rooted with sequences of Trochochaeta multisetosa. The numbers of corresponding samples from Table 1 are given in parentheses after names and sampling locations of the species. Average distances between sympatric conspecific individuals in italics, maximal values of average distances between different Spiophanes species in bold, and minimal values of average distances between different Spiophanes species underlined. (XLSX) S5 Table. Genetic distances between Spiophanes species-18S. Uncorrected pairwise average distances (p, in %) between specific clades for 18S (1656 bp) sequences used in the five-genes Bayesian analysis of Spiophanes spp. rooted with sequences of Trochochaeta multisetosa. The numbers of corresponding samples from Table 1 are given in parentheses after names and sampling locations of the species. Average distances between sympatric conspecific individuals in italics, maximal values of average distances between different Spiophanes species in bold, and minimal values of average distances between different Spiophanes species underlined. (XLSX) S6 Table. Genetic distances between Spiophanes species-28S. Uncorrected pairwise average distances (p, in %) between specific clades for 28S (287 bp) sequences used in the five-genes Bayesian analysis of Spiophanes spp. rooted with sequences of Trochochaeta multisetosa. The numbers of corresponding samples from Table 1 are given in parentheses after names and sampling locations of the species. Average distances between sympatric conspecific individuals in italics, maximal values of average distances between different Spiophanes species in bold, and minimal values of average distances between different Spiophanes species underlined. (XLSX) S7 Table. Genetic distances between Spiophanes species-Histone 3. Uncorrected pairwise average distances (p, in %) between specific clades for Histone 3 (297 bp) sequences used in the five-genes Bayesian analysis of Spiophanes spp. rooted with sequences of Trochochaeta multisetosa. The numbers of corresponding samples from Table 1 | 2020-07-02T10:29:21.963Z | 2020-07-01T00:00:00.000 | {
"year": 2020,
"sha1": "90de3ca3a53f2127f516b580a1d4ea0b77ff99d2",
"oa_license": "CCBY",
"oa_url": "https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0234238&type=printable",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "f2b092dfe227a48a7fca4c9a4f74acf2dd9915de",
"s2fieldsofstudy": [
"Biology",
"Environmental Science"
],
"extfieldsofstudy": [
"Medicine",
"Biology"
]
} |
79516738 | pes2o/s2orc | v3-fos-license | A novel mutation c.2010delG of CLCN5 gene associated with Dent disease-1 in an 11-year-old male with nephrolithiasis and nephrocalcinosis
RETRACTED Dent’s disease-1 (CLCN5 gene) is a rare X-linked recessive tubulopathy characterized by low molecular weight proteinuria, hypercalciuria, nephrolcalcinosis or nephrolithiasis, proximal tubular dysfunction and renal failure in adulthood. Females are carriers and usually mildly affected. We present an 11-year-old child with nephrocalcinosis and nephrolithiasis with c.2010delG (or p.Asp671fs) mutation in CLCN5 gene which had not previously been reported in the Dent’s disease-1.
INTRODUCTION
N ephrocalcinosis is characterized by the deposition of calcium in the kidney paranchyma and tubules. It is associated with conditions that cause hypercalcemia, hyperphosphatemia, and the increased excretion of calcium, phosphate, and/or oxalate in the urine. According to the laboratory results, three groups can be formed in patients with nephrocalcinosis to make a differential diag-nosis; hypercalciuria with hypercalcemia, hypercalciuria without hypercalcemia and hyperphosphaturia [1].
Dent's disease-1 is a rare cause of hypercalciuria without hypercalcemia. It is characterized by low molecularweight proteinuria, hypercalciuria, nephrocalcinosis, nephrolithiasis, and chronic renal failure. In about 60% of patients with X-linked nephrolithiasis, a mutation in the CLCN5 gene is detected (Dent disease type 1), whereas in 15%, the disease is due to a mutation in the OCRL gene (Dent disease type 2) [2][3][4]. Due to be X-linked, males are affected more severely, but females are carriers and usually only mildly affected in both forms of Dent's disease [5].
CASE REPORT
An 11-year old boy was referred for nephrolithiasis and nephrocalcinosis. He denied any renal disease, trauma, diarrhea or constipation at past medical history. The grandmother had renal failure of unknown origine at medical family history. There was no pathological finding on the physical examination, including growth parameters and blood pressure according to age group. Laboratory findings were normal except hypercalciuria (9 mg/kg/d), 88% tubular phosphore reabsorbtion rate, low molecular weight proteinuria (B 2 microglobulin 5080 mcg/L, n <250 mcg/L) in 24 hour urine, bilateral 4 mm non obstructive nephrolithiasis and nephrocalcinosis at ultrasonography. Urinary oxalate, citrate, uric acid, serum bicarbonate, vitamin-D and parathormone values were also normal. Genetic analysis revealed mutation c.2010delG (p.Asp671fs) in the CLCN5 gene that was detected in frame shift and identified as a stop codon (Figure). The patient was given potassium citrate and thiazide for persistant hypercalciuria.
DISCUSSION
The clinical diagnosis of Dent's disease is based on the presence of LMW proteinuria, (elevation of B 2 -microglobulin, clara cell protein RBP -retinol-binding protein) and/or about five fold above the upper limit which is pathognomonic for Dent's disease), hypercalciuria (>4 mg/kg per days characteristic for Dent's disease) and diagnosis should include at least one of the presence: nephrocalcinosis, nephrolithiasis, hematuria, hypophosphatemia or chronic renal disease [6,7]. Our patient fulfilled the criteria of the group of hypercalciuria without hypercalcemia rather than the groups of hypercalciuria with hypercalcemia and hyperphosphaturia among the three groups of nephrocalcinosis. Distal renal tubuler acidosis, medullar sponge kidney, neonatal nephrocalcinosis, loop diuretics, inherited tubulopathies, chronic hypokalemia and beta thalassemia are the underlying diseases in association with the group of hypercalciuria without hypercalcemia at nephrocalcinosis [1]. Normal blood pH, low molecular weight proteinuria, hypercalciuria and lower tubular phosphore reabsorbtion rate (TPR) without hypercalcemia pointed out inherited tubulopathy and Dent's disease in our patient. Genetical analysis also confirmed the diagnosis with the mutation in CLCN5 gene.
The exact prevalence of Dent's disease is undefined; to date, >250 families have been described [8]. Hypercalciuria and nephrocalcinosis are prevalent at a rate of 95% and 75% in affected males, respectively. Progression to end-stage renal failure are at the 3 rd -5th decades of life in 30-80% of affected males [9]. In the absence of therapy targeting for the molecular defect, the current care of patients with Dent's disease is supportive, focusing on the prevention of nephrolithiasis and nephrocalcinosis. Thiazide diuretics can be used to treat hypercalciuria [10,11].
In summary, a new mutation c.2010delG (p.Asp671fs) in the CLCN5 gene that was first detected in frame shift and identified as a stop codon in our patient. Dent's disease should be kept in mind in nephrocalcinosis with hypercalciuria, low molecular weight proteinuria and normal blood pH at male patients. | 2019-03-17T13:12:31.749Z | 2018-04-27T00:00:00.000 | {
"year": 2018,
"sha1": "1ff6b3337dac389853dafe7a5f93dca0f11258d0",
"oa_license": "CCBY",
"oa_url": "https://www.ped-perinatology.ru/jour/article/download/648/622",
"oa_status": "GOLD",
"pdf_src": "Adhoc",
"pdf_hash": "09ca68427b8794fe13712e011c05235f4aac6412",
"s2fieldsofstudy": [
"Medicine",
"Biology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
263909713 | pes2o/s2orc | v3-fos-license | The power of the ‘universal’: caste and missionary medical discourses of alcoholism in the Telugu print sphere, 1900–1940
This article explores missionary medical discourses in three Telugu journals published in the early twentieth century, to analyse how caste pivoted denunciations of alcohol, especially toddy and arrack, in the Madras Presidency and the Hyderabad state. It argues that one women’s missionary journal, Vivekavathi, deployed medical knowledge to formulate subtle and occasionally explicit condemnations of toddy and arrack as unclean and unhealthy substances. The journal relied on universal medical and missionary, British and American knowledge frameworks to mark out Dalits and other marginalised castes as consumers of these local beverages. This stigma was conjured through medical narratives of marginalised castes as lacking in the knowledge of alcohol’s relation to digestion, toddy’s role in ruining maternal and child nutrition, the unhygienic environment of arrack shops and their propensity to ‘alcoholism’. However, this article also traces counter-caste voices who too invoked ‘the power of the universal’ to dispel caste stigma against marginalised castes. While both sets of voices deployed medical ‘enslavement’ to alcohol as an interpretive move, they differed in their social imperatives and political imaginaries, defined in caste terms. This article explores a third set of implications of the term ‘universal’ by analysing global medico-missionary narratives of alcohol in two other Telugu journals. On a methodological plane, this article also pushes for a hybrid reading of what counts for ‘scientific instruction’, where hymns, catechisms, parables and allegories are considered alongside conventional scientific experiments. In that sense, it upholds vernacular missionary publications as an invaluable resource for the social history of medicine.
They have also found out that alcohol intake disrupts the process of digestion.Dr Monroe (sic) from Hull has carried out an experiment.He diced the beef into small pieces, extracted its gastric juices, and divided two portions equally into three parts.He then added water into one container, toddy into another, and arrack into yet another container.He created a natural stomach situation for those containers at 100°and began to observe the changes they underwent (emphasis mine). 1 Toddy enslaves you to debt It is the origin of dangers It is the biggest cause of mistakes It is the instrument of pain On account of drinking toddy, Your hands and feet give up on you Your throat and passages Are afflicted by a stench When you drink arrack You are causing diseases When you smoke cigarettes Your chest and heart burn (emphasis mine). 2 They taste the local and foreign alcohols first, feel a growing attachment the second time, and become slaves to alcohol the third time.Compared to the other intoxicants we have discussed above, alcohol is much more dangerous.It disrupts the flow of the blood, impacts liver function, burns the heart, and kills you a little day by day.It decreases your life expectancy.People die suddenly because of their drinking habits (emphasis mine). 3is article delves into the politics of missionary medical temperance in the Telugu print world of the early twentieth century Madras Presidency, especially between 1900 and 1940.In so doing, it is invested in exploring how medical temperance writings in Telugu missionary publicationsthat touch on the toxicity of alcohol and its harm to human digestion, diet and nutritiondiscern the subject of caste without always naming it.The invocations of caste feature in a temperance world brimming with overlapping medical, social and cultural descriptions of toddy and arrack.Toddy refers to a fermented beverage common to South India and, depending on the region, is tapped from the palmyra, coconut, date or sago palm tree.Arrack is typically spirit distilled from toddy or grain, 'cane sugar, molasses or jaggari'. 4nstead of situating alcohol in colonial South India within nationalist and subaltern frameworks of anti-drink movements, this article locates it within two sites, the social history of medicine and a rare historical genre, namely, the vernacular missionary journal. 5While caste has pivoted the critique of nationalist and missionary discourses of alcohol, it has not featured in a systematic analysis of medical histories of alcohol in colonial India.
The three early-twentieth-century Telugu missionary journals discussed in this article are Vivekavathi, The Telugu Baptist and The United Church Herald (UCH) (Figures 1-3).While none of them was a health journal, they nonetheless provided health-driven commentaries on alcohol.Vivekavathi authors especially presented excerpts, summaries and interpretations of experiments and medical findings by scientists, chemists and physicians writing on alcohol in the UK and America.This article argues that what was presented as medical and scientific knowledge carried often subtle and occasionally explicit condemnations of toddy and arrack as unclean and unhealthy substances.Vivekavathi associated these beverages with Dalit and non-Dalit marginalised communities.This article simultaneously argues that the same medical debates and discussions in Vivekavathi produced subtle counter-caste voices that sought to banish this stigma from the lives of marginalised castes.
Methodologically, this article first brings the social history of medicine in conversation with Dalit Studies and Alcohol Studies.Secondly, it is attentive to the hybrid form of these journals, which packaged secular Enlightenment discourses in a web of Biblical language and medical truth claims. 6his hybrid form meant that medical temperance instruction was embedded in catechisms, hymns and allegories but at other times present in summaries of experiments.Finally, this article disaggregates terms like 'universal', 'enslavement', 'addiction' and 'medical temperance'.The 'universal' is explored in medical Enlightenment and Dalit terms, in Biblical narratives of the clean and abstinent body, scientific assertions about the physiology of alcohol and its impact on digestion, liver and nervous system.The power of 'the universal' is disparately at work in both hegemonic and counter-caste voices.It also features in globalizing narrativesespecially in the journals UCH and The Telugu Baptistwhere alcoholism was shorn of caste associations and compared globally with drug and opium addiction.'Enslavement' figures as a trope to free the body of its physiological addiction to toddy and arrack, liberate the Christian subject and free the drinker from fetters of caste stigma.In presenting 'alcoholism' as enslavement, the Telugu print sphere deployed a suite of overlapping terms and placed them in interwoven worlds of meaning.
Situating Vivekavathi within the vernacular missionary world
The arrival of Anglican and Protestant missionaries across the nineteenth century meant that evangelical Christianity's pathways into India engendered complex entanglements with empire.In that sense, it has been argued that there was no single 'language' or 'framework' that the Bible provided for 'both interpreting and challenging imperial modernity'. 7The vernacular print sphere stood testament to this in terms of its hybrid genre that interwove medical and spiritual insights.Temperance expositions in these journals interspersed scriptural commentaries of the body as 'the temple of God' with scientific debates on whether alcohol was a stimulant or depressant, food or poison.Tamil, Telugu and Hindu missionary publications asserted that alcohol was a demon drink but also held toddy and beer to be lacking in nutritive content. 8They insisted on alcohol's adverse impact on the physiology of the body, lactating mothers and child nutrition; its toxic chemistry; the unhygienic environments of toddy shops and the relationship between alcohol and heredity.
A notable illustration is a Tamil missionary publication termed Kutumba Matuvilakku Castiram [Domestic Temperance Science] authored by Y.G.Bonnell, meant for parents and teachers.Coming from a community of Nadars who were historically stigmatised for being toddy tappers, Bonnell strove subtly to fight against this taint of backwardness.9As a teacher affiliated to the United Free Church of Scotland Mission and later headmaster of many schools, he was in a unique place to educate peers and students. 10This work executed a detailed foray into how alcohol changes brain chemistry, affects the processes through which blood clots occur, alters nerve function, obstructs gastric juice secretion and affects the kidney's capacity to eject waste.Considered compositely, he wrote that alcohol wears down the immunity of the body to disease. 11n dispelling 'myths' that alcohol is nutritive food and in explaining why it is addictive, Bonnell resorted to temperance songs and parables that specifically castigate kallu or toddy and sarayam or arrack as dangerous.One such parable is the story of the swimmer who mistook a white bear in the middle of a river in spate to be a bundle of money.He jumps in to secure the bundle only to realise that it was a drowning creature desperate to cling to him to save its own life.He explained that the bear was alcohol that the alcoholic could not shake off and he warns the reader darkly against 'brandy as the red bear', 'arrack as a black bear' and 'toddy as a white bear'. 12his co-production of medical and spiritual narrative occurred in both colonial Telugu and English missionary print.In English, the prolific volumes of Abkari (the mouthpiece of the Anglo-Indian Temperance Association), the Indian Temperance Record or the Indian publication of the Women's Christian Temperance Union (WCTU) and Christian Literature Society (CLS) publications carried the voices of physicians alongside those of the medically informed missionary. 13However, the use of Christian narrative devices to impart medical pedagogy was more common to the Telugu print sphere.
Vivekavathi was an inter-denominational CLS monthly women's periodical, which ran from 1909 to 1934.During its entire run, it endorsed the civilising quality of the British empire and American and British missions.It prided itself in carrying what the historian Mahaboob Basha termed the 'white women's burden of civilising the native woman' 14 and endorsing reforms like widow remarriage and the abolition of child marriage.The editorial committee featured mostly white women and a couple of native Christians at any given time.As historian, Chakali Chandra Sekhar, who too worked on Vivekavathi, pointed out, white women missionaries from Protestant missions exercised tremendous agency in shaping this journal. 15That said, the contributors were significantly though not exclusively Indian Christian converts and many of them women.In choosing a term like Vivekavathi or 'the wise woman', the editors positioned themselves as a secular alternative to other Christian publications, such as UCH and The Telugu Baptist.
Basha wrote that the monthly periodical had thirty subscribers to begin with, a number that went up to 1 400 by October 1913. 16It had thirty-two pages in the first issue.Its annual subscription rate was twelve annas (or 1/16 of a rupee) by the end of the first year of October 1910.To cite one of Vivekavathi's editors, Elizabeth McCauley, subscribers and readers included converted Christians, merchants, women teachers, Brahmin widows and 'progressive Hindus'. 17McCauley implies that Dalit readers of Vivekavathi were a large subset of converted Christians.She infers Telugu Dalit readership of the journal from the high literacy rate of converted Christians from the Madras Presidency region in the 1911 census, on the one hand, and their contribution to the journal, on the other.She adds in no unmistakable terms that converted Christians, for the most part, were 'formerly the despised and oppressed outcastes'. 18iven that this was a women's journal, some articles were targeted at them, as for instance, the offerings of domestic temperance science in the column termed Sundara Gruhamulu or 'The House Beautiful.'Many others were more broadly categorised and heavily medicalised, these series included Arogya Vishayamulu, or 'Health Matters', Madyanishedhamu or 'Prohibition'/'Temperance' and Patrika Sampadakurali Vyasamulu/Sampadakulu or 'editorials'.These articles were particularly invested in defining the characteristics of the alcoholic.
Between 1909 and 1926, roughly sixty articles on subjects such as alcohol, inebriation, intemperance and domestic temperance science were featured in Vivekavathi.Articles on temperance added up to a total of five in the inaugural year of the journal, i.e. 1909, and were most plentiful between 1915 and 1920, totalling up to thirty-five.The monthly issue of October 1919 alone saw seven articles published on the subject.While it is almost impossible to state exactly why this period saw a jump in the number of articles on temperance, it could have converged with the lead up to the anti-liquor agitation in the Madras Presidency.While the latter picked up momentum in mid-1921, heated discussion of the rise in liquor prices and increased alcoholic consumption among labourers took place well before this, in the year 1917. 19However, Vivekavathi did not present temperance debates in a nationalist vein unlike a popular, contemporary Telugu newspaper, Andhra Patrika, choosing rather to approach them as weighty scientific and spiritual matters.
Even though it was not a mouthpiece of WCTU, Vivekavathi embraced the 'internationalism', 'American values' and the 'Anglo-American roots' of this organization. 20The journal carried several articles by British and American medical missionaries and WCTU figures like Maud Allen and Margaret Denning.It used the genre of the temperance hymn, very common to WCTUs and the Band of Hope's temperance campaigns. 21Authors also cited renowned British and American medical figures like Victor Horsley, Mary Sturge, Leonard Rogers, Frederick Treves and Henry Beaumont.Their medical findings about the harmful chemistry of alcohol and its role in altering the physiology of the human body were presented as universal truth claims.Tarangini Sriraman These medical forays interspersed with international examples of scientific experiments were in tune with the journal's intention to 'open up perspectives of things happening around the world, with the blessings of God and how they came into being'. 22While Vivekavathi served universal narratives on a Biblical platter, it did so within a deep-seated regional framework.The cultural modernity of temperance is to be situated not so much in Vivekavathi's turn to Western medicine and science as much as in its (mis)translations, interpretations and regional formulations.Medico-scientific theories became a vehicle for expressing hegemonic and stigmatising caste-mediated propaganda that associated toddy and arrack with marginalised castes, and Dalits in particular.Counter-caste narratives saw the same British and American medical theories as facilitating something radically different, a temperance campaign for Dalit and marginalised caste self-respect.The feminist historian, Shailaja Paik refers to the 'microtransformation' of 'upper caste' values for a radical politics of anti-caste self-assertion. 23This article builds on Paik's arguments to demonstrate that universal medical theories, espoused by upper castes, were also the site of counter-caste self-making.
Rather than convey 'enslavement' to drink in purely medical terms or overarching terms as 'a disease of the will', 24 hegemonic counter-caste voices mingled the social, spiritual and medical.The enslavement of the alcoholic was explained as being oblivious to the medical hazards of drinking for children and women, the toxicity of brandy and some peculiarly minute descriptions of toddy and arrack.These drinks were described as contributing to secondary medical conditions like ajeernamu [indigestion], kshayarogamu [tuberculosis], vaatarogamu [gout], moorkharogamu [epilepsy] and rompa [cold and chills].Counter-caste voices called for respecting one's body as the temple of God, upholding the Nazarite vow while making a case for the dismantlement of especially toddy and arrack's strangle-hold over oppressed lives in terms that were medical, social and Biblical. 25he last section of this article briefly examines two other Telugu missionary journals, Kraistava Samyukta Sangha Vartamani or UCH and The Telugu Baptist.Not much is known about these other two journals except that UCH was published in two separate editions (Telugu and Telugu-English).Its publisher was not CLS Press but Minerva Press.26 UCH occasionally re-published articles from Vivekavathi, and both journals emulated Vivekavathi's style of using bilingual titles.While they too deployed medical narrative to denounce alcohol, they were not inclined to comment on local power structures and the caste dynamics of consuming toddy and arrack.They thus held out less potential for counter-caste narratives.
Toddy and arrack in English and Telugu colonial print culture
The stigma underlying Vivekavathi's subtle and not so subtle aversions to toddy and arrack must be read within a longer socio-medical history of caste in the vernacular and English print sphere.Far from being innocuous, these graphic associations were laden with the historicity of caste churnings around alcohol in both the Tamil-and Telugu-speaking parts of the Madras Presidency and the Hyderabad state during the early twentieth century.
C. Rajagopalachari or Rajaji, as he was popularly known, was at the helm of the Tamil Nadu Congress Committee-driven nationalist agitation against alcohol in Madras Presidency.He sought to mobilise public opinion through his unremitting campaign against toddy renters and arrack 'vends' (or vending spaces) for the nationalist daily, The Hindu.He assiduously published excerpts of his conversations with 22 'Sampadakulu', Vivekavathi, 1, 1 (1909), 1-7. 23Paik, op.cit.(note 6), 72. 24Mariana Valverde explains that physiologists and psychiatrists in the early nineteenth century construed alcoholism as a disease that incapacitated the will.She argues that this was a poorly medicalised argument.Mariana scientists researching nutrition at the Pasteur Institute, Coonnoor.He reported that toddy was less nutritious than buttermilk because it did not contain as much protein and its vitamin content was nothing out of the ordinary. 27He also reinforced caste stigma when he raged at Hindu temples for leasing out coconut trees for toddy-tapping.He wrote about one such temple, the Triplicane temple in Madras city: 'This ancient temple was intended for a civilised form of worship; but you have converted this into something like those temples where goats and fowls are slaughtered'. 28etailed ethnographies of 'castes and tribes of South India' figured in the deeply Orientalised and racialised writings of anthropologists like Edgar Thurston.Like Rajaji, Thurston clubbed together familiarity with toddy and animal sacrifice.In addition, he coupled drinking toddy and arrack with demon worship and flesh-eating practices among castes and tribes whom he regarded as backward. 29issionary publications in English produced a rich corpus of richly medicalised writing condemning toddy and arrack as particular to backward castes. 30ut these associations were also common in the Telugu public sphere across nationalist, imperial and niche health journals.A Telugu weekly titled Andhra Harijan made associations between toddy and arrack and marginalised castes in explicit terms when authors referred not only to toddy trees as intrinsically unhealthy but also their byproducts as also evil. 31While this establishes a general pattern to make caste-laden associations between toddy and arrack, it is possible to go one step further and discuss the more overarching trend to anchor health norms to caste within Telugu journals and newspapers.A monthly Telugu health journal called Arogya Prakashika, which was, inter alia, invested in propounding medical advice surrounding plague and tuberculosis, featured several articles on alcohol.These articles were peppered with ample references to how alcohol ravages the nervous and digestive system, solidifies food instead of breaking it down, increases cholesterol in the body, hastens the heartbeat and lowers immunity to plague.Toddy and arrack featured in temperance narratives that equated them with castes that were regarded backward.An article gestures to the dangerous habits of toddy and arrack consumption among rickshaw pullers and midwives before they go to work. 32Another article discusses how arrack destroys the nervous system and causes mental disorder while simultaneously citing Hindu texts that warn that it is 'more sinful to consume alcohol than kill a cow'. 33This journal, much like Vivekavathi, featured overlapping scriptural and medical narratives, caste narratives and citations of British and American authors.But the scriptural narratives of Arogya Prakashika were drawn from Hindu texts rather than the Biblical cannon.
Reading caste in Vivekavathi's medical temperance campaigns
The medical associations between toddy, arrack and caste are stark in Arogya Prakashika but they are mostly veiled in Vivekavathi.There is, however, a more distinctive imperial function that such articles in Vivekavathi carry out.While Vivekavathi, as Basha points out, sought to raise the health standards of natives, especially women, they also executed a subtle caste-driven mission to educate marginalised castes and tribes perceived to be lacking in basic medical precepts. 34his becomes manifest in an editorial that describes an initiative to secure temperance pledgestermed vagdattamu in Telugufrom the parents of children in 'Mala villages', who were seen to be shockingly familiar with toddy and unfamiliar with nutrition. 35The article goes on to say, We have been working with schools for some time now and in this connection, we visited a big Mala village.When we asked the question, who among you drinks kallu, forty students including the very small children raised their hands. 36e same editorial cautions readers in a dual sense against parents offering drinks to their children and the consequences of early alcoholism among adults before and after marriage for children.It cites medical data about child and adult mortality both in the Rayalaseema region in the Madras Presidency and England.The tendency to cite drinking statistics with relation to life expectancy, maternal health and hereditary disease was very common in British and American medical and missionary temperance publications. 37The Vivekavathi editorial claimed that children of drinking families in England were likely to develop conditions of muteness, epilepsy and tuberculosis.In Rayalaseema, a southern region of the Hyderabad state, parents with no pre-existing drinking habits lose thirteen children out of one hundred on an average every year, while parents with mild drinking habits lose twenty-three out of one hundred and those with extreme drinking habits lose thirty-two out of one hundred . 38he empirical descriptions of the drunkard in the Telugu context as 'Mala' and drunkards as 'Mala villages' stand out in this editorial.The warning about the potential loss of nutrients in breastmilk among those familiar with kallu and sara [arrack] is coated with the caste conviction that it is Mala mothers who are particularly culpable of alcoholism.The editorial subtly warns readers, in pedagogical terms, against emulating them.
Another article distinguishes drinking brandy and whisky in clubs, parties and lavish dinners with kallu and sara consumed by 'coolies', 'fishermen', 'cotton fields' and 'factory workers' as well as those working for the Dora.This is a caste-laden Telugu term translated as powerful landlord or lordover the land, the fields or a part of the villagewhere Dalits and marginalised castes had to render slave labour and other services in a system of caste-based slavery called vettichakiri. 39In vettichakiri, Dalits had to render caste-based (slave) farm labour while other marginalised castes had to provide free caste services in their capacity as barbers, washermen, toddy tappers and potters.Those marginalised castes who tapped toddy had to set aside trees exclusively meant for landlords and their families.
Writing in a deeply casteist register, the Vivekavathi author decries these elite and by extension upper caste drinkers for succumbing to peer pressure and marginalised castes for drinking out of bodily exhaustion and ignorance of hygiene and health norms.This is manifest in descriptions of how toddy and arrack are usually paired with manchi vasana gottu padu kooralu or 'curries reeking of rotting vegetables' served near the liquor shop. 40he same article argues that alcohol physiologically enslaves: the alcohol-habituated body is one in which disease spreads rapidly and effortlessly.When a fatigued body encounters alcohol, it causes the activity of the naramulu or the nervous system to slow down, with the result that the person feels pleasure and believes that alcohol is good for him/her.This slowed-down physiological activity caused especially cotton field workers, factory workers and coolies to become drunkards and worship Saramma and Kalludevimade-up disparaging names for 'lower caste' goddesses dedicated to toddy and arrack and behave like foxes, tigers and pigs. 41These deeply stigmatising caste associations were packaged in universal observations pertaining to hygiene, the body's physiology and the altered chemistry of the mind.What is however unmistakable is the coupling of classesthat are usually from marginalised castessuch as coolies, factory and field workers with bodies that were slowed down and enslaved by toddy and arrack.
Vivekavathi also made lengthy forays into what they presented as cutting-edge medical science, replete with irrefutable findings on the physiology and physiological effects of alcohol.Several articles delve into the scientific virtues of temperance on the strength of experiments while foregrounding the vile character of toddy and arrack.For example, an unnamed author alludes to experiments conducted by a 'Dr Monroe (sic)' from Hull.This experiment focuses on what occurs to beef, when it is cut into portions and when it is immersed in water, toddy and arrack, each of which was mixed with gastric juice in separate containers and maintained at 100°, to simulate a natural stomach situation.The Telugu article summarises the findings thus: when immersed in water, the digestion of the beef was well under way in four hours, that it broke down into chunks after eight hours and that the beef dissolved into the gastric juices after ten hours.His findings with toddy showed that the beef turned black and gave rise to black foam after four hours, that it broke down after eight hours and that it did not dissolve even after ten hours.With arrack, his findings were that the beef 'remained the same' after four hours and eight hours and that it became gatti or 'rigid' after ten hours. 42here were some key differences between this Telugu version and the original Dr Henry Munroe text, titled 'The Physiological Action of Alcohol', and another archival text by an American physician, Dr William Hargreaves.These were both temperance texts that were not influential in the British and American medical and missionary landscape.This could be because these two temperance figures summarised and consolidated important experiments in chemistry and physiology but did not themselves pioneer pathbreaking research.These texts were, however, regarded as good reviews of temperance literature along the lines of the chemical composition of alcohol, alcohol as a narcotic, its impact on digestion, its role in diseases affecting the different organs and diseases affecting the mind and hereditary disease.
The Hargreaves version contained a detailed exposition of the Vivekavathi experiment with a table very closely resembling that used in the Telugu version. 43Far from experimenting on kallu and sara as described in the Telugu version, they featured experiments with the more generic 'alcohol' and 'pale ale'. 44In the Munroe version, the experiment involved a mixture of bread, meat and gastric juice, to which 'a glass of pale ale or a quantity of alcohol' was added. 45He instructs the pouring of this mixture into a phial, which could be immersed in a sand-bath maintained at a low heat of 98º.He recommends the occasional brisk shaking of the contents in tune with the stomach's movements.He states that 'at the end of seven or eight hours, or even some days, the food is scarcely acted upon at all'. 46n Hargreaves' text, Dr Munroe mixed finely minced meat with gastric juice from the stomach of a calf with water, alcohol and pale ale and kept the bottles at a temperature of 100°.His observations of what happened to the beef are similar in both the Hargreaves and the Telugu Vivekavathi texts with varying details of the different hours of observation.The descriptions of Dr Munroe's inferencesnamely that alcoholics destroy the solvent power of gastric juice and prevent digestionare similar in all three versions.
Assuming that the Vivekavathi author used the Hargreaves versionwhich was almost certainly the case given the similarity of tables (Figure 4)the substituting of alcohol and pale ale with toddy and arrack is unmistakable.One can argue that the Vivekavathi author used toddy and arrack to render the article more context-friendly and took creative liberties; Munroe did say in his original text that any alcohol can be used.However, there can be no escaping the stark motif of toddy and arrack as particularly needing to be shunned across Vivekavathi's socio-medical landscape.
In replacing pale ale and alcohol with toddy and arrack, Vivekavathi may have conflated Munroe's findings about pale ale set in a British context with the common views of physicians and tropical disease epidemiologists like Norman Chevers and Leonard Rogers.Historian Erica Wald writes that Chevers excoriated toddy and arrack in the colonial context as a potent 'cocktail of drugs and hallucinogens'. 47These drinks were seen to be all the more dubious owing to the local toxins they contained.Where claret and beer were seen to be befitting of higher European classes in a tropical environment, toddy and arrack were demonstrated to be routinely consumed by 'lower-orders' among British soldiers, sailors and Indians and that too in deeply unhygienic surroundings marked by open sewers. 48nother Vivekavathi author pays homage to two famous medical temperance figures, Victor Horsley and Mary Sturge, by deploying the same title of the text they authored, 'Alcohol and the Human Body'.This article cited the statistical medical figures in the original work, that in seven major hospitals of London, the expenses on alcohol were 12 000 pounds or 120 000 rupees in 1852 and those on milk were 3 000 pounds or 45 000 rupees.Over the next few years, London's hospitals spent less than 3 000 pounds on alcohol while increasing their milk expenditure to 8 000 pounds. 49The author also asked the same question as did Horsley and Sturge, i.e. 'what is alcohol?'The answers were exactly similar as well, with both texts explaining fermentation to be a process where the yeast plant acts on sugars such that it splits them into alcohol and carbon dioxide or carbonic acid gas.However, the Telugu author goes from here to illustrate fermentation saliently through different types of toddy sourced from the palmyra, date or coconut trees. 50This stood in stark contrast to Horsley and Sturge, who considered the medical impact of a panoply of drinks ranging from beers and wines to distilled liquors such as whisky, gin and schnapps on digestive processes. 51The caste-laden condemnations of toddy and arrack in Vivekavathi were thus marked in a universal language of medical discovery.
Counter-caste voices: The power of the 'universal' No narrative other than a strictly temperance-driven one was possible considering Vivekavathi's editorial stance.Given, however, that it was a field of universal claims and assertions, it was one that enabled counter-caste voices to express themselves.This article's usage of 'counter-caste' owes much to literary historian and Dalit Studies scholar, Satyanarayana's deployment of 'counter-publics' or oppositional, anti-caste voices who broke 'invisible' barriers of articulating their identities in secular and liberal spaces. 52These voices sought not so much to preserve caste stigmas against alcohol but to ensure that those from marginalised backgrounds are freed from them.In so doing, they too relied on universal Biblical and medical narratives.
The term 'universal' has a place both within Dalit Studies and critiques of 'Medical Enlightenment'.Scholarship around the latter demonstrates that the universal empiricist, rationalist and naturalist interventions of Enlightenment were to the effect of presenting medical knowledge as 'certitudes' and proven truths rather than mere 'conjecture'. 53But as Maurice Finochiaro argues, the 'empirical' was a tool in the arsenal both of those like Galileo and Copernicus and those who opposed him. 54In the early twentieth centurythe same timeframe of this article -German brewers and producers executed their own stakes in the 'Medical Enlightenment'.Michael Hau writes that they mobilized medical arguments that alcohol was harmless and that it boosted immunity to tuberculosis, nervous disease and related morbidities. 55edical Enlightenment scholarship however did not bring the universal in conversation with inequality whereas Dalit Studies addressed genealogies of the universal in pursuit of self-empowerment.
Historian and Dalit Studies scholar Sanal Mohan writes that CMS missionaries synthesised Enlightenment ideals with Biblical concepts of salvation in addressing caste-based slavery in Kerala.They deployed terms like 'rights-bearing individual' and 'suffering slave' while failing to discern slavery within the world of agrarian caste-based exploitation. 56Converted Dalit slaves, however, were able to appropriate 'universals' and embrace missionary pedagogic habits of 'new sartorial and bodily practices, acquisition of language, literacy and education' to regain their humanity. 57 historian and Dalit Studies scholar working on the Telugu print spheres, Chinnaiah Jangam, brings a different critique by arguing against the singular acceptance of the premises of contempt for 'universal principles of Enlightenment' underlying Postcolonial Studies. 58If this discipline rejected such principles owing to their Eurocentric heritage, Jangam shows that Dalits experienced the colonial state as well as missionary presence in Telugu public sphere differently.Given the state of agrarian slavery and violent segregation that Dalits lived in, it was the sustained administrative efforts of the colonial state and missionaries that yielded results of accessible public education for Adi-Andhras.He concludes that liberalism, when channelled to fight 'caste-based violation', created avenues of 'ideological power' to Dalits. 59ounter-caste voices in Vivekavathi were able to appropriate universals, such as 'alcohol as a narcotic', alcohol as 'causing bodily heat' and 'destroying the digestive tract', packaged in parables and catechisms.These universals were very enabling in calling out the enslaving stigmas that toddy and arrack subjected marginalised communities to.But they also turned WCTU songs into temperance songs that attacked the very local manifestations of debt, landlordism, caste stigma, inequality and loss of selfrespect.
Vivekavathi authors like Hariyatamma, Nagulaiah, M. Augustine Narasimhulaiah, Pendurti Joseph and O. David were driven by Dalit sensibilities in that they spoke out against Brahminical ritualism, dehumanising upper caste scriptures and caste hierarchies on spiritual grounds.For example, an author presented a dialogue between 'a Dhoby' (regarded the 'washerman' and 'low caste') and 'a Brahmin', which entails the former challenging the latter's backward beliefs about a sacred river being capable of wiping all Sin. 60Counter-caste voices also marked out the alcoholic's 'enslavement' to drink and local caste hierarchies of toddy and arrack even as they strove to disseminate medical knowledge about temperance.
They narrated analogies and stories that revealed subtle ways in which they raised questions, chiefly whether medical knowledge was the monopoly of upper castes and why scientific discoveries about alcohol elude Dalit and marginalised subjects.Nagulaiah, a Vivekavathi author, wrote that temperance contributed to healthy bodily practice and explicitly indexed sara as bringing about the loss of dignity and self-respect.Writing in deeply allegorical terms, where he conjured the self-harm that sara can cause, he narrated the story of the Yogi or Saint Samuelpresumably the same Samuel who features in the First Book of Samuelwho brought about the reform of an abusive drunkard husband and father.While the allegory was placed firmly in a Telugu context and featured plot elements removed from the First Book of Samuel, certain imagery from this part of the Old Testament cannot be missed.
Upon witnessing a woman suffer on account of her drunk husband, Samuel intervened to pay off the debt the latter incurs when he drinks sara without paying the sara seller.The husband subsequently had a dream involving four miceor chaturmooshikamuluwhere the first one was fat and aggressive; the second weak, malnourished and poor; the third in a dying condition and the fourth blind.Samuel explained to the young man that the first mouse represented the sara seller, the second represented his enervated and helpless wife, the third his children and the fourth blind one none other than the drunk 56 man. 61While Nagulaiah did not deploy medical wisdom in this temperance allegory, the imagery of the mice representing the fat sara seller, the dying children of the drunkard and the malnourished wife are hard to miss.Equally significantly, the point of this story was to explain how alcohol caused disease and not only loss of wealth, intelligence, health and life but also self-respect.Medical themes are present in this article when considered in a hermeneutic sense.In the King James Bible, the First Book of Samuel 6: 4 and 6: 5 gestures to mice as signifying 'plague', disease and povertymade explicit in 'mice that mar the land'.Nagulaiah's allegory brought these Biblical meanings to bear on a Telugu landscape, where sara dealers held their customers in thrall, commonly those from marginalised castes.Akin to what the historian R.S. Sugirtharajah indicates, non-Brahmin castes and converted Christians have appropriated the Old Testament for 'hermeneutic interpretation', where it resonated with their ambitions to fight caste. 62Even though Nagulaiah did not use the term 'caste' itself nor does he cite a caste name, it is unmistakable that in his allegory, alcoholism is enslaving in medical terms of disease and impending death and socio-economic terms of debt, both themes that recur in several Dalit and marginalised accounts.
An attentiveness to how alcohol enslaves and eats away at one's self-respect, even as it enervates the body and those around the drunkard, stand out here.This article in Vivekavathi cannot be seen in isolation; these interpretations of sara as an 'unclean' 63 beverage that enslaves Dalits find resounding echo in contemporaneous Telugu Dalit writings.Historians and Dalit Studies scholars, Chinnaiah Jangam and Sambaiah Gundimeda, provide detailed references to leading early-twentieth-century Adi-Andhra or Telugu-speaking Dalit intellectuals and reformer-activists.Gundimeda writes that the term Adi-Andhra was critical to the self-naming practices of Telugu-speaking Dalits in the Madras Presidencyit meant those who were the original inhabitants of the Andhra region. 64rominent Adi-Andhra activists, such as Bhagya Reddy Varma, Jala Mangamma, Jala Rangaswamy, Gangaiah Rayudu, Arige Ramaswamy and Kusuma Dharmanna, came up with minute socio-medical strictures against alcoholism as reinforcing spirals of debt, poverty and disease in their communities. 65angam indicates that these leader-intellectuals signalled the political charge of collective selftransformation that temperance held out to Adi-Andhras' assertions of self. 66He argues that their many intellectual and political legacies of publishing journals, initiating social reform, moulding the political arena and public opinion on matters including temperance underscored their role in shaping nationalist modernity.
The scientific basis of temperance was sharply conspicuous in contemporary Adi-Andhra writings.Gorantla Rangaiah, an author writing for The Bhagyanagar Patrika, a Dalit fortnightly newspaper, wrote in a medical voice against narcotic addiction.He expresses alarm at the rising trend in Hyderabad schools of using mattu vastuvulu or narcotic substances under peer pressure.He wrote that children could incur serious and early risks to their bodily and mental health when they 'rolled cigars, did snuff and got hold of bidis and cigarettes'. 67All this served as a precursor to the most dangerous health hazard of drinking alcohol.But perhaps most notable of Adi-Andhra temperance advocates was the medical voice of Kusuma Dharmanna and his impressively detailed medical tract called Madyapana Nishedhamu, the same title that O. David had earlier used for one of his temperance songs, even if these may be unrelated events.Kusuma Dharmanna's work, Madyapana Nishedhamu [Temperance], draws on the body as 'the house beautiful' analogy in that just as the house has its pillars, beams and walls, bones and muscles perform similar functions and the human the proud owner of the house. 68He shows how the well-oiled and well-functioning body is ruined when alcohol is introduced to it and how it causes havoc especially to respiratory, digestive, brain, muscle and other vital aspects.He discusses the nature of the human body, the water content in it and that far from being a food, alcohol is a poison that causes imbalance.He narrates the diseases that ensue from alcohol and warns against administering toddy to children.
Written roughly around the same timeframe, a few articles in Vivekavathi resemble Kusuma Dharmanna's medical pamphlet, in their tone, analogies and citation of scientific discoveries.Notable among these, although written prior to Dharmanna's tract, is Pendurti Joseph's medical temperance catechism (Figure 5). 69Both authors wrote in an entirely medical register to describe alcohol as an unnatural poison and outlined what fermentation does to alcohol.They described alcohol as a corpse preservative and cited experiments in which fish and snakes die when they are dropped in water with even one per cent alcoholic content.
Both Pendurti Joseph and Kusuma Dharmanna invoke the power of universal medical knowledge implicit in the latest scientific findings of the day on the toxicity, nutritional outcomes and bodily processes informing alcohol consumption.Their analogies demonstrated knowledge of the experiments and summaries in the writings of physicians such as Benjamin Ward Richardson, Victor Horsley and 68 Kusuma Dharmanna, Madhyapana Nishedhamu (Vijayawada: Prajashakti, 1930, reprint 2016).'The House Beautiful' analogy features in the White Ribbon series and amply in Vivekavathi pieces.WCTU author Margaret Denning's articles in this series were translated from English to Telugu and published in Vivekavathi.
Mary Sturge.Horsley and Sturge's famous work, Alcohol and the Human Body finds mention in quite a few medical temperance articles in Vivekavathi, especially on 'the medusa' or the fresh-water jelly fish. 70t is this performative register of bringing the latest medical findings to bear on the stigmas of Dalit bodies and livelihoods that Dharmanna definitely deployed and Joseph may have also undertaken.G. Subramanyam, in his introduction to Dharmanna's Madyapana Nishedhamuwhich won the best prize in a competition set by a regional temperance associationwrote that even if this was a medical text, the author was all too aware that alcoholism was not so much a personal addiction as a social disease that weighed Dalits down like a rugmata [sickness]. 71What is no coincidence is also that Dharmanna was an Ayurvedic doctor by profession who made extensive socio-literary interventions against untouchability, critiques of Hinduism and caste landlordism. 72hile not enough is known about Pendurti Joseph, other similar voices in Vivekavathi refrained from stigmatising Adi-Andhras or Dalits and were embedded in a critique of caste hierarchies.They could have emerged from similar considerations as Dharmanna's, namely that temperance was an article of Dalit selfrespect.These inflections set these authors apart from stigmatising voices whose disgust towards toddy and arrack occupied a scriptural-moral plane that upheld upper caste values of temperance.
Another voice that stands out in Vivekavathi is that of O. David, who authored many 'temperance songs' and 'temperance lyrics.'In one article, he implicitly condemned alcohol by laying out commandments for a healthy body.He wrote that consumption patients particularly needed fresh air, clean food, well-rested body, a reviving environment and a temperament of dwelling on positive things. 73On another instance, he suggested that temperance was critical because it 'enslaves' the alcoholic to debt (see original in Figure 6): In O. David's verse, enslaving debt and social ruin coalesced with a sensory assessment of disease.Alcoholism referred to a social state that was mirrored in the disease-prone and collapsing state of the body marked by chest-burn and a reeking throat.Neither O. David, Nagulaiah nor any other counter-caste voice in Vivekavathi challenges toddy and arrack as stigmatised objects, in medical or other terms.However, they do underline the trauma caused by social stigma to the alcoholic.
The form and content of these temperance songs pointed to a hybrid genre.On the one hand, they were inspired by the WCTU temperance hymns and featured the same titles such as 'Matanubhava Yodhulu' or 'Onward Christian Soldiers' (Figure 7).On the other, they were distinctive in that they contained themes of 'bondage' to alcohol. 75While the WCTU version conjured a marching army united in their path of the Cross and acceptance of Jesus as leader, the Telugu version featured, in addition, a passionate appeal to forgo debt and the stench of local beverages. 76If the WCTU hymns featured classical Western musical notations, O. David's hymns conformed to upper caste musical tastes in that these songs took the form of a keertana (classical Carnatic music form) set to taalam and raagam. 77However, the same song was unmistakably local in its appeal to forgo toddy and arrack.
The local here was embedded in anti-caste movements.Even if the form of the song was upper caste, the genre of the temperance hymn had Dalit roots.Temperance hymns were popularised by Adi-Andhra leaders like Jala Rangaswamy and Jala Mangamma. 78Chinnaiah Jangam states that writings by Gandhian Dalit leaders like Rangaswamy denouncing alcohol were definitely shaped by upper caste 'strictures imposed by reformist Hindu parameters on the Dalit worldview'. 79However, he argues against dismissing these writings for they recast history to enable a Dalit nationalist modernity.These complex social formulations were evident in Vivekavathi too.The same O.David decried kallu and sara by implicating them within a system of sexual slavery forced by Hindu society on Dalit communities like 'Malas' and 'Madigas'. 80In such a system, girls were forced by upper caste Hindus to become 'joginis', a term used usually to refer to Dalit women who were coerced into ritual prostitution.Requiring joginis to drink toddy and arrack was all too common, he writes.His angst against sexual slavery was reflected in contemporary Dalit women's writings. 81In the same article, he expresses consternation against the increasing enslavement of Dalits to these drinks owing to the ubiquitous natural feature of the eeta chettu or the date palm tree in every village in the princely state of Hyderabad.He implies that Dalit parents wrongly perceived date toddy to have nourishing qualities since they fed even their infants with this. 82Reading O. David's various articles together, it can be argued that he simultaneously rejected narcotic substances because they compromised every scientific principle of clean food, caused secondary diseases like tuberculosis and entrapped Dalits further in caste-based slavery.
Historiographies of medical temperance, alcoholism and enslavement
This article relies on scholarship gesturing towards the complementarity of missionary and medical languages in driving dominant discourses of 'alcoholism'.In this regard, historians have globally engaged with medical temperance within overlapping frameworks of vice, epidemic disease, social purity, hygiene and eugenics.The essays published in the edited volume Global Anti-Vice Activism see alcohol as a site of medicine's civilizational phobias and missionary anxieties manifest in panics around 'racial hygiene', 'racial poison' and 'venereal disease'. 83Nikolay Kamenov wrote that scientific discourses on 'alcoholism' in interwar Bulgaria and the Balkans region recast the WCTU missionary language of 'social and moral hygiene'. 84In his work, temperance visual culture took the form of posters, magic lanterns and pamphlets to warn against the addictive aspects of alcoholism.Roy Porter traced the disease concept of alcoholism to Georgian England and Christian evangelical narratives. 85n this article, however, missionary publications did not merely pave the way for a more medicalised language to emerge.They constantly cited Anglo-American medical narratives and resorted to heavily medicalised debates on digestion, toxicity, nutrition and addiction and featured minute descriptions of how alcohol affected the muscles, liver, brain and heart.But even in 1930, Telugu missionary journals did not use the medicalised terms, vyasanamu or addiction and vyasanaparudu or addict, but rather gestured towards 'alcoholism' through terms like tragudu or drunkenness, tragubothu or drunkard, nisha or intoxication, mattu or narcotic state and lobaduta or enslavement.
Drink historians have used the term 'slavery' in depicting subtle processes of the medicalization of alcoholism in Britain and Europe.Mariana Valverde and James Nicholls furnished commentaries on the institutional vigour, legislative power and disciplinary force undergirding medical theses about alcoholism.They discussed the medical framing of alcoholism in the nineteenth and twentieth centuries as 'slavery of the will' and 'a disease of the will' and inebriation as 'predisposition' to alcohol. 86Valverde wrote that in using this framework, the medical establishment displayed their incoherence as they could not arrive at a consensus on how to medicalise alcoholism.They could not class alcoholism as separate from 'the narcomania' of other drugs nor could they settle on 'the personality type' of the addict. 87They were also not univocal in their medical critique of the dangerous effects of alcoholism such as 'evolutionary degeneration', liver or brain damage. 88ddiction scholarship has underlined that alcoholisminitially termed 'habitual drunkenness'evolved through scientific practices of refutation and refinement.In Scottish physician Thomas Trotter's writing at the dawn of the nineteenth century, drunkenness was socially transmitted but featured as a pathology of mental habit. 89An American physician writing around the same time Benjamin Rush had a similar explanation about chronic spirit consumption giving rise to the 'disease of habituation'. 90This gave way to sharper and more specific formulations by the early nineteenth century French psychiatrist Jean-Etienne Esquirol, such as 'dipsomania' or 'drinking in excess' and monomania or 'compulsive drinking', which arose from 'partial insanity'. 91James Nicholls wrote that a Swedish physician, Magnus Huss, deployed terms like 'chronic' and 'acute alcoholism' to imply 'constant drinking' and 'bouts of extreme intoxication' and linked alcoholism to the nervous system.Thomas Sutton, an English physician argued in 1813 that 'delirium tremens' borne of excessive drinking was a disease in and of itself and an 'affection of the brain'. 92ompared to this, Telugu missionary publics were much more united in their critique of 'alcoholism' even if they too did not have a clear definition of the term.They found certain dietetic arguments against alcoholism to be more persuasive than others.Vivekavathi consistently argued against alcoholism on grounds of poor digestion, bad nutrition and toxicity.They argued that it caused lasting harm to the stomach and digestive fluids, the liver, heart and nervous system.While they did mention the brain and discuss insanity occasionally, they did not classify them much.They were selective in accepting dietetic arguments and overlooking deeply technical psychiatric arguments of dipsomania, monomania and delirium tremens within texts authored by physicians such as Victor Horsley and Mary Sturge, Henry Munroe and William Hargreaves, whom they cited.This could be because Vivekavathi authors found dietetic and generic neurological arguments to be socio-culturally familiar while they read clinical psychiatric definitions to be alienating.
Scholarship on medical temperance in India has not tended to alcoholism per se but has subsumed this question across historiographical discussions of subaltern, labour, military history and cultural studies.David Hardiman's work has been pioneering in that he situated drink scholarship within the Subaltern Studies tradition of foregrounding peasant and Adivasi worlds. 93His work and that of Indira Munshi Saldanha have demonstrated toddy and mahua arrack to inhabit Adivasi worlds as a customary, ritualistic, dietary, medicinal and healing practice.These drinks formed the centrepiece of tribal insurgencies against colonial excise regimes, 'landlords, usurers, and liquor dealers'. 94In his more recent work, Hardiman also documented the work undertaken by missionaries to de-couple Christianity from liquor.Alongside providing medical missions to help Adivasis and Dalits, they wished to enforce vows of abstinence and temperance norms.Hardiman's scholarship, while deploying ethnographic vernacular resources, is limited by an English language colonial and missionary archive.An engagement with the vernacular archive is replete with possibilities of discerning complex narratives of caste and nuanced overlaps between the languages of the medical and the spiritual, the empirical and the didactic and the regional and the global.
Historian Darinee Alagirisamy explained the political saliency of foregrounding nira or unfermented toddy as a healthy drink in colonial Madras Presidency. 95She wrote that the Tamil Nadu Congress strove hard to garner acceptance for nira as both indigenous and nutritionally superior to tea and coffee.David Fahey and Padma Manian outlined the international influences of the Victorian vegetarian movement and Gandhi's encounter with Liberal temperance advocates in England and South Africa on his nationalist discourse driven by 'self-purification'. 96 Harald-Fischer Tine presented the internal class biases about alcohol within British colonial society where it was deemed imperative to police the access of soldiers and sailors or the 'white underclass' to arrack for medical and racial reasons. 97Lucy Carroll and Eric Robert Colvard traced the deep-seated affinities of missionary and nationalist temperance work across India, the UK and America.This was evident when Congress workers and Anglo-Indian Temperance Associationan organization in India with both British and Indian temperance workers strategised collaboratively. 98They also documented the influence that doctors wielded in shared and separate forums and medical arguments about paralysis, stomach disorders and delirium tremens that featured in their temperance work.While all this work speaks in passing to medical questions of drinking, this article is distinctive in that it foregrounds caste as a structuring principle of discussions of alcoholism.Barring Lucy Carroll and Darinee Alagirisamy's work, this scholarship on alcohol and 93 David Hardiman, Missionaries and their Medicine: A Christian Modernity for a Tribal India (Manchester: Manchester University Press, 2017); Hardiman, op.cit.(note 5) 'From Custom to Crime'. 94Ibid., 21; Hardiman, op.cit.(note 5); Saldanha, op.cit.(note 5). 95Darinee Alagirisamy, op.cit.(note 5). 96David Fahey and Padma Manian, 'Poverty and Purification: The Politics of Gandhi's Campaign for Prohibition', The Historian, 67, 3 (2005), 489-506. 97Fischer-Tine, op.cit.(note 49). 98Carroll, 'The Temperance Movement in India'; Colvard, 'A World without Drink', op.cit.(note 5).Colvard, however, traces dissonances in these affinities in terms of missionaries pushing their own agendas.temperance in India does not systematically approach alcoholism as a historical reflection on caste.It also does not address itself to Indian language-based archival sources.
Scholarship on race, caste and slavery have illuminated the field of medical evolution.Samuel Roberts argues that the historical progression of diseases like tuberculosis and AIDS can illuminate structures of 'racial utilitarianism' or racialised practices and logics of responding to urban industrial capitalism. 99Jim Downs' (2021) work outlines how the bodily degradation of the black slave in the slaveship, plantation and battlefield carved out concrete opportunities for universal theories about overcrowding, smallpox and yellow fever. 100Similar and yet different in its social framing, scholarship in India has commented on the role of caste in determining medical practices.Kancha Ilaiah Shephard, David Arnold and Sohini Bhattacharya have demonstrated how medical science in late nineteenth century colonial India deeply benefited from enslaving caste labour and stigma.Shephard points to the historical trend of relying on Dalit labour to carry out graveyard work during the bubonic plague, as this was work that no one would perform. 101Arnold wrote that colonial post mortem practices of dissection evolved through the recruitment of a Dalit community of Doms who alone were willing, where all other castes refused, to assist medical practitioners in this practice. 102Sohini Chattopadhyay described the sanitary science of mortuaries to have drawn on the Dalit labour of Mahars and Doms all the more during colonial outbreaks of cholera and plague. 103This article synthesises these insights from the social history of medicine with those of Drink and Dalit Studies to delineate alcoholism as socio-medical enslavement within the context of the early-twentieth-century Telugu print sphere.
Disaggregating languages of medical temperance
Unlike Vivekavathi, the two other Telugu journals, The Telugu Baptist and UCH, decried alcohol itself rather than toddy and arrack specifically.Their usage of the term 'enslavement' did not index countercaste voices in that their discussions lacked local descriptions of caste inequality, landlordism, debt and stigma.What this also implies is that not all universal narratives of medical temperance carried the promise of counter-caste assertion.
These two journals provided Christian instruction for a clean and healthy body expressed in medical terms.They too referred to alcohol causing adverse physiological outcomes, but they deployed a less hybrid language than Vivekavathi.Their medical dictums were embedded in long scriptural exposition mainly from St Paul's Epistles and the Old Testament.Unlike Vivekavathi, the medical expositions in these journals shied away from commenting on the local power dynamics of alcohol.They took a more casual approach to all kinds of alcohol as bad, with articles seeking the separation of believers not so much from toddy and arrack as from all intoxicants like opium, cannabis and alcohol.
The term 'slave' features here too, but drinkers were presented as slaves to swadesha, videsha sarayilu [local and foreign alcohols] rather than kallu and sara.These journals re-published articles that featured in Vivekavathi and, in so doing, shared biases against toddy and arrack.However, they did not feature any original content that carried potent and unveiled messages against kallu and sara, in the name of temperance.
While medical temperance campaigns found several vehicles, such as catechisms, hymns and experiments in Vivekavathi, the other two journals preferred scriptural commentary.Sometimes, these entailed the invocation of one denomination as superior to another in terms of their internal spiritual strictures against alcohol. 104Where medical analogies were offered in these journals, they were always justified in Biblical terms, compared to Vivekavathi, where the Biblical was expressed often as a matter of form rather than argument.Vivekavathi used various Christian devices but did not cite scripture very much and sometimes omitted all allusions to Christianity in providing health insights.
These journals, on the other hand, turned resolutely to the Old Testament as well as verses from St Paul's Epistlesto warn believers against slavery to alcohol.One article categorically stated that alcohol disrupts blood flow and impacts liver function, burns the heart and decreases life expectancy, and it rapidly turned to the Book of Proverbs to warn, Be not among winebibbers; among riotous eaters of flesh: For the drunkard and the glutton shall come to poverty: and drowsiness shall clothe a man with rags (Proverbs 23: 20-21). 105e message of slavery and the need to escape it is more explicit in passages such as this: If it is a yes, is it right to make impure our body with tobacco, opium, weed, and alcohols where the almighty God, our father, would like to reside?So, as Christians, 'Stand fast therefore in the liberty wherewith Christ hath made us free, and be not shackled again by the yoke of slavery' (Galatians, 5: 1).As said here, we should not touch, taste or be enslaved by these poisonous substances and seek help from the Holy Father to help others from this plight. 106e Telugu Baptist conveyed bodily suffering in medical terms and pinpointed disease as the manifestation of a failure to recognise that the body is 'the temple of God', not to be corrupted with alcohol.Here, it is proclaimed that If any man defiles the temple of God, him shall God destroy; for the temple of God is holy, which temple ye are' (1 Corinthians 3: 17). 107ile strictures against wine-bibbers and riotous eaters of the flesh may be an allusion to the stigmas that caste presentsgiven that Dalits and other marginalised castes were vilified as beef-eaters and alcoholicsin the absence of context, this reference can neither be regarded as reinforcing or resisting this stigma.These scriptural commentaries did not amount to a call to end slavery among Dalits through the agency of temperance as much as an exhortation to Telugu society, a subset of Indian society, to embrace good habits such as forgoing both tobacco and alcohol.For only a 'clean body' and one that is purged of durvyaparamu or 'dissipation', which results from drinking wine (a placeholder for all alcohol), in the estimation of St Paul 108 could be the ideal abode of God.
These journals placed alcohol's physical afflictions side by side with those wrought by tobacco and opium, but they saw alcohol as way more dangerous.The Telugu Baptist made references to nallamandu or 'opium' addiction in China, where the dependence on the poppy seed derivate grew unchecked: 'the people of China start by imbibing opium doses of as little as a green lentil, but soon, this increases to 104 M.E.Archibald, a white editor of Vivekavathi, wrote for The Telugu Baptist as well.She described how she met a Christian who expressed joy that Baptists took their covenant with God seriously, unlike other denominations, and abhorred all intoxicants, alcohol and cigarettes alike.'Total Abstinance (sic) -Madyapaana Nishedhamu', The Telugu Baptist, 35, 10 (1911), 145-9.consuming opium in dosages that resemble a raw egg'. 109The consequences of opium addiction were explained to be day-time sleepiness, nausea, bodily lethargy, impaired mental faculties and poverty.The author regards tobacco consumption to be both unchristian, in that it involved squandering money that could go to the Church, and lethal, in that it impaired speaking, hearing and seeing faculties. 110But smoking was dangerous in that it represented a sliding scale of slipping into alcoholism.Alcoholism, however, outpaced the unmitigated health disasters of smoking tobacco and opium in that it dramatically reduced life expectancy: They taste the local and foreign alcohols first, feel a growing attachment the second time, and become slaves to alcohol the third time.Compared to the other intoxicants we have discussed above, alcohol is much more dangerous.It disrupts the flow of the blood, impacts liver function, burns the heart, and kills you a little day by day.It decreases your life expectancy.People die suddenly because of their drinking habits (emphasis mine). 111e alcoholic body, marked by the failing heart and liver and diminished blood circulation, was abstracted from a social entity and suffered the same physical consequences everywhere.Enslavement here was nothing more than alcoholic and more generally, narcotic addiction.In another article, the author has dual advice for students and teachers.While steering clear of enslavement to alcohol, teachers should educate students about alcohol's harmful effects by conducting scientific experiments to prove this.Students must read 'Ayurvedic and medical texts' to educate themselves similarly. 112CH carried calls for the need to educate people in the global gains vis-à-vis temperance, for instance, in Persia and Glasgow.It exhorted readers to use scientific and religious texts to disseminate temperance ideas.The immersion of alcohol within a global landscape of temperance points to the use of universals that shied away from critiquing local power dynamics. 113
Conclusion
While Alcohol Studies have examined the place of temperance in social movements in India and elsewhere, they have not studied this systematically within either medical debates or the fine print of vernacular language press and journals and much less within missionary medical discourses of such journals.And it is this unexplored niche of medical writing on alcohol and temperance in reference to caste in Telugu missionary journals of the early twentieth century that this article dwells on.
The form of medical temperance was carefully curated to convey a range of missionary lessons about alcohol's addictive qualities, toxicity, lack of hygiene, impact on digestion and poor nutritional content.Owing to its desire to present itself as 'secular' unlike Christian missionary and Hindu women's journals of its time, Vivekavathi at different points, carried medical pronouncements against alcohol that did not foray into the spiritual realm.On other occasions, it presented medical temperance by using spiritual aids such as allegories, catechisms and hymns.The journal invoked a wealth of forms of 'the universal'.The universal was present either in terms of Christian narrative devices or medical paradigms of experiments, with its attendant language of observation, results and inferences, or in a synchronised overlap of both.It was also manifest in Vivekavathi's close affinities to British and American medical and missionary publications and the journal's inclination to use the same narrative genres.But the hybrid form of Vivekavathi did not only reflect the distillation of medical with missionary insight but also the 109 Ibid., 50. 110Ibid. 111Ibid., 49.The Telugu term lobaduta is recurringly used; this can be contextually translated into 'succumbing' to addiction, or a physical enslavement to alcohol.appropriation of medical temperance to convey an intense social commentary about caste and castecoded enslavement to alcohol.The term 'enslavement'even where it had universal connotationstherefore was differently imbued.Just as it has been shown that there was no single Bible but rather the possibilities of an insurrectionary Bible, abolitionist Bible and pro-slavery Bible 114 , medical texts like Horsley and Sturge's Alcohol and the Human Body as well as Henry Munroe's The Physiological Effects of Alcohol saw multiple interpretations, mutations, mistranslations and selective translations.The socio-medical undercurrents of temperance in certain counter-caste narratives were empowering, while other manifestations were not in Vivekavathi, UCH and The Telugu Baptist.
Similarly, the terms 'toddy' and 'arrack' too can be placed in a spectrum in terms of their caste connotations in medical discourses.If at times they signalled and performed a crushing stigma, especially where the medical invocation of these drinks was contrived to shame marginalised castes, and Dalits in particular, into renouncing these drinks, not all medical references to them carried out the same discursive functions.Historically, Dalit writing very deliberately synthesised cutting-edge medical findings about alcohol into their own temperance expositions, in a quest for collective self-assertion.The counter-caste voices of Vivekavathi spoke in a similar mould to liberate marginalised castes from these stigmas, while preserving the stigma of these drinks per se.
Figure 4 .
Figure 4.Table describing experiment with toddy and arrack in Vivekavathi; Source: AP State Archives.
Figure 5 .
Figure 5.A temperance catechism authored by Pendurti Joseph explaining fermentation and likening alcohol to a corpse preservative.
Figure 6 .
Figure 6.A temperance hymn in Vivekavathi that melds social and medical conceptions of enslavement.
Figure 7 .
Figure 7.The Vivekavathi temperance hymns draw on WCTU temperance songs, but feature local interpretations of toddy and arrack.
Valverde, Diseases of the Will: Alcohol and Dilemmas of the Will (Cambridge: Cambridge University Press, 1998). 25O. David, Mitanubhava Yodhulu, Vivekavathi, 9, 12 (1918), 283. 26All three journals were easily available in the Andhra Pradesh State Archives in its older location in Hyderabad in Telangana state.These archives have since moved to Mangalagiri in Andhra Pradesh.
Table describing experiment with toddy and arrack in Vivekavathi; Source: AP State Archives. | 2023-10-13T13:01:50.477Z | 2023-10-01T00:00:00.000 | {
"year": 2023,
"sha1": "2c752172f7d151a7fd2b74db095bdc3f8c63af5a",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Cambridge",
"pdf_hash": "2c752172f7d151a7fd2b74db095bdc3f8c63af5a",
"s2fieldsofstudy": [
"Sociology"
],
"extfieldsofstudy": []
} |
169542664 | pes2o/s2orc | v3-fos-license | The Influence of Relationship Marketing Orientation (RMO) on Customer Retention in Travel Agency Services
The advancement in information technology had influenced customer attitude, behavior, and lifestyle in purchasing holiday and travel services. There are significant shift from utilizing travel agency to online travel agents. These change had pressured, travel agency to consider on refining their strategy to focus more on relationship marketing orientation (RMO) to achieve customer retention. The impact of RMO on the business or firm performance has been widely studied by previous literature in the organizational contexts. However, the empirical research from the customer perspective is rather limited. Therefore, this conceptual paper is proposed to fill in the RMO literature gap by focusing on the relationship between RMO and customer retention from the customer perspective. Academically, this study contributes to the literature of RM in business-to-customer context specifically from customer point of view and practically to the travel agency in comprehending RMO in enhancing their services in the travel agency sector.
Introduction
The tourism sector has been identified as one of the economic sectors that will contribute to Malaysian economic development in the future (Malaysian Investment Development Authority (MIDA), 2016). MIDA stated that, this sector would strive to achieve approximately RM168 billion by the year 2020, thus this sector will experience tremendous changes in strategy due to an increasing number of the travel agency targeting the same target market. This increasing numbers of the travel agency will lead to greater competition, which urged them to work harder in order to attract and retain customers.
In addition, the advancement of information technology has added another pressure to travel agency since it has enabled consumers to do-it-yourself their holiday and travel instead of using travel agency services. Although travel agency services provide one-stop center and online travel agents (i.e. Agoda.com, Trivago.com, Airbnb) only provide the platform for holidays and travel planning, the user-friendly technology interface has made it simpler, and easier for customers to do-it-yourself their holiday and travel planning. This is supported by Lang (2000) who said that these changes have pressured travel agency services to change their strategy in order to remain competitive in the marketplace. These online travel agents offer money saving for customers where, they are free to choose from a wide variety of accommodation, flight tickets, and tour promotions according to their budgets. These online travel agents are also convenience to customers since all transactions are done online. In contrast, travel agency requires customers to meet them to discuss their holiday and travel, this indirectly incur travel and time costs to customers. Nevertheless, customers will also need to pay extra for all the services provided by this travel agency. Overall, online travel agents provide more benefits compared to travel agency services, especially for budget customers who are looking for more value for every single cents that they spent.
Given the competitive marketplace, travel agencies in Malaysia are facing a massive challenge from online travel agents. Zare & Chukwunonso (2015) suggested that travel agency should possess strategic plan specifically in customer relationship management in order to stay competitive in the marketplace. Customer relationship is constituted of Relationship Marketing (RM), in which RM is focusing on enhancing the employee-customer relationship through Relationship Marketing Orientation (RMO). Where, RMO strategy will provide competitive advantage to the firm. This is supported by Bataineh et al. (2015) who found that the company needs to practice high relationship quality with the customer in order to achieve sustainable competitive advantage.
The question now is that, to what extent RMO will influence customer retention? Previous research regarding this issue is still scarce. Thus, there is an opportunity for the current research to further discover RMO. This research will shade-light about the importance of RMO in relation to customer retention in travel agency services. Given the limited attention on the study involving business to customer context in RMO and customer retention, this study will narrow the existing research gap in the relationship marketing (RM) literature. The following section will explore the literature and explain the theoretical foundations of the current research.
Review of the Literature Relationship Marketing Orientation (RMO)
Building relationship with customers is the key factor to attract and retain them (Tamuliene & Gabryte, 2014). Hence, applying RMO in the service industry is a possible way to gain competitive advantage, in which RMO will enhance customer retention. The term RM was first coined by Berry, 1995 who defined it as "attracting, maintaining and in-multi-service organizations enhancing customer relationship". In addition, RM is further defined as to produce quality improvement in order to generate customer's satisfaction that leads to customer retention (East, Hammond & Gendall 2006). Accordingly, Morgan & Hunt (1994) suggested RM as a focus on all marketing activities directed to establishing, developing, and maintaining successful relational exchanges. Sheth & Parvatiyar (1995) described RM as an integrated involvement of customers, suppliers, and industry partners into a firm marketing activity and development. Whereas, Grönroos (1996) extended the definition of RM as the activities to identify, establish, maintain, and enhance the relationship with customers and other stakeholders to achieve all parties' objectives and profits through mutual exchange. Consequently, Gummesson (1997) in his study extended the definition in terms of marketing as building relationships, networking, and interaction in RM perspective.
Furthermore, RMO was also referred to as a relationship involving employee and customer through service delivery. Based on previous studies, there are several dimensions of RM orientation adapted in different context of research. Sin et al. (2005) has developed RMO scale based on the importance of the RMO dimension. Six dimensions were suggested namely trust, bonding, communication, shared value, empathy, and reciprocity. Furthermore, these six dimensions are shown to have a substantial association with firm business performance. These six dimensions and scales developed by Sin et al. (2005) were adapted by other recent scholars to investigate its consequences on company performance (Kucukkancabas, Akyol, & Ataman, 2009), customer satisfaction (Hau & Ngo, 2012), banking performance effectiveness (Wongsansukcharoen, Trimetsoontorn & Fongsuwan, 2015), brand equity (Yoganathan et al., 2015) and marketing effectiveness (Ghani, Othman, Ibrahim, & Ismail, 2016). In contrast, a recent study by Kwan & Carlson (2016) focuses on RMO in influencing firm performance using different dimensions of RM consisting of bonding, communication, empathy, harmonious conflict resolution, shared value, trust, and reciprocity. However, these studies focus on business-tobusiness setting instead of business to consumer setting.
Customer Retention
In today's business landscape with intense competition, customer retention is seen as one of the success factors for firm's sustainability. The marketplace is full of challenges in both macro and micro environment, which leads to constant changes in customer behavior, attitude, and lifestyle. Therefore, in order to survive and remain competitive in the marketplace, the firms cannot rely only on attracting new customers but they need to retain existing ones. This is supported by, Al-Hersh, Aburoub, andSaaty, 2014 andFarquhar (2005) who emphasized that retaining customers will increase company's profitability and decrease acquisition costs.
Customer retention can be defined as an activity of converting new customer to regular customer by providing excellent customer service that enhances long-term customer satisfaction (Kotler & Armstrong, 2013). Meanwhile, Jeng & Bailey (2012) further defined customer retention as a process in which customer involvement in formal and/or non-formal relationship with the firm in a long term basis as well as re-patronizing the firm's product and/or services.
Previous study had suggested several antecedents of customer retention. Bojei et al. (2011) who studied customer retention in retail setting indicated that customer service, loyalty or rewards program, store community, personalization, and customization influenced customer retention in the retail store. Another study in the mobile telecommunication context, showed that customer retains to specific mobile telecommunication provider because of satisfaction on the services provided, the quality of relationship and the switching costs (Tamuliene & Gabryte, 2014). In a business-to-business setting the antecedents such as perceived value, reputation, trust and switching cost have positive influence on customer retention in the healthcare service context. This proved that by focusing on customer satisfaction in services, it would lead to customer retention not only in the business-to-customer relationship but also in the business-to-business relationship.
In addition, there are also past researches, which suggested RM elements as antecedents of customer retention. These RM elements are communication, knowledgeability, empowerment, personalization, ethical behavior, fees and technology (Rootman, Tait, & Sharp, 2011), and customer trust (Soimo & Wagoki, 2015). The entire RM element in both research were proven to influence customer retention in the banking industry.
Relationship Marketing Orientation and Customer Retention
Theoretically, the relationship between RMO and customer retention can be explained through stakeholder theory (Jones et al., 2010). Stakeholder theory implies that manager should possess value creation or excellent product and/or services among the stakeholders (i.e. customers, communities, employees, suppliers, financiers and other stakeholder groups) which will influence the business performance (Freeman, Wicks, & Parmar, 2004). Furthermore, business performance consists of sales growth, market share, customer retention and return on investment. Thus, the value and relationship with stakeholder is a critical part for the manager to consider for ongoing success. This stakeholder theory clearly explains that RM is one of the value creation elements where manager tries to perform in order to fulfill the stakeholders' needs and wants that lead to competitive advantage (Jones, 1995). Thus, value creation is conducted using RM elements in the employee-customer relationship, in which the RM elements will influence customer to retain with the firm. The stakeholder theory (Jones et al., 2010) also suggested that customer will retain with the organization when they experience positive value of RMO through employee-customer relationship.
Another common theory used to explain RMO is social exchange theory (Emerson, 1976). This theory explains that the exchange process takes place between two parties and there are mutual agreement in the transaction and/or exchange. Moreover, Cropanzano & Mitchell (2005), stated that social exchange theory applied reciprocity rules in the exchange, where there is a need to give something for the purpose of getting something in return. Among previous studies that applied social exchange theory in RM are Yoganathan et al. (2015), Brun, Durif, & Ricard (2014) and Sheth & Parvatiyar (1995).
Past studies have suggested several elements of RM as the antecedents of customer retention. These elements include RM factors (communication, expertise of seller, comparison level of alternatives, cooperation and dependence on seller) (Bataineh et al., 2015), online RM (financial bonding, social bonding, structural bonding) , customer trust (Soimo & Wagoki, 2015) and RM (communication, empowerment, personalization, ethical behavior, fees and technology). In addition, there are also previous studies that investigated RM in terms of other consequences such as customer satisfaction (Hau & Ngo, 2012), customer loyalty (Anabila et al., 2012) (Chiu et al., 2005), performance (Kanti & Dixit, 2014) and repeat purchase intentions (Mpinganjira, 2014).
Earlier study focusing on RMO by Sin et al. (2002) provided the important dimension of RMO and their measurement scale in the service context. Additionally, Sin et al. (2005) adopted these scales to investigate the influence of RMO on business performance in two different cultures namely China and Hong Kong. This study indicated that RMO has direct relationship on business performance (sales growth, market share, and customer retention). Recent studies of RMO by Wongsansukcharoen et al. (2015) and Kwan & Carlson (2016) who adapted the same scale to investigate performance, in which this performance also consists of customer retention. Thus, it can be concluded that the studies done by Sin et al. (2005), Wongsansukcharoen et al. (2015) and Kwan & Carlson (2016) proved that RMO leads to customer retention. Table 1 provides a summary of previous researches, which studied the links between relationship marketing and customer retention. Consequently, based on the discussion and rationale of the link between RMO and customer retention, the following research proposition is suggested: P1: There is relationship between relationship marketing orientation and customer retention.
Conceptual Framework
RMO is proposed to have a relationship on customer retention. Figure 1 presents the illustration of the framework. The six dimensions of RMO were adapted from Sin et al. (2005) since these dimensions have been proven to influence performance (i.e. customer retention). Recently, these dimensions were also adapted by Ghani et al. (2016) and Yoganathan et al. (2015) to investigate marketing effectiveness and brand equity. The results of both studies indicated that these six dimensions had significantly affects the dependent variables. Thus, the current study proposed to investigate the effects of RMO on customer retention from customer's point of views.
Figure 1: The proposed conceptual framework Conclusion
The review of the past and recent literature on RMO provides an overview and some suggestions that need to be further investigated. Consequently, this paper reviews the pinnacle of most common RM elements that were tested in previous studies and their common consequences. Since most of previous researches concentrated on business-to-business, future research should concentrate more on business-to-customer aspect to gain deeper insight and comprehend RM in firm-customer relationship. Given the new challenges such as information technology ix. Soimo & Wagoki (2015) Customer trust Customer Retention Trust is key in RM which affects customer retention.
x. Bataineh et al. (2015) Communication advancement, changes in customer attitude, behavior, lifestyle and intense competition, the effectiveness of RMO in service context should be investigated further to give a clearer picture for the service organization to adopt RMO. Large number of previous researches on the effects of RMO on performance tend to concentrates more on employees' perspective in an organization. Therefore, future research in RMO should emphasize on final customer's perspective to further enrich the RM literature.
Overall, customer retention is a necessity for every firm to make sure customers are satisfied and stay with the firm while spreading positive word mouth-to-mouth (Carter, 2010). RMO is one of the important key success factors in retaining customers as well as in achieving competitive advantage. Therefore, travel agency management should stress more on giving RMO training to their sales people so that they are equipped with the necessary relationship management skills and thus RMO dimension thrive in any business dealing involving employee-customer relationship.
This research wills open-up manager's eyes on the importance of RMO in influencing customer retention especially for travel agency services. It will also provide guidelines for researchers to develop new research model through past empirical literature reviews in short, RMO can facilitate excellent employee-customer relationship in the realization of RM in travel agency services. This research also corroborates RMO as an important topic for future research. | 2019-05-30T23:44:30.140Z | 2018-04-12T00:00:00.000 | {
"year": 2018,
"sha1": "629159d5847aa6a612ce080c5d0ee82d22544d59",
"oa_license": "CCBY",
"oa_url": "https://hrmars.com/papers_submitted/3979/the-influence-of-relationship-marketing-orientation-rmo-on-customer-retention-in-travel-agency-services.pdf",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "7ae78197fb94a416c0b2ad99cd9efccabd80e09b",
"s2fieldsofstudy": [
"Business"
],
"extfieldsofstudy": [
"Business"
]
} |
271184515 | pes2o/s2orc | v3-fos-license | Bixafen, Prothioconazole, and Trifloxystrobin Alone or in Combination Have a Greater Effect on Health Related Gene Expression in Honey Bees from Nutritionally Deprived than from Protein Supplemented Colonies
Simple Summary In many regions of the world, foraging bees encounter an environment with an inadequate supply of nutrients compared with the colonies’ demands, as well as exposure to pesticides used to combat plant pests and diseases, including fungicides which historically were considered innocuous for bees. The use of pesticide mixtures can increase the risk for bees, especially when limited foraging leads to nutritional deficiencies. In our study, we compared the effects of the commercial fungicides bixafen, prothioconazole, and trifloxistrobin, alone and in combination, on foraging honey bees that were collected from colonies with supplemented or restricted food availability. For this purpose, we assessed the expression of health-related genes, including antioxidant genes (SOD-1, CAT, and GPX-1), detoxification genes (GST-1 and CYP306A1), the storage protein gene vitellogenin, and immune system antimicrobial peptide genes (defensin-1, abaecin, hymenoptaecin, and apidaecin) using real-time PCR. All three fungicides caused alterations in the expression of these genes, especially the mixture with the three active ingredients, both in bees from colonies with supplemented feeding and those from colonies subjected to food restriction. The effect of fungicide exposure on gene expression was dependent on bees’ nutritional status, which suggests that adequate nutrition can help counteract the effects of fungicides on bees. Abstract The aim of this study was to evaluate whether alterations in food availability compromise the metabolic homeostasis of honey bees exposed to three fungicides alone or together. Ten honey bee colonies were used, with half receiving carbohydrate-protein supplementation for 15 weeks while another five colonies had their protein supply reduced with pollen traps. Subsequently, forager bees were collected and exposed by contact to 1 or 7 µg of bixafen, prothioconazole, or trifloxystrobin, either individually or in combination. After 48 h, bee abdomens without the intestine were used for the analysis of expression of antioxidant genes (SOD-1, CAT, and GPX-1), detoxification genes (GST-1 and CYP306A1), the storage protein gene vitellogenin, and immune system antimicrobial peptide genes (defensin-1, abaecin, hymenoptaecin, and apidaecin), through real-time PCR. All fungicide treatments induced changes in gene expression, with bixafen showing the most prominent upregulation. Exposure to 1 µg of each of the three pesticides resulted in upregulation of genes associated with detoxification and nutrition processes, and downregulation of immune system genes. When the three pesticides were combined at a dose of 7 µg each, there was a pronounced downregulation of all genes. Food availability in the colonies affected the impact of fungicides on the expression of the studied genes in forager bees.
Introduction
Bees are susceptible to numerous environmental stressors, including pesticides; among these there is little or no use restriction for fungicides in many countries [1,2].This is of concern because fungicides account for approximately 35% of the global pesticide market [3].The widespread use of fungicides to control and prevent damage due to pathogenic fungi often results in the development of fungicide-resistant strains, implying the need for new plant disease control strategies.An alternative to help overcome resistance has been to apply fungicide mixtures to combat one or more pathogens at various stages of the life cycle [4].
Lethal effects for bees resulting from exposure to fungicides are typically observed through exposure to relatively high doses (>100 µg/bee) [5][6][7].However, the actual impact on bee health needs to be evaluated with environmentally relevant doses of fungicides.Data on these doses can be difficult to obtain, especially for new products, for which there is little information concerning contamination levels on treated crops.Determining impact is more complicated when multiple active ingredients (a.i.) with fungicidal action are used in the same commercial product.These mixtures can lead to synergistic effects, which may not be observed when they are evaluated individually [8][9][10].
A commercial product that contains the active ingredients bixafen (bix), prothioconazole (pro), and trifloxystrobin (tri) is recommended for controlling fungal diseases in various major crops such as cotton, corn, sunflower, and soybean, with no restrictions on applications during the flowering period of these plants [11].Since these crops are visited by bees for pollen and nectar, these non-target organisms may be exposed to the fungicidal active ingredients by contact or ingestion of contaminated food [12,13].In Brazil, in 2022, 1981 t of bix, 5193 t of pro and 5195 t of tri were commercialized in various commercial fungicides containing one or more of these a.i.[14].
Bix is a fungicide belonging to the chemical group of carboxamides; it interferes with the mitochondrial respiratory chain by inhibiting the enzyme succinate dehydrogenase, which is responsible for electron transfer in complex II.Consequently, with less energy available, vital fungal functions are impaired, impacting reproduction and potentially leading to mortality [15,16].There are few studies addressing the effects of bix on pollinators.Another fungicide from the same carboxamide chemical group, boscalid, is more widespread, and known negative impacts on bee health due to exposure to this fungicide include reduced longevity [17] and decreased wing beat frequency, rendering bees lethargic [18].
The fungicide tri is a member of the strobilurin chemical group, which inhibit mitochondrial respiration in complex III of the respiratory chain.Consequently, the energy cycle within the fungus is interrupted, halting ATP production [19].Exposure of the Brazilian stingless bee with the common name Mandaçaia (Melipona quadrifasciata) to tri combined with the fungicide tebuconazole resulted in increased mortality; additionally, bees were repelled from sprayed tomato flowers.With fewer bees and reduced flower visits, there was a negative impact on fruit production [20].
The fungicide pro belongs to the triazolintione group and interferes with sterol biosynthesis.When pro binds to the enzyme sterol C14-demethylase, it blocks the sterol biosynthesis pathway, resulting in inhibition of production of phospholipids, accumulation of free fatty acids, and fungal death [16].This a.i. is marketed in over 60 countries and is the third best-selling fungicide worldwide [21].There is disagreement regarding the level of toxicity of this fungicide for honey bees [22,23].However, triazole fungicides, which have the same mode of action as the triazolintione group, can affect bee behavior and susceptibility to diseases [24,25].
Another factor inherent to conventional agriculture that can be detrimental to bees is decreased natural food resources.Large-scale monocultures, coupled with deforestation of native vegetation, result in reduced availability of pollen and nectar [26].Though some crops provide pollen, the amino acid composition of pollen from a single species may not meet the nutritional demands of the bees [27,28], leading to malnutrition and increased susceptibility to environmental stressors [29,30].
Our objective was to determine how individual and mixed fungicide ingredients affect the expression of genes related to honey bee health and whether nutritional condition of the colonies they came from affects gene function.We treated forager bees from well fed and from nutritionally deprived colonies with the fungicides bix, pro, and tri and measured the expression of antioxidant, detoxification, storage protein, and antimicrobial protein genes.
Location of the Experiment and Preparation of the Honey Bee Colonies
The study was conducted in the municipality of Dracena, state of São Paulo, at a latitude of 21 • 27 ′ 37 ′′ South and longitude of 51 • 33 ′ 21 ′′ West, and altitude 392 m.The local climate is characterized as tropical with a dry season.During the experiment, the average temperature was 26.9 • C, the average relative humidity was 63.1%, and there was 390 mm of rainfall.
Ten colonies of Africanized honey bees (Apis mellifera), headed by open mated queens, and kept in 10 frame Langstroth hives, were used.In the week prior to the start of the experiment, the colonies were standardized so that they presented at that time six frames of brood and four frames of food.They were divided into two groups of five hives: one group with supplemented feeding (SF) and the other with reduced food (RF).
Over a period of 15 weeks, the SF colonies were able to store all the food they collected in the field and were supplemented with 500 mL of syrup containing water and sucrose in equal proportions by weight, twice a week, as well as 100 g of protein paste composed of two parts multifloral bee pollen and one part of honey, provided once a week.Both pollen and honey were obtained from hives in the same apiary where the experiment was conducted.For the other group (RF), during the same period, there was no food supplementation, and pollen traps were installed at the hive entrances to reduce the availability of protein for the bees.
Bee Pollen Collection from Colonies with Reduced Feeding
In the RF group, bee pollen was collected daily and weighed using a semi-analytical balance, recording the quantities obtained weekly for each colony.To determine the efficiency of the pollen collectors (%), the number of forager bees going through the pollen traps was observed during 30-min intervals between 8 and 10 AM, with three repetitions per colony.The number of bees entering with pollen and the number of pollen pellets falling into the pollen trap drawer were recorded.The efficiency of the traps was calculated as the ratio of the number of pollen pellets collected, divided by two times the number of bees entering the collectors with pollen loads.
Exposure of Forager Bees to Single Fungicides or in Combination
Forager bees were collected from the hive entrances after 15 weeks of feeding management.The bees were placed in 0.25 L clear round plastic containers with perforated lids.In the laboratory, the bees were anesthetized in a freezer (−20 • C) for several minutes.
The immobilized bees were exposed by contact to bix, pro, and tri, by applying 0.5 µL of an acetone solution containing 1 or 7 µg of one or all three active ingredients in combination, using a micropipette, on the dorsal part of the thorax of the bees [31].The doses were determined based on a study estimating the dose at which a bee would be exposed by contact when applying a fungicide containing the three active ingredients according to agronomic recommendations [32].The acetone solvent alone was applied for the control.Acetone and the fungicide a.i.were purchased from Sigma-Aldrich, St. Louis, MO, USA.
Following contact exposure, groups of 20 forager bees from colonies that received the same feeding management and exposure dose to the fungicides were placed in new plastic containers and fed ad libitum with syrup containing two parts sucrose and one part water (weight/weight), maintained for a period of 48 h in an incubator with controlled temperature (33 • ± 1 • C) and humidity (70 ± 10%).
Gene Expression Assessment
After 48 h in the incubator, the bees were killed in a freezer for dissection, which was carried out with the aid of a stereo binocular microscope, fine tip scissors, and entomological forceps.We discarded heads, thoraxes, and intestines, leaving only the abdomens.Each sample consisted of six abdomens, with four samples per treatment (two colony feeding regimes × four fungicides × two doses and controls).
RNA Extraction and cDNA Synthesis
The gene expression analysis was performed by Real-Time qPCR, using the primers described in Table 1.
Table 1.A reference gene and genes related to bee health that were analyzed, and their respective primer sequences (GenBank accession numbers) and references.
After RNA isolation, quantification was made in ng by absorbance at 260 nm using a Nanodrop-1000 (Thermo Scientific, Waltham, MA, USA).An aliquot of RNA was processed in a 1% agarose gel and post-stained with GelRed ® (Sigma) to verify RNA integrity.The cDNA was constructed from the extracted RNA by reverse transcription using 1 µg of total RNA and iScriptTM III (Life Technologies, Waltham, MA, USA).The resulting cDNA was stored at −20 • C until use.Subsequently, the sequence was amplified by polymerase chain reaction (PCR) using target primers.
Determination of Relative Gene Expression
The determination of gene expression was carried out using the real-time polymerase chain reaction (qPCR-RT) method, and the results were analyzed by relative quantification using the 2 −∆∆Ct method [38,39].qPCR reactions were performed in 96-well plates using iTaqTM Universal SYBR ® Green Supermix (Bio-Rad, Hercules, CA, USA).Reactions were conducted following the manufacturer's instructions, with 20 µL individual reaction mixtures consisting of 10 µL of SYBRGreen PCR Master Mix (2×), 8 µL of 200 nM primers, and 2 µL of cDNA template (samples) [39].
All reactions were performed using the iCycler iQ™ real-time qPCR detection system (Bio-Rad Laboratories) under the following protocol: 95 • C for 5 min, 40 cycles of 95 • C for 30 s, 56 • C for 30 s, followed by analysis of the dissociation curve (melting curve) to verify amplification of a single product.The denaturation-hybridization-synthesis cycle temperatures used were according to Pfaffl et al. (2004) [38].
The threshold of the exponential phase, denoted as the Cycle Threshold (Ct), was detected during the temperature cycles, precisely quantifying the products of the amplification reaction by fluorescence emission.Expression data was used to calculate the Ct values.PCR efficiency and the relative expression ratio of target genes in experimental groups versus control groups were calculated according to the method of Pfaffl et al. ( 2004) [38].The comparative Ct method (2 −∆∆Ct method) was used to analyze the expression level of these genes relative to the RPL32 gene [33].
Statistical Analyses
The mean amount of bee pollen collected weekly from each colony was calculated for the RF colonies.The efficiency of the pollen traps was evaluated by the mean of 12 observations across four hives.Data dispersion was determined by calculating the standard error.
For the gene expression analysis, a completely randomized design with a factorial scheme was employed.The experiment included two nutritional levels (supplemented and reduced feeding), four fungicides (three active ingredients and a mixture of all three), and three doses (control, 1 µg/bee, and 7 µg/bee).Since the data did not follow a normal distribution, a generalized linear mixed model was used.This model considered feeding status and exposure status as fixed variables and replicates as random effects for each fungicide.The data were then compared using Tukey's test at a 5% significance level [40].
Feeding Management of Colonies
During the 15-week feeding management period, each of the five colonies with food supplementation fully consumed the 100 g of protein supplement and the 1000 mL of sucrose syrup offered weekly.One colony from the RF group succumbed at the 11th week and was not replaced.Therefore, to maintain the same number of colonies among the groups, one colony from the SF group was randomly selected and removed from the experiment.
In the RF group, 33.8 ± 4.9 g of bee pollen was collected weekly per colony.By observing foraging bees with pollen in their corbiculae going through the pollen traps, it was estimated that the efficiency of these collectors was 57.9 ± 5.1%.Thus, of all the pollen collected by the bees from colonies with reduced feeding, a mean of only 42.1% remained available for the bees.
Gene Expression
We observed a decrease in activation of all the genes when bees were exposed to 7 µg of each of the three a.i. in combination (BPT), except for GPX-1 and GST-1, for bees from both feeding management groups (Figures 1 and 2).With the lowest dose of each a.i. in combination (1 µg), upregulation of the genes was observed, except for those involved in the immune system, which were downregulated.When comparing the controls of bee samples from SF and RF colonies, no difference in the expression of any of the genes was observed (Figures 1 and 2).
The expression of SOD-1 was increased by exposure to bix (7 µg) in bees from the SF group.With pro, there was a decrease (with 1 µg) and an increase (with 7 µg) in the expression of this gene for bees from the RF group.Tri led to a decrease in the gene expression of SOD-1 with both doses for bees from the RF group and with the higher dose for bees from the SF group.With the mixture of fungicides, the expression of SOD-1 was increased with 1 µg per bee from the SF group and decreased with 7 µg per bee, in bees from both dietary managements (Figure 1).
For the catalase gene, bix promoted upregulation with both doses in both dietary managements, except in bees from the SF group treated with the lower dose.With pro, there was an increase in the expression of this gene with 7 µg in bees from the RF group.For tri at a higher dose, there was a decrease in catalase expression.With the mixture of the three fungicides, an increase and a decrease in the expression of this gene were observed with the lower and higher doses, respectively, in bees from both dietary managements (Figure 1).
The expression of GPX-1 was upregulated with the use of bix, except for exposure to the lower dose, in bees from the SF group.Pro and tri led to an increase in GPX-1 expression with the use of 7 µg and 1 µg per bee from the RF group, respectively.With the mixture of fungicides at the lower dose, there was an increase in the expression of this gene (Figure 1).
There was an increase in the expression of GST-1 when bees from both dietary managements were exposed to the higher dose of both bix and pro.For tri (with 7 µg), there was a decrease in the expression of this gene in bees from the RF group.With BPT, there was upregulation of GST-1 with a dose of 1 µg per bee from both dietary groups (Figure 1).
Upregulation of the CYP306A1 gene was observed with exposure of bees to all fungicides applied individually, except for the lower dose of bix in bees from the SF group, the lower dose of pro in bees from both dietary managements, and the higher dose of tri for bees from the SF group.BPT promoted upregulation and downregulation of CYP306A1 with the use of 1 µg and 7 µg, respectively, of each active ingredient in bees from both dietary managements (Figure 1).
The expression of vitellogenin was upregulated with the exposure of bees to bix, except for the lower dose applied to bees from the SF group.With pro, upregulation was observed with the dose of 7 µg per bee in both dietary managements.For tri (1 µg/bee), vitellogenin was upregulated in bees from the SF group (Figure 2).With BPT at the lower dose, vitellogenin expression increased in bees from both dietary managements.
Defensin-1 was upregulated by exposure to bix, except for bees from the SF group exposed to the lower dose of this active ingredient.With the lower and higher doses of pro, there was downregulation of this gene in bees from the RF and SF groups.For tri and BPT, there was a decrease in the expression of defensin-1, except for the higher dose of tri in bees from both dietary managements and for bees from the SF group exposed to the lower dose of BPT (Figure 2).
Figure 1.
Expression in the fat body of antioxidant and detoxification genes of honey bee foragers from colonies with supplemented or restricted feeding and exposed or not to bixafen (bix), prothioconazole (pro), trifloxystrobin (tri), and a mixture with all three a.i.(BPT).Dose 1 = 1 µg A. I. bee −1 ; dose 7 = 7 µg bee −1 .Shown are the means of four biological samples per treatment in triplicate.For each gene, means followed by the same letter within the same a.i. or mixture with all three a.i.do not differ significantly from each other (p < 0.05), according to the Tukey test.
The expression of vitellogenin was upregulated with the exposure of bees to bix, except for the lower dose applied to bees from the SF group.With pro, upregulation was Expression in the fat body of antioxidant and detoxification genes of honey bee foragers from colonies with supplemented or restricted feeding and exposed or not to bixafen (bix), prothioconazole (pro), trifloxystrobin (tri), and a mixture with all three a.i.(BPT).Dose 1 = 1 µg A. I. bee −1 ; dose 7 = 7 µg bee −1 .Shown are the means of four biological samples per treatment in triplicate.For each gene, means followed by the same letter within the same a.i. or mixture with all three a.i.do not differ significantly from each other (p < 0.05), according to the Tukey test.
Figure 2.
Expression in the fat body of vitellogenin and immune system antimicrobial peptide genes of honey bee foragers from colonies with supplemented or restricted feeding and exposed or not to bixafen (bix), prothioconazole (pro), trifloxystrobin (tri) or a mixture with all three a.i.(BPT).Dose 1 = 1 µg a. i. bee −1 ; dose 7 = 7 µg bee −1 .Shown are the means of four biological samples per treatment in triplicate For each gene, means followed by the same letter within the same a.i. or mixture with all three a.i.do not differ significantly from each other (p < 0.05), according to the Tukey test.Expression in the fat body of vitellogenin and immune system antimicrobial peptide genes of honey bee foragers from colonies with supplemented or restricted feeding and exposed or not to bixafen (bix), prothioconazole (pro), trifloxystrobin (tri) or a mixture with all three a.i.(BPT).Dose 1 = 1 µg a.i.bee −1 ; dose 7 = 7 µg bee −1 .Shown are the means of four biological samples per treatment in triplicate For each gene, means followed by the same letter within the same a.i. or mixture with all three a.i.do not differ significantly from each other (p < 0.05), according to the Tukey test.
The abaecin gene was downregulated when bees from both dietary managements were exposed to bix, except with the application of the higher dose in bees from the RF group.With pro, the only change was observed with the use of 1 µg per bee in the RF group.Tri promoted downregulation of abaecin, except for bees from the SF group exposed to the lower dose.Both doses of BPT resulted in a decrease in the expression of this gene in bees from both dietary managements (Figure 2).
Hymenoptaecin was upregulated when bees were contaminated with the highest dose of bix.With isolated use of the other two a.i., and the combination of the three a.i. at both doses, hymenoptaecin was downregulated in bees from both feeding management groups, except for bees from the SF group exposed to the lower dose of tri (Figure 2).
The expression of apidaecin was upregulated in bees from both dietary managements, except for those from the SF group exposed to the lower dose.This gene was downregulated with pro and BPT, except for the higher dose of pro.With 7 µg of tri per bee, there was upregulation of apidaecin in bees from the SF group (Figure 2).
Observing the effects of each fungicide on all genes, it was noted that bix promoted a decrease in gene expression only with the lower dose and only for abaecin in the SF group.For bees from the same dietary management exposed to the higher dose of bix, there was a change in the expression of all evaluated genes, with upregulation of all except abaecin, which was downregulated.For bees from the RF group exposed to the lower dose of bix, there was an increase in the expression of six genes (CAT, GPX-1, GST-1, CYP306A1, vitellogenin, defensin-1, and apidaecin).With 7 µg of bix per bee from the RF group, there was upregulation of all genes, except for SOD-1, which was similar to the control, and abaecin, which was downregulated (Figure 2).
Exposure of bees to the lower dose of pro promoted a decrease in the expression of the hymenoptaecin and apidaecin genes in bees from the SF group and of all antimicrobial peptide genes in the RF group.With the higher dose, in the SF group, there was a change in the expression of GST-1, CYP306A1, vitellogenin, defensin-1, and hymenoptaecin.In bees from the RF group exposed to 7 µg of pro, the expression of all genes was altered except defensin-1 and hymenoptaecin (Figure 2).
Tri promoted alteration of expression of six genes in bees from the SF group, with three genes affected by the lower dose (CYP306A1, vitellogenin, and defensin-1) and three by the higher dose (abaecin, hymenoptaecin, and apidaecin).For bees from the RF group, there was a change in the expression of all genes except CAT, GST-1, vitellogenin, and apidaecin, when these bees were individually exposed to 1 µg of tri.With the higher dose of tri, the expression of GPX-1, vitellogenin, defensin-1, and apidaecin was altered in bees from the RF group (Figure 2).
In the bees exposed to the mixture of the three fungicides (BPT), a greater number of gene expression alterations was observed.For bees from the SF group, there was no difference compared with the control (without fungicide) for defensin-1 (with 1 µg) and GST-1 (with 7 µg).For bees from the RF group, there was no alteration in the expression of the genes SOD-1 (with 1 µg), GPX-1 (with 7 µg), and GST-1 (with 7 µg).Considering all comparisons of the treatments within each group with the respective control, it was observed that bees from the RF group showed 50% more gene expression alterations than bees from the SF group.
Colony Feeding Management
The use of pollen traps reduced the availability of this protein food for the bees.The mean total amount of bee-collected pollen per colony during the 15 weeks was 507 g, which was equivalent to 57.9% of the total bee-collected pollen brought into the hives by the bees of this group.This quantity is relatively small considering that up to 20 kg per hive per year can be obtained [41].Maintaining pollen traps in hives for several weeks can result in a reduction in the quantities of collected pollen due to the population adjustment made by the bees [42].However, in this study, the amount of bee-collected pollen harvested was relatively small from the early days of food restriction, indicating low availability of pollen in the field [26].
Unlike other livestock, beekeeping does not require routine feeding of colonies, except during periods of heavy rains, severe winters, and scarcity of nectar and pollen in the field [43].Our study was conducted in an experimental apiary used for several years, with a known natural food flow, which is characterized by low availability of nectar and pollen in most months.
Among the five colonies subjected to food restriction management, one succumbed.The others survived, though with visibly reduced populations, relying on the nectar and pollen foragers managed to bring into the colonies, passing through the pollen traps.For the other group (SF), providing protein paste made of honey and bee-collected pollen, which are foods that bees naturally consume, may have contributed to the acceptance of this paste, which was completely consumed by the bees.Adequate diets can contribute to immunocompetence and resistance to pathogens [44,45] and pesticide tolerance [46].Bees from the SF group consumed all the energy syrup offered twice a week.This nectar substitute, in association with the protein supplement, may promote a search for more protein food in the field, increased queen egg laying, and worker hygienic behavior [43].
Gene Expression Changes in Bees from Colonies with Supplemented or Reduced Feeding
According to the safety data sheet of the commercial fungicide containing the active ingredients bix, pro, and tri [47], the contact LD 50 for honey bees is greater than 200 µg per individual.In our study, for treatments with the mixture of all three a.i., the doses used (1 or 7 µg per individual of each a.i.) can be considered sublethal, as they are approximately 10 and four times lower than the lethal dose reported by the manufacturer, respectively.
Christen et al. ( 2019) [48] evaluated the toxic effect of azoxystrobin, a fungicide belonging to the strobilurin group, which includes tri.They observed downregulation of genes encoding enzymes involved in metabolism, oxidative phosphorylation, and hormonal regulation, which could affect energy production, ontogeny, and behavior of bees.Tri had already been detected in bee-collected pollen obtained from cultivated plants and in bee-collected pollen from nearby wild plants [49].Bix was detected in samples of bee bread [50], and pro has been detected in bee-collected pollen [51][52][53].Information regarding bee exposure through contact is limited.
Studying gene expression in bees can aid in understanding the effects of exposure to stressors such as pesticides.In our study, genes related to oxidative stress and detoxification (SOD-1, CAT, GPX-1, GST-1, and CYP306A1), nutrition and longevity (vitellogenin), and immunity (defensin-1, abaecin, hymenoptaecin, and apidaecin) were chosen.The metabolism of bees, like other aerobic species, involves the formation of free radicals.When these molecules are not properly processed to favor cellular homeostasis, oxidative stress can occur [54].
Regulation and inactivation of free radicals are carried out by the antioxidant system.This process is natural due to the metabolism of oxygen and other substances metabolized by bees; however, biotic and abiotic stressors can disrupt cellular homeostasis, generating more reactive oxygen species than the individual can neutralize and eliminate [55,56].The fungicides chlorothalonil, azoxystrobin, and folpet, when administered to adult bees at sublethal doses in syrup, caused transcriptional alterations in genes encoding enzymes involved in oxidative phosphorylation and metabolism.Among the fungicides, chlorothalonil had the most significant impact, reducing the transcription of genes related to energy production, metabolism, and the endocrine system.Disrupted energy production may decrease foraging activity and cause hormonal imbalances, affecting the transition of nurse bees to foragers [48].
Sublethal doses of insecticides such as organochlorines and organophosphates [57,58] and neonicotinoids [59] may lead to increased production of antioxidant enzymes such as SOD-1, CAT, and GPX, aiming to maintain cellular homeostasis.However, higher pesticide doses and specific characteristics of the xenobiotic can impair the production of these enzymes [60].Bix and tri are inhibitors of the respiratory chain, and their action may be unfavorable for the production of antioxidant enzymes [61,62].
The gene CYP306A1 belongs to the cytochrome P450 family, which is involved in various cellular biosynthesis and detoxification processes that are particularly important when there is exposure to xenobiotics.This orthologous gene is known for its involvement in ecdysteroid biosynthesis and had its detoxification function described in the silverleaf whitefly Bemisia tabaci (Aleyrodidae) exposed to the insecticide imidacloprid [63].Upregulation of CYP6 genes is associated with resistance to pyrethroids and neonicotinoids in other insects.It is suggested that these genes may be useful for bees to increase detoxification activity when there is pesticide contamination [64].In our study, expression of the CYP306A1 gene was upregulated with all active ingredients used individually, except for pro and tri at the lowest dose, for bees from the SF and RF groups, respectively.
In another study [65], 121 pesticide contaminants found in bees were tested to determine their effect on the active site of CYP9Q1, a broadly substrate-specific P450 with high quercetin-metabolizing activity that also metabolizes pesticides.Six triazole fungicides were identified, all fungal P450 inhibitors, that fit into the catalytic site.In adults fed combinations of quercetin and the triazole myclobutanil, the expression of five out of six mitochondrion-related nuclear genes was downregulated.Adult bees consuming quercetin with myclobutanil metabolized less quercetin and produced less thoracic ATP.The authors highlight that, although fungicides lack acute toxicity, they may affect bee health by interfering with detoxification by quercetin, compromising mitochondrial function and ATP production.
Vitellogenin is a storage protein, and vitellogenin gene expression is affected by the nutritional state of the individual.It can be considered a pleiotropic gene because, in addition to assisting in lipid transport, it is related to longevity, immunomodulation, and regulation of oxidative stress [66].In our study, exposure to any one of the three a.i.promoted upregulation of vitellogenin, except for pro at the lowest dose and tri at the highest dose, for bees from the RF group.
Regarding the genes strictly linked to the bees' immune systems, there was evident alteration in gene expression.There was a change in relative immune gene expression in 63% of the comparisons with the respective control for the SF group of bees and 72% for the RF group.For the other genes, there was a change in expression in 44% and 65% of the comparisons, respectively, for SF and RF.Thus, it is understood that the immune system was more activated in well-fed bees to mitigate the metabolic stress caused by the fungicides, especially when used in a mixture, even at the lowest concentration of each (1 µg).Comparing the gene expression results obtained from bees from colonies subjected to different nutritional managements, it was observed that the mitigating effect from nutritional supplementation was less for genes related to the immune system.However, to substantiate the effects of nutrition on the health of bees exposed to fungicides through contact, additional assessments should be conducted, such as longevity, behavior, and other physiological variables.
Gene expression is a process that occurs at the expense of energy produced in cells.With a need for increased expression of a certain gene for the cell to continue functioning in homeostasis, more energy is demanded [67].However, even with a higher demand for gene expression so that the harmful effects of xenobiotics do not impair cellular functioning, energy availability can be a determining factor in the response to an intoxication challenge.Of the three a.i.evaluated in this study, bix and tri are known to act as inhibitors of cellular energy production [61,62,68].With the exposure of bees to these a.i. at the highest evaluated dose, cellular energy demand may not have been met, and thus, gene expression was impaired.As a consequence, bees may become more vulnerable to other stressors, given the fragility resulting from exposure to the fungicide with all three a.i.
The exposure of bees to multiple pesticides simultaneously can result in synergism compared with the effects of the individual active ingredients.The mixture of an insecticide (thiacloprid) with a fungicide (cyproconazole) led to synergistic toxicity in bees, including the expression of the genes CYP306A1, CYP6AS14, apidaecin, defensin-2, and vitellogenin [69].The response to the challenge with bix, pro, and tri, whether isolated or in combination, was somewhat different among bees that had supplemented versus restricted feeding for most comparisons.
Figure 1 .
Figure1.Expression in the fat body of antioxidant and detoxification genes of honey bee foragers from colonies with supplemented or restricted feeding and exposed or not to bixafen (bix), prothioconazole (pro), trifloxystrobin (tri), and a mixture with all three a.i.(BPT).Dose 1 = 1 µg A. I. bee −1 ; dose 7 = 7 µg bee −1 .Shown are the means of four biological samples per treatment in triplicate.For each gene, means followed by the same letter within the same a.i. or mixture with all three a.i.do not differ significantly from each other (p < 0.05), according to the Tukey test.
Figure 2 .
Figure 2. Expression in the fat body of vitellogenin and immune system antimicrobial peptide genes of honey bee foragers from colonies with supplemented or restricted feeding and exposed or not to bixafen (bix), prothioconazole (pro), trifloxystrobin (tri) or a mixture with all three a.i.(BPT).Dose 1 = 1 µg a.i.bee −1 ; dose 7 = 7 µg bee −1 .Shown are the means of four biological samples per treatment in triplicate For each gene, means followed by the same letter within the same a.i. or mixture with all three a.i.do not differ significantly from each other (p < 0.05), according to the Tukey test. | 2024-07-15T15:35:16.408Z | 2024-07-01T00:00:00.000 | {
"year": 2024,
"sha1": "96325a141efeb5f060089aba77a88f089a0b56e4",
"oa_license": "CCBY",
"oa_url": null,
"oa_status": null,
"pdf_src": "PubMedCentral",
"pdf_hash": "aed80c384cf6b05cf1f0fb2597bc97bdff0f3365",
"s2fieldsofstudy": [
"Environmental Science",
"Biology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
59460170 | pes2o/s2orc | v3-fos-license | Long-range correlations in ALICE at the LHC
Long-range correlations between particles separated by a pseudorapidity gap are a powerful tool to explore the initial stages and evolution of the medium created in hadron-hadron collisions. An overview of the long-range correlations measured by the ALICE detector in pp, p-Pb and Pb-Pb will be presented. This includes analyses of forward-backward, two- and multi-particle correlations with the use of the central barrel and forward detectors.
Introduction
Long-range correlations (LRC) are usually considered between particles separated by pseudorapidity gap, that is typically taken to be |∆η| 1.0 in order to suppress contribution from resonances and (mini)jets. LRC can be created only at the early stages of the collision [1] and arise in Color Glass Condensate [2] and string fusion [3] models. On later stages of the system evolution they can be modified by medium and final state interactions (hydrodynamic expansion, energy loss in medium, conservation laws).
How do we extract information about long-range correlations? Different analysis methods are being used, which have different sensitivity to various aspects of physical origin of LRC. Two-particle correlations with large separating η gaps between pairs of particles allows one to get rid of short-range correlations and obtain the correlation pattern, azimuthal profile of which can be decomposed into Fourier series. Another technique, which is sensitive to collective phenomena, uses multiparticle correlations, for example, elliptic flow can be measured by calculating second-order Fourier coefficient v 2 taking 4, 6, 8 or even all particles in event (usually . Even with 4 particles, non-flow contribution is already suppressed enough, allowing one to measure collective effects during evolution of the system created in a hadronic collision. Forward-backward correlation analysis is another method sensitive to even-by-event fluctuations of number and properties of particle-emitting sources elongated in rapidity.
LRC are measured in Pb-Pb, p-Pb and pp collisions in all major experiments at the LHC. This article collects some experimental highlights on LRC from ALICE.
ALICE detector
In ALICE experimental setup, charged primary particles are reconstructed with the central barrel detectors combining information from the Inner Tracking System (ITS) and the Time Projection Chamber (TPC). Both detectors are located inside the ALICE solenoid with a field of 0.5 T and have full azimuthal coverage for track reconstruction within a pseudo-rapidity window of |η| < 0.9. Charged particle identification (PID) over a broad momentum range is done using information from the TPC and the Time of Flight (TOF) detectors. The TPC provides a simultaneous measurement of the momentum of a particle and its specific ionisation energy loss (dE/dx) in the gas. The TOF detector surrounds the TPC and provides separation between π-K and K-p up to p T = 2.5 GeV/c and p T = 4 GeV/c, respectively, by measuring the arrival time of particles. VZERO detectors, two forward scintillator arrays with coverage −3.7 < η < −1.7 and 2.8 < η < 5.1, are used in the trigger logic and for the centrality and reference flow particle determination [4].
Two-particle correlations with pseudorapidity gap
The two-particle correlation function C(∆η, ∆ϕ) measured in central 0-10% Pb-Pb collisions, is shown in figure 1 (a), demonstrating such structures as "near-side" peak from jets and resonances, "near-side ridge" along ∆η at ∆ϕ ≈ 0 and broad away-side structure elongated in ∆η [5]. The measured anisotropy of particle production at n-th harmonic order is given by coefficient V n∆ , which can be obtained from triggered, pseudorapidity-separated (|η| > 0.8) pair azimuthal correlations using expression dN pairs and p a T are transverse momenta of trigger and associated particles. In very central events 0-2% (figure 1, b), the away side exhibits a concave, doubly-peaked feature (V 3∆ > V 2∆ ), that is usually attributed to fluctuations in initial state geometry which can generate higher-order flow components. It was checked whether a set of single-valued points v n (p T ) can be identified that describe the measured long-range anisotropy via the . If so, V n ∆ is said to factorize into single-particle Fourier coefficients. These pair anisotropies are found to approximately factorize into single-particle harmonic coefficients for p a T < 4 GeV/c. This factorization is consistent with the picture of collective response of the medium to anisotropic initial conditions. Figure 1. Left: two-particle correlation function C(∆η, ∆ϕ) for central 0-10% Pb-Pb collisions at 2-4 p T range. Right: C(∆ϕ) for particle pairs at |η| > 0.8 in most central 0-2% events. The Fourier harmonics for V 1 to V 5 are superimposed in color [5]. 4. v 2 for identified particles in different p T ranges Elliptic flow (v 2 ) in Pb-Pb collisions was measured with the Scalar Product method, which is a two-particle correlation technique, using a pseudo-rapidity gap of |η| > 0.9 between reference particles in VZERO scintillators and the identified hadrons of interest within TPC [6]. Dependence of v 2 on p T in centrality class 10-20% is shown in figure 2 (a) for π, K, p, Λ, φ, Ξ and Ω. Shift of the v 2 (p T ) dependences towards higher p T is observed for heavier particles due to interplay between elliptic and radial flow ("mass ordering") up to 3 GeV/c. Mass ordering is stronger in most central collisions because of stronger radial flow. Particles with p T > 3 GeV/c tend to group according to their type, i.e. mesons and baryons (except for φ meson). In this range of p T , quark coalescence was suggested to be the dominant hadronization mechanism, however, ALICE data exhibit deviations from the number of constituent quark (NCQ) scaling at the level of ±20%.
Mass ordering was observed also for π, K and p in p-Pb collisions, when v 2 was extracted from two-particle correlations in high-multiplicity events after subtraction of correlations in low multiplicity events (figure 2, b) [7]. Qualitatively similar picture in p-Pb as in Pb-Pb collisions suggests common origin of long-range correlations in A-A and in smaller systems. This question currently is under intensive debates and investigation. Further studies of long-range correlation structures were performed in p-Pb collisions using inclusive muons from muon spectrometer as trigger particles (with 0.5 < p T < 4 GeV/c, −4 < η < −2.5) and associated charged particles in ITS+TPC (|η| < 1, 0.5 < p T < 4 GeV/c) [8].
Reconstructed muons mainly originate from weak decays of π, K and mesons from heavy flavor (HF) decays. Figure 3 (a) shows yield of associated particles in (∆η, ∆ϕ) coordinates in 0-20% event classes after subtraction of 60-100% class. It can be seen that near-side ridge extends up to |∆η| ≈ 5 and it decreases from η = −1.5 to η = −5.0. Analysis was done at both beam directions p-Pb and Pb-p, and, assuming factorization of the Fourier coefficients at central and forward rapidity, extracted v 2 coefficients were found to be larger by 16 ± 6 when muons are measured in Pb-going direction, rather than p-going ( figure 3, b). Calculations using AMPT event generator showed qualitatively similar behavior at low p T . In order to get more insight into particle production mechanisms, conventional analysis of the FB multiplicity correlations between two η-intervals was extended into azimuthal dimension. Figure 5 for two energies √ s = 0.9 and 7 TeV shows that correlation pattern in η-ϕ space can be split into short-range peak and non-zero plateau which spreads over all studied η-ϕ separations of FB windows (underlying long-range correlation). This non-zero plateau is found to increase with the collision energy. In a simple model of independent particle emitters [10] this rise is attributed to increase of fluctuations in number of independent sources (strings).
Final remarks
Long-range correlations provide insights into initial stages and details of the evolution of hadronic collisions. The correlations are being studied at ALICE in pp, p-Pb, Pb-Pb collisions by several analysis methods. Two-particle correlations with separating η gap reveal azimuthal correlation structures, like double-hump shape in central Pb-Pb collisions. Clear mass ordering at p T 3 GeV/c in p-Pb, Pb-Pb events is consistent with the picture of hydrodynamic expansion of the medium in A-A collisions, and also states the question about origin of the flow in small systems. Analysis of muon-hadron correlations in p-Pb collisions shows that long-range doubleridge persists up to ∆η ≈ 5. Forward-backward correlations in pp provide access to fluctuations in number and properties of particle-emitting sources.
With the limited central barrel acceptance for tracking, ALICE can measure long-range correlations with other, non-tracking detecting systems. For example, η-dependence of the anisotropic flow in Pb-Pb collisions was measured recently with two-and four-particle correlations using forward detectors in a range −3.5 < η < 5 [11]. After upgrade in 2019-2020, ALICE will have a new Inner Tracking System with larger η coverage and also a new Muon Forward Tracker (MFT). Combined ITS+MFT acceptance −3.6 < η < 1.2 will significantly extend possibilities to study LRC of charged particles. | 2016-11-30T11:04:28.000Z | 2016-11-30T00:00:00.000 | {
"year": 2017,
"sha1": "ba727c15411685b764744d35130ded71bceb4b9e",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1742-6596/798/1/012056",
"oa_status": "GOLD",
"pdf_src": "Arxiv",
"pdf_hash": "050d689ceadfff0e1041cefd4e7f3c9c7b23c721",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
16722439 | pes2o/s2orc | v3-fos-license | Quality of Life in HIV Clinical Trials: Why Sexual Health Must Not Be Ignored
Currently more than 20 antiretrovirals are commercially available for treatment of HIV infection [1]. Using them in combination therapy, we are able to treat HIV-infected patients with highly potent regimens and suppress the virus below detectable levels. Correct adherence to these regimens is important to ensure that the viral load will remain undetectable and that the disease does not progress. Moreover, even if adherence is not perfect and the virus becomes resistant to a first-line regimen, there are other treatment options to suppress the virus again to undetectable levels. Given the choice of several different potent antiretroviral regimens, doctors and patients will tend to prefer those regimens that are easy to take and that have limited short-term and long-term side effects. One potential side effect of antiretroviral treatment that has received very little scientific attention so far is sexual dysfunction. Sexual dysfunction is defined as difficulty during any stage of the sexual act, including desire, arousal, orgasm, and resolution, that prevents the individual or couple from enjoying sexual activity [2], and, according to DSM-IV (the Diagnostic and Statistical Manual of Mental Disorders, fourth edition) criteria, causing ‘‘marked level of distress or interpersonal difficulty’’ [3]. Sexual dysfunction disorders are generally classified into four categories: sexual desire disorders, sexual arousal disorders, orgasm disorders, and sexual pain disorders [2,3]. Most data relating to the association between antiretroviral therapy and sexual dysfunction are based on cross-sectional studies or case series, and have emerged from industrialised countries [4–12]. Some studies have suggested that antiretrovirals, in particular certain protease inhibitors, may cause sexual problems or dysfunction [4–10]; however, other studies have not confirmed this observation [11,12]. Furthermore, sexual dysfunctions in people infected with HIV may be caused by many different factors. Both organic and psychological factors have been identified, including coping with HIV, pre-existing sexual dysfunctions, sex hormone abnormalities, neuropathy from HIV itself (or related treatment for HIVcaused illnesses), and other iatrogenic causes [13]. Sexual dysfunctions are conceptualised as one component of sexual health, which is an essential element of overall healthrelated quality of life (HRQOL). While it may include a variety of diverse issues, comprehensive definitions such as the World Health Organization’s (WHO) working definition define sexual health as a state of physical, emotional, mental, and social well-being, and not merely the absence of disease, dysfunction, or infirmity. Sexual health encompasses the possibility of having pleasurable and safe sexual experiences [14]. The essence of such lengthy definitions has been summarised in one single statement as ‘‘the enjoyment of sexual activity of one’s choice, without causing or suffering physTable 1. Overview of the Most Commonly Used Questionnaires in Clinical HIV Trials
Currently more than 20 antiretrovirals are commercially available for treatment of HIV infection [1]. Using them in combination therapy, we are able to treat HIV-infected patients with highly potent regimens and suppress the virus below detectable levels. Correct adherence to these regimens is important to ensure that the viral load will remain undetectable and that the disease does not progress. Moreover, even if adherence is not perfect and the virus becomes resistant to a first-line regimen, there are other treatment options to suppress the virus again to undetectable levels.
Given the choice of several different potent antiretroviral regimens, doctors and patients will tend to prefer those regimens that are easy to take and that have limited short-term and long-term side effects.
One potential side effect of antiretroviral treatment that has received very little scientific attention so far is sexual dysfunction. Sexual dysfunction is defined as difficulty during any stage of the sexual act, including desire, arousal, orgasm, and resolution, that prevents the individual or couple from enjoying sexual activity [2], and, according to DSM-IV (the Diagnostic and Statistical Manual of Mental Disorders, fourth edition) criteria, causing ''marked level of distress or interpersonal difficulty'' [3]. Sexual dysfunction disorders are generally classified into four categories: sexual desire disorders, sexual arousal disorders, orgasm disorders, and sexual pain disorders [2,3].
Furthermore, sexual dysfunctions in people infected with HIV may be caused by many different factors. Both organic and psychological factors have been identified, including coping with HIV, pre-existing sexual dysfunctions, sex hormone abnormalities, neuropathy from HIV itself (or related treatment for HIVcaused illnesses), and other iatrogenic causes [13].
Sexual dysfunctions are conceptualised as one component of sexual health, which is an essential element of overall healthrelated quality of life (HRQOL). While it may include a variety of diverse issues, comprehensive definitions such as the World Health Organization's (WHO) working definition define sexual health as a state of physical, emotional, mental, and social well-being, and not merely the absence of disease, dysfunction, or infirmity. Sexual health encompasses the possibility of having pleasurable and safe sexual experiences [14]. The essence of such lengthy definitions has been summarised in one single statement as ''the enjoyment of sexual activity of one's choice, without causing or suffering phys- The EuroQoL (EQ-5D) describes states of health in five dimensions: mobility, self care, usual activities, pain or discomfort, and anxiety or depression.
SF-36
The SF-36 includes one multi-item scale that assesses eight health concepts: limitations in physical activities because of health problems, limitations in social activities because of physical or emotional problems, limitations in usual role activities because of physical health problems, bodily pain, general mental health (psychological distress and well-being), limitations in usual role activities because of emotional problems, vitality (energy and fatigue), and general health perceptions.
WHOQOL-HIV
The WHOQOL-HIV is a 125 item questionnaire on the same six dimensions as the WHOQOL-100 questionnaire [20]:
PL o S CLINICAL TRIALS
The Essay section contains opinion pieces on topics of broad interest to a general medical audience.
ical or mental harm'' [15]. The latter may pose a particular challenge for people living with HIV/AIDS (PLHA) due to HIVrelated stigma and the psychological impact of HIV. As such, sexual health is clearly linked with perceived overall quality of life.
Ideally, we should assess sexual health, including sexual functioning, during randomised clinical trials (RCTs) comparing different antiretroviral regimens, early versus late initiation of antiretroviral therapy, and treatment interruption studies. This could be done by including questions on sexual health in quality of life questionnaires that are administered during such clinical trials. However, quality of life questionnaires that are currently used in HIV clinical trials usually do not include such questions.
Questionnaires Used in HIV Clinical Trials
The quality of life questionnaires used most widely in HIV clinical trials are the MOS-HIV questionnaire [16] and to a lesser extent the EuroQoL (EQ-5D) questionnaire [17] (Table 1).
The MOS-HIV is a brief (generally taking less than ten minutes) comprehensive health status measure based on questions from the Medical Outcomes Study, a large multi-site study of the effects of different ways of delivering medical care [21]. The MOS-HIV was one of the first disease-targeted quality of life questionnaires available for PLHA, and is widely used in clinical trials and other research and evaluation studies. The EQ-5D is a simple instrument assessing only five dimensions of health. In both questionnaires, MOS-HIV and EQ-5D, there are no questions concerning sexual health.
The most widely used health status instrument in other domains is the SF-36 [18] (which is also derived from the Medical Outcomes Study). Other questionnaires assessing quality of life are the quality of well-being scale (QWBS) [22] and the Health Utilities Index (HUI) questionnaire [23]. None of these questionnaires include sexual health questions.
The only questionnaire that does include sexual health questions is the WHO QOL-HIV questionnaire [19]. This questionnaire is based on the WHOQOL-100 questionnaire [20]. Among 120 questions included in this questionnaire, five questions address sexual health and one specifically sexual dysfunction: ''Are you bothered by any difficulties in your sex life?'' (Box 1).
One hundred and twenty questions pose quite a heavy burden on both the interviewer and the patient/interviewee. In addition, given the sensitive character of questions on sexual health, one may anticipate that these questions would be the first ones to be omitted from a questionnaire during an assessment.
Why Should Sexual Health Be Assessed During HIV Clinical Trials?
The importance of sex in most people's lives is illustrated by the fact that there is a growing market for drugs that increase sexual pleasure. The pharmaceutical industry has recently discovered this market, which has resulted in the development of popular erectile dysfunction medications such as sildenafil (Viagra), tadalafil (Cialis), and vardenafil (Levitra) (Figure 1). Evidence from studies in the United Kingdom and United States [24][25][26] show that these drugs are also commonly used among HIV-infected people, particularly by men who have sex with men (MSM).
A recent study done in the US reported that 11% of MSM used Viagra in the 12 months preceding the interview [26]. Only a proportion of these medications are used for erectile dysfunction; recreational use of these drugs to improve sexual performance is also common.
The annual revenues generated by sales of Viagra and Cialis were respectively over $1 billion and $747 million in 2005 [27,28]. Pharmaceutical companies even have been accused of inventing new diseases, such as sexual dysfunction in women, in order to create new drug markets [29,30]. Other drugs that are widely used to improve libido or sexual pleasure are often unapproved indications of approved drugs (e.g., testosterone and amyl nitrate or ''poppers'') or re-stricted (i.e., illegal) drugs (e.g., crystal methamphetamine) [26,31].
This demonstrates a clear discrepancy between the widespread medicalisation of sexual behaviour and the lack of research in this field. A systematic search of 9,979 abstracts of the 2006 international AIDS conference in Toronto using the keyword ''libido'' resulted in four abstracts, while searching for the term ''sexual dysfunction'' resulted in one abstract. However, a search for the keyword ''resistance'' produced 486 abstracts.
We believe that sexual dysfunction needs more attention from researchers developing clinical trials, since it seems to be quite important from the patients' perspective. Focus group discussions among PLHA performed in Europe have shown that sexual well-being in general was felt to be a crucial part of overall HRQOL, and that sexual dysfunction was perceived to decrease it significantly [32]. PLHA are confronted with many factors that may interfere with their sexual wellbeing: the psychological impact of the HIV infection itself, the stigma associated with the infection, hormonal abnormalities, fear of transmitting the infection to others, depression, illnesses, and the side effects of drugs such as antiretrovirals. With respect to the latter, PLHA often attribute sexual problems specifically to prescribed drugs which they are taking, and this may ultimately compromise both adherence and secondary prevention. In one study in Italy, sexual dysfunction was associated with poorer adherence to protease inhibitor-containing antiretroviral treatment regimens [33].
Moreover, many drugs have been shown to potentially interfere with sexual functioning (e.g., anti-hypertensive medication, anti-depressive medication, etc.) [34]; thus sexual health questions should be included in all QOL questionnaires during trials with new drugs.
Conclusion
Sexual health is an important contributing factor to overall quality of life; therefore we propose that specific ques- Moreover, a specific validated questionnaire needs to be developed for studying HIV-related sexual health problems, not only for use in industrialised countries but also in countries with limited resources.
Obviously, besides the importance of posing sexual health questions in HRQOL questionnaires, asking patients about their sexual health during consultations remains essential. While discussing such topics in the context of safer sex practices may increase the quality of life for individual patients, from a public health point of view it can be viewed as contributing to primary HIV prevention. & | 2016-05-04T20:20:58.661Z | 2007-03-01T00:00:00.000 | {
"year": 2007,
"sha1": "373a47fdd2aa2c4d1324f4b6ec02f8e65ab155ed",
"oa_license": "CCBY",
"oa_url": "https://journals.plos.org/plosclinicaltrials/article/file?id=10.1371/journal.pctr.0020008&type=printable",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "adf55f3923315edb526726882507380f008469fa",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
116528090 | pes2o/s2orc | v3-fos-license | Stable and unStable milling proceSS for nickel Superalloy aS obServed by recurrence plotS and multiScale entropy
Nickel alloys are often used in different branches of industry, from medical to aerospace industries. Nickel-based alloys are known to have very good strength and temperature resistance, therefore they are called superalloys. Due to their strength properties, superalloys are very difficult to machine. Additionally, the cutting of these materials can generate harmful vibrations during machining. This poses a serious problem for engineers. Undesired relative vibrations between the tool and the workpiece may deteriorate the quality of machined surfaces or even damage the machine tool and the workpiece. As a result, the cutting forces that depend on the tool geometry, material properties, feed rate and cutting speed can have a large amplitude, which leads to faster tool wear. These vibrations of the tool are known as chatter [39, 40]. In order to use the full capacity of a new fast machine and to achieve a potentially high material removal rate together with the desired surface quality, optimum machining parameters are necessary. The fundamental parameters which may improve efficiency of the cutting process include cutting depth and velocity. Usually, the selection of cutting depth and spindle rotational speed is made via Stability Lobes Diagram (SLD) which can be applied in high-speed machining (HSM) processes to optimize the maximum depth of cut at the highest available spindle speed. When the cutting depth exceeds the critical value, chatter vibrations can occur at some spindle speed, whereas if the cutting depth is below the critical value, the process is stable regardless of the spindle speed. Generally, in practice, the selection of optimal speed and depth of cut is difficult. Classical SLDs can be obtained by modal analysis of the tool-spindle system; nevertheless, in many research papers the milling process is first modelled and then the numerical results are used to determine SLD (for example [11, 15]. An alternative solution is to calculate stability lobes directly from delay-differential equations [22]. However, only a few papers report complete experimental verification of these stability lobes [24]. The literature reports numerous analytical, numerical and experimental methods for cutting stability prediction. For instance, Altintas and Budak [2] describe an analytical method for predicting milling stability lobes based on a mean of the Fourier series of dynamic milling coefficients. This method is fast but cannot predict additional stability regions and period doubling bifurcation for a low radial depth of cut [8]. To overcome this problem, a multi-frequency solution of chatter stability was developed by Budak and Altintas [3] and then extended by Merdol and Altintas [3, 27]. One of the most popular numerical methods for chatter prediction is the Finite Element Method (FEM) [1, 10, 23, 38]. Bayly et al. [4] propose the use of a temporal finite element analysis for milling and interrupted turning [13]. Moreover, Voronov [38] and especially Adetoro et al. [1] propose an improved model of the classical milling process. This new Andrzej Weremczuk marek BoroWiec michał rudzik rafał rusinek
Introduction
Nickel alloys are often used in different branches of industry, from medical to aerospace industries. Nickel-based alloys are known to have very good strength and temperature resistance, therefore they are called superalloys. Due to their strength properties, superalloys are very difficult to machine. Additionally, the cutting of these materials can generate harmful vibrations during machining. This poses a serious problem for engineers. Undesired relative vibrations between the tool and the workpiece may deteriorate the quality of machined surfaces or even damage the machine tool and the workpiece. As a result, the cutting forces that depend on the tool geometry, material properties, feed rate and cutting speed can have a large amplitude, which leads to faster tool wear. These vibrations of the tool are known as chatter [39,40]. In order to use the full capacity of a new fast machine and to achieve a potentially high material removal rate together with the desired surface quality, optimum machining parameters are necessary. The fundamental parameters which may improve efficiency of the cutting process include cutting depth and velocity. Usually, the selection of cutting depth and spindle rotational speed is made via Stability Lobes Diagram (SLD) which can be applied in high-speed machining (HSM) processes to optimize the maximum depth of cut at the highest available spindle speed. When the cutting depth exceeds the critical value, chatter vibrations can occur at some spindle speed, whereas if the cutting depth is below the critical value, the process is stable regardless of the spindle speed. Generally, in practice, the selection of optimal speed and depth of cut is difficult. Classical SLDs can be obtained by modal analysis of the tool-spindle system; nevertheless, in many research papers the milling process is first modelled and then the numerical results are used to determine SLD (for example [11,15]. An alternative solution is to calculate stability lobes directly from delay-differential equations [22]. However, only a few papers report complete experimental verification of these stability lobes [24]. The literature reports numerous analytical, numerical and experimental methods for cutting stability prediction. For instance, Altintas and Budak [2] describe an analytical method for predicting milling stability lobes based on a mean of the Fourier series of dynamic milling coefficients. This method is fast but cannot predict additional stability regions and period doubling bifurcation for a low radial depth of cut [8]. To overcome this problem, a multi-frequency solution of chatter stability was developed by Budak and Altintas [3] and then extended by Merdol and Altintas [3,27]. One of the most popular numerical methods for chatter prediction is the Finite Element Method (FEM) [1,10,23,38]. Bayly et al. [4] propose the use of a temporal finite element analysis for milling and interrupted turning [13]. Moreover, Voronov [38] and especially Adetoro et al. [1] propose an improved model of the classical milling process. This new sciENcE aNd tEchNology model takes into account the well-known model and combines it with considerations about the nonlinearity of cutting force coefficients and the axial immersion angle as a function of the axial depth of cut. The model is validated and the theoretical findings show a very good agreement with the experimental results [1]. Although experimental findings are very often compared with FEM results [13], the problem of experimental signal analysis remains an open question. Various methods of signal analysis are applied to recognize chatter vibrations in cutting operations, including multifractal and wavelet approaches [18], multiscale entropy [20], Hilbert Huang transform [32,19], recurrence analysis [33,31], flicker-noise spectroscopy [21] and audio signal analysis [38]. In [7] the delay space reconstruction method is applied to show that the workpiece motion is characterized by fractal geometry. The auto-bispectra suggest a quadratic phase coupling among the spectral peaks associated with the cutter frequency. Finally, the authors propose a mechanics-based model with impact to explain the obtained results. Their predictions agree well with the experimental observations. Other researchers have noticed that the dynamic behaviour of the tool-workpiece system depends on the tool position in the workpiece. This phenomenon and the problem of the influence of several modes on stability lobes is discussed in [37]. On the other hand, modal interaction can affect surface roughness, particularly in the milling of thin-walled structures [34].
Numerical analysis of the cutting process is also useful for stability analysis. However, the models of cutting processes are described by delay differential equations, which may pose some problems in numerical calculations. To accelerate the integration procedure and overcome the difficulties, semi-or full-discretization methods for predicting milling stability are proposed in several papers. The semi-discretization method developed by Insperger and Stepan is an efficient numerical technique for stability analysis of linear delayed systems. Nonetheless, Ding et al. [8] propose a more effective full-discretization method for prediction of milling stability. Insperger [12] proposes the so-called act-and-wait concept for continuous-time control systems with feedback delay.
In the literature one can find a number of publications with practical approach to the problem of cutting efficiency. Wojciechowski et. al. [42] propose a method ensuring the reduction of forces and the improvement of efficiency in the finish ball end milling of hardened 55NiCrMoV6 steel. The primary objective of the paper is optimal selection of milling parameters (cutting speed and surface inclination angle), as this will ensure minimal cutting force values and increased process efficiency at the same time. From a practical point of view, surface roughness is very often analysed. Roughness depends on different factors including feed direction, axial and radial run-out errors, and cutting tool geometry. In the paper [45] an algorithm considering the effects of static and dynamic factors on surface roughness is proposed. A new approach to surface roughness parameters estimation during finish cylindrical end milling is presented in [41]. The proposed model takes account of the influence of cutting parameters, the tool's static runout and dynamic phenomena related to instantaneous tool deflection.
An interesting contribution to the problem is a numerical and experimental study of the dynamics of flank milling operations at low cutting rates presented in [28].The paper focuses on both the properties of cutting vibratory phenomena and their impact on the roughness of the machined surface. The study is based on a one-degree-of-freedom model of a mechanical machining system. The system consists of a rigid cutter and a flexible workpiece. The cutting force model is based on a regenerative mechanism. The roughness of the surface machined at high speed revolutions is studied for both forced vibrations occurring during stable cutting and self-excited vibrations occurring during unstable cutting. It is shown that forced vibrations have only a very slight impact on product roughness, while self-excited vibrations lead to a significant increase in roughness. This paper contains the extended research on Inconel milling stability for various cutting speeds that correspond both to a stable and an unstable regions in SLD. A bit similar experiment performed for increasing the depth of cut at constant cutting speed is published in [16]. The main aim of the present paper is to investigate the effect of cutting speeds on cutting stability via multiscale entropy and the recurrence quantification technique. Moreover, the analysis of cutting forces is made with new recurrence plot quantifiers together with statistical indicators. The determination of proper stability indicators, irrespective of the rotational speed of the spindle (cutting velocity), is the main objective of the paper. The indicators could be applied in the future to build a chatter control system for detecting stability loss symptoms on the basis of statistical parameters, entropy or recurrence plot analysis.
The paper is organized as follows: section 2 presents the methodology of experimental research. Next, in section 3, a statistical analysis of cutting forces is performed, and then a recurrence analysis is presented in section 4. In section 5, the multiscale entropy method is applied to analyse milling process stability. Finally, section 6 contains conclusions.
Experiment and methodology
The experimental investigations were conducted on Inconel 718 cut on the Blue Bird MG6037PKK milling machine. The experimental setup, shown schematically in Fig.1, is composed of two subsystems: a modal analysis system and a dynamometer system.
Fig. 1. Experimental setup of CNC milling machine with acquisition system scheme
The former is used to measure tool-holder stiffness and damping coefficient (modal parameters). It consists of the PCB 086C03 modal hammer, PCB 352B10 accelerometer and NI9234 data acquisition card (DAQ). The latter is used to measure the cutting force components (F x , F y and F z ) with the Kistler 9257B piezoelectric dynamometer connected to the Kistler 5017B signal conditioner and the SCADAS Mobile LMS analyser. Both experimental rigs are integrated with the computer system. Measurements are conducted in two steps. First, an impact test is performed to obtain data for a stability lobes diagram (SLD). The modal hammer is used to excite the tool, and then the resulting vibrations are measured by the low mass accelerometer mounted on the tool tip. Next, modal parameters in the form of frequency response function (FRF) are implemented to the CutPro9 software to calculate and plot an SLD (Fig. 2a). In the second step of the experiment, the unstable lobes are verified for a series of the spindle speed and the depth of cut marked as the points in Fig. 2b. The test is performed on Inconel 718 by a 12 mm diameter end milling cutter with flutes, made of PCD (FENES DIN 6527-A 12 KNZ4 13). The radial depth of cut equals 12mm (slot milling), the feed per flute is 0.01mm. The applied milling parameters are listed in Tables 1 and 2.
sciENcE aNd tEchNology
The tool was changed after each test to provide identical cutting conditions and prevent tool wear which could affect process dynamics.
During the milling process, the forces F x , F y and F z are recorded with a sampling rate of 2 kHz; this value is a necessary minimum because the natural frequency of the spindle-tool system is about 740 Hz. On the one hand, this sampling rate meets the Nyquist-Shannon sampling theorem, and on the other hand, it is low enough to record the milling process for a sufficient period of time. In addition to this, the presence of very long time series poses difficulty in a recurrence analysis. In order to avoid the aliasing phenomenon, the Kistler measuring system is provided with an anti-aliasing filter. Stable cutting occurs in the region below the stability boundary, while unstable machining should occur above the lobes (Fig. 2). According to the diagram, the critical cutting depth a pcr is 0.133mm. For a p <a pcr the process should be stable all the time regardless of the spindle speed. The subsequent section contains a force signal analysis aimed at analysing whether the cutting process with the assumed depth of cut and spindle speed is stable or not. This will help determine stability indicators taken directly from the experiment.
Statistical analysis
The statistical analysis is performed for three directions of the cutting force: F x , F y and F z . Although F x is the most important because its direction is compatible with the feed direction, all the three components are marked in Fig. 3 illustrating the distribution of the maximum cutting force.
A typical behaviour can be observed, namely, with increasing the depth of cut the force components increase too. In the case of unstable cutting (n1a3, n1a4, n2a3, n2a4, n3a3, n3a4 in Fig.2b), the maximum forces rapidly increase to high values (see a p =0.15 and a p =0.2 in Fig.3). Examining the distribution of the standard deviation (Fig.4) one can observe that the greatest dispersion of the results occurs in the case of unstable points n=3000rpm, a p =0.2mm (n1a4) and n=4150rpm, a p =0.2mm (n3a4). For the points located in the stable area, the dispersion of the results is reduced. A large dispersion of the results for the force component F y from the stable area for the spindle speed n=3700rpm can also be observed. Moreover, an analysis of the mean value is performed via kurtosis (Fig. 5). In the case of point n1a1, the kurtosis of all force components (F x , F y , F z ) proves that the force distribution is close to the normal one. The analysis of the distribution of the individual force components around the mean value demonstrates that the largest concentration of the results was obtained for point n1a3 (n=3000rpm, a p =0.15mm).
sciENcE aNd tEchNology
The results of all tests (particularly, the y component) at the spindle speed of n=3700rpm show a great deviation from the average value. Considering all analysed force signals, the smallest asymmetry of results is obtained for the unstable points n3a3 and n3a4 which correspond to n=4150rpm, a p =0.15mm and a p =0.2mm, respectively (Fig.6).
Recurrence plot
In order to ensure a more detailed force signals analysis and thorough verification of stability regions, the recurrence plot (RP) technique is employed. The RP approach provides a qualitative interpretation of hidden patterns of dynamical systems based on phase space reconstruction. This technique was introduced by Eckmann et al. [9]. The general idea of phase space reconstruction rests on the assumption that any time series x i can be presented as a delayed vector S i in an m-dimensional space called reconstructed phase space: where m is the embedding dimension and d stands for the time delay.
Usually the embedding parameters, i.e., the time delay (d) and the embedding dimension (m), can be estimated via the average mutual information function (AMI) and the false nearest neighbours method sciENcE aNd tEchNology (FNN). More information about the AMI and FNN functions can be found in [14,29,30]. In this study, the Tisean software [44] is used to obtain the embedding parameters. The recurrence plot technique is a graphical method designed to locate recurring patterns, non-stationarity and structural changes. It shows all time instants at which a state of the dynamical system recurs. The recurrence plot can be expressed as the matrix: where Θ is the Heaviside step function, ε is a tolerance parameter (threshold), s i and s j are delay vectors.
If the trajectory in the reconstructed phase space returns at the time i into the neighbourhood of ε where it was before at the time j then M ij =1, otherwise M ij =0. These results are plotted as black (M ij =1) and white (M ij =0) dots, respectively. From a practical point of view, RPs can be presented in a more useful and certainly more convenient form via recurrence quantification analysis (RQA). RQA is a method which quantifies the number and duration of recurrences of a dynamical system presented by its state space trajectory. The measures of RQA were elaborated by Zbilut and Webber [44], and then developed and introduced to MATLAB by Marwan [25,26]. The recurrence quantification analysis can provide useful information even for short and non-stationary data, where other methods fail. RQA can be applied for various kinds of data to recognize dynamical behaviour. The F x component is chosen for the analysis because it is naturally dependent on cutting parameters. The idea of phase space reconstruction (necessary for RP) assumes that any time series from Fig. 7. Recurrence plots for cases of stable cutting (a)-n1a1, (b)-n2a1, (c)-n3a1, and unstable cutting (d)-n1a4, (e)-n2a4, (f)
sciENcE aNd tEchNology
an analysed process has the same part of information. Moreover, the amplitude of F x is the biggest therefore a ratio of noise to signal is the smallest. The recurrence plots of the force (F x ) for the selected stable cutting cases (n1a1, n2a1, n2a4, n3a1) and unstable cutting cases (n1a4, n3a4) are presented in Fig. 7. Although relatively small differences can be noticed between the stable and unstable cases, it is rather difficult to spot differences between the recurrence patterns obtained for stable cutting (Fig. 7a,b,c,e) and unstable cutting (Fig. 7d,f); therefore, the recurrence quantification analysis is applied here (Fig. 8). RQA gives more distinct information in the form of index about recurrence and system dynamics. The determinism (DET) presented in Fig. 8a illustrates well the differences between the depth of cut (a p ) and the rotational speed (n); however, the stable (n1a1, n3a1) and unstable (n1a4, n3a4) points can only be distinguished when n=3000rpm and n=4150rpm. For n=3700rpm, both points (n2a1, n2a4) are stable, therefore they should have a similar DET factor. Unfortunately, the DET is not able to find any differences in stability. Nonetheless, the DET factor points to an increase in the rotational speed (n). The average length of the diagonal lines (L) presented in Fig. 8b is the least significant stability factor. Moreover, the impact of the cutting depth (a p ) and the rotational speed (n) on L is insufficient. However, the Shannon's entropy (Lent) is also an efficient method for identifying cutting instability (Fig. 8c). Small values of Lent in the cases of stable cutting (n1a1, n3a1) considerably increase in the case of unstable cutting (n1a4, n3a4) when chatter occurs. Even in the case of n=3700rpm (green bar), when both points n2a1 and n2a4 are stable, the Lent shows an increase that is smaller than for n=3000rpm and n=4100rpm. This is because the increase is only caused by the change of the depth of cut and not by the change of stability.
Summing up, the DET shows the change of the cutting depth (a p ) and the rotational speed (n), while the Lent is additionally able to distinguish stable and unstable cutting for various rotational speeds. A new cutting stability indicator can be proposed which is defined as a ratio of the Lent for unstable to stable points (LentR). This ratio is vital because for n=3000rpm LentR=471, for n=4150rpm LentR=908, while in the case of stable cutting at n=3700rpm LentR=2.5.
Entropy
When analysing complex systems with unpredictable behaviour, it is useful to apply the multiscale entropy method. The application of this method improves the understanding of complex phenomena and such analysis is becoming more and more popular [5,6]. Multiscale entropy is used for determining the complexity of measured finitelength time-series signals.
MSE may be encumbered with some error, depending on the scale factor length τ. Therefore, in this paper composite multi-scale entropy (CMSE) is employed, which helps prevent the above-mentioned errors. The concept of composite multiscale entropy is based on the coarse-graining procedure that uses a coarse-grained time series as an average of the original data points within non-overlapping windows by increasing the scale factor τ according to Eq. (6): where τ=1, 2, 3, when τ=1, the vector y j =x i . The vector x is a row of one-dimensional time series. Figure 9 illustrates the averaged values. In this approach, the CMSE for each scale factor τ is calculated on the coarse-grained time series y k j Finally, SampEn is the logarithm of conditional probability that two sequences with a tolerance r from the points that remain within r to each other at the next point: In Eq. 9, the parameter m denotes the length of two likelihood occurrence chains that are similar toward each other in the tolerance r. For the analysis of the time courses m=2 is applied, whereas the tolerance of probability r=0.1σ x is applied where σ x is a standard deviation of the original time course vector. Finally, the CMSE is calculated according to the algorithm shown in Fig.10.
The CMSE signal analysis was applied to the measured signals of the cutting force in x-direction for different sets of two cutting process parameters: the cutting depth a p and the angular velocity of the spindle n. Figs. 11 and 12 show the CMSE results. The composite multiscale entropy analysis revealed that the force signals demonstrate different levels of regularity. The case of machining with the angular velocity of spindle set to n=4150rpm and the cutting depth a p =0.2mm produced the most regular signals. In other cases, the CMSE reaches higher values, which points to irregularity of the measured signals. With the cutting depth set to a p =0.05mm, vibration occurs in the process, irrespective of the spindle velocity n. In Figure 11a one can observe a similar CMSE for each scale factor τ, while Fig. 11b reveals significant differences for milling at n=3000rpm, n=3700rpm or n=4150rpm.
The above is clearly shown in Figs. 12a, b and Fig. 12c, where it is plotted with respect to an increased angular velocity. Consequently, the signals are compared between the cases of inside and outside lobes shown in Fig. 2b. This corresponds to the stable and unstable conditions of the milling process. One can notice that the unstable case shown in Fig. 2 (point n3a4) turns out to be a more regular condition for machining (stars line in Fig. 11b or circles line in Fig.12c) than the second set point n3a1 (stars line in Fig. 11a or black line in Fig. 12c). Moreover, when the angular velocity of the spindle is decreased to n=3700rpm (Fig. 12b) or n=3000rpm (Fig.12a), the irregularity of the signals is similar to the case n3a1 (n=4150rpm, a p =0.05mm in Fig.12c). The most significant difference can be observed for the highest analysed angular velocity (n3).
The system is sensitive to process parameters and the CMSE shows a clear discrepancy whether the process is stable or not. For the angular velocity n=3700rpm (Fig. 12b) or even n=3000rpm (Fig.12a), the CMSE sciENcE aNd tEchNology is similar, which indicates that the depth of cutting has no significant effect on machining stability. Finally, one can observe that the CMSE method provides an alternative approach to estimating conditions of the machining process.
Conclusion
The paper investigated the problem of the milling process stability for nickel alloy. First, a modal analysis was performed to obtain a stability lobes diagram; the diagram was then verified for selected points by the statistical method, recurrence technique and multiscale entropy method.
Based on the statistical analysis results, it is found that in the case of unstable cutting (points above the stability lobes), there is a significant increase in the maximum cutting force, particularly in the feed direction (x). Similar conclusions can be reached by analysing the standard deviation that is quite obvious in this case. However, it is interesting to observe that a greater symmetry and concentration of the analysed signals occur in the unstable points. Recurrence plots and especially RQA are useful for identification chatter vibrations in cutting processes. Given a variety of RQA indexes, only a few of them were tested here to select the best one. Undoubtedly, the Shannons entropy (Lent) is the best because it can distinguish stable and unstable cutting for various rotational speeds (cutting speeds). The Lent ratio (between unstable and stable cutting) is the best indicator of cutting instability. L-entropy is extremely high when chatter vibrations appear (unstable cutting), because chatter vibrations show a higher regularity than small-amplitude stochastic vibrations in stable cutting. The recurrence diagrams are not useful to a sufficient degree because it is difficult to estimate differences in the RPs and the differences are subjective.
Independently of the above, the composite multiscale entropy analysis revealed a certain degree of signal regularity, which may prove useful when estimating which can give a hint if the milling conditions are profitable for the process. As regards the analysed force signals, it was noticed that increasing the rotational speed of the spindle causes the system to behave in a more regular way, provided that the cutting depth has been selected properly. This observation can be useful when other methods such as RP and RQA fail or their results are questionable. Since some of the analysed indexes show quite significant differences between stable and unstable cutting, the proposed methods have potential to be employed in the future in a chatter control system. | 2019-04-16T13:27:45.837Z | 2018-03-20T00:00:00.000 | {
"year": 2018,
"sha1": "e301e71d2808dba2cf360da22de718b56bcef60c",
"oa_license": null,
"oa_url": "https://doi.org/10.17531/ein.2018.2.19",
"oa_status": "GOLD",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "fcaf3d58073957fb422327ce344321ccd3e2ca8c",
"s2fieldsofstudy": [
"Materials Science"
],
"extfieldsofstudy": [
"Materials Science"
]
} |
237048493 | pes2o/s2orc | v3-fos-license | Semantic Answer Similarity for Evaluating Question Answering Models
The evaluation of question answering models compares ground-truth annotations with model predictions. However, as of today, this comparison is mostly lexical-based and therefore misses out on answers that have no lexical overlap but are still semantically similar, thus treating correct answers as false. This underestimation of the true performance of models hinders user acceptance in applications and complicates a fair comparison of different models. Therefore, there is a need for an evaluation metric that is based on semantics instead of pure string similarity. In this short paper, we present SAS, a cross-encoder-based metric for the estimation of semantic answer similarity, and compare it to seven existing metrics. To this end, we create an English and a German three-way annotated evaluation dataset containing pairs of answers along with human judgment of their semantic similarity, which we release along with an implementation of the SAS metric and the experiments. We find that semantic similarity metrics based on recent transformer models correlate much better with human judgment than traditional lexical similarity metrics on our two newly created datasets and one dataset from related work.
Introduction
The evaluation of question answering (QA) models relies on human-annotated datasets of questionanswer pairs. Given a question, the ground-truth answer is compared to the answer predicted by a model with regard to different similarity metrics. Currently, the most prominent metrics for the evaluation of QA models are exact match (EM), F1-score, and top-n-accuracy. All these three metrics rely on string-based comparison. EM is a binary metric that checks whether the predicted answer string matches exactly the ground-truth answer. While this metric works well for short factual * Both authors contributed equally to this research. answers, such as names of persons or locations, it has some obvious flaws when it comes to comparing slightly differing short answers or longer, more elaborate answers. Even a prediction that differs from the ground truth in only one character in the answer string is evaluated as completely wrong. To mitigate this problem and have a continuous score ranging between 0 and 1, the F1-score can be used. In this case, precision is calculated based on the relative number of tokens in the prediction that are also in the ground-truth answer and recall is calculated based on the relative number of tokens in the ground-truth answer that are also in the prediction. An extension of this metric runs stop word removal and lowercasing before the comparison, for example, to disregard prepositions. As an F1-score is not as simple to interpret as accuracy, there is a third common metric for the evaluation of QA models. Top-n-accuracy evaluates the first n model predictions as a group and considers the predictions correct if there is any positional overlap between the ground-truth answer and one of the first n model predictions -otherwise, they are considered incorrect. The answer string itself is not compared for top-n-accuracy but the start and end index of the answer within the text document from where the answer is extracted, called context. If a dataset contains multi-way annotations, there can be multiple different ground-truth answers for the same question. The maximum similarity score of a prediction over all ground-truth answers is used in that case, which works with all the metrics above. However, a problem is that sometimes only one correct answer is annotated when in fact there are multiple correct answers in a document. If the multiple correct answers are semantically but not lexically the same, existing metrics require all correct answers within a document to be annotated and cannot be used reliably otherwise. Figure 1 gives an example of a context, a question, multiple groundtruth answers, a prediction and different similarity Question: How many plant species are estimated to be in the Amazon region? Context: The region is home to about 2.5 million insect species, tens of thousands of plants, and some 2,000 birds and mammals. To date, at least 40,000 plant species [. . . ] Ground-Truth Answer: "40,000" Predicted Answer: "tens of thousands" Exact Match: 0.00 F1-Score: 0.00 Top-1-Accuracy: 0.00 SAS: 0.55 Human Judgment: 0.50 Figure 1: Exact match, F1-score, and top-1-accuracy are no good metrics to evaluate QA models. They do not take into account semantic similarity of predictions and ground-truth answers but only their lexical similarity. SAS is close to human judgment of similarity.
scores. Besides EM, F1-score, and top-1-accuracy, we also list human judgment. The example shows that the existing metrics cannot capture the semantic similarity of the prediction and the ground-truth answers but are limited to lexical similarity.
Extractive QA is not the only QA task that requires evaluation metrics that go beyond stringbased matching. Abstractive QA requires generative models to synthesize an answer and they need to be evaluated differently, too. As of today, the evaluation of this task reuses metrics from the research field of natural language generation (NLG) but they are mostly string-based and not tailored to QA. Given the shortcomings of the existing metrics, a novel metric for QA is needed and we address this challenge by presenting SAS, a cross-encoderbased semantic answer similarity metric, and compare it with traditional lexical metrics and two recently proposed metrics for the more general task of semantic textual similarity (STS). To encourage and support research in this area, we release the annotated dataset 1 and the trained model 2 under the Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0).
The remainder of this paper is structured as follows: Section 2 describes related work on metrics used for the evaluation of QA and similar tasks. In Section 3, we present our new metric SAS along with two existing STS metrics. Three datasets, two newly created, three-way annotated datasets and one existing dataset from related work, are the basis of the experiments presented in Section 4, which compare four lexical and three semantic similarity metrics. We conclude in Section 5 with directions for future work.
Related Work
The related work discussed in this section goes beyond evaluation metrics explicitly meant for QA. The task can be generalized to estimating the semantic similarity of a pair of text inputs, which is often referred to as semantic textual similarity (STS). While there is a benchmark dataset for STS estimation (Cer et al., 2017), to the best of our knowledge, not even a single dataset has been created for the subtask of estimating semantic answer similarity. STS is closely related to paraphrasing, which, strictly speaking, refers to semantic equivalence. As a consequence of this strict definition, Bhagat and Hovy (2013) introduce the concept of approximate paraphrases as conveying a similar meaning. Measuring to what extent a text is an approximate paraphrase of another text has been addressed in several subfields of research on natural language processing and we list different approaches in the following.
A recent tutorial summarizes evaluation metrics used in NLG, including but not limited to the QA task (Khapra and Sai, 2021). The challenge of evaluating NLG has also been recently addressed by Gehrmann et al. (2021), who introduce the GEM benchmark comprising eleven datasets. Besides traditional, lexical similarity metrics, such as BLEU (Papineni et al., 2002), ROUGE (Lin and Hovy, 2003), and METEOR (Banerjee and Lavie, 2005), the benchmark also includes the two semantic similarity metrics BERTScore (Zhang et al., 2020) and BLEURT (Sellam et al., 2020). Both BERTScore and BLEURT are BERT-based ) metrics tailored to evaluating NLG.
BLEU (BiLingual Evaluation Understudy) (Papineni et al., 2002) is a metric used to evaluate the quality of machine translations by measuring the n-gram overlap of prediction and ground truth. Similarly, there is ROUGE (Recall-Oriented Understudy for Gisting Evaluation), which comes in different variations, such as using n-gram overlap (ROUGE-N) (Lin and Hovy, 2003) or the longest common subsequence (ROUGE-L) (Lin and Och, 2004) of prediction and ground truth. METEOR (Banerjee and Lavie, 2005) addresses the same task but aims to improve on BLEU, for example, by using a weighted harmonic mean of precision and recall of uni-gram overlap. It relies on WordNet (Miller, 1995) and, for this reason, can only be used for English. There are also slightly modified versions of BLEU and Rouge for yes-no answers and entity answers that introduce a bonus term that gives more weight to correct answers of that type Yang et al. (2018). Still, these modified versions rely on the standard implementations of BLEU and ROUGE, thus inheriting their shortcomings with regard to lexical vs. semantic similarity.
BERTScore (Zhang et al., 2020) is similar to F1-score in that it performs stop word removal and lowercasing before the comparison. TF-IDF is used to lower the influence of stop words on the score. In our work, we argue that stopword removal should not be an extra step of the metric. Instead, the metrics should be based on models that have been trained to recognize what words and phrases are more or less important when comparing the meaning of two answers. The main advantage of BERTScore over traditional metrics is that it compares contextual embeddings of tokens in the prediction and the ground truth instead of the actual tokens. As future work, Zhang et al. (2020) mention that BERTScore could be adapted for the evaluation of different tasks and in this paper, we discuss whether BERTScore is superior to other metrics for the evaluation of QA tasks. Chen et al. (2019) apply BERTScore as an evaluation metric for QA and find that METEOR has a stronger correlation with human judgment.
While evaluating the correctness of predicted answers is the by far the most popular QA evaluation task, there are also approaches to evaluate the consistency of the predicted answers (Ribeiro et al., 2019) or other desirable properties of opendomain QA models, such as efficiency, context awareness, fine granularity of answers, end-to-end trainability or ability to generalize to different input data (Ahmad et al., 2019). Perturbations can serve as a means to evaluate the latter (Shah et al., 2020) and, in the same way, perturbations of the training data allow training models that are more robust (Khashabi et al., 2020). The correctness of answers can be estimated with methods from natural language inference: Chen et al. (2021) con-vert answers to declarative statements and check whether the statement can be inferred from the relevant document (context). Nema and Khapra (2018) consider the task of evaluating the answerability of generated questions. They find that existing string-based evaluation metrics do not correlate well with human judgment and propose modifications of these metrics, which give more weight to relevant content words, named entities, etc. With the metrics and the underlying models presented in this paper, we present end-to-end deep learning approaches, which learn these features automatically if they increase the correlation between the automated metric and human judgment.
Semantic similarity metrics might also mitigate the influence of an annotator bias on the evaluation, which has been reported to be learned by models and is currently not recognized if the same annotators create both the training and test dataset (Geva et al., 2019;Ko et al., 2020). That bias could be, for example, the position of the answer within a document always being in the first few sentences or a specific style of phrasing the questions. With the help of semantic similarity metrics the position of the annotated answer within the context would not make a difference for the evaluation.
In line with the evaluation of QA models, the automated evaluation of models for question generation also relies on BLEU, ROUGE, and ME-TEOR, while human evaluation is limited to small datasets (Du et al., 2017). For the evaluation of conversational QA, Siblini et al. (2021) address the problems that arise from teacher forcing, which refers to earlier ground-truth answers being available to a model at each step in the conversation. The authors discuss ideas to mitigate this problem, such as using the model's own predicted answers instead of the ground-truth answers. However, this approach only considers the ground-truth user reaction to the ground-truth answer but not the predicted answer as other reactions are not available in offline training and evaluation. Last but not least, there is research on error analyses of QA models, which defines guidelines (Wu et al., 2019) or identifies challenges and promising directions for future work (Rondeau and Hazen, 2018;Wadhwa et al., 2018;Pugaliya et al., 2019). These publications present anecdotal evidence of predictions that are evaluated as wrong due to the limitations of lexical similarity metrics but are in fact correct.
Approach
We consider four different approaches to estimate the semantic similarity of pairs of answers: a biencoder approach, a cross-encoder approach, a vanilla version of BERTScore, and a trained version of BERTScore. This section describes each of these four approaches and the pre-trained language models they are based on.
Bi-Encoder The bi-encoder approach is based on the sentence transformers architecture (Reimers and Gurevych, 2019), which is a siamese neural network architecture comprising two language models that encode the two text inputs and cosine similarity to calculate a similarity score of the two encoded texts. The model that we use is based on xlm-roberta-base, where the training has been continued on an unreleased multi-lingual paraphrase dataset. The resulting model, called paraphrase-xlm-r-multilingual-v1, has then been fine-tuned on the English-language STS benchmark dataset (Cer et al., 2017) and a machine-translated German-language version 3 of the same data. The final model is called T-Systems-onsite/cross-en-deroberta-sentence-transformer and is available on the huggingface model hub. As the model has been trained on English-and German-language data, we use the exact same model for all three datasets in our experiments. An advantage of the bi-encoder architecture is that the embeddings of the two text inputs are calculated separately. As a consequence, the embeddings of the ground-truth answers can be pre-computed and reused when comparing with the predictions of several different models. This pre-computation can almost halve the time needed to run the evaluation.
SAS Our new approach called SAS differs from the bi-encoder in that it does not calculate separate embeddings for the input texts. Instead, we use a cross-encoder architecture, where the two texts are concatenated with a special separator token in between. The underlying language model is called cross-encoder/stsb-roberta-large and has been trained on the STS benchmark dataset (Cer et al., 2017). Unfortunately, there are only English cross-encoder models for STS estimation available. Therefore, we train a German cross-encoder model for STS estimation, which we release online. This model is based on deepset/gbert-base and we train it for four epochs with a batch size of 16 and the Adam optimizer on the machine-translated version of the STS benchmark that has previously been used to train a bi-encoder STS model for German. For the warm-up phase of the training, we use 10% of the training data and linearly increase the learning rate to 2e-5. While pre-computation is not possible with the cross-encoder architecture, its advantage over bi-encoders is that it takes into account both text inputs at the same time when applying the language model in a monolithic way rather than calculating encodings separately and comparing them afterward.
BERTScore vanilla or trained The BERTScore vanilla approach uses the task-agnostic, pre-trained language models bert-base-uncased for the Englishlanguage datasets and deepset/gelectra-base for the German-language dataset. In line with the approach by Zhang et al. (2020), we use the language models to generate contextual embeddings, match the embeddings of the tokens in the groundtruth answer and in the prediction and take the maximum cosine similarity of the matched tokens as the similarity score of the two answers. The optional step of importance weighting of tokens based on inverse document frequency scores is not applied. For the vanilla version, we extract embeddings from the second layer and for the trained version from the last layer. In contrast to the vanilla model, the BERTScore trained model uses a taskspecific model tailored to STS estimation. It is the same multi-lingual model that is used by the biencoder approach, called T-Systems-onsite/crossen-de-roberta-sentence-transformer.
Experiments
To evaluate the ability of the different approaches to estimate semantic answer similarity, we measure their correlation with human judgment of similarity on three datasets. This section describes the dataset creation, experiment setup, and the final results.
Datasets
The evaluation uses subsets of three existing datasets: SQuAD, GermanQuAD, and NQ-open. We process and hand-annotate the datasets as described in the following so that each of the processed subsets contains pairs of answers and a class label that indicates their semantic similarity. There 0 The two answers are completely dissimilar. "power steering" = "air conditioning" 1 The two answers have a similar meaning but one of them is less detailed and could be derived from the more elaborate answer. "Joseph Priestley" ≈ "Priestley" 2 The two answers have the same meaning. "UV" = "ultraviolet" are three similarity classes: dissimilar answers, approximately similar answers, and equivalent answers, which are all described in Table 1.
SQuAD We annotate the semantic similarity of pairs of answers in a subset of the English-language SQuAD test dataset (Rajpurkar et al., 2018). The original dataset contains multi-way annotated questions, which means there are on average 4.8 answers per question. Answers to the same question by different annotators often are the same but in some cases they have only a small overlap or no overlap at all. We consider a subset where 566 pairs of ground-truth answers have an F1-score of 0 (no lexical overlap of the answers) and 376 pairs have an F1-score larger than 0 (some lexical overlap of the answers). As we use the majority vote as the ground-truth label of semantic similarity in our experiments, we let two of the authors label each pair of answers while a third author acts as a tie-breaker labeling only those samples, where the first two labels disagree. The resulting dataset comprises 942 pairs of answers each with a majority vote indicating either dissimilar answers, approximately similar answers, or equivalent answers.
GermanQuAD To show that the presented approaches also work on non-English datasets, we consider the German-language GermanQuAD dataset (Möller et al., 2021). It contains a threeway annotated test set, which means there are three correct answers given for each question. After removing questions where all answers are the same, there are 137 pairs of ground-truth answers that have an F1-score of 0 and 288 pairs of answers have an F1-score larger than 0. We label these 425 pairs in the same way as the SQuAD subset resulting in 425 pairs of answers each with a majority vote indicating their semantic similarity.
NQ-open
The original Natural Questions dataset (NQ) (Kwiatkowski et al., 2019) was meant for reading comprehension but adapted the dataset for open-domain QA and it has been released under the name NQ-open. We use the test dataset of NQ-open as it contains not only questions and ground-truth answers but also model predictions and annotations how similar these predictions are to the ground-truth answer. There are three classes of definitely incorrect predictions, possibly correct predictions, and definitely correct predictions. Min et al. (2021) report in more detail on how these additional annotations were created. They resemble the three similarity classes we defined in Table 1. After filtering for only those questions that have exactly one ground-truth answer, we create pairs of ground-truth answers and model predictions accompanied with the label indicating the correctness of the prediction, which also corresponds to the similarity of the ground-truth answer and the predicted answer. There are 3,658 pairs of answers of which 3118 have an F1-score of 0 and 540 pairs that have an F1-score larger than 0. Table 2 lists the correlation between different automated evaluation metrics and human judgment using Spearman's rho and Kendall's tau-b rank correlation coefficients on labeled subsets of SQuAD, GermanQuAD, and NQ-open datasets. The traditional metrics ROUGE-L and METEOR have very weak correlation with human judgement if there is no lexical overlap between the pair of answers, in which case the F1-score and BLEU are 0. If there is some lexical overlap, the correlation is stronger for all these metrics but BLEU lags far behind the others. METEOR is outperformed by ROUGE-L and F1-score, which achieve almost equal correlation. All four semantic answer similarity approaches outperform the traditional metrics and among them, the cross-encoder model is consistently achieving the strongest correlation with human judgment except for slightly underperforming the trained BERTScore metric with regard to τ on English-language pairs of answers with no lexical overlap. This result shows that semantic similarity metrics are needed in addition to lexical-based metrics for automated evaluation of QA models. The former correlate much better with human judgment and thus, are a better estimation of a model's performance in real-world applications.
SQuAD
GermanQuAD NQ-open Embedding Extraction for BERTScore BERT-Score can be used with different language models to generate contextual embeddings of text inputs. While the embeddings are typically extracted from the last layer of the model, they can be extracted from any of its layers and related work has shown that for some tasks the last layer is not the best (Liu et al., 2019). The experiment visualized in Figure 2 evaluates the correlation between human judgment of semantic answer similarity and a vanilla and a trained BERTScore model. Comparing the extraction of embeddings from the different layers, we find that the last layer drastically outperforms all other models for the trained model. For the vanilla BERTScore model, the choice of the layer has a much smaller influence on the performance, with the first two layers resulting in the strongest correlation with human judgment. For comparison, Figure 2 also includes the results of a cross-encoder model, which does not have the option to choose different layers due to its architecture.
Conclusion and Future Work
Current evaluation metrics for QA models are limited in that they check for lexical or positional overlap of ground-truth answers and predictions but do not take into account semantic similarity. In this paper, we present SAS, a semantic answer similarity metric that overcomes this limitation. It leverages a cross-encoder architecture and transformer-based language models, which are pre-trained on STS BERTScore vanilla is pretrained only on Wikipedia, whereas BERTScore trained is fine-tuned on the STS benchmark dataset (Cer et al., 2017). datasets that have not been used in the context of QA so far. Experiments on three datasets demonstrate that SAS outperforms four lexical-based and three semantics-based similarity metrics regarding the correlation between the automated metrics and human judgment of the semantic similarity of pairs of answers. A promising path for future work is to analyze pairs of answers where SAS differs from human judgment and find types of common errors. Based on these findings, a dataset tailored to training models for estimating semantic answer similarity could be created. | 2021-08-16T01:15:36.326Z | 2021-08-13T00:00:00.000 | {
"year": 2021,
"sha1": "0c8ffa9e192a4f9a90dd57aefe47289a3357f51d",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "ACL",
"pdf_hash": "5f76a49a4703ed749609c9c0a051f34ee0b07306",
"s2fieldsofstudy": [
"Computer Science"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
240199770 | pes2o/s2orc | v3-fos-license | Influence of Indian lignite on gas solid hydrodynamics of a 210 MW CFB riser
The circulating fluidized bed combustion (CFBC) technology is better alternative to be adopted for generating power than PC fired boiler technology due to several advantages in which mainly fuel flexibility, environmentally friendly and low cost compared to combination of PC fired boiler with FGD for similar environmental performance. Unlike PC fired boilers, CFB boilers are suitable for low grade fuels like lignite. The huge availability of lignite in India reduces the import cost of bituminous coal which is regularly supplied as a fuel for PC fired boilers from other countries. The objective of the present paper is that the study of gas solid behaviour in 210 MW CFB riser by using mathematical equations. In this study, the key parameters of hydrodynamics (suspension density, local voidage, pressure drop) which are affecting the design and performance of CFB boiler are studied at different height levels of riser. Three Indian lignite samples (Tamilnadu, Gujarat and Rajasthan) which are successfully used in commercial operations of CFB boiler in India are assumed for this study with two mean particle diameters. This work is also determined and compared the heat transfer coefficients of boiler elements (water wall, wing wall and furnace superheater) placed in the riser.
Introduction
Circulating fluidized bed technology uses the principle that the crushed coal along with lime stone is introduced into the lower furnace where it is kept suspended and burnt partially upward flow of primary air. The secondary air is injected at the bottom of upper furnace where complete combustion takes place [1].Understanding of gas solid behaviour is an important section for design and evaluation of performance of CFB boilers. The solid particles and gas in the riser influence many parameters like principal dimensions of furnace, pressure drop along the furnace height, heat transfer coefficients, rate of solids circulation, bed density and surface area of boiler elements [1]. CFBC furnace is divided into lower dense zone and upper dilute zone based on the solid particles present. Dense zone is considered from distributor to secondary air ports. The dilute zone placed in between secondary air ports to riser exit. Hydrodynamics is required for this zone to identify changes of physical and chemical behaviour. The dilute can be divided into two vertical sections like annular and core regions. The more solid particles accumulate together and the gas velocity becomes low in the annular region which is connected to wall. The gas velocity is high and dispersion of solids takes place in core region [2]. There are many studies available in open literature for axial voidage and bed density which are decreased from bottom to top of the bed [3][4][5][6][7][8][9]. Kunii and Levenspiel [4][5] developed correlations for estimating suspension density and axial voidage at any section in the riser. Adanez et al. [3] developed equations for decay factor using particle diameter, equivalent diameter of riser, riser velocities. Allan S Issangaya et al. [10], Afsin Gungor et al. [11] and Todd S Pugsley et al. [12] discussed about radial voidage in their studies at different operating conditions. The local voidage is a function of mean bed voidage at respective section and solids concentration at the wall is almost 2.3 times the mean bed solid concentration [2]. The thickness of boundary layer decreases along with height and varies in decimetres for commercial CFB risers [13]. The suitable correlation of local voidage at all radial locations in the riser is proposed by Allan [10].
1234567890''""
Heat transfer coefficient is a function of bed temperature and suspension density. There are many correlations available for commercial CFB boilers to predict heat transfer coefficient of water wall in open literature [2, 14 -19] at different bed temperatures (550 0 C-940 0 C) and densities (2-80 kg/m 3 ). Dutta and Basu [14] studied about heat transfer coefficient of wing wall for 20 MW and 170MW CFB boiler and developed a correlation of heat transfer coefficient for wing wall. The primary super heater is placed at the middle of the furnace. Basu [1] compared the heat transfer coefficients of primary superheater with water wall based on bed temperature. The heat transfer coefficient of superheater is slightly higher than water wall when bed temperature varies from 830 to 860 0 C and slightly reduced when bed temperature crosses 880 0 C.
Even though many researchers explained the gas solid behaviour experimentally and mathematically in the furnace, a detailed mathematical model for low grade fuels like Indian lignite is not available properly till date. The present work focuses the mathematical study of hydrodynamic parameters in the riser of 210MW CFB boiler using three types of Indian lignite which are popularly used in commercial operations. General and suitable correlations which are available in open literature and used for commercial operations are assumed for this work. This study is concentrated 0.8mm and 0.9mm of mean particle diameters.
Selection of Indian lignite
The power sector in India is looking for more CFBC boilers which are best suited for low grade coals like Indian lignite compared to conventional (PC fired) technology. The Indian lignite coals are high moisture content and have sulphur content varies in between 0.3 and 6% [23]. Hence these coals are suitable for clean power generation in CFBC boilers. The lignite in India is predominantly available with total quantity of 44138MT in Tamilnadu, Rajasthan and Gujarat [24]. Three lignite samples which are successfully used for power generation are taken for this study. The ultimate analysis and 1234567890''""
The hydrodynamic model
The riser consists of lower and upper section. The upper furnace is divided into annular and core regions. The secondary air is injected above the refractory. The solids in the gases are recirculated at lower furnace through loop seal and return leg. The three boiler elements are placed in the riser. Generally water wall is placed in the upper furnace. Basu and Nag [2] said that the height of the water is equal to the upper furnace. The wing wall is used for taking evaporative load in the study. The wing wall is connected to ceiling of furnace and side walls. The superheater is located in the bottom of the upper furnace. The riser configuration with heat exchangers is shown in figure 1.
Figure 1. CFB Riser configuration with boiler elements
The following assumptions are taken in case of operating parameters for this model. The bed temperature is assumed as 850 0 C to reduce NOx emissions, erosion problems, losses of combustion and thermal stresses on water tubes [1]. Ca/S ratio is taken as 2 at which the sulphur capture efficiency reaches 90% and high lime stone reactivity takes place [1]. Fluidization velocity is assumed as 6 m/s above which the amount of carry over unburnt carbon particles, stack gas losses and erosion problems may increase [1]. Minimum solids circulation rate is assumed through the riser. The primary air ratio and combustion efficiency are assumed as 40% and 99% for combustion calculations.
There are many parameters to influence decay factor like properties of particle, operating conditions of riser. Adanez J et al [3] proposed a correlation to determine decay factor based on particle diameter The solids fraction at top of the bed can be determined by rate of solids circulation and superficial gas velocity. Voidage at furnace exit from Adanez J et al [3] The pressure drop across the bed, Todd S Pugsley et al [12] (10) The heat transfer coefficient of boiler elements in the furnace is given by The coefficients in equation (11) are taken from different studies based on commercial operations. Basu and Nag [2] for water wall, Dutta and Basu [14] for wing wall, CHENG Leming et al [31] for superheater are taken for k, α and β.
Results and discussions
Initially combustion and mass balance were carried out using fuel and lime stone analysis, operating parameters for 210MW CFB boiler. The hydrodynamic parameters for three samples of lignite at different mean diameter of 0.8mm and 0.9mm are discussed below.
Mean particle diameter influences bed to wall heat transfer coefficient which affects the combustor height as shown in figure 2. It was observed that larger particle diameter provided more suspension density above the refractory zone and transferred more heat to the water wall compared to smaller particle diameter. As a result, the reduction in surface area of water wall and height of upper furnace were taken place at 0.9mm of mean particle diameter particle. The height of furnace operated by Neyveli lignite occupied minimum height due to particle density and heat transfer coefficient compared to Giral and Surat lignite. Solids circulation rate developed in the furnace always related to the voidage present in the dense and fast beds. The raise in solids circulation rate takes place with reduction of mean bed voidage [3]. It was observed in this study that the rate of solids circulated to the cyclone from the riser was more in case of Neyveli lignite due to less mean bed voidage for the both the mean particle diameters. The average bed density of lower furnace was obtained more due to more gap between solid particles. As a result, the pressure drop per unit height was obtained more in the lower furnace (dense zone) than upper furnace (dilute zone). It was observed that the pressure drop in the risers operated by Surat and Giral lignite were obtained almost same due to negligible change in voidage and bed density. The pressure drop in the furnace operated by Neyveli lignite was more than other lignite due to high bed density. The figure 5 explains suspension density at ten sections of upper furnace at 6 m/s of superficial gas velocity. Generally bed is denser just below the upper section of riser. When secondary air is introduced, the solid particles dispersion is taken place and voidage is increased. Due to this, bed density is suddenly dropped at the beginning of upper furnace. The density of bed at respective section was determined by the equation (6). It was understood that solid particle decaying is more in furnace operated by larger mean particle diameter. As a result, the density drop was higher for larger particles compared to smaller particles at the bottom sections of upper furnace as shown in figure 5. Small changes were observed in voidage and densities from third section of upper furnace to riser exit. The reason is that the particles were already dispersed totally.
When the larger particle diameter is considered, the solids falling velocity near the wall is decreased. As a result, more particles were accumulated and reduction in voidage near the wall was taken place. In case of three samples of lignite, the local voidage near the wall was more in the furnace operated by Neyveli lignite due to more bed density as shown in figure 6. The heat transfer coefficient of boiler elements in the furnace is affected mainly by suspension density and bed temperature. The amount of heat is transferred to furnace heating elements by solid particles which are falling down (clusters) and gas which is flowing up. The figure 7 shows relationship between mean particle diameter and heat transfer coefficient for three samples of Indian lignite. Heat transfer coefficient of water wall was high for Neyveli lignite due to high particle and bed density. More heat was transferred at 0.9mm of mean particle diameter due to less mean bed voidage. In case of Giral and Surat lignite, it was observed that the heat transfer coefficients were not changed much due to negligible change of bed density and solids circulation rate. The wing wall heat transfer coefficients were reached almost 30% of the water wall heat transfer coefficients in this study by assuming a suitable correlation from Dutta and Basu [14]. The heat transfer coefficients of superheater were almost equal to the water wall heat transfer coefficients. | 2021-10-30T16:22:46.142Z | 2018-09-20T00:00:00.000 | {
"year": 2018,
"sha1": "df549411f38c03e686cebae9d1dfdbc30fff8821",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1757-899x/402/1/012047",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "ddc83e721e66e1e29121d0223f20432566d1ceb7",
"s2fieldsofstudy": [
"Environmental Science",
"Engineering"
],
"extfieldsofstudy": [
"Physics"
]
} |
251180591 | pes2o/s2orc | v3-fos-license | Artificial Intelligence Based Pain Assessment Technology in Clinical Application of Real-World Neonatal Blood Sampling
Background: Accurate neonatal pain assessment (NPA) is the key to neonatal pain management, yet it is a challenging task for medical staff. This study aimed to analyze the clinical practicability of the artificial intelligence based NPA (AI-NPA) tool for real-world blood sampling. Method: We performed a prospective study to analyze the consistency of the NPA results given by a self-developed automated NPA system and nurses’ on-site NPAs (OS-NPAs) for 232 newborns during blood sampling in neonatal wards, where the neonatal infant pain scale (NIPS) was used for evaluation. Spearman correlation analysis and the degree of agreement of the pain score and pain grade derived by the NIPS were applied for statistical analysis. Results: Taking the OS-NPA results as the gold standard, the accuracies of the NIPS pain score and pain grade given by the automated NPA system were 88.79% and 95.25%, with kappa values of 0.92 and 0.90 (p < 0.001), respectively. Conclusion: The results of the automated NPA system for real-world neonatal blood sampling are highly consistent with the results of the OS-NPA. Considering the great advantages of automated NPA systems in repeatability, efficiency, and cost, it is worth popularizing the AI technique in NPA for precise and efficient neonatal pain management.
Introduction
Pain assessment and management is one of the research hotspots in the neonatal care field. Although premature and full-term infants usually experience a high frequency of painful stimuli during hospitalization, neonatal pain assessment (NPA) has not been given adequate attention in clinical practice, where painful clinical procedures with a severe degree of pain are prevalent [1]. The operations, such as blood sampling, sputum suction, and indwelling needle puncture, are the major procedures in neonatal care that occur most frequently [2]. These painful procedures are mostly performed on infants within the first 3 days of admission, and no pharmacological interventions are applied for them in most cases [3]. Assessments of continuous pain occurred in less than one-third of neonates, and daily in only 10% [4]. As neonatal pain has a great impact on the short-term and long-term development of newborns [5,6], more attention should be paid to pain management in routine neonatal care.
Accurate NPA is the key to effective pain management. In clinical NPA, the facial expression is considered to be the most explicit indicator, on which a number of painassessment scales have been designed, such as the neonatal facial coding system (NFCS) [7], children and infants postoperative pain scale (CHIPPS) [8], premature infant pain profile (PIPP) [9], and neonatal infant pain scale (NIPS) [10]. These NPA scales mainly consider the facial-expression characteristics of newborns, and some of them also integrate factors such as crying, limb movement, and vital signs. Traditional scale-based on-site NPA (OS-NPA) is a process that requires dynamic monitoring by nurses rather than an instantaneous operation. Therefore, regular OS-NPA for pain management is time consuming and laborious. Meanwhile, the results of the OS-NPA could be affected by many factors, including subjective differences in observers [2], interruption from other clinical procedures, a lack of time [11], the gender difference [12], the interference of neonatal activity, inadequate pain-assessment tools [1], and the loss of some transient behaviors.
With the rapid development of artificial intelligence (AI), deep learning has been integrated into neonatal pain-expression-recognition technology [13]. The automatic recognition of neonatal pain expressions has undergone a process from static images to dynamic videos, and from theoretical experiments to system development [14], which has enabled AI-based NPA (AI-NPA). On the one hand, AI-NPA could make up for the shortcomings of the OS-NPA performed by medical staff, and it has the advantages of convenience and efficiency. On the other hand, AI-NPA requires a large number of accurately labeled neonatal pain data to build a model with strong anti-interference ability and high robustness for real-world data.
Unfortunately, the current public reported neonatal pain-expression databases [15][16][17][18] still suffer from a limited number of newborns and samples, deficient population information, limited types of pain stimuli that have large differences with painful clinical procedures, the coarse labeling granularity of pain, etc. Moreover, these state-of-the-art databases and many AI-NPA methods [19][20][21][22] focus on ideal neonatal pain samples, which are samples with restrictions on the neonatal activities and facial posture in the datacollection stage, or with manual screening to avoid disturbed neonatal pain data. These methods make it difficult to meet the clinical requirements regarding accuracy, and they are not feasible for processing neonatal pain data collected in real-world clinical scenes. Hence, there is still a large gap between the current AI-NPA methods and the actual clinical needs in terms of the pain-analysis objectives and scenarios, which has resulted in limited clinical applications.
In order to realize an automated NPA system for actual clinical needs, we previously established a video database of neonatal facial expressions during painful clinical procedures in neonatal wards [23], and we developed an AI-NPA method for real-world data. Our AI-NPA method is robust to facial occlusion and pose variations. Concretely, the method applied generative adversarial networks (GANs) to learn how to recover ideal facial images from real-world facial images with variant poses and occlusion for obtaining modified facial features that enable subsequent interference-adaptive fine-grained neonatal pain assessment. Based on our AI-NPA method, we further completed the design of an automated NPA system for neonatal pain on a mobile platform with an in-hospital server.
In this paper, we propose a prospective study to analyze the consistency of the NPA results given by the automated NPA system and OS-NPA in a real-world clinical operation scenario, thus verifying the clinical practicability of automated NPA systems. A total of 232 newborns who underwent blood-sampling operations in the neonatal wards of the Children's Hospital of Zhejiang University School of Medicine were recruited and participated in this experiment. Both the OS-NPA performed by two nurses and the AI-NPA given by our automated NPA system were applied during the implementation of four types of blood-sampling operations (i.e., venous, arterial, heel, and fingertip blood sampling) to give their own NPA results in the form of a pain score and pain grade with reference to the NIPS. The correlation and consistency of the OS-NPA and AI-NPA results, as well as the performance of our automated NPA system on real-world clinical neonatal pain data, were derived and analyzed.
The experiment results show that, according to the OS-NPA results on 232 newborns, the accuracy of the automated NPA system was 88.79%, with kappa values of 0.92 and 95.25%, and a kappa value of 0.90 for the NIPS pain score and pain grade, both with p < 0.001. This indicates that the NPA results have a high consistency between the onsite evaluation and the AI inference, even in the neonatal pain data in the real-world blood-sampling scenario. This study could provide a performance baseline for the clinical application of automated NPA systems.
Study Design
This prospective study was approved by the ethics committee of the Children's Hospital of Zhejiang University School of Medicine (2018-IRB-051), and parental informed consent was obtained. Newborns who underwent blood-sampling operations in the Department of Neonatology between 1 July 2018 and 30 June 2019 were studied, and the duration of the blood-sampling operations were controlled within 1 min. As shown in Figure 1, when the neonates needed four types of blood sampling (i.e., venous, arterial, heel, and fingertip blood sampling) for clinical diagnosis and treatment, two nurses used the NIPS to perform the respective OS-NPA. Meanwhile, a third nurse used our automated NPA system to shoot the responses of the newborns during the procedure at the bedside, and to retrieve pain scores and pain grades with reference to the NIPS generated by the system. The NIPS pain scores and pain grades of the OS-NPA performed by the two nurses and the AI-NPA performed by the automated NPA system were collected for subsequent statistical analysis.
Inclusion and Exclusion Criteria
Neonates who were hospitalized in the Department of Neonatology for more than 3 days and who underwent the abovementioned four types of blood sampling were included in this prospective study. Exclusion criteria included: serious illness conditions, such as serious birth injury, severe asphyxia, shock, metabolic encephalopathy, moderate and severe hypoxic-ischemic encephalopathy, and severe cardiopulmonary disease; newborns with conditions affecting facial-image acquisition, such as severe congenital malformations, facial malformations, facial nerve injury, and facial surgery. Neonates should not have used sedative or analgesic drugs within 72 h to avoid inaccurate NPA results in the experiment.
OS-NPA Performed by Nurses
Two experienced nurses were assigned to quantitatively assess the pain of the four types of blood sampling for newborns using the NIPS [10]. The pain indicators of the NIPS are defined by the following parameters, with a total score range from 0 to 7 points: facial expression (0-1 point), cry (0-2 points), breathing pattern (0-1 point), position of arms (0-1 point), position of legs (0-1 point), and state of arousal (0-1 point). This is suitable for acute pain assessment, and the pain severity can be classified into mild pain with a score of 0-2, moderate pain with a score of 3-4, and severe pain with a score of 5-7. The Spearman correlation analysis was used to analyze the independent pain-score results of the two nurses, and the correlation coefficient was 0.89 with p < 0.001, which indicates the high confidence of the OS-NPA results. Data with a difference in two independent pain scores were further reviewed by these two nurses. They checked the corresponding on-site video to obtain a consistently confirmed pain score as the final OS-NPA result.
AI-NPA Performed by the Automated NPA System
The automated NPA system was implemented with the client-server model. The application was designed to run on the mobile nursing personal digital assistant (MNPDA) device. When performing the AI-NPA, the nurse could use the application to record videos of the newborns' behavior during blood sampling. Specifically, the nurse pressed the shooting key 3 s before the blood-sampling operation and the end key 57 s after the operation to achieve a 60 s video for each operation. The heads and faces of the infants were required to be completely presented in the video frame during the whole process. Figure 2 presents the typical image sequence of the neonatal pain video. The video stream was then transmitted to the in-hospital server hosting the AI-NPA model for real-time NPA inference. The AI-NPA model of our automated NPA system in this study was pretrained on a larger expression-recognition database and was further modulated by a self-built neonatal facial-pain database based on our previous work [23]. Our AI-NPA method includes two parts: unsupervised feature modification and self-attention pain classification. In the unsupervised-feature-modification part, we apply a generative adversarial network (GAN) to modify the facial features toward the direction of face frontalization and deocclusion.
Specifically, we use the latent vector of the generator as the modified facial features. That is, we only need the encoder network of the generator to extract facial features during the testing phase. The generator of the GAN has two paths, focusing on the reasoning of the global shapes and the transformation of the local details [24]. The discriminator of the GAN is responsible for learning to distinguish between the output of the generator and the normal (frontal and nonoccluded) face images, which pushes the facial features into the manifold of a normal face. We further employed four loss functions: symmetry loss, adversarial loss, identity-preserving loss, and total variation regularization, to train the generator network and discriminator network jointly, as shown in Figure 3. For the self-attention pain classification, considering that the modified facial features still contain many useless, redundant, and even wrong features, we choose the self-attention mechanism to further filter and enhance the modified features. Specifically, we build an attention branch parallel to the residual branch, which is based on the bottom-up top-down structure [25], and which can output the same size-attention mask that softly weights the facial features.
In the pretraining stage, a total of 12,271 basic expression images in the RAF-DB [26] were applied to obtain the initial model parameters. We then trained this model on 508 neonatal pain images from our real-world neonatal pain database, and we adjusted the parameters to the optimal through the backpropagation algorithm. The samples of our neonatal pain database are shown in Figure 4. The AI-NPA model is deployed on an in-hospital server. The software platform of the server is Ubuntu 16.4.7 with Python 3.6.8, and the hardware configuration is Intel ® Xeon ® CPU E5-2678 v3 (12/24 cores/threads, 2.5/3.1 GHz base/turbo), 112 GB 2400 MHz DDR4 RAM, and two Nvidia ® GeForce ® RTX 2080Ti GPUs (1350/1454 MHz base/boost with 11GB GDDR 6 VRAM). After the video is transmitted to the server, the AI-NPA model detects the face region of each frame of the image in the video by a built-in algorithm, and it crops the facial image with the size of 224 × 224. Based on the NIPS, the pain scores and pain grades are automatically generated, paired with the images, and stored on the server. The final AI-NPA results are determined as the highest pain score and the corresponding pain grade during one blood-sampling operation, which are used for the subsequent validation of the accuracy of the automated NPA system. The AI-NPA results could be pushed to the application on the MNPDA for on-site nurses or could be checked by the attending doctor for further neonatal pain management, as shown in Figure 5.
Statistical Analysis
Descriptive statistics for the categorical variables were reported as numbers and percentages, and a chi-square test was used for the comparison between groups; statistics were reported as means and standard deviations for continuous variables, and a t-test or nonparametric test was used for the group comparison. To analyze the correlation and consistency of the AI-NPA and OS-NPA results, the Spearman correlation analysis, receiver operating characteristic curve (ROC), and kappa coefficient test were used. SAS 9.4 statistical software was used for data analysis, and p < 0.05 was considered statistically significant. Because the NIPS pain scores are discrete-integer results, we used the confusion matrix to show the number of cases in which the results were different between the OS-NPA and AI-NPA in detail.
Study Population
A total of 232 newborns were included in this study, with a mean gestational age of 33.93 ± 4.77 weeks, a mean age of 21 (8.5, 41) days, and a mean birth weight of 2250 ± 1010 g. The detailed demographic characteristics are listed in Table 1. Two nurses performed a total of 464 independent OS-NPAs during 232 neonatal blood-sampling operations for each newborn. As a result, one set of consensus NIPS pain scores and grades was given by the two nurses for every OS-NPA and was used as the gold standard.
Comparison of NIPS Pain Scores between OS-NPA and AI-NPA
The NIPS pain scores of the OS-NPA and AI-NPA were 5.06 ± 1.85 and 5.17 ± 1.84, respectively. The Spearman correlation analysis showed a correlation coefficient of 0.95 (p < 0.001) when comparing the NIPS pain scores of the OS-NPA and AI-NPA. The agreement between the two groups was compared, with a kappa value of 0.92 (0.88, 0.95) (p < 0.001). The accuracy of the pain score given by the automated NPA system was 88.79% (206/232), and the confusion matrix of the NIPS pain scores for the two methods is shown in Table 2. The background is highlighted for the diagonal of the confusion matrix.
Comparison of the NIPS Pain Grades between OS-NPA and AI-NPA
According to the pain-grade criteria of the NIPS, the OS-NPA showed mild pain in 27 patients (8.94%), moderate pain in 47 patients (20.26%), and severe pain in 158 patients (68.10%). The AI-NPA derived mild pain in 26 patients (11.21%), moderate pain in 39 patients (16.81%), and severe pain in 167 patients (71.98%). Compared with the on-site evaluation, the accuracy of the NIPS pain grade given by the AI-NPA was 95.25% (221/232), and the agreement between the two groups was compared, with a kappa value of 0.90 (0.84, 0.96) (p < 0.001). See Table 3 for details. The background is highlighted for the diagonal of the confusion matrix.
We further investigated the cases of the severe-pain grade. Using the results of the OS-NPA as the criterion, the sensitivity and specificity of the AI-NPA for identifying severe pain were 100% and 87.84%, respectively. The corresponding kappa values and areas under the ROC curve (AUC) of the AI-NPA for identifying severe pain were calculated with different types of blood-sampling operations, as listed in Table 4.
High Evaluation Consistency of the Automated NPA System
The kappa value was used to test the degree of agreement between the results of the different NPA methods. The kappa value ranged from −1 to U+1; the larger the value, the higher the degree of agreement between the two, where the kappa value of the consistency between the NIPS pain scores of the OS-NPA and AI-NPA was 0.92 (0.88, 0.95) (p < 0.001). Using the NIPS pain score of the OS-NPA as the gold standard, the accuracy of the pain score given by the automated NPA system was 88.79%. The comparison between the AI-NPA and OS-NPA showed that the correlation coefficient between the two was 0.95 (p < 0.001), which was highly positive. These results indicated that there was high consistency in the NIPS pain scores between the AI-NPA and OS-NPA.
Meanwhile, the accuracy of the NIPS pain grade given by the AI-NPA was 95.25%, with a kappa value of 0.90 (0.84, 0.96) (p < 0. 001). The AI-NPA was also highly consistent with the OS-NPA regarding the evaluation of the degree of pain. Because the guidelines recommend a stepped approach for neonatal pain management, including environmental, nonpharmacological, and pharmacological interventions, provided that a standardized approach is used to assess the pain [27], in addition to judging the presence or absence of pain, it is also necessary to distinguish the pain grade.
AI-NPA of Severe Pain
The statistical results in Table 4 show that most of the four types of blood-sampling operations in this study were with severe pain, and especially for the venous sampling, of which 88.8% belongs to severe pain. The AI-NPA of severe pain showed an AUC of 0.974, with a 95% confidence interval (0.952, 0.995), and the kappa value was greater than 0.75. The sensitivity and specificity of the AI-NPA for identifying severe pain were 100% and 87.84%, respectively. The relatively low specificity of the AI-NPA of severe pain was due to the fact that the AI assessment mostly focused on the facial expression, while the assessment parameter "crying" of the NIPS varied the most, with corresponding scores of 0-2 assigned to "not crying", "sobbing", and "crying", respectively.
In real-world complex clinical operation scenarios, neonatal crying is easily interfered with by noisy environments, the alarms of monitors, and the occlusion of respiratorysupport devices, and there was also an objective difference in the assessment of the "crying" by the nurses. Although the range of scores for "crying" is the largest of all the items in the NIPS, "facial expression" has been proven to be the most specific indicator of pain, and additional information such as "crying" and "physiological signals" could not provide stable performance increments for the AI-NPA methods [28][29][30][31]. The main bottleneck for highly accurate AI-NPAs is still the volume of the current training data.
While it is currently not possible to accurately assess pain scores or pain grades by AI technology, it is feasible to tune an AI model to make a tradeoff between the sensitivity and specificity of pain assessments according to the clinical requirements. Based on the views of our neonatal care experts, we prefer sensitivity so that as many neonates as possible who are suffering from pain can be screened for diagnosis and possible analgesia by physicians. Therefore, the overestimation of severe pain by our AI-NPA method is currently a purposeful choice driven by clinical needs. Although further improvement is required for our AI-NPA method, it is superior to those reported in the relevant literature [19][20][21][22]30,31], and especially in real-world scenarios. Nevertheless, the specificity of the automated NPA system for identifying severe pain in this study was completely acceptable in clinical practice.
Strengths and Limitations of the Automated AI-NPA System
The strengths of our automated AI-NPA system are threefold: It could enable 24 h real-time pain-status monitoring, saving laborious human assessment by medical staff; it is robust to occlusion and extreme facial-pose-change disturbance compared with other automated methods; recorded videos and their automated assessment results in the system could facilitate research and education related to clinical pain assessment.
The main limitation of our automated AI-NPA system is that the videos evaluated by the AI-NPA method are currently recorded by nurses using mobile nursing personal digital assistants. Considering that it is not feasible for medical staff to perform both the medical operation and video filming, the automated NPA system currently requires an additional nurse to record the video. This conflicts with our desire to significantly reduce the workload of medical staff through automated pain assessment, given the fact that healthcare resources are valuable and limited. Therefore, as the AI technology further matures and works with automatic face-tracking algorithms, we will automate the entire process by recording video with bedside cameras in the future.
Conclusions
Accurate NPA is the premise of standardized neonatal pain management. The AI technique in NPA based on facial recognition can provide convenience for medical staff. This study showed that the automated NPA system could obtain real-time dynamic painevaluation results for blood sampling in a real-world neonatal ward scene, which is helpful to implementing the stepped analgesia program for neonatal pain management. Combined with the concept of closed-loop management, the AI technology embedded in the electronicmedical-record (EMR) system in the future will realize the real-time medical intervention for pain by medical staffs. The EMR system can cooperate with the MNPDA to enable bedside nurses to implement care measures in accordance with an automatically generated nursing plan. The downstream health education system can further perform pain-knowledge education for newborns' family members so as to realize the traceability, standardization, and intelligence of the whole process of pain management. | 2022-07-31T03:01:58.936Z | 2022-07-29T00:00:00.000 | {
"year": 2022,
"sha1": "e2a630e4a4b0d6afb4976d4ca677f044b742445e",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "ScienceParsePlus",
"pdf_hash": "e2a630e4a4b0d6afb4976d4ca677f044b742445e",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": []
} |
245839015 | pes2o/s2orc | v3-fos-license | Human Brain Organoids as an In Vitro Model System of Viral Infectious Diseases
Brain organoids, or brainoids, have shown great promise in the study of central nervous system (CNS) infection. Modeling Zika virus (ZIKV) infection in brain organoids may help elucidate the relationship between ZIKV infection and microcephaly. Brain organoids have been used to study the pathogenesis of SARS-CoV-2, human immunodeficiency virus (HIV), HSV-1, and other viral infections of the CNS. In this review, we summarize the advances in the development of viral infection models in brain organoids and their potential application for exploring mechanisms of viral infections of the CNS and in new drug development. The existing limitations are further discussed and the prospects for the development and application of brain organs are prospected.
INTRODUCTION
The study of human viral infections is limited by the lack of functional models that simulate normal human physiology and pathophysiology (1). For decades, human pathogenic viruses have been studied using immortalized cell lines, primary cells isolated from the body, and a variety of animal hosts (2). Although these traditional models have made significant progress in helping to understand viral pathogenesis and host pathogen interactions, and have contributed to the development of vaccines and treatment strategies, these models may have limitations in reproducing interactions between pathogens and human hosts (1). Recently developed humanized mouse models are more valuable for human viral disease research, but they are expensive and difficult to maintain (3). A two-dimensional (2D) research strategy based on induced pluripotent stem cells (iPSC) has provided valuable information for the pathophysiology of neurological diseases, but lacks three-dimensional (3D) properties of their internal structures; thus, tissue structures composed of many different types of cells, and their complex environments and functions cannot be simulated (1,3,4). Although the 3D-based organoid culture research strategy is still in its infancy, it has provided a new method for modeling human disease (Table 1) (4). Organoids are derived from human stem cells and retain the genomic background, cellular organization, and function of their tissues of origin. Recent advances in stem cell biology have made it possible to grow organoids in vitro, mimicking a variety of human tissues, including the brain, lungs, and intestines (2).
A growing body of clinical data suggests that many pathogen infections eventually lead to a range of brain diseases. The development of the human brain presents several unique aspects, such as higher complexity and the expansion of neuronal output, which have proven difficult to study using traditional biological models (5). Thus, developing treatments for brain diseases is challenging because it is difficult to translate preclinical data from current in vitro and in vivo models into clinical practice (6). As a result, in vitro methods to simulate human brain development and disease have become a hot research area (2). As early as 1907, Harrison took frog nerve tubes and attached tissue fragments to a glass cap of coagulated serum or lymph, and thus established a hanging drop culture that allowed direct observation of growing nerves while still alive (7). In 2001, researchers produced the first 2D embryonic stem cells (ESCs) to simulate the development of embryonic neural tubes (8). In 2008, scientists modified ESCs to differentiate into dorsal forebrain progenitor cells, mimicking the main pattern of in vivo cortical development (9). Later, Pasca et al. generated pyramidal neurons from hiPSCs in three-dimensional cortical structures called human cortical globules (hCSs) (10). In 2013, Lancaster et al. combined the three-dimensional matrix and serum-free growth techniques when culturing iPSCs to generate a model of various brain regions with single-tissue methods (11).
Furthermore, researchers achieved neural induction by highdensity monolayer cultures or by embedding hiPSC clusters in gelprotein mixtures, such as Matrigel, and then cultured them in a rotating bioreactor, to mimic the 3D cortical-like structure of the brain, and produced pyramidal neurons (10). Recent advances in human iPSCs-derived brain-like organs offer promising approaches to the study of neurodevelopment, infectious diseases, highthroughput drug screening, and the genetic basis of various neuropathologies, and promote the application of precision medicine (12). Furthermore, the use of CRISPR-Cas9 gene editing techniques to reprogram iPSCs derived from patient somatic cells has brought new prospects to many areas of neurobiology, providing a wider range of options for culture systems. In theory, individualized therapy for many diseases is possible ( Figure 1) (12,13). Therefore, brain organs, or brainoids do provide a good opportunity for the study of brain diseases. But one limitation in these systems is that they can't build all the types of cells that make up the central nervous system. For example, there are astrocytes, oligodendrocytes, neurons, ependyma and microglia cells, but no immune cell and vascular structure in these systems. In this review, we discussed the use of brain organoids derived from human stem cells in the study of various viruses, such as Zika virus (ZIKV), SARS-CoV-2, and human immunodeficiency virus (HIV), as well as prion infections.
BRAIN ORGANOIDS AS A MODEL OF ZIKV INFECTION
ZIKV is an enveloped RNA virus of the Flaviridae family, the ZIKV genome contains one open reading frame (ORF) and 5′ and 3′ untranslated regions. The single polyprotein encoded the
organoidon-achips
The organ-on-a-chip systems allow the creation of dynamic and controllable microenvironments proper for viral infections and immune response analysis. With microfluidic devices used in organ-on-a-chip, it is possible to simulate human vascular endothelial dynamics. The organoid-on-a-chip can evaluate interactions with cells of the immune system and the cellular response to viral infections.
A single organ on a chip has limited capacity to meet the wide range of needs emerging from drug discovery. Non-adhesive culture conditions and cell interactions are difficult to achieve in microfluidic chips. As far as genomic stability is concerned, organ-on-a-chip systems cannot fully recapitulate cell types and their ratios in vivo and may introduce unexpected mutations during cultivation. ORF can be further cleaved into ten functional proteins, including three structural proteins (as components of envelope, precursor membrane, and capsid) and seven nonstructural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) (14). Mainly transmitted by Aedes aegypti, ZIKV was first discovered in the blood of an outpost rhesus monkey in Uganda in 1947. But ZIKV caused its first major epidemic in Micronesia, with patients developing fever, joint pain, rashes and FIGURE 1 | Application of induced pluripotent stem cells (iPSCs) and iPSCs-derived system. The enhanced culture and differentiation of iPSCs will improve the efficiency and quality of the organoids derived from iPSCs. The use of gene editing techniques such as CRISPR-Cas9 and specific small molecules will allow the generation of terminally differentiated patient cells and isogenic lines, reducing background genetic variation, and broadening the range of cells available for drug screening. Targeted differentiation schemes can be used to generate neural progenitor cells, which can be further induced into specific populations of neurons or glial cells. Two-dimensional (2D) cultures produce highly enriched cell populations to study the role of drugs in specific cell types. Three-dimensional (3D) cultures produce organoids that are similar in organization and function to the developing human brain and are used in high-throughput drug screening, genetic research, and infectious disease modeling, and to drive the development of precision medicine.
conjunctivitis until 2007. ZIKV was transmitted to Brazil in 2013 and found to be associated with microcephaly (15)(16)(17)(18). There is also evidence that ZIKV can be transmitted sexually and from motherto-child, and that ZIKV infection in pregnant women increases the prevalence of microcephaly and other neurodevelopmental disorders in newborns (19). Computed tomography images of several infants infected with ZIKV revealed intracranial calcification, severe brain atrophy, and cerebellar hypoplasia (20). In December 2016, the World Health Organization (WHO) declared ZIKV infection a public health emergency. However, the pathogenesis of ZIKV infection, how it affects neurons, and the efficacy of drugs have not been fully understood until recently, and the study of ZIKV in brain organoids has contributed to the rapid progress in elucidating the pathogenesis of ZIKV. Several studies have shown that ZIKV infection can reduce the growth of organoids by increasing the death and proliferation of developing neurons, this leads to a reduction in the volume of the neuronal cell layer, which disrupts the formation of the neurospheres and in turn, reduces the growth of organoids (21,22). Another study found that morphological changes of human neural progenitor cells (hNPCs) apparently took place and the fluorescence signal of dead cells was signifificantly increased in NS5-expressing hNPCs, consistent with ZIKV infection, ZIKV-NS5 could induce apoptotic cell death of hNPCs through regulating p53 activity. Interestingly, the expression of ZIKV-NS5 alone in hNPCs could induce the p53-mediated apoptosis, implying its contribution to the microcephaly caused by ZIKV infection (14). Their results demonstrated that ZIKV induced cell death in human iPS-derived NSCs, disrupted the formation of neurospheres, and reduced the growth of organoids (17). Fetal ZIKV infection has been identified as the cause of brain defects such as microcephaly, but live infected human fetal tissue is not available and postmortem tissue is of different quality and genetic background; clinical studies alone cannot provide sufficient insight into how ZIKV causes this damage (21). Using a developmental brain model based on ESCs, researchers studied the effects of ZIKV infection on patterns of DNA methylation throughout the genome of selected nerve cell types. ZIKV unexpectedly altered the DNA methylation status of neural progenitor cells, astrocytes, and differentiated neuron genes; up to 10% of these genes are associated with susceptibility to schizophrenia, bipolar disorder, or other diseases (23). The findings suggest that microcephaly and other large brain abnormalities in babies born after ZIKV infection during pregnancy may be just the beginning, there may also be a range of neuropsychiatric complications associated with delayed onset (23). The brain organoids are valuable in elucidating the mechanisms behind these phenomena. To understand how the virus travels through the blood to mature endothelial cells and to tissues, the researchers used primary human mononuclear cells, brain-like organoids from ESCs, slices of the cerebellum of organ-type mice, xenogenetic zebrafish models and human fetal brain samples, and found that ZIKV-infected monocytes showed higher expression of adhesion molecules, and ZIKV induced human primary monocytes to have higher adhesion, diffusion, and migration in a blood-brain barrier (BBB)-like model in vitro (24). Based on transcriptomic analysis, ZIKV infection resulted in an up-regulation of Toll-like receptor 3 (TLR3) expression that derailed cell fate and reduced the number of functional neurons, leading to microcephaly (17,25). At the molecular level, Yoon et al. found that the ZIKV-encoded NS2A protein disrupted the proliferation of radial glial cells in both the embryonic mouse cortex and the human forebrain, and ZIKV-NS2A disrupted the formation of the adhesive junction complex, resulting in aberrations of niche signals affecting cortical neurogenesis (26). Furthermore, because of the high expression of receptor tyrosine kinases in human radial glial cells (RGCs) and outer radial glial cells (oRGCs), AXL is thought to be the receptor protein for ZIKV entry in neural progenitor cells (NPCs) (21). Another study in 2016 found that ZIKV AXL receptors were highly concentrated in the radial glia of NSCs in the human embryonic cerebral cortex, providing a hypothesis that these cells are particularly vulnerable to ZIKV infection, it provides a candidate mechanism for ZIKV virus to induce microcephaly (27). A study showed that the RNAi mechanism in the hNPCs was fully capable of cleaving the dsRNAs formed during the replication of the ZIKV genome, demonstrating the direct antiviral activity of enoxacin, a well-known RNAi enhancer, and enoxacin treatment can completely prevent ZIKV induced microcephaly in brain organoids (28). These studies showed that brain organoids are of great value in studying the effects of ZIKV at the structural, cellular, and genetic levels. Brain organoids infected with ZIKV are also used in high-throughput drug screening. The researchers screened about 6000 compounds for drug reuse and found that treatment with emricasan and clonitazene restored ZIKV-negative hNPCs. It is suggested that infected cells could recover by inhibiting apoptosis to gain time (29). High-level chemical screening of cortical hNPCs derived from human multipotent stem cells showed that hippeastrine hydrobromide and amodiaquine dihydrate (AQ) could inhibit ZIKV infection (30). The methylene blue (MB), approved by the United States Food and Drug Administration (FDA), inhibits the interaction between viral protease NS3 and its cofactor NS2B, inhibits viral protease activity and viral growth, and protects 3D cerebellar organoids from ZIKV infection. Mechanistic studies have confirmed that MB plays a role in ZIKV infection and post-infection, with MB most effective for ZIKV two hours before infection (31). Furthermore, bioinformatics analysis showed that betulinic acid (BA) treatment activated the AKT pathway, therefore, BA exerts a cell protective effect on neural progenitor cells infected by ZIKV (32). The broadly active antiviral compound TH6744 interferes with the ZIKV life cycle of RNA replication, with the synthesis, posttreatment, or stability of viral proteins, and with the assembly, transport, or budding of viral particles in the brain organoid model. Therefore, TH6744 can not only alleviate the pathogenic phenotype induced by ZIKV, but also blocks the transmission of the virus (33).
BRAIN ORGANOIDS AS SARS-COV-2 INFECTION MODEL
According to national authorities, as of 11 October 2020, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of the 2019 coronavirus pandemic, was responsible for more than 37 million confirmed cases of COVID-19 and 1 million deaths (34,35). SARS-CoV-2 is a b coronavirus, derived primarily from bats and other species. The virus can cause respiratory and intestinal diseases in different animal species (36). Furthermore, clinical observations revealed neurologic symptoms and neuropsychiatric disorders in patients with COVID-19 (37). The latest clinical reports have shown that the number of patients with neurological symptoms continues to increase, and a considerable number of patients with cerebrovascular injury, emotional disorders, and other neurological complications have been reported (34,38). On 30 January 2020, the WHO declared COVID-19 a public health emergency of international concern (39). Although the SARS-CoV-2 pathogen responsible for COVID-19 has been detected in the brains of some patients, the detailed mechanisms through which it infects brain cells and affects their function are not known.
To study the susceptibility to SARS-CoV-2 infection, researchers used single-layer brain cells and region-specific brain organoids derived from human iPSC as cell models, and found that a small number of neurons and astrocytes were infected; whereas, the epithelial cells of the choroid plexus can be heavily infected (34,37,38,40,41). The study also found that SARS-CoV-2 infection-induced neuronal apoptosis depends on the type II interferon response rather than on the type I interferon response (42). In another study, researchers challenged brain organoids with SARS-CoV-2 spinous pseudoviruses and live viruses, and also demonstrated viral propensity in choroidal plexus epithelial cells, but rarely or never infect neurons or glial cells (40). Researchers have discovered that viruses in the blood can infect peripheral immune cells. These infected white blood cells can pass through the BBB, which is made up of special tight junctions between endothelial cells, pericytes, and astrocytes. In addition, the virus may cross the BBB or enter the cerebrospinal fluid (CSF) through direct interaction with the endothelium. Both mechanisms lead to changes in brain homeostasis and increase the production of cytokines in the central nervous system (CNS) (43) (Figure 2). Next, researchers studied the effect of SARS-CoV-2 infection on the function of choroid plexus (CHP) epithelial cells, and observed a sharp decrease in fluid accumulation in damaged organs, indicating a morphological disorder leading to the destruction of barrier integrity (37,38,40). The S1 subunit of SARS-CoV-2 was found to mediate barrier rupture in a 3D microfluidic model of the human BBB (44). At the molecular level, there was evidence of increased levels of Tau protein in the soma of SARS-CoV-2-positive neurons, and changes in the location of Tau and T231 phosphorylation of Tau protein in SARS-CoV-2-positive neurons. Therefore, the researchers concluded that the Tau protein is abnormally phosphorylated in response to viral-induced stress, which may in turn, trigger further cell death processes (38). Thus, brain organoids could model the pathologies affecting the CNS by COVID-19.
Researchers including Markus Hoffmann showed for the first time that Angiotensinogen 2 (ACE2) is the key receptor for SARS-CoV-2 entering cells (45). Several studies have shown that high levels of ACE2 are found in mature choroid plexus cells expressing rich lipoproteins, co-entry factors TMPRSS2 and TMPRSS4, with the abundance of ACE2 increasing over time, but it is less expressed in neurons and other cell types (38,40,43). An analysis of the data from Allen's brain atlas revealed that the CHP had the highest level of ACE2, which confirmed the findings of an in vitro brain-like organ model (40). High expression of ACE2 in choroid plexus organs may explain their susceptibility to SARS-CoV-2 infection (35). However, some researchers still have doubts about whether ACE2 is the main pathway for SARS-CoV-2 entry into neuron cells. One reason is that the level of ACE2 mRNA in CNS seems to be very low. However, recent studies have shown that ACE2 mRNA levels are not the best substitute indicator of ACE2 protein expression (46). ACE2 protein is widely expressed in neurons positive for microtubule-associated protein 2 (MAP2) and in cells near the organoid lumen (46). Further, 2D monolayer cultures of cortical neurons and brain organoids were used to assess ACE2 expression and infectivity of the SARS-CoV-2 pseudovirus. The colocalization of the mCherry signal with Angiotensinogen-2 (ACE-2) from pseudovirus-infected cells indicated that the SARS-CoV-2 spike protein can target cells expressing ACE2 in brain organs (43). Therefore, a detailed study of the brain distribution of ACE-2 may shed light on potential SARS-CoV-2 induced neurological changes. Targeted drug inhibition of the ACE2 receptor may represent an opportunity to treat COVID-19. The researchers found that the FDA approved antiviral drug sofosbuvir could inhibit SARS-CoV-2 RNA-dependent RNA polymerase (RdRp) and prevent neuronal death (41). Other studies examining the inhibitory effects of remdesivir on SARS-CoV-2 infection in human iPSCs-derived brain cells, also support the anti-SARS-CoV-2 potential of remdesivir and provide evidence for its use in the treatment of neurological complications in patients with COVID-19 (47). This will bring new hope for the treatment of COVID-19.
BRAIN ORGANOIDS AS MODELS FOR HIV INFECTION
First isolated in 1983, the HIV is a retrovirus virus that causes acquired immune deficiency syndrome (AIDS) (48). The main modes of transmission of the virus are sexual transmission, blood transmission, and mother-to-child vertical transmission (49,50). The brain manifestations of HIV infection with cognitive, behavioral, motor and autonomic dysfunction remain a problem in the daily management of HIV treatment. HIV infection may cause HIV-associated neurocognitive disorder (HAND) such as HIV encephalopathy and AIDS dementia (51). HIV is an infection caused by HIV-1 and HIV-2 lentiviruses, with HIV-1 infection rates much higher than HIV-2 (52). Infected individuals must, however, remain on antiretroviral therapy (ART) for the rest of their lives, because as soon as treatment is interrupted, viral loads return to pre-ART levels (53). Unfortunately, no clinically approved HIV-1 transcriptional inhibitors can block the ongoing transcriptional events in infected cells (53). In addition, HIV co-infection with Mycobacterium tuberculosis is a major public health concern worldwide, particularly in southern Africa (54). The genetic diversity of HIV-1 is striking, with more than 10 subtypes and an increasing number of recombinant strains around the world. However, it is not entirely clear whether subtypes have a biological effect on latency establishment and reversal. This complexity disturbs HIV vaccine development for which major epitopes of envelope proteins vary widely among different subtypes, thus affecting the efficacy of vaccines. To investigate HIV pathogenesis, researchers have developed a 3D human brain organ culture system (MG-hBORG) that adds HIV-infected microglia to human brain organ culture system (hBORGS), leading to inflammation and astrocyte damage, which is the main feature of the CNS damage (50). Furthermore, the study of brain cells reservoirs of HIV-1 that prevent a sterile cure is another interesting context in which brain organoids may be used in the future. But in the absence of suitable models, the pathogenesis of HIV-induced neurological disorders remains unclear. The generation of 3D brain organs provides a promising model system for studying the pathogenesis of HIV. Using CRISPR/ Cas9 technology to edit HIV-1 proviral genes, Kunze et al. constructed an adeno-associated virus vector (AAV9P1), which demonstrated that AAV9P1 could provide tools for evaluating HIV-1 astrocyte infection. Inactivating the HIV-1 LTR, inhibits HIV-1 previral expression, with a subsequent decline in LTR activity induced by the LTR mutation, not the elimination of the original virus, so the development of the AAV vector, HIV suppression sequence to pass to the astrocyte, astrocyte may be an alternative strategy to control persistent HIV infection (49). Through the study of brain organoids, we will better understand the pathogenesis of HAND and HIV encephalitis (HIVE).
BRAIN ORGANOIDS AS A MODEL OF HERPES SIMPLEX VIRUS INFECTION
The human herpesvirus (HSV) is one of the most common human pathogens, and is highly prevalent worldwide and capable of forming lifelong infections. Although HSV infection is usually confined to the skin and mucous membranes of the lips and genitals, it is also a common cause of sporadic encephalitis, and can lead to potentially fatal CNS infection (55). The main causes of hospitalization in 106 patients with HSV encephalitis (HSVE), were seizures, abnormal behavior, loss of consciousness, confusion of consciousness, or disorientation. Second, weight loss and symptoms of infection, including low-grade fever and rash, as well as neurological or psychiatric disorders such as aphasia, behavioral changes, and epileptic activity, must also be reviewed (56). HSV-1 infection is human-specific and particularly common in adolescents and adults, and pregnant women are at risk of mother-to-child transmission, which can lead to fetal necrotizing encephalitis and even perinatal mortality (55). Sporadic encephalitis caused by HSV-1 infection affects 2-4 of 100,000 people annually, and although antiviral therapy with acyclovir derivatives significantly reduces mortality to about 25%, but sporadic encephalitis survivors often have serious and long-term neurological sequelae. Moreover, the pathophysiological characteristics of neurodevelopmental disorders associated with HSV-1 infection remain unclear (57). Therefore, it is necessary to study the genetics and epigenetics of HSV-1 and to establish an in vitro HSV-1 infection model of the FIGURE 2 | SARS-CoV-2 entry into the central nervous system. Viruses in the blood may infect peripheral immune cells. These infected white blood cells can cross the blood-brain barrier (BBB), which is made up of special tight junctions between endothelial cells, pericytes, and astrocytes. In addition, the virus may cross the BBB, which may be modified by cytokines, or enter the cerebrospinal fluid (CSF) through direct interaction with the endothelium. Both mechanisms lead to changes in brain homeostasis and increase the production of cytokines in the central nervous system. human CNS to advance our understanding of the molecular mechanism of latency and reactivation of HSV-1. Alzheimer's disease is a progressive neurodegeneration neuropathology characterized by the presence of extracellular amyloid plaques composed of amyloid beta (Ab) peptides and intracellular neurofibrillary tangles. The scientists used human induced pluripotent stem cells (hiPSCs) to compare patterns of Ab42 accumulation in 2D (monolayer of neurons) and 3D neuron cultures (organoid brain) infected with HSV-1 (58,59). Their in vitro models showed that hiPSC-derived CNS neurons allowed HSV-1 infection and that 2D neuron cultures showed Ab42 immunoreactivity mainly in HSV-1-infected cells, rarely in uninfected cells or in infected cells that come in contact with antiviral drug. In contrast, 3D brain organoids showed that Ab42 was mainly concentrated in uninfected cells surrounding HSV-1infected cells. It is suggested that HSV-1 infection is a predisposing factor to the onset of Alzheimer's disease. Further results showed that HSV-1 induced the accumulation of Ab42 in monolayer culture of neurons, but antiviral therapy prevented it accumulation (58,59). In another study, researchers generated in vitro models of neurodevelopmental disorders including singlelayer neuronal differentiation of hiPSC, 3D shoots of neuroepithelial cells, and 3D brain organoids to study fetal brain development and potential effects of neuropathic HSV-1 infection, the results revealed that NSCs-infected with HSV-1 showed impaired neuronal differentiation, cortical and brain regionalization, and fetal neurodevelopmental disorder (55)(56)(57). Furthermore, it was emphasized that HSV-1 infection in the human brain showed a decrease in the thickness of cortical plate and a decrease in the expression of the LIM homeodomain transcription factor ISL1 (55). The 3D brain-like organ model also showed that HSV-1 infection can promote abnormal microglial activation and is accompanied by the induction of inflammatory factors such as Tumor necrosis factor (TNF)-a, interleukin (IL)-6, Il-10, and IL-4. HSV-1 infection inhibited cell population growth, at least in part by inducing apoptosis. The influence of HSV-1 infection on fetal brain development is highly dependent on the viral load of HSV-1 and the target cells (56).
In general, HSV-1 cannot be reactivated effectively, but it is known that it can be reactivated from peripheral neuron culture. HSV-1 is also difficult to reactivate due to latency in brain organoids, which is parallel to the inefficient reactivation of HSV-1 in the CNS (relative to peripheral nerves), therefore, brain organoids provide an effective model for studying the pathogenic mechanism of HSV-1 (57).
BRAIN ORGANOIDS AS MODELS OF PRION INFECTION
Prion disease is a deadly neurodegenerative disease that has attracted widespread attention because of its transmissibility. Sporadic Creutzfeldt-Jakob disease (sCJD) is the most common prion disease in humans. Hereditary prion diseases are relatively rare and are associated with mutations in prion protein genes. To date, more than 50 different point mutations, deletions, and insertions have been identified. Most are autosomal dominant and completely permeable. Prion diseases also occur in animals and are of great concern due to their potential to be transmitted to humans (60). Prion diseases are usually sporadic, usually due to genetic mutations in the gene that encodes the prion protein or to exposure to prion-contaminated material. Different molecular subtypes of prions have been found to influence the clinical and pathological phenotypes in sCJD. Pathological features of prion diseases include loss of neurons, activation of microglia and astrocytes, spongy changes, and aggregation of pruritic prion proteins. However, the pathogenesis of prion diseases is not fully understood and diagnosis is possible only when the disease has progressed to intermediate or advanced stages (61).
iPCS-derived human brain organoids can be used to simulate prion transmission and reproduction, providing a powerful research tool to study the pathology of human prion diseases caused by different subtypes and as cell models to test treatments (61,62). iPSC models derived from patients with Gerstmann-Sträussler-Scheinker syndrome (GSS) were used to study the molecular mechanisms involved in prion disease. A detailed analysis of immunoreactive cells showed that at the later stages of the human prion infection, proteins harboring the pathogenic mutation Y218N in culture, induced hypertrophic reactive astrocyte cells containing high levels of glial fibrillary acid protein and formed thick bundles. Finally, the nuclear staining analysis of the differentiation culture showed that chromatin condensation and apoptosis increased in cases with Y218N mutants. In order to replicate the human prion infection and pathogenesis, the researchers inoculated brain tissue homogenates of patients with different subtypes of spontaneous CJD to brain organoids, which revealed that the amount of prions invading brain organoids was influenced by the spontaneous CJD subtypes (62). The researchers also identified human APOE E4 as a risk factor for CJD (63). Furthermore, the researchers found that prions are transmitted primarily vertically, but studies in brain organoids have shown that prions can also be transmitted horizontally through vesicles without cell-to-cell contact (64). Therefore, brain organoids are an ideal model for studying the pathogenesis and transmission of prion.
BRAIN ORGANOIDS AS A MODEL FOR HUMAN CYTOMEGALOVIRUS INFECTIONS
Human cytomegalovirus virus (HCMV) infection is one of the main causes of morbidity and mortality in immunodeficient patients and the most common viral infection in developing human fetuses. Most HCMV infections are asymptomatic; however, when primary (and sometimes non-primary) HCMV infections occur in pregnant women, the virus can cause damage to the CNS (65)(66)(67). Thus, the development of the HCMV vaccine is considered a public health priority (66). However, due to the uniqueness of the epidemiology of HCMV infection in mothers, the general strategy of vaccine development is limited (67). It is particularly important to find an effective model for vaccine development. One study showed that the HCMV strain TB40/E can infect human brain organoids and induce growth and structural abnormality of human brain organoids. A neutralizing antibody targeting HCMV pentamer complex (PC) epitope can effectively prevent the infection of human brain organoids, and thus ensure the normal growth of human brain organoids and the formation of cortex. Further study showed that HCMV infection in human brain is related to platelet-derived growth factor receptor (PDGFR) and epidermal growth factor receptor (EGFR), and it does not seem to be dependent on integrins such as integrin a3, a5, or b3 (65). Thus, the use of brain organoids offers new hope for investigating pathogenic mechanisms and developing a vaccine for HCMV.
Challenges and Future Perspectives
Although brain organoids offer new hope for studying human neurological disorders, they are still in their infancy and still have many limitations. First, more than 2 days of cell culture or selforganized neural rosette formation is required, which adds to the cost of each culture, in addition to the specialized and complex culture conditions, thus model requirements are higher (1,68). Second, in the absence of an embryonic axis to guide the development of the fetal brain, as in other neural cultures, brain organoids cannot mimic the overall shape of a developing human brain. The heterogeneity and inconsistency of brain tissue cultures in vitro limit its reproducibility in the study of disease pathogenesis (68). Third, current brain organoids lack the components of a normal host microenvironment, such as immune cells and blood vessels. Due to the absence of immune components in brain organoids, it is difficult to reproduce a complex immune response and inflammatory processes following viral infection. Because brain organs do not have a vascular system, the size of the growing brain organs will be limited to 5-10 mm and may not integrate effectively with host tissues (1,67).
Despite these challenges, the brain organoid, or brainoid, model represents an innovative and effective in vitro approach that encapsulates key features of a normal brain in vivo that many conventional cell and animal studies have proven difficult to achieve, and also replicates the pathological processes that mimic neurological disorders at gross morphological, protein, and genetic levels (68,69). Genetic manipulation of brain organoids could be accomplished by stable editing of the hPSC genome or by direct infection of brain organoids by electroporation or retrovirus-associated viruses. Therefore, it is possible to study how individuals or combinations of ZIKVencoding proteins or noncoding RNAs derived from ZIKV affect the development of brain-like organs (70). Furthermore, coinfection with HCV-HIV or co-infection with HIV-Mycobacterium tuberculosis are very common in the clinic, although it remains to be studied whether co-infection has a more severe impact on the CNS; in this context, the brain organoids represent an effective model to investigate the effects of these co-infections (3,54).
The combination of human-specific processes with experimental plasticity models to explore the flexibility of spatiotemporal molecular mechanisms of disease has profound implications for patients with neurological disorders (71). In the future, the application of brain organoid bioengineering technology will greatly improve the vascularization of brain organoids, introducing the interaction of nerves and blood vessels, and prevent necrosis, to enhance physiological representation of the complexity of the human brain. Most current models do not effectively reflect the inflammatory response because they lack microglia. Therefore, an important addition to future models is the incorporation of immune cells and endothelial cells into brain organoids to fully understand how microbial fluctuations may regulate immune cell responses. In addition, multiple brain regions can be assembled into a complex system that ultimately reconstructs a complete brain-like organ that fully simulates a whole brain infection. Thus, highly regenerative brain organoids that better encapsulate the complexity of the human brain will be a powerful disease model with the capacity to comprehensively mimic brain disease, and will represent a revolutionary drug development strategy leading a new chapter in brain science research and personalized treatment.
AUTHOR CONTRIBUTIONS
FB, AL, and XS conceived and designed the study. PY, JK, and XX conducted the database search and screening. YZ, WC, and YF evaluated the data. ML, and JC conducted the quality assessment. XS drafted the manuscript. FB and AL revised and approved the manuscript. All authors contributed to the article and approved the submitted version. | 2022-01-11T14:16:54.854Z | 2022-01-11T00:00:00.000 | {
"year": 2022,
"sha1": "436f28a271ac863333fe1d6130bc8dd50794168c",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Frontier",
"pdf_hash": "436f28a271ac863333fe1d6130bc8dd50794168c",
"s2fieldsofstudy": [
"Biology",
"Chemistry",
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
229444014 | pes2o/s2orc | v3-fos-license | Interrogating the Role and Value of Cultural Expertise in Law
: It is common for litigation to draw upon expert evidence to assist a judge to arrive at a balanced decision. This paper examines the role of one type of expert evidence submitted to courts, namely cultural expertise (CE), which provides information on socio-cultural issues such as kinship, family, marriage, customs, language, religion, witchcraft and so on. This type of evidence is primarily the result of qualitative, ethnographic research. I begin by examining the views of experts who have provided CE to courts / mediators; I then look at how judges view and make use of CE, and finally I examine lawyers’ views on CE. To address gaps in published research, I interviewed British barristers to understand how they make use of experts in the cases they litigate. Finally, I have surveyed legal decisions made by all British appellate courts to arrive at an approximate idea of the extent to which CE has been submitted in English and Welsh courts. I conclude that the extent to which CE—and other types of socio-legal evidence—is submitted varies considerably depending upon the legal / evidentiary procedures followed in di ff erent jurisdictions and in di ff erent countries.
Introduction
Scholarship on the relation between anthropology and law has slowly recognized that characterizations of the two disciplines as being epistemologically incommensurable are unhelpful and unproductive. Anthropology seeks to address social life in all its variety. As anthropologists have argued, culture as a system of values and meaning is unbounded, dynamic and involves a process of redefining and contesting social norms and forms of power. Law, on the other hand, is a process by which those in power attempt to regulate conflict by deciding/adjudicating the limits of acceptable -restated as enforceable 'legal'-behavior and social norms. In this view, anthropology and law are mutually constitutive, even though many legal professionals perceive their role as being neutral, objective, and independent of cultural concerns and many anthropologists argue that anthropological evidence on cultural issues is routinely rejected by the law.
As the legal process has engaged with socio-cultural and technological change it has had to deal with a more socially diverse range of litigants/claimants, the law has increasingly accepted evidence from a growing range of 'experts'-from the social, cultural, behavioral, scientific, technical, judicial and other fields-who have provided extensive information about the impact of change on social and cultural life. Arguably, in grappling with social change, law has been forced to address the ubiquity of culture as it decides the rights and wrongs of individuals from diverse cultures, communities, and states.
In this paper, I assess the emerging role of 'cultural expertise' or CE in judicial proceedings in western Europe and North America, a process which has long antecedents but which has recently been spurred by the work of Livia Holden (2011Holden ( , 2020 and her European Research Council funded project
The Perspective of Cultural Experts
The earliest paper exploring the use of cultural experts was written by Clark (1953), who described the work of anthropologists and psychiatrists involved in civil litigation in the 1950s which overturned racial segregation laws in the U.S. Clark identified a number of difficulties facing experts involved in these cases and argued that 'it will be necessary for the professional societies among 1 See: https://culturalexpertise.net/ (accessed on 9 July 2020). My appreciation of CE has benefited greatly as a result of participating in this project as a research associate. 2 Holden has convened a number of seminars and conferences with experts, lawyers, and judges from across Europe which have been summarized but not analysed. I cite these problematic and somewhat contradictory 'data summaries' below. 3 See, for instance, rule 704 in the U.S. Federal Rules of Evidence at: https://www.law.cornell.edu/rules/fre/rule_704 (accessed on 15 September 2020). the social and psychological sciences to develop safeguards against possible ethical abuses; e.g., flagrant manifestations of prejudice, distortion of data and deliberately misleading interpretations' (pp. 9-10).
Subsequently Rosen (1977) wrote about the involvement of anthropologists in litigation between 1950 and the early 1970s. Rosen discussed anthropological involvement in civil litigation in the U.S. which sought to overturn racial segregation and which supported First Nation peoples. He identified a number of landmark legal cases-including Brown v The Board of Education, Plessy, Loving v Virginia, Wisconsin v Yoder and United States v State of Washington-in which anthropological evidence played an important role in striking down segregation and anti-miscegenation laws and upholding the importance of culture to secure the rights of religious minorities and native land claims. 4 Today, it is clear that anthropologists, historians, geographers and individuals from other disciplines have engaged with the law on a global scale with respect to legal claims made on behalf of, and sometimes against, individuals and social groups. With regard to the individual claims, anthropologists and others have provided expert evidence in cases involving crime (Fontein 2014), genocide (in the international criminal courts ;Eltringham 2013;Anders 2014;Wilson 2015), in claims affecting different diasporas, i.e., South Asians (Holden 2011;Menski 2013); Africans (Clarke 2017) and in asylum and immigration proceedings in the U.K. (Good 2004(Good , 2007(Good , 2008Hoehne 2016;J. R. Campbell 2017J. R. Campbell , 2020a, the U.S.A. (Berger et al. 2015;Ngin 2018) and Sweden (Rabo 2019). Anthropological evidence/CE is likely to be of value in asylum proceedings for two reasons. First, Immigration Judges (IJs) applying international humanitarian law are required to understand whether an asylum applicant has 'a well-founded fear' of persecution which requires them to assess an applicant's account of persecution and flight, i.e., his/her subjective and objective fears. Second, unlike courts in other jurisdictions 5 , asylum courts are allowed to admit hearsay evidence as well as other evidence which an appellant may wish to submit (though it may attach little weight to this evidence). The focus of the work by experts has been to use anthropological/CE reports-which is uniformly viewed as hearsay evidence-to obtain recognition/protection for individuals.
The second focus of expert evidence relates to disputes concerning indigenous people's land claims. In these cases, anthropologists and historians have provided ethnographic and ethno-historical data in 'native'/indigenous land disputes and in related political struggles in Alaska and Canada (Feldman 1980;Ray 2011), in Australia (Weiner 1999) and in central and South America (Loperana et al. 2020;Hale 2020). However, in recent years, this type of litigation has seen a growing 'refusal' by First Nations/native Americans to engage with the law in its terms in an attempt to refuse political solutions imposed by the state (Hale 2020;Loperana et al. 2020). This paper focuses on litigation by or on behalf of individuals.
Most of the literature on expert witnessing is written primarily by anthropologists concerned with legal proceedings in which 'culture is teleologized in courts of law, by being treated as 'objective evidence" (Good 2008, p. S47). Some anthropologists have tended to distance themselves from lawyers and judges who are said to be 'empiricist and positivistic' and who 'are trained to think in radically different ways' than anthropologists and social scientists (Geertz 1983;Good 2004, p. 129;Bens 2016). However, as Loperana et al. (2020, pp. 588-89) have argued, this distinction obscures the fact that both anthropology and law emerged as powerful forms of knowledge during colonialism. 4 Gormley (1955) notes the involvement of anthropologists in 'Indian tribal claims' dating back to 1895 and possibly earlier. He cites 'Choctaws et al. v. United States,34 C.Cls. 17,54 et seq.' 5 In the UK, sec. 114 of The Criminal Justice Act 2003 defines hearsay evidence as any 'statement not made in oral evidence in the proceedings.' Reliance on a statement made otherwise than while giving evidence to prove the truth of a fact asserted remains hearsay. Hearsay evidence is inadmissible and the rule applies: (a) to both examination in chief and cross-examination; (b) whether the statement was made by the witness personally or by some other person; (c) to any 'out of court' statement, whether oral, written or otherwise; and (d) to statements given as evidence of the truth of its contents-if the statement is given for any purpose which is relevant to the facts in issue in the case, it is admissible, for example, evidence given as to a person's state of mind, rather than what was actually said. Furthermore, as Eltringham (2013) and Wilson (2015) have argued, an assessment of anthropological involvement in the legal process requires a clear understanding of the adversarial legal process. This begins with lawyers identifying and instructing experts in an attempt to win a case, and culminates when a judge, having assessed the evidence and legal arguments, arrives at a decision. Engaging with the law as an expert may reify and strengthen the law by reinforcing the role of the state over an individual or a community; though occasionally experts may provide evidence which successfully challenges and overturns case law. 6 (cf. Rosen 1977). It is important to step back from Clarke's (2020, p. 585) statement on the different tasks of the two disciplines when she states that: 'Where the court's purpose is to establish hegemonic order through judgement, one of anthropology's purposes is to illuminate alternative cosmologies and possibilities rendered by diverse subjectivities'. Experience with the legal process teaches us that the law only perceives a defendant/litigant as either guilty or not guilty; judges in criminal and other jurisdictions have little patience with arguments about culture/cultural relativity, witchcraft/sorcery or which seek to challenge the common sense reasoning they rely on (when this occurs, judges quickly act to rule such arguments inadmissible; (Fontein 2014;Anders 2014)). I argue that the specific role played by the two disciplines needs to be examined on a case by case basis. Hoehne (2016, p. 253) has argued 'that it is not a fundamental epistemological divide, but rather massive power differentials that characterize the relationship between social anthropologists and legal practitioners' and that this difference 'sits uneasily with the professional, moral and ethical standards of the discipline'. Hoehne argues that anthropologists providing evidence in courts of law must adopt a 'strategic form' of essentialism, as opposed to a post-positivist position reflecting the contingent nature of anthropological knowledge if they 'are to fulfil the requirements of the legal process and maintain one's role as expert' (p. 257). Hoehne's point is echoed by H. Campbell et al. (2017, p. 333) who also make a case for 'strategic essentialism'-which the authors define as 'the pragmatic practice of defining cultural groups or their practices in ways that emphasize commonalities rather than differences while recognizing that a more expansive analysis would include greater nuance and complexity' 7 -via a well-informed understanding of immigration and criminal law. They argue that anthropologists should defend the rights of subaltern peoples caught up in powerful legal proceedings that may unfairly imprison, execute or deport them to their country of origin. The authors argue that anthropologists 'need to find a way to use our ethnographic expertise to not just defend individuals' caught up in legal proceedings 'but also leave a positive social and institutional [i.e., professional] record' (H. Campbell et al. 2017, p. 333). For these anthropologists, witnessing is a political act.
The issue of the legitimacy attached to an expert's evidence was raised by Rosen (1977, p. 555) who observed that most anthropologists 'may not understand how expert testimony fits together with judicial reasoning and legal precedent, and precisely how the court's investigation of the facts articulates with the form of knowledge he possesses.' In particular, and in relation to the rules governing testimony/evidence which allow 'expert witnesses' to testify about specific social and physical phenomena which they have not personally witnessed 8 , Rosen draws attention to four issues which anthropologists-and by definition other types of expert-must take into account if their work is to be effective. First there are 'concerns about the adequacy, context, and form of presentation of anthropological evidence in an adversary proceedings' (p. 556). Is testimony adequate given the issues 6 See the civil cases identified by Rosen (1977) and ST (Ethnic Eritrean-nationality-return) Ethiopia CG [2011] UKUT 00252(IAC), an asylum claim decided by the Upper Tribunal of the UK's Immigration and Asylum Chamber which overturned five precedents on the basis of expert/CE evidence. 7 As Clarke (2020, p. 587) has noted, the term 'strategic essentialism' was introduced by Gayatri Chakravorty Spivak to refer to 'a political tactic in which minority groups, nationalities, or ethnic groups mobilize on the basis of shared gendered, cultural, or political identities to represent themselves.' The contrast between how anthropologists describe minority groups in legal claims and how the latter describe themselves is important and is addressed by Hale (2020) and Loperana et al. (2020). 8 In the UK, experts are permitted to testify about relevant socio-cultural issues, culture, family, foreign law, and a growing range of new topics related to the rise of technology, for instance, video/surveillance film, DNA, drone technology and so on. raised by a case? Will evidence be inappropriate or distorted by adversarial proceedings? What evidentiary standards are anthropologists required to meet? Second, there are questions regarding 'the mutual effect that courts and anthropologists have on one another' (p. 557). Have anthropological concepts been affected or appropriated by litigation? How has anthropological testimony been shaped by adversarial argument and judicial reasoning? How can anthropologists balance their work as experts against wider professional obligations? Third, Rosen notes that there are serious questions regarding 'anthropologists conceptions of their role in such proceedings, and how the courts and the profession may contribute to appropriate reforms' (p. 557). For instance, do anthropologists provide crucial information for judicial decisions or is their information 'simply useful for rationalizing judgements that are founded on other, perhaps judicially less palatable bases?' Rosen asks, 'How should anthropology approach the ethical implications of expert testimony?' There are, therefore, important political and practical issues which can undermine the effectiveness of experts and the extent to which their testimony will be recognized by a court as constituting valid evidence (and be accepted by their profession) which relate to the jurisdiction in which a case is heard/tried and the epistemological approach followed by the expert. Two examples will have to suffice. While Immigration Judges (IJs) in the U.S. 'have broad discretion to conduct and control immigration proceedings and to admit and consider relevant and probative evidence, including witness testimony', their discretion is not without limit. In the 'Matter of J-G-T-Respondent' 9 heard by the Executive Office for Immigration Review in 2020, the Board of Immigration Appeals decided that, . . . in assessing whether to admit the testimony of a witness as an expert, an Immigration Judge should consider whether it is sufficiently relevant and reliable for the expert to offer an informed opinion, and if it is admitted, the Immigration Judge should then consider how much weight the testimony should receive. If a party challenges the expert's qualifications, it is generally best to allow the party, upon request, to voir dire the witness before the testimony is presented in full. In considering how much weight to give an expert's testimony, the Immigration Judge should assess how probative and persuasive the testimony is regarding key issues in dispute for which the testimony is being offered. However, to the extent that the record contains contradictory evidence, the Immigration Judge should explain why inferences made by the expert are reasonable and more persuasive than the other evidence presented.
Second, and in relation to transitional justice cases which are heard by International Criminal Tribunals, Jones (2015) raises important questions about the status of expert evidence and how it is evaluated. She notes that different researchers and organizations involved in providing expert evidence to international tribunals follow different epistemological approaches, provide different 'versions of reality' and are able to secure varying levels of legitimacy. Jones argues that in transitional justice, 'first-hand experience' of the countries in question 'and more ethnographic and empirical understandings 10 were generally dismissed by the bench of international judges' in favor of the 'foreign expertise' provided by international NGOs whose evidence more closely resembled the judges own 'legal and fact-based approach' (p. 296). For Jones, the key question concerns the legitimacy which the courts attach to the versions of reality provided by different experts: she asks, whose voices-not only which experts but which local narratives-are being heard? In this regard, Wilson's (2015) review of how qualitative research faired in international criminal trials suggests that social scientists should not attempt to challenge the authority of the court and that reports should be written in clear, non-technical and jargon-free language. His analysis indicates that evidence provided by social 9 This case is cited as Matter of JGT-, 28 I&N Dec. 97 (BIA 2020: https://www.justice.gov/eoir/page/file/1319951/download). 10 The author explores the value of narrative interviews in providing relevant evidence, but there is no reason to limit research to this method.
scientists was more frequently cited in judicial decisions than that of military and police experts, document verification experts, financial experts, engineers, and medical experts (p. 732). The final issue raised by Rosen is what he calls 'the cultural argument', though as discussed below, 'culture' is not the only issue which anthropological/CE experts are asked to address. As Rosen frames the issue, legal cases present problems of interpreting to the court the language and concepts of the party involved, and the relation between the legal issues posed and the relevance of anthropological findings. Working closely with counsel, anthropologists have also been instrumental in formulating highly creative arguments that may influence the course and result of a case (p. 567).
Rosen cites several successful 'creative arguments' which includes litigation about the use of peyote by Native Americans, snake-handling cults in the state of Tennessee, the issue of 'arctic hysteria' and 'witchcraft murders'. The creation of a 'cultural defence' in the U.S. arose later (see Dundes Renteln 2002. The issue of whether it is right that anthropologists should 'interpret' culture rather than allow litigants an opportunity to speak for themselves in legal proceedings is determined in part by whether the litigation strategy adopted by legal counsel empowers parties to speak for themselves and by the disciplinary training of the experts who submit evidence. In group legal claims, such as land right claims, there has been a shift away from the anthropologist acting as the 'expert' to one of facilitating local groups to speak for themselves (Loperana et al. 2020;Hale 2020).
The social sciences have still not addressed the professional and pragmatic issues raised by Clark and Rosen nearly forty years ago. A lack of preparedness for adversarial proceedings can result in shock and surprise when judges dismiss an experts' evidence or when the judiciary appropriates and 'misuses' anthropological concepts and evidence (Riles 2006). 11 Apart from a desire to be useful to the courts, most experts possess a limited awareness about how the courts 'judicialize culture' 12 and they are unaware that they may be required to disclose their sources of information. 13 While professional associations have engaged with the general ethical issues facing members of their discipline, their guidance has left individual members to negotiate thorny issues arising from multiple and cross-cutting obligations to one's informants, and responsibilities to funders, one's university or employer, the nation-state in which fieldwork was conducted, and the courts. 14 The general approach to ethics adopted by professional bodies has had two principle implications for those who engage in expert witnessing. First, relatively few university academics engage in applied research, including expert witnessing (J. R. Campbell 2020a). Second, the ethical approach adopted by those who provide CE reflects their individual understanding of the challenges they confront. While some anthropologists would agree that they have an obligation to support subaltern people confronted by the law (Holden 2019, p. 190-f), they do not necessarily believe that they should challenge the judiciary-for example, by disputing judicial reasoning or challenging legal arguments made by the state-or take a political stance with regard to witnessing.
The problems which arise when anthropologists attempt to raise 'the cultural argument' are complex. First, even in instances where the courts have accepted cultural evidence provided by 11 It appears that only a small number of anthropologists who provide expert evidence are familiar with adversarial proceedings, namely Rosen (1977Rosen ( , 1989, Good (2007), H. Campbell et al. (2017), Dundes Renteln (2004) and J. R. Campbell (2015Campbell ( , 2017Campbell ( , 2020b. 12 See (Rosen 1989;Good 2007;Howes 2005). 13 See, for instance, the guidance set out by the UK's Crown Prosecution Service (2019a). 14 See Manners (1956) for an informed and early discussion of the complex issues which arose when anthropologists provided expert evidence in Native American land claims. For the guidance provided by the American Anthropological Association see: https://www.americananthro.org/LearnAndTeach/Content.aspx?ItemNumber=22869&navItemNumber= 652. The ethical code of the American Historical Association only notes that historians should avoid any 'conflicts of interest' (see: https://www.historians.org/jobs-and-professional-development/statements-standards-and-guidelines-of-thediscipline/statement-on-standards-of-professional-conduct#Reputation). The American Association of Sociology also sets out very general ethical guidelines, see: https://www.asanet.org/sites/default/files/asa_code_of_ethics-june2018.pdf (sites were accessed on 20 August 2020).
an anthropologist and the argument has been generalized for use across a range of cases-the best documented example is the concept of a 'cultural defense' employed in U.S. courts-the way in which cultural evidence is handled indicates that trial judges fail to apply a sufficiently robust test to assess the veracity of a 'cultural claim', because they rely upon their own 'common sense' to decide cases (Dundes Renteln 2005; Wilson 2015; Maucec 2020). In short, there is a significant difference between the way that anthropologists and judges understand and employ the notion of a cultural defense. It is also the case that the value of a legal victory may be more pyric than real depending on how the state implements the decision and whether a successful litigant is subordinated to the state. In addition, subsequent legal cases may challenge and undermine previous legal victories. This is certainly what occurs in civil and in asylum claims where the direction of the law is dependent upon who possesses the resources to use it. As Nader (Nader 2001(Nader -2002 has argued, in legal systems dominated by the state the plaintiff role tends to atrophy and, over time, the legal system disadvantages individual plaintiffs/claimants. Secondly, 'cultural issues' may not be the principle or only issue raised in a legal case. This situation may arise as a result of the 'liberal' approach adopted by a court when it fails to understand, take cognizance of or reconcile conflicting concepts of culture and justice (Douglas 2005), when a court frames the issues narrowly giving short shrift to underlying cultural issues (Roy 2005), when a court seeks evidence about foreign law and practice (J. R. Campbell 2020a) or when judges exclude expert evidence (Currie 2005;Wilson 2015). Jones (1994) provides an excellent overview of the English judiciaries' attempt to regulate 'expert' witnesses in trial proceedings. She observed that when the judiciary realized that experts disagreed in the methods they adopted and the conclusions they provided to the court-even though disagreement was and remains an essential aspect of scientific enquiry-they saw experts as 'suffering from partisanship' (p. 97-f). The judicial view of science and of scientific experts was compounded when experts were called by both parties and/or when fees were paid to experts to provide evidence. This situation contributed to a 'distorted' view of science by the judiciary, namely that scientists should provide evidence that was impartial, disinterested, and neutral. However, when expert evidence deviated from the ideological construct created by the judiciary, judges used procedural rules to 'dismiss expert evidence as the weakest kind of testimony.'
The Perspective of Judges on the Value of CE
The principle 'stick' devised by the judiciary to 'beat the expert witness' took the form of new rules of evidence and procedural/admissibility rules which 'forced experts to fit their evidence into an artefact of the law's design', namely a written report which addresses issues identified by lawyers. In most jurisdictions, experts have an 'overriding duty' to assist the court-which overrides their obligations to the party which instructed them-and their reports must conform to a specific format. 15 In the United Kingdom, for instance, experts must not advocate on behalf of the claimant. 16 The rules give the lawyer who submits the evidence maximum control over the expert and her evidence, and they give judges the power to admit or refuse expert evidence. 17 Furthermore, the adversarial system relies on the parties to cross examine experts in a process which seeks to undermine the weight that will be placed on their evidence leaving it to the jury or the judge-depending on the jurisdiction in which the case is being heard-to assess the value of the proffered evidence (Golan 2008, p. 917-f). In effect, experts have been co-opted by the legal process.
The judiciary regulates expert witnesses in part through imposing evidentiary hurdles which vary between different jurisdictions and countries. For instance, the Australian judiciary rely upon 'the field of expertise rule' which has been expressed as 'whether the subject matter on which expert evidence is being adduced 'is such as to be the proper subject of expert testimony" or is a 'recognized field of specialist knowledge' (Redmayne 2001, p. 100). In the U.S., the Frye and Daubert 'tests' have been devised by the judiciary as an attempt to shift the focus of admissibility away from the credentials of the expert to the specific nature of the scientific knowledge s/he proposes to submit to the court. In this way the Federal courts created a new 'law of evidence' that was intended to prevent the expert from giving his or her opinion on the 'ultimate issue', i.e., the precise factual issues before the jury (Golan 2008, p. 921-f). All the different tests, however, including the hearsay doctrine which allowed judges to exclude 'opinion evidence', have proved unwieldy and have been applied inconsistently by judges. As Redmayne has argued, the tests 'are so flexible that they could be applied strictly or laxly, depending upon the court' to 'screen scientific evidence to ensure its reliability' (Redmayne 2001, pp. 104-5). Such rules exist in many countries and may be used to regulate and prevent expert evidence from being given. In this regard, Currie (2005, p. 81) reports a civil defamation case in Canada in which the trial judge decided that the expert evidence to be provided by a sociologist and an anthropologist about racism in the local police force was inadmissible because sociology and anthropology 'lack the precision and specificity which characterizes a science like chemistry or an area of technical expertise like engineering.' 18 Admissibility is less likely to be an issue when a judge, rather than a jury, is deciding a case. This is particularly true with respect to hearing minor criminal offences, most civil claims (which are increasingly expected to be arbitrated/mediated) and asylum and immigration appeals. 19 While it is the case that in some countries there is a right to a trial by jury for serious offences and in important civil claims, the reality is that the number of jury trials for all types of claims have declined substantially in North America (Kritzer 2004;Galanter 2004) and the U.K. (Dingwall and Cloatre 2006). 20 It is important to add that across Europe, and in many other countries, civil claims are heard by judges and trial for criminal cases is rare (Leib 2007(Leib -2008. In France and Belgium, which follow the civil law tradition, only serious criminal offences are heard by a jury (Germain 2019).
Even though courts use evidence/admissibility standards, many lawyers and judges distrust experts who are seen as 'hired guns' because it is believed that their evidence can mislead juries and judges and may result in wrongly deciding cases (Mosteller 1989;Mantle and Chenane 2014;Wortly and Ward 2019). Judicial views of experts are not without some justification given the extent to which expert evidence is internationally marketed for all types of litigation 21 and because experts are appointed who possess diverse forms of disciplinary training and who have different motivations, including personal gain for providing their expertise.
It is useful to step away from a focus on the rules which judges can use to regulate expert evidence to understand the importance of CE in different countries. This task requires us to distinguish between proposals made to the judiciary which advocate the adoption of an 'appropriate cultural test' (cf. 18 On appeal the judge's decision was set aside. 19 Ruggiu 2019) from research which explores how judges actually assess and make use of CE. There are relatively few published studies which explore the second issue.
Cooke's (2019) study of Finland confirms that while it is possible for the courts to accept CE, this has not happened (though why this does not occur is not discussed). Instead she focuses on the role of 'informal' expertise which occurs when community members are appointed by a judge to address cultural issues. Lopes et al. (2019) 'survey' of Portuguese law identifies and discusses a small number of legal cases relating to migrants whose culture may be relevant when they are charged with specific crimes, i.e., Female Genital Mutilation, forced marriage, specific types of sexual acts, placing children for adoption, the marriage of minors and a child's right to education. In Portugal, expert evidence is only introduced 'through the initiative of a judge' which rarely occurs. The authors found that while 'the courts were relatively indifferent to cultural factors', an extensive network of 'mediators' operated at the community level 'who were mobilized to defend ethnic minority rights' (p. 66).
The situation in Italy is slightly better known. For example, Holden's survey of judges, lawyers and experts suggests a fairly small number of experts are appointed and that experts provide evidence primarily in immigration law (44%), refugee and asylum law (22%), family law (19%) and civil law (16%). 22 Ciccozzi and Decarli (2019) state that while a wide range of experts are recognized in Italian law, 'in most cases, judicial actors are not aware of the cultural complexities of the groups involved in trials and have no on-site experience with the populations to which the parties belong' (p. 37). Furthermore 'the number of experts appointed in Italian trials has been very low'. The authors provide a detailed discussion of Ciccozi's involvement as an anthropological expert in the 'L'Aquila trial' which involved an assessment of whether expert scientific evidence concerning the risk of seismic activities was understood (and ignored) by residents of L'Aquila village with fatal consequences.
With regard to the situation in Sweden, Holden's survey of judges, lawyers and experts 23 found that CE was primarily relied upon in immigration law (13%), refugee and asylum law (22%), international human rights law (11%) and in family law (10%). The majority of judges and lawyers had used experts in less than 10 cases. Rabo (2019) sets out a more nuanced history of CE in Swedish law in which she examines cases concerned with the land rights of the indigenous Sami people, cases of discrimination against Roma people and cases-concerned with deportation, adoption, hate speech, divorce, spirit possession, FGM and murder-involving 'new immigrants'. She notes that the Swedish courts employ the 'free sifting/consideration of evidence (fri bevisprovning)' which, in principle, means that there are 'no limitations concerning the sources that can be used by the parties in a trial' (p. 2). Nevertheless, until recently, the direction of the law has been in support of the idea that Sweden is a homogeneous society which has meant that foreigners have been seen as 'inferior' and that they have suffered discrimination as a result of this attitude. Rabo discusses, in some detail, cases litigated by the Roma and the Sami against the state for discrimination in which these ethnic minorities have had to find their own experts and have fought long legal battles against the state, the police, private companies and individuals. With regard to the 'new immigrants'-the largest group are Syrian refugees/immigrants-it is clear that there has been some difficulty in finding appropriate experts to give evidence and that while most have only submitted reports to a court, they have not been called to give oral testimony. Rabo concludes by noting that the term 'culture' is highly ambivalent because it has been increasingly used by populists/demagogues and 'mainstream opinion makers'. Because the Swedish state 'is quite blind to its multi-cultural history', she argues that it is up to anthropologists/cultural experts and more progressive lawyers to work together to pursue justice for minorities.
The situation in Poland differs from the above cited cases. In Holden's survey 24 of judges, lawyers and experts, it was found that 58% of judges and 45% of lawyers had never instructed an expert to provide CE. When CE had been submitted, it was predominately in family law (15%), crime (14%), refugee and asylum law (14%) and in immigration (12%). Burdziej (2019a) provides a more in-depth study of CE in Poland. He identifies the need for CE to deal with growing cultural diversity facing the immigration services, detention centers and in education-to address issues relating to migrants-and in the military (to prepare Polish troops who are deployed overseas). He undertook a key word search of legal databases-which held four hundred eighty-eight thousand decisions by Poland's appellate courts, administrative courts, and the Supreme Court-which yielded four hundred cases in which some form of CE was submitted (p. 12). He found that in the rare cases which involved CE that the majority 'concerned aspects of Polish dominant culture, not intercultural issues' (p. 13). The principle cases identified were asylum cases, hate speech crimes, defamation cases, media ethics, copyright, and religion. Burdziej found a very small number of cases which turned on CE: namely an honor killing, an attempt to kill a man responsible for an immigrants' daughter's suicide and Islamic terrorism. 25 He thought that many cases which raised cultural issues were simply 'filtered' out of the legal system 26 because 'Poland's relative cultural homogeneity leads courts to perceive culture as a largely non-problematic issue, thus not requiring the assistance of experts' (p. 25).
Burdziej has written a second paper titled 'Judging the Communist Past' (Burdziej 2019b) in which he provides a fascinating analysis of the use of historical expertise submitted to Polish courts which hear cases involved in 'lustration' ('vetting') the right of individuals to hold public office who are accused of collaborating with the communist secret services', cases to change the names of public places named after prominent communists, the removal of Communist monuments and the 'withdrawal of veteran status' from soldiers of the communist security services. The evidence submitted by historians comes from their analysis of what is left of the Stasi (secret service) archives (many files were destroyed or 'privatized' for the purpose of blackmail). The work of this court has become highly politicized.
A special appellate court was established in 2006 as was a new law of lustration which defined key definitions of what acts amounted to collaboration. Between 2008 and 2015 the court issued 903 decisions: 459 individuals were found to have been collaborators. Observers have noted the excessive leniency of the court which has led Burdziej to conclude that 'the courts often prefer erring on the side of the defendant' (p. 6). Controversy surrounding the courts decisions have, in part, ensured that prosecutors who submit evidence are 'careful to limit their claims to factual statements on the contents of the (Stasi) archive and not their interpretation as evidence of collaboration' (p. 8). Burdziej argues that this historical expertise has become institutionalized by the state, through the creation of the Institute of National Remembrance, in an effort 'to secure victory in the struggles over the dominant interpretation of the country's past ' (p. 22).
In Denmark, which has 'a particularly monocultural, legally homogeneous, monolithic' legal system, Vinding (2019) argues that legal culture has changed very slowly as a result of the impact of migration, religion and EU legislation. Vinding undertook a key word search of the Danish legal database 'Karnov' for the period 2001 to 2018 which held approximately thirty-six thousand cases. He was able to identify one hundred and eight cases in which CE was submitted. The cases primarily concerned the Immigration Service and the Police Intelligence Service, and they concerned expulsion, terrorism, the prosecution of hijackers and cases brought by Iraqi nationals against the Danish Defense 24 See: https://culturalexpertise.net/wp-content/uploads/2020/04/Poland.pdf. 25 Terrorist criminal trials, and Closed Material Proceedings, need to understand the statements and concepts relied upon by the accused to justify their actions including the notion of jihad, Da'wah and other Islamic concepts. However, courts tend to disregard evidence on social, cultural, and contextual issues (see J. R. Campbell 2020b). 26 If a 'filtering out' is occurring, this would also involve the police and state/public prosecutors, and would suggest that an examination of judicial decisions alone is only able to provide a partial understanding of the legal process and the role of CE.
Forces. Vinding argues that there is a need for the Danish courts to accept CE, but also a considerable reluctance by the judiciary to abandon its traditional approach to deciding cases. Vinding's conclusions are supported by the discussion which Vetters and Foblets (2016) had with European judges. They concluded that judges are heavily influenced by their 'legal culture' and 'working environment' which predisposes them to adopt a pragmatic understanding of the 'culture' of immigrants 'that is not so much built on a clear-cut distinction between 'our' and 'their' culture, but rather focuses on notions of culture more directly related to the legal and organizational context within which judges operate' (p. 273). Based on a limited look at asylum proceedings in three countries-the U.K., Belgium, and Germany-the authors argued that judges are constrained by evidentiary norms from introducing CE.
Finally, a study by J. R Campbell (forthcoming) analyzes the role of CE in the British asylum system-including the rules which govern the submission of evidence-in first instance asylum hearings and on appeal to the Upper Tribunal of the U.K.'s Immigration and Asylum Chamber. These judges rely on CE/expert evidence to decide cases. Of the twenty-eight precedent-setting 'country guidance' cases decided by the Upper Tribunal between 2015 and 2019, he analyzed five and found that Immigration Judges use procedural rules to hedge and control expert evidence and that their assessment of this material is often flawed. This is particularly the case with qualitative data, which is treated as hearsay, but problems also occur with their preference for, but inability to adequately understand/ 'test', statistical data.
Campbell concluded that Immigration Judges lack the training to adequately assess/test testimonial, qualitative and statistical evidence. The problem is that they do not realize their limitations and end up preferring evidence which they assume is more objective or scientific while setting aside cultural evidence on language, culture, kinship and the importance of social relations and social networks. Legal conceptions of objective evidence used by IJs . . . do not assist them to assess social science research . . .
Holden reports on a survey of judges, lawyers and experts in the United Kingdom. 27 She reports that the majority of judges have never instructed an expert (33%; whereas 61% had instructed 10 or less reports) whilst 47% of lawyers had instructed less than 10 cases. Her findings are highly problematic, given that judges are not supposed to instruct experts-though they might be involved in agreeing which experts should give evidence-and solicitors but not barristers should instruct experts.
Research on international criminal courts provides a different picture of the role of CE. Maucec (2020) has, via the examination of a small number of cases, examined the limited availability of expert evidence on 'cultural' issues-i.e., efforts to mount a 'cultural defence', the meaning and use of 'fetishes and mystical power' and 'speech crime'-that have arisen in the International Criminal Court (ICC). Once again, however, the writer's belief in the value of cultural expertise is belied by the limited extent to which CE appears to have been submitted to and accepted by the ICC. Anders (2014, p. 427) has written about the Special Court for Sierra Leone and argued that the judges 'refused to adopt the anthropologists' arguments (against the evidence relied upon by the prosecution) and did not share their concerns about the methodology and conceptual framework employed by the prosecution experts.' Anders' argues that that anthropological evidence was rejected because the judges 'were reluctant to recognize the challenge posed by Sierra Leone's socio-cultural specificities to the application of international criminal law.' In sharp contrast, Wilson (2015) compiled and analyzed a database of 473 expert reports submitted to the International Criminal Court for 27 See: https://culturalexpertise.net/wp-content/uploads/2020/06/UnitedKingdom.pdf. the former Yugoslavia. He found that while 'scientific experts' were called to give evidence twice as often as were other experts, in fact 'social researchers (who provided both quantitative and qualitative data) have a much higher citation rate' than medical, military and police, document verification, financial and engineering experts (p. 732). Judges welcomed socio-cultural and linguistic evidence; however, their receptiveness to expert evidence 'depends on how jealously judges protect their preeminence as the triers of fact' (p. 741). Wilson's findings are supported by Eltringham (2013) who studied the International Criminal Court for Rwanda. There are clearly a number of factors which influence the reception of CE in international criminal courts, which, in addition to the issues already discussed, include judicial conceptions about culture as something practiced by 'distant (read backward) communities', the court's failure to appreciate its own organizational culture (as somehow objective and situated above socio-cultural specificities) and the judges responsibility to assign culpability (Fraser and Leyh 2020).
The above studies provide insight into the diverse ways that courts in different countries and jurisdictions view, instruct, admit and assess CE/expert evidence; however, the studies rely on a small number of published cases and fail to engage directly with judges. In what follows, I seek to fill key gaps in the research cited above, which has not adequately assessed the views of lawyers and the judiciary regarding the value of CE, nor has research assessed the full extent to which CE/socio-legal expert evidence has been submitted in national legal systems.
How British Barristers Make Use of 'Experts'
To obtain a clearer picture of the role of CE in the law, I interviewed five London-based barristers who undertake work on behalf of claimants (not the government). The zoom-interviews which ranged from fifty to ninety minutes in length were recorded and transcribed into Word. I asked barristers to identify and discuss at least three legal cases which they were personally involved in which relied upon different types of expert evidence. I also asked barristers to provide me with relevant case material or summaries of their cases which were published on Bailii. Because the published cases identify my informants, I have drawn upon but not cited this material in order to ensure the anonymity of my informants. The barristers I interviewed worked in the following jurisdictions: asylum and immigration law, family law and public law.
The Asylum and Appeal Act 1993 28 made it possible for asylum claimants to appeal against the decision of an Immigration Officer to the Immigration and Asylum Tribunal; at roughly the same time, legal aid became available to pay lawyers to litigate these cases. The appeal procedures were initially 'informal' and lawyers relied on reports published by international human rights organizations or the U.S. Department of State; it was rare for an expert to be instructed to provide a report and unheard of for an expert to be cross examined. Over time the situation changed: argument in the Tribunal has become increasingly 'legalized' and 'experts' have increasingly been instructed to provide evidence. Rather than provide its own expert evidence, the Home Office criticizes the experts instructed by claimants and attempts to undermine their reports (Home Office 2005, p. vi). In addition, once lawyers were appointed as Immigration Judges, it became increasingly common to cross-examine experts. This shift has meant that all forms of evidence are much more closely scrutinized in a process which has seen experts who rely on qualitative research and those whose reports are based on the analysis of documents in the public domain, i.e., reports by international organizations, to be deemed by IJs as insufficiently credible. Barristers argue that an analysis of country guidance decisions shows that IJs prefer experts who can analyze published statistics on casualties, death rates, risk of injury, mental health, and so on for a given population rather than experts who possess a good knowledge of an asylum applicant's country of origin.
Barristers rely upon the solicitor who instructs them in a case to identify and instruct relevant experts. Solicitors also set out the Terms of Reference which experts are required to address; barristers may amend the ToRs but they are seldom in contact with experts. Barristers who litigate immigration and asylum claims rely heavily on 'country' experts-anthropologists, sociologists, historians, linguists and journalists who provide CE, as well as experts on foreign law-whose claim to expert knowledge rests on having conducted academic research on the country from which an asylum applicant originates. To identify an expert, immigration solicitors rely on recommendations from other practitioners and/or they contact experts listed on an online directory of country experts. 29 Because some barristers are skeptical about the individuals listed on the directory, they undertake research to identify a suitable expert, they read case law, google university webpages and consult colleagues in their chamber. Once identified and instructed, experts vary in their ability to fully address the terms of reference, in terms of their knowledge of the specific issues raised by a case, and in terms of their willingness to revise their reports to address barristers' concerns before their report is submitted to the court. Barristers who identify experts for their cases are less likely to end up relying on a person who produces generic material and/or who 'recycles' material which they have submitted in other cases (such experts tend to carry little weight with IJs).
Apart from selecting an expert who can address the specific issues raised in a case, a barrister's litigation strategy reflects a number of factors. First, all barristers had very clear views regarding how judges sitting in the jurisdiction in which they worked were likely to assess evidence. Thus, all the barristers' I spoke with held negative views regarding how IJs in the Immigration and Asylum Tribunal assessed expert evidence. One barrister thought that senior IJs were unwilling to take decisions which might be understood by the Government as 'impeding the operation of immigration control.' For this reason, barristers suggested that IJs failed to adequately reconcile international legal frameworks with British law in case their decisions challenged government policy. In order to 'square the circle' without overturning case law, it was suggested that IJs diminished the weight attached to expert evidence. 30 Barristers expressed a much more positive view of the quality of judge craft in jurisdictions where judges had been appointed to a judicial post following a successful legal career, i.e., the Family Division, the High Court, the Court of Appeal and so on.
Barristers who litigated across different jurisdictions provided further insights. For instance, appeals in the Family Court are often linked to claims in the Immigration and Asylum Tribunal which creates a 'gordian knot'-i.e., the need to ensure that both courts consider the decision of the other court-which can result in delays and problems for appellants. This situation arises in: deportation cases which raise an Art. 8 claim relating to the appellant's right to family life or the Human Rights claims of a child whose parent is to be deported; in care or adoption proceedings in the Family Courts; and with respect to claims regarding the threat of FGM to children threatened with deportation (Clark-Platts n.d.). 31 A barrister working in both courts stated that in the Family Courts-unlike in immigration proceedings-both parties must agree on a single expert to provide a report and that judges in the Family Court 'are much more likely to make positive findings from that evidence' and their findings 'tends to undercut adverse findings by IJs 'who criticize expert evidence and who 'think and act like border guards'. The barrister also noted that in the Family Court, legal counsel for the Secretary of State tend to adopt 'a more muted role' in proceedings. Interestingly, some barristers chose not to instruct experts particularly in the IAT when IJs focus on expert evidence and may fail to appreciate the wider picture. This decision reflected their strategy in mounting a judicial review against a government department, or their view of relevant case law in which experts had been 'trashed' by IJs. For example, one barrister relied on evidence submitted by IT experts in previous 'test cases' in the IAT to successfully mount a judicial review. The evidence-which was available online and was published by a Government Committee-concerned a notorious Home Office decision to set aside the examination results of fifty-thousand overseas students in the U.K. who were attending British educational institutions and who were required to pass the 'Test of English for International Communication' (TOEIC). The Home Office had wrongly decided that student test results were falsified and it revoked the right of overseas students to remain in the U.K. and required them to leave without completing their studies (House of Commons 2019).
In a different case, a barrister acting on behalf of an intervenor in a case heard by the Supreme Court, adduced unpublished social science research to provide the court with 'a macro-perspective' regarding the Government's failure to decide policy without taking into consideration the impact of its decision on sections of the British public. In this case, the barrister decided not to call an expert to submit evidence or testify. A third case involved a judicial review against a 'Conclusive Grounds' decision by the Home Office which had refused to recognize a young woman as a victim of trafficking. While the barrister could have instructed a country expert, she chose instead to instruct an ex-police officer who had extensive knowledge of trafficking in southern Europe. The official provided extensive evidence about corruption in the border force of the applicant's country of origin and the limited efforts by the Home Office to secure intelligence from Europol or Interpol to verify the evidence it relied on. The expert evidence, together with written evidence provided by organizations supporting victims of trafficking, successfully challenged the Home Office policy to refuse to reconsider negative trafficking decisions. All three cases were successful, and all relied on very different types of expert evidence or on legal argument.
It should not be too surprising, given the need to identify specific types of experts to address distinctive legal or factual issues in cases heard in different legal jurisdictions, that a wide range of experts are instructed and that, apart from immigration and asylum proceedings, cultural expertise seems to play a limited role. Nevertheless, interviews with five barristers provide insight into their use of experts, but a limited basis on which to generalize.
Cultural Evidence in British Law: A Survey of Bailii
How widely is CE used in British law? If it is used, to what extent and in what legal jurisdictions is it used in? To address gaps in the research cited above-and keeping in mind Golan's (2008) conclusion that 'scientists' have been providing expert evidence in British courts since at least the 18th century-I undertook a key word search of decisions made by all seven appellate courts and the Supreme Court in England and Wales to identify all types of expert evidence involved in legal cases. 32 The cases are held on Bailii (the British and Irish Legal Information Institute) <https://www.bailii.org/> a legal data base created in 1999. The database states that some, but not all, legal decisions are entered onto it and that 'Bailii makes available on the Internet a collection of leading cases identified by the legal academic community to support legal education' (Quick Guide to Bailii 2020). Given the long history of the British courts, it is best to assume that Bailli does not contain all the legal cases that have been decided in England and Wales. For this paper, I surveyed all the cases listed on the database in 2019-which contained nearly forty one thousand cases-by undertaking a key word search using the words 'expert', 'expert evidence', 'culture' and 'cultural' across all legal jurisdictions. I then read the cases identified by the key word search to identify the expert evidence submitted in each case.
Case decisions reported on Bailii are appeals from first instance courts; even when expert evidence was submitted, the reported cases on Bailii do not fully summarize the case or the evidence initially submitted. If the appeal concerns an issue of fact, then a decision briefly summarizes the evidence offered by the expert, the police and so on. However, many appeals are concerned with issues of law, not issues of fact, which means that expert evidence was not submitted. As will become clear, the British appellate courts hear cases involving individuals, firms and so on who are based in and outside the U.K.
The Family Division of the High Court hears cases where a child who is the subject of legal proceedings must be protected and this protection is not possible under the Children Act 1989. The most common type of case is where a child is made a 'ward of the court'. 33 It can also hear cases about forced marriage, Female Genital Mutilation 34 , applications for financial relief where a divorce has taken place outside England and Wales and cases involving parental access to children. 35 The Family Court will normally hear all other cases about family issues, but may transfer some cases to the High Court if complex issues are involved.
In 2019, this court decided sixty-one cases. A key word search for the terms expert evidence/expert identified twenty-nine cases (forty-eight percent of all cases) in which the term was used. I found that the following types of experts were cited: expert accountant/forensic accountant, expert social worker, DNA/blood test evidence and handwriting experts. 36 A key word search for the terms culture/cultural identified thirteen cases (twenty-three percent). In these cases, testimony was provided by social workers, the police, and social services. In each case the term was used in a generic sense as in the culture of one of the parties in the case, cultural heritage, cultural needs (of children in care), cultural practices or cultural forms of abuse (i.e., by a foreign parent).
The Court of Protection makes decisions on financial or welfare matters for people who 'lack mental capacity' to make these decisions. The court is based in London where most cases are heard by District Judges and a senior judge. In 2019, this court decided fifty-eight cases. A key word search for the terms expert evidence/expert identified thirty-seven cases (sixty-four percent of all cases). I read four (eleven percent) cases and identified the following types of experts 37 who submitted evidence: medical experts, social workers, nursing experts and experts on foreign law. A key word search for the term culture/cultural identified ten cases (fourteen percent) and I read two cases (twenty percent). No expert evidence was provided in the first case; in the second case psychiatrists and social workers provided 33 This court also handles cases of international child abduction but only if the abduction falls under either The Hague Convention on the Civil Aspects of International Child Abduction or Brussels II Regulation (EC) No. 2201No. /2003 Once evidence of FGM has been found by a doctor/nurse, the case is reported to the police for prosecution. Given the extent to which particular ethnic/minority groups are targeted as likely to have their children 'cut', it is surprising there is no requirement for CE to be submitted to the court. In the only case that has been successfully prosecuted in the UK, there was evidence of 'witchcraft'. See: 'UK. First successful prosecution of FGM' at: https://www.loc.gov/law/foreign-news/article/ united-kingdom-first-successful-prosecution-for-female-genital-mutilation/. See Fontein (2014). 35 For example, in 2017 I provided CE/expert evidence in a case in the Family Court regarding parental access to the children of a divorced couple. During proceedings the High Court Judge mediated between the two parties to arrive at an arrangement that was acceptable to both parents and which safeguarded the interests of the children. 36 Ministry of Justice (2013, pp. 11-12) guidelines for experts make it clear that socio-legal experts are unlikely to be involved in the family courts. The Guidelines stipulate that experts should have: a 'working knowledge of the social, developmental, cultural norms and accepted legal principles applicable to the case presented at initial enquiry, and has the cultural competence skills to deal with the circumstances of the case'; that 'professional practice is regulated by a UK statutory body'. Furthermore, ' [I]f the expert's area of professional practice is not subject to statutory registration (e.g., child psychotherapy, systemic family therapy, mediation, and experts in exclusively academic appointments) the expert would be expected to demonstrate appropriate qualifications and/or registration with a relevant professional body on a case by case basis.
Registering bodies usually provide a code of conduct and professional standards.' 37 These are civil claims. The rules of admissibility state that '[T]he question of admissibility was held to turn on four considerations: (i) whether the proposed expert evidence would assist the court in its task; (ii) whether the witness has the necessary knowledge and experience; (iii) whether the witness is impartial in his or her presentation and assessment of the evidence; and (iv) whether there is a reliable body of knowledge or experience to underpin the expert's evidence.' See Ministry of Justice (2016).
evidence. The term 'culture' was used in a generic sense as in 'cultural grounds' (i.e., the orientation of one of the parties) or the 'cultural norms' of one of the parties. The England and Wales High Court (Commercial Division) hears complex national and international business disputes. 38 Many of these cases can also be heard by the Circuit Commercial Court. However, it generally hears the more complex cases, or cases where there is a large amount at stake. This court decided three hundred and thirty cases in 2019. A key word search for the terms expert evidence/expert identified a total of one hundred and seventy-three cases (fifty-two percent). A careful reading of ten of these cases (nearly six percent) showed that the following types of experts had submitted evidence: on ship sale and purchase (ship valuations), foreign law, valuation of costs, on ship cargo, arbitration, handwriting and on commercial transactions. A key word search for the terms culture/cultural identified twenty cases. I read three cases (fifteen percent) which indicated a generic use of the term 'culture' as in 'bank culture' (in dealing with cases of fraud) and the 'cultural fit' of a manager in a bank.
The Chancery Division of the High Court deals with: (a) disputes relating to business, property or land; (b) disputes over trusts; (c) competition claims under European or U.K. competition law; (d) commercial disputes (domestic and international); (e) intellectual property issues; and (f) disputes over the validity of a will ('probate disputes'). 39 This court decided seven hundred and two cases in 2019. A key word search for the terms expert/expert evidence identified two hundred and sixty-nine cases (thirty-eight percent of the total). I read twenty cases (seven percent) which indicated that the following types of experts were involved: accountancy, liquidators, and experts in foreign law. However, there was a much larger use of the term generically as in: witness evidence, claimant's evidence, oral evidence, documentary evidence, lack of/no evidence and 'evidenced in writing'. With regard to a key word search for the terms culture/cultural, a total of thirty-eight cases (just over five percent) used the term. I read four cases (ten percent) and all four used the term in a generic sense, as in 'litigation culture', 'cultural activities', the culture of a business and 'company culture'. No expert evidence was adduced in these cases.
The Court of Appeal Civil Division hears appeals against certain decisions from all three divisions of the High Court of Justice and their specialist courts, including the Administrative Court: the county courts and the Family Court. It also hears appeals against certain decisions by other Tribunals. 40 Of particular relevance to this paper, this court hears appeals from the Upper Tribunal of the Immigration and Asylum Chamber which relies on a wide range of experts including anthropologists, historians, linguists and others in deciding asylum and immigration appeals 41 and age-dispute claims relating to assessing the age of child asylum seekers. 42 This court also decides Judicial Review applications.
This court decided four hundred and seventy-seven appeals in 2019. A key word search using the term expert evidence/expert identified a total of two hundred and nine (forty-four percent) cases. I read ten of these cases (five percent) which identified the following types of expert who submitted evidence: child psychologist, social workers, orthopedic experts, medical evidence, transport experts, coroners, and experts in foreign law. A key word search for the terms 'culture'/cultural expert identified seventy-eight cases (sixteen percent). A reading of eight cases (sixteen percent) revealed that in only one case was evidence adduced, namely by a medical expert; I found that the term was used generically, e.g., cultural ties, cultural needs, culturally appropriate placement, cultural reasons, a child's culture and 'culturally integrated'.
The Court of Appeal (Administrative Division) reviews decisions made by people or bodies with a public law function, e.g., local authorities, and regulatory bodies. It hears judicial review applications made by other courts, tribunals and public bodies and it hears challenges to decisions made by certain people or bodies (e.g., ministers or local government) where legislation provides a right to challenge. 43 It also contains a specialist Planning Court which handles judicial reviews of decisions about planning permission and challenges to planning decisions. It is a specialist court within the Queen's Bench Division of the High Court of Justice, which is based at the Royal Courts of Justice, London and in Birmingham, Cardiff, Leeds, and Manchester. Cases may be heard by one High Court judge or by a 'Divisional Court' which consists of two or more judges, normally a High Court Judge and a Lord Justice of Appeal.
In 2019, this court decided six-hundred and eighty-two appeals. A key word search for the terms expert/expert evidence identified two hundred and sixty-one cases (thirty-eight percent of all cases). A careful reading of seventeen cases (six percent) identified the following types of experts who submitted evidence: medical, foreign law, firearms, planning, prison/independent psychologist, experts on human trafficking, the Environmental Agency, parole boards, planning, Migration Advisory Council expertise, Nature England, government regulatory experts, veterinary experts, midwives and experts on public health. A key word search for the term 'culture'/cultural identified ninety cases (thirteen percent of the total). A reading of nine cases (ten percent) revealed that in three cases expert evidence was submitted by a human rights monitor, an archaeologist, a psychiatrist and a psychologist. The term was also used in a generic sense, e.g., sub-culture, cultural norms, cultural heritage, cultural identity, cultural differences, cultural consideration, compliance culture and culture of benefit dependency.
The Court of Appeal (Criminal Division) hears appeals from the Crown Courts. It hears appeals against convictions and sentences (even if the conviction was in a magistrate's court) and confiscation orders imposed by the Crown Court. It also heard applications for permission to appeal and appeals from proceedings in the Crown Court (including cases referred by the Attorney General where there is concern that the sentence given by the Crown Court may have been too lenient). 44 It is based at the Royal Courts of Justice in London. Cases are heard by Lord Justices of Appeal or, in some cases, High Court judges.
In 2019 this court decided four hundred and six appeals. A key word search for the terms expert/expert evidence identified seventy-seven cases (nineteen percent). I read seven of these cases (nine percent) and identified the following types of experts who submitted evidence: the analysis of drugs, health and safety experts, psychiatrists, firearms experts, and medical experts. A key word search of cases for the terms culture/cultural identified fourteen cases (three percent). A careful reading of two cases (fourteen percent) found that no expert evidence was submitted and that the term was used generically, e.g., culture of secrecy and gang culture.
England and Wales Queen's Bench Division of the High Court hears disputes relating to personal injury; negligence; breach of contract; breach of a statutory duty; breach of The Human Rights Act 1998; 43 This court also hears: (a) applications for 'habeas corpus', (a legal procedure where the court rules on whether the detention of an individual is legal); (b) applications to prevent individuals from continuing to initiate groundless legal proceedings ( libel, slander and other torts; and non-payment of a debt and 'enforcement orders. Many of these cases can also be heard by the Chancery Division. 45 In 2019, this court decided five hundred and seventeen appeals. A key word search for the terms expert/expert evidence identified two hundred and forty-two cases (forty-seven percent). I read fifteen cases (six percent) which identified the following types of expert 46 who submitted evidence: on road accidents, handwriting, medical, neurologist, on the price of drugs, on general medical practice, diabetology, foreign law and clinical psychology. A key word search for the terms culture/cultural identified forty-five cases (nine percent). A careful reading of four cases (nine percent) failed to identify a case in which expert evidence was involved and found only a generic use of the term, as in 'cultural reasons', culture of the defendant, culture (of a foreign country) and institutional culture.
Finally, I reviewed decisions by the United Kingdom's Supreme Court which is the final court of appeal in the U.K. for civil and criminal cases from England, Wales, and Northern Ireland. It hears cases of the greatest public or constitutional importance. In 2019 the Supreme Court decided fifty-nine appeals. A key word search for the terms expert/expert evidence identified twenty-two cases (thirty-seven percent). I read four cases (eighteen percent) which identified three types of experts who submitted evidence: medical, handwriting, and judicial experts. A key word search for the terms culture/cultural identified a total of eight cases (thirteen percent); my reading of three cases (thirty-seven percent) found that references were made concerning the expertise of the Department for Digital, Culture, Media and Sport (a government department), and to cultural life and cultural integration (in relation to a deportation appeal).
A lengthy history of immigration has led the U.K. government and British courts to recognize and enshrine certain aspects of the culture of ethnic minorities into British law, e.g., with respect to the right of Sikhs to wear a turban, a 'kara' or a 'kurban' 47 , the right of Muslim women to wear a burqa 48 , and limited rights accorded to Roma peoples. 49 Legislation has also been adopted which bars ethnic and religious discrimination. 50 Civil law cases in the U.K., which in the past may have called upon anthropologists/socio-legal experts to provide evidence on matters of culture, family, gender, religion and marriage, are now dealt with by the police who enforce civil rights legislation by imposing civil penalties rather than prosecuting individuals in court.
In keeping with studies of the role of CE in other countries discussed above, in England and Wales, the majority of the appellate court decisions recorded on Bailii do not involve CE 51 or expert evidence of any kind (because they are concerned with deciding/interpreting points of law not factual issues). Even so, some form of expert evidence is submitted in all the appellate courts in the U.K. because they 45 This Court also handles: applications to 'enroll' (register) deeds, including changing your name by deed poll; registration of judgments obtained abroad so that they can be enforced in England and Wales; election petitions to challenge the results of Parliamentary, European Parliamentary and local government elections; applications for bail; serving documents overseas and obtaining evidence for foreign courts; registration and satisfaction of 'bills of sale'; and 'interpleader proceedings' where a High Court Enforcement Officer is attempting to recover goods to settle a debt and a third party claims to be the owner of the goods. 46 Experts are regulated by Part 35 of the Civil Procedure Rules. However, the Queen's Bench Guide (Ministry of Justice 2018, sec. 10.8) requires parties to obtain permission from the court before they secure an expert, and both parties are given a very strong steer that they should agree a single, joint expert. The Guide also makes it clear that, 'The most common form of written evidence is a witness statement' (sec. 10.9.2). 47 See: Grillo (2017) on litigation which expanded the right not to be discriminated against as set out in the Equalities Act 2010 and the Human Rights Act 1998. 48 See endnote 36, and the discussion on 'face coverings' in 'What are the rules on burqas and face coverings in the UK' at: https://fullfact.org/law/what-are-rules-burkas-and-niqabs-uk/ (accessed on 20 August 2020). 49 The rights of Roma are not protected by legislation in the UK, though litigation has secured a measure of protection against discrimination (see: Willers and Baldwin n.d.). 50 These rights are set out in the Equalities Act 2010. 51 The limited extent to which expert evidence is cited in appellate decisions does not indicate whether doctors, psychologists, experts on human trafficking provide 'socio-cultural'/CE evidence. Given the strict regulations on the submission expert evidence in British courts which require an expert to have an acknowledged 'expertise' in their area of work, it seems unlikely that other professionals are allowed to provide CE that is in any way analogous or similar to anthropologists, historians and other social scientists.
hear appeals from first tier courts. The survey of Bailii indicates that the term 'expert' is understood in a very wide sense to include nearly every type of expertise that arises in a modern, technologically complex society. 52 It is notable that none of the cases identified by a key word search involved anthropological/CE for several reasons. First, anthropological/CE evidence, when it is submitted, is heard by the lower courts which are deemed to possess greater expertise in dealing with evidence. Secondly, most first instance cases are not appealed to the appellate courts. Third, it is probably the case that a key word search of a database like Bailii does not identify many relevant cases because the search engine is unable to identify the issues relating to CE/socio-legal expertise. This, in turn, probably means that it is not possible to determine the extent to which the courts deal with CE and whether, over time, courts have accepted more or less CE/expert evidence. Finally, and equally importantly, a key word search of Bailii together with an analysis of the cases requires substantially more resources and time than individual researchers have at their disposal to undertake a comprehensive analysis of the database. Ideally, what is required are more in-depth studies of countries and legal jurisdictions by teams of researchers (supported by external funding). Such studies would enable us to move beyond the analysis of individual cases, as useful as these are for illuminating judicial decision-making, and surveys of legal databases to provide a more complete understanding the role of all legal actors-the police, state prosecutors, lawyers, experts and judges. It is important to know how decisions are made to prosecute or drop cases, e.g., by filtering out and preventing claims from reaching court and, once a claim reaches court, to understand the work of all legal actors involved in providing evidence and arguing and deciding claims in which CE is submitted.
Conclusions
An assessment of the contribution of cultural expertise in assisting judges or mediators to better understand the cultural issues which arise in disputes has proven to be a complicated task. A review of the burgeoning literature reveals several issues. First, anthropologists and a wide range of socio-legal experts have increasingly submitted CE/expert evidence in asylum, civil, criminal, family law, indigenous land claims and in claims heard by international criminal courts. It is clear that these experts have quite different reasons for engaging with the courts, ranging from the desire to protect subaltern peoples to profiting from their work. Many, perhaps most, have a limited understanding of the adversarial process.
Second, regardless of jurisdiction and country, judges act as gatekeepers (a fact that has not been recognized in many studies). They exercise this role by imposing evidentiary hurdles and admissibility rules, in their assessment of the evidence, and by deciding legal claims. The judiciaries power is only slightly mitigated by the discretion given the lower courts-particularly asylum and immigration hearings and in civil claims-to admit hearsay/opinion evidence which, occasionally, can have a decisive impact on a judicial decision.
Third, courts in different countries and in different jurisdictions appear to exercise considerable discretion with regard to admitting socio-legal/CE evidence. This probably arises from the extent to which they are insulated from, or are required to engage with, international humanitarian and criminal law. The failure of courts in Denmark, Poland and Sweden seem to indicate that their failure to engage with international law is due in part to their concern to maintain a 'mono-cultural' society. The insular focus of these courts may also explain why so little CE is submitted and thus why judges and lawyers are relatively unfamiliar with CE/socio-legal expertise.
An appreciation of the extent to which CE is submitted in different jurisdictions and countries is constrained by the limitations of existing research which has primarily been based on the analysis of a limited number of judicial decisions. Recently, researchers have attempted to provide a much wider 52 The role of certain of social workers, psychiatrists and lawyer/mediators in arbitration has been discussed by Brophy et al. (2013), Hallett (2018) and De Girolamo (2013), respectively. picture of CE by searching legal databases. These studies are problematic for two reasons. The first type of studies are illuminating but they tell us relatively little about the work of all the legal actors-the police, public prosecutors, lawyers, experts and judges-whose work shapes the adversarial process and determines the role and value of CE. Surveys, on the other hand, hold out the promise of situating CE/socio-legal expertise in the wider adversarial system, but findings provide at best a limited picture due to methodological and technical problems involved in searching databases. Finally, a greater appreciation of how the law works shows that the key differences between anthropology and law are not epistemological but reflect the need for the court to reach a decision. While experts are clearly regulated by the judiciary, it should be clear that the situation they face varies considerably by country and by jurisdiction and that the principal use of CE is in asylum and immigration law, civil claims, international criminal law and secondarily in international human rights law and family law.
Funding: This research received no external funding. | 2020-12-03T09:02:46.951Z | 2020-11-30T00:00:00.000 | {
"year": 2020,
"sha1": "2bf7fd89e5b75458e13f000c6ca580af5921ecd1",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2075-471X/9/4/29/pdf",
"oa_status": "GOLD",
"pdf_src": "Adhoc",
"pdf_hash": "89b629d7686c94d72ecb9b7d14bc074372b1aaa5",
"s2fieldsofstudy": [
"Law",
"History"
],
"extfieldsofstudy": [
"Sociology"
]
} |
231733989 | pes2o/s2orc | v3-fos-license | Beliefs and Risk Perceptions About COVID-19: Evidence From Two Successive French Representative Surveys During Lockdown
Background The outbreak of COVID-19 has been a major interrupting event, challenging how societies and individuals deal with risk. An essential determinant of the virus’ spread is a series of individual decisions, such as wearing face masks in public space. Those decisions depend on trade-offs between costs (or benefits) and risks, and beliefs are key to explain these. Methods We elicit beliefs about the COVID-19 pandemic during lockdown in France by means of surveys asking French citizens about their belief of the infection fatality ratio (IFR) for COVID-19, own risk to catch the disease, risk as perceived by others, and expected prevalence rate. Those self-assessments were measured twice during lockdown: about 2 weeks after lockdown started and about 2 weeks before lockdown ended. We also measured the quality of these beliefs with respect to available evidence at the time of the surveys, allowing us to assess the calibration of beliefs based on risk-related socio-demographics. Finally, comparing own risk to expected prevalence rates in the two successive surveys provides a dynamic view of comparative optimism with respect to the disease. Results The risk perceptions are rather high in absolute terms and they increased between the two surveys. We found no evidence for an impact of personal experience with COVID-19 on beliefs and lower risk perceptions of the IFR when someone in the respondent’s family has been diagnosed with a disease. Answers to survey 1 confirmed this pattern with a clear indication that respondents were optimistic about their chances to catch COVID-19. However, in survey 2, respondents revealed comparative pessimism. Conclusion The results show that respondents overestimated the probabilities to catch or die from COVID-19, which is not unusual and does not necessarily reflect a strong deviation from rational behavior. While a rational model explains why the own risk to catch COVID-19 rose between the two surveys, it does not explain why the subjective assessment of the IFR remained stable. The comparative pessimism in survey 2 was likely due to a concomitant increase in the respondents’ perceived chances to catch the disease and a decreased expected prevalence rate.
INTRODUCTION
The outbreak of COVID-19 has been a major interrupting event for economies all over the globe. This event challenged the way societies and individuals deal with risks. Over and above public policies, an essential determinant of the spread of the virus is a series of small-scale individual decisions, such as wearing face masks in public space, regularly washing hands or deciding how often to go the office, class, store or anywhere else. According to decision theory, those decisions depend on tradeoffs between costs (or benefits) and risks, and those tradeoffs are deeply rooted in individual preferences. Considering the wide literature that has been devoted to the understanding of individual preferences and attitudes toward risk, their heterogeneity and their determinants , one of the key figures of the classical representation of behavior under uncertainty is that beliefs -in addition to risk attitudes -are key to explain decisions whenever they involve money, life duration, health states, approval of friends or well-being of others (Savage, 1954).
Uncertainty especially affects decisions regarding health where most probabilities are ambiguous and not objectively known (Attema et al., 2018). Smoking is a typical example of decisions for which subjective assessments of mortality risks have received a lot of attention (Viscusi, 1990;Khwaja et al., 2007). Beliefs and, more generally, risk perceptions are rather challenging to evaluate, especially for a new disease such as COVID-19. First, risk perceptions are threat-specific most of the time and incorporate different kinds of information through deliberative, affective and experiential processes (Slovic et al., 2004;Ferrer and Klein, 2015). In case of a new disease, the amount of available information, whenever it is publicized numerically or derived from personal experience, is limited. Second, the range of available methods to elicit beliefs is restricted. Experimental studies, which measure beliefs with monetary stakes, offer an array of incentive-compatible elicitation methods (Trautmann et al., 2015), but these methods are known to be difficult to implement in large representative samples . Additionally, incentive compatibility has no bite for events with serious health consequences such as COVID-19. In such cases, survey studies to assess beliefs and risks generally involve introspective judgments to assess beliefs and risks (Viscusi, 1990;Coe et al., 2012;Carman and Kooreman, 2014). Third, there is no pre-existing measuring rod to judge the accuracy of risk perceptions. In particular, without objective and subjective benchmarks, it is difficult to know if a low risk perception to catch the disease actually reflects optimism or pessimism, whenever it is considered as an absolute or a relative measure (Shepperd et al., 2013). Thus, understanding, predicting and aiding individual decisions in face of COVID-19 mainly relies on answers on a series of questions (Fischhoff, 2020), such as "how much of the disease is in the community?" (i.e., beliefs about the prevalence rate), "what is my risk of exposure to the COVID-19?" or "what is the risk to die if infected?" [i.e., beliefs about the infection fatality ratio (IFR)]. Because individual decisions are likely to be impacted by others' decisions, and therefore by their beliefs, second-order beliefs ("how do the others perceive the risk?") might also be of importance.
Several recent studies have reported information on the disease risk perception of COVID-19, and its perceived impact on health during the lockdown phase due to especially in Italy where the virus reached Europe the earliest, but also in other countries [e.g., (Barattucci et al., 2020;Cerami et al., 2020;Faasse and Newby, 2020;Lanciano et al., 2020;Liu M. et al., 2020;Moyce et al., 2020)]. These studies have shown that people perceive the impact of COVID-19 on their (mental) health as high (Tull et al., 2020;Xiong et al., 2020), and that their risk perception of this disease is correlated with adoption of preventative health behaviors (Dryhurst et al., 2020), but its level is not so high (Commodari and La Rosa, 2020;Lanciano et al., 2020;Liu M. et al., 2020), and lower than their concern for the future and for economic and social consequences of the pandemic (Lanciano et al., 2020). In this paper, we add to this literature by investigating beliefs about the COVID-19 pandemic during lockdown in France, analyzing responses to survey items that ask for the individuals' judgments of the risks associated with COVID-19. Self-assessments about risks included the IFR for COVID-19, the surveyed individuals' own personal risk to catch (or catch again) the disease, the risk as perceived by others and, finally, the expected prevalence rate after the COVID-19 pandemic in France. Those four self-assessments were measured twice during lockdown. In addition to shedding light on the determinants of beliefs in a representative population sample and their dynamics during lockdown, the paper also provides measures of the quality of these beliefs with respect to available evidence at the time of the surveys. Finally, comparing own personal risk to catch COVID-19 to expected prevalence in the two successive surveys provided a dynamic view of comparative optimism with respect to the disease. The organization of the paper is as follows. The next Section introduces the theoretical background and our resulting hypotheses. The Section 'Materials and Methods' describes the data from the two successive surveys and how we computed measures of objective risks. 'Results' presents the statistical methods used to analyze the survey responses and investigates dynamics, heterogeneity and determinants of the self-assessment of beliefs, assesses the quality of beliefs through calibration. and reports the dynamics of comparative optimism during lockdown. Finally, the last Section discusses the results and concludes.
THEORETICAL BACKGROUND AND HYPOTHESES
The Bayesian learning model (Viscusi, 1991) assumes individuals have three risk information sources, with each source characterized by its informational content. The first source is the prior risk assessment, a fundamental element of any Bayesian model. In the absence of information, a natural prior is the uninformative prior. The second source is the experience of the individual. Experience regroups direct individual experience with the risk, such as catching COVID-19 and indirect experience, e.g., knowing a close family/friend ill from COVID-19. Lockdown was used as an essential element of personal experience. During lockdown, limited physical and social interactions reduced the number of observed cases of infection from COVID-19, as well as observed fatalities. Experience predicts a decrease in the perceived risks to catch COVID-19, expected prevalence or lethality of the disease during lockdown. Because limited social interactions reduced the exchanges of private information upon beliefs, restricted personal experience with others' beliefs predicts stability of these beliefs during lockdown. The third source of information is the risk information provided to the individual. Risk information includes public information about the risk and observable events processed as information by the individual. Typical sources of risk information are media coverage or public messages from the authorities. In the case of COVID-19, existing medical evidence covered by media identified age, pre-existing chronic illness, and -to a lesser extent-gender as risk factors for the severity and lethality of the disease. Location was an important risk information for the localization of the disease. Publicized information on the epidemic dynamics, with a peak occurring in the middle of lockdown, was another source of indirect risk information for individuals. A formal elaboration of this model can be found in the Supplementary Material.
Apart from the Bayesian learning model, several insights from psychology might help formulating our hypotheses. First, comparative optimism refers to the phenomenon that people believe the probability that a future negative event happens to themselves is lower than the probability that it happens to a similar other person, and vice versa for positive events (Shepperd et al., 2013). Several empirical studies found evidence supporting comparative optimism in a health context (Weinstein, 1980;Hoorens, 1994;Shepperd et al., 2013;Kuper-Smith et al., 2020). Second, the representativeness heuristic entails that someone evaluates a subjective probability by the degree of correspondence between the sample and the population (Kahneman and Tversky, 1972;Tversky and Kahneman, 1974). As such, it emphasizes the generic features of an event. This heuristic has also been observed frequently in health settings (Attanasio et al., 1998;Brannon and Carson, 2003), for example in the long-lasting belief of medical experts that ulcers were caused by stress, while in fact they are caused by bacteria (Gilovich and Savitsky, 2002). Finally, the familiarity bias stems from the availability heuristic and holds that events are judged as more frequent or more important if they are more familiar in memory (Tversky and Kahneman, 1974).
Evidence of this bias in health was reported by Sherman et al. (1985) and Pachur et al. (2012).
Consistent with the Bayesian learning model and the summarized literature, this study aims to test the following hypotheses. First, we test if people have well-calibrated (i.e., accurate) beliefs about the prevalence rate, their probability of getting infected with COVID-19, and the IFR, and if these are mediated by socio-demographic characteristics such as gender and age (Weinstein and Klein, 1995;de Zwart et al., 2009;Vaughan, 2011;Clifton et al., 2016;Moran and Del Valle, 2016;Clark et al., 2020;Raude et al., 2020) (H1). It could also be that the lockdown has decreased risk perceptions, because people have the feeling that lockdown helps control, which in turn lowers risk perceptions (Nordgren et al., 2007). Second, we test comparative optimism by comparing respondents' beliefs about these probabilities for themselves and for others (H2). The Bayesian model could justify this if beliefs reflect perfect ignorance at the beginning of the pandemic, since as people gather more information on their close environment, they should update their own beliefs (downward) more than their beliefs for the other, for which they gather less information. In addition, the familiarity bias suggests people might be excessively optimistic about their own beliefs (Kilka and Weber, 2000;Heideker and Steul-Fischer, 2015). Third, we test if these beliefs change during the pandemic by comparing the results from survey 1 and survey 2 (H3). The Bayesian model and the representativeness heuristic might differ here. While the Bayesian model predicts that individuals rationally update their beliefs based on new information obtained during the lockdown, the representativeness heuristic predicts individuals to underweight new information and stick to their prior risk assessment despite new evidence.
Participants
Between March 31 and April 27, 2020, we conducted surveys in two representative samples of the French population 18 years of age and over (n = 1,005 and n = 1,004). The study design was approved by the ethical committee of the University Hospital Institute Méditerranée Infection (#2020-018). Participants to the surveys were selected and interviewed by IFOP (Paris, France), a survey research company. Participants were drawn from an online research panel of more than 750 000 nationally representative households of the French general population, that is developed and maintained by IFOP. Data were collected using an online survey between March 31 and April 2 for the first survey and between April 23 and April 27 for the second survey. The first survey was conducted two weeks after the nationwide lockdown was introduced (lockdown was active between March 17 and May 11, 2020) and the second survey was conducted two weeks before the lockdown ended. Internal procedures at IFOP explain the small difference in the number of participants between the two surveys (n = 1,005 vs. n = 1,004).
Prior information on the panelists was used to determine eligibility and to draw a random sample, stratified to match French official census statistics for sex, age, occupational status, education level, size of town, and region. Table 1 provides details about the socio-demographic characteristics of the participants and health status.
Measures
The questionnaire collected data on socio-demographics (Table 1), self-assessments of risks about COVID-19, confidence in beliefs, opinions toward COVID-19 and seasonal influenza, personal information about COVID-19 and health characteristics. Other questions relating to habits during lockdown and attitudes toward vaccination were included in both surveys, while questions related to sleep problems (survey 1), cultural profiles (survey 2) and specific medical treatments (survey 2) were included in one of the two surveys. Those questions are not included in the current analysis, but the interested reader on these topics is referred to The COCONEL Group (2020) for a description of items and responses.
Quantitative Measures: Assessment of Risks About COVID-19 and Seasonal Influenza
We measured self-assessments of risks about COVID-19 with four items. The first item concerned the perception of the IFR for COVID-19 (Q1: "Out of 100 people who are infected with the Coronavirus , how many of them die from the disease?"). Literature on smoking risk perception suggests that a base population reference point is a more readily understood method for eliciting probabilistic information about death rates than explicitly dealing with probabilities or fractions (Viscusi and Evans, 1990). Survey 1 included a replication of question Q1 for seasonal influenza (Q1bis: "Out of 100 people who are infected with seasonal influenza, how many die from the disease?"). In order to avoid order effects between items Q1 and Q1bis, their order was randomized for each respondent. The second item was an assessment of own personal risk to catch -or catch again in case the respondent has already been ill from the Coronavirus [Q2: "What risk do you have [to catch/to catch again] the Coronavirus by the end of the epidemic (on a 0-100 scale)?"]. Third, we assessed how other people perceive their risk to catch the disease [Q3: "How do the French generally assess their risk of catching the Coronavirus by the end of the epidemic (on a 0-100 scale)?"]. For items Q2 and Q3, we chose a numerical format rather than a purely qualitative answer scale because the former allows more variability than the latter. It is also generally associated with better prediction of behaviors (Juster, 1966;De Bruin et al., 2011).
Confidence in Beliefs
Survey 1 included two questions about confidence in beliefs for items Q2 and Q3 ("How confident do you feel in your answer?": a. very high, b. high, c. moderate, d. low, e. very low").
Qualitative Measures: Opinions Toward COVID-19
We evaluated opinion toward COVID-19 with numerical 11point scales for which participants were asked to give a score between 0 and 10. Those included relative risk (Q5: "Compared to other French people how would you rate your own risk of catching the Coronavirus ? Give a score between 0 and 10: the score 0 indicates that you think you are much less at risk than other French people and the score 10 that you think you are much more at risk than other French people. The intermediate notes allow you to qualify your judgment."), concern about the disease (Q6: "Are you worried about [catching/catching again] the Coronavirus by the end of the epidemic? Give a score between 0 and 10: a score of 0 means that you are not at all worried about the possibility of [catching/catching again] the Coronavirus (COVID-19) at all, and a score of 10 means that you are very concerned. The intermediate notes allow you to qualify your judgment."), contagiousness (Q7: "how contagious is the Coronavirus ? Give a score between 0 and 10, with a score of 0 indicating that this disease is not very contagious and a score of 10 that it is very, very contagious. The intermediate notes allow you to qualify your judgment.") and seriousness (Q8: "how serious is the COVID-19? Give a score between 0 and 10: a 0 indicates that catching this disease is not at all serious and a 10 that it is very, very serious. The intermediate notes allow you to qualify your judgment.") of the disease. Survey 1 also included replications of items Q7 (contagiousness: Q7bis) and Q8 (seriousness: Q8bis) for seasonal influenza.
Health Related Items Regarding the Pandemic and Health Status
Specific items regarding the pandemic included whether participants had been diagnosed with -or ill from-COVID-19 and whether some of their relatives (family members or friends) had been infected. Participants were also asked to provide an expectation of the duration of the pandemic (in weeks on a 0-52 scale: "When do you think this epidemic will be truly over?"). In addition to socio-demographic characteristics, the first survey also included an item on self-reported general health status ("In general, how would you rate your state of health? Very good, fairly good, bad, quite bad") and an item related to chronic illness ("Do you suffer from a chronic, that is to say long-lasting, disease or health problem that requires medical attention (for example: diabetes, heart or respiratory disease)? Disregard temporary or temporary health problems, such as the flu.").
External Data Sources
We also collected the available data on the number of recorded diagnoses with COVID-19 and deaths from COVID-19 in hospital in France at the time of survey administration. Those data, the only ones recorded for COVID-19 at the time of the two surveys, are publicly available from Santé Publique France, the national public health agency (https://www.santepubliquefrance. fr/). Figure 1 shows a timeline of the lockdown in France alongside the two surveys and the number of daily deaths recorded in hospital (top panel) and the number of daily cases diagnosed in hospital (bottom panel). Figure 1 also shows that the epidemic peak occurred between the two surveys, both for daily deaths (top histogram) and for diagnosed cases (bottom histogram).
We considered two datasets. The first dataset includes the number of deaths and diagnosed cases by gender and administrative French metropolitan regions. The second dataset also includes the number of deaths and diagnosed cases but aggregated by age category and "department" (a sub-division of administrative regions). To obtain a consistent measure of location areas between the two datasets, we merged departments into administrative regions in the second dataset to obtain a dataset by age category and region. We fixed the date at the center of the time interval within each survey. We computed the IFR as the ratio of the deceased persons over the number of diagnosed cases.
Because these data are hospital data only and in a context where confirmations of COVID-19 infections mostly occurred when being hospitalized, the diagnosed cases corresponded to the most severe cases of the disease, leading to a computed IFR that can be thought as upper bounds of the effective IFR. To translate the number of diagnosed cases into population proportions to measure prevalence, we collected population data on January 1, 2020 on gender, age and location area from the INSEE (the French national institute of statistics). Still, with data being restricted to hospital reports only and recorded at the beginning of the epidemic, the recorded cases corresponded to the most severe cases of the disease. A consequence is that the computed prevalence rate can be thought as a lower-bound of the population risk to catch COVID-19. Table 2 shows the computed IFR and prevalence rate by gender, age category and region. For the sake of clarity, and following Salje et al. (2020), we categorized regions into three geographical main areas in France by importance of incidence: highest incidence regions (Paris region and north-eastern part of the country; namely, Île-de-France and Grand-Est), medium incidence regions (located in the center-east part of the country: Hauts de France, Bourgogne-Franche-Comté, Auverge-Rhône-Alpes, Corsica, Provence-Alpes-Côte d'Azur and Center-Val-de-Loire) and lowest incidence regions (all located in the west part of the country: Normandie, Bretagne, Pays de la Loire, Nouvelle-Aquitaine and Occitanie). To provide a sensible view of the differences between categories, Table 2 reports the values for a reference case (a prototypical individual with the lowest risk: a woman, aged 20-29 in Nouvelle-Aquitaine region) and oddratios for alternative characteristics. Table 2 shows that COVID-19 was more prevalent among men, older people and people living either in the medium and highest incidence regions located in the north-east part of France (Hauts de France, Bourgogne-Franche-Comté and Grand-Est), Corsica (Corse) or the Paris area (Île de France). For example, Table 2 shows that the computed prevalence rates among the elderly was around 13 times higher than the prevalence rate among people aged 20-29 and their IFR was 40 to 50 times higher.
Statistical Analysis
We measured the determinants of beliefs for each item Q1 to Q4 using a Generalized Linear Model with a logit link and a quasibinomial distribution. We explored the dynamics of beliefs by comparing answers to items Q1 to Q4 between the two surveys with a base GLM model with no covariates. A difficulty with selfassessment of beliefs on 0-100 scales is the usually high frequency of the answer "50" in the responses. This "50 blip" might be an important source of bias in belief measurement (de Bruin et al., 2002). To account for such a possible bias, we fitted a beta function to each distribution of items and subtracted the expected proportion of responses in the 45-55 category from the proportion actually observed to infer the number of excess 50s over those expected in the best fit distribution. In addition, we measured the significance of the difference between distributions with and without the "50" answers by regressing the demeaned answers to item Q1 to Q4 on an indicatrix of the "50" answer. We set statistical significance at p < 0.05.
We regressed the answers to items Q1-Q4 measured as proportions on the answers to socio-demographic items, healthrelated items, qualitative items related to COVID-19 and, when relevant, beliefs about the seasonal influenza. For each item, we reported the results of the regressions for both the first survey only and the two surveys pooled together. In order to ease interpretation of the regression coefficients, we report average marginal effects, with standard errors computed using the Delta method. All statistical analyses were performed using R We compared responses to items Q1-Q4 with available information at the date of the surveys to measure calibration of beliefs. We evaluated calibration by comparing the selfassessment of the IFR with the expected prevalence based on a paired t-test. Existing scientific evidence suggested that prevalence (the population risk to catch the disease) was expected to be higher than IFR (the population risk to die from the disease if infected), which justified the use of a one-sided test. We also measured calibration of beliefs based on the comparisons between the IFR from COVID-19 and from the seasonal influenza. This measure was possible in the first survey only, and statistical significance was assessed based on a (twosided) paired t-test. In addition, a comparison of answers to questions on contagiousness and seriousness for illness-seasonal influenza vs. COVID-19 provided a qualitative evaluation of the consistency of answers. For both, we used a linear regression without intercept. For contagiousness we regressed the answer to question Q7 on the answer to question Q7bis to obtain a measure of the assessed difference in contagiousness between the two diseases. We then compared this measure with existing evidence at the time of the survey. Our last calibration exercise used available hospital data described in Section 2.3 to measure the difference between beliefs about own personal risk to catch COVID-19 (item Q2) and public information on prevalence rates by gender, age category and location area. Lastly, we measured comparative optimism during lockdown by comparing answers to item Q2, the own personal risk perception to catch COVID-19, and answers to item Q4, the perceived prevalence for COVID-19. Significance levels were assessed based on paired Student t-tests.
Descriptive Statistics
Descriptive statistics are illustrated in Tables 1 and 3 and in Figure 2. Figure 2 shows the distribution of answers to items Q1 to Q4 for the two surveys. For each item, the left part of the violin plot shows the distribution of answers for the first survey and the right part of the violin shows the distribution of answers for the second survey.
Calibration of Beliefs
Calibration of reported beliefs was assessed based on both the consistency of answers and their comparison with available information at the time of the surveys.
Infection Fatality Ratio vs. Prevalence Rate
We first evaluated calibration by comparing the self-assessment of the IFR with the expected prevalence. Available scientific evidence suggests that COVID-19 is highly contagious and only severe forms of the disease led to fatal issues. Notably, the prototypical case of the Diamond Princess cruise ship suggested an IFR at least ten times lower than the prevalence rate. On 3rd February 2020, an outbreak of COVID-19 was reported on the Diamond Princess cruise ship, with initially 10 persons confirmed to be infected with the virus. The outbreak of COVID-19 led 3711 crew and passengers to be quarantined for three weeks. By the end of February, 7 persons had died among the 705 persons diagnosed and tested positive, giving an IFR of 0.99% and an observed prevalence rate of 19%. The IFR from the Diamond Princess cruise ship was much lower than the values reported in March 2020, which were closer to 3-4% .
The average IFR provided by the respondents was equal to 16.46 in survey 1 and to 16.1 in survey 2, around two to three times lower than the reported expected prevalence rate (45.05 in survey 1, 33.93 in survey 2). In both surveys the differences between the answers to the corresponding items Q1 and Q4 were highly significant (one-sided paired t-tests, both P-values < 0.01).
COVID-19 vs. Seasonal Influenza
Survey 1 contained a replication of the assessment of the IFR for the seasonal influenza. The number of deaths from COVID-19 during the first wave of the epidemic corresponded to at least twice the usual mortality from seasonal influenza (Équipes de surveillance de la grippe, 2019). Usual estimates of the IFR for seasonal influenza in France are less than 0.5%. As a source of comparison, Rajgor et al. (2020) reported a common IFR of 0.1% for influenza and -based on the above mentioned case of the Diamond Princess cruise ship -a rate of 1% for COVID-19. This is much less than the rough upperbound measure for the IFR in France, according to the hospital data for severe cases of COVID-19: Table 1 shows that the available public information about COVID-19 corresponded to an IFR around 10% at the time of survey 1. Figure 3 shows the comparisons between assessments for COVID-19 and for seasonal influenza in survey 1. Figure 3-A shows that for most respondents (n = 455), the IFR of COVID-19 was higher than that of seasonal influenza. Otherwise, similar IFRs were reported by 103 respondents, whereas 97 reported lower rates. The difference between the two IFRs was highly significant (paired Student t-test, P value < 0.01).
Answers to qualitative questions on seriousness and contagiousness of seasonal influenza vs. COVID-19 confirmed the good calibration of beliefs about COVID-19. Figure 3 shows the distribution of answers to the qualitative question on contagiousness (panel B) and seriousness (panel C) for seasonal influenza (on the x-axis) and COVID-19 (on the y-axis). A linear regression without intercept for reported contagiousness showed that participants estimated contagiousness of COVID-19 to be 1.17 (standard error 0.01), i.e., higher for COVID-19 than for seasonal influenza. This is in line with available evidence on contagiousness at the time of survey 1. For example, Biggerstaff et al. (2014) report a median reproduction number (R0) for seasonal influenza to be around 1.3, while a review by Liu Y. et al. (2020) reports a median R0 for COVID-19 of 2.8, and evidence on the Diamond Princess cruise ship ( Rocklöv et al., 2020) shows that isolation and quarantine lowered the R0 to 1.78 (1.37 higher than that of seasonal influenza). Additionally, a minority of n = 22 individuals provided qualitative answers to questions Q7, Q7bis, Q8, and Q8bis revealing they thought COVID-19 to be less serious and less contagious than seasonal influenza.
Calibration of Own Personal Risk to Catch COVID-19
and Socio-Demographic Characteristics Figure 4 shows the comparison of the average answers to question Q2 (own personal risk to catch COVID-19) with the available information on prevalence at the time of the surveys by gender, age category, location and survey date. For all sociodemographic groups, the own personal risk to catch COVID-19 increased between the two surveys. Own personal risk was well-calibrated with respect to location: it was lower in low-incidence regions and higher in high-incidence regions. Things were different for gender and age: women expressed higher risks than men, although the available scientific evidence systematically showed a different pattern. The same applied to age: younger people view themselves as being at higher risk than older people and this despite the rather large difference in prevalence between age classes. Considered jointly, these findings are summarized in the following observation: H1: Risk perceptions in absolute terms were rather high for the IFR, the expected prevalence and own risk to catch the disease. Comparison with socio-demographic risk factors shows poorlycalibrated beliefs with respect to age and gender. In relative terms, risk perceptions were well-calibrated: IFR was two to three times lower than expected prevalence: COVID-19 was perceived as more serious and more contagious than seasonal influenza.
Comparative Optimism During Lockdown
The comparison of answers to item Q2 on the assessment of own personal risk to catch COVID-19 and answers to item Q4 on the expected prevalence of the disease offers a direct measure of comparative optimism (Weinstein, 1982). Figure 5 shows the distribution of answers as a function of the survey number.
During the first survey, usual findings from the literature were confirmed, notably that individuals were optimistic about their own chances to catch COVID-19: their probability to encounter the negative event was judged to be lower than others' probability, as measured by the expected prevalence (average difference of -10.56 points, two-sided paired t-test, P-value < 0.01). The second survey shows a rather different picture, with relative pessimism as a dominant trait. At that time, respondents judged their own probability to catch the COVID-19 to be higher than the expected prevalence (average difference of 12.31 points, twosided paired t-test, P-value < 0.01). Figure 5 shows that the reversal in comparative optimism between the two surveys was the consequence of a simultaneous increase in own personal risk to catch COVID-19 and a decrease in expected prevalence. According to the GLM on the optimism index (defined as the difference between expected prevalence and the own risk to catch COVID-19, re-scaled between 0 and 1) a large significant difference between the two surveys was found (P-value < 0.01).
We summarize these findings in the following observation: H2: Comparative optimism was observed in the first survey, but it turned into comparative pessimism in the second survey. Hence, comparative optimism decreased during lockdown.
Changes in Beliefs During Lockdown
Heterogeneity was much lower for IFRs than for the other items for which Figure 2 shows a large dispersion of the answers on the measurement scales. This difference in heterogeneity was expected due to the lower value associated to the IFRs. According to the GLM (see Table 5 in the Supplementary Material for details), no differences between the two surveys were found for the IFRs (Q1, P value = 0.915) and the assessment of how other people perceive their risk to catch the disease (Q3, P value = 0.617). For their own risk to catch COVID-19 (Q2) and the expected prevalence (Q4), we found large significant differences between surveys (both P values < 0.01) with an 11.5 percentage points increase in the own personal risk to be infected and an 11.1 percentage points decrease in the assessed expected prevalence.
Respondents became more pessimistic about their own risk to catch the disease and more optimistic regarding the expected prevalence. Visual inspection of Figure 2 shows a modal answer at 50 for answers to questions Q2 and Q3. Such a pattern is consistent with Bruine de Bruin et al. (2002) and might have biased the answers. Following (de Bruin et al., 2002) we fitted a beta function to each distribution to assess the importance of this "50-blip". Table 4 shows the impact the "50s" blip on the answers to questions Q1-Q4, without correction for poststratification. For items Q1 and Q4, it is clear from Table 4 that the "50-blip" did not operate. For item Q3, even if the percentage of excess 50s is relatively high (from 14 to 16% of the sample), the proximity of the average answer to 50 makes the possible overestimation of the risk immaterial. The same occurs for the answers to item Q2 in survey 2. The highest impact of the 50blip was found for question Q2 in survey 1, which results in a slight overestimation of own personal risk to catch COVID-19. Removing the "50" answers in item Q2 did however not change the conclusion about significance of the difference between the two surveys (P value < 0.01). Table 5 in the Supplementary Material shows the results of the Generalized Linear Model regressions for items Q1 to Q4 for the first survey only and the two surveys pooled together after the inclusion of control variables. A significant impact of time on reported answers was confirmed for items Q2 and Q4. As described in Section 4.2.3, being male had a strong impact on reported beliefs: men were more optimistic than women on IFR and expected prevalence. The Generalized Linear Model regressions show that the impact of gender on own risk to catch COVID-19 becomes no longer significant after controlling for socio-demographic variables, health indicators and qualitative opinions. Beliefs about others' risk perception was not gender-specific. Age was also an important explanatory variable for self-assessment of beliefs, with elderly people being more optimistic than younger people for both IFRs, own personal risk to catch the disease and expected prevalence. Results on others' risk perception showed a lower impact of age, except for respondents aged 18 and 19 who revealed a significantly lower assessment of others' risk perceptions. Education appeared to be an important determinant of the assessment of IFR, with higher educational levels being associated with lower assessments of the IFR from COVID-19. The same applies to income category, with a tendency for a higher income to reduce beliefs about COVID-19. Regarding employment status, we found some evidence when both surveys were pooled together that being an employee, in either the public or the private sector, was associated with a higher assessed IFR compared to being inactive. The impact of location area was particularly important in survey 1 and almost vanished when both surveys were pooled together. A striking figure of the data is that living in a higher incidence region significantly decreased the assessment of own personal risk to catch COVID-19 and increased the assessment of others' risk perception. Location area had however little impact on expected prevalence or predicted IFR. Regarding health issues, no systematic pattern arose from the surveys. Having health problems was associated with decreased IFRs, but only the passage from very good to good health was significant. Personal experience with COVID-19 increased all reported answers, but it failed to reach significance. On the opposite, having relatives who had been ill from COVID-19 significantly reduced the reported IFRs. Consistently, respondents who reported a high relative risk to catch COVID-19 also reported a higher personal risk to catch the disease, an association reassuring for the quality of the data. In both surveys, they also reported higher values for others' risk perceptions and higher expected prevalence. Respondents who were more worried about catching COVID-19 were also more likely to report high IFRs and high risk to catch the disease. No significant pattern emerged in the two other items Q3 and Q4. The perceived contagiousness of COVID-19 had a strong positive impact on reported answers in both surveys. On the opposite, the perceived seriousness of the disease never reached significance, nor did the level of confidence in reported beliefs. Finally, the duration of the epidemic was significantly associated with the expected prevalence, a result that further validates the quality of the data, and the expected IFR for the seasonal influenza was associated with higher reported levels for most items: IFR (for COVID-19), others' perceived risk and expected prevalence.
Considered jointly, these findings are summarized in the following observation: H3: Results shows individuals reacted to risk information: perceived own risk to catch COVID-19 increased between survey 1 and survey 2 while beliefs on how other people perceive their risk to catch the disease remained stable. IFR also remained stable between the two surveys while expected prevalence decreased. Age, gender and education -but not health-related items-were important explanatory variables for beliefs.
DISCUSSION AND CONCLUSION
In this paper, we set out to investigate how people form beliefs in pandemic risk settings, by implementing two surveys in the French population just after the outbreak of COVID-19. Contrary to several previous studies on this topic, which relied on convenience sampling (Cerami et al., 2020;Wong et al., 2020), we did so using a large representative sample of the general public. Based on two successive surveys conducted during lockdown, before and after the first epidemic peak, we report self-assessments about risks including the IFR for COVID-19, the own personal risk to catch (or catch again) the disease, the risk as perceived by others and the expected prevalence rate in the French population. Our main findings were that risk perceptions in absolute terms were rather high, with the average belief reaching approximatively 16% for the IFR, i.e., three to ten times the clinical figures available at the time of the survey, and ranging from 34 to 45% for expected prevalence. Own personal risk to catch the disease increased significantly between the two surveys from 35% to 46%, whereas the perception of others' risks remained rather stable. Compared to available evidence, such numbers show that respondents overestimated the probabilities to catch or die from COVID-19. This finding is not unusual in the literature (Ferrer and Klein, 2015) and does not necessarily reflect a strong deviation from rational behavior. As shown by Viscusi (1985), overestimating small risks fatalities rationally occurs in a Bayesian model when learning is based on partial information. In the case of the outbreak of an unknown disease, information costs are high, and one would expect that posterior Bayesian probabilities adjust slowly to their objective counterparts. While such a rational model explains why own personal risk to catch COVID-19 rose between the two surveys, it could not explain why the subjective assessment of the IFR remained stable before and after the epidemic peak.
We found mixed evidence regarding the effect of gender on the propensity to overestimate risks. In bivariate analysis, women appeared to overestimate risks more than men, but in multivariate analysis, the opposite result emerged in the first survey, while the effect was insignificant for both surveys combined. Previous evidence on this comparison is also ambiguous, with one study reporting men to be more reluctant to wear face-masks than women (Haischer et al., 2020), and another study finding mixed evidence (Howard, 2020). Concerning health protection behavior in case of pandemic outbreaks such as COVID-19, the evidence is more conclusive in that men engage less in it than women (Moran and Del Valle, 2016;Clark et al., 2020).
The theoretical part of the paper builds on the Bayesian learning model to predict a number of effects of lockdown on beliefs. In the words of Gigerenzer (2020), the Bayesian learning model is limited because it remains an as-if theory of mind. Given several sources of information and a risk assessment, the Bayesian learning model describes the optimal solution for a rational individual. One advantage of the Bayesian learning model is to make clear predictions on how experience and available information impact individuals' risk perceptions. The model is far from being perfect, especially in the health domain.
In an extensive review of medical decision-making studies, Blumenthal- Barby and Krieger (2015) found that only 10% of the studies disconfirmed the presence of a bias or heuristic in the patient population under investigation. Among the dozens of biases identified in the literature in health-related decision making, the availability heuristic and the representativeness heuristics are particularly central to risk assessment (Dawson and Arkes, 1987). Because COVID-19 was a new, unknown disease, few specific instances were available to the individual to form their judgments of likelihood. One exception is seasonal influenza. Participants to survey 1 clearly identified COVID-19 as more contagious and more dangerous than seasonal influenza. It is therefore unlikely that the availability heuristic had played a major role in risk assessment. We found large, significant, increases in the assessments of both the own risk to catch COVID-19 and expected prevalence. This shift shows individuals accounted for new information on the disease and updated their prior, hence suggesting the Bayesian learning model provided a better account of the data than the representativeness heuristic. The Bayesian learning model has several downsides, among which the lack of explanations for the psychological processes leading to risk assessments and its inability to account for some important characteristics of ecological rationality (Marewski and Gigerenzer, 2012). One alternative is to replace Bayesian learning with fast and frugal heuristics adapted to fit risk judgments in a given decision-making context. Whereas the Bayesian learning model supposes that more information is always best, the fast and frugal heuristics can achieve superior performance when information is ignored (Hoffrage et al., 2000). Indeed, instead of weighting and averaging all sources of information as the Bayesian learning model, fast and frugal heuristics retain one single predictor, a source of information (e.g., experience) or even a selected part of a source (e.g., own personal experience with COVID-19) to update beliefs. Unfortunately, our dataset did not contain enough information to be able to identify, at the individual level, the first-order predictor of risk perceptions.
Several factors might explain why respondents had higher assessments of their own risk during lockdown and kept their appraised fatality rates constant. First, health risks, and especially IFRs, are often overestimated (Viscusi, 1991;Skinner et al., 1998;Carman and Kooreman, 2014). Second, highly publicized small risks are also often overestimated at the individual level (Lichtenstein et al., 1978;Viscusi and Hamilton, 1999). In this respect, the large media coverage of the contagiousness of the disease during lockdown could have contributed to the increase of perceived risks to catch COVID-19. In the same fashion, the epidemic peak made the risk to catch the disease more immediate, a factor that led risk perceptions to become more pessimistic (Shepperd et al., 2000). Third, risk perceptions are reflective of not only numerical information, through deliberative processes, but also of experiential factors such as the information derived from personal experiences or events that occurred in the individual's inner circle. We found no significant evidence for an impact of personal experience with COVID-19 on beliefs, although lower risk perceptions of the IFR were reported when some of the respondent's relatives (family members or friends) had been diagnosed with a disease. The latter result contradicts evidence described by Chen and Kaphingst (2011) in the context of lung cancer. Personal experience with health issues (health status or experience of a chronic illness) did not show a unilateral impact on beliefs: only chronic illness had a significant, and negative, impact on own risk perception to catch COVID-19.
Over and above the amount of absolute pessimism about IFRs or own personal risk to catch the disease that was found in the data, a comparison of the latter with the expected prevalence rate allowed us to study the dynamics of comparative optimism -or pessimism-during lockdown in France. Comparative optimism arises when an individual gives a comparative risk estimate that is lower than that of a relevant comparison group. The typical metric to identify comparative optimism is an average of these estimates lower than the comparison group (10, Table 1). Such a behavioral trait has been regularly found in the health domain for breast cancer and prostate cancer (Clarke et al., 2000), pneumonia or influenza (Weinstein, 1987). Answers to survey 1 confirmed this pattern with a clear indication that respondents were optimistic about their chances to catch COVID-19. However, after the epidemic peak, respondents in survey 2 revealed comparative pessimism due a concomitant increase in their perceived chances to catch the disease and a decrease in their expected prevalence rate. Such a pattern is consistent with the results of Burger and Palmer (1992) who found a decrease in comparative optimism of Californian students about natural disasters after the 1989 earthquake. One possible explanation is that the lockdown and highly pessimistic information at that time had forced the respondents to focus on information about their own vulnerability to the virus.
The survey items were based on two different question formats that might have impacted the results. The first format, corresponding to the items on the IFR and the expected prevalence, used a base population reference point to elicit beliefs. The second format, which aimed to assess own risk to catch COVID-19 and risk perception by others, used singular terms and was more open ended than the first format. We deliberately chose those two different formats because the first one explicitly refers to the distribution of a characteristic in the population (death for the IFR and diagnostic for the prevalence rate), whereas the second one was meant to emphasize features specific to the individual perspective. One concern with the use of numerical formats to express beliefs is that the answer "50" may either reflect epistemic uncertainty or an inability to translate one's feelings into a number rather than a true numerical assessment of beliefs. As Bruine de Bruin et al. (2002) showed, this is particularly the case for items described in singular terms as compared to those items described in distribution terms. As expected in our data, more "50"s were observed with the singular terms than with the distributional terms. However, because the average answer was close to 50 anyway, we found no significant impact of the "50"s on the answer, except for the own risk to catch COVID-19 in survey 1.
The strict nationwide lockdown at the time of the two surveys imposed several constraints on data collection that might have created specific features of the data. First, due to traveling restrictions for non-essential workers, the poll institute could not organize phone surveys. Although the IFOP panel included more than 750 000 households, the usual sampling methods for representative phone surveys were not available. While this did not impact the representativeness of the data samples, it makes comparisons with traditional, phone-based, surveys on health risks conditional on the survey method. Another technical limitation was the impossibility to build a within-subject design in a longitudinal survey. As a consequence, our investigations on beliefs dynamics were entirely based on between-subject comparisons, lacking precise information away from the observed differences in average patterns.
As a conclusion, this study shows well-calibrated beliefs about COVID-19, especially when it comes to seriousness and contagiousness of the disease and to respondents' own risk of COVID-19. While these findings were obtained in the early pandemic stage characterized with much limited knowledge of the virus and insufficient means to struggle against the outbreak (mainly, lack of medical face masks and screening tests for the general population), they pointed out the importance of eliciting and analyzing individual beliefs over time in a pandemic context where the spread of COVID-19 is highly attributable to individual risky behaviors. | 2021-02-02T18:26:53.176Z | 2021-02-01T00:00:00.000 | {
"year": 2021,
"sha1": "fa945dc0050455ae7c0ae92102d49803e12df401",
"oa_license": "CCBY",
"oa_url": "https://www.frontiersin.org/articles/10.3389/fpsyg.2021.619145/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "fa945dc0050455ae7c0ae92102d49803e12df401",
"s2fieldsofstudy": [
"Sociology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
98597792 | pes2o/s2orc | v3-fos-license | Near-field mapping of three-dimensional surface charge poles for hybridized plasmon modes
Near-field mapping of three-dimensional surface charge poles for hybridized plasmon modes Yu Huang,1 Emilie Ringe,2,a Mengjing Hou,1 Lingwei Ma,1 and Zhengjun Zhang3,b 1State Key Laboratory of New Ceramics and Fine Processing, School of Materials Science and Engineering, Tsinghua University, Beijing 100084, P. R. China 2Department of Materials Science and Nanoengineering & Laboratory for Nanophotonics, Rice University, 6100 Main Street, Houston, TX 77005, USA 3Key Laboratory of Advanced Materials (MOE), School of Materials Science and Engineering, Tsinghua University, Beijing 100084, P. R. China
I. INTRODUCTION
Metallic nanoparticles can undergo light-driven collective oscillations of the conduction electrons known as localized surface plasmon resonances (LSPRs).By virtue of being small, such particles are able to guide and concentrate light at the sub-wavelength scale 1 and provide extremely large, localized enhancements of local electric fields. 2 These unique properties provide possibilities for applications of plasmonic nanoparticles in a wide variety of fields such as plasmonic waveguiding, 3 surface-enhanced spectroscopies, [4][5][6][7] chemical and biological sensing, 8,9 as well as single molecule detection. 10,11Understanding and predicting the fundamental physics governing LSPRs are both necessary to realize and fully optimize potential devices.In particular, mapping the electric fields near plasmonic particles is of primary importance as they determine the enhancement factors and particle-molecule as well as particle-particle coupling.7][18] In addition, experimental mapping of surface plasmons with nanoscale resolution is inherently difficult due to the diffraction limit of light. 19,204][25][26] However the interpretation of plasmonic EELS data can be difficult, even more so when trying to correlate with near-or far-field light scattering as these techniques do not capture the same information, in particular in the presence of dark modes 27,28 or hot spots between coupled nanoparticles. 29or better mapping and understanding of plasmon resonances, experimental optical data on plasmonic nanostructures are typically complemented by comparisons obtained from numerical or analytical approaches, including discrete dipole approximation (DDA), finite difference time domain (FDTD), and boundary element method (BEM).For straightforward geometries, simple plasmon modes such as dipoles can readily be recognized by computing the polarization vector directions 30 or electric field directions 31,32 and then deducing surface charge distributions.Alternatively, artificial surface charges and currents can be attached to boundary elements to map out the plasmon modes using BEM. 23,33The relative merits and limitations of such methods have been reviewed elsewhere. 29,34,35n this work we describe and discuss a new approach to directly obtain maps of the surface plasmon modes at the near-field resonance energies via full electrodynamics calculations.To taken into account spectral deviation between the near-and far-field plasmon resoanances, a near-field spectroscopy is proposed to extract the near-field LSPR energies.Firstly conventional plasmon mapping strategy is applied for futher comparisons.Then the induced surface charge distributions at the near-field resonance energies are calculated by Gauss' law during FEM calculations.Instead of conventional mapping, the three-dimensional (3D) geometry of surface charge poles is used to determine the symmetry of the plasmon resonance modes.Furthermore, by using a nanosphere dimer we demonstrate that complex structures can also be addressed with this approach.Even higher-order hybridized plasmon modes are presented clearly.
II. COMPUTATIONAL METHOD AND THEORY
Finite element method (FEM) 36,37 calculations were performed in COMSOL Multiphysics (installed on a Quad Intel Xeon CPU, 64 GB RAM workstation). 38Specifically, we solve the vector wave equation for the time-harmonic electric field where µ r and ε r are the relative permeability and permittivity, E = (E x , E y , E z ) is the electric field, and k 0 = 2π/λ is the incident wave vector with λ the wavelength of the incident plane wave.For far-field scattering properties, extinction spectra are calculated by integrating the time-averaged extinction Poynting vectors S ext (i.e.electromagnetic power flow) over an auxiliary surface enclosing the nanoparticle 39,40 where E inc , E sca , H inc and H sca are the incident and scattered electric and magnetic field respectively, C ext is the extinction cross section, |W inc | = 1/2cε 0 E 0 2 is the power flow per unit area of the incident plane wave, E 0 (set at 1 V/m here) is the modulus of E inc , c is the velocity of light and ε 0 is the permittivity of vacuum.
Plasmonic nanoparticles are commonly used as enhancing surfaces for spectroscopy, such that computing the near-field enhancement factor for various resonant modes is a valuable tool for spectroscopic applications.The electromagnetic (EM) enhancement factor of SERS is approximately |E 4 |/|E inc | 4 when the Stokes shift is small. 37Assuming Raman probe molecules are arranged randomly and uniformly on the surface of metallic nanoparticles, the averaged EM enhancement factor can be evaluated by averaging the volume integral of |E| 4 /|E inc | 4 within a certain distance above the metal surface (here we take 2 nm, see Fig. S1). 41,42lotting the EF as a function of energy provides valuable information about the near-field enhancement at various wavelengths that can be used to optimize enhancing substrates for specific resonant molecules.For an entire spectrum (i.e.∼100 spectral points) of our model, the computational time is 24∼36 h.The excitation energy producing the largest near-field enhancement can be extracted from the EF plots (as shown in Fig. 1).Classical Gauss' law is applied to calculate the surface charge density ρ at these highly enhancing energies, since plasmonic metals such as Ag and Au are good electrical conductors and hence almost all the induced charge distributes on the particle surface.The Gauss' law in the integral form is: where Φ E is the electric flux through the metal surface S, Q = S ρ dS is the total charge, thus the surface charge density is: where n = (n x , n y , n z ) is the outward normal vector of the particle surface.
III. RESULTS AND DISCUSSION
Fig. 1(a) shows the extinction spectrum of a Ag sphere with radius a = 60 nm calculated by FEM and Mie theory, respectively. 40The extinction efficiency,Q = C ext πa 2 was calculated in 5 nm intervals from 300 to 800 nm; silver was modeled by a Lorentz-Drude dispersion function fitting experimental data. 43,44Mie theory and FEM results are in excellent agreement, with the maximum relative deviation 41 where Q FEM and Q Mie are the extinction efficiency calculated by FEM and Mie theory, respectively.Two peaks are present in both the extinction and the EF spectrum.They are the result of two kinds of plasmon modes.E-field distributions at near-field LSPR energies (λ = 470 and 375 nm) indicated by the red arrows in Fig. 1(a) are presented in Fig. 1(b) and 1(c).The incident light is set along the z-axis with x-axis polarization.The sphere center is placed at the origin where the transient incident field is set to be E 0 • exp(−iϕ), ϕ is the phase.The patterns of E-fields and their directions show a pair of surface charge poles at λ = 470 nm (Fig. 1(b)) and two pairs of surface charge poles at λ = 375 nm (Fig. 1(c)), revealing a dipole and quadrupole plasmon mode, respectively.As the conventional mapping method, a similar analysis can be used to determine the nature of the plasmon resonances for other highly symmetric structures.
Interestingly, for the quadrupole mode, the near-field enhancement peaks at a much lower energy (λ = 470 nm) than the far-field extinction spectrum (λ = 425 nm).4][15][16][17] The diviation is known to depend on the size of the particle, with larger particles displaying more marked shifts, owning to the retardation from the resonance in far-field scattering. 13,30,45For large enough particles, this shift has been observed to be comparable to the resonance half-width. 45To achieve the highest local electric field enhancement, near-field EF spectroscopy (Equation ( 4)) is thus used.
The surface charge distribution within a full oscillation at a specific resonant frequency is calculated using Gauss' law in the differential form (Equation ( 5)); results of this novel approach are shown in Fig. 2 for the two modes previously discussed.The 3D surface charge poles can easily be located when the metallic sphere reaches a maximum transient charge polarization.This occurs at ϕ = 0.34π and ϕ = 1.34π for the dipole mode (Fig. 2(a)) and at ϕ = 0.80π and ϕ = 1.80π for the quadrupole mode (Fig. 2(b)).These poles alternate between negative and positive as the conduction electrons are driven by the oscillating electric field of light.The response of the electron cloud to the incident electric field is however delayed owing to retardation effects, which can be easily observed as a phase delay between the maximum transient charge polarization and the maximum incident field polarization.
Using Gauss' law to map 3D surface charge poles from numerical results on electric fields is a completely general approach and can readily be applied to complex and arbitrary systems such as coupled nanoparticles, patterned substrates, or rough surfaces.Here, we provide specific examples to demonstrate this broad applicability.Let's take a sphere dimer, an important system in sensing and enhanced spectroscopy owing to the electric hot spot at the nanoparticle junction. 18,29,46Results for a dimer composed of two Ag nanospheres (radius a = 60 nm) separated by 2 nm are shown in Fig. 3.The gap size is fixed to be 2 nm in order to avoid the invalidation of classic electrodynamics for the reason that below 1 nm the quantum and nonlocal effects need to be considered. 47Given the symmetry of the system, three unique polarization/propagation geometries of the incident plane wave exist: (I) propagating along z-axis with x-axis polarization, (II) propagating along x-axis with z-axis polarization, and (III) propagating along z-axis with y-axis polarization.The extinction and EF spectra for these three geometries are reported in Fig. 3 Electric field distributions and 3D surface charge pole maps at the EF peak energies mentioned above, are plotted together in Fig. 4, highlighting the strong correlation between resonant mode geometry and local field enhancement.The resonant modes are labeled (i) to (vi) as identified in Fig. 3(c).Depending on the nature of the coupling between the particles, modes can be bonding, non-bonding, or antibonding following the plasmon hybridization model. 48Fig. 4 shows these types of hybridization: (i) bonding dipole mode, (ii) bonding second-order standing-wave mode, (iii) asymmetric bonding dipole mode, (iv) asymmetric bonding quadrupole mode, (v) antibonding dipole mode, (vi) antibonding quadrupole mode.The four bonding modes sustain "hot spots", highly enhancing regions at the interparticle junction, desirable for spectroscopy applications, 2,4 while the two antibonding modes show poor near-field enhancement.The maximum local EF for the bonding modes is as high as 10 9 while it is lower than 10 4 for the antibonding modes.
In particular, the bonding second-order standing-wave mode (ii) is easily confused with the bonding quadrupole mode (iv).If Fig. 4(ii) were simply a two-dimensonal plot similar to Fig. 1(b) and 1(c), it would be recognized as four charge poles for each sphere judging from the boundary charge distributions.Actually seen from the 3D surface charge plot, the two poles in the middle part of the sphere merge into a big pole, thus only three poles charge are present for each sphere.This mode is the second-order standing-wave plasmon mode commonly observed in plasmonic nanorods 20,49 at an energy just above that of the dipole.The plasmon hybridization theory predicts a red shift in wavelength (i.e., decreased resonance energy) for the bonding mode and a small blue shift (i.e., increased resonance energy) for the antibonding mode with respect to the isolated, single particle resonance energy.The red shift is obvious in all of our results, but our 5 nm calculation spacing makes it difficult to assess the presence of a slight blue shift.We thus performed more finely spaced (1 nm) calculations in the energy region of interest (370-380 nm) for the antibonding dipole mode (vi).The single particle quadrupole EF and extinction spectrum maximum was found to be 375 nm and 372 nm, respectively; these values were 376 and 375 nm for mode (vi).A slight blue shift was thus indeed observed: 1 nm in the EF spectrum (376 -375 nm) and 3 nm in the extinction spectrum (375 -372 nm) (see Fig. S4). 41eyond assessing mode geometry and observing near-field and far-field differences, this FEMbased approach is also useful because of its applicability to arbitrary experimental geometries.In The inset in (ii) shows the two small but strong poles located in the junction area.Higher order modes for I and II are reported in the Supplemental Material. 41dition to the particle geometry and structure, the incident light direction (α) can be manipulated.An example is shown in Fig. 5, where α = 45 • .In this case, the calculated extinction spectrum reveals three peaks, at λ = 660, 475, and 415 nm, coinciding with the peak of modes (i), (iii) and (iv) in Fig. 3(b) (See Fig. S5). 41The EF spectrum displays two peaks: the peak at λ = 660 nm corresponds to the bonding dipole mode (i), while the broad peak around λ = 460 nm is a superposition of other bonding modes including (ii), (iii), and (iv) (see Fig. S6). 41This correspondence confirms that the results of α = 45 • excitation is a combination of the two unique excitation geometries I and II discussed previously.This mode combination can easily be discerned graphically, as shown in Fi. excellent correspondence, it is worth mentioning that the explanation surrounding mode mixing is an approximation and is not quantitative at this point.
IV. CONCLUSION
In conclusion, we have introduced and applied a FEM-based tool to calculate and map 3D surface charge poles directly and thus to determine complicated or hybridized plasmon resonance geometries in arbitrary shapes.For a nanosphere dimer, a map of the electric field enhancement alone could not discern between a second order standing wave bonding mode and a quadrupole bonding mode; the model presented here clearly shows the difference.This approach is based on Gauss' law and allows the investigation of near-field enhancement while providing a better understanding of plasmon resonance symmetry and surface charge distribution.This new tool is expected to be of wide appeal given that it can provide the comprehensive physical image of a plasmonic system necessary for fundamental studies and spectroscopy applications.
FIG. 1 .
FIG. 1.(a) Extinction efficiency Q of a = 60 nm radius Ag nanosphere calculated by Mie theory (black line), finite element method (hollow dots), and near-field E F spectrum (red line), calculated with 5 nm steps.The Mie and FEM spectra feature two peaks at λ = 435 nm and 375 nm, while the near-field displays peaks at λ = 470 nm and 375 nm.Slice plot of E-field amplitude distribution and transient E-field directions for (b) λ = 470 nm, ϕ = 0.34π (dipole) and (c), λ = 375 nm, ϕ = 0.80π (quadrupole).The incident light propagates along the z-axis with x-axis polarization.The arrow length is proportional to the transient E-field value.The surface charge poles are indicated by "+" and "-" marks.
FIG. 3 .
FIG. 3. (a) Dimer structure composed of two Ag nanospheres with 60 nm radius and 2 nm inter-particle separation.The three unique excitation geometries are indicated by I, II, and III.(b) Extinction cross section C ext for the three geometries.The positions of extinction peaks due to low-order plasmon resonances are λ = 660 (i) and 430 nm (ii) for geometry I, λ = 480 (iii) and 420 nm (iv) for geometry II, and λ = 430 (v) and 375 nm (vi) for geometry III.(c) Near-field E F spectra of the three geometries.Low-order plasmon resonance energies are λ = 660 (i) and 455 nm (ii) for geometry I, λ = 470 (iii) and 425 nm (iv) for geometry II, and λ = 470 (v) and 375 nm (vi) for geometry III. | 2019-01-03T07:59:14.591Z | 2015-10-22T00:00:00.000 | {
"year": 2015,
"sha1": "c357d2b7100062d48ca341e59d125b8e8fcefec6",
"oa_license": "CCBY",
"oa_url": "https://pubs.aip.org/aip/adv/article-pdf/doi/10.1063/1.4934722/12988566/107221_1_online.pdf",
"oa_status": "GOLD",
"pdf_src": "ScienceParseMerged",
"pdf_hash": "c357d2b7100062d48ca341e59d125b8e8fcefec6",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Chemistry"
]
} |
245165071 | pes2o/s2orc | v3-fos-license | A Novel Clinical Score Predicting the Presence of Fatty Pancreas
Background: Fatty pancreas (FP) has become an increasingly encountered entity in recent years. Several studies have shown an association with several disease states. Aims: we aimed to generate a simple non-invasive scoring model to predict the presence of FP. Method: We performed a retrospective cross-sectional analysis at Galilee Medical Center. Inclusion criteria included patients who underwent endoscopic ultrasound (EUS) for hepatobiliary indications and who had either hyperechogenic pancreas consistent with FP or no sonographic evidence of fatty pancreas. Results: We included 569 patients. Among them, 78 patients had FP by EUS and 491 patients did not have FP. On univariate analysis, obesity (odds ratio (OR) 5.11, p < 0.0001), hyperlipidemia (OR 2.86, p = 0.0005), smoking (OR 2.02, p = 0.04), hypertension (OR 2.58, p = 0.0001) and fatty liver (OR 5.94, p < 0.0001) were predictive of FP. On multivariate analysis, obesity (OR 4.02, p < 0.0001), hyperlipidemia (OR 2.22, p = 0.01) and fatty liver (OR 4.80, p < 0.0001) remained significantly associated with FP. We developed a diagnostic score which included three parameters that were significant on multivariate regression analysis, with assignment of weights for each variable according to the OR estimate. A low cut-off score of ≤1 was associated with a negative predictive value (NPV) of 98.1% for FP, whereas a high cut-off score of ≥2 was associated with a positive predictive value (PPV) of 35–56%. Conclusion: We recommend incorporating this simple score as an aid to identify individuals with FP.
Introduction
Fatty pancreas (FP) is characterized by fat deposition within the pancreatic parenchyma [1], with a prevalence ranging from 16-35% according to abdominal imaging [2,3]. FP is an increasingly encountered incidental finding on abdominal imaging including endoscopic ultrasound (EUS) [4,5]. While the first observation of FP was reported as early as the 1930s, the clinical implications of this finding were not deeply investigated for several decades, as most clinicians perceived this condition as a benign condition without any future clinical consequences. In recent years, FP was demonstrated to be associated with several comorbid diseases [6][7][8][9]. To date, studies that addressed this coexistence reported only disease states that were shown to be associated with FP, such as diabetes mellitus (DM) [10][11][12], metabolic syndrome [6,13], cardiovascular disease [14] and fatty liver [15]. Recent data have revealed a more severe associated comorbid diseases. Previous studies showed a significant association of pancreatic intra-epithelial neoplasia with pancreatic fat accumulation (OR of 6.1; p = 0.001) [7] and FP (OR 17.86; 95% CI 4.935-88.12) [16]. Additionally, FP was shown to promote the malignant potential and spreading of tumors [17]. Notably, recently published data demonstrated that FP was associated with pre-cancerous main duct intraductal papillary mucinous neoplasm (M-IPMN) (OR 2.69, 95% CI 1.05-6.9, p = 0.04) [18] and represented a risk factor for acute pancreatitis [1]. All of these studies reported sole association with FP.
However, none have tried to develop a clinical score encompassing the most significant predictive factors for the presence of FP. Therefore, the aim of the present study is to develop a clinical score based on clinical data that predicts the presence of FP.
Study Design
A retrospective data analysis of all patients who underwent EUS for hepatobiliary indication at Galilee Medical Center from 1 January 2015 to 1 January 2020 was conducted. Extracted data included demographics (age and gender) and background diseases. The study was approved by the local institutional ethics committee. Written informed consent was waived due to the retrospective non-interventional study design.
Methods
All EUS examinations were performed by an experienced advanced endoscopist and were reviewed and approved by another EUS expert. Patients were sedated with intravenous midazolam and propofol according to the decision of the endoscopist.
Study Definitions
Obesity was defined as having a body mass index (BMI) > 30 kg/m 2 . Cirrhosis was defined as end-stage liver disease accompanied by symptoms related to cirrhosis and synthetic function impairment. Obstructive lung disease was defined as a respiratory disease characterized by airway obstruction and demonstrated by reduced FEV1/FVC < 0.7 using spirometry. Congestive heart failure was defined as a reduced heart pump function diagnosed by echocardiogram. Ischemic heart disease was defined by typical symptoms of angina pectoris coupled with atherosclerotic coronary arteries. Diabetes mellitus was defined as having both typical symptoms including polydipsia and polyuria coupled with hemoglobin A1c level ≥ 6.5%. Hypertension was defined as having systolic blood pressure at ≥140 mm Hg and/or diastolic blood pressure at ≥90 mm Hg following 2-3 visits at 1-4-week intervals. Hyperlipidemia was defined as having a serum cholesterol level of ≥240 mg/dl. Fatty liver was diagnosed as having a hyperechogenic texture in the same EUS exam and in the absence of significant alcohol consumption.
Fatty Pancreas Diagnosis in Our Cohort
The diagnosis of FP in our cohort was performed by EUS examinations as demonstrated by the hyperechogenicity of the pancreatic parenchyma compared to the kidneys or the spleen [19] using linear echoendoscope (Pentax, Tokyo, Japan), model 3870.
Statistical Analysis
The main end point of our study was to predict the presence or absence of FP by a combination of simple clinically relevant parameters. A univariate descriptive statistic was used to compare patients with and without FP. Data was reported as frequencies (percentages) for non-continuous categorical variables and as mean ± standard deviation for quantitative continuous variables. Univariate and multivariate logistic regression models were used to evaluate the effect of baseline parameters on the presence of FP by reporting odds ratios (OR) and confidence intervals (CI). Finally, we generated a score predicting FP by attributing weights to the parameters that were significant on multivariate analysis according to the OR estimates. The overall diagnostic accuracy of the scoring system was determined by a receiver operating characteristics (ROC) curve. A threshold for statistical significance was set at a p value < 0.05. All analyses were performed by an experienced statistician using statistical analysis software (SAS vs. 9.4 Copyright (c) 2016 by SAS Institute Inc., Cary, NC, USA).
Demographics and Baseline Characteristics
Overall, we included 569 patients. The indications of the EUS examinations were as follows: abdominal pain with liver enzyme abnormality (374 patients), investigation for acute pancreatitis episodes with unknown etiology (84 patients) and suspected pancreatobiliary malignancies (111 patients). Among them, 78 patients had FP by EUS (group A), as compared to 491 patients who did not have FP (group B). The average age in group A was 62.1 ± 14.1 vs. 62.9 ± 14.1 years in group B. The prevalence of obesity, hyperlipidemia, hypertension and fatty liver were higher in group A (53.8%, 24.4%, 48.7% and 84.6%, respectively) compared to group B (19.3%, 10.2%, 26.9% and 47.2%, respectively). Notably, there was no difference in the gender predominance between the two groups. Table 1 demonstrates the demographics and the clinical characteristics of the study cohort. Notably, there was no difference in the rate of side branch intraductal main papillary mucinous neoplasm (IPMN) (16.7% in group A vs. 20.6% in group B, p = 0.2) or for mixed type IPMN (2.6% in group A vs. 2.7% in group B, p = 0.4). Main duct IPMN was commoner in group A (10.3%), compared to 3.3% in group B (p = 0.02). Moreover, there was no difference in the rate of pancreatic adenocarcinoma or neuroendocrine tumor in group A (12.8% and 5.1%, respectively) vs. 8.2% and 5.9% in group B (p = 0.09 and 0.4), respectively.
Model Building
For the generation of a diagnostic score, we assigned weights for each variable according to the OR estimates in the multiple logistic regression model. Accordingly, obesity, hyperlipidemia and fatty liver received 1 point each, making a maximal score of 3 points. The ROC for this score is 0.77 (OR 3.54, 95% CI 2.56-4.89, p < 0.0001) (Figure 1). Further analysis to assess the diagnostic performance of the score in terms of sensitivity and specificity, positive predictive value (PPV) and negative predictive value (NPV) revealed that a score of ≥2 was associated with a very high specificity and variable range of sensitivity, while a score of ≤ 1 was associated with a high sensitivity and excellent NPV (
Model Building
For the generation of a diagnostic score, we assigned weights for each variable according to the OR estimates in the multiple logistic regression model. Accordingly, obesity, hyperlipidemia and fatty liver received 1 point each, making a maximal score of 3 points. The ROC for this score is 0.77 (OR 3.54, 95% CI 2.56-4.89, p < 0.0001) (Figure 1). Further analysis to assess the diagnostic performance of the score in terms of sensitivity and specificity, positive predictive value (PPV) and negative predictive value (NPV) revealed that a score of ≥2 was associated with a very high specificity and variable range of sensitivity, while a score of ≤ 1 was associated with a high sensitivity and excellent NPV (Table 3).
Discussion
Although FP was shown previously to be an incidental and benign finding encountered on abdominal imaging studies [20], the clinical consequences of this condition have rarely been investigated and discussed among clinicians as its diagnosis was not routinely translated into clinical decision-making. To date, from the studies that have reported several disease associations with FP, none have tried to incorporate clinical parameters into a score that predicts the presence of FP. Herein, we generated a simple, easily available score with high accuracy based on the presence of three variables, including: obesity, hyperlipidemia and fatty liver with a good ROC of 0.77 and a very high specificity of 82.1-98.4%. Among patients with a score of 1, FP was absent in 98.1%, while among patients with a score of ≥2, FP was present in 35-56%.In fact, besides the incidental identification of patients with FP, this score enables clinicians to proactively identify and diagnose patients with FP, especially in the current era where data that report an association of FP with severe and life-threatening diseases are accumulating as demonstrated by recent studies [1,18]. In particular, this refers to the association of FP with pancreatic adenocarcinoma [20] and main-duct IPMN [18]. To summarize, this was the first time to show that incorporating several predictors of fatty pancreas in one score improves the probability of identifying this disorder. Similarly, incorporating multiple predictors in a score was shown to improve the probability of predicting common bile duct stone [21]. Therefore, our developed score built from modifiable variables opens the door to developing further accurate scores for FP in order to optimize the identification of patients with FP and to apply appropriate intervention.
Therefore, this work provides clinicians with a simple score for identifying FP presence, as its identification is of paramount importance for assignment of treatment and follow-up plan for this potentially reversible disorder.
The limitations of this study its single-center nature and retrospective design and that we don't have histological confirmation of pancreatic fatty infiltration; however, the straightforward endoscopic ultra-sonographic diagnosis of fatty pancreas as demonstrated by hyperechogenicity makes our diagnosis of FP reliable.
In conclusion, we have developed a scoring system for predicting the presence of FP. We recommend incorporating this score into clinical decision-making regarding the identification of such patients, especially as emerging evidence correlates FP to several pancreatic and extra-pancreatic diseases. Additional prospective studies should be performed to validate our findings and to better define the long-term consequences of FP.
Author Contributions: T.K. and W.S. contributed to the study design and concept. T.K., A.M. and W.S. contributed to data analysis and writing the initial draft. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data are available with the corresponding authors at the Gastroenterology department at Galilee Medical Center and will be available upon reasonable request.
Conflicts of Interest:
Authors declare no conflict of interest regarding this manuscript. | 2021-12-16T16:10:15.142Z | 2021-12-01T00:00:00.000 | {
"year": 2021,
"sha1": "95e1bc6a71c2f32b7b315cc0f15fdd83265c7fe2",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2077-0383/10/24/5843/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "3a448d44a7a74e04d66efeddabc70717493f384f",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
71714688 | pes2o/s2orc | v3-fos-license | Synergistic Roles of Biphasic Ethylene and Hydrogen Peroxide in Wound-Induced Vessel Occlusions and Essential Oil Accumulation in Dalbergia odorifera
The heartwood of Dalbergia odorifera (D. odorifera), named “Jiang Xiang” in traditional Chinese medicine, is highly valuable. Mechanical wounding induced the production of “Jiang Xiang” in D. odorifera. Ethylene and hydrogen peroxide (H2O2) are proposed to play vital roles in wound signaling. However, little is known about the role of ethylene or H2O2 in the wound-induced formation of vessel occlusions and biosynthesis of “Jiang Xiang” in D. odorifera. In this study, the pruning of D. odorifera saplings resulted in the synergistic biosynthesis of biphasic ethylene and H2O2, which was followed by formation of vessel occlusions and “Jiang Xiang” in the pruned stems. In this process, the H2O2 production stimulated higher biosynthesis of ethylene. Treatments with aminoethoxyvinylglycine (AVG), an inhibitor for ethylene biosynthesis and ascorbate acid (AsA), a scavenger of H2O2, markedly reduced the production of ethylene and H2O2, respectively, and the corresponding the percentage of vessels with occlusions (PVO), oil content, and the amount of “Jiang Xiang” formed. These results indicate that ethylene and H2O2 might be important wound signals in D. odorifera that induce vessel occlusions and formation of “Jiang Xiang,” and thus ethylene and H2O2 might play vital roles in “Jiang Xiang” formation in pruned stems of D. odorifera.
INTRODUCTION
Dalbergia odorifera T. Chen (D. odorifera) is a medium-sized tree from the Leguminosae family, and famous for its heartwood, one of the best rosewoods ( Figure 1A) in the world. The heartwood of D. odorifera, named "Jiang Xiang" in traditional Chinese medicine, has been widely used as Chinese Pharmacopeia for centuries to stop bleeding, regulate the "Qi, " dissipate blood stasis, and relieve pain (Yukihiro et al., 1985;Cheng et al., 1998;Wang et al., 2000;Sugiyama et al., 2002;Choi et al., 2009;Liu et al., 2017). However, the heartwood of D. odorifera forms relatively slowly when trees are more than 6-years-old, and the percentage of heartwood over total stemwood is very small (Ma et al., 2017); thus, study on the promotion of heartwood formation is encouraged (Cui et al., 2017). (C) Represents mechanical wound induced "Jiang Xiang" (discolored wood). Meng et al. (2010) found that mechanical wounding induced the production of "Jiang Xiang" in D. odorifera (Figures 1B,C). Similar results were also observed in Santalum album (Wei et al., 2000) and Aquilaria sinensis (Chen et al., 2012;Wang et al., 2016). Xylem cells destined to form tracheids or vessel members, which provide the conduit for water and mineral transportation, undergo apoptosis (Lucas and Liu, 2017). Vessel occlusions develop in the conduit lumen in response to mechanical wounding in many species for wound sealing and reducing the risk of pathogen intrusion (Saitoh et al., 1993;Dute et al., 1999). Although xylem vessels are primarily dead cells at maturity, they are in contact and communication, especially via pits, with living perivascular parenchyma cells that surround vessels. Perivascular parenchyma cells are active in regulating xylem vessels contents. After wounding or fungal pathogen infection, heartwood substances accumulate in the perivascular parenchyma cells, and are released into the infected vessel lumen to format vessel occlusions for restricting the vessel ingress of the fungus (Zhang et al., 2010).
Ethylene is an important regulator of plant development and growth, and is known to be associated with plant defense (Hou et al., 2013;Savatin et al., 2014). Ethylene production in response to wounding has been demonstrated in a wide range of species. Mechanical wounding induced mandarin (Citrus unshiu) (Hyodo and Nishino, 1981), tomato (Solanum lycopersicum) (O'Donnell et al., 1996), winter squash (Cucurbita maxima) (Kato et al., 2000) and lettuce (Lactuca sativa) (Ke and Saltveit, 2010) to regulate the expression of ACS and ACO genes in ethylene synthesis. Sun et al. (2007) found that pruning induced ethylene release and tylose development in grape (Vitis vinifera) stems. Van Doorn et al. (1991) also reported that tylose development was reduced in the cut stem of lilac (Syringa vulgaris) flowers treated with ethylene inhibitors.
Our field experiment showed that direct injection of ethylene and H 2 O 2 into the stems of D. odorifera resulted in vessel occlusions and formation of "Jiang Xiang, " and it was observed that mechanical wounding induced the production of "Jiang Xiang" in D. odorifera. Ethylene and H 2 O 2 have been proposed to play vital roles in wound signaling, but little information is available on the roles of ethylene and H 2 O 2 in the wound-induced formation of vessel occlusions and biosynthesis of "Jiang Xiang" in D. odorifera. Thus, this study was conducted to investigate whether ethylene or H 2 O 2 are involved in wounding-induced vessel occlusions and "Jiang Xiang" formation in D. odorifera.
Plant Materials and Experimental Design
Three-year-old saplings of D. odorifera were selected to grow in a greenhouse at the Research Institute of Tropical Forestry, Chinese Academy of Forestry, Guangzhou City, Guangdong, China. The saplings were grown under day/night temperatures of 31 ± 3/24 ± 4 • C, respectively. The saplings were 1.68 ± 0.32 m in height, with a stem diameter of 2.64 ± 0.21 cm at a height of 10 cm above the ground.
The sapling stems were cut through at 10 cm above the ground in air, water, 0.5 mM aminoethoxyvinylglycine (AVG) (an inhibitor of ethylene biosynthesis) or 1 mM ascorbic acid (AsA) (the special scavenger of H 2 O 2 ). The cut ends were soaked in the water, AVG or AsA for 2 h before being exposed to air.
Ethylene Measurement
Ethylene production in pruned stems was estimated by measuring the ethylene evolved from the cut stem end. To collect gas evolved from the cut, a 5-cm-long rubber tube was attached immediately after each treatment and gas was collected in a 5-mL syringe. Each ethylene sample was collected for half an hour before measurement. The ethylene concentration in the accumulated gas in the syringe was measured at 0.5, 1,2,4,6,9,12,15,18,24,30,36 and 48 h after treatment, respectively, using an analytical gas chromatograph (Shimazawa, Japan). Five replicate stems were used for the ethylene measurement with each treatment and ethylene production is reported as the concentration in the 5-mL headspace. Because of slight differences in the diameter among the shoots, ethylene concentration data were normalized to a shoot with a 2.64-cm diameter (approximate mean diameter).
The gas chromatograph was equipped with Column SE-54 (30 m × 0.32 mm × 0.25 µm) and column temperature was 70 • C. The temperatures of the H + -FID flame detector and vaporization chamber were 100 • C. Helium was used as carrier gas at a flow rate of 50 mL min −1 . Hydrogen was used as fuel gas at a flow rate of 60 mL min −1 . The flow rate of air was 400 mL min −1 and the split ratio was 10:1.
Hydrogen Peroxide Assay
A 1.1-cm-thick stem was collected from the end of each treated stem. Then, a 1-mm-thick section at the end of the cut stem was abandoned, and five 2-mm-thick sections were collected at the positions of 2, 4, 6, 8, and 10 mm from the pruning end. Stem sections were taken at 0, 1, 2, 6, 12, 24 and 36 h after pruning. They were transferred immediately on cutting to liquid nitrogen and stored at −80 • C for measurement of endogenous H 2 O 2 . For each sampling, five replicate stems were used for the H 2 O 2 measurement with each treatment (a total of 140 stems). The samples were immersed in liquid nitrogen and ground to a powder with a pestle and mortar. H 2 O 2 was analyzed by enzyme-linked immunosorbent assay (ELISA) (Sangon Biotech, Shanghai, China) as described previously with minor modification (Hong et al., 2017). Briefly, the samples, standards, and HRP-labeled detection antibody were added successively in the microporous plate pre-coated with H 2 O 2 capture antibody, then incubated and thoroughly washed. The substrate tetramethylbenzidine (TMB) was converted to blue by the peroxidase catalysis and finally converted to yellow under the action of an acid. H 2 O 2 concentrations in the samples were positively correlated with the color intensity. The absorbance (OD value) was measured with a microplate reader at a wave-length of 450 nm to calculate the H 2 O 2 concentration in the sample.
Vessel Occlusion Assessment
Sections were prepared from stems as described in section "Hydrogen Peroxide Assay". For each sampling, five replicate stems were used for assessment of vessel occlusions with each treatment (a total of 140 stems). The samples were collected and fixed in formalin acetic acid-alcohol (FAA) for assessment of vessel occlusions at 0, 1, 2, 3, 4, 5 and 6 weeks, respectively. The percentage of vessel occlusions (PVO) was assessed as previously described (Sun et al., 2007). Briefly, transverse sections (20 µm in thickness) were prepared with a sliding microtome (Leica RM2255, Germany). The sections were temporarily mounted with a cover slip in water, then observed under light microscopy (Olympus BX51, Japan) equipped with a digital camera (Pixera Pro 600ES, United States). Five areas, each containing 20-30 vessels, were chosen randomly for analysis. The analysis of each replication (cut stem) included 100-150 vessels.
Essential Oil Assessment
The samples were collected as mentioned above for assessment of essential oil at 0, 1, 2, 3, 4, 5, and 6 weeks, respectively. For each sampling, five replicate stems were used for assessment of essential oil with each treatment (a total of 140 stems). The samples were immediately immersed in liquid nitrogen and ground to a powder with a pestle and mortar. For each treatment, 1 g powdered stem samples were weighed, immersed in 30 ml petroleum ether and shaken for 24 h. After filtration and concentration (Concentrator 5301, Eppendorf, Germany), essential oil was obtained and the oil content was also calculated.
Essential oil from the 6 week harvest was used for GC-MS with an Agilent 6890 N-5975 I system with an Innowax DB-5MS column (30 m × 0.25 mm, 0.25 µm film thickness). Helium was used as carrier gas at a flow rate of 1 mL min −1 . Oven temperature was programmed to 70 • C for 1 min, raised to 250 • C at a rate of 8 • C min −1 , and kept constant at 250 • C for 15 min. Mass spectra were recorded at 70 eV with the mass range m/z 35 to 450. The identification of essential oil components was done by computer matching against NIST and Wiley GC-MS Library or comparing the retention times of oil components with standard samples.
Statistical Analysis
Data were calculated based on combined averages from five individual saplings (n = 5) per treatment. The plots for the graphs were generated in SigmaPlot 10.0 (Systat, United States). The significance of differences among treatments was evaluated with Duncan's multiple range tests using the data processing software SPSS 17.0 (IBM, NY, United States).
Wounding Induced Enhanced Ethylene Production
Whether stems were pruned in air or in water, production of ethylene increased in response to wounding. Ethylene production increased in a biphasic manner with peaks at about 6 h and 18 h after pruning and an intervening decline to the initial level at 12 h (Figure 2). The second peak was about 3-times FIGURE 2 | Effect of different pruning treatments on ethylene concentrations. Ethylene concentrations were normalized to a cut surface with a 2.64-cm diameter (approximate mean diameter 10 cm above ground). Error bars represent ± S.D, n = 5. greater than the first peak and 15-times greater than the initial concentration. By 30 h after pruning, the ethylene concentration was again at the initial level where it remained with a slight diurnal oscillation (Figure 2).
In contrast, the pattern of ethylene production from the pruned end of stems treated with AVG or with AsA was dramatically different from the treatments in air and water. These two treatments almost eliminated the first rise in ethylene concentration; the second increase was greatly reduced too. The second peak value was 360.82 and 347.30 nL L −1 in stems pruned in air and in water, respectively, while it was only 79.41 and 106.22 nL L −1 in stems pruned in AVG and in AsA, respectively (Figure 2). Ethylene production induced by wounding was greatly suppressed in the presence of AVG and AsA.
Wounding Induced Enhanced H 2 O 2 Production
Wound-induced H 2 O 2 production was measured from 0 to 36 h after pruning at 5 depths from the wound surface. H 2 O 2 also produced in a biphasic manner, peaked at about 2 h and 12 h after pruning and declined to the initial level at 6 h and 24 h (Figure 3). The second peak of H 2 O 2 production was about 2.5 times of the first peak and almost 40 times higher than the initial concentration. Generally, H 2 O 2 production decreased with distance from the wound surface (Figure 3). The peak in H 2 O 2 concentration at 12 h after pruning decreased with distance from the wounded surface, ranging from 98.60 nmol g −1 FW at 2 mm to 29.65 nmol g −1 FW at 10 mm ( Figure 3A).
Wound-induced H 2 O 2 production of the treatment with AsA ( Figure 3D) was greatly reduced compared to the treatments in air and water (Figures 3A,B). However, no significant difference was found in H 2 O 2 production between the treatments in air, water and AVG (Figures 3A-C). The peak in H 2 O 2 concentration at 12 h after pruning was 98.60, 96.75 and 97.68 nmol g −1 FW in stems pruned in air, water and AVG, respectively. By contrast, it was only 40.82 nmol g −1 FW in stems pruned in AsA (Figure 3). Thus, wound-induced H 2 O 2 production was greatly suppressed in the presence of AsA, but not under the influence of AVG.
Wound-Induced Vessel Occlusions and Essential Oil at Different Depths From the Cut Surface
Prior to pruning, the D. odorifera stems had essentially no occlusions. After pruning in air, occlusions were observed in the vessel lumens, and continued production of vessel occlusions sealed some vessel lumens. A significant difference in vessel occlusions was observed between depths from the cut surface (Figure 4). The percentage of vessels with occlusions (PVO) kept increasing in the first 4 weeks after pruning. After that, the changes were small. PVO increased much faster at 2, 4 and 6 mm compared to 8 and 10 mm, and PVO at 4 mm increased at a faster rate than other distances. Four weeks after pruning, the PVO at 2 and 4 mm sections reached 60.98 and 77.01%, respectively, while they were only 16.43 and 12.20% for 8 and 10 mm sections, respectively ( Figure 5A). The pattern of essential oil content was almost synchronous with that of PVO ( Figure 5B).
Nine major volatiles were identified in the essential oil extracted from the samples at different depths from the wound surface. However, none of these volatiles was found in the control samples (stems prior to pruning). The relative amount of volatiles was highest at 4 mm, and decreased significantly above and below this depth ( Table 1). These results showed that wounding induced the most vessel occlusions and essential oil at 4 mm from the cut surface.
Wound-Induced Vessel Occlusions and Essential Oil in Stems of Different Treatments
In the stems pruned in air or water, occlusions were observed in a large number of vessels in secondary xylem (Figures 6A,B), while occlusions were absent in most of the vessels in the stems treated with AVG or AsA (Figures 6C,D). The PVO and essential oil content were markedly affected by the AVG and AsA treatments. At 6 weeks after pruning, PVO at 4 mm section of stems treated in air, water, AVG or AsA reached 80.24, 76.92, 19.6 and 27.46%, respectively ( Figure 7A). Consequently, the essential oil content was 0.45, 0.47, 0.13 and 0.18% for stems treated in air, water, AVG or AsA, respectively ( Figure 7B). These results indicated that suppression of wound-induced ethylene or H 2 O 2 production significantly reduced vessel occlusions and essential oil content.
The kind and quantity of the volatiles in the treatment pruned in AVG or AsA also decreased significantly compared to the treatments pruned in air and water ( Table 2). Similar volatiles were detected in the treatments pruned in air and water. Nine volatiles identified in wild "Jiang Xiang" was found in The same letters indicate that the difference is not significant at the 0.05 level (p > 0.05, Duncan's), "-" means not detected. All data are shown as mean ± standard deviation. both treatments pruned in air and water and relative amounts of them were not statistically different. However, the volatiles of the treatment pruned in AVG or AsA were significantly different from the treatments in air and water. Only two and four volatiles were identified in stems pruned in AVG and AsA, respectively, which had fewer compounds than those in air and water ( Table 2).
DISCUSSION
One of the prominent phenomena of plants in the response to external stimuli is the production pattern of ROS. Biphasic H 2 O 2 production with first a minor burst followed by a second major burst has been reported in several plants. In this study, there were two bursts of H 2 O 2 production in D. odorifera stems after wounding, which was also observed in Arabidopsis and ryegrass (Le Deunff et al., 2004;Song et al., 2006). In addition, there were less vessel occlusions and amount of "Jiang Xiang" at 2 mm depth or at ≥ 6 mm due either high or low concentrations of H 2 O 2 resulting in parenchyma cells necrosis. The medium H 2 O 2 concentration at 4 mm depth induced the most vessel occlusions and "Jiang Xiang." This result was consistent with the fact that H 2 O 2 plays a dual role in plants: it acts as a signal molecule at low concentrations and it leads to necrosis at high levels (Gill and Tuteja, 2010). 2 | Chemical composition and relative amount of essential oil at 4 mm from the cut surface in stems pruned in air, water, AVG or AsA after 6 weeks treatment and prior to pruning (CK).
Along the same pattern, ethylene production also burst twice in response to wounding. A similar pattern of woundinduced ethylene production was described in grape (Sun et al., 2007). Also, the hormone jasmonic acid was induced in a biphasic manner in response to wounding in pea seedlings (Yang et al., 2009). When wounding was conducted in air and in water, similar ethylene or H 2 O 2 production was observed. It is understandable that treatments with ethylene or H 2 O 2 inhibitors significantly inhibited ethylene or H 2 O 2 production in wounded stems, respectively. In contrast, it is intriguing thing that AsA, the special scavenger of H 2 O 2 , also greatly inhibited ethylene production. This result indicated that the reduction in H 2 O 2 inhibited ethylene production induced by wounding. Meanwhile, wound-induced H 2 O 2 production was not suppressed in the presence of AVG, a specific inhibitor for ethylene biosynthesis. In addition, the first H 2 O 2 burst at 2 h (Figure 3) and a significant ethylene production peak at 6 h (Figure 2) were observed. Thereafter, the second H 2 O 2 burst and another ethylene production peak occurred at 12 h and 18 h, respectively (Figures 2, 3). These results suggest that mechanical wounding induced biphasic H 2 O 2 and ethylene production and each H 2 O 2 burst was followed by an ethylene production peak. Hence, it appears that H 2 O 2 production was required for ethylene production. Similar biphasic ROS or ethylene production was observed in plants subjected to other stresses such as ozone (Xie et al., 2011) and pathogen infection (Wi et al., 2012). After inoculation with Pseudomonas syringae, ethylene production in tobacco leaves followed a biphasic pattern reminiscent of H 2 O 2 production (Mur et al., 2008). Wi et al. (2012) further provided evidence that a pathogeninduced oxidative burst was required as an upstream regulator of ethylene production in both phases. They suggested that the rapid transient increases in ROS generation followed by ROS-induced ethylene production act as a determinant of a hypersensitive response, whereas the later massive ROS burst and subsequent massive ethylene production act as a positive determinant of pathogen expansion and cell death in compatible interactions. Wi et al. (2012) also hypothesized that the late massive ROS burst stimulated higher biosynthesis of ethylene, which could promote disease susceptibility. In this process, the biphasic production of ethylene and ROS might be regulated by a self-amplifying loop (Wi et al., 2012). The issue of ROS signal specificity has recently received considerable attention. One possibility is that the specific features of ROS signaling could be perceived and decoded into specific responses, which determine gene expression patterns (Mittler et al., 2011). Despite these results, detailed studies on biphasic H 2 O 2 or ethylene production have not been performed.
Nine major volatiles in wild "Jiang Xiang" (Ma et al., 2017) were identified in the essential oil extracted from stems pruned in air or water, none of which was found in the control samples (stems prior to pruning). As suppression of wound-induced ethylene or H 2 O 2 production markedly reduced vessel occlusions and oil content, the kind and quantity of the volatiles in the treatments pruned in AVG or AsA decreased significantly compared to the treatments pruned in air and water. These results showed that wound-induced ethylene and H 2 O 2 production were accompanied by vessel occlusions and accumulation of "Jiang Xiang" in pruned stems of D. odorifera. In other words, ethylene and H 2 O 2 played vital roles in woundinduced vessel occlusions and accumulation of "Jiang Xiang" in D. odorifera.
This study demonstrated that wound-induced ethylene and H 2 O 2 played vital roles in vessel occlusions and accumulation of "Jiang Xiang" in D. odorifera. They might regulate vessel occlusions and formation of "Jiang Xiang" in pruned stems of D. odorifera. Mechanical wound or fungal pathogen infection can induce structural barriers (including callose deposition, cell wall thickening and vessel occlusions), phytoalexins production, and induced expression of defense-related genes (Ahuja et al., 2012). For example, after fungal infection, rapid vessel occlusions by tyloses and formation of fungitoxic terpenoid aldehydes were observed in resistant cotton (Zhang et al., 1993;Daayf et al., 1997). It is hypothesized that pruning caused the production of wound signals (ethylene and H 2 O 2 ). In this process, the H 2 O 2 production stimulated higher biosynthesis of ethylene. Then synergistic biosynthesis of biphasic ethylene and H 2 O 2 resulted in the formation of structural barriers (vessel occlusions) and formation of phytoalexins ("Jiang Xiang") that may contribute to physical barriers and chemical inhibition of microbes within vessels to prevent their spread. The PVO and amount of secondary substances increased with duration after pruning, and ultimately "Jiang Xiang" was formed in pruned stems of D. odorifera.
CONCLUSION
It was concluded that mechanical wounding resulted in the synergistic biosynthesis of biphasic H 2 O 2 and ethylene, vessel occlusions and "Jiang Xiang" formation in D. odorifera. In this process, the H 2 O 2 production stimulated higher biosynthesis of ethylene. Suppression of ethylene or H 2 O 2 reduced the amount of vessel occlusions and "Jiang Xiang" production significantly. These results indicate that ethylene and H 2 O 2 play vital roles in vessel occlusions and formation of "Jiang Xiang" in D. odorifera in response to mechanical wounding, and thus ethylene and H 2 O 2 could be applied to induce formation of "Jiang Xiang" in D. odorifera. Further study is needed to explore the molecular mechanism of ethylene and H 2 O 2 regulation on formation of "Jiang Xiang" in D. odorifera.
DATA AVAILABILITY
All datasets generated for this study are included in the manuscript and/or the supplementary files.
AUTHOR CONTRIBUTIONS
ZC wrote the manuscript. ZY helped in the experiments. DX revised the manuscript. | 2019-03-08T14:05:33.359Z | 2019-03-08T00:00:00.000 | {
"year": 2019,
"sha1": "31b90827518a16088dc0f3090079e3c6b00f01d1",
"oa_license": "CCBY",
"oa_url": "https://www.frontiersin.org/articles/10.3389/fpls.2019.00250/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "31b90827518a16088dc0f3090079e3c6b00f01d1",
"s2fieldsofstudy": [
"Environmental Science",
"Medicine",
"Biology",
"Chemistry"
],
"extfieldsofstudy": [
"Medicine",
"Biology"
]
} |
22862960 | pes2o/s2orc | v3-fos-license | The Structural Basis of Serpin Polymerization Studied by Hydrogen/Deuterium Exchange and Mass Spectrometry*
The serpinopathies are a group of inherited disorders that share as their molecular basis the misfolding and polymerization of serpins, an important class of protease inhibitors. Depending on the identity of the serpin, conditions arising from polymerization include emphysema, thrombosis, and dementia. The structure of serpin polymers is thus of considerable medical interest. Wild-type α1-antitrypsin will form polymers upon incubation at moderate temperatures and has been widely used as a model system for studying serpin polymerization. Using hydrogen/deuterium exchange and mass spectrometry, we have obtained molecular level structural information on the α1-antitrypsin polymer. We found that the flexible reactive center loop becomes strongly protected upon polymerization. We also found significant increases in protection in the center of β-sheet A and in helix F. These results support a model in which linkage between serpins is achieved through insertion of the reactive center loop of one serpin into β-sheet A of another. We have also examined the heat-induced conformational changes preceding polymerization. We found that polymerization is preceded by significant destabilization of β-sheet C. On the basis of our results, we propose a mechanism for polymerization in which β-strand 1C is displaced from the rest of β-sheet C through a binary serpin/serpin interaction. Displacement of strand 1C triggers further conformational changes, including the opening of β-sheet A, and allows for subsequent polymerization.
␣ 1 -Antitrypsin (␣ 1 -AT) 2 is a member of the serpin class of protease inhibitors. Serpins inhibit their target proteases via a unique mechanism (Fig. 1) (1). Cleavage of a serpin's reactive center loop (RCL) by a protease triggers a massive conformational change in which the RCL inserts into the central -sheet A, becoming a sixth strand. This process disrupts the protease active site and traps the acyl-enzyme intermediate, rendering the protease inactive. The active, loop-expelled form of ␣ 1 -AT is significantly less stable than the loop-inserted form, and as a result of this metastability, the serpin structure is readily disrupted by mutations that result in inappropriate conforma-tional changes. It has been established that a number of serious diseases associated with serpin mutations are caused by the formation and accumulation of serpin polymers (2). The details of the polymerization mechanism are therefore of considerable interest.
Serpin polymerization has been studied extensively by a variety of methods, including circular dichroism, fluorescence spectroscopy, electron microscopy, and x-ray crystallography (3)(4)(5). The most generally accepted model of pathological serpin polymerization is one in which the RCL of one serpin anneals between -strands 3 and 5A of another (1). Two major lines of evidence support this model. The first is that RCLmimicking peptides have been shown to anneal between strands 3 and 5A and have also been shown to block polymerization in vitro (6). The second comes from fluorescence spectroscopy. Distances between donor-acceptor Förster resonance energy transfer pairs were measured at multiple sites in polymerized ␣ 1 -AT, and subsequent modeling found that a structure in which the RCL of one molecule anneals between strands 3 and 5A of another was most consistent with the distance constraints (7). In contrast, although several crystal structures of polymerized serpins have been determined, none of them reveal this type of linkage. Linkage mechanisms observed to date in crystal structures include several types of RCL/-sheet C interactions and an RCL/ strand 6A interaction (8 -10). Furthermore, electron microscopy has detected a head-to-head disulfide-bonded interaction (11). The structure of pathogenic serpin polymers has thus not been unambiguously resolved, although detailed knowledge of the structure of these polymers is critical for understanding the basis of serpin-linked diseases and for designing effective therapeutics.
Hydrogen/deuterium exchange measured by mass spectrometry (HXMS) has proven to be a powerful method for probing the structure and dynamics of proteins and protein assemblies that are not accessible to other techniques, and several recent studies have employed HXMS to gain insight into the structural organization of amyloid fibrils (12)(13)(14). HXMS provides information on dynamics and solvent accessibility throughout the entire structure of a protein and does not require the attachment of exogenous labels (15). In this work, we employed HXMS to probe the local dynamics and solvent accessibility of polymeric wild-type ␣ 1 -AT and to examine the structural changes that occur in monomeric ␣ 1 -AT prior to polymerization. Our results reveal the structure of polymerized ␣ 1 -AT at a new level of detail and suggest a mechanism for polymerization.
MATERIALS AND METHODS
Expression and Purification of Wild-Type ␣ 1 -AT-Wild-type ␣ 1 -AT was expressed and purified as described previously (16). The activity assay on purified ␣ 1 -AT was done as described previously (16). For all samples, the stoichiometry of inhibition was determined to be 1.0.
Native Gel and Silver Staining-Each monomeric and polymeric sample (500 ng each) was loaded onto a native gel (imidazole/HEPES (pH 7.4)) and run at 15 mA in electrophoresis running buffer containing 43 mM imidazole and 35 mM HEPES (pH 7.4). During the run, the electrophoresis apparatus was immersed in ice to prevent sample denaturation and polymer dissociation. Proteins were visualized by silver staining.
Electron Microscopy-A solution containing polymerized ␣ 1 -AT was placed on a carbon-coated electron microscopy grid and visualized by negative staining.
Peptide Mapping by High Performance Liquid Chromatography (HPLC)-Tandem Mass Spectrometry-Peptide mapping experiments were performed as described previously (16) using an LCQ-DECA mass spectrometer (Thermo Electron Corp.). Sequencing of each peptic peptide was carried out using the search algorithm SEQUEST (Thermo Electron Corp.) and by manual inspection.
Hydrogen/Deuterium Exchange of Monomeric Wild-type ␣ 1 -AT and Polymers-Hydrogen/deuterium (H/D) exchange of monomeric wild-type ␣ 1 -AT at 25°C was performed by 10-fold dilution of the sample with 10 mM sodium phosphate (pD 7.8), 50 mM NaCl deuterium buffer. The exchange reaction was allowed to proceed for different time periods, and the reaction was quenched by dropping the pH by the addition of cold 100 mM NaH 2 PO 4 (pH 2.4) buffer containing 2 M guanidine hydrochloride. The quenched sample was immediately frozen in a dry ice/ethanol bath. The final protein concentration was 0.1 M. Similarly, H/D exchange of monomeric wild-type ␣ 1 -AT (0.1 mg/ml) at 45°C was carried out by adding prewarmed 10 mM sodium phosphate (pD 7.8), 50 mM NaCl deuterium buffer. Each sample was incubated in the deuterium buffer for a different time period, and the reaction was quenched as described above. Deuterium labeling times at 45°C were 10, 20, 30, 50, 60, 80, 90, 100, 200, 300, 400, 500, 700, 1000, 2000, 3000, and 5000 s. The final protein concentration was 0.1 M. Temperature dependence of H/D exchange between 25°C and 45°C samples was corrected using the HXpep software package, and accordingly, 25°C samples were incubated in the deuterium buffer for six times longer than the corresponding 45°C samples. For pulse labeling experiments, ␣ 1 -AT was preincubated in H 2 O at either 25 or 45°C and then pulsed with 10-fold dilution into D 2 O at the same temperature, incubated for 10 (at 45°C) or 60 (at 25°C) s, and quenched.
Wild-type ␣ 1 -AT polymer was prepared by incubating the protein (0.1 mg/ml) in 10 mM sodium phosphate (pH 7.8), 50 mM NaCl at 45°C for 4 days. The presence of the polymer was confirmed by electron microscopy and a native gradient gel. The incubated sample was aliquotted and added to prewarmed 10 mM sodium phosphate (pD 7.8), 50 mM NaCl deuterium buffer. The exchange reaction was quenched at different time points by the addition of cold 100 mM NaH 2 PO 4 (pH 2.4) containing 2 M guanidine hydrochloride. The sample was immediately frozen in a dry ice/ethanol bath. The final protein concentration was 0.1 M.
Isotope Analysis by HPLC-Electrospray Ionization Mass Spectrometry-Deuterated protein was digested with porcine pepsin and analyzed by HPLC-mass spectrometry as described previously (16). Briefly, the frozen sample was thawed and digested with 5 g of porcine pepsin on ice for 5 min. Peptic peptides were separated using a micropeptide trap and a C18 HPLC column and eluted with an acetonitrile gradient (2 to 45% over 12 min) at a flow rate of 50 l/min. The centroid mass of each peptide was determined using the MagTran software package. Because we are concerned with differences in deuterium uptake rather than absolute values, the percent exchange for each peptide was calculated directly without a correction for back-exchange. The percent exchange was plotted against deuteration time, and a trend line was drawn using ORIGIN (Origin-Lab). Peptide identification was as described previously (16).
Structure and Dynamics of the Monomeric Intermediate
Preceding Polymerization-Amide hydrogens that are solvent-exposed and not hydrogen-bonded will rapidly exchange with deuterium when a protein is incubated in D 2 O. Hydrogens that are protected from exchange by local structure can exchange if this structure is transiently broken by fluctuations. Hydrogen exchange thus provides a measure of both solvent exposure and conformational flexibility in proteins (17). When coupled with pepsin digestion, HXMS experiments quantify region-specific deuterium uptake (15). The peptides analyzed in this study encompass 71% of the entire amino acid sequence of wild-type antitrypsin covering the most secondary structural regions (Fig. 2). HXMS experiments were performed to monitor the confor- According to a previous study, the active form of ␣ 1 -AT, upon incubation at 45°C, undergoes a conformational change to a partially disrupted structure in which -strands 3 and 5A are separated (3). This conclusion was based on changes seen in intrinsic tryptophan fluorescence and changes in the CD spectrum during the first 4000 s of incubation. This initial conformational change was concentration-independent, but further incubation led to the concentration-dependent formation of polymers. Deuterium labeling was performed at 45°C, with deuteration times ranging from 10 to 5000 s. Under our experimental conditions, ␣ 1 -AT at 45°C retained its monomeric form during the deuterium labeling time (Fig. 3A). Deuterium uptake of monomeric ␣ 1 -AT at 25 and 45°C was monitored. To account for the intrinsic temperature dependence of H/D exchange, shorter labeling times were employed at 45°C relative to 25°C (the temperature dependence and the resulting incubation time differences were calculated using the program HXpep) to achieve comparable "labeling strength" at both temperatures. Deuterium-labeled ␣ 1 -AT was digested with pepsin, and the percent exchange for each peptide was determined. Fig. 4 shows exchange versus time at 25 and 45°C for 25 peptides covering most structural regions of ␣ 1 -AT. As mentioned above, it has been proposed that serpin polymerization is preceded by the accumulation of a species in which -strands 3 and 5A are separated (3). This proposed separation will break a number of backbone hydrogen bonds and expose the amide hydrogens to solvent. Amide hydrogen atoms in these regions should therefore readily undergo exchange. However, in a peptide derived from residues 325-338 that encompasses -strand 5A, we observed, within experimental error, no significant difference in exchange at 25 and 45°C (Fig. 4).
The largest increase in deuterium uptake was observed in -strand 3C (residues 209 -227). At the longest incubation times, the intermediate at 45°C showed a percent exchange increase of 16% relative to ␣ 1 -AT at 25°C. -Strand 3C formed amide hydrogen bonds with -strand 2C (residues 282-290) and -strand 4C (residues 204 -209). Peptide 190 -208, which covers the majority of -strand 4C, showed no increase in exchange at 45°C, indicating that it is primarily strand 3C/2C interactions that are destabilized by heating. Regrettably, we were unable to identify a peptide covering -strand 2C and so have no reporter for the dynamics of that region. Helix C also showed a substantial increase in deuterium uptake at longer incubation times (14%). In addition, much of -sheet B, -strand 1C, helices G and H, and much of helix A showed smaller but still significant increases in exchange at the longer time points (increases of between 7 and 9%). Of particular interest is the peptide spanning residues 352-372 that contains -strand 1C. This region showed an 8% increase in exchange at 45°C. Although this increase is modest, this peptide contains very few residues involved in secondary structure (see Fig. 2), and thus, the observed increase in exchange is consistent with substantial destabilization of those secondary structure elements contained in this region (all of strand 1C and two residues of strand 4B).
Because ␣ 1 -AT was in D 2 O during the incubation at 45°C, the observed increases in deuterium uptake could be due to either increased local fluctuations or local unfolding. To determine which mechanism was responsible for the observed increases in exchange, ␣ 1 -AT was incubated (at 45 or 25°C) for 5000 s in H 2 O buffer and then briefly pulse-labeled in D 2 O buffer. This type of pulse labeling experiment is well established as a means for detecting local and global protein unfolding (18). When corrected for temperature, we observed no significant difference in deuterium uptake at 25 and 45°C (Fig. 5). This finding indicates that ␣ 1 -AT does not adopt a partially unfolded conformation during the initial incubation phase at 45°C. The observed increases in deuterium uptake are therefore due to decreased stability and increased flexibility in local regions of ␣ 1 -AT.
Structure and Stability of Polymeric ␣ 1 -AT-After the initial conformational change (during the ϳ4000-s unimolecular phase), subsequent incubation at 45°C resulted in polymerization, with the process reaching completion in ϳ48 h. We found that incubation of ␣ 1 -AT at 45°C for 4 days at pH 7.8 resulted in a "ladder" pattern on a native gel, which is characteristic of serpin polymers (Fig. 3A). Additionally, electron microscopy revealed "bead" polymers (Fig. 3, B and C) similar to those observed in previous studies. ␣ 1 -AT polymers were labeled with deuterium for different time periods to investigate the stability and solvent accessibility of monomers within the poly-mers in a region-specific manner. Because dissociation of the polymer occurs much slower than our deuterium labeling time (as a previous study demonstrated (19)), dilution of the polymeric sample into D 2 O does not affect the structural integrity of the polymer within our experimental time scale. Fig. 6 shows the results of deuterium labeling for both the monomer and polymer at 45°C. Results for 25 peptides are shown. Regions of ␣ 1 -AT that become shielded from solvent upon polymerization should show reduced exchange compared with the monomer, and Fig. 6 indeed shows a large decrease in exchange in the polymer. The extent of reduction in percent exchange in the polymer relative to the monomer is color-coded and mapped onto the crystal structure of ␣ 1 -antitrypsin in Fig. 7. Most regions in the polymeric form have deuterium uptakes comparable to those of the intermediate. However, there are several regions with notable reductions in the polymer compared with the intermediate. If the polymer is formed by the loop/sheet A polymerization mechanism, peptides derived from the RCL and the center of -sheet A should show significant reductions in exchange. Indeed, a specific region within the RCL had a large reduction in the percent exchange in the polymeric form (Fig. 6). This region consists of residues 339 -351 and showed a 34% reduction in exchange relative to the monomer (at 10 s of incubation in D 2 O). The other half of the RCL, residues 352-372, had only a 9% decrease in percent exchange in the polymer (see "Discussion"). Although the results did not indicate the opening of -sheet A in the monomeric intermediate, a large reduction in percent exchange was observed in a peptide derived from residues 325-338 (16% reduction) containing -strand 5A. These reductions in percent exchange in residues 339 -351 (the RCL) and 325-338 (-strand 5A) suggest that these two regions form the annealing site. In addition, a noticeable decrease in the percent exchange was observed in regions nearby the annealing site including -strands 1A and 3A and helices E and F. A significant decrease in exchange was also observed in -strand 3C (17% reduction). At 45°C, this region is highly unstable in the monomer, and this reduction may reflect restabilization induced by polymerization.
DISCUSSION
The current proposed mechanism of serpin polymerization suggests that polymerization is preceded by the accumulation of a partially unfolded form (often referred to as M*) in which -strands 3 and 5A are separated and that this separation readily allows the annealing of the RCL of a second serpin to form a sixth strand. Following this initial dimerization, subsequent RCL/sheet A linkages with additional serpins lead to the formation of polymers (2, 3). There are two chief pieces of evidence that support the proposed structure of M*. The first is the crystal structure of the ␦-conformation of antichymotrypsin, which reveals partial RCL insertion and partial separation of strands 3 and 5A (20). The second is the change in tryptophan fluorescence that is observed during the initial unimolecular phase that precedes polymerization when ␣ 1 -AT is incubated at 45°C. It is believed that the change in fluorescence is due mainly to changes in the environment of Trp 194 , which is buried and located in the breach region at the top of sheet A.
The separation of -strands 3 and 5A that is hypothesized for M* would break five hydrogen bonds and expose the amide hydrogens involved to solvent. This should result in significant increases in deuterium uptake in peptides covering stands 3 and 5A, and we did not observe such increases, even after 24 h of incubation at 45°C. This lack of increased deuterium uptake in strands 3 and 5A during the first 5000 s of incubation indicates that there is no accumulation of the proposed M* intermediate during this period. Although the hypothesized structure of M* is based in part on the structure of the ␦-conformation of ␣ 1 -antichymotrypsin, it should be noted that the RCL of antichymotrypsin is five residues longer than that of ␣ 1 -AT, allowing it to partially insert into -sheet A without displacing strand 1C. In the case of ␣ 1 -AT with its shortened RCL, it is not clear that a partially inserted conformation would be stable.
We did observe minor increases in flexibility throughout much of the structure and significant increases in flexibility in -sheet C and helix C. Trp 194 is close in space to several residues from region 215-227, which shows the largest increase in exchange upon heating. In particular, the side chain of Trp 194 is very nearly in van der Waals contact with the side chain of Met 221 . Possibly the increase in the flexibility of the local environment is responsible for the changes seen in the fluorescence of Trp 194 at 45°C. Furthermore, increased exchange in residues 64 -77 indicates moderate destabilization of helix C, which may account for the very small change in the CD signal previously seen during the first ϳ5000 s of incubation at 45°C (3).
Many studies, including the current H/D exchange study, support a model of polymerization in which the RCL of one serpin inserts between -strands 3 and 5A of another serpin. If no detectable unfolding occurs during the period preceding polymerization, then how does RCL/sheet A polymerization take place? The increased flexibility seen in -sheet C at 45°C suggests a mechanism. Several studies have shown the importance of strand 1C displacement for polymerization. In particular, an engineered disulfide bond between strands 1 and 2C was sufficient to completely block polymerization (21). Increased flexibility in -sheet C, including strand 1C, at 45°C indicates that interactions between strand 1 and the rest of sheet C are weakened upon mild heating. With these weakened interactions, strand 1C will be more easily displaced from the rest of sheet C. Thus, at 45°C, ␣ 1 -AT initially populates not the M* intermediate, but a species in which -sheet C is significantly destabilized (during the first ϳ5000 s). Following this, there is a lengthy lag phase prior to the appearance of polymers. If strand 1 were simply removed from sheet C by thermal fluctuations, it is not clear why the resulting species would not quickly proceed to the latent form, which is incapable of forming polymers. We suggest instead that strand 1C is displaced by an interaction with another serpin (Fig. 8). This displacement of strand 1C is the initial event that triggers the opening of sheet A and the formation of the M* intermediate. This intermediate could then quickly proceed to form a dimer with the nearby second serpin, which will go on to form polymers. Zhou and Carrell (22) have recently demonstrated that it is dimers, rather than monomers, that initiate and propagate serpin polymerization. They proposed a structure of the "activated," polymerization-promoting dimer that requires the complete removal of strand 1C from -sheet C in one of the two monomers (22).
Polymerization results in changes in deuterium uptake throughout much of the structure of ␣ 1 -AT. We distinguish between two types of changes. Amide hydrogens that are solvent-exposed and not hydrogen-bonded will exchange completely within ϳ10 s under our labeling conditions, and therefore, differences in deuterium uptake at the earliest time points can be attributed primarily to changes in solvent exposure upon polymerization. Changes in the rate of deuterium uptake over longer time periods can be attributed to changes in stability/ flexibility. The largest change in solvent accessibility upon polymerization is seen in the N-terminal portion of the RCL. Peptide 339 -351, which includes RCL residues P14 -P8, shows a large decrease in exchange at 10 s in the polymer relative to the monomer. Less dramatic changes are seen in the C-terminal portion of the RCL. This would seem to conflict with the observation that peptides corresponding to P7-P3 can anneal between the lower portions of -strands 3 and 5A and that this is sufficient to block serpin polymerization in vitro, suggesting that these residues, located in the C-terminal half of the RCL, play a critical role in polymer formation (we note, however, that in the case of wild-type ␣ 1 -AT, the peptide FLEAIG, corresponding to RCL residues P7-P2, is not an effective blocker of polymerization) (23). However, the peptic fragment that includes these residues (fragment 352-372) is quite large and contains, in addition to part of the RCL, both -strand 1C and a portion of strand 4B. Interpretation of the results for this peptide is therefore not straightforward. As mentioned previously, considerable evidence indicates that polymerization requires the removal of strand 1 from -sheet C. If this is the case, the changes in deuterium uptake in peptide 352-372 will reflect the net result of both losing protection in -strand 1C and gaining protection due to the insertion of the RCL into -sheet A. The relatively small changes in deuterium uptake seen in this region may be due to these two opposing factors. Although our results are not conclusive regarding the C-terminal half of the RCL, they very strongly demonstrate that the N-terminal half is involved in newly formed interactions in the polymer.
The other region in which large differences in deuterium uptake are seen at the earliest incubation times consists of residues 160 -172. This region comprises the C-terminal half of helix F and part of the loop connecting it to -sheet A. Both this study and a previous study (16) found that this region shows very little protection from exchange in the monomer, indicating that the structure in this region is at best marginally stable. In the polymer, exchange at 10 s is reduced by ϳ36%. This increase in protection could indicate either that part of this region is buried in an intermolecular interface or that interactions in the polymer stabilize the C-terminal portion of helix F. Protein engineering has demonstrated that destabilizing mutations in the C-terminal (but not N-terminal) half of helix F accelerates polymerization (24). The dramatic reduction in deuterium uptake in the polymer relative to the monomer strongly supports an important role for helix F in polymerization.
In addition to changes in solvent accessibility, H/D exchange also reveals changes in conformational stability/flexibility upon polymerization. Significant decreases in the rates of deuterium uptake are seen in -strands 3 and 5A, which is consistent with this region being stabilized by the annealing of the RCL of another serpin to form strand 4A. Helices D and E, -strands 1A and 3C, and the N-terminal portion of helix F are also substantially stabilized in the polymer. Most likely the large increase in stability seen in the C-terminal portion of helix F propagates to the N-terminal portion, explaining the decreased rates of exchange seen there. Helix E is very close to the C-terminal portion of helix F and is also in van der Waals contact with helix D. The stabilization of the top of helix F might result in stabilizing contacts with helix E, which are in turn propagated to helix D. Although -strand 3C is clearly stabilized upon polymer formation, the mechanism for this stabilization is unclear.
Unlike wild-type ␣ 1 -AT, the Z-mutant spontaneously forms polymers at 37°C (25). Presumably the Glu 342 3 Lys mutation, which destroys a salt bridge, introduces an unfavorable Lys/Lys electrostatic interaction at the top of -sheet A and promotes facile separation of strands 3 and 5A. The Z-mutant may therefore populate a monomeric strand-separated structure rather than require an interaction with a second serpin to induce strand separation. As the polymerization of wild-type ␣ 1 -AT is frequently used as a model for the polymerization of pathogenic mutant serpins, potential differences in the polymer- ization mechanisms of wild-type and Z-mutant ␣ 1 -AT require investigation.
In addition to crystal structures that suggest multiple possible linkage mechanisms, there is evidence from polymerization kinetic studies indicating that different serpins may form structurally distinct polymers. It was shown that when preformed polymers of the serpin ␣ 1 -antichymotrypsin were added to a solution of monomers, the rate of polymerization was increased through a "seeding" mechanism similar to that seen for amyloid proteins (26). However, preformed polymers of other serpins, including ␣ 1 -AT and antithrombin, did not seed polymerization of antichymotrypsin. This serpin-specific seeding suggests that the polymer structures, and possibly the polymerization mechanisms, of different serpins may differ in significant ways. Comparison of the extensive data on wild-type ␣ 1 -AT polymerization with not just ␣ 1 -AT mutants but also with other serpins will be important in establishing the commonalities and differences in serpin polymerization.
The serpinopathies are similar to amyloid diseases in that protein conformational change and subsequent polymerization are strongly implicated as an underlying cause of pathology. It is therefore instructive to compare our H/D exchange results on polymerized ␣ 1 -AT with H/D exchange studies of amyloid fibrils. At the level of overall morphology, it has already been observed that serpin polymers differ from amyloids. Serpins generally adopt short polymers with a beadlike structure rather than the long fibrils formed by amyloids. We found that there are also differences at the microscopic level. Amyloid fibrils have generally been found to contain a core region showing extreme protection from H/D exchange (Ͼ95% of amide hydrogens in these regions remain unexchanged even after several days), indicating the almost complete exclusion of solvent from the core (27). Furthermore, some amyloid proteins, such as human prion protein, undergo a complete alteration in their tertiary and secondary structures upon fibrilization (14). In contrast, the monomers in polymeric ␣ 1 -AT retain a large degree of conformational flexibility. In fact, excepting the annealing site and helices E and F, the distribution of relative flexibility in the structure is similar to that seen in the monomeric form. These observations are consistent with a model in which serpins do not undergo a major change in their tertiary fold upon polymerization, instead remaining largely native-like apart from local changes required for the formation of loop/ sheet linkages.
Protein fibrilization/polymerization is often described in terms of the accumulation of a partially unfolded aggregationprone intermediate arising from destabilization of the native structure, followed by interprotein interactions leading to the formation of amyloid fibrils or polymers. Although we do not detect the accumulation of a partially unfolded structure, the increased rates of H/D exchange seen at 45°C, particularly in -sheet C, do indicate that the native structure of ␣ 1 -AT is destabilized by moderate temperatures and that this destabilization (presumably by facilitating the facile displacement of stand 1C) is an important step in the polymerization process. Factors other than temperature, mutations in particular, can also destabilize the native conformation. This is clearly reflected in the fact that many mutant serpins readily form polymers at physiological temperatures. Now that H/D exchange has elucidated the details of wild-type ␣ 1 -AT polymerization, comparative studies on the polymerization of pathological mutants such as Glu 342 3 Lys (the Z-mutant) can be pursued.
Both mild denaturant and mild temperature increases have been shown to induce the polymerization of ␣ 1 -AT (28). Our current and previous H/D exchange results indicate that these two perturbations have dramatically different effects on ␣ 1 -AT structure: mild denaturant causes a transition to a molten globule state with a solvent-accessible core (29), whereas mild temperature increases result in increased structural fluctuations but not local or global unfolding. Therefore, subtle changes in the environment allow ␣ 1 -AT to explore a wide range of different structural states. For a full understanding, serpin polymerization must be considered in the context of a complex energy landscape in which folding, stability, dynamics, and polymerization are intertwined. | 2018-04-03T05:08:20.257Z | 2008-11-07T00:00:00.000 | {
"year": 2008,
"sha1": "23b9d8fd67ccfb8ff810c7c851f5b65fb92026c6",
"oa_license": "CCBY",
"oa_url": "http://www.jbc.org/article/S0021925820646543/pdf",
"oa_status": "HYBRID",
"pdf_src": "Highwire",
"pdf_hash": "ce8b29dec0294ab4ea06b35b2e463488a043b59d",
"s2fieldsofstudy": [
"Chemistry",
"Medicine"
],
"extfieldsofstudy": [
"Chemistry",
"Medicine"
]
} |
219949197 | pes2o/s2orc | v3-fos-license | The Effect of Optically Induced Dielectrophoresis (ODEP)-Based Cell Manipulation in a Microfluidic System on the Properties of Biological Cells
Cell manipulation using optically induced dielectrophoresis (ODEP) in microfluidic systems has attracted the interest of scientists due to its simplicity. Although this technique has been successfully demonstrated for various applications, one fundamental issue has to be addressed—Whether, the ODEP field affects the native properties of cells. To address this issue, we explored the effect of ODEP electrical conditions on cellular properties. Within the experimental conditions tested, the ODEP-based cell manipulation with the largest velocity occurred at 10 Vpp and 1 MHz, for the two cancer cell types explored. Under this operating condition, however, the cell viability of cancer cells was significantly affected (e.g., 70.5 ± 10.0% and 50.6 ± 9.2% reduction for the PC-3 and SK-BR-3 cancer cells, respectively). Conversely, the exposure of cancer cells to the ODEP electrical conditions of 7–10 Vpp and 3–5 MHz did not significantly alter the cell viability, cell metabolic activity, and the EpCAM, VIM, and ABCC1 gene expression of cancer cells. Overall, this study fundamentally investigated the effect of ODEP electrical conditions on the cellular properties of cancer cells. The information obtained is crucially important for the utilization of ODEP-based cell manipulation in a microscale system for various applications.
Among the techniques for microparticle manipulation in a microfluidic system, microparticle manipulation using DEP mechanism has been widely adopted for various applications (e.g., microparticle enrichment [18,19], microparticle patterning [10,11], cell-type classification [20], or rare cell isolation [3,4]). DEP-based microparticle manipulation, first presented in the 1950s [21], is well-described elsewhere [22], and can be briefly described as follows. When dielectric microparticles (e.g., biological cells) are suspended in a solution in which an electric field is exerted, charges can be electrically polarized on the microparticles' surface. The interaction between the induced charges on the microparticles and the electric field specifically exerted on them can generate a DEP force on such microparticles [22]. In practice, therefore, the scientists can delicately control the electric field exerted on biological cells via a designed microelectrode array, to manipulate these cells in a manageable manner. Although the DEP technique is proven feasible for the fine manipulation of cells, this technique normally requires a costly, time-consuming, and technically demanding microfabrication process to create a unique metal microelectrode layout that is specific to the application [23,24]. In addition, cell manipulation using the DEP mechanism might be technically challenging for general scientists, due to its complexity. To tackle the technical issues, the optically induced dielectrophoresis (ODEP)-based technique could open up a new horizon for the manipulation of microparticles in a more efficient, flexible, and user-friendly manner [23,24].
Microparticle manipulation based on the ODEP mechanism was first proposed in 2005 [23]. Similar to the DEP mechanism described earlier, an electric voltage is applied between the top and bottom substrates of an ODEP system to generate a uniform electric field in the solution layer sandwiched between the two substrates. Under this circumstance, the dielectric microparticles suspended in the solution layer are electrically polarized. In contrast to the DEP mechanism, however, the key technical merit of the ODEP mechanism is that it can easily and quickly create or modify an electrode layout through the control of optical patterns, by acting as a virtual electrode [23,24]. When the photoconductive layer of an ODEP system is briefly projected with light, it can lead to a significant reduction in electrical impedance in the light-illuminated area. This phenomenon can cause the exerted voltage to decrease across the solution layer inside the light-projected region. This phenomenon, therefore, generates a locally nonuniform electric field within an ODEP system. For microparticle manipulation, scientists can simply utilize a commercial projector to project optical images on an ODEP system, by which the interaction between the generated nonuniform electric field and an electrically polarized microparticle is used to manipulate the microparticle. In terms of operation, one can control the light image projected onto the ODEP system and thus the nonuniform electric field created to manipulate dielectric microparticles within the ODEP system.
The use of ODEP mechanism in microfluidic systems has been successfully applied for the high-accuracy positioning and assembling of metallic beads [25], dynamic analysis of cancer-immune microenvironment [26], or the isolation and purification of rare cell species in clinical samples (e.g., Raji cells [7], circulating tumor cells (CTCs) [5,9], and CTC clusters [6]) in a higher performance manner than their conventional counterparts. Although the incorporation of the ODEP mechanism in a microscale system has provided a powerful tool for the biological cell-relevant studies or applications, one fundamental question has to be answered before its widespread applications-Whether the cell manipulation using the ODEP mechanism influences the cellular properties. If this is the case, the use of such a mechanism for cell manipulation could complicate the subsequent biological studies or applications. To the best of our knowledge, however, this fundamental issue has not yet been well explored.
To investigate the effect of ODEP-based cell manipulation (the operating conditions explored-(1) the magnitude of AC voltage: 5-10 V; (2) frequency of AC voltage: 1-5 MHz, (3) operating time: 3 min) on the properties (e.g., cell viability, metabolic activity of cells, or gene expression of cells) of biological cells (e.g., PC-3 and SK-BR-3 cancer cells), we performed various experiments. Within the experimental conditions investigated, the ODEP-based cell manipulation with the best performance (e.g., highest cell manipulation velocity) occurred at 10 Vpp and 1 MHz, for the two cancer cell types explored. Under this ODEP electrical condition, however, the cell viability of the cancer cells tested was significantly affected (e.g., 70.5 ± 10.0% and 50.6 ± 9.2% reduction for the PC-3 and SK-BR-3 cancer cells, respectively), possibly due to the electrical lysis of cells. Conversely, the exposure of cancer cells to the ODEP field with particular conditions (e.g., 7-10 Vpp, 3-5 MHz, and 3 min exposure time) might not significantly alter the cell viability, cell metabolic activity, and gene expression of cancer cells. Within this ODEP operating condition, the highest maximum velocity (e.g., 106.7 ± 30.6 µm s −1 and 100.0 ± 20.0 µm s −1 for the PC-3 and SK-BR-3 cancer cells, respectively) of a light image that can manipulate cells occurred at the voltage magnitude and frequency of 10 Vpp and 3 MHz, respectively. The above-mentioned cell manipulation velocities are technically sufficient for various applications. As a whole, this study fundamentally investigated the effect of ODEP electrical conditions on the cellular properties of cancer cells. The information obtained is crucially important for the utilization of ODEP-based cell manipulation in a micro-scale system for various applications.
Microfluidic Chip and Experimental Setup
To explore the effect of ODEP on the properties of biological cells, we designed a simple microfluidic chip. Figure 1a schematically presents the top-view layout of the microfluidic chip, mainly encompassing a microchamber (L = 6.0 mm, W = 4.0 mm, and H = 50.0 µm) connecting two microchannels (L = 2.0 mm, W = 1.0 mm, and H = 50.0 µm). In this work, an ODEP field was applied to the microchamber (Figure 1a). The structure of the microfluidic chip is illustrated in Figure 1b. Briefly, the microfluidic chip consists of two custom-made polydimethylsiloxane (PDMS) (Sylgard ® 184, Dow Corning, Midland, MI, USA) adapters for tubing connection (A), an indium-tin oxide (ITO) glass (10 Ω, 0.7 mm; Innolux Corp., Miaoli County, Taiwan) (B), a processed adhesive tape (L298, Sun-yieh, New Taipei City, Taiwan; thickness: 50 µm) containing hollow microchamber and microchannel structures (C), and a bottom ITO glass (D) with a coating layer of photoconductive material (a 1.0 µm-thick hydrogenated amorphous silicon (a-Si:H) layer). The fabrication processes (e.g., PDMS replica mold, metal mold-punching fabrication, plasma-enhanced chemical vapor deposition (PECVD)-based thin-film technology, and plasma oxidation-aided bonding) were well-described previously [8,27,28]. After each substrate was fabricated (Figure 1b), substrate A was bonded with substrate B through the surface treatment of plasma oxidation. This step was followed by the assembly with the substrate D with the aid of the processed double-sided adhesive tape (substrate C) ( Figure 1b). In operation, the prepared cell suspension was manually loaded into the microchamber, using a pipette and a tip. For cell manipulation using the ODEP mechanism, the commercially available ODEP-based cell isolation equipment (Celnostics, Ace Medical Technology Co., Ltd., Taipei City, Taiwan) was used to achieve the ODEP-based cell manipulation in the proposed microfluidic chip. Within this equipment, briefly, a function generator (AFG-2125, Good Will Instrument Co., Ltd., New Taipei City, Taiwan) was utilized to apply an AC bias between the two ITO glasses. A computer-interfaced digital projector (EB-X05, Epson, Nagano, Japan) was used to illuminate light images onto the photoconductive material (i.e., the Substrate D) of microfluidic chip to generate the ODEP force on the cells. An illustration of the important operating modules of the equipment is schematically presented in Figure 1c (a photograph of the overall experimental setup is provided in a Supplementary Figure S1).
The Assessment of the ODEP Manipulation Force Generated on Cells
The working principle of ODEP for cell manipulation is described in the Introduction. The ODEP force generated on a cell can theoretically be described by Equation (1) [24,27,28]: In Equation (1), r (cellular radius), ε0 (vacuum permittivity), εm (relative permittivity of the surrounding solution), ∇|E| 2 (gradient of electric field squared), and Re[fCM] (real part of the Clausius-Mossotti factor (fCM)) are the important parameters [24,28]. The fCM is described by Equation (2) [29][30][31]: where * and * represent the complex permittivity of the cell and the surrounding solution, respectively. For a single-cell model, the complex permittivity of the cell and the surrounding solution can be further described by Equations
The Assessment of the ODEP Manipulation Force Generated on Cells
The working principle of ODEP for cell manipulation is described in the Introduction. The ODEP force generated on a cell can theoretically be described by Equation (1) [24,27,28]: In Equation (1), r (cellular radius), ε 0 (vacuum permittivity), ε m (relative permittivity of the surrounding solution), ∇|E| 2 (gradient of electric field squared), and Re[f CM ] (real part of the Clausius-Mossotti factor (f CM )) are the important parameters [24,28]. The f CM is described by Equation (2) [29][30][31]: where ε * cell and ε * m represent the complex permittivity of the cell and the surrounding solution, respectively. For a single-cell model, the complex permittivity of the cell and the surrounding solution can be further described by Equations (3) and (4): In Equations (3) and (4), C * mem represents the complex cell membrane capacitance, ε * int represents the complex permittivity of the cellular interior (i.e., cell cytoplasm), R represents the radius of the cellular interior, d represents the thickness of cell membrane, ε represents the relative permittivity of the cell membrane, cellular interior, or surrounding solution (denoted by the subscript mem, int, or m, respectively), σ represents the conductivity of the cell membrane, cell interior, or surrounding solution (denoted by the subscript mem, int, or m, respectively), j represents the imaginary vector ( √ −1), and ω represents the angular frequency (i.e., ω = 2πf ) of the applied AC field, respectively [29][30][31]. Under a given solution condition, overall, the magnitude and frequency of the electric voltage applied could play important roles in the ODEP force generated on a particular cell [24,28]. In this work, the effect of electric conditions (i.e., the magnitude and frequency of the electric voltage exerted) on the ODEP force generated on cells was experimentally evaluated. In the evaluation, the ODEP manipulation force, a net force between the ODEP and the friction force, acting on the manipulated cells was then experimentally assessed, based on the method described previously [8,27,28]. In a steady state, the ODEP manipulation force acting on a cell was balanced by the viscous drag of fluid acting on such a cell under a continuous flow condition. As a result, the hydrodynamic drag force of a moving cell was used to evaluate the net ODEP manipulation force of a cell according to Stokes' law (Equation (4)): In Equation (5), r (cellular radius), η (fluidic viscosity), and v (the velocity of a moving cell) are the important parameters. Under the given solution and cellular size conditions, overall, the ODEP manipulation force of the manipulated cell could then be experimentally assessed through the measurement of the maximum velocity of a moving optical image that can manipulate such a cell [8,27,28]. In practice, briefly, a light bar image with different moving velocities (e.g., from low to high velocities) was used to manipulate a cell (e.g., attracted and pulled a cell). Through this process, the maximum velocity of a moving optical image that can manipulate such a cell was then determined. In this work, therefore, the above-mentioned velocity was utilized as an index for the evaluation of the ODEP manipulation force generated on a specific cell under a particular electric condition. Based on this, the effect of electric conditions (e.g., magnitude of AC electric voltage: 7-10 Vpp and frequency of AC electric voltage: 1-5 MHz) on the ODEP manipulation of the cells tested (e.g., PC-3 and SK-BR-3 cancer cells) was evaluated. Briefly, the cell sample tested was prepared in a cell suspension (cell density: 10 6 cells mL −1 ), followed by loading into the microchamber of the microfluidic chip ( Figure 1a). The maximum velocity of a moving light bar (L: 1.3 mm W: 100.0 µm) that could manipulate these cells was then assessed [27,28].
Evaluation of the Properties of Cancer Cells Treated with Varied ODEP Operating Conditions
For the analysis of the ODEP effect on the cellular properties, the cancer cells tested (e.g., PC-3 and SK-BR-3 cancer cell lines, two of the commonly-used cancer cell lines in cancer-related studies [32,33]) were first treated with the ODEP fields under different conditions for 3 min, followed by assaying their cellular properties, including cellular viability, cellular metabolism activity, and gene expression. In this study, the biological assays were carried out at 1.5 ± 0.2 h after the ODEP exposure treatment. In brief, the background medium of the prepared cancer cell suspension (cell density: 5 × 10 6 cells mL −1 for PC-3 cancer cells, and 3 × 10 6 cells mL −1 for SK-BR-3 cancer cells) was first replaced by a 9.5% (w/v) sucrose buffer solution (relative permittivity:~76.19 [34], fluid viscosity: 1.0389 × 10 −2 g s −1 cm −1 [35], osmolality: 270-290 mOsmol kg −1 , and conductivity: 1-5 µS cm −1 ), the commonly used low conductivity buffer solution for ODEP-based cell manipulation [5,8,9,27,28]. The processed cell suspension sample was then loaded into the microchamber of a microfluidic chip (Figure 1a). In this process, however, a small portion of cells might retain in the microchannel area. In this work, the cell manipulation using ODEP was first performed to quickly (operation time-within 5 s) transport these cells to the microchamber area before ODEP treatment. At the microchamber zone, 40 rectangular light bar images (L: 4.0 mm, W: 100.0 µm, the interval between light bars: 50.0 µm) were then illuminated to provide an ODEP field on the cancer cells tested. In this study, ODEP fields with different electric conditions (magnitude of AC electric voltage: 7-10 Vpp and frequency of AC electric voltage: 1-5 MHz) were used to treat the cancer cells tested. After exposure to the ODEP field for 3 min, the treated cells were then harvested for the subsequent bioassays. In the process, briefly, a suction-type syringe pump was utilized to collect the treated cancer cells through the hole for harvesting cells (Figure 1a). Before the following bioassays, the collected cells were kept in the form of cell suspension in the original sucrose buffer solution under the thermal condition of 25 • C.
In this work, the commonly-used Cell Counting Kit-8 (CCK-8) (CK04-05, Dojindo, Kumamoto, Japan) and ATP Colorimetric/Fluorometric assays (K354-100, BioVision Inc., Milpitas, CA, USA) were utilized to assay the cell viability and cell metabolic activity of cancer cells, respectively. The bioassays were carried out according to the manufacturer's instructions. For further analysis of whether cell manipulation using ODEP could affect cellular gene expression, the expression levels of genes in cancer cells treated with different ODEP fields were experimentally evaluated. As the ODEP-based cell manipulation was commonly used for the isolation and purification of circulating tumor cells (CTCs) [5,9], the important CTC-related gene expressions were thus explored in this work. For the cancer cells (PC-3 and SK-BR-3 cancer cells) tested, the mRNA levels of epithelial-to-mesenchymal transition (EMT)-associated genes [EpCAM (Hs00158980_m1) and VIM (Hs00958111_m1)], the multidrug resistance-associated protein 1 (MRP1) gene [ABCC1 (Hs01561502_m1)], and the housekeeping gene [GAPDH (Hs02758991_g1)] were experimentally quantified. The bioassay was based on a method described previously [8,9,27]. In brief, RNA was extracted from the cancer cells tested using a bromochloropropane (BCP)-based TRI Reagent procedure (Thermo Fisher Scientific, San Jose, CA, USA [36]). This process was followed by the reverse transcription using a SuperScript ® IV Reverse Transcriptase Kit (Thermo Fisher Scientific, San Jose, CA, USA). The mRNA level was subsequently quantified using a StepOne™ Real-Time PCR System (Thermo Fisher Scientific, San Jose, CA, USA).
Statistical Analysis
In this study, data were obtained from three separate experiments, and are presented as the mean ± standard deviation (n = 9). To compare the results from different operating conditions, we used one-way ANOVA and Tukey's honestly significant difference (HSD) post-hoc test for the statistical analysis.
Effect of the Electric Conditions on ODEP-Based Cell Manipulation
In this study, the effect of the ODEP electrical conditions (e.g., magnitude of AC electric voltage: 7-10 Vpp, and frequency of AC electric voltage: 1-5 MHz) on the cell manipulation of the cancer cells (e.g., PC-3 and SK-BR-3 cancer cells) tested were experimentally evaluated. Figure 2 reveals the quantitative relationship between the maximum velocity of a moving light bar that can manipulate the cancer cells tested (thus, the ODEP manipulation force of the cancer cells; Equation (5)) and the frequency of electric voltage applied under different voltage magnitude conditions. The results (Figure 2a,b) showed that the highest values of the maximum velocity of a light image that can manipulate the cancer cells, both occurred at 10 Vpp and 1 MHz for the two cancer cell types explored (i.e., 475.0 ± 25.0 µm s −1 and 458.3 ± 38.2 µm s −1 for the PC-3 and SK-BR-3 cancer cells, respectively). When the voltage frequency was lower than 1 MHz, some undesirable phenomena such as cell adhesion (10 Vpp and 500 kHz) (Figure 2c), cell aggregation (10 Vpp and 750 kHz) (Figure 2d), and cell lysis (10 Vpp and 100 kHz) (Figure 2e), would occur that could significantly affect the cellular properties [28]. Within the experimental conditions explored, moreover, the appearance rate of the above-mentioned phenomena was near 100% when the operating conditions were set at the particular electric conditions as indicated. These resulting phenomena might be due to the enhancement of electrically induced cell deformability, mutual DEP, and electrical lysis of cells at a voltage frequency near or below 1 MHz [26]. When the electric condition was set at 7 Vpp and 5 MHz (Figure 2a,b), conversely, we observed that the maximum velocities of a light image that can manipulate the cancer cells tested both reached their lowest levels for the two cancer cell types tested (i.e., 3.3 ± 5.8 µm s −1 and 10.0 ± 10.0 µm s −1 for the PC-3 and SK-BR-3 cancer cells, respectively) within the experimental conditions investigated. When the voltage magnitude and frequency were lower and higher than 7 Vpp and 5 MHz, respectively, the above-mentioned velocities were further decreased, which might significantly affect cell manipulation using the ODEP mechanism. Based on the facts described above, the optimal window of electric conditions for effective cell manipulation using ODEP was set at 7-10 Vpp and 1-5 MHz for the voltage magnitude and frequency, respectively. The ODEP operating conditions within such ranges were also reported for the ODEP-based cell manipulation for different applications [7,8,27,28]. Within this electric condition range, moreover, the maximum velocity of a light image that could manipulate the cancer cells tested decreased significantly with an increase of voltage frequency (Figure 2a,b), which was also consistent with previous findings [28]. Furthermore, under a given voltage frequency condition, the increase in voltage magnitude was found to increase the maximum velocity of a light image that could manipulate the cancer cells tested. This phenomenon could be explained by Equation (1), in which the ODEP force generated on a cell is proportional to the electric field squared. In addition to the values of the maximum velocity of a light image (and thus the ODEP force) investigated in the above-mentioned studies, it might also be valuable to explore the values of electric-field distribution in ODEP-based cell manipulation.
Biosensors 2020, 10, x FOR PEER REVIEW 7 of 14 and 5 MHz (Figure 2a,b), conversely, we observed that the maximum velocities of a light image that can manipulate the cancer cells tested both reached their lowest levels for the two cancer cell types tested (i.e., 3.3 ± 5.8 μm s −1 and 10.0 ± 10.0 μm s −1 for the PC-3 and SK-BR-3 cancer cells, respectively) within the experimental conditions investigated. When the voltage magnitude and frequency were lower and higher than 7 Vpp and 5 MHz, respectively, the above-mentioned velocities were further decreased, which might significantly affect cell manipulation using the ODEP mechanism. Based on the facts described above, the optimal window of electric conditions for effective cell manipulation using ODEP was set at 7-10 Vpp and 1-5 MHz for the voltage magnitude and frequency, respectively. The ODEP operating conditions within such ranges were also reported for the ODEP-based cell manipulation for different applications [7,8,27,28]. Within this electric condition range, moreover, the maximum velocity of a light image that could manipulate the cancer cells tested decreased significantly with an increase of voltage frequency (Figure 2a,b), which was also consistent with previous findings [28]. Furthermore, under a given voltage frequency condition, the increase in voltage magnitude was found to increase the maximum velocity of a light image that could manipulate the cancer cells tested. This phenomenon could be explained by Equation (1), in which the ODEP force generated on a cell is proportional to the electric field squared. In addition to the values of the maximum velocity of a light image (and thus the ODEP force) investigated in the above-mentioned studies, it might also be valuable to explore the values of electric-field distribution in ODEP-based cell manipulation.
Effect of the ODEP Field on the Cellular Viability of Cancer Cells
In this study, PC-3 and SK-BR-3 cancer cells, two of the commonly used cancer cell lines in the cancer-related studies, were used as model cells to investigate the effect of the ODEP-based cell manipulation on cellular properties. In the operations, the prepared cancer cell suspension was loaded into the microchamber of the ODEP microfluidic chip (Figure 1a), followed by exposure to the ODEP field with varied electric conditions (i.e., 7-10 Vpp and 1-5 MHz for the voltage magnitude and frequency, respectively), as determined previously (Figure 2). In our preliminary tests, the impact of ODEP exposure time on the cell viability of PC-3 cancer cells was investigated. Results (Supplementary Figure S2) showed that the cell viability of such cells had no significant difference (p > 0.05) compared to the control (i.e., the cancer cells without exposure to the ODEP field) within the 5 min exposure time, when the voltage magnitude and frequency were set at 10 Vpp and 5 MHz, respectively. When the ODEP exposure time was as long as 10 and 15 min, conversely, the cell viability of the PC-3 cancer cells might be significantly affected (p < 0.05) (e.g., 43.7 ± 5.5% and 51.3 ± 10.3% reduction for the ODEP exposure time of 10 and 15 min, respectively) in comparison to the control. Based on the preliminary results, therefore, the ODEP exposure time was set at 3 min, and within the time period, general cell manipulation using the ODEP mechanism was normally carried out (e.g., the rapid separation of Raji cells [7] or isolation and purification of CTCs [5,9]). In practice, 40 rectangular light bar images were illuminated on the microchamber zone. During the period of ODEP exposure, the cancer cells tested were attracted within the light bars, as shown in Figure 3(aII,aV). After the exposure to the ODEP field, we observed microscopically that the treated cancer cells within the microchamber were aligned in accordance with the original light bar images, as exhibited in Figure 3 (aIII,aVI). Then, the ODEP-treated cancer cells were collected for the following bioassays.
Biosensors 2020, 10, x FOR PEER REVIEW 9 of 14 significantly downregulated (p < 0.05) (the observed cell viability: 49.4 ± 9.5%) when the magnitude of the voltage was higher than 10 Vpp. Overall, the lower cell viability of cancer cells found at the higher voltage condition could be explained by the fact that the electrical lysis of cells is prone to occur when a cell is exposed to a high-voltage magnitude [37,38]. Taken together, the increase of voltage magnitude or decrease of voltage frequency, respectively, could raise the maximum velocity of a light image that can manipulate cells (i.e., the ODEP manipulation force of cells; Equation (5), Figure 2). Nevertheless, the increase in voltage magnitude or decrease in voltage frequency, respectively, could accordingly increase the tendency of electrical lysis of cells, which could affect cellular viability. Within the experimental conditions explored (i.e., 7-10 Vpp and 1-5 MHz), the determinant electrical condition that could significantly affect cell viability was a low-voltage frequency of 1 MHz, when the voltage magnitude range was 7-10 Vpp.
Effect of the ODEP Field on the Cellular Metabolic Activity and Gene Expression of the Cancer Cells Tested
As suggested by the previous results (Figure 3b), cell manipulation using the ODEP mechanism at a low voltage frequency of 1 MHz could be lethal to the cancer cells tested. Such an ODEP electrical condition was ruled out in subsequent studies. In the following evaluations, the In this study, the Cell Counting Kit-8 (CCK-8) was utilized to investigate the viability of cancer cells treated with various ODEP fields. The results (Figure 3(bI)) revealed that the cell viability of PC-3 cancer cells had no significant difference (p > 0.05) compared to that of the control cells (i.e., the cancer cells without exposure to the ODEP field), when the voltage magnitude and frequency were set at 7-10 Vpp and 3-5 MHz, respectively. A similar result was also found for the SK-BR-3 cancer cells tested (Figure 3(bII)). However, once the voltage frequency was set at 1 MHz, the cell viability of the PC-3 and SK-BR-3 cancer cells decreased significantly (p < 0.05) compared to that of the control, irrespective of a voltage magnitude of 7, 8.5, or 10 Vpp. This finding could be explained as follows. If the equivalent circuit of a biological cell is assumed to be a lumped circuit, the electrical property of the cell membrane will be similar to the electric capacitance [28]. When the voltage frequency is altered from high to low frequency (e.g., from 10 MHz to 100 kHz), the electrical impedance of a cell membrane might increase. This phenomenon could, therefore, result in the shift of the exerted electric voltage from the cell cytoplasm to the cell membrane of a cell [28]. This fact might in turn lead to an increase in the transmembrane potential in a cell membrane when the voltage frequency is changed from high to low frequency. The above-mentioned phenomena could explain the low cell viability caused by cell electrical lysis (Figure 2e) when the voltage frequency is set at low conditions (e.g., 1 MHz) (Figure 3b).
Under the low voltage frequency condition of 1 MHz (Figure 3(bI)), moreover, the cell viability of the PC-3 cancer cells significantly decreased (p < 0.05) when the magnitude of voltage was higher than 8.5 Vpp (observed cell viability: 40.4 ± 3.6%). A similar result was also found for the SK-BR-3 cancer cells tested (Figure 3(bII)), in which the cell viability of the SK-BR-3 cancer cells was significantly downregulated (p < 0.05) (the observed cell viability: 49.4 ± 9.5%) when the magnitude of the voltage was higher than 10 Vpp. Overall, the lower cell viability of cancer cells found at the higher voltage condition could be explained by the fact that the electrical lysis of cells is prone to occur when a cell is exposed to a high-voltage magnitude [37,38]. Taken together, the increase of voltage magnitude or decrease of voltage frequency, respectively, could raise the maximum velocity of a light image that can manipulate cells (i.e., the ODEP manipulation force of cells; Equation (5), Figure 2). Nevertheless, the increase in voltage magnitude or decrease in voltage frequency, respectively, could accordingly increase the tendency of electrical lysis of cells, which could affect cellular viability. Within the experimental conditions explored (i.e., 7-10 Vpp and 1-5 MHz), the determinant electrical condition that could significantly affect cell viability was a low-voltage frequency of 1 MHz, when the voltage magnitude range was 7-10 Vpp.
Effect of the ODEP Field on the Cellular Metabolic Activity and Gene Expression of the Cancer Cells Tested
As suggested by the previous results (Figure 3b), cell manipulation using the ODEP mechanism at a low voltage frequency of 1 MHz could be lethal to the cancer cells tested. Such an ODEP electrical condition was ruled out in subsequent studies. In the following evaluations, the effect of the ODEP fields with varied electrical conditions (i.e., voltage magnitude and frequency: 7-10 Vpp and 3-5 MHz, respectively) on the metabolic activity and gene expression of cancer cells were explored. For the former, the ATP synthesis of cells was used as an indicator based on the previous studies [39]. The results (Figure 4a) showed that the ATP level of the ODEP-treated PC-3 cancer cells showed no significant difference (p > 0.05) compared to that of the control (i.e., the cancer cells without exposure to the ODEP field), within the experimental conditions explored. A similar result was also observed for the SK-BR-3 cancer cells (Figure 4b). The findings in this study could indicate that the ODEP field could not affect the ATP synthesis of a cell within the electric conditions tested. This result was, to some extent, similar to that found in a previous DEP-based study [40].
cancer cells without exposure to the ODEP field), within the experimental conditions explored. A similar result was also observed for the SK-BR-3 cancer cells (Figure 4b). The findings in this study could indicate that the ODEP field could not affect the ATP synthesis of a cell within the electric conditions tested. This result was, to some extent, similar to that found in a previous DEP-based study [40]. In addition to the effect on cellular metabolic activity, the ODEP effect on the gene expression of cancer cells was investigated. In this study, the gene expression of EMT-associated genes (e.g., EpCAM and VIM) [8] and the MRP1 gene (e.g., ABCC1) [8] in the PC-3 and SK-BR-3 cancer cells was quantified and then normalized to the gene expression of GAPDH. The results (Figure 5a) showed that the gene expression of the above-mentioned genes in the PC-3 cancer cells was not significantly different (p > 0.05) from that of the control cells (i.e., the cancer cells without exposure to the ODEP field) within the ODEP operating conditions (i.e., 7-10 Vpp, 3-5 MHz, and 3 min exposure time) studied. A similar outcome (Figure 5b) was also found for the SK-BR-3 cancer cells. In addition to the effect on cellular metabolic activity, the ODEP effect on the gene expression of cancer cells was investigated. In this study, the gene expression of EMT-associated genes (e.g., EpCAM and VIM) [8] and the MRP1 gene (e.g., ABCC1) [8] in the PC-3 and SK-BR-3 cancer cells was quantified and then normalized to the gene expression of GAPDH. The results (Figure 5a) showed that the gene expression of the above-mentioned genes in the PC-3 cancer cells was not significantly different (p > 0.05) from that of the control cells (i.e., the cancer cells without exposure to the ODEP field) within the ODEP operating conditions (i.e., 7-10 Vpp, 3-5 MHz, and 3 min exposure time) studied. A similar outcome (Figure 5b) was also found for the SK-BR-3 cancer cells.
cancer cells without exposure to the ODEP field), within the experimental conditions explored. A similar result was also observed for the SK-BR-3 cancer cells (Figure 4b). The findings in this study could indicate that the ODEP field could not affect the ATP synthesis of a cell within the electric conditions tested. This result was, to some extent, similar to that found in a previous DEP-based study [40]. In addition to the effect on cellular metabolic activity, the ODEP effect on the gene expression of cancer cells was investigated. In this study, the gene expression of EMT-associated genes (e.g., EpCAM and VIM) [8] and the MRP1 gene (e.g., ABCC1) [8] in the PC-3 and SK-BR-3 cancer cells was quantified and then normalized to the gene expression of GAPDH. The results (Figure 5a) showed that the gene expression of the above-mentioned genes in the PC-3 cancer cells was not significantly different (p > 0.05) from that of the control cells (i.e., the cancer cells without exposure to the ODEP field) within the ODEP operating conditions (i.e., 7-10 Vpp, 3-5 MHz, and 3 min exposure time) studied. A similar outcome (Figure 5b) was also found for the SK-BR-3 cancer cells. Previous studies similar to this work revealed that the exposure to a DEP field (e.g., 20 Vpp, 250 kHz, and 1 h exposure time) could downregulate the expression of cell differentiation-related genes (e.g., ADGRE5/CD97, RUNX2, and NES) in mesenchymal stem cells (e.g., UE7T-13 cells) [41]. It was also reported that the exposure to a DEP field (e.g., 21 Vpp, 5 MHz, and 15 min exposure time) might not affect the cell morphology, cell oxidative respiration rate, and cell cycle dynamics of fibroblast-like cells (e.g., BHK-21/C13 cells) [40]. After exposure to the DEP field, however, the gene expression of the c-fos protein (commonly activated by environmental stress [42]) was observed to be upregulated [40]. The reasons behind the discrepancies found in the previous and current studies might be complicated and could be related to the cell species, electrical conditions, and the genes explored. Further experiments will be required to systematically investigate this issue. In addition to the gene expressions investigated in this study, it might also be valuable to explore the expression profiles of the genes regulating the signal pathways downstream of voltage-gated ion channels.
Conclusions
The ODEP-based microsystem provides a powerful tool for biological cell-relevant studies or applications. Before its widespread application, one fundamental issue has to be addressed-Whether cell manipulation using the ODEP mechanism influences the native properties of cells. To address this issue, we carried out various experiments. In this study, the impact of the ODEP electrical conditions ((1) the magnitude of AC voltage: 5-10 V; (2) frequency of AC voltage: 1-5 MHz), (3) operating time: 3 min) on the properties (e.g., cell viability, metabolic activity of cells, or gene expression of cells) of biological cells (e.g., PC-3 and SK-BR-3 cancer cells) was explored. Within the experimental conditions studied, the ODEP-based cell manipulation with the highest performance (e.g., highest cell manipulation velocity) occurred at 10 Vpp and 1 MHz for the two cancer cell types explored. Under this ODEP electrical condition, however, the cell viability of the cancer cells tested was significantly affected (e.g., 70.5 ± 10.0% and 50.6 ± 9.2% reduction for the PC-3 and SK-BR-3 cancer cells, respectively) possibly due to cell electrical lysis. Conversely, the exposure of cancer cells to the ODEP field with a particular condition range (i.e., 7-10 Vpp, 3-5 MHz, and 3 min exposure time) might not significantly alter the cell viability, cell metabolic activity, and the gene expression of cancer cells. Within this ODEP operating condition, the highest maximum velocity (i.e., 106.7 ± 30.6 µm s −1 and 100.0 ± 20.0 µm s −1 , for the PC-3 and SK-BR-3 cancer cells, respectively) of a light image that could manipulate cells occurred at the voltage magnitude and frequency of 10 Vpp and 3 MHz, respectively. The above-mentioned cell manipulation velocity ranges were technically sufficient for various applications. Overall, this study fundamentally investigated the effect of ODEP electrical conditions on the cellular properties of cancer cells. The information obtained is crucially important for the utilization of ODEP-based cell manipulation in a microscale system for various applications.
Conflicts of Interest:
The authors declare no conflict of interest. | 2020-06-18T09:07:04.393Z | 2020-06-01T00:00:00.000 | {
"year": 2020,
"sha1": "5ecbe9b669e3d7933353fa08c3a8da6b7b55765a",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2079-6374/10/6/65/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "04898fb0c6d8d337bd94c796ba80fa49ffd68f10",
"s2fieldsofstudy": [
"Biology"
],
"extfieldsofstudy": [
"Medicine",
"Chemistry"
]
} |
56126002 | pes2o/s2orc | v3-fos-license | Metals releases and disinfection byproduct formation in domestic wells following shock chlorination
Introduction Conclusions References
Introduction
Shock chlorination is an in-situ method for disinfecting water wells contaminated with pathogens and nuisance bacteria.Much guidance is available for treatment of domestic wells (e.g.Schnieders, 2005;Driscoll, 1986) and http://www.unce.unr.edu/publications/files/nr/2006/FS0668.pdf, last accessed April 2010).The guidelines recommend CT values (free chlorine concentrations × resting time of the solution) that are very high relative to those used for public water supply treatment, which is appropriate given that treatments occur sporadically, usually in response to perceived problems with water or the health of those who consume water from a well.The procedure typically involves Table 1.Well characteristics and water physical and chemical characteristics immediately prior to shock chlorination; DTW is depth to water from the land surface, WD is total depth of the well, Static vol.refers to the standing volume of water in a well after water level recovery following the end of pumping.The measurements reported are also referred to as IP (Initial Purge) in Table 4. Initial Purge refers to purging prior to treatment, to replace water stored in the casing with water from the surrounding saturated formations.Well diameter, case material, depth to water from the land surface and well depth were recorded at each site.The screened interval length was obtained from well logs available on the Nevada Division of Water Resources (http://water.nv.gov/, last accessed April 2010).
US fish and wildlife
Well The odor threshold for chlorine gas in air is approximately 0.3 mg L −1 (Amoore and Hautala, 1983).The relationship between the amount of chlorine degassing from a solution and the amount that can be detected by smell varies, depending upon the individual sensitivity and the degassing rate, which is in turn related to mixing dynamics within a water body, contact time, temperature of water, and changes in temperature and barometric pressure.Because of these several factors, guidelines about purging based on scent are subjective and the consistency and efficacy of application may vary considerably between people.
Shock chlorination may also increase the concentrations of lead and other trace elements following treatment (Seiler, 2006) and change arsenic concentrations (Gotkowitz et al., 2008).Both lead and arsenic have toxicological effects, with an Action Level and Maximum Contaminant Level of <15 µg L −1 and 10 µg L −1 , respectively.
This paper describes the changes in concentration of Pb, Cu, As, U, gross-α and gross-β radiation, HAA5, THM, and free chlorine (semi-quantitatively measured) from shock chlorination of four domestic water wells.The study also sought to demonstrate that simple test strips for semiquantitative measurement of concentration of free chlorine (used for pool and spa maintenance) indicates when purging has returned concentrations of metals and disinfection byproducts to pre-treatment background levels.Test strips may be a better indicator than the scent of water that posttreatment purging is complete, because they provide a semiquantitative indicator of total chlorine concentrations.
Study site
Four wells (Table 1) were selected in the Lahontan Valley, in Nevada (Fig. 1).Each well was used for domestic supply prior to being retired for water right acquisition by the US Fish and Wildlife Service within the Stillwater National Wildlife Refuge.Three were cased with steel (ASTM A135 SCH40 ERW low-carbon steel, based on inspections at the sites) and one was cased with polyvinyl chloride casing (specifications unknown).These wells were chosen for two reasons.First, they conformed to standards and practices commonly used for domestic water well construction and had well logs available through the Nevada State Engineer's office.Second, because they were no longer in service we did not run the risk of exposing people to metals or disinfection by-products released or created by shock chlorination.Well logs for each indicated that the exteriors of screened intervals on the casing were packed with gravel for wells 182 (PVC-cased), 167N (steel-cased) and 142 (steel-cased).The construction log for well 51 (steel-cased) contained incomplete information about how the screened interval was finished.The wells pumped water from a stratum of Quaternary valley-fill sands in the inland terminus of the Carson River from depths of less than 15 m (50 ft) from the land surface.Infiltration from irrigation and the Carson River has been identified as the main source of recharge (Glancy, 1986).Although the groundwater system in the Lahontan Valley is nominally comprised of three geochemically separated systems (shallow (<15.2 m (50 ft) from the land surface); intermediate (15.2-<152.4m (50-<152.4ft); deep (≥152.4m (≥500 ft)), aquifer material composition, yields and chemistry vary highly throughout the region (Glancy, 1986).Water in the aquifer has high but spatially variable concentrations of arsenic (as much as 2100 mg L −1 (Walker et al., 2005)) and uranium (as much as 290 mg L −1 (http://www.atsdr.cdc.gov/HAC/PHA/fallonleukemia2/fln p1.html, last accessed April 2010) from contact with sediments and from long-term evapoconcentration (Welch and Lico, 1998).Depth to water (DTW) in the four wells ranged from 1.7 to 5.5 m (5.5 to 18.0 ft) with total well depth ranging from 8.0 to 9.5 m (26.7 to 31.0 ft) (Table 1).
Well treatment and purging
A 1 / 2 horsepower portable jet pump fitted with a 7.6 m (25 ft) long, 1.9 cm (0.75 in) interior diameter suction line was used for each trial.The pump and suction line were rinsed with distilled, deionized water between uses and allowed to air dry.Each well was chlorinated and purged as a separate experiment, to avoid cross-contamination of wells.The pump had a fiberglass-reinforced thermoplastic housing and impeller, with Buna-N seals and ceramic bearings.The outflow line was fitted with a GPI electronic inline flow meter, a flow control valve, and a tee that divided flow between a discharge hose and a flow-through chamber for www.drink-water-eng-sci.net/4/1/2011/ Drink.Water Eng.Sci., 4, 1-8, 2011 a YSI model 556MPS (Yellow Springs Instruments, Yellow Springs OH), for real-time measurement of temperature (T ( • C)), pH, oxidation-reduction potential (ORP (mV)), and conductivity (C (mS cm −1 )).The multi-probe was calibrated immediately prior to each stage of field trials, using pH 4.00, 7.00 and 10.00 standards (Fisher Scientific Buffer-Pac, Cat# SB105), a conductivity standard (1000 mS cm −1 at 25 • C (Yellow Springs Instrument Co. Cat #3167) and an oxidation-reduction potential reference solution (Equipco Inc., part #3682500).
Prior to conducting trials, each well was purged of stagnant water at a rate of 9.5 to 18.9 liters per minute (2.5 to 5.0 gpm) until temperature, pH, ORP, and conductivity readings stabilized (<5% variation in continuous readings), then purged an additional four times the well's standing volume of water.The standing well volume (WV) was estimated as WV = H × A, with H as the measured height of the water column in the well after the level stabilized following pumping and A as the cross sectional area of the interior of the well casing.Using this as a basis for purging following treatment differs from the dynamic well volume, which represents the steady state water level achieved during pumping.This differs from the static water level due to drawdown during pumping.The water levels used as a reference in this paper are standing well volumes (WV), which do not reflect the influence of pumping.Pre-chlorination (designated as IP) water samples for all wells were collected for As, Cu, Pb, U, gross-α and gross-β radiation, total organic carbon, and for a steel-cased well and a PVC-cased well, water samples were collected for HAA5 and THM analyses.Table 1 presents IP conditions in each well.
Shock chlorination
Following initial purging, each well was chlorinated to an estimated 200 mg L −1 as total Cl using household bleach (unscented, regular strength Clorox ® labeled as containing 6% sodium hypochlorite (NaOCl), which is 64.6% Cl by mass) by adding 8.9 ml (0.3 fl oz) bleach per liter (0.26 gallons) of well volume (Table 1).Pump discharge was circulated back into the well casing for 15 min to disperse bleach into solu- tion.Wells were re-capped for a resting period of at least twelve hours.
Sampling
Immediately prior to and during purging temperature, pH, oxidation-reduction potential, and conductivity were measured.Immediately prior to purging post-chlorination (PC) samples from all wells were tested for As, Cu, Pb, U, and gross-α and gross-β radiation.Samples from two wells (142 and 182) were tested for disinfection byproducts (HAA5 and THM).All metals and disinfection by-products samples were collected using a PTFE dip bailer.
Post chlorination purging samples were collected for As, Cu, and Pb at intervals defined by the volume of water purged from each well, including 1 / 2 , 1, 2, 3, and 4× WV (designated as 1/2, 1, 2, 3, 4).After 4 WV, samples were collected for As, Cu, Pb, U, gross-α and gross-β radiation, HAA5 and THM (Table 2).All samples were unfiltered.Samples were collected, immediately placed in a portable cooler with blue ice to avoid exposure to sunlight and changes in temperature and submitted within 24 h to the Nevada State Health Laboratory (University of Nevada School of Medicine, Reno, NV -a certified drinking water analysis laboratory) for analysis (Table 3).Samples for arsenic, copper, lead and uranium were collected in 500 ml high density polyethylene bottles with 5.0 ml of 15% nitric acid as a preservative in containers provided by the laboratory.After sample collection, the final concentration of nitric acid was 0.15% nitric acid.
Chlorine test strips
Free chlorine was measured semi-quantitatively with test strips for swimming pool and spa maintenance (Arch Chemicals, Inc. -HTH line).The test strips indicated free chlorine concentrations in ranges rather than absolute numbers, similar to pH indicator strips.In order to determine the range, a user dips the strip in a solution and compares colors appearing in segregated rectangles with a key.The ranges reported include undetectable (indicated as 0 on the test strip), >0-1 mg L −1 , >1-2 mg L −1 , >2-3 mg L −1 , >3-5 mg L −1 , >5-10 mg L −1 and >10 mg L −1 .We tested the ac-curacy of the strips using dilutions of sodium hypochlorite solution and determined that they were adequate for distinguishing between the classes noted above (data not shown).
Conductivity, pH, oxidation-reduction potential and temperature
Post-chlorination conductivity measurements were elevated from initial values, and returned to near initial values following purging 4 WV (Table 4).pH rose in wells 167N and 51 and decreased slightly in wells 182 and 42.Increases in pH conform to observations that concentrations of chlorine of 200 mg L −1 can be expected to increase pH by up to two units (Schnieders, 2005).Oxidation-reduction potential increased above pre-chlorination levels in all wells, as would be expected with the addition of an oxidizer.Water in well 142 returned to background levels for temperature, pH, conductivity, and oxidation-reduction potential after four WV had been purged, though measurements from the other wells indicated that 4 WV of purging was not sufficient to return to background conditions ( levels, though temperature and pH were <16% of starting values.This suggests that the oxidizing effects of chlorination led to short-term changes in the immediate vicinity of the well screen or that were stagnant zones within the well, which would have required more purging to eliminate.
Chlorine concentration
Decreases of free chlorine concentration were hypothesized to be an indicator of purging of mobilized trace metals and disinfection byproducts.The concentration of free chlorine decreased in wells 51, 167N and 142 to >1-2 ppm free chlorine after 4 WV were pumped (Table 5).Site 182 required that 5 WV be purged before free chlorine decreased to >3-5 ppm (data not shown).
Mobilization of trace metals
Concentrations of lead and copper in well water increased following shock chlorination, which corresponds with results reported by Seiler (2006).Lead concentrations increased up to thirteen-fold and copper concentrations increased up to four-fold following treatment.Concentrations of both decreased to initial levels within 2 WV of purging.The return to background levels corresponded with the decline in free chlorine to >3-5 mg L −1 (Table 5).
All wells contained arsenic in concentrations that exceeded the MCL (0.010 mg L −1 ) prior to treatment.The decline in arsenic concentration in the chlorine solution prior to purging was similar to results reported by Gotkowitz et al. (2008) and would be expected with the observed changes pH and oxidation-reduction potential.However, the decline was followed by an increase in arsenic above background levels during pumping.After 4 WV, water from one well returned to initial arsenic concentrations, while the others remained 3-12% higher than initial concentrations.
Uranium and radionuclides
Samples were analyzed for uranium and gross-α and gross-β radionuclides (Table 6).Uranium concentrations increased at sites 51 and 142 but remained the same or decreased in wells 167N and 187.Concentrations returned to approximately the same as starting levels in all wells after purging 4 Drink.Water Eng.Sci., 4, 1-8, 2011 www.drink-water-eng-sci.net/4/1/2011/ WV.Gross-α concentrations changed from the IP to PC sampling steps, increasing in wells 51 and 167N, decreasing in well 142 and remaining approximately the same in well 182.Gross-β concentrations remained approximately the same in all wells at each sampling stage, with concentrations appearing to decrease slightly between the IP and 4 WV samplings.
The maximum change in gross-β concentrations was approximately −9% in well 142.
Total organic carbon and disinfection byproducts
Total organic carbon (TOC) concentrations in wells 182 and 142, collected prior to treatment, are shown in Table 7. Water from wells obtained after treatment, but prior to purging contained concentrations of THM up to ten times the MCL (Table 7).Following purging of 4 WV, concentrations in well 142 decreased to below the detection limit, indicating that the increase in concentration was temporary and could be remediated by purging 4 WV after treatment.
Well 182 retained disinfection byproducts and free chlorine, with the concentration of free chlorine (indicated by test strips) >10 mg L −1 after purging 4 WV.Testing for free chlorine at the fifth well volume purged indicated a concentration of >3-5 ppm.
Conclusions
This study confirmed temporary mobilization and changes in concentration of trace metals from shock chlorination treatment, as previously reported by Seiler (2006) and Gotkowitz et al. (2008).It also demonstrated that disinfection byproducts can be formed and can persist beyond 3 WV of purging.The concentration of disinfection byproducts exceeded drinking water MCLs in the two wells tested.In one well, concentrations decreased to below the detection limits for HAA5 and THM with purging.In the other, concentrations of disinfection byproducts remained elevated, with final concentrations of THM and HAA5 at highest levels after purging 4 WV.Metals, though mobilized by shock chlorination, decreased to near pre-treatment background levels after 4 WV were purged in all wells.
The concentrations of DBP precursors and peak concentrations of HAA5 and THM differed in the two wells sampled.Although concentrations of THM and HAA5 were lower in well 182, samples collected after 4 WV indicated that DBPs persisted.This may have been due to differences in well construction and interaction between chlorine solutions and aquifer materials.With respect to well construction, the depths of wells and depths to water from the land surface for wells 182 and 142 were similar, but lengths of screened intervals and static volumes differed (Table 1).The screened interval in well 182 spanned 3.3 m (11 ft), compared with the screened interval of 1.5 m (5 ft) in well 142.Well 182 also had approximately 1.3 m of casing below the screened interval, which may have served as a reservoir for residual chemicals because volume exchange in this portion of the casing would be difficult to accomplish with a surface pump.The stagnant portion of well 182 may also have contained sediments, which would have further prolonged release of HAA5 and THM into pumped water.In contrast, screens in the other wells were mounted at the bottom of the casing.Although this does not assure that complete mixing took place throughout post-treatment purging, it is more likely that purging led to complete exchanges of water within the well casing than in well 182.It is also possible that disinfection byproducts formed within the saturated formations immediately adjacent to the well screen.
Although well logs indicated that the screened intervals were set in sands, well yields appeared to be very different based on the discharge rates used for purging and drawdowns observed.In order to maintain a steady discharge rate from well 142, the pumping rate was <20% of the purging rate used in well 182, even though lift distances from the water table to the surface were slightly less for well 142 than well 182.This suggests that the saturated formations in the screened intervals pumped for well 142 were less permeable than those in well 182, which in turn suggests that aquifer materials may have been composed of fine-textured soils including silts and clays.In the absence of significant advective mixing, formation of disinfection byproducts beyond the immediate well volume would be limited to chemical dispersion.In well 182, however, it is likely that a larger volume of aquifer material was exposed to chlorine.This hypothesis is, in part, supported by the persistent indication of oxidizing conditions in water withdrawn after purging 4 WV.A prior study in the region (Fram et al., 2005) demonstrated that free chlorine released significant amounts of carbon from aquifer materials.They concluded that carbon was likely released from clay minerals in proportion to the amount of available chlorine in solution.This suggests that aquifer material and interactions with high concentrations of available chlorine were a potential source of precursors to HAA5 and THM formation.
The mixing procedure used to disperse chlorine throughout the entire well volume was unlikely to have created a gradient that would advect chlorine into aquifer materials, but chemical diffusion through the well screen may have taken place during the resting phase.Given the differences in screened intervals, this may have led to accumulation of DBPS's in aquifer materials in contact with the screened interval of well 182, which could have required more thorough purging to remove than those in contact with the shorter screened interval in well 142.Although purging appeared to be adequate based on number of well volumes removed and stability of temperature and pH, test strips indicated the presence of free chlorine.Consequently, the chemical reactions that led to metals releases and DBP formation may have been due to chemical diffusion during the resting time following introduction of bleach and likely due to reactions with aquifer materials and gravel packing in the immediate vicinity of the screened interval of the well casing.
Publications about shock chlorination recommend purging varying numbers of standing well volumes post treatment, before returning a well to service.Guidance also is based on detecting the scent of chlorine in water.Without metering equipment most domestic well owners have no accurate means of determining when pumping is sufficient to remove a specific number of well volumes.Also, determining the sufficiency of purging by scent is subjective and may not be consistent, especially with respect to avoiding exposure to metals and disinfection byproducts.The use of chlorine test strips shows promise as a simple and accurate means of de-termining when purging is complete, though this should be verified with further experimental work.This technique has the potential to be a reliable guideline for public health officials and informal educators (including county-based educators of Cooperative Extension), who provide advice about proper procedures for shock chlorination, especially when well owners cannot measure discharge rates and discharge volumes.
Figure 1 .
Figure 1.Location of study area, with locations of test wells indicated in circles.
Table 2 .
Sampling intervals used, with associated chemical constituents measured.
Table 3 .
Methods used for sample analysis."EPA"refersto a standard analytic method published by the US Environmental Protection Agency (available at http://www.epa.gov/sam/index.htm,lastaccessedApril 2010), used by the Nevada State Health Laboratory."SM" refers to standard method 5310 C (Persulfate-ultraviolet or Heated Persulfate Oxidation Method), published in Standard Methods for the Examination of Examination of Water and Wastewater(Clesceri et al., 1998).Total Organic Carbon samples were analyzed using a Shimadzu Spectrophotometer at the University of Nevada.
* The analytic detection limit is based on linear regression analysis of the calibration curve, conducted with three replicates each of 4 standards, ranging from 0.274-20.270mg L −1 .
Table 4 .
Physical and chemical characteristics of water in test wells observed during post-chlorination purging (T : temperature, C: conductivity, ORP: oxidation-reduction potential, ND: not measured).Final column displays the percentage difference between values at IP (prior to chlorination) and post-chlorination, with 4 well volumes purged.
Table 5 .
Trace metal concentrations from pre-treatment (PT) through purging, expressed in number of well volumes pumped.The Maximum Contaminant Level for arsenic is 0.010 mg L −1 .Action Levels for lead and copper are 1.3 and 0.015 mg L −1 , respectively.Samples that contained analytes in concentrations less than the reporting limit (RL) are reported as "<RL."The reporting limit for lead varied according to results of internal laboratory quality control assessments and chemical quality of water samples.
Table 7 .
Total Organic Carbon (TOC) concentrations following initial purging of test wells (IP), and HAA5 and THM concentrations following IP, post chlorination (PC) and after 4 well volumes (WV) had been purged. | 2018-10-21T15:14:34.809Z | 2010-06-04T00:00:00.000 | {
"year": 2010,
"sha1": "f9d7631c108eefebd0567480959254b6cd224f29",
"oa_license": "CCBY",
"oa_url": "https://dwes.copernicus.org/articles/4/1/2011/dwes-4-1-2011.pdf",
"oa_status": "GOLD",
"pdf_src": "ScienceParseMerged",
"pdf_hash": "0e7db00847d8775f623965027875252f3ffa48e9",
"s2fieldsofstudy": [
"Engineering"
],
"extfieldsofstudy": [
"Chemistry"
]
} |
130302558 | pes2o/s2orc | v3-fos-license | The Impact of Community-based Afro-alpine Tourism on Regional Development
Abstract Can existing Afro-alpine tourism promote poverty mitigation and resolve regional disparities? This article explores the significance of alpine tourism in the Mt Kenya region based on analysis of the state of the art and official statistical data along with own surveys, mapping activities, and household observations. The results show that economic benefits from mountaineering tourism in the Mt Kenya region are smaller than commonly calculated, and that low and inconsistent incomes are distributed unevenly. There are clear parallels to the critical situation in the Rwenzori Mountains in Uganda: Alpine tourism does not reduce regional income disparities and largely fails to promote sustainable development. The article also takes a closer look at the development effects of community-based tourism, drawing from the example of the Mt Kenya Guides and Porters Safari Club (GPSC), a community-based tourism organization operating from Naro Moru, at the fertile western foot of Mt Kenya. Results show that this form of tourism stabilizes the livelihoods of rural households, contributes to community welfare, and reduces the vulnerability of families. The GPSC's democratic organizational structure with elected and regularly rotating offices prevents the enrichment of only few members and ensures even distribution of benefits to all members and to the whole community. Overall, however, there is not enough tourism in the study area to initiate sustainable regional development in the foreseeable future.
Introduction and main hypothesis
For many years, tourism was considered a magic formula for promoting regional development and reducing poverty in developing countries (Ashley et al 2000;Harrison 2001;Mowforth and Munt 2003;UNWTO 2007;Telfer and Sharpley 2008). This applies in particular to the glaciated highlands of East Africa, which-in top-down processes-were all declared national parks over the past decades. The state national park authorities of Tanzania, Kenya, and Uganda attach extraordinary economic importance to Afro-alpine tourism in the scenically very attractive highest mountains of Africa: Kilimanjaro (5895 m), Mt Kenya (5199 m), and Rwenzori (5109 m) (see http://www.kws.org, http://www.tanzaniaparks.com, and http: //www.uwa.or.ug).
In contrast, the present contribution and numerous other studies (eg Sinclair 1998;Hall 2007;Zhao and Richie 2007;Saarinen et al 2009Saarinen et al , 2011Van Der Duim et al 2012) take a critical and differentiated approach. Analyses of the effects of alpine tourism on nature reserves in tropical countries (eg Vorlaufer 1995;Banskota and Sharma 1998;Mü ller-Bö ker 2000) as well as specialized studies of Afroalpine tourism in Kenya and Uganda (Osmaston et al 1998;Erhard 2000;Erhard and Steinicke 2006;Steinicke 2011;Steinicke and Neuburger 2012) typically show that, at best, high-mountain tourism benefits state organizations and big tour operators, whose headquarters are usually located in the respective capital. In comparison, the development impulses received by the destinations visited are insignificant: Both in the Mt Kenya region (Erhard 2000) and in the Rwenzori Mountains (Craddock Williams 1998;Steinicke 2011), alpine tourism offers only marginal and insecure monetary income. Despite this rather unfavorable balance, which is also observed in other tourist regions of the global South (Harrison 2001), in many areas the local population has tried to increase benefits to the community in organized fashion by distributing income from tourism directly to the families, avoiding external absorption of profits. This idea, termed ''communitybased tourism,'' has been adopted in development
Systems knowledge
Mountain Research and Development (MRD) An international, peer-reviewed open access journal published by the International Mountain Society (IMS) www.mrd-journal.org collaboration in recent years and has been used as a regional development strategy (Scheyvens 2002;Blackstock 2005;Okazaki 2008).
From the less than voluminous state of research on community-based tourism in the high mountains of tropical Africa, as well as observations based on it, the following main hypothesis can be deduced for the Mt Kenya region: Existing Afro-alpine tourism can neither promote poverty mitigation nor resolve regional disparities. However, strategies of community-based tourism stabilize the livelihoods of rural households, even though they are inadequate measures for advancing regional development.
This work is intended as a contribution toward more detailed analysis of community-based tourism in the tropical mountains of Africa, as previously provided for East Africa by Manyara et al (2006) and Manyara and Jones (2007). The following sections explore the significance of alpine tourism in the Mt Kenya region and present the example of the Mt Kenya Guides and Porters Safari Club (GPSC), a community-based organization in Naro Moru, at the semiarid, but fertile, western foot of Mt Kenya. Naro Moru Location is characterized by scattered settlements and, according to the 2009 population census, has a total population of 21,533; Kamburaini Sublocation, which is an important starting point for climbing the volcano, has 6414 inhabitants in 1813 households (KNBS 2010). Based on the above main hypothesis, the study was designed to answer the following questions for Naro Moru and, more specifically, for Kamburaini: N How does alpine tourism in general affect the regional economy at the foot of Mt Kenya? Does tourism in the Mt Kenya region act as a driver of regional development and resolve sociospatial disparities? N Does income from community-based tourism stabilize households' livelihoods? How substantial is this income for the members of the cooperative in Naro Moru? N Which aspects of community-based tourism-eg ''participation'' or ''empowerment''-are relevant to development and people's livelihoods? Can they promote both poverty mitigation and future prospects for the next generation?
Methods
Apart from evaluating the latest research and analyzing official statistics, this study involved the collection of primary data by a range of methods, including open, qualitative interviews and focus group discussions with guides and porters and the GPSC Steering Board; interviews with experts (District Officer, Chancellor and Chief of Location, Minister of Agriculture of Nyeri District, Chief of Sublocation, and hotel managers of Naro Moru lodges); analysis and evaluation of the GPSC's books and accounts; inspection of the visitors' books at four gates of the Mt Kenya National Park; and various mapping activities. The local population's points of view were recorded in a variety of interviews conducted by 9 students of a Master's program in Geography. The students were invited to stay with families in Naro Moru for several days and share in their daily routines, getting to know the families' livelihoods, their family structures and ties, as well as their working and living conditions. This intensive interviewing activity covered 27 households; interviews and focus group discussions with guides and porters provided deeper insights into another 48 households. Tourism, regional development, and poverty reduction: some comments on underlying theoretical concepts The ongoing integration of tourism in national and international strategy papers for regional development and international cooperation is based on the fact that, at the global level, tourism has been one of the most dynamic sectors of the economy over the last decades (UNWTO 2006;OECD 2010;Conrady and Buck 2011). Governments of developing countries, international development agencies, and NGOs see tourism promotion as an opportunity to initiate development processes, particularly in peripheral regions that have no other resources and potentials (Baumhackl et al 2006;Telfer and Sharpley 2008;OECD 2010). Accordingly, in the 1990s they began to apply the concept of communitybased tourism, and since the 2000s a model that provides for the poor-so-called pro-poor tourism-has been practiced as a suitable future blueprint for tourism in developing countries (Hall 2007). While the former focuses on the participation and the empowerment of local communities, the latter aims at modifying economic distribution processes to improve the income of the impoverished population. However, both scholars and politicians have discussed controversial views on the potential of this type of tourism as a developing factor. Experiences gained from the implementation of these concepts in the past decades paint a contradictory picture that leaves doubts as to their effectiveness (Ashley et al 2001;Hall 2007;Okazaki 2008). Many authors question the statement postulated by the UNWTO (2007) that ''tourism exchanges benefit primarily the countries of the South'' (Chok et al 2007;Scheyvens 2007;Schilcher 2007). Some deplore the limited effects of tourism on ecological and socioeconomic sustainability, while others demonstrate the poverty-reducing outcomes that community-based and pro-poor tourism can have at the local level; most agree, however, that under neoliberal conditions, with uneven distribution of income and unequal power structures at the regional and international scales, tourism does not initiate any fundamental changes.
The potential of tourism per se as a poverty-reducing factor is rarely questioned; the discussion is, rather, about adequate ways of ensuring its long-term economic viability and its social and environmental sustainability. Suggestions include, for example, establishing high-priced eco-tourism or providing targeted support to small, medium, and micro-enterprises in southern Africa (Harrison 2001;Baumhackl et al 2006;Rogerson 2007;Saarinen et al 2009;Mitchell and Ashley 2010). Local initiative and bottom-up approaches are often underlined as crucial to achieving a fair and long-term effect (Zapata et al 2011).
One of the aims of this study was to review the poverty-reducing effects of community-based tourism. This was done based on the livelihoods approach, which was deemed suitable because it puts the focus on securing the daily survival of households and places the different factors of their standard of living in the foreground (Carney 2002;Lohnert and Steinbrink 2005). The approach starts from the premise that everyday management of a destitute household is based not only on monetary income from a variety of sources but also on various forms of capital, that is, social, human, natural, physical, and financial resources, which household members use in very complex, efficient, and at the same time dynamic combinations (Bohle 2001). Access to this capital, in turn, depends on parent structures and conditions, which generate inclusion and exclusion processes.
As the present study focuses on the impact of tourism at both the regional and the household levels, application of the livelihoods approach implies connecting the analyses of processes at the macro-and microlevels. At the regional level, we illuminate the effects that tourism has on the regional economy in general (job creation, regional added value, etc) and on the internal structuring of the tourism sector (corporate structure, development of upstream and downstream sectors, etc). At the community and household level, we analyze to what extent the GPSC as a community-based tourism organization contributes to community welfare, and to what degree the local population's integration in the organization ensures and stabilizes households' livelihoods.
Tourism around Mt Kenya
Since the 1990s, Kenya as a tourist destination has experienced an expansion of mass tourism. In Kenya, unlike its neighboring countries, this is connected with the successful implementation of government policies to promote a diversified offering (Job and Metzler 2003). Although the country was able to reverse the decline in visitor numbers recorded in 2007/2008, and in 2010 even reached its previous record result of nearly 1.5 million tourist arrivals (KNBS 2011), the term ''mass tourism'' should be used with caution and put into relation: The total number of visitors to Kenya is only slightly higher than the number of foreign tourists visiting a single Austrian valley known for its ski slopes (such as the Ziller or the Oetz valleys). Nevertheless, tourism accounts for approximately 9% of the Kenyan GDP, and more than 1 million people are (formally or informally) employed in this sector (UNWTO 2006). Despite the diversification of attractive offers to tourists, demand is still focused on coastal tourism and on game drives in the numerous national parks in savanna areas. With an average of 25,600 visitors annually (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009), Mt Kenya is one of the least visited national parks in Kenya; Nakura National Park, for example, has 213,800 visitors annually (Kenya Ministry of Tourism 2011). The small number of tourists visiting Mt Kenya shows that mountaineering and hiking constitute only a tiny fraction of Kenya's tourism (Figure 1).
A socioeconomic profile of the Mt Kenya region
Situated directly at the equator, the region around Mt Kenya is a favored area for agriculture. It has fertile volcanic soils and an adequate water supply from 2 rainy seasons as well as relief precipitation (especially in the eastern and southern part) and, to a lesser extent, glacier melting. The landscape is smooth, and the temperatures are mild due to the high altitude. Already in precolonial times, Nilotic (mainly Maasai) and Bantu peoples competed for this land with its exceptionally favorable conditions; and in colonial times the British claimed the northern and western part of the area as ''White Highlands'' for white immigrants (Morgan 1963). In administrative terms, the Mt Kenya region can easily be delimited. Like the sectors of a pie chart, 8 districts converge at Mt Kenya's highest peak (Nelion, 5199 m). With the exception of Kyeni East Division in Nyeri North District, all 7 remaining districts are clustered mainly around the volcanic areas of Mt Kenya. The region is inhabited predominantly by Bantu peoples: Kikuyu in the west and northwest, Embu in the south, and Meru in the east and northeast. Even today, some 70% of the more than 1.7 million residents around the volcano make their living from the primary sector, which contributes approximately 80% of households' income (Kenya Ministry of State for Planning 2008; KNBS 2010). Although the poverty rate is still over 30% (Table 1), this region of Kenya is economically advantaged and therefore a popular in-migration area. The current high rates of population influx, particularly in the northern and western parts of the Mt Kenya region, have been highlighted in all surveys. On the other hand, at 3.8 the total fertility rate in the region is significantly lower than the national average (5.0). While the Kikuyu still constitute the majority of the region's population (Sim and Ronald 1979), several experts interviewed already consider that it has a multiethnic structure. The population of the divisions of Kyeni East and Timau increased by one third-from 133,600 to 178,000-in the 2 census periods from 1999 to 2009. In addition to small-scale subsistence farming and minor business enterprises in the small towns, the medium-and large-scale horti-and floriculture in Timau Division has offered substantial job opportunities since the early 1990s (Schuler 2004; Figure 2). North and west of Mt Kenya is also where the most important gates to the National Park are located and where the most popular mountain climbing routes begin. Accordingly, tourist activities, guides' and porters' cooperatives, and lodges that cater to (affluent) foreign visitors are concentrated around these entry points (Figure 3). Naro Moru is located in this area as well.
The preference for the western and northern slopes is associated with distinct topographical and climatic reasons: The smooth morphology in the rain shadow offers relatively easy access to the summit of Mt Kenya. Furthermore, it should be noted that the western and northern foot slopes-the former ranchland of the Laikipian ''White Highlands''-are semiarid, with less than 600 mm of annual precipitation in some parts, while the southern and eastern foot slopes benefit from sufficient precipitation and thus constitute appropriate agricultural land (Ojany and Ogendo 1988: 57, 135).
The importance of alpine tourism in the Naro Moru area
The approximately 25,600 annual visitors to the Mt Kenya National Park predominantly come from Europe (51%) and Africa (30%), followed by North America (10%). Most African visitors are Kenyans traveling in large groups; school excursions account for a substantial portion of domestic visitors. Half of the European tourists are British. A significant share is associated with the British Army base at Nanyuki, which regularly organizes trips to Mt Kenya. Other large visitor groups come from Germany, the Netherlands, France, Austria, Switzerland, and the Czech Republic. The easily accessible Naro Moru Gate to the National Park is not the one with the most entries to the summits (Figure 3). To the GPSC, however, it is of particular interest because of its proximity to the organization's office.
The results of our initial analyses show a low significance of tourism in the investigation area. In addition, it is important to consider that domestic tourism is based mainly on student excursions that are organized by official authorities and hence do not have a substantial impact on Naro Moru's economy. Even though social institutions (orphanages, hospital wards, etc) and schools are funded through revenues from the National Park, none of the schools in the vicinity focus on tourism with a view to training better qualified local personnel.
For the region as a whole, tourism has only a limited effect on socioeconomic development. This is due to the fact that a small number of large tour operators and lodges around Mt Kenya are monopolizing tourism in the region (Figure 2).
Local tour operators have established themselves in recent years, but competition has increased significantly, relativizing the economic potential for local companies. Within the tourism sector very few permanent positions are created, predominantly at the management level; many of them are filled by executives from large tourism centers such as Nairobi or Mombasa. Employment opportunities for the local population are usually temporary and low income and typically include basic service jobs at hotels or as guides and porters for multiday tours. Erhard (2000) calculated that only 2% of Naro Moru's population work in tourism, and its townscape is hardly affected by tourism.
The low importance of tourism as an economic sector is also reflected in the Nyeri North District Development Plan 2008-2012 (Kenya Ministry of State for Planning 2008). Neither in this publication nor in the Nyeri District Statistics Office (KNBS 2010) is tourism in Naro Moru an issue. It was therefore no surprise that during our interviews the District Officer and other representatives of the public administration such as the Councilor were unable to provide any information about it-apart from referring to the two international hotels of Naro Moru, the Naro Moru River Lodge and the Mountain Rock Lodge. The Naro Moru River Lodge is a large tourism establishment situated on the outskirts of the town. With 133 beds, it receives the highest numbers of visitors among local accommodators and is thus the most significant employer of local people in the tourism sector: More than 50 persons from the area work there permanently, while the smaller Mountain Rock Lodge employs between 20 and 30 local people.
The tourism sector plays an important role only for about 120 of the total 1813 households in the neighboring rural sublocation of Kamburaini. These are the guides' and porters' households (see below). Overall, therefore, we found a type of tourism on the slopes of Mt Kenya that is evident neither spatially nor in numbers. There are even months in the rainy seasons when hardly any visitors are encountered on the individual routes. But even in July/August and December/ January, when most tourist entries are recorded, only some huts along the routes notice an increase in tourist activities.
Trickle-down effects from tourism on other economic sectors are moderate. While some visitors purchase small amounts of locally grown staple foods, the 2 large hotels import most of the goods from Nairobi. Nevertheless, a small number of locally based companies in the fields of transportation, maintenance, and construction do benefit from tourist activity. In spite of these moderate benefits, the tourism industry cannot kick-start regional development in Naro Moru based on its intensity alone. Moreover, there is no evidence that visitor numbers will increase in the near future.
In terms of poverty, Naro Moru is not particularly advantaged, as a comparison with other locations around Mt Kenya and even within Kieni East Division shows ( Table 1).
The GPSC as a community-based tourism organization
The development of the Mt Kenya Guides and Porters Safari Club (GPSC) can be traced back to the period of decolonization. Already in the 1960s, climbers as well as hikers in the Mt Kenya area used guiding and portage services by the local population. Guides and porters were subsequently under the management of the Naro Moru River Lodge. In 1970, however, the idea emerged to bring these service providers together in a registered and community-based organization. Approximately a dozen similar cooperatives developed around the Mt Kenya massif. Each of these guides' and porters' organizations encompasses 50 to 150 members. In 1995, the GPSC decided to purchase its own plot in the Sublocation, and from 2005 onward, with financial support from the Austrian Embassy, it built its own office, a dining hall and kitchen, a hostel, and a water tank.
The GPSC offers tours along various routes on Mt Kenya and of varying duration. Depending on the size and needs of tourist groups, the organization assembles teams of guides, porters, and cooks who are paid by days and service (in US$, exchange rate February 2011: guides 7.60, cooks 7.00, porters 5.50). A team is composed of 1 or 2 porters per tourist, as well as 1 guide and 1 cook per 5 tourists ( Figure 4). As a rule, these are members of the GPSC, who are called up according to a rotation system. In order to continually improve services, new applicants must go through a 2-year probationary period before they are permanently admitted. In addition, the association strives to train its members in higher-standard services, mountain-climbing skills, and various tourism businesses.
The demand for tours is influenced above all by the seasonality of alpine tourism. The months from April to June and from October to December usually see fewer than half the number of orders registered during the high-season months from July to September and from January to March. The GPSC receives the vast majority of orders from the surrounding lodges and tourism agencies, who prefer booking with the GPSC due to the reliability of its members ( Figure 5).
The GPSC's dependence on these business clients represents a major risk: From 2001 to 2010, over 80% of orders came from only 3 major tour operators, while less than 15% were individual bookings. The latter, however, earn the GPSC much higher revenues, as no commissions need to be paid to any intermediaries. For this reason, in recent years the GPSC has worked to increase the share of individual bookings by expanding and diversifying its services (tours in other national parks, lodging and provisions on its own plot); this goal was achieved in the season of 2009-2010. However, while tourism in Kenya currently shows an overall positive development, the GPSC's added value is on a downward trend. Overall, tourism in Kenya is extremely fragile due, among other things, to political instability. Reports on tribal instability and insecurity in the international mass media triggered a regressive phase prior to the riots in 2007/2008. A steady increase in members to the current number of 150 forced the GPSC to cease accepting new applicants. The GPSC was founded as a self-ruled organization focusing on the welfare of its members and collective help among them. Its grass-roots style of decision-making fosters member participation. While a committee of 9 elected members is responsible for managing the organization, important decisions are made and elections are held at the annual general meetings. The GPSC's ambitions and organizational structure are oriented to the social well-being of its members and the community, which is fostered through a welfare fund, microcredits, support for individual members facing hardship, and various other forms of assistance. This is reflected in the organization's balance sheet, particularly in the high proportion of expenditures-usually 80-90%-paid directly to the members in the form of salaries, bonuses, and other compensations (Table 2). Furthermore, in dry periods the GPSC offers water from its tank free of charge not only to its members, but to all households of the community. The conscious rejection of a profit-oriented business model has caused the GPSC's annual expenditures to exceed revenues in some years. The high expenses for welfare services are thus preventing the organization from building up financial reserves for additional infrastructure and general maintenance work. As a result, the GPSC remains dependent on ''external'' support for major investments (eg from Austria).
As rural households in the region largely practice small-scale subsistence agriculture, with only a very small production surplus for sale, working as guides or porters plays a prominent, but varying role in their livelihoods ( Figure 6). Depending on the assets available in each household, the GPSC contributes in different ways to reducing household vulnerability: For households with small land properties (household types A and B) GPSC jobs represent one of few, or even the only source of monetary income. During the peak season, porters and guides usually have 1-3 opportunities per month to work on a multiday tour. Based on an average tour duration of 3 days and a season of about 7 months per year, this allows them to earn between US$340-480 (J250-350, February 2011 exchange rate) per year, depending on their activities. These earnings are mostly invested in improving their households' human capital, such as in education for their children and in capacity-building for adult household members. Besides the income, households have access to the large social network of the GPSC committee and can thus extend their social capital. This seems to be very important in cases of emergency, when the GPSC's welfare fund provides access to money and committee members advocate the respective household's interests with the local authorities, the public administration, and the private sector. Even families of low vulnerability (household type C) benefit from GPSC activities, as the organization's infrastructure secures water supply in times of water scarcity. Hence, by improving financial (income), human (education), physical (livestock), natural (water), and social (network) capital, the GPSC has a positive effect on households, reducing their vulnerability.
However, the social status of guides and porters in the community is relatively low, because their work is hard, and they are exposed to all weather conditions and to the risks of the mountains. Consequently, the majority of porters interviewed do not recommend their sons and daughters to become a porter or guide. Most of them stated that they want their children to be well educated and get ''better'' jobs and living conditions-a view that is confirmed by members of the younger generations (Figure 7).
Synthesis and conclusion
The example of the Mt Kenya region shows the multiple layers of impact that alpine tourism has in mountain areas of tropical Africa. Because of low numbers of visitors in the Mt Kenya National Park and the monopolies of a few large commercial tour operators, development endeavors at the regional level are hardly noticeable, neither with supply and service companies, nor in terms of employment effects. Nevertheless, the few available jobs in tourism are very popular, because they provide access to foreign currency and/or generous gratuities from tourists from rich countries, as well as to contacts abroad, which the local people see as a potential springboard for foreign jobs or training. The large (exclusive) hotels are the guides' and porters' organizations' most important link to clients, which leads to very one-sided relationships of dependency. In contrast, community-based tourism, as shown by the example of the GPSC, stabilizes rural households' livelihoods and contributes to community welfare. Its particularities-a not-for-profit business model, bottomup organization, and a democratic organizational structure with elected and regularly rotating officesprevent the inequitable enrichment of a small group of members and ensure that benefits are evenly distributed among all members and the whole community. Due to the high reputation that the GPSC enjoys with both the large hotels and the local authorities-in July 2011 almost all representatives of public institutions followed the GPSC's invitation to attend the inaugural opening of its new buildings-it can be assumed that the concerns of the GPSC and its members are heard by the political decision makers. On the one hand, this empowerment secures uncomplicated access (in terms of bureaucracy) to the National Park as a base for tourism. Beyond that, it can contribute to more rapid emergency response, for example, in the event of a crisis, such as a drought or ethnic conflict. Overall, our results confirmed the main hypothesis underlying this research project. The results are also similar to those recently presented by Lapeyre (2011), who analyzed a tourism community-publicprivate partnership in a rural area in Namibia.
For rural households, employment as a guide or porter is one of the few sources of monetary income in the region; it supports the families' livelihoods. By strengthening the different forms of capital it also reduces households' vulnerability. The occupation of guide or porter is not deemed an attractive prospect for the future, however, as income from it is used primarily for educating the family's children in order to enable them to achieve higher-quality employment. In this sense, tourism may contribute to migration out of the region. At the same time, career opportunities within the tourism sector are scarce, and other segments of the economy neither provide highly qualified jobs nor respond to development impulses from tourism.
As shown by the example of the Mt Kenya region, Afro-alpine tourism as a whole provides only limited impulses to regional development. Even though community-based tourism helps to alleviate poverty and to trigger empowerment processes, the generation of equitable welfare for households and the community depends on the respective organizations' internal participatory and democratic structures. Communitybased tourism alone cannot initiate sustainable regional development.
Comparing these results for the Mt Kenya region with findings on other mountain areas of tropical Africa, it can be assumed that they share similar structures and processes. Studies on tourism in the Rwenzori Mountains indicate a much less developed tourism sector, demonstrating its proportionately low impact at the local and regional levels. Although no relevant research results are available for the popular climbers' destination Kilimanjaro-recent scientific publications mostly look at ecological and climatic issues (Bloemer 2002;Bart et al 2003;Nü sser 2009;Wangui et al 2012)-the basic situation is likely to be comparable: The local population around the 3 highest mountain massifs in Africa has remained at the level of subsistence economy and does not have any attractive prospects within the region (Engelhard 1990;Erhard 2000). | 2019-04-25T13:02:39.905Z | 2012-11-01T00:00:00.000 | {
"year": 2012,
"sha1": "70d71770665cb3bec884a4e68175c3158dc4533d",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.1659/mrd-journal-d-11-00102.1",
"oa_status": "GOLD",
"pdf_src": "BioOne",
"pdf_hash": "70d71770665cb3bec884a4e68175c3158dc4533d",
"s2fieldsofstudy": [
"Geography",
"Environmental Science",
"Economics"
],
"extfieldsofstudy": [
"Geography"
]
} |
12913805 | pes2o/s2orc | v3-fos-license | Emerging epidemic viral encephalitides with a special focus on henipaviruses
In the last few decades, there is an increasing emergence and re-emergence of viruses, such as West Nile virus, Enterovirus 71 and henipaviruses that cause epidemic viral encephalitis and other central nervous system (CNS) manifestations. The mortality and morbidity associated with these outbreaks are significant and frequently severe. While aspects of epidemiology, basic virology, etc., may be known, the pathology and pathogenesis are often less so, partly due to a lack of interest among pathologists or because many of these infections are considered “third world” diseases. In the study of epidemic viral encephalitis, the pathologist’s role in unravelling the pathology and pathogenesis is critical. The novel henipavirus infection is a good example. The newly created genus Henipavirus within the family Paramyxoviridae consists of two viruses, viz., Hendra virus and Nipah virus. These two viruses emerged in Australia and Asia, respectively, to cause severe encephalitides in humans and animals. Studies show that the pathological features of the acute encephalitis caused by henipaviruses are similar and a unique dual pathogenetic mechanism of vasculitis-induced microinfarction and parenchymal cell infection in the CNS (mainly neurons) and other organs causes severe tissue damage. Both viruses can cause relapsing encephalitis months and years after the acute infection due to a true recurrent infection as evidenced by the presence of virus in infected cells. Future emerging viral encephalitides will no doubt continue to pose considerable challenges to the neuropathologist, and as the West Nile virus outbreak demonstrates, even economically advanced nations are not spared.
Introduction
The last few decades saw continuing outbreaks of human infections involving many pathogenic viruses from practically all major families of viruses. Some of these viruses are known and endemic while others are novel. Some of the known viruses are emerging again to cause more outbreaks and in places not previously known for outbreaks to occur. Many of these viruses cause severe human disease and may affect many different organ systems. The most devastating epidemic in the recent past is of course due to the human immunodeficiency virus, which causes infection of many organ systems including the central nervous system (CNS). The influenza viruses (H5N1, H1N1, etc.) pose a serious health threat to millions of people by causing severe respiratory syndromes [1,3].
Viral encephalitis due to known arboviruses, such as Japanese encephalitis (JE) virus and tick-borne encephalitis (TBE) virus continue to cause widespread infections and deaths in endemic countries. In the Indian subcontinent, thousands of JE virus infections, still occur despite the availability of vaccines [26]. In the Far East, Russia and eastern Europe, TBE recurs regularly to cause deaths [51]. Because TBE is occurring in increasing numbers and in previously unaffected areas, it is now considered to be emerging. West Nile virus (WN), another known arbovirus usually found in Africa, Europe and Asia, recently emerged for the first time to cause severe disease in humans and animals in North America in 1999 [9,44,48]. In a short span of a decade [48], it has spread to the whole of North America to be the most important emerging epidemic encephalitis. The virus has caused neuroinvasive disease in more than 11,000 people with a mortality of 9%.
The enteroviruses are still responsible for outbreaks of CNS infections despite the worldwide retreat of poliovirus. The most important of these is probably Enterovirus 71 because of the very high mortality in patients who develop encephalomyelitis, although fortunately this complication is rare [40,81]. Nonetheless, it is emerging in many previously unaffected areas, particularly in Asia, where typically, children develop CNS disease in a background of epidemic hand, foot and mouth disease [90].
Another important group of viruses that has recently emerged to cause epidemic encephalitis are the henipaviruses. This group of viruses from the genus Henipavirus (family Paramyxoviridae) comprises the Hendra (HeV) virus and Nipah virus (NiV). Like other paramyxoviruses, henipaviruses are enveloped, negative-strand RNA viruses. Because of the novel nature of this new genus, the high mortality and the unique pathogenesis of the disease, this review will focus specially on henipavirus infection as previous reviews has focused mainly on NiV alone [81,86]. The pathology and pathogenesis shall be highlighted and compared with other viral encephalitides. For other important epidemic viral encephalitides, there are existing good reviews [26,32,51].
HeV was first isolated after an outbreak in horses and two humans in the town of Hendra, Australia in 1994 [56,72]. Following this, there were more outbreaks in horses and humans and to date a total of six cases with three fatalities have been reported [35,60,64]. All the human cases have had close contact with infected horses which are now thought to be the intermediate hosts for HeV transmission [52,80]. HeV cases have not been reported outside Australia.
The first NiV outbreak started in northern Malaysia in 1998 [7], involving about 27 patients with 15 fatalities [8]. The outbreak which started in pig farms then spread to other farms in the south following transfer of infected pigs [54]. This area became the second and most severely affected epicenter [7], and the virus was named after Kampung Sungai Nipah (Nipah river village) located here. A prevalence of 265 cases of acute NiV encephalitis with 105 fatalities in Malaysia has been reported [62], but if asymptomatic or mildly symptomatic cases are also included, the total number is probably more than 350 cases [86]. Infected pigs exported to Singapore spread the infection to some abattoir workers [7,63]. After 1999, in Malaysia and Singapore, there are no reports of new NiV outbreaks until 2001 onwards when several recurrent outbreaks of NiV in Bangladesh and India [36,38,49] have involved more than 120 people.
Henipavirus transmission to humans
The natural host of henipaviruses has been confirmed to be the fruit bat (Pteropus species), and bat-to-human transmission may be direct or indirect via intermediate hosts [13,34]. In HeV transmission, the horse is the intermediate host [64,72]. As virus could be detected in oronasal secretions and urine from infected horses, contact with these secretions appears to be the most likely route of transmission [21,80]. Although person-to-person HeV transmission has not been reported, involvement of the lung and kidney in acute infection and the presence of virus in nasopharyngeal secretions strongly suggest this possibility [64,85]. The mode of bat-to-horse transmission remains unclear, but may be due to the ingestion of feed or pasture contaminated by bat-derived foetal tissues or urine [80]. Outside Australia, HeV has not been isolated from bats.
In the first known NiV outbreak in Malaysia and Singapore, the intermediate host is the pig as it is clear that direct contact with infected pigs or fresh pig products was responsible for viral transmission. In Malaysia, a higher prevalence of infection was found among pig farmers, abattoir workers, pork sellers and army personnel involved in culling of pigs [2,10,62,65,69]. Widespread surveillance of pig populations and culling of sick animals stopped the Malaysian epidemic [8], while the banning of pig imports and abattoir closure stopped the outbreak in Singapore [8,10]. It was suggested that bat-to-pig transmission could result from pigs ingesting half-eaten, contaminated fruits dropped by bats near farms [13].
Person-to-person transmission is very rare in Malaysia. In a large serological survey of health staff, serum neutralisation tests were all negative [55]. However, a nurse who had previously cared for a NiV-infected patient, seroconverted but remained asymptomatic although the brain MRI showed a few discrete lesions typically seen in acute NiV encephalitis [75,76].
In contrast to the Malaysian outbreak, the Bangladeshi/ Indian outbreaks showed a high incidence of person-toperson transmission either to health care workers or other people who had contact with patients [31,36,39]. Personto-person transmission could be explained by the presence of virus in patients' secretions [14]. Moreover, no animal has been positively identified as intermediate hosts so far, although in some outbreaks, patients were reported to have had contact with sick animals (pig, cow and goat) [49]. Date palm sap, a local delicacy in Bangladesh, has been implicated in some cases [50]. Palm sap drip collected overnight into open pots tied onto palm trees allow foraging fruit bats to feed and contaminate the sap thus transmitting virus to people who drink the raw sap.
Severe HeV infection can manifest either as a neurological or a pulmonary syndrome, but since only very few patients have been involved, it is not well characterised as NiV infection. Neurological signs include confusion, motor deficits and seizures while the pulmonary syndrome comprises an influenza-like illness, hypoxaemia, diffuse alveolar shadowing in chest X-rays [64,72].
Severe NiV infection is characterised predominantly by an acute febrile encephalitic syndrome. In a cohort of 90 patients with acute NiV encephalitis, the main presenting features were fever, headache, dizziness, vomiting, and reduced level of consciousness [27]. In fact, more than 50% of patients have some degree of reduced consciousness. Clinical signs, such as areflexia, hypotonia, abnormal pupillary response, tachycardia, hypertension, abnormal doll's eye reflex and segmental myoclonus, suggested involvement of the brainstem and upper cervical cord. Segmental myoclonus was characterised by focal, rhythmic jerking of the diaphragm and muscles in the limbs, neck and face. Meningism and generalised tonic-clonic convulsions were also observed.
A pulmonary syndrome appears to occur in a minority of patients. In the same cohort only 14% was reported to have unproductive cough [27]. In another Malaysian hospital series, 24% of patients had abnormal findings in the chest X-rays, but none had severe lung disease [11]. In the Singapore series of 11 patients, 3 were clinically thought to have atypical pneumonia with abnormal chest X-rays [63].
Brain MRI scans in acute henipavirus encephalitis show multiple, disseminated, small discrete hyperintense lesions mainly in the cortex, subcortical and deep white matter [47,64,70]. In three Bangladeshi patients with apparent acute NiV encephalitis and available brain MRI findings, only one patient showed the same discrete hyperintense lesions while other two patients showed multiple confluent lesions [66].
Specific anti-henipavirus IgM and IgG antibodies that can be detected in the serum and CSF in most patients are critical to diagnosis. More is known about seroconversion after NiV infection than HeV infection. Overall, antibodies are more likely to be positive in serum than CSF. In NiV infection, IgM seroconversion by day 4 is about 65%, and by day 12, 100%. IgM can persist for at least 3 months. There is 100% IgG, seroconversion by day 25 [67] and IgG levels may persist for several years [12]. Using either serology or IHC in autopsy tissues, a positive diagnosis can be made in a majority of NiV cases, and when these diagnostic methods are used in combination, the diagnosis can be confirmed in all cases [87]. Specific neutralising antibodies, IgM or IgG have been reported in HeV infected patients [35,60,72].
Complications and sequelae in henipavirus infection
Mortality in HeV infection is about 50%, while mortality in severe NiV infection is about 40% (Malaysia) to 70% (Bangladesh/India) [36,38,62]. In many of the Malaysian patients who recovered, there were no apparent serious sequelae [27]. Neuropsychiatric sequelae have been reported [59]. Fatal intracerebral haemorrhage is a rare complication [27]. It is not known if post-infectious encephalomyelitis occurs following acute henipavirus infection.
Henipavirus infection may be complicated by relapsing encephalitis following apparent recovery. One case of relapsing HeV encephalitis and more than 20 cases of relapsing NiV encephalitis (probably \10% of survivors) have been reported thus far [12,60,74]. Some cases of relapsing NiV encephalitis only had mild symptoms such as fever and headache during the acute phase, and hence have also been called ''late-onset'' encephalitis. The single case of relapsing HeV encephalitis occurred about 13 months after exposure, while an average of 8 months elapsed before relapsing NiV encephalitis occurred. Clinical and radiological findings suggest that relapsing NiV encephalitis is distinct from acute NiV encephalitis [70,74]. The brain MRI in relapsing henipavirus encephalitis shows patchy, confluent hyper-intense cortical lesions.
Pathology of henipavirus infection
The macroscopic features of the HeV-infected brain have not been reported while the features in the NiV-infected brain are non-specific; no discrete lesions could be identified. Our current knowledge of the microscopic pathology of henipavirus infection is based on autopsy tissues: 2 autopsies of HeV infection and more than 30 autopsies of NiV infection [85,87]. In general, the microscopic features in these two infections appear to be very similar, hence they will be discussed together.
Acute henipavirus infection
One of the most important targets in acute henipavirus infection is the endothelium and smooth muscle of blood vessels. True vasculitis characterized by varying degrees of segmental endothelial ulceration, karyorrhexis, intramural necrosis and inflammatory cells is observed in blood vessels of the brain (Fig. 1a), lung, kidney, heart and many other major organs. Milder subendothelial inflammation (endothelitis) may also be seen (Fig. 1b). In some vessels particularly in NiV infection, occasional endothelial multinucleated giant cell (Fig. 1d) can be detected (about 30% of cases). Thrombosis is often associated with vasculitis and some vessels may be completely obliterated by thrombotic plugs (Fig. 1a). Viral antigens (Fig. 1c), RNA and nucleocapsids can be detected in endothelium and multinucleated giant cells, and vascular smooth muscle. In NiV infection, there is a suggestion that CNS vascular susceptibility is the highest compared to vessels in other organs. Typically, small vessels, e.g. capillaries, small arteries and venules show evidence of vasculitis, but not the larger vessels. Focal haemorrhages may be observed near vascular lesions.
In the CNS, the main pathological findings are vasculitis (with or without thrombosis), parenchymal necrosis and evidence of viral infection in neuroglial cells. Vascular lesions in both grey and white matter are seen throughout the CNS and these are often associated with discrete necrotic or more subtle vacuolar plaque-like lesions. The necrotic plaque-like lesions are characterized by varying degrees of necrosis, neuropil vacuolation/oedema and mild inflammation (Fig. 1e). In neuronal areas, there may be some neuronal loss as well. In the white matter, welldeveloped plaques consist of eosinophilic, necrotic material similar to axonal spheroids seen in diffuse axonal injury (Fig. 1f). Inflammatory cells when present comprise neutrophils, macrophages, lymphocytes, plasma cells and reactive microglia. In the HeV-infected molecular layer of the cerebellum, subtle vacuolar plaques are paler staining and consist of small fine vacuoles associated with an increase of CD68-positive macrophages/microglia which may be difficult to detect without IHC. In some cases, focal neuronophagia, microglial nodule formation, clusters of foamy macrophages, perivascular cuffing and meningitis can be found. Some neurons show the rare perineuronal Viral inclusions, antigens, RNA and nucleocapsids are observed mainly in neurons (soma and processes), although the very rare ependymal cell or astrocyte may be involved [28,41,87]. Neuronal viral inclusions can be found in the cytoplasm and nuclei, mostly near vasculitic vessels or necrotic plaques. Cytoplasmic inclusions are usually small, discrete, eosinophilic, and sometimes multiple (Fig. 2a, b). Nuclear inclusions are less commonly found and occupy most of the nucleus. Inclusions were reported in 62% of cases in one series. Viral antigen/RNA positive neurons (Fig. 2c) found at the periphery of necrotic/vacuolar plaques may form concentric or eccentric rings. Some plaquelike, groups of positive neurons are not associated with prominent necrosis, vacuolation or oedema (Fig. 2d). Necrotic plaques in the white matter are generally not associated with viral antigens/RNA. In acute NiV encephalitis, the level of viral antigens peaks at about 6-10 days, and are largely cleared after approximately 14 days. In a case of resolving acute NiV encephalitis who died several months after the infection, the brain showed randomly distributed, discrete slit-like, oval or spherical lesions in the grey and white matter. Most lesions consist of foamy macrophages, lymphocytes and reactive gliosis (unpublished observations). There was no evidence of vasculitis, but some residual perivascular cuffing remained.
In the lung, apart from vasculitis, there is parenchymal inflammation, necrosis, intra-alveolar macrophages/ inflammatory cells, type II pneumocyte proliferation, alveolar membranes and haemorrhage. Occasionally, intraalveolar multinucleated giant cells with nuclear inclusions are noted. In the kidney, the rare focal glomerulitis with or without thrombosis or necrosis, and focal inflammation around necrotic tubules may be observed. The heart and lymph node may also show features of infection. Viral antigens are demonstrable in the lung parenchyma including alveolar type II pneumocytes, intra-alveolar macrophages, and renal glomeruli and tubules.
Relapsing henipavirus encephalitis
The pathological features of relapsing henipavirus encephalitis are based on an autopsy case of HeV [85] and two autopsies of relapsing NiV encephalitis [74,87].
Macroscopically, relapsing NiV encephalitis shows varying degrees of confluent softening and necrosis in the cerebral cortex and subcortical areas such as the thalamus and basal ganglia. The microscopic features of relapsing HeV and NiV encephalitis again appear to be similar and the pathology is confined to the CNS. Other non-CNS organs are essentially normal and vasculitis is absent throughout.
In affected neuronal areas (cerebral cortex, basal ganglia brainstem, etc.), confluent and extensive parenchymal necrosis, oedema and inflammation are seen with some spill over into adjacent white matter. Inflammatory cells consist of macrophages, lymphocytes and some plasma cells, with prominent perivascular cuffing. In many areas, severe neuronal loss is replaced by reactive glial and prominent vascular proliferation. Although viral inclusions can be found, the most prominent inclusions are found in relapsing NiV encephalitis. Focal viral antigens/RNA and nucleocapsids are demonstrated mainly in surviving neurons (Fig. 2e, f), ependyma and possibly in other glial or inflammatory cells as well. Neuronophagia and prominent microglial nodules are rarely observed. Severe meningitis is found in many areas. Vasculitis, endothelial syncytia or thrombosis as seen in acute henipavirus encephalitis are absent. Blood vessels are all negative for antigen/RNA.
Comparative pathology of viral encephalitides
Certain pathological features in henipavirus encephalitis may be distinctive enough to suggest the diagnosis, particularly if this can be confirmed by IHC, ISH, serology, RT-PCR and virus culture. Perhaps the most unique finding in acute henipavirus encephalitis is the multinucleated endothelial cell found in about 30% of NiV cases. This feature has not been described in other viral encephalitides. Extensive vasculitis found in all cases is probably a more useful feature to diagnose acute henipavirus encephalitis. In viral encephalitis, CNS vasculitis is rarely encountered with the exception of varicella-zoster and herpes simplex which may be associated with granulomatous angiitis [23,71], a feature not seen in acute henipavirus infection. In the case of varicella zoster, usually larger vessels are implicated, but the type of vascular lesions may be variable [43]. Other non-viral pathogens including rickettsiae, and Neisseria [73] may be associated with vasculitis. In rickettsial encephalitis, vasculitis and necrosis are more subtle and less prominent [79]. Other CNS changes in acute henipavirus encephalitis, such as perivascular cuffing, parenchymal inflammation and neuronophagia, are rather non-specific features and can be found in other viral encephalitides [20].
Vasculitis, a key event in the pathogenesis of acute henipavirus infection, follows from vascular endothelial and smooth muscle cell involvement resulting in thrombosis, vascular occlusion, ischaemia, microinfarction, and probably thromboembolism as well. These vascular lesions contribute to the formation of necrotic/vacuolar plaques that seem to correlate with the multiple discrete lesions seen in brain MRI studies [47,70]. However, neuronal infection also contributes to plaque formation especially in the cerebral grey matter and brain stem. This dual pathogenetic mechanism in the CNS and other organs appear to be unique to henipaviruses. Necrotic/vacuolar plaques in the white matter are caused mainly by ischaemia/microinfarction since glial cells are far less susceptible to infection. Vasculitic vessels probably cause a breach in the blood-brain barrier to facilitate virus escape into the parenchyma to infect neurons. Inter-neuronal spread further into the periphery of the plaque may contribute to dissemination. In subacute sclerosing panencephalitis (SSPE), it is believed that endothelial infection facilitates measles virus entry into the brain although vasculitis is absent [15].
The neuron is often the main, if not the only target cell, of most neurotropic viruses, including JE virus [16], TBE virus [24], West Nile virus [30], rabies [42], measles, herpesviruses and enterovirus 71 [83]. Neuronal infection most likely leads to viral cytolysis and cell damage in various critical parts of the CNS leading to an encephalitic syndrome. Hence, significant pathological changes such as viral inclusions and virus localisation/detection by various means are expected to be found in relation to the neuron. Since viral inclusions can be found in many other viral encephalitides, it is not that helpful for the diagnosis of henipavirus encephalitis. Compared to endothelium and neurons, glial and epithelial cells are rarely involved. This may be related to the density of ephrin B2 and B3 recently found to be the receptors for henipavirus although this has not been studied [4,57,58]. A recent paper suggests that the blood vessel may also be a target in JE, but this has yet to be confirmed [25]. Macrophages are the main infected cells in subacute HIV encephalitis [22].
IHC is very useful to detect viral antigens to confirm the diagnosis of henipavirus infection and indeed in numerous other viral encephalitides as well. Both polyclonal and monoclonal antibodies to NiV and HeV have been produced and many of them are cross reactive and sensitive enough to detect both infections [53,77,87,89]. Unfortunately, most antibodies suitable for IHC are proprietary and not commercially available. Needless to say, a good knowledge of target cells is needed to assess the IHC assay, and in the case of henipavirus, viral antigen localisation in neurons and blood vessels are useful for diagnosis. If available, ISH which detects viral RNA is also a useful adjunct to tissue diagnosis of henipavirus infection [84].
One of the most interesting complications of acute henipavirus infection is relapsing encephalitis. Fortunately, it is relatively rare and not uniformly fatal. The presence of viral inclusions, nucleocapsids, antigens and RNA confirms relapsing henipavirus encephalitis as a recurrent infection rather than post-infectious encephalitis [74]. It is assumed that if recurrent viruses were from extra-CNS sites, then viraemia (and vasculitis) have to occur to enable virus to enter the CNS similar to acute henipavirus encephalitis. Hence, the absence of vasculitis and extra CNS organ involvement indirectly suggests reactivation of latent viral foci from within the CNS, introduced during the acute infection. Vasculitis-induced thrombosis, ischaemia and microinfarction do not appear to play a role in contrast to acute henipavirus encephalitis.
The risk factors for relapsing henipavirus encephalitis are unknown. Clinically, relapsing henipavirus encephalitis does share some similarities with SSPE. However, the latter typically sets in several years after the acute infection and may be more often fatal. Nonetheless, like relapsing henipavirus encephalitis, SSPE is not invariably fatal and recurrences have been reported [12,17,74]. Virus genomic mutations reported in SSPE is one possible mechanism for henipaviruses to remain latent and escape the immune response [6]. So far, no viral mutations have been found in relapsing henipavirus encephalitis [82]. Measles virus is well known to cause immune suppression [29] and being paramyxoviruses themselves, henipaviruses may similarly cause immune suppression that might impact on the development of relapsing encephalitis. Further investigations are much needed to unravel the pathogenesis of relapsing henipavirus infection.
Conclusion
It is apparent that NiV and HeV, both from the same Henipavirus genus, share many common clinicopathological features suggesting that the pathogenesis of the human diseases, respectively, are essentially the same.
The outbreaks of henipavirus infections in Asia and Australia are prime examples of emerging zoonotic infectious diseases that continue to occur worldwide, many of them associated with severe epidemic encephalitis. As far as newly emerging or novel viruses are concerned, one of the most important natural host is the bat. As many as 40 viruses have recently been isolated from bat species in as many years, though not all have been shown to be pathogenic to humans. These include Ebola virus, Australian bat lyssavirus, SARS coronavirus, Menangle virus, Tioman virus, henipaviruses [5,18,88]. Because of the worldwide range of bats and their ability to fly to large areas of human habitat, they are very effective for virus dissemination under the right conditions. In the case of henipaviruses, the range of pteropid bats includes Southeast Asia, China, Japan, Oceania, the Indian subcontinent, Australia and Africa, and there is evidence of bat infection in many of these countries [13,19,33,34,37,39,45,46,61,68,78]. Hence, future henipavirus outbreaks can be expected in these regions.
Needless to say, the role of pathologists in understanding the pathology and pathogenesis of emerging viral encephalitides is critical. The relative lack of interest in these infections among pathologists especially in more developed countries is perhaps not surprising, as they are often regarded as tropical or ''third world'' diseases. However, with increasing air travel and ability of viruses to emerge in previously unaffected areas, e.g. West Nile virus in North America, it is quite clear that greater efforts should be made to study these diseases and to improve diagnostic capabilities. Further understanding of the pathology and pathogenesis of emerging epidemic viral encephalitides should continue to contribute significantly to the development of therapeutic strategies and vaccines. | 2017-08-02T19:38:07.176Z | 2010-07-23T00:00:00.000 | {
"year": 2010,
"sha1": "ed883041a0d3e747865c7854f961ac149cd31690",
"oa_license": "implied-oa",
"oa_url": "https://europepmc.org/articles/pmc7086526?pdf=render",
"oa_status": "GREEN",
"pdf_src": "PubMedCentral",
"pdf_hash": "ed883041a0d3e747865c7854f961ac149cd31690",
"s2fieldsofstudy": [
"Medicine",
"Biology"
],
"extfieldsofstudy": [
"Biology",
"Medicine"
]
} |
244840711 | pes2o/s2orc | v3-fos-license | Evaluation of the Abbott BinaxNOW COVID-19 Test Ag Card for rapid detection of SARS-CoV-2 infection by a local public health district with a rural population
SARS-CoV-2 RT-PCR, the gold standard for diagnostic testing, may not be readily available or logistically applicable for routine COVID-19 testing in many rural communities in the United States. In this validation study, we compared the BinaxNOW™ COVID-19 Test Ag Card with SARS-CoV-2 RT-PCR in 214 participants who sought COVID-19 testing from a local public health district in Idaho, USA. The median age of participants was 35 and 82.7% were symptomatic. Thirty-seven participants (17.3%) had positive RT-PCR results. Results between the two tests were 94.4% concordant. The sensitivity of the BinaxNOW™ COVID-19 Test Ag Card was 67.6% (95% CI: 50.2–81.9%), and the specificity was 100.0% (95% CI: 97.9–100.0%). The positive predictive value (PPV) for the BinaxNOW™ COVID-19 Test Ag Card was 100.0% (95% CI: 86.2–100.0%), and the negative predictive value (NPV) was 93.6% (95% CI: 89.1–96.6%). Although the sensitivity of BinaxNOW™ COVID-19 Test Ag Card was lower than RT-PCR, rapid results and high specificity support its use for early detection of COVID-19, especially in settings where SARS-CoV-2 RT-PCR testing is not readily available. Rapid antigen tests, such as the BinaxNOW™ COVID-19 Test Ag Card, may be a more convenient tool in quickly identifying and preventing COVID-19 transmission, especially in rural settings.
Introduction
Early detection of SARS-CoV-2, the virus that causes COVID-19, is essential for slowing community transmission. Real-time reverse transcription-polymerase chain reaction (RT-PCR) assays for COVID-19 diagnosis have excellent sensitivity and specificity but require laboratory instrumentation and sending specimens to a laboratory for testing can culminate in delayed turnaround times for results [1]. Rapid, point-of-care tests that quickly identify individuals with SARS-CoV-2 are particularly useful in public health settings to limit the spread of infection [2]. The BinaxNOW™ COVID-19 Test Ag Card (BinaxNOW™ COVID-19 Test) is a rapid antigen test in a card format that detects SARS-CoV-2 protein antigens present in a nasal swab specimen. This test received an Emergency Use Authorization (EUA) from the Food & Drug Administration (FDA) on August 26, 2020 [3]. The BinaxNOW™ COVID-19 Test is a lateral flow immunoassay used at the point of care with results read visually after a 15-minute incubation period. Use of the BinaxNOW™ COVID-19 Test has increased the availability of COVID-19 testing in Idaho, where laboratory-based COVID-19 testing remains limited for many geographically isolated, rural communities. Additional data on the performance of the BinaxNOW™ COVID-19 Test are needed to help providers interpret results and determine whether confirmatory testing should be recommended [4]. A prospective, observational study at a local public health district in Southwestern Idaho was conducted to evaluate the performance of BinaxNOW™ COVID-19 Test compared with the gold standard of SARS-CoV-2 RT-PCR. Demand for COVID-19 testing was evaluated over time, by monitoring testing rates throughout the region as well as within the Southwest District Health clinic.
Materials and methods
Through this evaluation, we determined the percent concordance, sensitivity, specificity, positive predictive value, and negative predictive value of BinaxNOW™ COVID-19 Test and SARS-CoV-2 RT-PCR. Examination of BinaxNOW™ COVID-19 Test performance characteristics employed a modified, more restrictive positive test interpretation criteria, where only bands that extended across the full width of the test strip were scored as positive, as recommended by Pilarowski and colleagues [5].
Study site and participant enrollment
This study was conducted by Southwest District Health, one of Idaho's seven public health districts, that serves six counties with a total estimated population of 283,930 residents. All counties in Southwest District Health are at least partially classified as rural by the Health Resources & Services Administration's Federal Office of Rural Health Policy [6]. Subjects were identified among clients who scheduled an appointment for COVID-19 testing with Southwest District Health Employee Health/Clinic Services in Caldwell, Idaho, or through any mobile testing locations facilitated by Southwest District Health between November 2020 and May 2021. Prior to enrollment and performance of any study-specific procedure, a signed informed consent form was obtained for each participant. Interpreter services were offered when a language barrier was present. For minors, written consent from a parent or guardian was required for enrollment, as well as verbal assent from the minor. Eligible participants were those who sought COVID-19 testing from the Southwest District Health Clinic, were able to provide informed consent, and had not been diagnosed with COVID-19 in the 90 days prior to specimen collection. All participants were screened using a standardized questionnaire to collect demographic data and to identify symptom, exposure, and immunization status. Demographic data including age, gender, race, ethnicity, and contact information were collected and participants were asked about their symptoms and previous vaccinations, recent travel, and whether they had been exposed to a person diagnosed with COVID-19 in the previous 14 days. until resistance is met at the level of the turbinate, rotating five times against the nasal wall, and repeating in the other nostril using the same swab [7]. Results were visually read promptly 15 minutes after the test card was closed. A second specimen was collected from each participant for SARS-CoV-2 RT-PCR testing, using the same swab collection procedure. Second specimens were immediately placed in 3 mL virus stabilization tubes containing viral transport media and stored between 2˚C −8˚C until transported to the Idaho Bureau of Laboratories (IBL) for RT-PCR testing using the TaqPath TM COVID-19 Combo Kit (ThermoFisher Scientific A47814). Results were determined by the automated ThermoFisher Interpretive Software TM and confirmed by IBL staff. Results were available in 1-3 days of specimen collection. A subset of positive specimens was sequenced with the Oxford Nanopore Technologies (ONT) MinION Mk1c using the ONT protocol "PCR tiling of SARS-CoV-2 virus with rapid barcoding (SQK-RBK110.96) Version: PCTR_9125_v110_revD_24Mar2021" developed by Freed and Silander [8]. Assemblies were performed using CLC Genomics version 21.0.4 long read support module and lineage calls made by the Pangolin COVID-19 Lineage Assigner version 3.1.5 [9]. The resulting consensus sequences were uploaded to GISAID [10].
Data governance and analysis
A unique study identification number was assigned to each participant in the order of enrollment. Paper screening forms were archived and stored in a locked filing cabinet, while results from RT-PCR testing were reported using an electronic portal and added to the study database. The study database is stored securely on Southwest District Health servers, with access restricted to study personnel. Data analysis was conducted using Excel for data storage, with R 4.0.3 software used for statistical analysis. Sample size calculations were based on the primary outcome, percent concordance of BinaxNOW™ COVID-19 Test and SARS-CoV-2 RT-PCR results, using precision-based calculations. For an estimated concordance of 90%, a sample size of 200 participants would produce a two-sided 95% confidence interval of width ± 4.2%. Secondary outcomes include determining the sensitivity, specificity, positive predictive value, and negative predictive value for the BinaxNOW™ COVID-19 Test, which are reported as percentages with two sided 95% confidence intervals using an alpha of 0.05. COVID-19 testing demand was monitored throughout the study enrollment window by the Idaho Department of Health & Welfare. Data are based on electronic laboratory reports (ELRs) received from laboratories that report both positive and negative results. Data are also based on the date the specimen was collected, not the date the lab result was received. Records were extracted weekly and shared with Idaho's Public Health Departments.
Ethics and confidentiality
This project was reviewed by the Idaho Department of Health and Welfare, Division of Public Health Research Determination Committee and was deemed public health surveillance activity. The project was determined as non-research and exempt from ethical review by the Institutional Review Board.
Participant confidentiality is strictly held in trust by the participating investigators, staff, and institutions. All research staff received training in Health Insurance Portability and Accountability Act (HIPAA) and privacy and confidentiality as part of their routine duties. Participant information was stored securely, de-identified, and used for analysis as described above. Documents that identify the participant (e.g., the signed informed consent) were maintained in a locked file cabinet at Southwest District Health with access limited to the principal investigator.
Of the 214 paired samples, 5.6% (n = 12) had discordant results. All discordant results were due to a false negative from the BinaxNOW™ COVID-19 Test. A total of 83.3% (n = 10) of participants with discordant results reported symptoms at the time of sample collection. The median (IQR) days between symptom onset and sample collection for symptomatic participants with discordant results were 2.5 (1-4) days (n = 10), with a greater variability in days since symptom onset compared with symptomatic participants with concordant results with a median (IQR) of 2 (1-3) days (n = 167) since symptom onset (Fig 1). There was no significant difference in days since symptom onset when comparing concordant and discordant results (p = 0.439) using the student's t-test (Table 1).
A total of 14.5% of participants received at least one dose of a COVID-19 vaccination, and 12.1% of participants were fully immunized. Of those who were fully immunized, 70.3% (n = 19) showed no symptoms. One fully immunized participant (3.7%, n = 1), who did not report any symptoms at time of sample collection, tested positive for COVID-19 through RT-PCR with concordant BinaxNOW™ COVID-19 Test results.
The Idaho Bureau of Laboratories sequenced 33 of the 37 positive samples ( Table 2). Of the 33 samples successfully sequenced, 25 produced unambiguous lineage calls and 8 produced sequences with some ambiguity as assigned by PangoLEARN [9]. During the enrollment period, the most prevalent SARS-CoV-2 variants were Alpha (B. Of the 37 RT-PCR positive specimens collected, the cycle threshold (Ct) values ranged from 16 to 38, with a median value of 28 (Table 3). The median Ct value for samples with concordant results was 24, compared to a median Ct value of 36 for discordant samples (p<0.001) ( Throughout the enrollment period, Southwest District Health clinic saw a median (IQR) RT-PCR turnaround time of 1 (0-2) days. During the same time period, the region as a whole saw a median wait time of approximately 2 (1-4) days for RT-PCR results to reach the patient. The discrepancy in turnaround time is likely due to the Southwest Health District Clinic
PLOS ONE
Evaluation of the Abbott BinaxNOW™ COVID-19 Test Ag Card for rapid detection of SARS-CoV-2 infection having immediate access to Electronic Lab Reports (ELRs) while many private clinics throughout the region relied on private web based patient portals or mailed RT-PCR results to inform patients of their results. Participant enrollment varied during the study period (Fig 2). Study enrollment paused between February 6, 2021 and March 25, 2021 (MMWR week 6-12) due to leadership staff turnover, new training and implementation, as well as prioritization of vaccination efforts. During the study period, the demand for COVID-19 testing across the health district declined over time.
Discussion
The BinaxNOW™ COVID-19 Rapid Antigen Test was evaluated in people seeking free testing conducted by a local public health district in Idaho serving a rural population. Results were compared to the gold standard of SARS-CoV-2 detection by RT-PCR. The prevalence of COVID-19 among all study participants was 17.3%, indicative of high levels of community transmission of SARS-CoV-2 during the study time frame. Overall, results were highly concordant (94.4%) and no false positives were identified. Sensitivity of the BinaxNOW™ COVID-19 Test was 67.6% overall, and lower for asymptomatic participants. Results from our study were Variant of Concern (VOC) is classified by the CDC as a variant for which there is evidence of increased transmissibility, more severe disease, significant reduction in neutralization by antibodies generated during previous infection or vaccination, reduced effectiveness of treatments or vaccines, or diagnostic detection failures [11]. 2 Variant of Interest (VOI) is classified by the CDC as a variant with specific genetic markers that have been associated with changes to receptor binding, reduced neutralization by antibodies generated against previous infection or vaccination, reduced efficacy of treatments, potential diagnostic impact, or predicted increase in transmissibility or disease severity [11]. https://doi.org/10.1371/journal.pone.0260862.t002
PLOS ONE
Evaluation of the Abbott BinaxNOW™ COVID-19 Test Ag Card for rapid detection of SARS-CoV-2 infection consistent with larger studies assessing the performance of the BinaxNOW™ COVID-19 Test.
A study conducted by in Pima County, Arizona in November of 2020 found similar results, with a 64.2% sensitivity for symptomatic participants, 35.8% sensitivity for asymptomatic participants, and a 100% specificity for all participants regardless of symptom status [12]. Another study evaluating drive-through COVID-19 testing in Massachusetts demonstrated the Binax-NOW™ COVID-19 Test to have a similar specificity of 100% for symptomatic adults and 99.6% for asymptomatic adults. However, a higher sensitivity for symptomatic adults (96.5%), compared to asymptomatic adults (70.2%) was observed. The Massachusetts study also demonstrated that sensitivity was significantly lower when BinaxNOW™ COVID-19 Tests were
PLOS ONE
Evaluation of the Abbott BinaxNOW™ COVID-19 Test Ag Card for rapid detection of SARS-CoV-2 infection conducted at temperatures below the manufacturer's recommended range [13]. Although BinaxNOW™ COVID-19 Tests were conducted indoors for our study, samples were collected in an outdoor drive through tent; the indoor testing area was close to a doorway leading to the outdoor testing site, and temperatures, from sample collection to test results, were not monitored. One objective of our study was to inform recommendations on whether confirmatory testing by RT-PCR is needed following rapid antigen testing with the BinaxNOW™ COVID-19 Test. The lack of false positives observed in this study, along with specificity above 99% reported from other studies, indicate that positive results by BinaxNOW™ COVID-19 Test may not require confirmation by RT-PCR. However, lower sensitivity and the possibility of a false negative result strongly supports that confirmatory testing should be recommended in some circumstances, especially when the results of the BinaxNOW™ COVID-19 Test is inconsistent with the clinical context (e.g. among symptomatic individuals or those with a known exposure to someone with COVID-19), as per current CDC recommendations [14].
The main benefits of rapid antigen testing are its accessibility, and that it can be conducted outside of a laboratory setting with rapid availability of results at the point of care. In our study, the median turnaround time for RT-PCR results was approximately 24 hours, though early in the pandemic, rural areas of Idaho experienced significant turnaround times for RT-PCR results, with most results taking more than a week [15]. The BinaxNOW™ COVID-19 Test might be particularly useful in rural settings without local RT-PCR capacity.
Among symptomatic participants with false negatives from the BinaxNOW™ COVID-19 Test, we saw a greater variability in days since symptom onset (see Fig 1), however differences in days since symptom onset did not differ between discordant and concordant samples, likely due to a small number of symptomatic participants with discordant results (n = 10). Because viral loads are generally higher at symptom onset and decline over time, the BinaxNOW™ COVID-19 Test might be less able to detect SARS-CoV-2 in individuals with specimen collection several days after symptom onset. Additionally, the BinaxNOW™ COVID-19 Test was less sensitive among asymptomatic participants, whereas viral loads have been shown to be comparable between symptomatic and asymptomatic COVID-19 patients [16]. In this study, 62.1% of asymptomatic participants did not report a known exposure in the 14 days prior to sample collection. Since asymptomatic participants without a known exposure to COVID-19 are unable to define a timeline of their exposure, we can assume there is greater variability in the days since exposure, and consequently, will have a greater variability in SARS-CoV-2 viral load [17]. Generally, rapid antigen tests perform better when viral loads are high, which is also when people with SARS-CoV-2 are more likely to be infectious.
Demand for COVID-19 testing declined rapidly during our enrollment period. There was a decline in SARS-CoV-2 RT-PCR testing demand throughout Idaho as well as a decline in new participants. We can attribute this to a decline in COVID-19 prevalence throughout the region and an increase in COVID-19 immunization during this time.
There were several limitations to this study, including small sample size, which may have affected analysis of the performance of the BinaxNOW™ COVID-19 Test, compared to RT-PCR. Participants were identified among people seeking testing from the local public health district and they may not be representative of the community as a whole. An additional limitation may have been the temperature variability of the testing location, which may have affected performance of the BinaxNOW™ COVID-19 Test, though the primary methodology of COVID-19 testing (rapid antigen and RT-PCR) in the United States mirrors this process, with samples being collected from patients outdoors or within their personal vehicle and this study may inform these practices.
Rapid antigen testing, including the BinaxNOW™ COVID-19 Test, is an important tool in detecting SARS-CoV-2, particularly in rural areas with limited access to high level laboratories that are capable of conducting SARS-CoV-2 RT-PCR. While the BinaxNOW™ COVID-19 Test is less sensitive than SARS-CoV-2 RT-PCR, it is highly specific, and can be used as a tool to rapidly detect SARS-CoV-2 in community settings. Community testing programs run by local health departments using rapid antigen tests can improve access to COVID-19 testing and to help limit the spread of SARS-CoV-2. As additional waves of novel coronavirus variants emerge, rapid antigen testing may continue to serve as a valuable alternative testing modality to RT-PCR tests.
Supporting information S1 Table. Characteristics of study participants and SARS-CoV-2 detection method and result. (DOCX) | 2021-12-04T05:11:26.985Z | 2021-12-02T00:00:00.000 | {
"year": 2021,
"sha1": "6c3a55f608eaf2ca7b2b1b5d5c0f8ff216326f67",
"oa_license": "CCBY",
"oa_url": "https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0260862&type=printable",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "6c3a55f608eaf2ca7b2b1b5d5c0f8ff216326f67",
"s2fieldsofstudy": [
"Medicine",
"Environmental Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
13500976 | pes2o/s2orc | v3-fos-license | Age-specific gender differences in early mortality following ST-segment elevation myocardial infarction in China
Objective To assess whether younger, but not older, women in China have higher in-hospital mortality following ST-Segment Elevation Myocardial Infarction (STEMI) compared with men, and whether this relationship varied over the last decade or across rural/urban areas. Methods We analysed a nationally representative sample of 11 986 patients with STEMI from 162 Chinese hospitals in 2001, 2006 and 2011, in the China PEACE-Retrospective AMI Study and compared in-hospital mortality between women and men with gender–age interactions in multivariable models. Results The overall in-hospital mortality rate was higher in women compared with men (17.2% vs 9.1%, p<0.0001; unadjusted OR 2.07, 95% CI 1.85 to 2.33). The unadjusted OR for mortality in women, compared with men, was 2.20 (95% CI 1.59 to 3.04), 2.21 (95% CI 1.74 to 2.79), 1.37 (95% CI 1.15 to 1.65) and 1.25 (95% CI 0.97 to 1.63) for ages <60, 60–69, 70–79 and ≥80 years, respectively. After adjustment for patient characteristics, hospital characteristics and year of study, the OR for mortality comparing women with men was 1.69 (95% CI 1.01 to 2.83), 1.64 (95% CI 1.24 to 2.19), 1.15 (95% CI 0.90 to 1.46) and 0.82 (95% CI 0.60 to 1.11) for ages <60, 60–69, 70–79 and ≥80 years, respectively. The gender–age interaction for mortality was statistically significant (p=0.009), even after adjustment for a wide range of confounders, and did not vary over time or across rural/urban areas. Conclusions Among a Chinese population with STEMI, gender differences in early mortality were age-dependent and greatest in the younger groups <70 years of age. Trial registration number http://www.clinicaltrials.gov (NCT01624883).
Background
We intended study hospitals to reflect diverse sites of care in China. As hospital volumes and clinical capacities differ between urban and rural areas as well among the 3 official economicgeographic regions of China, we separately identified hospitals in 5 strata: Eastern-rural, Central-rural, Western-rural, Eastern-urban, and Central/Western-urban regions. We considered an area urban if it is part of a downtown or suburban area within a direct-controlled municipality (Beijing, Tianjin, Shanghai, Chongqing) or 1 of 283 prefectural-level cities. We considered surrounding county-level regions, including counties and county-level cities, to be rural. Within this framework, China is composed of 287 urban regions and 2010 rural regions. We considered Central and Western urban regions together given their similar per capita income and health services capacity as shown below: We identified cases for study inclusion using a stratified 2-stage cluster sampling design. In the first stage, we identified hospitals using a simple random sampling procedure within each of the 5 study strata. In the 3 rural strata, the sampling framework consisted of the central hospital in each of the predefined rural regions (2010 central hospitals in 2010 rural regions). Within each rural region, the central hospital is the largest general hospital with the greatest clinical capacity for treating acute illness. In each of the 2 urban strata, the sampling framework consisted of the highest-level hospitals in each of the predefined urban regions (833 hospitals in 287 urban regions). Hospital level is officially defined by the Chinese government based on clinical resource capacity. For example, secondary hospitals have at least 100 inpatient beds and the capacity to provide acute medical care and preventive care services to populations of at least 100,000, while tertiary hospitals are large referral centers in provincial capitals and major cities.
Population, Economy, and Hospitals in Different Geographic Strata of Mainland China
We excluded military hospitals, prison hospitals, specialized hospitals without a cardiovascular disease division, and traditional Chinese medicine hospitals. Since the majority of hospitals in China are publicly owned and administered, hospital closure is rare, and hospital number has remained stable over the past decade. We therefore decided to select representative hospitals from 2011 to reflect current practices and trace this cohort backward to 2006 and 2001 to describe temporal trends.
In the second stage, we drew cases based on the local hospital database for patients with acute myocardial infarction at each sampled hospital. We ordered each hospital's list of eligible cases by date of admission and selected cases using systematic random sampling with equal probabilities. We selected a case at random, after which we selected every k th case based on sample size requirements, where k is the sampling interval.
In each of the 5 study strata, we determined the sample size required to achieve a 2% precision for describing the primary outcome, in-hospital mortality, which we had estimated to be approximately 9% in urban hospitals and 7% in rural county-level hospitals.
The following Equation 1 can be used to define the sample size required (n) for a given proportion of the primary outcome (P), desired precision (d), and specific choice of α.
Equation 1: However, because random cases sampled within the same hospital are likely to be more similar to one another than to random cases from another hospital, the effective sample size is reduced. Consequently, a design effect adjustment should be introduced as follows: Where the design effect (deff) is given by where is the intraclass correlation for the statistic in question and ′ is the average number of sampled cases within each hospital. ′ is also known as the cluster size.
With this framework in mind, to achieve a precision of 2% with an α of 0.05 in each of the 3 rural strata, assuming an intraclass correlation of 0.02 and design effect of 1.8, we would need to sample 1150 medical records among hospitals with an average cluster size of 40. Analogously, to achieve a precision of 2% with an α of 0.05 in each of the 2 urban strata, assuming an intraclass correlation of 0.02 and design effect of 2.2, we would need to sample 1750 medical records among hospitals with an average cluster size of 60. These cluster sizes in rural and urban settings appeared reasonable based upon our previous survey of treatment for acute coronary syndromes at more than 1000 hospitals in 2010, which demonstrated that the median volume of hospitalization for acute myocardial infarction was approximately 180 cases per year in urban hospitals and 95 cases per year in rural county-level hospitals. Assuming an anticipated participation rate of 85% among selected hospitals, we approached 35 hospitals for participation in each stratum for a total of 175 hospitals (70 urban and 105 rural). We doubled cluster sizes for 2011 to improve precision in the description of hospital-level treatment patterns and outcomes.
Consequently, the total expected sample volume with the above assumptions was approximately 6,950 cases in 2001, 6,950 cases in 2006, and 13,900 cases in 2011.
CHINA PEACE-RETROSPECTIVE AMI STUDY QUALITY ASSURANCE AND QUALITY CONTROL STRATEGIES IN MEDICAL RECORD SAMPLING
As the China PEACE Retrospective Study of Acute Myocardial Infarction was designed to study a nationally representative hospital cohort, we selected hospitals based on random sampling rather than previous collaboration or longstanding experience with retrospective data collection. As we anticipated that many hospitals would have little previous experience with clinical research, we provided participating sites with substantial support to ensure adherence with multiple quality control strategies for identifying the universe of hospitalizations for acute myocardial infarction and the subset sampled for the China PEACE-Retrospective AMI Study.
For all participating hospitals, we first held a local investigator meeting to provide in-depth information on study design and operating procedures. We trained sites on how to identify their universe of hospitalizations for acute myocardial infarction. We specified 3 separate approaches. Our first preference was that sites query their electronic database of hospitalizations for patients discharged with a principal discharge diagnosis of acute myocardial infarction using ICD-9 codes 410 or ICD-10 codes I21. If such a database was not available, we required that site coordinators manually search the written hospitalization log in the hospital archiving office for hospitalizations for acute myocardial infarction. Site coordinators reviewed the original medical record in cases where the diagnosis of acute myocardial infarction was uncertain. Finally, as the least preferred option, we asked that study coordinators manually search the written log of hospitalizations of each hospital ward for cases for acute myocardial infarction. Site coordinators reviewed the original medical records in cases of uncertain diagnosis.
To verify compliance with the search strategy, research staff from the study coordinating center visited 46 study sites to repeat the search process and confirm that the list of hospitalizations with acute myocardial infarction was complete. These 46 sites provided approximately 60% of all hospitalizations for acute myocardial infarction from which we sampled cases for the PEACE Retrospective AMI Study.
After we drew case samples at each hospital using systematic random sampling and assigned each record a unique study ID, we required local investigators to gather the original record, assign it its study ID, scan it, and transmit the scanned copy to the coordinating center. Upon receipt of the scanned record, coordinating center staff verified the accuracy of the study ID and ensured that the record itself was complete and legible. To facilitate this process, the coordinating center provided each study site with a high-speed scanner. In addition, coordinating center staff provided on-site assistance for almost 50 study sites that provided approximately 50% of sampled cases.
To ensure transparency in all sampling performed by the NCCD, we have recorded all sampling procedures including the contents of the sampling framework database that contains all eligible cases for sampling in their predefined sequence and the seeds used in random number generation.
We assigned the Front Page and Laboratory Results sections of the medical records to qualified non-medical abstractors, as these sections contain almost exclusively hard data elements. We performed a second abstraction of all records abstracted by non-medical abstractors to ensure accuracy.
Abstractors with formal medical training abstracted all sections of the medical chart containing soft data elements. These include the admission note, daily progress notes, procedure notes, diagnostic testing reports, medication administration record, physician orders, nursing notes, and discharge summary. We randomly audited approximately 5% of the abstracted records. If the records were not abstracted with 98% accuracy, all medical records in the audited batch were considered unqualified and were re-reviewed by a different abstractor. Discrepancies in abstraction were resolved by review of the original medical record.
To minimize abstraction errors, abstractors started by abstracting only printed medical records. After gaining experience, these individuals were allowed to begin abstracting hand-written records. In addition, a physician was always present in the room with abstractors or was available online to answer questions and address areas of concern as they arise. Common problems have led to updates and further customization of the data dictionary and web-based data management program built for the China PEACE-Retrospective AMI Study into which data were directly entered. This computer program has been customized to expedite the identification of medications that may have more than one trade name. Furthermore, medical records belonging to the same hospital and year were assigned to a broad group of reviewers to avoid potential residual disparities in quality among different abstractors
Data Management and Cleaning
Ongoing data cleaning is performed in a systematic manner. Data is regularly queried for invalid and illogical values as well as for duplicate record entry. Invalid values are identified as outliers in continuous data distributions. Duplicate records are identified by the presence of similar study identification numbers, hospital identification numbers, medical record identification numbers, and the date of discharge. Once a potential error is found, the issue is resolved after tracing and reviewing the relevant records.
Frequent sources of error result in additional intensive training for abstractors and supervisors, if necessary. The study data dictionary is updated, as needed.
Study ID
Coding Instructions: Indicate the identification number for the study.
Medical Record Number
Coding Instructions: Indicate the identification number for this record.
Hospital Number
Coding Instructions: Indicate the identification number for this hospital. Urban residents/worker medical insurance: the medical insurance provided by the government for people who live in urban areas; also is the type of insurance provided by a company or factory for their employees. Farmer medical insurance: a new medical insurance type funded by the government for farmers. Business insurance: Insurance paid for by individuals to insurance companies. Comprehensive arrangement for serious disease: Insurance for people living in urban areas who are employed by a company or factory. Also may apply to persons who are retired. Specifications vary by city. Covered diseases are different than that covered by urban residents/worker medical insurance. Poverty assistance insurance: Medical aid provided by the government for povertystricken individuals to cover a portion of expenses Free medical care: The government provides free medical treatment for a specific group of people such as for disabled members of the armed forces. Self-pay/no insurance: Indicated for those without medical insurance or for whom certain diseases are not covered by insurance. Patient will pay for these costs. Patients may have more than one insurance type at a given time. Abstractor should select all that apply Source: Definition per China team.
Height
Coding Instructions: Indicate the patient's height in centimeters Note(s): Measurement from the transferring facility is acceptable Target Value: The first value between arrival at this facility and discharge Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Admission Weight
Coding Instructions: Indicate he patient's weight in kilograms Note(s): Measurement from the transferring facility is acceptable Target Value: The first value between first medical contact and discharge Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Marital Status
Coding Instructions: Indicate the marital status of the patient Target Value: N/A Selections: Married Single Divorced Widowed Unrecorded Supporting Definitions: Married -A formal union with a person of the opposite sex, typically as recognized by law, by which they are considered husband and wife. Single-Never been married. Divorced -Previous marriage has been legally dissolved by a court or another legal body. Widowed-A person who has lost his/her spouse to death and has not remarried.
Time of Symptom Onset (precise)
Coding Instructions: Indicate the time the patient first noted ischemic symptoms that lasted greater than or equal to 10 minutes. If an estimated symptom onset time is recorded, code 'symptom onset time estimated'. Target Value: The first time symptoms lasted greater than 10 minutes on the symptom onset date Selections: (none) Supporting Definitions: (none) Note: If the symptom onset time is not specified in the medical record, it may be recorded as 0700 for morning; 1200 for lunchtime; 1500 for afternoon; 1800 for dinnertime; 2200 for evening and 0300 if awakened from sleep. Source: Adapted from AHA Get with the Guidelines ACTION registry
CPR/Chest Compressions Prior to Arrival
Coding Instructions: Indicate if cardiopulmonary resuscitation (CPR) in any form, including chest compression, was performed for the patient, prior to arrival Target Value: Any occurrence between first medical contact and arrival to this facility Selections: (1) No (2) Yes Supporting Definitions: CPR comprises of a series of interventions performed for patients with sudden cardiac arrest in order to restore the perfusion and oxygenation of vital organs. Chest or abdominal compression, ascertainment of patent airways, rescue breathing, as well as electrical cardioversion and defibrillation are the cornerstones of CPR. Medications (including epinephrine, lidocaine, amiodarone, and atropine) may or may not be used during CPR.
Cardiac Arrest Prior to Admission
Coding Instructions: Indicate if the patient had an episode of cardiac arrest prior to admission to this facility. Note(s): Evaluated by ED personnel and either (1) received attempts at external defibrillation or chest compressions or (2) were pulseless but did not receive attempts to defibrillate or cardiopulmonary resuscitation (CPR). Target Value: Any occurrence prior to admission to this facility. Selections: (1) No (2) Yes Supporting Definitions: 'Sudden' cardiac arrest is the sudden cessation of cardiac activity so that the victim becomes unresponsive, with no normal breathing and no signs of circulation. If corrective measures are not taken rapidly, this condition progresses to sudden death. Cardiac arrest should be used to signify an event as described above that is reversed, usually by CPR, and/or defibrillation or cardioversion, or cardiac pacing. Sudden cardiac arrest is not the same as sudden cardiac death. Sudden cardiac death describes a fatal event. Source: ACC/AHA/HRS 2006 Key Data Elements and Definitions for Electrophysiological Studies and Procedures; adapted from AHA Get with the Guidelines ACTION registry
Antiplatelet Therapy in the Emergency Department Prior to Admission
Coding Instructions: Indicate whether the patient was given aspirin, clopidogrel, ticlopidine or other antiplatelet therapy by a health provider (EMS, transferring hospital, etc.) in the emergency department prior to admission. Target Value: Any occurrence between arrival to the ED and admission Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Fibrinolysis in the Emergency Department Prior to Admission
Coding Instructions: Indicate whether the patient was given a fibrinolytic by a health provider (EMS, transferring hospital, etc.) in the emergency department prior to admission. Target Value: Any occurrence between arrival to the ED and admission Selections: (1) No (2) Yes. If yes, please specify the name of the fibrinolytic agent that was used. Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Glucose Insulin Potassium (GIK) Solution in the Emergency Department Prior to Admission
Coding Instructions: Indicate whether the patient was given GIK by a health provider (EMS, transferring hospital, etc.) in the emergency department prior to admission. Target Value: Any occurrence between arrival to the ED and admission Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Definition per CORE team
CPR/Chest Compressions in the Emergency Department Prior to Admission
Coding Instructions: Indicate if cardiopulmonary resuscitation (CPR) in any form, including chest compression, was performed for the patient in the emergency department prior to admission Target Value: Any occurrence between arrival to the ED and admission Selections: (1) No (2) Yes Supporting Definitions: CPR comprises of a series of interventions performed for patients with sudden cardiac arrest in order to restore the perfusion and oxygenation of vital organs. Chest or abdominal compression, ascertainment of patent airways, rescue breathing, as well as electrical cardioversion and defibrillation are the cornerstones of CPR. Medications (including epinephrine, lidocaine, amiodarone, and atropine) may or may not be used during CPR.
External Defibrillation in Emergency Department Prior to Admission
Coding Instructions: Indicate if there is documentation of electrical defibrillation in the emergency department prior to admission Target Value: Any occurrence between arrival to the ED and admission Selections: (1) (2) Yes Supporting Definitions: Heart failure is defined as physician documentation or report of any of the following clinical symptoms of heart failure, described as: unusual dyspnea on light exertion, recurrent dyspnea occurring in the supine position, fluid retention; or the description of rales, jugular venous distension, pulmonary edema on physical exam, or pulmonary edema on chest x-ray presumed to be due to cardiac dysfunction. A low ejection fraction without clinical evidence of heart failure does not qualify as heart failure. Source: Acute Coronary Syndromes Data Standards (JACC 2001(JACC 38: 2114, The Society of Thoracic Surgeons; adapted from AHA Get with the Guidelines ACTION registry
Cardiogenic Shock on Presentation to This Facility
Coding Instructions: Indicate if the patient was in a state of cardiogenic shock on presentation to this facility. Note(s): Transient episodes of hypotension reversed with IV fluid or atropine do not constitute cardiogenic shock. The hemodynamic compromise (with or without extraordinary supportive therapy) must persist for at least 30 minutes. Target Value: Any occurrence between first medical contact and arrival at this facility Selections: (1) No (2) Yes Supporting Definitions: Cardiogenic shock is defined as a sustained (>30 minutes) episode of systolic blood pressure <90 mm Hg, and/or cardiac index <2.2 L/min/m2 determined to be secondary to cardiac dysfunction, and/or the requirement for parenteral inotropic or vasopressor agents or mechanical support (e.g., IABP, extracorporeal circulation, ventricular assist devices) to maintain blood pressure and cardiac index above those specified levels. c. Development of pathological Q-waves in 2 or more contiguous leads in the ECG (or equivalent findings for true posterior MI). d. Imaging evidence of new loss of viable myocardium or new regional wall motion abnormality. e. Documentation in the medical record of the diagnosis of acute myocardial infarction based on the cardiac biomarker pattern in the absence of any items enumerated in a-d due to conditions that may mask their appearance (e.g., perioperative infarct when the patient cannot report ischemic symptoms; baseline left bundle branch block or ventricular pacing). 2. ECG changes associated with prior myocardial infarction can include the following (with or without prior symptoms): a. Any Q-wave in leads V2-V3 >=0.02 seconds or QS complex in leads V2 and V3. b. Q-wave >=0.03 seconds and >=0.1 mV deep or QS complex in leads I, II, aVL, aVF, or V4-V6 in any two leads of a contiguous lead grouping (I, aVL, V6; V4-V6; II, III, and aVF). c. R-wave >=0.04 seconds in V1-V2 and R/S >=1 with a concordant positive Twave in the absence of a conduction defect. 3. Imaging evidence of a region with new loss of viable myocardium at rest in the absence of a non-ischemic cause. This can be manifest as: a. Echocardiographic, CT, MR, ventriculographic or nuclear imaging evidence of left ventricular thinning or scarring and failure to contract appropriately (i.e., hypokinesis, akinesis, or dyskinesis). b. Fixed (non-reversible) perfusion defects on nuclear radioisotope imaging (e.g., MIBI, thallium). 4. Medical record documentation of prior myocardial infarction. Source: Joint ESC-ACC-AHA-WHF 2007 Task Force Consensus Document "Universal Definition of Myocardial Infarction"; adapted from AHA Get with the Guidelines ACTION registry
History of Heart Failure
Coding Instructions: Indicate if there is a previous history of heart failure. Note(s): A previous hospital admission with principal diagnosis of heart failure is considered evidence of heart failure history. Target Value: Any occurrence between birth and arrival at this facility Selection: (1) No (2) Yes Supporting Definitions: Heart failure is defined as physician documentation or report of any of the following clinical symptoms of heart failure described as unusual dyspnea on light exertion, recurrent dyspnea occurring in the supine position, fluid retention; or the description of rales, jugular venous distension, pulmonary edema on physical exam, or pulmonary edema on chest x-ray presumed to be cardiac dysfunction. A low ejection fraction without clinical evidence of heart failure does not qualify as heart failure.
Source: Acute Coronary Syndromes Data Standards (JACC 2001(JACC 38: 2114, The Society of Thoracic Surgeons; adapted from AHA Get with the Guidelines ACTION registry
History of Percutaneous Coronary Intervention
Coding Instructions: Indicate if the patient had a previous percutaneous coronary intervention (PCI) of any type (balloon angioplasty, stent or other). Note (s): Timeframe does NOT include the current admission. Target Value: Any occurrence between birth and arrival at this facility Selections: (1) No (2) Yes Supporting Definitions: A percutaneous coronary intervention (PCI) is the placement of an angioplasty guide wire, balloon, or other device (e.g. stent, atherectomy, brachytherapy, or thrombectomy catheter) into a native coronary artery or coronary artery bypass graft for the purpose of mechanical coronary revascularization. Source: Adapted from AHA Get with the Guidelines ACTION registry
History of Atrial Fibrillation or Flutter
Coding Instructions: Indicate if there is a previous history of atrial fibrillation or flutter Note(s): Code "No" if patient was first diagnosed with atrial fibrillation or flutter after reperfusion during this admission. If there is no prior documentation of atrial arrhythmias, it is acceptable to code "No" Target Value: Any occurrence between birth and arrival at this facility Selections: (1)
History of Chronic Lung Disease
Coding Instructions: Indicate if the patient has a history of chronic lung disease. A history of chronic inhalation reactive disease (asbestosis, mesothelioma, black lung disease or pneumoconiosis) qualifies as chronic lung disease. Radiation induced pneumonitis or radiation fibrosis also qualifies as chronic lung disease. A history of atelectasis is a transient condition and does not qualify. Notes: Code "No" if patient was first diagnosed with chronic lung disease after reperfusion during this admission. Target Value: Any occurrence between birth and arrival at this facility Selections: (1) No (2) Yes Supporting Definitions: Chronic lung disease can include patients with chronic obstructive pulmonary disease, chronic bronchitis, or emphysema. It can also include a patient who is currently being chronically treated with inhaled or oral pharmacological therapy (e.g., betaadrenergic agonist, anti-inflammatory agent, leukotriene receptor antagonist, or steroid). Patients with asthma or seasonal allergies are not considered to have chronic lung disease. Source: Adapted from AHA Get with the Guidelines ACTION registry (2) Yes Supporting Definitions: Mark "Yes" if the patient received any surgery, or any chemotherapy, or any radiotherapy, or any palliative or hospice services because of a cancer diagnosis, or any combination of these, related to a diagnosed cancer. Source: Definition per CORE team
Major Surgery Within the Past Four Weeks
Coding Instructions: Indicate if the patient has a history of major surgery within the past four weeks prior to the index admission Target Value: Any occurrence within four weeks prior to index admission. Selections: (1) No (2) Yes Supporting Definitions: Major surgery is defined as any surgery that requires general anesthesia, or that involves opening the great cavities of body (cranium, chest, abdomen, pelvic cavity), or those in which severe bleeding is likely to occur, or that is likely to be life-threatening. Examples include: 1. Cranial surgery (including decompression following a bleed, tumor removal, or others) 2. Thoracic surgery (including coronary bypass surgery, heart valve surgery, other heart surgery, any lung surgery, any mediastinal surgery) 3. Abdominal surgery (any surgery related to organs in the abdominal cavity, including the stomach, bowels, liver, kidneys, or others) 4. Pelvic surgery (including Cesarean section surgery, vaginal hysterectomy, cystocele surgery, rectocele surgery, or others) 5. Any surgery not meeting the above criteria but being performed under general anesthesia 6. Any surgery not meeting the above criteria but lasting more than 60 minutes (e.g. orthopedic surgery lasting for >60 minutes, other types of surgery lasting for >60 minutes).
The definition is meant to be distinct from minor surgery. Minor surgery would be defined as any invasive procedure in which only the skin or mucus membranes and connective tissues are resected.
Cardiac CT Ejection Fraction (EF) Value
Coding Instructions: Indicate the value of EF in percent documented in the CT scan report. Target Value: Any occurrence between arrival at admitting hospital and discharge Selections: (none) Supporting Definition: The left ventricular ejection fraction is the percentage of the blood emptied from the left ventricle at the end of the contraction. The left ventricular ejection fraction can be accessed via invasive (i.e. LV gram) or noninvasive (i.e. Echo, MR, CT or Nuclear) testing. Source: ACC Clinical Data Standards, the Society of Thoracic Surgeons; adapted from VIRGO registry
Multi-Gated Acquisition (MUGA) Scan
Coding Instructions: Indicate whether the patient underwent MUGA or not Target Value: Any occurrence between arrival at admitting hospital and discharge Selections: (1)
Initial Creatine Kinase (CK) Value
Coding Instructions: Indicate the value of the initial CK. Notes: Initial CK level corresponds to the first sample obtained within the first 24 hours of care. It may also have been collected at the transferring hospital. Target Value: Any occurrence between first medical contact and 24 hours after arrival at first facility Selections: N/A Supporting Definition: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Initial CK Upper Limit of Normal (ULN)
Coding Instructions: Indicate the ULN of the initial CK sample. Note(s): If a range is given for ULN values, record the highest number in the range. Examples: If the reference range given is 0.0-1.5, record ULN as 1.5. If the reference range given is < 1.5, record ULN as 1.5 as well. (1) IU/L (2) % (3) (mg/mL)/IU (4) ng/mL
MEDICATIONS ADMINISTERED
Note: All the information about medications was entered into a central medication database. For each medication, we abstracted name, dose, and route of administration. The following definitions describe specific pre-defined questions pertinent to medication administration for acute myocardial infarction.
Aspirin in the First 24 Hours
Coding Instructions: Indicate if aspirin was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g. transferring facility or EMS). Target Value: Any occurrence between 24 hours before first medical contact and 24 hours after first medical contact Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from VIRGO registry
Documented Reasons for Non-prescription of Aspirin in the First 24 Hours
Coding Instructions: If the medication was not prescribed, indicate whether the reason was documented.
Target Value: Any documentation in the chart regarding contraindications for the prescription of the medication in the first 24 hours. Selections: (1) No (2) Yes -Allergy (3) Yes -Other (specify) Source: Definition per CORE team
Aspirin in the First 24 Hours-Start Date
Coding Instructions: Indicate the date that aspirin was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact. Selections: (none) Supporting Definitions: (none) Source: Adapted from VIRGO registry
Aspirin in the First 24 Hours-Start Time
Coding Instructions: Indicate the time that aspirin was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact. Selections: (none) Supporting Definitions: (none) Source: Adapted from VIRGO registry
Aspirin in the First 24 Hours-Dose
Coding Instructions: Indicate the cumulative dose of aspirin. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact. Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Aspirin at Discharge
Coding Instructions: Indicate if aspirin was continued or prescribed. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: Any occurrence on discharge Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Aspirin at Discharge-Dose
Coding Instructions: Indicate the dose of aspirin prescribed at discharge. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: The highest value on discharge Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Documented Reasons for Non-Prescription of Aspirin at Discharge
Coding Instructions: If the medication was not prescribed, indicate whether the reason was documented. Target Value: Any documentation in the chart regarding contraindications for the prescription of the medication at Discharge. Selections: (1) No (2) Yes -Allergy (3) Yes -Other (specify) Source: Definition per CORE team
Warfarin during the First 24 Hours
Coding Instructions: Indicate if warfarin was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g. transferring facility or EMS). Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact. Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from VIRGO registry
Warfarin at Discharge
Coding Instructions: Indicate if Warfarin was continued or prescribed. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: Any occurrence on discharge Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Clopidogrel in the First 24 Hours
Coding Instructions: Indicate if clopidogrel was administered, regardless of location of care (e.g. transferring facility or EMS). Target Value: Any occurrence between 24 hours before first medical contact and 24 hours after first medical contact transferring facility or EMS). Target Value: Any occurrence between 24 hours before first medical contact and 24 hours after first medical contact Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Ticlopidine in the First 24 Hours-Start Date
Coding Instructions: Indicate the date the initial dose of ticlopidine was administered, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Ticlopidine in the First 24 Hours-Start Time
Coding Instructions: Indicate the time that ticlopidine was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Ticlopidine in the First 24 Hours-Dose
Coding Instructions: Indicate the cumulative dose of ticlopidine. Target Value: The cumulative dose between first medical contact and 24 hours after first medical contact Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Ticlopidine at Discharge
Coding Instructions: Indicate if ticlopidine was continued or prescribed at discharge. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: Any occurrence on discharge Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Ticlopidine at Discharge -Dose
Coding Instructions: Indicate the dose of ticlopidine prescribed at discharge. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: The highest value on discharge Selections: (none) Supporting Definitions: (none) "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: The highest value on discharge Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Fibrinolysis in the First 24 Hours
Coding Instructions: Indicate if fibrinolytics were administered, regardless of location of care (e.g. transferring facility or EMS). Target Value: Any occurrence between 24 hours before first medical contact and 24 hours after first medical contact Selections: (1) No (2) Yes Supporting Definitions: If fibrinolytics are used, please name the fibrinolytic agent, the Bolus dose (IU), the Second Bolus dose (IU/hour), or Maintenance dose (IU/hour), where applicable, or mark "Unknown" if doses are not documented. Source: Adapted from VIRGO registry
Refusal of Fibrinolysis
Coding Instructions: Indicate if there is documentation of patient refusal for fibrinolysis Target Value: Any occurrence from the beginning of hospital stay to discharge Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Definition per CORE team
Fibrinolysis in the First 24 Hours -Start Date
Coding Instructions: Indicate the date the initial dose of fibrinolytic administered, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Fibrinolysis in the First 24 Hours -Start Time
Coding Instructions: Indicate the time that a fibrinolytic was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact. Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Documented Reasons for Non-prescription of Fibrinolysis in the First 24 Hours
Coding Instructions: If the medication was not prescribed, indicate whether the reason was documented. Target Value: Any documentation in the chart regarding contraindications for the prescription of the medication in the first 24 hours. Selections: (1) No (2) Yes -Allergy (3) Yes -Other (specify) Source: Definition per CORE team
Fibrinolysis after the First 24 Hours
Coding Instructions: Indicate if fibrinolytics were administered, regardless of location of care (e.g. transferring facility or EMS). Target Value: Any occurrence after the first 24 hours from the first medical contact Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Fibrinolysis after the First 24 Hours-Start Date
Coding Instructions: Indicate the date the initial dose of fibrinolytic was administered, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value after the first 24 hours from the first medical contact Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Fibrinolysis after the First 24 Hours -Start Time
Coding Instructions: Indicate the time that a fibrinolytic was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value after the first 24 hours from the first medical contact Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Anticoagulant Use
Coding Instructions: First, please indicate if an anticoagulant is being used or not (YES/ NO). Next, please specify the type of anticoagulant(s) administered, including: coumarins, unfractionated heparin, low molecular weight heparins (enoxaparin, dalteparin, or fondaparinux), direct thrombin inhibitors (such as lepirudin, bivalirudin, and argatroban). Please indicate the start date and time of first administration of the anticoagulant medication(s). For anticoagulants that are used in the first 24 hours from first medical contact, depending on the type of the anticoagulant, please also specify: (1) Was a bolus used? If yes, provide the date, time and dose.
(2) Was an infusion used? If yes, provide the date, time and dose.
(3) Was a subcutaneous injection used? If yes, provide the date, time and dose.
(4) Provide the number of daily injections where applicable. Source: Adapted from VIRGO registry
Glycoprotein IIb/IIIa Inhibitor Use
Coding Instructions: First, please indicate if glycoprotein IIb/IIIa inhibitors were used (YES/ NO). Next please specify the data and time of first administration and the name of drug being used: abciximab, tirofiban or eptifibatide. If used during the first 24 hours from the time of first medical contact, please indicate the date and time, as well as bolus and maintenance doses. Source: Adapted from VIRGO registry
Beta-blockers in the First 24 Hours
Coding Instructions: Indicate if a beta-blocker was administered, regardless of location of care (e.g., transferring facility or EMS).
Target Value: Any occurrence between 24 hours before first medical contact and 24 hours after first medical contact Selections: (1) Code 'no' if a there is no documented reason that the patient was given a sublingual, IV, or short acting formula of one of these medications within the first 24 hours. Source: Adapted from AHA Get with the Guidelines ACTION registry
Beta-blockers in the First 24 Hours-Start Date
Coding Instructions: Indicate the date the initial dose of beta-blocker was administered, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Beta Blockers in the First 24 Hours-Start Time
Coding Instructions: Indicate the time that the beta blocker was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g., transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact.
Angiotensin Converting Enzyme Inhibitor in the First 24 Hours
Coding Instructions: Indicate if an angiotensin converting enzyme inhibitor was administered, regardless of location of care (e.g., transferring facility or EMS). Target Value: Any occurrence between 24 hours before first medical contact and 24 hours after first medical contact Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Angiotensin Converting Enzyme Inhibitors in the First 24 Hours-Start Date
Coding Instructions: Indicate the date the initial dose of angiotensin converting enzyme inhibitor was administered, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact. Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Angiotensin Converting Enzyme Inhibitors in the First 24 Hours-Start Time
Coding Instructions: Indicate the time that the angiotensin converting enzyme inhibitor was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact.
Angiotensin Receptor Blockers in the First 24 Hours-Start Date
Coding Instructions: Indicate the date the initial dose of angiotensin receptor blocker was administered, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact. Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Angiotensin Receptor Blockers in the First 24 Hours-Start Time
Coding Instructions: Indicate the time that the angiotensin receptor blocker was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g., transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact. Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Angiotensin Receptor Blockers at Discharge -Dose
Coding Instructions: Indicate the dose of angiotensin receptor blocker prescribed at discharge. The name of the agent used should also be mentioned. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: The highest value on discharge Selections: (none) Supporting Definitions: (none) Source: Definition per CORE team
Angiotensin Receptor Blockers at Discharge
Coding Instructions: Indicate if an angiotensin receptor blocker was continued or prescribed at discharge. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target
Statins in the First 24 Hours-Start Date
Coding Instructions: Indicate the date the initial dose of statin was administered, regardless of location of care (e.g. transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact. Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Statins in the First 24 Hours-Start Time
Coding Instructions: Indicate the time that the statin was administered in the first 24 hours before or after first medical contact, regardless of location of care (e.g., transferring facility or EMS). If administered more than once, code the first date/time it was administered. Target Value: The first value between 24 hours before first medical contact and 24 hours after first medical contact. Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Statins at Discharge
Coding Instructions: Indicate if a statin was continued or prescribed at discharge. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: Any occurrence on discharge Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Documented Reasons for Non-Prescription of Statin at Discharge
Coding Instructions: If the medication was not prescribed, indicate whether the reason was documented. Target Value: Any documentation in the chart regarding contraindications for the prescription of the medication at Discharge. Selections: (1) No (2) Yes -Allergy (3) Yes -Other (specify) Source: Definition per CORE team
Statin at Discharge -Dose
Coding Instructions: Indicate the dose of statin prescribed at discharge. The agent used should be named, too. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: The highest value on discharge Selections: (none) Supporting Definitions: (none) Source: Definition per CORE team
Non-statin Lipid Lowering Agents in the First 24 Hours
Coding Instructions: Indicate if a non-statin lipid-lowering drug was administered, regardless of location of care (e.g. transferring facility or EMS). Target Value: Any occurrence between 24 hours before first medical contact and 24 hours after first medical contact Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Non-statin Lipid Lowering Agents at Discharge
Coding Instructions: Indicate if a non-statin lipid-lowering agent was continued or prescribed at discharge. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: Any occurrence on discharge Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Nitrates in the First 24 Hours
Coding Instructions: Indicate if a nitrate was administered, regardless of location of care (e.g. transferring facility or EMS). Target Value: Any occurrence between 24 hours before first medical contact and 24 hours after first medical contact Selections: (1) No (2) Yes -If yes, name the drug. Supporting Definitions: (none) Source: Definition per CORE team
Nitrate at Discharge
Coding Instructions: Indicate if a nitrate was continued or prescribed at discharge. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: Any occurrence on discharge Selections: (1) No (2) Yes -If yes, name the drug and the dose at discharge Supporting Definitions: (none) Source: Definition per CORE team
Ranolazine at Discharge
Coding Instructions: Indicate if ranolazine was continued or prescribed at discharge. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: Any occurrence on discharge Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Definition per CORE team
Calcium Channel Blockers at Discharge
Coding Instructions: Indicate if a calcium channel blocker was continued or prescribed at discharge. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: Any occurrence on discharge Selections: (1) No (2) Yes -If yes, please name the agent being prescribed. Supporting Definitions: (none) Source: Definition per CORE team
Traditional Chinese Medications (TCM) at Discharge
Coding Instructions: Indicate if a TCM was continued or prescribed at discharge. Note(s): Discharge medications do not need to be recorded for patients who were discharged to "Other acute care hospital", "Hospice", or "Left against medical advice (AMA)." Target Value: Any occurrence on discharge Selections: (1) No (2) Yes -If yes, please name the drug. Supporting Definitions: (none) We will note the following seven categories TCPM (traditional Chinese patent medication) commonly used for AMI in China based on their main ingredient. 1. Danshen or Ginseng or Red Ginseng 2. Ginkgo 3. Sanqi (Panax notoginseng) 4. Hirudin 5. Erigeron breviscapus Extract(Dengzhan Hua 6. Lipid-lowing TCPM (Xuezhikang and Taizhian) 7. Others(Jiuxinwan and Gegengsu) Source: Definition per China team
Diagnostic Catheterization or Diagnostic Coronary Angiography
Coding Instructions: Indicate if the patient had a diagnostic coronary angiography procedure. Target Value: Any occurrence between arrival at first facility and discharge Selections: (1) No (2) Yes Supporting Definitions: Diagnostic coronary angiography is defined as the passage of a catheter into the aortic root or other great vessels for the purpose of angiography of the native coronary arteries or bypass grafts supplying native coronary arteries. Source: NCDR; adapted from AHA Get with the Guidelines ACTION registry
Catheterization Laboratory Arrival Date and Time
Coding Instructions: Indicate the date the patient arrived to the cath lab, as documented in the medical record, as well as the time of arrival Target Value: The first value between arrival at this facility and discharge Selections: (none) Supporting Definitions: Indicate the time (hours: minutes) using the military 24-hour clock, beginning at midnight (0000 hours). If an arterial sheath is already in place, use the time of the introduction of a catheter or the time the sheath was exchanged. Source: Adapted from AHA Get with the Guidelines ACTION registry and from VIRGO registry
Time of Arterial Access
Coding Instructions: Indicate the time arterial access was obtained. Note(s): Time of Procedure: The time arterial access was obtained is defined as the time at which local anesthetic was first administered for vascular access, or the time of the first attempt at vascular access for the cardiac catheterization (use whichever is earlier). Target Value: N/A Selections: (none) Supporting Definitions: Indicate the time (hours: minutes) using the military 24-hour clock, beginning at midnight (0000 hours). Source: Adapted from AHA Get with the Guidelines CATH-PCI registry Arterial Access Site Coding Instructions: Indicate the primary location of percutaneous entry. If more than one entry site was used, choose the site that was used to perform the majority of the procedure. Target Value: N/A Selections: Choose one of the following: a. Femoral; mark "Femoral" if percutaneous puncture of either femoral artery. b. Radial; mark "Radial" if percutaneous radial approach. c. Brachial; mark "Brachial" if either a cutdown or percutaneous puncture of either brachial artery. d. Other; mark "Other" if percutaneous entry other than femoral, brachial, or radial approaches to the cardiovascular system. Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
Arterial Dominance
Coding Instructions: Indicate the dominance of the coronary anatomy (whether the posterior descending artery comes from the right or left vessel system).
Selections: Selection Text Definition
Left dominance is present when the posterior descending artery (PDA) and posterolateral artery (PLA) arise from the left circumflex artery. Right dominance is present when the posterior descending artery (PDA) and posterolateral artery (PLA) arises from the right coronary artery. Co-dominance is present when the right coronary artery supplies the posterior descending artery (PDA) and the circumflex supplies the posterolateral artery (PLA). Thus, there is approximately equal contribution to the inferior surface of the left ventricle from both the left circumflex and right coronary arteries. If not reported, select "unrecorded". Supporting Definitions: (none) Source: NCDR; Adapted from the CathPCI registry
Stenosis Percent
Coding Instructions: Indicate the best estimate of most severe percent stenosis in any coronary artery. Provide the most severe stenosis for the vessel that is primarily providing perfusion to the myocardium in that territory. (Ex. If a patient's mid LAD is 100% and a graft provides revascularization to that territory of the heart, code the % stenosis of the graft. If the same patient has an open graft, and a 70% stenosis of the 2 nd diagonal, code 70% since that is the most severe stenosis % for that territory of the myocardium.) In instances where multiple lesions are present, enter the single highest percent stenosis noted. If no stenosis, then enter 0%. Stenosis: Stenosis represents the percentage diameter reduction, from 0 to 100, associated with the identified vessels. Percent stenosis at its maximal point is estimated to be the amount of reduction in the diameter of the "normal" reference vessel proximal to the lesion. Source: NCDR; Adapted from AHA Get with the Guidelines ACTION registry
CABG Time
Coding Instructions: Indicate the time of the coronary artery bypass graft (CABG) surgery. Target Value: The first value between arrival at this facility and discharge Selections: (none) Supporting Definitions: (none) Note(s): The time of the procedure is the time to the nearest minute (using 24-hour clock), that the skin incision or its equivalent was made to start the surgical procedure Source: Adapted from VIRGO registry
In-Hospital Bleeding
Coding Instructions: Indicate if the patient had a bleeding event during hospitalization. Target Value: Any occurrence mentioned in the chart from arrival to the facility to discharge Selections: (1) No (2) Yes Supporting Definitions: If yes, please indicate the time and date. Source: Definition per CORE team
Location of Bleeding
Coding Instructions: Indicate the location of bleeding. Target Value: N/A Selections: (1) Access site (2) Intracranial (3) Intraocular (4) Intraspinal (5) Retroperitoneal (6) Pericardial (7) Gastrointestinal (8) Genitourinary (9) Other (specify) Supporting Definitions: Access site bleeding is marked when bleeding happens at the site of vascular access. Intracranial bleeding includes intracerebral and subdural bleeding. Please record each site separately if more than one applies. Source: Definition per CORE team
Hypovolemic/Hemorrhagic Shock after Bleeding
Coding Instructions: Indicate if bleeding led to hypovolemic or hemorrhagic shock. Target Value: Any occurrence during the hospital stay Selections: (1) No (2) Yes (3) Unknown Supporting Definitions: In patients who bled during the hospital stay, evidence in the chart may suggest hypovolemic shock: A) Physician report of hypovolemic shock in the chart B) Post-bleeding systolic hypotension (peak systolic pressure<90mmHg) or a reduction of >40mmHg in systolic blood pressure plus evidence of organ hypoperfusion, which is not responsive to administration of plasma expanders or packed RBCs. Source: Definition per CORE team
Interventions for Management of Bleeding
Coding Instructions: Indicate the intervention(s) used to manage bleeding. Target Value: Any occurrence after the bleeding event. Selections: (1) whole blood transfusion (2) packed red cell transfusion (3) local compression (4) surgical intervention, including open surgery, closure or endoscopic interventions (5) others (6) none (7) unrecorded Supporting Definitions: Access site bleeding is marked when bleeding happens at the site of vascular access. Intracranial bleeding includes intracerebral and subdural bleeding. Note(s): Please record each intervention separately if more than one was applied. Source: Definition per CORE team
Blood Transfusion
Coding Instructions: Indicate if the patient was transfused with whole blood or any of its components. Target Value: Any occurrence between first medical contact and discharge Selections: (1) No (2) i.e. development of blood clots in the deep veins of lower extremity or upper extremity) and pulmonary embolism ([PE]; i.e. migration of the clots to the pulmonary arteries. The clots can clog the pulmonary arteries, impairing the gas exchange in the lungs. PE is associated with symptoms such as dyspnea and chest pain and can be fatal). For any patient that was diagnosed with DVT, or with PE, or with both during the index admission, please mark "Yes". Source: Definition per CORE team
In-Hospital Deep Vein Thrombosis (DVT)
Coding Instructions: Indicate if in-hospital DVT was diagnosed during the hospital stay Target Value: Any occurrence from the beginning of the hospital stay to discharge Selections: (1) No (2) Yes -If yes, please indicate the date Supporting Definitions: DVT refers to development of blood clots in the deep veins of lower extremity or upper extremity migration of the clots to the pulmonary arteries. The symptoms and signs include extremity pain, warmness, swelling, a palpable venous cord, and tenderness. The diagnosis is made by ultrasonography or venography. Mark "Yes" if there is physician documentation for the diagnosis. Source: Definition per CORE team
In-Hospital Pulmonary Embolism (PE)
Coding Instructions: Indicate if in-hospital PE was diagnosed during the hospital stay Target Value: Any occurrence from the beginning of the hospital stay to discharge Selections: (1) No (2) Yes -If yes, please indicate the date Supporting Definitions: PE refers to a clinical condition resulting from migration of the clots (rarely other material) to the pulmonary arteries. The clots can clog the pulmonary arteries, impairing the gas exchange in the lungs. PE is associated with symptoms such as dyspnea, hemoptysis and chest pain and can be fatal. The diagnosis could be made by ventilationperfusion (V/Q) scanning, computed tomography pulmonary angiography, or conventional pulmonary angiography. Mark "Yes" if there is physician documentation for the diagnosis. Source: Definition per CORE team
In-Hospital Infection
Coding Instructions: Indicate if in-hospital infection occurred during the hospital stay Target Value: Any occurrence from the beginning of hospital stay to discharge Selections: (1)
In-Hospital Cardiogenic Shock
Coding Instructions: Indicate if cardiogenic shock occurred during the hospital stay Target Value: Any occurrence from the hospital stay to discharge Selections: (1) No (2) Yes-If yes, please indicate the date and time Supporting Definitions: Cardiogenic shock is defined as a sustained (>30 minutes) episode of systolic blood pressure <90 mm Hg, and/or cardiac index <2.2 L/min/m2 determined to be secondary to cardiac dysfunction, and/or the requirement for parenteral inotropic or vasopressor agents or mechanical support (e.g., IABP, extracorporeal circulation, ventricular assist devices) to maintain blood pressure and cardiac index above those specified levels. Source: Acute Coronary Syndromes Data Standards (JACC 2001(JACC 38: 2114, Adapted from AHA Get with the Guidelines ACTION registry
In-Hospital Cardiac Rupture
Coding Instructions: Indicate if there is physician documentation of in-hospital cardiac rupture Target Value: Any occurrence from the beginning of hospital stay to discharge Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Definition per CORE team
In-Hospital Recurrent Angina/Recurrent Myocardial Infarction
Coding Instructions: Indicate if there is physician documentation of in-hospital recurrent angina or recurrent myocardial infarction Target Value: Any occurrence from the beginning of hospital stay to discharge Selections: (1) No (2) Yes Supporting Definitions: Physician documentation of recurrent angina, or recurrent myocardial infarction, or both, warrants a "Yes" answer. Source: Definition per CORE team
In-Hospital Ventricular Tachycardia/Ventricular Fibrillation
Coding Instructions: Indicate if there is physician documentation of in-hospital ventricular tachycardia or ventricular fibrillation Target Value: Any occurrence from the beginning of hospital stay to discharge Selections: (1) No (2) Yes Supporting Definitions: Physician documentation of ventricular tachycardia or ventricular fibrillation, or both, warrants a "Yes" answer. Source: Definition per CORE team
In-Hospital Papillary Muscle Rupture
Coding Instructions: Indicate if there is physician documentation of in-hospital papillary muscle rupture. Target Value: Any occurrence from the beginning of hospital stay to discharge Selections: (1) No (2) Yes Supporting Definitions: (none) Source: Definition per CORE team
In-Hospital Ventricular Septal Perforation
Coding Instructions: Indicate if there is physician documentation of ventricular septal perforation. Target Value: Any occurrence from the beginning of hospital stay to discharge Selections: (1) No (2) Yes Supporting Definitions: (none)
In-Hospital Acute Renal Failure
Coding Instructions: Indicate if there is physician documentation or report of in-hospital acute renal failure Target Value: Any occurrence from the beginning of hospital stay to discharge Selections: (1) No (2) Yes Supporting Definitions: Acute renal failure is defined as physician documentation of acute renal failure Source: Definition per CORE team
In-Hospital Peripheral Embolization
Coding Instructions: Indicate if there is physician documentation or report of in-hospital peripheral embolization Target Value: Any occurrence from the beginning of hospital stay to discharge Selections: (1)
In-Hospital Cardiac Arrest
Coding Instructions: Indicate if the patient experienced an episode of cardiac arrest in your facility. Target Value: Any occurrence between arrival at this facility and discharge Selections: (1) No (2) Yes Supporting Definitions: 'Sudden' cardiac arrest is the sudden cessation of cardiac activity so that the victim becomes unresponsive, with no normal breathing and no signs of circulation. If corrective measures are not taken rapidly, this condition progresses to sudden death. Cardiac arrest should be used to signify an event as described above that is reversed, usually by CPR, and/or defibrillation or cardioversion, or cardiac pacing. Sudden cardiac arrest is not the same as sudden cardiac death. Sudden cardiac death describes a fatal event.
In-Hospital Cardiac Arrest Date
Coding Instructions: Indicate the date of the cardiac arrest. Target Value: The first value between arrival at this facility and discharge Selections: (none) Supporting Definitions: (none)
In-Hospital Cerebrovascular accident (CVA)/Stroke
Coding Instructions: Indicate if the patient experienced a stroke or CVA in your facility. Target Value: Any occurrence between arrival at this facility and discharge Selections: (1) No (2) Yes Supporting Definitions: Stroke: A stroke or cerebrovascular accident is defined as loss of neurological function caused by an ischemic or hemorrhagic event with residual symptoms at least 24 hours after onset or leading to death. Source: Acute Coronary Syndromes Data Standards (JACC 2001(JACC 38: 2114
In-Hospital Stroke Date
Coding Instructions: Indicate the date of onset of stroke. If a stroke occurs during sleep, last awake time may be used. Target Value: The first value between arrival at this facility and discharge Selections: (none) Supporting Definitions: (none) Source: Adapted from AHA Get with the Guidelines ACTION registry
In-Hospital Hemorrhagic Stroke
Coding Instructions: Indicate if the patient experienced a hemorrhagic stroke with documentation on imaging. Target Value: Any occurrence between arrival at this facility and discharge Selections: (1) No (2) Yes Supporting Definitions: Hemorrhagic Stroke: A hemorrhagic stroke requires documentation on imaging (e.g. CT scan or MRI of hemorrhage in the cerebral parenchyma, or a subdural or subarachnoid hemorrhage). Evidence of hemorrhagic stroke obtained from lumbar puncture, neurosurgery, or autopsy can also confirm the diagnosis. Source: Acute Coronary Syndromes Data Standards (JACC 2001(JACC 38: 2114
In-Hospital Unspecified Stroke
Coding Instructions: Indicate if the patient experienced an unspecified stroke with documentation on imaging. Target Value: Any occurrence between arrival at this facility and discharge Selections: (1) No (2) Yes Supporting Definitions: An unspecified stroke can be defined as stroke without documentation on imaging. Evidence of stroke obtained from clinical symptoms. Source: Definition per CORE team
In-Hospital (New Onset) Heart Failure
Coding Instructions: Indicate if there is physician documentation or report on development of heart failure during hospital stay.
Target Value: Any occurrence between first medical contact and arrival at this facility Selections: (1) No (2) Yes Supporting Definitions: Heart failure is defined as physician documentation or report of any of the following clinical symptoms of heart failure described as unusual dyspnea on light exertion, recurrent dyspnea occurring in the supine position, fluid retention; or the description of rales, jugular venous distension, pulmonary edema on physical exam, or pulmonary edema on chest x-ray presumed to be cardiac dysfunction. A low ejection fraction without clinical evidence of heart failure does not qualify as heart failure. Source: Acute Coronary Syndromes Data Standards (JACC 2001(JACC 38: 2114, The Society of Thoracic Surgeons
In-Hospital Atrial Fibrillation or Flutter
Coding Instructions: Indicate if patient was diagnosed with atrial fibrillation or flutter during this admission. Target Value: Any occurrence between arrival at this facility and discharge Selections: (1) No (2) Yes Supporting Definitions: Fibrillation of > 30 seconds, which presents as supraventricular complexes at an irregular rhythm and no obvious P waves on ECG; or Flutter presents as identically recurring regular sawtooth flutter waves on ECG and evidence of continual electrical activity. Note(s): Code "No" If there is no documentation of atrial arrhythmias during hospitalization, it is acceptable to code "No" Source: Adapted from AHA Get with the Guidelines ACTION registry
Smoking Cessation Counseling
Coding Instructions: Indicate if there was documentation in the medical record that smoking cessation advice or counseling was given during this admission. Note | 2017-04-05T04:12:11.535Z | 2014-12-15T00:00:00.000 | {
"year": 2014,
"sha1": "fa9c0aa20f7630f8d3a29a578ae17df37526884a",
"oa_license": "CCBYNC",
"oa_url": "https://heart.bmj.com/content/101/5/349.full.pdf",
"oa_status": "HYBRID",
"pdf_src": "PubMedCentral",
"pdf_hash": "45c39c5190809da0bba21a7775977c0c0ee16aff",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
233402082 | pes2o/s2orc | v3-fos-license | One-dimensional Co–Cu–Fe–Ni–Zn high-entropy alloy nanostructures
We streamlined a way towards first experimental demonstration of synthesizing one-dimensional Co–Cu–Fe–Ni–Zn high-entropy alloy (HEA) nanowires (NWs) by an economical electrochemical technique in aqueous medium. HEA NWs have a face-centred cubic structure with nanocrystalline features, including twins along the nanowire of uniform length ∼50 µm and diameter 100 ± 20 nm. Control over the formation and uniform stoichiometry (required range for HEAs ∼ 5–35 atom %) throughout the HEA NWs and its nanocrystallinity was achieved using pulse electrodeposition. These Co–Cu–Fe–Ni–Zn nanowires are a new family of HEAs containing both magnetic and non-magnetic elements in addition to its new 1D morphology. GRAPHICAL ABSTRACT IMPACT STATEMENT A breakthrough in attaining single FCC phase high-entropy alloy nanowires through aqueous electrochemical route which facilitates new possibilities for unexplored applications of high-entropy alloys at lower dimensions.
Introduction
Unique one-dimesional (1D) morphology like nanowires of pure metals, alloys and their oxides, sulphides, nitrides have become imperative in electronics [1,2] such as transistors, quantum devices, nanosensors, etc. In addition, due to their exceptional functionalities through a wide concentration range, they are being used in several device applications such as in energy [3] and biomedicine [4]. On the other hand, studies reported on metallic nanowires are restricted to a very small compositional space not exceedingly more than three to four elements [5,6]. Furthermore, the current evolution of new highentropy alloys with remarkable performance has opened the possibilities of exploring enormous combination of unfamiliar alloys in bulk [7][8][9]. Recently, new initiatives have been taken in downsizing these HEAs into thin films [10] and nanoparticles [11] but not towards other new morphologies. In addition, there are very limited HEA families that have been investigated by conventional methods of synthesis due to complexities in mixing of multiple elements in a large compositional space [12,13]. Therefore, a demand to develop a new combination of alloys is still high, preferably high-entropy alloys with various morphologies having unprecedented functionalities which would contribute to a significant advancement for the next-generation materials. In anticipation, a few attempts have been made on multicomponent oxides by dealloying [14]; however, there is no investigation found on HEA NWs especially on a new alloy system in a single step.
With the inclination on fabricating nanowires, very few approaches have been preferred such as chemical methods [15][16][17], lithography [18], printing [19] etc. Moreover, several aspects, such as chemical stability, crystallinity, and composition of the HEA NWs, must be taken into consideration, while nominating a suitable technique among the numerous available methods [20][21][22]. Out of the existing methods, the available electrochemical approach is conterminous for synthesizing nanowires [22,23] and it is recently being used in synthesizing multicomponent alloy thin films. In this regard, template-assisted electrochemical approach has several advantages such as simplicity of operation, low cost, ease of availability, reliability and scalability with a new combination of concentrated elements [22,24]. In addition, unlike the conventional techniques, it is a single-step process, where excellent crystallinity, single-phase nanowires with wider tunability in terms of composition range, morphology, and properties can be attained at relatively low temperatures.
In this work, we demonstrate the synthesis of a new family of quinary alloy nanowires, a formidable task notably by conventional techniques for developing HEAs. This study reveals the prominent pulse electrodeposition technique, yet is seldom used for HEAs, to synthesize unique Cobalt (Co)-Copper (Cu)-Iron (Fe)-Nickel (Ni)-Zinc (Zn) HEA NWs assisted by an anodic aluminium oxide (AAO) template. Several parameters were taken into account and we were successful in synthesizing Co-Cu-Fe-Ni-Zn nanocrystalline nanowires with single FCC phase in a single step at above-room temperature using an aqueous electrolyte. As the current strategy of fabrication of HEA NWs has not been implemented to date, the pursuit of suitable application is a major challenge. However, these nanowires composed of five elements in the form of single-phase alloy with high aspect ratio can be explored in the new direction where multi-functionality is of major importance over the conventional nanowires.
Materials and methods
The AAO wafers with gold contact have been purchased from InRedox, USA (AAO Wafer diameter 10 ± 0.1 mm, thickness 50 ± 2 μm, Au layer back contact thickness 0.5 ± 0.1 μm, pore diameter 100 ± 20 nm). The AAO wafer as cathode and electrolytic nickel as anode have been used for synthesis of HEA NWs. An aqueous electrolyte containing metal BO 3 ) have been added to the electrolyte as complexing agent and buffer, respectively. No surfactants or other additives like brighteners, refiners, etc. were added to the electrolyte. Pulse electrodeposition technique has been employed to synthesize HEA NWs at ∼ 45°C at an electrolyte pH ∼ 3.5. After the deposition, HEAdeposited AAO wafers were dissolved in NaOH to collect the nanowires. HEA NWs have been collected with the help of a strong magnet followed by washing and dispersion in an organic solvent for further microstructural characterization.
Characterization
Nanowires were initially characterized by JEOL JSM 7800F field emission gun scanning electron microscope having Zirconia/Tungsten filament, equipped with an EDAX Energy-dispersive X-ray spectroscopy (EDS) detector to confirm their morphology and composition. Furthermore, these nanowires were characterized by transmission electron microscopy (TEM) for a detailed analysis of crystallinity, phase, etc. Nanowires for TEM characterization were prepared by dispersing them in ethanol, followed by ultrasonication. After drying on a lacey carbon copper grid, the nanowires were inspected using a JEOL ARM-200F analytical transmission electron microscope equipped with an imaging Cs-corrector under an acceleration voltage of 200 KeV. The probe convergence semi-angle was approximately 30 mrad. Selected area electron diffraction (SAED) patterns were acquired at a camera length of 20 mm. Several highresolution TEM images were obtained with an image corrector in a parallel beam mode. High-angle annular dark field scanning TEM (HAADF-STEM) images were recorded using an annular-type detector with collection semi-angle ∼ 100-269 mrad. STEM-EDS spectra and elemental mapping were acquired as well.
Results and discussion
Experimental scheme (Figure 1) represents the complete process for preparation, deposition and collection of HEA NWs. Morphological and microstructural details of nanowires, containing all five elements with their stoichiometry, were determined by field emission gun scanning electron microscopy (FEGSEM) and scanning transmission electron microscopy (STEM) assisted with EDS.
Observations made from FEGSEM ( Figure 2) revealed that, the synthesized nanowires are very uniform in length and diameter with smooth surface finish. Also, composition of the nanowires along their length is invariant with the presence of all five elements in the HEA composition range (5-35 at. %) throughout the wire (Table 1). Furthermore, the uniform composition along the length of the nanowire shows the reliability of the method for complex alloy systems like HEAs. Usually, Febased alloys, specifically containing Fe, Co and Ni, follow anomalous behaviour during the deposition [25] which dominates the composition range in the order of Fe > Co > Ni irrespective of the experimental parameters.
Contrarily, due to the geometrical constraints, such as porous structure of AAO template with pore diameter of 100 ± 20 nm, the composition of alloy directly followed the electrolytic concentration unlike in deposition of thin films. On the other side, duration of the electrodeposition for synthesizing present HEA NWs is a few minutes only unlike the existing studies on electrochemical synthesis of nanowires [26]. In addition, due to the growth in porous structure, several phenomena, which usually occur during the thin film depositions, probably may not occur in the nanowire deposition. For further determination of crystallinity and phase, the nanowires were characterized by TEM.
Representative microstructure of the HEA nanowire is shown in Figure 3. The low magnification TEM in Figure 3(a) clearly displays the nanowire length is typical of the order of several microns in size and the diameter is around 100 nm. A bright field TEM image, taken from the individual nanowire of Figure 3(a), contains several polycrystalline grains, appearing in dark contrast in Figure 3(b). The SAED pattern also confirmed that nanowire is indeed polycrystalline with random orientation. Furthermore SAED rings (Figure 3(b)) reveal that individual nanowires are of face-centred cubic (FCC) structure. Polycrystalline features observed in nanowires are due to the application of pulse during the deposition where new nucleation occurs during every pulse ON time. High-resolution TEM of nanowire revealed extremely fine size nanotwin lamellae within a grain, as shown in Figure 3(c). These twins are oriented along 111 planes, as shown in the inserted fast Fourier transform Figure 3(c) that lie parallel to nanowire surface and growth direction, and spacing is extremely fine in the range of 1-5 nm.
pattern (FFT) in
Furthermore, STEM-EDS chemical mapping reveals that the nanowire contains HEA elements of Fe, Co, Ni, Cu and Zn (Figure 4(a-f)). These five elements are homogenously distributed throughout the nanowire with a defined HEA composition. The chemical composition, obtained from STEM-EDS of several HEA nanowires, is shown in Table S1, which is consistent with SEM-EDS of Table 1.
Several reports explained the growth mechanism of single/polycrystalline metal nanowires by 2D/3D nucleation/growth, instantaneous and progressive growth based on the overpotential and kinetics of nucleation [27,28]. However, the growth mechanism of the HEA NWs, especially for an alloy system with five elements is very complex. In the present study, the application of pulse, ON and OFF time plays a significant influence in forming nanowires with uniform composition etc. It is well understood that the growth of the nanowires begins from the Au seed layer which is a back contact for the AAO template. During the pulse condition, potential gradient and mobility of metal ions and their competition led to the growth process [26]. The hydrates/complexes of five elements further reduce its mobility during the reduction process. Therefore, application of pulse in addition to the electrolyte concentration resulted in the formation of nanocrystalline FCC containing multiple elements with uniform composition unlike in conventional DC deposition. But, after OFF time, followed by forward pulse ON time enables the new nucleation having uniform distribution, and filling efficiency of elements throughout the pore homogeneously. On the other side, twin formation in thin films for FCC materials during the pulse electrodeposition is very common [29][30][31]. In the present deposition, probably due to the five dissimilar ions that fill a very small pore size of 120 nm during pulse leads to a large amount of stress development. Simultaneously during every forward pulse (ON), the stacking arrangement of atoms gets disturbed and may not follow the regular stacking sequence which finally results in faults in the wire which further leads to the formation of twins along the length of the wire [31]. Although the formationgrowth mechanism of HEA NWs is yet to be understood, however, the twin formation in HEAs during the deposition can be explained by two reasons: Primarily, due to the evolution of stresses/strains and their relaxation [29,30] during the pulses and secondly, owing to the elements having dissimilar crystal structures such as FCC, BCC (body-centred cubic) and HCP (hexagonally close packed), leading to the generation of dislocations/faults in the stacking sequence of the nanowire, which probably resulting in the formation of nanotwins.
These findings are uncommon for the high-entropy alloy morphologies in the primary attempt, where their large aspect ratio, alloying elements and ultimately a single-phase alloy with highly stable crystal structure. On the other side, the biggest advantage of this study is that these nanowires consist of five transition elements more specifically distinctive entities of iron, cobalt, and nickel present in the alloy makes HEA NWs magnetic. Having magnetic (Fe, Co, Ni), non-magnetic (Cu, Zn) transition elements in HEA NWs and corrosion/oxidation-resistant elements (Ni, Cu, Zn), these NWs are probably an efficient alternative for the existing nanowires with limited functionality in healthcare/device applications. Further modification/engineering of these nanowires in terms of size, aspect ratio, composition, etc., can improve their synergistic effect in revealing their distinctive unexplored functionalities.
Conclusions
We emphasized our study in the direction of the unexplored 1D morphology of high-entropy Co-Cu-Fe-Ni-Zn alloy and reported the primary observations. Nanocry stalline nanowires with FCC structure are reported by an electrochemical approach in a single step using aqueous electrolyte. Application of pulse for the preparation of these nanowires enables the possibility of control over its formation simultaneously with composition and crystallinity which are shown from the microstructural analysis by TEM, STEM etc. Moreover, the formation of twins in HEA NWs of five elements with dissimilar crystal structures may aid in good ductility in this new morphology. The unique combination of Co-Cu-Fe-Ni-Zn HEA nanowires developed by an electrochemical approach facilitates the design of various multifunctional novel nanostructures in the area of HEAs. In addition, being the foremost observation of HEA NWs, this can significantly influence both basic and applied research perspectives of advanced nanomaterials research which is different from its bulk counterparts. | 2021-04-27T13:22:13.182Z | 2021-07-03T00:00:00.000 | {
"year": 2021,
"sha1": "f9157d17d3f606c3160f205dcf2d04b32c743b5c",
"oa_license": "CCBYNC",
"oa_url": "https://www.tandfonline.com/doi/pdf/10.1080/21663831.2021.1896588?needAccess=true",
"oa_status": "GOLD",
"pdf_src": "TaylorAndFrancis",
"pdf_hash": "f9157d17d3f606c3160f205dcf2d04b32c743b5c",
"s2fieldsofstudy": [
"Materials Science"
],
"extfieldsofstudy": []
} |
256775748 | pes2o/s2orc | v3-fos-license | Modeling dynamic behavior of dielectric elastomer muscle for robotic applications
Recently, as a strong candidate for artificial muscle, dielectric elastomer actuators (DEAs) have been given the spotlight due to their attractive benefits from fast, large, and reversible electrically-controllable actuation in ultra-lightweight structures. Meanwhile, for practical use in mechanical systems such as robotic manipulators, the DEAs are facing challenges in their non-linear response, time-varying strain, and low load-bearing capability due to their soft viscoelastic nature. Moreover, the presence of an interrelation among the time-varying viscoelasticity, dielectric, and conductive relaxations causes difficulty in the estimation of their actuation performance. Although a rolled configuration of a multilayer stack DEA opens up a promising route to enhance mechanical properties, the use of multiple electromechanical elements inevitably causes the estimation of the actuation response to be more complex. In this paper, together with widely used strategies to construct DE muscles, we introduce adoptable models that have been developed to estimate their electro-mechanical response. Moreover, we propose a new model that combines both non-linear and time-dependent energy-based modeling theories for predicting the long-term electro-mechanical dynamic response of the DE muscle. We verified that the model could accurately estimate the long-term dynamic response for as long as 20 min only with small errors as compared with experimental results. Finally, we present future perspectives and challenges with respect to the performance and modeling of the DE muscles for their practical use in various applications including robotics, haptics, and collaborative devices.
Introduction
Artificial muscles have been given great attention as the next-generation actuators for robotics and collaborative systems owing to their versatility in motion and inherent flexibility enabling the robots to be suitable for use in a safe working environment. As one of the promising candidates for artificial muscles, dielectric elastomer actuators (DEAs) have been intensively studied (Madden, 2011). Generally, the DEA is composed of a DE membrane and a couple of thin compliant electrodes. These electrodes are established on both surfaces of the DE membrane through various methods, including brushing, spraying, and sputtering. When an electrical voltage is applied across the compliant electrode, the DEA, pre-stretched and clamped onto a rigid frame, is compressed by electrostatic force in the thickness direction and simultaneously expands to the in-plane direction . The electrically-induced actuation performance is scalable by following a simple electrostatic model, which can be expressed by the equation below.
where p is the effective pressure, ε 0 is the permittivity of free space, ε r is the relative permittivity of the dielectric elastomer, V is the voltage input, and z is the thickness. Since the relative permittivity is one of the crucial parameters to determine the achievable area strain of the DEAs, researchers have adopted diverse dielectric materials including acrylic elastomers, silicone elastomers, and polyurethanes for the DEAs Madden, 2011;Youn et al., 2020). Due to the attractive benefits from large strain actuation, high energy density, and mechanical resilience coming from a highly flexible structure, the DEAs have been exploited for various applications such as haptic devices, soft grippers, wearable interfaces, and locomotion (Frediani et al., 2014;Yun et al., 2019;Youn et al., 2020;Chen et al., 2021;Hwang et al., 2021;Youn et al., 2021). For better usability in these applications, the integration of stiff frame, fiber, fluid, and pneumatics into a dielectric elastomer has been also attempted to improve the actuation performance of the DEAs (Huang et al., 2012;Zhang C et al., 2015;Liu et al., 2017;Hau et al., 2018;Cao et al., 2020;Costa et al., 2020;Song and Cha, 2020;Guo et al., 2021;Hwang et al., 2021). Depending on the purpose, researchers have proposed DEAs with different designs such as multilayer stack, extender, bending beam, diaphragm, and rolled structures for eliciting linear elongations, out-of-plane expansions, or bending movements from the fundamental in-plane expansion behavior responding to electrical stimulation .
Of these proposed configurations, we investigate rolled DEAs in detail. Rolled DEAs, as the name suggests, are fabricated by rolling dielectric elastomer films into a cylindrical shape (Pei et al., 2003). From the rolling process, multiple DE layers are created and the actuator can withstand higher load and more actuation cycles compared to single-layer DEAs (Youn et al., 2020). When a spring core is embedded in the fabrication stage, the configuration is termed the spring-rolled DEA. Owing to the use of the compressed spring structure enabling axial stretching via a spring being restored from its compressed state, spring-rolled DEA could produce electrically controllable and linear tensile stroke while maintaining its structure without the need for an external frame or clamping mechanism (Pei et al., 2004).
However, despite the evolution of DEAs with various aspects, difficulty in their fine control is a huge obstacle to exploiting DEAs for practical use in robotic applications. Although many models have been proposed to establish a control strategy, long-term and/or dynamic actuation responses of the DEAs rarely allow accurate estimation due to the presence of inherent non-linear and complex time-dependent viscoelastic behavior, which is interrelated with dielectric and conductive relaxations (Zhao et al., 2011;Tran et al., 2018). Adding to the complexity, the DEA is also affected by cyclic relaxations. With repeated loading and unloading in dynamic actuation, the decrease in relaxation modulus becomes less with the increasing number of loading cycles (Sahu et al., 2015;Sahu and Patra, 2016).
In this paper, studies on the rolled DEAs are reviewed with respect to fabrication methodologies and models for predicting their actuation responses. Moreover, we propose a new model integrating hyperelastic and viscoelastic models that is capable of estimating the long-term viscoelastic creep deformation as well as temporal and frequency responses of the spring-rolled DEA. Lastly, we discuss practical considerations and perspectives for implementing DE actuators in diverse application fields.
Dielectric elastomer muscle
For rolled DE configurations, commercially available 3 M VHB acrylic elastomers, as well as silicone elastomers, are widely used thanks to their soft and elastic nature, and relatively high dielectric properties . As a compliant electrode for the DEA, carbon-based conductive materials and metallic nanowires have been mainly adopted due to their cost-effectiveness, facile coating process in large areas, and high tolerance to electrical excitation up to several kilovolts (Kovacs et al., 2008;Rajamani et al., 2008;Kiil and Benslimane, 2009).
Generally, as illustrated in Figure 1, depending on the material adopted for the DE membrane, the rolled DEAs can be prepared by following two different processes. Figure 1A presents a stepwise fabrication process of a silicone-based rolled DEA. Both Ecoflex and Sylgard, commercially-available polydimethylsiloxane (PDMS), have been widely exploited for the silicone-based DE membrane (Youn et al., 2020). Unlike pre-cured elastomers, silicone elastomers allow thickness control, as they can be solidified at an arbitrary shape because liquid elastomers can maintain chain mobility at room temperature for over an hour, which provides enough time for structural forming via bar coating or spin-coating. A single-layered DEA is prepared by curing the thin liquid elastomer, followed by establishing a thin conductive layer on both surfaces and the DEA can be a multi-layered structure by repeating the fabrication process. Particularly, their rolled configuration can be achieved by cutting the membrane DEA into a strip, rolling, and then integrating a couple of rigid covers onto both ends of the structure.
In parallel, 3 M VHB acrylic elastomer-based rolled DEAs can be constructed by following the fabrication process illustrated in Figure 1B. Firstly, two acrylic DE membranes are cut from a 3 M VHB tape batch. Then, the membranes are bi-axially pre-stretched and conductive layers are established. The pre-stretching is done to enhance actuation performance by improving the material's electromechanical properties, such as its electrical breakdown strength, response time, and pull-in stability Sahu et al., 2020;Sahu et al., 2021). Then, the stretched DE membranes are attached. Next, the spring length is matched or compressed to match the width of the brushed electrode pattern. Finally, the spring-rolled DEA is established by rolling the spring structure along the circumferential direction of the stacked DE membrane. With the spring release, the wrapped membrane is further pre-stretched, until the force equilibrium is reached between the compressed spring and the stretched membrane. The deduction in the membrane's thickness in the fabrication stage further increases the actuator's linear stroke performance and enhances its breakdown strength, ensuring actuation reliability. (Amin and Assal, 2018;Li et al., 2018).
Applications of the silicone-based DEAs are depicted in Figures 1C-E. The wing flapper shown in Figure 1C is driven by a siliconbased rolled DEA, which is made of BJB TC-5005 silicone as adopting graphite powder for compliant electrodes, and it is reinforced with a Frontiers in Bioengineering and Biotechnology frontiersin.org shell structure. This structure is bio-inspired from insects' thoracic mechanism, and it serves to maintain DEA's pre-stretch, provide external structure, and function as a mechanical amplifier. The prototype dynamically produces a strain of 6.1% at 1 Hz under an electric voltage of 6 kV (Lau et al., 2014). In Figure 1D, the DEA for a micro-aerial vehicle is a combination of a DE membrane based on a silicone mixture of Ecoflex 0030 and Sylgard 184 and a compliant electrode using stamped carbon nanotubes (CNTs). The DEA operates as indirect flight muscles in insects. For its take-off, the vehicle requires a driving frequency of 350 Hz and a voltage of 1.3 kV. At this actuation condition, the resulting strain is around 7% (Chen et al., 2019). In Figure 1E, an array of rolled DEAs is embedded onto a foam-based wearable sleeve in order to communicate and deliver various information to the wearer. The actuator is also composed of a DE membrane based on the aforementioned silicone mixture and a compliant electrode using CNTs. Under 1 kV, the actuator produces strain as high as 12.5% in a frequency ranging from 10 to 200 Hz (Zhao et al., 2020). Compared to acrylic actuators, siliconebased DEAs exhibit high bandwidth, but low to moderate deformation (Chen et al., 2019). Therefore, they are often used in fast-wing flapping motion or vibration applications.
Figures 1F-H depict applications of the acrylic elastomer-based spring-rolled DEAs. Figure 1F shows a walking robot using the spring-rolled DEA (Pei et al., 2003). Here, the bending motion is realized by patterning the electrodes. By activating the front or the back electrode, the DEA is able to bend either forward or backward respectively. Six spring-rolled DEAs, with a maximum bending angle of 60°at 5.9 kV, are attached to a hexagonal frame, and their locomotion was achieved by using a dual tripod gait. Similarly, the DEA in the crawler shown in Figure 1G has patterned electrodes as well. By activating one side, bending is induced and by activating both sides, elongation of the DEA is achieved. The rolled DEA is fabricated with 3 M VHB 4910 with carbon grease as its electrode. At 6.5 kV activation voltage, it shows displacement of 16.5 mm (15% strain) and 108 degrees of bending angle. The locomotion was tested under 1 Hz of actuation frequency, and by integrating with electro-adhesion pads as its feet, it can crawl over inclined surfaces as well (Guo et al., 2022). Figure 1H shows the gripper application of rolled DEAs. It is fabricated from VHB 4910 and patterned carbon grease electrodes. Using the collective bending motion of three rolled DEAs, the gripper's gripping force is reported as 228.3 mN under a step input voltage of 5 kV (Li J. et al., 2019). Frontiers in Bioengineering and Biotechnology frontiersin.org As acrylic-based rolled DEAs exhibit larger deformation and higher energy density compared to silicone-based DEAs, they are often integrated with locomotion robots or gripper applications, where large movement or load-bearing capability is of concern rather than its operation bandwidth (Chen et al., 2019).
3 Time-dependent modeling of dielectric elastomer muscle
General modeling methods
Since the development of spring-rolled DEA, several modeling approaches have been proposed in order to estimate the electromechanical actuation response of DEAs. The modeling approaches can be categorized into two main types, stress-strain models and analytical electromechanical models. In the case of the stress-strain modeling method, the actuation behavior is analyzed by considering the internal physics of the actuator, such as the stress-strain relationship and energy equilibrium within the actuator components. In the beginning stages, the forceactuation stroke relationship of spring-roll DEAs was modeled by considering their geometry. Based on the simulation, it was reported that the output force of the DEA is proportional to its roll stiffness (Pei et al., 2004). Then, a model adopting a stress-strain relationship with an assumption that the output force could be induced by multiple factors including the spring was proposed. As key factors, the elasticity of the materials, pre-strain, and electrostatic force were derived from Young's modulus, Poisson's ratio of the material, and the Maxwell stress equation, respectively. Based on the comparison of numerical simulation and experimental results, it was reported that the model could accurately estimate actuation behaviors only for low voltage operations, where the strain is below 4% (Jung et al., 2006). In parallel, another model was developed that formulated a relationship between activation voltage and force output by inserting spring force and electrostatic force terms, which were derived from the Maxwell equation. However, there was a discrepancy between the estimated strain and the achieved strain. Similar to the aforementioned model, the discrepancy tended to increase with activation voltage. To address this issue, they added a correction factor to fit the experimental data (Zhang et al., 2006).
Meanwhile, the analytical electromechanical modeling method considered the step voltage response and frequency response of the DEAs. These responses are analyzed by using conventional electromechanical system models, which were expressed as a combination of parameters achieved from experiments. The electrical system could be expressed with the parameters of inductance (L), resistance (R), and capacitance (C). The parameters of mass (m), damping coefficient (c), and spring constant (k) were used for the mechanical system. For an analytical modeling method using the second-order mass-springdamper model, the model parameters including damped frequency, damping ratio, and damping coefficient are obtained from the step voltage response of the DEA. This work focused on the methodology that could extract system parameters from transient measurements under a fixed voltage level (Tryson et al., 2010). In a wider scope, a dynamic response model formulated by integrating the mechanical mass-spring-damper model with the electrical model composed of R and C was proposed. The expanded analytical model could allow the formulation of static step response as well as frequency dependence in the actuator performance. However, the model requires practical verification and comparison with experimental results (Zhao et al., 2018).
Modeling hyperelasticity under stationary load
As introduced in the previous section, the stress-strain modeling method exhibited limitations in accurately estimating large strain actuation. As the solution, hyperelasticity models have been proposed since the models consider the non-linear nature of dielectric elastomers resulting from long and flexible polymer chains (Moscardo et al., 2008). Various hyperelasticity models have been adopted to estimate the strain energy, such as the neo-Hookean, Gent, Yeoh, and Mooney-Rivlin models (Kovacs et al., 2007;Moscardo et al., 2008;Tryson et al., 2009;Awadalla and Anees, 2011;Sarban et al., 2011;Wang et al., 2016;Amin and Assal, 2018;Li L. et al., 2019). Based on the free energy analysis, Li et al. (2018) proposed a model for the spring-rolled DEA as describing the strain energy of the material with the Gent model. In their work, the actuation performance of three DEAs with different pre-stretch conditions was measured, and model fitting of experimental data was also implemented to obtain their shear modulus and stretch limit. Although the model exhibited consistency with one experimental condition through the whole range of deformation, it rarely showed consistency for different pre-stretch conditions. Recently, another free energy modeling method of the spring-rolled DEA was reported, which describes the strain energy of the material with the Yeoh form. This model was also verified by comparing theoretical data with the experiment (Amin et al., 2020). However, although both hyperelastic modeling approaches contributed to improving modeling capability, the models still could allow accurate estimation of the actuation performance in the specific condition only.
Modeling time-dependency under dynamic load
Although it is reported DEAs are capable of producing electrically-controllable strain, their actuation strain to the applied electric field is non-linear and time-dependent. DEAs also suffer from the viscoelastic creep phenomenon originating from dissipative relaxations within the material. These time-dependent and rate-dependent phenomena are not observed in stationary load conditions, but they prevail under dynamic loading cases. Therefore, addressing the viscoelastic behavior of the DEAs along with hyperelastic modeling is a key factor in improving accuracy in the estimation of their stationary and dynamic actuation performance. In viscoelastic modeling methods, the elastomer could be described by rheological models. As can be seen in the conventional 2-branch general Maxwell model (Zhang J et al., 2015;Li et al., 2021), rheological models are expressed with springs and dampers connected in series and/or in parallel branches. By adding a damper to the system, rheological models could include the rate of deformation to describe the viscoelastic time-dependent relaxation.
Frontiers in Bioengineering and Biotechnology frontiersin.org
Through simulation, it was demonstrated that the spring-rolled DEA's axial stretch undergoes creep deformation with time until the damper is fully relaxed and eventually reaches equilibrium (Zhang C et al., 2015). An expanded rheological model with five parallel branches is also introduced. The model contains the Kelvin-Voigt model with three generalized Maxwell units, which are expressed by a spring and a damper in parallel. The proposed model describes material hyperelasticity through the free energy analysis with the Gent model. Based on experimental comparison among rheological models with different numbers of branches, they verified that the addition of parallel branches to the rheological model improves its prediction capability (Nguyen et al., 2022). However, the verification was limited to the static response of the actuator.
Meanwhile, in our previous study, we investigated the dynamic behavior of the rolled DEA with a compressed spring core for a period of as long as 20 min (Jeong and Kyung, 2021). Unlike the previous estimation via simulation, we found from the observation that viscoelastic creep deformation of the DEA continues even after 20 min of operation, under dynamic voltage input and mechanical loading. Also, we were able to observe ratedependent behavior, where the actuation stroke decreased with increasing actuation frequency. Considering the creep deformation of the DEA, we established a model that could estimate its long-term dynamic behavior by addressing the behavior with the Gent hyper-elastic model and with the 5branch rheological model, as shown in the inset of Figure 2B. In this research, the simulated model response followed the experimental result for the entirety of the actuation period of 20 min. Compared with the conventional 2-branch model, it can be seen that the rheological model of a greater number of branches should be utilized for long-term estimation. However, in the previous research, there were 15 unknown parameters, and they had to be model-fit to the experimental data. This requires high computing power and a long time to build the model, which is inadequate for application in practical situations. To address such limitations, amendments are made to the previous model and the improved model is presented below.
Here k is the spring constant, P is the external load, ϕ is the applied voltage, L s is the initial spring length, L 1 is the compressed spring length, L θ is the width of the elastomer, L rad is the thickness of the elastomer, ε is the dielectric constant of the elastomer, λ is the stretch ratio in the axial direction, J is the physical stretch limit of the material, ξ i is the stretch ratio of the dashpot component, μ i is the shear modulus, η i is the coefficient of viscosity, and the subscript i represents each branch in the rheological model. a, b, w are geometric parameters and they are shown in Figure 2A.
Detailed model derivations are included in the Supplementary Material S1. Further experiments are conducted to verify the proposed model. The spring-rolled DEA is fabricated following the steps outlined in Section 2 using 3 M VHB F9473PC. As the elastomer's thickness is 254 μm, instead of manually pre-stretching it before rolling, it was prestretched after using the compressed spring's restoration. To observe Frontiers in Bioengineering and Biotechnology frontiersin.org and verify the estimation capability of the model, a robotic arm assembly was prepared and the DEA was installed to control the arm, as shown in Figure 2A. An encoder was placed at the hinge and the DEA's deformation was calculated from the angular data. For the experiment, a sinusoidal voltage input of 0.5 Hz frequency and 4000 V amplitude was applied and the actuator's response was measured for 1,200 s. The parameters on the left-hand side of the first equation can be obtained from the experimental setup and the fabrication stage, refer to Table 1. μ 1 and J can be derived from initial conditions, the initial actuator length, assuming no load is applied to the system. Then, only eight unknown parameters remain, which comprise the time constants of each rheological model branch, τ μ i /η i . In the modelfit process, in order to minimize computational burden, μ i and η i are assumed to be powers of 10, and μ i ≤ η i , so that τ < 1, as the rise time of the actuator response is more than 20 min. For the value optimization, the set of parameters that yields the lowest cost function, compared to the experimental data, is selected. The cost function is composed of root-mean-square error and actuation stroke error. The optimized parameters for the data in Figure 2B are μ 2 10 5 , μ 3 10 5 , μ 4 10 5 , μ 5 10 6 , η 2 10 7 , η 3 10 7 , η 4 10 7 , η 5 10 7 . Figure 2B clearly depicts that the actuator's relaxation phenomenon (black) is time-dependent and prolongs without reaching an equilibrium when actuated under the dynamic loading condition. The figure verifies that the proposed model addresses both the hyperelasticity and time-dependency of the dielectric elastomer, and as a consequence, it is able to accurately estimate the long-term dynamic response of the actuator. The proposed model's performance on a different loading condition has been tested and verified as well. Its detailed modeling process and results are included in the Supplementary Material S1. Figure 2C depicts the simulation result of the actuator when the 2-branch rheological model is used. The optimal model parameters are μ 2 10 6 , and η 2 10 7 . It can be observed that for the long-term actuation and under dynamic loading, the conventional 2-branch model is insufficient to accurately model the complex long-term response of the DEA. Therefore, the introduction of multiple branches to the rheological model is necessary, if the DEA were to be operated in practical and long-term applications. As mentioned, the optimization method was selected to minimize the computational burden and the results prove its potential. However, this method has its limitations. For a more accurate estimation of the long-term response, the number of branches in the rheological model can be increased. This in turn increases the number of unknown parameters. As a consequence, despite using the optimization method presented, the computation time for selecting optimal values increases. In Section 4, we further discuss the tradeoff relationship between accuracy and efficiency that follow in the multiple parameter estimation methods.
Future perspectives and challenges 4.1 Artificial intelligence-based data-driven modeling of non-linear response
In the current stress-strain modeling method, the rheological model describing the viscoelasticity of the actuator contains unknown parameters that need to be determined. These parameters are found empirically from experiments. Considering DE muscles' long-term behavior and their application, the rheological model with multiple branches needs to be used. However, the current parameter optimization process is bound to result in a greater computational burden with the increase in the number of parameters. In addition to the increased burden, the optimization process is prone to fall into local minima, and thus a thorough examination is required.
C Data-driven modeling: Considering the complexity of multiple parameter models, approaching the problem with artificial intelligence-based data-driven modeling seems appropriate. Li J. et al. (2019) proposed a model-free control method for DEAs, based on deep reinforcement learning. For the demonstration, simple DEA configurations, circular and rectangular actuators that undergo in-place expansion, are selected and tested. The proposed control method exhibits low trajectory RMSE and shows the potential for the deep learning algorithm to be used in DEAs. For the pursuit of increasing accuracy, the general rolled DEA model proposed in this work could be integrated into the deep learning algorithm. Such integration would allow for the estimation and control of complex structured DEAs, including rolled DEAs. Also, the limitation of increasing the number of branches, and thus unknown parameters, in the rheological model would be overcome. In addition, the study of finding the optimal rheological model for dielectric elastomers could follow.
C Real-time multiple parameter estimation with computation efficiency: Increasing the accuracy of the model is of great issue. However, as can be seen in experiments by Tryson et al. (2009), DE actuator's performance varies per sample. Actuators fabricated in the same fashion do not yield the same results. Moreover, the physical properties of elastomers, such as inertia distribution, viscosity, and stiffness, also vary under dynamic deformation and environmental factors. Due to the variance caused by the coupling of multiple parameters, the modeling and parameter optimization process needs to be carried out per sample for an accurate estimation. In addition, the deep learning method may yield high accuracy, but at the cost of a long model training time. For the model to be used in practical applications, a real-time estimation method needs to be established. An example of a real-time estimation can be witnessed in the haptics field. A data-driven haptic rendering of viscoelastic deformable objects, including Ecoflex silicone is proposed. A random forest regression method was utilized, and the relationship between force and displacement was trained. As a result, the viscoelastic behavior was replicated in the relationship (Bhardwaj et al., 2019). In haptic applications, high perceptual realism is important. Therefore, instead of applying complex physics models and obtaining high accuracy, similarity is pursued and computational efficiency is achieved. As a result, the model is capable of handling large and complex datasets with good accuracy, and at an update rate above 1 kHz, which is sufficient for real-time computation. Such an approach for similarity could also be integrated into the field of realtime DEA modeling and could become an effective and robust solution for actuator control.
Practical considerations for application
Owing to their actuation capability, producing a large and controllable electrically-induced strain with fast response speed, and soft and lightweight configuration, DEAs have been given great attention as a promising candidate for artificial muscle. Advances in DEA technology with respect to material and structural design could contribute to eliciting outstanding actuation performance on par with or exceeding that of a natural muscle. Moreover, the recent development of rolled DEAs opened up opportunities for exploiting them for practical applications since DE muscles could produce much larger force and energy output than conventional membrane DEAs under the same electric voltage.
However, there are still technical challenges in several aspects. In fabrication, multi-layered stacking for most silicone elastomers requires a complicated process, superposing each membrane DEA while suffering from low yield and non-uniformity of layers resulting from curing the liquid resin cast. Operating voltage for the DEAs is also limited due to electro-mechanical instability (EMI) impeding recovery from a deformed state, electrical safety for human contact, and the necessity of a compact power source for miniaturization of the control system. Particularly, for practical use in the fields of robotics, tactile display, and human-assistive devices, the following aspects need to be considered further.
• Robotic kinematic application: DE muscles have been utilized to operate robotic systems. A robotic hand, where DE muscles move fingers through a pulley mechanism (Jung et al., 2006), and a cabledriven hand exoskeleton for rehabilitation, where DE muscles assist finger extension (Amin et al., 2020), have been proposed. In such robotic kinematic applications, it is important to amplify the DE muscle's mechanical strength and/or its displacement. For the amplification, adopting a multi-layered DE structure and increasing the actuation voltage may be required to improve its performance. Externally, the actuator can be reinforced with exoskeletons, that can serve as mechanical amplifiers. In addition, the actuator's durability and robustness upon long-time dynamic loading need to be guaranteed.
• Tactile/haptic display: Proposed tactile or haptic feedback devices include the force feedback glove (Zhang et al., 2006), Braille displays (Ren et al., 2008;Levard et al., 2011), and haptic sleeve (Zhao et al., 2020). In tactile displays, their actuation bandwidth and wearability, including safety, are major factors that require attention. As humans are sensitive to stimulations up to 200 Hz, elastomer material needs to be selected considering its frequency response. Thus, silicone-based elastomers are favorable among various material candidates. In addition, for the device to be wearable, its circuitry needs to be compact and wireless. Therefore, work on miniaturizing and insulating the high-voltage circuit is necessary. For safety, it is recommended to use either thin-film elastomers to lower the DE muscle's operation voltage (Ji et al., 2020), embed wiring within the device, or insulate contact points with an additional elastomer layer.
• Bio-inspired robot: DE muscles are frequently used for soft robots, as their actuation resembles the motions of nature and animals. Several designs have been proposed, ranging from locomotion robots mimicking inchworms (Wang et al., 2016;Li L. et al., 2019;Guo et al., 2022), to eyeball muscles for humanoid robot control , and to flying robots inspired from insect wings (Lau et al., 2014;Chen et al., 2019). These robots require either fast movement or large force output. For a wider bandwidth, a more elastic material, such as silicones, should be selected. If its force exertion is inadequate, it can be enhanced through mechanical amplifiers, stacking, or arranging actuators in an array. Similar to tactile display, ultimately for their free-roaming, methods to develop untethered robots should be further investigated. In addition, most reported applications operate between two states. That is on and off. As proposed in previous sections, an accurate model needs to be developed to control these robots.
Data availability statement
The raw data supporting the conclusion of this article will be made available by the authors, without undue reservation.
Author contributions
SJ wrote the first draft of the manuscript, and HM, SY, and K-UK contributed to manuscript revision, read, and approved the submitted version. | 2023-02-12T16:13:54.868Z | 2023-02-10T00:00:00.000 | {
"year": 2023,
"sha1": "47d0d2e1390f4d421bcefeb2c4e09e31f51665f7",
"oa_license": "CCBY",
"oa_url": "https://www.frontiersin.org/articles/10.3389/fbioe.2023.1006346/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "48072f5ac48a66dbca743709142baef41d7f2f31",
"s2fieldsofstudy": [
"Engineering",
"Materials Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
222164734 | pes2o/s2orc | v3-fos-license | Analysis of Performance Parameters of the Smash in Male and Female Professional Padel
The aim of this study was to analyze the distribution and effectiveness of the different types of smash in professional padel according to the area and direction of the strokes and the gender. Through systematic observation, 1.015 smashes from eight finals (four men’s and four women’s) of the professional matches were analyzed. The smashes were categorized into four types of smash: tray, flat, topspin and off the wall. The results showed both men’s and women’s that the tray is the most used smash by padel players, presenting a percentage of point continuity of almost 90%. The flat and topspin smashes are the strokes that achieve the highest percentage of winning points (near 60%), although this efficiency decreases significantly when the players move away from the net area (p < 0.05), especially in the flat smash. Men perform a higher percentage of winning smashes than women, mainly in the flat smash (p = 0.02). Furthermore, with regards to direction, flat and off the wall smashes are predominantly down the line strokes and women perform significantly more cross court topspin smashes than men (p = 0.005). The results shown could be used to design tasks and exercises by padel coaches at professional players.
Introduction
Padel is a racket sport that was born in Mexico approximately 50 years ago [1] and has experienced enormous growth in the last decade both in the number of players and in the facilities for practicing it [2,3]. Currently it is practiced in more than 40 countries and it has an international tournament circuit in which the best padel players in the world participate [4]. This greater professionalization of padel has also produced an increase in scientific publications [5], especially those related to the performance analysis of the sport [6,7]. In this respect, most research has focused on three fundamental aspects: temporal parameters [8][9][10][11], players' movements and distance covered on the court [12][13][14][15] and game actions [16,17]. The results of these investigations have an enormous transfer and practical application in the design of training sessions adapted to the characteristics of the competition [18,19].
These studies, carried out on professional players, have determined that there are two basic tactical positions in padel: the offensive position, where the players play close to the net and the defensive position, where the players play near the baseline of the court [20]. Previous studies indicate 1.
Tray: Offensive stroke, without a bounce, which is made over the head and on the dominant side of the player. In this shot, before hitting the ball, the player opens the face of the racket pointing upwards and hits with a slice effect. The impact point on the ball is lower than in the other smashes.
2.
Flat smash: Offensive stroke, without a bounce, which is made over the head and on the dominant side of the player. In the execution of this shot, the player hits the ball with a lot of power at the highest possible point, with a flat stroke (no effect), so that after bouncing on the opposite side, the ball could go out of court or return to the other side after rebounding against the wall. 3.
Topspin smash: Offensive stroke, without a bounce, made over the head and hitting the ball from the non-dominant side (behind the player's head). In the execution of this shot the player hits the ball with a lot of power, with a topspin effect, accelerating the shot by arching the back so that, after bouncing the ball on the wall of the opposite side, it goes out over one of the sides walls of the court. 4.
Off the wall smash: Offensive stroke, with a bounce, which is made above the head and on the dominant side of the player. This shot is made when the player, after receiving a lob, lets the ball bounce on his/her side and waits for the bounce on his/her back wall to make a smash. Depending on the player's aim, this shot can be done with a flat or slice effect.
• Shot effectiveness: The classification proposed by Courel-Ibáñez and Sánchez-Alcaraz (2017) was used to determinate smash effectiveness, distinguishing between the winner (the attacking player wins the point by making a smash), error (the attacking player fails the smash and loses the point) or continuity (the point continues after the smash).
•
Hitting zone: The court was divided into six zones, depending on the court side (right or left) and the distance to the net (net, middle and baseline) when the player hit the ball ( Figure 1). Baseline zone, from wall to the serve line (3 m); middle zone, from the serve line to 1/3 court area (3.5 m); net zone, from 1/3 court area to net (3.5 m). • Shot direction: Two possible trajectories were distinguished: down the line and cross court ( Figure 1). on the opposite side, the ball could go out of court or return to the other side after rebounding against the wall. 3. Topspin smash: Offensive stroke, without a bounce, made over the head and hitting the ball from the non-dominant side (behind the player's head). In the execution of this shot the player hits the ball with a lot of power, with a topspin effect, accelerating the shot by arching the back so that, after bouncing the ball on the wall of the opposite side, it goes out over one of the sides walls of the court. 4. Off the wall smash: Offensive stroke, with a bounce, which is made above the head and on the dominant side of the player. This shot is made when the player, after receiving a lob, lets the ball bounce on his/her side and waits for the bounce on his/her back wall to make a smash. Depending on the player's aim, this shot can be done with a flat or slice effect.
Procedure
Data were collected through systematic observation, carried out by two observers who have a degree in Sports Science and are specialized in padel. Observers were specifically trained in the use of the observational instrument during two weeks. The training focused on the clear identification of the variables (type of smash, shot effectiveness, hitting zone and shot direction) and the use of the observational instrument software. At the end of the training process, each observer analyzed the same two sets in order to calculate the inter-observer reliability through the Multirater Kappa Free [30], obtaining values above 0.80. To ensure the consistency of the data, intra-observer reliability was evaluated at the end of the observation process, obtaining minimum values of 0.89. The kappa values obtained revealed a very high degree of agreement (>0.80) [31]. All the analyzed matches
Procedure
Data were collected through systematic observation, carried out by two observers who have a degree in Sports Science and are specialized in padel. Observers were specifically trained in the use of the observational instrument during two weeks. The training focused on the clear identification of the variables (type of smash, shot effectiveness, hitting zone and shot direction) and the use of the 4 of 10 observational instrument software. At the end of the training process, each observer analyzed the same two sets in order to calculate the inter-observer reliability through the Multirater Kappa Free [30], obtaining values above 0.80. To ensure the consistency of the data, intra-observer reliability was evaluated at the end of the observation process, obtaining minimum values of 0.89. The kappa values obtained revealed a very high degree of agreement (>0.80) [31]. All the analyzed matches were retransmitted in streaming and later hosted on the World Padel Tour website (https://www.worldpadeltour.com/) [32], from where they were downloaded for the observation, collection and analysis of the data. LINCE specialized software was used for this process of recording and data collection [33].
Data Analysis
Firstly, a descriptive analysis of the data was carried out and the mean (M), standard deviation (SD), frequency (n) and percentage (%) were calculated on the whole sample. A comparison was made of the statistics of distribution and efficiency of the smashes according to gender and the playing side of the court using Pearson's Chi-Square test. Column proportions were compared using Z tests on the effectiveness of the smash according to the area and gender of the players. A significance level of p < 0.05 was established, which was adjusted according to Bonferroni in the Z tests. The associations among the categories of the variables were performed with corrected standardized residuals (CSR). The effect size was calculated using Crammer's V [34]. All data was analyzed with the IBM SPSS 25.0 statistical package for Macintosh (IBM Corp: Armonk, NY, USA). Figure 2 shows the results of the distribution percentages of the different smashes analyzed according to the type of smash, direction of hitting and court area comparing by gender. According to smash frequency, significant differences were found between genders (χ 2 = 26.423; gl = 3; p < 0.001; V = 0.161). The female player performed significantly more tray strokes than male (CSR = 4.3. Otherwise, the male player made more flat (CSR = 3.7) and top-spin smashes (CSR = 3.0) than female players. Depending on the area of the court, most of the smashes were made in the middle area of the court for both genders. In relation to the direction of the smash, male players performed significantly more down the line strokes than female (χ 2 = 4.281; gl = 1; p = 0.039; CSR = 2.1; V = 0.065).
Results
(https://www.worldpadeltour.com/) [32], from where they were downloaded for the observation, collection and analysis of the data. LINCE specialized software was used for this process of recording and data collection [33].
Data Analysis
Firstly, a descriptive analysis of the data was carried out and the mean (M), standard deviation (SD), frequency (n) and percentage (%) were calculated on the whole sample. A comparison was made of the statistics of distribution and efficiency of the smashes according to gender and the playing side of the court using Pearson's Chi-Square test. Column proportions were compared using Z tests on the effectiveness of the smash according to the area and gender of the players. A significance level of p < 0.05 was established, which was adjusted according to Bonferroni in the Z tests. The associations among the categories of the variables were performed with corrected standardized residuals (CSR). The effect size was calculated using Crammer's V [34]. All data was analyzed with the IBM SPSS 25.0 statistical package for Macintosh (IBM Corp: Armonk, NY, USA). Depending on the hitting side, there was a trend towards a higher percentage of smashes in the middle zone in both genders. On the left side of the court, the smashes made in the different areas of the court were similar between male and female (χ 2 = 2.371; gl = 2; p > 0.05; V = 0.065). By contrary, in the right side of the court male and female players played different (χ 2 = 9.497; gl = 2; p < 0.05; V = 0.145). Male players made more smashes in the net (CSR = 1.8) and in the baseline (CSR = 1.5), and females played more smashes in middle area (CSR = 2.4). Figure 3 shows the percentage of winning points depending on the type of smash and the area of the court by gender. In general lines, male and female players make more winning smashes in the net area than the others. Moreover, male players made more flat and top-spin smashes than females regardless of the hitting area. On the contrary, the female players made more winner shots with the tray stroke only at the net area.
Results
Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 5 of 9 Figure 3 shows the percentage of winning points depending on the type of smash and the area of the court by gender. In general lines, male and female players make more winning smashes in the net area than the others. Moreover, male players made more flat and top-spin smashes than females regardless of the hitting area. On the contrary, the female players made more winner shots with the tray stroke only at the net area. The division by the type of smash (Table 1) showed differences in the distribution of hitting efficiency according to the gender of the players in the flat smash (χ2 = 7.833; gl = 2; p < 0.05; V = 0.187). The men performed a significantly higher percentage of winning flat smashes (CSR = 2.7) and less percentage of continuity of this stroke than the women (CSR = −2.8). No significant differences were found according to gender for the tray (χ2 = 3.332; gl = 2; p > 0.05), topspin (χ2 = 2.376; gl = 2; p > 0.05) and off the wall smashes (χ2 = 1.333; gl = 2 p > 0.05). Furthermore, the flat smash and topspin smash were the two smashes that produced more winners, while the tray and off the wall smashes were the smashes recording a higher percentage of continuity. The smash with the highest percentage of errors was the off the wall smash. Table 2 shows the differences in the percentages of distribution of the smash direction The division by the type of smash (Table 1) showed differences in the distribution of hitting efficiency according to the gender of the players in the flat smash (χ 2 = 7.833; gl = 2; p < 0.05; V = 0.187). The men performed a significantly higher percentage of winning flat smashes (CSR = 2.7) and less percentage of continuity of this stroke than the women (CSR = −2.8). No significant differences were found according to gender for the tray (χ 2 = 3.332; gl = 2; p > 0.05), topspin (χ 2 = 2.376; gl = 2; p > 0.05) and off the wall smashes (χ 2 = 1.333; gl = 2 p > 0.05). Furthermore, the flat smash and topspin smash were the two smashes that produced more winners, while the tray and off the wall smashes were the smashes recording a higher percentage of continuity. The smash with the highest percentage of errors was the off the wall smash. Table 2 shows the differences in the percentages of distribution of the smash direction regarding players' gender and type of smash. Players' gender significantly determined the direction of the topspin smash (χ 2 = 7.203; gl = 1; p < 0.01; CSR = 2.7; V = 0.31). Thus, while men equally distributed the direction of their topspin smashes, women performed more than 80% of their topspin smashes cross court. No significant differences were found between men and women in the direction of the flat smash (χ 2 = 2.635; gl = 1; p > 0.05), tray (χ 2 = 0.260; gl = 1; p > 0.05) and off the wall smash (χ 2 = 0.152; gl = 1; p > 0.05). In general, both men and women hit a higher percentage of down the line flat and off the wall smashes, while the tray was directed equally down the line and cross court.
Discussion
Smash stroke is the game actions that produce the highest percentage of winner shots in padel [12,27] and has a great influence in match outcome. For that, the aim of this study was to analyze the distribution and effectiveness of the different types of smashes in professional padel according to court area, smash direction and players' gender. The results obtained showed that the tray stroke is the most widely used smash type, mainly for female players (Figure 2). In addition, this hit means, in almost 90% of cases, the continuity of the point, which would indicate an important technical domain for the players, since there are hardly any errors [26], although it is also easier for the opponents to defend [17]. On the other hand, the distribution of the smashes by court area showed that approximately 60% of the smashes were made in the middle area. These data confirm that the defending players seek to make high and deep lobs forcing the attacking players to hit the ball in situations far from the net, reducing the possibilities of achieving a winner [16,22]. In addition, on the right-side male players performed a greater number of smashes at the net and baseline areas than female. In the baseline area these differences could be due to anthropometric and strength differences that would allow the male players to make effective smashes further from the net. Regarding the smash effectiveness according to the hitting area (Figure 3), data showed that there is a direct relationship between the distance to the net in the hit and its effectiveness (mainly from the net area with the others). Thus, as the players hit closer to the net, the number of winners increased significantly, especially when using topspin and flat smashes. This could be explained by a favorable position for the attackers (decreasing the possibility of an error), and a shorter reaction time for the defenders. In this sense, coaches should include a high percentage of tray strokes in the training session (with the aim of keeping opponents away from the net), as well as working shots close to the net as a means of finishing the point.
Previous studies in professional padel have shown that players who are in defensive positions predominantly use the lob stroke to recover the net [7,26,35]. However, the attacking players try to maintain the offensive position [22], due to the greater probabilities of winning the point when they are in situations close to the net [16,17,22]. Therefore, in order to maintain the attacking position, it seems that players try to hit most of the balls sent by the defenders without bouncing, thus decreasing the frequency of the off the wall smash, using safer smashes like trays. However, the greater continuity of the hits for a good defense after the smash makes it necessary for the attacking players to look for other types of tactical actions that allow them to provoke successful situations in the point. Thus, it seems that varying the directions of the shots has been one of the fundamental tactical principles to achieve success in racket sports [25,36]. The results of this study showed that both men and women vary the directions of the down the line and cross court smashes in a balanced way, which would produce more movement on the part of the opponents swinging from one side of the court to the other, which may imply that they hit in more unfavorable situations making more mistakes [12,13].
Gender differences showed that men made a significantly higher percentage of winners than women (10% more topspin smashes and almost 20% more flat smashes). These results could be due to the anthropometric and strength differences between elite men and women players [37,38]. The results of these studies show that men padel players are taller, with greater muscle percentage and higher levels of vertical jump and grip strength than the women players, which would allow them to use the powerful smash successfully in positions further from the net. Regarding the hitting directions, it was observed that the women performed a significantly higher percentage of cross court flat and topspin smashes than the men. These differences may have a tactical explanation. While the men finish off a powerful smash down the line aiming to bring the ball to their field after the bounce on the back wall, the women do the cross court smashes with the aim of getting the ball to go over the side wall of the court (3 m high) [20]. In this way, the down the line smash, if not done with a lot of power, can cause the ball to bounce off the back wall at the opposite end with less force, offering a very favorable position for the return by the defending players [29]. With the results, coaches should work mostly on the flat stroke for the men and women, although women have a lower chance of winning than men.
In line with our hypothesis, males and females use one type of spike more than another, and with different performance (mainly in flat and top-spine strokes). The results of this study have important practical applications for the training of padel players, facilitating the design of tasks and exercises, as well as preparing them for competition taking into account the differences between the men's and women's categories. Moreover, knowledge of the effectiveness of the different types of smash depending on the court area in which the player is located will allow the training of perceptual and decisional mechanisms during the game by the player and the application of feedback about the behaviors by the coach [39,40]. However, this study has certain limitations that need to be taken into account when interpreting the results. For example, the use of some contextual variables such as the score (winning, drawing or losing) or the importance of the point (key moment) can influence decision-making in moments of pressure (choking) affecting performance [41]. In addition, other variables that influence smash effectiveness such as the position of the opponents or the speed of the smash have not been evaluated [29].
Conclusions
The tray was the most commonly used smash by padel players. Female players used more tray and less flat and topspin smashes than male players. Tray represents a percentage of point continuity of almost 90%. The flat and topspin smashes were the shots that achieved the highest percentage of winners, although this efficiency decreased significantly when the players moved away from the net area, especially in the flat smash. Regarding gender, men performed a significantly higher percentage of winning smashes than women. In addition, with regard to direction, flat smashes and off the wall smashes were predominantly down the line strokes and women performed significantly more cross court topspin smashes than men. | 2020-10-06T13:34:06.524Z | 2020-09-25T00:00:00.000 | {
"year": 2020,
"sha1": "dcaf5279715b1e3421e44c447640664f52da9dbc",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/1660-4601/17/19/7027/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "4da3a0aa65bcf33f92cda220e4c185e01880d14b",
"s2fieldsofstudy": [
"Education"
],
"extfieldsofstudy": [
"Medicine",
"Mathematics"
]
} |
90958845 | pes2o/s2orc | v3-fos-license | ANTIMICROBIAL ACTIVITY OF MORINGA ON EAR, NOSE AND THROAT ASSOCIATED FUNGI, AND VANCOMYCIN RESISTANT COCCI ISOLATED AT AMINU KANO TEACHING HOSPITAL, KANO, NIGERIA
This study was aimed at evaluating the antimicrobial activity of Moringa on ear, nose and throat associated fungi and vancomycin resistant cocci. The plant material was extracted with methanol and petroleum ethe and screened for phytochemical contents. The microbial isolates were obtained from females and males patients (both adults and children) attending ear, nose and throat clinic at Aminu Kano Teaching Hospital. Coccal bacteria and fungi were isolated accordingly. The cocci were screened for vancomycin resistance. The antimicrobial assay was carried out using gradient double (12.5-100mg/mL) assay. The MIC, MBC/MFC and Brine shrimp toxicity test were also conducted. Staphylococcus aureus, Streptococcus pyogenes, Streptococcus pneumoniae, Candida albicans and Aspergillus fumigatus were isolated. Up to 21.4% of S. aureus were vancomycin resistant, 20% of S. pneumoniae isolated were vancomycin resistant and 16.7% S. pyogenes were vancomycin resistant. The plant extracts showed zones of inhibition of 08mm-20mm at concentrations ranging from 12.5100mg/mL. The most susceptible organism to both extracts was C. albicans and the least susceptible was S. aureus. The MIC of the methanol extracts ranged from 0.78 to 50mg/mL but MBC/MFC ranged from 6.25 to 200mg/mL. The MIC of the petroleum ether was at 50 to 200mg/mL and the MBC/MFC was from 200 to 800mg/mL. The brine shrimp lethality assay showed LC50 value of 93.48μg/mL for Moringa methanol extract while the LC50 value for Moringa petroleum ether extract was 3.691μg/mL. Moringa methanol extract (100mg/mL), showed appreciable activity against the fungal isolates and vancomycin resistant cocci associated with Ear, Nose and Throat symptoms while Moringa petroleum ether extract showed activity only on the fungal isolate C. albicans. The study demonstrated that Moringa methanol extracts was more active than Moringa petroleum ether extracts. The search for novel cytotoxic ingredient in Moringa should be encouraged.
INTRODUCTION
Medicinal plants have been used from ancient time for their medicinal values as well as to impact flavor to food. Nowadays, the crude extracts and dry powder samples from medicinal and aromatic plants species are used for the development and preparation of alternative medicine and food additives (1) . Moringa, native to parts of Africa and Asia, is the sole genus in the flowering plant family Moringaceae. Important medicinal properties of the plant include antipyretic, antiepileptic, antiinflammatory, antiulcerative (2) , antihypertensive (3) cholesterol lowering (4) antioxidant (5) anti diabetic, hepatoprotective (6) , (antibacterial and antifungal activities (5). Vancomycin is an antibiotic used to treat a number of bacterial infections (7). It is recommended intravenously as a first-line treatment for complicated skin infections, bloodstream infections, endocarditis, bone and joint infections, and meningitis caused by methicillinresistant Staphylococcus aureus (8). In addition to natural circumstances, misuse of vancomycin has led to vancomycin resistance. The reasons for clinical failure of vancomycin are many and have been hypothesized to include poor penetration of the drug to certain tissues (9, 10).Wide varieties of Ear, Nose and Throat diseases are usually presented to the Otorhinologist (head and neck surgeons) (11). The pattern of these diseases may vary from community to community or hospital to hospital based on the availability of specialist personnel or facilities for the management of such diseases which are either congenital or acquired in origin. Ear, nose and throat diseases are serious public health problems with universal distribution affecting all age groups (12).
One of the research problems facing chemotherapy today is that microorganisms are now gaining resistance to vancomycin, which has been considered to be the reference standard for the treatment of bacterial infection. In Nigeria today, ear, nose and throat-related conditions constitute a major burden of infections. Yet the majority of the citizens are ignorant of ENT treatment options. Disease of the ear, nose and throat (ENT) affect the functioning of adults as well as children, often with significant impairment of the daily life of affected patients (13). Due to the emergence of vancomycin resistance which is the last resort antibiotic where other antibiotics have failed and ignorance of the severity of ear, nose and throat infections, there is need for an easy, effective and affordable means to cure infections of the ear, nose and throat (ENT). Moringa is known for its numerous medicinal properties one of which is its antimicrobial activity (14). It is very common worldwide to find people consuming this plant in combination or alone as remedies against symptoms believed to be associated with the selected microorganisms targeted in this work. There is a need to find out if this plant has potent antimicrobial activity against ENT fungi and vancomycin resistant cocci. This plant is easy to afford. Moringa can be included in our foods and drinks e.g. tea and soups (15).
Collection and Identification of Plant Materials
The Moringa leaves were collected from Naibawa in Kano state. It was identified and compared to voucher specimen with voucher number (Moringa BUKHAN 0011) at the department of Plant Biology Bayero University, Kano Herbarium with the assistance of Baha'uddeen Said Adam(16).
Processing and Extraction of Plant Materials
The Moringa leaves were thoroughly washed with distilled water and air dried in a shady environment for two weeks and made into powdered form using a clean pestle and mortar, then it was sieved through a mesh to obtain fine powder of approximately 20µm particle size. The powder was stored at room temperature in sealed container until required for use as demonstrated(17). [ Accordingly, one hundred grams (100g) of the powdered plant material was extracted separately with methanol and petroleum ether using soxhlet apparatus as demonstrated (18).
Confirmation of the Bioactive Components of the Plants
Phytochemical screening was carried out to confirm the bioactive components of the plant as follows:
Test for Alkaloids
This was carried out qualitatively as demonstrated (19). Using a pipette, 1.0 ml of the extracts was placed in two separate test tubes. Using a dropper, three drops of Meyer's reagent was added separately. A white precipitate with Meyer's reagent indicated the presence of alkaloids.
Test for Saponins
This was carried out as demonstrated by the method reported (20). 0.5g each of the extracts was placed in a test-tube, 5.0ml of sterile distilled water was added to the extract in the test-tubes and shaken vigorously. A froth that persisted for 15 minutes was an indication of the presence of saponins.
Test for Steroids
This was carried out as demonstrated by the method (19). 2g of each the extracts was placed in a test tube and evaporated to dryness. The extract was then dissolved in acetic anhydride followed by the addition of chloroform and then concentrated sulphuric acid was added by the side of the test tube. Appearance of a brown ring at the interface of the two liquids and the appearance of violet colour in the supernatant layer indicated the presence of steroids in the extract.
Test for Reducing Sugar
This was carried out as demonstrated by the method (20). Here, 1g each of the extracts was weighed and introduced into separate test tubes. The extracts were diluted with 2.0ml each of dimethyl sulphoxide (DMSO) and sterile distilled water respectively. Fehling's solution was added to the solution obtained, and then the mixture was warmed. A brick-red precipitate at the bottom of the test tubes indicated the presence of reducing sugar.
Test for Tannins 2ml of each of the plant was diluted with distilled water in separate test tubes and 2-3 drops of 5% ferric chloride (fecl3) were added. A green-black or blueblack colouration indicated the presence of tannin as demonstrated (19).
Test for Flavonoids
This was carried out as demonstrated (21). A 4mg weight of the extracts and a piece of magnesium ribbon were added together followed by concentrated HCL drop-wise. A colour change from crimson to magenta indicated the presence of flavonoids in the extracts.
Test for Terpenoids 0.5ml of the extracts was added to 2ml of chloroform, 3ml of concentrated H2S04 was added to form a layer. A reddish brown colouration at the interface indicates the presence of terpenoids as demonstrated by the method (21) .
Test for Anthraquinone 0.5ml of the extract was taken into a dry test-tube and 5ml of chloroform was added and shaken for 5mins. The extract was filtered and drops of ammonia solution were added. A pink violet or red colour in the ammonical layer (lower layer) indicates positive results. This is as demonstrated by the method (21).
Tests for Phenol
Few drops (0.5%) of dilute ferric chloride solution was added 0.5 ml of each of the extracts, the formation of a dark green colour shows the presence of phenol according to the method (22).
Collection and Identification of Test Isolates
The isolation was carried out in Aminu Kano Teaching Hospital after ethical clearance has been approved. The isolation was carried out under the supervision of a medical laboratory technician. Thirty three specimens were collected from patients attending ENT clinic in any age group. The organisms were isolated from the ear, nose and throat swabs. The specimen was cultured on Sabouraud Dextrose Agar (Manufacturing date, 2016; Expiring date, 2018) for the isolation of fungi. After 3-7days of incubation the fungi isolates were identified macroscopically and microscopically with the help of scheme (23).
The specimens were cultured on Chocolate agar for the isolation of cocci bacteria. The cocci were identified macroscopically in the culture plates after 24 hours of incubation, after which gram staining was carried out. This was followed by catalase and coagulase test to confirm the species of Staphylococci. Optochin, bacitracin disc and bile solubility test were used to further confirm the species of streptococci. The identified cocci were subjected to vancomycin sensitivity disc (30µg) and the cocci that were found to be resistant to the vancomycin were used for this study as demonstrated (24).
Preparation of Extracts Concentrations
This was carried out according to the method described by the method (25). Stock solution of moringa, methanol and petroleum ether crude extracts were prepared by dissolving 0.6g of each of the plant extracts in 6mL of dimetyhylsulphoxide (DMSO) in glass vial bottles. Therefore, each stock solution had concentration of 100000µg/mL (100mg/mL). The stock solution was double-diluted to get three varied extracts concentrations in addition to it to make them four different concentrations of 100mg/mL, 50mg/mL, 25mg/mL and 12.5mg/mL (26).
Standardization of Inoculum
The isolates were adjusted to 0.5 McFarland standard (1.5 X 10 8 CFU/mL) turbidity for bacteria isolates and 1 x 10 6 spores/mL for the fungi isolates by adding sterile normal saline. McFarland standards were used as a reference to adjust the turbidity of microbial suspension so that the number of microorganisms will be within a given range. For the preparation of the 0.5 McFarland standard, 0.05mL of barium chloride (BaCl2) (1.17% w/v BaCl2.2H2O) was added to 9.95mL of 0.18M H2SO4 (1.0% w/v) with constant stirring. To aid comparison the standard was compared against a white background with a contrasting black line (27).
Preparation of Antibiotic Dilution
The antibiotic ciprofloxacin and fluconazole were purchased from Lamco pharmacy Kano State, Nigeria and was reconstituted by dissolving 3g of the ciprofloxacin and fluconazole powder in 100ml of distilled water so as to get a concentration of 30mg/mL. The prepared dilution of the antibiotics was used for subsequent antimicrobial test as positive control (21). [21] Antimicrobial Assay The bioassay was carried out using the agar well diffusion method described by cheesbrough (2006). 0.1mL of the standardized inoculums (1.5 x 10 8 CFU/ml) of Staphylococcus aureus was inoculated onto sterile prepared Mueller Hinton Agar and was spread with a sterile swab while Streptococcus pneumoniae was inoculated on chocolate agar and Streptococcus pyogenes was inoculated on sterile blood agar plates.
Aspergillus fumigatus and Candida albicans were inoculated on sterile Sabouraud dextrose agar plates. Six wells were made with a 6mm sterile cork borer into the agar plates containing the bacterial and fungal inoculums and 0.lmL of the four different concentrations from the stock solution of the extracts at concentrations (100, 50, 25, and 12.5mg/mL) were introduced into their respective wells. 0.1mL of DMSO was introduced into the fifth well to serve as negative control while 0.1mL of 30mg/mL of ciprofloxacin was introduced into the sixth well to serve as a positive control for the bacterial isolates and fluconazole was used for the fungal isolates. The inoculated plates were left to stand for about 30 minutes to allow diffusion of extract before incubating at 37•C for 24 hours for the bacterial isolates and the fungal isolates were incubated at 37 o C for 3 days. The zones of clearance produced around the wells after incubation was observed and measured using a vernier caliper and recorded (in mm). Each of the experiment was conducted thrice and mean results were taken for the test organisms (28).
Determination of Minimum Inhibitory Concentration, Minimum Bactericidal and Fungicidal Concentration
The minimum inhibitory concentration (MIC) is defined as the lowest concentration of the antimicrobial agent that inhibited visible growth of microorganisms after overnight incubation (Andrews, 2002). The doubling macro dilution broth method was used to determine the MIC. Two (2) mL of the reconstituted crude extract at a concentration of 100000µg/ml was added to 2mL sterile Mueller Hinton broth for the bacterial isolates, 2mL of the reconstituted crude extract was added to 2mL of Sabouraud dextrose broth for the fungal isolates. Two (2) mL of this extract concentration was transferred to another test-tube and this dilution continued until the 12 th test-tube was reached, giving extract concentrations ranging from 800-0.39mg/mL in different test tubes. 0.1mL of an 18h culture of bacteria and 3 days culture of fungi previously adjusted to 0.5 MacFarland standard was inoculated into each of the test tubes and the contents were thoroughly mixed. A test tube containing the broth and extract was used as positive control while a test tube containing the broth and bacteria/fungal inoculum was used as negative control. The inoculated culture tubes were incubated at 37 o C and were observed for growth after 24 hours for the bacterial isolates and 3days for the fungal isolates. The lowest concentration of extract showing no visible growth when compared with the control was considered as the MIC as demonstrated by the method (27).
The minimum bactericidal/fungicidal concentration is the lowest concentration of antimicrobial agent that prevented the growth of an organism. 0.1mL aliquot from the tubes that showed no visible bacterial/fungal growth from the determination of minimum inhibitory concentration was inoculated on a sterile Mueller Hinton Agar for 24 hours at 37•C for the bacterial isolate while the fungal isolates were inoculated on sterile Sabouraud dextrose agar at 37 o C. The lowest concentration in which no growth occurred was taken as the minimum bactericidal concentration (MBC/MFC) as demonstrated (27).
Assay for LC50 of the Plant Extracts by Brine Shrimp Lethality Test
The eggs were hatched in a hatching chamber containing ocean sea salt water (75ml). Natural light was allowed to penetrate into the hatching chamber for 48 hours so that the eggs will hatch into the shrimp larvae. Twenty milligram (20mg) of the extracts were separately dissolved in 2ml of methanol and equally a positive control of which 20mg of the extracts was dissolved in 2ml of distilled water. 500, 50 and 5ml of the solution equivalent to 1000, 100 and 10 µg/mL respectively was transferred into vials. A negative control which is simply the solvent (methanol) without the test extracts was also prepared. Each concentration were tested in triplicate, therefore each extracts had 9 test tubes. The methanol in the test tubes containing the extracts were allowed to evaporate to dryness for about 48 hours at room temperature and were subjected to test for their activity against Brine shrimp larvae (Artemia salina). To each test sample vial, sea water was added and a drop of DMSO solvent was added in order to facilitate the solubility of each test samples in the sea water. Ten (10) shrimps were transferred using a Pasteur pipette and natural sea water was added to make a total volume of 5ml. After 24 hours, the number of surviving shrimps at each concentration was counted and recorded. Lc50 values were determined at 95% confidence intervals by analyzing the data on a computer loaded with a "Finney Programme." The Lc50 values of the brine shrimps obtained for extracts of the plants studied were recorded (29).
Phytochemical Moringa Plant Extracts
Some Phytochemical components of Moringa plant extracts are presented in Table 1. The data showed that, phenol, steroids, reducing sugar, flavonoid, terpenoids, tannins anthraquinone were present in both Moringa methanol and petroleum ether extracts. Alkaloid and saponin were present in Moringa methanol but absent in moringa petroleum ether extracts.
Inhibitory Activity of the Moringa Extracts on ENT Associated Fungi and Vancomycin-Resistant Cocci
The inhibitory activity of Moringa on the test organisms is presented on Table 2. Moringa methanol extract showed zones of inhibition ranging from 8 -20mm at concentrations ranging from 12.5 -100mg/mL on the test organisms. However Moringa petroleum ether extract was not active on all the organisms except Candida albicans and at a concentration of 100mg/mL with a 09mm zone of inhibition.
MIC and MBC/MFC of the Moringa Extracts on the Test Organisms
The MIC and MBC/MFC of the methanol and petroleum ether extracts on the test organisms is presented in Table 3. From the data presented, the MIC for the test organism ranged from 0.78 to 50mg/mL while the MBC/MFC ranged from 6.25 to 200mg/mL. The MIC of the petroleum ether extracts ranged from 50 to 200mg/mL while the MBC/MFC ranged from 200 to 800mg/mL.
Assay for the LC50 of Moringa extracts by Brine Shrimp Lethality Test
Brine shrimp lethality toxicity assay of the plant extracts is presented in table 4. The brine shrimp results in this study are interpreted as follows: LC50 <1.0 µg/mL -highly toxic; LC50 1.0-10.0 µg/mLtoxic; LC50 10.0-30.0 µg/mL -moderately toxic; LC50 >30 <100µg/mL -mildly toxic, and > 100µg/ml as non-toxic (Moshi et al., 2010). From the data presented in table 4, the LC50 for MME is 93.48µg/mL while the LC50 MPE is 3.691µg/mL, From this result, MPE is toxic while MME is mildly toxic.
In this study Moringa methanol was found to be more active than moringa petroleum ether extract. Moringa methanol inhibited the growth of all the organisms it was tested on. Moringa petroleum ether extract on the other hand inhibited only the growth of Candida albicans at a concentration of 100mg/mL which gave a 09mm zone of inhibition. However when the concentration was increased from 200-800mg/mL it was found to inhibit the growth of the organisms and it also had bactericidal and fungicidal effect. A similar study was conducted and moringa leaf petroleum ether extract was found to be active at this same concentration as reported (26). It was also revealed in this study that moringa methanol extract possessed all the phytochemicals it was screened for. Alkaloids and saponins were absent in moringa petroleum ether extracts, this could be the reason for its poor activity. However, a similar research carried out (30), reported the presence of Alkaloid and Saponin but in low amount. Alkaloids are naturally occurring chemical compounds containing basic nitrogen atoms. They often have pharmacological effects and are used as medications and recreational drugs (31). Saponins cause hemolysis of red blood cells (32).
Other solvents and extraction method
should be used for the extraction of the plant material to see if the performance of the extracts will be better. 4. The plant extracts should be evaluated in vitro to ascertain their activity on ear, nose and throat fungi and vancomycin resistant cocci and also to evaluate their effect on vital organs of the body.
5.
Combination of Moringa with plants with lower toxicity should be encouraged.
CONCLUSION
The present study deduced that, Moringa methanol extract was more active than Moringa petroleum ether extracts, although Moringa petroleum ether extract was experimentally more active on C. albicans than on vancomycin resistant coccal bacteria. Aspergillus fumigatus was the most predominant fungus while Staphylococcus aureus was the most predominant cocci associated with ENT at the time of this study. Moringa methanol extract was more active on S. aureus, S. pyogenes, S. pneumoniae, Aspergillus fumigatus and C. albicans. | 2019-04-02T13:08:25.527Z | 2017-06-22T00:00:00.000 | {
"year": 2017,
"sha1": "5f606db45bcbafcabae5de3e7a4279f98119d540",
"oa_license": null,
"oa_url": "https://www.ajol.info/index.php/ajcem/article/download/157977/147577",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "e9331659eedafffbdd6b4c7c0ba777724cea85ac",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Biology"
]
} |
84896126 | pes2o/s2orc | v3-fos-license | Exploring the Acceptor Substrate Recognition of the Human β-Galactoside α2,6-Sialyltransferase
Human β1,4-galactoside α2,6-sialyltransferase I (ST6GalI) recognition of glycoprotein acceptors has been investigated using various soluble forms of the enzyme deleted to a variable extent in the N-terminal half of the polypeptide. Full-length and truncated forms of the enzyme have been investigated with respect to their specificity for a variety of desialylated glycoproteins of known complex glycans as well as related proteins with different carbohydrate chains. Differences in transfer efficiency have been observed between membrane and soluble enzymatic forms, indicating that deletion of the transmembrane fragment induces loss of acceptor preference. No difference in substrate recognition could be observed when soluble enzymes of similar peptide sequence were produced in yeast or mammalian cells, confirming that removal of the membrane anchor and heterologous expression do not alter enzyme folding and activity. When tested on free oligosaccharides, soluble ST6GalI displayed full ability to sialylate free N-glycans as well as various N-acetyllactosaminyl substrates. Progressive truncation of the N terminus demonstrated that the catalytic domain can proceed with sialic acid transfer with increased efficiency until 80 amino acids are deleted. Fusion of the ST6GalI catalytic domain to the N-terminal half of an unrelated transferase (core 2 β1,6-N-acetylglucosaminyltransferase) further showed that a chimeric form of broad acceptor specificity and high activity could also be engineered in vivo. These findings therefore delineate a peptide region of ∼50 amino acids within the ST6GalI stem region that governs both the preference for glycoprotein acceptors and catalytic activity, thereby suggesting that it may exert a steric control on the catalytic domain.
Human 1,4-galactoside ␣2,6-sialyltransferase I (ST6GalI) recognition of glycoprotein acceptors has been investigated using various soluble forms of the enzyme deleted to a variable extent in the N-terminal half of the polypeptide. Full-length and truncated forms of the enzyme have been investigated with respect to their specificity for a variety of desialylated glycoproteins of known complex glycans as well as related proteins with different carbohydrate chains. Differences in transfer efficiency have been observed between membrane and soluble enzymatic forms, indicating that deletion of the transmembrane fragment induces loss of acceptor preference. No difference in substrate recognition could be observed when soluble enzymes of similar peptide sequence were produced in yeast or mammalian cells, confirming that removal of the membrane anchor and heterologous expression do not alter enzyme folding and activity. When tested on free oligosaccharides, soluble ST6GalI displayed full ability to sialylate free N-glycans as well as various N-acetyllactosaminyl substrates. Progressive truncation of the N terminus demonstrated that the catalytic domain can proceed with sialic acid transfer with increased efficiency until 80 amino acids are deleted. Fusion of the ST6GalI catalytic domain to the N-terminal half of an unrelated transferase (core 2 1,6-N-acetylglucosaminyltransferase) further showed that a chimeric form of broad acceptor specificity and high activity could also be engineered in vivo. These findings therefore delineate a peptide region of ϳ50 amino acids within the ST6GalI stem region that governs both the preference for glycoprotein acceptors and catalytic activity, thereby suggesting that it may exert a steric control on the catalytic domain.
Glycosyltransferases belong to a family of hundreds of enzymes that catalyze the transfer of a sugar from a sugar nucleotide donor to glycoproteins and glycolipids in eukaryotic cells and to polysaccharides in bacteria. Many of these enzymes are in charge of assembling cell-surface glycoconjugates, for which the extent of sialylation has been increasingly recognized as the molecular determinant of a wide array of biological recognition processes as different as viral and bacterial adhesion, immune response, and neuronal outgrowth (1,2). These events equally involve all types of glycoconjugates and different glycosidic linkages, but the key specificity of such complex molecular recognition is still unclear to us. It thus appears of primary importance to elucidate how glycan expression at the cell surface is controlled by glycosyltransferases. Especially within a transferase family sharing similar catalytic properties, the determinants governing specificity for their substrates and their underlying respective acceptor(s) need to be identified to understand the structural information that may be encoded by these enzymes through such a remarkable diversity.
Sialyltransferases constitute a family of at least 15 distinct glycosyltransferases that catalyze the transfer of sialic acid from CMP-NeuAc to glycoproteins or glycolipids during biosynthesis in animals and man (3,4). A number of them have been cloned across the animal kingdom, including bacteria, suggesting a broad function throughout evolution (1). In eukaryotes, they share with the other glycosyltransferases a typical type II architecture with a short cytoplasmic N-terminal tail, a transmembrane fragment followed by a stem of variable length, and a C-terminal catalytic domain facing the lumen of the Golgi apparatus (5). Unlike other glycosyltransferase families, they display significant sequence homology in their catalytic domain, viz. three conserved peptide sequences designated as sialyl motifs L and S (6) and VS (7). Mutagenesis of the L motif, consisting of ϳ50 amino acids present in the middle of the catalytic domain, showed that it is largely involved in donor recognition (8), whereas the S motif, located closer to the C terminus, contributes to the binding of both donor and acceptor substrates (9). Sialyltransferases are generally named and classified according to the nature of the monosaccharide substrate and the type of linkage formed (10). Within a subfamily catalyzing the same glycosidic bond, sialyltransferase members may exhibit very low amino acid sequence identity as well as major differences in acceptor substrate recognition, i.e. in their requirement for simple or more complex acceptor structures (11). A typical example demonstrating such keen specificity in substrate recognition resides in the ST6GalNAc 1 group.
ST6GalNAcI, likely to be involved in the synthesis of the sialosyl-Tn antigen, exhibits the broadest specificity on (NeuAc␣2,3) 0 -1 (Gal1,3) 0 -1 GalNAc-O-Ser/Thr, whereas ST6GalNAcIII was shown to be specific for NeuAc␣2,3Gal-1,3GalNAc␣/-R without discriminating between ␣and -linked GalNAc (11,12). As a result, this enzyme can synthesize the GD1a antigen (12). ST6GalNAcII presents an intermediate specificity and is active on (NeuAc␣ 2,3) 0 -1 Gal1,3GalNAc-O-Ser/Thr (13), and isoforms I and II are considered to be responsible for mucin synthesis. The recently cloned ST6GalNAcIV displays a similar substrate specificity to that of ST6GalNAcIII, but prefers O-glycans to glycolipids (14). Another candidate for GD1a synthase activity has been identified in mouse and is referred to as ST6GalNAcV (15). Comparison of peptide sequences reveals interesting features since ST6GalNAcIII appears to be virtually devoid of a stem region, whereas ST6GalNAcI exhibits the longest variable region. It thus appears that in this group, the broader the acceptor specificity, the longer the stem region.
Other examples showing distinct acceptor preferences can be found among sialyltransferase families. Among the six enzymatically distinct human ␣2,3-sialyltransferases cloned to date, galactoside ␣2,3 sialyltransferase ST3GalI and ST3GalII display similar catalytic activity for Gal1,3GalNAc substrates common to O-linked glycans and glycolipids. ST3GalI prefers glycoproteins to glycolipids, whereas ST3GalII prefers glycolipids (16). In addition, ST3GalIII was found to utilize Gal1,3GlcNAc more efficiently than Gal1,4GlcNAc in vitro (17). Other studies also demonstrated that mouse ␣2,8 sialyltransferase II requires the presence of internal fucose on Nglycans as well as the polypeptide region to sialylate variants of the neural adhesion molecule at special IgG-like domains (18). It is thus believed that these enzymes should display specific structural features to transfer sugars to appropriate acceptors and to exhibit such exquisite specificity.
Very recently, three x-ray crystallographic structures of eukaryotic glycosyltransferases have been reported (19 -21) showing common structural features, which are as follows: (i) the occurrence of a structurally very similar catalytic domain consisting of a three-layer ␣//␣-fold and (ii) the presence of a large pocket on one face capable of accommodating both the donor and acceptor substrates, surrounded by conserved motifs identified for a given glycosyltransferase family (22). It appears that the recognition of donor substrate is mediated by residues that are found in the most conserved regions, but very little experimental background has so far delineated those residues that specify acceptor recognition. For several ␣2,3-fucosyltransferases, a region located in a hypervariable segment at the N-terminal end of the catalytic domain was shown to account for differences in acceptor recognition (23)(24)(25). It has also been demonstrated that soluble forms of some transferases are less efficient in vivo than their membrane-bound counterparts in glycosylating endogenous acceptors (26), whereas no difference was observed for other classes of enzyme (27). These findings suggest that acceptor recognition may reside in variable regions, and as a result, we investigated whether or not it could be located within the stem region. We therefore compared the activity of membrane and soluble forms of human ST6GalI (EC 2.4.99.1) for glycoprotein acceptors of known glycan structure and show here that deleting the stem region of this transferase results in loss of acceptor preference together with increased transfer efficiency. We thus tentatively conclude that, in vivo, the juxtamembrane region of ST6GalI restricts enzyme specificity and activity.
EXPERIMENTAL PROCEDURES
Materials-Asialofetuin and human protein acceptors, neuraminidase, CMP-NeuAc, p-nitrophenyl phosphate, and cacodylic acid were purchased from Sigma. Synthetic acceptor substrate (LacNAc-O-spacer-biotin) was purchased from Syntesome. Biotinylated Sambucus nigra lectin was from EY Laboratories. Streptavidin-alkaline phosphatase conjugates were purchased from Jackson ImmunoResearch Laboratories, Inc. Restriction enzymes, purified rat liver ␣2,6-sialyltransferase, and Taq polymerase (high fidelity) were from Roche Molecular Biochemicals. FLAG-cytomegalovirus plasmid was from Eastman Kodak Co. pcDNA3 was from Invitrogen. fluorescent-assisted carbohydrate electrophoresis gels were from Bio-Rad. All culture reagents were from Life Technologies, Inc. Oligonucleotide primers were from and plasmid sequencing was carried out by Eurogentec.
Cloning of Human ST6GalI and Expression of Soluble and Membrane Variants-The cDNA encoding ST6GalI was isolated from an HL-60 cDNA library and further cloned in the FLAG-cytomegalovirus expression vector containing the preprotrypsinogen signal peptide upstream from the FLAG sequence. Sequencing of the insert showed the same sequence as the one published earlier (28). The full-length sequence was also deleted of the same N-terminal portion for the soluble form to be constructed in stably transfected CHO-K1 cells.
Various truncated forms of human ST6GalI lacking the transmembrane fragment were cloned into the pFLAG expression vector. The ⌬1-28 form of ST6GalI was stably transfected in CHO-K1 cells for comparison with the membrane ST6GalI form, and a chimeric form was constructed with the N-terminal (amino acids 1-52) membrane portion of the rat core 2 1,6-GlcNAc-transferase and ⌬1-70 of human ST6GalI and also tagged at the N terminus. The six soluble forms of ST6GalI deleted to a variable extent (⌬1-28, ⌬1-35, ⌬1-48, ⌬1-60, ⌬1-80, and ⌬1-100) were transiently expressed in COS cells as summarized in Scheme 1.
The cell culture media were collected over a 48 -72-h period and further concentrated Ͼ10-fold by Centriprep centrifugation. Each batch was assayed for activity on asialofetuin to calibrate enzyme production and stored in small aliquots of similar activity at Ϫ20°C. When necessary, the amount of the tagged protein was further estimated by competitive immunoassays using anti-FLAG antibodies. The stable cell line was found to secrete ϳ10 ng of FLAG-ST6GalI/10 6 cells/h. The specific activity of this enzymatic form was estimated to be 7.5 units/mg of protein.
Human soluble ST6GalI was produced in Pichia pastoris after truncation of the N terminus and the transmembrane fragment as described (29). This deletion was therefore very similar to the FLAG-ST6GalI construct. The activity present in the supernatant was 0.8 milliunits/mg of yeast protein.
Transfection-Transfection of cells was carried out using FLAGcytomegalovirus plasmids containing the different ST6GalI constructs and the LipofectAMINE (Life Sciences) procedure as recommended by the manufacturer. Semiconfluent cells were routinely transfected with fluorescein isothiocyanate; hCG, human chorionic gonadotropin; hTSH, human thyroid-stimulating hormone; dansyl, 5-dimethylaminonaphthalene-1-sulfonyl. SCHEME. 1. Scheme of various human ST6GalI constructs. The N-terminal peptide portion of human ST6GalI was progressively truncated up to residue, 100, which was predicted to correspond to the boundary between the stem and the catalytic domain (indicated by shading). The nomenclature used indicates the number of amino acids deleted. The chimeric construct C2GnT-ST6GalI⌬1-70 (where C2GnT is core 2 1,6 N-acetylglucosaminyltransferase) was obtained by joining the N-terminal peptide segment of human core 2 6-N-acetylglucosaminyltransferase (amino acids 1-52) to the truncated ST6GalI form (⌬1-70). The checkered boxes indicate the transmembrane domain (TMD). ϳ5 g of plasmid DNA. Stably transfected CHO-K1 cells were selected using Geneticin at a concentration of 500 g/ml over a time period of at least 4 weeks. Transient expression was routinely performed in COS cells using a lower selection pressure of the antibiotics.
Solid-phase Assay for ST6GalI Acceptor Specificity-The preference of ST6GalI for glycoprotein acceptors was assayed according to a solidphase procedure described earlier by us (30) using the sialic acidspecific lectin S. nigra (SNA) as reported by Mattox et al. (31). Briefly, coating of microtiter plates was carried out using increasing amounts of protein acceptors in 100 l of 50 mM PBS (pH 7.5) overnight at 4°C. After saturating wells with 2% bovine serum albumin and washings, coated glycoproteins were treated with neuraminidase from Clostridium perfringens overnight at room temperature (30). The extent of sialic acid removal was assessed on each plate using fetuin (1 g) for comparison with a standard preparation of asialofetuin taken as a reference. Washings were performed with PBS containing 0.1% bovine serum albumin and 0.05% Tween 20. Incubation with the various forms of ST6GalI was carried out in 50 mM cacodylate buffer (pH 6.5) containing 32 M CMP-NeuAc and 2 mM MnCl 2 in a final volume of 100 l for 1-2 h at 35°C. Cell supernatants containing soluble forms of ST6GalI were concentrated Ͼ10-fold using Centriprep columns. Immunopurification of sialyltransferase variants was performed by incubating the concentrated supernatant overnight with anti-FLAG antibodies coupled to agarose beads and further released by 25 g of FLAG peptide at 4°C prior to use. Intracellular full-length membrane ST6GalI forms were extracted with 0.5% Triton X-100 in 50 mM PBS for 45 min at 0°C, and a 10,000 ϫ g supernatant was immediately tested for activity.
Sialic acid linked in the ␣2,6-position was measured by amplifying SNA binding to the reaction product using a streptavidin-alkaline phosphatase conjugate as previously described (31). All assays were performed in quadruplicates, and the data are expressed as the means of these values. Nonspecific binding was assessed using non-transfected cell media or extracts and was generally found to give an absorbance of 0.2-0.3. This value was deduced from the assays in all graphs. Since different forms of the ST6GalI enzyme were assayed on several acceptors containing glycans exhibiting variable content of available galactose residues, the activity of each transferase form had to be calibrated. For the commercial enzyme, 1 unit of activity is the amount of enzyme catalyzing the transfer of 1 mol of NeuAc to asialofetuin/min at 37°C. In the first step, we standardized this activity to a standard calibrator preparation of this protein under the above conditions to determine the comparative activity of soluble forms of the rat liver enzyme. Under these conditions, solid-phase assays were designed to contain at least 0.7 milliunits of recombinant soluble forms when tested on asialofetuin, reaching an absorbance of 0.8 Ϯ S.D. When various deletion variants were compared, the activity of each enzyme was adapted to this absorbance per 1 g of acceptor/60 min, a period of time of linear velocity over which sialic acid is transferred as a function of time.
Liquid-phase Assays of Secreted Soluble ST6GalI Variants-Standard reactions were conducted at 37°C for 15-60 min in a final volume of 50 l in the presence of 50 M CMP-NeuAc, 3.5 M CMP-[ 14 C]NeuAc (110,000 cpm), 1 mM LacNAc-O-spacer-biotin or asialofetuin (2 mg/ml) as acceptor, and 2 mM MnCl 2 in 50 mM cacodylate buffer (pH 6.5). The reaction was initiated by addition of the cell supernatant. In assays with the LacNAc derivative, the reaction was stopped by addition of 3 ml of cold water, and the mixture was applied to a Waters Sep-Pak C 18 reverse-phase chromatography cartridge. After washing with 15 ml of water, the radiolabeled product was eluted with 3 ml of methanol and counted with 3 volumes of liquid counting scintillant (Ultima-FloM, Packard Instrument Co.). In assays with asialofetuin, the reaction was stopped by addition of 0.5 ml of cold water and applied to Nanosep TM In all assays, consumption of substrates was kept below 15-20% to ensure accurate initial rate measurements, and each test was done in duplicate.
Fluorescence-assisted Carbohydrate Electrophoresis-To analyze sialic acid transfer to N-glycans of different structure, carbohydrate chains from asialofetuin and native or recombinant glycoprotein hormones were released by N-glycanase and further labeled with a 2-ami-nobenzoic acid chromophore as recommended by Bio-Rad. Labeled glycans were incubated with or without soluble ST6 forms from CHO cells or yeast. The reaction mixture was then precipitated with ethanol overnight at Ϫ20°C. The 15,000 ϫ g supernatants were evaporated and further dissolved in water. The labeled glycans were then separated by gel electrophoresis and detected at 540 nm using a Bio-Rad GlycoImager. Size calibration was carried out using a glucose ladder and an N-glycan library provided by the manufacturer.
Fluorescence Microscopy-Double labeling with FITC-SNA and anti-FLAG monoclonal antibody M2 was performed as follows. FITC-SNA labeling was assayed essentially as reported (32). Briefly, cells were fixed for 5 min with 0.5% paraformaldehyde in calcium-and magnesium-containing PBS and incubated with 50 g/ml lectin for 60 min on ice. For intracellular distribution of FLAG-tagged enzymes, FITC-SNAstained cells were fixed with 3.5% paraformaldehyde in PBS, permeabilized with 0.5% Triton X-100 in PBS containing 1% fetal calf serum, and incubated with 10 g/ml anti-FLAG monoclonal antibody M2 for 1 h. They were then rinsed three times and incubated with rhodamine isothiocyanate-conjugated goat anti-mouse IgG antibody for 1 additional h, and staining was visualized by confocal microscopy on a Leica instrument (Uniblitz Sutter Instrument Co.). Confocal images were processed with Metamorph Imaging System Version 3.5. Volumes were originally raytraced as 24-bit TrueColor images and transferred to Adobe Photoshop as 24-bit RGB TIFF files.
Protein Sequence Analysis-Protein sequences were retrieved from the GenBank TM /EBI Data Bank and analyzed using BLAST programs (33) and ClustalW (34). Secondary structure predictions were performed using programs available on the NPS and JPred servers.
Differential Activity of Full-length and Soluble Forms of
ST6GalI-The specificity of the rat enzyme has been extensively characterized (5-9); and since sequence analysis of rat liver and human ST6GalI showed 88% identity, it is assumed that the specificity of both enzymes is very similar. Of particular interest is the earlier observation by us that recognition of the human enzyme for LacNAc residues includes the branch downstream of the ␣-linked mannose (35), confirming the earlier observation that, like the rat enzyme, ST6GalI can discriminate glycan acceptors on the basis of their branching pattern (36). Preliminary testing showed that the purified liver preparation was more stable than the native form of the human enzyme extracted from transfected cells using a detergent. We therefore used the rat liver transferase to compare the acceptor preference of the same soluble forms of human ST6GalI expressed in two different systems for various glycoproteins of known glycan structure.
First, we investigated the influence of the branching pattern on enzyme specificity based on the earlier results that fetuin, human transferrin, and human ␣ 1 -acid glycoprotein are efficient acceptors for the native enzyme (37). These glycoprotein acceptors all contain complex-type N-glycans with terminal LacNAc groups, but differ in branching pattern and sialylated bonds. Human transferrin contains biantennary glycans sialylated in the ␣2,6-position; fetuin contains 2,2,4-triantennary glycans with both ␣2,3and ␣2,6-linked sialic acids; and ␣ 1 -acid glycoprotein contains 2,2,4,6-tetraantennary N-glycans with both ␣2,3and ␣2,6-sialic acid. The presence of ␣2,6-linkages was confirmed by SNA binding to native preparations, and loss of lectin binding confirmed the removal of sialic acid by the neuraminidase treatment (data not shown).
We thus first compared the specificity for these acceptors of native and soluble forms of ST6GalI produced in mammalian cells or yeast. When the desialylated glycoproteins were incubated with membrane ST6GalI in the presence of CMP-NeuAc and the reaction products were probed by lectin binding (Fig. 1), efficient transfer was recorded in all cases, with affinity values similar to those reported previously, with the membrane enzyme having a slightly better recognition of asialofetuin compared with other desialylated glycoproteins (Fig. 1A). These results with the rat transferase are similar to those previously ST6GalI Acceptor Recognition reported (37), and this assay has been repeated here for comparative purposes. Interestingly, the soluble forms of similar length produced in CHO cells (Fig. 1B) or yeast (Fig. 1C) equally recognized these acceptors, independently of the expression system. These findings show that the soluble enzymes did not differ in binding asialoglycoprotein acceptors and therefore suggest that differential acceptor recognition by the full-length transferase has been altered by truncation of the membrane anchor.
Transfer Efficiency of Soluble FLAG-ST6GalI for the Glycoprotein Hormone Family-To understand whether or not the enzyme may recognize a determinant present in the polypeptide backbone that could influence branch specificity, we investigated a family of structurally related proteins sharing bi-or triantennary glycans differing in terminal glycosylation. Glycoprotein hormones are heterodimers sharing a common ␣-subunit and differing in their -subunit (38,39). Placental hCG contains biantennary glycans; urinary hCG contains mostly triantennary oligosaccharides with ␣2,3-sialic acid as terminal sugars; and follicle-stimulating hormone is known to display triantennary glycans terminating in both ␣2,3and ␣2,6-sialic acids. Pituitary hTSH contains biantennary glycans largely capped with sulfated GalNAc, but also terminating in ␣2,3-linked sialic acid. When produced in CHO cells, recombinant hTSH was found to exhibit triantennary glycans with low ␣2,3-sialic acid content (38,40).
When incubated in the presence of soluble FLAG-ST6GalI, both urinary hCG ( Fig. 2A) and human pituitary follicle-stimulating hormone (Fig. 2B) behaved as efficient acceptors in their native state, indicating that their glycans display sites available for receiving additional sialic acid. As anticipated, desialylated forms were found to be better acceptors, demonstrating that their branching pattern has no influence in the transfer reaction and that the soluble enzyme does not discriminate LacNAc groups originally sialylated in the ␣2,3-position. It was concluded that like the plasma proteins tested above, these glycoprotein acceptors do not appear to contain any structural features governing sialic acid transfer in the ␣2,6-position to specific lactosaminyl groups. These data are in total agreement with previous structural work on hCG remodeling using soluble bovine ST6GalI (41).
Pituitary hTSH was anticipated to be a very poor acceptor because of its high content of sulfated sugar (Fig. 3A). However, when recombinant hTSH was tested as an acceptor, the native preparation could readily be over-sialylated to the same extent as the hormone treated with neuraminidase (Fig. 3B). In this regard, this glycoprotein hormone reacted similarly to asialofetuin and asialo-follicle-stimulating hormone, indicating that the recombinant product could be extensively sialylated in vitro. Similar data were obtained with the ST6GalI⌬1-28 form produced in yeast (data not shown), confirming again that 2. In vitro sialylation of gonadotropins. A, a highly purified preparation of urinary hCG (E) was treated with (f) or without (q) neuraminidase before incubation with soluble FLAG-ST6GalI (f and q) produced in CHO-K1 cells as described under "Experimental Procedures." B, a highly purified preparation of native human pituitary follicle-stimulating hormone (hFSH) (‚) was treated with (ࡗ) or without (OE) neuraminidase before incubation with soluble FLAG-ST6GalI produced in CHO-K1 cells as described under "Experimental Procedures." heterologous expression per se does not affect the catalytic properties of the transferase. Rather, it is likely that the transferase deleted of its membrane portion is no longer able to recognize features present in the various acceptors necessary to specify sialic acid transfer.
Sialic Acid Transfer to Free N-Glycans and Derivatized Disaccharides-To further investigate to what extent soluble forms of ST6GalI are able to sialylate N-linked glycans, we analyzed if these enzymatic forms could achieve the completion of complex oligosaccharides. Glycans released from glycoprotein acceptors by endoglycosidase treatment were incubated with the soluble forms produced in CHO cells or yeast and further derivatized for separation on gels according to their sialic acid content. Fig. 4 presents evidence that the glycans originally present in asialofetuin (lane 2) could be sialylated to a similar extent as the most negatively charged originally present in the fetuin preparation (lane 5). Urinary hCG glycans (lane 4) were similarly completed (lane 3) by the enzyme, further confirming that the protein acceptor is totally dispensable for sialic acid transfer. Whether these glycans were of low sialic acid content in their native form or desialylated by neuraminidase treatment did not affect enzyme efficiency. Therefore, it appeared again that the soluble transferase fully recognized the entire panel of glycan structures whatever the expression system used, yeast or CHO cells (lanes 7 and 8).
To further elucidate the contribution of the glycan moiety to the specificity of soluble ST6GalI, we studied the capacity of soluble enzymes to transfer sialic acid to Gal1,4GlcNAc-O(CH 2 ) 6 -NH-dansyl and Gal1,3GlcNAc-O(CH 2 ) 6 -NH-dansyl derivatives. When the reaction was carried out in solution, dansylated substrates were allowed to follow the reaction as a function of time by fluorescence-assisted carbohydrate electrophoresis since only the sialylated trisaccharide products mi-grated on the gel because of the negative charge. Fig. 5 shows that both types of dansylated substrates can be sialylated over a period of 2-24 h. The benzyl-Gal1,4GlcNAc and octyl-Gal1,3GlcNAc glycosides were also tested by solid-phase assays to determine if coating the acceptors on a solid phase can affect the kinetics of sialic acid transfer. Soluble FLAG-ST6GalI⌬1-28 could sialylate both disaccharides as a function of time over a 3-h period as detected by lectin binding (data not shown). Again, no difference in substrate recognition could be observed for the soluble forms whether they had been expressed in CHO cells (Fig. 6A) or yeast (Fig. 6B). Analysis of the dansylated products by mass spectrometry after methylation indicated that both compounds were sialylated on the galactose residue through an ␣2,6-linkage, but some substitution could be detected on the GlcNAc residue. 2 These data thus confirmed that the soluble catalytic domain of ST6GalI is able to recognize a substrate as small as a disaccharide. They are in good agreement with earlier observations on the full-length transferase showing that, in vitro, the 6-OH of galactose and the amide group of the GlcNAc residue are the only groups essential for efficient transfer of sialic acid (42).
Effect of Deleting the Stem Sequence of ST6GalI on Acceptor Preference-Various protein sequence analysis methods were used to tentatively determine the boundary between the stem and the catalytic domain to identify the minimal size of the catalytic domain of the transferase. Comparison of known ST6GalI peptide sequences from different species showed the highest similarity in a region starting at about residue 100 down to the C-terminal end. In addition, predictive methods based on hydrophobic cluster analysis indicated that this region is composed of alternating, regular secondary structure elements, whereas the preceding segment (amino acids 50 -100) appeared to be less organized and composed mostly of coiled regions. We therefore decided to progressively delete the N terminus up to residue 100 to assess to what extent this segment is involved in enzyme activity and to elucidate whether or not this portion may interact with the catalytic domain and affect acceptor preference. It was expected that selective acceptor recognition would be abolished, whereas cat-2 J. C. Michalski and C. Ronin, unpublished data.
ST6GalI Acceptor Recognition
alytic properties would be maintained. To compare the activity of the truncated ST6GalI variants, the activities of all enzymatic forms were matched using asialofetuin under conditions in which sugar transfer was displayed as suboptimal lectin binding and increased as a function of time. Under these conditions, sialic acid transfer was quantitatively similar for all dose-response curves and independent of enzyme concentration in the medium. Since multiple -Gal residues are present in different glycan structures for each glycoprotein acceptor, no classical kinetic parameters can be accurately measured in this approach, but only relative activity.
Five additional deletions in the stem regions were carried out and tested for activity; but for clarity, only those starting from amino acids Ϫ28 to Ϫ60 are shown in Fig. 7. Four variants with deletions over the first 80 amino acids were found to be enzymatically active, whereas truncation of ϳ100 amino acids abolished activity. Of interest, the transfer reaction was found to be increased as the deletion augmented from ⌬1-28 to ⌬1-60 (Fig. 7, A-C). For all three glycoproteins, optimal transfer was reached at lower acceptor concentration for the shortest enzymatic variant. Additional truncation between amino acids Ϫ60 and Ϫ80 did not further modify this property, but a more extensive deletion of 100 amino acids led to inactivation (data not shown). Half-maximal activity of the five active variants over the three acceptors indicated that transfer efficiency was augmented by a factor of 3-5 when the stem region was shortened (Table I). However, since these acceptors differ in LacNAc content, no conclusion could be drawn as to whether the removal of the N-terminal region affected acceptor recognition and/or the catalytic properties of the transferase.
To obtain a better understanding of the structural basis of increased transfer efficiency, we determined the kinetic parameters of the shortest and longest soluble truncated forms in 1-4) or Gal1,3GlcNAc (lanes 6 -9) over a time period of 2 h (lanes 1 and 6), 4 h (lanes 2 and 7), 8 h (lanes 3 and 8), or 24 h (lanes 4 and 9). The products were analyzed by gel electrophoresis. Lane 10 represents the glucoselabeled ladder. ST6GalI Acceptor Recognition liquid-phase assays using a radioactive donor and similar enzyme concentrations. Kinetic analyses were performed using both a synthetic acceptor (Gal1,4-GlcNAc-O-spacer-biotin) and a glycoprotein acceptor (asialofetuin). As shown in Table II, there appeared to be no difference in affinity between both enzymatic forms for the common donor and the two acceptors. They both exhibited K m values of ϳ18 M for CMP-NeuAc, 1 mM for LacNAc, and 150 M for asialofetuin, in the same order of magnitude as those reported in the literature for the rat enzyme (42). Thus, removing ϳ60 amino acids upstream from the catalytic domain did not alter the recognition parameters, but rather modified the velocity of sialic acid transfer. Catalytic efficiency was found to be augmented by a factor of 3 for CMP-NeuAc, by a factor of 36 for LacNAc, and by a factor of 10 for asialofetuin. To ensure that truncation did not modify biosynthesis and secretion of the variants during transfection, Western blot analysis of the intracellular content was carried out using the same amount of total proteins and showed no difference in the various truncated forms of ST6GalI. Similarly, cell supernatants showed no significant variation in the secretory rate of the variants in the medium (data not shown). Gel electrophoresis carried out under mild reducing conditions did not display FLAG-tagged proteins of high molecular weight, ruling out that enzyme activity can be related to association with unknown proteins or enzyme oligomerization. In addition, confocal microscopy using FITC-SNA binding further showed that each transfected cell exhibited a higher content of sialylated glycoconjugates at the cell surface for the ⌬1-80 variant compared with the ⌬1-28 variant, 3 further indicating that both forms were also differentially active on endogenous acceptors.
Altogether, these data demonstrate that releasing ST6GalI as a soluble enzyme significantly altered the conformation of the transferase to broaden its acceptor specificity and to accelerate the transfer reaction. It is therefore proposed that in the full-length enzyme, the stem region interacts with the catalytic domain to expose a specific recognition site for glycoprotein acceptors. Removing this modulatory region probably released steric constraints and induced a conformational change that opened up the active site and thus facilitated sugar transfer.
To elucidate whether or not membrane anchorage can indeed contribute to such an interaction, a chimeric protein was constructed by fusing the shortest and most active truncated variant to an unrelated N-terminal moiety of another transferase. The catalytic domain of ST6GalI was thus extended with the N-terminal portion of the core 2 6-GlcNAc-transferase tagged with the FLAG epitope (Scheme 1), and the chimeric form was stably transfected in CHO cells. The activity was then compared for these full-length membrane enzymes both in vivo and in vitro. Fig. 8 (A and B, panels 2) shows the subcellular immunolocalization of these transferases using anti-FLAG antibodies. Both enzymes were localized to the Golgi apparatus very similarly. The respective activities of the full-length forms were then compared in vivo and in vitro. As shown in Fig. 8 (A and B, panels 1), both ST6GalI and the chimeric transferase proved to be able to sialylate endogenous cell-surface glycoconjugates to a large extent. However, full-length ST6GalI led to SNA binding distributed in the Golgi and periplasmic compartment, whereas the cell surface exhibited spotted labeling (Fig. 8A, panel 1). In striking contrast, the cell surface of cells expressing the chimera was intensely highlighted by the fluorescent lectin (Fig. 8B, panel 1). It was concluded that the chimeric form was less restrictive and more efficient in sialylating cell-surface acceptors than the wild-type enzyme.
To tentatively compare the acceptor recognition of the fulllength, chimeric, and secreted forms of ST6GalI, detergent extracts of different CHO clones were assayed in solid-phase assays for their respective activity for asialofetuin (Fig. 9A) or asialotransferrin (Fig. 9B). We observed that human fulllength ST6GalI had lower activity for transferrin like the rat enzyme presented in Fig. 1A. The chimeric enzyme in which the N-terminal anchor and the stem region were substituted generated a transfer activity equally efficient for both fetuin and transferrin acceptors. The half-maximal transfer rate was found to be similar to that for the corresponding truncated variant, again reproducing increased efficiency compared with that of the full-length enzyme. These findings further support the conclusion that the membrane anchor does not participate by itself in acceptor specificity, but may rather constrain the stem region through lipid packing and somehow sustain its interaction with the catalytic domain of the transferase.
DISCUSSION
This study was aimed at exploring the acceptor preference of the human ST6GalI sialyltransferase and infers that recognition of the glycan, followed by sugar transfer, is governed by a short peptide sequence present in the juxtamembrane portion of the protein. Among glycosyltransferases, the ␣2,6-sialyltransferase has been extensively characterized because this enzyme terminates most of the N-glycans in human serum proteins, thereby controlling both their duration in blood and metabolic clearance (1). Like most of the glycosyltransferases, the rat liver transferase is a type II membrane Golgi protein of a well characterized domain structure. It has been widely assumed that the stem region protrudes into the lumen because proteolytic degradation has been shown very early on to release an active soluble form missing 62 amino acids (36). Within a homologous glycosyltransferase group, the length and primary structure of catalytic domains are relatively well conserved, and variability in the molecular size of these enzymes is often attributable to differences in the length of the stem region. This polypeptide portion tethers the catalytic domain to the membrane anchor and generally displays highly variable amino acid sequence. It is thought to consist of a flexible stretch of little secondary organization. The longest stem reported so far (416 amino acids) was observed in N-acetylgalactosaminyltransferase, a region rich in hydroxyamino acids containing many potential glycosylation sites (43). Conversely, ST6GalNAcIII was found to have a very short stem region, if any (12). Earlier on, it was proposed that the stem region may contribute to Golgi localization and/or enzyme oligomerization (44). It has also been very recently shown that the ST6GalI stem region is directly involved in Golgi retention and contains at least two possible cleavage sites responsible for enzyme secretion (45). However, because this portion has been found to be dispensable for enzyme activity in vitro, it has not been considered to be functionally related to the catalytic properties of the transferase, but rather to facilitate its access to the glycan acceptor by acting as a spacer arm.
Until recently, very little information was available concerning the acceptor preference of glycosyltransferases, but ongoing cloning of many new enzymes and characterization of their catalytic activity unraveled a very complex redundancy. It appears that each class of transferases can be divided in subfamilies displaying several activities capable of constructing the same sugar linkage on distinct glycoconjugates and that ex-hibit high but not necessarily exclusive preference for an acceptor. To date, a key step is therefore to understand the structural features that allow each glycosyltransferase within a family to discriminate among their glycoprotein or glycolipid acceptors. Consequently, it is also an intriguing question to understand how a cell equipped with a wide array of such enzymes can control the many combinatorial possibilities of glycosylation of its cell-surface glycoconjugates and/or secretory glycoproteins. Since full-length and soluble transferases transfected in mammalian cell lines were found to be differentially active at glycosylating cell proteins (26), it has been suggested that membrane anchorage and intracellular localization are also important for enzyme activity. As an example, transfecting the full-length rat ␣2,6-sialyltransferase in mammalian cells was shown to result in limited addition of ␣2,6linked sialic acid, indicating a weak competition with endogenous enzymes (46). Indeed, the mode of action of glycosyltransferases in vivo is far from being fully identified, but it is widely admitted that the remarkable diversity of the cell glycosylation machinery should contribute to the synthesis of glycan structures with exquisite specificity. This study has deciphered the minimal catalytic domain of ST6GalI and showed that it is active in vivo in both membrane and soluble forms. Preliminary work has also revealed that all truncated variants of ST6GalI (⌬1-28 to ⌬1-80) are fully active on cellsurface acceptors in vivo and clustered in the Golgi apparatus together with endogenous sialylated acceptors. 3 These findings further confirm that discrimination of cell-surface glycoconjugates heavily relies on intracellular compartmentalization of the transferase. While this work was in progress, a second Golgi retention signal was reported for ST6GalI (45), which is in total agreement with our observation that, in the soluble enzymes, Golgi localization and activity are still interdependent for the newly synthesized glycoproteins trafficking in the trans-Golgi network to acquire their final glycan structure.
Using site-directed mutagenesis, several laboratories have elucidated the location of substrate-binding sites of sialyl-and fucosyltransferases over the last few years. Structure-function relationships of the rat liver enzyme ST6GalI have been well advanced, especially the role of the two sialyl motifs in catalysis (6,8,9). In this regard, it was shown that LacNAc was less effective than glycoprotein acceptors, partly because enzyme specificity extends over glycan branching (35,36,41). Although the S sialyl motif is involved in the recognition of both the donor and the lactosaminyl substrates (9), it is anticipated that regions located elsewhere in the protein should also contribute to the display of glycan specificity through more distal interactions. In this regard, recent work on homologous ␣3fucosyltransferases demonstrated that residues important for discriminating related substrates are located within a hypervariable segment of the catalytic domain close to the stem region (23)(24)(25). Using subdomain swapping with human ␣1,3fucosyltransferases III and VI, a hypervariable region of 11 amino acids (positions 103-153) at the N terminus of the catalytic domain has been found to contain structural information for Gal1,3GlcNAc-R versus Gal1,4GlcNAc specificity (23). When the stem region was deleted in human fucosyltransferases V (amino acids 76 -374) and III (amino acids 62-361), which then exhibited 93% identity, two regions in the N-terminal half of the catalytic domain were also identified to induce an ␣1,3to ␣1,4-switch, representing a difference in 8 and 12 amino acids between the two protein sequences, respectively (24). These studies therefore delineate the existence of discrete subsites within this family of enzymes sharing similar ␣1,3/4activity that are responsible for the various aspects of enzyme specificity. Recently, site-directed mutagenesis demonstrated that, in bovine fucosyltransferase b-encoded ␣1,3-transferase, a single amino acid substitution at position 115 is sufficient to switch the substrate specificity, whereas a mutation at position 116 regulates the transferase efficiency of bovine ␣1,3-fucosyltransferase and human ␣1,3/4-fucosyltransferase III (25). This work lends further support to the possibility that, in fucosyltransferases, the stem region may participate in transferase activity by creating a highly specific acceptor site in the spatial vicinity of the catalytic domain. Such a modulatory region allows several transferases involved in Lewis antigen synthesis to discriminate structurally related disaccharide substrates while catalyzing a similar fucose transfer through highly homologous catalytic domains.
Regarding ST6GalI, our data also favor the participation of the stem region in the selective recognition of glycan acceptors. Removal of this region proved to alter interactions with the catalytic domain that are essential for restricting access to acceptor subsite and for accommodating well defined glycans. This in turn affects the catalytic domain probably through further conformational change to readily proceed with sugar transfer, as if the enzyme has been turned on to a more open functional state. Since constructing a chimeric molecule containing the stem and anchor portions of an unrelated transferase (core 2 6-GlcNAc-transferase) showed that this latter transmembrane fragment does not reverse this effect, we also concluded that the role of the stem region may be specific for the transferase.
While this work was in progress, the crystal structures of three eukaryotic glycosyltransferases were elucidated (19 -21). These proteins were crystallized in the presence of UDP, thereby revealing the portion of the protein interacting with the nucleotide moiety and the involvement of the DXD motif previously identified in many glycosyltransferases (22). Despite the lack of sequence identity, these enzymes adopt a similar topology (␣//␣-fold), with the key amino acids involved in UDP binding apparently located at equivalent positions. Since only the catalytic domains of these glycosyltransferases have been crystallized, there is still no indication of a possible role of the stem region in acceptor recognition. In addition, whether or not sialyltransferases can adopt a similar topology is still difficult to assess for at least two reasons: (i) they use a different sugar donor (CMP); and (ii) they do not contain the DXD, motif but instead contain other specifically conserved motifs (6 -7, 22). Nevertheless, since acceptor-binding subsites may involve regions upstream of the catalytic domain for at least ␣3-fucosyltransferase and, as reported here, for ST6GalI, there are definitely structural features to be searched in the variable region that control the transfer reaction. Even though the stem region is predicted to be highly flexible, it often contains cysteine residues as well as several N-and/or Oglycosylation sites, which can contribute to a local conformation stabilized by membrane anchorage. Since truncated variants exhibited unchanged affinity for the donor and acceptor substrates, but displayed increased velocity parameters, it was concluded that they lost peptide elements crucial for the active site that possibly maintain the catalytic domain in a closed state. Alternatively, removal of this modulatory region may induce local changes that facilitate positioning of critical amino acids within the catalytic site to proceed with the transfer reaction. Further work based on crystallization of the fulllength transferase in the presence of its acceptor will thus be required to delineate both the location of the acceptor-binding site and the catalytic mechanism. Accordingly, variation in stem length and sequence within a glycosyltransferase subfamily may reflect variation in acceptor preference among transferases sharing similar properties like glycan specificity, recognition of the underlying glycosylation site, and/or the nature of the glycoconjugate acceptor.
Whether or not the stem region plays a physiological role in regulating glycosyltransferase action is still unknown. Since enzyme diversification occurs during evolution (2), it may very well be that modulatory sites could have been added to restrain catalytic properties of transferases to well defined classes of glycoconjugates and to create acceptor preference. Such a fine mechanism may allow site-specific sialylation like that found in the glycoprotein hormone family, for which alternate sulfate signaling and sialylation control hormonal activity and pulsatility (38). Also, the ␣2,8-sialyltransferase subfamily was found to exhibit exquisite glycosylation very specific for neural cell adhesion molecules during development (47) since polysialylation by ␣2,8 sialyltransferase II was demonstrated to require core fucosylation of neural-cell adhesion molecule variants, showing the highest preference of all transferases known to date, i.e. for both the glycan and protein acceptors (18). Furthermore, mice produce four isoforms of ␣1,3-galactosyltransferase that differ in the length of this region, indicating that polymorphism of a transferase polypeptide may physiologically occur, possibly introducing new regulatory features (48). Such a length polymorphism has also been described to occur in ST3GalIII (49). Moreover, it has been found that two naturally occurring ␣2,6-sialyltransferases differing by a single amino acid in their catalytic domain can differ in their enzymatic properties (50). If these findings prove to be applicable to all classes of glycosyltransferases, such functional complexity may very well have important implications during biosynthesis of cell-surface glycoconjugates in coordinating the action of multiple enzymes in a tissue-specific manner and in displacing the physiological balance in favor of those enzymes overexpressed under pathological conditions. | 2019-03-22T16:07:27.646Z | 2001-06-15T00:00:00.000 | {
"year": 2001,
"sha1": "b5bd29c3786786e176ca3786a1f3e3a12b561a00",
"oa_license": "CCBY",
"oa_url": "http://www.jbc.org/content/276/24/21608.full.pdf",
"oa_status": "HYBRID",
"pdf_src": "Adhoc",
"pdf_hash": "3df6c0fbb181cb188c344d8ba7328807f458d2dd",
"s2fieldsofstudy": [
"Biology",
"Chemistry"
],
"extfieldsofstudy": [
"Chemistry"
]
} |
55564710 | pes2o/s2orc | v3-fos-license | Screening for Dermatophytosis and other Skin Infections in Orphanages in and around Kalaburgi, India
1 Assistant professor, Department of Microbiology, ESIC Medical College, Gulbarga, India 2 MBBS Student, ESIC Medical College, Gulbarga, India 3 Professor & HOD, Department of Microbiology, ESIC Medical College, Gulbarga, India 4 Associate professor, Department of Microbiology, ESIC Medical College, Gulbarga, India 5 Assistant professor, Department of Microbiology, ESIC Medical College, Gulbarga, India *Corresponding author
Introduction
Superficial mycoses are infections of skin, hairs and nails caused by dermatophytes, yeasts and non-dermatophyte molds. Among these, dermatophytes are responsible for the largest number of cases. Dermatophytes are divided into three genera: Trichophyton, Epidermophyton and Microsporum. Distribution of the dermatophytes varies with the geographical area and course of time (Agarwal et al., 2014).
The high humidity and temperature provides a fertile ground for their abundant growth (Sen et al., 2006).
Dermatophytes can survive solely on outer cornified layers of the skin. The ability of certain fungi to adhere to particular host arises from numerous mechanisms and host factors, including the ability to adapt to the human body (Tainwala et al., 2011).
Dermatophytosis is a common contagious skin disease caused by fungi known as Dermatophytes. Dermatophytosis is most common in tropical and subtropical countries. Kalaburgi is a district in Karnataka where the hot and humid climate favours the fungal infections. Dermatophytoses in orphans reflects the status of health hygiene and personal cleanliness of a community. The present study aimed at screening Dermatophytosis and evaluating the pattern of the same. 175 orphans were screened for dermatophytosis. Samples were collected with aseptic precautions and processed for KOH mount and fungal culture (Sabarauds dextrose agar and Dermatophyte test medium). The isolates were confirmed by performing LPCB mount. Among 175 orphans 6(3.42%) were positive by KOH mount and culture (Sabarauds dextrose agar& DTM) had 4(2.28%) positivity. The prevalence of other diseases such as Pityriasis alba, Tinea versicolor, Scabies and Pediculosis capitis were 6.8, 2.2, 12.5, 0.5% respectively. This study detects epidemiological pattern, predominant organism causing dermatophytosis and other skin infections in orphanages.
Article Info
India is a large subcontinent with remarkably varied topography, situated within the tropical and subtropica belts of the world. Its climate is conducive to the acquisition and maintenance of mycotic infections Since dermatophytosis occurs most frequently during the monsoon, the present study was planned during this period (Singh Suman et al., 2003).
Accccording to World Health Organization (WHO), the prevalence rate of superficial mycotic infection worldwide has been found to be 20-25%, that of india is found to be 21-22% (Lakshmanan et al., 2015;Suman Saurabh et al., 2013).
Surveillance for fungal infections is important to define their burden and trends, to provide the infrastructure needed to perform various epidemiological and laboratory studies, and to evaluate interventions (Bassiri-Jahromi et al., 2009).
The aim of the study was to know the prevalence of cutaneous fungal infection and other infections especially Dermatophytosis. By knowing the prevalence we can estimate the treatment problem and take measures to prevent the spread of infection.
Materials and Methods
Source of the study: The study was conducted on students present in orphanages in and around kalaburgi Study period: The study was conducted for a period of 2 months from June to July 2015 Inclusion criteria: All the children present in the orphanages between the age group of 5 to 20 years were screened for Dermatophytosis.
Exclusion criteria: Students on antifungal treatment for dermatophytosis.
Study site: Orphanages in and around kalaburgi
Duration of study: 2 months (june & july)
Methodology
The present study was conducted in 5 orphanages in kalaburgi. Ethical consideration was obtained from the institution. Sample size was 175. We had prepared a proforma which included the data regarding the socio-demographic information such as age, sex, Address, history of presenting illness, General physical examination and local examination which was filled by the investigator. Each child was be examined in bright ambient light to search for disease of the skin and its appendages. The findings are recorded in the same form. If there are suspected dermatophytosis like scaly lesions on skin, alopecia, discolouration/pitting nails the specimen were collected under aseptic precautions with blunt end of sterile blade for skin lesions, hairs were plucked, nail clippings were collected and immediately sent to microbiology laboratory for further processing. In the laboratory If the specimen is skin scrapings/hair/nail clippings, KOH mount was performed and simultaneously inoculated on Saborauds dextrose agar and Dermatophyte test medium. The cultures were incubated at room temperature for 28 days. The tubes were observed for growth daily and if found, the day on which appearance of colony was noted and LPCB mount were done to identify the fungi. If there was no growth upto 28 days the specimen was considered to be negative Finally the data was gathered using the proforma clinical examination, and culture details are compiled, coded and entered in excel spread sheet. And this data was tabulated and were analyzed using chi square test to study the association between various factors. P value < 0.05 will be considered statistically significant. P value < 0.001 is highly significant. The data will be represented as number and percentage.
Results and Discussion
Out of 175 screened orphan children Males were were about 80 (45.71%) and females were 95 (54.28%). (Table 1) Dermatophytes was highly significant with prevalence of 3.43%. Prevalence of other skin diseases like Pyteriasis alba, T.versicolor and scabies were 6.86%, 2.29%, 12.5% respectively. (Table 2 and Table 3) Among 6 dermatophytes 6 were positive by KOH, 4 were positive by Sabaraud's dextrose agar and Dermatophyte test medium. (Table 4) Among 4 grew on culture media 3 were Trichophyton mentagrophytes and 1 was Trichophyton verrucosum. (Table 5) The present study covered the spectrum of dermatophytosis in a specialised population group of orphanages. The study included 175 children between age group 5-20 years among which Males were 80 and females were 95 in number.
The prevalence of dermatophytosis in our study was 3.42%.The low prevalence might be due to the target population and age group we have selected. There are hardly any studies of dermatophytosis on orphanages.
The KOH mount had prevalence of 3.42% in comparison with Culture (Sabarauds dextrose agar& DTM) which had 2.28%.
Among 6 dermatophytic infections 3 were Tinea capitis, 2 were T.corporis and 1 was T.cruris. This shows that prevalence of Tinea capitis was more due to overcrowding, poor hygiene and other factors.
The prevalence of overall fungal infections in our study was 12% which explains high prevalence due to environmental factors like semi-arid region, hot and moist climate.
Other disease apart from fungal infection was mainly comprised of Scabies and Pediculosis capitis which had 12.5 and 0.5% prevalence and the disease was observed in government orphanage. The incidence was increasing in that place due to inappropriate treatment and awareness about the disease. The prevalence of dermatophytosis in our study group was 3.4% which is lower than the studies mentioned in the above table.
The reason for low prevalence might be the target population we have selected and the age group. In conclusion, our study was conducted on orphanages which is a neglected group on which not much study has been done. Though the dermatophytes detected were less, commonest was Tinea capitis followed by Tinea corporis and Tinea cruris. The prevalence of fungal infections was comparatively high, scabies accounted more number of cases among overall skin infections in our study.
Comparision of prevalence of fungal infections with other studies
Apart from studying the prevalence of disease we have to look at factors responsible for the same and conduct educational programme creating awareness about the common contagious skin illnesses and how to prevent from getting infection. Regular checkup by the doctors on this under priviliged population can be recommended to government authorities with follow up. | 2018-12-31T14:10:58.655Z | 2016-09-15T00:00:00.000 | {
"year": 2016,
"sha1": "14fcc3a5c897dc2430643fb7de16b98e24965e41",
"oa_license": null,
"oa_url": "https://www.ijcmas.com/5-9-2016/M.%20Ravish%20Kumar,%20et%20al.pdf",
"oa_status": "GOLD",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "7c85052853ed67a8116ef91ae961b6372f7b9f4c",
"s2fieldsofstudy": [
"Medicine",
"Biology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
255772607 | pes2o/s2orc | v3-fos-license | Effect of the COVID‐19 pandemic on the frequency of emergency department visits in Portugal: An interrupted time series analysis until July 2021
Abstract Objectives This study aims to evaluate the effect of the COVID‐19 pandemic on the frequency of emergency department (ED) visits in Portugal between March 2020 and July 2021. Methods We used data on the monthly number of visits for all public hospitals’ EDs from mainland Portugal between January 2017 and July 2021. We studied the impact of the pandemic overall, by type of ED (general, pediatric, and obstetric) and by Manchester Triage System color (red, orange, yellow, green, and blue) using an interrupted time series analysis. The prepandemic period corresponded to the months from January 2017 to February 2020 and the pandemic period to the months from March 2020 to July 2021. Results We observed over 26 million ED visits, the majority in general EDs (74.0%) and triaged yellow (48.4%) or green (38.4%). During the pandemic period, ED visits decreased 45.7% (95% confidence interval [CI]: –39.8% to –51.2%) and pediatric ED visits decreased by 72.4% (95% CI: –64.6% to –78.6%). A decrease was observed for all colors but tended to be progressively smaller as the priority increased. There was an increase in ED visits during the pandemic period (2.3%; 95% CI: 1.4% to 3.2%), eventually returning to prepandemic values. Conclusion Our data indicate a considerable and long‐lasting effect of the COVID‐19 pandemic affecting mainly pediatric and milder cases, which were returning toward prepandemic values as the pandemic progressed. In a country with frequent use of EDs, the health system may need to be prepared to respond to prepandemic baseline ED demand, together with additional demand because of long‐term sequels of COVID‐19 cases and delayed care for chronic and acute conditions.
Background
The COVID-19 pandemic declared on March 11, 2020, 1 brought an abrupt decrease in emergency department (ED) visits. In March 2020 usage of EDs had fallen 48% in Portugal. 2 A decrease was also observed in the United States in March and April 2020, ranging from 30.9% to 45%. 3,4 In Northern Italy, between March and May 2020, the lockdown period showed the largest decrease in ED visits: 66.2%. 5 As the pandemic evolved, a trend for reduced ED use persisted, according to studies from Canada, 6 China, 7 Italy, 8 Scotland, 9 and the United States. 10 In the National Health Service Scotland, between January and June 2020, ED visits decreased 40.7% (95% confidence interval [CI]: -47.7% to -33.7%). However, these ED visits have been increasing, returning to the 2018-2019 average baseline. 9 A study in Italy, including 147,446 ED visits between January and August 2020, showed decreases of 25.4%, 25.3%, and 23.5% in the early, mid-, and late post-wave periods, respectively. In contrast, the authors found a fall of 66.4% during the first wave, greater than the decrease observed in the second and third periods. 8 When considering the evolution of ED visits, a study from China described an overall decrease of 22.6% (95% CI: -27.53% to -17.36%) in the period January 23-September 6, 2020. 7 Including data until September 23, 2020, a study from Canada described an incident rate ratio (IRR) of ED visits of 0.65 (95% CI: 0.62-0.67). 6 Based on data up to November 15, 2020, a study from the United States reported that ED visits had partially returned to baseline but were still 23% below the number of visits expected (IRR: 0.77; 95% CI: 076-0.78). 10 In Portugal, ED care has an open-door policy 11 and payment of a small flat rate is required, but around 60% of the population is exempted (eg, for health or socioeconomic reasons). 12 There is a National Health System with universal access, 12 including primary care, but demand for urgent care is highly concentrated in acute care providers 11 and very frequent when compared to other countries. 13 15 In a pandemic context, uncertainty, fear of contagion, civil responsibility, mobility restrictions, and concern about placing an unnecessary burden on the health system may have driven patients to delay or avoid seeking care.
Goals of this investigation
This study aims to evaluate the effect of the COVID-19 pandemic on the frequency of ED visits in Portugal between March 2020 and July 2021.
Study design
To estimate the effect of the COVID-19 pandemic on the number of ED visits, we used an interrupted time series analysis. 7
Data sources
We used publicly available information from the Portal da Because population estimates for 2021 were unavailable, we assumed no variation between 2020 and 2021.
Selection of participants
We extracted the total frequency of ED visits per type of ED and triage color from the open-source data. For our studied period (January 2017-July 2021), there were records of 26,384,594 ED visits overall.
To avoid differences in included hospitals throughout the period, we excluded those that did not report data for at least 95% of months in the prepandemic or pandemic period (38 and 17 months). In the triage color analysis, we excluded ED visits triaged white, corresponding to a problem suitable to resolution in a non-ED setting, 20 and non-triaged. Psychiatric visits were recorded for only 1 hospital and were not included in type of ED analysis.
Exposure
Our exposure was the COVID-19 pandemic in Portugal starting on March 2020.
Outcomes
As outcome variables, we considered the number of ED visits, total and stratified by type of ED where the patient was treated (general, pediatric, and obstetric) and triage category (red, orange, yellow, green, and blue). The Manchester Triage System is used in Portugal to categorize patients in terms of the maximum waiting time until being seen by a doctor, from immediate medical attention (red) to 240 minutes (blue). 20 All of these outcomes were measured by the total number of ED visits per month.
Statistical analysis
We used absolute and relative frequencies to describe the number of ED visits by type of ED and triage color. We then used measures of central tendency (median) and dispersion (interquartile range) to describe the monthly frequency of ED visits in the prepandemic and pandemic periods. These periods were not compared, but we estimated the IRRs and the respective 95% CI with the interrupted-time series, providing the effect and precision.
Owing to overdispersion of the outcomes, we fitted a quasi-Poisson generalized linear model, with time as a variable to model the trend before the pandemic. Because we hypothesized that the pandemic would have an immediate effect and usage would slowly increase over time, we added a level and slope change in our model. The level change corresponds to the effect of the pandemic, and a change in slope corresponds to the comparison between the monthly trend before the pandemic and after. We accounted for seasonal effects by adding the month as a spline variable and autocorrelation by adding first-order lagged residuals. 21 We used the population in mainland Portugal as an offset term to adjust for changes in population size over the years. 22 Figures 1 and 2).
LIMITATIONS
This study has some limitations. First, visits recorded as non-triaged were excluded (10.0% of total), which may distort the values of the decrease/increase observed for each triage color analysis if the group composition changed after the pandemic. Second, available data did not have the granularity to identify subgroups with greater changes in ED visits, namely based on age and underlying reason (eg, behavioral, cardiovascular, or trauma). Third, inconsistencies and delays in reporting data from hospitals to the national database may influence our results. We observed a growth in the number of visits in July 2021 compared to June 2021, but possible delays in reporting would lead to an underestimation of the growing trend during the pandemic. Fourth, because we could not isolate COVID-19 cases treated at ED, some of the described decreases may be underestimated. This additional usage for COVID-19 may include initial cases and treatment of sequels in "long-COVID" cases. 24 Fifth, our results may not apply to other countries, partly owing to differences in baseline frequency of ED care, how the incidence evolved during the pandemic, measures adopted by authorities, and social response to the pandemic. However, the data allowed us to conduct a study at the national level for a long pandemic period (18 months until July 2021) and have more robust estimates of ED demand decrease.
DISCUSSION
Our study on the use of EDs in the Portuguese National Health Ser- The COVID-19 pandemic has affected the demand for emergency care globally. A study in Italy described that ED visits decreased 66.2% in March-May 2020, and reductions from 30.9%-45% were described in March-April 2020 in the United States. 3,5 Data for January-June 2020 from Scotland indicate a fall of 40.7%. 9 The scale of these reductions is comparable to or even higher than what we described in Portugal. However, these studies focus on the first pandemic wave period, when sudden drops occurred in a few weeks. Other studies, which also included later phases of the pandemic, showed a smaller reduction in ED visits than we observed for Portugal. Two studies from China and Canada that included data until September 2020 observed a 22.6% and 35% drop, respectively. 6,7 In Italy, a 23.5% reduction was observed in the late post-wave period (August 18-31), 8 whereas in the United States, the reduction until November was 23%. 8,10 In our data, April 2020 and February 2021 were the months with the 2 lowest frequencies for most of the series subgroups, so the inclusion of a longer period, during which there was a new drop in use, may justify the larger decrease we observed in our study.
In our study, the pandemic led to a decrease of 45% in ED visits, which is a sign of a major and long-lasting change in ED care, in a country known for its frequent use of emergency care 13 and where ED is sometimes seen as a gateway to enter the health system. 11 Despite a smaller change than for other triage colors, the most urgent cases (red and orange; for which there is a higher risk of mortality) also A previous study reports that people avoided going to ED because of fear of being infected and lack of confidence in response to conditions other than COVID-19. 27 We observed that during the initial phase of the pandemic (March and April 2020) and when incidence reached a peak (January 2021), there was a comparatively lower frequency of ED visits. We hypothesized that fear and lack of confidence were more intense in the initial phase of the pandemic-people recoiling from a sudden and unknown threat-and during a high-incidence period, leading to larger decreases in ED use in those periods. there is still concern about the possibility that the health system failed to capture children in need of ED care and whether there were groups disproportionately affected.
As described in a recent systematic review, 31 we also found greater decreases for less severe patients (triaged yellow and green were available to respond to COVID-19, whose effects may span long periods. A survey conducted by the World Health Organization at the end of 2020 with 112 countries found that primary care was predominantly affected during the pandemic, with many countries reporting disruptions in routine scheduled visits. 33 The decision to visit ED is multifactorial. 34 However, investing in the education of patients to recognize when they need urgent care may help reduce avoidance of necessary care during a pandemic or other crises and contain inappropriate demand in the post-pandemic period. This might be extremely important considering the epidemiological evolution of the pandemic. In January 2022 the incidence increased in Portugal (3892 new cases per million inhabitants), 14 so it is important to reassure those needing urgent care that visiting ED is safe, but at the same time to ensure that non-urgent cases receive appropriate care while avoiding ED crowding.
In summary, our data indicate a considerable and long-lasting effect of the COVID-19 pandemic in ED visits, affecting mainly pediatric and milder cases, which were returning toward prepandemic values as the pandemic progressed. In a country with frequent use of ED, the health system may need to be prepared to respond to pre-pandemic baseline ED demand, together with additional demand because oflong-term sequels of COVID-19 cases and delayed care for chronic and acute conditions. | 2023-01-14T05:12:34.592Z | 2023-01-11T00:00:00.000 | {
"year": 2023,
"sha1": "2ceeb6fc313fb12f33e70110547d1283fdf8aaab",
"oa_license": "CCBYNCND",
"oa_url": "https://doi.org/10.1002/emp2.12864",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "2ceeb6fc313fb12f33e70110547d1283fdf8aaab",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": []
} |
253236977 | pes2o/s2orc | v3-fos-license | The $\mathfrak{sl}_{2}(\mathbb{R})$ coalgebra symmetry and the superintegrable discrete-time systems
In this paper, we classify all the variational discrete-time systems in quasi-standard form in $N$ degrees of freedom admitting coalgebra symmetry with respect to the generic realisation of the Lie-Poisson algebra $\mathfrak{sl}_{2}(\mathbb{R})$. This approach naturally yields several quasi-maximally and maximally superintegrable discrete-time systems, both known and new. We conjecture that this exhausts the (super)integrable cases associated with this algebraic construction.
This paper is devoted to the classification and the study of a class of discretetime systems in N degrees of freedom admitting coalgebra symmetry with respect to the Lie-Poisson algebra sl 2 (R). We make use of the notion of coalgebra symmetry for discrete-time systems we recently introduced in [29]. The main outcome of this paper is that the coalgebra symmetry approach can be fruitfully used to systematically produce superintegrable discrete-time systems in an analogous way as its continuous counterpart introduced in [7,13]. In particular, within this paper we introduce several seemingly new discrete-time systems, including an N degrees of freedom maximally superintegrable discretisation of the celebrated Smorodinski-Winternitz system [21,23], one of the first maximally superintegrable systems ever introduced [44].
To be more specific, we consider a class of discrete-time systems we call the systems in quasi-standard form which we define as the discrete Euler-Lagrange equations (dEL) of the following class of discrete Lagrangians (dLagrangian) [41]: , q(t ) = q 1 (t ), . . . , q N (t ) .
Here h > 0 is a (fixed) constant and t ∈ hZ, while the functions q k (t ) are not supposed to be defined for all t ∈ R. Following the tradition of [42,[55][56][57] we choose such notation to avoid confusing double indexing in the formulae. Furthermore, we suppose that the functions k = k (ξ) are smooth and locally invertible functions in a given open domain of R. That is, the dLagrangian (1.1) is supposed to be a well-defined function on U 1 ×U 2 ⊂ (R N ) ×2 , where U i , i = 1, 2 are open subsets of R N . After we fix the explicit form of the functions k we will specify the form of the sets U i , i = 1, 2, see Theorem 3.1.
Computing the discrete Euler-Lagrange equations associated to the dLagrangian (1.1) we have that a system in quasi-standard form has the following explicit expression: for k = 1, . . . , N . The number N is called the degrees of freedom of the system, and will be denoted throughout the paper with the shorthand notation d.o.f.. Because of the local invertibility assumptions on the functions k we have that equation (1.2) is step-by-step solvable to determine the next iterate, as it was done in [57] in a particular case. In general, we will not show the solved equations to avoid cumbersome expressions.
Remark 1.1. Before going further we give a few remarks on the terminology we chose to adopt and the meaning of the parameters.
• The fixed parameter h > 0 has the meaning of time step between two different stages of the evolution of the system. If h → 0 we have the so-called continuum limit.
To be more precise, the limit h → 0 gives a (possibly trivial) continuous-time system which needs to be discussed case-by-case. Indeed, each discrete-time system need an ad-hoc scaling of the parameters. We will discuss the continuum limits of the discretetime systems we found in this paper in Section 5. • We choose to call "system in quasi-standard form" the systems (1.2) because if k (ξ) = ξ for all k = 1, . . . , N the system is in the so-called standard form, as defined in [32,54].
From the general theory of discrete variational systems, see [17,61] we have that they are naturally symplectic. Indeed, we can introduce the following canonical momenta: and write the the dEL equations (1.2) in canonical form as: (1.4b) with k = 1, . . . , N . Since in the right hand side of equation (1.4b) q(t + h) is present, the iteration step in equation (1.4) is intended as a two-step procedure. That is, it must be accomplished in the following way: We will denote the iteration from step t to t + h by ϕ h . As a map we will assume that ϕ h : U 1 ×U 2 → U 1 ×U 2 , where the U i are properly chosen open subsets of R N . Once the sets U i , i = 1, 2, are fixed there is a one-to-one correspondence between the discrete-time system in the form (1.4) and its map form. So, in this paper we will use interchangeably the words "discrete-time system" and "map". From the general theory of discrete-time variational system the equations (1.4) preserve the canonical Poisson bracket: see [17,57,63]. Here, by preservation of a Poisson bracket we mean that the following relation holds true: q i (t + h), q j (t + h) = q i (t ), q j (t ) , (1.7a) p i (t + h), p j (t + h) = p i (t ), p j (t ) , (1.7b) q i (t + h), p j (t + h) = q i (t ), p j (t ) . (1.7c) This implies that the map ϕ h is a Poisson map. In particular, since the canonical Poisson bracket (1.6) has maximal rank we have that the map ϕ h is a symplectic map.
From the general theory of integrable symplectic maps, we recall the following definitions: • A symplectic map ϕ h : U 1 × U 2 → U 1 × U 2 possessing N functionally independent invariants in involution with respect to a non-singular Poisson bracket for all values of h > 0 is said to be a Liouville integrable map. • A Liouville integrable map ϕ h : U 1 ×U 2 → U 1 ×U 2 possessing N + k, with k > 0, functionally independent invariants in involution with respect to a non-singular Poisson bracket for all values of h > 0 is said to be a superintegrable map.
functionally independent invariants in involution with respect to a nonsingular Poisson bracket for all values of h > 0 is said to be a quasimaximally superintegrable (QMS) map. [29] not all discrete-time systems admitting more than N non-commuting invariants are integrable with respect to other commonly accepted notions of integrability for discrete-time systems, e.g. algebraic entropy [14].
For a complete overview on the integrability of Poisson and symplectic maps we refer to [17,61,63], the review part of the thesis [60], and our previous paper [29].
Before recalling the construction of coalgebra symmetry for discrete-time systems we make a final observation on continuum limits, i.e. the limits as h → 0 + . Given a (super)integrable discrete-time system, if a continuum limit is known then it is expected to be (super)integrable as well. However, we observe that this does not necessarily follow from the (super)integrability of the associated discrete system. For instance, in [27] were presented several examples where in the continuum limit one invariant of a system in two d.o.f. is lost, yet the continuous-time system possesses two invariants. If both the discrete-time system for h > 0 and its continuum limit as h → 0 are (super)integrable, we say that the discrete system is a (super)integrable discretisation.
We now briefly recall the definition of coalgebra symmetry for discrete-time systems we introduced in [29]. The concept of coalgebra was formulated in quantum group theory [18,19]. Precisely, a coalgebra is a pair of objects (A, ∆) where A is a unital, associative algebra and ∆ : That is, ∆ satisfies the following condition: and it is an algebra homomorphism from A to A ⊗ A: The map ∆ is called the coproduct map. When there is no possible confusion on the coproduct map, it is customary to denote the coalgebra simply by A.
In [29] we gave the following definition: Definition 1.3. A Poisson map ϕ h is said to possess the coalgebra symmetry with respect to the Poisson coalgebra (A, ∆) if for all N ∈ N the evolution of generators A i , i = 1, . . . , K in a N degrees of freedom realisation of the Poisson coalgebra is: (i) closed in the Poisson coalgebra, that is: (ii) it is a Poisson map with respect to the Poisson algebra A, i.e.: (iii) assuming that the Poisson algebra A admits r independent Casimir functions {C 1 (t ), . . . ,C r (t )}, these are preserved as invariants by the map ϕ h , i.e.: This definition allows us to provide an analogue of the construction of the invariants for Hamiltonian systems given in [7,13], a result stated in [29,Theorem 3.3].
The plan of the paper is the following: in Section 2 we remind the main properties of the Lie-Poisson algebra sl 2 (R) and its generic symplectic realisation.
In particular, we discuss the construction of the corresponding left and right Casimir invariants. In Section 3 we classify all systems in quasi-standard form admitting coalgebra symmetry with respect to the generic realisation of the Lie-Poisson algebra sl 2 (R). This result is contained in Theorem 3.1. In SECTION 4 we study the Liouville-Poisson integrability of the dynamical system of the form (1.10) associated to the generic realisation of the Lie-Poisson algebra sl 2 (R). To be more specific, we impose the existence of an additional polynomial invariant and prove in Theorem 4.1 that it exists if the degree of the invariant is 1, 2, or 3. For invariants of higher degree we conjecture, based on the evidence obtained for degrees 4 and 5, that no polynomial invariants exist. In Section 5 we identify the systems we found in Section 4 in the explicit symplectic realisation. This gives us several integrable systems, which we put in the context of the literature. In particular, we find a discrete-time maximally superintegrable version of the Smorodinski-Winternitz system. We note that the maximal superintegrability of this discrete-time model follows from a peculiar construction of the sl 2 (R) coalgebra which is possible only in this specific case. We also find a QMS reduction of the discrete-time Wojciechowski system, introduced in [56], and a generalisation of a N degrees of freedom autonomous discrete-time Painlevé I equation we obtained in our earlier work [29]. In Section 6 we provide some concluding remarks and discuss the further possible developments.
THE sl 2 (R) LIE-POISSON COALGEBRA
The three-dimensional Lie-Poisson algebra sl 2 (R), spanned by the generators J := {J − , J + J 3 }, is characterized by the following Lie-Poisson brackets: and it is endowed with the Casimir invariant: . This Lie-Poisson algebra can be endowed with a "natural" coproduct map called the primitive coproduct [59]. Its explicit action on the basis generators and the unit element is given by (µ = ±, 3): and extends to polynomial elements through the homomorphism property: For example, the coproduct of the Casimir function (2.2) is computed as: where we used the definition primitive coproduct (2.3) and the properties of the tensor products. From the definition of the Casimir invariant (2.2) we obtain the final result: Note that the Casimir of sl 2 (R) ⊗2 genuinely contains new information because it is not just two tensor copies of the casimir C , but additional terms are present.
After giving this summary of the abstract properties related to the sl 2 (R) Lie-Poisson coalgebra we need a way to "embed" these properties in the setting of (continuous or discrete-time) dynamical systems. To this end, we use the concept of symplectic realisation of a Lie-Poisson algebra as expressed in the following definition: Remark 2.2. We remark that on a symplectic manifold (M , ω), by Darboux theorem [2] and its discrete-time analog [17] we can introduce two sets of canonically conjugated variables (ξ, π) ∈ U ξ ×U π ⊆ (R N ) ×2 , where ξ are the canonical coordinates, and π their corresponding canonical momenta, defined on some open sets U ξ ,U π ⊆ R N , and satisfying the canonical Poisson relations (analogous to (1.6)): Throughout the paper, we will denote the discrete-time variables in lowercase Latin letters and the continuous time variables in capital Latin letters. The Greek letters ξ and π will denote variables that can be either continuous or discretetime. The difference between the discrete-time and the continuous-time cases is that in the continuous-time setting the dynamics is specified by a smooth function H = H (ξ, π), while in the discrete-time setting by a symplectic map ϕ h : U ξ ×U π → U ξ ×U π , as discussed in the Introduction.
A one degree of freedom symplectic realisation of sl 2 (R) on (ξ 1 , π 1 ) ∈ R + × R is given by: for arbitrary b 1 ∈ R. Indeed, from the canonical commutation relations (2.11) it is readily verified that: Moreover, this realisation is generic, since N = 1, m = 3 and r = 1. So, following the literature (see for example [11,16]), we will call this realisation the generic one d.o.f. symplectic realisation of sl 2 (R). That said, we will identify the generators with the corresponding functions on the right hand side of (2.12). The image of the Casimir functions (2.2) through the realisation (2.12) reads: Once a realisation has been constructed, the primitive coproduct can be used to raise the number of degrees of freedom. For example, from formula (2.3) we obtain the following functions on R 2 + × R 2 : once two copies of the one degree of freedom symplectic realisation (2.12) are considered, each of them associated to the corresponding site in the tensor product space 1 ⊗ 2 with associated free parameters b i ∈ R, i = 1, 2.
The functions (2.15) provide a two degrees of freedom symplectic realisation for the same Lie-Poisson algebra sl 2 (R), now w.r.t. the canonical Poisson bracket in R 2 + ×R 2 . The crucial point is that at this level the image of the Casimir element, which is now given by (2.6), turns out to be: that is, it is no longer a constant but a function. Moreover, by construction, this function Poisson commutes with the generators in the two degrees of freedom symplectic realisation (2.15). So, by applying the coproduct map iteratively, and extending its definition through the following generalization of the coassociativity property: where ∆ [1] := Id and ∆ [2] := ∆, one ends up with the N degrees of freedom symplectic realisation: where the canonical variables are defined on R N + × R N , and b k ∈ R, k = 1, . . . , N are N associated arbitrary constants. In equation (2.18) and the following, we omit the symbol (D ⊗ · · · ⊗ D), because we will not consider different realisations. At this level, the crucial fact is that a total number of (2N − 3) left and right Casimir invariants can be obtained from the left and right embedding of the m-th order (2 ≤ m ≤ N ) coproduct on the Casimir function (2.2): The image of these elements under the N d.o.f. symplectic realisation (2.18) reads as: where we indicated the N (N − 1)/2 rotation generators as: These (2N −3) quadratic (in the momenta) functions Poisson commute with the generators (2.18) by construction. Moreover, they turn out to be functionally independent.
Remark 2.3. We remark that if we restrict to the case N = 2 the left and right Casimir functions collapse to the same expression: which is nothing but (2.6), the latter leading to the invariant (2.16) at a fixed realisation, as expected. This extends to any N , in fact for m = N the two expressions collapse to: (2.23) where the action of the N th-order coproduct is given by (2.17). This is why formulae (2.20) give us 2N −3 functionally independent invariants and not 2N −2.
So, if the variables (ξ, π) := (Q, P) ∈ U Q ×U P ⊂ R N + × R N are continuous, we can conclude that the family of Hamiltonian systems: where h is any smooth function of the generators of the Lie-Poisson algebra sl 2 (R), is QMS. This is because the above Hamiltonian is automatically endowed with the 2N − 3 functionally independent invariants (2.20). On the other hand, in the discrete-time setting, there is no exact equivalent of the Hamiltonian function and the notion of coalgebra symmetry is replaced by Definition 1.3. Thus, the problem of characterising discrete-time (symplectic) systems admitting sl 2 (R) as a hidden coalgebra symmetry is much less trivial. In the case of systems in quasi-standard form, this problem is solved in Theorem 3.1. Moreover, due to the absence of the Hamiltonian, these systems are not naturally born QMS, and in fact not even Liouville integrable. The problem of integrability is tackled in Section 4.
Remark 2.4. We remark that the symplectic realisation with b i = 0, for i = 1, . . . , N , is connected to radially symmetric systems. In this particular case, the left and right Casimir invariants are: which are nothing but the Casimir invariants associated with rotation subalgebras so(m) ⊆ so(N ). In the continuous setting, the Hamiltonian function (2.24) is given by: As expected, rotational symmetry is sufficient to provide quasi-maximal superintegrability. Notice that if we restrict to natural Hamiltonian systems defined in Euclidean N -space, i.e. we take: then as a consequence of Bertrand's Theorem [2,15], only two MS subcases arise. Namely, the Harmonic and Kepler-Coulomb (KC) systems, with the corresponding potentials given by: respectively. In [29,Proposition 4.2] it was proved that discrete-time radial systems in standard form admit sl 2 (R) coalgebra. In the present paper in Corollary 3.3 the converse is proved.
In general, the presence of non-central terms has the effect of breaking the radial symmetry, but by preserving quasi-maximal superintegrability, which is kept thanks to the existence of the new integrals obtained through the image of the left and right Casimir invariants under the realisation (2.18). Let us conclude the Section by mentioning that the same choice of potentials (2.28), but now in terms of the new realisation involving non-central terms, would result respectively in the N d.o.f. Smorodinsky-Winternitz system [21] and the N d.o.f. generalized KC system, the latter being the N d.o.f. generalization of the (fourth-order) superintegrable Hamiltonian introduced in [62].
Although the many advances pursued in the framework of superintegrable systems with an arbitrary number of d.o.f. using the method described above, which has also been applied to the analysis of many other Lie-Poisson coalgebras [4,16], the problem of finding MS subcases is usually left open. This is due to the fact that additional integrals, not directly obtainable following this algebraic approach, may arise. From this perspective, even for the sl 2 (R) Lie-Poisson algebra, we have recalled how it is possible to reach quasi-maximal superintegrability at most, as a total number of 2N −3 functionally independent integrals can be constructed, besides the Hamiltonian, from the left and right Casimir invariants. Thus, for MS subcases, the search for an additional missing functionally independent constant would be required.
CLASSIFICATION RESULTS
We state and prove the following classification result: Remark 3.2. The "if" part of this theorem is a generalisation of [29,Proposition 4.2], where it is proved that radial systems in standard form admit coalgebra symmetry. The converse will be explicitly stated in Corollary 3.3.
Proof. Consider the N d.o.f. realisation of sl 2 (R) given in equation (2.18) with respect to the discrete canonical variables (q(t ), p(t )). Then, the evolution of the J + (t ) under the system (1.4) is: .
The right-hand side needs to be independent from q k (t + h). To ensure this independence, we can differentiate equation (3.2) with respect to q l (t + h) and put the result identically equal to zero: Interpreting the equation in terms of the variable ξ l = q l (t )q l (t + h), we can integrate equation (3.3) to: Plugging back into equation (3.2) we obtain: Imposing that the right-hand side of the equation is in C ∞ (sl 2 (R)) we easily obtain that C k = C for all k. Note that, when integrating we obtain a condition of the following form: but the additive constant D l is inessential to equations of motion (1.2) and can be safely put to zero. Then, using the scaling: we see that we can choose without loss of generality C = 1. This proves the first part of the theorem, represented by the first identity in (3.1). Moreover, from this follows that the dLagrangian and the associated pairs of canonical variables or an open subset of it depending on the form of the function V = V (q(t )).
We fixed the form of the coefficients k , so now we proceed to derive the full system in order to check the conditions of Definition 1.3. By direct computations we have: While it is possible to prove that the system (3.8) satisfies condition (ii), is it clear that conditions (i) and (iii) are not satisfied in general. As outlined in the last section of [29] we start by imposing that condition (iii) holds. Evaluating the translation of the Casimir function (2.2) we have: To have the preservation of the Casimir functions, we impose that the term in square brackets is identically zero: Since V does not depend on p we can take coefficients with respect to it. In this way we obtain that V has to satisfy the following system of equations: Apply Lagrange's identity [65]: to equation (3.11b) to obtain: This readily implies: Not all equations in (3.14) are independent: one can only consider the subset with i = 1 and j = 2, . . . , N . The solution of this system of N −1 partial differential equations is given by V = V (q 2 ), as in formula (3.1). The introduction of V = V (q 2 ) makes equation (3.11a) vanish identically meaning that condition (iii) is satisfied. We observe then that further restrictions on the domain of the symplectic map can only come from the singularities of the function V = V (χ).
Finally, the sl 2 (R) associated dynamical system has the following form: Hence, condition (i) is satisfied, and this ends the proof of the theorem.
From Theorem 3.1 we have that the most general form of the sl 2 (R) coalgebrically symmetric Lagrangian is: , whose associated discrete Euler-Lagrange equations are: From formula (1.3) the symplectic form of the system is: We conclude this Section with a corollary which represents the inverse of [29, Proposition 4.2]: possesses coalgebra symmetry with respect to sl 2 (R) if and only if it is radially Proof. Observe that, from equation (3.16), if b k = 0 for all k = 1, . . . , N the corresponding system becomes in standard form. So, since Theorem 3.1 is a necessary and sufficient condition the statement follows.
Remark 3.4. We remark that the dynamical system associated with the sl 2 (R) (3.15) is independent from the value of the constants b k . So, from the coalgebra point of view systems in standard form, that is b k = 0 for all k = 1, . . . , N , and in quasi-standard form can be considered in the same way: the (super)integrability of a system in standard form can be transferred to a system in quasi-standard form and vice-versa.
POLYNOMIAL INVARIANTS OF THE ASSOCIATED DYNAMICAL SYSTEM
As discussed in [29] in the discrete-time setting the existence of a rank one coalgebra symmetry alone it is not enough to guarantee the Liouville integrability of the underlying discrete-time system. This is different from the continuum setting, when this has been proven to be true, see [5,11].
So, in this section, we will discuss the integrability of the system (3.15) assuming the existence of an additional polynomial invariant: with the corresponding functions f (ξ) = V (ξ): Remark 4.2. We remark that through degeneration of parameters in (4.3) the functions f 2 and f 3 are both connected to the function f 1 . Indeed, if λ 2 → 0 then f 2 degenerates to f 1 , provided we make the identification κ = λ 1 /λ 3 . Similarly, if τ → 0 then f 3 degenerates to the particular case of f 1 with κ = ±1. A graphical representation of this degeneration scheme is given in Figure 1. Moreover, we observe that there are countably many values of κ such that the associated evolution map, i.e. the map: from equation (3.15) with f 1 = κ/2 is periodic. In Appendix B we show how to find these values, but we shall not discuss these degenerate cases further.
Proof. The proof of this result is through direct computation by considering the three cases d = 1, 2, 3. Indeed, the system already possesses one Casimir function, so if we are able to prove the existence of a functionally independent invariant, we have proven integrability. Case d = 1. This case is prototypical for the other two which can be proven directly, so we will give all the details of the proof. If d = 1 the invariant has the following form: Then, applying the translation by the step h and using the form of the system (3.15) we obtain: We impose that I 1 (t + h) = I 1 (t ). This yields a polynomial expression of degree one in J + and J 3 . Taking coefficients with respect to these two variables, we obtain (after removing the non-zero common factors): The first equation in (4.7) implies a 1,0,0 = a 0,1,0 with no other possibilities: If a 0,1,0 = 0 the first equation in (4.8) implies that f = constant. This is not restricting because a 0,1,0 = 0 yields I 1 ≡ 0, which must be discarded. Then, we write f (ξ) = κ/2, with κ = 0, and the system reduces to the single algebraic equation: (4.9) κa 0,1,0 + a 0,0,1 = 0.
Scaling away the constant a 1,0,0 we proved formula (4.2a), while formula (4.3a) was already established. The functional independence of the set {I 1 ,C } is trivial, so the system is clearly integrable. This concludes the case d = 1.
This brings (4.16) into the form of (4.3b), and the invariant (4.11) into the following form: Now, note that the coefficient of the a 1,1,0 is exactly the Casimir function C of sl 2 (R), see formula (2.2). From Theorem 3.1 this invariant is admitted by the system (3.15) regardless of the function f = f (ξ). So, this invariant is trivial and we can safely discard it by putting a 1,1,0 = 0 in (4.18) (we are already taking into account its existence). This proves formula (4.2b). The functional independence of the set {I 2 ,C } is trivial, so the system is clearly integrable. This concludes the case d = 2.
Case d = 3. Let us consider now the case I 3 (t ) in formula (4.1) (we omit the explicit form because it is rather cumbersome). We employ the same strategy we employed before: from the identity I 3 (t +h) = I 3 (t ) equate all the coefficients in J + , J 3 to zero starting from the higher degree. Taking the coefficients of degree 3 in J + , J 3 in I 3 (t + h) = I 3 (t ) we have the following equations: Since the only possible solution for f from this set of equations is the constant, we can safely put all the coefficients of f to zero and obtain: (4.20) a 0,0,3 = a 0,1,2 = a 0,2,1 = a 0,3,0 = a 1,0,2 = a 2,0,1 = a 3,0,0,0 = 0.
Inserting these values into I 3 (t + h) = I 3 (t ) the degree three terms disappear, and we can consider the degree two ones: (4.21) Taking the coefficients with respect to J − the first equation in (4.21) yields: Since a 1,2,0 = 0 we perform the scaling a 0,1,1 = −2a 1,2,0 τ. This proves formula (4.3c). To conclude the proof we notice that using the results as displayed in Appendix A and the scaling just presented we have: Now, note that the coefficient of the a 1,1,0 is exactly the Casimir function C of sl 2 (R), see formula (2.2). From Theorem 3.1 this invariant is admitted by the system (3.15) regardless of the function f = f (ξ). So, this invariant is trivial and we can safely discard it by putting a 1,1,0 = 0 (we are already taking into account its existence). Then, analogously to the case d = 1 we can scale away the constant a 1,2,0 and we arrive at the expression (4.2c). The functional independence of the set {I 3 ,C } is trivial, so the system is clearly integrable. This concludes the case d = 3 and hence concludes the proof of the theorem.
For invariants of degree d > 3, by direct computation it is possible to prove the following result: Then, with a technique analogue to the one that we used to prove Theorem 4.1 we can see that for d = 4, 5 no different integrable systems are found. The calculations are of increasing complexity, so we don't show them here. However, this leads to the following conjecture:
STUDY OF THE OBTAINED SYSTEMS
In this section, we present a more detailed study of the realisation in canonically conjugated coordinates (q(t ), p(t )) of the integrable systems we found in the previous section. 5.1. Degree one invariant. We consider the system (4.3a) in the realisation of sl 2 (R) (2.11) with canonical coordinates (q(t ), p(t )). The equation assumes the following form: From formula (1.3) the symplectic form of the system is: The case when b k = 0 for all k = 1, . . . , N is a linear system and it was considered in [29, Example 1], so we will consider only the case when b k = 0.
Remark 5.1. We remark that the system (5.1) is symmetric with respect to the transformation: or its symplectic form (5.2) is symmetric under the transformation: p 1 (t ), . . . , −p i (t ), . . . , p N (t ) . (5.4b) This implies that the system can be defined outside its natural domain as described Theorem 3.1 by reflection. Reflecting with respect to each coordinate axis we obtain that the symplectic map (5.2) can be defined on the set The coalgebraic invariant (4.2a) has the following explicit form 1 : From the coalgebra construction, we get the following two sets of invariants (5.7) By induction on the d.o.f. N , it is easy to see that both sets are functionally independent. As expected, this implies that the system (5.1) is Liouville integrable. Similarly the set (5.8) [2] , . . . , C [N ] , C [2] , . . . , C [N −1] , 1 We added a cosmetic 1/2 factor to better compare with the continuous case.
is made of functionally independent functions which implies the system (5.1) is QMS. As remarked in [29] this is the best we expect for general discretetime systems with sl 2 (R) coalgebra symmetry. However, this case is special because it is possible to consider a different set of invariants constructed from the following elements: where: Those N elements, at fixed realisation, result in the following functions: This gives us another set of commuting invariants: (5.12) which proves Liouville's integrability again. Note that: Then, the following set of invariants: (5.14) S 1 = H [1] , . . . , H [N ] , C [2] , . . . , is a set of 2N − 1 functionally independent invariants. For explicit commutation relations involving left and right Casimir invariants, also with these additional constants, we refer the reader to [37][38][39].
is equivalent through Gauss elimination to the upper triangular matrix: Since . Noting that solutions in one orthant evolve inside the same orthant, while the coordinate lines q k = 0 k=1,...,N are not accessible by the evolution we have that these sets exhaust all the points of the phase space, because: where the set U was defined in formula (5.5). So, this proves the functional independence of the invariants everywhere they are defined. This allows us to conclude that the dSW system (5.1) is MS. In Figure 2 we show an orbit of (5.1) near the fixed point Next, we note that the system (5.1) is a discretisation of the Smorodinsky-Winternitz (SW) oscillator (caged isotropic harmonic oscillator) [21,23]. Indeed, under the following scaling: in equation (5.1) we obtain that the leading order as h → 0 is SW oscillator in Lagrangian form: In the same way, writing the invariant (5.2a) using the definition of the canonical momentum (3.18b) andQ k = P k we obtain: That is, at order 4 the Hamiltonian of the SW oscillator appears. The SW system is a well-known MS system whose invariants can be produced through the coalgebra symmetry method [11]. In particular, we note that the other invariants (2.20) and (5.10) are preserved in form through the scaling (5.24). Indeed, noting that in the limit h → 0: A natural generalisation of the SW system arises if we consider anisotropy. That is, considering the following dynamical system: This system is clearly integrable considering the separation of variables in Cartesian coordinates. Following [22], see also [28,52], the case when ω = ω (l 1 , . . . , l N ) where (l 1 , . . . , l N ) ∈ Z N and the integers l i are coprime is MS. In the discrete-time setting we consider the Lagrangian: Then, using the analogue invariants (5.10) we prove that the dEL equations of the Lagrangian (5.31) are integrable. Finally, under the scaling, the associated dEL equations: k , reduce to the Hamiltonian system associated to (5.30).
Differently from the continuum case, the experimental evidence seems to suggest that the system (5.31) is not MS. Indeed, from an experimental study of the orbits of the system (5.31) it is evident that they do not lie on a closed curve, but rather they are a space-filling curve, dense in the phase space. For instance, Figure 3 shows an orbit of the system (5.31) in two d.o.f. in the scaling (5.33) with parameters such that ω 1 /ω 2 = 2/3 ∈ Q. However, despite the frequency ratio is rational after M = 1 000 000 iterations a rectangle in the phase space is almost completely filled.
The orbit is taken in a neighbourhood of the fixed point: The above remarks imply that for ω ∈ ωZ N , the system (5.31) is an integrable, but not a MS discretisation of the anisotropic SW system (5.30). At the same time, the discussion made suggests that there might be some MS subcases of the system (5.31) for given values of the parameters c k . The search for this kind of MS system is outside the possibilities given by the coalgebra approach and will be the subject of future research. Here we limit ourselves to noting that in [28] the conditions for the MS of [22] for the anisotropic SW oscillator (5.30) were derived using a perturbative multiple scales approach. This hints at the possibility of a similar procedure to isolate the MS subcases using, for instance, the multiple scales method for discrete-time systems in the spirit of [30]. 5.2. Degree two invariant. We consider the system (4.3b) in the realisation of sl 2 (R) (2.11) with canonical coordinates (q(t ), p(t )). The equation assumes the following form: .
From formula (1.3) the symplectic form of the system is: Remark 5.2. We remark that if λ 3 = 0 and b k = 0 for all k = 1, . . . , N , this system is a generalisation of the McMillan map [43] introduced in [42]. Indeed in such a case if λ 3 = 0 considering the scaling q k → |λ 3 |q k , and λ 1 = λ 3 σ we obtain: .
We discussed the coalgebra symmetry properties in [29, Subsection 6.2], where the method was used to prove that the system is QMS. We finally recall that the system (5.37) is an example of N d.o.f. integrable system in standard form [32,54], which was later extended in a series of papers [55,56]. For a general review on N d.o.f. integrable systems in standard form we refer also to [58,Chapter 25].
To complete the study on this system, including continuum limits and relation to known discrete-time integrable systems we distinguish the cases λ 3 = 0 and λ 3 = 0. Case λ 3 = 0. For b k = 0 for some k ∈ {1, 2, . . . , N } and λ 3 = 0 we can scale away this constant through the reparametrisation (λ 1 , λ 2 ) = (λ 3 σ 1 , λ 3 σ 2 ) and obtain the system: . Remark 5.3. We remark that the discrete-time system (5.38) is symmetric with respect to the coordinate transformation (5.3), thus allowing us to extend the domain of definition for negative values of the coordinates. However, from the right hand side of equation (5.38) we might need to restrict the domain of definition of this discrete-time system. Indeed, if σ 2 > 0 we need to exclude the N -sphere S N (σ −1 2 ) = q(t ) = σ −1 2 . So, in the end we have that the domain of definition of the discrete-time system (5.38) is: where the set U is given in equation (5.5).
In this case the coalgebraic invariant (4.2b) has the following explicit form 2 : From the coalgebra construction, we get the following two sets of invariants is made of functionally independent functions which implies the system (5.38) is QMS.
Coming to the continuum limit, we see that applying the scaling (5.24) as h → 0 we obtain from the dEL equations (5.38): while for the invariant (5.40) we have: The classical continuum system defined by H QMS W clearly possesses coalgebra symmetry with respect to sl 2 (R), and in fact it turns out to be a QMS deformation of the SW system: The continuum first integrals are again given by formula (5.28). 2 We remove the common factor λ 3 and add a cosmetic 1/2 factor to mimic the continuum case.
THE sl 2 (R) COALGEBRA SYMMETRY AND THE SUPERINTEGRABLE DISCRETE-TIME SYSTEMS 27 The system (5.45) is a particular case of the Wojciechowski system defined by the following Hamiltonian [66]: In general, this system is proven to be integrable both through direct construction of the integrals or by the existence of a Lax pair, see [66]. This system was discretised in [57] as: . Clearly, this system is a generalisation of the system (5.35) with λ 3 = 0, using the identification c k = σ 2 /2 and making the proper scaling in the coordinates q k and the parameters b k 3 . So, we have that the system (5.38) is a coalgebraic QMS subcase of the discrete-time Wojciechowski system (5.47).
Remark 5.4. We remark that there might be superintegrable subcases of the Wojciechowski system (5.46) if ω i /ω j ∈ Q for some i , j ∈ {1, . . . , N }. At present, we did not prove the existence of these superintegrable subcases, but we limit ourselves to notice that the proofs of integrability of both [57,66] do not work in the QMS case. So, it is reasonable to believe that the presented first integrals are not exhaustive of all the integrable cases and there might be intermediate cases between the Liouville integrable case (N invariants) and the QMS case (2N − 2 invariants). The problem of the existence of superintegrable subcases of the Wojciechowski system (5.46) and its (possible) superintegrable discretisation will be the subject of future research.
Case λ 3 = 0. Let us assume now that λ 3 = 0 and that b k are arbitrary (possibly also zero for all k = 1, 2, . . . , N ). In such a case, we can rescale λ 1 = −λ 2 σ and the system (5.35) reduces to: Remark 5.5. We remark that the discrete-time system (5.48) is symmetric with respect to the coordinate transformation (5.3), thus allowing us to extend the domain of definition for negative values of the coordinates. In this case the right hand side does not give any additional restriction, so the system can be defined on the whole set U given in equation (5.5).
The coalgebraic invariant (4.2b) has the following explicit form 4 : From the coalgebra construction, we get the same two sets of invariants as for λ 3 = 0 (5.41) case with H 2 (λ 3 = 0) replaced by H 2 (λ 3 = 0). The two sets are clearly functionally independent. In the same way replacing H 2 (λ 3 = 0) with H 2 (λ 3 = 0) in (5.42) we obtain a set of 2N −2 functionally independent invariants for (5.48). Summing up, we proved that the system (5.48) is Liouville integrable and moreover is QMS.
Differently from the λ 3 = 0 the continuum limit of the case λ 3 = 0 is not known. Analogously to [29,Subection 5.3] it is possible to prove that no scaling of the form where σ(h) is an analytic function of its argument, balances the terms in the systems (5.48).
5.3. Degree three invariant. We consider the system (4.3c) in the realisation of sl 2 (R) (2.11) with canonical coordinates (q(t ), p(t )). The equation assumes the following form: From formula (1.3) the symplectic form of the system is: Remark 5.6. We remark that if b k = 0 for all k = 1, . . . , N , the system (5.51) with the plus sign is a particular case of an autonomous version of the discrete-time Painlevè I equation [25,32] we introduced in [29, Section 5] with the parameter β = 0.
The coalgebra symmetry properties of the system (5.53) were discussed in [29], where we showed that the system is QMS. We also recall that the continuum limit of the system (5.53) is unknown.
Considering b k = 0 for some k ∈ {1, . . . , N } the coalgebraic invariant (4.2c) has the following explicit form 5 : From the coalgebra construction, we get the following two sets of invariants (5.55) is made of functionally independent functions which implies the system (5.51) is QMS. Based on the above discussion we have that the system (5.51) for b k = 0 for some k ∈ {1, . . . , N } is a novel QMS system, generalising the already known system (5.53).
We finally note that at present no continuous analogue of the system (5.51) is known. Indeed, analogously to the system (5.53) mentioned in Remark 5.6 and to [29,Subection 5.3], it is possible to prove that no scaling of the form (5.50) with τ(h) analytic functions of its argument balances the terms in the systems (5.51).
CONCLUSIONS
In this paper, we gave the necessary and sufficient conditions for a discretetime system in quasi-standard form (1.2) to admit coalgebra symmetry with respect to the generic symplectic realisation of the Lie-Poisson algebra sl 2 (R) (2.18) in N degrees of freedom. In particular, from Theorem 3.1 we see that the most general form of this system is the one considered in [57]. In this sense, we see that the coalgebra symmetry approach naturally selects the functional form for the functions k = k (ξ), k = 1, . . . , N .
As already discussed in [29] the systems in quasi-standard form (1.2) admitting coalgebra symmetry with respect to the generic symplectic realisation Lie-Poisson algebra sl 2 (R) (2.18) in N degrees of freedom naturally possess 2N − 3 functionally independent invariants. However, in general these systems are not Liouville integrable. This is because we are missing an exact discretetime equivalent of the Hamiltonian. So, to characterise the Liouville integrable cases we searched for polynomial invariants in the variables of the algebra sl 2 (R) of increasing degree for the associated dynamical systems (3.15). It turns out that such invariants do exist for degrees 1, 2, and 3, while for degrees 4, and 5 no new systems arise. This led us to conjecture that the only non-trivial Liouville integrable systems with a polynomial invariant are those admitting an invariant of degrees 1, 2, and 3. The proof of this conjecture is particularly challenging because, in general, it does not seem possible to resum the expression (4.27) to: which will make the claim easier to prove. Unfortunately, since in general not even the degree of the function f = f (ξ) is known it is not possible to apply the "destructive" test of algebraic entropy [14,24,26] 6 . For the very same reason it is not possible to apply the "constructive" singularity confinement test [24,25]. Under the (reasonable) assumption that f (ξ) is a rational function, one could try to apply a Nevanlinna theory for maps [1]. However, at present, such a theory is not enough developed to treat a system of the form (3.15). So, this topic will be the subject of further research.
In the sequent section we carefully analysed the obtained systems in the realisation (2.18) and proved explicitly their Liouville integrability and quasimaximal superintegrability. In addition, we found a maximally superintegrable system given by the discrete Lagrangian (5.1) which is a discretisation of the well-known SW system [21,23]. We also discussed the possible generalisation (5.31) discretising the so-called caged anisotropic oscillator [22], which we showed to be Liouville integrable. Unfortunately, differently from its continuous counterpart, the system (5.31)does not appear to be maximally superintegrable from a numerical study of its orbits. Since this kind of analysis goes outside the applicability of the coalgebra method, this topic will be the subject of future research. Besides these cases, we found the coalgebraic subcase of the discretetime Wojciechowski system (5.35), whose general case was constructed in [57]. Finally, we found a non-birational generalisation of a system we proposed in [29], which we deem to be new. In Table 1 we give a compendium of the known Liouville integrable cases of equation (1.2).
Summing up, in this paper we proved that the coalgebra symmetry method can be applied to systematically produce N d.o.f. discrete-time superintegrable systems.
An interesting open problem is the existence of a coalgebraic discretisation of the Kepler-Coulomb system: This model is MS, through the existence of an additional integral of motion called the Laplace-Runge-Lenz vector [40,44]. It would be interesting to show if it is possible to construct such an invariant in the discrete-time setting. 6 Quoting from [64] "Algebraic entropy. pro: it is canonical (invariant by birational changes of coordinates), and the vanishing of the entropy may serve as a characterisation of integrability, as the sign of catastrophic drop of the complexity, con: destructive rather than constructive, since it gives a yes/no answer to the question "is this model integrable?"". Italics from the original text. Another open problem is the generalisation to non-Euclidean manifolds. For the coalgebra approach to these systems, see for example a series of papers by Italian-Spanish school [3,[8][9][10][11] culminating in the proof of a Bertrand-like theorem on curved space [12], linked to the so-called Perlick classification [51]. For a different, more geometric perspective on the subject, see the recent classification in [34,35]. Finding a connection between these two approaches and building the discrete-time analogue will be subject of future research.
Finally, we note that the present construction can be applied to other Lie-Poisson algebras, especially the ones related to classified Lie algebras for which the Casimir functions are known, see [36,[45][46][47][48][49][50]. In the continuous setting this was done by Ballesteros and Blasco [4,16]. We note that the most natural extension would be the two-photon or h 6 Lie-Poisson algebra [67], a Lie-Poisson algebra containing many other interesting Lie-Poisson algebras as subalgebras, including sl 2 (R), whose associated Hamiltonian integrable systems have been discussed in [6,16]. If V = κξ/2 then the system (3.15) becomes: Hence, the eigenvalues are: The eigenvalues are different for all κ = ±2. This readily implies that the matrix is diagonalisable for every κ = ±2. When κ = ±2 the matrix is not diagonalisable because the only eigenvalue is µ 1 = 1, which has geometric multiplicity one.
Then for κ = ±2 we have: The periodicity condition is that there exists a L ∈ N such that M L = I 3 , where I 3 is the identity matrix. That is, the solution can be written as: (B.11) κ = ±2 cos kπ L , k = 0, . . . , L − 1.
However, this is not definitive: we need to discard k = 0 because it yields κ = ±2 which is not acceptable, and since the cosine function is antiperiodic of antiperiod π we can choose the sign plus in (B.11). So we denote the final expression of the solutions as κ k,L and its expression is: (B.12) κ k,L = 2 cos kπ L , k = 1, . . . , L − 1.
We underline that from formula (B.12) we have κ k,L < 2, and for all values of k and L except for the numbers κ k,L are irrational numbers. In particular, this implies that, for every L prime we will obtain new solutions. This implies that the set of values of κ k,L such that the dynamical system is periodic is a countably infinite set.
That is, these cases correspond to the degenerate case of the function f 3 as τ → 0, see Figure 1. | 2022-11-01T01:16:03.436Z | 2022-10-31T00:00:00.000 | {
"year": 2022,
"sha1": "0f1ef53d07b4cf539fa8100735c6a146f3964548",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "0f1ef53d07b4cf539fa8100735c6a146f3964548",
"s2fieldsofstudy": [
"Physics",
"Mathematics"
],
"extfieldsofstudy": [
"Physics",
"Mathematics"
]
} |
15638818 | pes2o/s2orc | v3-fos-license | Redox proteomic analysis of the gastrocnemius muscle from adult and old mice
The data provides information in support of the research article, “Differential Cysteine Labeling and Global Label-Free Proteomics Reveals an Altered Metabolic State in Skeletal Muscle Aging”, Journal of Proteome Research, 2014, 13 (11), 2008–21 [1]. Raw data is available from ProteomeXchange [2] with identifier PDX001054. The proteome of gastrocnemius muscle from adult and old mice was analyzed by global label-free proteomics and the relative quantification of specific reduced and reversibly oxidized Cysteine (Cys) residues was performed using Skyline [3]. Briefly, reduced Cysteine (Cys) containing peptides was alkylated using N-ethylmalemide (d0-NEM). Samples were desalted and reversibly oxidized Cys residues were reduced using tris(2-carboxyethyl)phosphine (TCEP) and the newly formed reduced Cys residues were labeled with heavy NEM( d5-NEM). Label-free analysis of the global proteome of adult (n=5) and old (n=4) gastrocnemius muscles was performed using Peaks7™ mass spectrometry data analysis software [4]. Relative quantification of Cys containing peptides that were identified as reduced (d(0) NEM labeled) and reversibly oxidized d(5)–NEM labeled was performed using the intensity of their precursor ions in Skyline. Results indicate that muscles from old mice show reduced redox flexibility particularly in proteins involved in the generation of precursor metabolites and energy metabolism, indicating a loss in the flexibility of the redox energy response.
a b s t r a c t
The data provides information in support of the research article, "Differential Cysteine Labeling and Global Label-Free Proteomics Reveals an Altered Metabolic State in Skeletal Muscle Aging", Journal of Proteome Research, 2014, 13 (11), 2008-21 [1]. Raw data is available from ProteomeXchange [2] with identifier PDX001054. The proteome of gastrocnemius muscle from adult and old mice was analyzed by global label-free proteomics and the relative quantification of specific reduced and reversibly oxidized Cysteine (Cys) residues was performed using Skyline [3]. Briefly, reduced Cysteine (Cys) containing peptides was alkylated using N-ethylmalemide (d0-NEM). Samples were desalted and reversibly oxidized Cys residues were reduced using tris(2-carboxyethyl)phosphine (TCEP) and the newly formed reduced Cys residues were labeled with heavy NEM( d5-NEM). Label-free analysis of the global proteome of adult (n¼ 5) and old (n¼4) gastrocnemius muscles was performed using Peaks7™ mass spectrometry data analysis software [4]. Relative quantification of Cys containing peptides that were identified as reduced (d(0) NEM labeled) and reversibly oxidized d(5)-NEM labeled was performed using the intensity of their precursor ions in Skyline. Results indicate that muscles from old mice show reduced Table Subject area Biology More specific subject area Redox proteomics and aging Type of data Excel sheets, figures How data was acquired QExactive mass spectrometer coupled with ultimate 3000 RSLC nano system (Thermo Scientific).
Data format
Raw and processed data.
Experimental factors
Initial blocking of free thiols with d(0)NEM, reduction of reversibly oxidized Cys residues and labeling of newly reduced residues with d(5) NEM.
Experimental features
Label-free analysis of global proteome and relative quantification of oxidation state of redox sensitive Cysteine residues.
Value of the data
A number of metabolic and muscle diseases are associated with aberrant redox regulation. The combination of label free analysis and redox proteomic data provides additional information on the functional proteome.
This is the first article to our knowledge that combines differential Cys labeling with global labelfree analysis of the proteome.
Data provides valuable information on the changes that occur in the skeletal muscle proteome with aging.
Skeletal muscle aging is an ideal model for redox proteomics [5] and our findings suggest that skeletal muscle aging is associated with a decrease in redox flexibility.
1. Data, experimental design, materials and methods
Sample preparation
Adult (12 months) and aged (25 months) C57BL/6 male mice were purchased from Charles River and housed in the Specific Pathogen-Free (SPF) Facility at the University of Liverpool for at least 2 weeks before use. All experiments were performed in accordance with United Kingdom Home Office guidelines under the United Kingdom Animals (Scientific Procedures) Act 1986. Animals were sacrificed by cervical dislocation and gastrocnemius muscles were immediately dissected. A gastrocnemius muscle from each mouse was placed immediately in a thiol blocking buffer containing (25 mM d(0) NEM, 50 mM ammonium bicarbonate, pH 8) for redox proteomic analysis. Briefly protein extracts for redox analysis were prepared in the presence of thiol blocking buffer containing d(0) NEM under anaerobic conditions. Homogenized protein lysates were cleared by centrifugation at 15,000g for 10 min at 4 1C and protein concentrations were calculated by Bradford assay (BioRad) using BSA as a standard. Protein extracts for redox analysis were desalted using Zeba spin desalting columns (Thermo) and protein concentrations were re-calculated as before. 200 mg in 160 ml of 25 mM ammonium bicarbonate of the desalted protein extract was denatured by addition of 10 μL of 1% w/v RapiGest™ (Waters, Manchester, UK) in 25 mM ammonium bicarbonate followed by incubation at 80 1C for 10 min. Reversibly oxidized Cys residues were reduced by the addition of 10 ml of 100 mM TCEP and incubated at 60 1C for 10 min. Newly reduced Cys residues were subsequently alkylated with 10 ml of 200 mM d(5) NEM and incubated at room temperature for 30 min. Trypsin (Sigma, Poole, UK) was reconstituted in 50 mM acetic acid and 2 μg added to the samples followed by incubation overnight at 37 1C. The digestion was terminated and RapiGest ™ removed by acidification (3 μL of TFA and incubation at 37 1C for 45 min) and centrifugation (15,000 Â g for 15 min).
Sample analysis
The data-dependent label-free analysis was performed using an Ultimate 3000 RSLC ™ nano system coupled to a QExactive ™ mass spectrometer (Thermo Scientific). The sample (5 mL corresponding to 200 ng of protein) was loaded onto the trapping column (Thermo Scientific, PepMap100, C18, 75 μm  20 mm), using partial loop injection, for 7 min at a flow rate of 4 μL/min with 0.1% (v/v) TFA.
The sample was resolved on the analytical column (Easy-Spray C18 75 mm  500 mm  2 mm column) using a gradient of 97% A (0.1% formic acid) 3% B (99.9% ACN 0.1% formic acid) to 60% A 40% B over 120 min at a flow rate of 300 nL/min. Data dependent acquisition consisted of a 70,000 resolution fullscan MS scan (AGC set to 106 ions with a maximum fill time of 250 ms). The 10 most abundant peaks were selected for MS/MS using a 17,000 resolution scan (AGC set to 5  104 ions with a maximum fill time of 250 ms) with an ion selection window of 3 m/z and a normalized collision energy of 30. To avoid repeated selection of peptides for MS/MS the program used a 30 s dynamic exclusion window.
Data processing and quantification
Raw spectra were converted to mascot generated files (mgf) using Proteome Discoverer software (Thermo Scientific). The resulting mgf files were searched against Uniprot Mouse database sequence database (12/05/2012, 16376 sequences) using an in-house Mascot server (Matrix Science, London, UK). Search parameters used were: peptide mass tolerances, 10 ppm; fragment mass tolerance, 0.01 Da, 1þ, 2þ and 3þ ions; missed cleavages, 1; instrument type, ESI-TRAP. Variable modifications included were: d (0) NEM, d(5) NEM, mono-, di-and tri-oxidation of Cys residues and oxidation of methionine. Label-free relative quantification software Peaks7 ™ was used to analyze RAW data files against the same mouse protein database used for identifications with Mascot. Proteins were considered significantly changed between adult and aged samples using a -10logP score of 20 (equivalent to a p value of 0.01), a fold change Z1.5, using a quality value of 0.8 and FDR set to 1%. Supplementary file 1 contains additional information on the search and result parameters used including a Volcano plot for peptides, the distribution of feature vector ratio by quality, distribution of feature vector ratio by intensity, retention time shift distribution and M/Z Shift distribution. A volcano plot of the expression of proteomic data is provided in Fig. 1 and a list of proteins identified and quantified is included in Supplementary Table 1. An example of a peptide containing a redox sensitive Cys residue (Cys385 from aconitase) is presented in Fig. 2. Analysis of redox peptides identified with both d(0) NEM and d(5) NEM was performed using Skyline and the relative quantification of the reversible oxidation state of Cys residues (reduced:reversibly oxidized) was calculated from the intensity of precursor ions and is included in Supplementary Table 2. | 2016-05-16T19:34:16.005Z | 2015-07-02T00:00:00.000 | {
"year": 2015,
"sha1": "dace970578903b4a420ed39613228fb5588bc0fa",
"oa_license": "CCBYNCND",
"oa_url": "https://doi.org/10.1016/j.dib.2015.06.012",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "dace970578903b4a420ed39613228fb5588bc0fa",
"s2fieldsofstudy": [
"Biology",
"Medicine"
],
"extfieldsofstudy": [
"Chemistry",
"Medicine"
]
} |
256879223 | pes2o/s2orc | v3-fos-license | A Comparative Study on the Meat Quality, Taste and Aroma Related Compounds between Korean Hanwoo and Chikso Cattle
The aim of this study was to compare the meat quality and taste-and-aroma-related components of beef between breeds. For this purpose, Hanwoo and Chikso steers (n = 7 per breed) raised under identical conditions until 30 months old were used. After 24 h of slaughter, longissimus lumborum (LL) and semimembranosus (SM) muscles were collected and analyzed for technological quality, free amino acids, metabolites, and volatile compounds. The Chikso meat showed lower values for shear force and color traits (lightness, redness, and yellowness) compared to Hanwoo (p < 0.05). The Chikso presented a higher amount of sweetness-related free amino acids (alanine, proline, and threonine) in the LL muscle, while Hanwoo had a higher amount of methionine and glutamine associated with umami taste (p < 0.05). A total of 36 metabolites were identified and quantified in the meat samples; out of them, 7 compounds were affected by breed (p < 0.05). Regarding aroma compounds, a significantly higher amount of fat-derived aldehydes associated with fatty and sweet notes was found in Hanwoo, whereas a higher amount of pyrazines associated with roasty notes was found in Chikso (p < 0.05). Thus, under identical feeding conditions, breed showed a significant effect on the quality and taste-and-aroma-related components that may influence the eating quality of beef between the two breeds studied.
Introduction
The Chikso breed, together with three other Korean native cattle breeds, has been registered with the Domestic Animal Diversity Information System of the Food and Agriculture Organization. The Chikso breed is characterized by its unique brindle coat color [1], which is completely different from the other registered cattle breeds (e.g., brown Hanwoo, Jeju black cattle, and Korean black cattle). Historically, the Chikso breed was used mainly as draft and pack animals. Compared to the other popular commercial beef cattle breeds (e.g., Hanwoo) raised in Korea, the Chikso breed is generally maintained at a smaller population size, around 4000 heads in 2016 [2]. In recent years, due to the increasing demand for safe meat products derived from native cattle breeds in South Korea, the Chikso breed has been recognized as a valuable breed and has received more attention from beef producers [3]. In contrast to the Hanwoo breed, which is commonly raised following the standardized production process and fed grain feed in feedlots of commercial farms, the Chikso breed is currently raised by small-scale producers in several localities (e.g., Gyeongbuk and Kangwon provinces) in the country [4]. Until now, almost all studies on the Chikso breed have only focused on the genetic diversity aspects [3,4].
Meat quality can be defined in different ways, from product yield to a set of properties (e.g., color, texture, water holding capacity, eating attributes, etc.) that together identify what we appreciate about meat when buying or eating it and using it as raw material for processing into meat products [5]. It is well known that there are many factors influencing the quality of meat [6]. Among others, animal characteristics (e.g., genetics or breed and feeding diets) significantly affect the physiochemical composition of muscle tissues, which subsequently affects the meat quality [7].
More to the point, flavor, consisting of odors and tastes, is among the leading factors determining the eating quality of meat and its products by consumers [8,9]. Tastes (sweetness, saltiness, bitterness, sourness, and umami) of cooked meat are contributed by nonvolatile constituents or taste-active compounds (free amino acids and metabolites) of fresh meat [10][11][12]. Otherwise, these constituents, together with others (lipids), are the major contributors to the formation of aroma characteristics via the Mallard reaction and/or thermal oxidations during the cooking/heating of meat [13]. Researchers have discovered that both pre-and post-harvest management can influence the flavor precursors of fresh meat and, as a result, the flavor characteristics of cooked meat [14].
To the best of our knowledge, limited scientific information regarding the quality characteristics of Chikso beef is available. Also, there is a lack of comparative studies on meat quality between this breed and other commercial cattle breeds under identical rearing conditions. Therefore, the aim of this study was to determine the meat quality, taste, and aroma-related components of Chikso beef and compare them with those of Hanwoo beef under identical raising conditions.
Sampling
A total of 14 steers (7 for Hanwoo and 7 for Chikso) were used in the present investigation. All the animals were raised under identical commercial conditions and fattened in feedlots with a grain-based diet. During the growing phase, they were fed twice/day with formula feed (3.0-7.5 kg) and rice straw (3.0-4.0 kg). At the fattening phase (from 18 months of age), they were fed ad libitum with a high rate (90%) of concentrate feed (11-12% crude protein) and only 10% rice straw up to 30 months of age. The average body weight at slaughter was around 680 and 740 kg for the Chikso and Hanwoo steers, respectively. The animals were slaughtered at an abattoir of the National Institute of Animal Science (NIAS), Wanju-gun, Korea. All procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of NIAS (Approval No. NIAS 20001992). The next day (24 h post-mortem), two representative muscles including: Longissimus lumborum (LL) and semimembranosus (SM) (n = 7 per breed and muscle type) were removed from the left side of all the carcasses and immediately used for analyses. After trimming all visual fat and connective tissues, each muscle was cut into sub-samples depending on the analysis. The samples for analysis of quality attributes (color, pH, water holding capacity, cooking loss, and shear force) were immediately used after cutting, while the rest of the meat samples were stored at −20 • C for free amino acids, metabolites, and aroma-flavor compound analysis.
Meat Color and pH Measurement
The color and pH were measured on the same samples (in the form of 2.5-cm-thick steaks). For the color, it was measured on 5 different areas on the transverse section of each sample using a Minolta Chroma Meter CR-400 with a D65 illuminant and 2 • observer (Minolta Camera Co., Osaka, Japan). Prior to the measurement, all the samples were kept at 4 • C for blooming for 30 min. The results were expressed as CIE L* (lightness), CIE a* (redness), and CIE b* (yellowness). Immediately after the color measurement, the pH was measured in triplicate by inserting the solid-state probe of the pH meter (pH*K 21 m, NWK-Technology GmbH, Kaufering, Germany) deeply into the meat samples. Before use, the pH device was calibrated with pH 4.0 and 7.0 standard solutions (NWK Technology, City, Country, Germany).
Cooking Loss and Warner-Bratzler Shear Force (WBSF)
The cooking loss and WBSF were measured on the same piece of each sample according to the method described by Hoa et al. [15]. Briefly, a steak (weighing 200 g and having a 2.5 cm thickness) was taken from each muscle sample, trimmed of its outer fat, placed in a polyethylene bag, and cooked in a 73 • C preheated water bath until the core temperature reached 72 • C. After cooking, the cooked samples were cooled in ice water for 30 min, removed from the bags, and blotted dry with paper towels. The sample weight before and after cooking was recorded to determine the cooking loss (pre-cooking weight minus post-cooking weight divided by the pre-cooking weight and multiplied by 100). The WBSF of cooked meat samples was measured using an Instron Universal Testing Machine (Model 4465, Instron Corp., High Wycombe, UK) at a crosshead speed of 200 mm/min and a 50 kg load cell. For this, 5 strips per cooked sample were made parallel to the muscle fiber direction using a 0.5-inch metal corer. WBSF values (kg-force) were obtained by completely cutting the strips with the device.
Water Holding Capacity (WHC)
The WHC of each meat sample was measured in duplicate using the procedure as described by Hoa et al. [15]. Briefly, after grinding using a mini grinder (Hanil Co., Chungcheongnam-do, Korea), an aliquot (0.5 g) of meat was taken and placed in a 2 mL ultra-centrifugal filter unit, which was then inserted into an ultra-centrifugal filter device (Millipore Corp., Bedford, MA, USA). After heating for 20 min at 80 • C in a water bath, the samples were cooled at 4 • C for 10 min and then centrifuged at 2000× g for 10 min. The initial weight of the ultracentrifugal filter unit before cooking and its weight with the cooked sample were recorded to determine the water loss. Also, the total moisture and fat contents in each fresh meat sample, determined by using a Food Scan Lab 78810 (Foss Tecator Co., Ltd., Hillerod, Denmark), were used to determine its WHC. Finally, the WHC was calculated using the equation developed by Laakkonen et al. [16] as follows: 2) FF = 1 − Total fat content 10 W1: Weight of sample and centrifugal filter unit before heating. W2: Weight of sample and centrifugal filter unit after heating and centrifuging. S: Sample weight. 2) FF: Fat factor; 1: Constant.
Tastes-Related Compounds (Free Amino Acids, FAAs and Metabolites)
The content of FAAs in the meat samples was analyzed using the protocol described by Cho et al. [17] with minor modifications. Briefly, 2.5 g of each sample was weighed, placed in a conical tube, and homogenized with 5 mL distilled water at 1200× g for 1 min. The homogenate was filtered with Whatman filter paper (Whatman Inc., Clifton, NJ, USA), and an aliquot of 100 µL filtrate was taken and mixed with 900 µL methanol containing 0.1% formic acid. Next, the samples were centrifuged at 13,000× g at 4 • C for 10 min, and the supernatant was again filtered through a 0.45-µm filter membrane (Millipore Ltd., Cork, Ireland). After derivatizing with AccQ-Tag TM (Waters Co., Milford, MA, USA), according to the manufacturer's instrument, the samples were used for FAA analysis. The FAA composition was separated on an amino acid column ( The UPLC conditions were set as follows: Separation temperature at 37 • C, flow rate of 0.4 mL/min, and solvent ingredient: initial 100% B, linear change to 83% B for 6.5 min, linear change to 100% A for 3.5 min, and then linear change to 100% B for 2 min, and maintained for an additional 5 min. The results were expressed as milligrams per 100 g of meat (mg/100 g meat). Each sample was determined in duplicate.
For the analysis of metabolites, the meat samples (2.5 cm thick in steak form) were cooked on a frying pan at around 180 • C for approximately 4 min. During the frying, the meat was turned at 1-min intervals. The extraction and separation of metabolites were carried out following the protocol as described in our previous study [17]. Briefly, triplicate aliquots (20 mg) of meat were weighed and extracted with an acetonitrile/water (1:1, v/v) mixture on ice for 10 min. After centrifuging at 3000× g for 10 min at 4 • C, the supernatant was collected and freeze-dried. The metabolites were analyzed using a 600 MHz Agilent NMR spectrometer (Agilent Technologies, Palo Alto, CA, USA) equipped with a 600 MHz 4-mm gHX NanoProbe (Agilent Technologies, Santa Clara, CA, USA) at a 1 H frequency of 599.93 MHz. The analytic conditions used were the same as those described by Cho et al. [16], and the acquired 1 H-spectra were identified using the Chenomx 600 MHz library database and Chenomx NMR Suite 7.1 professional. The concentration (mM/kg) of each metabolite was determined using the known concentration of an international standard (3-trimethylsilyl-2,2,3,3-tetradeuteropropionic acid-d4 (TSP-d4, Sigma-Aldrich, St. Louis, MI, USA). The results were expressed as mM/kg of meat sample.
Volatile Flavor Compounds
The aroma volatile compounds in the meat samples were extracted using a solid-phase micro-extraction (SPME) method as described in our previous work [18]. To strengthen the chemical reactions (e.g., Mallard reaction) for the formation of volatile flavor compounds, the fresh meat samples were ground into small particles before cooking. The ground meat samples were then cooked on a frying pan at around 180 • C for 2 min (the meat was stirred throughout the frying process). Immediately after cooking, the samples (1.0 g each) were taken and placed into a 20-mL headspace vial and tightly capped with a magnetic screw cap. For quantification, 1.0 µL of an internal standard (2-methyl-3-heptanone at 816 mg/mL in methanol) was also added. The extraction of aroma volatiles was done using a 75 µm SPME assembly of CAR/PDMS fiber (black, autosampler type, Supelco, Bellefonte, PA, USA) connected to a SPME auto-sampler (model: PAL RSI 85) of gas chromatography (model: 8890 GC system) and mass spectrophotometry (5977B MS, Agilent Technologies) at 60 • C for 50 min. The extracted volatiles were desorbed at the injection port at 250 • C for 5 min and then separated on a DB-5MS column (30 m × 0.25 mm i.d. × 0.25 µm film thickness; Agilent J & W Scientific, Folcom, CA, USA). The GC/MS conditions set were the same as those used in the above-cited reference [18]. The peaks were identified by comparing their mass spectra with those in the Wiley registry library (Agilent Technologies) and/or by comparing their retention times with those of external standards. The quantification of each compound was carried out by comparing its peak area percentage with that of the internal standard (1.0 µL of 2-methyl-3-heptanone, 816 mg/mL in methanol). The results were expressed as µg/g of meat sample.
Statistical Analysis
Statistical analysis was conducted using SAS Enterprise software (version 7.1; SAS Institute, Inc., Cary, NY, USA). For each muscle type, data was separately analyzed using the General Linear Model procedure, where the cattle breed was considered a fixed effect and the quality attributes, free amino acids, metabolites, and aroma compounds were considered dependent variables. The mean difference was compared using Duncan's Multiple Range Test, and significant differences were set at a 5% level. Data was presented as means ± standard deviation.
Technological Quality
The mean values for technological quality (pH, water holding capacity, cooking loss, and shear force) of the beef samples from two breeds are shown in Table 1. The ultimate pH of meat is usually determined by the extent of pH decline within 24 h of slaughter. As we know, postmortem glycolysis (the process that converts glycogen into energy under anaerobic conditions) is the main process resulting in increased H + ion accumulation and pH decline in meat. The rate and extent of post-mortem glycolysis vary depending on animal species, genetics, feeding systems, and stress caused during fasting and transport [19]. Researchers have also reported that any abnormal rate of postmortem glycolysis may result in inferior meat quality [20,21]. Our results showed that no differences in pH values occurred between the two cattle breeds (p > 0.05). This means that the rate of glycolysis or pH decline was similar in both breeds studied. Warren et al. [22] reported that under identical feeding conditions, breed does not influence the ultimate pH of beef longissimus lumborum muscles. Contrastingly, Xie et al. [23] showed a wide variation in pH values of beef longissimus dorsi muscles among Limousin, Simmental, Luxi, Qinchuan, and Jinnan. Means within a row (in each muscle type) with different superscripts ( a , b ) differ significantly (p < 0.05).
Water-holding capacity is the ability of fresh meat to retain moisture, which is an important technological quality trait. The results showed that the breed only affected the WHC of LL muscle, with a higher value in the Chikso compared to the Hanwoo (p < 0.05). It is reported that ultimate pH is the major factor affecting the WHC of meat; low ultimate pH (near the isoelectric point of proteins) causes protein denaturation, which in turn lowers the WHC [24]. Since the pH showed no difference between the two breeds, there might be other factors (e.g., quality and quantity of protein or types of muscle fiber) affecting the WHC of beef samples rather than the ultimate pH.
Tenderness is an important factor determining the eating quality of meat [6]. For decades, the WBSF has been considered a highly reliable technique for evaluating beef tenderness [25,26]. Our results showed that in both muscles, the Hanwoo had a significantly (p < 0.05) lower WBSF value compared to the Chikso. According to the classification of beef tenderness based on the WBFS values by Belew et al. [26], the LL and SM muscles from Hanwoo can be considered "tender" (3.2 < WBSF < 3.9 kg) and "intermediate" (3.9 < WBSF < 4.6 kg) cuts, respectively. Whereas, both the muscles from Chikso can be considered "tough" (WBSF > 4.6 kg) cuts. The lower shear force values in the Hanwoo beef may be attributed to its higher intramuscular fat (IMF) content compared to the Chikso (our analysis showed that the IMF was approximately 3 times higher in the Hanwoo compared to the Chikso), because the level of IMF is negatively correlated to the WBSF values [27,28]. Otherwise, Chikso is associated with increased physical activity, which may result in tougher meat.
Color Traits
Color is known as an indicator of the freshness and wholesomeness of meat, and it largely affects the purchasing decisions of the meat by consumers [21,29]. The values of color traits of LL and SM muscles from Chikso and Hanwoo are presented in Table 2. It was observed that the cattle type affected all the color traits, in which the Chikso meat exhibited lower L* (lightness), a* (redness), and b* (yellowness) values compared to the Hanwoo meat (p < 0.05). This signifies that the Chikso meat seemed to be darker compared to the Hanwoo meat. Hughes et al. [30,31] conducted two studies to investigate the light scattering of beef longissimus muscles among color groups; these authors found that the light, medium, and dark groups had the L* (lightness) values of 35.3-37.3, 34.5-35.0, and 28.2-29.7, respectively. According to the beef classification based on color by these authors, Hanwoo and Chikso meat belonged to the light and medium color groups, respectively. Swatland [32] reported that dark beef usually has a swollen muscle fiber structure with less light scattering compared to light muscles. On the other hand, the color of beef is strongly affected by genetics, the feeding system, and its chemical composition (e.g., level of fat, moisture, and pigment myoglobin) [33][34][35]. Beef meat with a higher fat content is associated with a lighter and redder color [36]. According to our analysis, the IMF content in the Hanwoo was approximately 3 times greater than that in the Chikso. Therefore, the greater fat content of the Hanwoo may be responsible for its lighter color. Furthermore, the red color of meat is mainly determined by the pigment myoglobin and its biochemical states [29,34] as well as the ultimate pH [30]. In the present study, all the samples were collected at the same time (24 h postmortem), at the same temperature (4 • C), bloomed for the same period of time (30 min), handled under identical conditions, and had similar pH values ( Table 1). The paradoxical color results, therefore, may be due to the differences in muscle fiber structures, fat content, and pigment proteins between the two cattle breeds. Similar to the current finding, Aviles et al. [33] reported that breed has an effect on beef color in that local cattle breeds usually present darker meat with a higher haematin content (a dark blue pigment) compared to meat production-specialized breeds.
Free Amino Acids (FAA) and Metabolites
Taste is an important sensory trait of muscle foods. Meat tastes include saltiness, sweetness, sourness, and umami, and FAAs are known as meat's taste-active components [37,38]. The concentration of FAAs in beef from Chikso and Hanwoo is presented in Table 3. It was observed that alanine, glutamine, leucine, and glutamate were the most predominant FAAs in both the muscles from Chikso and Hanwoo. Supporting the present results, Jayasena et al. [39] and Cho et al. [40] reported a similar trend for these FAAs in the longissimus muscles of Hanwoo steers. Based on their similar taste qualities, Kato et al. [40], Dashdorj et al. [41], and Frank et al. [12] categorized the FAAs into different classes such as sweetness (threonine, glycine, alanine, serine, and proline) and umami taste (alanine, glutamine, glutamic acid, aspartic acid, lysine, methionine, and serine). Results show that in the LL muscle, alanine, proline, and threonine, and in the SM muscle, methionine and glutamine, were affected by the breed (p < 0.05). The concentrations of alanine, proline, and threonine (associated with sweet taste) were higher in Chikso's LL muscle, while methionine and glutamine (associated with umami taste) were higher in Hanwoo's SM muscle (p 0.05). Similar to the present results, Koutsidis et al. [11] and Frank et al. [12] found an effect of breed on FAAs in the LL beef muscle. In a study conducted by Dashdorj et al. [41], nearly half of FAAs in beef longissimus dorsi muscles were affected by breed (Hanwoo and Angus) when using different feeding diets. In a recent study conducted by Cho et al. [17], almost all FAAs detected in the same beef muscles were affected by gender, with meat from cows having a higher FAA content compared to steer meat. Metabolites are the final or intermediate products of the metabolic process in muscle tissues after slaughter, and many of them have been reported to contribute to the flavors of or act as precursors for the development of aromas in cooked meat [42,43]. The representative image of the 1 H NMR spectrum showing the presence of polar metabolites in the LL muscles of both cattle breeds is shown in Figure 1. In addition, as shown in Table 4, a total of 36 metabolites were identified and quantified in both muscles of two breeds. Out of them, lactate was the most predominant metabolite (261-236 mmol/kg), followed by creatine, carnosine, and glucose in both the muscles of the two breeds. Lactate and glucose have been reported to be formed from the glycolytic pathway in post-mortem muscles [44]. Adenosine 5monophosphate (AMP), inosine, and hypoxanthine are the nucleotide-derived metabolites formed from the breakdown of adenosine triphosphate (ATP) by endogenous enzymes, while carnitine, carnosine, creatine, and creatinine are formed from proteolysis [45]. The remaining compounds were amino acids, which are also formed from the proteolytic process by endogenous enzymes in meat [45]. Our statistical analysis shows that the breed affected the concentrations of 4 compounds (hypoxanthine, inosine, phenylalanine, and O-acetylcarnitine) in LL muscles, with a higher amount for the Hanwoo (p < 0.05). In the SM muscle, anserine, isoleucine, and trimethylamine were affected by the breeds (p < 0.05). Almost all of the metabolites found in our samples have also been reported in beef LL muscles from various breeds (Hanwoo and Santa Gertrudis-Brahman cross steers) [15,17,45] and pork [46]. Research conducted to examine the effect of sex metabolites in Hanwoo beef has shown that cow meat has higher acetate, creatine, creatinine, glycine, tyrosine, etc. than steer meat [17]. Hoa et al. [15] showed that the castration method also affects the metabolites in Hanwoo LL and SM beef muscles. Thus, this study examined for the first time the effect of breed on the FAAs and metabolites, with the results indicating that the variations in their concentrations may lead to a difference in the quality of cooked meat between the two breeds. Means within a row (in each muscle type) with different superscripts ( a , b ) differ significantly (p < 0.05).
Volatile Flavor Compounds
As a part of flavor, aroma (sensed by the nose) is among the most important factors determining the eating quality of cooked meat [9]. Survey studies on customer satisfaction for beef have revealed that flavor and tenderness contribute equally to overall like ratings [47,48]. The aroma and flavor are contributed by a variety of volatile compounds that are formed from various chemical reaction processes such as the thermal oxidation of fat, the Mallard reaction between amino acids and reducing sugars, and the interaction of intermediates between these two pathways during the cooking/heating of meat [13]. Studies have stated that the quality and quantity of volatile flavor compounds are largely affected by the contents of precursors (e.g., amino acids, sugars, fat, etc.) in the raw meat [9,11]. The concentration (µg/g) of volatile compounds in the cooked Chikso and Hanwoo meat samples is shown in Table 5. A total of 47 compounds-21 aldehydes, 7 alcohols, 4 ketones, 6 sulfur-and nitrogen-containing compounds, and 9 hydrocarbons-were detected and identified in the cooked LL muscles from both breeds. It is well recognized that almost all aldehydes (except the Strecker aldehydes) are formed in cooked meat from the thermal oxidation of fatty acids during cooking [13,49,50]. With their low odor detection thresholds, aldehydes appear to be important contributors to cooked meat aromas [9,13]. Out of the aldehydes, 8 compounds, including pentanal, hexanal, E,2-hexenal, heptanal, benzaldehyde, E,E,2,4-nonadienal, nonanal, and decanal, were affected by the breed, with a significantly (p < 0.05) higher amount in the Hanwoo compared to the Chikso. These aldehydes have been characterized as exerting pleasant aromas such as fatty, sweet, and fruity odor notes in cooked pork and beef [8,12]. The higher aldehyde content in Hanwoo beef could be the result of its higher IMF content, as mentioned above. Because the thermal oxidation of fat is the major pathway for the formation of these aldehydes during cooking [49,50], similar to the current finding, Hoa et al. [15,27] reported that beef with a higher IMF content has a higher amount of volatile flavor compounds. According to research on the effects of breed on flavor compounds of cooked beef, Wagyu beef with a higher IMF content has more fat-derived flavor compounds associated with fatty and sweet odors than the lower-fat Angus beef breed [12]. Beside the fat-derived aldehydes, some of the identified compounds (e.g., 2-methyl pentanal, 2-methyl propanal, 3-methyl butanal, and 2-methyl butanal) are known to be formed from the Strecker degradation of amino acids (e.g., isoleucine and leucine) during the cooking of meat [13,49]. Our result shows that all of these Strecker aldehydes were not affected by the breed (p > 0.05). This is probably due to the similar amounts of their corresponding precursors, such as FAAs (Table 3), and metabolites (Table 4), between the two breeds.
Alcohols, which have a low odor detection threshold, contribute significantly to the development of cooked meat flavor [8,13]. The result shows that all the identified alcohols (except 2-heptanol and 1-octen-3-ol) were affected by the breed, with a significantly higher amount in the Hanwoo meat (p < 0.05). Alcohols are known to be the products of the thermal oxidation of fatty acid-derived products during the cooking of meat [50]. Therefore, the higher amounts of the alcohols in Hanwoo meat may be related to its higher IMF content. 1-pentanol, hexanol, 1-hexanol, and 1-heptanol have been found to have desirable odor notes (oily, fatty, and green) in cooked meat [8].
Nitrogen-and sulfur-containing heterocyclic compounds derived from the Mallard reaction between amino acids and reducing sugars, with their low odor detection threshold, have been reported to contribute to the desirable aromas in cooked meat [13]. By using the gas chromatography-olfactometry technique, Frank et al. [12] demonstrated that the pyrazines contribute to the meaty and roasty odor of grilled beef. In the present study, 3 sulfur-containing compounds and 3 pyrazines were detected in the cooked LL muscles from both breeds. These identified compounds have also been found in cooked beef in the literature [15,17]. Out of them, 2 compounds, including methylpyrazine and 2-ethyl-3,5-dimethylpyrazine, were affected by the breed, with the Chikso meat having a higher amount compared to the Hanwoo (p < 0.05). The lower amount of identified pyrazines in Hanwoo meat could be due to the effect of fat-derived aldehydes, whose abundance can modulate the Mallard reaction pathway by inhibiting the formation of nitrogen-and sulfur-containing heterocyclic compounds [13,49]. Means within the same row with different superscripts ( a , b ) are significantly different (p < 0.05). ID (1) : Identification method: The compounds were identified by mass spectra (MS) from a library or external standards (STD).
Conclusions
This study demonstrates that under identical feeding conditions, the breed showed a significant effect on quality attributes such as shear force (tenderness), water holding capacity, color, and taste-and aroma-related components. Some sweetness-related free amino acids (alanine, proline, and threonine) were found at a higher concentration in the Chikso LL muscle, while a higher amount of methionine and glutamine associated with umami taste were found in the Hanwoo SM muscle. A total of 36 metabolites were identified and quantified in the meat samples; out of them, 7 compounds were affected by the breed. Hanwoo meat, with a higher amount of fat-derived aldehydes, may have a higher intensity of fatty aromas, while Chikso meat may be associated with a roasty aroma due to the presence of a higher amount of pyrazines. The insights into the effect of breed on the muscle fiber properties, shelf life, and consumer perception of beef under identical feeding conditions will be investigated in our future study. | 2023-02-16T16:03:40.778Z | 2023-02-01T00:00:00.000 | {
"year": 2023,
"sha1": "4b8d2bb5048257847acc22472283c29273aa72c7",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2304-8158/12/4/805/pdf?version=1676296877",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "89b111ff1e51656a945af4ccd31dc9b403f4848b",
"s2fieldsofstudy": [
"Agricultural And Food Sciences"
],
"extfieldsofstudy": []
} |
266572561 | pes2o/s2orc | v3-fos-license | Optimization of a Protocol for Isolating Cell-free DNA From Cerebrospinal Fluid
A standardized protocol for the isolation of cell-free DNA (cfDNA) from cerebrospinal fluid (CSF) is lacking. Therefore, we established a cfDNA isolation protocol optimized for clinical CSF specimens, integrating acceptable modifications and using artificial CSF generated from remnant CSF spiked with reference cell-free tumor DNA (ctDNA). We compared the isolation yields of in vitro diagnostic (IVD)-certified column-based (CB) and magnetic bead-based (MB) isolation. Furthermore, we modified both methods, including pre- and post-elution steps. To confirm ctDNA integrity and quantify the variant allele frequency after isolation, we performed droplet digital PCR (ddPCR) targeting IDH1 R132C in the reference ctDNA. MB isolation had a higher yield than CB isolation (P<0.0001), and post-isolation vacuum increased the final concentration in both methods, with little effect on cfDNA integrity. Our study provides a protocol to maximize CSF-ctDNA concentrations in IVD testing and future studies.
and verified it using five actual clinical CSF samples.
Artificial CSF samples were generated by adding Seraseq ctDNA Mutation Mix v2 AF 2% (SeraCare, Milford, MA, USA) as reference ctDNA to non-bloody remnant CSF.All samples were non-malignant CSF samples obtained via lumbar puncture between April 2021 and September 2023.The study was approved by the Institutional Review Board (IRB) of the Soonchunhyang University Seoul Hospital, Korea (IRB No. 2020-04-037-007).
Fifty-nine 2-mL aCSF samples containing ctDNA were prepared by spiking 1 µL or 2 µL of reference ctDNA into 118 mL of pooled CSF supernatant to obtain two levels of samples (Supplemental Data Figure S1).Twenty-nine aCSF samples were used for column-based (CB) isolation and 30 for magnetic beadbased (MB) isolation (Supplemental Data Figure S2).CB isolation was conducted using a Cobas cfDNA Sample Preparation Kit (Roche Diagnostics, Pleasanton, CA, USA).The elution step in the manufacturer's protocol was modified to two 50-µL elution steps from one 100-µL elution step.For comparison, aCSF-ctDNA was isolated from 14 samples per the manufacturer's instructions and from the remaining 15 samples using the modified method (Supplemental Data Figure S3).MB isolation was conducted using a Chemagic cfDNA 2k Kit H24 (Perkin Elmer, Hamburg, Germany) on a Chemagic 360 instrument (Perkin Elmer).The amount of beads was varied to 50 µL, 75 µL, and 100 µL among the 30 aCSF-cfDNA samples, and the addition of poly(A) RNA buffer reagent was evaluated.We concentrated the CSF-cfDNA using a HyperVACMAX VC2200 centrifugal vacuum concentrator (Hanil Scientific, Gochon-eup, Korea).The concentrated cfDNA pellets were transferred into the elution buffer for analysis.
Assuming that the baseline CSF-cfDNA was isolated together with reference ctDNA, we quantified the baseline CSF-ctDNA before reference ctDNA spiking, using a cell-free DNA ScreenTape assay (Agilent, Santa Clara, CA, USA).We determined the amount of cfDNA, considering that the ScreenTape assay has a sizing range of 100-5,000 bp.We calculated the percent isolation yield (%yield) using the following formula: %yield= aCSF sample quantity ×100% Baseline CSF quantity+Seracare AF2% quantity To confirm the presence of spiked ctDNA and quantify the variant allele frequency (VAF) after isolation, we performed droplet digital PCR (ddPCR) targeting IDH1 R132C (COSMIC ID: COSM28747) using 26 CB samples and 28 MB samples that had the minimal concentration required for PCR.The PCRs were run on a QX200 instrument (Bio-Rad, Hercules, CA, USA).The inherent VAF in the reference ctDNA was 2.53%.To quantify droplet copies and VAF, we used an IDH1 mutation assay (Bio-Rad).
Statistical analysis was performed using R Studio version 2022.02.3+492 bit (The R Foundation for Statistical Computing, Vienna, Austria) and GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA).
The highest %yield achieved with the two cfDNA isolation methods using aCSF was 82.5 ± 19.2 (56.5-116.9)% for twostep elution in case of CB isolation and 162.3 ± 47.7 (120.9-261.8)% for 75-µL bead volume with input poly(A) in case of MB isolation, compared with the values obtained using the theoretical concentration.In clinical CSF samples, the %yield was 194.7 ± 55.9 (125-261.8)% for MB isolation with a 75-µL bead volume in the presence of poly(A) (Fig. 1A-1C).
The CV (%) was 23.3% for CB and 29.4% for MB, respectively.Thus, MB isolation showed a more consistent and higher %yield than CB isolation (Fig. 1D).
To assess whether vacuum concentration affected ctDNA size, we compared the average ctDNA size before and after vacuum concentration.The concordance rate of 0.78 was obtained, which implies a moderate level of agreement (Fig. 1G) based on the intraclass correlation coefficient (2.1).
CSF contains lower levels of genomic nucleic acids and cfDNA than plasma but is rich in factors that act as inhibitors in nucleic acid amplification tests [3,4].Because of the blood-brain barrier, CSF-ctDNA can be used as a biomarker that is superior to plasma ctDNA in certain CNS tumors.However, the limited CSF acquisition volume and several contraindications for collection hamper the investigation of pre-analytical variables in CSF-ctDNA.To the best of our knowledge, this study included the largest number of clinical CSF samples.
Previous studies on CSF-cfDNA isolation mainly used silica CB technologies [5,6], particularly, the QIAamp circulating nucleic acid kit.The efficiency and recovery rate of CB isolation of CSF-cfDNA are superior to those of MB methods when combined with a post-vacuum step [7].As the CB kit used in our study was www.annlabmed.orghttps://doi.org/10.3343/alm.2023.0267intentionally developed without a post-vacuum concentration step, CSF-cfDNA isolation using the MB method had a better yield and lower CV than CB isolation.It is conceivable that, at low concentrations, the efficiency of CB methods is lower than that of MB methods [8].
Increasing the number of elution steps with half the elution volume improved the yield in CB isolation, indicating that additional elution increases the incubation time and the probability of reaction with the elution solution [9].Therefore, to increase the yield, we recommend eluting twice with half the volume of elution solution.This approach is often used to increase the yield of plasma cfDNA [10].
In the MB method, a bead volume of 75 µL showed a better yield than 50 µL or 100 µL for isolation from a 2-mL CSF sample.A bead volume of 50 µL is recommended in the manufacturer's protocol for plasma cfDNA isolation [11,12]; however, we consider 50 µL to be insufficient for the isolation of CSF-cfDNA because of the scarcity of cfDNA in CSF.However, excess bead volume ( > 75 µL in our study) is also not beneficial [13].
The addition of poly(A) buffer reportedly enhanced the amount of ctDNA isolated [14]; however, we observed no significant effect.We assume that this is because the CSF samples used in our experiment had substantially low ctDNA concentrations, and the effect of poly(A) be insignificant, given the nature of short cfDNA.
Although post-vacuum concentration did not enrich the CB eluates (140.8%) and MB eluates (156.6%) as much as expected, it did not affect the average size of the cfDNA, which is consistent with previous findings [15].
The effects of the yield difference between the CB-and MB-based CSF-cfDNA isolation methods were significant in terms of the number of total events in ddPCR but not in terms of VAF recovery.We speculate that this reflects an influence of residual genomic DNA in the baseline CSF samples that could have undergone droplet generation so that wild-type cytosine-containing sequences would have been amplified at higher levels than thymine mutant sequences, resulting in lower-than-expected VAF recovery [16].
Our study was limited in that we did not broadly compare cfDNA isolation methods or compare manual and automated methods.However, MB isolation of CSF-cfDNA, particularly when automated, has advantages in terms of isolation yield and low CV, improving workflow efficiency and ensuring a consistent output [17].To date, manufacturers have not released complete CSF-cfDNA isolation instructions; however, we experimentally established a protocol for in vitro diagnostic use.Our study provided an optimized protocol for reliable research and clinical testing. | 2023-12-29T06:16:52.947Z | 2023-12-28T00:00:00.000 | {
"year": 2023,
"sha1": "88c11510ac7ac725a91abb36900fc98677a02426",
"oa_license": "CCBYNC",
"oa_url": "https://www.annlabmed.org/journal/download_pdf.php?doi=10.3343/alm.2023.0267",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "4058d3cec7177545302ca33d6b7fe768346e989d",
"s2fieldsofstudy": [
"Medicine",
"Biology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
257766377 | pes2o/s2orc | v3-fos-license | Parameter Efficient Local Implicit Image Function Network for Face Segmentation
Face parsing is defined as the per-pixel labeling of images containing human faces. The labels are defined to identify key facial regions like eyes, lips, nose, hair, etc. In this work, we make use of the structural consistency of the human face to propose a lightweight face-parsing method using a Local Implicit Function network, FP-LIIF. We propose a simple architecture having a convolutional encoder and a pixel MLP decoder that uses 1/26th number of parameters compared to the state-of-the-art models and yet matches or outperforms state-of-the-art models on multiple datasets, like CelebAMask-HQ and LaPa. We do not use any pretraining, and compared to other works, our network can also generate segmentation at different resolutions without any changes in the input resolution. This work enables the use of facial segmentation on low-compute or low-bandwidth devices because of its higher FPS and smaller model size.
Introduction
Face parsing is the task of assigning pixel-wise labels to a face image to distinguish various parts of a face, like eyes, nose, lips, ears, etc. This segregation of a face image enables many use cases, such as face image editing [20,46,57], face e-beautification [37], face swapping [16,35,36], face completion [23].
Since the advent of semantic segmentation through the use of deep convolutional networks [31], a multitude of research has investigated face parsing as a segmentation problem through the use of fully convolutional networks [13,14,25,26,28,29]. In order to achieve better results, some methods [14,28] make use of conditional random fields (CRFs), in addition to CNNs. Other methods [25,27], * Work done during internship at Adobe Figure 1. The simple architecture of Local Implicit Image representation base FP-LIIF: A light convolutional encoder of modified resblocks followed by a pixel only MLP decoder focus on a two-step approach that predicts bounding boxes of facial regions (nose, eyes, hair. etc.) followed by segmentation within the extracted regions. Later works like AGRNET [48] and EAGR [49] claim that earlier approaches do not model the relationship between facial components and that a graph-based system can model these statistics, leading to more accurate segmentation.
In more recent research, works such as FaRL [59] investigate pretraining on a human face captioning dataset. They pre-train a Vision Transformer (ViT) [8] and finetune on face parsing datasets and show improvement in comparison to pre-training with classification based pre-training like ImageNet [44], etc., or no pre-training at all. The current state-of-the-art model, DML CSR [58], tackles the face parsing task using multiple concurrent strategies including multi-task learning, graph convolutional network (GCN), and cyclic learning. The Multi-task approach handles edge discovery in addition to face segmentation. The proposed GCN is used to provide global context instead of an average pooling layer. Additionally, cyclic learning is carried out to arrive at an ensemble model and subsequently perform self-distillation using the ensemble model in order to learn in the presence of noisy labels.
In this work, we perform face segmentation by taking advantage of the consistency seen in human facial structures. We take our inspiration from various face modeling works [1,12,61] that can reconstruct a 3D model of a face from 2D face images. These works show it is possible to create a low-dimensional parametric model of the human face in 3D. This led us to conclude that 2D modeling of the human face should also be possible with low dimension parametric model. Recent approaches, like NeRF [34] and Siren [47] demonstrated that it is possible to reconstruct complex 3D and 2D scenes with implicit neural representation. Many other works [2,11,43,56] demonstrate that implicit neural representation can also model faces both in 3D and 2D. However, to map 3D and 2D coordinates to the RGB space, the Nerf [34] and Siren [47] variants of the models require training a separate network for every scene. This is different from our needs, one of which is that we must map an RGB image into label space and require a single network for the whole domain. That brings us to another method known as LIIF [3], which is an acronym for a Local Implicit Image Function and is used to perform image super-resolution. They learn an approximation of a continuous function that can take in any RGB image with low resolution and output RGB values at the sub-pixel level. This allows them to produce an enlarged version of the input image. Thus, given the current success of learning implicit representations and the fact that the human face could be modeled using a low-dimension parametric model, we came to the conclusion that a low parameter count LIIF-inspired model should learn a mapping from a face image to its label space or segmentation domain. In order to test this hypothesis, we modify a low-parameter version of EDSR [24] encoder such that it can preserve details during encoding. We also modify the MLP decoder to reduce the computing cost of our decoder. Finally, we generate a probability distribution in the label space instead of RGB values. We use the traditional cross-entropy-based losses without any complicated training mechanisms or loss adaptations. An overview of the architecture is depicted in Figure 1, and more details are in Section 3. Even with a parameter count that is 1/26 th compared to DML CSR [58], our model attains state-of-the-art F1 and IoU results for CelebAMask-HQ [21] and LaPa [29] datasets. Some visualizations of our outputs are shared in Figure 3 and Figure 4.
To summarise, our key contributions are as follows: • We propose an implicit representation-based simple and lightweight neural architecture for human face semantic segmentation.
• We establish new state-of-the-art mean F1 and mean IoU scores on CelebAMask-HQ [21] and LaPa [29]. • Our proposed model has a parameter count of 1/26 th or lesser compared to the previous state-of-the-art model. Our model's SOTA configuration achieves an FPS of 110 compared to DML CSR's FPS of 76.
Face Parsing
Since face parsing intrinsically involves capturing the parametric relationship between the facial regions, the existing methods in face parsing aim at modeling the spatial dependencies existing in the pixels of the image. Multiple deep learning-based models with multi-objective frameworks have been proposed to handle spatial or inter-part correlations and boundary inconsistencies and capture the image's global context. Liu et al. [29] proposed a two-head CNN that uses an encoder-decoder framework with spatial pyramid pooling to capture global context. The other head uses the shared features from the encoder to predict a binary map of confidence that a given pixel is a boundary which is later combined with the features from the first head to perform face parsing. EAGRnet [49] uses graph convolution layers to encode spatial correlations between face regions into the vertices of a graph. Zheng et al. [58] combine these approaches to build DML-CSR, a dual graph convolution network that combines graph representations obtained from shallow and deeper encoder layers to capture global context. Additionally, they employ multi-task learning by adding edge and boundary detection tasks and weighted loss functions to handle these tasks. They also use an ensemble-based distillation training methodology claiming that it helps in learning in the presence of noisy labels. They achieve state-of-the-art performance on multiple face-parsing datasets. Recently a transformer-based approach has also achieved state-ofthe-art performance but with the help of training with additional data. FaRL [59] starts by pre-training a model with a face-image captioning dataset to learn an encoding for face images and their corresponding captions. Their image encoder, a ViT [8], and a transformer-based text encoder from CLIP [42] learn a common feature encoding using contrastive learning. They then use the pre-trained image encoder and finetune on various face-related task datasets to report state-of-the-art numbers. We compare our performance numbers with a non-pre-trained version of FaRL [59] because we wanted to test our model on only the task-related dataset, and using additional image-caption data was out-of-the scope of this work.
While these approaches handle the image as a whole to predict a single mask segmenting all the components Figure 2. Encoder Architecture: It has three res-block groups. The first two (2,6) res-block groups, followed by a strided convolution per group, are mainly used to reduce the spatial dimensions of the activation maps. The final group of res-blocks creates the grid of features vectors Z. Notice each res-block group has a group-level residual connection. simultaneously, some approaches model the individual classes separately. These approaches, called the local methods, claim that focusing on the facial components (e.g. eyes, nose, etc.) results in more accurate predictions but at the expense of efficient network structure in terms of parameter sharing. Luo et al. [32] propose a model which segments each detected facial part hierarchically into components and pixel-wise labels. Zhou et al. [60] built interlinking CNNs to perform localization followed by labeling. Lin et al. [25] propose an RoI-Tanh operator- Our proposed method is a whole image-based method that uses a single encoder-decoder pair to parse faces using implicit neural representations on the global image scale.
Parametric Human Face Models and Implicit representations
Parametric models for the human face have been explored for a long time since the pioneering work of [39]. Principle component analysis (PCA) was used to model face geometry and appearance in [1] to create a 3D Morphable model (3DMM) of the face. Many of the approaches where the 3D face model is estimated from the 2D image, estimate coefficients of the pre-computed statistical face models [41,50,51]. Other methods use regression over vox- Figure 5. ResBlock Modification: Comparison of the residual block design in EDSR with our modification. We add an Instance normalization after each convolution in the residual block. els or 3D meshes [10,53] of face to arrive at a detailed face model. Many approaches have also used 3DMM type models with deep learning techniques [6,7,9,18,45].
With the emergence of Implicit Neural Representation [33,34,38,47], a new approach to parameterizing signals has been gaining popularity. Instead of encoding signals as discrete pixels, voxels, meshes, or point clouds, implicit neural representation can parameterize these as continuous functions. A lot of work in the field has been around 3D reconstruction and shape modeling [4,15,30,33,34,38,55]. Deepsdf [38] encodes shapes as signed distance fields and models 3D shapes with an order of magnitude lesser parameters. Neural Volume [30] and NeRF [34] introduced 3D volume rendering by learning a continuous function over 3D coordinates. These works led to the use of implicit representation in the domain of human body or face rendering like [43,56], that use implicit representation to model human heads and torso in a detailed manner. Others like Pi-Gan [2] and NerFACE [11] used it in the domain of faces. NerFACE [11] can extract a dynamic neural radiance field face from a monocular face video and be used with the parameters of a 3DMM. Besides 3D modeling, implicit neural representation has also been used in 2D image-to-image translation. Local Implicit Image function (LIIF) [3] proposed an implicit neural representation-based super-resolution model that treats images in the continuous domain. Based on these approaches of low-dimensional parametric face models, stunning performance of implicit neural representation in 3D reconstruction, and 2D imageto-image translation, our choice of method for exploring face segmentation gravitated towards the implicit representation approach of LIIF. The 2D texture-less appearance of the face segmentation mask prompted us to explore a lowparameter version of the LIIF model for face parsing.
Methodology
The human face has a regular and locally consistent structure because the various features on the human face, like eyes, nose, mouth, .etc, would maintain their relative position. We use this uniformity to design a lightweight model for face parsing. We adopt the LIIF [3] framework to learn a continuous representation for segmentation for locally consistent structures of human faces.
Segmentation as Local Implicit Image Function
An image I in LIIF is represented by a 2D grid of features Z ∈ R H×W ×D such that a function f θ can map each z ∈ R D in Z to another domain. Here, f is an MLP, and θ are its parameters. This can be represented by eq 1: where, x ∈ X is a 2D coordinate, and s is the signal in the domain we want to convert our image I into. The coordinates are normalized in the [−1, 1] range for each spatial dimension. In this paper, s is the probability distribution among a set of labels, i.e., P (y|I, x), where y denotes the class label. So for a given image I, with latent codes z and query coordinate x q , the output can be defined as P (y|x q ) = f θ (z, x q ). So using the LIIF approach, we can write where z * is the nearest z to the query coordinate x q and v * is the nearest latent vector coordinate.
Other methods mentioned in LIIF [3], such as Feature Unfolding and Local Ensemble, are also used. Feature Unfolding is a common practice of gathering local information or context by concatenating the local z in the 3 × 3 neighborhood for each z in Z. To illustrate, the feature unfolding of a Z of dimension (H × W × D) would end up as (H × W × 9D). Local Ensemble is a way to address the discontinuity in f θ along sharp boundaries. An average of f θ is calculated for each pixel according to the four nearest neighbors of z * . This also bakes in a voting mechanism in the model at a per-pixel level.
Image Encoder
We now describe our image encoder that takes as input an RGB image of size 256 × 256 and generates an output volume of latent vectors of size 64 × 64. Our encoder is a modified version of EDSR [24] as shown in Figure 2. We modify all the resblocks by appending an instance normalization block [52] after every convolution layer, Figure 5. We create 24 resblocks Figure 2, and all convs have a size of 3 × 3 and filter depth of 64 channels unless otherwise stated. The input is first passed through a conv before passing it into resblock-groups.
We have three resblock-groups. We added the first two to extract and preserve the fine-grained information from the image while the activation volume undergoes a reduction in the spatial dimensions because of the strides conv. The third group of resblock is used to generate the image representation Z. Each of the resblock-groups are a series of resblocks followed by a residual connection from the input, Figure 2. The output of the first resblock-group that contains two resblocks is passed to a 3 × 3 conv with a stride of 2. This is passed to the second resblock-group which has six resblocks. This is again followed by a conv of stride 2. The output of this second downsampling is passed through the third resblock-group containing 16 resblocks. This generates a feature volume of size 64 × 64 × 64, which is then passed to the LIIF decoder.
LIIF decoder
The task of the decoder is to predict the segmentation labels at each pixel, which depends on the local context and the global context. Therefore, to provide global context during per-pixel prediction, we first extract the global context by doing an average pool of the latent volume along the spatial dimensions, as shown in Figure 6. Additional local context is added by passing the latent volume through a 3 × 3 unfolding operation, which increases the channel size to 64 × 9. The unfolded volume is then sent through a two-layer reduce channel MLP (RCMLP) with depths of 256 and 64. This makes the next upsampling operations computationally cheaper. The resulting volumeẐ of size 64 × 64 × 64 is bilinearly upsampled to the output size and concatenated with the previously extracted global feature and two channels of positional encoding. The positional encodings are x,y coordinates ranging from −1 to 1 along the spatial dimension of the feature volume. This volume of latent vectors is flattened and passed through a 4-layer MLP of 256 channels each to predict logits for the segmentation labels. Next, we perform a LIIF-like ensemble of these predictions using multiple grid sampling overẐ. Note that the regular conv can't be used to directly replace the unfolding operation because grid sampling result would differ for aẐ derived from a w×h×9D volume compared to a w×h×D volume because of the different neighbors of the D and 9D channels.
Loss
The logits are passed through a softmax and then guided with a cross-entropy loss L cce and edge-aware crossentropy loss L e cce . The edge-aware loss is calculated by extracting the edges of the ground truth label map with an edge detection kernel (Fig. 7) and calculating cross-entropy only on these edges. The final loss can be defined as: where λ is the additional weight for the edge-cross-entropy.
Datasets
We use three face datasets to perform our experiments, LaPa [29], CelebAMask-HQ [21] and Helen [19]. LaPa is a face dataset with more in-the-wild photos having varying poses and occlusions.
Implementation Details
We implement our training and evaluation pipeline on PyTorch [40] Table 4. Model size comparison: The table shows the parameter count, GFlops, and FPS for each of the models and the relative size of each model compared to FP-LIIF Figure 8. Few test samples from CelebAMask-HQ dataset illustrating noisy ground truth data and our prediction for the same. In the top row headgear has been marked as hair and in the bottom row strands of hair are not clearly segmented in the ground truth mask.
on Python 3.8.5. We train FP-LIIF on 4 Nvidia A-100 GPUs with a mini-batch size of 33 and 64 for CelebAMask-HQ and LaPa respectively. The network is optimized for 400 epochs using Adam [17] with an initial learning rate of 5e-4. The learning rate is decreased by a factor of 10 after every 20 epochs. The λ for edge-cross-entropy was set to 10 and 40 for CelebA and Lapa, respectively. Temperature scaling of softmax is also done with τ = 0.5. The images and masks in the datasets were resized to 256 × 256 using bicubic sampling before being used for training and evaluation. During training various data augmentations are applied like random affine transformations of rotation by 30°, shear between 0 to 20, scaling between 0.5 to 3 followed by random cropping. Color jitter is also applied to the input image with a brightness between [0.5,
Evaluation Metrics
To keep our evaluation consistent with other works, we primarily use a class-wise F1 score and a mean F1. In addition to that, we use mean intersection over union (mIoU) to compare with DML CSR. The background class is ignored in all these metrics.
Baselines
We compare our FP-LIIF performance with several baselines, like Wei et al. [54] (figures taken from [59]), AGR-NET [48], EAGR [49], DML CSR [58], FARL [59] from scratch, i.e., no pre-training. The results on LaPa are reported in Table 1, results comparing performance on CelebAMask-HQ are in Table 2 and results on Helen are in Table 3. Finally, a comparison of model size, Gflops and FPS is made in Table 4.
Results
According to Table 1's LaPa results, FP-LIIF performs better overall in mean-F1 and in classes such as eyes (lefteye and right-eye), brows (left-brow and right-brow), and skin. Table 2 demonstrates that our approach performs better than the baselines on CelebA in terms of mean-F1 and also at the class-level F1 of skin, nose, eyes (left-eye, righteye), lips(upper-lips, lower-lips), hair, necklace, neck, and cloth. We have also included a row of results demonstrating our performance when we change our output size to 512 × 512. The results show that even without training for a higher resolution output, our network seamlessly generates decent segmentation results at a higher resolution with nominal degradation in LaPa while still matching the current SOTA of 92.38 by DML CSR. Table 2 demonstrate superior performance at 512 resolution with a mean-F1 of 86.14, which is 0.07 higher than DML CSR. We achieve these results without training on multiple resolutions, i.e., we train on just 256 × 256 and the network seamlessly scales to multiple resolutions. Our results on the Helen dataset, which has a small number of training samples (2000), are in Table 3. Our performance is close to SOTA despite training on non-aligned face images. Last but not least, in Table 4, we comprehensively compare the model sizes, GFlops and FPS of all our baselines. With only 2.29 million parameters, FP-LIIF is the most compact face-parsing network available; it is 65 times smaller than FARL and 26 times more compact than DML CSR. Our fps, evaluated on an Nvidia A100 GPU, stood at 110 frames per second, whereas DML CSR's performance was at 76 frames per second, and EAGR and AGR-Net demonstrated fps of 71 and 70, respectively. Additional comparative analysis of our results with DML CSR is included in the supplementary.
Ablations
To evaluate the effect of several components we conduct the following ablations.
Network without LIIF Decoder: We replaced the LIIF decoder with a Conv U-Net type decoder. The total parameter count of this model is 9.31M params (3x FP-LIIF Comparison with lightweight segmentation model: Table 9 shows the results for face segmentation on LaPa using SFNet [22] which is a recent lightweight segmentation network for cityscapes. Figure 9. Comparison with SFNet [22], another lightweight segmentation network.
Low Resource Inference
One of the practical advantages of FP-LIIF is that it enables face parsing on low-resource devices. The primary prerequisite to enable low resource inference is that a model's inference should be low in compute and therefore have a high frame per second (FPS) count. Our ability to predict segmentation masks at multiple resolutions enables us to meet the demand for low inference costs. To achieve this, we can instruct the network to perform a low-resolution prediction, and the result can be upscaled to a higher resolution. Table 6 shows the FPS for lower-resolution inference on a single Nvidia A100 GPU. However, the shorter inference time should not result in poor quality output when upsampled to a higher resolution. Therefore we compare our upscaled outputs with the ground truth and present the findings in Table 1, 2. Here, we can see that our 192×192 or 128×128 segmentation output, when upscaled to 256×256, leads to a minimal loss in quality, as can be seen in both classwise and overall F1 scores. In Table 1
Limitation
Encouraged by the performance of this low parameter FP-LIIF network in face segmentation, we tested its effectiveness at semantic segmentation in a more generic domain like Cityscapes [5]. We chose this dataset because it lacked the structural regularity that we exploited in this work and segmentation using the current architecture should not be feasible. As expected, the mIoU score on the validation was reported at 62.2, which is 20+ points lower than SOTA models reporting scores in the range of ∼85 mIoU.
Conclusion and Future Work
This work presents FP-LIIF, an implicit neural representation-based face parsing network. We exploit the human face's regular and locally consistent structure to propose a low-parameter, local implicit image function-based network that can predict per-pixel labels for a human face image. Our model is 1/26 th or lesser in size compared to the current SOTA models of face parsing but outperforms them on mean-F1 and mean-IoU over multiple datasets like CelebAMask-HQ and LaPa. This network can also generate outputs at multiple resolutions, which can be very useful in reducing the inference time of segmentation by generating output at a lower resolution. These properties make it feasible to use this architecture on low-resource devices.
Future work will address misprediction in regions with fewer class labels. We would also extend Implicit neural representation-based segmentation to domains with a regular and consistent structure, like medical imaging, human body parts, etc., and to domains where structure uniformity is not guaranteed, like in the wild images. The quantitative results shown in Table 1, 2 points that even though FP-LIIF fares better in mean F1, the best classwise performance is scattered across multiple models. But at the same time, the gap between the best classwise scores and FP-LIIF's classwise scores is marginal. Therefore, we try to further identify the problematic areas and include visualizations of FP-LIIF's worst-performing results compared to DML CSR in F1 in Figure 11, 10. It can be seen from Figure 11 that rows a) and c) have negligible differences, and in the remaining rows, both are performing poorly in the problematic regions of hair and face. In Figure 10's rows b) and d), the F1 scores for these are debatable because of incorrect labeling in the ground truth. In the remaining rows, the underrepresented class of hat and earrings are bringing down our performance. Therefore the current setup of FP-LIIF is affected by a lack of data as compared to DML CSR. This can also be corroborated by Table 2. It is also necessary to point out that the ground truth data of CelebAMask-HQ is noisy ( Figure 8) and can cause problems in training and testing.
From the point of view of inference time, FP-LIIF could be used to generate segmentation at a lower resolution, and Figure 11. Few results where DML CSR performed better than FP-LIIF on LaPa dataset the generated output scaled at the required higher resolution to improve inference time and hence increase fps. The generation of lower-resolution segmentation does not require any additional training and is an outcome of being an implicit neural representation network. The 128-resolution version of FP-LIIF clocked an fps of 294 compared to the regular version of resolution 256, which runs at 120 fps. This makes our model more conducive for low compute devices.
Variance in performance over mulplte runs
We also calculate the mean and variance of our model's F1 score for Lapa, CelebAMask-HQ and Helen in Table 7 should be noted that other state-of-the-art works do not report these mean and variance over multiple runs and therefore direct comparison of these numbers is not possible. | 2023-03-28T01:22:38.128Z | 2023-03-27T00:00:00.000 | {
"year": 2023,
"sha1": "c293b29b17020ccdb11ee5cf7b80d8d436e6b78d",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "c293b29b17020ccdb11ee5cf7b80d8d436e6b78d",
"s2fieldsofstudy": [
"Computer Science"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
251252874 | pes2o/s2orc | v3-fos-license | Dynamics of groups of automorphisms of character varieties and Fatou/Julia decomposition for Painlev\'e 6
We study the dynamics of the group of holomorphic automorphisms of the affine cubic surfaces \begin{align*} S_{A,B,C,D} = \{(x,y,z) \in \mathbb{C}^3 \, : \, x^2 + y^2 + z^2 +xyz = Ax + By+Cz+D\}, \end{align*} where $A,B,C,$ and $D$ are complex parameters. We focus on a finite index subgroup $\Gamma_{A,B,C,D}<{\rm Aut}(S_{A,B,C,D})$ whose action not only describes the dynamics of Painlev\'e 6 differential equations but also arises naturally in the context of character varieties. We define the Julia and Fatou sets of this group action and prove that there is a dense orbit in the Julia set. In order to show that the Julia set is ``large'' we consider a second dichotomy, between locally discrete and locally non-discrete dynamics. For an open set in parameter space, $\mathcal{N} \subset \mathbb{C}^4$, we show that there simultaneously exists an open set in $S_{A,B,C,D}$ on which $\Gamma_{A,B,C,D}$ acts locally discretely and a second open set in $S_{A,B,C,D}$ on which $\Gamma_{A,B,C,D}$ acts locally non-discretely. After removing a countable union of real-algebraic hypersurfaces from $\mathcal{N}$ we show that $\Gamma_{A,B,C,D}$ simultaneously exhibits a non-empty Fatou set and also a Julia set having non-trivial interior. The open set $\mathcal{N}$ contains a natural family of parameters previously studied by Dubrovin-Mazzocco. The interplay between the Fatou/Julia dichotomy and the locally discrete/non-discrete dichotomy plays a major theme in this paper and seems bound to play an important role in further dynamical studies of holomorphic automorphism groups.
Every line parallel to the x-axis intersects the surface S A,B,C,D at two points (counted with multiplicity) and one can therefore define an involution s x : S A,B,C,D → S A,B,C,D that switches them: s x (x, y, z) = (−x − yz + A, y, z) . Consider also the finite index subgroup Γ ≡ Γ A,B,C,D = g x , g y , g z < Γ * , where g x = s z • s y , g y = s x • s z , and g z = s y • s x .
Date: August 3, 2022. 1 The dynamics of the action of groups Γ * A,B,C,D and Γ A,B,C,D on S A,B,C,D and their individual elements have several deep connections, including to the dynamics of mapping class groups on character varieties, to the monodromy of the Painlevé 6 differential equation, and to the aperiodic Schrödinger equation; see, for example, [6] for a nice description of these connections. We will provide more details about the former two connections in Section 2.
The previous works arising from dynamics on character varieties can be traced back to Markoff Theorem and to works on Markoff Surfaces, see [5] and [50]. To the best of our knowledge, deeper investigations on these dynamics follow two main trends: (i) Global dynamics on the 2-dimensional real (singular) surfaces for real parameters A, B, C, D, as initiated by Goldman [24] (see also [23]). (ii) Study of domains in the complex surface S A,B,C,D on which the group Γ A,B,C,D acts properly discontinuously, as initiated by Bowditch [5] and later studied extensively by several authors. We refer the reader especially to the works by Tan, Wong, and Zhang [54], Maloni, Palesi, and Tan [38], and Hu, Tan, and Zhang [27].
Based on motivations from the monodromy of the Painlevé 6 differential equation, the previous dynamical results follow three main trends: (i) Finite orbits under Γ A,B,C,D correspond to algebraic solutions of Painlevé 6. They were classified by Dubrovin and Mazzocco [16] and by Lisovyy and Tykhyy [36]. See also [7, Section 4] for a classification of bounded orbits. (ii) Study of individual mappings from Γ and Γ * displaying rather interesting dynamics by Iwasaki and Uehara [32], including uniformly hyperbolic ones by Cantat [6]. These mappings share many features in common with the complex Hénon mappings. (iii) Proof by Cantat and Loray [7] that except for the case Picard parameters (A = B = C = 0, D = 4), the action of Γ on S A,B,C,D preserves neither an (multi) affine structure nor a (multi) holomorphic foliation. This corresponds to the Malgrange irreducibility of Painlevé 6, as explained in their paper.
To the best of our knowledge, the previous dynamical works related to the aperiodic Schrödinger equation primarily involve delicate issues about the iteration of a single real mapping γ ∈ Γ A,B,C,D on the real slice of S A,B,C,D . There is a huge literature on the subject and we refer the reader to the papers by Casdagli and Roberts [13,49] and references therein for an introduction. For more contemporary works, see, for example, the paper of Damanik, Gorodetski, and Yessen [14] and the paper of Yessen [56].
Main goal of our work: Our paper is devoted to questions about the "pointwise complex dynamics of the whole group", i.e. to the orbits of individual points, their closures, and more generally to the nature of subsets of the complex surface S A,B,C,D that are invariant under Γ * and Γ.
We focus on the "chaotic" part of the dynamics (i.e. the dynamics on the Julia set J A,B,C,D , as defined in Section 1.5) which is therefore complementary to the extensive work previously done by Bowditch, Tan and his collaborators, and others about the domains on which the dynamics is properly discontinuous.
Let us point out that the complex dynamics on the SL(2, C) character varieties (i.e. of Γ A,B,C,D on S A,B,C,D ) is described as being "extremely mysterious and non-trivial" by Goldman [24, p. 461] and by Tan, Wong, and Zhang [54, p. 762]. Our goal in this paper is to provide a better understanding of this global dynamics, especially of the chaotic part of it which is far less developed than regions where the action is properly discontinuous.
The fact that mapping class group actions on character varieties can be formulated as a complex dynamical systems, which incidentally is the object of this paper, allows one to see the present work as joining a recent trend of papers where methods of holomorphic dynamics were used to investigate certain natural dynamical systems, see [6], [10], and [11].
Also, from a different perspective, the complex dynamical system in question describes the transverse dynamics of the celebrated Painlevé 6 equation. This provides additional motivation to describe the corresponding dynamics in its entirety as accurately as possible. In particular, the splitting of dynamics in two regions with contrasting dynamical behavior -a region called Fatou where the dynamics is simple (e.g. properly discontinuous) and another region named Julia where the dynamics is chaotic -also accounts for the somehow "double nature" of Painlevé 6: in the vast literature on the subject, it is possible to find a thread where this equation is viewed as part of integrable system and another one where it is regarded as a complicated dynamical systems. The simultaneous existence of large Fatou and Julia sets (both with non-empty interiors cf. Theorem G) justifies and conciliates both perspectives. In particular, the results obtained in this paper -and especially those involving the Julia set -are in line with general programs aimed at the dynamical study of Painlevé equations, cf. [30] and [46].
It is important to remark that we are considering the action of Γ A,B,C,D on the (non-compact) affine surface S A,B,C,D ⊂ C 3 . Indeed the elements of Γ A,B,C,D extend only as birational mappings of the compactification S A,B,C,D ⊂ CP 3 , with both indeterminate and super-attracting/collapsing behavior at infinity (see Section 8). This lack of compactness, the wild behavior of birational mappings at infinity, and the "attracting nature of infinity for Γ A,B,C,D " makes several aspects of this dynamical system rather challenging. It also seems to rule out the possibility of ergodictheoretic methods like those recently used by Cantat and Dujardin [8], in the case of automorphism groups of compact surfaces. This will be discussed more in Section 1.8.
1.2. Some "preferred" parameters. Throughout the paper we will refer to the following specific parameters and parametric families.
Picard Parameters: (A, B, C, D) = (0, 0, 0, 4). For these parameters, Picard proved that the Painlevé equation has explicit first integrals and countably many algebraic solutions. This is related to the curious fact that the action of Γ 0,0,0,4 is semi-conjugate to an action on (C\{0}) 2 by monomial mappings. In particular, for these parameters everything can be computed rather explicitly.
Punctured Torus Parameters: (A, B, C, D) = (0, 0, 0, D) for any D ∈ C. These parameters correspond to dynamics on the character variety of the once punctured torus; see, e.g. [6,Sec. 1.1]. For real D and the corresponding real surfaces, this is the family studied by Goldman [24].
Notice that both the Markoff and Picard parameters are included within the Punctured Torus Parameters. Meanwhile, the Picard parameters are in the closure of the Dubrovin-Mazzocco parameters, corresponding to a = −2, but the Markoff parameters are not.
1.3. Two relevant dynamical dichotomies. With the goal of producing interesting invariant sets and finding points with complicated orbit closures we introduce two dynamically invariant dichotomies.
Whereas the action of Γ (or of Γ * ) is genuinely non-linear, i.e., it cannot be embedded into the action of some finite dimensional topological group, it still appears to share some basic properties/issues with actions of countable subgroups of finite dimensional Lie groups. This typically happens on some (proper) open subsets of the surface S A,B,C,D and, for this reason, the notions of locally discrete vs. locally non-discrete dynamics of Γ will come in handy. Meanwhile, to deal with the non-linear nature of the global dynamics we will adapt the Fatou/Julia theory to the group Γ. The core of this paper lies in the interplay between these four notions. It is probably fair to say that these group actions provide a setting where characteristics of linear and of non-linear dynamics nicely blend together and this very phenomenon lends further interest to their study. 1.4. Locally non-discrete/discrete dichotomy. Let M be a (possibly open) connected complex manifold and consider a group G of holomorphic diffeomorphisms of M . The group G is said to be locally non-discrete on an open set U ⊂ M if there is a sequence of maps {f n } ∞ n=0 ∈ G satisfying the following conditions (see for example [48]): (1) For every n, f n is different from the identity.
(2) The sequence of maps f n converges uniformly to the identity on compact subsets of U . If there is no such sequence f n on U we say that G is locally discrete on U .
Remark that for an action by a finite dimensional Lie group, local non-discreteness on some open set implies that the corresponding sequence of elements converges to the identity on all of M , i.e. that the action is globally non-discrete. However, in our context the non-linearity of the mappings allow for local non-discreteness to occur It follows from the definitions that F A,B,C,D is open while J A,B,C,D is closed. Furthermore both sets are invariant under Γ. (For some parameters S A,B,C,D may be singular, but this is not an issue: we will see in Remark 3.1 that such singular points are always in J A,B,C,D .) It is worth to emphasize that the Julia set is non-empty for every choice of parameters; see, for example, Lemma 4.3. However, the Fatou set can be empty for some parameter values; indeed this happens for the Picard Parameters (Theorem D(i), below).
1.6. Main results. Classical results from the holomorphic dynamics of rational maps of the Riemann sphere assert that there is a dense orbit in the Julia set (topological transitivity) and that repelling periodic points are dense in the Julia set. We search for analogous statements for the action of Γ A,B,C,D on S A,B,C,D .
Theorem A. For any parameters (A, B, C, D) there is a point p ∈ J A,B,C,D such that i.e., there is a dense orbit of Γ in J A,B,C,D .
In our setting of group actions, the natural analog of having a dense set of repelling periodic points consists of looking for a dense set J * A,B,C,D ⊂ J A,B,C,D of points whose stabilizers contain a hyperbolic element. A negative answer to this question is provided by Theorem D, below, which states, in particular, that this is not the case for the Picard Parameters (0, 0, 0, 4).
Whereas we leave it as an open question to characterize for which parameters (A, B, C, D) the set J * A,B,C,D is dense in J A,B,C,D , we are still able to provide an affirmative answer to this type of question up to replacing "hyperbolic derivative" by derivative conjugate to a "shear map". This is the content of Theorem B below. Often parametric families of rational maps of the Riemann sphere have some mappings with connected Julia set and other mappings with disconnected Julia set. In our context we have a slightly surprising general topological property of Julia sets, namely: Theorem C. For any parameters A, B, C, D the Julia set J A,B,C,D is connected.
Let us now transition from results that hold for all parameters to results that only hold for certain parameters. In contrast to the Picard parameters, for which the corresponding Fatou set is empty, Teichmüller theory can be applied to show that the Fatou set is non-empty for certain values of the parameters, including the Markov parameters (A, B, C, D) = (0, 0, 0, 0). A more powerful approach stems from issues related to the Bowditch Conjecture and the Bowditch BQ Conditions, as introduced by Bowditch [5] and Tan, Wong, and Zhang [54] and later studied by Maloni, Palesi, and Tan [38], Hu, Tan, and Zhang [27], and several others. See Section 2.2 for more details. Their methods can be adapted to prove the following theorem: Theorem E. We have the following: (1) Punctured Torus Parameters: For any complex D not equal to 4 the Fatou set F 0,0,0,D is non-empty. (2) Dubrovin-Mazzocco Parameters: For any a ∈ (−2, 2) the Fatou set F A(a),B(a),C(a),D(a) is non-empty. Moreover, the result carries over to an open neighborhood in C 4 of any such parameter.
We will denote the subset of the Fatou set obtained in the proof of Theorem E by V BQ ≡ V BQ (A, B, C, D) and call it the Bowditch set because any point p ∈ V BQ satisfies the BQ Conditions. As a matter of fact, the action of Γ on V BQ is properly discontinuous, as was shown in [54] and [38]. (We will also show this in the proof of Theorem F, below.) Because of this context, the Bowditch set seems to be of considerable interest and, in particular, the following rephrasing of Statement (1) in Theorem E seems to be new: However, let us note that we recently discovered that Statement (2) of Theorem E is an immediate consequence of a stronger statement [38,Theorem 5.3]. In general, a complete classification of all (A, B, C, D) ∈ C 4 for which V BQ (A, B, C, D) is non-empty seems to be a delicate question.
Because the Fatou set can be non-empty it is interesting to determine how "large" the Julia set is, especially if one wishes to apply Theorems A and B. For this reason we study the interplay of the Fatou/Julia dichotomy with the locally non-discrete/discrete dichotomy. We will first discuss the locally non-discrete/discrete dichotomy and then relate it to the Fatou/Julia dichotomy.
In Section 7, it will be shown the existence of a very large set of parameters (A, B, C, D) for which Γ A,B,C,D is locally non-discrete on some open subset of the surface S A,B,C,D . We therefore have the following theorem about the coexistence of both locally discrete and locally non-discrete dynamics: Theorem F contrasts with the behavior of the much studied locally discrete and locally nondiscrete subgroups of circle diffeomorphisms, see Section 1.9 for further detail. Remark. The invariant set B A,B,C,D ⊂ S A,B,C,D of topological dimension 3 "persists" over the open subset of parameters P ⊂ C 4 . The existence of persistent invariant sets of topological dimension 3 for the action of a large group (Γ is free on two generators) strongly hints at a fractal nature for B A,B,C,D .
As in the case of Theorem E, Teichmüller theory can also be used to establish the existence of B A,B,C,D for certain values of the parameters. In these cases, B A,B,C,D is the boundary of the Bers component and hence is very irregular; see for example [39, Appendix A].
With the notation of Theorem F, we expect that for each (A, B, C, D) ∈ P we have U ⊂ J A,B,C,D . However, at present, we can only prove this under one further (weak) assumption, namely: (P) any fixed point of any γ ∈ Γ A,B,C,D \ {id} is in J A,B,C,D . We prove in Proposition 9.8 that there is a countable union of real-algebraic hypersurfaces H ⊂ C 4 such that this Property P holds if (A, B, C, D) ∈ C 4 \ H.
Theorem G. Let P ⊂ C 4 be the open neighborhood of the Markoff Parameters and of the Dubrovin-Mazzocco parameters, a ∈ (−2, 2), given in Theorem F and let H ⊂ C 4 be the countable union of real-algebraic hypersurfaces provided by Proposition 9.8. For any (A, B, C, D) ∈ P \ H we have: Here, U and V BQ are the (non-empty) open subsets from the statement of Theorem F.
Corollary to Theorem G. For any (A, B, C, D) ∈ P \ H there exists p ∈ U ⊂ J A,B,C,D such that i.e. the orbit of p has closure of real dimension four, but it is not all of the surface S A,B,C,D since the orbit of p misses the Fatou component V BQ .
Remark 1.1. Any γ ∈ Γ A,B,C,D preserves an invariant holomorphic 2-form Ω whose equation is given in Section 3.3. This forces that the eigenvalues of Dγ at any fixed point p ∈ S A,B,C,D are resonant, λ 1 = 1/λ 2 , making it rather unlikely that such a fixed point is in the Fatou set. Indeed, the possibility that Hypothesis (P), above, fails for any parameters (A, B, C, D) is a known challenging in holomorphic dynamics. It is explained at length in the paper [40] by Mc-Mullen, see especially the remark on p. 220 of that paper. 1.7. Strategy for Proof of Theorem G. Let us briefly describe the strategy for proving Theorem G, which is arguably the most elaborate result in our paper. It is straightforward to prove that any Fatou component V is Kobayashi hyperbolic. In particular, if Γ is locally non-discrete on any open subset of V then it is locally non-discrete on all of V . The idea is then to show that a region U where Γ induces a "complicated enough" locally non-discrete dynamics is not compatible with the structure of a (Kobayashi hyperbolic) Fatou set. This region must hence be contained in J A,B,C,D and this yields Julia sets with non-empty interior. In order to rule out the possibility that this region U intersects an unbounded Fatou component, the following theorem will be needed. Theorem H. Suppose that for some parameters A, B, C there is a point p ∈ C 3 and ǫ > 0 such that for any two vertices v i = v j ∈ V ∞ , i = j, there is a hyperbolic element γ i,j ∈ Γ A,B,C satisfying: Then, for any D, we have that B ǫ/2 (p) ∩ S A,B,C,D is disjoint from any unbounded Fatou components of Γ A,B,C,D . Here, K(ǫ) > 0 denote the constant given in Proposition 7.1.
We refer the reader to Proposition 8.2 for the definition of hyperbolic element γ along with the corresponding points Ind(γ) and Attr(γ). Hypothesis (A) requires that the six elements γ i,j have sufficiently rich "combinatorial behavior" at infinity while Hypothesis (B) requires that these six elements are sufficiently close to the identity on the ball B ǫ (p). Note that the conditions of Theorem H are explicit and easy to check. In particular, for any (A, B, C, D) ∈ P they imply that U is disjoint from any unbounded Fatou component.
The idea of the proof of Theorem H is that if B ǫ (p) were in an unbounded Fatou component V then we use local non-discreteness to produce a sequence of elements converging uniformly on compact subsets of V to the identity and we use the prescribed combinatorial behavior at infinity to show that this same sequence of elements sends compact subsets of V uniformly to infinity.
Having ruled out the possibility that an unbounded Fatou component intersects U we then use the following theorem to prove that no bounded Fatou component intersects U .
Theorem K. Suppose that (A, B, C, D) ∈ C 4 \ H and that V is a bounded Fatou component for Γ A,B,C,D . Then the stabilizer Γ V of V is cyclic.
It is a standard result that the group Aut(V ) of holomorphic automorphisms of a Kobayashi hyperbolic manifold V is a real Lie group. To prove Theorem K we use that Γ A,B,C,D preserves a volume form (see Section 3.3) to show that the closure G = Γ V of Γ V is a compact Lie group. Checking that any element of Γ V has infinite order we conclude that G has positive dimension. Supposing that Γ V is non-cyclic we can conclude that it is non-Abelian and moreover that there are non-commuting elements arbitrarily close to the identity. This gives that the connected component of the identity, G 0 , is non-Abelian. Since G 0 is compact, it must therefore have real-dimension 3.
The assumption that no element of Γ has a fixed point in V gives that G 0 acts freely on V and thus that V /G is a manifold of real dimension 1. This allows us to derive a contradiction to the fact that the Julia set is connected (Theorem C).
1.8. Relationship to the recent work of S. Cantat and R. Dujardin. A recent work by S. Cantat and R. Dujardin [8] studies the dynamics of subgroups of the automorphism group of (compact) real and complex projective surfaces (or, more generally, Kähler surfaces). They obtain deep results on the structure of stationary measures for these subgroups, including criteria to determine when they must be invariant by the group in question (stiffness) as well as detailed descriptions of the resulting invariant measures. Whereas their results clearly have serious implications on the corresponding dynamics, the situation seems to be genuinely more subtle in the case of groups of birational maps. Relevant examples include the case of the actions of Γ and Γ * since, as shown in the present paper, for certain values of the parameters these actions display both non-empty Fatou set and Julia sets with non-empty interior. This phenomenon indicates that the problem of invariance of stationary measures, and of their subsequent description, will hardly allow for a reasonably "compact" classification. Also, the nature of the Fatou components constructed in our Theorem E shows that even the convergence of stochastic processes will no longer be automatic which, in turn, seriously limits the applications of stationary measures in topological problems such as description of non-compact invariant closed sets.
1.9. Further comments on the locally non-discrete/discrete dichotomy. The notion of locally non-discrete group (or even pseudogroup) first appeared in [47] after previous works by A. Shcherbakov, I. Nakai, and E. Ghys, see [53,43,22]. The notion was further elaborated in [37]. The main common tool of [47] and [37] is the construction of certain vector fields obtained as a sort of limit of certain dynamics in the group/pseudogroup in question which allow for a detailed analysis of the corresponding group dynamics. In turn, the method put forward in those papers ensures the existence of the desired vector fields for locally non-discrete (pseudo-) groups provided there is a local expansion in the dynamics: typically, we would like some element in group to have a fixed point where all its eigenvalues are of modulus greater than 1. This issue poses major difficulties to extend these methods to situations where the dynamics does not include a local expansion, as it happens for example in [48]. In particular, the issue about local expansion becomes particularly challenging when the dynamics of Γ (or Γ * ) is discussed. Indeed, whereas there are large sets of parameters for which the action of Γ is locally non-discrete, this action always preserves a volume-form on S A,B,C,D (cf. Section 3.3) so that the dynamics of elements in Γ never includes an expansion. Clearly, this prevents us from using the mentioned technique of "vector fields" and hence forced us to look for new methods to provide insight into the dynamics in question.
It should be pointed out that the structure of locally non-discrete groups of diffeomorphisms of the circle is rather well developed, see for example [18], [1], [12], and the references therein. However, in that case the relatively simple nature of the topological dynamics of groups acting on (real) 1-dimensional manifolds leads to the fact that coexistence of local discreteness and local non-discreteness is impossible. More precisely, in Diff ω (S 1 ), if a non-Abelian group G ⊂ Diff ω (S 1 ) is locally non-discrete on an interval I ⊂ S 1 , then every point p ∈ S 1 has a neighborhood where G is locally non-discrete, see [18]. This dramatically contrasts with our Theorem F which asserts that the same group action Γ A,B,C,D can simultaneously have open sets U and V on which it acts locally non-discretely and locally discretely, respectively. When combined with the complicated behavior of the pointwise dynamics of Γ A,B,C,D , in particular the existence of invariant sets with topological dimension 3 (Corollary of Theorem F), we find very rich dynamics in the system that we study.
Let us also mention that V. Kleptsyn and his collaborators have found examples of locally discrete subgroups of Diff ω (S 1 ) that are not conjugate to Fuchsian groups, up to finite covering [1]. In our context, the Picard parameters provide an analogous non-trivial example of a group that is "purely locally discrete", i.e., there is no open set U ⊂ S A,B,C,D , U = ∅, where the group acts in a locally non-discrete way.
1.10. Structure of the paper. We present in Section 2 a discussion of the motivations for a dynamical study of Γ A,B,C,D , with emphasis on the connections to dynamics on character varieties and to the Painlevé 6 differential equation. In Section 3 we present several basic properties of the group Γ ≡ Γ A,B,C,D and of the surface S ≡ S A,B,C,D that will be used throughout the paper. In Section 4 we study properties of the "parabolic" elements of Γ and use them and Montel's Theorem to prove Theorems A, B, and C. Section 5 is devoted to a careful study of the Picard parameters (A, B, C, D) = (0, 0, 0, 4) and a proof of Theorem D. In Section 6 we prove Theorem E about existence of the unbounded Fatou components, following the techniques from [5,54,38,27]. In Section 7 we produce examples of locally non-discrete dynamics and that will be needed for Theorem F. In Section 8 we collect several important properties of how Γ acts near infinity that will be needed later in the paper, including a proof of Proposition 9.8, which plays a key role in Theorem G. Section 9 is devoted to a proof of Theorem H about unbounded Fatou components and Theorem K about bounded Fatou components. We finish the proofs of our theorems with Section 11 where we prove Theorems F and G.
Acknowledgments:
We are grateful to Micha l Misiurewicz for ideas which helped with the proof of Proposition 5.7. We also thank Eric Bedford, Philip Boalch, Serge Cantat, Jeffrey Diller, Romain Dujardin, Simion Filip, Yan Mary He, John Hubbard, Bernard Julia, Mikhail Lyubich, Jean-Pierre Ramis, and Ser Peow Tan for interesting comments and discussions regarding this work. The second author is grateful to his colleagues Pavel Bleher, Alexander Its, Eugene Mukhin, Vitaly Tarasov, and Maxim Yattselev for their feedback on several early versions of this work. This work was supported by CNRS, through IEA "Dynamics for groups of birational maps acting on surfaces", and by the US National Science Foundation through grant DMS-2154414 and DMS-1348589.
Dynamics on Character Varieties and the Painlevé 6 equation.
The most compelling motivations for studying the dynamical systems considered here comes from dynamics in character varieties and from the sixth Painlevé equation: the two motivations being closely related as will soon be seen.
2.1.
Motivations from Dynamics on Character Varieties. We begin with character varieties as the dynamical system discussed in this paper is equivalent to the action of the mapping class group of a surface on the space of SL(2, C)-representations of its fundamental group, up to conjugation. This contains, in particular, the standard action of mapping class groups on Teichmüller spaces and hence can be viewed as belonging to the setting of "higher Teichmüller theory" as well; see, e.g., [55].
The dynamics we are interested in can be cast in the broader framework of (natural) dynamics on character varieties as was first pointed out in W. Goldman's papers [24] and [23]. Let us briefly explain how character varieties arise in this context. Consider, for example, the case of a punctured torus. Its fundamental group Π is isomorphic to the free group F 2 on two generators. The space of representations of Π in SL(2, C) is clearly identified with SL(2, C) × SL(2, C) which, in turn, is acted upon by SL(2, C) via simultaneous conjugation. The corresponding character variety is, by definition, the (categorical) quotient of the latter action. Next, note that automorphism group of Π also acts on the space of representations by pre-composition and this action descends to the character varieties. However, on the character variety, the action of inner automorphism of Π becomes trivial so that the action Aut(Π) factors through an action of the group Out (Π) consisting of outer automorphisms of Π.
Whereas the previous construction essentially makes sense for representations of Π in any group G, the fact that we are dealing with G = SL(2, C) can further be exploited as follows. As pointed out, the character variety in question is identified with pair of matrices in SL(2, C) × SL(2, C) up to simultaneous conjugation. A classical result of Fricke, however, shows that the latter space can be parameterized by C 3 we inherit an action of Out (Π) on C 3 which, in addition, coincides with the action on C 3 of the group Γ * obtained by setting A = B = C = 0. Incidentally, the latter group coincides with the group of polynomial automorphisms of C 3 preserving the (Markoff) polynomial x 2 + y 2 + z 2 − xyz − 2, cf. [26] and see [45] for a general result on the polynomial nature of similar actions. Finally, note that an analogous construction applies to the quadruply-punctured sphere. In the latter case, the corresponding action of Out (Π) recovers the action of Γ and Γ * with all the parameters A, B, C, and D, see [2] for a detailed account. Other than [2], more or less comprehensive versions of the previous discussion appear in a number of papers including [24], [23], [6], [7], [26].
2.2.
Properly Discontinuous Dynamics, quasi-Fuchsian representations, and Bowditch BQ Conditions. We now elaborate how the dynamical context explained in Section 2.1 leads to the existence of parameters with non-empty Fatou sets.
Consider the general case of the 4-holed sphere Σ 0,4 and note that its fundamental group can be identified with the free group on three generators Π ∼ = F 3 . Furthermore, since the orbits of the action of SL(2, C) on itself by conjugation are also of dimension 3, we see that the space of representations from Π to SL(2, C) (up to conjugation) has 6 complex dimensions. A far more accurate description is possible: the space of representations can be identified with the quartic hypersurface of C 7 given by S = {(a 1 , a 2 , a 3 , a 4 , x, y, z) ∈ C 7 : x 2 + y 2 + z 2 + xyz = Ax + By + Cz + D} , where A, B, C, D are as follows: A = a 1 a 4 + a 2 a 3 , B = a 2 a 4 + a 1 a 3 , C = a 3 a 4 + a 1 a 2 , and (4) D = 4 − [a 1 a 2 a 3 a 4 + a 2 1 + a 2 2 + a 2 3 + a 2 4 ] . In particular, by fixing the variables a 1 , . . . , a 4 , we obtain the surface S A,B,C,D . The parameters a 1 , . . . , a 4 are identified with the traces of the matrices in SL(2, C) arising from the loops around the holes of Σ 0,4 .
In turn, recalling that the group of outer automorphisms of the fundamental group of Σ 0,4 is isomorphic to the mapping class group of Σ 0,4 , the latter acts on the space Rep qf (Σ 0,4 ) = Teich (Σ 0,4 ) × Teich (Σ 0,4 ) diagonally with respect to its standard action on Teich (Σ 0,4 ). In particular, the action of the mapping class group on Rep qf (Σ 0,4 ) is properly discontinuous. Now, note that the 4-holed sphere consists of two pair of pants joined by the "waist". Hence the real dimension of Teich (Σ 0,4 ) is 6 so that the real dimension of Rep qf (Σ 0,4 ) is 12 and therefore Rep qf (Σ 0,4 ) has has non-empty interior in S. Finally, if the parameters a 1 , . . . , a 4 are chosen so that the corresponding surface S A,B,C,D intersects Rep qf (Σ 0,4 ) and this intersection contains an open set U ⊂ S A,B,C,D , then the preceding shows that Γ acts properly discontinuously on U and hence that U is contained in the Fatou set of the action of Γ on S A,B,C,D .
Note that it is easy to find parameters a 1 , . . . , a 4 so that the resulting surface S A,B,C,D does not intersect Rep qf (Σ 0,4 ). Recalling that a 1 , . . . , a 4 are the traces of matrices in SL(2, C) arising from loops around the holes, it is enough to force one of these matrices to be an elliptic element conjugate to an irrational rotation. An alternative argument leading to open sets of parameters a 1 , . . . , a 4 for which S A,B,C,D is disjoint from Rep qf (Σ 0,4 ) can be formulated by using the well-known Jorgensen inequality.
It is easy to prove that if (x, y, z) ∈ Rep qf (Σ 0,4 ) then (1) and (2) hold, but the converse is known as the Bowditch Conjecture. See [52,34,51] for recent related works and more details.
The proof of Theorem E provided in this paper will be an adaptation of ideas from the papers [54,38] and also the later paper by Hu, Tan, and Zhang [27]. To clarify this connection, let us begin by recalling that in the most standard notation, the sixth Painlevé equation takes on the form where α, β, γ, and δ are complex parameters. K. Iwasaki in [29] provided a useful alternative way to represent the parameters involved in this equation by denoting them as κ 1 , κ 2 , κ 3 , and κ 4 where the following formulas hold: Now, as a non-autonomous differential equation of second order, P6 can equivalently be viewed as a vector field Z VI on C 3 having the form where the function H α,β,γ,δ (x, y, z) is obtained from the right side of (6) by substituting z for dy/dx. In particular, the variable x can naturally be identified with "time" in the standard form (6). Let D denote the foliation defined on C 3 defined by the local orbits of Z VI . The foliation D is holomorphic since it can be defined by a polynomial vector field obtained from Z VI by multiplying it by the denominator appearing in the rational expression for the function H α,β,γ,δ (x, y, z). It is immediate to check that the foliation D on C 3 is transverse to the fibers of the standard fibration (projection) π x of C 3 to the x-axis, away from the invariant fibers sitting over {x = 0} and {x = 1}. In [44], Okamoto obtained a much nicer birational model for the foliation D. By compactifying the fibers of π x in a suitable Hirzebruch surface, performing a number of well chosen blow up maps, and removing a certain resulting divisor the following setting for the foliation D (see for example [44], [31]): Owing to the fourth condition, paths contained in C \ {0, 1} can be lifted in the leaves of D so as to yield a homomorphism ρ from the fundamental group of C \ {0, 1} to the group of holomorphic diffeomorphisms Diff (F ) of F . In other words, ρ(π 1 (C \ {0, 1}) ⊂ Diff (F ) is the holonomy group of D whose action on F encodes the transverse dynamics of the initial Painlevé 6 equation.
Whereas the a mapping RH (κ 1 ,κ 2 ,κ 3 ,κ 4 ) can hardly be computed due to its highly transcendent nature, it still provides a holomorphic conjugation between the dynamical systems in question. In particular, at least on compact parts of F , conjugation invariant properties of the dynamics of Γ can immediately be translated into dynamical properties of Painlevé 6, and conversely. This applies in particular to our results involving the dynamics of Γ in its Julia set. The group of automorphisms of C 3 generated by s x , s y , and s z is isomorphic to the free product Z/2Z * Z/2Z * Z/2Z. In particular, the group of automorphisms generated by g z , g y , and g z is free on two generators. However, a stronger statement holds: for every choice of the parameters A, B, C, and D, the group Γ * is isomorphic to the indicated free product and the group Γ is free on two generators, when viewed as a group of automorphisms of S A,B,C,D .
The non-existence of additional relations between the maps s x , s y , and s z even when restricted to a particular surface S A,B,C,D is a consequence of El-Huiti's theorem in [17] -albeit not an immediate one. For details the reader can check [6] and [7].
3.2. Singular points of S A,B,C,D . We begin by considering the natural compactification of C 3 in CP 3 so that CP 3 = C 3 ∪ Π ∞ , where Π ∞ ≃ CP 2 is the plane at infinity in CP 3 . Next, denote by S A,B,C,D the closure of S A,B,C,D in CP 3 . By using the standard affine atlas for CP 3 , it is straightforward to check that S A,B,C,D ∩ Π ∞ consists of three projective lines forming a triangle In particular, there follows that S A,B,C,D ∩ Π ∞ locally coincides with the axes {u = v = 0} and {u = w = 0}. An analogous use of the remaining coordinates shows that the third side of the mentioned triangle coincides with the projective line of Π ∞ ≃ CP 2 which is missed by the domain of the affine coordinates (v, w) ≃ (u = 0, v, w). A slightly less immediate computation with the above equation for S A,B,C,D also shows that S A,B,C,D is smooth on a neighborhood of Π ∞ . Therefore S A,B,C,D is singular if and only if the affine surface S A,B,C,D is so (and the corresponding singular sets always coincide). Since S A,B,C,D is Stein, this implies that the singular set of S A,B,C,D consists of isolated points and, in fact, it is finite since it is contained in a compact part of S A,B,C,D (S A,B,C,D is smooth near Π ∞ ). The surface S A,B,C,D is singular if and only if at least one of the following conditions is satisfied (see, e.g. [2,29]): • We have a i = ±2 for at least one i ∈ {1, 2, 3, 4}.
• The coefficients a 1 , . . . , a 4 satisfy the relation Since Γ must preserve Sing (S A,B,C,D ), it that Γ has a finite orbit whenever Sing (S A,B,C,D ) = ∅. The existence of a finite orbit for Γ is naturally yields some insights in the dynamics of Γ, as it will be seen in the course of this work.
and a simple calculation shows that s * x Ω = −Ω and similarly for s y and s z . Therefore, the elements of Γ preserve Ω and hence also preserve the associated real volume form Ω ∧ Ω. It assigns infinite volume to the whole surface S A,B,C,D but a finite volume to a sufficiently small neighborhood each singular point of S A,B,C,D .
Let p be a smooth point of S A,B,C,D . An immediate consequence of the existence of the invariant volume form is that if p is fixed for γ ∈ Γ A,B,C,D then det(Dγ(p)) = 1. In particular, if p is a hyperbolic fixed point for γ, then it must be of saddle type.
3.4. Pencils of rational curves. Recall that π x : Let S x 0 denote the closure of S x 0 in S A,B,C,D . Since S x 0 has degree two, it is uniformized by the Riemann Sphere provided that S x 0 is smooth. The statement remains valid in the case where S x 0 is singular up to passing from S x 0 to its normalization. Denoting by π y and π z the projections of C 3 to C respectively defined by π y (x, y, z) = y and π z (x, y, z) = z, the fibers Π y 0 , S y 0 , Π z 0 , and S z 0 are analogously defined.
Clearly the collection of rational curves obtained from S x 0 , x 0 ∈ C, defines a rational pencil in S A,B,C,D . By performing finitely many blow-ups, this pencil becomes a singular rational fibration D x , with connected fibers, over a suitable Riemann surface S A,B,C,D . Naturally, there are analogous pencils D y and D z defined on S A,B,C,D with the help of the collection of rational curves S y 0 and S z 0 contained in S A,B,C,D .
Let us close this section with an elementary lemma, whose simple proof will be omitted.
Lemma 3.2. For all but finitely many values of x 0 ∈ C, the rational curve S x 0 is smooth and intersects the plane at infinity Π ∞ of CP 3 in two distinct points. Analogous statements hold for the rational curves S y 0 and S z 0 .
For the pencil D x on S A,B,C,D , we fix a (minimal) blow-up procedure turning D x into a (singular) rational fibration D x . Clearly there are only finitely many values of x 0 ∈ C corresponding to singular fibers of this fibration. We then define a finite set B + x ⊂ C by saying that x 0 ∈ B x if at least one of the following conditions fails to hold: • The rational curve S x 0 is smooth and intersects the plane at infinity at two distinct points.
• The fibration D x is regular on a neighborhood of the fiber sitting over x 0 . We also define B x ⊆ B + x to be the set of points at which the first of the two conditions above fails to hold. The corresponding sets for the coordinates y and z will similarly be denoted by B + y , B y and by B + z , B z .
Dynamics of parabolic maps and proof of Theorems A, B, and C
We begin by studying the generators g x = s z • s y , g z = s x • s z , and g z = s y • s x of Γ. Here these mappings are explicitly written as The map g x preserves the coordinate x and hence each affine plane Π x 0 ⊂ C 3 . Since, in addition, g x clearly preserves the (affine) surface S A,B,C,D , it follows that g x individually preserves each one of the curves S x 0 (and hence also the rational curves S x 0 ⊂ S A,B,C,D ). Similar conclusions hold for the maps g y and g z .
We can now complement the discussion in Section 3 revolving around Lemma 3.2. Owing to these statements, we know that the rational curve S x 0 is smooth and intersects the divisor at infinity Π ∞ of CP 3 in two distinct points for all but finitely many values of x 0 ∈ C. It is then natural to consider the automorphism of S x 0 induced by g x . The following proposition can be found in [7] (see in particular Proposition 4.1 in the paper in question). The proof is elementary, in the spirit of most of the discussion in the previous section.
Proposition 4.1. Let x 0 ∈ C be such that the rational curve S x 0 is smooth and intersects the divisor at infinity Π ∞ in two distinct points. Then the restriction g x 0 of g x to S x 0 is a Möbius transformation whose two fixed points are at infinity. Furthermore, we have: • If x 0 ∈ (−2, 2) then g x 0 is elliptic. It is periodic if and only if x 0 = ±2 cos(θπ) with θ rational.
The analogous statements hold for restrictions of g y to the fibers S y 0 and of g z to the fibers S z 0 .
Remark 4.2. The case x 0 = ±2 deserves an additional comment for the sake of clarity. First note that the affine singular points of the curves S x 0 are contained in the affine curve given by In fact, the curve above parameterizes the set of points where the intersection of S A,B,C,D and Π x 0 is not transverse. From this, we see that this curve intersects the plane at infinity for x 0 = ±2.
In particular, the affine curve S 2 (resp. S −2 ) is always smooth. On the other hand, this curve intersects the plane Π ∞ ⊂ CP 3 at a single point (with multiplicity 2) which, depending on the coefficients, may or may not be a singular point of S 2 (resp. S −2 ). The restriction g 2 of g x to S 2 is as follows: (1) When S 2 is smooth, then g 2 is a parabolic map whose single fixed point coincides with the intersection of S 2 with Π ∞ . (2) Otherwise S 2 consists of the union of two projective lines intersecting each other at a point in Π ∞ . Each line is then preserved by g 2 and, in each of these lines, g 2 induces a parabolic map whose fixed point coincides with their intersection. The analogous statement holds for g −2 and S −2 . Proposition 4.1 yields the following lemma: Analogous statements hold for the g y and g z mappings.
Proof. Suppose that U is an open neighborhood of a point p ∈ S x 0 . Because x 0 ∈ (−2, 2) \ B x , Proposition 4.1 ensures that g x restricted to the fiber S x 0 is elliptic with both fixed points lying in Π ∞ . In particular, the iterates g n According to Proposition 4.1, the restriction of g x to the fiber S x 1 is hyperbolic with both fixed points at infinity. Therefore the orbit g n x (q) tends to infinity. Hence U cannot be contained in the Fatou set of g x since it contains points with both bounded and unbounded orbits. The lemma follows.
We will need the following elementary lemma, whose simple proof is omitted.
Fix any x 0 ∈ C. Then, for all but finitely many choices of y 0 the fibers S x 0 and S y 0 intersect transversally. When the fibers intersect transversally, they do so at two distinct points.
, the corresponding rational curve S x intersects Π ∞ transversally and at two distinct points. In other words, S x \ S x consists of two distinct points provided that S x ⊂ T ǫ (x 0 ). On the other hand, up to a birational transformation, the projective curves S x define a (regular) holomorphic fibration over the disc D ǫ . Because the fibers are rational curves and therefore pairwise isomorphic as Riemann surfaces, the theorem of Fischer and Grauert [20] implies that this fibration is holomorphically trivial, i.e., it is holomorphically equivalent to D ǫ × CP 1 . Since S x \ S x consists of two points, there also follows that as complex manifolds.
The hypothesis that S y 0 intersects S x 0 transversally implies that their intersection consists of exactly two (distinct) points (Lemma 4.4). We can then reduce ǫ > 0, if necessary, so that S y 0 intersects each S x from T ǫ (x 0 ) transversally in two points, each of them varying holomorphically with x. These points will be referred to as the two branches of S y 0 in T ǫ (x 0 ). Since each branch is the graph of a holomorphic function on x, we can pick either one of them and construct a further holomorphic diffeomorphism to make it correspond to the point 1 ∈ C \ {0}. By means of this construction, T ǫ (x 0 ) \ S y 0 becomes identified with an open set in D ǫ × C \ {0, 1}. The lemma follows since the latter manifold is Kobayashi hyperbolic as the product of two hyperbolic Riemann surfaces.
x and let y 0 be any point chosen so that S x 0 and S y 0 intersect transversally. For any open U ⊂ S with U ∩S x 0 = ∅ there is an iterate n such that g n x (U )∩S y 0 = ∅. Analogous statements hold when x and y are replaced with any two distinct variables from {x, y, z}.
Proof. Owing to Lemma 4.5, we fix a tube T ǫ (x 0 ) as in (10) which is Kobayashi hyperbolic. Let then U be a non-empty open set of S A,B,C,D intersecting S x 0 . Up to trimming U , we can assume that U ⊂ T ǫ (x 0 ). Since g x preserves the S x fibration and fixes the points in S x ∩ Π ∞ (g x has no poles), it follows that g n x (U ) remains in T ǫ (x 0 ) for every n ∈ Z. Now assume for a contradiction that g n x (U ) ⊂ T ǫ (x 0 ) \ S y 0 for every n. Since T ǫ (x 0 ) \ S y 0 is Kobayashi hyperbolic, this implies that {g n x } forms a normal family on U . This is, however, impossible since U intersects S x 0 and S x 0 ⊂ J (g x ) (Lemma 4.3). Thus there must exist n such that g n x (U ) ∩ S y 0 = ∅ and the lemma follows. Remark 4.7. The above proof actually shows slightly more than the statement of Lemma 4.6. Once ǫ > 0 is chosen sufficiently small so that T ǫ (x 0 ) intersects S y 0 in two branches (each a graph of a holomorphic function of x) and once U is chosen sufficiently small so that U ⊂ T ǫ (x 0 ) then for each branch of S y 0 in T ǫ (x 0 ) there is an iterate n so that g n x (U ) intersects the branch in question. Lemma 4.6 admits a useful quantitative version that can directly be proved, namely: converging to some q ∈ S x 0 and assume, in addition, the existence of δ > 0 such that the argument of x j lies in an interval of the form [δ, π − δ] ∪ [π + δ, 2π − δ] for every j. Then for every j large enough, there corresponds n j ∈ Z such that g Analogous statements hold when x is replaced by y or z.
Proof. Recall that g x preserves every leaf of the foliation D x induced by the collection of rational curves S x ⊂ S A,B,C,D . The assumption on x 0 implies that g x on the curve S x 0 corresponds to an elliptic element that is conjugate to an irrational rotation. Thus, up to saturating U by the dynamics of g x , there is no loss of generality in assuming that U is a disc bundle over an annulus Furthermore the "contracting coefficient" of g x j tends to 1 as q j → q. In particular, for j large enough, no fundamental domain of g x j on S x j ≃ CP 1 can contain the annulus A. The lemma follows at once.
The preceding lemmas are summarized by the proposition below.
Given two open sets U 1 , U 2 ⊂ S, both of which intersect S x 0 , there is an iterate n such that g n x (U 1 ) ∩ U 2 = ∅. Analogous statements hold when x 0 is replaced with y 0 or z 0 .
Proof. As in the proof of Lemma 4.6, we work within the tube T ǫ (x 0 ) defined in Equation (11). Since U 2 is open and U 2 ∩S x 0 = ∅ they have infinitely many points of intersection. We can therefore pick a point (x 0 , y 0 , z 0 ) ∈ U 2 ∩ S x 0 so that S y 0 intersects S x 0 transversally (Lemma 4.4). We can make ǫ > 0 smaller, if necessary, so that S y 0 ∩ T ǫ (x 0 ) is expressed as two smooth branches, each of them coinciding with the graph of a holomorphic function of x. Let S y 0 (ǫ) denote the branch passing through (x 0 , y 0 , z 0 ). Since U 2 is open, it follows that U 2 ∩ S y 0 (ǫ) is an open neighborhood of (x 0 , y 0 , z 0 ) in S y 0 (ǫ). This means that we can reduce ǫ > 0 even further, if necessary, so that we can assume the entire branch S y 0 (ǫ) is contained in U 2 .
With the above setting, the combination of Lemma 4.6 and Remark 4.7 ensures the existence of n such that g n x (U 1 ) ∩ S y 0 (ǫ) = ∅. Since S y 0 (ǫ) ⊂ U 2 , the proposition follows.
Recalling that the sets B + x , B + y , and B + z are all finite, Lemma 4.4 allows us to choose points 2) \ B z and to form the "grid" so that (i) every pair of curves (fibers) have transverse intersection (possibly empty), and (ii) each x 0 , x 1 , y 0 , y 1 , z 0 , and z 1 is of the form ±2 cos(θπ) for some irrational θ.
We will also say that any pair of irreducible components of G with empty intersection in S A,B,C,D are "parallel", e.g. S x 0 and S x 1 are parallel.
Proof. It is clearly enough to show the existence of γ ∈ Γ with γ(U ) ∩ S x 0 = ∅. If U already has non-trivial intersection with S x 0 then there is nothing to prove. Otherwise, there are two cases to be considered: Case 1: U intersects an irreducible components of G that is not parallel to S x 0 . Without loss of generality, we can suppose U intersects S y 0 . Lemma 4.6 then implies the existence of an iterate g n y of g y so that g n y (U ) ∩ S x 0 = ∅. Case 2: U intersects the component S x 1 that is parallel to S x 0 . In this case, we first apply Lemma 4.6 to find an iterate g n x of g x such that g n x (U ) ∩ S y 0 = ∅. Hence the problem is reduced to the situation treated in Case 1 so that it suffices to proceed accordingly.
Proof. Using Propositition 4.10 we can find some γ 1 ∈ Γ so that γ 1 (U 1 ) and U 2 both intersect the same irreducible component of G. The result then follows immediately from Proposition 4.9.
Remark 4.12. Concerning Corollary 4.11, note that the non-empty intersection γ(U 1 ) ∩ U 2 may be disjoint from G (and hence it might be disjoint from J A,B,C,D as well). Proposition 4.13 below is the last ingredient in the proof of Theorem A. Notwithstanding its very elementary nature, this proposition is likely to find further applications in the study of the dynamics associated with the group Γ. Proof. Assume aiming at a contradiction that γ(U ) is disjoint from G for every γ ∈ Γ. Then, for every γ ∈ Γ we have that where ι : S A,B,C,D ֒→ C 3 denotes the inclusion. Applying Montel's Theorem to each coordinate, this implies that the whole group Γ forms a normal family on U . This contradicts the assumption that U ∩ J A,B,C,D = ∅ and establishes the statement.
Proof of Theorem A. The proof is based on Baire's argument. First note that J A,B,C,D has the Baire property since it is a complete metric space as a closed subset of a manifold. The topology in J A,B,C,D is the one inherited from the topology of S A,B,C,D and hence it is second countable, i.e., there is a countable basis {V k } ∞ k=1 for the topology of J A,B,C,D . By definition the open sets V k ⊂ J A,B,C,D are given by where U k is an open set of S A,B,C,D . To prove the theorem, it suffices to show that the restriction to J A,B,C,D of the action of Γ is topologically transitive in the basis {V k } ∞ k=1 . Namely, given k 1 and k 2 , we need to prove the existence of γ ∈ Γ such that γ(V k 1 ) ∩ V k 2 = ∅. In fact, assuming the existence of these elements γ, for every n ∈ N, consider the set formed by all points in J A,B,C,D whose orbit intersects V n . This set is clearly open since V n is so. It is also dense in J A,B,C,D since it intersects non-trivially every set V k defining a basis for the topology of J A,B,C,D . Taking then the intersection over n ∞ n=1 γ∈Γ we obtain a G δ -dense subset of J A,B,C,D . By definition, the Γ-orbit of any point in this intersection visits all the open sets V k so that these points have dense orbits in J A,B,C,D .
It remains to show that for any two open sets V k 1 and V k 2 from our basis there exits γ ∈ Γ satisfying γ(V k 1 ) ∩ V k 2 = ∅.
We start by working with the corresponding open sets U k 1 and U k 2 of S. We first use Proposition 4.13 to find γ 1 , γ 2 ∈ Γ such that γ 1 (U k 1 ) and γ 2 (U k 2 ) each hit the grid G. We can then use Proposition 4.10 to find γ 3 ∈ Γ such that γ 3 • γ 1 (U k 1 ) and γ 2 (U k 2 ) intersect the same irreducible component of G. Without loss of generality, we suppose it is S x 0 .
Since S y 0 is transverse to S x 0 , we can choose a sequence of points {q j } ∞ j=1 ⊂ S y 0 converging to q ∈ S x 0 ∩ S y 0 that satisfies the hypotheses of Lemma 4.8. Moreover, by Lemma 4.3, S y 0 ⊂ J A,B,C,D so each element of the sequence is in J A,B,C,D . We therefore find a point q N ∈ J A,B,C,D and γ 4 , γ 5 ∈ Γ with γ 4 (q N ) ∈ γ 3 • γ 1 (U k 1 ) and γ 5 (q N ) ∈ γ 2 (U k 2 ). In other words, Since J A,B,C,D is invariant under Γ, this proves that 14. After finding γ 1 , γ 2 ∈ Γ such that γ 1 (U k 1 ) and γ 2 (U k 2 ) each hit the grid G, it is tempting to use Corollary 4.11 to find γ 3 with However, this does not necessarily prove that γ 3 (γ 1 (V k 1 ))∩γ 2 (V k 2 ) = ∅ because the intersection (14) need not be in G and hence it potentially might not contain any points of J A,B,C,D ; see Remark 4.12. This is why we use the "quantitative" Lemma 4.8 instead.
Proof of Theorem B. Let us define a new "grid" G ′ using Equation (12), but this time we will use x 0 = y 0 = z 0 = 0 and These values are chosen so that As in Proposition 4.13, if U is any open set that intersects J A,B,C,D non-trivially then there is an element γ ∈ Γ with γ(U ) intersecting G ′ . In fact, the proof of Proposition 4.13 does not use the choices that x 0 , x 1 ∈ B x , y 0 , y 1 ∈ B y , or z 0 , z 1 ∈ B z that were made in the construction of our original grid G so that it applies equally well to G ′ . Conjugating by γ, if necessary, it then suffices to prove that shear fixed points are dense in our newly chosen grid G ′ . We will prove it for S x 0 and S x 1 and leave the completely analogous proofs for S y 0 , S z 0 , S y 1 and S z 1 to the reader.
Every point of S x 0 is a fixed point for g 2 x . We will show that all but finitely many of them are shear fixed points. Clearly this assertion is, in turn, equivalent to showing that the derivative D(g x ) 2 has two dimensional generalized eigenspace associated to eigenvalue 1 but only one eigenvector associated to eigenvalue 1. Considering g 2 x as a mapping from C 3 → C 3 we have If z = C/2 or y = B/2 then this matrix has generalized eigenspace of dimension 3 associated to the eigenvalue 1 but only two eigenvectors, namely e 2 = [0, 1, 0] and e 3 = [0, 0, 1]. Therefore, it suffices to prove that for every p ∈ S x 0 bar some finite set. Taking the gradient of the defining equation for S A,B,C,D yields that Restricted to x = x 0 = 0 we can only have e 2 ∈ T p S A,B,C,D if y = B/2 and we can only have e 3 ∈ T p S A,B,C,D if z = C/2. Combined with x = 0 each of these conditions amounts to at most two points of S x 0 . Therefore, all but at most finitely many points of S x 0 are shear fixed points of g 2 x . The situation for S x 1 is essentially the same, except that one must work with g 4 x . We leave the details to the reader.
Proof of Theorem C. Let p 1 and p 2 be two arbitrary points in J A,B,C,D . We will show that for any neighborhoods U 1 of p 1 and U 2 of p 2 , there is a path in J A,B,C,D connecting from U 1 to U 2 . Theorem C will immediately follow.
The grid G given in (12) is path connected and, by virtue of Lemma 4.3, we have G ⊂ J A,B,C,D . Therefore, it suffices to find a path connecting from inside of U 1 to G and a path connecting from inside of U 2 to G. As the situation is symmetric, it suffices to consider U 1 .
Since p 1 ∈ U 1 ∩J A,B,C,D , Proposition 4.13 gives some γ ∈ Γ such that γ(U 1 )∩G = ∅. Let C be an irreducible component of G with γ(U 1 ) ∩ C = ∅. Since we have chosen the irreducible components of G to be smooth, C is biholomorphic to C \ {0}.
Consider now the Riemann surface γ −1 (C) contained in S A,B,C,D ⊂ C 3 . Since γ is a holomorphic diffeomorphism of S A,B,C,D , there follows that γ −1 (C) is again biholomorphic to C \ {0} and hence it is uniformized by C and contained in C 3 . We claim that γ −1 (C) intersects the grid G. Indeed, if we had γ −1 (C) ∩ G = ∅, the uniformization map from C to γ −1 (C) would yield a (non-constant) holomorphic map from C to C 3 each of whose coordinates omits two values in C. Picard theorem would then imply that this map must be constant and this is impossible.
Finally, note that γ −1 (C) ⊂ J A,B,C,D since J A,B,C,D is invariant by Γ and C ⊂ G ⊂ J A,B,C,D . Furthermore, by construction, γ −1 (C) also intersects U 1 . Since γ −1 (C) is path connected, we can therefore find a path contained in γ −1 (C) ⊂ J A,B,C,D going from U 1 to G. The proof of Theorem C is complete. From our point of view, this case is also very interesting as it will soon be clear. More importantly, however, the information collected in the course of this discussion will enable us to prove Proposition 9.8 in Section 9. Albeit a somewhat technical statement, Proposition 9.8 plays an important role in the proofs of Theorems G and K.
Throughout this section we will typically drop the parameters from our notation, writing S ≡ S 0,0,0,4 , Γ ≡ Γ 0,0,0,4 , J ≡ J 0,0,0,4 , and so on. The singular locus of S will be denoted by S sing . To prove Proposition 5.1 we will use the existence of a semi-conjugacy between the action of Γ and the group action of monomial mappings on C * × C * , which is described in the sequel (see for example [7,Section 1.5]). Consider the group: The group Γ(2) is generated by the matrices and it is straightforward to check that the mapping sending these matrices to g x , g y , and g z is an isomorphism from Γ(2) to Γ. For every M ∈ Γ(2), the image of M in Γ by the above mentioned isomorphism will be denoted by f M .
Associated with a matrix M = {m ij } ∈ Γ(2) is a monomial mapping η M : C * × C * → C * × C * given by It turns out that Φ is a degree two orbifold cover. Furthermore, a straightforward verification shows that that the critical points of Φ are precisely the four points (u, v) = (±1, ±1) while the corresponding critical values are the four points of S sing .
Proposition 5.2. Φ semi-conjugates the action of Γ(2) on C * × C * to the action of Γ on S. More specifically, given M ∈ Γ(2) and (u, v) ∈ C * × C * , we have Proof. This can be checked by a direct calculation in terms of the generators (18) of Γ(2) and of the generators g x , g y , and g z of Γ.
Proof of Proposition 5.1. Suppose that there is an open U ⊂ S and a sequence M n ∈ Γ(2) \ {id} such that f Mn | U converge locally uniformly to the identity on U . Making U smaller, if necessary, we can assume that U is evenly covered by Φ and that V is one of the two connected components of Φ −1 (U ). Then, the monomial maps η Mn converge locally uniformly to the identity on V . However, this is impossible because Γ(2) is a discrete subgroup of GL(2, Z). Since the action is question is linear, from the discrete character of Γ(2) there follows that this action on pairs (log |u|, log |v|) is locally discrete as well.
The semiconjugacy from Proposition 5.2 also allows us to completely determine the Julia set associated with Picard parameters. Namely, we have: The proof of Proposition 5.3 will, however, require the following lemma: Lemma 5.4. The set formed by the union over all hyperbolic elements of Γ(2) of the corresponding eigendirections is dense in P 1 (R).
Proof. Conjugating by elements of Γ(2), the problem reduces to considering directions between (1, 0) T and (1, 1) T . Let p be any prime number and let 1 ≤ q ≤ p − 1 be any even number. Then, there exist positive integers a and b such that bp − aq = 1. Since q is even, b is odd. If a is odd then we can replace a and b by p + a and q + b, respectively. This allows us to assume that p a q b ∈ Γ(2).
Choosing p sufficiently large, it is a consequence of the Perron-Frobenius Theorem that this matrix has an eigenvector whose direction is arbitrarily close to (p, q) T . Finally, by appropriately choosing q, every vector (v 1 , v 2 ) T between (1, 0) T and (1, 1) T can be approximated. The lemma follows.
Proof of Proposition 5.3. It suffices to prove that the Julia set J for the monomial action on C * ×C * is all of C * ×C * . Indeed, suppose there is an open U ⊂ S contained in the Fatou set for the action of Γ on S. Making U smaller, if necessary, we can assume that U is evenly covered by Φ and that V is one of the two connected components of Φ −1 (U ). Because of the semi-conjugacy Φ, the assumption on U would imply that V is in the Fatou set for the monomial action of Γ(2) on C * × C * . On the other hand, the unit torus T 2 ⊂ C * × C * is invariant under the monomial action of Γ(2) with the hyperbolic elements of Γ(2) corresponding to Anosov mappings η M : T 2 → T 2 . Therefore, T 2 ⊂ J. Moreover, for each hyperbolic M ∈ Γ(2) the hyperbolic set T 2 has stable and unstable manifolds under η M , namely: In the remainder of this section, we will focus on points having non-trivial stabilizers in Γ. As mentioned, the discussion below will allow us to prove Proposition 9.8.
Let S(R) = S ∩ R 3 denote the real slice of S. It is well known that S(R) \ S sing consists of one bounded component and three unbounded components (see [2]). Let S(R) 0 denote the closure of the bounded component of S(R) \ S sing . It is straightforward to check that For every M ∈ Γ(2) and any fixed point p ∈ C * × C * of η M , there exist suitable local coordinates (complexified angular coordinates) in which Dη M (p) = M . In particular, the eigenvalues of Dη M (p) and of M coincide. Moreover, by noticing that (η M ) k = η M k for every integer k > 1, there follows that the analogous statement holds for periodic points as well.
Note also that when M ∈ Γ(2) is parabolic, η M may have fixed points away from of the real torus T 2 ⊂ C * × C * . However, by the preceding discussion, the derivative Dη m (p) at such a fixed point will always have eigenvalues equal to ±1. When M ∈ Γ(2) is hyperbolic, any fixed point p of η M is on T 2 and the fixed point is a saddle, with one of the eigenvalues of Dη M (p) having absolute value less than one and the other having absolute value greater than one. Proof. The critical values of Ψ are precisely the singular point of S, which are permuted by elements of Γ. Therefore, Ψ −1 (p) = {q 1 , q 2 } and they will either each be a fixed point for η M or they will form a period two cycle for η M . In either case (Df M (p)) 2 and Dη M (q 2 )Dη M (q 1 ) will be conjugate matrices and hence have the same eigenvalues. Meanwhile Dη M (q 2 )Dη M (q 1 ) and M 2 have the same eigenvalues, so the result follows.
Proposition 5.6. For the Picard parameters, whenever M is a hyperbolic matrix, every fixed point of the corresponding mapping f M not lying in S sing must be a hyperbolic saddle. In addition, these fixed points are all located on S(R) 0 = Φ(T 2 ).
In particular, there is no dense J * 0,0,0,4 ⊂ J 0,0,0,4 consisting of points with hyperbolic stabilizers. Proof. It follows from Lemma 5.5 and from the discussion in the paragraph before this lemma that a hyperbolic fixed point of an arbitrary element in Γ, in fact, must be a fixed point of some mapping f M , where M hyperbolic. Every such fixed point is therefore a hyperbolic saddle and is contained in S(R) 0 . Clearly S(R) 0 is a proper subset of S which, in turn, coincides with J in view of Proposition 5.3. The proposition follows.
Proof of Theorem D. It follows directly from the combination of Propositions 5.1, 5.3, and 5.6.
Recall that, by construction, every mapping f M : S → S is the restriction of a polynomial diffeomorphism of C 3 which will be denoted by F M : C 3 → C 3 . These maps F M leaves invariant all the surfaces of the form S 0,0,0,D , with D ∈ C. From this it follows that if p ∈ S 0,0,0,4 is a fixed point of f M then two of the eigenvalues of the 3 × 3 matrix DF M (p) are the same as those of Df (p) and the third eigenvalue is 1, provided that p is a regular point of S 0,0,0,D . Owing to Lemma 5.5, we conclude that two of the eigenvalues of DF M (p) have the same absolute values as the eigenvalues of M and the remaining eigenvalue is 1.
The following proposition describes the eigenvalues of DF M (p) at singular points p of S. The fact that the eigenvalues of M are squared is essentially the same phenomenon that occurs for the classical one-dimensional Chebeyshev map, and we are grateful to Micha l Misiurewicz for explaining it to us. Proof. The proof relies upon the semiconjugacy Ψ from Proposition 5.2. We apply it to points (u, v) = (e iθ , e iφ ) ∈ T 2 and abuse notation slightly by writing (x, y, z) = Ψ(θ, φ) = (−2 cos θ, −2 cos φ, −2 cos(θ − φ)).
Here we focus on the singular point p = (−2, −2, −2) = Ψ(0, 0). The minor adaptations required for the other singular points essentially amount to some sign modifications in the equations below and thus can safely be left to the reader.
Since M ∈ Γ(2) we have det(M ) = 1. Using this, the characteristic polynomial for M 2 is P M 2 (x) = x 2 − (m 2 11 + m 2 22 + 2m 12 m 21 ) x + 1. Now let N = DF M (p) and write N = {n jk } 1≤j,k≤3 . Next note that F M (S D ) = S D for every D ∈ C which amounts to saying that F M preserves the polynomial Q(x, y, z) = x 2 + y 2 + z 2 + xyz in the sense that P • F M (x, y, z) = P (x, y, z). Also the restriction of F M to S D preserves the volume form associated to Ω. By integrating this volume form over the the surfaces S D in the transverse direction, we see that F M also preserves a volume form on C 3 . Since p is a fixed point of F M , the preceding conditions combine to ensure that det(N ) = 1 and, in addition, that one of the eigenvalues of N equals 1. Hence,the characteristic polynomial of N is P N (x) = x 3 − (n 11 + n 22 + n 33 ) x 2 + (n 11 + n 22 + n 33 ) x − 1 .
This will imply that P N (x) = P M 2 (x)(x − 1) therefore completing the proof of Proposition 5.7.
To begin, consider the x-coordinate of the semi-conjugacy (19): Now it suffices to take the limit as θ goes to 0 to conclude that m 2 11 = n 11 + n 13 . Similarly, setting θ = 0 and doing the analogous computation involving partial derivatives with respect to φ yields m 2 12 = n 12 + n 13 . Finally, the analogous computation with φ = θ lead to (m 11 + m 12 ) 2 = n 11 + n 12 . The three previous equations can be solved for n 11 to find n 11 = m 2 11 + m 11 m 12 .
Combined with the fact that det(M ) = 1, Equations (21) and (22) imply that the condition expressed by (20) holds. The proof of the proposition is completed.
Corollary 5.8. Let (A, B, C, D) be the Picard Parameters (0, 0, 0, 4). Assume that M ∈ Γ(2) is hyperbolic. Denote by f M : S → S the element of Γ associated with M and let F M : C 3 → C 3 be the corresponding extension of f M to C 3 . If p ∈ S is a fixed point of F M then DF M (p) has one eigenvalue of modulus less than one, one eigenvalue equal to one, and one eigenvalue of modulus greater than one.
Existence of Fatou Components and Proof of Theorem E
Before proving Theorem E we will show that the action of the bigger group Γ * A,B,C on C 3 always has non-empty Fatou set F A,B,C . Given parameters (A, B, C), Proposition 6.1 will provide a point p ∈ C 3 and ǫ > 0 so that the ball B ǫ (p) of radius ǫ around p is contained in F A,B,C . In particular, to obtain 4-tuples of parameters (A, B, C, D) ∈ C 4 for which the action of Γ A,B,C,D on S A,B,C,D has non-empty Fatou set F A,B,C,D , it will be enough to select D so that S A,B,C,D ∩ B ǫ (p) = ∅; see Corollary 6.2. Although we focus the application of Proposition 6.1 and of Corollary 6.2 to Theorem E, they apply to many other settings. The idea for the proof of Proposition 6.1 comes from the papers of Bowditch [5], Tan, Wong, and Zhang [54], Maloni, Palesi, and Tan [38], and Hu, Tan, and Zhang [27]. See Section 2.2 for more details.
Proof. Let q ∈ B ǫ (p) and note that this implies that each coordinate of q has modulus larger than 2 + √ r. We will show that if w 1 w 2 . . . w k is any reduced word in s x , s y , and s z then, for each 0 ≤ j < k, each coordinate of w 1 w 2 . . . w j (q) has modulus less than or equal to the modulus of the corresponding coordinate of w 1 w 2 . . . w j+1 (q). Here, the word w 1 w 2 . . . w j is to be interpreted as the identity map provided that j = 0. This claim clearly implies that for every γ ∈ Γ * we have Applying Montel's Theorem to each coordinate implies that the action of Γ * is normal on B ǫ (p) so that the statement follows.
To check the claim, we begin by fixing q = (x, y, z) ∈ B ǫ (p). Consider the involution s x and the points q and s x (q). Clearly the coordinates y and z of these two points coincide. In turn, to show that the modulus of the x coordinate of q is strictly smaller than the modulus of the x coordinate of s x (q), note that where the second inequality follows from the assumption that ǫ ≤ R + 1 − √ 4 R + r + 1. Indeed, this condition can be reformulated as R − ǫ + 1 ≥ √ 4 R + r + 1 which, by taking squares on both sides, leads right away to the inequality in question. Naturally, analogous estimates hold with respect to the coordinates y or z when s x is replaced by s y and s z . Therefore, we have shown that for every point q ∈ B ǫ (p), applying s x , s y , or s z to q strictly increases the modulus of one of the coordinates while leaving the other two coordinates unchanged. Now consider any reduced word w 1 w 2 . . . w k in s x , s y , and s z . Suppose for a contradiction that there is some q ∈ B ǫ (p) and some 0 ≤ j < k so that w 1 . . . w j (q) has some coordinate of modulus strictly larger than the corresponding coordinate of w 1 . . . w j+1 (q). If there are more than one such j where this happens we choose the smallest one, which satisfies j ≥ 1 by virtue of the calculation in the previous paragraph. Note also that taking j to be minimal implies that each coordinate of w 1 . . . w j (q) has modulus greater than or equal to the minimal modulus of a coordinate of q which, in turn, exceeds 2 + √ r = 2 + max{|A|, |B|, |C|}.
In particular, the preceding ensures that min{|x ′ |, |y ′ |, |z ′ |} ≥ 2 + √ r. On the other hand, since the word w is reduced and all generators s x , s y , s z are involutions, we must have w j = w j+1 . Without loss of generality, we can then suppose w j = s x and w j+1 = s y . This yields Our assumption on j implies |x| ≤ |x ′ | and |y ′ | > |y ′′ |. Therefore, We now split the discussion in two cases. Assume first that |x ′ | ≥ |y ′ |. Then, the second inequality from (25) gives In turn, moving |B| to the right side, dividing by |x ′ |, and recalling that |x ′ | ≥ 2 + √ r leads to This is impossible since min{|x ′ |, |y ′ |, |z ′ |} ≥ 2 + √ r. If we consider now the case where |y ′ | > |x ′ |, we just need to use the first inequality from (25) to similarly show that Thus, in any event, we obtain a contradiction that proves our initial claim. As already pointed out, the ball B ǫ (p) must therefore lie in the Fatou set of the action of Γ * A,B,C on C 3 . The proof of Theorem E will be a quick application of Corollary 6.2 combined with the following elementary lemma whose proof we leave to the reader. Let M be a (possibly open) connected complex manifold and consider a group G of holomorphic diffeomorphisms of M . The group G is said to be locally non-discrete on an open U ⊂ M if there is a sequence of maps {f n } ∞ n=0 ∈ G satisfying the following conditions (see for example [48]): (1) For every n, f n is different from the identity.
(2) The sequence of maps f n converges uniformly to the identity on compact subsets of U .
If no such sequence of maps {f n } ∞ n=0 ∈ G exists then G is said to be locally discrete on U . The reader will note that a group G may be locally non-discrete on one open set U and locally discrete on a disjoint open set V .
It is convenient to begin our discussion with Proposition 7.1 below. This proposition constitutes a simple general result of which specific variants appear in [48, p. 9-10] and in [37,Sec. 3] while the main idea dates back to Ghys [22]. Both Proposition 7.1 and Lemma 7. (1) and (2) to make sense and Condition (1) has been modified because of the possibility that U not be connected.
Condition (0) is required for Conditions
Consider a ball B ǫ (0) ⊂ C n of radius ǫ > 0 around the origin of C n . Assume we are given local holomorphic diffeomorphisms F 1 , F 2 : B ǫ (0) → C n and denote by G the pseudogroup of maps from B ǫ (0) to C n generated by F 1 , F 2 . Naturally the inverses of F 1 , F 2 are respectively denoted by F −1 Note that the construction of these elements is so far purely formal in the sense that the domain of definition (contained in B ǫ (0)) of diffeomorphisms in S(n) viewed as elements of G may be empty. Nonetheless, we have: then the following hold: (1) For every n and every γ ∈ S(n), the domain of definition of γ as element in G contains the ball B ǫ/2 (0) ⊂ C n of radius ǫ/2 around the origin.
Note that, in general, the above proposition falls short of implying that the pseudogroup G is locally non-discrete since the sets S(n) may degenerate so as to only contain the identity map.
As mentioned, variants of Proposition 7.1 can be found in the literature and, to the best of our knowledge, the idea goes back to Ghys in [22]. Otherwise, its proof is nearly identical to that from [48, p. 9-10], so we provide only a sketch. If max sup for some K > 0, then for any τ > 0 satisfying 4K + τ < ǫ the commutator [f 1 , f 2 ] is defined on the ball of radius ǫ − 4K − τ and satisfies Starting with S(0), we choose τ = τ 0 = K = K 0 and ǫ 1 = ǫ − 8K so that the preceding yields for every γ ∈ S(1). Now inductively setting for every γ ∈ S(n), where The proposition then follows by choosing K sufficiently small that ǫ n ≥ ǫ/2 for all n.
The following simple lemma complements Proposition 7.1.
Lemma 7.2. Let F 1 , F 2 : B ǫ (0) → C n be as above with F 1 (0) = F 2 (0) = 0. Assume that their derivatives at the origin satisfy where τ > 0 is some uniform (universal) to be determined later (the norm used here is the standard norm on linear operators on C n ). Then, there is some 0 < δ < ǫ such that F 1 and F 2 satisfy the hypotheses of Proposition 7.1 on B δ (0).
Proof. There follows from the proof of Proposition 7.1 that the relation between ǫ and K = K(ǫ) given by (28) is linear. Thus for ǫ = 1, K = K(1) becomes a uniform (universal) constant. We then choose τ = K/2 so that τ is also universal.
In view of the above remark, we proceed as follows. Fix δ < ǫ and consider the homothety Λ δ : C n → C n sending (x 1 , . . . , x n ) to (δx 1 , . . . , δx n ). Set F j,δ = Λ −1 δ • F j • Λ δ and note that F 1 , F 2 satisfy the conditions of Proposition 7.1 on the ball of radius δ if and only if we have (30) max sup We will now check that Estimate (30) is always satisfied provided that δ is small enough. Clearly, it suffices to consider the case of F 1,δ . Owing to Taylor formula, we have provided that δ is small enough. The result is proved. Proposition 7.3. Suppose that (A 0 , B 0 , C 0 ) are parameters for which there are two non-commuting elements F 1 , F 2 ∈ Γ sharing a common fixed point p ∈ C 3 . Assume also that the derivatives of F 1 , F 2 at p satisfy inequality (29). Then the following holds: (1) There exists r > 0 such that for all parameters (A, B, C) sufficiently close to (A 0 , B 0 , C 0 ) the group Γ acting on C 3 is locally non-discrete on the open ball B r (p) ⊂ C 3 .
(2) Let D 0 be chosen so that p ∈ S A 0 ,B Recall from the introduction that El-Huiti's theorem implies that Γ is the free group on two generators (one can choose any two of the three mappings g x , g y , and g z as generators). Therefore, two elements γ 1 , γ 2 ∈ Γ commute if and only if there exists a ∈ Γ so that γ 1 and Γ 2 are both powers of a. This is an immediate consequence of the Nielsen-Schreier Theorem which states that any subgroup of a free group is free.
The proof of Proposition 7.3 will require a simple algebraic lemma: is a non-trivial reduced word in a, b, a −1 , and b −1 . Since a and b do not commute this word does not reduce to the identity in F n .
Proof of Proposition 7.3. Lemma 7.2 implies the existence of some δ > 0 such that for parameters (A 0 , B 0 , C 0 ) the mappings F 1 and F 2 satisfy the hypotheses of Proposition 7.1 on B δ (p). Moreover the conditions of Proposition 7.1 are open in the C 0 topology (on B δ (p)). This implies that for all parameters (A, B, C) sufficiently close to (A 0 , B 0 , C 0 ) the mappings F 1 and F 2 continue to satisfy the hypotheses of Proposition 7.1 on B δ (p). Thus, for these parameters (A, B, C), the elements in the iterated commutators S(n) converge uniformly to the identity on the ball B δ/2 (0) ⊂ C 3 . Since F 1 and F 2 do not commute, Lemma 7.4 can inductively be applied to ensure that each set S(n) contains at least two non-commuting elements. Therefore, for every n ≥ 0, there are elements in S(n) that are different from the identity which, in turn, proves that the pseudogroup generated by F 1 and F 2 is locally non-discrete on B r (p). Statement (2) then follows immediately because elements of Γ different to the identity cannot coincide with the identity when restricted to any surface S A,B,C,D . In other words, for every choice of parameters (A, B, C, D), any non-trivial reduced word in g x and g y (or in any two of the three mappings g x , g y , and g z ) induces a mapping of S A,B,C,D that is different from the identity.
If we let h x = g 2 x , h y = g 2 y , and h z = g 2 z , then each of these maps is tangent to the identity at the origin. Notice that when D = 0 the surface S 0,0,0,D passes through (0, 0, 0). Therefore, applying Proposition 7.3 to the non-commuting pair of elements h x and h y yields: It is immediate to check that the three above singular points satisfy these equations.
Meanwhile, we must check that these three points lie on the surface S a . By symmetry, it suffices to check for p 1 = (x 1 , y 1 , z 1 ) and, in this case, we have so that p 1 ∈ S a . Note that there are other common fixed points for s x , s y , and s z but they are located away from S a . Proposition 7.7. For every fixed a ∈ (−2, 2) the group Γ a acting on S a has locally non-discrete dynamics in some neighborhood of each of the singular points p 1 , p 2 and p 3 .
Moreover, the Dubrovin-Mazzocco parameters at this fixed value of a yield an interior point of N D. More precisely, given parameters (A, B, C, D) sufficiently close to the Dubrovin-Mazzocco parameters in question, the group Γ is locally non-discrete on the intersection of S A,B,C,D with B r (p 1 ) ∪ B r (p 2 ) ∪ B r (p 3 ) for some r > 0.
Note that the parameter a = −2, which we do not consider a Dubrovin Mazzocco parameter, corresponds to the Picard Parameters (A, B, C, D) = (0, 0, 0, 4) for which Γ is locally discrete on all of the corresponding surface, as was shown in Section 5.
Proof. The entire system is symmetric under permutations of the coordinates (x, y, z) so that it suffices to consider the case of p 1 . Note that p 1 is fixed by every element in Γ (in fact, by every element of the group generated by s x , s y , and s z ).
Let us first show that a suitable iterate of g x = s z • s y is close to the identity on some suitable neighborhood of p 1 . A direct calculation yields that the eigenvalues of D p 1 g x are For a ∈ (−2, 0) ∪ (0, 2) we see that λ 1 and λ 2 form a complex conjugate pair of eigenvalues, each of which has modulus one. In this case, D p 1 g x is conjugate to a rotation in suitable coordinates. In particular, for any τ > 0 there is k sufficiently large so that To find a second mapping satisfying the hypotheses of Lemma 7.2, it suffices to consider the map h obtained by conjugating g k x by, say, g y . Since p 1 is fixed by g y as well, there follows that D p 1 h is conjugate to D p 1 g k x and that the conjugating matrix does not depend on k (it is simply the matrix given by Dg y (p 1 )). Therefore, up to making k larger if needed, we can assume that both g k x and h satisfy the hypotheses of Lemma 7.2. Since g k x and h do not commute in Γ, the result then follows from Proposition 7.3.
In the special case that a = 0 another direct calculation yields that (D p 1 g x ) 2 = Id, in which case g 2 x is tangent to the identity at p 1 . Letting h = g y g 2 x g −1 y we obtain a second mapping tangent to the identity at p 1 . Since, as previously seen, these two maps do not commute, the desired result in this special case follows again from Proposition 7.3.
Dynamics near infinity
In this section we will collect a few important, if slightly technical, results on the dynamics of the group Γ as well as the dynamics of (individual) hyperbolic maps near ∆ ∞ . The corresponding results, especially Proposition 8.4 and Lemma 8.6, will play major roles in the proofs of Theorems H and K to be supplied in the forthcoming sections.
Let us begin with a general review of the behavior of a given map in Γ = Γ A,B,C,D near infinity. Up to compactifying C 3 into the projective space, we begin by recalling that the closure of any surface S A,B,C,D intersects the hyperplane at infinity Π ∞ ⊂ CP 3 in a triangle ∆ ∞ . In homogeneous coordinates, this triangle is given by As previously seen, for every choice of the parameters (A, B, C, D), the surface S A,B,C,D is smooth on a neighborhood of ∆ ∞ . Furthermore, S A,B,C,D is tangent to the plane at infinity exactly at the vertices in V ∞ and, hence, it has transverse intersection with the plane at infinity elsewhere in ∆ ∞ . In the sequel, to abridge notation, the parameters A, B, C, D will be dropped whenever there is no possibility of misunderstanding. Thus, we will sometimes write Γ for Γ A,B,C,D , S for S A,B,C,D , and S for the closure of S.
Because S is tangent to the plane at infinity at the vertices of ∆ ∞ , near v 1 we can use the affine coordinates (Y /X, Z/X, W/X) on CP 3 and express the surface S so that W/X becomes a holomorphic function of (Y /X, Z/X). In other words, S is locally given as the graph of a holomorphic function in the variables (Y /X, Z/X) on a neighborhood of v 1 . This local representation will be referred to as the "standard coordinates" on S in a neighborhood of v 1 . The analogous construction leads to "standard coordinates" on S near v 2 and near v 3 , respectively.
Inside the plane at infinity, consider a neighborhood W 1 of v 1 where S is the graph of some holomorphic function h 1 as indicated above. Neighborhoods W 2 and W 3 respectively of v 2 and v 3 are similarly defined along with the corresponding holomorphic functions h 2 and h 3 . We state below some immediate consequences of what precedes in terms "variation with parameters". Lemma 8.1. Consider parameters A 0 , B 0 , C 0 and D 0 and apply the above construction to the surface S A 0 ,B 0 ,C 0 ,D 0 . Up to reducing the neighborhoods W i , i = 1, 2, 3, there exists a neighborhood W ⊂ C 4 of (A 0 , B 0 , C 0 , D 0 ) such that all of the following holds: (1) For every (A, B, C, D) ∈ W, the surface S A,B,C,D is (locally) the graph of a function h i defined on W i . Furthermore, for i = 1, 2, 3, the functions h i vary continuously with the parameters.
(2) For every (A, B, C, D) ∈ W, the intersection of the surface S A,B,C,D with the plane at infinity over the set ∆ ∞ \ (W 1 ∪ W 2 ∪ W 3 ) is uniformly transverse. Furthermore, the slopes vary continuously with the point and with the parameters.
We will also need some facts about birational extensions to S of elements in Γ and, more generally, in Γ * . Recalling that Γ * is generated by the involutions s x , s y , and s z , an element γ in Γ * is said to be cyclically reduced if its reduced spelling in the "letters" s x , s y , and s z is not conjugate in Γ * to another element in Γ * having strictly smaller length when spelled in the same "letters".
We also recall that a meromorphic self-map from a complex surface to itself is said to be algebraically stable if it does not contract a hypersurface to its indeterminate set, cf. [21,15]. In the present case where elements of Γ act on the surface S, to be algebraically stable amounts to saying that γ does not contradict any of the sides of ∆ ∞ to an indeterminacy point which, in turn, necessarily lies in ∆ ∞ as well.
With this terminology, the following definition/proposition and remark summarize [ Definition/Proposition 8.2. For any parameters A, B, C, D and any γ ∈ Γ we have (i) γ is said to be hyperbolic if and only if it conjugate to a cyclically reduced word in s x , s y , and s z that contains all three mappings. (ii) A hyperbolic map γ ∈ Γ possesses a single indetermination point which coincides with a vertex of ∆ ∞ and will be denoted by Ind(γ).
Attr(γ). The vertices Ind(γ) and Attr(γ) may or may not coincide. In particular, the map γ is algebraically stable if and only if Ind(γ) = Attr(γ). (iv) Alternatively, γ : S S is algebraically stable if and only if it is a cyclically reduced composition of s x , s y , and s z of length at least two.
(v) If γ is algebraically stable and hyperbolic then γ is holomorphic around Attr(γ) and, in fact, Attr(γ) is a superattracting fixed point of γ. Moreover, the roles of Ind(γ) and Attr(γ) are interchanged if we pass from γ to γ −1 , i.e., Attr(γ) = Ind(γ −1 ) and Ind(γ) = Attr(γ −1 ). (vi) An element γ is said to be parabolic if it is conjugate in γ to one of the maps g x , g y , or g z .
Every element of Γ different from the identity either is hyperbolic or is parabolic. Note that the properties described in Definition/Proposition 8.2 are independent of the parameters A, B, C, D. They depend only on the spelling of γ as an element in the abstract group generated by three letters a, b, and c with the relation abc = id (up to the substitutions a = g x , b = g y , and c = g z ). In particular, the points Ind(γ) and Ind(γ −1 ) = Attr(γ) are independent of the parameters and this plays an important role in the statement of the following proposition.
Proposition 8.4. Consider an hyperbolic element γ ∈ Γ and, for a choice of parameters A 0 , B 0 , C 0 , and D 0 , let f A 0 ,B 0 ,C 0 ,D 0 be the resulting birational map of the (compact) surface S A 0 ,B 0 ,C 0 ,D 0 . Then there exist a neighborhood U ∆ of ∆ ∞ ⊂ CP 3 and a neighborhood U 0 ⊂ C 4 of (A 0 , B 0 , C 0 , D 0 ) ∈ C 4 such that the following holds: • Proof. It suffices to prove the proposition for an algebraically stable hyperbolic γ since, every hyperbolic element in Γ is conjugate in Γ to an algebraically stable one and all elements in Γ preserve the infinity. Let then f A 0 ,B 0 ,C 0 ,D 0 be an algebraically stable hyperbolic map as indicated above. The statement amounts to showing the existence of a neighborhood V 0 ⊂ S A 0 ,B 0 ,C 0 ,D 0 of ∆ ∞ ⊂ S A 0 ,B 0 ,C 0 ,D 0 satisfying the two conditions below: (1) f A 0 ,B 0 ,C 0 ,D 0 has no fixed point (actually periodic point) in V 0 \ ∆ ∞ .
(2) The neighborhood V 0 can be chosen to vary continuously with the parameters A, B, C, and D.
To prove the first assertion, we consider the points Attr(f A 0 ,B 0 ,C 0 ,D 0 ) and Ind(f A 0 ,B 0 ,C 0 ,D 0 ) in ∆ ∞ . As mentioned, these points are distinct and do not depend on the choice of the parameters. To abridge notation, we then set P = Attr(f A 0 ,B 0 ,C 0 ,D 0 ) and Q = Ind(f A 0 ,B 0 ,C 0 ,D 0 ). We also recall that P is a super-attracting fixed point of f It remains to prove that the above neighborhood can be assumed to vary continuously with the parameters. For this, we first consider the (local) description of the surfaces S A,B,C,D provided by Lemma 8.1. The lemma in question shows that the surface S A,B,C,D converge towards S A 0 ,B 0 ,C 0 ,D 0 as (A, B, C, D) → (A 0 , B 0 , C 0 , D 0 ) on a neighborhood of ∆ ∞ ⊂ CP 3 . In fact, on a neighborhood of the points P and Q, this follows from the local structure of these surfaces as graphs of holomorphic functions. Conversely, away from these neighborhoods of P and Q, the statement follows from the (uniform) transverse intersection of these surfaces and the plane at infinity of CP 3 . In particular the maps f A,B,C,D converge uniformly to f A 0 ,B 0 ,C 0 ,D 0 on a neighborhood of ∆ ∞ . Thus, the statement is reduced to show the existence of ǫ > 0 such that for all parameters (A, B, C, D) sufficiently close to (A 0 , B 0 , C 0 , D 0 ), the ball of radius ǫ around P still is contained in the basin of attraction of P with respect to f A,B,C,D . Here, the same argument applies to Q and f −1 A,B,C,D and the notion of "ball" is relative to some auxiliary metric, for example, the Euclidean metric in the "standard coordinates" of Lemma 8.1.
The last claim can be proved as follows. The germ of f A,B,C,D is a rigid, reducible, superattracting germ in the sense of [19] (cf. [6]). It actually falls in the "class 6" of the classification provided in [19] and hence it is conjugate to a monomial map, owing to results of Dloussky and Favre. Since the monomial map does not depend on the parameters, the claim follows from checking directly in the argument of [19] that the conjugating map depends continuously on the initial map.
As a consequence of Proposition 8.4, we also obtain the following result (cf. Lemma 16 from [32]). Proof. As a composition of the algebraic maps g x , g y , and g z , f A,B,C,D is itself an algebraic map from C 3 to C 3 . Thus the set of its fixed points is an algebraic set of C 3 so that its intersection with S A,B,C,D is an algebraic subset of S A,B,C,D . If the latter set is not constituted by isolated points, then it must contain a curve C (as already seen it cannot coincide with all of S A,B,C,D ). However, by the maximum principle, the C must accumulate on ∆ ∞ and this contradicts Proposition 8.4.
The remainder of this section will be devoted to the proof of Lemma 8.6 below which is crucial for the understanding of the structure of unbounded Fatou components as it will be seen later on.
We resume the terminology used at the beginning of the section. Thus (X : Y : Z : W ) are homogeneous coordinates on CP 3 . For any A, B, C, D, the surface S = S A,B,C,D is tangent to the hyperplane at infinity at each of the vertices v 1 , v 2 , and v 3 of ∆ ∞ . Again, near, say v 1 , we can use the affine coordinates (Y /X, Z/X, W/X) on CP 3 and express the surface so that W/X is a holomorphic function of (Y /X, Z/X). Similar descriptions apply to the other two vertices, and the resulting coordinates were called "standard coordinates" around the vertices in question.
Next, let U ∞ (i) be an open neighborhood of v i in S so that U ∞ (i) is contained in the graph of h i : W i → C, i = 1, 2, 3. Without loss of generality, we assume that the three neighborhoods U ∞ (i), i = 1, 2, 3, are pairwise disjoint. If p ∈ U ∞ (i), we denote by dist(p, v i ) the Euclidean distance between p and v i in the standard coordinate chart on U ∞ (i). Let For any p ∈ U ∞ , we define dist(p, V ∞ ) to equal dist(p, v i ) where U ∞ (i) is the unique one of the three neighborhoods containing p.
Let S(0) be a set consisting of six hyperbolic elements γ i,j ∈ Γ, i, j ∈ {1, 2, 3}, i = j. Like in Proposition 7.1, for every natural number n we will consider the inductively defined sets of iterated commutators S(n). For every n the set S(n + 1) contains every possible commutator of any two distinct elements of S(n). Lemma 8.6. Given parameters A, B, C, D, assume there are six hyperbolic elements γ i,j ∈ Γ as above such that for every pair i = j ∈ {1, 2, 3} we have Ind(γ i,j ) = v i and Attr(γ i,j ) = v j . Let S(0) be the set consisting of the six elements γ i,j and let {S(n)} be the corresponding sequence of inductively defined subsets of Γ. Then, up to reducing the neighborhood U ∞ ⊂ S A,B,C,D , for any point q ∈ U ∞ , there exists a constant 0 < λ < 1 and a sequence {η n } ∞ n=0 ⊂ Γ satisfying the two conditions below: (i) η n ∈ S(n) for every n ≥ 0, and (ii) dist(η n (q), V ∞ ) ≤ λ 4 n for every n ≥ 0.
Proof. Note that we might have asked the elements γ i,j of the set S(0) to satisfy γ i,j = γ −1 j,i though this is not necessary. Also, in view of Definition/Proposition 8.2, all elements γ i,j are algebraically stable and γ i,j is holomorphic around Attr(γ i,j ) = v j .
The proposition will be proved by induction. In fact, we will prove a rather stronger statement. Namely, there exists λ, 0 < λ < 1, such that for each integer n ≥ 0, we have: If n is even then for every pair of distinct i, j ∈ {1, 2, 3} there exists some γ If n is odd then for each i ∈ {1, 2, 3} there exists some τ i ) ±1 (U ∞ \U ∞ (i)) ⊂ U ∞ (i), and for any q ∈ U ∞ \ U ∞ (i) we have dist (τ The base of the induction is n = 0, in which case for each pair i = j we let γ (0) i,j = γ i,j be the corresponding element of S(0). Fix then a pair i = j and let k ∈ {1, 2, 3} be such that k = i and k = j. Consider standard local coordinates (u 1 , u 2 ) around v k and let (v 1 , v 2 ) stand for standard local coordinates in a neighborhood of v j . By hypothesis we have Ind(γ i,j ) = v i and Attr(γ i,j ) = v j . Therefore, item (iii) of Definition/Proposition 8.2 gives that γ i,j (∆ ∞ \ {v i }) = v j . Hence, if γ i,j is expressed in local coordinates under the form (v 1 , v 2 ) = γ i,j (u 1 , u 2 ), both coordinates of γ i,j (u 1 , u 2 ) will vanish along both axes {u 1 = 0} and {u 2 = 0}. Similarly, if we express γ i,j from the (v 1 , v 2 ) coordinates to themselves, then both coordinates of γ i,j (v 1 , v 2 ) vanish along both axes {v 1 = 0} and {v 2 = 0}. This implies that for any 0 < λ < 1, we can choose the neighborhoods U ∞ (k) and U ∞ (j) sufficiently small so as to ensure that for any q ∈ U ∞ (k) ∪ U ∞ (j) the estimate holds. After choosing a sufficiently small neighborhood U ∞ (i) of v i and perhaps making U ∞ (k) smaller, the same reasoning applies to show that for any q ∈ U ∞ (k) ∪ U ∞ (i) we have Repeating for all six distinct pairs i = j we obtain sufficiently small neighborhoods U ∞ (1), U ∞ (2), and U ∞ (3) such that (34) and (35) hold. If we then let U ∞ (1), U ∞ (2), and U ∞ (3) be round balls of equal sufficiently small radius in the standard local coordinates, both estimates (34) and (35) will continue to hold. In addition, for all six distinct pairs i = j we have . Therefore we can assume that (E1) and (E2) hold when n = 0. For the remainder of the proof we keep the neighborhood U ∞ = U ∞ (1) ∪ U ∞ (2) ∪ U ∞ (3) fixed and inductively prove that (E1) and (E2) hold on U ∞ for every even n and that (O) holds on U ∞ for every odd n.
Suppose now that n is even and that the collection of six elements γ (n) i,j ∈ S(n) exist and satisfy (E1) and (E2). For each i ∈ {1, 2, 3}, we will prove the existence of an element τ (n+1) i ∈ S(n+1) satisfying (O).
Fix then i ∈ {1, 2, 3} and let j, k ∈ {1, 2, 3} \ {i} be the other two indices. We define i,k . Using (E1) and (E2) we can see that the above composition is holomorphic on all of V ∞ \ V ∞ (i) and that it maps V ∞ \ V ∞ (i) into V ∞ (i). Moreover, for any q ∈ V ∞ \ V ∞ (i) we have that i,k (q) ∈ V ∞ (i). Again using (E1) and (E2) we have the each of the four mappings in the commutator used to define τ (n+1) i contracts distance to V ∞ by a factor of λ 4 n and hence dist τ Therefore the same proof as in the previous paragraph applies to (τ (n+1) i ) −1 after switching j and k. We conclude that (O) holds for n + 1.
Suppose now that n is odd and that the collection of three elements τ (n) i ∈ S(n) exist and satisfy (O). We will prove that all six elements γ (n+1) i,j ∈ S(n + 1) satisfying (E1) and (E2) exist.
For any distinct i, j ∈ {1, 2, 3} let k ∈ {1, 2, 3} \ {i, j} be the remaining element. Now define Using (O) we can see that the above composition of mappings is holomorphic on U ∞ \ U ∞ (i) and that it maps U ∞ \ U ∞ (i) into U ∞ (j). Again using (O), each of these four mappings contracts distance to V ∞ by a factor of λ 4 n and hence dist γ We conclude that (E1) holds for n + 1.
To see that (E2) holds for n + 1, note that Therefore the same proof as in the previous paragraph applies to (γ (n+1) i,j ) −1 after switching i and j. We conclude that (E2) holds for n + 1.
Therefore, we conclude that statements (E1) and (E2) hold for every even n ≥ 0 and that (O) holds for every odd n ≥ 0. Since the Fatou set is invariant, one can consider the stabilizer Γ V ≤ Γ of V , which consists of those elements of Γ that map V to V . The purpose of this section is to establish several general properties of Fatou components. By combining these properties with the previous material, the proofs of Theorems H and K will quickly be derived in the next section. We begin with Proposition 9.1 below which ensures the hyperbolic character of these components (cf. Lemma 4.5). (12), which is a subset of the Julia set J A,B,C,D . Hence V does not intersect G and we can therefore consider the inclusion Clearly the image of ι is contained in a product of hyperbolic Riemann surfaces which is naturally a Kobayashi hyperbolic domain. The fact that holomorphic mappings do not increase the Kobayashi psuedometric then implies that V is also Kobayashi hyperbolic as well; see [33,Proposition 3.2
.2].
Let then V be a given component of F A,B,C,D and let Aut(V ) denote its group of holomorphic automorphisms. By building on the general theory of topological transformation groups of Gleason, Montgomery, and Zippin [41], Cartan was able to show that the automorphism group of a bounded domain in C n is a finite-dimensional real Lie group. In turn, Kobayashi [33] was able to extend Cartan's theorem to general (Kobayashi) hyperbolic manifolds. Owing to Proposition 9.1, Corollary 9.2 below summarizes these results in the case of a Fatou component.
Recall that the action ϕ : G × M → M of a group G on a manifold M is said to be proper if the preimage by ϕ of any compact set of M is again compact in G × M . (2) For every point p ∈ V , the stabilizer G p of p in G is such that its local action around p is conjugate to the local (linear) action of a closed subgroup of SU (2) on a neighborhood of (0, 0) ∈ C 2 .
Proof. As a closed subgroup of a Lie Group, G is itself a real Lie Group. Moreover, since Γ is locally non-discrete on the open set U ⊂ V , modulo reducing U , there are elements {γ n } ∞ n=1 ⊂ Γ that converge uniformly to the identity on U . In particular, for n large enough, we have γ n (V ) ∩ V = ∅ and thus γ n (V ) = V since F is invariant under Γ. Hence, up to dropping finitely many elements in the sequence in question, we can assume without loss of generality that {γ n } ∞ n=1 ⊂ Γ V for every n ∈ N. Next we have: Claim. The sequence {γ n } ∞ n=1 actually converges to the identity uniformly on compact subsets of V , i.e. as elements of Aut(V ).
Proof of the claim. Consider a relatively compact open set U ′ ⊂ V with U ⊂ U ′ . The claim amounts to checking that {γ n } ∞ n=1 converges uniformly to the identity in U ′ . If this were not the case then, up to passing to a subsequence, there would exist ε > 0 such that Since {γ n } ∞ n=1 is contained in a normal family on V , we can extract a limit map γ ∞ defined on U ′ and thus satisfying sup x∈U ′ γ ∞ (x) − x ≥ ε so that γ ∞ does not coincide with the identity on U ′ . However, γ ∞ must coincide with the identity on U ⊂ U ′ since {γ n } ∞ n=1 converges uniformly to the identity on U . The resulting contradiction proves our claim.
Since {γ n } ∞ n=1 converges to the identity on compact subsets of V , it follows that the elements of Aut(V ) obtained by restricting them to V actually converges to the identity as elements of Aut(V ) equipped with its Lie group structure, cf. Corollary 9.2. Thus Γ V is a non-discrete subgroup of Aut(V ) and hence its closure must be a Lie group with strictly positive dimension.
It remains to check the second assertion. For this, recall that Γ preserves the real volume form associated with the holomorphic volume form Ω given in (9). This implies that the group of derivatives of elements of G p is a subgroup of SL(2, C). On the other hand, Corollary 9.2, Part (3), informs us that G p must be compact. Since SU (2) is a maximal compact subgroup [4] of SL(2, C), it follows from the classical Bochner Linearization Theorem that the local action of G p around p ∈ V is conjugate to the (local) linear action of a closed subgroup of SU (2) on a neighborhood of the origin.
The possibility of having a point p ∈ V whose stabilizer G p is conjugate to all of SU (2) is a challenge for us as it raises quite a few technical issues. To avoid get involved in a much longer argument and keep us focused on the situations of primary interest, we will work only with the following set of parameters: : every fixed point of every element of Γ A,B,C,D \ {id} is in J A,B,C,D }. In Propositions 9.8 and 9.10, it will be seen that the set C 4 good is "quite large" and contains several parameters of interest. In particular, it will be shown that C 4 \ C 4 good is at worst a countable union of (proper) real-algebraic subsets of C 4 and, in particular, it has null Lebesgue measure.
First, however, we have a simple and well-known lemma. Proof. Since f is an isometry of the Kobayashi metric, the statement would be immediate if the Kobayashi metric were a Riemannian metric which, however, is not always the case. To overcome this difficulty, we locally replace the Kobayashi metric by the Bergman one as follows. For small r > 0, let B r (p) denote the ball of radius r with respect to the Kobayashi metric. Since a holomorphic map cannot increase the Kobayashi distance, there follows that f (B r (p)) ⊂ B r (p). The analogous argument applies to f −1 allows us to conclude that f induces an automorphism of B r (p). Now, if r > 0 is small enough, then B r (p) can be identified with a bounded domain in some space C n so that the Bergman metric is well defined. Thus f induces an isometry of the resulting Riemannian metric on B r (p) and hence coincides locally with the identity. The lemma then follows. (2) For any p ∈ V , the stabilizer G p of p in G coincides with the identity, i.e. G p = {id}.
Proof. Beyond the proof of Proposition 9.3, it remains to show Claim (2). Suppose for contradiction the existence of a point p ∈ V for which there is g ∈ G, g = id, satisfying g(p) = p. Note that g may lie in Γ V \ Γ V so that the statement does not follow immediately from the definition of the parameter set C 4 good . Since g = id, Lemma 9.4 implies that Dg(p) is not the identity either. On the other hand, Dg(p) is conjugate to a matrix in SU (2) so that the preceding discussion shows that Dg(p) − id is, in fact, invertible.
On the other hand, g lies in G = Γ V so that there are elements γ n ∈ Γ with γ n → g locally uniformly on V . Moreover, since the functions are holomorphic, this implies C ∞ convergence on compact subsets of V . Since Dg(p) − id is invertible, it follows from the implicit function theorem (for Banach spaces) that for sufficiently large n the mappings γ n have fixed points p n converging to p. Therefore Dγ n (p n ) → Dg(p) = id, implying that γ n = id for sufficiently large n. Thus, we have found non-trivial γ n ∈ Γ having fixed points in the Fatou component V , contradicting the choice of parameters (A, B, C, D) ∈ C 4 good . We conclude that G p = {id}. Recall from Definition/Proposition 8.2 that, bar the identity, all elements of Γ are split in hyperbolic maps and parabolic maps. Furthermore an element of Γ is parabolic if and only if it is conjugate to a non-trivial power of the generators g x , g y or g z . Proof. Since parabolic maps are conjugate to a non-trivial power of one of the generators g x , g y , or g z , it suffices to prove the statement for a non-trivial power of, say, g x . It follows from Proposition 4.1 and Lemma 3.2 that for all but finitely many values of x 0 ∈ C \ [−2, 2] the action of g x on the fiber S x 0 is loxodromic, with two distinct fixed points at infinity. Consider an iterate g ℓ x for some ℓ = 0. Any point on any such S x 0 will have orbit under g ℓ x that tends to infinity. These points form an open dense subset of S A,B,C,D , implying that any point having bounded orbit under g ℓ x (and hence any fixed point of g ℓ x ) must be in the Julia set. Now we will need a significantly more elaborate result. is not a hyperbolic saddle. In our setting, Condition (4) is equivalent to requiring that tr(Df A,B,C,D (x, y, z)) ∈ [−2, 2]. Therefore, H is a semi-algebraic subset of C 7 , i.e. it is a set given by finitely many polynomial equations and polynomial inequalities with real coefficients.
Notice that H = pr (H), where pr(A, B, C, D, x, y, z) = (A, B, C, D). It follows from the Tarski-Seidenberg Theorem [3, Theorem 2.2.1] that H is also semi-algebraic. By definition, the dimension of a semi-algebraic set is the (real) dimension of the (real) Zariski closure of the set. Therefore, if we prove that dim(H) ≤ 7 it will follow that H is contained in a finite union of real-algebraic hypersurfaces of C 4 , which is sufficient for our purposes.
The Cylindrical Algebraic Decomposition Theorem [3, Theorem 2.3.6] asserts that a semialgebraic set can be decomposed into finitely many sets, each of which is homeomorphic to [0, 1] d i for some d i . Moreover, the dimension of the set (in the sense of the previous paragraph) equals the maximum of the d i , see [3,Section 2.8]. In particular, if dim(H) = 8, then H would have non-empty interior. We will prove that this is not the case.
First recall that the fixed points of f A,B,C,D are all isolated (see Corollary 8.5 or Lemma 16 in [32]). Let us first prove the following: Claim.
There is an open U ⊂ C 4 \ (H ∪ NS).
Proof of the claim. Consider a sequence of parameters (A n , B n , C n , Z n ) ∈ C 4 \ NS converging to the Picard Parameters (0, 0, 0, 4). Suppose for contradiction that for every n the mapping f An,Bn,Cn,Dn has a fixed point p n ∈ S An,Bn,Cn,Dn that is not a hyperbolic saddle. Then, since f An,Bn,Cn,Dn preserves the volume form ω, both eigenvalues of Df An,Bn,Cn,Dn (p n ) have modulus equal to 1.
For sufficiently large n the fixed points p n remain away from some fixed neighborhood of ∆ ∞ (Proposition 8.4). Therefore, we can extract a subsequence so that p n k converges to some point p ∞ ∈ S 0,0,0,4 . Since f A,B,C,D is continuous and depends continuously on the parameters we have that p ∞ is a fixed point of f 0,0,0,4 . For any (A, B, C, D) ∈ C 4 let F A,B,C denote the extension of f A,B,C,D as an automorphism of C 3 . The points p n k and p ∞ are also fixed points for F An k ,Bn k ,Cn k and F 0,0,0 , respectively. For each k all three of the eigenvalues of DF An k ,Bn k ,Cn k (p n K ) have modulus equal to 1, with the third eigenvalue corresponding to a direction transverse to S An,Bn,Cn,Dn . Since the eigenvalues of a matrix depend continuously on its entries and since the derivative of DF A,B,C (q) depends continuously on the parameters (A, B, C) and on the point q, we conclude that each eigenvalue of DF 0,0,0 (p ∞ ) has modulus equal to 1. Since p ∞ ∈ S 0,0,0,4 this contradicts Corollary 5.8.
We conclude that there is some n such that every fixed point of f An,Bn,Cn,Dn is a hyperbolic saddle. Since we chose the parameters (A n , B n , C n , D n ) ∈ C 4 \ {NS} the surface S An,Bn,Cn,Dn is also smooth. Both of these are open conditions and therefore they hold on some small neighborhood U of (A n , B n , C n , D n ) ∈ C 4 . The claim follows at once. It is an complex algebraic subset of C 7 . The projection M = pr( M) ⊂ C 4 onto the first four coordinates is therefore constructible, see [42]. Alternately, up to replacing the initial algebraic set by the corresponding projective scheme, the so-called main theorem of elimination theory tells us that the resulting projection on the coordinates (A, B, C, D) yields an algebraic set M . In any case, the fundamental result to be used here is the fact that the Zariski-closure of the constructible set M must coincide with its closure for the standard topology, see [42]. Since it was shown that the complement of M has non-empty interior in the standard topology, it follows that M is contained in a proper Zariski-closed subset of C 4 .
Consider the Zariski-open set Z = C 4 \ (NS ∪ M ), where M stands for the closure of the constructible set M . In particular, Z is not empty. Suppose that (A 0 , B 0 , C 0 , D 0 ) ∈ Z and that Moreover, the initial fixed point p A 0 ,B 0 ,C 0 ,D 0 can actually be (globally) continued along paths c : [0, 1] → Z. Indeed, as the parameters vary in Z, two fixed points of f A,B,C,D cannot collide since they are all simple. Furthermore, they cannot hit ∆ ∞ either, owing to Proposition 8.4.
Suppose for contradiction that the set H had non-empty interior. We can therefore choose some (A 0 , B 0 , C 0 , D 0 ) and some ǫ > 0 such that B ǫ ((A 0 , B 0 , C 0 , D 0 )) ⊂ H ∩ Z. Since f has finitely many fixed points we can reduce ǫ > 0, if necessary, so that there is some fixed point p(A, B, C, D) varying holomorphically over B ǫ ((A 0 , B 0 , C 0 , D 0 )) such that tr(Df Proof. Let S(0) be the set of all six elements γ i,j ∈ Γ satisfying the hypotheses of Theorem H. As in Proposition 7.1 and Lemma 8.6, for every natural number n we will consider the inductively defined sets of iterated commutators S(n) where, for every n, the set S(n + 1) contains every possible commutator of any two distinct elements of S(n).
We assume aiming at a contradiction that for some parameter D there is an unbounded Fatou component V ⊂ S A,B,C,D for Γ A,B,C,D such that V ∩ B ǫ/2 (p) = ∅.
Let p ′ ∈ B ǫ/2 (p) ∩ V and let δ > 0 be sufficiently small so that B δ (p ′ ) ∩ S A,B,C,D ⊂ B ǫ/2 (p) ∩ V . Since the elements γ i,j satisfy Hypothesis (B), it follows from Proposition 7.1 that there is some integer N ≥ 0 such that for any γ ∈ S(N ′ ), with N ′ ≥ N , we have γ(p ′ ) ∈ B δ (p ′ ) and hence that For a fixed neighborhood U ∞ of the vertices of ∆ ∞ as above, Lemma 8.6 gives us a sequence of elements {η n } ∞ n=0 ∈ Γ satisfying Assertions (i) and (ii) of the lemma in question. We claim that V intersects U ∞ non-trivially. Since V is unbounded, there is a sequence {q k } ∞ k=1 ⊂ V which accumulates to ∆ ∞ . Passing to a subsequence, if necessary, we can suppose that it converges to some q ∞ ∈ ∆ ∞ . If q ∞ ∈ V ∞ then the claim holds. Otherwise, we have that γ 3). Applying this to the singleton set {r}, we find that η n (r) → r. In contrast, Assertion (ii) from Lemma 8.6 implies that dist(η n (r), V ∞ ) → 0, providing a contradiction.
We conclude that any Fatou component for Γ A,B,C,D that intersects B ǫ/2 (p) ∩ S A,B,C,D nontrivially must be bounded. It should be noticed that, in each of the examples from Section 7 where we prove that Γ A,B,C,D is locally non-discrete on some open U ⊂ S A,B,C,D , the proof was carried out by producing non-trivial elements of the sets S(n) of iterated commutators that converge to the identity on U as n tends to infinity. The theorem above asserts that, if in addition we have (A, B, C, D) ∈ C 4 good , then this set U does not intersect any bounded Fatou component of Γ A,B,C,D . Indeed, if V were to be a Fatou component intersecting U then for all sufficiently large n the elements of S(n) would stabilize V so that Γ V would not be Abelian. Proof. Since G is a closed subgroup of the Lie group Aut(V ) it is a real Lie group. Thus we only have to show that G is compact.
We first notice that every element of G = Γ V preserves the holomorphic volume form Ω defined in (9). Indeed, by construction, Ω is invariant by elements in Γ V and the condition of preserving Ω is clearly closed so that it has to hold for the closure of Γ V . Since V is bounded, the total (real) volume of V defined by means of Ω is finite. Hence, we can find a relatively compact open set K 0 ⊂ V such that vol Ω (K 0 ) > 1 2 vol Ω (V ). This implies that for every g ∈ G we have g(K 0 )∩K 0 = ∅. Let K = K 0 . Since Aut(V ) acts properly on V , {α ∈ Aut(V ) : α(K) ∩ K = ∅} is a compact subset of Aut(V ). It follows that the closed subset G is compact as well.
Proof of Theorem K. As an abstract group, Γ is isomorphic to the congruence group Γ(2) that is defined in (17). Since any Abelian subgroup of a non-elementary Fuchsian group is cyclic, it suffices to prove that Γ V is Abelian.
We begin by pointing out that every element in Γ V is hyperbolic. Indeed, parabolic elements are conjugate to one of the generators g x , g y , or g z , and hence are such that every open set of S A,B,C,D contains points whose orbit under (any) parabolic element is unbounded. Thus Γ V cannot possess parabolic elements. Finally, it cannot possess elliptic elements either since Γ contains no elliptic element; see Definition/Proposition 8.2 and Remark 8.3. Now, suppose for contradiction that there are two non-commuting elements η, τ ∈ Γ V . By using again the isomorphism between Γ and Γ(2), we see that η and τ correspond to hyperbolic elements in Γ(2) which do not commute. In particular, their iterates also do not commute since two hyperbolic elements of SL(2, Z) commute if and only if they have the same axes of translation in H 2 .
Owing to Lemma 10.1, the closure G = Γ V is a compact Lie Group. Since η and τ are hyperbolic elements, they have infinite order. We can therefore find subsequences of the iterates η n k and τ n ℓ converging to the identity and such that the elements of this subsequence are pairwise different. Thus the subgroup of G generated by η and τ non-trivially accumulates on the identity which implies that the dimension of G itself as a real Lie group is strictly positive. Furthermore, since η n k and τ n ℓ do not commute for any k and ℓ, we also have that the identity component G 0 of G is non-Abelian. On the other hand, the only compact connected real Lie groups of dimension one or two being Abelian (tori), there follows that, in fact, we have dim R (G) ≥ 3, where dim R (G) stands for the dimension of G as real Lie group.
Conversely, the condition that the parameters (A, B, C, D) are in the set C 4 good implies that for any point p ∈ V the orbit G 0 (p) of p under G 0 is diffeomorphic to G 0 . In particular, dim R (G 0 ) ≤ 4. However, if we had dim R (G 0 ) = 4, then G 0 (p) would be four dimensional, implying that G 0 (p) = V . This is clearly impossible since G 0 is compact and V is open. Summarizing, we must have dim R (G) = 3.
The action of G 0 on V is smooth, proper, and free (since (A, B, C, D) lies in C 4 good ). It follows that the quotient space V /G 0 can be given a structure of a smooth manifold with dim R (V /G 0 ) = dim R (V ) − dim R (G 0 ) = 1 in such a way that the quotient map π : V → V /G 0 is a submersion. Thanks to the classical result of Ehresmann, this gives V the structure of a fiber bundle V → V /G 0 where the fibers are diffeomorphic to G 0 , see, for example, [35]. In particular, the base V /G 0 is of dimension 1.
As a one-dimensional smooth manifold, there follows that V /G 0 is either S 1 or R. The former case is impossible because V is non-compact, while the total space of a fiber bundle with compact base and compact fibers is compact.
We will now show that the possibility of having V /G 0 = R cannot occur either. For this, note first that our assumption on parameters implies that S A,B,C,D is smooth and hence the closure S A,B,C,D is smooth in CP 3 . It is therefore biholomorphic to the blow-up of CP 2 at six points, implying that S A,B,C,D is simply connected. If we choose some point p 0 ∈ V then the orbit of G 0 through p 0 gives an embedding of G 0 into S A,B,C,D . Since S A,B,C,D is simply connected, there follows that S A,B,C,D \ G 0 (p 0 ) has two connected components U 1 and U 2 , see, for example [9, Proposition 7.1.1]. Moreover, one of these components, say U 1 , contains the triangle at infinity ∆ ∞ and the other component U 2 is bounded in S A,B,C,D .
By Theorem C, the Julia set J A,B,C,D is connected. The Julia set is also unbounded since it contains the fibers S x 0 for x 0 ∈ (−2, 2) by virtue of Lemma 4.3. Therefore, J A,B,C,D ⊂ U 1 and U 2 ⊂ V .
Recalling that π : V → R stands for the bundle projection, we clearly can assume without loss of generality that π(p 0 ) = 0. The fiber bundle structure implies that V \ G 0 (p 0 ) has two connected components. One of them corresponds to U 1 ∩ V and the other to U 2 ⊂ V . Since π is non-zero on each component we can suppose that π(U 1 ∩ V ) = (0, ∞) and π(U 2 ) = (−∞, 0). However, notice that U 2 ∪ G 0 (p 0 ) is closed and bounded (in C 3 ) and hence it is compact. This implies that π attains a minimum value on U 2 ∪ G 0 (p 0 ) which, in turn, contradicts the fact that π(U 2 ) = (−∞, 0).
We conclude that any two elements of the stabilizer Γ V must commute, and hence that Γ V is cyclic.
Coexistence: Proof of Theorems F and G
In this section we combine the previous results to prove Theorem F about the coexistence of local discreteness and non-discreteness for an open set of parameters and Theorem G about coexistence of Fatou set and Julia set with non-empty interior for the same set of parameters, after removing countably many real-algebraic hypersurfaces H; see Proposition 9.8.
Proof of Theorem F. Lemma 7.5 and Proposition 7.7 give that there is an open neighborhood P 1 ⊂ C 4 that contains (0, 0, 0, 0) and that contains each of the Dubrovin-Mazzocco parameters (A(a), B(a), C(a), D(a)) for a ∈ (−2, 2) such that for each (A, B, C, D) ∈ P 1 we have that Γ A,B,C,D is locally non-discrete on a non-empty open set U ⊂ S A,B,C,D . Moreover, the proofs these results are obtained by showing that for arbitrarily large n there are non-trivial elements of the sets S(n) of iterated commutators of "level n" from Proposition 7.1. Therefore, for each of these parameters values we have non-commuting elements arbitrarily close to the identity on U .
Meanwhile, Theorem E and the proof of Proposition 6.1 ensure the existence of an open set P 2 ⊂ C 4 containing (0, 0, 0, 0) and each of the Dubrovin-Mazzocco parameters (A(a), B(a), C(a), D(a)) for a ∈ (−2, 2) such that for each (A, B, C, D) ∈ P 2 we have: (i) A point p = (u, u, u) ∈ C 3 and an ǫ > 0 with the following property. For every point q in the ball B ǫ (p) and any non-trivial γ ∈ Γ A,B,C one of the coordinates of γ(q) has modulus greater than |u| + ǫ. In particular, any non-trivial γ ∈ Γ A,B,C satisfies γ(q) ∈ B ǫ (p). (ii) Some point q 0 ∈ B ǫ (p) ∩ S A,B,C,D which must therefore be in F A,B,C,D . Let V BQ be the Fatou component containing q 0 from Property (ii). We will first show that Γ A,B,C,D is locally discrete on V BQ . Suppose for a contradiction the existence of a non-empty open set W ⊂ V BQ where the action of Γ A,B,C,D is non-discrete. In other words, there exists a sequence {f j } ⊂ Γ A,B,C,D , f j = id for all j ∈ N, such that the restrictions of the elements f j to W converge uniformly to the identity. Owing to the fact that {f j } is a normal family on V BQ , there follows that a subsequence {f j k } actually converges to the identity on compact subsets of V BQ , cf. the claim in the proof of Proposition 9.3. However, we must then have that f j k (q 0 ) → q 0 and, in particular, for sufficiently large k that f j k (q 0 ) ∈ B ǫ (p). This contradicts Property (i) above.
Therefore, the local non-discreteness of U and local discreteness on V BQ coexist in the dynamics of Γ A,B,C,D for every (A, B, C, D) ∈ P = P 1 ∩ P 2 . In other words, these groups are locally nondiscrete without being "globally non-discrete".
It only remains to check that Γ A,B,C,D acts properly discontinuously on V BQ . Let K ⊂ V BQ be a compact set and, aiming at a contradiction, assume the existence of an infinite sequence {f j } of pairwise distinct elements in Γ A,B,C,D such that K ∩ f j (K) = ∅ for all j. Thus, since K is compact, we can find a subsequence j k and points p and q in K such that the sequence {y k = f j k (p)} k∈N converges to q. Up to enlarging K, we can assume without loss of generality that q lies in the interior of K. Setting l k = j k+1 − j k , it follows that g k = f l k sends y k to y k+1 and both points converge towards q as k → ∞.
On the other hand, since V BQ is contained in the Fatou set, normality implies that the derivatives of the elements g k are uniformly bounded in K. Hence, for k large enough, g k sends some fixed neighborhood U q of q to a bounded subset of K. Again, normality implies that a subsequence {g k(i) } i of {g k } converges uniformly on (compact subsets of) U q to a non-constant map g ∞ : U q → K. Moreover, since each of the mappings g k preserves the volume form Ω, so does the limit g ∞ , thus implying that the limit is locally invertible. Up to reducing the size of U q we can suppose that g ∞ is actually invertible. There follows that the sequence of maps h i = g −1 k(i+1) • g k(i) converges uniformly to the identity on compact subsets of U q . This contradicts the local discreteness of Γ A,B,C,D on V BQ . The proof of Theorem F is complete. Here, H is the countable union of real-algebraic hypersurfaces provided by Proposition 9.8. Recall from Theorem F that there are non-commuting pairs of elements of Γ A,B,C,D arbitrarily close to the identity on U . Therefore, Theorem K gives that U is disjoint from any bounded Fatou component of Γ A,B,C,D . In fact, to ensure that U is disjoint from any bounded Fatou component is the only place in the proof where the parameters in H need to be removed from the set of parameters P.
We will now use Theorem H to show that U is also disjoint from any unbounded Fatou component, and this for every (A, B, C, D) ∈ P. Since this requires more specific details, the discussion will be split into two cases in order to make the argument more clear. Also, in the sequel, we are allowed to reduce the size of the open set U , if necessary. | 2021-04-20T01:15:53.553Z | 2021-04-19T00:00:00.000 | {
"year": 2021,
"sha1": "ce3da39c0e398b19720c8e35fdb41f92853ab3ea",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "65b33ee94cb9894ac5747a60c7b223891833c851",
"s2fieldsofstudy": [
"Mathematics"
],
"extfieldsofstudy": [
"Mathematics",
"Physics"
]
} |
206518469 | pes2o/s2orc | v3-fos-license | Prosthesis Prescription Protocol of the Arm (PPP-Arm): The implementation of a national prosthesis prescription protocol
Background and aim: In order to create more uniformity in the prescription of upper limb prostheses by Dutch rehabilitation teams, the development and implementation of a Prosthesis Prescription Protocol of the upper limb (PPP-Arm) was initiated. The aim was to create a national digital protocol to structure, underpin, and evaluate the prescription of upper limb prostheses for clients with acquired or congenital arm defects. Technique: Prosthesis Prescription Protocol of the Arm (PPP-Arm) was developed on the basis of the International Classification of Functioning and consisted of several layers. All stakeholders (rehabilitation teams, orthopedic workshops, patients, and insurance companies) were involved in development and implementation. A national project coordinator and knowledge brokers in each team were essential for the project. Discussion: PPP-Arm was successfully developed and implemented in nine Dutch rehabilitation teams. The protocol improved team collaboration, structure, and completeness of prosthesis prescriptions and treatment uniformity and might be interesting for other countries as well. Clinical relevance A national protocol to prescribe upper limb prostheses can be helpful to create uniformity in treatment of patients with upper limb defects. Such a protocol improves quality of care for all patients in the country.
Background
In the Netherlands, annually about 40 people lose an upper limb or hand and about 50 children are born with a transverse upper limb reduction. The group of clients with an arm deficiency is thus small, but often needs long-term care because they usually have a normal life expectancy. It is important that the care of this small but intensive client group is of a high quality because all aspects of functioning may be influenced by the limb loss. Furthermore, care is associated with high costs, especially when prostheses or adaptive devices are provided. Specialized care by a multidisciplinary team is necessary.
In the Netherlands, each rehabilitation team previously followed its own prosthesis prescription process. As a result, upper limb prosthesis care differed among the diverse rehabilitation institutions in, for example, the use of educational materials, structure, and content of the prescription letter for insurance companies and advice regarding types of devices. There was a need for a standardized method of prescribing upper limb prostheses in order to improve quality of care. In 2009, the national Working Group Amputation and Prosthetics of the Arm of the Dutch Society of Rehabilitation Specialists took the initiative to develop the Prosthesis Prescription Protocol of the Arm (PPP-Arm). 1 Aim PPP-Arm aims to structure, underpin, and evaluate the prescription of an adaptive device or a prosthesis and to create a uniform, nationally applicable, prescription policy. The intention was to use the protocol in all nine Dutch centers dealing with these patients.
Technique
The protocol has been created through the collaboration of various rehabilitation centers, patients, orthopedic workshops, and insurance companies ( Figure 1). Representatives of these groups formed the Working Group PPP-Arm.
In October 2011, the paper version of the protocol was approved by the Assembly of the Dutch Society of Rehabilitation Specialists. In order to make the protocol efficient and user-friendly, a digital version of the protocol was developed in 2012.
After obtaining funding, a national project coordinator was appointed. Furthermore, in each team, a knowledge broker was appointed, who was responsible for the implementation of the protocol within his own center.
The protocol is based on the World Health Organization's criteria of the International Classification of Functioning (ICF) and consists of the following layers ( Figure 2): 1. Establishing patient's demands; 2. Preparation of treatment requirements; 3. Establishing device requirements; 4. Selection, try-out, and final decision; 5. Delivery of the device; 6. Use of the device (instructions and training); 7. Evaluation. 2 To fill in the digital protocol, the practitioner has to log in on a specific website (Orthofirm) with a personal username and password. The data of the client can be entered and stored in a structured manner in the various folders and tabs ( Figure 3). The layers of the protocol provide the various practitioners (rehabilitation physician, therapist, and orthopedic technician) the opportunity to register relevant information quickly and easily. In the folder "Care request," there is room to describe the client's care request and provide information about the medical diagnosis. In the next folder, the medical analysis of the need for care is registered. In this folder, there are several tabs based on the ICF: anatomical features, body functions, activities, and participation. Most questions have multiple choice options with the possibility to add relevant information.
Results of several questionnaires and measurement instruments can be registered in the protocol, including Canadian Occupational Performance Measurement (COPM), Assessment of Capacity for Myoelectric Control (ACMC), and Dutch version of Quebec User Evaluation of Satisfaction with assistive Technology (D-QUEST). The folder "Program of demands" helps to identify which solution fits best the client's demands and describes the requirements of the device.
On the last page of the folder "Program of demands," the decision is made whether a prosthesis should be prescribed or not: the Go/No Go decision. The conclusion may be that a prosthesis is not a suitable solution for the client's needs. In that case, a "No Go decision" will be chosen. If it is concluded that a prosthesis or adaptive device is the proper solution for the client's functional problems, a "Go decision" is registered and motivated.
In some cases, it is difficult to decide which type of prosthesis should be advised, for example, when a client has bilateral upper limb amputations or when a multi-articulated hand is considered. In these situations, a trial period might be an option: the client will be able to try a temporary prosthesis for several weeks at home. After this trial period, the final decision on the prosthesis choice can be made and motivated. In the folder "Components," information about the components of the advised prosthesis can be added by the technician. 4 After filling in all folders of the protocol, a thorough and structured analysis has been performed to achieve a suitable solution for the client's needs. The protocol does not provide a score of some kind to sort devices. After completing all the layers of the protocol in a standardized way, it becomes clear which needs and goals the patient has and which prosthesis components may be helpful to achieve these goals and needs. A list of possible options for prosthesis components can be found in the folder "Components." A prescription report can be generated for the insurance company. This report consists of a main page with the client's care request, medical diagnosis, explanation, and motivation for the prosthesis choice and the components of the prosthesis. An appendix can be added with other information concerning body functions, activities, and participation and measurement data.
After the prosthesis is delivered, the protocol can be used for instructions, training, and evaluation. In the folder "Evaluation," the care request and need for care can be evaluated when the client has been using the prosthesis for a while. The questions are based on the ICF. Measurement data of the follow-up can be registered here and help to formulate a conclusion.
We also have developed the folder "Information." In this folder, there are over 100 links to educational materials for clients on for example: phantom sensations, upper limb prostheses, personal care, and peer contact. This information includes photographs, movies, brochures, and PowerPoint presentations. 3
Discussion
A national prosthesis prescription protocol for upper limb prosthesis users, "PPP-Arm," was successfully developed and implemented in nine Dutch rehabilitation teams. The protocol is used to structure, underpin, and evaluate the prescription of an adaptive device or a prosthesis. The protocol created a uniform, nationally applicable, prescription policy.
Several advantages of the protocol were experienced during the execution of the project: The main disadvantage mentioned was the time investment needed to learn how to use the protocol. The use of PPP-Arm has been evaluated using a questionnaire answered by the participating rehabilitation teams and by the insurance companies. The most important conclusions were as follows: • • PPP-Arm is used by all participating centers • • The quality of the prosthesis applications has improved: PPP-Arm provides a more structured report with all the information that is needed for the insurance company to judge the application. • • PPP-Arm has produced several positive developments: better collaboration of different team members, more structure and completeness in the prescriptions, more treatment uniformity in the country, and implementation of a temporary prosthesis trial period.
By developing PPP-Arm, we have managed to create a national uniform and structured method to advise and evaluate the prescription of upper limb prostheses. This might be interesting for other countries as well. An English version is available for usage in other countries.
Key points
• • Rehabilitation teams for upper limb prosthesis users work more uniformly using a national prescription protocol. • • A national project coordinator and knowledge brokers in each team are prerequisites for successful protocol development and implementation.
• • All stakeholders should be involved in protocol development and implementation: rehabilitation teams, orthopedic workshops, patient organizations, and insurance companies. • • The ICF framework can be used to structure a prescription protocol for upper limb prosthesis users.
Author contribution
All authors contributed equally in the preparation of this manuscript.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. | 2018-04-03T04:18:17.355Z | 2018-01-04T00:00:00.000 | {
"year": 2018,
"sha1": "b650d886421b79706de0498cb604ebf70b8d1701",
"oa_license": "CCBYNC",
"oa_url": "https://journals.sagepub.com/doi/pdf/10.1177/0309364617747962",
"oa_status": "HYBRID",
"pdf_src": "PubMedCentral",
"pdf_hash": "b650d886421b79706de0498cb604ebf70b8d1701",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
235279696 | pes2o/s2orc | v3-fos-license | Utilization of volcanic ashes for geopolymer based on alkaline activator and solid-liquid ratio
The volcanic ashes are an abundant natural resource in Indonesia, but they are still little is used optimally, such as geopolymer raw material. Geopolymers are a class of inorganic polymer that be able by the reaction of an aluminosilicate material with an alkaline solution. The research aims to investigate the synthesis of geopolymers using two types of volcanic ash (Mt. Merapi and Mt. Sinabung). The synthesis of geopolymer was carried out at variation of solid/liquid ratio at 65% : 35% and 70% : 30%, with two alkaline solutions (NaOH and KOH) under different alkaline concentration (8, 9, and 12 M). A multi-analytical approach is proposed: chemical (XRF) and spectroscopic (FT-IR) analyses. The Results showed that geopolymer is influenced by volcanic ash-type and SiO2/Al2O3 ratio. Geopolymers are formed when volcanic ash of Sinabung mix with NaOH 10 NaOH with ratio of 65% : 35%, and the polymers are relatively stable. The FTIR spectra of the synthesized geopolymers showed broad absorbance bands, between wave 972-962 cm−1 and 931-976 cm−1 assigned to the internal vibrations of Si–O–Si, and Si–O–Al respectively. Both volcanic ash materials from the Merapi and Sinabung volcanoes can be utilized for making geopolymer, suitable for both engineering and agriculture applications.
Introduction
Geopolymers are a class of inorganic polymers prepared by chemical reactions of alkaline solutions and aluminosilicate material [1]. Geopolymers have attracted the attention of researchers for several years due to they are environmental friendly materials [2]. The geopolymerization process influenced by the chemical and mineralogical composition of aluminosilicate raw materials, particle size distribution and specific surface area of the raw material, curing temperature, the composition of alkaline solution, and liquid to solid ratio [3].
The use of raw materials commonly for the synthesis of geopolymer was kaolinite, metakaolin, and fly ash [4][5][6]. Volcanic ash is one of aluminosilicate-type whose percentage in Al 2 O 3 and SiO 2 can allow their utilization in the synthesis of geopolymers [7]. Among the materials of aluminosilicate for synthesis geopolymer, volcanic ash has used as raw materials for polymers [7][8][9]. On the other hand, volcanic ash is one of more efficient material, because of low CO2 emission and good physical and mechanical properties for geopolymer materials [10].
Indonesia is an active volcanic region with over 100 volcanoes [11], i.e., Mount Merapi in Central Java and Mount Sinabung in North Sumatera. Their volcanic activities generate vast deposits of [12], while volcanic deposits of Mt. Sinabung was approximately 3 x 10 8 m 3 [13]. [14] reported that some characteristics of these volcanic ash can be considered as pozzolan material, due to their had high silica (49.33% -61.13% SiO 2 ) and alumina (15.93% -17.78% Al 2 O 3 ) content. Hence, volcanic ashes can allow readily soluble in alkali to produce geopolymers. Nevertheless, the little amount of this raw material is used for the production of geopolymers. The possibility of utilization of volcanic ash for the synthesis of effective geopolymers could be of economic importance for countries with vast deposits of this raw material.
The objective of this work is to investigate the synthesis of geopolymers using two types of volcanic ash (Mt. Merapi and Mt. Sinabung) as source material and second to study the properties of geopolymers with solid/liquid ratio, molar ratio, and alkaline types. The products were characterized by Infrared spectroscopy (FTIR).
Material
Volcanic ashes were collected from the volcanic deposit of Indonesia (Mount Merapi and Sinabung). Mount Merapi is at Central Java about 30 km north of Yogyakarta city located 7° 32,5' S and 110° 26,5' E, and having a height of 2986 m. While Mount Sinabung is located in Karo District, North Sumatera, geographically on 3° 10' 16.7" N and 98° 23' 24.66" E and altitude 2,460 m.
The volcanic ashes were dried at 105° C for 72 hours, and ground order to have a fine powder of <200 µm. Their chemical composition was determined by X-ray fluorescence. The alkaline solution used alkaline hydroxide (NaOH and KOH), followed by the addition of sodium silicate (Na 2 SiO 3 ).
Geopolymer synthesis
The synthesis of geopolymer was conducted at the variation of solid-liquid ratio 65:35, 70:30 and molarity of NaOH/KOH; 8, 9 and 10 M. The solid consists of volcanic ash and Al 2 O 3 . This volcanic ashes between Mt. Merapi and Mt sinabung have a higher SiO 2 /Al 2 O 3 moral ratio [14], then the need for the addition of reactive aluminum such as Al 2 O 3 . Geopolymer pastes were obtained by mixing alkaline solution with volcanic ashes (Figure 1).
Alkaline solution (NaOH and KOH) was prepared using NaOH pellet with distilled water. The solution was stored for 24 hours. Alkali solution was prepared by mixing sodium silicate (Na 2 SiO 3 ) with NaOH or KOH solution. The volcanic ashes and Al 2 O 3 as the starting material, was mixed with alkali solution. The paste was then poured into 4 x 4 x 4 cm 3 and then vibrated for 5 min to remove air bubbles during pouring. After casting, the samples were cured at 70° C for 7 days. Before curing the samples in oven at 70° C, they were first cured in open air at ambient temperature for 30 min to avoid cracks due to rapid water evaporation. Then, the geomaterial formation of geopolymer was used FTIR spectroscopy. Figure 1. Schematic illustration of geopolymer preparation process.
Geopolymer structure
Total chemical properties (Table.1) of volcanic ash showed that the SiO 2 / Al 2 O 3 molar ratio is 5.26 for volcanic ash of Mt. Sinabung (Vsg) and 5.84 for volcanic ash of Mt. Merapi (Vmr) [14]. These value are content basic from raw material of geopolymer. After geopolymer volcanic ash is formed the change in the ratio of SiO 2 / Al 2 O 3 becomes 3.88 for Vsg 65/30 NaOH 10, and 3.66 for Vmr 65/30 NaOH 10. It showed that geopolymer have occured optimally [15,16] reported that the optimum value geopolymers was reported to vary from 3.3 to 4.5.
The type of volcanic ash influenced the formation of geopolymers ( Figure 2). Geopolymer had cavities caused by volcanic glass content contained in the volcanic ash mineral. Volcanic glass content will result in the cavity [17]. While the hardening time influenced CaO content in the raw material. The higher the CaO content, the faster the geopolymer hardening time. Mt. Merapi (CaO; 6.22%) very quickly hardens with a curing time of 2 days at 70° C, while Mt. Sinabung (CaO; 5.87%) which requires more curing time for 4 days at 70° C. However, high CaO content affected the fragility and strength of geopolymers [18]. Geopolymer, which has high cao content will easily crack, such as Vmr.
The use of the ratio of solid and liquid 65:35 produced better geopolymers compared to 70:30. The high comparison of liquid and solid caused the more unreactive of geopolymerization reaction. Hence, its resulted in concrete solids with low density but high porosity [5]. The alkaline molarity (NaOH and KOH) affects geopolymers, the optimum conditions of geopolymer were obtained at 10 M NaOH and 10 M KOH. Geopolymer with NaOH solution as activator has better geopolymer than geopolymer with KOH solution as an activator [19].
FTIR spectra
The infared spectra of volcanic ashes and geopolymers presented in Figure 3 [8,10]. It is also observed that FTIR main band systematically switches to lower wave with concomitant increase in band intensity as molaritas alkali activator ratio increases in the system [21]. The absorption of sharp peak around 1600 cm -1 and 2985 -3321 cm -1 are assigned to stretching (-OH) and bending (H-O-H) vibrations of bound water molecules [22], which gives evidence of adsorbed water in the geopolymers. The absorption bands between 1400 -1500 cm -1 are attributed to stretching vibrations of O-C-O bond indicating the presence of sodium bicarbonate that is suggested to occur due to the atmospheric carbonation [23]. The presence of sodium carbonate may disrupt the polymerization process. It was showed by solid/liquid (65:35) less steep compared to (70:30). Figure 3 and 4, showed band of KOH solution higher than NaOH for geopolymer products. Geopolymers with KOH as activator has rather lower strength than geopolymer with NaOH solution as an activator. The difference between KOH and NaOH is due to ionic size where Na + is having smaller ionic size compares to K + . Na + will be more active thus will enhance the dissolution process of alumino-silicate minerals [19].
Conclusions
The geopolymer of volcanic ash depends mainly on SiO 2 /Al 2 O 3 molar ratio and chemical composition from raw material. The results obtained showed that the strength of geopolymers increases with increasing amount of amorphous phase and decreases with increasing SiO 2 /Al 2 O 3 molar ratio of the amorphous phase. Geopolymerization phase increases with alkali solution, i.e, NaOH. However, it decreases with KOH. Geopolymers with very good and successfully prepared from Vsg 65/35 NaOH 10 formed geopolymers that are relatively stable. | 2021-06-03T00:17:00.927Z | 2021-01-01T00:00:00.000 | {
"year": 2021,
"sha1": "8d4f7daf53301bbb768e1872bd8d41a7a69e65e8",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1755-1315/708/1/012058",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "8d4f7daf53301bbb768e1872bd8d41a7a69e65e8",
"s2fieldsofstudy": [
"Materials Science",
"Environmental Science"
],
"extfieldsofstudy": [
"Physics"
]
} |
258865214 | pes2o/s2orc | v3-fos-license | Approximating Multiobjective Optimization Problems: How exact can you be?
It is well known that, under very weak assumptions, multiobjective optimization problems admit $(1+\varepsilon,\dots,1+\varepsilon)$-approximation sets (also called $\varepsilon$-Pareto sets) of polynomial cardinality (in the size of the instance and in $\frac{1}{\varepsilon}$). While an approximation guarantee of $1+\varepsilon$ for any $\varepsilon>0$ is the best one can expect for singleobjective problems (apart from solving the problem to optimality), even better approximation guarantees than $(1+\varepsilon,\dots,1+\varepsilon)$ can be considered in the multiobjective case since the approximation might be exact in some of the objectives. Hence, in this paper, we consider partially exact approximation sets that require to approximate each feasible solution exactly, i.e., with an approximation guarantee of $1$, in some of the objectives while still obtaining a guarantee of $1+\varepsilon$ in all others. We characterize the types of polynomial-cardinality, partially exact approximation sets that are guaranteed to exist for general multiobjective optimization problems. Moreover, we study minimum-cardinality partially exact approximation sets concerning (weak) efficiency of the contained solutions and relate their cardinalities to the minimum cardinality of a $(1+\varepsilon,\dots,1+\varepsilon)$-approximation set.
exist for general multiobjective optimization problems. Moreover, we
Introduction
Many real-world optimization problems require taking into account multiple conflicting objective functions. Since solutions that optimize all objectives simultaneously do usually not exist, the solutions of interest are the so-called efficient (or Pareto optimal ) solutions. A solution is called efficient if any other solution that is better in some objective is worse in at least one other objective. The images of the efficient solutions in the objective space are called nondominated points. A main goal in multiobjective optimization is to determine the set of all nondominated points (the nondominated set) and, for each of them, one corresponding efficient solution. One major difficulty, however, is intractability, that is, the fact that the nondominated set (and the efficient set) may be exponentially large for discrete problems (see, e.g., [1]), and typically infinite for continuous problems.
This provides a strong motivation to consider approximations of multiobjective optimization problems. Approximations for general multiobjective problems have been investigated since the 1990s [2,3]. Their existence and cardinality for general multiobjective problems have been systematically investigated in the seminal work of Papadimitriou and Yannakakis [4]. They show that, for any instance of a multiobjective optimization problem with a constant number of positive-valued, polynomial-time computable objective functions and any ε > 0, there exists a (1 + ε, . . . , 1 + ε)-approximation set (also called a (1+ε)-approximation set or an ε-Pareto set) whose cardinality is fully polynomial, i.e., polynomial in the encoding length of the instance and 1 ε . Additionally, they prove that a (1 + ε)-approximation set can be computed in (fully) polynomial time for every ε > 0 if and only if a certain auxiliary problem Gap δ (the gap problem) can be solved in (fully) polynomial time for every δ > 0.
While an approximation guarantee of 1 + ε for any ε > 0 is the best solution quality one can expect for singleobjective problems (apart from solving the problem to optimality), even better approximation guarantees than (1 + ε, . . . , 1 + ε) can be considered in the multiobjective case since the approximation might be partially exact, i.e., exact in some of the objectives (and (1 + ε)-approximate in the others). In fact, it has recently been shown in [9] that, under similar assumptions as in [4], there exist polynomial-cardinality one-exact (1 + ε)-approximation sets that achieve an approximation guarantee of 1 in one of the objective functions (without loss of generality the first one) and 1 + ε in the others.
In this paper, we systematically investigate which types of partially exact approximation sets of polynomial cardinality are guaranteed to exist for general multiobjective problems. In particular, we show that, for any constant number p of objective functions and any k ≤ ⌈ p 2 ⌉, polynomial-cardinality (1+ε)-approximation sets exist that approximate each feasible solution exactly in k of the objectives if we allow the objectives in which the approximation is exact to differ between different feasible solutions. We call these quasi-k-exact (1 + ε)-approximation sets. Their existence contrasts the fact that the efficient and nondominated sets are already intractable for many biobjective problems, which implies that polynomial-cardinality (1 + ε)-approximation sets that are exact in the same objectives for all solutions can in general be exact in at most one objective [9]. We obtain our general result by providing a general existence proof that generalizes the existence proofs for (1 + ε)-approximation sets and one-exact (1 + ε)-approximation sets provided in [4] and [9], respectively.
Moreover, we investigate the question of (weak) efficiency of the solutions contained in minimum-cardinality partially exact (1 + ε)-approximation sets of different types and relate their cardinalities to the minimum cardinality of a (1 + ε)-approximation set for a given instance. In particular, we show that, for every instance, the minimum cardinality of a quasi-1-exact (1 + ε)-approximation set equals the minimum cardinality of an ordinary (1 + ε)-approximation set 1 . This contrasts the result from [9] stating that the minimum cardinality of a one-exact (1 + ε)-approximation set can be an arbitrary factor larger than the minimum cardinality of an ordinary (1 + ε)approximation set. By generalizing the corresponding proof from [9], however, we show that, for k ≥ 2, the cardinality of a quasi-k-exact (1+ε)-approximation set can again be an arbitrary factor larger than the minimum cardinality of an ordinary (1 + ε)-approximation set.
Finally, we present several results concerning the polynomial-time computability of partially exact approximation sets. In particular, we show that the (fully) polynomial-time solvability of the gap problem does not suffice for the computation of any type of partially exact approximation set in general.
While we focus on results that are applicable to general multiobjective optimization problems under weak assumptions, there also exists a large body of work on approximations for specific multiobjective problems. For an up-todate survey on both general and problem-specific approximation results, we refer to [10].
Preliminaries
Multiobjective optimization problems can contain objective functions that are to be minimized or maximized (or even a combination of both). However, we only consider the case of minimization problems in this article, i.e., all objectives are to be minimized. This is without loss of generality here since all our arguments can be straightforwardly adapted to the case where some or all objective functions are to be maximized. Definition 2.1 (Multiobjective Optimization Problem) A multiobjective optimization problem Π is given by a set of instances. Each instance I consists of a (finite or infinite) set X I of feasible solutions and a vector f I = (f I 1 , . . . , f I p ) of p objective functions f I i : X I → Q for i = 1, . . . , p to be minimized. The feasible set X I might not be given explicitly.
Here, the number p of objective functions in a multiobjective optimization problem Π is assumed to be constant. Moreover, as is common in the context of approximation of multiobjective optimization problems, we assume that the objective functions take only positive, rational values and are polynomialtime computable (cf. [4,6,8,9]). Additionally, we use the following standard assumption 2 : Assumption 1 For any multiobjective optimization problem Π, there exists a polynomial P such that for any instance I of Π, there exists a constant M I ≤ P(enc(I)) such that, for any x ∈ X I and any i ∈ {1, . . . , p}, we have enc(f I i (x)) ≤ M I , where enc(I) denotes the encoding length of the instance I and enc(f I i (x)) denotes the encoding length of the value f I i (x) ∈ Q >0 in binary. This, in particular, implies that, for any instance I and any objective function value In the following, we sometimes blur the distinction between the problem Π and a concrete instance I = (X I , f I ) and we drop the superscript I indicating 2 Given the previously-stated assumptions on the objective functions, this general assumption also used in [6,8,9] holds for large classes of problems -in particular, for combinatorial problems. Together with the previously-stated assumptions, it guarantees the existence of all versions of approximation sets considered here and in [4,6,8,9]. We note, however, that specific types of approximation sets can also exist for some classes of problems for which this assumption does not formally hold (e.g., for multiobjective linear programming problems, see [4]).
the dependence on the instance in X I , f I , M I , etc. whenever the instance is clear from the context.
For multiobjective optimization problems, the solutions of interest are the so-called efficient solutions for which any other solution that is better in some objective is worse in at least one other objective: Definition 2.2 For an instance I of a multiobjective optimization problem, a solution x ∈ X dominates another solution In this case, we call the corresponding image f (x) ∈ f (X) ⊆ Q p a (weakly) nondominated point. The set X E ⊆ X of all efficient solutions is called the efficient set (or Pareto set) and the set Y N := f (X E ) of nondominated points is called the nondominated set. Similarly, the set X WE ⊆ X of all weakly efficient solutions is called the weakly efficient set and the set Y WN := f (X WE ) of weakly nondominated points is called the weakly nondominated set.
One way to circumvent the problem that the nondominated set often has cardinality exponential in the input size (see, e.g., [1]) is the concept of approximation. Here, instead of requiring to select at least one corresponding efficient solution for each nondominated point, the requirement is relaxed so that each image point is only required to be dominated "approximately" by some image of a selected solution. This idea is formalized as follows: Let (X, f ) be an instance of a multiobjective optimization problem and let α = (α 1 , . . . , αp) ∈ R p with α i ≥ 1 for all i ∈ {1, . . . , p}. We say that a feasible solution x ∈ X α-approximates another feasible solution When fixing a desired vector α of approximation guarantees (or a set of possible desired vectors), approximation as in Definition 2.3 can equivalently be expressed as a relation R ⊆ X × X, where (x, x ′ ) ∈ R (also denoted as xRx ′ ) if and only if x ′ is α-approximated by x for some desired vector α. In the following, we consider general approximate dominance relations that are monotonic in the following sense: Note that any monotonic relation R is, in particular, reflexive, i.e., xRx for all x ∈ X. We are particularly interested in the following monotonic approximate dominance relations, which give rise to ordinary (1 + ε)-approximation sets as introduced in [4] and the different types of partially exact approximation sets considered here: Definition 2.5 Let (X, f ) be an instance of a multiobjective optimization problem and let ε > 0. We define the following relations on X × X: where k components α i are equal to 1 and the other p − k are equal to 1 + ε.
x ′ :⇔ x αapproximates x ′ , where α 1 = 1, k − 1 of the other components α i are equal to 1, and the remaining p − k are equal to 1 + ε. 3 The relations will also be referred to as partially exact (approximate dominance) relations in the following.
Given an approximate dominance relation, approximation sets for multiobjective optimization problems are defined as follows: Let (X, f ) be an instance of a multiobjective optimization problem and let R be a relation on X × X. A set P ⊆ X of feasible solutions is called an R-approximation set for (X, f ) if, for any feasible solution x ′ ∈ X, there exists a solution x ∈ P such that xRx ′ (i.e., x R-dominates x ′ ).
In the following, we will often use that, since monotonic relations R are reflexive, R-approximation sets for monotonic relations correspond to Rdominating sets for X, i.e., to subsets P ⊆ X of feasible solutions such that any feasible solution x ′ / ∈ P is R-dominated by some solution x ∈ P . This correspondence, in particular, holds for the monotonic relations We are interested in approximation sets whose cardinality is bounded by a polynomial in the encoding size of the given instance (X, f ). For the re- that depend on a given value ε > 0, 3 Note that this means that ⪯ 1,(k) we are specifically interested in approximation sets whose cardinality is fully polynomial, i.e., polynomial in the encoding size of the instance and in 1 ε . In the literature, several auxiliary problems have been defined for the computation of (fully) polynomial-cardinality approximation sets [4,6,8,9], which are also used in the following. The first such problem, whose (fully) polynomialtime solvability has been shown to characterize the (fully) polynomial-time compatibility of (1 + ε)-approximation sets in [4] is the gap problem: Definition 2.7 (Gap Problem) Given an instance (X, f ) of a p-objective optimization problem, a point b ∈ Q p , and some δ > 0, the problem Gap δ (b) is the following: Either return a feasible solution x ∈ X with f i (x) ≤ b i for i = 1, . . . , p, or answer correctly that there does not exist any feasible solution The following auxiliary problem, which scalarizes the multiobjective problem via budget constraints on all but one objective function, is widely used both in practice and in the theoretical literature on multiobjective optimization: Given an instance (X, f ) of a multiobjective optimization problem and bounds b i > 0, i = 2, . . . , p, for all objective functions except the first one, the problem Constrained(b 2 . . . , bp) is the following: Either answer that there does not exist a feasible solution . . , p, or return a feasible solution that minimizes f 1 among all such solutions, i.e., return x ∈ X with Note that Constrained is hard to solve even for the biobjective version of most relevant problems such as shortest path. A way to circumvent the hardness of Constrained is to consider solutions that violate the given bounds slightly, while requiring an objective value that is at least as good as the objective value of any solution that respects the bounds [6]: Definition 2.9 (DualRestrict) Given an instance (X, f ) of a multiobjective optimization problem, bounds b i > 0, i = 2, . . . , p, for all objective functions except the first one, and some δ > 0, the problem DualRestrict δ (b 2 , . . . , bp) is the following: Either answer that there does not exist a feasible solution Note that Constrained can be viewed as the limit case where δ = 0 in DualRestrict.
The auxiliary problems Constrained and DualRestrict can also be defined such that, instead of the first one, some other objective is to be optimized subject to budgets on the remaining objectives. In the following, we sometimes indicate which objective is optimized by an upper index. For example, DualRestrict i denotes the DualRestrict problem with a bound on all objectives but the i-th one.
Existence of Approximation Sets
We start by investigating the existence of R-approximations of (fully) polynomial cardinality under weak assumptions on the approximate dominance relation R. For the (1 + ε)-dominance relation R = ⪯ ε , this is established in the seminal work of Papadimitriou and Yannakakis [4], who show the existence of (1 + ε)-approximations of fully polynomial cardinality O(( M ε ) p−1 ). We now generalize the existence proof from [4] to more demanding approximate dominance relations R that refine ⪯ ε , which, as shown afterwards in Theorem 3.2, include the partially exact approximate dominance relations ⪯ 1 ε and ⪯ (k) ε .
Theorem 3.1 Let (X, f ) be an instance of a p-objective optimization problem and let ε > 0. Consider a monotonic relation R ′ defined on X × X and let R :=⪯ ε ∩ R ′ . If there exists a constant c (independent of the instance size and 1 ϵ ) such that every nonempty subset X ′ ⊆ X admits an R ′ -dominating set P X ′ ⊆ X ′ of cardinality at most c, then there exists an R-approximation set of cardinality O M ε p−1 .
in which all images of feasible solutions are contained. We create a grid by subdividing this hyperrectangle into smaller hyperrectangles such that, in each dimension, the ratio of the larger to the smaller coordinate is 1 + ε. Thus, the total number of subdivisions in each dimension is log 1+ε . By the assumption on R ′ , any subset X ′ of feasible solutions admits an R ′ -dominating set P X ′ ⊆ X ′ of cardinality at most c. Moreover, any two solutions x and x ′ with images in the same hyperrectangle of the grid satisfy x ⪯ε x ′ by construction of the grid. Thus, if X ′ is the preimage of a nonempty hyperrectangle of the grid, then the R ′ -dominating set P X ′ is actually an R-dominating set for X ′ . Consequently, forming the union of the sets P X ′ over all subsets X ′ that are preimages of nonempty hyperrectangles of the grid yields an R-approximation set. Moreover, since R is monotonic, it suffices to consider only weakly nondominated nonempty hyperrectangles.
To bound the cardinality of this R-approximation set, we note that, for each hyperrectangle [l 1 , u 1 ] × · · · × [lp, up] of the grid, we have u i = (1 + ε) · l i for all i ∈ {1, . . . , p}, so the vectors l = (l 1 , . . . , lp) and u = (u 1 , . . . , up) of lower and upper bounds of the hyperrectangle lie on a straight line through the origin. Calling the set of all hyperrectangles of the grid whose bounds lie on the same straight line through the origin a diagonal, we see that (1) at most one hyperrectangle on each diagonal is weakly nondominated and nonempty, and (2) the number of diagonals is Hence, since the cardinality of the constructed R-approximation set is at most c times the number of weakly nondominated nonempty hyperrectangles and c is a constant, its cardinality is also in O M ε p−1 .
□
We remark that the proof of Theorem 3.1 also yields an R-approximation set of (fully) polynomial cardinality if the cardinality of the R ′ -dominating set P X ′ for each subset X ′ ⊆ X is only required to be polynomial in the instance size (and in 1 ε ) instead of constant. In this case, the cardinality of the obtained R-approximation set would be upper bounded by O M ε p−1 times the worst-case cardinality of P X ′ taken over all preimages X ′ of weakly nondominated nonempty hyperrectangles of the grid.
Proof 1) Choosing R ′ := X ×X in Theorem 3.1 yields R =⪯ ε . Moreover, R ′ is clearly monotonic and, for any nonempty subset X ′ ⊆ X, any solution x ∈ X ′ constitutes an R ′ -dominating set, so the result follows from Theorem 3.1.
Again, R ′ is clearly monotonic. Moreover, given any nonempty subset X ′ ⊆ X, any solution x ∈ X ′ such that f 1 (x) = min x ′ ∈X ′ f 1 (x ′ ) constitutes an R ′ -dominating set (note that, by Assumption 1, the number of possible images is finite, so the minimum is guaranteed to exist). 3) Given k ≤ ⌈ p 2 ⌉, choosing ε . Since R ′ is clearly monotonic, it remains to show that there exists an R ′ -dominating set of constant cardinality for any nonempty subset X ′ ⊆ X. To show this, we use a result from [11] about dominating sets in k-majority tournaments. A k-majority tournament is a directed graph on a finite vertex set V where, given 2k − 1 linear orders of V , there exists a directed arc from u to v if and only if u is ranked higher than v in at least k of the orders. The result from [11] states that any k-majority tournament admits a dominating set of cardinality O(k log k).
In our situation, given some nonempty subset X ′ ⊆ X, each objective f j defines a reflexive, transitive, and strongly connected order on f (X ′ ) by ranking f (x) higher than f (x ′ ) if and only if f j (x) ≤ f j (x ′ ). This order can be made antisymmetric (i.e., turned into a linear order) by breaking ties within any subset of elements that have the same f j -value via an arbitrary linear order on this subset. Since k ≤ ⌈ p 2 ⌉, this construction yields p ≥ 2k − 1 linear orders on f (X ′ ). Moreover, by Assumption 1, the number of possible images the objective space is finite, so f (X ′ ) is a finite set. Consequently, the directed graph on f (X ′ ) in which there exists a directed arc from f (x) to f (x ′ ) if and only if f (x) is ranked higher than f (x ′ ) in at least k of the orders is finite and contains a k-majority tournament on f (X ′ ) (since, in case that p > 2k − 1, the additional p − 2k + 1 linear orders only lead to additional arcs in the graph). Consequently, by the result of [11], the graph contains a dominating set of constant cardinality O(k log k). By construction of the graph, choosing one corresponding solution x for each image in this dominating set yields an R ′ -dominating set for X ′ . □ In the remainder of this section, we show that the existence results for one-exact (1 + ε)-approximation sets and quasi-k-exact (1 + ε)-approximation sets with k = ⌈ p 2 ⌉ from Theorem 3.2 are tight in the sense that polynomialcardinality (1 + ε)-approximation sets with additional exact components do not exist in general. This is made precise for the case of one-exact (1 + ε)approximation sets in the following theorem. Theorem 3.3 There exist instances of many p-objective combinatorial optimization problems, including the multiobjective versions of shortest path, minimum spanning tree, assignment, minimum s-t-cut, knapsack, and traveling salesman, that do not admit a polynomial-cardinality one-exact, quasi-2-exact (1 + ε)-approximation set for any ε > 0.
Proof The biobjective versions of all of these problems are known to be intractable, i.e., there exist instances for which the cardinality of the nondominated set is exponential in the instance size. From such an intractable biobjective instance, construct a p-objective instance where objective f 1 corresponds to the first objective function of the biobjective instance and the other p − 1 objective functions correspond to the second objective function of the biobjective instance. Then, at least one solution corresponding to each nondominated point of the biobjective instance is required in any one-exact, quasi-2-exact (1 + ε)-approximation set of the p-objective instance.
□ Figure 1 Existence of quasi-k-exact (1 + ε)-approximation sets of fully polynomial cardinality. For each weakly nondominated nonempty hyperrectangle in the objective space as in the proof of Theorem 3.1, a dominating set for the supergraph of a k-majority tournament defined on the images in the hyperrectangle is selected to obtain a quasi-k-exact (1 + ε)-approximation set. There exists at most one weakly nondominated nonempty hyperrectangle in each diagonal of the grid. A quasi-k-exact (1 + ε)-approximation set is given by preimages of the bold points. Weakly nondominated nonempty hyperrectangles are indicated by shaded boxes. One diagonal is indicated by the bold lines. The constructed supergraph of a k-majority tournament is illustrated for one hyperrectangle.
The above theorem shows that, for many important problems, one-exact (1 + ε)-approximation sets are the best polynomial-cardinality partially exact approximation sets one can hope for when requiring to be exact in one fixed objective. In fact, we are not aware of many (non-trivial) problems that admit polynomial-cardinality one-exact, quasi-2-exact (1 + ε)-approximation sets. One notable exception is the biobjective minimum cut problem, for which it is known that the cardinality of the efficient set is polynomial in the instance size [12]. Since the efficient set corresponds to a two-exact (1 + ε)approximation set in the biobjective case, this in particular shows the existence of a polynomial-cardinality one-exact, quasi-2-exact (1 + ε)-approximation set.
We now consider quasi-k-exact (1 + ε)-approximation sets, that is, approximation sets that approximate each feasible solution exactly in k objectives (and with a factor of (1 + ε)-approximate in all other objectives), but allow the objectives in which the approximation is exact to differ depending on the approximated solution. Here, a similar argument as in the proof of Theorem 3.3 shows that, for many multiobjective combinatorial optimization problems, polynomial-cardinality quasi-k-exact (1 + ε)-approximation sets for k > ⌈ p 2 ⌉ are not guaranteed to exist for all instances.
Theorem 3.4 There exist instances of many p-objective combinatorial optimization problems, including the multiobjective versions of shortest path, minimum spanning tree, assignment, minimum s-t-cut, knapsack, and traveling salesman, that do not admit a polynomial-cardinality quasi-k-exact (1 + ε)-approximation set for any k > ⌈ p 2 ⌉ and any ε > 0.
Proof As noted in the proof of Theorem 3.3, the biobjective versions of all of these problems are known to be intractable. From such an intractable biobjective instance, construct a p-objective instance where the first ⌈ p 2 ⌉ objectives corresponds to the first objective of the biobjective instance and the other ⌊ p 2 ⌋ objectives correspond to the second objective function of the biobjective instance. Then, for k > ⌈ p 2 ⌉, at least one solution corresponding to each nondominated point of the biobjective instance is required in any quasi-k-exact (1 + ε)-approximation set of the p-objective instance. □ Figure 2 summarizes the known tractability and intractability results for different types of approximation sets.
(1 + ε)-approximation set An arrow indicates that one type of approximation set fulfills the requirements of the other. The dashed line marks the boundary between tractability and intractability in general pobjective optimization problems. New results obtained in this paper are shown in bold. The result on (1 + ε)-approximation sets is shown in [13], and the results on one-exact and twoexact (1 + ε)-approximation sets are shown in [9].
Minimum-Cardinality Approximation Sets
Theorem 3.2 implies that an asymptotic upper bound for the minimum cardinality of an ordinary, a one-exact, and a quasi-k-exact (1 + ε)-approximation and it is easy to see that this bound is tight, i.e., there exist instances where even a minimum-cardinality ordinary (1 + ε)-approximation set consists of Θ M ε p−1 solutions. For a specific instance, however, as pointed out in [6,8], there might still exist (1 + ε)approximation sets of vastly different cardinalities. This motivates to study properties of (1 + ε)-approximation sets of minimum cardinality. Moreover, it raises the question whether the minimum possible cardinality can increase for a specific instance when considering partially exact (1 + ε)-approximation sets. As shown in [9], this is indeed the case when considering one-exact (1 + ε)approximation sets: there exist instances for which the minimum cardinality of a one-exact (1 + ε)-approximation set is an arbitrary factor larger than the minimum cardinality of an ordinary (1 + ε)-approximation set.
Hence, in this section, we systematically study minimum-cardinality ordinary and partially exact (1 + ε)-approximation sets. In Subsection 4.1, we first investigate the (weak) efficiency of solutions in minimum-cardinality approximation sets and show that minimum-cardinality ordinary (1+ε)-approximation sets and quasi-1-exact (1 + ε)-approximation sets can consist of dominated solutions only. For minimum-cardinality one-exact (1 + ε)-approximation sets, however, we show that one contained solution must be weakly efficient, but not more than one in general. Afterwards, in Subsection 4.2, we relate the minimum cardinalities of different types of partially exact approximation sets to the minimum cardinality of an ordinary (1 + ε)-approximation set. Here, we show that requiring the (1 + ε)-approximation set to be quasi-1-exact never changes the minimum cardinality for any instance. In contrast, requiring to be one-exact or quasi-k-exact for any k ≥ 2 can increase the minimum cardinality by an arbitrarily large factor.
Efficiency of Minimum-Cardinality Approximation Sets
In this subsection, we investigate the question of (weak) efficiency of solutions in minimum-cardinality ordinary and partially exact (1 + ε)-approximation sets.
Since any dominated solution in any type of partially exact or ordinary (1+ ε)-approximation set can always be replaced by an efficient solution dominating it without impairing the obtained approximation guarantee, it follows that there always exist minimum-cardinality (1+ε)-approximation sets of each type that consist of efficient solutions only. In contrast, it is well known that, even in the biobjective case, there are ordinary (1+ε)-approximation sets of minimum cardinality that consist of strictly dominated solutions only. The following proposition generalizes this result to quasi-1-exact (1 + ε)-approximation sets.
For one-exact (1 + ε)-approximation sets, the definition implies that each such set must contain at least one weakly efficient solution, namely a solution with minimum f 1 -value. The following result, however, shows that, even in the biobjective case, there are minimum-cardinality one-exact (1 + ε)-approximation sets in which all other solutions are strictly dominated. Proposition 4.2 For any ε > 0 and any positive integer n ∈ N + , there exists an instance of a biobjective optimization problem that admits a one-exact (1 + ε)approximation set of minimum cardinality n + 1 in which all but one of the solutions are strictly dominated.
Proof Given ε > 0 and n ∈ N + , let δ > 0 so that (1 + δ) 2n = 1 + ε and consider the following instance of a biobjective optimization problem: The feasible set is given as X := {x 0 , . . . , x n ,x 1 , . . . ,x n ,x 0 , . . . ,x n } and We now show that {x 0 , x 1 , . . . , x n } is a minimum-cardinality one-exact (1 + ε)approximation set, even though x 0 is the only solution in this set that is not strictly dominated. To this end, first note that Thus, x 0 is not ⪯ 1 ε -dominated by any other solution, so x 0 must be contained in every one-exact (1 + ε)-approximation set. Moreover, the following statements hold: (2) and (3)), (1) and (3)), and, therefore, also In total, this implies that {x 0 , x 1 , . . . , x n } is a one-exact (1 + ε)approximation set, even though x 0 is the only solution in this set that is not strictly dominated. Moreover, for each i ∈ {1, . . . , n}, in order to ⪯ 1 ε -dominatex i , at least one of the three solutions x i ,x i , andx i must be contained in any one-exact (1 + ε)approximation set. Consequently, any one-exact (1 + ε)-approximation set has cardinality at least n + 1. □ The final result of this subsection, which will be used when analyzing the minimum cardinality of quasi-1-exact (1 + ε)-approximation sets in the following subsection, states that any (1 + ε)-approximation set consisting only of weakly efficient solutions must actually be quasi-1-exact. Proof Let P be a (1+ε)-approximation set in which all solutions are weakly efficient. Let x ′ ∈ X be an arbitrary feasible solution and let x ∈ P be a solution that ⪯ εdominates x ′ . If x ′ is not ⪯ (1) ε -dominated by x, i.e., we do not have f i (x) ≤ f i (x ′ ) for any i ∈ {1, . . . , p}, we immediately obtain that x is strictly dominated by x ′ , which contradicts the weak efficiency of x. □
Relations Between Minimum Cardinalities
We now relate the minimum cardinalities of different types of partially exact approximation sets to the minimum cardinality of an ordinary (1 + ε)-approximation set. Concerning one-exact (1 + ε)-approximation sets, it is shown in [9] that, even for biobjective problems, the minimum cardinality of a such a set can be an arbitrary factor larger than the minimum cardinality of an ordinary (1 + ε)-approximation set for certain instances. In contrast to this, we now use Proposition 4.3 to show that requiring a quasi-1-exact (1 + ε)-approximation set instead of only an ordinary (1 + ε)-approximation set never changes the minimum cardinality for any instance. Theorem 4.4 For any instance of a multiobjective optimization problem and any ε > 0, there exists a minimum-cardinality ordinary (1 + ε)-approximation set that is quasi-1-exact. In particular, the minimum cardinalities of an ordinary (1 + ε)approximation set and of a quasi-1-exact (1 + ε)-approximation set coincide for every instance.
Proof Let P * be a smallest ordinary (1 + ε)-approximation set. Turn P * into a quasi-1-exact (1 + ε)-approximation set P ′ of the same cardinality that consists of weakly efficient solutions only by replacing any strictly dominated solution x ∈ P * by a weakly efficient solution that strictly dominates it (this does not impair the approximation guarantee). By Proposition 4.3, the set P ′ is then indeed a quasi-1exact (1 + ε)-approximation set. □ Contrasting the result on the minimum cardinality of quasi-1-exact (1 + ε)approximation sets shown in Theorem 4.4, we now show that, for k ≥ 2, the minimum cardinality of a quasi-k-exact (1 + ε)-approximation set can be an arbitrary factor larger than the minimum cardinality of an ordinary (1 + ε)approximation set -even in the case of three objectives (p = 3). The proof of the following theorem generalizes the construction used in the proof of Theorem 2 in [9]. Theorem 4.5 For any ε > 0 and any positive integer n ∈ N + , there exist instances of 3-objective optimization problems such that |P (2), * | > n·|P * |, where P (2), * denotes a minimum-cardinality quasi-2-exact (1 + ε)-approximation set and P * denotes a minimum-cardinality ordinary (1 + ε)-approximation set.
Computation of Approximation Sets
After studying the existence of (fully) polynomial-cardinality approximation sets and the impact of using partially exact approximate dominance relations on the minimum cardinality of approximation sets, we now consider the computability of such sets. As shown in [4], a (1+ε)-approximation set can be computed in (fully) polynomial time for every ε > 0 if and only if the auxiliary problem Gap δ (b) can be solved in (fully) polynomial time for every vector b and every δ > 0. Hence, a natural question is whether (fully) polynomial-time solvability of Gap δ (b) is also sufficient for the (fully) polynomial-time computability of partially exact (1 + ε)-approximation sets. The following theorem gives a negative answer to this question even for biobjective problems and the least-demanding type of partially exact (1 + ε)-approximation sets. The proof is partly similar to the proof of Theorem 1 in [8], where it is shown that polynomial-time algorithms only generating solutions by solving Gap δ (b) cannot approximate the minimum cardinality of an ordinary (1 + ε)-approximation set to a factor better than 3 in the biobjective case (such algorithms are called generic in [8]). Proof Given some (rational) ε > 0, consider two instances I 1 , I 2 with feasible sets X I1 = {x 1 } and X I2 = {x 1 , x 2 } and the objective function values given by where l(ε) is a positive integer that is exponential in the input size and in 1 ε . We show that no algorithm that generates feasible solutions only by solving Gap δ (b) for values δ ≥ 1 l(ε) can distinguish between the two instances (i.e., detect whether x 2 is part of the feasible set). Since any smaller value of δ would be exponential in the input size and in 1 ε by definition of l(ε), and any quasi-1-exact (1 + ε)-approximation set for instance I 2 must include x 2 , this will show the claim.
So consider a call of Gap δ (b) for some point b and some δ ≥ 1 l(ε) . We distinguish two cases in order to show that Gap δ (b) can always return either x 1 or NO as a correct answer for both instance I 1 and instance I 2 . If b i ≥ f i (x 1 ) for i = 1, 2, then Gap δ (b) can return x 1 for both instances. Otherwise, we have b j < f j (x 1 ) = 1 + 1 l(ε) for some j ∈ {1, 2}. Using that δ ≥ 1 l(ε) , this implies that Consequently, for both instances, there exists no solution with j-th objective value less than or equal to bj 1+δ , so Gap δ (b) can return NO. This shows that an algorithm that generates feasible solutions only by solving Gap δ (b) for values of δ ≥ 1 l(ε) cannot detect whether x 2 is part of the feasible set as claimed. □ Theorem 5.1 shows that harder-to-solve auxiliary problems than the gap problem must be used for computing partially exact (1+ε)-approximation sets in polynomial time. We now use Proposition 4.3 to show that, for biobjective problems, several known algorithms for computing (small-cardinality) (1 + ε)-approximation sets based on such auxiliary problems actually yield quasi-1-exact (1 + ε)-approximation sets.
Theorem 5.2 1) For biobjective optimization problems for which
Constrained 1 and Constrained 2 are solvable in (fully) polynomial time, a quasi-1-exact (1 + ε)-approximation set with minimum cardinality can be computed in (fully) polynomial time for every instance. 2) For biobjective optimization problems for which DualRestrict 1 δ or DualRestrict 2 δ is solvable in (fully) polynomial time, a quasi-1-exact (1 + ε)-approximation set with cardinality at most twice the cardinality of a smallest quasi-1-exact (1 + ε)-approximation set can be computed in (fully) polynomial time for every instance.
Proof 1) If Constrained 1 and Constrained 2 are solvable in (fully) polynomial time, an ordinary (1 + ε)-approximation set with with minimum cardinality can be computed in (fully) polynomial time for every instance by a greedy procedure as shown in [6]. Since this set contains only weakly efficient solutions [14], it is actually a quasi-1-exact (1 + ε)-approximation set of minimum cardinality by Proposition 4.3. 2) If DualRestrict 1 δ or DualRestrict 2 δ is solvable in (fully) polynomial time, an ordinary (1 + ε)-approximation set with cardinality at most twice the cardinality of a smallest ordinary (1 + ε)-approximation set (which equals the minimum cardinality of a quasi-1-exact (1+ε)-approximation set by Theorem 4.4) can be computed in (fully) polynomial time for every instance as shown in [6,14]. Moreover, the (1+ε)-approximation sets returned by the algorithms in [6,14] contain only weakly efficient solutions [14]. Consequently, by Proposition 4.3, these sets are actually quasi-1-exact (1 + ε)-approximation sets. □ We remark that, as shown in [9], the same results can be achieved even for one-exact (1 + ε)-approximation sets, but modified algorithms targeted specifically at the computation of one-exact (1 + ε)-approximation sets are necessary in this case. Moreover, it has been shown in [9] that, for general pobjective problems, one-exact (1 + ε)-approximation sets can be computed in (fully) polynomial time if and only if DualRestrict 1 δ (b 2 , . . . , b p ) can be solved for any choice of the bounds b 2 , . . . , b p and any δ > 0 in (fully) polynomial time.
We finish this section by considering the case where all feasible solutions are given explicitly in the input. For this case, we now present two related approaches for the polynomial-time computation of R-approximation sets for approximate dominance relations R as in Theorem 3.1 under the assumption that checking whether a given solution x ∈ X R ′ -dominates another given solution x ′ ∈ X (for R ′ as in the theorem) is possible in polynomial time. This class of relations, in particular, includes quasi-k-exact (1+ε)-dominance for any k ≤ ⌈ p 2 ⌉ as well as one-exact (1+ε)-dominance and ordinary (1+ε)-dominance. The first approach is based directly on the proof of Theorem 3.1. Given that checking whether a given solution x ∈ X R ′ -dominates another given solution x ′ ∈ X (for R ′ as in the theorem) is possible in polynomial time, we can generate the directed graph corresponding to the approximate dominance relation R ′ in each weakly nondominated nonempty hyperrectangle considered in the proof in polynomial time and compute a dominating set in each of these graphs. While computing a minimum-cardinality dominating set is NP-hard in general, a simple greedy algorithm can be used to compute a dominating set that is larger by at most a factor 1 + log|V |, where V denotes the vertex set [15]. This yields a polynomial-cardinality set in our situation since |V | is bounded by |X| in all cases, which is polynomial in the input size since X is given explicitly in the input.
A second related approach, which additionally yields a bound on the cardinality of the obtained R-approximation set, consists of constructing the directed graph on X corresponding to the approximate dominance relation R (which is possible in polynomial time given that R ′ -dominance of solutions can be checked in polynomial time) and to compute a dominating set of cardinality at most 1 + log|X| times the minimum cardinality of dominating set in this graph using the algorithm from [15]. The resulting set then corresponds to an R-approximation set of cardinality at most 1 + log|X| times the minimum cardinality of an R-approximation set.
Conclusion
In this paper, we explore the borderline between tractability and intractability when approximating multiobjective optimization problems using more demanding approximate dominance relations than (1 + ε)-dominance. We show that, under very weak assumptions, there always exist (1 + ε)-approximation sets of fully polynomial cardinality such that every feasible solution is, in fact, 1-approximated with respect to at least half of the objectives functions if we allow these objective functions to differ between different feasible solutions. If a polynomial-cardinality approximation set is required to be exact in one specified objective function, however, exactness cannot be ensured in a second objective function in general, even if this second objective function is allowed to differ depending on the approximated solution. This leads to the "frontier of intractability" being located in between one-exact (1+ε)-approximation sets and one-exact, quasi-2-exact (1 + ε)-approximation sets and in between quasi-⌈ p 2 ⌉exact (1+ε)-approximation sets and quasi-(⌈ p 2 ⌉+1)-exact (1+ε)-approximation sets.
While first positive and negative results concerning the polynomial-time computability of quasi-k-exact (1 + ε)-approximation sets have been obtained in Section 5, an important question for future research is to obtain further computability results that extend beyond the biobjective case or the situation where the feasible set is given explicitly in the input. Specifically, since both the polynomial-time computability of ordinary (1 + ε)-approximation sets and of one-exact (1 + ε)-approximation sets have been characterized by the polynomial-time solvability of certain auxiliary problems (Gap and DualRestrict, respectively), it would be interesting to investigate whether a similar characterization via an auxiliary problem can also be obtained for quasi-k-exact (1 + ε)-approximation sets for k ≤ ⌈ p 2 ⌉. As our results shown, such an auxiliary problem would have to be strictly harder to solve than Gap. | 2023-05-25T01:16:19.910Z | 2023-05-24T00:00:00.000 | {
"year": 2023,
"sha1": "054dd03613329bc586c64298f53160fdda45fc83",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "054dd03613329bc586c64298f53160fdda45fc83",
"s2fieldsofstudy": [
"Mathematics",
"Computer Science"
],
"extfieldsofstudy": [
"Computer Science",
"Mathematics"
]
} |
69989308 | pes2o/s2orc | v3-fos-license | Optimization of Automobile Suspension System Using Hybrid GSA Algorithm
In this paper Hybrid Genetic- Simulated Annealing Algorithm (GSA) is studied and applied to the suspension system of an automobile with two different case studies. Mathematical modelling of an automobile is done for passive suspension system. GA, SA and GSA were simulated using programming in MATLAB. First case study is based on the data of researcher in the field where GA was applied for optimization of automobile suspension system. Here SA, as well as Hybrid GSA are applied and compared with the results of GA. Convergences of the three methods are also plotted. In second case study an automobile model MARUTI 800 is considered. All the necessary parameters are calculated through measurements. The results shows required seat acceleration for current available springs of the automobile in field. For both the case studies the Hybrid GSA shows best results among the three. This can be performed before actual manufacturing of springs and automobile to check whether desired seat acceleration and ride comfort will achieve or not. If it is satisfied with the value of seat acceleration, decision would be made regarding manufacturing of springs with those values of stiffness and damping constants.
Introduction
The Suspension system of an automobile is often a most important part of it, due to its safety features, shock absorption and stability during running conditions. It is basically the combination of springs, shock absorbers and mechanism that connects the body of the Automobile to the chassis. The suspension system can be active or passive. Active suspension system has components which needs power for their working operation dislike passive suspension. Now a days a semi-active suspension system have been used which is combination of active and passive suspension system. Genetic algorithm (GA) is the optimization technique which imitate the process of natural genetics [1,2] . GA initially considers the population of each parameter from its history and selects two parameters for its processing which includes mutation, crossover where the feasibility of the parameters is checked for their survival in the field of application. Some researchers [2] have explain the Genetic algorithm optimization and applied it to the suspension system of an automobile in view of optimum seat acceleration. Mathematical modelling were done based on equivalent system of an automobile. The parameters were optimized for their best utilization. Few investigators have used GA for solving the poor linkage version of the problem where the order of genes is randomized, and crossover is unable to extract and recombine good building block [3] . The GA can be applied to various system by changing the pattern of genes. Simulated annealing (SA) is the optimization algorithm which is based on the way a metal get heat treatment by the process of annealing through heating at certain temperature and slowly cool down in environment to relieve the internal stresses [4] . With the reference to previous experience of the material highest possible set point is decided and named as initial temperature and then neibourhood point will be checked more precise and optimum value of an output. Optimization using GA works on the globalization of the parameters whereas it is using SA by localization of the parameters, this inspires to combine the two concepts to get the more effective solution. Several attempts have been made to hybridize the GA and SA and proven to be good. Balram Suman had used four different Simulated Annealing algorithm to solve multiobjective optimization problem [4] . Sanjay Bisht gives a conceptual basis for Hybrid Genetic and simulated annealing optimization applied in the field of Defense [5] . Younis Elhaddad et al. had applied the Hybrid Genetic and Simulated Algorithm to Traveling salesman problem [6] . A.H. Gandomi et al. gives a concept of Hybrid Optimization applied to study the Behavior of Bolted joint [7] . M.J. Thoresson et al. had proposed a methodology for the efficient determination of gradient information, when performing gradient based optimization of an off-road vehicle's suspension system [10] . M. Yogeswarana et al. discussed a machine loading problem in a flexible manufacturing system (FMS), with bi-criterion objectives of minimizing system imbalance and maximizing system throughput in the occurrence of technological constraints such as available machining time and tool slots [11] . M. Lozano et al. had explain that the flexible architecture of evolutionary algorithms allows specialized models to be obtained with the aim of performing as other search methods do, but more satisfactorily [12] . Suspension system of an automobile is the combination of different springs and their location with respect to mass center of an automobile, vibration absorbing dampers. The important Parameters used to design an automobile suspension system are: 1. Springs stiffness. 2. Deflection of the springs. 3. Placement of spring relative to C.G of the automobile. 4. Vibration reducing units such as Dampers. 5. Amplitude of vibration of tires due to contact with road surface. These design variables have linked with ride comfort. It is tried here that these Design parameters shall be selected for minimum seat acceleration i.e. maximum ride comfort. In this paper an equivalent mathematical model of an automobile is considered with specified design parameters. Considering the variation in the values of the parameters and other dimensions of the automobile a Programming is done in MATLAB for GA, SA and GSA for two case studies separately. To validate the results a similar automobile model is developed in MSC.ADAMS (a dynamic analysis software package) with the obtain values of spring stiffness and damping constants from programming.
Mathematical Modelling
The equivalent model of an automobile is developed considering the automobile suspension system with 5 degrees of freedom which represents the independent generalized co-ordinates with following co-ordinates q = [q1, q2, q3, q4, q5, qA, qB] The suspension of an automobile is controlled by the force vibration equation: Where, ̈ݍ is the Acceleration of seat which decides the ride comfort of an automobile, =̇ݍ Deflection of the springs with time, q= Deflection of springs. This equation is being apply for each Wheel in terms of all parameters and rearrange it to get .̇ݍ For analyzing any suspension system above said equation is to be solved at each point of consideration. Ride comfort is chosen to be the criteria for optimization of suspension system, which depends on the amplitude of Seat acceleration and is expressed as,
1234567890''""
International Assuming that all dampers and springs behave linearly, ̈ݍ Seat acceleration, is to be optimized by selecting other parameters in that equation. Elaboration of the same can be shown as, With dimensions a=2.03m, b=0.25m and d=0.76m and Vehicle velocity V=25 m/s. Maximum value of seat acceleration cannot exceed the acceleration due to gravity, since it violate the principles of gravity. Constraint 1: g1= f -10 m/s 2 ≦0 -----(5) Following are some constraints that are placed on the system so that it's movement will be restricted, so that the tire deflections and relative spaces between bodies are restricted by, Constraint 5: g5= ቚq 2, (t)-q 3 (t)+aq 5 (t)ቚ ≤ 0.12m -----(9) Constraint 6: g6=หq 5 (t)-q 4 (t)-(b+d)q 5 (t)-q 4 (t)ห ≤ 0.12m.
----(10) 0.050 meter and 0.12 meter are the restrictions placed on the difference of respective deflections of the automobile considered whose values may be change when automobile model is different.
Case Study I
In this case study data used by A.E. Baumal [2] et al. is considered. Dimensions of automobile Upper bound and lower bound on design variables, probability of mutation, probability of crossover, mass of automobile and it's mass moment of inertia etc. all the values are used by them. As this analysis requires large computation work that could not be done manually, which is done here using a MATLAB program. The researchers did optimization of suspension system by Genetic Algorithm. Here it is done by Genetic Algorithm, Simulated Annealing Algorithm and Hybrid Genetic and Simulated Annealing Algorithm (GSA). Results have been validated by literature as well as MSC.ADAMS.
Case Study II
An automobile MARUTI 800 is consider for the application of the hybrid optimization technique.
Case study 1
Case study 1 is based on the data from literature, therefore results obtain through current model should come close to the literature, so as to validate them. Also to validate the results an equivalent model of an automobile is developed in dynamic analysis software package MSC.ADAMS. Results obtain from the dynamic analysis are compared with those obtain through analytical method in MATLAB. The number of population is taken as 500 and number of generation is taken as 500. These number of population and number of generation can be varied for different Run. Thousands of values for both number of population (NP) and number of generation (NG) can be used. Minimum Seat acceleration () is obtain from Hybrid GSA algorithms, which can be seen from table 1. It is also said that, to get optimum ride comfort i.e. Minimum seat acceleration, springs with stiffness K1, K2 and K3 and dampers with damping constant C1, C2 and C3 should be selected. Changes can be made in the number of population (NP), number of generation (NG), probability of mutation and probability of crossover as per our priority. Running the MATLAB program number of times it is come to know that the seat acceleration is minimum for the probability of crossover 0.7 and probability of mutation 0.015. Genetic algorithm is heuristic search technique which works globally to find the minimum value of certain objective. Simulated annealing is also an heuristic search technique but it works locally rather that global like GA. Search of certain goal can be made optimum by combining these two method to Get added benefit of the two. The number of iteration is taken as 1000. This number of iteration can be varied for different Run. Seat acceleration is minimum for Run 3. It can be said that to get minimum seat acceleration, take springs with stiffness K1, K2 and K3 and Dampers with damping constant C1, C2 and C3. In above shown program NG is taken as 100, NP is taken as 100 and number of iteration is taken as 100. This NP, NG and NI can be varied for different Run. Thousand of value for these NP, NG and NI can be taken.
Case study 2
From the results it is seen that seat acceleration is obtain minimum from Hybrid GSA algorithm. As compare to GA and SA, GSA gives much more desirable minimum objective function. So combining GA and SA the result is improved. From above results it can be said that hybridizing two algorithms proves better than the individuals. We can change the values of number of population and number of generation to get different run. On combining the two techniques a new method evolves i.e. Hybrid Genetic-Simulated annealing algorithm. These all statements are come true after looking the results tabulated below. Hybrid
Conclusion
Hybrid GSA algorithm is systematically applied to the suspension system of automobile through two different cases. Following conclusions can be sorted: 1. As the Hybrid GSA is integration of two different algorithm, it not only combines their steps but also improves their results. It is seen through both the case studies that the most suitable ride comfort is achieved using Hybrid GSA. 2. A suitable combination of design variables is obtain for maximum ride comfort. 3. MATLAB program can be used to apply it on any automobile suspension system, as here done for MARUTI 800. 4. The suspension system of an automobile can be design considering the obtain parameters combination for desired ride comfort. | 2019-02-19T14:06:28.441Z | 2018-06-01T00:00:00.000 | {
"year": 2018,
"sha1": "a49885d5bda9d5a4615bbb725ea60f1d6f2ede82",
"oa_license": null,
"oa_url": "https://doi.org/10.1088/1757-899x/377/1/012149",
"oa_status": "GOLD",
"pdf_src": "IOP",
"pdf_hash": "d987d0311bc4c9efae0774375d9ea4865e8e2db2",
"s2fieldsofstudy": [
"Engineering"
],
"extfieldsofstudy": [
"Physics",
"Computer Science"
]
} |
259899387 | pes2o/s2orc | v3-fos-license | Using PIM-Taiwan, PRISCUS, and Beers criteria to assess potentially inappropriate medication use among older adults with 90-day rehospitalization: a population-based study in Taiwan
Background: Multimorbidity and polypharmacy increase the risk of hospitalization in older adults receiving potentially inappropriate medication (PIM). The current study compared the ability of PIM-Taiwan, PRISCUS, and Beers criteria to predict 90-day rehospitalization in older patients with and without PIM. Methods: The retrospective cohort study used Taiwan’s Longitudinal Health Insurance Database to retrieve quarterly information about prescribed medication for adults aged ≥65 years hospitalized between 2001 and 2018. We analyzed the association of PIM with 90-day rehospitalization using logistic regression. Results: The study cohort included 206,058 older adults (mean age: 72.5 years). In the analysis, 133,201 (64.6%), 97,790 (47.5%), and 147,450 (71.6%), were identified as having PIM exposure in PIM-Taiwan, PRICUS, and Beers criteria, respectively. PIM-Taiwan criteria found exposure to PIM affecting the cardiovascular (adjusted OR [aOR] 1.37, 95% confidence interval [CI] = 1.32–1.41), gastrointestinal (aOR 1.26, 95% CI = 1.23–1.30), central nervous (aOR 1.11, 95% CI = 1.08–1.14), and respiratory (aOR 1.16, 95% CI = 1.12–1.20) systems significantly increased the risk of 90-day rehospitalization, after adjustment for covariates. In PRISCUS criteria, exposure to PIM affecting the respiratory (aOR 1.48, 95% CI = 1.41–1.56), central nervous (aOR 1.12, 95% CI = 1.09–1.15), and cardiovascular (aOR 1.20, 95% CI = 1.16–1.24) systems significantly increased the risk. In Beers criteria, exposure to PIM affecting the cardiovascular (aOR 1.37, 95% CI = 1.32–1.41), gastrointestinal (aOR 1.38, 95% CI = 1.35–1.42), central nervous (aOR 1.18, 95% CI = 1.15–1.21), endocrine (aOR 1.10, 95% CI = 1.06–1.15), and respiratory (aOR 1.09, 95% CI = 1.04–1.13) systems significantly increased the risk. Patients with 90-day rehospitalization had higher rates of the potentially harmful drug-drug interaction (DDI) pairs of serotonin syndrome (n = 19; 48.8%), QT prolongation (n = 4; 30.8%), extrapyramidal symptoms (EPS) (n = 102; 24.5%), and hypokalemia (n = 275; 20.1%). Conclusion: Beers criteria was more efficient in predicting 90-day rehospitalization among older adults experiencing PIM in Taiwan than either PIM-Taiwan or PRISCUS. The risk of 90-day rehospitalization was associated with the potentially harmful DDI classes of serotonin syndrome, QT prolongation, EPS, and hypokalemia.
Introduction
World population aging is growing rapidly in both developed and developing countries. In Taiwan, the proportion of the older adults increased from 7% in 1993 (an aging country) to 14% in 2018 (an aged country), and is predicted to reach 20% in 2025 (a superaged country) (National Development Council, 2023). This accelerated aging brings a heavy burden of healthcare for older adults in Taiwan (Chen, 2010). In addition, multimorbidity and polypharmacy among older adults are associated with high rates of unplanned hospitalization, particularly for adverse drug reactions (ADRs) due to age-related alterations in pharmacokinetics, pharmacodynamics, and drug-disease interactions (Leendertse et al., 2008;Olivier et al., 2009;Hamilton et al., 2011;Dalleur et al., 2012). Therefore, adequate drug treatment is paramount to forestall preventable hospitalizations in older adults, particularly the frail and those with multiple morbidities.
Potentially inappropriate medication (PIM) use is not a rare event in older adults, and is associated in this population with ADRs, hospitalization, functional decline, and even death (Lau et al., 2005;Passarelli et al., 2005;Lin et al., 2008;Bradley et al., 2012). For this reason, many countries have built their own criteria systems for PIM, because national pharmacotherapeutic guidelines vary in terms of the specific drugs approved (Chang et al., 2014). The different criteria may lead to wide variations in reported PIM prevalence and their associated health-related outcomes (Chang et al., 2014). Of all PIM criteria the Beers criteria, first established in 1991, is the most widely used to detect PIM in older adults as an indicator of geriatric healthcare quality (Wenger and Shekelle, 2001;Pugh et al., 2006;Chang et al., 2011;Dimitrow et al., 2011;Corsonello et al., 2012). In Taiwan, the PIM-Taiwan, introduced in 2012, outlines the relevant and country-specific PIM criteria (Chang et al., 2012;Chang et al., 2019). The PIM-Taiwan criteria have proven their applicability in several cross-sectional studies among older Taiwanese adults (Chang et al., 2014;Chang et al., 2015;Chang et al., 2018). In comparison with the Beers criteria and PRISCUS criteria, PIM-Taiwan can detect a similar number of PIMs across different populations in Taiwan. PIM users had higher health resource utilization and higher costs of medications than non-PIM users (Chang et al., 2014;Chang et al., 2015). Therefore, PIM-Taiwan criteria can be an important tool to reduce PIM-related adverse events in older Taiwanese adults. In Europe, the PRISCUS criteria, based on expert knowledge developed in Germany, helps physicians make individualized therapeutic decisions for their patients (Holt et al., 2010;Fromm et al., 2013;Schubert et al., 2013). Most studies related to PIM criteria have focused on the prevalence of PIM in limited smallto-medium-sized samples, such as inpatient populations or nursing home residents (Reich et al., 2014;Garcia et al., 2015). However, few studies have assessed the association of PIM with health-related outcomes using nationwide samples (Mekonnen et al., 2021).
Based on National Health Insurance (NHI) claims data, the current study was conducted to estimate the prevalence and changes in PIM among older patients before and after a hospitalization as measured by PIM-Taiwan, PRISCUS, and Beers criteria. We sought to investigate the association of PIM via these criteria systems with the risk of 90-day rehospitalization in this population. Furthermore, we also analyzed the risk of hospitalization associated with different classes of potentially harmful drug-drug interactions (DDIs).
Data source and study population
The retrospective cohort study was performed using claimbased data from the 2010 Longitudinal Generation Tracking Database (LGTD 2010), the Catastrophic Illness Registry Dataset (CIRD), the Taiwan Cancer Registry (TCR) database, and Multiple Causes of Death data from 2000 to 2018. Detailed descriptions of the aforementioned 3-dataset sample and study procedures have previously been published and the representativeness of LGTD and TCR has been validated (Chang et al., 2015;Chiang et al., 2015;Hu et al., 2016;Chiang et al., 2017;Center for Biomedical Resources of NHRI, 2023). The LHTG 2010 is a subset of the National Health Insurance Research Database (NHIRD) composed of 2 million randomly-sampled beneficiaries (nearly 10% of the total Taiwanese population) drawn in 2010. The NHIRD was developed and managed by Taiwan's National Insurance program, which was introduced in 1995 and provides for approximately 99.9% of Taiwanese residents. The LGTD 2010 includes comprehensive claims data for reimbursements for ambulatory, inpatient, emergency, and Chinese medicine visits, and is used to gather information on prescriptions and comorbidities (Center for Biomedical Resources of NHRI; Chiang et al., 2017;Hu et al., 2016). The CIRD contains information about catastrophic illness in patients who suffered from at least one of 30 specific severe medical conditions such as cancer, chronic mental illness (schizophrenia, bipolar I disorder, major depressive disorder), hemodialysis, systemic autoimmune diseases, and stroke (Chang Frontiers in Pharmacology frontiersin.org 02 (Chang et al., 2015). The Multiple Causes of Death data provides information on the cause(s) and date of death. We enrolled patients aged 65 years or older at the first hospitalization from 2001 to 2018 in Taiwan, excluding those with a history of concomitant cancer or catastrophic illness based on the CIRD and TCR records. Patients with at least one hospitalization recorded during the 2001-2018 study period were considered for the first hospitalization. We followed those patients from the first hospitalization until 90 days after discharge or until rehospitalization within this period.
Measurements Definition of study outcome
The study outcome of 90-day rehospitalization was defined as one of the following events within 90 days after the first discharge: readmittance to the hospital, emergency department visit, hospitalization visa in the emergency department, or death.
Potentially inappropriate medication (PIM) exposure
We assessed medication use within 30 days before the first hospitalization (prehospitalization) and within 30 days after the first discharge (post-discharge). Postdischarge was defined as the 30 days following the discharge date, or the date of readmittance within 30 days after discharge, whichever came first. The medications used were classified as PIM or not PIM according to three criteria systems: the updated 2018 PIM-Taiwan criteria (Chang et al., 2018), the updated 2019 Beers criteria (By the American Geriatrics Society Beers Criteria ® Update Expert Panel, 2019), and the entire PRISCUS criteria (Holt et al., 2010) (Supplementary Tables S1-S3) ( Table 1). Exposure to PIM was considered if the prescribed drug was represented in any of the criteria, and PIM items were also counted. PIM was identified in the database via Anatomical Therapeutic Chemical code and further classified by human body systems, including cardiovascular, endocrine, gastrointestinal, genitourinary, musculoskeletal, central nervous, and respiratory systems; and the additional classifications (not used in all criteria systems) of sex hormones, anti-infective, and blood and bloodforming organs.
Drug-drug interactions (DDIs)
The DDI pairs were taken from those listed by McEvoy et al. (2017) and the National Institute of Health and Care Excellence (NICE) clinical guidelines (Chang et al., 2015). The DDI pairs in Micromedex ® provided information on the severity of contraindication; UptoDate ® provided DDI pairs in categories in X (Shariff et al., 2021). The DDI pairs were classified by human body systems based on the potential DDI-induced ADRs, including the cardiovascular system (bradycardia, QT prolongation, ventricular arrhythmias, and hypotensive), central nervous system (hypotensive, extrapyramidal symptoms [EPS], and central nervous system toxicity), blood (bleeding, hypoglycemia, hyperkalemia, hypokalemia, thrombotic events, and change in drug concentrations), gastrointestinal system (gastrointestinal lesions), and musculoskeletal system (myopathy). The patients' DDI pairs were determined for the prehospitalization and post-discharge periods. The number of DDI pairs was also identified.
Covariates
Patients' demographic (age and gender), socioeconomic (residential areas of NHI division, monthly income of NHI registration), Charlson Comorbidity Index (CCI), and type of comorbidity were retrieved at the first hospitalization. The insurance income ranking was used as a proxy for the monthly income level. A patient had a high or low monthly income if they received New Taiwan dollars (NT$) < 22,000 or NT$ ≥ 22,000 monthly, respectively (in January 2023, US$ 1 = approximately NT$ 30). Patients' comorbidities were identified within the year preceding the first discharge to calculate the CCI score (Charlson et al., 1987) and further categorized into four groups (i.e., 0, 1, 2, and ≥3 comorbidities).
To determine polypharmacy, we selected only those drugs supplied for more than 28 days and prescribed within 30 days of the first discharge. Polypharmacy was defined as the concomitant use of 5 or more drugs (Lu et al., 2015).
Statistical analysis
Logistic regression was used to evaluate the odds ratios (ORs) of clinical outcomes in the patients prescribed PIM after discharge, after adjusting for the covariates age at hospitalization, gender, geographic region, income, CCI score, admitted year, DDI (post-discharge), and polypharmacy. The results were reported as adjusted odds ratios (aORs) with 95% confidence intervals (CIs). Data management, computation, and analysis were performed using SAS software, version 9.4 (SAS Institute, Inc., Cary, NC, United States).
Study population characteristics
As shown in Table 1, a total of 206,058 older patients (mean age: 72.5 years) were enrolled in this study. Of these, 50.2% were female, 33.4% lived in the Taipei district, and 75.3% had high income. In the analysis, PIM-Taiwan, Beers, and PRICUS criteria identified 133,201 (64.6%), 147,450 (71.6%), and 97,790 (47.5%), respectively, as having PIM exposure. The following percentage of patients had a CCI score of 0: 48.6% (in PIM-Taiwan), 47.9% (in Beers), and 46.9% (in PRISCUS). However, those assessed via PRISCUS criteria had a higher proportion of CCI scores of 2 and ≥3 (13.5% and 10.1%, respectively) than those assessed via PIM-Taiwan (12.8% and 9.3%, respectively) or Beers criteria (12.9% and 9.5%, respectively). In all three cohorts, more than 10% had the comorbidities of cerebrovascular disease, chronic pulmonary disease, peptic ulcer disease, and diabetes without chronic complications. In all three cohorts, more than one-fourth of the patients had DDIs. PIM-Taiwan, Beers, and PRICUS criteria identified polypharmacy in 32.1%, 32.5%, and 36.7% of the cohort, respectively.
PIM patterns
As shown in Figure 1, all three PIM criteria showed a lower percentage of patients with PIM exposure at post-discharge than before hospitalization. Compared to before hospitalization, fewer Frontiers in Pharmacology frontiersin.org patients post-discharge had PIM exposure in all categories (i.e., ≥10, 6-10, or 1-5 items). As shown in Table 2
FIGURE 1
The distributions of potentially inappropriate medication at the prehospitalization and those at the postdischarge. PIM, potentially inappropriate medication; Prehospitalization, within 30 days before the first hospitalization; Postdischarge , within 30 days after the first discharge.
Frontiers in Pharmacology frontiersin.org
Discussion
Main study findings To our knowledge, this is the first nationally representative evaluation of the association between exposure to PIM and healthrelated outcomes among older adults in Taiwan. We compared exposure to PIM and the risk of 90-day rehospitalization in this population using the PIM-Taiwan, PRISCUS, and Beers criteria systems. The PIM criteria identified a high percentage of older adults who had been prescribed at least one PIM at the first hospitalization; meanwhile, the greatest incidence of PIM was found by Beers, then PIM-Taiwan, and then PRISCUS criteria. The findings proved that Beers criteria are more efficient in predicting 90-day rehospitalization among older adults experiencing PIM in Taiwan than either PIM-Taiwan or PRISCUS criteria.
Findings in the context of other studies
This retrospective cohort study, examining 206,058 older adults at the first hospitalization, found varying PIM rates The study outcome of 90-day rehospitalization was the detection of the following events within 90 days after the first discharge: rehospitalization, emergency department visit, hospitalization via the emergency department, and death.
Frontiers in Pharmacology frontiersin.org from 48% to 72% using the PRICUS, PIM-Taiwan, and Beers criteria systems. Our findings are similar to those of previous studies (Chang et al., 2014;Storms et al., 2017). Approximately one-half of the older patients in our study had a CCI score of 0 at the first hospitalization; however, those assessed via PRISCUS criteria had a higher proportion of CCI scores of 2 and ≥3 than those evaluated using the other two criteria. The data showed that about one-half of our older participants felt well in physical condition; the results implied that the PRICUS criteria had greater sensitivity than the other two criteria in identifying participants with poor physical condition (Fromm et al., 2013;Chang et al., 2019). In addition, over one-third of our patients had polypharmacy (taking >5 medications) at the first hospitalization. Our finding was similar to that of the other study in Australia (36%) (Lu et al., 2015). As was shown in Figure 1, older adults in Taiwan had fewer PIM prescribed after their first hospital discharge. According to Table 2, older adults in Taiwan after their first discharge had PIM exposure most commonly in the musculoskeletal, central nervous, gastrointestinal, and cardiovascular systems. The result of high PIM exposure in the musculoskeletal and gastrointestinal systems is not consistent with the results of other studies (Ye et al., 2016;Saboor et al., 2019). This discrepancy may be related to differences in medication prescribing habits in different regions. Furthermore, we initially assessed the association of PIM with 90-day rehospitalization among older adults in Taiwan. Table 2 shows that the older patients with 90-day rehospitalization still had PIM exposure in the musculoskeletal, central nervous, gastrointestinal, and cardiovascular systems after their hospitalization. We further evaluated the association of exposure to PIM with the risk of 90-day rehospitalization. According to the Beers criteria, exposure to PIM affecting the cardiovascular, gastrointestinal, central nervous, endocrine, and respiratory systems was positively associated with an increased risk of 90-day rehospitalization among older adults in Taiwan; hence, lack of PIM exposure was negatively associated with the risk. In PIM-Taiwan, PIM exposure affecting the cardiovascular, gastrointestinal, central nervous, and respiratory systems was positively associated with the risk; conversely, exposure to PIM affecting the sex hormones, genitourinary, endocrine, and musculoskeletal systems was negatively associated with the risk. In PRISCUS criteria, exposure to PIM affecting the respiratory, central nervous, and cardiovascular systems was positively associated with the risk of rehospitalization; conversely, exposed to PIM affecting the genitourinary and musculoskeletal systems had a negative association. Our findings tended to corroborate those of most studies. In a systematic review and meta-analysis including 63 studies, the pooled estimates for PIM and all-cause hospitalization were not statistically significant (aOR 1.11, 95% CI 0.76-1.63; adjusted Hazard Ratio 1.02, 95% CI 0.89-1.18) (Garcia et al., 2015). In brief, Beers criteria was proven more efficient in predicting 90day rehospitalization among older adults experiencing PIM in Taiwan, compared to PIM-Taiwan and PRISCUS criteria. Approximately 10%-30% of all hospitalizations in older adults are due to ADRs, of which almost 50% are potentially preventable (Leendertse et al., 2008;Salvi et al., 2012;Parameswaran Nair et al., 2016;Oscanoa et al., 2017). PIM criteria can help physicians and pharmacologists detect DDIs early and reduce DDI-induced ADRs, even though PIM does not seem to be an important cause of ADRs in older adults (Laroche et al., 2007). The DDI pairs have been classified according to major human body systems based on the potential for DDI-induced ADRs. Although the entire cohort of patients had a higher rate of DDIs than those with 90-day rehospitalization after the first discharge, the rate of ≥3 DDI
Musculoskeletal system
Myopathy 6,278 (3.1) 7,576 (3.7) 1,031 (13.6) DDI, drug-drug interaction; EPS, extrapyramidal symptoms; Prehospitalization, within 30 days before the first hospitalization; Post-discharge, within 30 days after the first discharge. a The study outcome of 90-day rehospitalization was the detection of the following events within 90 days after the first discharge: re-hospitalization, emergency department visit, hospitalization via the emergency department, and death. b The denominator is the study cohort (n = 206,058). c The DDI, class was classified by human body system, based on the potential DDI-induced adverse drug reactions. d To protect the personal data of patients, any number less than three cannot be provided by the Health and Welfare Data Centre.
Frontiers in Pharmacology frontiersin.org 08 pairs occurring in those with 90-day rehospitalization obviously increased from 11.3% in the whole cohort to 18.3% in those rehospitalized. This result showed that the risk of 90-day rehospitalization in older adults may be associated with the number of DDI pairs occurring, but not the number of those who experienced a DDI. We further analyzed the DDI classes in our study; the entire cohort of patients after discharge and those who experienced 90-day rehospitalization had the same 5 most common classes of DDI: bleeding, hypotension, myopathy, bradycardia, and change in drug concentrations. A greater rate of 90-day rehospitalization was associated with the potentially harmful DDI classes of serotonin syndrome, QT prolongation, EPS, and hypokalemia, as shown in Table 4. Notably, physicians and pharmacists should be well aware of the association of the above drug-related adverse effects with rehospitalization in older adults, despite their low incidence.
Study strengths and limitations
The primary strength of the current study was its use of a large population-based cohort that enabled analysis of the association between three PIM criteria and rehospitalization among older adults experiencing PIM. The findings may help prevent medicationrelated hospitalization in older adults, particularly in ethnic Chinese older adults. Furthermore, this study captured almost all prescribed medications reimbursed by the NHI, meaning our database contains detailed nationwide information on medications, including the use of drug combinations.
Of course, we acknowledge several possible limitations in this study. First, it is impossible to clearly demonstrate causality between a high incidence of PIMs and the outcome measure of rehospitalization based on this retrospective data. Second, patients' adherence to medications could not be evaluated, due to the limitations of the prescription claims databases. However, medication nonadherence would most likely have resulted in a nondifferentiated exposure misclassification, leading to possible underestimation of PIM prevalence and rehospitalization. Third, the use of some drugs, such as over-the-counter medications, alternative remedies, and herbal supplements, could not be included, because these drugs are not covered by the NHI. Fourth, three different lists were used in three different cohort of older adults. This method may have some disadvantages such as case-to-case variance bias. Although those cohorts had minor variations, we considered these factors as variables to be adjusted in the regression analysis. Finally, we did not have access to other potentially confounding factors for hospitalization among older patients such as disease severity, biochemistry data, and patient habits, such as alcohol use and tobacco consumption.
Conclusion
This study made several important findings. First, all three instruments found high PIM rates in older adults in Taiwan, but the greatest incidence of PIM was found by Beers, then PIM-Taiwan, and then PRISCUS criteria. Second, the 5 common potential DDIinduced adverse effects in older adults in Taiwan were bleeding, hypotension, myopathy, bradycardia, and change in drug concentrations. Despite their low incidence, the potential DDIinduced adverse effects of serotonin syndrome, QT prolongation, EPS, and hypokalemia were associated with high rates of rehospitalization among older adults. Finally, Beers criteria proved most efficient in predicting 90-day rehospitalization in older adults with PIM in Taiwan, compared to PIM-Taiwan and PRISCUS criteria. The current study findings permit several recommendations to be made. Physicians and pharmacists should seek to prevent and carefully manage ADR in older adults via the PIM criteria. In particular, they should look for PIM particularly in the DDI classes of serotonin syndrome, QT prolongation, EPS, and hypokalemia, which have the greatest potential for serious adverse events, despite their low incidence.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, data are available from the National Health Insurance Research Database (NHIRD) published by Taiwan National Health Insurance (NHI) Bureau. Due to legal restrictions imposed by the government of Taiwan in relation to the "Personal Information Protection Act," data cannot be made publicly available. Requests for data may be sent as a formal proposal to the NHIRD (http://nhird.nhri.org.tw).
Ethics statement
The study was performed in accordance with the Declaration of Helsinki. All methods were performed in accordance with the relevant guidelines and regulations. Ethical approval was obtained from the Ethics Committee of the Dalin Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Chia-Yi, Taiwan (IRB No. B10901009) and Kaohsiung Medical University, Kaohsiung, Taiwan (IRB No. KMUHIRB-E(I)-20190112). The need for informed consent was waived because we used de-identified medical information from the NHIRD. | 2023-07-15T15:21:03.315Z | 2023-07-13T00:00:00.000 | {
"year": 2023,
"sha1": "eb54d74a0321759dfd02a9426d02f0ba86e54103",
"oa_license": "CCBY",
"oa_url": "https://www.frontiersin.org/articles/10.3389/fphar.2023.1194537/pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "45c26ba9038cdf249cfadcb0e1658b23b6c1f9cc",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
118992946 | pes2o/s2orc | v3-fos-license | Entropy per particle spikes in the transition metal dichalcogenides
We derive a general expression for the entropy per particle as a function of chemical potential, temperature and gap magnitude for the single layer transition metal dichalcogenides. The electronic excitations in these materials can be approximately regarded as two species of the massive or gapped Dirac fermions. Inside the smaller gap there is a region with zero density of states where the dependence of the entropy per particle on the chemical potential exhibits a huge dip-and-peak structure. The edge of the larger gap is accompanied by the discontinuity of the density of states that results in the peak in the dependence of the entropy per particle on the chemical potential. The specificity of the transition metal dichalcogenides makes possible the observation of these features at rather high temperatures order of 100 K. The influence of the uniaxial strain on the entropy per particle is discussed.
I. INTRODUCTION
We devote our work to the memory of Alexei Alexeyevich Abrikosov. One of the topic of his research at the end of the last millennium [1] was the unusual magnetoresistance, linear in magnetic field and positive, observed in nonstoichiometric silver chalcogenides. His approach was based on the assumption that these substances are gapless semiconductors with a linear energy spectrum discovered by himself with coauthors in sixties [2]. This work of Abrikosov had drawn attention of two authors of the present work (VG and SS) to the various realizations of the Dirac fermions in condensed matter systems. It was impossible to foresee that the discovery of graphene in 2004 would make the Dirac fermions in condensed matter one of the hottest topics of research for decades.
Another lesson that one may learn studying the scientific heritage of Alexei Abrikosov is to focus on the theoretical results that are closely related to experiment. He always taught that the article must be finished by the formula, which can be checked by experimentalist. Following his advice here we present a study of the entropy per particle s = ∂S/∂n (S is the entropy per unit volume and n is the electron density) for which a witty approach for experimental measurement was discovered by Kuntsevich et al. [3]. In spite of the fundamental character of entropy that characterizes thermodynamics, heat transfer, thermoelectric properties of many-body systems, it is always hard to measure it directly. The recent experiment [3] is not an exception as the quantity measured directly in a 2D electrong gas is the temperature derivative of the chemical potential, ∂µ/∂T . The key idea of the authors of experiment [3] is that modulation of the sample temperature changes the chemical potential and, hence, causes recharging of the gated structure, where the 2D electrons and the gate act as two plates of a capacitor. Therefore, ∂µ/∂T is directly determined in the experiment from the measured recharging current. The Maxwell relation is then allows to equate both derivatives It was theoretically predicted [4] that in a quasi-twodimensional electron gas (2DEG) with parabolic dispersion, the entropy per electron exhibits quantized peaks when the chemical potential crosses the size quantized levels. The amplitude of such peaks in the absence of scattering depends only on the subband quantization number and is independent of material parameters, shape of the confining potential, electron effective mass, and temperature.
Very recently we studied [5] the behavior of s as a function of chemical potential, temperature and gap magnitude for the gapped Dirac materials. A special attention was paid to low-buckled Dirac materials [6,7], e.g. silicene [8] and germanene [9]. The dispersion law in these materials writes where η = ±1 and σ = ±1 are the valley and spin indices, respectively. Here v F is the Fermi velocity, k is the wavevector, and the valley-and spin-dependent gap, ∆ ησ = ∆ z − ησ∆ SO , where ∆ SO is the material dependent spin-orbit gap caused by a strong intrinsic spinorbit interaction. It can have a relatively large value, e.g. ∆ SO ≈ 4.2 meV in silicene and ∆ SO ≈ 11.8 meV in germanene. The adjustable part of the gap ∆ z = E z d, where 2d is the separation between the two sublattices situated in different planes, can be tuned by applying an electric field E z . Accordingly, the density of states (DOS) reads where the function f (ǫ) is assumed to be a continuous even function of energy ǫ and in the case of the discussed materials N = 2 and f (ǫ) = |ǫ|/(π 2 v 2 F ). The DOS (3) has 4 discontinuities at the points ǫ = ±∆ i , where i = 1 corresponds to η = σ = ±1 with ∆ 1 = |∆ SO − ∆ z | and the second one with i = 2 corresponds to η = −σ = ±1 with ∆ 2 = |∆ z + ∆ SO |.
One of the main results predicted in [5] is that for µ = ±∆ 2 (∆ SO , ∆ z > 0 was assumed) there is a peak of the height s = ±2 ln 2/3 in entropy per particle when T → 0. The calculation of [5] shows that a peak at µ = ±∆ 2 can still be seen for the temperature, T ∼ 10 −2 ∆ 1 for ∆ 2 = 2∆ 1 . Taking ∆ 2 ∼ ∆ SO , one estimates that the necessary temperature is the order of a few Kelvins.
Layered transition-metal dichalcogenides (TMDCs) represent another class of materials that can be shaped into monolayers, where similar effects might be observed. Single layer TMDCs with the composition MX 2 (where M = Mo, W is a transition metal, and X = S, Se, Te is a chalcogen atom) are truly two-dimensional (2D) semiconductors with a large band gap of the order of 1 eV to 2 eV (see, e.g. Refs. [10,15]). Consequently, one may expect that the peaks in entropy per particle can be seen at much higher temperatures.
The paper is organized as follows. We begin by presenting in section II the model describing single layer TMDCs. Since the full description of strained TMDCs is very complicated, the effect of a uniform uniaxial strain is taken into account only via scalar potential spinindependent parts of the Hamiltonian. In section III we discuss the DOS and present an analytical expression for the entropy per particle in TMDCs. The results for the obtained behaviour of the entropy per particle are discussed in section IV and conclusions are given in section V.
II. MODEL
The low-energy excitations in monolayer TMDCs can be described by the following model Hamiltonian density [11][12][13][14][15] where τ = ±1 is the valley index, H τ D is the linear in momentum in Dirac-like part [16] and H 2 is the quadratic part. The Dirac Hamiltonian contains free massive Dirac fermion, H τ 0 , and spin-orbit term H τ SO , H τ D = H τ 0 +H τ SO . The first term is σ σ σ are the Pauli matrices acting in the 2 × 2 "band" space, σ 0 is the unit matrix, the Fermi velocity v F = at/ ∼ 0.5 × 10 6 m/s with t being the effective hopping integral and a is the lattice constant, the major band gap ∆ ∼ 1 eV to 2 eV. The inversion symmetry breaking results in the the spin-orbit part of the Hamiltonian where s z is the Pauli matrix for spin, 2λ v ∼ 150 meV to 500 meV is the spin splitting at the valence band top caused by the spin orbit coupling, 2λ c is the spin splitting at at the conduction band bottom. The DFT calculations [15] show that absolute value 2λ v ≫ |2λ c | ∼ 3 meV to 50 meV and the sign of λ c depends on the compound, λ c > 0 for MoX 2 and λ c < 0 for WX 2 compounds. The quadratic part of the Hamiltonian, H 2 , contains the following diagonal terms where m e is the free electron mass, and α = β are constants of the order of 1. Finally, as discussed in [11][12][13][14][15] more accurate approximations also include the trigonal warping terms. The spin-up and spin-down components are completely decoupled, thus the spin index σ = ±1 is a good quantum number. Neglecting the quadratic term (7) we obtain the dispersion laws for conduction and valence bands This spectrum closely resembles that of described by Eq.
(2) of massive fermions in low-buckled Dirac materials except to the first valley-and spin-dependent term in Eq. (8). In the first approximation one can neglect the conduction band splitting and take λ c = 0 to arrive at the simplest model [16], where the conduction bands remain spin degenerate at K and K ′ points and have small spin splitting quadratic in k, whereas the valence bands are completely split, Single layer TMDCs can sustain deformations higher than 10% [17,18]. The experimental possibility to tune the band gap with strain has been proven for MoS 2 in [19][20][21][22] and in WS 2 [23][24][25]. The full description of strained TMDCs is much more involved than that of graphene and includes five different fictitious gauge fields as well as scalar potentials entering spin-independent and spin-dependent parts of the Hamiltonian [26]. Below we restrict ourselves by a qualitative estimate of the strain effect on the properties of TMDCs and consider only the scalar potential term in the spin-independent Hamiltonian (5), viz.
where ε ↔ is the strain tensor. The explicit expressions for the diagonal terms D ± are provided in [26] and here we only keep the linear in strain contributions neglecting the higher order terms with α + 2 = −3.07 eV and α + 2 = −1.36 eV. The corresponding parameters for the spin-dependent part are smaller by the three orders of magnitude, so that the corresponding term can be safely neglected. Assuming that the strain is a uniform uniaxial one, we can express D ± via ε ≡ ε xx (ε > 0 for tensile strain) and the Poisson's ratio, ν, [27] as follows D ± = α ± 2 ε(1 − ν). Thus in the present toy model the effect of strain is reduced to renormalization of the chemical potential, and the gap Setting ν = 0 one may estimate that 1% tensile strain shifts µ by 22 meV and ∆ by −17 meV, respectively.
III. ENTROPY PER PARTICLE
As it was mentioned above, the entropy per particle is directly related to the temperature derivative of the chemical potential at the fixed density n (see Eq. (1)). The latter can be obtained using the thermodynamic identity At thermal equilibrium, the total density of electrons is where f FD (x) = 1/[exp(x) + 1] is the Fermi-Dirac distribution function and we set k B = 1. Note that in the presence of the electron-hole symmetry it is convenient to operate with the difference n between the densities of electrons and holes instead of the total density of electrons, as usually done for graphene [5]. One can show that in a close analogy with graphene and low-buckled Dirac materials the DOS for TMDCs described by the approximate spectrum (8) is Here we denoted Obviously for λ c = 0 the resulting DOS corresponds to the spectrum (9). The DOS (16) differs from the one described by the equation (3) by the presence of the energy shift, ǫ i , in the modulus and in the argument of the θ-function. As a consequence the quantization of the entropy per particle, s = ±2 ln 2/3, obtained in [5] for the low-buckled Dirac materials does not occur in TMDCs.
The behavior of the DOS given by Eq. (16) is illustrated in Fig. 1. To be specific, we took the values ∆ = 1.79 eV, 2λ v = 0.43 eV corresponding to the compound WS 2 . The constant 2λ c for WS 2 is −0.03 eV [15]. In order to demonstrate the role of this parameter we choose the larger value of λ c . Furthermore, we consider three possible cases: λ c = 0 is shown by the dash-dotted (red) line, long dashed (green) line is for λ c = 0.05 eV, dotted (blue) line is for λ c = −0.05 eV. Note that in general ab initio density functional theory calculations [15] predict that λ c > 0 and λ c < 0 correspond to MoX 2 and WX 2 compounds. Going from the negative Obviously for λ c = 0 the last two discontinuities become degenerate ǫ + −1 = ǫ + 1 = 0.895 eV. For a finite λ c their ordering depends on the sign of λ c .
The peculiarities of DOS in TMDCs beyond the Dirac approximations are discussed in [28,29]. The quadratic part of the Hamiltonian (7) results in the curving of the linear in energy pieces seen in Fig. 1. Such curving is not essential and the does not change the discontinuous character of the DOS function that is responsible for the peaks in s(µ).
An advantage of the linearized approximation is that it resembles the case of gapped graphene and allows to obtain rather simple analytical results. For example, one can derive the analytical expression for the particle density (carrier imbalance) [30] and find the derivative ∂µ/∂T using Eq. (14). Its generalization for the lowbuckled Dirac materials was made in [5] (see also [31]). The expression for the particle density in TMDCs beyond the Dirac approximation is discussed in [29], but it is not very practical for obtaining the derivative ∂µ/∂T . Differentiating Eq. (15) with respect to T and µ and shifting the variable of integration ǫ → ǫ + ǫ i for each term in the DOS (16) one obtains and (18) where µ i = µ − ǫ i is the shifted chemical potential. Since the corresponding integrands in Eqs. (17) and (18) become formally the same as in the case of the low-buckled Dirac materials [5] we arrive at the final expressions and n µ (µ, ∆, T ) = 1 is antisymmetric under change µ → −µ and symmetric under ∆ → −∆. The last property is checked using the identity for the dilogarithm function
IV. RESULTS
Basing on obtained Eqs. (14), (19) and (20) one can investigate the dependence s(µ) for the different cases. in [5], one can see that overall shape of s(µ) is similar for TMDCs and low-buckled Dirac materials, although the details are different. For example, inside the gap for µ ∈ [ǫ − 1 , ǫ + 1 ] the dependence of s on the chemical potential exhibits a huge dip-and-peak structure in the temperature vicinity of the point µ = (λ v + λ c )/2. (The value i = 1 corresponds to the smaller gap in Eq. (16)). This feature is even more pronounced and sharp in TMDCs than in the other materials due to the larger ratio ∆/T . However in the low-buckled Dirac materials this structure was present in the temperature vicinity of the Dirac point, µ = 0, because the whole dependence s(µ) was an antisymmetric function of µ. This is obviously not the case of TMDCs. As discussed in [5] the peak inside the gap is mainly due to the specific dependence of the chemical potential on the electron density.
The presence of the second larger gap, ∆ 2 > ∆ 1 , in silicene and similar materials results in the emergence of the peak in s(µ) near the points µ = ±∆ 2 . Similarly the discontinuities of the DOS given by Eq. (16) associated with a larger gap i = 2 also result in the peaks in s(µ). They are shown in the inserts in Fig. 2, because they are much smaller in height. As explained above the value of s at the peaks in the low temperature limit is not equal to the quantized value ±2 ln 2/3 expected for the low-buckled Dirac materials [5]. It is essential that both peaks can still be seen at rather high temperatures. The peak on the right starts to smear at T = 80 K, while the peak on the left can still be seen.
It is shown in Fig. 1 that for λ c = 0 the two discontinuities of the DOS merge at µ = ǫ + −1 = ǫ + 1 . Then the positive peak in s(µ) disappears as can be seen on the dash-dotted (red) in in Fig. 3. As in Fig. 2 the vertical lines correspond to the singularities of the DOS. There is only one singularity for the dash-dotted (red) line at µ = ǫ + −1 = ǫ + 1 = 0.895 eV. For nonzero λ c there are two singularities shifted from this point to the left and right by |λ c | = 0.05 eV. In this case the peak at the larger energy µ = ∆/2 + |λ c | is restored as can be seen on the dotted (blue) line for λ c < 0 and long dashed (green) line for λ c > 0. Finally we consider how a uniform uniaxial strain would affect the results shown in Fig. 2. We use Eqs. (12) and (13) to model the dependence of chemical potential and gap ∆ on the strain, respectively. The dependence s(µ) is shown for three values of the strain: the dotted (green) line is for ε = 0, the dashed (red) line is for ε = 2% and the solid (blue) line is for ε = 4%. As expected, the presence of strain results in the movement of the peaks in s(µ).
V. CONCLUSION
In the present work we had derived a general expression for the entropy per particle as a function of the chemical potential, temperature, and gap magnitude for the single layer transition metal dichalcogenides subjected to the uniform uniaxial strain. The spectrum of quasiparticle excitations of these materials is similar to that of the low-buckled Dirac materials, viz. there is the valleyand spin-dependent gap ∆ τ σ = [∆ − τ σ(λ v − λ c )]/2 in the spectrum. The difference from the latter is that the whole spectrum is also shifted by a valley-and spindependent constant ǫ τ σ = τ σ(λ v + λ c )/2. This introduces the hole-electron asymmetry in the band structure of TMDCs and makes the resulting DOS (16) asymmetric function of the energy. When a small spin splitting at the conduction band bottom, λ c , is taken into consideration the DOS (16) has 4 discontinuities: 2 for the negative and 2 for the positive energies. The positions of these discontinuities are not just at the energies ±|∆ τ σ | with τ = σ = ±1 and τ = −σ = ±1 due to the energy shift ǫ τ σ . It is demonstrated that inside the smaller gap there is a region with zero density of states where the dependence of the entropy per particle on the chemical potential exhibits a huge dip-and-peak structure. The edge of the larger gap is accompanied by the discontinuity of the density of states that results in the peak in the dependence of s on the chemical potential. The specifics of the transition metal dichalcogenides makes the found features to be of the "high temperature" nature, since they can be observed at rather high temperatures up to 100 K.
Since the Seebeck coefficient is related to the temperature derivative of the chemical potential, the strong peaks in the entropy per particle also indicate the same kind of singularities in the Seebeck coefficient in these materi-als. The latter can be expected at the edge of the gaps and has the origin similar to the electronic topological transitions [32][33][34]. | 2018-04-20T07:57:51.000Z | 2017-12-25T00:00:00.000 | {
"year": 2017,
"sha1": "d79c5a22a521def6f0d983ed64b8ef8a492393f0",
"oa_license": null,
"oa_url": "http://arxiv.org/pdf/1712.09118",
"oa_status": "GREEN",
"pdf_src": "Arxiv",
"pdf_hash": "d79c5a22a521def6f0d983ed64b8ef8a492393f0",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
268802315 | pes2o/s2orc | v3-fos-license | Comparative Genomic Analysis of PEBP Genes in Cucurbits Explores the Interactors of Cucumber CsPEBPs Related to Flowering Time
The family of phosphatidylethanolamine-binding proteins (PEBPs) participates in various plant biological processes, mainly flowering regulation and seed germination. In cucurbit crops, several PEBP genes have been recognized to be responsible for flowering time. However, the investigation of PEBP family members across the genomes of cucurbit species has not been reported, and their conservation and divergence in structure and function remain largely unclear. Herein, PEBP genes were identified from seven cucurbit crops and were used to perform a comparative genomics analysis. The cucurbit PEBP proteins could be classified into MFT, FT, TFL, and PEBP clades, and further, the TFL clade was divided into BFT-like, CEN-like, and TFL1-like subclades. The MFT-like, FT-like, and TFL-like proteins were clearly distinguished by a critical amino acid residue at the 85th position of the Arabidopsis FT protein. In gene expression analysis, CsaPEBP1 was highly expressed in flowers, and its expression levels in females and males were 70.5 and 89.2 times higher, respectively, than those in leaves. CsaPEBP5, CsaPEBP6, and CsaPEBP7 were specifically expressed in male flowers, with expression levels 58.1, 17.3, and 15.7 times higher, respectively, than those of leaves. At least five CsaPEBP genes exhibited the highest expression during the later stages of corolla opening. Through clustering of time-series-based RNA-seq data, several potential transcription factors (TFs) interacting with four CsaPEBPs were identified during cucumber corolla opening. Because of the tandem repeats of binding sites in promoters, NF-YB (Csa4G037610) and GATA (Csa7G64580) TFs appeared to be better able to regulate the CsaPEBP2 and CsaPEBP5 genes, respectively. This study would provide helpful information for further investigating the roles of PEBP genes and their interacting TFs in growth and development processes, such as flowering time regulation in cucurbit crops.
Introduction
Phosphatidylethanolamine-binding proteins (PEBPs) are widespread in plants and contain a conserved domain peculiar to eukaryotes, despite sharing an ancient functional unit with PEBPs in bacteria, archaea, and animals [1,2].In the model plant Arabidopsis, AtPEBP proteins are mainly classified into three evolutionary branches, including FLOW-ERING LOCUS T (FT)-like, TERMINAL FLOWER1 (TFL1)-like, and MOTHER OF FT AND TFL1 (MFT)-like clades.In the AtFT-like clade, overexpression of both FT and TWIN SISTER OF FT (TSF) promotes earlier flowering [3].Among the three members of the TFL1like clade, TFL1 is a major negative regulator of inflorescence meristem development that considerably delays flowering time, while BROTHER OF FT AND TFL1 (BFT) play a role by acting redundantly with it [4].Also, ARABIDOPSIS THALIANA CENTRORADIALIS (ATC), an additional paralogous of TFL1, exhibits a weak ability to complement the early flowering and terminal flower formation of tfl1-1 mutant phenotypes, implying that ATC has a comparable function to TFL1 and BFT.MFT-like genes, which are considered to be the evolutionary origin of FT-like and TFL-like clade genes, regulate the transcripts of several key genes during seed germination but also have a mild FT-like activity to encourage flowering in Arabidopsis [5,6].In any case, PEBP family proteins are pivotal participants in controlling the plant flowering time.
FT and TFL1 proteins, with the opposite functions on flowering, have two highly conserved short motifs, DPDxP and GxHR, which presumably contribute to the conformation of the ligand binding pocket [1,7].The amino acid residues encoded by the fourth exons of FT and TFL1 that are essential for conferring biological specificity on the two proteins can be divided into four segments (Segment A-D).Segment B is especially important for the determination of functional specificity between FT and TFL1, as is the adjacent segment C that contains the LYN/IYN triplet motif [8].Additionally, the distal CCAAT box (TGTG(N2-3)ATG) and proximal CORE1/2 domains in the FT promoter region are found to be required cis-elements for binding by CONSTANS (CO) involved in photoperiodic pathways [9,10].These findings offer valuable information for investigating FT-and TFL1-like specific interactors in other crops.
The Cucurbitaceae members (cucurbits) contain many important vegetable and fruit crops, such as cucumber (Cucumis sativus L.), melon (Cucumis melo L.), watermelon (Citrullus lanatus L.), pumpkin (Cucurbita moschata L.), cushaw (Cucurbita argyrosperma L.), zucchini (Cucurbita pepo L.), and winter squash (Cucurbita maxima Duch.), etc. [11].The flowering time of the cucurbit crops is strongly correlated with early maturity and production, which is of great significance to the global or local economy [12].To date, only a few floweringrelated genes have been characterized in several cucurbit crops, including CsFT and CsTFL1 in cucumber, CmoFTL1 and CmoFTL2 in Chinese pumpkin, and FT in watermelon, all of which belong to the PEBP family [13][14][15][16].However, the regulatory mechanisms of the FT and TFL-like genes have not been well studied, and the roles of other PEBP members in regulating flowering time remain unclear.
The aim of this study was to identify the PEBP family genes in the seven cucurbit crops: cucumber, melon, watermelon, pumpkin, cushaw, zucchini, and winter squash.This investigation intends to comparatively analyze their chromosomal locations, phylogenetic relationships, and protein-conserved domains.Clustering of time-series-based RNA-seq data and promoter cis-element analyses were performed to explore the functions of the cucumber PEBP genes and their potential interacting transcription factors (TFs).These results could enhance our understanding of the cucurbit PEBP family members and provide available clues for delving deeper into their roles, especially in flowering time regulation.
Evolution of PEBP Genes in Arabidopsis and Seven Cucurbit Crops
To investigate the evolutionary relationships of the PEBP proteins, a comparative Neighbor-Joining (NJ) tree was conducted using the amino acid sequences encoded by the 72 PEBP genes from Arabidopsis and seven cucurbit crops (Figure 2, Table S1).According to the NJ tree, a total of 72 PEBPs fall into four major branches, which were designated as MFT clade, FT clade, TFL clade, and PEBP clade.Further, the TFL clade could be divided into BFT-like, CEN-like, and TFL1-like subfamilies (Figure 2).Both MFT-like and TFL1-like subfamilies consisted of 17 cucurbit PEBP members, including at least two PEBPs from each cucurbit crop.The FT/TSF-like subfamily was comprised of one ClaPEBP, one CmePEBP, and one CsaPEBP, as well as two CarPEBPs, two CmaPEBPs, and two CmoPEBPs.The member number of the FT/TSF-like subfamily seemed to be consistent with the ploidy of different crops.The remaining BFT-like, CEN-like, and PEBP_bact-like subfamilies only had one PEBP protein from each of the seven cucurbits, except for the absence of any CpePEBP in the BFT-like subgroup (Figure 2).
A syntenic map illustrated the conserved syntenic blocks across the seven cucurbit crops, and 62 of the 65 PEBP genes were mapped to these orthologous blocks, indicating a high degree of macrosynteny (Figure 3, Table S3).In total, 289 orthologous gene pairs were identified, with the number of genes matched varying from one to four.Among them, most one-to-one orthologous genes belonged to the PEBP genes from MFTlike, BFT-like, or PEBP-like subfamilies, e.g., CmaPEBP5/6/8/9, CmoPEBP4/5/7/8, and CpePEBP2/4/6/10.The one-to-four orthologous genes mainly come from the TFL1-like subfamily.The FT/TSF-like members matched one or two PEBPs as orthologous genes.Interestingly, CmePEBP3-CsaPEBP6 seemed to be a unique orthologous gene pair (Figure 3, Table S3), which likely resulted from an independent whole-genome duplication event between melon and cucumber.
Additionally, 28 orthologous gene pairs were identified between Arabidopsis and seven cucurbit crops, including five in cucumber, three in melon, five in watermelon, three in pumpkin, four in cushaw, five in zucchini, and three in winter squash (Figure S1, Table S4).Except for ClaPEBP1-AtATC/AtTFL and CsaPEBPE8-AtATC/AtTFL, the rest were oneto-one orthologous gene pairs.All of these gene pairs fall into the MFT, TFL (BFT-like, CEN-like, and TFL-like), and PEBP clades, indicating that they might be derived from a common ancestor of Arabidopsis and seven cucurbit crops.A syntenic map illustrated the conserved syntenic blocks across the seven cucurbit crops, and 62 of the 65 PEBP genes were mapped to these orthologous blocks, indicating a high degree of macrosynteny (Figure 3, Table S3).In total, 289 orthologous gene pairs were identified, with the number of genes matched varying from one to four.Among them, most one-to-one orthologous genes belonged to the PEBP genes from MFT-like, BFT-like, or PEBP-like subfamilies, e.g., CmaPEBP5/6/8/9, CmoPEBP4/5/7/8, and CpePEBP2/4/6/10.The one-to-four orthologous genes mainly come from the TFL1-like subfamily.The FT/TSF-like members matched one or two PEBPs as orthologous genes.Interestingly, CmePEBP3-CsaPEBP6 seemed to be a unique orthologous gene pair (Figure 3, Table S3), which likely resulted from an independent whole-genome duplication event between melon and cucumber.Additionally, 28 orthologous gene pairs were identified between Arabidopsis and seven cucurbit crops, including five in cucumber, three in melon, five in watermelon, three in pumpkin, four in cushaw, five in zucchini, and three in winter squash (Figure S1, Table S4).Except for ClaPEBP1-AtATC/AtTFL and CsaPEBPE8-AtATC/AtTFL, the rest were oneto-one orthologous gene pairs.All of these gene pairs fall into the MFT, TFL (BFT-like,
Comparative Analysis of the PEBP Family Gene Structure
Multiple sequence alignments showed that all the cucurbit PEBP proteins from the MFT clade, FT clade, and TFL clade had conserved PEBP domains and DPDxP motifs (Figures 4 and S2).Also, the 85th amino acid 'Y' of AtFT, marked with a green rectangle, could clearly distinguish the MFT-like (W), FT-like (Y), and TFL-like (H) proteins, which was a key amino acid residue that determined FT-like and TFL-like functions.Most cucurbit FT-like and TFL-like proteins had amino acid residues marked with a red rectangle, indicating they might interact with the 14-3-3 protein.The other two amino acid sequences, 'LGRQTVYAPGWRQN' and 'L/IYN' were found to be highly conserved in the segments B and C of each cucurbit FT-like protein (Figures 4 and S2), which were the critical motifs for determining the FT activity and the FT/TFL1 function.These results suggested that the FT-like clade from seven cucurbits was the most conserved, followed by the TFL-like clade, while the MFT-like clade was more variable.In gene structure analysis, the 58 cucurbit PEBP genes from the MFT, FT, and TFL clades were identified to possess a similar number of exons, varying between three and six.Among them, approximately 75% (44) of PEBP genes had four exons (Table S1).The PEBPs ranged from 233 bp to 6958 bp in gene length.From these, it could be inferred that the variations in the length and number of exons and introns may lead to the functional diversity of PEBP family members.
Cis-Acting Elements in the PEBP Promoter Regions
Within the putative promoter regions (2000 bp upstream from the transcription start site) of PEBP genes, there were five categories of cis-acting elements predicted, including the defensive or stress-related element, the developmental-related element, the hormoneresponsive element, the light-responsive element, and the TF binding element (Figure 5, Table S5).Among them, 40 of the 65 cucurbit PEBP promoters had no developmentalrelated element, indicating that they might be specific PEBP members involved in growth and development regulation.Exactly like the CsaPEBP1 promoter, it was the only one of eight cucumber PEBP members with this element (CAT-box), which functioned in meristem expression (Table S5).Moreover, the TF binding elements (e.g., AP-1, MYB, MYC, and HD-Zip) were identified in each PEBP promoter, accounting for almost 50% of the total (Figure 5, Table S5), indicating that they might have the potential to regulate PEBP genes.
FT is a critical integration factor in the pathway of flowering initiation signal (florigen) in Arabidopsis, and the distal CCAAT box and proximal CORE domains in the promoter region are required for its proper expression [10].Herein, we extracted the 7000-bp region of 11 FT/TSF-like genes from seven cucurbits and searched for the CCAAT box and CORE1/2 domain.Results showed that seven to sixteen CCACA boxes were present in all FT/TSF-like promoters, and the CORE1/2 domain was detected except for CmePBEP1, CmoPEBP2, and CpePEBP8 (Figure S3).The above cis-acting regulatory elements would provide some information for the tissue-specific expression and interaction models of the In gene structure analysis, the 58 cucurbit PEBP genes from the MFT, FT, and TFL clades were identified to possess a similar number of exons, varying between three and six.Among them, approximately 75% (44) of PEBP genes had four exons (Table S1).The PEBPs ranged from 233 bp to 6958 bp in gene length.From these, it could be inferred that the variations in the length and number of exons and introns may lead to the functional diversity of PEBP family members.
Cis-Acting Elements in the PEBP Promoter Regions
Within the putative promoter regions (2000 bp upstream from the transcription start site) of PEBP genes, there were five categories of cis-acting elements predicted, including the defensive or stress-related element, the developmental-related element, the hormoneresponsive element, the light-responsive element, and the TF binding element (Figure 5, Table S5).Among them, 40 of the 65 cucurbit PEBP promoters had no developmentalrelated element, indicating that they might be specific PEBP members involved in growth and development regulation.Exactly like the CsaPEBP1 promoter, it was the only one of eight cucumber PEBP members with this element (CAT-box), which functioned in meristem expression (Table S5).Moreover, the TF binding elements (e.g., AP-1, MYB, MYC, and HD-Zip) were identified in each PEBP promoter, accounting for almost 50% of the total (Figure 5, Table S5), indicating that they might have the potential to regulate PEBP genes.
Spatio-Temporal Expression Patterns of CsaPEBP Genes
Cucumber is a typical model for unisexual flower development, and two PEBP family genes, CsaPEBP2 (CsFT) and CsaPEBP5 (CsTFL1), have been reported to influence flowering time [14,15].To further investigate the roles of cucumber PEBP members, the expression patterns of eight CsaPEBPs were analyzed in various tissues (leaf, stem, tendril, male, female, and fruit).Reverse transcription-polymerase chain reaction (RT-qPCR) analysis indicated that each CsaPEBP gene displayed distinct expression patterns.In detail, Csa-PEBP2 and CsaPEBP8 were expressed in all tissues (Figure 6b,h).CsaPEBP1 was highly expressed in flowers, and its expression levels in females and males were 70.5 and 89.2 times higher, respectively, than those in leaves (Figure 6a).CsaPEBP5, CsaPEBP6, and Csa-PEBP7 were specifically expressed in male flowers, with expression levels 58.1, 17.3, and 15.7 times higher, respectively, than those of leaves (Figure 6e-g).CsaPEBP3 and Csa-PEBP4 exhibited higher expression levels in the stem (Figure 6c,d).Furthermore, the FT is a critical integration factor in the pathway of flowering initiation signal (florigen) in Arabidopsis, and the distal CCAAT box and proximal CORE domains in the promoter region are required for its proper expression [10].Herein, we extracted the 7000-bp region of 11 FT/TSF-like genes from seven cucurbits and searched for the CCAAT box and CORE1/2 domain.Results showed that seven to sixteen CCACA boxes were present in all FT/TSF-like promoters, and the CORE1/2 domain was detected except for CmePBEP1, CmoPEBP2, and CpePEBP8 (Figure S3).The above cis-acting regulatory elements would provide some information for the tissue-specific expression and interaction models of the cucurbit PEBP genes.
Spatio-Temporal Expression Patterns of CsaPEBP Genes
Cucumber is a typical model for unisexual flower development, and two PEBP family genes, CsaPEBP2 (CsFT) and CsaPEBP5 (CsTFL1), have been reported to influence flowering time [14,15].To further investigate the roles of cucumber PEBP members, the expression patterns of eight CsaPEBPs were analyzed in various tissues (leaf, stem, tendril, male, female, and fruit).Reverse transcription-polymerase chain reaction (RT-qPCR) analysis indicated that each CsaPEBP gene displayed distinct expression patterns.In detail, CsaPEBP2 and CsaPEBP8 were expressed in all tissues (Figure 6b,h).CsaPEBP1 was highly expressed in flowers, and its expression levels in females and males were 70.5 and 89.2 times higher, respectively, than those in leaves (Figure 6a).CsaPEBP5, CsaPEBP6, and CsaPEBP7 were specifically expressed in male flowers, with expression levels 58.1, 17.3, and 15.7 times higher, respectively, than those of leaves (Figure 6e-g).CsaPEBP3 and CsaPEBP4 exhibited higher expression levels in the stem (Figure 6c,d).Furthermore, the expression levels of CsaPEBP1, CsaPEBP2, CsaPEBP5, CsaPEBP6, and CsaPEBP7 were found to match the RNAseq data reported by Li et al. [17], indicating that these genes could play stable roles during the growth and development of cucumber, especially in the development of floral organs.In cucumber, the male or female corolla development could be divided into fo stages, including green bud, green-yellow bud, yellow bud, and flowering based on t color and shape (Figure 7a).To explore the roles of several CsaPEBP genes in the proce of male and female flower opening, RT-qPCR analysis was performed to examine th expression levels at each stage.As shown in Figure 7b, during the opening of the m flower, the expression levels of CsaPEBP1 and CsaPEBP7 had obviously increased at flo ering (M4), CsaPEBP5 and CsaPEBP6 showed higher expression at the yellow bud (M while CsaPEBP2 maintained a high expression level at both M3 and M4.In the progressi of female corolla opening, the expression of CsaPEBP1 and CsaPEBP5 exhibited a simil gradually upregulated trend, with a peak at flowering (C4), the peak expression of C PEBP2 occurring at the yellow bud (C3), while CsaPEBP6 and CsaPEBP7 were express at a higher level at C3 and C4.Unfortunately, the expression patterns of the five CsaPE Figure 6.Expression profiles of CsaPEBP1-CsaPEBP8 genes (a-h) in cucumber leaf, stem, tendril, male, female and ovary.Grey columns indicate the relative gene expression levels determined using RT-qPCR, and red lines represent the transcript abundance changes calculated using the Reads Per Kilobase per Million mapped reads (RPKM) method.Lowercase letters (for example, a, b, c, etc.) above the columns represent significant differences at the 0.05 level based on Tukey's test.
In cucumber, the male or female corolla development could be divided into four stages, including green bud, green-yellow bud, yellow bud, and flowering based on the color and shape (Figure 7a).To explore the roles of several CsaPEBP genes in the process of male and female flower opening, RT-qPCR analysis was performed to examine their expression levels at each stage.As shown in Figure 7b, during the opening of the male flower, the expression levels of CsaPEBP1 and CsaPEBP7 had obviously increased at flowering (M4), CsaPEBP5 and CsaPEBP6 showed higher expression at the yellow bud (M3), while CsaPEBP2 maintained a high expression level at both M3 and M4.In the progression of female corolla opening, the expression of CsaPEBP1 and CsaPEBP5 exhibited a similar, gradually upregulated trend, with a peak at flowering (C4), the peak expression of CsaPEBP2 occurring at the yellow bud (C3), while CsaPEBP6 and CsaPEBP7 were expressed at a higher level at C3 and C4.Unfortunately, the expression patterns of the five CsaPEBP genes in the ovary did not exhibit sufficient regularity.These results suggested that CsaPEBP1, CsaPEBP2, CsaPEBP5, CsaPEBP6, and CsaPEBP7 might be potential factors involved in regulating cucumber flowering, mainly functioning at the later stages: the yellow bud and flowering.
Screening for TFs That Interact with CsaPEBPs
FT or FT-like proteins, belonging to the PEBP family are well-known integrators in the flowering networks in multiple plants.Also, they can be directly activated by some TFs, or form complexes with TFs to trigger the onset of flowering [18].In the current work, the clustering of time-series analysis between 1251 differentially expressed TFs and 4 CsaPEBP genes (CsaPEBP1/2/5/7) was performed from the published RNA-seq data of the corolla opening.Results showed that the expression patterns of CsaPEBP1, CsaPEBP2, CsaPEBP5, and CsaPEBP7 were consistent with those of 66, 20, 50, and 23 TFs, respectively (Figure 8a, Table S6), with correlation coefficients varying from 0.8 to 1.0 (Figure 8b).Further, the TFs with a strong positive correlation (correlation index > 0.99) were selected for screening interactors of CsaPEBPs (Table S6).
Screening for TFs That Interact with CsaPEBPs
FT or FT-like proteins, belonging to the PEBP family are well-known integrat the flowering networks in multiple plants.Also, they can be directly activated by TFs, or form complexes with TFs to trigger the onset of flowering [18].In the current the clustering of time-series analysis between 1251 differentially expressed TFs and PEBP genes (CsaPEBP1/2/5/7) was performed from the published RNA-seq data corolla opening.Results showed that the expression patterns of CsaPEBP1, CsaPEBP2 PEBP5, and CsaPEBP7 were consistent with those of 66, 20, 50, and 23 TFs, respec (Figure 8a, Table S6), with correlation coefficients varying from 0.8 to 1.0 (Figure 8b ther, the TFs with a strong positive correlation (correlation index > 0.99) were select screening interactors of CsaPEBPs (Table S6).Multiple TF binding sites (TFBSs) were identified in the promoter regions of th PEBP1, CsaPEBP2, CsaPEBP5, and CsaPEBP7 genes, which could correspond to the TFs mentioned above (Figure 8c, Table S7).Based on the sequence similarity, the Multiple TF binding sites (TFBSs) were identified in the promoter regions of the CsaPEBP1, CsaPEBP2, CsaPEBP5, and CsaPEBP7 genes, which could correspond to the most TFs mentioned above (Figure 8c, Table S7).Based on the sequence similarity, the TFBSs with 100% similarity to Arabidopsis ones were retained, as follows: one ERF, three NAC, and one HSF in the CsaPEBP1 promoter; eight TALE in the CsaPEBP2 promoter; two Dof, and five NF-YB TALE in the CsaPEBP5 promoter; eight WRKY and 79 GATA in the CsaPEBP7 promoter (Figure 8d).These elements provided critical evidence for whether TFs directly regulate CsaPEBP genes.Especially, the tandem repeats of NF-YB and GATA might enhance the binding abilities of Csa4G037610 and Csa7G064580 to the promoters of CsaPEBP2 and CsaPEBP5, respectively, thereby regulating their expression.
Discussion
Proper flowering time for crops such as cucurbits means a greater yield and more robust adaptation [19].Mining the vital genes related to flowering time would be beneficial to gain a better understanding of their genetic regulatory mechanisms, facilitating the cultivation of cucurbit varieties with superior growth periods and adaptations.
Comparative genomics analysis is acknowledged as a powerful approach to exploring the evolution and function of genes.In the current work, we examined the inter-and intra-specific differences between seven cucurbit species using a comparative genomics analysis of their PEBP genes.By integrating transcript expression profiles with promoter element prediction, these findings would empower us to discover the novel genes and their potential interactors involved in cucumber flowering.
Conservation of the PEBP Family in Seven Cucurbits
At the whole-genome level, eight CsaPEBPs, nine CmePEBPs, seven ClaPEBPs, 10 CmoPEBPs, 11 CarPEBPs, 10 CpePEBPs, and 10 CmaPEBPs were identified from cucumber, melon, watermelon, pumpkin, cushaw, zucchini, and winter squash, respectively.The number of PEBP family members in seven cucurbits was similar to Arabidopsis (7) but fewer than rice (19) and maize (25) [20,21].This validated the theory that monocots had three to four times more PEBP family members than dicots did [22].In actuality, cucurbits have a slightly higher number of PEBP genes than Arabidopsis, perhaps as a result of having a larger genome.Interestingly, out of the seven cucurbits, the tetraploid ones (pumpkin, cushaw, zucchini, and winter squash) have a larger number of PEBP genes than the diploid cucurbit ones (cucumber, melon, and watermelon), which is similar to the findings from Yi et al. [19].
According to Arabidopsis, a total of 65 PEBP proteins from seven cucurbits could be grouped into four main clades as follows: MFT, FT, TFL (including BFT-like, CEN-like, and TFL1-like subclades), and PEBP clades.Among these, the PEBPs from the MFT, FT, and TFL clades exhibited high sequence homology with Arabidopsis PEBPs, sharing the conserved PEBP domains and DPDxP motifs.Significantly, the MFT-like, FT-like, and TFL-like proteins may be distinguished by the 85th amino acid of AtFT, which explains why FT and TFL1 homologs functioned differently in several plants [22].Consistently, the cucumber FT gene (CsaPEBP2) has been reported to be a key gene controlling early flowering, but CsTFL1 (CsaPEBP5) inhibited terminal flower formation [14,15,23].For the gene structure, most cucurbit PEBP genes contained four exons and had large variations in gene length, which might lead to the functional diversity of PEBP family members.
Gene duplication events are an important way for plant evolution to have high efficiency [12].For the PEBP family in the seven cucurbits, whole-genome duplication might contribute more to gene diversity.Together with interspecies collinearity analysis and evolutionary relationships, the orthologous gene pairs with Arabidopsis were conducive to investigating the biological functions of PEBP genes in cucurbits.Notably, all of these gene pairs belonged to the MFT, TFL, and PEBP clades, indicating that they may have shared an ancestor with Arabidopsis and seven cucurbit crops.Moreover, the majority of collinear gene pairs belonged to one-to-one relationships, which was comparable to other crops, such as tomato [24].Additionally, it was also noticed that the members from the PEBP clade, which encoded the PEBP domain present in bacteria and archaea, were distinct in both the conserved domain and gene structure, unlike PEBP members from other clades (Figure S4, Table S1).Nevertheless, their phylogenetic relationships and conserved sequence regions provided evidence for the ancient common origin of a basic protein functional unit.
Functional Diversification of Cucumber PEBP Genes
Despite extensive sequence conservation, PEBP genes are well-known to be the key regulators of various biological processes, especially FT/TFL-like genes in the regulation of floral transition [1,25].Considering that the exact molecular function of the cucurbit PEBP genes was, in most cases, still unclear, CsaPEBP genes from cucumber, a well-established model system for unisexual flower development, were chosen for expression pattern profiles.Both in RNA-seq and RT-qPCR analyses, eight CsaPEBP genes exhibited divergent expression levels in six tissues of plants.This resembled the reports on tomato, bamboo, and Rosaceae tree species [22,24,26].More particularly, CsaPEBP5, CsaPEBP6, and CsaPEBP7 in male flowers, as well as CsaPEBP3 and CsaPEBP4 in the stem, appeared to exhibit strong tissue-specific high expression.Even if the genes belonging to the same clade might function differently, such as CsaPEBP1 and CsaPEBP6 from the MFT-like subclade and CsaPEBP5 and CsaPEBP8 from the TFL-like subclade.For the genes expressed in either male or female flowers, CsaPEBP1, CsaPEBP2, CsaPEBP5, CsaPEBP6, and CsaPEBP7 might act at the later stages of corolla opening.CsaPEBP2 and CsaPEBP5 have been reported to influence flowering in cucumber, namely CsFT and CsTFL1 [14,15], and correspondingly, their expression trends were reversed during the female corolla opening process.Summarily, these variable expression patterns suggested functional diversification of CsaPEBP genes, which was supported by the variations in their gene structure.
Novel TFs That Potentially Interact with CsaPEBPs
In the plant flowering regulatory network, TFs generally function in conjunction with flowering-related genes to regulate flowering time.CO (Zinc finger), CIB1 (bHLH), and FLC (MADS-box) have been recognized as the classical TFs that directly bind to the FT and its homolog promoter and induce their expression [27], and recently several novel TFs have also been confirmed to do so.For example, Arabidopsis B3-Domain TF VAL1 regulates the floral transition by repressing FT [28].StABL1, a TF central to abscisic acid signaling, binds to FT homologs to promote flowering and, consequently, a short life cycle in potatoes [29].This demonstrated that other TFs also have the potential to directly interact with FT or its homologs.Considering this, a strategy combining clustering of time-series and promoter cis-element analysis was carried out.After screening, several TFs were obtained as possible roles for directly binding FT or other PEBPs during flowering response.Except for the ERF TF reported in Arabidopsis, there is still little information on the direct interaction between FT and the other TFs [30].
In addition, the tandem repeats of NF-YB and GATA binding sites occurred in the promoter regions of CsaPEBP2 and CsaPEBP5, which may mean a stronger binding ability between TFs and CsaPEBPs.Although these findings need to be further verified, they can still provide novel evidence for understanding the flowering regulatory mechanism in cucurbit crops.
Multiple Alignments, Phylogenetic and Synteny Relationships Analysis
Using the MUSCLE method with default settings, the PEBP family members from all seven cucurbit crops, as well as Arabidopsis, were sequenced by parameters.The outcome file was imported into the GeneDoc (version 2.7.0) software for visualization [33].Based on the multiple sequence alignment, an NJ phylogenetic tree was conducted using MEGA 11 (version 11.0.13)software in accordance with the p-distance model, as well as 1000 bootstrapping replicates and other default options [34].The NJ tree was displayed and optimized using an online tool iTOL [35].
Conserved syntenic blocks within and between the genomes of seven cucurbit species and Arabidopsis were identified and visualized using MCScanX (version 1.1) and TBtools-II (version 2.0) software [36,37].
Gene Localization and Structure Analysis
The chromosomal localization and exon/intron organization of the PEBP genes were investigated based on the GFF/GTF annotation files of seven cucurbit crop genomes from CuGenDB.The gene localization information was visualized using an advanced Circos plot in TBtools-II [37].Then the exon number and length of each PEBP gene were counted.
Investigation of Cis-Elements in the Promoter Region
The 2000 bp DNA sequences upstream at the transcription start site (ATG) were extracted from the CuGenDB database as the promoter regions of PEBP genes.These sequences were uploaded to the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/, accessed on 7 November 2023) to identify the putative cisacting elements.To predict TFBSs, the FASTA file of the promoter sequences was submitted to the PlantPAN4.0website (http://plantpan.itps.ncku.edu.tw/plantpan4/index.html, accessed on 19 February 2023) [38].
Plant Sampling, RNA Preparation and RT-qPCR Analysis
The inbred cucumber line 'D1008' was used as the experimental material and was planted at a greenhouse in the Horticulture Experiment Station, Northeast Agricultural University (Harbin, China).The cultivation conditions were maintained at 28 • C/18 • C (day/night) and a 16 h day/8 h night cycle.At the fruiting stage, a total of five tissues (leaf, stem, tendril, female flower, and male flower) were separately collected.Otherwise, female and male flowers were sampled from 1, 3, 4, and 5 days after labeling (DAL; labeling was made when an ovary became visible), which represented the four stages of cucumber flowering: green bud, green-yellow bud, yellow bud, and flowering.Five individual plants were pooled to form one biological replicate, and three biological replicates were prepared for each tissue.All samples were stored at −80 • C for gene expression detection.
Extraction of total RNA was performed using the RNAprep Pure Plant Kit with DNase I (Tiangen, Beijing, China), and then the qualified RNA was used for the synthesis of first-strand cDNA using the HiScript 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China).For each gene, the RT-qPCR protocol with cross-intron primers was executed at a qTOWER3 system (Analytik Jena, Jena, Germany) in triplicate (Table S8).The cucumber elongation factor 1 alpha (CsEF1α) gene, as an internal control, was used to calculate the relative transcription levels of target genes using the 2 −∆∆Ct method.The statistical analysis of gene expression was performed by using SPSS software (Version 18.0).
Clustering of Time-Series and Correlation Analysis Based on RNA-Seq Data
The RNA-seq data published by Sun et al. [40] were acquired from the NCBI Gene Expression Omnibus database (GEO accession: GSE76358), which referred to the corollas at 1, 3, 4, and 5 DAL from the normal ovary.All protein sequences of the differentially expressed genes were submitted as queries to the PlantTFDB database [41] to screen the TFs.Also, the CsPEBP genes in this study were matched in the RNA-seq data using a BLASTp search.Finally, a total of 8 cucumber PEBP genes and 1251 TF genes were identified and shown in Table S9.Gene expression levels were measured using transcripts per million (TPM) values.
Clustering of time-series analysis was performed using the Short Time-series Expression Miner (STEM) program [42], and the maximum number of model profiles was set to fifty.
Pearson correlation coefficients of expression levels between CsPEBPs and TFs were estimated using GraphPad Prism 9 and visualized as a radar chart.
Conclusions
In short, 65 PEBP genes were comparatively identified and characterized in seven cucurbit crops, including their evolutionary, structural conservation, and functional diversification.A distinct PEBP clade was noticed, encoding the conserved domain in bacteria and archaea.CsaPEBP1 was highly expressed in male and female flowers, while CsaPEBP5, CsaPEBP6, and CsaPEBP7 were active in cucumber males.All four genes might function at the later stages of corolla opening.Furthermore, several cucumber TFs were inferred to be partners of FT or other PEBPs; of these, Csa4G037610 and Csa7G64580 might have more vital binding ability due to the tandem repeats of NF-YB and GATA in CsaPEBP2 and CsaPEBP5 promoters.Together, these outcomes enrich our understanding of the primary characteristics of the PEBP family in cucurbit crops and provide novel thoughts for flowering regulation through TFs.
Figure 1 .
Figure 1.Localization and synteny of the PEBP genes in seven cucurbit crops.(a) ClaPEBPs in the watermelon genome.(b) CmePEBPs in the melon genome.(c) CsaPEBPs in the cucumber genome.(d) CarPEBPs in the cushaw genome.(e) CmaPEBPs in the winter squash genome.(f) CmoPEBPs in the pumpkin genome.(g) CpePEBPs in the zucchini genome.Gene pairs with a syntenic relationship were joined by the red lines.Tandem duplicated genes were laid in a red box.
Figure 1 .
Figure 1.Localization and synteny of the PEBP genes in seven cucurbit crops.(a) ClaPEBPs in the watermelon genome.(b) CmePEBPs in the melon genome.(c) CsaPEBPs in the cucumber genome.(d) CarPEBPs in the cushaw genome.(e) CmaPEBPs in the winter squash genome.(f) CmoPEBPs in the pumpkin genome.(g) CpePEBPs in the zucchini genome.Gene pairs with a syntenic relationship were joined by the red lines.Tandem duplicated genes were laid in a red box.
Figure 2 .
Figure 2. Evolutionary tree of FT/PEPB-related proteins in Arabidopsis and seven cucurbit crops.
Figure 3 .
Figure 3.A syntenic map of PEBP genes among seven cucurbit crops.
Figure 4 .
Figure 4. Pattern diagram of PEBP family proteins from seven cucurbit crops.Red rectangle indicates amino acid residues that interact with 14-3-3 protein.The green rectangle indicates a key amino acid residue that determines MFT-like, FT-like and TFL-like proteins.Black boxes represent the conserved DPDxP, GxHR motif and L/IYN.Underlines represent segments A, B, C and D, respectively.
Figure 4 .
Figure 4. Pattern diagram of PEBP family proteins from seven cucurbit crops.Red rectangle indicates amino acid residues that interact with 14-3-3 protein.The green rectangle indicates a key amino acid residue that determines MFT-like, FT-like and TFL-like proteins.Black boxes represent the conserved DPDxP, GxHR motif and L/IYN.Underlines represent segments A, B, C and D, respectively.
Figure 5 .
Figure 5. Prediction of cis-acting elements in the cucurbit PEBP gene promoters.
Figure 5 .
Figure 5. Prediction of cis-acting elements in the cucurbit PEBP gene promoters.
Int. J. Mol.Sci.2024,25, x FOR PEER REVIEW 9 of expression levels of CsaPEBP1, CsaPEBP2, CsaPEBP5, CsaPEBP6, and CsaPEBP7 we found to match the RNA-seq data reported by Li et al.[17], indicating that these gen could play stable roles during the growth and development of cucumber, especially in t development of floral organs.
Figure 6 .
Figure 6.Expression profiles of CsaPEBP1-CsaPEBP8 genes (a-h) in cucumber leaf, stem, tend male, female and ovary.Grey columns indicate the relative gene expression levels determined us RT-qPCR, and red lines represent the transcript abundance changes calculated using the Reads P Kilobase per Million mapped reads (RPKM) method.Lowercase letters (for example, a, b, c, e above the columns represent significant differences at the 0.05 level based on Tukey's test.
Int. J. Mol.Sci.2024,25, x FOR PEER REVIEW 10 of 17 genes in the ovary did not exhibit sufficient regularity.These results suggested that Csa-PEBP1, CsaPEBP2, CsaPEBP5, CsaPEBP6, and CsaPEBP7 might be potential factors involved in regulating cucumber flowering, mainly functioning at the later stages: the yellow bud and flowering.
Figure 7 .
Figure 7. Expression patterns of CsaPEBP genes in the process of cucumber flowering.(a) Progression of corolla opening in the normal male (M; top) and female (C, corolla and O, ovary; bottom), including four stages: green bud (1), green-yellow bud (2), yellow bud (3), and flowering (4).Bar = 1 cm.(b) Heatmap (left) showing different expression levels of five CsaPEBP genes in male and female flowers at four developmental stages.Changes in gene expression levels are shown in color as the scale.Mini Line Chart (right) shows the mean expression trend of each gene.
Figure 7 .
Figure 7. Expression patterns of CsaPEBP genes in the process of cucumber flowering.(a) Progression of corolla opening in the normal male (M; top) and female (C, corolla and O, ovary; bottom), including four stages: green bud (1), green-yellow bud (2), yellow bud (3), and flowering (4).Bar = 1 cm.(b) Heatmap (left) showing different expression levels of five CsaPEBP genes in male and female flowers at four developmental stages.Changes in gene expression levels are shown in color as the scale.Mini Line Chart (right) shows the mean expression trend of each gene.
Figure 8 .
Figure 8. Screening for TFs that interact with CsaPEBPs.(a) Time-series clusters of CsaPEB TFs based on RNA-seq data.The horizontal axis represents four developmental stages in the las of a female flower: green bud (C1), green-yellow bud (C2), yellow bud (C3), and flowerin Trend lines are color-encoded with gray shades denoting high membership values of genes b ing to the time-series cluster, and the red lines indicate the expected trend.(b) Radar chart Pearson correlation coefficient between expression levels of TFs and CsaPEBPs.(c) Number o scription factor binding sites (TFBSs) in CsaPEBP gene promoters.(d) Putative TFs interactin CsaPEBPs.
Figure 8 .
Figure 8. Screening for TFs that interact with CsaPEBPs.(a) Time-series clusters of CsaPEBPs and TFs based on RNA-seq data.The horizontal axis represents four developmental stages in the corollas of a female flower: green bud (C1), green-yellow bud (C2), yellow bud (C3), and flowering (C4).Trend lines are color-encoded with gray shades denoting high membership values of genes belonging to the time-series cluster, and the red lines indicate the expected trend.(b) Radar chart of the Pearson correlation coefficient between expression levels of TFs and CsaPEBPs.(c) Number of transcription factor binding sites (TFBSs) in CsaPEBP gene promoters.(d) Putative TFs interacting with CsaPEBPs. | 2024-04-01T15:57:15.808Z | 2024-03-29T00:00:00.000 | {
"year": 2024,
"sha1": "b8cf77528f1a58b057504e7a940c8e1dfc0c60eb",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/1422-0067/25/7/3815/pdf?version=1711701152",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "75c6f620a5cdfbe61fd6b8b077b2c633e68b8615",
"s2fieldsofstudy": [
"Biology",
"Environmental Science"
],
"extfieldsofstudy": []
} |
38279881 | pes2o/s2orc | v3-fos-license | Evaluation of dietary electrolyte balance on nursery pig performance1
Abstract Increasing dietary electrolyte balance (dEB) has been reported to linearly improve pig growth performance up to approximately 200 to 250 mEq/kg. However, recent data indicate that increasing dietary dEB reduced growth performance of nursery pigs. To attempt to solve this discrepancy, a total of 2,880 weanling pigs (327 × 1,050; PIC, Hendersonville, TN; 5.2 kg initial BW) were used to determine the effects of increasing dEB on nursery pig performance. Pens of pigs were blocked by BW and gender on arrival. Within block, pens were randomly assigned to one of four dietary treatments. There were 30 pigs per pen (60 pigs per double-sided feeder) and 12 replications (feeder) per treatment. Dietary treatments were fed in two phases. The phase 1 diet was based on corn–soybean meal, contained dried distillers grains with soblubles (DDGS), spray-dried whey, and specialty protein sources, and was fed from days 0 to 8. The phase 2 (days 8 to 21) diets contained corn, soybean meal, and DDGS with reduced amounts of specialty protein sources. Dietary electrolyte balance was determined using the following equation: dEB = [(Na × 434.98) + (K × 255.74) − (Cl × 282.06)] mEq/kg. The dEB of the four phase 1 diets were 84, 137, 190, and 243 mEq/kg, and dEB of the four phase 2 diets were 29, 86, 143, and 199 mEq/kg. After feeding experimental diets for 21 day, a common, commercial corn–soybean meal diet was fed to all pigs from days 21 to 35 and contained a dEB of 257 mEq/kg. During days 0 to 8, increasing dEB increased (quadratic, P < 0.05) ADG, ADFI, and G:F. From days 8 to 21, increasing dEB improved ADG (quadratic, P = 0.022) and ADFI (linear, P = 0.001), resulting in an improvement (quadratic, P = 0.001) in G:F. Overall (days 0 to 21), increasing dEB increased (linear, P < 0.05) ADG, ADFI, and improved (quadratic, P < 0.001) G:F. When a common diet was fed to all pigs from days 21 to 35, there was a linear reduction in ADG and G:F with increasing dietary dEB, but no effect of ADFI. For the overall nursery period (days 0 to 35), increasing dEB from days 0 to 21 increased (linear, P < 0.001) ADG and final BW, which was the result of increased (quadratic, P < 0.05) G:F and marginally greater (linear, P = 0.077) ADFI. In conclusion, increasing dietary dEB up to 243 and 199 mEq/kg (in phases 1 and 2, respectively) in nursery diets improved growth performance of weanling pigs.
INTRODUCTION
Electrolytes are key minerals that can be defined as chemical substances that separate when dissolved in fluids to form positive (cation) and negative (anion) ions. Previous research has indicated that cations and anions are closely linked to the alkalinity and acidity of body fluids (Mushtaq and Pasha, 2013). In particular, the monovalent minerals Na, K, and Cl are considered strong ions because of their ability to exert significant shifts in acid-base homeostasis (Mongin, 1981). Therefore, maintaining an appropriate dietary electrolyte balance (dEB) is critical for pigs to achieve optimal growth performance.
Typically, ingredients such as calcium carbonate, calcium phosphate, and sodium chloride are used in swine diets to meet mineral requirements, but they also contribute to the dEB, thus potentially altering acid-base homeostasis and growth performance of pigs (Patience et al., 1987;Haydon et al., 1990;DeRouchey et al., 2003). Traditionally, the optimal dEB for pigs is reported to be approximately 250 mEq/kg (NRC, 2012), but limited research exists in this area. Recently, Guzmán-Pino et al. (2015) used CaCl 2 to reduce dEB and reported that nursery pigs had reduced ADG and G:F when dEB exceeded 150 mEq/kg; in addition, ADG was improved by 48.7% by decreasing dEB from 269 to 16 mEq/kg. Because many common U.S. nursery diet formulations exceed 150 mEq/kg, these results are important. However, the results from Guzman-Pino et al. (2015) contradict those who reported improved growth performance with increasing dietary dEB (Patience et al., 1987;Dersjant-Li et al., 2001;DeRouchey et al., 2003). To solve the discrepancies among literatures, this study aimed to evaluate the impact of dEB on growth performance of nursery pigs in a commercial scale research facility.
General
The Kansas State University Institutional Animal Care and Use Committee approved the protocol used in this experiment. The experiment was conducted at a commercial nursery in southeast MN. Pigs were housed in pens (1.82 × 3.35 m) that were equipped with a double-sided, five-hole stainless steel dry feeder and one cup waterer for ad libitum access to feed and water. The facility was equipped with a computerized feeding system (FeedPro; Feedlogic Corp., Willmar, MN) that delivered and recorded daily feed additions of experimental diets as specified. Diets were manufactured at commercial feed mills for pelleted diets (Hubbard Feeds, Mankato, MN) and mash diets (Bixby Feed Mill, Blooming Prairie, MN).
Experimental Design
A total of 2,880 weanling pigs (327 × 1,050; PIC, Hendersonville, TN; initially 5.2 kg BW) were used in a 35-d growth trial with 12 replications (feeder) per treatment. Each fence line feeder provided feed to one pen of barrows and one pen of gilts and each pen initially contained 30 pigs. On arrival, pens of pigs were blocked by average individual pig BW and gender. Within blocks, feeders (combination of two pens) were randomly assigned to one of four dietary treatments. Dietary treatments were fed in two phases. The phase 1 diet was based on corn-soybean meal and contained dried distillers grains with solubles (DDGS; >6 and <9% oil, NRC, 2012), spray-dried whey, and specialty protein sources and was fed from days 0 to 8 (Table 1). The phase 2 diets contained corn-soybean meal-DDGS with reduced amounts of specialty protein sources and were fed from days 8 to 21 (Table 2). Diets 1 to 4 were formulated to contain increasing dEB ranging from 84 to 243 mEq/kg during days 0 to 8 and 29 to 199 mEq/kg from days 8 to 21 after weaning. The lowest dEB diets were achieved by adding 1.17% and 1.25% CaCl 2 from days 0 to 8 and from days 8 to 21, respectively. Dietary Ca concentrations were maintained in the three highest dEB diets, but increased in the low dEB diet because of the increasing level of CaCl 2 . The following equation derived by Mongin (1981): dEB, mEq/kg = (Na × 434.98) + (K × 255.74) − (Cl × 282.06) was used to determine dEB. After feeding experimental diets for 21 d, a common diet was fed from days 21 to 35 (Table 3) to all pigs and was a typical nursery diet fed in commercial production with a dEB of 257 mEq/kg. Nutrient loading values of the ingredients were adopted from NRC (2012), and diets were formulated to contain similar metabolizable energy with the exception of available P which were based on NRC (1998). Days 0 to 8 diets were prepared in pellet, whereas days 8 to 35 diets were provided in meal form. The ingredient mineral, electrolyte concentrations, energy, AA, and standardized ileal digestible AA coefficients for ingredients reported by the NRC (2012) were used in formulating diets. Pigs and feeders were weighed on days 0, 8, 15, 21, and 35 of the trial to determine ADG, ADFI, and G:F.
Diet Sampling and Analysis
Complete diet samples in phases 1 and 2 were obtained from a minimum of six individual feeders per treatment, blended, and frozen at −20 °C for subsequent analysis. Composite samples of diets were split using a riffle splitter (Humboldt Mfg. Co., Norridge, IL) and processed through a 1-mm screen in a Willey mill (Thomas Scientific, Swedesboro, NJ) prior to analysis. All diet samples were submitted for analysis of DM (method 935.29; AOAC International, 2012), CP (method 990.03; AOAC International, 2012), and ether extract (method 920.39; AOAC International, 2012) for preparation and analyzed using an ANKOM XT20 Fat Analyzer (Ankom Technology, Fairport, NY); Ca, P, and K (method 968.08; AOAC International, 2012) for preparation using ICAP 6500 (ThermoElectron Corp., Waltham, MA); Na (method 990.08; AOAC International, 2012); and Cl (method 969.10; AOAC International, 2012; Ward Laboratories Inc., Kearney, NE).
Statistical Analysis
Feeder (one pen of barrows and one pen of gilts) served as the experimental unit. Data were analyzed as a randomized complete block design with dietary treatment as a fixed effect and block serving as the random effect in the model. Preplanned linear and quadratic polynomial contrasts were used to determine the effects of increasing dEB. The coefficients for the unequally spaced linear and quadratic contrasts were derived using PROC IML in SAS. Results were considered significant at P ≤ 0.05 and marginal effects between P > 0.05 and P ≤ 0.10. The PROC GLIMMIX procedure in SAS (SAS Institute, Inc., Cary, NC) was used for statistical analysis.
RESULTS
Analysis of experimental diets indicated that most nutrients were similar to formulated values for phase 1 diets (Table 4). Analyzed values for Na, Ca, K, and Cl were greater across dietary treatments in phase 2 (Table 5) than formulated values; however, the treatment pattern of dEB targeted was ultimately maintained across dietary treatments fed from days 0 to 8 and from days 8 to 21.
DISCUSSION
For the current experiment, dEB was determined by examining the interrelationship among the monovalent micromineral ions Na, K, and Cl: dEB, mEq/kg = (Na × 434.98) + (K × 255.74) -(Cl × 282.06) (Mongin, 1981). Previous literature examining the effects of dEB on pigs have generally demonstrated improvements in growth rate when dEB was increased with the NRC (2012) reporting the optimal electrolyte balance for pigs to be approximately 250 mEq/kg. However, there are discrepancies within the literature in regard to an optimal dEB range for growing pigs. Patience et al. (1987) conducted an experiment in which six levels (−85, 0, 100, 175, 277, and 341 mEq/kg) of dEB were fed to growing pigs (initially 15 kg BW) for 28 d with the supplemental calcium chloride included in the three low dEB diets and sodium bicarbonate used in the three high dEB diets. The authors reported that performance was optimized when pigs were fed a diet with a dEB of 175 mEq/kg. Similarly, Dersjant-Li et al. (2001) conducted an experiment in which pigs (initially 9 kg BW) were fed three dEB concentrations (−100, 200, and 500 mEq/kg) by including CaCl 2 in the low dEB diet and NaHCO 3 in the other two dEB diets. They observed greater ADG, ADFI, and G:F when pigs were fed diets with a dEB of 200 or 500 mEq/kg compared with −100 mEq/kg.
In contrast, Patience and Chaplin (1997) compared diets containing −20, 104, and 163 mEq/kg of dEB fed to pigs (initially 35 kg BW). For that study, supplemental salts CaCl 2 , NaHCO 3 , and KHCO 3 were included to alter dEB. Results indicated a tendency for improved growth when pigs were fed a dEB diet of −20 mEq/kg compared with pigs fed either 104 or 163 mEq/kg. Reason for the discrepancy between this and the aforementioned studies is unclear but may have been related to limit feeding. Feed intake was equalized across all treatments by feeding an amount that was equal to the pigs with the lowest feed intake compared with ad libitum feed intake allowed in other studies. Recently, Guzmán-Pino et al. (2015) conducted an experiment to determine the influence of dEB on growth performance of nursery pigs (initially 13 kg BW). In their study, four dEB concentrations (16, 133, 152, and 269 mEq/kg) were fed to pigs from 21 to 37 d after weaning with dEB altered by including CaCl 2 and NaHCO 3, respectively. The authors observed that increasing dEB decreased ADG and G:F when dEB exceeded 150 mEq/kg with a 48.7% improvement in ADG by decreasing dEB from 269 to 16 mEq/kg.
In this study, decreasing dEB in nursery diets resulted in a reduction in ADG and final BW, which was the result of marginally lower ADFI and poorer feed efficiency. A possible explanation for the lower feed intake in pigs fed the low dEB diet could be contributed to the CaCl 2 used to lower dEB. Yen et al. (1981) indicated that dietary CaCl 2 limited feed intake in pigs through Cl-induced metabolic acidosis. Furthermore, sensory tests using humans have indicated calcium chloride itself is perceived as producing a bitter and metallic off-flavor (Lawless et al., 2003(Lawless et al., , 2004. In addition, previous preference work conducted by Guzmán-Pino et al. (2015) examining a low (−16 mEq/kg; with CaCl 2 ) and high (388 mEq/kg; without CaCl 2 ) dEB diet indicated that pigs fed the low dEB diet had decreased preference compared with the high dEB diet. However, when similar diets were used in a growth performance trial, performance decreased when pigs were fed levels above 150 mEq/kg of the diet. Personal communication with the authors indicated that a similar unprotected CaCl 2 source, and the same equation was used to calculate the dEB as in our study. In addition, diets fed by Guzmán-Pino et al. (2015) were not analyzed for Na, K, and Cl, thus making it difficult to assess whether dEB concentrations were similar to estimated values. Another possible explanation includes starting the experimental diets at weaning (the study herein) vs. 21 d after weaning (Guzmán-Pino et al., 2015). Concentrations of Na, K, and Cl were generally lower in the diets used by Guzmán-Pino et al. (2015) than by Lei et al. (2017) or in the present study.
In agreement with the present study, Lei et al. (2017) reported that weanling pigs (initially 8 kg BW) fed diets with increasing dEB (0, 83, 166, and 250 mEq/kg) had improved ADG and ADFI. The authors altered dEB with the addition of CaCl 2 or NaHCO 3 . Furthermore, increasing dEB resulted in improvements in apparent total tract digestibility of DM and N in pigs fed the high dEB diets (166 and 250 mEq/kg) compared with the low dEB diet (0 mEq/kg). Although digestibility measurements were not quantified in the current study, the improvement in feed efficiency as dEB increased could be a result of an improvement in nutrient digestibility.
It is worth noting that the majority of literatures discussed earlier, including the present study, only considered monovalent ions (Na, K, and Cl) in the calculation of dEB as it is most commonly used in swine diet formulation. More comprehensive estimation of dEB has been proposed that takes into account the contribution of divalent ions, such as Ca, Mg, S, and P. Future research is in need to investigate whether the discrepancies among studies may be a result of differences in dEB originating from divalent ions.
Moreover, interestingly, we observed a linear improvement in ADG and G:F during phase 3 (days 21 to 35) in pigs previously fed low dietary dEB. It is possible that pigs fed lower dietary dEB during days 0 to 21 with decreased growth performance had compensatory gain during the post-treatment period when dietary dEB was increased to 257 mEq/kg. However, further research is needed to verify this hypothesis.
In conclusion, results from this study indicate that increasing dEB up to 243 and 199 mEq/kg of diet from weaning to day 21 after weaning resulted in an increase in ADG and final BW, which was the result of a tendency for an increase in ADFI and G:F. | 2017-10-24T11:30:17.687Z | 2018-07-14T00:00:00.000 | {
"year": 2018,
"sha1": "05abc90ff300dc862ec9cb01b6bf75a263b41169",
"oa_license": "CCBYNC",
"oa_url": "https://academic.oup.com/tas/article-pdf/3/1/378/28250427/txy090.pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "59eae96dc3af2d8066a177e82f1d091d3959d017",
"s2fieldsofstudy": [
"Agricultural And Food Sciences"
],
"extfieldsofstudy": [
"Medicine",
"Chemistry",
"Biology"
]
} |
259145017 | pes2o/s2orc | v3-fos-license | Stochastic dynamics of multi-waterfall hybrid inflation and formation of primordial black holes
We show that a hybrid inflation model with multiple waterfall fields can result in the formation of primordial black holes (PBHs) with an astrophysical size, by using an advanced algorithm to follow the stochastic dynamics of the waterfall fields. This is in contrast to the case with a single waterfall field, where the wavelength of density perturbations is usually too short to form PBHs of the astrophysical scale (or otherwise PBHs are overproduced and the model is ruled out) unless the inflaton potential is tuned. In particular, we demonstrate that PBHs with masses of order 1020 g can form after hybrid inflation consistently with other cosmological observations if the number of waterfall fields is about 5 for the case of instantaneous reheating. Observable gravitational waves are produced from the second-order effect of large curvature perturbations as well as from the dynamics of texture or global defects that form after the waterfall phase transition.
Introduction
The seeds of cosmological large-scale structures are generated by curvature perturbations during inflation, and their signals are imprinted and observed as the temperature anisotropies of the cosmic microwave background (CMB).The spectrum of curvature perturbations is not exactly flat, which is consistent with the prediction of slow-roll inflationary scenario [1].However, there are several puzzles in the formation of small-scale structures, such as the formation of supermassive black holes (SMBHs) [2,3] and black holes (BHs) observed by the LIGO-Virgo-KAGRA colllaboration [4,5].Although their evolution is highly non-linear, it is said that their number densities are in tension with the standard scenario of BH formation in astrophysics.One therefore needs to consider a primordial origin of such BHs called primordial black holes (PBHs) [6][7][8] (see also the recent review article on PBH [9]).It is also discussed that astrophysical size PBHs of order 10 17-23 g can explain the dark matter (DM) in the Universe [10] (see, e.g., Refs.[11][12][13][14]).Among several scenarios of the formation of PBHs, it would be minimal to consider the formation of those small-scale structures by generating large curvature perturbations in the small scales during inflation (see, e.g., Refs.[15][16][17][18][19][20] and also references in Ref. [9]). 1 The curvature perturbations are converted to density perturbations after inflation, and the dense regions collapse to form PBHs when the corresponding wavelength enters the horizon [6][7][8].
To generate large curvature perturbations, a bump or at least a very flat region for inflaton potential is required [16-18, 20, 31-49].This can be naturally realised by hybrid inflation models because a waterfall field has a vanishing mass at the waterfall phase transition [50].Remarkably, it can generate large density perturbations at small scales that result in the formation of PBHs [51][52][53].Since large density perturbations are generated during the waterfall phase transition, the waterfall field experiences the stochastic dynamics that requires detailed numerical simulations (see Refs. [54][55][56][57][58][59][60][61][62][63] for the first papers on the subject) and was omitted in some literature.Several studies on hybrid inflation models reveal that the stochastic effect on the waterfall fields significantly alters the predictions of hybrid inflation models (see, e.g., Ref. [64]).Ref. [65] provided an analytic method to calculate the spectrum of curvature perturbations in a model with a single waterfall field, considering the stochastic effects under some approximations.Their analytic formula is partially consistent with the numerical simulations for the stochastic dynamics performed in Ref. [66].These studies are performed in models with a single waterfall fields, and we thereby concluded that the peak amplitude of curvature perturbations is too large to provide consistent astrophysical consequences even though the stochastic effect is considered. 2 Our recent work [67] shows that this problem can be evaded by considering a higher-order term for the inflaton potential, though it requires a tuning in a parameter.As another aspect, the hybrid inflation model with a single waterfall field suffers from the so-called domain-wall problem because the Z 2 symmetry is spontaneously broken after the waterfall phase transition.This problem can be evaded by introducing a small explicit Z 2 -symmetry breaking term, which however may affect the dynamics of waterfall phase transition (see, e.g., Ref. [68]).
In this paper, we analyze a hybrid inflation model with multiple waterfall fields, considering the stochastic effect on the waterfall fields.They are assumed to be symmetric under O(N ) symmetry.The idea to introduce multiple waterfall fields was considered in Ref. [69], although they did not consider the stochastic dynamics of waterfall fields.We show that the spectrum of curvature perturbations can be reduced with a fixed peak frequency for this case.Alternatively, the peak frequency can be reduced for a fixed peak amplitude of curvature perturbations.This can result in the formation of PBHs at astrophysically interesting mass scales when the number of waterfall fields is sufficiently large.We employ a sophisticated numerical method proposed and developed in Refs.[70][71][72][73][74] to incorporate the stochastic dynamics of multiple waterfall fields.We also confirm that the numerical results can be qualitatively understood by the analytical calculation of Ref. [65], extending it for the case with multiple waterfall fields.In particular, the peak amplitude of curvature perturbations is proportional to the inverse of the number of waterfall fields.We provide a semi-analytic formula for the spectrum of curvature perturbations, which is useful to consider the formation of PBHs.From the formula, we particularly demonstrate that PBHs with mass of order 10 20 g can form after hybrid inflation consistently with other cosmological observations if the number of waterfall fields is about 4 or 5 for the case of instantaneous reheating.Moreover, the domain wall problem is absent for the case with O(N ) symmetric waterfall fields.The spontaneous symmetry breaking of O(N ) symmetry after the waterfall phase transition results in the formation of texture (for N = 4) or global defects (for N ≥ 5).Gravitational waves (GWs) are generated via the dynamics of those structure as well as the second-order effect of large curvature perturbations.We demonstrate that the GW signals can be observed by future GW experiments, including LISA.
The rest of the paper is organised as follows.In Sec. 2, we first summarise the formalism for the stochastic equations for O(N ) symmetric fields and write down the Fokker-Planck and Langevin equations.These equations are solved analytically and numerically in Sec. 3, in hybrid inflation models with multiple waterfall fields.We provide a semi-analytic formula for the spectrum of curvature perturbations, where numerical coefficients are fitted by the results of numerical calculations.Furthermore, in Sec. 4 we use it to consider the formation of PBHs and the prediction of GW spectrum.Sec. 5 is devoted to discussion and conclusions.
Stochastic equations for symmetric fields
Let us first see the Fokker-Planck (FP) and Langevin equations that describe the stochastic dynamics of O(N ) symmetric fields during inflation.One will find that the system reduces to the one only for the radial mode in the slow-roll limit with an effective centrifugal force proportional to the number N of the fields due to the diffusion term.If the N fields correspond to the waterfall fields in the hybrid inflation, such a centrifugal force prevent them from staying around the origin of the potential and accordingly suppresses the curvature perturbation as we will see.
We employ the slow-roll approximation throughout this paper.The Langevin equation for N (canonical) inflaton fields ψ i (i = 1, 2, • • • , N ) (which we will identify as waterfall fields in the context of hybrid inflation model) is expressed as (see, e.g., Ref. [72]) where N is the e-folding number as the time variable, M Pl is the reduced Planck mass, V is the inflatons' potential, is its first derivative, and ξ i (N ) is an independent stochastic noise: 2) The corresponding FP equation for the inflatons' probability distribution This is equivalent to the following adjoint FP equation for the distribution P FPT (N | ψ) of the first passage time N from a certain field space point ψ to the end-of-inflation surface: Note that the forward e-folding number N is a deterministic time variable while the backward one N is stochastic as it depends on the realisation of the inflatons' stochastic dynamics.
Let us then suppose that the inflatons are symmetric and the potential (and hence all physical variables in the slow-roll limit) depends only on the radius ψ r = ψ 2 i .In this case, P FPT should solely be a function of N and ψ r .Therefore, the adjoint FP equation reduces to (refer also to Ref. [75]) (2.5) One finds a new term ∝ V N −1 ψr ∂ ψr from the Laplacian V ∂ 2 i .It can be understood as an effective single field system with a new centrifugal force dictated by the FP equation or the corresponding Langevin equation with a normalised noise We below see the effect of this centrifugal force in the hybrid inflation model.
Multi-waterfall hybrid inflation
In order for a general discussion, we consider the (1 + N )-field hybrid inflation with a phenomenological potential [65,66] characterised by the five dimensionful parameters Λ, M , ϕ c , µ 1 , and µ 2 .Here and hereafter, the summation in terms of i is implicit for ψ 2 i (≡ i ψ 2 i ).The inflaton potential is expanded around the critical point ϕ c and the higher-order terms than the quadratic one are neglected as we are mainly interested in the dynamics around the waterfall phase transition.Among the inflatons ϕ and ψ i (i = 1, 2, • • • , N ), the latter is often called waterfall fields as they terminate inflation by their tachyonic instability in the standard scenario.For a large ϕ, the waterfall fields stay around the origin and inflation occurs by the potential energy (≃ Λ 4 ).The Hubble parameter during inflation is given by H inf ≃ Λ 2 /( √ 3M Pl ).When the field ϕ reaches the critical point ϕ c , ψ i becomes tachyonic (called the waterfall phase) and starts to roll down toward the potential minimum ψ 2 i = M 2 .Depending on the parameters, the waterfall phase does not necessarily end quickly but can last as a second phase of inflation.The inflation ends when the slow-roll condition is violated.It is often controlled by the second slow-roll parameter along the waterfall direction In particular, the potential of interest is symmetric with respect to ψ i and therefore the model is effectively a two-field system of ϕ and ψ r = ψ 2 i with the centrifugal force disucssed around Eq. (2.7) in the slow-roll limit.We below study this two-field system.
Analytic estimations
The spectrum of curvature perturbations can be estimated by an analytic method proposed in Ref. [65] (refer also to Refs.[66,67]) for the hybrid inflation model.We extend their analysis to the case with multiple waterfall fields, including the leading-order effect of quadratic term for the inflaton potential.
We first estimate the distribution of the waterfall fields at the critical point, which determines the dynamics in the waterfall phase and then the final curvature perturbation.It follows the Langevin equations, with the independent noise The stochastic noise for the inflaton ϕ is usually negligible for the waterfall dynamics, whereas that for the waterfall fields is important around the waterfall phase transition.We denote the e-folding number at the time of the waterfall phase transition as N c .Neglecting the stochastic noise for ϕ, it is solved as for N ≈ N c .Eq. (3.2) derives the evolution equation for ⟨ψ 2 r ⟩ as where we assume ⟨ψ 2 r ⟩ ≪ M 2 and define With an approximation V ≃ Λ 4 , it can be solved as where we adopt the asymptotic condition x 0 e −t 2 dt is the error function.ψ r,c is defined by hence represents the amplitude of ⟨ψ 2 r ⟩ at the phase transition (N = N c ).This is enhanced by a factor of N for multiple waterfall fields.
The solution Eq. (3.7) approaches to 2ψ 2 r,c exp 8 2) ≡ N cl .The time scale N cl represents the time after which the classical dynamics (namely the effect of the first term in the right-hand side of Eq. (3.2)) begins to dominate.We will observe later that Π should be 10-40, so that N c = 7-28 for the case we are interested in.The stochastic effect is therefore important even well after the waterfall phase transition.As will be observed below, large curvature perturbations are generated during this regime, so that the numerical simulation for the FP equation is necessary to calculate the spectrum of curvature perturbations.This will be done in Sec.3.2.In this subsection, we nevertheless derive an analytic rough estimation on the curvature perturbation by omitting the stochastic effect and solving the classical equation of motion in the waterfall phase with the initial condition of ψ r = ψ r,c at N = N c , following Refs.[65][66][67].We will see that it indeed provides some information of the spectrum of the curvature perturbation.
Here we quote some results presented in Ref. [67] with some corrections by a factor of N including the leading-order correction from the quadratic potential for the inflaton.The e-folding number from the time of the waterfall phase transition N c to the end of inflation N end is calculated as where c (< 1) and χ 2 are given by c ≡ The δN approach gives the spectrum of curvature perturbations as where the backward e-folds N k is defined by the time when k = aH and ∆N 1 is the solution to the following equation: The peak amplitude reads where we neglect O(c/χ 2 ) terms in the second line.We will see in Sec. 4 that the parameter χ 2 is ∼ 10 almost constantly for the parameters of interest and also µ 2 is fixed to ≃ 10M Pl by the CMB observation.Therefore, the peak amplitude P (peak) R is uniquely determined by the corresponding peak wavelength (through N PT ) except for the suppression due to N .If N is equal to unity or two, the peak amplitude is too large for a desirable amount of PBHs with astrophysical scales [65,66].This challenge can be evaded when we consider multi-waterfall fields because of the factor of 1/N in Eq. (3.15).
As we commented above, the stochastic effect is not negligible for several e-foldings after the waterfall phase transition.Nevertheless, the above analytic form works reasonably well as we confirm with the detailed numerical calculations in the stochastic formalism in the next subsection.We will use the analytic formula (3.12) as a fitting function of the numerical result, regarding N PT and P (peak) R as fitting parameters rather than giving them by Eqs.(3.9) and (3.14).
Numerical implementation of the stochastic-δN formalism
In the full stochastic approach to inflation, the power spectrum of the curvature perturbation can be calculated in the so-called stochastic-δN formalism [70][71][72][73][74].The essential idea is that the probability distribution of the stochastic backward e-folds N is nothing but that of the curvature perturbation thanks to the δN formalism [76][77][78][79].The power spectrum is obtained by extracting the variance of N corresponding to the scale of interest.Specifically, it is given by the field-space integration (see Eq. (3.19) of Ref. [74]) While the backward probability P bw is formulated by a combination of the probability density function of ϕ (the solution of the FP equation) and that of N (the solution of the adjoint FP equation) in a rigorous way (see Ref. [73]), it is easy to sample the solution of the Langevin equation in a practical, numerical simulation.There, the backward probability is approximated as where S is the sampling number, ϕ i (N ) is the ith solution of the Langevin equation, and N i is the first passage time from the initial field value ϕ 0 for that solution.Its derivative can be further approximated by the finite difference: with a certain small step ∆N .Substituting these approximations back into the formula (3.16), the power spectrum is approximated as where Note that the average δN 2 ϕ 0 → ϕ ± i (k) is defined only for sample paths passing the point ϕ ± i (k), which is a non-trivial condition in a two or more than two fields case.We further try to approximate it by the average without a condition.Let us first define the total perturbation for paths passing ϕ ± i (k) by δN (ϕ 0 ) Mean squares of both sides read where we used the Markovian property of the Langevin equation for the last line.We then expect that the difference between δN 2 (ϕ 0 ) ϕ + i (k) and δN 2 (ϕ 0 ) ϕ − i (k) will be suppressed as O(∆N 2 ).This would be statistically justified because for typical points ϕ ± i (k), δN 2 (ϕ 0 ) ϕ ± i (k) asyptotes to a function only of ϕ 0 and becomes stationary against the change between ϕ + i (k) and ϕ − i .Therefore, one finds at the leading order in ∆N .
The variances on the right-hand side are found as a solution of the adjoint FP equation.In fact, recalling that the average ⟨f (N (ϕ))⟩ of an arbitrary function f (N (ϕ)) is defined by the probability density P FPT as the adjoint FP equation with the boundary condition P FPT (N | ϕ) = δ(N ) on the end-of-inflation surface ∂Ω leads to the following recursive partial differential equations: with the boundary condition ⟨N (ϕ)⟩ ϕ∈∂Ω = ⟨δN 2 (ϕ)⟩ ϕ∈∂Ω = 0.In this paper, we numerically solve these partial differential equations in the Jacobi method with use of the C ++ package, StocDeltaN [80] (see also the author's GitHub page [81]).
Results
We show the numerical results of the stochastic-δN formalism in this subsubsection.We fixed M and ϕ c by M = ϕ c / √ 2 = 10 16 GeV while µ 1 is determined by Π parameter through its definition (3.6).Λ and µ 2 are fixed by Eqs.(4.4) and (4.7) due to the CMB constraint.In the δN formalism, the end surface should be given by a uniform-density slice.We first solve the classical slow-roll equations of motion (Eq.(3.2) without the noise terms) from the initial condition ϕ = ϕ c and ψ = ψ r,c until η ψψ ≡ M 2 Pl V ψr ψr V reaches −2 (for sufficient violation of the slow-roll condition) for the first time and define the end-of-inflation energy density for the stochastc-δN scheme by the potential value at that time.
Fig. 1 shows the contours of ⟨N (ϕ, ψ r )⟩ and ⟨δN 2 (ϕ, ψ r )⟩ for Π 2 = 100 and N = 1 or 100 as solutions of the recursive partial differential equations (3.25) obtained by the StocDeltaN package.One finds that inflatons' sample trajectories represented by red lines pass the larger-ψ r region for N = 100 than N = 1 due to the noise-induced centrifugal force.The interesting feature is that though ⟨N ⟩ is not so sensitive to the number of waterfall fields N , the perturbation variance ⟨δN 2 ⟩ indeed decreases in the large N case.Therefore, one can expect that a large N can reduce P R , keeping its peak scale large enough.
Fig. 2 shows the resultant power spectra of curvature perturbations for several Π 2 and N .We adopt the sampling number S = 10 5 and the e-folds step ∆N = 0.5.Each error bar represents the propagated error in the finite difference equation (3.22) from the estimated standard error in the sampling averages 1 S S i=1 ⟨δN 2 (ϕ ± i (k))⟩.We also show the fitting formula (3.12) by the dotted lines with the corresponding colours.The fitting parameters N PT and P (peak) R are chosen by the numerically obtained peak wavenumber and peak amplitude.One finds that the fitting formula reasonably works well, particularly for a large Π 2 and a large N where the slow-roll approximation and the classical approximation (to neglect the noise terms) are better respectively.Table 1 shows the numerical results of the fitting parameters N PT and P (peak) R for Π 2 = 50, 100, 200, 500, 1000, and 2000 with N = 1, 3, 10, 100, and 1000.Those results are also shown in Fig. 3.The shaded region on the top-left panel represents the analytical result Eq. (3.14) multiplied by a factor of 1/3 for Π 2 ∈ (50, 2000).One finds that the analytic formula reasonably works again.Furthermore, the behaviour P (peak) R ∝ 1/N is confirmed by the numerical result.The thin dashed (solid) curves on the bottom panel represent the analytic results Eq. (3.9) in the leading ∼ Π (next-to-leading ∼ Π 2 ) order approximations with c = 0. Though the leading order approximation is not enough, the next-to-leading order approximation is well consistent with the numerical result.One can also see that N PT is almost independent of N and hence the degeneracy between N PT and P chosen by the numerically obtained peak wavenumber and peak amplitude.
Parameters of interest for astrophysical PBH formation
In this section, we discuss which parameters should be chosen to generate the desired mass of PBHs consistently with other observed quantities.We then apply our result of the previous section to calculate the PBH mass function.We also show a GW spectrum predicted in our model.for given Π 2 and N .50,2000).The thin dashed (solid) curve on the bottom panel represents the analytic formula Eq. (3.9) in the leading ∼ Π (next-to-leading ∼ Π 2 ) order approximations with c = 0.
CMB constraints
Since perturbations at the CMB scale exits the horizon before the waterfall phase transition, they are generated by the quantum fluctuation of the inflaton ϕ.The amplitude of curvature perturbations is therefore given by the standard textbook formula where is a slow-roll parameter.This should be consistent with the observed one: where k * (= 0.05 Mpc −1 ) represents the wavenumber at the pivot scale [1].This implies or In addition, using Eq.(4.3) and Eq.(3.11), we obtain Thus we expect χ 2 ∼ 10.
The spectral index in our model is provided by We can consider µ 2 ≃ 10M Pl to explain the observed spectral index of n s = 0.9649±0.0042[1].
Constraint from self-ordering scalar fields
Our model has the global O(N ) symmetry, which is spontaneously broken to O(N − 1) at the waterfall phase transition.For the case with N ≤ 4, topologically non-trivial field configurations form after the spontaneous symmetry breaking (SSB).For higher N , no stable topological defect exist, but the Nambu-Goldstone (NG) modes are randomly distributed over the horizon scale and their gradient energy is proportional to the SSB scale squared.When those modes come into horizon at a late time, they tend to move in order to reduce the gradient energy [82].The dynamics of NG modes as well as topological defects can affect the CMB temperature anisotropy.Especially for the case with N = 2 and 3, where cosmic strings (for N = 2) and monopoles (for N = 3) form at the phase transition, Ref. [83] discussed a constraint on the SSB scale based on the Planck results such as M < 2.9 × 10 15 GeV for N = 2, M < 6.4 × 10 15 GeV for N = 3, ( We expect that the constraint for the case with large N is weaker by a factor of a few.Therefore we are interested in the case with M ≲ 10 16 GeV.Note that this constraint can be evaded if one introduces a tiny but nonzero explicit O(N ) breaking term in the Lagrangian because all NG modes are aligned by that term.The dynamics of self-ordering scalar fields also emits gravitational waves [84][85][86] (see also Ref. [87] in a different context).It has a flat spectrum in the frequency range of interest.Its amplitude is given by where Σ N is a numerical factor that approaches to unity for a large N limit.For example, Σ N ≃ 4.1 (1.7) for N = 4 (8) [88].This should be added to the gravitational wave signals from second-order effect from large curvature perturbations, which we will discuss shortly.
PBH formation
PBHs form when an overdense region with an O(1) density perturbation enters the horizon.
A sizable amount of PBHs can form when [8] 4 From the top-right panel of Fig. 3, the number of waterfall fields should be O(1-10).
The perturbation of comoving wave length λ enters the horizon when a(t)λ(t) = 1/H(t) ≡ 1/H HC .Assuming that this happens during the radiation dominated epoch, we obtain the corresponding e-folding number N PT at which the relevant perturbation exits the horizon: where we use t end ∼ 1/H inf and t RH ∼ 1/H RH .The PBH mass is given by the total energy enclosed within the Hubble horizon at the horizon crossing: where γ (= O(1)) is a numerical constant [8].We thus obtain where we assume γ = 0.2.From the bottom panel of Fig. 3, we are interested in Π ∼ 10-20.
We should take care of the tail of the spectrum for the curvature perturbations so that the spectrum around the CMB scale is not affected.Since the width of the spectrum ∆N 1 is of the same order with the peak frequency N PT , the latter one cannot be arbitrary large.For the case of instantaneous reheating (i.e., H RH = H inf ), we obtain Π
Summary of parameters of interest
The parameters we are interested in are roughly given by Λ ≃ 5.4 × 10 15 GeV Π 10 ) M ≲ 10 16 GeV, ( This is the number of waterfall fields that is requied to generate PBHs with a mass of interest.
Results
We show an example that provides PBHs with a mass of order 10 20 g, where we choose N = 5, Π 2 = 100, and µ 2 = 10M Pl with ϕ c / √ 2 = M = 10 16 GeV.Interporating the numerical results in Table 1, we obtain N PT ≃ 14 and P GeV.The peak amplitude and frequency are determined by the interpolation for numerical results of stochastic-δN formalism, whereas the shape of the spectrum is given by the analytic result Eq. (3.12).results in Table 1, whereas the shape of the spectrum is given by the analytic result Eq. (3.12) to show a smooth curve.The shaded region above the dashed curve on the top of the figure is excluded by the overproduction of PBHs (see Ref. [9] and references therein).The redshaded region in the left corner of the figure is excluded by the CMB observation [112][113][114][115].The blue and green shaded region in the upper left corner is excluded by constraint for µ and y distortions [116,117]. 5The constraint and future sensitivity curves for gravitational wave experiments are shown as the other dense and light-shaded regions, respectively (see Ref. [119] and references therein).
In the right panel of Fig. 4, we show the PBH mass function generated by the collapse of overdense regions.The peak mass for the PBH is about 5.3 × 10 19 g, which is within the window for the PBH DM.Note that, in order to keep the model parameters simple, we chose the threshold value δ c (∼ 0.3) so that the total PBH energy density is equal to the observed DM density rather than fixing δ c and fine-tuning the parameters.Hence this mass function should be understood as a schematic illustration.In fact, the threshold value with respect to the density contrast is known to be non-universal but depend on the profile of the overdensity and anyway one has to go beyond the Press-Schechter approach for a precise evaluation of the PBH abundance.The non-Gaussianity of the curvature perturbation will also affect the PBH abundance.We leave all these possible corrections for future works (see also the comments in footnote 4).The shaded regions are excluded by overproduction of PBHs (see, e.g., Ref. [9] and references therein for detail).
Fig. 5 shows the spectrum of stochastic gravitational waves predicted in our model.As we discussed around Eq. (4.9), gravitational waves are generated from the dynamics of re-alignment of NG modes when the modes enter into the horizon.In the figure, we plot the signal as the red dashed line for the case with M = 10 16 GeV and N = 5.We assume Σ N = 4. Stochastic gravitational waves are also generated by the second-order effect from large curvature perturbations.The solid blue curve represents the prediction for the case with the above parameter sets.The dense shaded regions are excluded by current experiments and the light-dense regions are future sensitivity curves for planned experiments.
Discussion and conclusions
We have analytically and numerically calculated the spectrum of curvature perturbations in the hybrid inflation model with O(N )-symmetric waterfall fields.The numerical simulation is based on the stochastic-δN algorithm proposed and developed in Refs.[70][71][72][73][74].We formulate the stochastic dynamics under the O(N ) symmetry and demonstrate that the resulting amplitude of curvature perturbations is reduced by a factor of order √ N for a fixed peak frequency.This factor is important to generate PBHs within a desirable mass range from a collapse of the overdense region at the horizon entry.It has been observed that the PBHs with a mass of order 10 20 g can be generated consistently with other cosmological constraints if the number of waterfall fields is about 5 for the case of instantaneous reheating.Such PBHs can explain all energy density of dark matter in the Universe.We have also calculated the spectrum of stochastic gravitational waves generated by the second-order effect from large curvature perturbations.The resulting GW signals can be observed in future GW experiments, including LISA.
The extended sector for the waterfall field is also motivated by avoiding the domain wall problem that arises in a model for a single waterfall field with Z 2 symmetry.The O(N ) symmetry is spontaneously broken by the vacuum expectation value of waterfall fields after the waterfall phase transition.This implies that stochastic NG modes are produced, which results in a continuous re-ordering of NG modes after the phase transition for the case of N ≥ 5 [82].The cosmological effect of those dynamics is qualitatively similar to the case with N = 2, 3, and 4, where topologically non-trivial field configurations form after the spontaneous symmetry breaking.Those dynamics particularly affect the CMB spectrum and give an upper bound on the symmetry-breaking scale.Moreover, it generates GWs with a flat spectrum [84][85][86], which would be observed by future GW experiments as well as the pulsar-timing array experiments.This is an interesting smoking-gun signal of the present model.
One may consider that any global symmetries should not be exact to ensure consistency with quantum gravity.The NG boson then has a small mass from an explicit symmetrybreaking term, and the stochastic modes are ordered by the mass term at a late time.In this case, the constraints from the re-ordering of NG modes may be evaded and the GW spectrum from the dynamics of self-ordering NG modes is suppressed at a small frequency.
Although we have specifically considered the case with global O(N ) symmetry for waterfall fields, the qualitative results, such as the scaling of the amplitude of curvature perturbations as ∝ N −1 , are expected to hold for other symmetries such as SU(N ).Moreover, one can consider the gauged O(N ) or other symmetries instead of the global ones.The stochastic effect on the NG modes as well as the radial mode should be the same in a covariant gauge at least in the regime where the mass of the gauge boson is much smaller than the Hubble parameter during inflation (see, e.g., Refs.[120,121]).Since the vacuum expectation value of the waterfall fields is comparable to the Hubble parameter at the waterfall phase transition, our calculation can be justified at least for the case with a sufficiently small gauge coupling constant.The gauge field becomes heavier after the waterfall phase transition than the Hubble parameter.The stochastic structure of NG modes in the covariant gauge can be understood as the massive gauge fields in a unitary gauge.The heavy gauge fields are then supposed to decay into light fields.Therefore the constraint from CMB anisotropy and prediction of GW signals from NG modes are absent for N ≥ 4 in this case.Still, the GW signals from the second-order effect of large curvature perturbations are the prediction of the model.
)weighted by the backward probability distribution P bw (ϕ | N ), the probability such that the inflatons take certain field values ϕ = {ϕ, ψ r } at the time N e-folds before the end of inflation.Ω denotes the field-space region where the slow-roll inflation can be realised.ϕ 0 is certain initial field values of the inflatons and δN (ϕ 0 → ϕ * ) is a fluctuation in the e-folds N (ϕ 0 → ϕ * ) from ϕ 0 to ϕ * , i.e., δN (ϕ 0 → ϕ * ) ≡ N (ϕ 0 → ϕ * ) + ⟨N (ϕ * )⟩ − ⟨N (ϕ 0 )⟩, where N (ϕ) is a first passage time from ϕ to the end of inflation.Note that this formula requires the leading-order slow-roll approximation, N k ≃ − ln(k/k f ), where k f is the comoving Hubble scale aH at the end of inflation.
by introducing N .
Figure 3 .
Figure 3.The numerical result for the fitting parameters P (peak) R (top) and N PT (bottom) as functions of Π and N .On the top-left panel, the shaded region represents the analytical result Eq. (3.14) multiplied by a factor of 1/3 for Π 2 ∈ (50, 2000).The thin dashed (solid) curve on the bottom panel represents the analytic formula Eq. (3.9) in the leading ∼ Π (next-to-leading ∼ Π 2 ) order approximations with c = 0.
2 ≲ 250 for ϕ c / √ 2 =
M = 10 16 GeV, which leads to N PT ≲ 20 and M PBH ≲ 2 × 10 25 g.For the case of the lowest possible reheating temperature of H RH ∼ 1 MeV 2 /M Pl , we obtain Π 2 ≲ 200 for ϕ c / √ 2 = M = 10 16 GeV, which leads to N PT ≲ 18 and M PBH ≲ 1 × 10 27 g.The upper bounds on Π or N PT do not change much for different values of ϕ c and M (with Π fixed).Because of the relations of Eq. (4.13) and Π ≡ M √ µ 1 ϕ c /M 2 Pl , a larger PBH mass can be generated for a smaller M and ϕ c under the same upper bound on N PT .
Figure 4 .
Figure 4. Spectra of curvature perturbations (left panel) and PBH mass function (right panel) for the case with N = 5, Π 2 = 100, µ 2 = 10M Pl , and ϕ c / √ 2 = M = 10 16GeV.The peak amplitude and frequency are determined by the interpolation for numerical results of stochastic-δN formalism, whereas the shape of the spectrum is given by the analytic result Eq. (3.12).
Figure 5 .
Figure5.Spectra of stochastic gravitational waves generated by the dynamics of NG modes (red dashed line) and the second-order effect from large curvature perturbations (solid blue curve).
Table 1 .
Numerical results of the fitting parameters N PT and P PT given by Eq.(4.13).The first two conditions, Eqs.(4.14) and (4.15) are very rough estimation and are shown for illustrative purposes.More accurate conditions should be read by interpolating the numerical results in Table1.We added a sub-Planckian condition for the critical point: ϕ c ≲ M Pl .The parameters we have are µ 1 , µ 2 , M , ϕ c , Λ, and N .Four of them, µ 1 , µ 2 , Λ, and N , are (approximately) determined by the first four conditions.We have two free parameters, M and ϕ c , which have the upper bounds. | 2023-06-14T01:15:56.711Z | 2023-06-12T00:00:00.000 | {
"year": 2023,
"sha1": "83a531590eaff5721564e3e968956cf8754c78e0",
"oa_license": "CCBY",
"oa_url": "https://iopscience.iop.org/article/10.1088/1475-7516/2023/11/089/pdf",
"oa_status": "HYBRID",
"pdf_src": "ArXiv",
"pdf_hash": "83a531590eaff5721564e3e968956cf8754c78e0",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
267311937 | pes2o/s2orc | v3-fos-license | Quantum algorithms in particle physics
We motivate the use of quantum algorithms in particle physics and provide a brief overview of the most recent applications at high-energy colliders. In particular, we discuss in detail how a quantum approach reduces the complexity of jet clustering algorithms, such as anti-kT , and show how quantum algorithms efficiently identify causal configurations of multiloop Feynman diagrams. We also present a quantum integration algorithm, called QFIAE, which is successfully applied to the evaluation of one-loop Feynman integrals in a quantum simulator or in a real quantum device.
Introduction
Quantum algorithms have emerged recently as a promising avenue to efficiently tackle complex problems in the field of particle physics [1].Events occurring at high-energy colliders, such as the CERN's Large Hadron Collider (LHC), are typically analysed by factoring them according to the characteristic energy scale of their subprocesses.In this respect, recent applications of quantum algorithms in particle physics have considered specific aspects of the collision.For example, quantum algorithms have been explored for track reconstruction [2,3,4] and jet clustering [5,6,7,8], including analysing the formation of jets in a medium [9,10,11].Applications include also the simulation of parton showers [12,13,14], quantum machine learning [15,16,17] and the determination of parton densities [18].On the most theoretical side, quantum algorithms have been used to evaluate helicity amplitudes [19] and the colour algebra [20] of elementary processes, and have been applied in the quest for selecting the causal configurations of multiloop Feynman diagrams [21,22,23,24].This concerns also quantum integrators [25,26,27], including their application to loop Feynman integrals [28].
As quantum computers continue to advance, further applications are expected to appear offering a more complete picture.In general, it is assumed that the main advantage of a quantum approach is the potential speedup gain and the possibility of solving problems whose complexity scales exponentially or superpolynomially.These problems would become intractable at some point with classical computers.While this is true if the quantum principles of superposition and entanglement are exploited in the design of the quantum algorithm, it is also true that the current development of quantum technologies and hardware devices is still very limited due to the low coherence time of qubits and the inevitable quantum noise.
The most interesting aspect of quantum algorithms in particle physics, in my opinion, is related to the original motivation of Richard P. Feynman [29], suggesting that quantum effects, in this case particle collisions, should be better simulated with a quantum system.This presentation is devoted to describe in detail three recent applications of quantum algorithms in high-energy physics.Jet clustering in Section 2, the bootstrapping of the causal representation of multiloop Feynman diagrams in the loop-tree duality in Section 3, and the quantum integration of loop Feynman integrals, in Section 4.
Jet clustering
Clustering is one of the most frequent classical problems in many fields, such as machine learning and computational geometry.In particular, jet reconstruction in particle physics is fundamental in the majority of experimental analyses.The most widely used jet clustering algorithm at the LHC is anti-k T [30], which corresponds to the class of hierarchical or sequential jet recombination algorithms.In this type of clustering algorithm a distance is defined and at each step of the clustering process the pair of particles separated by the smaller distance is recombined into a new pseudoparticle until all pairs of pseudoparticles are separated by a distance greater than a minimum reference distance.
Having to determine at each stage of clustering what is the absolute minimum distance between all pairs of particles is computationally expensive and the computation time increases rapidly with the multiplicity of the event.The computational complexity of the classical anti-k T is O(N 3 ), where N is the number of particles to cluster.However, the FastJet [31] implementation reduces this complexity to O(N log (N )) by identifying each particle's geometrical nearest neighbour and optimizing the clustering by means of Voronoi diagrams.
A quantum version of anti-k T with simulated LHC data has been presented in Refs.[7,8].Without any optimization, this quantum counter- part of anti-k T would require O(N 2 log (N )), and could achieve the same complexity as FastJet by applying just geometric nearest-neighbour optimization.A comparison of the classical and quantum versions of anti-k T is illustrated in Fig. 1, where one can hardly appreciate any difference.There is, however, a fundamental difference between them as a quantum algorithm determines the minimun distance in a probabilistic way.Consequently, while the classical version of anti-k T is deterministic, i.e., given a set of particle data the clustering ordering will always be the same, in the quantum version the clustering ordering may be altered.This probabilistic determination of the mininum distance reduces the complexity of the algorithm.The final jet reconstruction is nevertheless quite similar because of the prevalence of collinear radiation as predicted by QCD.In Ref. [7], we have also presented quantum versions of k T [32,33] and of the general purpose clustering algorithms K-means [34,35] and Affinity Propagation [36].
Causal configurations of multiloop Feynman diagrams
A Feynman propagator describes the propagation of a particle between two interaction points in space-time in either directions.Therefore, we can formally write where the two propagation states are represented by |0⟩ and |1⟩.The total number of states in a Feynman diagram is 2 n , where n is the total number of Feynman propagators.However, not all of these states are physical.Configurations in which a particle returns to the departure point, i.e., those configurations in which a particle describes a closed cycle, require traveling back in time and thus breaking causality.
2023˙MTTD˙rodrigo printed on January 30, 2024 Fig. 2. Quantum circuit used to bootstrap the causal LTD representation of a three-vertex multiloop Feynman diagram using amplitude amplification.
In Feynman's representation, these nonphysical configurations are intrinsically present in the integrand and inevitably lead to nonphysical singularities.An unconventional approach is the loop-tree duality (LTD) [37,38] where a manifestly causal representation exists [39,40,41,42,43,44,45,46] and nonphysical singularities are absent, giving rise to numerically more stable integrads.On the one hand, the LTD representation fixes the directions of propagation allowed by causality.On the other hand, once the propagation directions are fixed, the manifestly causal LTD representation can easily be bootstrapped.
Given a multiloop Feynman diagram, our goal is to select from among all possible configurations obtained by identifying each Feynman propagator with a qubit, those configurations allowed by causality, and thus reconstruct its LTD representation.In this sense, there is an interesting connection with graph theory since the causal configurations correspond to directed acyclic graphs, which are commonly studied in many other fields.
We have presented two different quantum algorithms to identify the causal configurations of a multiloop Feynman diagram.The first one is based on Amplitude Amplification or Grover´s algorithm [21,22].The second one, selects the causal configurations by minimizing a Hamiltonian in a Variational Quantum Eigensolver (VQE) approach [23,24].Different energy levels of the Hamiltonian correspond to configurations with different number of cycles, and by construction, the ground-state is directly associated to the subspace of states related to directed acyclic graphs.
In Fig. 2, we present the quantum circuit used to select the causal configurations of a three-vertex multiloop Feynman diagrams in the Grover's approach.Note that the number of loops is not relevant because the only causal configuration of a group of propagators connecting two interaction vertices is the one in which all propagators are aligned in the same direction.The qubits |q i ⟩, which are initially set in an uniform superposition, represent the three propagators.The ancillary qubits |c i ⟩ and |a i ⟩ are binary and loop clauses, respectively, that probe if a combination of qubits are in the same state.The qubit |out⟩ is the Grover's marker.Measurements are done over classical bits.
In both the VQE and Grover´s based algorithms, the main difficulty lies in the large number of causal configurations to be identified with respect to the total number of possible states, approximately half of all configurations, which requires several adaptations to implement the quantum algorithms efficiently.In both cases, we successfully identified all the causal configurations of Feynman diagrams with up to six interaction vertices, or four loops.In the case of the nonplanar diagram that appears for the first time at four loops, the u-channel [41], it was necessary to resort on the quantum simulator Quantum Testbed (QUTE) provided by Fundación Centro Tecnológico de la Información y la Comunicación (CTIC) that supports simulations with 33 qubits.VQE requires less quantum resources than Grover's algorithm (i.e.fewer qubits and shorter circuits), still a larger number of executions must be performed in order to achieve high success rates.
Loop Feynman integration
The next steep after selecting the causal configurations of a loop Feynman diagram and constructing its LTD representation is to integrate over the loop momenta.To this purpose, we have proposed a new quantum integration algorithm called Quantum Fourier Iterative Amplitude Estimation (QFIAE) [27] and applied this algorithm to benchmark one-loop Feynman integrals [28], including its implementation in a real quantum device.
The main novelty introduced by QFIAE consists of efficiently decomposing the integrand of the target function into its Fourier series using a Quantum Neural Network (QNN).Then, each trigonometric component of the Fourier series is quantumly integrated using Iterative Quantum Amplitude Estimation (IQAE) [47].The Fourier decomposition encodes the target function with a minimum number of quantum arithmetic operations, and takes advantage of the fact that the trigonometric functions are more suitable for integration in a quantum approach.Moreover, the QNN is central to achieve a potential quadratic speedup.
There are several reasons to choose the LTD representation.The absence of nonphysical singularities makes the integrand more stable numerically.The number of integration variables is independent of the number of external particles.The intregration domain is defined by the loop three-momenta, which is Euclidean.Finally, a local renormalisation of ultraviolet (UV) singularities easily renders the integrand UV finite.
For example, the LTD representation of a locally renormalised one-loop bubble Feynman integral, with internal masses m 1 and m 2 , and external where i,0 ± p 0 , and the on-shell energies are given by q , assuming the external momentum has vanishing spatial components, p = (p 0 , 0).The last factor of the integrand that depends on q (+) UV,0 = (ℓ 2 + µ 2 UV − ı0) 1/2 implements the local UV renormalisation, where µ UV is the renormalisation scale.The integration measure reads ℓ ≡ d 3 ℓ/(2π) 3 .
Numerical results for several one-loop Feynman integrals using QFIAE in a quantum simulator have been presented in Ref. [28], in full agreement with known analytic expressions.Moreover, for the one-loop tadpole integral, QFIAE was successfully implemented in real devices, including several strategies to mitigate quantum noise during execution.Deviations of less than 4% were observed, which represents a noteworthy achievement for the current state of quantum computing technology.
Conclusions
Quantum algorithms represent a potential paradigm shift in computational methodologies for particle physics.Although still in its infancy, interest in quantum technologies in the field of high-energy physics is growing and several interesting applications covering different aspects of collider events have already emerged.We have discussed in detail three of these applications, which focus on jet reconstruction and loop Feynman integrals.At the moment, most of these applications can only be tested in quantum simulators, while their implementation in real quantum devices is limited by quantum noise.Besides a potential speedup, quantum algorithms in particle physics represent an appealing quantum way to analyse data and provide theoretical predictions.After all, quantum field theory is quantum by definition.
Fig. 1 .
Fig. 1.Jet reconstruction of an event in the classical (left) and quantum (right) versions, respectively, of the anti-k T jet clustering algorithm with radius R = 1. | 2024-01-30T06:45:04.757Z | 2024-01-29T00:00:00.000 | {
"year": 2024,
"sha1": "e55fcca40dc255c8929643e1da3f04c1bb24a29a",
"oa_license": null,
"oa_url": "https://www.actaphys.uj.edu.pl/S/17/2-X1/pdf",
"oa_status": "GOLD",
"pdf_src": "ArXiv",
"pdf_hash": "e55fcca40dc255c8929643e1da3f04c1bb24a29a",
"s2fieldsofstudy": [
"Physics",
"Computer Science"
],
"extfieldsofstudy": [
"Physics"
]
} |
14245104 | pes2o/s2orc | v3-fos-license | Expression of a2 Vacuolar ATPase in Spermatozoa is Associated with Semen Quality and Chemokine-Cytokine Profiles in Infertile Men
Background A number of laboratory tests have been developed to determine properties of spermatozoa quality but few have been adopted into routine clinical use in place of the WHO semen analysis. We investigated whether Atp6v0a2 (a2 isoform of vacuolar ATPase) is associated with abnormal semen quality and changes in chemokine-cytokine profiles in infertile men. Patients and Methods Semen samples were collected from 35 healthy donors and 35 infertile men at the Andrology laboratory from August 2011 to June 2012. The levels of Atp6v0a2 mRNA and protein, and its localization in spermatozoa were determined. a2NTD (the N-terminal portion of Atp6v0a2) and secreted chemokine-cytokine profiles in seminal fluid were measured. Results Atp6v0a2 protein (P<0.05) and mRNA (P<0.05) in spermatozoa from infertile men were significantly lower than those from fertile men. Fluorescent microscopy revealed that Atp6v0a2 is mainly expressed in the acrosomal region. Infertile men’s seminal fluid had significantly lower G-CSF (P<0.01), GM-CSF (P<0.01), MCP-1 (P<0.05), MIP-1α (P<0.01) and TGF-β1 (P<0.01) levels when compared to the seminal fluid from fertile men. Seminal fluid a2NTD levels were significantly correlated with G-CSF (P<0.01), GM-CSF (P<0.01), MCP-1 (P<0.05), MIP-1α (P<0.01) and TGF-β1 (P<0.01) which are key molecules during the onset of pregnancy. Conclusion These results suggested that a critical level of Atp6v0a2 is required for the fertile spermatozoa and its decreased level in spermatozoa could be used to predict male infertility. This study provides a possibility that Atp6v0a2 could be potentially used as a diagnostic marker for the evaluation of male infertility.
Introduction
The vacuolar (H+)-ATPase (V-ATPase) is a multi-subunit enzyme that couples ATP hydrolysis to the pumping of protons across plasma membranes. It is ubiquitously expressed in eukaryotic cells, where it participates in the acidification of highly differentiated organelles, including the Golgi apparatus, lysosomes, endosomes, and secretory vesicles [1,2,3]. In addition, the V-ATPase is also found at high density in the plasma membrane of specialized epithelial cells that are involved in active proton transport and pH regulation of extracellular compartments. Those plasma membrane V-ATPases have important roles in such processes as renal acidification [3], bone resorption [4] or spermatozoa capacitation [5]. In a murine study, V-ATPases in the apical membrane of epididymal cells, which are also controlled by reversible endocytosis and exocytosis, are required for spermatozoa maturation, viability and pH homeostasis [5]. In addition, the a2 isoform of V-ATPase (Atp6v0a2) is located specifically in the acrosomal membrane of murine spermatozoa to regulate an acidic intra-acrosomal pH, which is necessary for processing protease zymogen, essential for fertilization [6]. In agreement with this previous study, Atp6v0a2 was highly expressed in the acrosomal region of the capacitated murine spermatozoa but not detected in non-capacitated spermatozoa from the caudal epididymis [7]. This study provided a new insight into a possible association of Atp6v0a2 with male infertility.
Although seminal fluid has been conventionally viewed as transport media for spermatozoa traversing the female reproduc-tive tissues, it is now known to have broader biological actions in regulating female fertility. Seminal fluid contains a complex array of cytokines, chemokines, and other bioactive molecules [8,9]. Seminal fluid induces pro-inflammatory cytokines and chemokines such as GM-CSF, IL-6, IL-8, MCP-1, MIP-3a, and IL-1a in the female reproductive tract [10]. Particularly, IL-1 has a potential role in the regulation of blastocyst implantation during early pregnancy [11,12]. IL-1 enhances V-ATPase activity, and increased level of IL-1 may feedback to down regulate the innate immune response, which is essential for implantation [13]. We have shown that Atp6v0a2 can regulate IL-1b as well as IL-1a with little or no subsequent increase in TNF-a secretion [14,15]. In addition, sperm capacitation induces the release of a2NTD, which is N-terminal portion from Atp6v0a2. We have shown that a2NTD induces maternal inflammatory cytokines such as LIF, IL-1b, TNF-a and MIP-1a, and exposure of the uterus to spermatozoa accompanied by seminal fluid enhances pregnancy success rate in the murine model [7].
In approximately 30% of couples, male factor infertility is the only cause of infertility, and in another 20% to 30% of couples, it is a contributing factor for their infertility [16]. Semen analysis is the most commonly used diagnostic tool for male infertility. Recently, the World Health Organization (WHO) has issued standards for abnormal semen analysis in 2010 [17]. However, these standards are not quantitative and do not identify abnormal parameters related to the underlying causes of infertility. To issue these standards, semen obtained only from fertile men were used, and there were no ''threshold values'' for spermatozoa concentration, motility, and morphology to differentiate men as subfertile, of indeterminate fertility, or fertile [18]. Thus, none of these parameters can predict the fertile capacity of spermatozoa or pregnancy outcome with a great deal of confidence. Unfortunately, most clinical laboratories still rely on semen analysis only based on standards to determine plan of care. Indeed, even with techniques such as IVF or IVF with intracytoplasmic spermatozoa injection (ICSI), pregnancy success rates are still remain at 25-30% [19]. This could be partly related to lack of our understanding of the molecular pathology of spermatozoa and semen. Therefore, if a new biomarker could be associated with spermatozoa from infertile men, this would provide a better method to predict fertilization capacity of spermatozoa and pregnancy outcome.
Based on the findings from our lab and others, in this study, we aim to investigate Atp6v0a2 expression and localization in human spermatozoa, and examine the possibility of Atp6v0a2 as a useful biomarker for male factor infertility.
Study Population and Semen Analysis
We evaluated 35 spermatozoa samples from infertile men aged who were diagnosed with male infertility and undergoing assisted reproduction techniques (ART) such as IVF/ICSI at the Andrology Lab, Chicago, IL, USA. Duration of infertility was at least 12 months or more, and azospermic men were excluded from the study. Additionally, semen samples were obtained from 35 fertile donors who had at least one or more live born infant before they had vasectomy for male sterilization. All the collected samples were evaluated at Department of Microbiology and Immunology, the Chicago Medical School at Rosalind Franklin University of Medicine and Science, IL, USA. The study was approved by the IRB of the Rosalind Franklin University of Medicine and Science. Written informed consent was obtained from all study patients and controls prior to enter the study, which was approved by the local Institutional Review Board (IRB). Before the implementation of the ART, semen samples were collected in sterile containers by masturbation after 3-5 days of sexual abstinence. After liquefaction of the semen at room temperature (22uC) for 30 minutes, the samples were assessed for ejaculate volume, spermatozoa concentration, motility and morphology. Morphology of spermatozoa was studied using Kruger's strict criteria [20]. The samples were classified according to 2010 WHO semen analysis guidelines [17]. An aliquot of each semen sample from normal fertile donors and infertile men was collected by centrifugation at 4006g for 5 minutes to obtain spermatozoa-free seminal fluid as previous reported [21] and then frozen at 280uC until analysis was made. Absence of spermatozoa was verified by H&E staining of seminal fluid. The spermatozoa pellet from each infertile man or normal fertile donor was used to determine the levels of Atp6v0a2 by the Real-time PCR and Western blot.
Isolation of Motile and Immotile Spermatozoa from Semen by Density Centrifugation
PureSperm gradients 40% and 80% (Nidacon international, Gothenburg, Sweden) were used for the separation of motile and immotile spermatozoa. After 20 minutes of centrifugation at 3006g, motile spermatozoa were harvested from the lowest layer and immotile spermatozoa from the next layer up. After purification, the motility was $95% in the motile fraction and ,10% in the immotile fraction.
Real-time Quantitative RT-PCR
Total RNA was isolated from the spermatozoa pellet recovered from normal fertile donors and infertile men by using the RNeasy mini kit (Qiagen, Valencia, CA, USA) with on-column DNase digestion using an RNase-free DNase set (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. RNA was stored at 280uC until further use. The cDNA was generated from 1 mg of total cellular RNA by reverse transcription at 50uC for 60 min using random hexamer primer of transcription first strand cDNA synthesis kit (Roche Diagnostics GmBH). Duplex real time-PCR was performed by TaqMan analysis using the Applied Biosystems Step One Real-time PCR system for Atp6v0a2 and a housekeeping gene b-actin. The gene expression of target gene was measured and normalized to the b-actin housekeeping gene. Realtime PCR was performed using universal PCR master mix reagent (Applied Biosystems, Foster city, CA) according to the manufacturer's instructions. The prevalidated Taqman gene expression assays for Atp6v0a2 (Hs00429389_m1) and b-actin (Hs01060665_g1) for TaqMan assays were obtained from Applied Biosystems.
ELISA for a2NTD
Recombinant N-terminal portion from Atp6v0a2 (a2NTD) was expressed and purified from Escherichia coli and subjected to endotoxin removal column chromatography (Proteome Resources, Aurora, CO, USA). The levels of a2NTD were measured in spermatozoa free-seminal fluid recovered from normal fertile donors and infertile men by using sandwich enzyme-linked immunosorbent assay (ELISA). First, chicken anti-a2NTD (Proteome Resources, Aurora, CO, USA) was added to polystyrene trays overnight. After discarding this antibody, wells were washed with PBST three times. 100 ml of spermatozoa free-seminal fluid samples were added to each well with 25 ml of incubation buffer and incubated overnight at 4uC. Each well was washed with wash buffer three times. Then incubated with 50 ml of biotinylated chicken anti-a2NTD at RT for 2 h. Wells were washed with wash a2 Vacuolar ATPase Reflects Semen Quality PLOS ONE | www.plosone.org buffer four times. 100 ml of streptavidin-HRP working solution was added to each well and incubated at room temperature for 30 min. Then the wells were washed with wash buffer four times and incubated with 100 ml of stabilized chromogen solution (R&D system, Minneapolis, MN, USA) at RT for 30 min in the dark. The reaction was stopped with 100 ml of stop solution, and then the absorbance was read at 450 nm. The concentration of a2NTD was measured by plotting linear curve.
Western Blotting
Spermatozoa pellet from normal fertile donors and infertile men were washed two-times with PBS. Equal amounts of protein (50 mg) from total spermatozoa lysates were resolved on 4-20% SDS-PAGE and blotted onto PVDF transfer membranes. The membranes were blocked at room temperature for 1 hour in protein free blocking TBS-T buffer. The blots were then probed with rabbit anti-Atp6v0a2 polyclonal antibody (3 mg/ml) (anti-a2V) (SAB2100187; Sigma-Aldrich, St Louis, MO, USA) or mouse anti-human b-actin (Sigma-Aldrich, St Louis, MO, USA) followed by donkey anti-rabbit IRDye-800CW (1:20,000), and goat anti-mouse IRDye-680CW (LI-COR Biosciences, Lincoln, Nebraska, USA) respectively. Fluorescent blots were imaged on the Odyssey Infrared Imaging System (LI-COR Biosciences, Lincoln, Nebraska, USA).
Immunofluorescence Staining
Motile and immotile spermatozoa were isolated as described above and then fixed in 4% paraformaldehyde for 20 min RT and then air-dried onto slides. Spermatozoa were permeabilized and blocked with 0.1% Triton X-100 and 3% BSA for 10 min at room temperature, then washed with PBS. Slides were then incubated with primary anti-a2V antibody (1:50), diluted in 1% BSA overnight at 4uC. Slides were washed with PBST three times for 10 min each, then incubated with donkey anti-rabbit IgG secondary antibody conjugated with FITC (1:40, 30 min, at 37uC) (sc-2090; Santa Cruz Biotechnology, Santa Cruz, CA, USA), then washed and mounted in Vectashield with DAPI (Vector Laboratories, Burlingame, CA, USA) and stored at 4uC in the dark. The stained slides were viewed under a fluorescent microscope (Nikon Eclipse 80i; Nikon Inc, Tokyo, Japan). The fluorescence of FITC was monitored using a B-2 filter with a 495 nm band pass barrier filter.
Multiplex Assays
All procedures were following manufacturer's instructions (Milliplex MAP kit, Millipore Crop, Billerica, MA, USA) and samples were analyzed on the MAGPIX instrument. The concentration of the analytes was determined in spermatozoa free-seminal fluid recovered from normal fertile donors and infertile men by using the Bio-Plex Manager version 5.0 and MAGPIX xPONENT software, respectively. The assays were run in triplicate to confirm the results. Analytes were normalized to total protein concentration which was estimated with a Bradford assay, a colorimetric protein assay. Twelve analytes were determined: IL-1b, IL-2, IL-4, IL-6, IL-10, IL-12, IFN-c, monocyte chemo-attractant protein 1(MCP-1), macrophage inflammatory protein-1a (MIP-1a), TNF-a, granulocyte colonystimulating factor (G-CSF), and granulocyte-macrophage colonystimulating factor (GM-CSF).
TGF-b ELISA
TGF-b level were analyzed in spermatozoa free-seminal fluid recovered from normal fertile donors and infertile men by using Quantikine TGF-b ELISA kit (R&D systems, Minneapolis, MN) in accordance with the protocol specified by the manufacturer.
Statistical Analysis
A detailed statistical analysis was performed using SPSS version 20. Mann-Whitney U-test was applied to compare Atp6v0a2 mRNA and protein levels, and student t-test for the comparison of spermatozoa parameters, seminal fluid cytokine and chemokine levels of the two groups. Spearman's test was used to determine correlations between a2NTD and cytokine and chemokine levels. A probability of ,0.05 and ,0.01 was considered to be significantly different.
Characteristics of Study Population
The semen profiles of infertile men and normal fertile donors are shown in Table 1. Semen from controls had no abnormal parameters. On the other hand, semen from study group had at least one or more abnormal findings in spermatozoa volume, concentration, motility or morphology. Semen volume (P,0.01), sperm concentration (P,0.05) and sperm motility (P,0.01) of study group were significantly decreased in the infertile men when compared to the normal fertile donors.
Atp6v0a2 mRNA and Protein Expression in Human Spermatozoa
The Atp6v0a2 gene expression was analyzed by real time PCR in the spermatozoa recovered from infertile men and normal fertile donors (n = 18 each). The level of Atp6v0a mRNA was significantly decreased in the spermatozoa recovered from infertile men when compared to the normal fertile donors (P,0.05) (Fig. 1A). Western blotting analysis was confirmed that Atp6v0a2 protein was significantly lower (P,0.05, n = 5) in the spermatozoa recovered from infertile men when compared to the normal fertile donors (Fig. 1B).
Levels of Secreted a2NTD in Seminal Fluid
The concentration of secreted a2NTD was measured in the spermatozoa-free seminal fluid recovered from infertile men and normal fertile donors. Secreted a2NTD protein was significantly higher in normal fertile donors group (57.9611.0 ng/ml, n = 20) when compared to infertile men group (45.7616.3 ng/ml, n = 20; P,0.01) (Fig. 1C). Similarly, western blot analysis of spermatozoa free-seminal fluid with anti-a2NTD shows that a2NTD was detected in ,20 kDa band, and its level was significantly higher in normal fertile donors group compared to infertile men group (P,0.05, n = 6 each) (Fig. 1D).
Seminal Fluid Cytokine and Chemokine Profiles
The concentrations of MCP-1 (P,0.05), GM-CSF (P,0.01), G-CSF (P,0.01), MIP-1a (P,0.01) and TGF-b1 (P,0.01) was significantly decreased in the spermatozoa-free seminal fluid recovered from infertile men group when compared to the normal fertile donors group. In contrast, there were no significant differences in IFN-c and IL-12 between spermatozoa-free seminal fluid from both studied groups (Fig. 2). TNF-a, IL-1b, IL-2, IL-4, IL-6 and IL-10 could not be detected in spermatozoa free-seminal fluid since they were below the sensitivity level of the assay and presumably very low. Previously, we have shown a2NTD induces the secretion of various chemokine and cytokines in the female reproductive tract [7]. To study whether secreted a2NTD of seminal fluid is correlated with secretion of various cytokines and chemokines, linear correlation analysis was performed. Significantly positive correlations were observed between the level of a2NTD and G-CSF (P,0.01), GM-CSF (P,0.01), MCP-1 (P,0.05) MIP-1a (P,0.01) and TGF-b1 (P,0.01) in spermatozoa free-seminal fluid ( Table 2).
The Expression Levels and Immunolocalization of Atp6v0a2 Protein in Immotile and Motile Spermatozoa Collected from Normal Fertile Donor
To determine the association of Atp6v0a2 protein with the sperm motility, Atp6v0a2 protein level was checked in the motile and immotile spermatozoa. Western blot analysis shows that the expression of Atp6v0a2 protein was significantly higher (P,0.01; n = 6 each) in motile spermatozoa compared to immotile spermatozoa from normal fertile donors (Fig. 3A and 3B). These results suggest that the increased Atp6v0a2 protein was associated with motility of human spermatozoa.
Immunohistochemistry analysis was performed to examine the expression and localization of Atp6v0a2 protein in motile and immotile spermatozoa from normal controls. Atp6v0a2 protein was not detected in immotile spermatozoa ( Fig. 3C; panel 1) however, in motile spermatozoa Atp6v0a2 protein was abundantly detected and localized in the acrosomal region ( Fig. 3C; panel 2).
Discussion
Spermatozoa deliver not only the male genome to the ova but several crucial molecules such as mRNA and proteins which are required for fertilization and early embryonic development [22,23,24]. Our previous studies suggested a possibility that Atp6v0a2 in spermatozoa may have an important role in fertilization [7]. In this study, we demonstrate that Atp6v0a2 is highly expressed in normal human spermatozoa but only weakly expressed in spermatozoa from infertile men. Additionally, Atp6v0a2 was highly expressed in motile spermatozoa but not immotile spermatozoa when normal spermatozoa were used. These findings suggest that Atp6v0a2 in spermatozoa plays an important role in sperm motility and is crucial for successful pregnancy.
Recently, it was shown that V-ATPases at the plasma membrane regulate the pH of intracellular organelles which activate spermatozoa motility [25] and is solely responsible for the acrosomal-region-acidification [6]. It is possible that Atp6v0a2 could play a role in spermatogenesis as pH-sensor [26]. The present results clearly indicated that Atp6v0a2 is more abundantly expressed in motile spermatozoa than in viable immotile spermatozoa. Similar to the study of Jaiswal etal., which investigated murine capacitated spermatozoa, Atp6v0a2 was predominantly localized in the acrosomal region of human normal spermatozoa. Atp6v0a2 is likely to be involved as a pH modulator in motile spermatozoa. Furthermore, we reported that Atp6v0a2 plays a role in the successful implantation of a murine embryo. Previously, we found that the expression of Atp6v0a2 is remarkably greater in the egg fertilized by spermatozoa than the unfertilized egg [7]. Therefore, it is speculated that Atp6v0a2 in human spermatozoa may be a vital molecule not only controlling intracellular pH, but also initiating embryogenesis.
There is an early inflammatory stage followed by an antiinflammatory stage and both are required to for a successful pregnancy [27]. Previously, we reported that capacitated spermatozoa initiates inflammation in the maternal side by the release of the immune regulatory molecule a2NTD, which in turn induces the expression of inflammatory cytokines such as LIF, IL-1b, and TNF-a [7]. In other studies, we showed that the key portion of the a2V protein, a2NTD (the N-terminal portion) is released from the activated monocyte and induces the cytokine secretion [15]. In this Table 2). Chemokines are a family of more than thirty chemo-attractant cytokines involved in leukocyte migration, angiogenesis and cell activation. They play important roles in events associated with inammation and immune defense [28,29], and probably have similar roles in the male reproductive tract. MCP-1, which recruits monocytes, memory T-cells, and dendritic cells to sites of tissue injury, infection, and inflammation [30,31], can control Th2 polarization to maintain the maternal immune tolerance toward the allogeneic fetus [32,33]. Previously, we have reported that MCP-1 can induce expression of uterine Atp6v0a2 in animal model [7]. In this study, we demonstrated that MCP-1 in seminal fluid from normal fertile donors was significantly increased compared to seminal fluid from infertile men as well as a2NTD. Thus, it is possible that MCP-1 and a2NTD have a synergetic effect to establish immunological milieu at the implantation site, though further research is necessary to determine their roles in maternal-fetal tolerance.
Over the past 20 years, it has been convincingly demonstrated that the family of colony-stimulating factors (CSF) plays a pivotal role in reproduction by modulating immune milieu of reproductive tract [34,35]. Hence, we speculate that the family of CSFs in seminal fluid may have an important role in successful fertilization. In fact, G-CSF has been reported to impair human trophoblast cell growth and function in vitro by inducing placental granulocy-tosis, and cause abortion in murine model [36]. In addition, Robertson et al., reported that GM-CSF is a key factor regulating fertility and targets both dendritic cells in reproductive tract and developing embryos [35]. In this study, seminal fluid from normal fertile donors was characterized by higher levels of G-CSF and GM-CSF as compared with seminal fluid from infertile men, suggesting a yet undefined role of those. Therefore, these findings raises possibility that a2NTD in seminal fluid is associated with contribution to the establishment of early inflammatory phase which is crucial for the successful pregnancy as well as G-CSF and GM-CSF.
TGF-b is an important signaling cytokine in seminal fluid as it is well-known for its immunedeviating properties [37,38] and plays a key role in preventing a negative immune response to spermatozoa in the female reproductive tract [39,40]. In addition, Loras et al., reported that TGF-b1 concentration was related to infertility [41]. In this study, we show that the level of TGF-b in spermatozoa freeseminal fluid from infertile men was significantly lower than those from normal fertile donors. There was a strong positive correlation coefficient between a2NTD and TGF-b1 (r = 0.602). It may be speculated that seminal fluid TGF-b1 is essential for initiation of immune tolerance to seminal antigens and preventing aberrant immunity to spermatozoa. These proteins appear to play a vital role in priming of a specific female immune response which is necessary for successful embryo implantation.
In this study, we analyzed whether Atp6v0a2 is associated with spermatozoa status diagnosed with the WHO semen analysis and showed that lower Atp6v0a2 expression in spermatozoa from infertile men was consistent with ''abnormal'' based on WHO semen analysis (i.e., parameters were lower than the standards indicated for semen volume, number and motility and morphology) [42]. With regard to spermatozoa motility, we found that Atp6v0a2 in motile spermatozoa was expressed at higher levels than in immotile spermatozoa. This result strongly suggested that Atp6v0a2 is potentially useful for the evaluation of spermatozoa motility whether semen is diagnosed normal or abnormal based on WHO semen analysis standards.
In summary, Atp6v0a2 protein and mRNA in spermatozoa from infertile men was significantly lower compared to that of normal spermatozoa. Fluorescent microscopy revealed that higher expression of Atp6v0a2 mRNA and protein in the acrosomal region of motile spermatozoa than immotile spermatozoa. Secreted a2NTD in seminal fluid from normal fertile donors was significantly higher than infertile men. Furthermore, a2NTD expression correlated with levels of cytokines and chemokines which are key molecules during the onset of pregnancy. In conclusion, a critical level of Atp6v0a2 may be necessary for the normal function of spermatozoa and decreased levels of Atp6v0a2 may predict male infertility. | 2016-05-12T22:15:10.714Z | 2013-07-30T00:00:00.000 | {
"year": 2013,
"sha1": "1dac764e3900cea861fa355ed156681a227536a2",
"oa_license": "CCBY",
"oa_url": "https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0070470&type=printable",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "1dac764e3900cea861fa355ed156681a227536a2",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Biology",
"Medicine"
]
} |
151748749 | pes2o/s2orc | v3-fos-license | Using Aesthetic Response-A Poetic Inquiry to Expand Knowing , Part II : Theoretical Perspectives on Arts-based Research
Apart from being inspired from both an interpretive and a constructivist tradition, research methods based in aesthetics can thrive from a clear rationale concerning its perceptual building-blocks in both the intersubjective and intra-psychological domains. This article aims to address the complexity of sharing implicit processes and tacit knowledge in the arts-based inquiry. Layers of this inquiry is reflected along with theoretical perspectives of such undertakings. The article also offers a theoretical rationale for why to add and acknowledge important perceptual and affective building blocks in arts-based research (ABR). Through theories from expressive arts therapy, heuristic inquiry, attachment theory and contemporary affective neuroscience some thoughts on the embodied felt sense as a perceptual hub is shared. Based in contemporary attachment theory and psychotherapy research, a rationale is given for why engaging in ABR can offer clinicians and researchers a deepened understanding of the studied phenomena. Our undertakings are presented in part 1 of these two articles. From this embodied perspective, the described arts-based inquiry can be considered as a privileged way to nuance and enlarge understanding in both the intersubjective and intra-psychological domain, which could be particularly helpful to ABR researchers who are informed by a psychodynamic perspective.
The value of the aesthetic response, as a source of knowledge, cannot be underestimated.Ammaniti and Gallese (2014) noted, "Before and below mind-reading is intercorporeality as the main source of knowledge we directly gather about others" (p.16).They also see this embodiment as a prerequisite for sharing and understanding symbolic representations.Our activated mirror neurons impact our ability to empathize with unconscious states of others.Intercorporeity, rather than the symbolic representation, therefore is placed at the core of emotional resonance.Intercorporeity means the experience of being embodied and it is never a private affair, but is always mediated by our continual interactions with other human and nonhuman bodies (Weiss, 1999, p. 5).Csordas (2008) and Allegranti (2013) pointed out that our existence in relation to others -our inter-subjectivity -is something tangible and bodily.Intersubjectivity is not only declarative or explicit but an analogical map of the other within us, and needs to be read on implicit levels from inside ourselves (Lyons-Ruth et al., 1998).Gerge (2016) exemplified this in her analyses of clinicians' drawings of a worrying or reassuring clinical meeting.The arts-based inquiry offers each and every one of us possibilities to read, by experiencing those maps and work with the embodied aesthetic response (Panhofer et al., 2011), thus allowing a deepened sharing over time and space.
The Aesthetic R The Aesthetic Response esponse
The function of an aesthetic response (Robbins, 1971(Robbins, , 1987) ) is, according to Conrad (2010), to stabilize the brain's semi-randomly generated neural circuits.The neuronal circuits are selectively steadied if they succeed in making sense out of raw sensory input.Under the brain's developmental process, the aesthetic response of humans serves the function of calming the circuits or networks that successfully mediate perception and interpretation.These will become more agile, in the process of making something special.According to Conrad (2010), this is triggered by structures in art and nature, which provoke the sense-making, a deliberate human action.
When discussing the theoretical underpinnings of how self-identity is revised through portraiture in palliative care, Carr (2014) stated: The relationship between the 'container' and the 'contained' is an intersubjective one, and this combined experience becomes in the portrait a concrete, sensory and symbolic form (Langer, 1953), offering a unique way to hold, contain and safeguard this attuned experience.(pp.57-58) Carr's (2014) clinical standpoint is based in the tradition focused on the 'art' within art therapy (Allen, 1992(Allen, , 2001;;Malchiodi, 1999;McNiff, 1986), thus a therapeutic endeavour closely connected with ABR.
The artistic expressions constitute in therapy an expanded field, in which new structuring of the inner world is possible.This leads to, that conflict-loaded material, i.e. overwhelming affect and negatively charged images, can be externalized, perceived and processed.(Gerge, 2010, p. 69) This endeavour can be as valuable in research as in clinical work.The aesthetic sensibility needed in arts-based inquiry (Viega, 2016a) can be considered especially wellsuited for professionals trained in the creative arts therapies, for examples see (Gerge, 2016;Gerge, Wärja, & Pedersen, 2017).
In Phenomenology of Perception Merleau-Ponty (1963) described the body's encounter with the phenomenon, as it is through the body we understand the world (Bullington, 1999).Damasio (1995Damasio ( , 1999) ) stated that a thought always is a description of an embodied state.By using our theory of mind (Frith & Frith, 1999) and our embodied empathy (Rothschild, 2000) we can imagine ourselves as the other and authentically meet the other; an important prerequisite for change-inducing meetings in psychotherapy (Gerge, 2011(Gerge, , 2015) ) and research (Gerge, 2016).This capacity rests on an embodied felt-sense experience (Gendlin 1964(Gendlin , 1978)), which is the core of the aesthetic response, empathy, image formation, and at its end-point, an acknowledged perception.We think such a standpoint is valuable also in research.
The Experienc
The Experience of the F e of the Felt Sense elt Sense Gendlin (1978) developed Merleau-Ponty's (1963, 1973) ideas to show how interactions are more fundamental than perception.The bodily experience was essential to the work of Merleau-Ponty, who referred to the internal total awareness as body schema (1963, p. 113-114); see definition at the end of the article.This embodied felt sense is supposed to ground our conscious awareness, in line with Damasio's (1995Damasio's ( , 1999) ) notion -a thought is an ongoing description of a state in the body.In everyday language, we lack words to name these crucial processes, however in the therapeutic practice of Focusing (Gendlin, 1978), it is described as an embodied tacit knowledge.This is a special kind of internal bodily awareness, a body-sense focused on meaning making, which offers a potential research tool for collecting and describing data from non-verbal sources.What is described can further on be analysed.
Att Attachment Theory and C achment Theory and Cont ontempor emporary Neur ary Neuroscienc oscience in ABR e in ABR The capacity of being together is mediated over our mirror-neurons (Rizzolatti, Fadiga, Fogassi, & Gallese, 1999).Our brains are in Gallese's words "we-centric" (Gallese, 2009).Our capacity to resonate -a necessary prerequisite for reasoning -can be used as a vehicle for deepening our understanding of the topics of interest we want to study.These processes will always be partly preverbal.
Recently, McCaffrey & Edwards (2015) discussed ABR as non-narrative, "as many processes and products in arts creation transcend literal meanings" (p.516).We fully agree with their reflection, though we think that the concept non-narrative has to be problematized, as narrate, apart from referring to speech and writing can be to tell by means of images.Our conscious awareness resides in larger orders of affective and cognitive narratives, which are based in experiential implicit bodily processes referred to as the "neuronal architecture which supports consciousness" by Damasio (1999, p. 15).In 1992, Daniel Stern introduced the term "proto-narrative envelope" ( pp. 291-295).This envelope he stated, contains experiences organized within the structure of a narrative and are built from the child's experiences of happened or imagined events, as these are unfolding.However, it refers to a story without words or symbols, a plot visible only through the perceptual, affective, and motoric strategies to which it gives rise.Stern (1995) stressed how early experiences of mother-child interactions "have a beginning, a middle, and an end and a line of dramatic tension; they are tiny narratives … 'proto-narrative envelopes'" (Stern, 2002, p. 6).In 2004, Stern changed the concept proto-narrative envelope to "lived story" (Stern, 2004, p. xiii).This concept is related to an emotional narrative that is felt rather than told as a cognitively constructed story for both infants and adults (Stern, 2004) and is a prerequisite for a moment of meeting, (see further definitions at the end of the article).The concept lived story and our contemporary neuro-affective understanding of human recognition processes (Gallese, 2009;Hass-Cohen & Findlay, 2015), make statements such as the arts-based inquiry is nonverbal or non-narrative highly questionable.On the other hand, they point to possible ways in which music, experienced on preverbal levels can give us images and evoke stories (Stern, 2000(Stern, , 2004(Stern, , 2010)).From this perspective arts/artifacts can be seen as frozen moments in time, even if the person who has produced the artifact is not present.
As shown in the method description of part 1, (Gerge et al., 2017), we want to highlight the short time frame of a half of a minute, to be used when becoming touched and initiating an aesthetic response (even though conducting an artwork or performance might take somewhat or much longer time).An extended moment of meeting will not be longer than 30 seconds according to Stern (2004).We propose that this limited time frame is an essential building block of the felt-sense experience.It is a prerequisite for the aesthetic response, even at times when the final arts-based inquiry involves composing a musical symphony or creating a mural painting (taking months to prepare).
We hypothesize that the similar processes as in attachment modulation are active also when we relate to an artifact, for examples see part 1 (Gerge et al., 2017), where poetic responses were achieved.In poetic transcription (Faulkner, 2009;Furman, 2006;Leavy, 2015) the inquiry is derived from a grounded theory perspective, where selected words and phrases from the informants become building blocks of the poems conducted.In poetic transcription, according to Faulkner (2009), notes of written state-ments of the informants are directly cited, and then put into a structure chosen by the researcher.In our work, poetic statements were created as aesthetic responses in line with the primary signification of the word poetry (poiesis) meaning "to create".When working with the aesthetic response in an arts-based inquiry, we consider it possible that the implicit information, the proto-narrative in the art piece, can be listened to in a similar way as when one is present with another human being.In this way, ABR has the potential to resemble the regulatory processes of the early dyad where the concept moment of meeting (MoM) is crucial (Sander, 2002;Stern, 2004).These time-framed important building blocks of early development and attachment continues to be active in human perception all through our lives.As adults we also form lived stories in our on-going endeavour to understand the world and ourselves.Organisation of meaning is implicit as Lyons-Ruth (1999) stated.She highlighted that we do "not require reflective thought or verbalization to be known" (p.578).
The process of being known, or feeling felt, is well described in attachment research (Schore, 1994(Schore, , 2003a(Schore, , 2003b(Schore, , 2003c(Schore, , 2003d;;Sander, 2002;Stern, 2004).We consider these communicative aspects of attachment research also relevant in the experiential being with in relating to an art piece in art/creative arts therapy or ABR.The processes that build the experience of being known are scientifically validated in recent neuroaffective research on mirroring (Ammaniti & Gallese, 2014;Gallese, 2009) and intersubjectivity (Siegel, 2010).Sander (2002) highlighted the communicative aspects of affects and affect regulation.In line with Schore (1994), he described affects as observable states and thus highly communicative.Such perspective is emphasized in contemporary social psychological research (Wetherell, 2015).Daniel Stern (1985) developed Silvan T. Tomkins' affect theory (1962Tomkins' affect theory ( , 1963) ) and added the concept vitality affects, where the musicality and dynamic quality of being together (and being with emotions and affects) and relate on preverbal levels as building blocks of perception were highlighted (Ammaniti & Ferrari, 2013;Stern, 1985Stern, , 1994Stern, , 2000)).Building on nonlinear systems theory, Sander (2002) and Stern (1985Stern ( , 2010) ) developed our understanding on how human beings can fit together.By accurate preparatory attunement (recognition processes) special moments of shared experience (MoM) can generate strong feelings of connection between people.These processes are well described in psychotherapy research (Hughes, 2007), including music therapy (Blom, 2014;Coomans, 2016) and art therapy (Hass-Cohen & Findlay, 2015).Also, when moving into a relationship with an artifact, we can open up to a transformative meeting.By using the timeframe of an extended MoM in profound meetings with other human beings and/or their artifacts, we consider that a certain rigor can be brought into the multi-layered themes of implicit processing, which will be activated every time our attachment system is activated (Trevarthen & Aiken, 2001).This processing will also be activated every time we allow ourselves to be touched, which is an essential part of the arts-based inquiry.
Research on attachment points to the role of shared interconnectivity described by (Gallese & Ferri, 2014;Bowlby, 1968;Gaensbauer, 2016;Schore, 2003dSchore, , 2009;;Siegel, 1999) as a prerequisite for development, empathy, and mentalization (Fonagy & Luyten, 2015).Kenny (1989Kenny ( , 2006) described an intuitive level of togetherness that encompasses therapist and client in clinical work.Kenny's (2006) meta-theory on what might happen in the encounter between client and therapist is highly relevant for ABR.She highlighted the need for an existential phenomenology; immersion, tacit knowing, and intentional verification of the felt sense experience (Douglass & Moustakas, 1985) to make sense of what one experiences.This neatly fits into contemporary attachment research (Gallese 2009;Narvaez et al., 2013).Kenny (2006Kenny ( , 2015) ) labelled the process of making sense as an act of creativity, empathy, and letting oneself be touched by the other/the other's expression, and what might be called existential beauty.She also stated "Perhaps my words will never be able to describe the beauty of these moment" (Kenny, 2006, p. 192), thus addressing the implicit nature of togetherness.This is in line with Stern's reflection "one can not get to the lived experience and stay there while talking about it" (2004, p.xiii).But of course we ought to try, and we are genetically hard-wired to make sense of, and share implicit information, both for our survival as individuals and as species.
Art a Art as "Making Specia s "Making Special l" and Implicit P " and Implicit Pr roc oces essing sing Dissanayake (1988Dissanayake ( , 2000Dissanayake ( , 2003) ) argued that art serves as the key social role of "making special" when experiencing phenomena and thus can help us both invest and research certain phenomena with special significance.In Jungian psychology this process is described as crystallization (Henderson, 1964).Knill, Barba, and Fuchs (1995) spoke of crystallization as "the basic human need to crystallize psychic material; that is, to move toward optimal clarity and precision of feeling and thought" (p.30).This is a key function of art-making and art experiencing and can be considered an important principle when making sense of information in the aesthetic domain.In this way the artistic process has great similarities with the pragmatic and eclectic approach to qualitative research (Gordon, 1999;Leavy, 2015).
According to Faulkner (2005), cited in Leavy (2015), The poetic criteria related to qualitative and artistic criteria are; "artistic concentration, embodied experience, discovery/surprise, conditional, narrative truth, and transformation" (p.97).In line with this, Leavy (2015, citing Richardson, 1997and Pelias, 2011) put forward what we might speak of crystallization instead of triangulation.She continued, "poetry is both a style of representation as well as a vehicle through which the research community can engage in larger questions about the nature of social research, truth and knowledge" (p.97).In this way the arts-based research or inquiry can help us grasp the essence of the studied phenomena and experiences.It can add resonance to reason.The ABR potential experience of resonance and touching "essence" supposedly adds to the experience of deepened sharing.This encompasses also to lower the guard, to make oneself vulnerable, and open to a potential experience of awe (Gerge, 2016).
C Concluding W oncluding Wor ords ds
Our aim in this paper was to describe a vital procedure including a timeframe in line with contemporary theory of the phenomenon of the moment of meeting (MoM) (Sander, 2002;Stern, 2004).The use of the method and its six steps; relate, resonate, respond, reflect, react, and results was exemplified in "Using Aesthetic Response, a Poetic Inquiry to Expand Knowing.Part I: The Rx6-Method" (Gerge et al., 2017).The Rx6 method is grounded on a theoretical framework of inter-subjectivity and implicit processing, in line with relational psychodynamics.Engaging in ABR can offer clinicians and researchers of all orientations a deepened, expanded, and embodied understanding of the studied phenomena.This can be considered a prerequisite for heightened empathy.We consider the ABR-approach especially important when working with the indwelling procedures of implicit processing, eg., when using tacit knowledge (Polanyi, 1958), in research and psychotherapy.
We hope that the MoM can be considered a perceptual tool to make sense of what we encounter in the experience of being with another human being, and/or with her/ his artifacts, or -when being in an I-Thou relation (Buber, 1962(Buber, /1993)).The MoM theory helped us conceptualize the perceptual building blocks of coming close to another human being.This understanding is necessary in attachment work (Schore, 1994;2003a, 2003b, 2003c;Trevarthen, 1993), which is well documented and implemented in psychotherapy with psychodynamic orientation (Schore, 2003d;Terr, 2003), music therapy (Blom, 2014;Coomans, 2016), and art therapy (Carr, 2014).Though we propose that this approach can be of value for clinicians and researchers of other orientations as well.
The theory of MoM, (Sander, 2002;Stern, 2004) and consciously working with intermodal transfer can build loops of ABR over an extended time (days, weeks and months) (Gerge, 2016).We regard the MoM as a grounding basic beat -a pulse of perception -a building block of the flow of human experience in processes where the medium is the message (Lyons-Ruth, 1999); as in experiential psychotherapy (Gerge, 2015) or ABR (Gerge, 2016;Gerge et al., 2017).Although Ledger and McCaffrey (2015) argued that it is too early to create definitions as these might limit the possibilities for innovations that ABR might bring to music therapy (p.453), we consider the structure, including the time frame suggested here, as an important reflective tool, both as a method, and for methodologies "steeped in aesthetics" (Viega, 2016b, p. 5).In putting focus on perception, as described in attachment research (Sander, 2002, Stern, 2004;Narvaez et al., 2013), contemporary affective neuroscience (Ammaniti & Gallese, 2014) and clinical work, (Hass-Cohen & Carr, 2008;Hughes, 2007), we can begin to structure ABR.From a solid hub of theoretical frameworks (with spokes in all directions), various modalities and art disciplines can radiate.From such a hub, seen as an "orienting lens" for understanding a phenomenon (Creswell, 2009;Shannon-Baker, 2015), a variety of sound research and innovative applications might rise.In this way ancient roots of creativity can meet and be nourished from heuristic research perspectives (Moustakas, 1990(Moustakas, , 1994)), expressive arts therapy (Estrella, 2005;Levine & Levine, 1998, 2005;Richardson, 2016) and attachment theory, including the theories of the lived story (Stern, 2004), and the theory of MoM (Sander, 2002;Stern, 2004).This will further enrich qualitative research, the discursive practice (Wetherell, 2015), and psychotherapy research including the arts-based therapies.
To conclude, first we propose that clinicians, conducting clinical research, scholars in psychology and health care research, including rehabilitation medicine, consider using ABR as a means to expand their understanding.Second, these processes will deepen their compassion for and understanding of their research participants and their processes of change.With Chenail (2008) we answered the question concerning ABR "But is it research?"with a firm YES.
Gl Glos oss sary of T ary of Terms erms
Body schema: Merleau-Ponty (1963, p. 113-114) referred to the internal total awareness as body schema.This embodied awareness is conceptualised as an on going bodily interaction that opens us to a sense of the world beyond what we conventionally call perception.The body-sense can encompass perceptions and emotions but also memories of past situations and options of what to do next.
Present moment or now moment: Daniel Stern (2004) defined the "present moment" as a lived story with a beginning and an end, intentional characters, together with a "temporal contour along which the experience forms during its unfolding" (p.219).The present moment is defined as lived through as it unfolds and is not distanced by language or abstract explanation from those experiencing it.The now moment consists of an emerging interpersonal process that is unpredictable-hence "sloppy," consisting of the present moment in life-as-lived, which Stern generalizes from attachment research to psychotherapy.The concept is elaborated from The Interpersonal World of the Infant (Stern, 1985).
Moment of meeting:
The now moment is often followed by a moment of meeting (Stern, 2004).When this occurs, new ways of being-with-the-other unfolds through the offered time-framed togetherness, where we can "read in the behavior of the other a reflection of their own experience" (p. 220-221).This leads to the possibility of events becoming intersubjectively conscious, and further on verbalized and narrated experiences. | 2018-12-11T03:53:41.707Z | 2017-02-13T00:00:00.000 | {
"year": 2017,
"sha1": "04fad313dbd22e9ac820452b2c2fd804c9caa8c5",
"oa_license": "CCBY",
"oa_url": "https://voices.no/index.php/voices/article/download/2353/2161",
"oa_status": "GOLD",
"pdf_src": "ScienceParseMerged",
"pdf_hash": "04fad313dbd22e9ac820452b2c2fd804c9caa8c5",
"s2fieldsofstudy": [
"Psychology"
],
"extfieldsofstudy": [
"Psychology"
]
} |
16956582 | pes2o/s2orc | v3-fos-license | Intensified wind pollination mediated by pollen dimorphism after range expansion in an ambophilous biennial Aconitum gymnandrum
Abstract Pollination systems and associated floral traits generally differ between core and marginal populations of a species. However, such differences are rarely examined in plants with a mixed wind‐ and bumblebee‐pollination system, and the role of wind pollination during range expansion in ambophilous plants remains unclear. We compared floral traits and the contributions of bumblebee and wind pollination in refugium and marginal populations of the ambophilous plant Aconitum gymnandrum. We found that most floral traits differed between the two populations, and those traits associated with the shift to wind pollination were pronounced in the marginal population. Bumblebee visitation rates varied significantly, but were generally low in the marginal population. Wind pollination occurred in both populations, and the efficiency was lower than that of bumblebee pollination. Two types of pollen grains, namely round and fusiform pollen, were transported to a stigma by bumblebees and wind, but fusiform pollen contributed to wind pollination to a larger degree, especially in the marginal population. Our results suggest that wind pollination was enhanced by pollen dimorphism in the marginal population of A. gymnandrum, and wind pollination may provide reproductive assurance when bumblebee activity is unpredictable during range expansion, indicating that ambophily is stable in this species and shift in pollination system could be common when plants colonize new habitats.
| INTRODUCTION
The geographic distribution of species is often the result of a complex history of range shifts, for example, the contraction or expansion of species distribution due to the glacial-interglacial cycles (Abbott & Brochmann, 2003). The current distribution of species at mid-to high latitudes could be attributed to range expansion from a "refugium" population at lower latitudes following global warming after the Pleistocene (Hewitt, 2000). Therefore, many extant plant species may have experienced range expansions and contractions in response to historical warming and cooling events (Svenning, Eiserhardt, Normand, Ordonez, & Sandel, 2015). After range expansion, marginal populations may be of smaller size and lower density than core populations owing to founder effects (Hardie & Hutchings, 2010) and the resultant vulnerability to environmental fluctuations (Frankham, 2005), and pollinator availability may be less reliable in marginal populations than that in core populations. Accordingly, under selective pressure from pollinator scarcity and unpredictability, the pollination system of marginal populations may differ substantially from that of core populations (Busch, 2005;Geber & Moeler, 2006;Herlihy & Eckert, 2005).
Uniparental reproduction, consisting of asexuality and selfing, could overcome low plant density and pollinator scarcity (Eckert, Samis, & Dart, 2006). Such a strategy may be an important evolutionary trend in marginal populations because it provides reproductive assurance during colonization and establishment at low density when outcrossing is limited (Barrett, 2002;Geber & Moeler, 2006;Schoen, Morgan, & Bataillon, 1996). However, reduced population genetic diversity owing to founder effects in marginal populations (Eckert, Samis, & Lougheed, 2008) would be further intensified after continuous uniparental reproduction, exposing populations to in the risk of extinction (Frankham, 2005). Therefore, selection for outcrossing would be favored in marginal populations, and wind pollination provides a solution to the conflict between small population size and maintenance of outcrossing in marginal populations. Wind pollination is considered to play an important role in conferring reproduction assurance under pollinator scarcity (Culley, Weller, & Sakai, 2002;Friedman & Barrett, 2009). For example, in the self-incompatible and insect-pollinated Linanthus parviflorus, pollination mediated by airborne pollen accounts for more than 60% of seed production from open-pollinated flowers, providing one of the few examples supporting the hypothesis of reproductive assurance by wind pollination when pollinators are unreliable (Goodwillie, 1999).
Given that mechanisms of pollen dispersal differ between windand animal-pollinated plants, the floral traits are also dissimilar.
Generally, few ovules, unisexual flowers, reduced volatile emissions, and a high degree of sexual dimorphism in color and scent are often associated with the transition from animal pollination to wind pollination (Friedman & Barrett, 2008Welsford, Hobbhahn, Midgley, & Johnson, 2016). Furthermore, other floral traits, such as reduced perianth and nonadhesive pollen, both of which could facilitate the pollen dispersal by wind, might also be expected in wind-pollinated flowers. However, it remains unclear how these traits alter between core and marginal populations of plant species with a mixed animal-and wind-pollination system (ambophily), especially for those species that have experienced a recent range expansion. In addition, wind-and animal-mediated pollen dispersal may result in selection for different pollen size in plants with different pollination systems (Welsford et al., 2016) Thus, the contemporary flora presents an opportunity to examine evolutionary shifts in pollination system in plants that have experienced range expansion on the QTP. Aconitum gymnandrum (Ranunculaceae) is a biennial herb native to the QTP with a mixed wind-and bumblebee-pollination system (Duan, Zhang, He, & Liu, 2009;Zhang, Duan, & Liu, 2006). Phylogeographic analysis of the genetic structure of A. gymnandrum across its distribution range resolved two distinct lineages, generally congruent with eastern and western groups of populations, but evidence for range expansion from a refugium population after the Last Glacial Maximum was only identified in the eastern lineage based on haplotype diversity (Wang et al., 2009). Therefore, in the present research, we examined
| Study species and populations
Aconitum gymnandrum is the only species classified in Aconitum subg. Gymnaconitum (Ranunculaceae) (Li, 1988). The sepals and petals of A. gymnandrum are blue-purple and showy, but the flowers exhibit several traits, that is, degenerate sepals and exposed anthers and stigmas (Figure 1a), which are unique in the genus Aconitum (Wang et al., 2001). Recently, the subgenus Gymnaconitum was raised to generic status (Wang, Liu, Yu, Gao, & Chen, 2013).
Aconitum gymnandrum is widely but discontinuously distributed on the inner QTP and generally grows in degraded alpine grasslands and abandoned agricultural lands at altitudes ranging from 2230 to 4300 m a.s.l. (Wang et al., 2009). An individual plant consists of several branches, including one main terminal branch and multiple lateral branches. Each branch can be considered as an independent raceme. Flowers of A. gymnandrum are characterized by protandry and herkogamy (Zhang et al., 2006) and are pollinated by bumblebees and wind, but contribution of wind pollination to seed set is small in comparison with bumblebee pollination because of the high visitation rate and pollination efficiency of bumblebees (Duan et al., 2009).
On the basis of the genetic structure of A. gymnandrum (Wang et al., 2009), we selected a refugium population in Tongren County In addition, using SEM micrographs with the scale bar as a reference, we measured the diameter (D) of round pollen and the long axis (a) and short axis (b) of the fusiform pollen from mixed anthers sampled from different flowers and fixed in FAA as described above. We calculated the size of round and fusiform pollen grains based on the surface area formula for a circle (1/4 × 3.14 × D 2 ) and ellipse (1/4 × 3.14 × a × b), respectively.
| Bumblebee visitation and pollination efficiency
Pollinator observations and pollination efficiency were investigated from 2012 to 2014 in both populations on sunny days without strong wind. Bumblebees are the main animal pollinators of A. gymnandrum (Duan et al., 2009;Zhang et al., 2006). We treated all bumblebees as a single pollinator functional group without considering differences in bumblebee species. At the time of peak blooming in the populations, we labeled two to three inflorescences randomly and recorded the flowering stage and number of flowers on each inflorescence. When the stigma was receptive, we removed the bags and observed these flowers on sunny days. After each flower was visited once by a bumblebee, we re-bagged the flower and fixed the stigma in FAA after 3 h. If no visitation was observed, the flower was discarded. In the laboratory, the number and type of pollen grains deposited on the stigma were determined under a stereomicroscope after the stigma was washed, rehydrated, softened in 8 N NaOH for 8 h, and stained with 1% aniline blue solution.
| Airborne pollen and pollination efficiency
To examine the frequency of each pollen type in airborne pollen of the refugium and marginal populations, each slide covered with Vaseline on one side was affixed to a pole and placed vertically in both populations from 10:00 to 16:00 on sunny days from 2012 to 2014. All slides were put randomly in both populations with direction of the covered slide facing the prevailing wind, and the height of each slide was equal to the middle of the main raceme in each population. After collection, the slides were examined under a stereomicroscope, and the type and number of pollen grains affixed to the slide were determined.
To determine the potential and actual contributions to seed production by airborne pollen, 1,465 flowers from the third node of the inflorescence on different plants in the male phase were emasculated F I G U R E 1 (a) Flowers of Aconitum gymnandrum, with a lower sepal and lateral sepal indicated by red and yellow arrows, respectively. (b) Two types of pollen grains of Aconitum gymnandrum, defined as round and fusiform pollen and covered with fine nylon nets of 1-mm mesh in each study year.
When the stigma had turned brown, which is an indicator of loss of stigma receptivity (Zhang et al., 2006), we collected the stigma from 771 flowers and fixed each in FAA. The type and number of pollen grains on each stigma were determined in the laboratory using the above-mentioned methods. The remaining 694 flowers were collected before fruit dehiscence, and the number of seeds was counted.
| Statistical analysis
Independent t-tests were used to compare the floral traits between the refugium and marginal populations. A general linear model was employed with the aim of identifying differentiation in the pollination system of A. gymnandrum after range expansion from a glacial refugium population.
| Floral traits
Plants in the refugium population were taller and produced a greater number of flowers than plants in the marginal population, but the flower density per inflorescence was higher in the marginal population than in the refugium population (Table 1). No significant difference was observed in the distance between the two lower sepals, but the minimum width of the lateral sepals was markedly smaller in the marginal population than that in the refugium population (Table 1). The numbers of stigmas and ovules per flower were higher in the marginal population than those in the refugium population, but the opposite was true for the number of anthers and total number of pollen grains (Table 1). Thus, the P/O ratio was higher in the refugium population than in the marginal population. Importantly, pollen dimorphism was observed in both populations (Figure 1b). The number of round pollen grains was significantly higher in the refugium population than in the marginal population, but no significant difference was observed in the number and the percentage of fusiform pollen grains between the two populations. The size of round and fusiform pollen grains was 340.44 ± 15.71 μm 2 (Mean ± SE, N = 30) and 164.51 ± 3.70 μm 2 (N = 30), respectively. The difference in size between the two types of pollen grains was significant (t = 14.23, p < .001).
| Bumblebee visitation and pollination efficiency
Visitation rates of bumblebees were affected significantly by year, population, and year × population interaction (Figure 2; Table 2).
Specifically, visitation rates of bumblebees were higher in the refugium population than those in the marginal population in both years and were lower in 2012 than those in 2014 in both populations.
F I G U R E 2 Visitation rates of bumblebees (mean ± SE) in the refugium (filled dots) and marginal (open dots) populations in 2012 and 2014
T A B L E 1 Floral traits (mean ± SE, with sample size in parentheses following the trait name) of the refugium and marginal population of Aconitum gymnandrum on the Qinghai-Tibet Plateau After a single visit by a bumblebee, the number of pollen grains deposited on the stigma per flower was 922.76 ± 47.08 (Figure 3), and the number of round and fusiform pollen grains deposited differed significantly (Table 3). Importantly, the number of round pollen grains was significantly higher than that of fusiform grains, and the number of round and fusiform pollen grains varied significantly between the 2 years ( Figure 3). The percentage of fusiform pollen grains was relatively stable in the marginal population between the 2 years, but varied significantly in the refugium population ( Figure 4a; Table 4), although <20% fusiform pollen grains were deposited on the stigma by bumblebees in both populations in both years.
| Airborne pollen and efficiency of wind pollination
The number of both round and fusiform pollen grains was higher in the refugium population than that in the marginal population, and the number of fusiform pollen grains was generally higher than that of round pollen grains (Figure 5a; Table 3). However, the percentage of fusiform pollen grains varied significantly between years and populations ( Figure 5b).
The percentage of airborne fusiform pollen grains deposited on the stigma by wind was generally more than 50% (Figure 4b; Table 4).
The percentage of fusiform pollen grains was generally higher in the marginal population than that in the refugium population across the 3 years (Table 4). The seed number resulting from flowers pollinated by airborne pollen grains varied significantly across years, but no significant difference between the two populations was observed ( Figure 6, Table 2).
| Intensified wind pollination after range expansion
Evolution of a wind-pollination system is generally associated with open habitats, unisexual flowers, dioecy, the uniovulate condition, small plain flowers, and a lack of nectar (Friedman & Barrett, 2008).
Furthermore, a number of other species traits, including condensed inflorescences, flowers with exposed anther and stigma and great pollen production, are more common in wind-pollinated plants than in animal-pollinated plants (Culley et al., 2002;Weller, Sakai, Culley, Campbell, & Dunbar-Wallis, 2006 exaptation in transitions to wind pollination (Welsford et al., 2016).
Nevertheless, it is generally accepted that floral traits differ markedly between wind-and animal-pollinated plants, but the traits associated with the shift to wind pollination have been rarely examined at an intraspecific level. The present findings on the ambophilous Aconitum gymnandrum suggested that most of the measured floral traits differed significantly between the marginal and refugium populations.
Importantly, condensed inflorescences, reduced width of the lateral sepals, and a higher number of stigmas in the marginal population were strongly suggestive of an evolutionary trend toward wind pollination after range expansion, because these traits facilitate pollen export and deposition by wind (Culley et al., 2002;Weller et al., 1998).
Reproductive assurance has been proposed to be an important selective pressure by maintaining outcrossing in unpredictable animal-pollinator environments (Dafni & Dukas, 1986;Gomez & Zamora, 1996;Goodwillie, 1999;Karrenberg, Kollmann, & Edwards, 2002;Totland & Sottocornola, 2001). In geographically marginal populations, pollinator services are generally deemed to be less reliable than those of core populations because of the small population size and density due to founder effects (Aizen & Feinsinger, 1994;Groom, 1998). This assertion was supported by the present observations because the bumblebee visitation rate in the marginal population was considerably lower than that in the refugium population of A. gymnandrum ( Figure 2). Furthermore, the bumblebee visitation rate was highly variable in both populations in different years (Table 2). Therefore, wind pollination may provide reproductive assurance and maintain outcrossing in the biennial A. gymnandrum when bumblebees are infrequently active in the marginal population, although the pollination efficiency of bumblebees is significantly higher than that of wind in this species (Duan et al., 2009).
The dispersal agents for pollen grains differ between wind-and animal-pollinated plants; thus, the range in pollen grain size of wind-pollinated plants (17-58 μm) tends to be smaller than that of animal-pollinated plants (5-200 μm), although the difference in average pollen grain size is not significant among these groups (Wodehouse, 1935). In A. gymnandrum, two types of pollen grains that differ in morphology were observed, but the size of round and fusiform pollen grains was generally small (<30 μm) (Figure 1b).
Examination of pollinated stigmas indicated that both types of pollen grains were transported by bumblebees and wind in both populations (Figures 3 and 5), but the ratio between the two pollen types was distinct. When flowers were visited once by a bumblebee, the proportion of fusiform pollen grains on the stigma per flower was <15%. However, the proportion of fusiform pollen grains on the stigma of wind-pollinated flowers was more than 50% in the two populations across the 3 years. These results suggest that fusiform pollen grains could be dispersed more easily by wind than round pollen grains, and thus, we infer that wind pollination is mainly mediated by fusiform pollen in the ambophilous A. gymnandrum, although round pollen grains also contributed to wind pollination to a lesser degree and the percentage of fusiform pollen grains accounted for less than 10% of the pollen produced by a flower. Collectively, we conclude that wind pollination was enhanced in the marginal population of A. gymnandrum, which was generally mediated by fusiform pollen grains and facilitated by floral traits associated with wind pollination.
montanum, the lateral sepals are considered to be important for successful pollination (Fukuda, Suzuki, & Murata, 2001), and the distance between the two sepals is under selection based on the body size of bumblebees (Brink, 1980). However, the sepals of A. gymnandrum are markedly degenerate in comparison with other Aconitum species, leading to the different floral architecture in A. gymnandrum (Wang et al., 2001(Wang et al., , 2013. Specifically, the two lower sepals of A. gymnandrum occupy the position of the two lateral sepals of other Aconitum species. Thus, the distance between the two lower sepals in A. gymnandrum, rather than the distance between the two lateral sepals in other Aconitum species, might be under selection by bumblebees, which might be mirrored by the observation that the distance between the two sepals showed no difference between the two populations. The high pollination efficiency of bumblebees (Figure 2) in the present experiments strongly indicated that A. gymnandrum depends mainly on bumblebees for seed production, and the actual contribution to seed production by wind pollination was minor in comparison with bumblebee pollination, despite the stable occurrence of wind pollination in this species.
The frequency of ambophily is generally very low (Culley et al., 2002) but might be underestimated. For example, some plants that were once considered to be only wind pollinated or insect pollinated were identified to be pollinated by both wind and insect (Anderson, Overal, & Henderson, 1988;Gong et al., 2015;Peeters & Totland, 1999). Therefore, ambophily might represent an adaptation to different environments that vary in conditions favoring wind or animal pollination. Generally, for plants inhabiting open or alpine locations where animal pollinators are rare, wind pollination could be more common and dependable (Goodwillie, 1999;Totland & Sottocornola, 2001). In contrast, in closed habitats or low altitudes where animal pollinators are frequent, animal pollination could ensure reproductive assurance, and thus, ambophily is flexible and important to assure seed production. However, it is still uncertain whether ambophily is a stable stage or a transitional condition to either full wind pollination or animal pollination (Culley et al., 2002;Friedman & Barrett, 2009). Our observations of pollen dimorphism and the associated differentiation (to a certain degree) in pollen dispersal agents suggest that ambophily is a stable stage in A. gymnandrum because pollen dimorphism might be genetically controlled, which needs be demonstrated in future researches. Furthermore, selective pressures favoring floral traits associated with pollination by wind (e.g., degenerate sepals and pollen dimorphism) and by bumblebees (e.g., stable distance between the lower sepals) also support the hypothesis that ambophily is stable in A. gymnandrum. Collectively, wind pollination could overcome the shortage of bumblebee to assure seed production of this biennial in conditions (e.g., low temperature during glacial stage) or locations (e.g., new habitats) where bumblebee service is limited, and the mixed pollination system might be greatly beneficial for colonization of the high-altitude Qinghai-Tibet Plateau by A. gymnandrum in comparison with other Aconitum species. | 2018-04-03T01:07:08.987Z | 2016-12-20T00:00:00.000 | {
"year": 2016,
"sha1": "92a8a7fcaf040095c83af15a32a7d73bae813faa",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.1002/ece3.2636",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "92a8a7fcaf040095c83af15a32a7d73bae813faa",
"s2fieldsofstudy": [
"Environmental Science"
],
"extfieldsofstudy": [
"Biology",
"Medicine"
]
} |
17847800 | pes2o/s2orc | v3-fos-license | Search for Single Top Quark Production Using Likelihood Discriminants at D0 in Run II
We present an improved search for single top quarks in two production modes, s-channel (tb) and t-channel (tqb). The search is performed in the electron+jets and muon+jets decay channels, with one or more b-tagged jets, on nearly 370 pb^{-1} of D0 Run II data collected between August 2002 and October 2004. Impact-parameter based b-quark tagging is used to select signal-like events. We use a likelihood discriminant method to separate signals from backgrounds. The resulting expected/observed 95% confidence level upper limits on the single top quark production cross sections are 3.3/5.0pb (s-channel) and 4.3/4.4pb (t-channel).
INTRODUCTION
The top quark was originally discovered in 1995, at the Fermilab Tevatron pp Collider Run I by the CDF and DØ collaborations [1]. It was observed in its tt production mode via the strong interaction (qq → g → tt). Within the Standard Model, another production mode via the electroweak interaction is possible. This mode is called single top quark production as only one top quark is produced with another b quark through the Wtb vertex. As a consequence, a measurement of the single top quark production cross section can be used to constraint the magnitude of the CKM matrix element V tb and study the properties of the Wtb coupling. The two main Feynman diagrams for sand t-channel single top quark production at the Tevatron Run II are given in Fig. 1.
Single top quark production has not yet been observed and is more challenging than tt production due to smaller cross sections (2.86 pb in total, with s-channel cross section to be 0.88 ± 0.14 pb and t-channel cross section to be 1.98 ± 0.30 pb) and a much larger, less discriminable background. We present a new analysis of ∼ 370 pb −1 of DØ Run II data using a likelihood discriminant method to separate signals and backgrounds and we derive 95% confidence level upper limits to the s-and t-channel single top quark production cross sections.
The dominant Feynman diagrams for single top quark production at the Tevatron pp Collider: s-channel, tb final state (left diagram) and t-channel, tqb final state (right diagram). In this note, we use the simplified notation tb and tqb which implicitly includes all possible charge conjugations.
The DØ detector for Run II, completely described in [3], consists of a central tracking system, a liquid-argon/uranium sampling calorimeter and an iron toroid muon spectrometer. The central tracking system is composed of a silicon microstrip tracker (SMT) and a central fiber tracker (CFT), both located into a 2T superconducting solenoidal magnet.
The SMT detector has about 800000 individual strips and its design is optimized for tracking and vertexing capabilities allowing heavy flavour tagging. The calorimeter is longitudinally segmented into electromagnetic and hadronic layers and is housed into three cryostats: a central barrel covering |η| ≤ 1.1 and two end-caps that extend coverage up to |η| ≤ 4. The muon system resides beyond the calorimeter and consists of a layer of tracking detectors and scintillation counters before the toroidal magnet, followed by two similar layers after the toroid. Tracking in the muon system relies on wide or mini drift tubes depending on the acceptance (up to |η| = 2).
ANALYSIS OVERVIEW
This analysis focuses on the final state topology where the top quark decays into a b quark and a W boson, which subsequently decays leptonically (W → eν, µν). The general selection is designed to reject misreconstructed events and to select a signal-like data sample that is well reproduced by Monte Carlo backgrounds samples. The b-tagging requirement enhances the discrimination between signals and the W +light jet and multijet backgrounds. We estimate, mostly from Monte Carlo samples, the yields for the main backgrounds after the final procedure which includes trigger and selection effects as well as b quark jet tagging. Due to the presence of two dominant classes of background with different kinematic properties (W +jets and tt-like events), two likelihood discriminant variables are built. The estimated yields and likelihood distributions for the backgrounds are confronted to the number of observed events in data to extract 95% confidence level upper limits using a Bayesian fit.
LIKELIHOOD DISCRIMINANT METHOD
After the event selection, a final discriminating variable is constructed in order to efficiently characterize the signal type events and reject the background type ones, based on the shapes of the mostly uncorrelated input variables. Examples of likelihood filters outputs for signal and backgrounds are given in Fig. 2 and Fig. 3. The number of observed events is consistent with the background prediction for both muon and electron channels and for all b-tagging schemes, within the total uncertainties. We therefore set upper limits at the 95% confidence level, using a Bayesian approach [4]. The | 2007-07-11T03:33:11.000Z | 2007-07-11T00:00:00.000 | {
"year": 2007,
"sha1": "a884713f3e204b39a828977ffa2598e41a4f201e",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "fce594362aceeb333b1efa4609f55c413dd0371d",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
119210490 | pes2o/s2orc | v3-fos-license | Line identifications in the spectrum of $\chi$ Lupi A
We present new abundance determinations for the sharp-lined HgMn star \chi{} Lupi A from archival FEROS spectra. Selected unblended lines with accurate atomic data have been synthesized to derive new abundances for \chi{} Lupi A. These spectra show evidence of the presence of blue-shifted lines of the companion \chi{} Lupi B. The synthesis of the spectrum of \chi{} Lupi B confirms that this star is cooler, probably an early A star with normal abundances.
Introduction
Previous studies of HgMn star χ Lupi A have reported overabundances of iron-peak elements in its atmosphere and pronounced overabundances of heavy elements. The last extensive abundance analysis from optical spectra is that of Wahlgren et al. (1994). The low rotational velocity of this star facilitates continuum placement and line synthesis. It also favors a radiative atmosphere little mixed by rotation. Overabundances and underabundances probably reflect an efficient action of radiative acceleration on these heavy elements which have rich transitions, accumulating these elements in the line forming region.
The aim of this work is therefore to provide determinations of new abundances of heavy elements, using upgaded atomic data. As our spectra obviously show the presence of the blue-shifted lines of the companion χ Lupi B, we have also attempted to model the lines of χ Lupi B.
Observed spectra and reduction
The observed FEROS spectrum (R = 48000) of χ Lupi has been retrieved from the ESO archive. This FEROS spectrum spans a wide wavelength range from 3700Å up to 7500Å. The exposure time of the spectrum is 50 seconds and the signal-to-noise ratio is 325.
Synthetic spectrum computations and abundance determinations
The fundamental parameters have been derived using the UVBYBETA program (Napiwotzki et al. 1993). For χ Lupi A, this yields T eff = 10608 ± 200 K, v sin i = 5.0 ± 0.5 km s −1 , log g = 3.98 ± 0.25 dex. We have derived a microturbulence velocity of ξ = 0.10 ± 0.20 km s −1 by requesting that strong and weak lines of Fe II yield the same iron abundance.
We computed a model atmosphere with the ATLAS9 code (Kurucz 1993) with 72 parallel layers assuming Local Thermodynamical Equilibrium (LTE), Radiative Equilibrium (RE) and Hydrostatic Equilibrium (HE). Synthetic spectra were computed using SYNSPEC49/SYNPLOT (Hubeny & Lanz 1995) code by using as first solution the solar abundances. In order to compute the composite spectrum consisting of the spectra of A and B components, we modified SYNPLOT interface for binary stars, into a new interface which we call SYNPLOTBIN. This interface computes the flux spectrum of the components individualy using SYNSPEC49, combines them and then normalizes them using the theoretical continuum fluxes for the given atmospheric parameters.
For χ Lupi A, we find distinct underabundances of He, C, nearly solar abundances for O, Mg, Al, S, Ca, Sc, and Fe. We find mild overabundances for P and most of the iron-peak elements. We find pronounced overabundances for the the Sr-Y-Zr triad, Ba and Hg (about 100000 ). Using the found abundances, we are preparing the first list of identifications for all lines absorbing more than 2 % in the spectrum of χ Lupi A from 3700Å up to 7500Å. The modelling of the the lines of χ Lupi B suggests this star is a A3 dwarf with a normal surface composition and we confirm the atmospheric parameters of T eff =9200 K and log g= 4.00 found by Wahlgren et al. (1994). Figure 1 shows the composite synthetic spectrum for χ Lupi A and B superimposed onto the observed spectrum. At the time of the observation, χ Lupi A was redshifted with an orbital radial velocity of 15 km s −1 while χ Lupi B was blueshifted with a velocity of −56 km s −1 .
Conclusions
Using the upgraded atomic data, we have derived a new set of abundances using a high resolution, high signalto-noise FEROS spectrum of χ Lupi A+B. The synthesis of the blue-shifted lines of χ Lupi B confirms that it is a superficially normal A3 dwarf. | 2018-10-05T07:48:06.000Z | 2018-10-05T00:00:00.000 | {
"year": 2018,
"sha1": "df3f6c792ba375b1f9b6b44345569d07123afbcc",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "df3f6c792ba375b1f9b6b44345569d07123afbcc",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
269407971 | pes2o/s2orc | v3-fos-license | Ginsenoside Rh2 shifts tumor metabolism from aerobic glycolysis to oxidative phosphorylation through regulating the HIF1-α/PDK4 axis in non-small cell lung cancer
Background Ginsenoside Rh2 (G-Rh2), a steroidal compound extracted from roots of ginseng, has been extensively studied in tumor therapy. However, its specific regulatory mechanism in non-small cell lung cancer (NSCLC) is not well understood. Pyruvate dehydrogenase kinase 4 (PDK4), a central regulator of cellular energy metabolism, is highly expressed in various malignant tumors. We investigated the impact of G-Rh2 on the malignant progression of NSCLC and how it regulated PDK4 to influence tumor aerobic glycolysis and mitochondrial function. Method We examined the inhibitory effect of G-Rh2 on NSCLC through I proliferation assay, migration assay and flow cytometry in vitro. Subsequently, we verified the ability of G-Rh2 to inhibit tumor growth and metastasis by constructing subcutaneous tumor and metastasis models in nude mice. Proteomics analysis was conducted to analyze the action pathways of G-Rh2. Additionally, we assessed glycolysis and mitochondrial function using seahorse, PET-CT, Western blot, and RT-qPCR. Result Treatment with G-Rh2 significantly inhibited tumor proliferation and migration ability both in vitro and in vivo. Furthermore, G-Rh2 inhibited the tumor’s aerobic glycolytic capacity, including glucose uptake and lactate production, through the HIF1-α/PDK4 pathway. Overexpression of PDK4 demonstrated that G-Rh2 targeted the inhibition of PDK4 expression, thereby restoring mitochondrial function, promoting reactive oxygen species (ROS) accumulation, and inducing apoptosis. When combined with sodium dichloroacetate, a PDK inhibitor, it complemented the inhibitory capacity of PDKs, acting synergistically as a detoxifier. Conclusion G-Rh2 could target and down-regulate the expression of HIF-1α, resulting in decreased expression of glycolytic enzymes and inhibition of aerobic glycolysis in tumors. Additionally, by directly targeting mitochondrial PDK, it elevated mitochondrial oxidative phosphorylation and enhanced ROS accumulation, thereby promoting tumor cells to undergo normal apoptotic processes.
Introduction
Lung cancer, the most common cause of cancer-associated deaths worldwide (Hirsch et al. 2017), accounts for 85% of non-small cell lung cancer (NSCLC).Despite many advances within the treatment of NSCLC in the past two decades, the overall cure and survival rates are relatively low (Wang et al. 2021;Miller and Hanna 2021).Currently, the main treatments for NSCLC have involved surgical treatment, postoperative radiotherapy, and targeted therapy, but all are ineffective.Aerobic glycolysis caused by an altered tumor microenvironment is an important factor contributing to malignant tumor progression (Chu et al. 2016;Long and Suresh 2020).Thus, new adjuvant treatment strategies are urgently needed.Currently, the advancement of Chinese medicine is providing novel insights into the treatment of lung cancer.In recent years, various forms of complementary/alternative medicine have been used to treat cancer, among which traditional Chinese medicine has been widely accepted as the dominant form of complementary and alternative therapy (Xiang et al. 2019).Some herbal medicines are now being widely studied due to their low toxicity and high effectiveness.
Ginseng, one of the oldest and most widely known herbal medicines, has long been used as a tonic to enhance immunity and slow down aging.Ginsenosides are the active ingredients found in Chinese ginseng.Previous studies have indicated that ginsenoside Rh2 (G-Rh2) can reduce neurological damage in neurological diseases (Chen et al. 2022), possess anti-inflammatory activity against rheumatoid arthritis (Tang et al. 2021), anti-diabetic (Alolga et al. 2020) and anti-fatigue (Luo et al. 2019) properties, anti-arrhythmic effects in cardiovascular diseases (Sun et al. 2016), and antitumor effects (Xiaodan and Ying 2022;Li et al. 2020a, b;Man et al. 2021).G-Rh2, a major active component of ginsenosides, has shown promising advantages in cancer treatment (Li et al. 2020a, b).For example, G-Rh2 induces apoptosis in breast cancer cells by upregulating endoplasmic reticulum stress (Liu et al. 2022); inhibits the growth of liver cancer cells (Chen et al. 2021), and triggers apoptosis in colorectal cancer cells through activation of the P53 pathway (Li et al. 2011).G-Rh2 exists in two structures: S-and R-type.The 20(S)-Rh2 monomer, which is the primary component isolated from ginseng, has demonstrated its major effect on tumor therapy (Dong et al. 2011).Therefore, this study focused on 20(S)-Rh2.Since the specific mechanism of action of G-Rh2 in NSCLC was unclear, this study was aimed to investigate how G-Rh2 inhibited malignant progression in NSCLC.
In the presence of oxygen, tumors produce adenosine triphosphate (ATP) through glycolysis instead of mitochondrial oxidative phosphorylation, a phenomenon known as the "Warburg effect" (Vander Heiden et al. 2009).This metabolic process can result in the production of large amounts of lactate.Tumor cells gain advantages in survival and proliferation within the tumor microenvironment by undergoing metabolic reprogramming (Hsu and Sabatini 2008).On the other hand, normal cells use glycolysis under aerobic conditions to convert glucose into pyruvate, which is further metabolized in mitochondria through the tricarboxylic acid cycle (TCA) and oxidative phosphorylation, ultimately producing ATP (Abdel-Wahab et al. 2019).Sodium dichloroacetate (DCA) serves as an inhibitor of pyruvate dehydrogenase kinase (PDKs) (Semba et al. 2016;Khan et al. 2017;Fujiwara et al. 2013), which stimulates the process of mitochondrial oxidative phosphorylation by inhibiting PDK and activating the pyruvate dehydrogenase complex (PDC) (Zhou et al. 2015).Currently, DCA has been used in clinical practice as an anticancer drug targeting glycolysis (Pathania et al. 2009).However, due to the hepatotoxicity (Stacpoole et al. 2008) and neurotoxicity (Stacpoole 1989;Schaefer 2006;Yount et al. 1982;Spruijt et al. 2001) associated with high doses of dichloroacetate, it is not suitable for long-term clinical use.Therefore, finding a way to minimize the dosage of DCA while simultaneously increasing its inhibitory effect on tumor glycolysis and enhancing mitochondrial function is necessary.
The hypoxia-inducible factor HIF1-α is one of the major drivers of glucose metabolism in cancer cells.It regulates metabolic energy shifts by promoting the expression of several glycolytic enzymes (Zheng et al. 2021;Lai et al. 2013).HIF1-α inhibits pyruvate dehydrogenase (PDH) activity, which affects mitochondrial function by inducing the expression of PDK.This, in turn, promotes the production of lactic acid and the ability of aerobic glycolysis.Changes in PDK4, a key enzyme in both aerobic glycolysis and mitochondrial glucose oxidation, directly impact mitochondrial function as well (Leclerc et al. 2017;Kinnaird et al. 2016).As a downstream signaling molecule of HIF1-α, the role of HIF1-α/ PDK4 in glycolysis and mitochondrial function is critical (Leclerc et al. 2017;Lu et al. 2021;Li et al. 2023).
Because ginseng supports the body's positive energy, ginsenosides Rh2 might potentially induce a reversion PDK, it elevated mitochondrial oxidative phosphorylation and enhanced ROS accumulation, thereby promoting tumor cells to undergo normal apoptotic processes.
to normal oxidative phosphorylation by inhibiting aerobic glycolysis in tumors.However, further exploration is needed to understand the potential mechanism.Therefore, the objective of the current study was to clarify the mechanisms underlying the effects of G-Rh2 on aerobic glycolysis and oxidative phosphorylation in NSCLC.
Our preliminary research has found that G-Rh2 can potentially serve as an inhibitor of PDK4, thereby inhibiting tumor glycolysis, promoting mitochondrial oxidative phosphorylation, contributing to the benign transformation of tumors, and inducing tumors to enter the normal apoptotic program.Additionally, when combined with DCA, it showed the advantage of increasing efficacy and reducing toxicity.These findings provided evidence that tonic herbs did not directly kill tumors and might present a novel approach in clinical tumor therapy.
Cell culture
The authenticated human lung adenocarcinoma cell lines A549 and PC9 were obtained from the Type Culture collection of the Chinese Academy of Sciences and were stored in the Shanghai Key Laboratory of Molecular Imaging.The A549 cell line was cultured at 37 ℃ under 5% CO 2 in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Gibco) and 1% penicillin-streptomycin solution (Gibco).The PC9 cell line was cultured at 37 ℃ under 5% CO 2 in RPMI-1640 supplemented with 10% FBS (Gibco) and 1% penicillin-streptomycin solution (Gibco).
Cell proliferation assays
Cell proliferation viability was assessed using a CCK-8 assay kit.Cells were seeded at a density of 5.0 × 10 5 cells/ well in a 96-well plate and incubated at 37 ℃ for 24 h to allow adhesion to the plate.After treating the cells with different concentrations of G-Rh2 for 24 h, 10% cell counting kit-8 (CCK-8) solution was added and incubated for 1 h at 37 ℃.The optical density was measured at 450 nm using a microplate reader to determine the IC50 value.Subsequently, the proliferation rate was determined after G-Rh2 treatment.
Ethynyl deoxyuridine (EdU) analysis
Cell proliferation viability was assessed using an EdU cell proliferation assay kit (UE, China).Cells were seeded at a density of 3.0 × 10 5 cells/well in a 24-well plate and incubated at 37 ℃ for 24 h to allow adhesion to the plate.Subsequently, the cells were treated with different concentrations of G-Rh2 for 24 h.Cells were labeled with EdU reagent and incubated for 2 h, fixed in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and stained with Hoechst 33,342.Images were obtained using an inverted fluorescence microscope (Olympus) after sample preparation.The proliferation rate was determined by calculating the ratio of EdU-positive cells to the total number of Hoechst 33,342-positive cells.
Colony formation assay
Cell proliferation viability was assessed using colony formation assay.Cells were seeded at a density of 1.0 × 10 3 cells/well in a 12-well plate and incubated at 37 ℃ for 24 h to allow adhesion to the plate.Subsequently, the cells were treated with different concentrations of G-Rh2 for 24 h.The medium was replaced with normal medium, and the cells were cultured for 1 week.The cells were then fixed in 4% paraformaldehyde, stained with 1% crystal violet, and photographed.The colony formation rate was calculated as (clone numbers/plated cell numbers) × 100%.
Apoptosis and cell cycle assays
Apoptosis and the cell cycle were analyzed using flow cytometry.Cells were seeded at a density of 3.0 × 10 5 cells/well in a 12-well plate and incubated at 37℃ for 24 h to allow adhesion.Then, they were treated with different concentrations of G-Rh2 for 24 h.Detection was performed using an apoptosis detection kit.The cells were stained with an Annexin V-fluorescein isothiocyanate conjugate and propidium iodide at room temperature for 15 min in the dark.The apoptotic rate was determined using a NovoCyte 2000 Flow Cytometer (Agilent Technologies, Inc, CA, USA).
Cell cycle detection was performed using a cell cycle detection kit (Beyotime, China).Cells were fixed in 70% ethanol for 30 min at 4℃ and stained with propidium iodide in a water bath maintained at 37℃, away from light.Subsequently, the apoptotic rate was determined using a NovoCyte 2000 Flow Cytometer (Agilent Technologies, Inc, CA, USA).
Cell migration assays
Cell migration ability was assessed using transwell assays.Cells were seeded at a density of 5.0 × 10 4 cells/well in the upper chamber of a transwell culture insert (8 μm pore size; Corning, NY, USA), and then treated with different concentrations of G-Rh2 for 24 h.The lower chamber was filled with DMEM supplemented with 15% FBS.After incubation at 37℃ for 24 h to allow adhesion and migration, the cells were fixed in 4% paraformaldehyde and stained with 1% crystal violet.Subsequently, the cells on the upper surface of the membrane were removed using cotton swabs, and images were captured using an inverted fluorescence microscope (Olympus).
Wound healing assays
Cell migration ability was assessed using wound healing assays.Cells were seeded at a density of 4.0 × 10 5 cells/ well in a six-well plate, and incubated at 37℃ for 24 h until they reached 100% confluence.The surface was then scratched using a 200 µl pipette tip, and the cells were treated with various concentrations of G-Rh2 for 24-48 h.Wound images were taken at 0, 24, or 48 h of incubation using an inverted fluorescence microscope (Olympus).
RNA isolation and quantitative reverse transcriptionpolymerase chain reaction (RT-qPCR)
The primers utilized for qPCR were as follows:
Glucose intake, pyruvic acid and lactification assays
Cells were plated at a density of 4.0 × 10 5 cells/well in a six-well plate, incubated at 37℃ for 24 h to allow adherence, and then treated with different concentrations of G-Rh2 for 24 h.Subsequently, the cells were cultured in DMEM without FBS for serum starvation.The samples were assayed using a glucose detection kit (Sigma), lactate detection kit (Nanjing Jiancheng Bio), and pyruvate content detection kit (Solarbio, China).After adding the detection reagent, the optical density values were measured at 540 nm, 530 nm, and 520 nm using a microplate reader.A standard curve was generated by analyzing different standard dilutions, and the resulting OD values were utilized to determine glucose intake, lactification, and pyruvic acid production.
ATP content assays
Cells were plated at a density of 4.0 × 10 5 cells per well in a six-well plate.They were incubated at 37 ℃ for 24 h to allow adherence and then treated with different concentrations of G-Rh2 for 24 h.Afterward, 200 µl of lysate was added to each well of a six-well plate and centrifuged at 12,000 g for 5 minutes at 4 ºC.The supernatant was retained, and the ATP content was detected using an ATP assay kit (Beyotime, China).The RLU value was measured using a luminometer.
Reactive oxygen species (ROS) assays
Cells were plated at a density of 3.0 × 10 5 cells per well in a six-well plate, incubated at 37 ℃ for 24 h to allow adhesion and subsequently treated with different concentrations of G-Rh2 for 24 h.The levels of ROS were detected using a ROS detection kit (UE, China).The DCFH-DA probe was diluted with serum-free medium, and the cells were further incubated in a cell incubator at 37 °C for 30 min in the dark.The fluorescence intensity was then measured using a NovoCyte 2000 Flow Cytometer (Agilent Technologies, Inc, CA, USA) or captured using a fluorescence microscope (Leica, US).
γ-count assays
Cells were plated at a density of 4.0 × 10 5 cells per well in a 12-well plate, incubated at 37℃ for 24 h to allow adhesion and then treated with different concentrations of G-Rh2 for 24 h.Subsequently, 18 FDG-PBS solution (4 µCi/ml) was added to each well and incubated at 37℃ for 1 h.Next, a 0.5 M NaOH solution was added to lyse the cells, after which they were placed in the detector tube.CPM values were measured using a γ-counter and standardized according to the cell counting method to analyze the experimental results.
PET-CT assays
Nude mice (4-6 weeks old, weighing 15-20 g) that had been subcutaneously transplanted with A549 cells were treated with G-Rh2.The treatment involved intraperitoneal injection of 20 mg/kg every other day for a duration of 2 weeks.A dose of 2 mCi/ml was administered via tail vein injection at a volume of 100 µl per mouse.After allowing 30 min for metabolism, the mice underwent PET-CT scanning using equipment from Mediso in Hungary.The region of interest (ROI) for the tumor was then delineated, and the maximum standardized uptake value (SUVmax) was calculated to analyze the experimental results.
Tumor mouse model establishment
BALB/c nude mice (4-6 weeks old, weighing 15-20 g) were purchased from Shanghai Jihui Laboratory Animal Care Co.Ltd.This experiment was approved by The Animal Ethics Committee of Shanghai University of Medicine and Health Science.
A subcutaneous tumor model in nude mice was established by injecting 100 µl of PBS solution containing 5 × 10 6 A549 cells into the right upper limb.After 1 week, the mice were divided into two groups: (a) the control group receiving 100 µl of PBS solution, and (b) the G-Rh2 treatment group receiving a dose of 20 mg/kg body weight.Tumor volume and mouse weight were measured every other day.
To establish a model of tail vein metastasis, luciferaselabeled A549 cells (3 × 10 5 in 100 µl of PBS) were injected into the tail vein.Two weeks after the injection, G-Rh2 (intraperitoneal injection at a dose of 20 mg/kg, every other day) was administered for 2 weeks.The lung metastasis of tumor cells was then detected by measuring the fluorescence signal in the lungs using an animal bioluminescence imaging system 15 min after intraperitoneal injection of D-fluorescein sodium salt.The number of metastatic foci in lung tissues was determined through hematoxylin and eosin (HE) staining.
To establish the lymphatic metastasis model, luciferase-labeled A549 cells (3 × 10 5 in 100 µl of PBS solution) were injected into the paw pads of nude mice.Treatment with G-Rh2 at a dose of 20 mg/kg (intraperitoneal injection, every other day for a total of 14 days) was initiated 2 weeks after the injection.Two weeks later, the lymphatic metastasis of tumor cells was detected by measuring the fluorescence signal in the lower leg using an animal bioluminescence imaging system.This measurement was taken 15 min after intraperitoneal injection of D-fluorescein sodium salt, allowing us to determine the number of metastatic tumor cells in the plantar region.
Proteomic analysis
A549 cells were seeded at a density of 2.0 × 10 6 cells/well in a 10 cm dish and incubated at 37℃ for 24 h to allow adherence.After that, the cells were either treated with G-Rh2 for 24 h or left untreated.Subsequently, the cells were lysed using RIPA buffer containing 1mM phenylmethanesulfonyl fluoride.The protein concentration was determined using a BCA kit (Yeasen, China).The protein was stored at -80℃ until further experiments.For total protein extraction, 250 µg of protein was obtained and then reduced with 10 mM DTT at 37℃ for 1 h.Subsequently, it was alkylated with 50 mM IAA at 25℃ in the dark for 1 h.Next, 1.5 mL of prechilled (-20℃) 100% acetone was added, and the samples were centrifuged at 14,000 g three times for 15 min to remove impurities.The samples were then diluted with 500 µL NH 4 HCO 3 .Peptides were desalted with SPE C18 cartridges (Thermo Fisher Scientific) and lyophilized under vacuum prior to nano-HPLC-MS/MS analyses.Three biological replicates were analyzed to evaluate the reliability of protein identification, reproducibility, and accuracy.
Immunohistochemical analysis
The tumors from both the control and ginsenosidetreated groups were dissected.The tumor tissue was then embedded in paraffin and cut into approximately 5 μm thick sections.These sections were deparaffinized and fixed in 4% paraformaldehyde.Subsequently, they were equilibrated with an equilibration buffer and incubated at 37 °C for 60 min to allow for terminal reactions.To block the endogenous peroxidase, 3,3'-diaminobenzidine (DAB) was used for color development.The slides were sealed with 100% glycerin, and the stained tissue was observed and photographed under a microscope.
Immunofluorescence analysis
The dissected lymph node tissues were fixed using 4% paraformaldehyde, sectioned, and placed in a blocking solution to prevent non-specific binding.Then, the tissues were permeabilized and blocked.The primary antibody was incubated overnight at 4 °C, followed by incubation with the secondary antibody at room temperature in the dark.The slides were counterstained with DAPI and sealed using an anti-quenching agent.Finally, the staining was photographed under a fluorescence microscope.
Extracellular acid ratio (ECAR) and oxygen consumption rate (OCR) analysis
Drug-treated A549 cells were incubated overnight for 24 h at a density of 15,000 cells per well.The cells were spread flat on Seahorse XFe96 Cell Culture Microplates (Agilent, Palo Alto, CA, USA) using an 80-µl cell suspension.Probe plates (Seahorse XFe96 Flux Assay Kit) were hydrated by incubating them with a hydration standard solution (XF Calibrant) at 37 °C overnight.Before the experiment, the medium was replaced with an assay solution (DMEM with 10 mM glucose, 1 mM pyruvate, and 2 mM glutamine, pH 7.4, Agilent).The cells were then incubated in a non-CO 2 incubator for 1 h at 37 °C prior to the start of the assay.The assay was conducted using a Seahorse XF Glycolytic Rate assay Kit (Agilent) and Seahorse XF Cell Mito Stress Test Kit (Agilent) as per the manufacturer's instructions.Rotenone/antimycin A (0.5 µM), 2-DG (50 mM), oligomycin (1 µM), FCCP (0.5 µM), and rotenone/antimycin A (0.5 µM) were added to the specific dosing wells following the manufacturer's instructions.The OCR and ECAR data were measured and plotted using Seahorse XF96 software.
Statistical analysis
All results were statistically analyzed using GraphPad Prism (GraphPad Software, Inc., CA, USA).T-tests were used for comparisons between two groups, and one-way analysis of variance (ANOVA) was used for comparisons among three or more groups.The data are presented as mean ± standard deviation (SD).P-values of ≤ 0.05 were considered statistically significant.
G-Rh2 inhibited tumor cell proliferation and promotes apoptosis in vitro
To assess the inhibitory effects of G-Rh2 on the proliferative ability of tumor cells in vitro, we conducted proliferation assay experiments using NSCLC A549 and PC9 cells.The chemical structure formula of G-Rh2 is shown in Fig. 1A.Initially, we performed the CCK8 experimental assay, which revealed that G-Rh2 treatment for 24 h inhibited the proliferation ability of A549 and PC9 cells in a concentration-dependent manner.The IC50 values were determined as 41.13 µg/ml for A549 and 34.16 µg/ ml for PC9 (Fig. 1B, C).Thus, the concentration of G-Rh2 at 40 µg/ml for A549 and at 35 µg/ml for PC9 was selected for subsequent experiments.G-Rh2 treatment was performed for 0, 24, 48, or 72 h at the abovementioned concentrations, while the control group received vehicle alone.The data presented in Fig. 1D and E confirmed that G-Rh2 inhibited the proliferative ability of NSCLC cells in a time-dependent manner.EdU assays further confirmed that G-Rh2 significantly inhibited NSCLC cell proliferation, as evidenced by reduced numbers of EdU-positive cells with increasing concentrations of G-Rh2 (Fig. 1F-I).Moreover, colony formation assay demonstrated that G-Rh2 treatment resulted in fewer cloned spots compared to the control group (Fig. 1J-M).Furthermore, cell cycle detection using flow cytometry revealed that treatment with the IC50 concentration of G-Rh2 arrested the cell cycle in the G0/G1 phase (A549: 56.15% blocked; PC9: 57.94% blocked) compared to the control treatment (A549: 35.55% blocked; PC9: 30.43% blocked), as well as in the S phase (A549: 24.57% blocked; PC9: 11.5% blocked) compared to the control treatment (A549: 39% blocked; PC9: 38.23% blocked) (Fig. 1N-P).Moreover, flow cytometry apoptosis detection showed that G-Rh2 treatment for 24 h promoted apoptosis in both A549 and PC9 cells in a dose-dependent manner (Fig. 1Q-T).These results indicated that G-Rh2 inhibited tumor cell proliferation, induced G1 phase cell cycle arrest, and promoted tumor cell apoptosis in vitro.
G-Rh2 inhibited tumor growth and promoted apoptosis in vivo
Considering the substantial inhibitory effect of G-Rh2 on tumor cell proliferation observed in in vitro experiments, we constructed a model of primary tumor foci to verify whether G-Rh2 also showed significant inhibition of tumor growth in animal experiments conducted in vivo.The results of the animal experiments demonstrated that intraperitoneal administration of 20 mg/kg G-Rh2 didn't decrease the weights of the mice, thus suggesting that 20 mg/kg G-Rh2 was non-toxic.Additionally, the in vivo tumor formation experiments revealed that G-Rh2 treatment significantly suppressed tumor growth, as evidenced by reductions in both tumor volume and weight (Fig. 2A-D).Furthermore, G-Rh2 treatment resulted in a significant decrease in the expression of the tumor growth marker Ki67, as observed through immunohistochemistry. Conversely, TUNEL staining indicated that G-Rh2 promoted tumor apoptosis (Fig. 2E-H).These findings indicated that G-Rh2 obviously suppressed the D, E) Effects of G-Rh2 on A549 (D) and PC9 (E) cell proliferation analyzed through CCK8 assays after IC50 treatment for 0, 24, 48, or 72 h.Data are expressed as mean ± SD. (F-I) EdU detection data showing A549 and PC9 cell proliferation ability following G-Rh2 treatment at 0, 30, or 40 µg/ml for A549 cells and 0, 30, or 35 µg/ml for PC9 cells for 24 h.Data are expressed as mean ± SD. (J-M) Colony formation assay, showing A549 and PC9 cell proliferation after G-Rh2 treatment at 0, 30, or 40 µg/ml for A549 cells and 0, 30, or 35 µg/ml for PC9 cells.Data are expressed as mean ± SD. (N-P) Flow cytometry measurement of the percentages of A549 and PC9 cells in G0/G1, S, and G2/M phases.Data are presented as the mean ± SD. (Q) Western blot detection of the cell cycle-associated protein Cyclin A2. (R-U) Flow cytometry analysis showing A549 and PC9 cell apoptosis after 24-hours treatment with 0, 30, 35, 40, or 45 µg/ml G-Rh2 Data are expressed as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.001 vs. the control group growth of primary tumor foci in vivo while concurrently inducing apoptosis.
G-Rh2 inhibited tumor cell metastasis in vitro
To examine the potential inhibitory effect of G-Rh2 on the migratory ability of tumor cells in vitro, we performed assays using transwell and wound healing assays.The transwell assays confirmed the inhibitory effect of G-Rh2 on the migration of NSCLC cells (Fig. 3A-D).Similarly, the wound healing assays verified that G-Rh2 restrained the rate of wound healing, as indicated by the measured widths of the wounds (Fig. 3E-G).These results demonstrated that G-Rh2 inhibited both invasion and migration of NSCLC cells.
G-Rh2 inhibited distal tumor metastasis in vivo
Previous studies have demonstrated the inhibitory effects of G-Rh2 on tumor cell migration and metastasis, and we constructed a metastatic foci model to verify the impact of G-Rh2.Specifically, we constructed a mouse tail vein lung metastasis model and demonstrated that G-Rh2 obviously inhibited the invasion of NSCLC cells from the tail vein to the lungs based on the results of in vivo fluorescence intensity analysis of tail vein transfer experiments (Fig. 4A, B).Moreover, through Western blot analysis of proteins associated with epithelial mesenchymal transition (EMT), we observed that G-Rh2 treatment led to an increase in E-cadherin levels while decreasing N-cadherin levels (Fig. 4C).Additionally, HE staining demonstrated that G-Rh2 treatment significantly decreased the number of foci in lung tissues (Fig. 4D-F).These results indicated that G-Rh2 inhibited the metastasis of NSCLC by inhibiting the EMT process.
Plantar lymph metastasis model assays demonstrated that G-Rh2 inhibited tumor lymph metastasis based on an observed decrease in fluorescence intensity within the lymph nodes of the feet and legs (Fig. 4G, H).Immunofluorescence analysis showed that G-Rh2 treatment obviously inhibited the expression of CD31 (Fig. 4I), indicating that G-Rh2 inhibited tumor lymph angiogenesis.Collectively, these results confirmed the in vivo inhibition of tumor metastasis by G-Rh2.
G-Rh2 inhibited tumor aerobic glycolysis by regulating HIF1-α /PDK4
To verify the relationship between G-Rh2 and HIF1-α, we conducted proteomic analysis to identify the differentially expressed proteins in A549 cells before and after G-Rh2 treatment.The analysis identified 176 upregulated and 586 downregulated proteins in G-Rh2-treated A549 cells (Fig. 5A, B).Furthermore, clusters of orthologous genes (COG) analysis indicated that the differentially expressed genes were primarily associated with signal transduction mechanisms (Fig. 5C).
KEGG enrichment analysis confirmed that G-Rh2 treatment is associated with enrichment in the metabolic pathway and oxidative phosphorylation pathway.The heatmap showed that the differences in PDK4 proteins in the metabolic pathways were more significant (Fig. 5D, E).Subcellular analyses showed that, apart from unclear enrichment, the majority of genes re-executing their functions were predominantly located within mitochondrial structures (Fig. 5F).GO analysis (Fig. 5G) also confirmed that the differentially expressed proteins after ginsenoside treatment were mainly enriched for mitochondrial redox reactions.
KEGG enrichment analysis of the proteomic findings demonstrated that G-Rh2 inhibits the growth of NSCLC through the glucose metabolic pathway.HIF1-α is known to regulate the expression of PDK4, a pivotal enzyme during the glucose oxidation process.Inhibition of PDK4 promotes the aerobic glucose oxidation process.Molecular docking modeling predicted strong binding of G-Rh2 to HIF1-α (Fig. 5H), with H-bonds formed with Lys190 Trp189, and a docking score of -5.604 kcal/ mol.Western blot assays confirmed that G-Rh2 inhibited the expression of PDK4 (Fig. 5I), and demonstrated the inhibitory effect of G-Rh2 treatment on the expressions of HIF1-α as well as VEGF in A549 cells (Fig. 5J).We examined the expression of the downstream target genes of HIF1-α.RT-qPCR results showed the inhibitory effect of G-Rh2 on the expression of its downstream target genes PKM2, GLUT1, PGK1, PDK4, and LDHA, among which the inhibitory effect of PDK4 was the most significant, indicating that G-Rh2 inhibited the HIF1-α pathway (Fig. 5K).In vitro assays of glucose intake and lactate production were performed.The G-Rh2-treated group showed significant inhibition of glucose intake as well as lactate production in NSCLC cells (Fig. 5L, M).Western blot detection proved that G-Rh2 treatment downregulated the expression of of the expression of glycolysis-related proteins HK2, LDHA, PFKFB2, and PKM2 (Fig. 5N), indicating the inhibitory impact of G-Rh2 on the glycolytic process of tumors.In vivo, G-Rh2 inhibited glucose uptake in tumor tissues, as evidenced by a comparison of the SUVmax values with the ROI regions (Fig. 5O, P).The γ-counter detected significantly inhibited glucose intake of G-Rh2-treated cells, based on a comparison of CPM values (Fig. 5Q).These results suggested that G-Rh2 regulated HIF1-α and downregulated PDK4 expression, thereby inhibiting aerobic glycolysis in NSCLC.
G-Rh2 targeted inhibition of PDK4 expression promoted mitochondrial oxidative phosphorylation
G-Rh2 has been shown to target HIF-1α and downregulate PDK4 expression, thus inhibiting tumor glycolysis.Consequently, G-Rh2 may be a potential PDK4 inhibitor.Therefore, we selected DCA, an inhibitor of PDK, for comparison.DCA competitively inhibits all PDKs.PDKs can lead to inactivation by phosphorylation, inhibiting the entry of pyruvate into the TCA cycle and promoting aerobic glycolysis.DCA inhibits the expression of PDKs, promotes PDH activity, and catalyzes the conversion of pyruvate to acetyl-CoA, which then enters the TCA cycle and completes the oxidative phosphorylation process.Western blot analysis of PDK4 expression in A549 cells treated with G-Rh2 and DCA revealed that both substances similarly downregulated PDK4 expression (Fig. 6A).Molecular docking modeling predicted a strong binding between G-Rh2 and PDK4 (Fig. 6B), forming hydrogen bonds with Leu306, Leu334, Glu254, Gly295, and His266.The docking score was − 6.687 kcal/ mol.Subsequently, in vitro assays of glucose uptake, lactate production, and pyruvic acid production in A549 cells treated with G-Rh2 and DCA revealed that both substances reduced glucose uptake, lactate production, and pyruvic acid production; promoted PDH activity; and inhibited ATP production (Fig. 6C-G).These findings indicated that G-Rh2 inhibited tumor glycolysis to a similar extent as DCA, suggesting that the inhibition of PDK4 by G-Rh2 complemented the effect of DCA.ROS levels were measured in different treatment groups and revealed that both G-Rh2 and DCA increased ROS production in A549 cells, suggesting that both treatments inhibited PDH activity by suppressing PDK4 expression, promoted oxidative phosphorylation of glucose, inhibited glycolysis, and increased ROS production (Fig. 6H-J).The bioenergetic profiles of A549 cells were examined using a Seahorse XFe96 Analyzer.The ECAR values of glycol PER in A549 cells treated with G-Rh2 and DCA were much lower in contrast with the control group, indicating that both substances inhibited the glycolytic capability of NSCLC cells (Fig. 6K, L).The basal OCR values and maximal respiration values were significantly higher in the G-Rh2 and DCA groups compared to the control group, indicating that both treatments promoted mitochondrial respiration in NSCLC cells and facilitated the oxidative phosphorylation process (Fig. 6M, N).Treatment of A549 cells and PDK4 overexpressing A549 cells with G-Rh2 revealed that PDK4 overexpression significantly increased the ECAR value of glycol PER, promoting aerobic glycolysis and reversing the inhibitory impact of G-Rh2 on A549 cells (Fig. 6O, P).Additionally, after treating A549 cells and PDK4 overexpressing A549 cells with G-Rh2, PDK4 overexpression significantly suppressed basic OCR values and maximal respiration values, which inhibited aerobic oxidative processes and reversed the ability of G-Rh2 to promote oxidative phosphorylation in A549 cells (Fig. 6Q, R).Therefore, G-Rh2 might act as a potential PDK4 inhibitor, activating PDH activity, facilitating the conversion of pyruvate to acetyl-CoA, reversing aerobic glycolysis in tumors, and promoting oxidative phosphorylation.Moreover, G-Rh2 focused on the regulation of glucose metabolism by modulating the target protein PDK4.
G-Rh2 in combination with DCA reversed the biometabolic behavior of the tumor and reduced DCA dosage
After confirming that G-Rh2 may directly target PDK4 as a potential inhibitor, we chose the PDK inhibitor DCA for combination therapy.DCA is a broad-spectrum PDK inhibitor (Bonnet et al. 2007) that is not commonly used in clinical settings due to its neurotoxicity.Co-administration of G-Rh2 at the IC50 concentration with a low concentration of DCA significantly inhibited cell viability, as revealed by CCK8 detection of A549 cell proliferation (Fig. 7A).Colony formation assay further confirmed that combined treatment with G-Rh2 and DCA resulted in fewer cloned spots compared to the control group (Fig. 7B, C).When G-Rh2 at the IC50 concentration was combined with a low concentration of DCA, transwell and wound healing assays demonstrated a significant inhibition of A549 cell migration (Fig. 7D-G).Flow cytometry apoptosis detection showed that G-Rh2 and DCA treatment significantly promoted apoptosis in A549 cells (Fig. 7H, I).The above experiments demonstrated that G-Rh2, as a potential PDK4 inhibitor, could change the metabolic behavior of tumors from aerobic glycolysis to oxidative phosphorylation.Furthermore, when combined with a low dose of DCA, it could reduce the toxicity of DCA and facilitate the shift in metabolic behavior.
Discussion
As we all know, ginseng is a high-quality traditional Chinese herb in the Wujiake family.It is widely known for its effects of tonifying vital energy, calming the mind, and improving the immune system.G-Rh2, one of the primary active components found in ginseng, has been extensively studied and shown to possess various pharmacological properties.It has demonstrated the ability to inhibit tumor progression, alleviate side effects caused by radiation and chemotherapy (Li et al. 2020a, b), and enhance immune function (Xiaodan and Ying 2022;Lee et al. 2018) in the treatment of both malignant and non-malignant diseases (Guan and Qi 2023).Several reports suggest that ginseng is widely used to treat advanced tumors (Yang et al. 2022;Xiao et al. 2019;Wang et al. 2018)In our current study, we also observed a corresponding tumor-suppressive effect of G-Rh2.It suppressed the proliferation as well as the invasion of NSCLC, and enhanced apoptosis at higher doses.However, this was not its primary function.Surprisingly, our study revealed for the first time that G-Rh2 could modulate the metabolic behavior of tumors, specifically reversing the shift from tumor aerobic glycolysis to oxidative phosphorylation, and ultimately promoting the "benignization"" of tumors.Thus, it may serve as a crucial deterrent against malignant tumor progression.
The Warburg effect, as a hallmark of cancer, means that tumors rely on glycolysis for energy, either aerobically or anaerobically (Hsu and Sabatini 2008;Liberti and Locasale 2016;Koppenol et al. 2011).Pyruvate enters the TCA cycle via PDH, and PDK facilitates the conversion of aerobic oxidation to aerobic glycolysis by inhibiting PDH activity (Leclerc et al. 2017).PDK is highly expressed in several cancers (Atas et al. 2020), which have four isoforms of PDK1-4.PDK1 is a pivotal glycolytic enzyme linked to tumor proliferation (Du et al. 2016;Deng et al. 2013) and metastasis (Dupuy et al. 2015;Zhou et al. 2019).Previous studies have shown that PDK1 can activate the PI3K/Akt/mTOR pathway to promote tumor proliferation and invasion (Jiang et al. 2021;Sambandam et al. 2019); Additionally, phosphorylation of PDK1, an upstream regulator of C-Myc, activates PLK1-MYC signaling and promotes cancer proliferation (Chinen et al. 2014).PDK1 regulation by HIF-1α promotes tumor glycolysis and malignant progression (Peng et al. 2018;Kim et al. 2006).DCA is an inhibitor of PDK1.It activates the PDC through inhibition of PDK1.The activation of PDC catalyzes pyruvate into acetyl coenzyme A, which is later converted into citrate to enter the TCA cycle and promotes the oxidative phosphorylation process in mitochondria (Zhou et al. 2015).The above demonstrates the effectiveness of targeting PDK1.PDK2 promotes multiple tumor drug resistance by inhibiting tumor mitochondrial function (Kitamura et al. 2021;Liang et al. 2020) and negatively regulating macrophage polarization (Li et al. 2017).Less research has been done in this area regarding tumors.High PDK3 expression can drive glycolysis in tumor-resistant cells (Xu et al. 2019), increasing the hypoxic response of tumor cells and promoting glycolytic processes (Li et al. 2022).PDK4 is the most widely distributed PDK subtype (Li et al. 2020a, b), an essential mitochondrial matrix enzyme in cellular energy metabolism, highly expressed in various malignant tumors (Leclerc et al. 2017).It is a key factor linking glycolysis and the TCA cycle.Targeting PDK4 to inhibit tumor glycolysis and restore mitochondrial oxidative phosphorylation function has become a hotspot for tumor therapy and an important target for clinical treatment, attracting high attention.PDK4 plays a key role in both metabolic diseases and cancer, and impaired mitochondrial function is one of the causes of several metabolic diseases.Increased activity of PDK4 in skeletal muscle inhibits glucose oxidation, thereby exacerbating blood glucose levels, and is therefore highly expressed in diabetes (Li et al. 2023).In atherosclerosis, high expression of PDK4 promotes vascular calcification (Kulkarni et al. 2012), and its absence also inhibits platelet function and arterial thrombosis (Ma et al. 2020).High expression of PDK4 in pulmonary arterial hypertension causes pulmonary artery wall cells to undergo a transition from glucose oxidation to aerobic glycolysis (Lu et al. 2021).In cancer, PDK4 is a key enzyme in aerobic glycolysis and mitochondrial oxidative reactions.Its high expression inhibits mitochondrial oxidative respiration, metabolically reprograms and promotes lactate production, which increases the glycolytic capacity of tumors, thus promoting malignant progression (Zheng et al. 2021;Li et al. 2020a, b;Flora et al. 2023;Dou et al. 2023).Currently, DCA has been used for treating mitochondrial metabolic diseases (Stacpoole et al. 2008;Barshop et al. 2004;Michelakis et al. 2017), malignant tumors (Vander Heiden 2010;Michelakis et al. 2010),etc.However, the hepatotoxicity and neurotoxicity of DCA at high doses limit its clinical use (Stacpoole 1989;Stacpoole et al. 1998;Tao et al. 2000).PDK1 and PDK4 have similarities and have been widely analyzed.
Here, we found that G-Rh2, as a fostering drug, had no severe toxic effects (Guan and Qi 2023) but an inhibitory effect against the PDK4 target for the first time.It could act synergistically with lower doses of DCA to inhibit both sites of PDK, acting as a potentiator and detoxifier.
Based on the above findings, this study suggests for the first time that the vital function of G-Rh2 is to act on the PDK4 locus.The key results demonstrated that G-Rh2 could inhibit tumor glycolysis, promote oxidative phosphorylation, stimulate ROS production, and induce apoptosis.As a result, tumors changed their biological behavior due to metabolic shifts and showed a benign trend.Combining low-dose DCA with PDK reinforces its inhibitory effect and promotes the transition from glycolysis to aerobic oxidation in tumors.Previous studies have shown that G-Rh2 can inhibit the malignant progression of NSCLC through metabolic modulation of the immune environment and other pathways (Qian et al. 2019;Ma et al. 2023;Li et al. 2018).Some studies have also discovered that G-Rh2 has an impact on aerobic glycolysis and mitochondrial oxidative phosphorylation in tumors.It activates the apoptotic program of tumor cells and demonstrates a potentiating and toxicity-reducing effect when used in combination with chemotherapeutic agents (Liu et al. 2020(Liu et al. , 2021)).Compared to the toxic effects of DCA, ginseng has been applied in clinical practice for thousands of years and is an important remedy to support the body's well-being.Our study further proved that instead of solely killing the tumors, the primary aim of Rh2 was to promote the oxidative phosphorylation function of tumors, change the behavior of tumor metabolism, promote tumors to shift to benign metabolic profile, and converge towards favorable biological behavior, ultimately leading to the activation of normal apoptotic programs.Therefore, G-Rh2 might server as a highly promising target drug with significant clinical potential.
Although we have demonstrated the potential of G-Rh2 to inhibit the malignant progression of NSCLC, there has not been an in-depth study on the specific mechanism of action in regulating PDK4.We are continuing our research to identify the direct targets of G-Rh2 against tumor mitochondrial function and glucose metabolism, hoping to provide new and effective targets as well as novel ideas for clinical tumor control.
Conclusion
This study investigated the regulatory targets and mechanisms of G-Rh2 for NSCLC.For the first time, we confirmed how G-Rh2 changed the biometabolic behavior of tumors by regulating PDK4, leading to a favorable trend towards normal apoptotic programming.We confirmed that G-Rh2 was capable of targeting and down-regulating HIF-1α's expression, subsequently reducing the expression of PDK4.It further suppressed aerobic glycolysis, promoted mitochondrial aerobic oxidative process, and stimulated the production of ROS to promote tumor cells to enter the normal apoptotic program through PDK4 regulation.The combination of G-Rh2 with DCA, an inhibitor of PDKs, dramatically reversed the biometabolic behaviors of tumors, and further reduced the dosage of DCA, thereby decreasing the toxicity of DCA.Ginseng, known primarily as a deficiency tonic, does not mainly function through direct tumor destruction.Instead, altering the biometabolic behavior of tumors offers a novel strategy for future treatment options.The effects of G-Rh2 at the PDK4 locus suggest G-Rh2 could be a potential therapeutic agent in clinical settings.In conclusion, our results confirmed the ability of G-Rh2 to target at the PDK4 locus, providing molecular evidence for altered tumor biometabolic behavior.
(
See figure on previous page.)Fig. 1 G-Rh2 inhibited tumor cell proliferation and promoted apoptosis in vitro.(A) Chemical structure of G-Rh2.(B, C) Viability of human A549 (B) and PC9 (C) cells assessed by CCK8 assays after 24-hours treatment with different concentrations of G-Rh2.(
Fig. 2
Fig. 2 G-Rh2 inhibited tumor growth and promoted apoptosis in vivo.(A) Representative images of A549 tumor formation in the xenografts of nude mice.(B) Measurement of mouse weights after G-Rh2 treatment for 14 days.(C) Summary of tumor volumes recorded every 2 days.Data are expressed as mean ± SD. (D) Determination of tumor weights 14 days after intraperitoneal injection of G-Rh2.Data are expressed as mean ± SD. (E-H) Immunohistochemical analysis showing the percentages of Ki67-positive and TUNEL-positive cells.Data are expressed as mean ± SD. **P < 0.01, ***P < 0.001, ****P < 0.001 vs. the control group
(
See figure on previous page.)Fig. 5 G-Rh2 inhibited tumor aerobic glycolysis by regulating HIF1-α/PDK4.(A-B) Volcano plots showing 176 upregulated and 586 downregulated differentially expressed proteins in A549 cells after G-Rh2 treatment compared to the control group,.(C) COG analysis histogram after G-Rh2 treatment, compared to the control group.(D) Clustered heat map showing a portion of A549 cells after G-Rh2 treatment, compared to the control group.(E) KEGG pathway enrichment of proteomics in A549 cells.(F) Subcellular location analysis after G-Rh2 treatment, compared to the control group.(G) GO pathway enrichment of proteomics in A549 cells.(H) Molecular docking modeling predicting the binding of G-Rh2 to HIF-1α.(I) Western blot detection of PDK4 protein expression.(J)Western blot detection of HIF1-α and VEGF protein expression.(K) RT-qPCR detection of downstream genes of HIF1-α.(L) Glucose intake in A549 cells after G-Rh2 treatment.(M) Lactification in A549 cells after G-Rh2 treatment.(N) Western blot detection of glycolytic protein expression.(O, P) PET-CT detection of the concentration of 18 FDG in A549 tumors, revealing glucose intake after G-Rh2 treatment in xenografts of nude mice.(Q) γ-count data showing glucose intake of A549 tumors after G-Rh2 treatment in xenografts of nude mice.Data are expressed as mean ± SD.**P < 0.01, ***P < 0.001, ****P < 0.001 vs. the control group
Fig. 6
Fig. 6 G-Rh2 targeted inhibition of PDK4 expression promoted mitochondrial oxidative phosphorylation.(A) Western blot detection of PDK4 protein expression following G-Rh2 and DCA treatment.(B) Molecular docking modeling predicts the binding of G-Rh2 to PDK4.(C) ATP production in A549 cells after G-Rh2 and DCA treatment.(D) Glucose intake in A549 cells after G-Rh2 and DCA treatment.(E) Lactification in A549 cells after G-Rh2 and DCA treatment.(F) Pyruvic acid production in A549 cells after G-Rh2 and DCA treatment.(G) PDH activity in A549 cells after G-Rh2 and DCA treatment.(H-J) Fluorescence detection and flow cytometry analysis of ROS levels in A549 cells after G-Rh2 and DCA treatment.Data are expressed as mean ± SD. (K, L) ECAR measured with Seahorse XF in A549 cells and A549 cells treated with G-Rh2 and DCA.2-DG, 2-deoxyglucose.(M, N) OCR measured with Seahorse XF in A549 cells and A549 cells treated with G-Rh2 and DCA.FCCP, carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone.(O, P) ECAR measured with Seahorse XF in A549 and PDK4 overexpressing A549 cells treated with G-Rh2.(Q, R) OCR measured with Seahorse XF in A549 cells and PDK4 overexpressing A549 cells treated with G-Rh2.**P < 0.01, ***P < 0.001, ****P < 0.001 vs. the control group | 2024-04-28T06:17:03.419Z | 2024-04-26T00:00:00.000 | {
"year": 2024,
"sha1": "df9922037733145386535cda2abd4bcd0b996fbf",
"oa_license": "CCBY",
"oa_url": null,
"oa_status": null,
"pdf_src": "PubMedCentral",
"pdf_hash": "0eb2a7d63f268efb11bdc8d81dbde248039f83c7",
"s2fieldsofstudy": [
"Medicine",
"Biology"
],
"extfieldsofstudy": [
"Medicine"
]
} |
11655989 | pes2o/s2orc | v3-fos-license | Transcriptomic and proteomic profiling of maize embryos exposed to camptothecin
Background Camptothecin is a plant alkaloid that specifically binds topoisomerase I, inhibiting its activity and inducing double stranded breaks in DNA, activating the cell responses to DNA damage and, in response to severe treatments, triggering cell death. Results Comparative transcriptomic and proteomic analyses of maize embryos that had been exposed to camptothecin were conducted. Under the conditions used in this study, camptothecin did not induce extensive degradation in the genomic DNA but induced the transcription of genes involved in DNA repair and repressed genes involved in cell division. Camptothecin also affected the accumulation of several proteins involved in the stress response and induced the activity of certain calcium-dependent nucleases. We also detected changes in the expression and accumulation of different genes and proteins involved in post-translational regulatory processes. Conclusions This study identified several genes and proteins that participate in DNA damage responses in plants. Some of them may be involved in general responses to stress, but others are candidate genes for specific involvement in DNA repair. Our results open a number of new avenues for researching and improving plant resistance to DNA injury.
Background
Maintenance of genome stability is of critical importance for all organisms. Genomic DNA is continuously subject to many types of damage resulting from endogenous factors (production of reactive oxygen species, stalled replication forks, etc.) or the action of exogenous agents (radiation, naturally occurring radioisotopes, chemical mutagens such as heavy metals, etc.) [1]. Doublestrand DNA breaks (DSBs) are one of the most serious forms of DNA damage, potentially causing chromosomal translocations and rearrangements [2]. In response to DSBs, cells initiate complex signalling pathways that activate DNA repair, cell-cycle arrest, and eventually cell death [3]. DSBs repair is mediated by two basic mechanisms: homologous recombination (HR) and non-homologous end joining (NHEJ) [4]. In HR, an intact copy of the damaged region (a sister chromatid, for example) acts as a template to repair the break. In NHEJ, DSBs are simply rejoined largely independently of the DNA sequence. Bacteria and yeast usually employ HR whereas mammals and plants usually use NHEJ.
In addition to the direct repair of DNA breaks, additional responses are activated during DNA-damage stress. For example, DNA damage in plant cells usually induces the accumulation of signal transduction intermediates such as nitric oxide, ROS or ethylene [5,6] and produces changes in the cytosolic-free Ca 2+ [7]. It also induces cell cycle arrest, the inhibition of DNA and RNA synthesis, and a rapid protein turnover via the proteasome [8,9]. Additional reported effects are a reduction in the photosynthesis-related proteins [10], the accumulation of protective proteins such as pathogenesis-related protein-1 [11], the accumulation of protecting pigments [12], an increase in the expression of senescence-and cell death-associated genes [13] and the activation of different cellular detoxification mechanisms [14]. The regulation of all these responses is complex and involves different levels of regulation, including the modulation of transcriptional activity [15], post-transcriptional mechanisms (RNA processing, RNA silencing, etc.) [16][17][18] and post-translational modifications (protein phosphorylation, ubiquitination, SUMOylation, etc.) [19]. These processes are based on signal transduction initiated by sensor proteins that recognise the damage in the DNA and activate the transducers, which send the signal to the effector proteins [20]. The network of transcriptional, post-transcriptional and posttranslational modifications ensures temporally and spatially appropriate patterns of stress-responses.
DNA topoisomerase I (TOPI) regulates the topological state of DNA by cleaving and re-joining one DNA strand and allowing DNA relaxation [21]. TOPI activity is essential in dividing cells to release the torsion created by the progression of DNA replication forks. The presence of active TOPI is essential for embryo development in Drosophila and mouse [22]. In plants, TOPI plays a similar basic role and, for example, the disruption of the two TOPI encoding genes in Arabidopsis thaliana is lethal [23]. Camptothecin (CPT) is a plant alkaloid that specifically binds TOPI, stabilising the complexes formed between DNA and TOPI [24]. The collisions between the trapped TOPI-CPT complexes and the replication fork during DNA replication produce DSBs which induce DNA damage responses [25]. In consequence, actively dividing cells are much more sensitive to CPT than non-dividing cells, a property that has been exploited in the treatment of cancer [24]. However, non-dividing cells are also sensitive to CPT as collisions of the RNA polymerase machinery with the TOPI-CPT complexes, although less frequent, are also able to produce DSBs [26]. CPT-mediated TOPI-DNA complexes can be degraded via the 26S proteasome pathway so, at low CPT concentrations, cells can survive [27]. However, in actively dividing cells the high number of collisions may exceed the capacity of the cells to eliminate TOPI-DNA complexes and the DNA repair capability of the cells and, under these circumstances, cell death is initiated. CPT has a similar effect on TOPI in plant and animals. For example, CPT inhibits, in vitro, the activity of TOPI extracted from maize immature embryos [28], produces the abortion of shoots and roots in Arabidopsis [23], and induces cell death in tomato cell cultures [29].
In this study, we profiled proteins and genes whose expression is changed in immature maize embryos as a consequence of the DNA damage produced by CPT. Immature embryos contain a high proportion of actively dividing cells and, in consequence, are particularly sensitive to CPT. The combination of microarray and twodimensional gel electrophoresis protein analysis allowed us to identify molecular events that are regulated during DNA repair responses in plants at different levels: transcriptional, post-transcriptional, translational and posttranslational. We identified candidate genes and proteins which may be specifically involved in the DNA repair responses.
Results
Camptothecin induces DNA damage responses in maize immature embryos but not an extensive cell death process Maize caryopses were collected 15 days after pollination and their dissected embryos incubated, in the dark, in culture medium with or without 50 μM camptothecin (CPT). After 8 days of culture, the germination rates of treated and untreated embryos were not significantly different (24% ± 5 in control and 20% ± 6 in treated embryos) and their morphological characteristics were similar.
CPT is a DNA damaging agent that induces DNA repair responses [24], while ribonucleotide reductase (RNR) is an enzyme that provides dNTPs for DNA repair, with RNR genes being induced in response to DNA damage [30]. In order to check if, under our conditions, CPT is able to induce DNA repair responses in maize embryos, we used a maize gene ZmRNR2 probe encoding the ribonucleotide reductase, in northern blot hybridization ( Figure 1A and 1B). There was a high level of accumulation of the ZmRNR2 mRNA after 3 days of CPT treatment and, although reduced, the accumulation was maintained after 8 days of treatment (Figure 1A). The induction of ZmRNR2 was much higher in the embryo axis than in the scutellum ( Figure 1B).
Nucleases are involved in DNA damage responses [31] and in cell death [32]. In plants, cell death-related nucleases have been classified according to their cationic cofactors, as Ca 2+ or Zn 2+ -dependent [33]. The ability of CPT to induce nuclease activities in maize embryos was tested using in-gel nuclease activity assays ( Figure 1C and 1D). An increase in the activity of a Ca 2+ -dependent nuclease of about 32 kDa was clearly evident after 3 days of CPT treatment using an assay buffer containing 1 mM CaCl 2 , being only slightly reduced after 8 days of treatment ( Figure 1C), and was higher in the embryo axis compared to scutellum ( Figure 1D). In contrast, no zinc-dependent nuclease activity was detected using 1, 2 or 5 mM Zn 2+ (results not shown).
The CPT-induced Ca 2+ -dependent nuclease could be involved in DNA repair but also in programmed cell death (PCD). PCD is usually characterised by internucleosomal genomic DNA fragmentation, producing, after gel electrophoresis, a characteristic DNA ladder pattern [34]. The results of electrophoresis of genomic DNA extracted from treated embryos was not significantly different to that observed with untreated embryos, showing a certain DNA ladder ( Figure 1E). The same analyses using DNA extracted separately from embryo axis and scutellum clearly show that the DNA ladder was only present in the scutellum sample ( Figure 1F). Degradation in genomic DNA extracted from scutellum has been previously observed in maize [34]. Cells in the scutellum close to the embryo axis undergo PCD as a normal part of seed development and this may explain the observed DNA ladder [35]. Exposure to 50 μM CPT did not, however, produce a significant change in the DNA integrity in the embryo axis or scutellum. This suggests that the CPT-induced Ca 2+ -dependent nuclease is involved in DNA repair and not in cell death.
In situ detection of fragmented DNA (TUNEL), as a sensitive technique to detect the initial steps of genomic DNA degradation, was used to analyse the effects of CPT on maize embryo DNA integrity ( Figure 2). In accordance with published data [35], untreated embryos only showed TUNEL positive nuclei in the scutellum, close to the embryo axis (Figures 2A and 2C). There was no increase in the number of positive nuclei in the scutellum of CPT treated embryos ( Figure 2B and 2D). The embryo axis of untreated embryos did not show any TUNEL staining ( Figure 2E). On the contrary, in CPT-treated embryos, some cells in the embryo axis showed TUNEL stained nuclei ( Figure 2F). However, the proportion of cells with stained nuclei was not high, which may explain why we did not observe extensive genomic DNA degradation in gel electrophoresis.
These results indicate that, under the conditions used here, CPT induced DNA repair responses in maize embryos but not an extensive cell death process. In-gel nuclease activity assay of total protein extracts (10 μg) of immature maize embryos treated with 50 μM CPT for three (E3D) and eight (E8D) days. The nuclease activity is detected as a non-stained halo in a polyacrylamide gel containing DNA stained with ethidium bromide. The deduced weight of the proteins with nuclease activity is indicated on the left (kDa). (D) Ingel nuclease activity assay of total protein extracts (10 μg) of dissected embryo axis (EA) and scutellum (SC) of immature maize embryos treated with 50 μM CPT for three days. The deduced weights of the proteins with nuclease activity are indicated on the left (kDa). (E) Integrity of nuclear DNA (4 μg) of immature maize embryos treated with 50 μM CPT for three (E3D) and eight (E8D) days, assayed by electrophoresis on 1.5% agarose gels. (F) Integrity of nuclear DNA (4 μg) of dissected embryo axis (EA) and scutellum (SC) of immature maize embryos treated with 50 μM CPT for three days, assayed by electrophoresis on 1.5% agarose gels.
Transcriptional responses to CPT-induced DNA damage
A global picture of the changes in gene expression produced during CPT treatment was obtained using the Affymetrix™ GeneChip Maize Genome Array. In this experiment, control and 3-day CPT-treated embryos were compared ( Figure 3). Ninety-three probe sets were found to have significantly increased or decreased signal in response to CPT, 39 up-regulated (Table 1) and 54 down-regulated ( Table 2). The probe set corresponding to the ZmRNR2 gene, previously used as a control for DNA damage response, was among the up-regulated genes. A quantitative real-time RT-PCR approach was used to validate the expression of 10 genes identified as differentially expressed in the microarray analysis, including 7 up-and 3 down-regulated genes ( Figure 4). Realtime PCR results were in very good agreement with the microarray data, although there were higher fold-changes using real time RT-PCR, which may be due to differences in the dynamic range and sensitivity of the two methods, as has been previously suggested [36].
The molecular roles of many of the altered genes remain unknown (31% of the up-regulated and 44% of the down-regulated). These genes may be involved in the control and/or execution of DNA damage responses ( Figure 5). DNA replication, recombination and repair (18%) and defense and stress responses (15%) were the two most abundant functional categories among the upregulated genes. Among down-regulated genes, the two most abundantly represented categories were signal transduction and gene expression (22%) and cell growth and division (17%). The functional category of DNA replication, recombination and repair was significantly more represented among the induced genes while the cell growth and division category was significantly more represented among the repressed genes ( Figure 5).
CPT treatment induced the expression of genes involved in DNA repair and DNA damage responses as, for example: -Two subunits of the ribonucleotide reductase: involved in the DNA repair processes [30].
-RAD51: encodes a protein required for meiosis and HR repair [37]. Maize mutants in two RAD51 maize genes are hypersensitive to radiation [38]. The Arabidopsis gene AtRAD51a is transcriptionally up-regulated by DSB-inducing agents and seems to be required for HR repair after bleomycin treatment [39].
-Rpa2: encodes a protein that is part of a heterotrimeric protein complex that specifically binds singlestranded DNA (ssDNA) and plays multiple roles in DNA metabolism, including DNA repair and recombination [40]. RPA genes are transcriptionally induced in Aspergillus nidulans exposed to CPT [41].
-TBPIP1: encodes a protein involved in chromosome pairing and segregation [42]. In humans, TBPIP1 enhances the strand exchange mediated by RAD51 [43]. In Arabidopsis, the TBPIP1 gene is transcriptionally induced by DNA damage [44].
-XRI-1: encodes a protein essential for meiosis and that plays a role during HR in Arabidopsis [45]. This gene is highly and rapidly transcriptionally induced by X-ray radiation and is also highly induced by other DSBs-inducer agents [44]. The encoded protein is probably part of the meiotic recombination complex MND1/ AHP2, which collaborates with RAD51 in the DNA strand invasion during recombination [46].
-Acetyltransferase, GNAT family protein: some yeast GNAT family members are involved in DSBs repair [47].
-Rph16: encodes a protein similar to RAD16 and is involved in the nucleotide excision repair of UV damage [48].
CPT treatment repressed the expression of genes involved in cell cycle, cell division and cell growth ( Table 2). For example: -Three cyclins: IaZm, IIZm and IIIZm.
-Shugosin-1: encodes a protein involved in the maintenance of centromeric cohesion of sister chromatids during meiosis and mitosis. Depletion of the human Sgo1 gene produces mitotic cell cycle arrest [49].
-TPX2: encodes a protein necessary for mitotic fuse formation in vertebrates [50]. The inhibition of the Arabidopsis TPX2 gene blocks mitosis [51].
-Knolle: encodes a syntaxin-like protein that acts during cytokinesis vesicle fusion and mediates cell-plate formation [52]. Knolle expression is repressed by gamma radiation in Arabidopsis [15].
-Patellin-5: patellins are involved in vesicle trafficking events. The Arabidopsis patellin PATL1 has been associated with the formation of the cell-plate during cytokinesis [53].
-Microtubule-associated protein RP/EB family member 3: encodes a protein that binds to the end of the microtubules and is important in maintaining the structure of the mitotic spindle [55].
-Growth regulating factor 8-like: encodes a protein involved in leaf and cotyledon growth in Arabidopsis [56]. -Rough sheath1: encodes a protein involved in cell differentiation [57].
-Frizzy-like protein/WD-repeat cell cycle regulatory protein: encodes a protein similar to the tomato CCS52B that probably is involved in cell-cycle control during mitosis [58].
Alterations in maize embryo proteome in response to CPT
Equal amounts of total protein extracted from control and from CPT-treated maize embryos were fractionated using 2-D gel electrophoresis ( Figure 6A and 6B). At least three-fold increase/decrease and t-test p < 0.05 were used as the criteria to select differentially accumulated polypeptides. In response to CPT treatment, 455 spots showed quantitative or qualitative (presence/ absence) variations between the two gels, with the intensity decreasing in 169 and increasing in 286. Some examples of up-or down-accumulated spots are shown in figure 6C. Forty-three of the spots with significant differential expression on gels were chosen for identification by MS/MS mass spectrometry. Interpretable MS/MS spectra were obtained for 31 spots. The location of these in the gels is shown in Figures 6A and 6B.
The identified proteins belong to a variety of functional categories (Table 3). For example, CPT alters the accumulation of two enzymes involved in glycolytic metabolism: glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and triosephosphate isomerase 1. Interestingly, in human neuronal cells, CPT also produces changes in the accumulation of GAPDH [59]. In plants, the accumulation of both proteins has been described in response to different types of stress [60][61][62][63][64].
Some of the identified proteins are involved in antioxidant responses. Antioxidant activity protects against ROS accumulation, which can be produced by a variety of stresses, including DNA damage [65]. In mammals, CPT induces the accumulation of antioxidant enzymes in the nucleus [66]. The accumulation of two proteins involved in pathogenesis responses, PR1 and Betv1, was observed in response to CPT. They are also induced by abiotic stresses such as heavy metals [67] and UV radiation [68]. The accumulation of at least two 26s proteasome regulatory subunits is altered in response to CPT, one increased and the other reduced. Interestingly, CPT-TOPI-DNA complexes may be degraded by the ubiquitin-dependent pathway [27].
We observed changes in the accumulation of several proteins involved in RNA metabolism or RNA binding proteins: -RraA: an RNaseE inhibitor which may be involved in the degradosome complex in E. coli [69,70]. Regulation of RraA by DNA damage stress could be related to changes in the regulation of RNA homeostasis.
-DEAD-family RNA helicase: DEAD-RNA helicases act in RNA metabolism promoting either RNA synthesis or decay [71]. Some have been associated with abiotic stress [72].
-Glycine-rich RNA-binding protein 8: some glycinerich RBPs in Arabidopsis (GR-RBPs) are significantly induced by cold, drought and salinity, whereas others are repressed by other sources of stresses [73].
We also observed an increase in the accumulation of two spots corresponding to eukaryotic elongation factors. Experiments in yeast and mammals demonstrate that translation initiation factor 5A (eIF5A) is actually involved in mRNA nucleus-cytoplasm export and not translation, specifically regulating genes involved in cell growth and proliferation, and in cell death [75]. In Arabidopsis, AteIF5A/AtFBR12 (At1g26630) promotes PCD associated with the hypersensitive pathogen response [76], and AteIEF5A-1 (At1g13950) has been associated with PCD during xylogenesis [77]. Thus, regulation of eiF5A by CPT suggests it is involved in cell cycle and PCD regulation.
Lack of correlation between CPT-induced changes in protein abundance and changes in mRNA accumulation
This study provided data on the most differentially expressed genes in control and CPT-treated embryos, and the most differentially accumulated proteins, allowing us to compare the datasets. The genes encoding 24 of the 31 identified proteins are represented in the microarray, but there was no significant change in expression in response to CPT (Table 4). This was confirmed by northern blot hybridizations using probes corresponding to nine of these genes, with no significant differences in the hybridization intensities observed (Figure 7).
Discussion
Our aim was to identify new elements involved in cellular responses to genomic damage in plants, using CPT as a toxic agent and applying transcriptomic and proteomic approaches to identify the genes, proteins and cellular mechanisms involved. We identified a series of genes and proteins whose expression/accumulation significantly change in response to CPT, although the identified genes do not correspond to the identified proteins. These differences may be a consequence of the different sensitivity of the methods. Moreover, the level of protein accumulation does not necessarily agree with the level of mRNA expression. This incongruent expression between mRNAs and proteins has been observed by other groups, in other species and experimental conditions [78][79][80] and is most likely a result of the biology of gene expression which includes various levels of regulation during protein synthesis: post-transcriptional, translational, and post-translational. Thus, integrated analysis of both mRNAs and proteins is crucial to gain further insights into complex biological systems.
The basic mechanism of action for CPT has been well-studied and characterised in animal cells [24]. CPT generates replication-mediated DSBs in DNA which in turn induce DNA repair, cell cycle arrest and, under certain circumstances, cell death. Under our conditions, CPT did not induce extensive cell death in maize embryos, as demonstrated by TUNEL staining which only appeared in some cells in the embryo axis after CPT-treatment. At the developmental stage analyzed here, cells in the scutellum divide at a very limited rate, but cells in the embryo axis divide rapidly. This difference may explain the higher sensitivity to CPT of the cells in the embryo axis.
Two basic mechanisms of DSBs DNA repair have been described: homologous recombination and non-homologous end joining [4]. Our transcriptomic analysis identified the induction of some genes already known to be involved in DNA repair. Interestingly, most of them are involved in the HR repair pathway, suggesting that this is the main mechanism for DSBs repair in maize embryos, at least in response to CPT. CPT also produces an increase of a 32 kDa calcium-dependent nuclease activity. However, this nuclease is unlikely to be involved in the extensive fragmentation of the genomic DNA observed in different cell death processes as extensive DNA fragmentation was not observed. Nucleases are also involved in most DNA repair mechanisms, including HR [81]. These data suggests that the 32 kDa nuclease activity observed may be involved in the DNA repair process.
CPT induces reversible or permanent cell-cycle arrest in G2-M phase in human and other cells [82] and produces major alterations in the expression of cell-cycle regulatory genes [83]. We found that CPT reduces the expression of several mitosis-related genes. In addition, we observed a reduction in the accumulation of the histone H2B involved in the structure of chromatin, and changes in the accumulation of two eukaryotic translation initiation factors which seem to also be involved in the cell-cycle process [84]. These results suggest that, in maize embryos, one of the cellular responses to CPT is the arrest of cell division.
In addition to more specific processes, DNA damage induces general stress mechanisms in maize embryos. For example, we observed changes in the expression and accumulation of proteins involved in ROS processing (glutathione S-transferase, Class III peroxidase precursor, chloroplast Cu/Zn superoxide dismutase, cytosolic ascorbate peroxidase), enzymes involved in glycolic metabolism (glyceraldehyde-3-phosphate dehydrogenase, triosephosphate isomerase 1) and in pathogen responses (pathogenesis-related protein1 and Bet v I allergen). Pathogen resistance is increased after DNA damage induction, indicating a cross-link in DNA damage (and maybe other abiotic stresses) and defense responses [85].
An increasing number of studies combining proteomics and transcriptomics clearly demonstrate that mRNA and protein accumulation are not always correlated [86][87][88]. For instance, in yeast 73% of the variance in protein abundance is explained by the translation mechanism and only 27% due to variations in mRNA concentration [89]. Protein abundance is influenced by several factors at the post-transcriptional, translational, and post-translational levels. For example, there is a time lag between transcription and translation in which introns are excised and the transcripts are moved from the nucleus to the cytoplasm, and translation rates may be influenced by ribosome, tRNA and amino acids availability, codon usage or accessory protein binding association [90]. In addition, protein abundance is also influenced by post-translational processes such as glycosylation, phosphorylation and proteolytic processing.
Our proteome analysis indicated differences in the abundance (up and down) of the encoded proteins of 24 genes whose mRNA levels do not significantly change in response to CPT (Table 4). It is possible that CPT induces the transcription of some genes only during the first hours of treatment, and after three days of treatment the mRNA levels are similar to the control but the abundance of the encoded protein is higher. Differences in the translation rate may also explain the lack of correlation. In animals, post-transcriptional regulation of gene expression during the stress response means specific stress-induced transcripts receive the highest transcriptional priority [91]. Interestingly, some of the spots identified in the proteomic analysis correspond to proteins associated with RNA metabolism and RNA binding proteins, and may be involved in the regulation of mRNA translation. Table 3. (C) Portions of 2-D gels showing spots (arrows) that were differentially abundant between untreated (CON) and treated (CPT) embryos. Many post-translational processes affect the position of a protein in 2D gels such that the protein appears as differentially accumulated in a proteomic analysis. We have identified changes in genes and proteins involved in protein modification and post-translational regulation. For example, the accumulation of at least two 26s proteasome regulatory subunits is altered in response to CPT and the expression of the proteasome inhibitor-like protein PI31 is increased. Ubiquitin/proteasomemediated protein degradation plays a central role in the regulation of several aspects of plant development and stress responses [92] and our data indicate that it may also be involved in regulating DNA damage responses. In fact, there are evidences that CPT-TOPI-DNA complexes may be degraded by the ubiquitin-dependent pathway in mammals and yeast [27,93]. Our data suggest that a similar situation may occur in plants. Moreover, the expression of embryonic flower 2 is repressed, a gene encoding a protein homologous to Drosophila Polycomb genes which mediate the epigenetic control of homeotic gene expression [94].
The role of several of the genes identified in the transcriptomic analysis is unknown. These genes may play a role in DNA damage detection and repair mechanisms, especially those genes that are only induced in response to genomic damage and not in response to other types of stress. Unfortunately the data currently available in maize does not allow us to determine which of them are specifically induced by DNA damage, but many of the maize identified genes have clear homologues in Arabidopsis (Table 1 and 2). Microarray analyses in Arabidopsis have been used to study the effects of several abiotic stresses, including two DNA damage agents, bleomycin [4] and gamma radiation [15]. Examining microarray databases [95] we identified eight Arabidopsis genes homologous to maize CPT-induced genes and exclusively induced by DNA damage: At5g02220, of unknown function; At1g13330, encoding TBP-1 tat binding protein; At5g48720, encoding an X-ray induced gene required for post-meiotic stages of pollen development and for male and female meiosis; At3g27060 and At2g21790, encoding the ribonucleotide reductase (RNR) small and large subunit, respectively; At5g20850, encoding AtRAD51; and two genes, At5g18270 and At3g04060, encoding NAC transcriptions factors. NAC proteins constitute one of the largest families of plantspecific transcription factors, and the family is present in a wide range of land plants [96]. These two NAC
Conclusions
The integration of microarray and proteomic analyses provides new data on DNA damage responses in plants. This is a complex process involving DNA repair and arrest of cell-cycle, but also general stress responses. Post-translational processing and the regulation of mRNA translation seem to have an important role in DNA damage responses.
Plant material and treatments
Maize (Zea mays L cv W64A pure inbred line) was grown under controlled conditions (16 h light, 28°C). Immature embryos (15 days after pollination) were extracted in sterile conditions and placed on MS plates (4.4% (w/v) Murashige and Skoog medium, 0.8% (w/v) Gelrite) supplemented (or not) with 50 μM camptothecin (Sigma-Aldrich) and maintained in a growth chamber at 26°C in darkness.
Histological analysis
Embryos were collected, fixed in ethanol-formaldehydeacetic acid (80:3.5:5) for 1 h at room temperature, followed by 1 week at 4°C, and then stored in 70% ethanol at 4°C. Fixed samples were embedded in paraplast, de-waxed with Histo-Clear II (National Diagnostic, UK), re-hydrated in an ethanol series and equilibrated in 0.02 M citric acid-0.16 M Na 2 HPO 4 , pH 7.0. TUNEL assays were done using the In Situ Cell Death Detection kit (Roche) according to the supplier's protocol for difficult tissues. In negative controls, the TdT enzyme was omitted, and the positive controls were treated with DNase I for 10 min. Experiments were repeated three times.
RNA extraction and quantification
Total RNA was isolated from frozen samples using the lithium chloride method. DNase digestion of contaminating DNA in the RNA samples was done using RNase-Free DNaseI. Final RNA purification was performed using the RNeasy Mini Kit (Qiagen) according to standard protocols. RNA was quantified with a Nano-Drop ND-100 spectrophotometer (NanoDrop Technologies). RNA quality was assessed with a 2100 Bioanalyzer from Agilent Technologies.
Affymetrix GeneChip hybridization/Microarray analysis
Gene expression was analyzed using the Affymetrix GeneChip ® Maize Genome Array, which contains probe sets to interrogate 13,339 genes, performing four independent biological replicates. cDNA synthesis, probe labeling, array hybridization and data analysis were as described by Bannenberg and col. [97], in the Genomics Service of the Centro Nacional de Biotecnologia (CNB-CSIC, Madrid). Raw data and normalised data were deposited at the ArrayExpress data library (http://www. ebi.ac.uk/arrayexpress/) under accession number E-MEXP-2702. Differential expression was considered following the p < 0.05 and 2.0 fold change as the criteria of significance. Functional categories of the genes were determined based on Gene Ontology data. We used the Fisher's Exact Test (p ≤ 0.05) and ANOVA (p ≤ 0.01) to determine the significant differences in the functional categories among up-and down-regulated genes.
Real time quantitative RT-PCR
To validate the expression changes found in the microarray experiments, transcript levels of the ten selected genes were quantified by the ABI Prism 7700 (Applied Biosystems, Foster City, CA, U.S.A) as described by Mascarell-Creus and col. [98]. The oligonucleotides chosen to amplify the selected genes were designed using the Primer Express Software (Applied Biosystems) and are listed in table 5. The actin gene was used as the internal control.
In-gel digestion of proteins and MS and MS/MS spectra Proteins were in-gel digested with trypsin and tryptic peptides were extracted and analyzed by MALDI-TOF/ MS (4700 Proteomics Analyzer, Applied Biosystems) or LC-ESI-QTOF (Q-TOF Global, Micromass-Waters) mass spectrometers in the Proteomics Platform (PCB) of the University of Barcelona as previously described [100]. The oligonucleotides were designed using the Primer Express software (Applied Biosystems). | 2017-08-03T01:38:35.973Z | 2011-05-19T00:00:00.000 | {
"year": 2011,
"sha1": "fe24c36501fd8b73ed2cd663096187468ecb72d6",
"oa_license": "CCBY",
"oa_url": "https://bmcplantbiol.biomedcentral.com/track/pdf/10.1186/1471-2229-11-91",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "f66c65363e1d901f2f209bee04870f4008fbfd7e",
"s2fieldsofstudy": [
"Biology",
"Environmental Science"
],
"extfieldsofstudy": [
"Medicine",
"Biology"
]
} |
10296693 | pes2o/s2orc | v3-fos-license | Infiltrative Hepatocellular Carcinoma With Portal Vein Tumor Thrombosis Treated With a Single High-Dose Y90 Radioembolization and Subsequent Liver Transplantation Without a Recurrence
Background Infiltrative hepatocellular carcinoma with macrovascular invasion is a relatively rare presentation and usually fatal disease. Methods Both patients exceeded Milan and University of California-San Francisco (UCSF) criteria, and per Barcelona Clinic Liver Cancer group guidelines, they were enrolled in a prospective open-label radioembolization phase II trial that gave them optimized lobar doses of Yttrium-90 as solely the first-line therapy without concomitant or additional pharmacological or locoregional therapies. Results Three months after radioembolization, the patients demonstrated no residual viable disease on surveillance imaging. The patients were then followed up with serial imaging for 2 years in 3-month intervals, without documenting recurrence or extrahepatic disease. Finally, both patients underwent transplantation and after more than 20 months of imaging surveillance, no locoregional or systemic recurrence have been observed. Conclusions We present, to our knowledge, the first 2 reports of transplantation after successfully downstaging infiltrative disease with portal vein tumoral thrombosis, which traditionally poses as a relative contraindication for resection or transplantation.
Background.Infiltrative hepatocellular carcinoma with macrovascular invasion is a relatively rare presentation and usually fatal disease.Methods.Both patients exceeded Milan and University of California-San Francisco (UCSF) criteria, and per Barcelona Clinic Liver Cancer group guidelines, they were enrolled in a prospective open-label radioembolization phase II trial that gave them optimized lobar doses of Yttrium-90 as solely the first-line therapy without concomitant or additional pharmacological or locoregional therapies.Results.Three months after radioembolization, the patients demonstrated no residual viable disease on surveillance imaging.The patients were then followed up with serial imaging for 2 years in 3-month intervals, without documenting recurrence or extrahepatic disease.Finally, both patients underwent transplantation and after more than 20 months of imaging surveillance, no locoregional or systemic recurrence have been observed.Conclusions.We present, to our knowledge, the first 2 reports of transplantation after successfully downstaging infiltrative disease with portal vein tumoral thrombosis, which traditionally poses as a relative contraindication for resection or transplantation.
(Transplantation Direct 2017;3: e206; doi: 10.1097/TXD.0000000000000707.Published online 18 August, 2017.)I nfiltrative hepatocellular carcinoma (HCC) is a relatively rare liver carcinoma subtype accounting for an estimated 7-20% of all diagnosed HCC cases.Compared with the more common nodular subtype, infiltrative HCC lesions usually progress aggressively and are associated with a worse prognosis.Morphologically, this subtype has been characterized by an ill-defined, diffuse and cirrhotomimetic phenotype which may be limited to 1 liver segment but also spread throughout the entire liver parenchyma. 1,2On magnetic resonance imaging (MRI), this variant usually manifests as a geographic area of increased signal on T2-weighted images with variable degrees of arterial enhancement, portal venous phase washout, and portal vein tumor thrombosis (PVTT). 1,37][8] However, several studies investigated the safety and efficacy of locoregional therapies such as, for example, Yttrium-90 (Y 90 ) radioembolization and concluded that this may be a safe and viable treatment option for these patients. 9,10e herein report on 2 patients diagnosed with infiltrative HCC with portal vein tumor thrombus PVTT (American Joint Committee on Cancer [AJCC] stage IIIB) and were therefore, beyond criteria for transplantation.Per Barcelona Clinic Liver Cancer group guidelines, they were enrolled in a prospective open-label radioembolization phase II trial and underwent orthotopic liver transplantation (OLT) after downstaging the disease and exhibiting complete response to therapy using a single high-dose lobar radioembolization followed by a 2-year observation period without concomitant or additional pharmacological or locoregional therapies.
Both underwent hepatology evaluation and screening MRI, which demonstrated a 7.3 Â 10.3 cm HCC with infiltrative features predominantly involving segments 6 and 7 with minimal involvement of segments 5 and 8 in case 1; and a 4.5 Â 8.3 cm HCC with infiltrative features predominantly involving segments 5 and 8 with minimal involvement of segment 6 and 7 in case 2. In addition, there was thrombosis of the right anterior and posterior portal veins, with imaging features of tumoral thrombosis present in both patients' tumors (Figures 1-3).An additional chest computer tomography performed in both patients was unremarkable.
Because of the advanced initial tumor stage of both patients with T3bN0M0 (stage IIIb) per the modified AJCC/ Union for International Cancer Control staging system, the tumors were deemed nonresectable and noneligible for liver transplantation at a multidisciplinary tumor board meeting, and the decision to offer locoregional therapy was made.After interventional oncology consultation was performed, and per Barcelona Clinic Liver Cancer guidelines, the patients were enrolled in a prospective open-label radioembolization phase II trial immediately after presentation without undergoing any prior systemic or concomitant therapies.
After analyzing the pertinent imaging in conjunction with the nuclear medicine staff, a determination was made on the liver volume to be treated (case 1 = 900mL, case 2 = 1350 mL) as well as an estimated involvement of the total liver volume (case 1 = 40%, case 2 = 20%).Technecium-99 m macroaggregated albumin hepatic shunt study was then performed demonstrating the pulmonary shunt fraction (case 1 = 12.3%, case 2 = 14.3%).The starting dose was calculated to deliver 120 Gy to the tumors in both case 1 and case 2. The estimated lung dose was calculated (case 1 = 15.9Gy, case 2 = 27.8Gy).Based on these calculations, a dose was ordered (case 1 = 69.5 mCi [2.57GBq], case 2 = 106.1 mCi [3.93 GBq]) and subsequently, the patients' right hepatic artery was selectively catheterized delivering an approximated dose of Y 90 microspheres (case 1 = 63.99 mCi, case
RESULTS
After radioembolization, the patients demonstrated no residual viable tumor by modified Response Evaluation Criteria in Solid Tumors and European Association for the Study of the Liver criteria (Figures 1-3).The AFP values dropped significantly 3 months post-radioembolization.Case 1 dropped to an AFP of 84 ng/mL, and case 2 dropped to 127 ng/mL.The patients were then followed clinically for 2 years (case 1) and 1.5 years (case 2), with MRI examinations performed every 3 months.During that period, no disease recurrence or extrahepatic disease progression was observed.In both cases, the right liver lobe became atrophic, and there was early response to therapy by modified Response Evaluation Criteria in Solid Tumors and European Association for the Study of the Liver criteria.
After these periods, the patients were both listed with a Model for End-Stage Liver Disease score of 22 exception points (as per the Organ Procurement and Transplantation Network allocation system in force in 2013-2014) and rapidly underwent OLT at our institution under the presumption of no tumor viability per imaging findings as well as no disease progression during the observation period.Both patients underwent OLT with standard technique undergoing no complications.Pathological explant (Figure 4) examination concurs with serial MR examinations demonstrating an atrophic right lobe with complete histopathologic tumor necrosis (Figures 5).Additionally, sphere deposition in the tumor bed (Figure 6A) and in the portal vein from case 2 (Figure 6B) was observed.After 24 months (case 1) and 20 months (case 2) of imaging surveillance after OLT, no locoregional or systemic recurrence have been observed (Figure 7).
DISCUSSION
To the best of our knowledge, these 2 cases represent the first reports of OLT after successful downstaging of AJCC stage IIIB disease that included PVTT and extended diseasefree survival with complete response to therapy using a single high-dose lobar radioembolization followed by a 2-year observation period (1.5 years in case 2) without concomitant or additional pharmacological or locoregional therapies.
Infiltrative HCC with PVTT is a rare presentation and typically yields poorer prognosis than the focal/nodular subtype and lower survival rates. 1,3,11Reported survivals yield 75.4% and 46.0% at 1 and 3 years in patients with focal/nodular HCC in comparison to 33.3% and 13.6% survival at 1 and 3 years in infiltrative HCC. 11Another study demonstrated a 4.0-month median overall survival (OS) and survival rates at 3 months, 6 months, and 1 year of 63%, 30%, and 8%, respectively. 64][5]8 Sorafenib has been shown to increase survival with reported median OS rates of 7.5 months in patients with infiltrated HCC in comparison to those who did not receive tumor therapy (3 months).However, sorafenib still cannot be deemed a curative therapy option because there have been no reports of effective downstaging and subsequent OLT. 6,7t should be noted that sorafenib has not been evaluated specifically in infiltrative HCC with PVTT, so no conclusion can be made regarding its use in this patient population.
Transarterial chemoembolization (TACE) has been previously studied in patients with infiltrative HCC demonstrating mixed results.One study showed no benefit in treating infiltrative HCC with chemoembolization, reporting increased morbidity and mortality with decreased OS. 12 However, evidence does show TACE as well tolerated, extending median survival to 12 months as compared with 3 months with supportive measures.Better results have been achieved in patients with initial AFP levels less than 400 ng/mL and bilirubin levels less than 2.0 mg/dL. 1 Others have also reported TACE to be a safe and effective treatment method in this patient population. 13esearch has shown that the use of transarterial radioembolization for the treatment of HCC with PVTT is an effective therapy that yields a higher response rate (50-70%) 14,15 with a median OS of 13.0 months 14 and an improved median progression-free survival (11.0 months) as compared with similar patients treated with sorafenib (4.1 months). 15,16Recently, Y 90 radioembolization therapy for infiltrative HCC with portal venous thrombosis demonstrated a median OS of 13 months and a median time to progression of 9 months. 10ere, ECOG performance status and Child-Pugh class have been deemed to be independent predictors of time to progression and in addition to hepatobiliary toxicity (grade 2 or higher), they were seen to be predictors of OS. 10 Downstaging of infiltrative HCC to within Milan criteria has been controversial as some consider the criteria to be too stringent.][19] Despite the controversy, tumor downstaging to meet Milan criteria for OLT in selected patients has been associated with excellent post-transplant outcomes. 17Cohort studies comparing the use of TACE and Y 90 radioembolization for downstaging showed that radioembolization was more successful than TACE (58% vs 31%). 20Although these data are encouraging, little is known about downstaging stage IIIB disease with radioembolization and the long-term outcomes.
These 2 presented cases support attempting to downstage infiltrative disease with PVTT using Y 90 radioembolization before OLT as a potential curative treatment option.Possible rationale to explain the results in these 2 cases might be the degree of cirrhosis with relatively well-preserved liver function (Child-Pugh A5) as well as good preprocedural performance status, both conditions known to be favorable for response to treatment with radioembolization.These cases may also be optimal examples of advanced HCC with PVTT to be treated with high-dose radioembolization because both were unilobar, similar in size, and with ipsilateral branch PVTT.Limitation of this report includes the absence of preradioembolization pathology because all imaging and laboratory workup corroborated the diagnosis of infiltrative HCC before radioembolization therapy.
CONCLUSIONS
These 2 presented cases support that a single treatment of high-dose lobar radioembolization as the first-line therapy in patients with unilobar infiltrative HCC with ipsilateral PVTT could be a safe and efficacious treatment strategy to successfully downstage the tumor for successful liver transplantation and potential cure of the disease.
FIGURE 1 .
FIGURE 1. Case 1, Imaging findings.Pretreatment axial contrast enhanced T1-weighted MRI image on arterial phase (A) demonstrates infiltrative HCC in a geographic area of arterial enhancement involving the posterior segments of the right lobe that includes the right portal vein (arrow), which represents tumor thrombus.Post-Y 90 Bremsstrahlung fused SPECT-CT scan (B) demonstrates increased Y 90 tracer activity within the treated infiltrative HCC.On the same slices and on the 1-month postprocedural scan (C), note the reduction of enhancement of both the infiltrative mass.After 1 year, note the significant liver atrophy of the right liver lobe and the sustained complete treatment response without a recurrence (D).SPECT-CT, single-photon emission computed tomography.
FIGURE 2 .
FIGURE 2. Case 2, Imaging findings.Pretreatment axial contrast enhanced T1-weighted MRI image on arterial phase (A) demonstrates a geographic area of arterial enhancement (arrows) involving predominantly the anterior segments of the right lobe.On the same slice and on the 3-month postprocedural scan (B), note the absence of enhancement of the infiltrative mass (arrow).Note the significant resultant liver atrophy.
FIGURE 3 .
FIGURE 3. Case 2, Imaging findings.Postprocedural Bremsstrahlung fused SPECT-CTscan (A), demonstrates increased tracer activity within the expected location of the infiltrative lesion and within the right portal vein.On the correlative T1-weighted MRI image on portal venous phase on the preprocedural scan (B), the tumor thrombosis involves the right portal vein (arrow).In the correlative T1-weighted MRI image on portal venous phase on the 3-month postprocedural scan (c), note the partial recanalization of the portal vein (arrow).
FIGURE 4 .
FIGURE 4. Gross pathologic specimens.The surface of both livers is diffusely cirrhotic (A) case 1; (B) case 2 with significant atrophy of the right liver lobe (arrows) and hypertrophy of the left liver lobe (star).
2 = 105.2mCi).The administered radiation dose was equivalent to 112.1 Gy to the liver and 14.84 Gy to the lungs in case 1.The dose was equivalent to 193.5 Gy to the liver and 26.8 Gy to the lungs in case 2.
FIGURE 5 .
FIGURE 5. Histological examination of the treated lesions.Case 1 slides (A) Hematoxylin-eosin stain (10Â) demonstrates complete pathologic necrosis (arrowhead) of the tumor bed with glass beads in the necrotic parenchyma (arrow) as well as a partially recanalized segmental portal vein (star).Case 2 slides (B) Hematoxylin-eosin stain (10Â) demonstrates atrophic fibrotic parenchyma (arrowhead) with complete pathologic necrosis of the tumor bed with glass beads (arrow). | 2018-04-03T01:01:55.032Z | 2017-08-18T00:00:00.000 | {
"year": 2017,
"sha1": "ca91a993bebe5cd5c6002242c322a0288843cdc3",
"oa_license": "CCBYNCND",
"oa_url": "https://doi.org/10.1097/txd.0000000000000707",
"oa_status": "GOLD",
"pdf_src": "ScienceParsePlus",
"pdf_hash": "2b5e76f9ed5a28510d6494226049cc2d95d0757b",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
97054152 | pes2o/s2orc | v3-fos-license | Thermal conductivity reduction in graphene with silicon impurity
We present a molecular dynamics investigation on the thermal conductivity of silicon-doped graphene and the resulting change in phonon properties. A significant reduction in the thermal conductivity is observed in the presence of silicon impurity even at a small concentration of silicon atoms. Conductivity values continued to decrease with an increase in silicon concentration. The increase in the scattering rate, which is measured by the reduction or broadening of the peaks of the van Hove singularities, is the most significant factor contributing to the large conductivity reduction. An analysis with scattering time models shows that the mass displaced by the silicon impurity plays a significant role in reducing the conductivity, especially at a moderate concentration. The non-mass effect, which comes from the change of the sp2 C–C bonds to the sp3 Si–C bonds, is less strong or comparable with the mass change effect. For high impurity concentrations, the shape of the graphene is severely distorted and the irregularity of the ripples increases, which could contribute to the reduction in conductivity.
Abstract We present a molecular dynamics investigation on the thermal conductivity of silicon-doped graphene and the resulting change in phonon properties. A significant reduction in the thermal conductivity is observed in the presence of silicon impurity even at a small concentration of silicon atoms. Conductivity values continued to decrease with an increase in silicon concentration. The increase in the scattering rate, which is measured by the reduction or broadening of the peaks of the van Hove singularities, is the most significant factor contributing to the large conductivity reduction. An analysis with scattering time models shows that the mass displaced by the silicon impurity plays a significant role in reducing the conductivity, especially at a moderate concentration. The nonmass effect, which comes from the change of the sp 2 C-C bonds to the sp 3 Si-C bonds, is less strong or comparable with the mass change effect. For high impurity concentrations, the shape of the graphene is severely distorted and the irregularity of the ripples increases, which could contribute to the reduction in conductivity. 1
Introduction
Since graphene, a two-dimensional graphite, was fabricated by exfoliation from graphite [1], there has been significant research interest in graphene synthesis and graphene properties [2][3][4]. Numerous works have reported the remarkable properties of graphene, including extraordinary high electron mobility [5], thermal conductivity [6], stiffness and strength [7], as well as large surface area to volume ratio [8], and an unusual electronic structure [9]. However, some of the properties of graphene are not well suited to practical applications, which lead to intense research into new methods of functionalization or property tuning [10]. In applications where the thermal requirements are critical, it is important to regulate the change of the thermal properties. For graphene to be a high-performance thermoelectric (TE) candidate, it is desirable to lower its thermal conductivity as much as possible, which would increase its TE figure of merit [11]. However, it is often the case that the thermal transport properties of graphene are altered as a side effect of tuning other properties. Graphene's excellent heat transfer properties make it an ideal material for addressing challenges associated with the cooling loads of electric devices [6,12]. Compared to carbon nanotubes (CNTs), graphene has lower contact resistance and superior thermal heat transfer properties. However, for practical applications, the zero bandgap of pristine graphene must be addressed with proper functionalization. The insertion of impurity or dopant atoms is the most effective method for addressing the bandgap concerns [10,13]. On the other hand, it can negatively impact thermal conduction properties. Therefore, controlling the impact of impurity on heat transfer [14][15][16] becomes important for advancing graphene-based technology.
An understanding of the origin of graphene's high thermal conductivity is challenging because the models of two-dimensional and nanoscale physics are limited. The contribution of phonon scattering to the thermal conductivity is dominant over the electronic contribution [12]. Sadeghi et al. [17] have reviewed current challenges surrounding the thermal transport properties. As regards the heat transport of tuning graphene, investigations of isotope [18], nitrogen [16], and vacancy [19] have been performed.
Recently, an elaborate model that corrects the Klemens' model, which was originally developed for general bulk materials and subsequently adopted for graphene, has been suggested [20].
The silicon-doped graphene has been recently synthesized [21], and density functional theory (DFT) calculation had shown the possibility of enhanced on/off efficiency [13]. It was proposed that bonding stability allows the addition of different atoms bound to Si atoms [22]. In the present study, single-layered graphene with silicon as impurity has been investigated using non-equilibrium MD (NEMD) and equilibrium MD (EMD) to demonstrate the reduction in graphene's thermal conductivity and the relevant phonon behavior, respectively. MD has already been adopted in investigating the thermal conduction of graphene relating to rectification [14,23] and thermal property management by controlling chirality, shape, and type of defect [14][15][16][24][25][26]. To evaluate the scattering contributions in the Klemens' model from the MD results, Callaway type model calculations are conducted. Based on these calculations, additional consideration is given on the long-length behavior of the thermal conductivity.
Numerical methods
In classical MD, Newton's second law is applied to all atoms in the calculation domain, and the force F i , acting on an atom i, is expressed as the gradient of potential energy according to where r i is the position vector of the atom i, and the intermolecular potential / ij (r) is given by Tersoff potential model [27,28]. The potential is used for C-C, Si-Si, and C-Si interactions for silicon-doped graphene in a single form with multiple sets of parameters. All of the parameters used in the current study follow values previously tested and recommended in some applications by Tersoff [27]. The current programs have been validated through tests that reproduce previous results [29,30]. The Newtonian equation is discretized in time using the Verlet algorithm with a fixed time step of Dt = 0.2 fs for both NEMD and EMD, and both methods have previously been systemically applied to silicon bonds with the Tersoff potential [31]. The procedure used by Wei et al. [32] is employed for NEMD calculations, where the half-sized domain, which is used for time-saving purposes, is the only difference from general NEMD calculations. The atomic arrangement of carbon and silicon is depicted in Fig. 1. The length of the model sheets ranges from 10.65 to 63.9 nm, and the width of all the simulated sheets is fixed at 3.9 nm. The silicon atoms are regularly distributed in convenience for the stability of the calculation and the consistent comparison of results for different sheet lengths. This artificial arrangement is not expected to have a significant impact on the results compared with using a random distribution. Five impurity concentrations, ranging between 0 and 10 %, are simulated. Periodic and free boundary conditions are applied in the y and z directions, respectively. The general procedures of NEMD (as described in [31]) are used in the simulations. The atoms start from their initial conditions and undergo the equilibration processes at a temperature of 300 K with 1,000,000 time steps. Then, the sheets are exposed to heat flux by adding and removing kinetic energy at the source and sink.
The thermal conductivity is calculated using Fourier's law from the resulting temperature gradient and the applied heat flux as where heat flux, q, is applied as De/A not exceeding 23.0 9 10-5 eV/nm 2 for each time step. The same concentrations are used for both the EMD and NEMD calculations. The DOS calculated using EMD reveals the change of phonon property. The domain for the EMD calculations has x and y dimensions of 10.65 and 11.81 nm, respectively, and the periodic boundary conditions are extended to the x direction to simulate an infinite system. The Fourier transform over velocity autocorrelation function is given by which is considered to be equivalent to the DOS [33].
Results and discussion
After a sufficient number of time steps to adapt to the heat flux, steady temperature profiles are established. Figure 2 presents a temperature profile obtained at a 63.9-nm-long pure graphene sheet subjected to a thermal energy of 15.3 9 10 -5 eV/nm 2 . With the exception of the regions near the sink and source, a linear profile is generated, from which the thermal conductivity can be calculated. The thermal conductivity values calculated for all conditions are shown in Fig. 3. The thermal conductivity increases with increasing sheet length, as expected because the phonon mean path depends on the sheet dimension. This can simply be explained using kinetic theory, where c, v p , l, and s represent specific heat, phonon velocity, mean free path, and scattering time, respectively. As the impurity level of the graphene increases, the conductivity decreases. Because of the extremely large computational cost, most of MD calculations employ an extrapolation scheme to estimate the bulk thermal conductivity based on smaller scale simulations, where Matthiessen's rule has been favored. In this case, where the free path of infinite size l ? and the free path confined by distance between sink and source L contribute to the effective free path of finite length l eff . According to the NEMD procedure, the effective conductivity approaches the bulk conductivity as 1/L approaches zero (i.e., at the infinite length). Thus, the inverse relationship shown in Fig. 3 facilitates the estimation of infinite conductivity k ? . For the pristine graphene sheet, the extrapolation gives k ? = 774 W/mK, which is in agreement with recent NEMD calculations on the same potential model [16,32,34]. The large deviation from linearity at the maximum concentration (10 % silicon atoms) is related to the resolution of conductivity, i.e., the conductivity value is so small that the inverse value appears large. The conductivity reduction ratios compared to pure infinite sheet, which is calculated from Eq. (5), are presented in Fig. 4. A drastic reduction in the thermal conductivity is observed. A 2 % impurity concentration results in more than an 80 % reduction, while a 10 % impurity concentration reduces the conductivity to \2 % of that in pure graphene. These reduction rates are larger than those obtained by simple mass substitution of isotopic atoms [18] or by nitrogen doping of graphene [16].
Using the NEMD model, an accurate fitting scheme is necessary to estimate the bulk thermal conductivity other than Eq. (5). There are three conjectures on the behavior of the thermal conductivity: (1) As in MD simulation, the bulk thermal conductivity is usually estimated by the Matthiessen's rule preferred in general bulk (three dimensional) materials , which assumes a finite k at infinite length [31,32]; (2) the k diverges in lower dimensions especially with logarithmic law (log dependence) for two dimensional lattices, which, in graphene, is known to be valid up to L \ *9 lm, and then diverges [35]; and (3) the k follows log dependence when L is smaller than about 100 lm, and for longer L, it converges when the thermal transport reaches the diffusive regime [36].
Barbarino et al. [37] calculated the thermal conductivity for the graphene whose length is in the range of 0.83-100 lm using approach-to-equilibrium MD (AEMD) with the reactive empirical bond order (REBO) potential, which can accommodate a large computational domain size with relatively shorter computing time. Their results support the third conjecture mentioned above.
We assumed Matthiessen's rule to estimate k ? , which is turn out to be 774 W/mK. This is much smaller value compared to experimentally observed literature value of about 3000 W/mK [38,39]. However, if we use the size dependence proposed by Barbarino et al. for fitting, instead of Matthiessen's rule, the thermal conductivity becomes about 2700 W/mK. The k/k (100 mm) at 60 nm is 0.1, and its value is 0.6 at about L = 3 mm where most experimental data are available; thus, 6 times of our k = 440 W/ mK at L = 60 nm yields about 2700 W/mK.
The DOS for the same configuration and concentration rate as in the NEMD cases is also obtained using EMD calculations. The calculated DOS of pure sheet is consistent with the dispersion curve obtained from Boltzmann transport calculations using the same original potential [40,41]. Figure 5 presents the dispersion curve obtained from Boltzmann transport calculations and the DOS obtained from the EMD. Most of the singular points in DOS shown in Fig. 5 correspond to the frequencies of the symmetric points in the dispersion curve. Each of six phonon branches causes a high DOS peak at the M points. The nonzero and higher DOS at zero frequency (C 1-3 point) is related to the quadratic dispersion relation of outof-plane phonon modes, which is a direct consequence of the two dimensionality of graphene [42]. At K 1-2 where the optical and acoustic modes are contacted, the valley between the two M peaks is formed. This is analogous to the electronic band structure, where a zero bandgap is formed by the contact.
The DOS for the pristine and impure graphenes is presented in Fig. 6. To allow the DOS calculated for the different impurity concentrations to be compared, the frequency, x, has been corrected to adjust for the mass changes that result from the replacement carbon with silicon atoms [15], using with subscripts 0 and d denoting the host and dopant fraction, respectively. Without the correction, the real confinement frequency is reduced from 73 GHz for the pure case to 70 GHz at 10 % impurity concentration. The most remarkable change in the DOS distribution is that the increased concentration rate reduces or broadens all of the M point peaks of the pure sheet. In Adamyan's calculations [43] for sheets doped with aluminum atoms, which have a mass comparable to silicon atoms, the peaks are lower near the van Hove singularities. Generally, peak broadening indicates that the scattering rates increase [33,44] for all phonons.
The contraction of the frequency regime, as shown in Eq. (6), suggests that the effect of the phonon velocity being reduced by the impurities can account for only a limited or small portion of the full conductivity reduction [see Eq. (4)]. Hence, the distribution of DOS indicates that the primary source of the reduction is the increased scattering rate. Both calculation models predict that the additional peaks arise at the same frequencies, K 1-2 and K 4-5 . These additional peaks accompany the splitting of the intersected modes, which is analogous to the bandgap opening in the doped graphene [43,45]. It has been suggested that the shapes of the additional peaks are the result of optical phonons that occupy a narrow band with relatively low phonon group velocity [45]. This conversion of peaks into modes that do not contribute to the heat transport could enhance the reduction in thermal conductivity. At the maximum concentration in Fig. 6e, the DOS is significantly different from the lower concentration models in that the peaks disappear and the gap separating the inplane acoustic modes from the other modes is bridged because of the growth of the additional distribution.
The cause of imperfect scattering due to defects in the crystalline structure can be analyzed with a few constituent elements in the Rayleigh scattering model, as shown by Klemens [46]. When adjusted to two-dimensional material, such as graphene [47], the phonon relaxation time is given by where c d is the concentration of defect atom. The elemental parameters S 1 , S 2 , and S 3 are related to differences in mass, velocity, and radial spacing, respectively [46]. For example, S 1 accounts for only mass changes, such as in isotopes. While the first term is the contribution of pure mass change, the latter two are commonly derived from the change of interatomic force. The scattering rate is larger Thermal conductivity reduction in graphene with silicon impurity 1197 than the simple mass change when the change of the interatomic force involves different atomic species. Hence, the conductivity of graphene doped with heavy atoms such as silicon is lower than that of enriched isotope or of graphene doped with lighter atoms. Figure 7 shows the instantaneous atomic arrangements. Even in the pure sheet, graphene has intrinsic ripples on the order of one nanometer, which is comparable to that of real suspended graphene [48]. The level of perturbation is getting higher than the natural ripple (Fig. 7a) (for pure graphene) with increasing silicon concentration. This perturbation occasionally results in irregular twisting of the smooth ripples. Non-planar shapes such as ripples and areas of high curvature can cause reduced thermal conductivity [49]; reductions by folding [50] and curvature [16] have been reported. While the large-scale irregular ripples occur when the impurity concentration is relatively high, as shown in Fig. 7, another source of strain induced by the change of bond type is present at even low impurity concentrations. Figure 8 depicts tetrahedra comprised of four atoms. The tetrahedra comprised of four carbon atoms, i.e., those that do not include a Si impurity atom, are planar and show a propensity for sp 2 bonding, as shown in Fig. 8a. The average bond angle and length are 120.0°and 0.146 nm, respectively, which are typical sp 2 values. When the silicon impurity is at the center of the tetrahedra, with carbon neighbors, as shown in Fig. 8b, tetrahedra are no longer within two-dimensional plane, as the silicon atoms rise or sink away from the plane formed by the three adjacent carbon atoms. The average bond angle and length for every Si-C bond for all of the impure cases are 102.3°and 0.176 nm, respectively. These values indicate that the silicon impurity converts the two-dimensional sp 2 bonds into sp 3 bonds. The sp 3 bond induces an increase in out-ofplane motion, and its disturbance to in-plane heat transfer is well known [51]. It is acknowledged that the addition of sp 3 bonds reduces the in-plane thermal conductivity of graphene [52,53]. Even a simple substitution of the bond type without the insertion of the impurity [52] can lead to a large reduction, of a similar order of magnitude to that which results from nitrogen doping [16], when sp 2 orbitals are preserved in N-C bonds of the N-doped graphene. The thermal conductivity reduction due to either simple sp 3 substitution or of lighter atomic substitution is commonly about 77 % at 5 % impurity concentration as in the references [16,52]. The effect of sp 3 replacement on pure strain (S 3 ) is expected to be negligible for interstitial silicon compared with large mass substitution of a vacancy [54]. The second term, S 2 , which indirectly reflects changes in the intermolecular angle and distance due to the strain, for example, in Tersoff model, is expected to have a contribution comparable to the mass effect.
Callaway type calculations are performed using the same constituent models and calculation parameters as Alofi and Srivastav [47]. In Fig. 9, the solid symbol indicates the results obtained from the calculation and open symbol denotes the data obtained by the MD simulation. The conductivity values from the MD calculations are rescaled by half for comparison. The length dependence shown in Fig. 9a for pure graphene agrees with the recent reports that the thermal conductivity increases logarithmically but converges at a large (millimeter) scale [37]. The parameter S in the Klemens' imperfection model is The mass parameter S 1 that results from the mass change is fixed because of the known mass of silicon, but non-mass terms, S 23 = S 2 ? S 3 , are difficult to measure and are therefore calculated by fitting the thermal conductivity, which is obtained from the present MD calculation. In particular, at a moderate concentration of 0.63 %, the value of S 23 = 0.15 gives the best fit to the thermal conductivity [a curve with open circle symbol in Fig. 9b]. While the mass parameter S 1 is 0.38 at 0.63 % of silicon atoms, the S 23 parameter is related to the change of intermolecular force and its resultant strain. The value S 23 = 0.15 is used for all of the impurity ratios. This choice of parameters is valid for the low concentrations, as shown in Fig. 9b, which demonstrates that the focused length scale is close to the range of MD calculation. At moderate concentrations, the mass effect (S 1 = 0.38) is larger than or comparable to non-mass effect (S 23 = 0.15). This is because the mass of the silicon atom is nearly two times larger than the mass of the host atom. In the Callaway type model calculation, the reduction ratios in long-length limit are increased compared with those for the small sizes. According to Eq. (5), the thermal conductivity at 0.63 % impurity is 28 % of the pristine graphene, as shown in Fig. 4; this ratio decreases to 22 % at 10 lm in the Callaway model. The same tendency is found for the isotopic defect; some MD calculations [26,55] based on the small size have reported to underestimate the reduction in comparison with experimental and theoretical values [18,36].
For higher concentrations, the value of S 23 = 0.15 fitted for moderate concentration does not sufficiently account for the reduction in the thermal conductivity. The predicted thermal conductivity is smaller than that predicted using MD, and the deficiency in S is not negligible. A new kind of scattering model is needed to describe the effect of the irregular ripple, as shown in Fig. 7c. The strain evaluated for the sp 3 bond follows the Klemens' scattering formula, as shown in Fig. 7b. In this case, every defect has the same strain rate and interatomic relation, so the defect density is proportional to the scattering rate, which the Klemens' model also implies. However, the ripple, which is not subject to a single point defect, requires an understanding of the principle of formation and the resultant effect on thermal transport.
Summary and conclusions
Molecular dynamics simulations are performed to investigate the effect of silicon impurity on the thermal conductivity of graphene sheets. The results obtained are in good agreement with previously published phonon property calculations. NEMD reveals a drastic reduction in thermal conductivity with even a small concentration of silicon impurity. The calculations show that only 0.63 % atomic replacement decreases the thermal conductivity to 30 % of that of pure graphene. Impurity concentrations of 3 % reduce the conductivity to less than one-tenth of the pure graphene conductivity. These results suggest that silicon impurity is more effective at reducing the conductivity than the isotopic dopant; experiments have shown that the thermal conductivity of graphene isotopically enriched with 1.1 % of 13 C is reduced to 63 % of pure graphene [18]. Further, our results suggest that silicon impurity is superior to nitrogen doping, which has been observed to reduce the thermal conductivity to 30 % of pure graphene at of 3 % nitrogen impurity [16]. For higher silicon concentrations, the present NEMD shows that the conductivity decreases to 1.7 % of the pure graphene conductivity at 10 % silicon. The reduction rate may increase when the logarithmic length dependence is applied, which has been recently supported by experiments and calculations.
The DOS obtained from EMD calculations shows the phonon properties expected in the presence of crystalline imperfectness. The peak broadening or reduction becomes more significant at higher concentration, which is related directly to the reduction in phonon heat transfer. The broadening implies an increase in the scattering rate, which results in a reduction in thermal conductivity. The increase in scattering by imperfection is shared with all of the acoustic and optical branches of phonons.
From the scattering time model analysis, it can be seen that the contribution of the mass element (S 1 ) of the silicon impurity plays a significant role in reducing the conductivity, especially at a moderate concentration. The nonmass effect comes from the change of the sp 2 C-C bonds to the sp 3 Si-C bonds. For high impurity concentrations, the shape of the graphene is severely distorted and the irregularity of the ripples increases, which could contribute to the reduction in conductivity. We propose that the ripples have a different scattering source than that which is normally modeled in Klemens' formula. | 2019-04-06T13:11:43.355Z | 2015-09-24T00:00:00.000 | {
"year": 2015,
"sha1": "f30994c0f00f6eee33d21505064e518e44d20827",
"oa_license": "CCBY",
"oa_url": "https://link.springer.com/content/pdf/10.1007/s00339-015-9489-1.pdf",
"oa_status": "HYBRID",
"pdf_src": "Adhoc",
"pdf_hash": "142cf143a9b9d88404a29db92a084fa1750f227d",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Materials Science"
]
} |
59943613 | pes2o/s2orc | v3-fos-license | Unique Patterns and Biogeochemical Relevance of Two-Component Sensing in Marine Bacteria
Marine microbes must manage variation in their chemical, physical, and biological surroundings. Because they directly link bacterial physiology to environmental changes, TCS systems are crucial to the bacterial cell. This study surveyed TCS systems in a large number of marine bacteria and identified key phylogenetic and lifestyle patterns in environmental sensing. We found evidence that, in comparison with bacteria as a whole, marine organisms have irregular TCS system constructs which might represent an adaptation specific to the marine environment. Additionally, we demonstrate the biogeochemical relevance of TCS systems by correlating the presence of the PMT9312_0717 response regulator protein to phosphate concentrations in the South Pacific. We highlight that despite their potential ecological and biogeochemical relevance, TCS systems have been understudied in the marine ecosystem. This report expands our understanding of the breadth of bacterial TCS systems and how marine bacteria have adapted to survive in their unique environment.
A bacterium's survival is dependent on its ability to respond to changes in its environment. This is especially true for marine microbes, which experience changes in nutrient availability, light, temperature, and community structure that can occur on time scales as short as hours (1). Two-component sensory (TCS) systems, which modulate gene expression on short time scales, are the most common sensory systems in prokaryotes (2). The organism's complement of two-component-system genes may thus reveal details about its lifestyle, ecological niche, and physiological complexity.
Canonically, two-component sensory systems are composed of a single histidine kinase and response regulator protein pair (Fig. 1). The histidine kinase contains a sensory domain that is activated by a specific stimulus, which may be a small molecule, nutrient, or physical property such as light or temperature. Upon activation, the histidine kinase is autophosphorylated on a conserved histidine residue, inducing conformational changes that enhance interactions with the response regulator protein (3). Through a highly specific protein-protein interaction, the phosphate moiety is transferred from the histidine kinase to a conserved aspartate residue on the response regulator (4). This typically stimulates binding of the response regulator to a DNA promoter region, resulting in transcription of genes in the downstream operon (5). A variation on this model is the hybrid histidine kinase, in which the kinase and response regulator are located on a single protein. An intermediary protein or set of proteins is involved in transmitting the signal, allowing multiple levels of control and a fine-tuned physiological response (3).
Two-component sensory systems represent a direct link between the environment and the physiology of the prokaryotic cell and, as such, may be important players in biogeochemical cycles. One well-studied example is the Pho system, composed of the histidine kinase PhoR and response regulator PhoB. The Pho system is common in marine bacteria and regulates genes that are involved in phosphate acquisition (6). In Prochlorococcus, activation of PhoR by low intracellular phosphate stimulates transcription of alkaline phosphatase. Alkaline phosphatase cleaves phosphate from organic matter, providing a source of phosphate that would otherwise be inaccessible (7). As the vector linking microbial physiology (i.e., the expression of alkaline phosphatase enzyme) to ocean chemistry (i.e., phosphate availability), the Pho system may be a key regulator of phosphorus cycling in the ocean. Other two-component sensory (TCS) systems may also play important roles in mediating microbe-environment interactions, though we do not yet understand their biogeochemical contexts.
In this study, we surveyed the TCS system genes of 328 diverse marine bacteria, identifying phylogenetic and lifestyle factors that correlate with greater numbers of sensory genes. We compared these marine bacteria to a curated reference collection of 1,152 bacterial genomes, most derived from the GEBA initiative (8). This allowed us to identify key differences in how TCS systems are structured in marine bacteria versus bacteria in general. To demonstrate the importance of TCS systems to marine biogeochemistry, we examined the distribution of a putative phosphate-sensing Prochlorococcus response regulator (PMT9312_0717) in metaproteomes from the tropical Pacific. We highlight gaps in our knowledge of marine TCS and emphasize the importance of TCS to our overall understanding of marine bacteria.
RESULTS
Lifestyle influences on TCS gene abundance. We examined the genomes of 328 diverse marine bacteria publicly available in the JGI IMG data warehouse (see Table S1 in the supplemental material). The data set emphasizes cultivable and oceanographically important organisms such as Prochlorococcus, Synechococcus, Pelagibacter, Alteromonas, and Roseobacter. In total, 15 phyla and 183 genera are represented from a variety of habitats, including coastal ecosystems, the open ocean, hydrothermal vent systems, host-associated environments, and marine sediments. All genomes were labeled high-quality finished or permanent drafts. TCS system genes were identified using highly conserved protein family domains for histidine kinases or response regulators as described in Materials and Methods.
We began by examining one half of the two-component system-the sensory histidine protein kinase (HPK). The number of histidine kinases in each genome ranges from 1 (Pelagibacter) to 174 (Desulfovibrio inopinatus DSM 10711) and is related to genome size; we found on average 0.902 histidine kinases (HPK) per 100 proteinencoding genes (Fig. 2). A K-means clustering analysis was performed on the proportion of the genome devoted to histidine kinases (expressed for human readability as the number of histidine kinases per 100 protein-encoding genes) versus genome size. This analysis revealed that the number of histidine kinases per 100 protein-encoding genes is dependent on both phylogeny and lifestyle. The first cluster contains oligotrophic picoplankton such as Prochlorococcus, Pelagibacter, and open ocean Synechococcus. These organisms have very small genomes and few histidine kinases per 100 FIG 1 Overview of two-component-system signaling in (A) a traditional histidine kinase-response regulator system and (B) a hybrid histidine kinase system. In panel A, the phosphorylation is transferred from the histidine kinase to the response regulator by a direct protein-protein interaction. In panel B, the phosphorylation is transferred to an internal receiver domain on the histidine kinase, then to one or more histidine phosphotransfer (Hpt) proteins, and finally to the terminal response regulator. Two-Component Sensing in Marine Bacteria protein-encoding genes. Cluster 2 contains Rhodobacter, Vibrio, coastal Synechococcus, and Alteromonas and Pseudoalteromonas genomes, which are larger and have more histidine kinases per 100 protein-encoding genes. Cluster 3 is composed mainly of Alteromonas and Pseudoalteromonas genomes that have genome sizes similar to those of cluster 2 organisms but more histidine kinases per 100 protein-encoding genes. Cluster 4 organisms have the greatest numbers of histidine kinases per 100 proteinencoding genes; many have specific lifestyle traits such as particle association, parasitism, and mat formation. A number are sulfate-reducing Deltaproteobacteria.
Marine organisms are often classified by their nutritional preferences as copiotrophs (adapted to high-nutrient conditions), oligotrophs (organisms adapted to low-nutrient conditions), or as organisms that fall between those classifications (9)(10)(11)(12). This provides a framework for understanding properties such as growth rate, cell size, and genome size (Table 1). We classified marine bacteria as copiotrophs or oligotrophs based on published isolation and laboratory growth conditions and found that the copiotrophs have significantly more histidine kinases per gene than the oligotrophs (P ϭ 3e Ϫ15 by Student's t test) (Fig. 3B).
Unusual patterns in marine TCS sensing genes. We compared the TCS system genes of marine bacteria with those of 1,152 reference bacteria (Table S2). The reference data set includes bacteria from diverse lineages and habitats, including terrestrial, freshwater, host-associated, and marine bacteria, and is representative of trends in the two-component systems of bacteria as a whole. The phylogenetic distribution of genomes in the reference data set is broad (see Fig. S1 in the supple- The number of histidine kinases per 100 protein-encoding genes in the genome for organisms unambiguously designated copiotrophs or oligotrophs. Error bars represent 95% confidence intervals of the average value within the copiotroph (n ϭ 74) and oligotroph (n ϭ 34) categories. The copiotrophs were shown to have significantly more histidine kinases per gene than the oligotrophs by a Student's t test (P ϭ 3e Ϫ15 ). mental material). An important caveat is that both the marine and reference data sets contained mainly cultured organisms and may not represent natural diversity (13). Based on the traditional understanding, the numbers of histidine kinase and response regulator genes are expected to be equal. Indeed, we found the average response regulator/histidine kinase (RR/HPK) ratio in the reference data set to be 0.99 (Fig. 4). The RR/HPK ratio was found to be slightly higher in the marine bacteria (1.03), suggesting that there are a small number of "extra" response regulators in the genomes. The difference between the marine and reference data sets is significant based on a one-way analysis of variance (ANOVA) test [P ϭ 0.009, F(372, 1151) ϭ 6.73].
To better understand the origin of the high RR/HPK ratio, we examined the locations of the TCS genes in a subselection of marine genomes (Fig. 5). Genomes were selected for their oceanographic relevance, quality, and representation in the literature. Typically, the TCS system histidine kinase and response regulator genes are located in the same operon (14). "Orphan" genes are defined here as genes that are more than four average gene lengths away from another TCS gene. These genes may participate in regulatory networks with other TCS systems (15).
Orphan TCS genes were identified in all seven of the genomes that we examined, consistent with the high RR/HPK ratios described above. For example, the Pelagibacter ubique HTCC1016 genome has three modular TCS systems plus one orphan HPK (a KipI family gene) and one orphan RR (RegB). Notably, KipI is known to participate in regulatory networks (15). Larger genomes have more orphan genes. For instance, Alteromonas sp. Alt199 has 26 orphan histidine kinases (30% of the histidine kinase genes) and 8 orphan response regulators (14% of the response regulator genes). Crocosphaera sp. WH8501 has 8 orphan histidine kinases (15%) and 9 orphan response regulators (17%). It is difficult to ascertain the function of orphan genes due to the lack of experimental evidence associated with them. These systems are ideal for future study.
Hybrid systems occur when the histidine kinase and the response regulator are located on a single protein; here they were identified as genes containing both a histidine kinase HAMP domain and a response regulator receiver domain. In marine bacteria, the proportion of hybrid histidine kinases relative to total histidine kinase content ranged from zero to 61% (Fig. 6). Marine bacteria have significantly more hybrid histidine kinases than reference bacteria (P ϭ 3e Ϫ10 by a Student's t test).
Patterns in Proteobacteria and Cyanobacteria. We examined the TCS systems of Proteobacteria and Cyanobacteria to explore differences among phylogenetically related organisms. Proteobacteria tend to have many histidine kinases (Fig. 7E). As before, Histidine kinase genes and response regulator genes that are within four genes of another TCS gene are represented as long blue and green lines, respectively. Orphan histidine kinases and response regulators are represented as short cyan and red lines, respectively. There are many orphan TCS genes in marine bacteria, including in oligotrophs such as Pelagibacter and Prochlorococcus.
FIG 6
Percentage of histidine kinases that are hybrids in marine (orange) and reference (blue) bacteria. The marine bacteria have a greater percentage of hybrid histidine kinases than the reference bacteria. Error bars represent a bootstrapped 95% confidence interval. The difference is statistically significant by a Student's t test (P ϭ 3e Ϫ10 ).
we found that the Deltaproteobacteria devote a large portion of their genome to histidine kinases (see Fig. 3). In the Cyanobacteria, we observed more variation, with Prochlorococcus and Synechococcus tending to have few histidine kinases and the nitrogen-fixing diazotrophs having more (Fig. 7C).
Proteobacteria tend to have RR/HPK ratios greater than 1 (1.03 on average), suggesting the possibility of the presence of extra response regulators in the genomes. The RR/HPK ratios of the Cyanobacteria are more variable than those of the Proteobacteria. Prochlorococcus and Synechococcus, for instance, have very high RR/HPK ratios (1.22 and 1.23, respectively), indicating that there are many extra response regulators in the genome, while Trichodesmium organisms have a low RR/HPK ratio (0.75), indicating the possibility of the presence of extra histidine kinases.
Biogeochemical relevance of two-component sensory systems. To investigate whether TCS proteins can be used as biomarkers of oceanographic processes, we examined the distribution of a putative Prochlorococcus MIT9312 phosphate-sensing response regulator (PMT9312_0717) in metaproteomes of the tropical Pacific (Fig. 8). Data acquisition and analysis were previously described by Saito et al. in November 2011 (16). We identified two unique peptides from this protein. Using the open-source Metatryp software package (17), we determined that one peptide is specific to Prochlorococcus and that the other is specific to the order Synechococcales. Thus, the distribution presented here can be thought of as a general picocyanobacteria signal (Fig. 8D).
PMT9312_0717 was confined to the upper euphotic zone and was less prevalent in phosphate-rich waters near the equator (Fig. 8A). Notably, protein abundance was inversely correlated to phosphate concentration; the relationship can be modeled with a simple power law (r 2 ϭ 0.64) ( Fig. 8B and C). Prochlorococcus is highly abundant throughout the transect, while the protein is not, suggesting that PMT9312_0717 is specifically involved in phosphate regulation (Fig. S2). However, as with many orphan The RR/HPK ratio of the Proteobacteria tends to be greater than 1. Note that Crocosphaera, Synechococcus, and Trichodesmium have more histidine kinases per 100 protein-encoding genes than Prochlorococcus, which is adapted to highly oligotrophic conditions. The picocyanobacteria Prochlorococcus and Synechococcus have particularly high RR/HPK ratios.
Two-Component Sensing in Marine Bacteria
TCS systems in marine bacteria, we were unable to identify a histidine kinase with corresponding relationships to phosphate concentrations. Thus, the exact regulatory function and mechanism of PMT9312_0717 remain a mystery, despite its probable biogeochemical relevance.
DISCUSSION
Two-component sensory systems allow bacteria to directly sense their internal and external surroundings and therefore play key roles in bacterial "intelligence" (18,19). Because TCS systems are the most common regulatory systems in bacteria, studying them can provide insight into how individual cells might interact with their surroundings. As a result, TCS systems are of potentially significant ecological and biogeochemical importance, and yet they have been little studied in this context (7,16). In this study, we surveyed the TCS systems of 328 diverse marine bacteria and described patterns in marine two-component sensing. We identified evidence of regulatory networking that may distinguish marine bacteria from other organisms and demonstrated that the abundance of a TCS system protein can be linked to oceanographic patterns. Lifestyle influences TCS gene abundance. As in prior surveys, we found that the number of histidine kinases in the genome is largely determined by genome size (2,4). However, the proportion of the genome encoding histidine kinases varies and is related to lifestyle and niche. The concept of copiotrophy versus oligotrophy is a common theme in microbial oceanography due to pronounced patterns in nutrient abundance versus scarcity, respectively, found in ocean environments (9)(10)(11)(12). It is similar to K versus r selection in that oligotrophs, which are adapted to low-nutrient conditions, grow significantly slower than copiotrophs, which are adapted to high-nutrient conditions. Oligotrophs have smaller genomes and cell sizes that allow the organism to thrive in resource-limited environments ( Table 1).
The small genomes of oligotrophs are thought to be the result of genome streamlining processes that are in turn driven by the need to conserve energetic and elemental resources (nitrogen and phosphorus in particular) (9,11,19). Our analysis suggests that the nutritional and energetic costs of maintaining TCS systems in oligotrophic environments outweigh the regulatory benefits. Lack of TCS genes has previously been observed in extremely oligotrophic organisms such as Pelagibacter, which may rely on simpler regulatory structures (such as one-component systems or riboswitches) which require fewer energetic and nutritional resources (20)(21)(22). The extent to which other marine oligotrophs may utilize simplified regulatory systems instead of or in addition to two-component systems is not yet known, but this survey found that the presence of few histidine kinases per gene is a hallmark of oligotrophy.
A previous effort to identify genomic signatures of oligotrophy found that histidine kinases were a weak indicator of copiotrophy versus oligotrophy (19). However, the study compared only two microbes, in contrast to the broader study across a larger number of diverse marine bacteria described here. The inability to adapt to rapid changes in nutrient availability and other environmental conditions may explain why oligotrophs are unable to survive in nutrient-rich environments (23). A recent comparison of coastal and open ocean strains of Synechococcus cyanobacteria supports this notion, where a coastal strain was found to have a more dynamic proteome response to iron scarcity than an oligotrophic strain (24).
In contrast to oligotrophs, certain organisms devote a comparatively large portion of their genome to histidine kinase genes. These bacteria often have specific traits such as mat formation that may require coordinated physiological alterations (25,26). Many are sulfate or nitrate reducers from deep sea sediments or hydrothermal vents, which might represent particularly dynamic environments. Two-component systems may be particularly important for redox regulation in these organisms (27).
Unique patterns in marine TCS systems: RR/HPK ratios, orphan genes, and hybrid systems. We identified unique trends in the TCS system genes of marine bacteria. Specifically, we found that marine bacteria tend to have (i) higher RR/HPK ratios, (ii) many orphan TCS genes, and (iii) more hybrid histidine kinases than bacteria examined in a reference data set. We discuss each of these observations in turn.
Marine bacteria have significantly higher response regulator:histidine kinase ratios (RR:HPK) than bacteria as a whole, suggesting that they have extra response regulator genes. By considering the RR/HPK ratio, we found that approximately 1 in 50 response regulator genes lacks a histidine kinase partner (Fig. 4). The origin of the extra response regulators is puzzling. Some, lacking a histidine kinase partner, may have no regulatory function and could have been the result of incomplete genetic innovations or horizontal gene transfers (HGT). However, this explanation is not consistent with the tendency toward genome streamlining in the nutrient-limited ocean environment (9,11). An intriguing alternative is that some of the extra response regulators might participate in regulatory networks in which multiple response regulators interact with a single histidine kinase. Such networks are a topic of increasing study and have been implicated in coordinating nutrient acquisition (specifically, acquisition of phosphate and iron), sporulation processes, stress response, and circadian rhythms (28-31).
Colocalization of TCS genes is thought to provide for concerted transcription of the sensory genes and better success in HGT (15,32). Marine bacteria seem to be an exception to this rule, having many orphan genes in their genomes (Fig. 5). Orphan TCS genes are present in even the most streamlined genomes (i.e., that of Pelagibacter ubique), suggesting that they play important biochemical roles. The nonmodularity of TCS systems in marine bacteria suggests that the genes are not acquired through HGT but are instead acquired through gene duplication and genetic remodeling (32). TCS systems created in this way are thought to be more likely to participate in regulatory cross talk than systems that are acquired through horizontal gene transfer (3). Indeed, orphan genes are often involved in essential regulatory networks in model organisms (33)(34)(35)(36). Alternatively, it is possible that in situations in which regulatory networks occur, the relationship between HPK and RR is not as specific as in normal twocomponent systems. This lack of specificity could allow nonmodular TCS genes to become fixed in the genome. Most of what we know about two-component systems is based on studies of modular systems from model organisms; additional studies on non-model organisms may thus reveal new mechanisms of acquisition and action of TCS genes.
In hybrid histidine kinases, the phosphorylation signal is relayed by one or more histidine phosphotransfer (Hpt) proteins before it reaches a terminal response regulator (Fig. 1). The added complexity of the phosphorelay is thought to provide multiple points of regulation, allowing fine-tuned physiological responses. For example, a hybrid histidine kinase regulates glycan utilization in Bacteroides thetaiotaomicron by integrating both intracellular metabolism and extracellular substrate signals (37). Hybrid histidine kinases are associated with physiological and behavioral complexity, being especially prevalent in higher eukaryotes (38). Marine bacteria such as Proteobacteria and Bacteroidetes can have many hybrid TCS systems ( Fig. 5; see also Table 1), concurrent with a tendency toward metabolic complexity and particle association, which may drive multicellular behaviors (28).
Together, the presence of extra response regulators, the nonmodularity of TCS systems, and the prevalence of hybrid TCS systems suggest increased regulatory complexity in marine bacteria. This may confer advantages in the ocean environment, where bacteria are often chronically nutrient limited. In the ocean, nutrient availability is determined by diffusion rates, resulting in covariance in the distribution of organic nitrogen, carbon, phosphorus, and trace nutrients, especially at the microscale (28,39). Nutrient colimitation has been demonstrated in a number of marine environments and may be more prevalent than originally thought (40)(41)(42). Marine organisms appear to have specific physiological responses to colimitation, for example, proteome restructuring and cell size decreases in iron and phosphate colimited Trichodesmium cells (43). Although the regulatory systems for colimitation in marine microbes have yet to be elucidated, a precedent for regulatory networking has been identified in nonmarine organisms such as Edwardsiella tarda, in which the Pho and Fur systems interact with one another (31). The irregularities in marine bacterial TCS genes suggest that similar regulatory networks may underpin concerted responses to multiple environmental perturbations.
Comparison of Proteobacteria and Cyanobacteria. Proteobacteria and Cyanobacteria are perhaps the most abundant and well-studied cells in the ocean. Comparing them demonstrates the impact of lifestyle traits on the number of histidine kinases in the genome. With the exception of the oligotrophic Pelagibacter species, the Proteobacteria have a larger proportion of genes encoding histidine kinases (0.95 per 100 protein-encoding genes on average) than the Cyanobacteria (0.57 per 100 proteinencoding genes on average) ( Fig. 7A and C). This may be related to their tendency toward copiotrophy, which is associated with large numbers of TCS system genes. Within the Cyanobacteria, nitrogen-fixing organisms have more histidine kinases. Diazotrophs are not subjected to the same genome streamlining pressure as other marine Cyanobacteria owing to the fact that they can access an unlimited supply of atmo-spheric nitrogen. However, the complexities of the diazotrophic lifestyle may necessitate a large number of regulatory genes. For instance, nitrogen fixation rates are known to respond to many environmental parameters such as iron nitrogen, phosphorus, dust, and light availability (44)(45)(46)(47)(48). Reflecting this, marine diazotrophs have two-component systems regulating nutrient availability, complex circadian rhythms, and redox state.
Evidence for regulatory networking is prevalent in both the Proteobacteria and Cyanobacteria (Fig. 7B and D). Most Proteobacteria and picocyanobacteria have elevated RR/HPK ratios that suggest the presence of branched regulatory networks in which one histidine kinase communicates with multiple response regulators. Branched regulatory networks could allow multiple operons to be affected by a single sensory input, with each chemical sensor triggering multiple downstream effects, providing greater metabolic flexibility and dynamism. This could provide for fine-tuned responses to multiple stimuli, such as nutrient colimitation. In nutrient-limited environments, regulatory networking may provide an advantage for cellular resource conservation. For instance, a single stimulus can trigger multiple physiological reactions without the need to express an entire two-component system for each operon. Consistently, we found that the oligotrophic genomes from organisms such as Pelagibacter and Prochlorococcus have especially high RR/HPK ratios (Table 1).
Notably, the diazotrophic organism Trichodesmium has low RR/HPK ratios, suggesting regulatory networking in which multiple histidine kinases interact with a single response regulator. This may allow integration of multiple environmental signals in regulating a single physiological response and may underpin extensive proteomic changes such as coordination of carbon and nitrogen fixation processes over the course of the diel cycle (49). However, identifying these networks is challenging because the majority of two-component-system genes in Trichodesmium (and many marine bacteria) have not been characterized.
Two-component systems as potential biogeochemical biomarkers. Having so far considered TCS genes, we next turned to the gene products and their relationship to oceanographic processes. We studied a Prochlorococcus/Synechococcus PhoB-like response regulator protein, PMT9312_0717, in the tropical Pacific Ocean. PMT9312_0717 is one of the most abundant TCS system proteins in this transect. It is an orphan, and its partner histidine kinase is not known. In the metaproteomics analysis, we found that the protein abundance is inversely correlated with inorganic phosphate concentration, particularly at concentrations below 1 M phosphate such as are encountered near the nutricline (Fig. 8). The abundance of Prochlorococcus cells is high throughout the transect, while that of PMT9312_0717 is not and instead follows trends in phosphate concentrations (see Fig. S2 in the supplemental material). This implies that the abundance of the protein is related to oceanographic processes, suggesting that the protein self-regulates its production. Increased abundance of TCS systems in response to low dissolved phosphorus concentrations has been previously observed for phosphateregulating systems (7,16). Measurement of levels of protein phosphorylation (i.e., the activity of the TCS system), while technically difficult, could provide additional information about the function of the protein.
In addition to this protein distribution, genomic evidence also suggests that PMT9312_0717 is involved in phosphate regulation. PMT9312_0717 is present in many Prochlorococcus species, including both high-and low-light ecotypes, as well as in Synechococcus species. It is similar to both the Prochlorococcus sp. 9312 phosphatesensing histidine kinase PhoB (37% identity) and the analogous Synechococcus sp. WH8102 PhoP protein (WP_025362545.1; 37% identity) (16) but is a distinct protein.
Because just a few point mutations can change the function of a response regulator, we hypothesized that this protein participates in phosphate regulation but in ways that differ from those seen with PhoB/PhoP. Corroborating its possible role in phosphate sensing, PMT9312_0717 is located near a DedA family alkaline phosphatase-related gene (e.g., PMT9312_0712) in multiple strains of Prochlorococcus. TCS systems have amino acids known as specificity residues that govern the histidine kinase-response regulator interaction (4). The specificity residues of the PMT9312_0717 and PhoB genes in Prochlorococcus sp. MIT 9312 are not shared, indicating that the PMT9312_0717 is unlikely to interact with the phosphate-sensing PhoR histidine kinase.
TCS systems underpin many of the physiological changes observed in laboratory and field perturbations of marine bacteria. They are involved in nutrient acquisition, detoxification, quorum sensing, and other topical themes in marine microbiology. However, our knowledge of these systems in marine species is limited. Sequence-based identification of TCS systems is difficult because histidine kinases and response regulators share highly conserved catalytic domains. Thus, while it is possible to identify TCS genes, identifying their physiological functions is tricky. Few two-component systems have been experimentally verified in marine species, despite the fact that just a few amino acid substitutions can drastically change physiological function (4).
The environmental drivers behind gain/loss of two-component sensory system genes and protein synthesis processes are not well understood. For instance, distribution of the phosphate-sensing phoR-phoB two-component genes in Prochlorococcus, while initially hypothesized to be correlated to phosphate availability, cannot be consistently linked to large-scale oceanographic patterns (7,50). An important consideration is that two-component sensory systems act on short time scales-seconds, minutes, or hours (51,52). The presence of a TCS system thus suggests the need to continuously monitor the stimulus. For nutrient-sensing regulators, it may be more accurate to suggest that distribution of the TCS system is related to varying (not necessarily chronically depleted) nutrient concentrations such as are found in surface waters. This is corroborated by our finding that the amount of PMT9312_0717 protein in the water column increases significantly near nutricline like concentrations. Given this circumstantial and yet consistent evidence, a detailed biochemical characterization of PMT9312_0717 is an intriguing topic for future study. However, on the basis of this and previous work, it is clear that TCS system protein abundances can contain valuable biogeochemical information (15).
Conclusions. Two-component sensory systems reveal characteristics of both individual cells and the ecosystems in which they live. In this way, they represent a unique opportunity to link microbial physiology to the environment. We know relatively little about TCS systems in marine bacteria, but it is clear that the distribution of TCS genes and proteins is dependent on both the traits of the organism and its surrounding environment. For instance, we found that oligotrophs have significantly fewer histidine kinases per gene than copiotrophs and that diazotrophy is associated with greater numbers of TCS system genes. Importantly, we found that marine microbes may have adapted unique ways to sense their environments using complex regulatory networks. Additional characterization of these networks may provide us with a greater appreciation for both the uniqueness of the ocean environment and the breadth of sensory systems used by prokaryotes. Detailed biochemical characterization of marine twocomponent systems is greatly needed and has great potential to advance our understanding of microbial life and its connections to global biogeochemical cycles.
MATERIALS AND METHODS
Marine and reference bacteria data sets. We acquired the genomes of 328 marine bacteria available in the JGI IMG data warehouse (see Table S1 in the supplemental material). All of the genomes were high-quality finished genomes or permanent drafts. We note that permanent draft genomes could be missing some RR or HPK genes due to incomplete assembly, but this is not expected to significantly affect results. The marine data set is phylogenetically diverse; the best represented phyla are the Cyanobacteria, Proteobacteria, Firmicutes, Bacteroidetes, and Actinobacteria (see Fig. S1 in the supplemental material). The bacteria were isolated from various habitats that included coastal ecosystems, the open ocean, hydrothermal vent systems, marine sediments, and microbial mats. Most of the genomes that we examined were from bacterial isolates in culture; as such, bacteria of particular oceanographic interest such as Prochlorococcus, as well as easily cultivated organisms such as Alteromonas, form a large fraction of the data set. While we attempted to capture maximal phylogenetic and habitat level variability, this data set does not necessarily represent the actual levels of bacterial diversity nor cell abundance in the ocean environment. Rather, the intention was to allow us to examine the marine bacteria for which we have good-quality genomic information at this time.
The reference bacteria data set was largely derived from the organisms associated with the GEBA-I initiative, which provides a collection of bacterial genomes spanning the breadth of known phylogenetic diversity (Table S2) (8). The GEBA-I database includes genomes of both bacteria and archaea; we used only the bacterial genomes here. We observed that Cyanobacteria are underrepresented in the GEBA-I data set relative to the marine data set described above. To facilitate comparisons between the marine and reference data sets, we included 180 high-quality, nonmarine Cyanobacteria genomes from the JGI IMG warehouse in the reference data set.
Identification of two-component-system genes. We identified TCS system genes by protein family (pfam) domain annotations (53). The pfam domain annotation was performed for each genome as part of their initial ingestion into the IMG warehouse. Comparisons were thus facilitated by the fact that all genomes were subjected to similar annotation analysis pipelines in the JGI IMG portal.
Histidine kinases are composed of two domains-an HATPase domain and a phosphoacceptor domain-which represent separate entries in the pfam domain database. We tested both the HATPase and phosphoacceptor domains by comparing the number of histidine kinases identified with the results of an extensively curated previously published survey of TCS genes in bacteria (18). We found that the HATPase domain was better conserved, identified more histidine kinases, and resulted in histidine kinase identifications comparable to those in the previous study. Thus, we used the HATPase domain for histidine kinase identification subsequently. The specific pfams used were pfam02518 (HATPase_c), pfam13581 (HATPase_c_2), pfam13589 (HATPase_c_3), pfam14501 (HAT-Pase_c_5), and pfam07536 (HWE_HK). The HATPase domain is also present in proteins such as DNA gyrase (pfam00204), HSP90 (pfam00183), and MutL (pfam13941); we removed genes containing these three pfams from the analysis.
Response regulators are composed of a phosphoreceiver domain (sometimes known as a REC domain) and an output domain. The output domain can vary significantly and determines the biological function of the response regulator (52,54). We used the protein family domain for the response regulator receiver (pfam00072) to identify response regulators in our data sets. When possible, we compared the results of this pfam-based identification of response regulators to that of a previously published survey and found that the results of the pfam-based analysis were comparable (18).
Hybrid histidine kinases occur when the response regulator phosphoacceptor domain is present on the same protein as the histidine kinase sensory and phosphoacceptor domains. We define them here as genes containing both a HATPase domain (pfam02518, pfam13581, pfam13589, pfam14501, or pfam07730) and the response regulator phosphoreceiver domain (pfam00072). Thus, the hybrid histidine kinases are present in both the histidine kinase data and response regulator data for a given genome.
Statistical and meta-analyses. All analyses were conducted in Python 3.6. The entire data analysis and visualization pipeline, including statistical analyses and machine learning algorithms, can be recreated by accessing the scripts at https://github.com/naheld/patterns_TCS_sensing_marine_bacteria. A fully executable cloud environment is provided courtesy of the Binder project (https://mybinder.org/). We used the Bokeh (https://bokeh.pydata.org/en/latest/) and seaborn (https://seaborn.pydata.org/) libraries to generate visualizations and perform statistical analyses. For ratio and statistical analyses, genomes containing 0 histidine kinases and/or 0 response regulators were excluded.
For K means clustering analyses, we counted the number of histidine kinases, including hybrid genes, and normalized the data by dividing by the number of protein-encoding genes. For human readability, we express this value as the number of histidine kinases per 100 genes. To eliminate the effect of scaling, we subjected the data to unit normalization. We then performed the clustering analysis to identify groups of genomes with similar signaling repertoires in relation to genome size. We used the Silhouette method implemented in the Scikit Learn Cluster module to select the optimal number of clusters (4) by maximizing the average Silhouette coefficient values (55). Clustering analysis was performed with the SciKit Learn K-means clustering algorithm.
Locations of TCS genes. The locations of TCS genes in the genomes of marine bacteria were examined by plotting the gene starting locations on a linearized depiction of the circular bacterial genome (a number line). The average gene size was calculated for each genome examined. A gene was considered to be an orphan if its start point was farther than four average gene lengths from another TCS gene.
PhoB protein distribution analysis. We examined the distribution of response regulator PMT9312_ 0717 in marine metaproteomes from the METZYME expedition in the tropical Pacific. Analysis of these metaproteomes has been previously described (16). Briefly, microbial biomass was collected with in situ particle collection pumps (McLane Labs) on the KM1128 METZYME research expedition. Proteins from the 0.2-uM to 3-uM size fraction were subjected to SDS detergent extraction and in-gel trypsin digestion as previously described (16). Peptides were analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS) using 1-dimensional chromatography on a Thermo Fusion Orbitrap mass spectrometer (16). Spectral counts for the protein were generated by mapping against 6 metagenomes sampled from the METZYME expedition. These were sequenced at JGI and assembled using metaSPAdes (56). Genes were predicted and annotated using the pipeline described previously by Dupont et al. (57). Peptide-tospectrum matches (PSMs) were identified by SEQUEST and restricted to a 10-ppm peptide mass threshold and a 99.0% protein probability threshold (16). Two unique peptides were identified for PMT9312_0717; we performed redundancy analysis using the openly available Metatryp software (17). Corresponding inorganic phosphate analyses were conducted by Joe Jennings at Oregon State University as previously described (58). | 2019-02-14T22:07:50.587Z | 2019-02-05T00:00:00.000 | {
"year": 2019,
"sha1": "4784d2a3b373df10532d2aa59eb7dce8920a8858",
"oa_license": "CCBY",
"oa_url": "https://msystems.asm.org/content/msys/4/1/e00317-18.full.pdf",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "4784d2a3b373df10532d2aa59eb7dce8920a8858",
"s2fieldsofstudy": [
"Environmental Science",
"Biology"
],
"extfieldsofstudy": [
"Biology",
"Medicine"
]
} |
53636754 | pes2o/s2orc | v3-fos-license | Why Leaders Fail in Introducing Values-Based Leadership ? An Elaboration of Feasible Steps , Challenges , and Suggestions for Practitioners
The recent debate on modernizing leadership and organizations has placed a strong emphasis on values and ethics. This article elaborates on the benefits and challenges in the integration of values into leadership actions, most notably in regards to appraisals of Values-Based Leadership (VBL). The article contributes to a more specified understanding of the interrelation of stages in the VBL and the pitfalls in each step. A key argument is that if the critical issues and challenges in the introduction of the VBL are not identified, this would lead to unintended consequences in organizations, such as insignificant value-statements, inappropriate use of values, and illegitimate leadership practice. Contrary to many developmental suggestions of more virtuous and ethical or values-based leadership qualities, this article proposes leadership actions for leaders to undertake an effective, sound, and sustainablevalues-based leadership practices.
Introduction
Over the last decade, much has been written about values, ethics, and integrity from a normative perspective, mostly suggesting what leaders should do and how leaders ought to behave.Both the positive and negative sides of many organizations and personal cases have been analyzed in order to develop a better understanding of ethics and values in leadership (Bass & Steidlmeier, 1999;Cagle & Baucus, 2006;Graber & Osborne Kilpatrick, 2008;Buchko, 2007;Mussig, 2003;Pruzan, 1998).Accordingly, while efficiency and profitability are viewed as a leader's primary objectives, there is a long held view that leaders also have responsibility for ensuring standards of moral and ethical conduct (Barnard, 1938;Cullen, Victor & Stephens, 1989;Resick, Hanges, Dickson & Mitchelson, 2006).
Especially the Values-Based Leadership (VBL) evoked the role and importance of ethics and values in leadership (De Hoog & Den Hartog, 2008;Treviño, Brown & Hartman, 2003;Brown & Treviño, 2006;Treviño, Weaver & Reynolds, 2006).As a result, our knowledge about advances and pitfalls on the leadership forms and patterns has deepened.However, despite the lively academic and practical analysis, the question of getting a grip of challenges related to the VBL has become more critical.Especially, how we can identify problems in putting organizational values into practice and how different acts in leadership effect, for example, on organizational behavior and decision-making, problem solving in organization, motivation, and sustainable organizational leadership.
In these increasingly complex and ambiguous times, leaders and personnel are often unsure how to act and what to prefer.Messick and Bazerman (1996, 9) underline that: "Executives today work in a moral minefield.At any moment, a seemingly innocuous decision can explode and harm not only the decision maker but also everyone in the neighborhood.".Thus, to overcome troubles caused by 'only ends justify the means' philosophy which powers results-driven cultures, values and ethics may offer more predictable, stable, and sustainable base for leadership.As Moorman and Grover (2009) note ethics and values can provide a certain 'warranty' on integrity and future prospects in organizations.Moreover, ethics and values in leadership are thought to be uniquely important because of the impact leaders have on the conduct of others and on organizational performance and effectiveness as well as sustainability of leadership.
To address the better understanding of the VBL, this literature based article debates VBL by presenting a simplified three-step sketch, asking what problems leaders face in each step and why.Especially, if our desire is to integrate values into business and leadership, then to understand VBL as sequent parts or steps might be one potential perspective in terms of successful and sustainable leadership of organizations.Moreover, each step in the VBL highlights challenges what leaders may consider if they pursue towards a sound and effective VBL.In the conclusion, recommendations to avoid unexpected effects and ways to overcome highlighted challenges will be presented.In part, giving details to the VBL and discussing the potential pitfalls of VBL, it helps enhancing focus on ethics education and further leadership research.Pruzan (1998) argues that we should actively introduce the notion of organizational and stakeholders' values into the leadership culture and develop a values-based perspective on management.Accordingly, Van Wart (1998, pp. 319) notes that the art of values management for practitioners has already become the leading skill necessary for private and public managers.
On Importance of Values-Based Leadership in Organizations
A reason why many leaders and academics has fascination on shared concept of values is because values are seen as the underlying attitudes and beliefs that help determine individual behavior, both personnel and leaders (Barnard, 1938, p. 279;Treviño & Brown, 2004, p. 75).In such way, values are a means of influencing behaviors without the need to resort to formal structures, systems, strategies, or control mechanisms.Values would also provide a means of directing the organization in a desired way without having to resort to authoritarianism (Buchko, 2007, p. 38) and using tight or confusing rules (Mills & Spencer 2005, p. 26).Moreover, the overall consensus seems to be that values are an important factor in the successful management of large organizations (e.g.Mintzberg et al., 2005;Hofstede, 2005) and in creating a competitive edge (Blanchard & O'Connor, 1997).
As such, introducing values into business and leadership is not a new thought.The concept of values as central to organizations and organized societies has a long history in the sociology of organizations, as well as, to understanding guiding principles of institutions, organizations, and individuals (Schwartz, 1992;Cummings & Worley, 2001).
Values-Based Leadership (VBL) refers to leadership based on foundational moral principles or values such as integrity, empowerment, and social responsibility (Reilly & Ehlinger, 2007, p. 246).Brown and Treviño (2006, pp. 595) reviewed recent research that systematically conceptualizes VBL constructs and defined the VBL as "the demonstration of normatively appropriate conduct through personal actions and interpersonal relationships, and the promotion of such conduct to followers through two-way communication, reinforcement, and decision-making.". Brown, Treviño and Harrison (2005) add to this definition that it includes making the leader a legitimate and credible role model with normatively appropriate virtues such as honesty, trustworthiness, fairness, and care.The definition, as they observe it, implies also that values-based leaders should not only draw attention to ethics and make it salient in the organizational environment, but that they should also engage stakeholders and subordinates interpersonally in the process.
VBL also concerns to carry out well-known managerial functions; it is setting (ethical) standards and goals, rewarding on achieving desired outcomes (ethical conduct) and penalizing those who do not follow the standards (Treviño, Brown & Hartman, 2003).Mussig (2003, pp. 73) argues that "values-driven leadership sets the function of the relationship as putting values into practice" and "the function of the leader may be to bring values to the relationship.".Others have noted the importance of shared values in creating a strong organization culture (Minztberg et al., 2005;Schein, 1985), motivating behavior by providing direction and emotional intensity to action (Schwartz, 1992), representing standards to judge and justify actions (Mills & Spencer, 2005), and socialization activities and individuals to organization and leadership (Grojean et al., 2004).
A part of the VBL actions is to guide organizational members towards goals, which benefit the organization, its members, stakeholders, and society (Kanungo, 2001).In such, VBL is positively related to satisfaction with the leaders, perceived leader effectiveness, follower's job dedication and willingness to report ethical violations (Brown, Treviño & Harrison, 2005).Some refer also the fact that introducing the VBL foster greater accountability, increased organization valuation, and gaining competitive edge, attracting and retaining staff and investors, and enhancing the organization's reputation within the corporate world (Buckley et al., 2001;Pruzan, 1998).In this vein, VBL is also criticized, because sometimes for instance, top executives might see ethics as "good business" in terms of enhanced image, reputation, and a source of competitive advantage (Buckley et al., 2001).Some studies of a values-based dimension of leadership have been embedded within the transformational, transactional and charismatic leadership domains (Bass & Steidlmeier, 1999;Bass & Avolio, 2000;Brown, Treviño & Harrison, 2005).The relationships are not clear-cut, and the VBL likely use both transformational and transactional leadership approaches to influence followers' behavior (Treviño, Brown & Hartman, 2003).Even though ambiguity of links, VBL tends to include similar inspirational motivation component that transformational leadership has (Bass & Avolio, 2000).Treviño, Brown and Hartman (2003) found that VBL entails a transactional component that involves setting standards and expectations of ethical conduct for followers, hold their subordinates accountable using the rewards and punishment systems that are available (also Resick, Hanges, Dickson & Micthelson, 2006).
Sketching Steps for Values-Based Leadership and It's Unintended Outcomes
As discussed above, there is evident need for the Values-Based Leadership (VBL).However, it is an ambiguous and challenging construct.To understand better the practice and execution of VBL, three simplified steps of VBL (Figure 1) are taken into consideration.The sketch presented in Figure 1 comes with combination of the existing models of Rest (1986), Treviño (1986), Jones and Ryan (1997) and Wittmer (2001) on the VBL and de Woot's (1996) model on change management.
Figure 1. Steps of values-based leadership and anticipated outcomes
In figure 1, the upper line represents an ideal proceeding of VBL.The lines beneath illustrates situations (blank boxes) where one of the step is not taken into account, and the practice thereby leads some unintended outcomes in VBL.The lowest line represents situation where every leaders ignore every steps and aspects of VBL, and then apparent outcome is likely to be VBL, i.e. for instance, integrity violations, misbehaving the personnel, ignoring personal and organizational rules, values and virtues, unfair behavior, telling lies, disloyalty, etc.
In the following sections, the questions of how and why boxes are left blank are discussed in detail.Still, brief comments on each step can be made here.Sensitivity, particularly sense of values is the capability to recognize the values dimension in leadership, and this must take place before one can consciously engage his/her VBL.Otherwise, decision or judgment is made of other than values-based grounds.Values awareness refers here to judgment, especially to situations where values-based decisions are made.For instance, a leader decides what the "right" thing to do is.The third step, ethical competence is the drive to live in accordance with ethical principles and values, and apply them in daily situations, introduced in the values awareness step (Jones & Ryan, 1997;Wittmer, 2001).No doubt, there are numerous content variables that influence the VBL process and leaders' behavior and managerial actions.Among those identified by researchers are individual, situational, organizational, legal, ethical, educational, gender, cultural and climate context variables (see Mencl & May, 2009;McDevitt, 2007;Hood, 2003;Wittmer, 2000;Treviño, Weaver & Reynolds, 2006;Treviño, 1986;Rest, 1986;Loe, Ferrell & Mansfield, 2000).
Sense of values
Values The three steps represent the consistent and widespread supported categories in VBL literature, consisting values sensitivity, sense of values, and ethical competence.In that sense, using these three steps are not entirely new, and as Jones (1991) andO'Fallon andButterfield (2005) review of ethical leadership models show, although models' emphases and terminology differ, most models represent sequentially related components.
However, applying de Woot's (1996) model of change management which is not used in VBL field provides an idea to illustrate the proceeding of VBL with blank boxes.De Woot's (1996) schema helps us especially to reveal why each step is vital in VBL, and more importantly, why solving one step is not enough for overall success of VBL.For example, despite that leaders may tend to be very sensitive to ethical or value dilemmas, they might lack values awareness, and thereby end in a situation where any ethical or value principle tends to fit and be applicable.Similarly, leaders may have sense of values and great values awareness about the emerging issues, but they lack the skills to apply values in practice, i.e.VBL competence.
Sense of Values
"Recognizing a situation as an "ethical" one is the first, critical step in the process of ethical decision making.… In short, ethical deliberation implies ethical detection."(Wittmer, 2000, p. 181).Sense of values is an important attribute because leaders must consider the welfare of others, both inside and outside the organization (Wittmer, 2000;Petrick & Quinn, 1997;Rest, 1986).Sensitivity involves an interpretive process wherein the individual recognizes that organizational values or a moral principle is relevant to the circumstances (Rest, 1986).
More specifically, sense of values is a leaders' capacity "that enables professionals to recognize, interpret and respond appropriately to the concerns of those receiving professional services" (Weaver, Morse & Mitcham 2008, p. 607).Therefore, if one does not recognize an issue as having values, moral or ethical content, the value judgment in leadership process is not likely to be engaged (Brown & Treviño, 2006), and thereby, we might end in a situation where we have a blank box in VBL (Figure 1).Then, an outcome can be that VBL is a series of unconscious acts.Of course, there are several issues that might cause this.
First, individually, our perceptive skills vary in giving values-based meaning to situations as a function of the complexity of factors involved and our own skills and abilities (Lord & Maher, 1990;Wittmer, 2000).According to the social cognition theory, our individual perceptions make up the schema or lens through which we see the world (Fiske & Taylor, 1984)."… a person who fails to recognize a moral issue will fail to employ moral decision-making schemata and will make the decision according to other schemata, economic rationality, for example" (Jones, 1991, p. 380).Caldwell et al. (2002) argue that leader's belief and values determine his/her values framework, but leadership traits, styles and situational factors also influence leader behaviors.
Secondly, a magnitude of consequences is important to sense of values (Brown & Treviño, 2006), and especially, a leader's capacity to evaluate and to considerer consequences.Ethically intense situations draw observers' attention to leaders, and when intense situations are handled correctly, ethical intensity will generally be positively associated with perceptions of VBL.Jones (1991) also includes the magnitude of consequences of the act, but he adds that ethical intensity is a function of the issue-related moral imperative present in a moral situation.It varies with several factors: 1) the extent of social consensus that the act is moral/immoral, 2) the probability that the act will take place and cause the harm/good predicted, and 3) the social, cultural, physical, and psychological proximity of the victims of the act in question.
Thirdly, sometimes leaders are very well aware of values dimensions and are very sensitive to values-related or ethical issues, but still values is left outside leadership.What is striking in VBL and scandals in the corporate world is that, although all of the individuals reported reaching appropriate values sensitivity, only a few actually act on the basis value and ethical judgments and will act as whistle-blowers.There is a great variety of organizational, cultural and individual related factors, which occasionally explain wrong-doing or not choosing the "right thing" despite being readily apparent.Sometimes leader may enforce unethical acts or acts against shared values, but if an act was intended to benefit the organization at the expense of some third party, an attempt to rig a competitive bid, for example, the leader's decision might be judged in positive terms.(Jones & Ryan, 1997).
Jones and Ryan (1997) use the term 'moral approbation' to describe the manifoldness how individuals recognize values-related issues and how this turn value or moral recognition into moral behavior.Basically, individuals derive arguments from five sources: philosophy, religion, biology, socialization, and cognitive development.
However, the relative weights of each source are unknown and individuals have different philosophical and religious orientations.As well, they are biologically different and are socialized differently, but still, it seems that combination of forces from these five sources leads most individuals to seek moral approbation.Furthermore, one might argue that if a leader cannot recognize manifold grounds, he/she is likely miss the step of sense of values.
Leaders often compare their actual selves with their "ideal" selves (what they would like to be) and their "ought" selves (what they think they should be), and sometimes behave in a manner that reduces the discrepancy.Yet, part of the judgment in values sensitivity derives from organizational ethics, especially from codes of ethics, authority structures, and ethical culture on behavior.Some comes from individual, organizational, and group values, as well as, individual moral grounds, organizational policies and practices undertaken in organization.
Still, they conclude that organizational life and leadership is too complicated to be interpreted in terms of a single aspect.(Jones & Ryan, 1997.)In the VBL, we may fail in this step if we are not able to particularize the source and magnitude of values-related issues (Weaver et al., 2008).Otherwise, the problem with values and recognizing them is just one of the issues that leaders and personnel have to cope with, and the issue acquires no ethical or values-related meaning.To particularize the issue requires interpretation which is based on social cognition and the leader's values and reasoning, as well matching the given interpretation with organization vision and strategy.When it is particularized, personnel and stakeholders often feel that the issue is taken into consideration.
Values Awareness
Values awareness is seen here as the second step in the VBL.As Treviño, Weaver and Reynolds (2006, pp. 954) refer to Rest's (1986) study where he observed that "once an individual becomes aware of an ethical issue, ethical judgment process should be more likely to be triggered".In other words, without sensitivity, situations and issues are not interpreted as values-grounded or we assume that no values dimensions are involved.Moreover, one might expect that an outcome of such a VBL where a we do not recognize the importance of values discretion (i.e.sense of values) but we do not have solid values-ground, the VBL might turn out as drifting and implying a mentality of 'anything goes' (cf. Figure 1).When values awareness exists in a leadership, values-based and moral reasoning (e.g. in Kohlberg's sense) become an "intentional, effortful, and controllable" process (Haidt, 2001).
Leaders demonstrate values awareness by having a concern for 1) the collective good of the group, 2) the impact of both means and ends, 3) the long-term and not just the short-term, and 4) the perspectives and interests of multiple stakeholders (Treviño, Brown & Hartman, 2003).Leaders also have responsibility for instituting standards of ethical conduct and moral/organizational values that guide the behavior of followers, of course, alongside the greater emphasis given to productivity and financial objectives (Grojean, Resick, Dickson & Smith, 2004), and these tasks require awareness of values-based dimensions.
Thus, it is obvious that values awareness is a complex issue in leadership.What particularly can cause a blank box in the second step in the VBL?A certain kind of start is Stephens and Lewin's (1992, pp. 2) argument that "Perhaps unethical choices in organizations are often made not because of human evil or unethicality, but because ethical decision making is cognitively complex and strongly affected by organizational design.".Jones and Ryan (1997) argue that leaders take their moral cues from their immediate peers or the larger society.
Leaders are also limited in their capacity to process information and therefore rely on decision making heuristics to simplify the process, particularly in complex situations such as values-based judgment.
Treviño, Weaver and Reynolds (2006) argue that value-based judgment, which is often based on values awareness, is associated with several factors.First, values awareness has been strongly associated with age and education level, especially in terms of cognitive moral development (see also Weber, 1996).Secondly, they argue that particularly the type of harm and the magnitude of consequences tend to explain the intensity of values awareness.Thirdly, context is associated with values awareness.For instance, Weber (1990) finds that values awareness tend to be lower when individuals respond to work-related dilemmas compared to non-work dilemmas.
Moreover, Weaver et al. (2008) summarize six often pop-up difficulties in values awareness.They also provide ways to overcome difficulties: 1) Uncertainty, when values awareness is seeking other cues and viewpoints, examine personal vulnerabilities and challenge pre-existing conviction and new ideas; 2) Vulnerability, when in values awareness attention should be paid on ongoing contact with the client to identify cues and evoke deeper understanding of the client's experience; 3) Receptivity, which emphasizes that values awareness, is openness to learning from others.Receptivity is enhanced by freedom from competing obligations; 4) Responsiveness, which is in values awareness a sense of the demand of another's or anticipation of harm to the other.It is enhanced by having support from colleagues; 5) Courage, in values awareness is challenging fixed conceptions of responsibility and normality, placing values in tension with one another, and to assume responsibility for consequences; 6) Relationship, which increases possibilities to develop values awareness and recognize the unique, irreplaceable characteristics of customer and personnel.
Leaders are often held responsible for making codes and values visible and interpreting them if necessary.For example, Paarlberg and Perry (2007) found that managers play important role in interpreting broad strategic values into goals that are meaningful for employees in their daily work and routines.Leaders also have the primary role in communicating and demonstrating the true importance of ethical values to the organization's members (Grojean, Resick, Dickson & Smith, 2004) and to motivate and commit personnel to ethics codes and organizational values.
Thus, as Grojean, Resick, Dickson, and Smith (2004) argue it is important for leaders to have awareness of personal values as they influence the choices they make and the behaviors in which they engage.To gain ethical-conscious leadership, it involves integration of personal ethical values and the needs of the stakeholders inside and outside of the organization, especially in the development of an ethical framework.This kind of focus requires that values is the cornerstone of how they conduct decisions and set strategies by practicing ethical behavior in their personal life, in their organizational duties, and in their relationships (Sims & Brinkmann, 2002).
A utility of shared values and ethical codes according to Meglino and Ravlin (1998) Just as important in understanding the impact of ethics initiatives and infrastructures on behavior, however, is the question of their integration with routine organizational functions (Weaver, Treviño & Cochran, 1999).
Leaders not only directly influence the behavior of members, but their actions also influence the perceptions of members which lead to norms and expectation of appropriate conduct that become ingrained in the organization's climate (Grojean, Resick, Dickson & Smith, 2004).Leader's actions both directly and indirectly establish the ethical tone of an organization by the actions that are encouraged, rewarded, and demonstrated.According to Brown et al. (2005) values-based leaders provide followers with voice, i.e. allowing followers a say in decision making and listening to their ideas and concerns.De hoogh and Den Hartog (2008) call this power sharing and employee empowerment.Treviño's et al. (1999) empirical data showed that the degree to which individuals openly talk about ethics in an organization is a good predictor of ethical conduct in that organization.Gortner (2001, pp. 524) adds that along with articulation goes sensitization which can help in developing mutual understanding, in creating a respect for the perceptions and ideas of others even though there is disagreement, and ultimately, in finding ways to solve the seemingly unresolvable.
However, there is a threat of overemphasizing values and ethics, as well as sanctions for misbehavior.Bartol and Locke (2000) suggest caution in using rewards to foster desired behavior, especially that personnel do not sacrifice the overall desired outcomes.Essentially, because of balancing organization's desired outcomes and ethics, leaders are responsible for engaging subordinates in in-group exchange activities to enact ethical codes.Despite the consensus that rewards and punishments have an impact on ethically relevant behavior the relationship of rewards and punishments to ethical behavior is not so simple.Often, introducing more sanctions or rewards does not lead to increase in ethical awareness (Treviño & Youngblood, 1990).Sometimes leaders have to think about situation where weak sanctions can be worse for ethical behavior than no sanctions at all (Tenbrunsel & Messick, 1999).
In addition, strong sense of values and ethics, values awareness may lead over-emphasizing their importance.If leaders pay attention to much that is wrong or bad, or what is according to values and what is not, managers capacity for focusing on those wrong or bad things which are 'really important' can significantly diminish.Thus, for leaders ethical awareness is necessarily selective.Some situations may even require that leaders operate in the moral sphere in a decisive and uncompromising way (Smilansky, 1996, p. 1415).
Executives need courage, confidence, and moral strength to make unpopular decisions.When these leaders' traits are not tempered with modesty, openness, and integrity, ethical problems can arise.Messick and Bazerman (1996) list three illusions related to leader's personal characteristics and what leaders should re-evaluate.First, illusion of favorability in which the leader has an unrealistically positive view of the self.The challenge here is that leaders edit and filter information about themselves to maintain a positive image and undermine critics about their honesty, fairness and other ethical traits.Secondly, leaders might keep up the illusion of unrealistic optimism, where they believe themselves to be relatively immune from risks.Thirdly, the illusion of optimism is often supported by the illusion of control, especially that leaders feel that they are able to control the environment and their organization.
Competence to Put Values into Practice
Values awareness, including generally applicable principles, ethical codes, values and guidelines constitute the foundation of the moral orientation and are necessary in all external and internal actions of leadership in organizations.In real-life situations we have to be able to apply these general principles and act according to them.
However, it is seldom obvious what constitutes right or wrong (Kavathatzopoulos, 2003) and often there is no ready-made solution for emerging ethical problems.Instead, issues related to values are often encumbered by ambiguity.Therefore, leader actions that clarify ethical issues help to reduce ambiguity (Grojean, Resick, Dickson & Smith, 2004, p. 229).In other words, it is argued here that being sensitive to value dilemmas and recognizing ethical issues in leadership, is not enough: they are just precursory steps to the third step, a competence to put values into practice.
What exactly is competence, especially competence to put values into practice?Kavathatzopoulos (2003) defines it as a psychological skill, a leader's ability to treat values conflicts in the best possible way for all parties concerned, knowing how to think, how to analyze actual cases, how to make decisions and how to solve problems on the basis of values.This may also hold quite well in the VBL.It also implies self-confidence and willingness to execute difficult decisions, and to support and sustain values-based positions.Furthermore, it illustrates that the leader has an ability to apprehend values-based situations and to realize responsibilities (Kälvemark Sporrong et al., 2007, p. 826), being aware of values and having consistent attitudes and the willingness to realize them (as proposed in Figure 1).
As for VBL competence, in most cases a leader's acts are weighted in terms how well he/she is able to interpret and communicate values and ethical codes to the personnel.Interpretation and communication are needed because even the individual's immediate peers or the top executives, as well as an espoused values and code of ethics could give conflicting or ambiguous signals (Jones & Ryan, 1997).Often, organizational values are abstract, illustrating good and wanted outcomes and behavior, but they do not guide practitioners in particular tasks and efforts (Brytting & Trollestad, 2000;Wittmer, 2001).Also, in situations where tasks are ill-defined, and standards of practice are not well established, interpretation of values provided by leaders is important (Brown, Treviño & Harrison, 2005).
Employees align values through social processes and routine interactions between employees, managers, and customers and other stakeholders.In these processes, leaders and middle managers play key role in integrating individual values through the interpretation of strategic goals.A performance management system, an organizational structure, and control and rewards provide a formal opportunity to both articulate values and signify which employee values are important to the organization's mission (Paarlberg & Perry, 2007).
One problem in the VBL competence is that leaders have difficulties to see the context of the problem, how to control the problem, or finding their own way to handle it.Then, the question is how to create functional problem-solving strategies and apply critical thinking, and still, follow organizational values.Yet, sometimes solutions may be antithetical to certain appreciated values in the organization.Still, in leadership one may come to the conclusion that it is necessary to violate a principle which one upholds strongly, in order to preserve something more important.Such ethical controversies heighten the complexity of the values competence and make the VBL much more difficult (Kavathatzopoulos, 2003;Smilansky, 1996).
It is strongly proposed in the literature that values-based leaders need a good image and legitimization, and they are especially weighted in real-life situations.Leaders are obligated to furnish a moral example for their subordinates and leaders are a central source of such guidance (Brown et al., 2005).Through role modeling, values-based leaders promote altruistic behavior and they can earn the confidence and loyalty of their followers.Role modeling influences employees via two routes: increasing trust in leaders, and facilitating value congruence (i.e.increasing 'fit' into the organization) (Grojean, Resick, Dickson & Smith, 2004, p. 229).
In other words, leaders who demonstrate actions that are consistent with the organization's values and mission are likely to be viewed as more trustworthy.De Hoogh and Den Hartog (2008, p. 300) propose that employees will be more positive, hopeful, and optimistic about their organization and contribute to organizational success when their leaders act in an ethical manner.If the leader's moral integrity is in doubt, the leader will more likely fail to influence followers to achieve organizational goals (Kanungo, 2001).Weaver, Treviño and Agle (2005) found that if a leader is seen as an ethical role model at work, interviewed individuals associated characteristics such as willingness to turn mistakes into learning experiences and humility to ethical leadership.Moreover, as Jones and Ryan (1997, pp. 672) puts it, "human beings are very rarely physically forced to do anything.… the personal costs of failure to comply with organizational directives can be quite high and agents may feel psychologically compelled to behave unethically.".
Role-modeling and successful cases representing that ethical conduct are needed because most lower-level employees in large organizations rarely interact with senior managers.Then, they must make inferences about the attributes of such leaders based upon available information (e.g.public relations information and organizational outcomes) and images rather than direct experience (Lord & Maher, 1990).Even then in today's climate of high-profile business scandals, much attention focuses on the role top executives play in setting the ethical tone of their organizations and research also have demonstrated the importance of top management commitment in fostering ethical practices.But employees often are influenced most by their peers, groups, friends and immediate supervisors.Moreover, research has shown that top-down initiatives to foster organizational ethics are limited in impact (Weaver, Treviño & Agle, 2005, p. 314).If a leader is only peripherally involved, or his role is modest, he may consider the situation as ethically distant because he will be judged less harshly than those who are directly involved or whose role is great (Jones & Ryan, 1997).
Then, the question is how to influence the ethical quality of follower decisions if a leader is not physically present?In the research, several ways are presented.First, if values-based leaders represent attractive role models, but also in terms of their assigned role, their status, they are an important source of ethical guidance for their employees (Brown & Treviño, 2006).Secondly, Brown and Treviño (ibid,p. 607) refer to Dukerich, Nichols, Elm and Vollrath's (1990) study that leader moral reasoning can influence moral reasoning in work groups.Thirdly, followers know that the leader will be holding them accountable for their decisions and will use rewards and discipline to do so (Brown, Treviño & Harrison, 2005).Fourthly, Lord and Brown (2001) show that leaders can impact on man subordinate processes by influencing their self-concept.Here, leaders mediate culture and other exogenous factors as well as a range of psychological matching (i.e.observational learning, imitation, and identification), which in turn influence the subordinate's behavior.
Competence in the VBL is formed in a social context, thus there is a need to create arenas where the collective values competence of a workplace and among leaders can be maintained and developed (Kälvemark Sporrong et al., 2007).The attitudes and behaviors of peers in the workplace affect individuals' ethical behavior, with frequency and intensity of interaction peers make that influence stronger.De Hoogh and Den Hartog (2008, pp.300) comment that too little attention has been given to top management teams in VBL.Top management team is involved in strategic decisions and directing the organization toward desired goals.If for example, despotic or self-interest guide working within a team, it gives a negative image for the rest of the organization.This also highlights the importance of values approval from one's peers and how other persons' ethical behavior serves as an influential role model for an individual's own ethical behavior in single situations.For that reason, sometimes despite an organization's efforts to highlight an executive's stance towards ethics, the role models people look to tend to be among those with whom they have close working relationships (Treviño, Weaver & Reynolds, 2006).Thus, in leadership it is important to find and embed ethical practices and behaviors in the all processes and functions of the organization.
Conclusion and Implication
The discussion on the Values-Based Leadership (VBL) was declined here into three steps with emphasis on leaders' perspective which are needed in effective and sound values-based actions and leadership.It was asked how and why problems occur in VBL, and further, the idea of the blank boxes illustrated problems and pitfalls that we face in each step and which might accumulate as real problems in outcomes of VBL (cf.figure 1, section 3).The steps also illustrated the debate on the interrelation of steps.
In sum, three following conclusion are feasible.First, if sense of values is missing, the importance of values-based judgment or values dimensions is not taken into the consideration of the leadership, and therefore, leaders are not able to give ethical meaning for the issue.This follows that leaders are not capable of solving ethical problems.Discussions revealed that a key to better values sensitivity is related to the way how a leader is able to particularize and specify an ethical issue and share it with personnel.
Secondly, values awareness represented here a leader's consciousness of values and ethics, and the capabilities to fit them to organization's operations and strategies.If it is missing, leadership suffers from drifting without values and ethical principles; most values tend to fit, but choices and decisions cannot be restored to any of the appreciated values.Thus, to reach values awareness it requires choices and prioritizations between competing values and particularization of values in leadership goals and functions.Also, values awareness in leadership means that a leader constantly upholds key values and applies values and ethics in his/hers acts.
Thirdly, in VBL competence it is tested in practice how well leaders are able to use values and ethics in leadership.A common problem is that values and ethics are known, but they are not present in decision making or prioritizations.In sum, when values are applied in leadership, leaders should be receptive to various signals and have the ability to reflect and benchmark their decisions to former decisions.In the same way, leaders are responsible for giving operational meaning for abstract values and ethical principles and the coordination of value-intensive work in all organization.In addition, competence includes facilitation and mentoring of value-guided working and maintaining open dialogue on values and ethics.
Moreover, when difficulties are recognized in each step of VBL, it is possible to suggest some developmental actions to values-based leaders (see Table 1).The steps of VBL are at the left side, and some suggestions for the leaders are on the right side.Emphasis here is on leadership acts, not on virtuous or moral aspects of leadership and values/ethics.In other words, suggestions are presented for practitioners.
Table 1.Suggestions towards better values-based leadership practices Steps of VBL Suggestions for leaders' tasks.
Sense of values
Increase personnel's knowledge on values and ethics by using round-table discussions.
Listen to stakeholders' needs and turn them into values and ethical principles.Find a balance between enduring values and short-term operational goals.
Values awareness
Create and facilitate discussions on persistent ethical principles and value transformation.
Institutionalize values and ethical codes and procedures, especially when the organization expands.Incorporate values and ethics in strategically important functions and services, and monitoring, feedback and evaluation systems.
Competence to put values into practice
Communicate values and create trust in ethical codes.Coordinate competing values and find opportunities for consensus.Operationalize values and provide task-specific and self-related feedback.Establish value-platforms in several levels of organization hierarchy where values can be shared, debated, and agreed.Further research is needed in order to understand ethical leadership benefits and pitfalls, and also how to institutionalize ethical leadership practices in a variety of organizations.Further studies would also be fruitful in investigating unintentional outcomes of VBL.Perhaps, using a multidimensional evaluation model and stakeholder approach would benefit the research, and help moving towards comprehensive constructions of VBL in organizations, and thereby, making values and ethics more familiar to a variety of stakeholders, not just the CEO-office business.
is that if people hold similar value systems, this enables them to communicate more clearly, predict each other's behavior, and more efficiently to coordinate activities, resulting in reduced role conflict and ambiguity and increased satisfaction with the interpersonal relationship.Codes assist in raising the general level of awareness of values and ethical climate in organizations, but achieving a high level of awareness requires interpretation and communication of values and codes.Moreover,Paarlberg and Perry (2007, pp.405) conclude that "organizations cannot influence employee behavior by communicating "the values of organization," as articulated by the top leadership through formal presentations or the distribution of laminated cards.". | 2018-11-06T16:49:46.433Z | 2012-04-26T00:00:00.000 | {
"year": 2012,
"sha1": "723d299020cb7bedd180976b213aee810df6077a",
"oa_license": "CCBY",
"oa_url": "https://ccsenet.org/journal/index.php/ijbm/article/download/15718/11163",
"oa_status": "HYBRID",
"pdf_src": "Anansi",
"pdf_hash": "723d299020cb7bedd180976b213aee810df6077a",
"s2fieldsofstudy": [
"Business"
],
"extfieldsofstudy": [
"Political Science"
]
} |
258950235 | pes2o/s2orc | v3-fos-license | Bioaffinity Nanoprobes for Foodborne Pathogen Sensing
Bioaffinity nanoprobes are a type of biosensor that utilize the specific binding properties of biological molecules, such as antibodies, enzymes, and nucleic acids, for the detection of foodborne pathogens. These probes serve as nanosensors and can provide highly specific and sensitive detection of pathogens in food samples, making them an attractive option for food safety testing. The advantages of bioaffinity nanoprobes include their ability to detect low levels of pathogens, rapid analysis time, and cost-effectiveness. However, limitations include the need for specialized equipment and the potential for cross-reactivity with other biological molecules. Current research efforts focus on optimizing the performance of bioaffinity probes and expanding their application in the food industry. This article discusses relevant analytical methods, such as surface plasmon resonance (SPR) analysis, Fluorescence Resonance Energy Transfer (FRET) measurements, circular dichroism, and flow cytometry, that are used to evaluate the efficacy of bioaffinity nanoprobes. Additionally, it discusses advances in the development and application of biosensors in monitoring foodborne pathogens.
Introduction
Food safety has become a global concern, given the frequency and severity of foodborne disease outbreaks recently, and the grave effects associated with them. Most of these outbreaks are caused by foodborne pathogens, which are bacteria, viruses, and parasites that contaminate food, causing diseases and in some cases deaths. Some of the predominant foodborne pathogens include Listeria monocytogenes, Campylobacter, Staphylococcus aureus, Salmonella typhimurium, and Escherichia coli [1]. The advancement of food supply chains, which has become a giant network connected to all parts of the world, makes the spread of foodborne illnesses more rapid and the strain on socio-economic development disastrous. Although there have been great advancements in the methods used to detect pathogens, there are still increasing outbreaks of foodborne diseases showing that the methods of analysis are no longer enough.
The conventional culture-based methods are very cheap and easy to use but take up to several days to produce results and require extra biochemical or molecular tests to confirm that the species match the pathogen of interest, making them highly unsuitable for on-site detection [2]. Other methods have also been developed, such as the polymerase chain reaction (PCR), nucleic acid, and immunoassay-based methods. These methods are able to fix the time constraint of the culture-based methods but are usually expensive and require specific reagents, complicated sample pretreatment, and experienced personnel for the analysis, making the possibility of commercialization nearly impossible [3,4]. Hence, the urgent need for inexpensive, easy-to-use, but accurate, and rapid detection methods that do not require specialized expertise or equipment to run. Nanosensors, which are a product of biosensor technology incorporated into nanotechnology, are the newest and most advanced detection technology being developed by scientists [4,5].
A nanosensor is basically a compact analytical device with dimensions below 100 nanometers (nm) that detects the presence of biomolecules and nanoparticles or monitors physical and chemical parameters on a nanoscale. It can be a sensor with bioreceptor
Antibodies
Antibodies, which are also known as immunoglobulins (Ig), are defined as large proteins having 'Y' shapes and are naturally produced in the body of an animal in response to antigens. Their main purpose is to serve as a defense mechanism of the immune system through phagocytosis, complement-mediated lysis, neutralization of infectivity, and antibody-dependent cellular cytotoxicity (ADCC) [13][14][15]. The structure of antibodies can be split into two; the first is the antigen-binding fragment (Fab), and the second is the constant or crystallizable fragment (Fc). The Fab region establishes the idiotype of the antibody and possesses affinity towards the target, transmitting a neutralizing effect to it once it binds to it. The Fc region controls other immune-associated activities including macrophage and complement binding, as well as defining the isotopes of the antibody [14,16]. All antibodies comprise four polypeptide chains held together by disulfide bonds. Two of these chains are heavy chains and the other two are light chains, which come together to create the Y-shaped structure [17,18]. Typically, antibodies are classified into five classes (IgA, IgD, IgE, IgG, and IgM) depending on the composition of their heavy chain constant region. These antibodies occur in the form of monomers (IgD, IgE, and IgG), dimers (IgA), or pentamers (IgM) [14,16]. There are two general types of antibodies, which are monoclonal antibodies and polyclonal antibodies. They have the same basic structure and function, but their differences have to do with their production and specificity. Table 1 provides a summary of the differences between monoclonal and polyclonal antibodies. Table 1. Comparison between monoclonal and polyclonal antibodies [19].
Category Monoclonal Antibodies Polyclonal Antibodies
Synthesis Synthesized by one clone Synthesized by numerous clones
Production requires both in vitro and in vivo systems
Production is strictly in vivo (animal host is a must) Production requires trained personnel Highly skilled personnel are not needed Short-term production is expensive but long-term production is cheap.
Short-term production is cheap but long-term production is expensive due to animal maintenance and deaths.
Homogeneity
They are homogenous in nature, making it easy to characterize their chemical nature and an easy choice for conjugation to different probes.
They are difficult to characterize since they are not homogenous.
Specificity
Highly specific They are specific but exhibits cross reactivity Degradation Vulnerable to degradation under slightly harsh conditions. Less vulnerable to degradation.
Affinity Purification An excellent tool for affinity purification.
They are not a good choice for affinity purification Scientists, for the past decades, have used antibodies as the predominant receptor for biosensors due to the natural antigen-antibody interaction. In the development of antibodies to be used as bioreceptors, there are three factors that are sought. The first is sensitivity, which is the ability of the antibody to be able to recognize and quantify the target molecule even when the concentrations are low. This is mostly a problem for foodborne pathogens since they usually occur in trace amounts and yet are very potent. The second is specificity, which is the ability to differentiate the pathogen strain even in the presence of other strains or pathogens. The last thing is high affinity, which is the ability to form a complex with its target that is strong enough to allow further analysis [20,21]. Selecting antibodies with all three characteristics is quite a task, but because monoclonal antibodies have very impressive specificities [22], scientists have been able to develop them for a myriad of foodborne pathogens over the years.
Recombinant technology has been infused into the development of antibodies to make the process more efficient. This technology allows the production of antibodies from synthetic antibody repertoires without the immunization of animals [23]. Plückthun and Skerra developed a method that uses vectors present in bacterial systems to create fully functional, completely folded antibody fragments [24], as opposed to the traditional method of generating the fragments by proteolytic cleavage alone [25].
Antibodies as bioaffinity nanoprobes have a wide variety of applications grouped into diagnostic and therapeutic medicine, agrobiotechnology, food safety, environmental protection, and many others. These antibodies cannot work independently, but they have to be incorporated into technologies such as colorimetric, electrochemical, voltammetric, and optical biosensors, or even conventional assays including the enzyme-linked immunosorbent assay (ELISA) to be able to function completely as a nanoprobe.
In the case of biosensors, their success depends partly on the ability to immobilize the antibody while maintaining their original activity. This immobilization step is so crucial because it affects the sensitivity and overall performance of the sensor, as well as the detection limit [17]. The antibodies can be immobilized on the solid sensor surface through methods such as:
•
Adsorption using electrostatic or hydrophobic interactions; • Entrapment; • Covalent coupling using amine coupling, thiol coupling, or coupling through glycan moiety; • Affinity: immobilization is performed through intermediate proteins such as in the case of avidin-biotin [13].
Enzymes
Enzymes are macromolecules that act as catalytic agents, meaning they accelerate the rate of biological or chemical reactions without being consumed or taking part in the reaction. They are usually proteins, but some RNA molecules called ribozymes have been found to possess catalytic abilities. They function by lowering the activation energy of the reaction as they stabilize the transition state. Some enzymes rely on other small non-protein molecules called coenzymes to function fully. Enzymes are far more efficient than most of the inorganic catalysts available. This is seen in their impressive specificity. Inorganic catalysts increase the rate of a bunch of chemical reactions in the system while enzymes target specific reactions. Even when they target more than one reaction, the reactions are usually of the same type and their reactants have similar structural traits. Because of the specificity of enzymes, they are able to selectively differentiate between substances (substrates or analytes), even those that are optical isomers.
As per the International Union of Biochemistry and Molecular Biology's proposal, enzymes are classified into six primary groups, which include oxidoreductases that catalyze redox reactions, transferases that facilitate the transfer of atoms from a donor substrate to an acceptor, hydrolases that catalyze the breaking of bonds through the addition of water, lyases that cause the cleavage of bonds through methods other than hydrolysis, isomerases that catalyze the conversion of isomers, and ligases that catalyze the binding of molecules [26][27][28]. During a reaction, the enzyme binds to the substrate at a specific location to form a complex. These locations in the enzyme's structure, called the active binding sites, take up only a small portion of the enzyme's total size and are mostly filled with water in the absence of binding. They are often grooves and crevices that the substrate binds to in order for the reaction to be accelerated. Since enzymes are mostly proteins, they are made up of amino acids which form the primary, secondary, tertiary, and quaternary structures. The conformation of amino acids within the active sites plays a crucial role in stabilizing the specific binding of substrates and thereby determining the enzyme's specificity [27,29,30]. Enzymes are isolated traditionally from natural sources; that is, from the organisms that provide an abundant or easily isolated source [31].
The unique characteristics of enzymes, that is, their ability to specifically recognize substrates and catalyze their transformation, giving rise to a signal, make enzymes a perfect bioaffinity probe, fit for use in a biosensor or nanosensor. Biosensors using enzymes as bioaffinity probes were the earliest biosensors to be developed. The amperometric enzyme electrode for glucose sensing utilizing a soluble enzyme electrode was first designed by Clark and Lyons in 1962 [32]. Since then, scientists have grown keen on the use of enzymes as bioaffinity probes in sensing even for the detection of foodborne pathogens. In the use of enzymes as bioaffinity probes, the analyte, specifically the foodborne pathogen, can be recognized by three means.
•
The first option is that the concentration of the enzyme can be estimated by measuring the catalytic transformation of the analyte, which is metabolized by the enzyme. • Secondly, the enzyme is inhibited or activated by the analyte, hence the concentration of the analyte is proportional to the decrease in enzymatic product generation.
•
The last option is by tracking the alteration in the characteristics of the enzyme.
The catalytic impact of the enzyme, upon which the theory of analyte detection is based, is also dependent on multiple characteristics, inclusive of the concentration of the analyte, pH, temperature, and the presence of either a competitive or non-competitive inhibitor [31]. The success of an enzyme in a nanosensor depends on its ability to be held on tightly or carried by a solid surface. The process of attaching the enzyme to the solid surface is referred to as immobilization. Just like antibodies, enzymes can be immobilized through entrapment, affinity attachment, and nonspecific covalent attachment [31]. Aside from using enzymes as the main bioaffinity probe in a sensor, they can also serve as labels in immunoassays (antibody-based biosensing), as in the case of alkaline phosphatase and horseradish peroxidase (HRP) [31,33]. Enzymes as bioaffinity nanoprobes have applications in several fields, including food safety, environmental monitoring, heavy metal detection, and health, and can be used for so long as they are not consumed.
Even though enzymes have excellent specificity and are perfect as nanoprobes, they have their own limitations which include it being expensive and difficult to find new active and efficient enzymes and difficulty in improving the sensitivity and adaptation to other functions [34].
Aptamers
The origin of the term "aptamer" can be traced back to the Latin word "aptus", meaning "to fit", and the Greek word "meros", meaning "part" [35]. Aptamer is used to describe DNA or RNA oligonucleotides that are short and single-stranded, as well as peptides that can recognize their targets with exceptional affinity, selectivity, and specificity [36]. DNA and RNA aptamers were first unearthed in the year 1990 by two independent teams: Ellington and Szostak in the preparation of RNA molecules that targeted organic dyes [37] and Tuerk and Gold in T4 DNA polymerase [38]. In 1996, Colas et al. also introduced peptide aptamers, as they reported short structures of peptides with the ability to detect cyclin-dependent kinase 2 [39]. Aptamers are selected through a meticulous, repetitive procedure, consisting of a series of selection and amplification, popularly known as "Systematic Evolution of Ligands by Exponential Enrichment (SELEX)." This process involves three main stages, and they are shown in Figure 1. on tightly or carried by a solid surface. The process of attaching the enzyme to the solid surface is referred to as immobilization. Just like antibodies, enzymes can be immobilized through entrapment, affinity attachment, and nonspecific covalent attachment [31]. Aside from using enzymes as the main bioaffinity probe in a sensor, they can also serve as labels in immunoassays (antibody-based biosensing), as in the case of alkaline phosphatase and horseradish peroxidase (HRP) [31,33]. Enzymes as bioaffinity nanoprobes have applications in several fields, including food safety, environmental monitoring, heavy metal detection, and health, and can be used for so long as they are not consumed. Even though enzymes have excellent specificity and are perfect as nanoprobes, they have their own limitations which include it being expensive and difficult to find new active and efficient enzymes and difficulty in improving the sensitivity and adaptation to other functions [34].
Aptamers
The origin of the term "aptamer" can be traced back to the Latin word "aptus", meaning "to fit", and the Greek word "meros", meaning "part" [35]. Aptamer is used to describe DNA or RNA oligonucleotides that are short and single-stranded, as well as peptides that can recognize their targets with exceptional affinity, selectivity, and specificity [36]. DNA and RNA aptamers were first unearthed in the year 1990 by two independent teams: Ellington and Szostak in the preparation of RNA molecules that targeted organic dyes [37] and Tuerk and Gold in T4 DNA polymerase [38]. In 1996, Colas et al. also introduced peptide aptamers, as they reported short structures of peptides with the ability to detect cyclin-dependent kinase 2 [39]. Aptamers are selected through a meticulous, repetitive procedure, consisting of a series of selection and amplification, popularly known as "Systematic Evolution of Ligands by Exponential Enrichment (SELEX)." This process involves three main stages, and they are shown in Figure 1. The first step is the incubation of the target molecule with the library. In this step, the target is incubated into a large, random pool of about 10 15 single-stranded nucleic acid sequences where there is an interaction between the target and nucleic acids [41]. The next step is the segregation of the nucleic acid-target complexes from the unbound sequences and the discarding of the unbound sequences [42]. The first step is the incubation of the target molecule with the library. In this step, the target is incubated into a large, random pool of about 10 15 single-stranded nucleic acid sequences where there is an interaction between the target and nucleic acids [41]. The next step is the segregation of the nucleic acid-target complexes from the unbound sequences and the discarding of the unbound sequences [42].
The final step is the amplification of the sequences that formed the complexes by polymerase chain reaction (PCR) in the case of DNA or reverse transcription PCR (RT-PCR) for RNA. This amplified group of sequences becomes the new initial library for the next cycle [35].
The selection method is repeated until some oligonucleotide sequence(s) with exceptional specificity and selectivity are obtained. These become the selected aptamers, and it usually takes at least 8-15 rounds of SELEX to achieve, yet the whole process takes a few weeks [42]. Aptamers can be selected to detect a wide variety of targets including proteins, bacteria, viruses, protozoa, small chemicals, metal ions, antibiotics, parts of cells, and even whole cells [43]. They are also applied in so many fields such as the medical sector for diagnosis and therapy, food quality, environmental safety, research, and bioanalysis. Aptamers have become a highly sought-after choice in the development of biosensors due to the mouthwatering advantages they possess. Some of these advantages are excellent affinity, sensitivity, and selectivity towards targets, low cost of production, shorter production time, low toxicity, easy modification, ability to easily permeate tissues due to small size, stability in extreme conditions, and the ability to retain their original conformation when favorable conditions are restored [44][45][46]. Aptamers can be described as a prominent successor of antibodies in bioanalytics and nanosensor development as they provide solutions for most of their limitations and a competitive affinity and limit of detection of targets.
Non-Antibody Binding Proteins
Non-antibody binding proteins, also known as synthetic binding proteins, are proteins with a non-immunoglobulin fold generated by non-antibody scaffolds. These scaffold domains are obtained by creating a random library through targeted mutagenesis in a loop region or another acceptable surface area. Variants are then selected against a specific target using phage display or other molecular selection methods [47]. While several protein scaffold options have been proposed, only a few have been proven to provide specificities for various target types and offer practical advantages. The scaffold domains that have been found to produce these proteins include Anticalins, Lipocalin, Sso7d protein, Darpins, Fibronectin type 3, Affibodies, and ankyrin repeat protein [47,48]. Non-antibody binding proteins offer advantages such as low molecular weight, which facilitates tissue penetration; high thermal stability, with approximately 70% of the available scaffolds having denaturation temperatures between 37 and 120 • C; ease and cost-effectiveness of production compared to antibodies; longer shelf life; and robustness [49]. These scaffold proteins enable the generation of chemically consistent proteins that can be tailored to detect various analytes without significantly affecting the biosensor configuration, while also enhancing the packing density of the recognition element [50]. These benefits, combined with their ease of expression, justify their use as a viable alternative to traditional antibodies or their recombinant fragments [47].
Molecularly Imprinted Polymers
Molecular imprinting is a template-guided process that creates selective pockets within a three-dimensional polymeric matrix. By removing the template from the polymer, functional porous materials with high-affinity binding sites, known as molecularly imprinted polymers (MIPs), are obtained. These binding pockets have configurations and functionalities that match those of their target molecules [51,52]. The synthesis of MIPs involves a three-step process: • Incubation: Monomers are incubated with a dummy, epitope, or template molecule, which facilitates the formation and stabilization of non-covalent interactions between the functional monomers and the template.
• Polymer Formation: The polymer is formed around the template with the help of cross-linkers, resulting in the creation of a network structure. • Template Removal: Suitable solvents are used to remove the templates, leaving behind specific binding sites that are complementary to the template molecule [53,54].
MIPs offer several advantages over antibodies, making them a promising alternative. These advantages include structure predictability, chemical and thermal stability, longer shelf life, cost-effectiveness and ease of production, minimal batch-to-batch variation during mass production, and high sensitivity. Due to these properties, MIPs find applications in various fields such as food safety, environmental science, therapeutics, and more [51,53,55]. The ongoing research and development in molecular imprinting techniques continue to enhance the selectivity, stability, and sensitivity of MIPs, further expanding their potential applications in various scientific and technological fields.
SPR
Surface plasmon resonance (SPR) is one of the most prominent sensitive and qualitative, label-free techniques used to monitor binding events and to measure the relations between biomolecules such as protein and protein, protein and DNA/RNA, enzymesubstrate/inhibitor, and receptor-drug [56,57]. It is a spectroscopic method that measures the refractive index changes very at the surface of thin metals such as gold, silver, and aluminum films as a result of biomolecular interactions. Generally, when incident light strikes the metal surface at a given angle (incidence angle), the photons induce an excitation of the free electrons in the surface coating of the metal, causing them to oscillate. The movement of the electrons is called plasmon and it is always parallel to the surface of the metal [56]. A typical SPR equipment comprises a source of monochromatic polarized light and a thin film of metal (most often gold) supported by a glass prism in combination with a photodetector, which is represented in Figure 2. [53,54].
MIPs offer several advantages over antibodies, making them a promising alternative. These advantages include structure predictability, chemical and thermal stability, longer shelf life, cost-effectiveness and ease of production, minimal batch-to-batch variation during mass production, and high sensitivity. Due to these properties, MIPs find applications in various fields such as food safety, environmental science, therapeutics, and more [51,53,55]. The ongoing research and development in molecular imprinting techniques continue to enhance the selectivity, stability, and sensitivity of MIPs, further expanding their potential applications in various scientific and technological fields.
SPR
Surface plasmon resonance (SPR) is one of the most prominent sensitive and qualitative, label-free techniques used to monitor binding events and to measure the relations between biomolecules such as protein and protein, protein and DNA/RNA, enzyme-substrate/inhibitor, and receptor-drug [56,57]. It is a spectroscopic method that measures the refractive index changes very at the surface of thin metals such as gold, silver, and aluminum films as a result of biomolecular interactions. Generally, when incident light strikes the metal surface at a given angle (incidence angle), the photons induce an excitation of the free electrons in the surface coating of the metal, causing them to oscillate. The movement of the electrons is called plasmon and it is always parallel to the surface of the metal [56]. A typical SPR equipment comprises a source of monochromatic polarized light and a thin film of metal (most often gold) supported by a glass prism in combination with a photodetector, which is represented in Figure 2. When the polarized light goes through the glass prism, an evanescent wave that passes through the thin film of metal is generated from the internally reflected light. If the intensity of the reflected light is monitored with time in relation to the angle of incidence, a minimum reflected light will be achieved at an incident angle referred to as an SPR angle. This SPR angle depends on certain optical characteristics of the system such as the refractive index within close proximity of the metal film surface [59]. The sample solution, containing the target molecule most of the time, flows across the SPR surface after the bioaffinity probe (antibodies, aptamers, enzymes) has been immobilized unto the solid surface [60]. When there is any form of interaction at the surface of the metal, an alteration When the polarized light goes through the glass prism, an evanescent wave that passes through the thin film of metal is generated from the internally reflected light. If the intensity of the reflected light is monitored with time in relation to the angle of incidence, a minimum reflected light will be achieved at an incident angle referred to as an SPR angle. This SPR angle depends on certain optical characteristics of the system such as the refractive index within close proximity of the metal film surface [59]. The sample solution, containing the target molecule most of the time, flows across the SPR surface after the bioaffinity probe (antibodies, aptamers, enzymes) has been immobilized unto the solid surface [60]. When there is any form of interaction at the surface of the metal, an alteration in the refractive index will be triggered, resulting in a change in the SPR angle and producing a signal that can be detected [61,62]. The amount of analyte that is bound to the bioaffinity probe is measured by observing the intensity of the reflected light or the shifts in the resonance angle, making it a real-time analysis method [63]. The gold metal and its glass support make up the SPR sensor chip and it is upon this chip that ligands, which in this sense are the bioaffinity probes, are immobilized, sometimes with the help of a polymer matrix.
There are so many different chemical mechanisms that are used for immobilization. Some of them are aldehyde, amino, carboxyl, hydroxyl, and thiol group coupling. Sometimes immobilization cannot be performed directly; hence, certain molecules can be used as capture surfaces to enhance the immobilization process. These are biotin, histidine-tagged, and glutathione-S-transferase fusion proteins [57,61]. Immobilization of the bio affinity probe on the sensor surface is by far the most important step of the SPR analysis since the success of the binding analysis somewhat depends on the response generated by the immobilized ligand [64]. The SPR equipment generates output data in the form of a sensorgram, which is a plot of response units (RU) with respect to time, and Figure 3 gives a pictural view of what a typical sensorgram looks like. in the refractive index will be triggered, resulting in a change in the SPR angle and producing a signal that can be detected [61,62]. The amount of analyte that is bound to the bioaffinity probe is measured by observing the intensity of the reflected light or the shifts in the resonance angle, making it a real-time analysis method [63]. The gold metal and its glass support make up the SPR sensor chip and it is upon this chip that ligands, which in this sense are the bioaffinity probes, are immobilized, sometimes with the help of a polymer matrix.
There are so many different chemical mechanisms that are used for immobilization. Some of them are aldehyde, amino, carboxyl, hydroxyl, and thiol group coupling. Sometimes immobilization cannot be performed directly; hence, certain molecules can be used as capture surfaces to enhance the immobilization process. These are biotin, histidinetagged, and glutathione-S-transferase fusion proteins [57,61]. Immobilization of the bio affinity probe on the sensor surface is by far the most important step of the SPR analysis since the success of the binding analysis somewhat depends on the response generated by the immobilized ligand [64]. The SPR equipment generates output data in the form of a sensorgram, which is a plot of response units (RU) with respect to time, and Figure 3 gives a pictural view of what a typical sensorgram looks like. The sensorgram starts with a baseline, indicative of the response before the start of any form of interaction. When the analyte solution flows close to the surface of the sensor, the analysis enters an association phase which shows the binding of the analyte to the immobilized ligand. This is seen in the steady rise in response on the sensorgram to a point where the complex attains equilibrium and the curve flattens out. Right after the equilibrium phase, a drop in response is observed, which indicates the dissociation phase. This is the stage where the ligand-analyte complex separates, as the SPR system stops the flow of the analyte solution and switches to the flow of a running buffer. More often than not, the complex does not dissociate completely and a regeneration solution, which is usually a mild alkaline or acidic solution, is used to regenerate the sensor surface for subsequent analysis [57,66,67]. Most of the SPR equipment available, such as the Biacore equipment, is able to generate a table of the raw data, the association constant (ka), the dissociation constant (kd), and, most importantly, the equilibrium dissociation constant (Kd), The sensorgram starts with a baseline, indicative of the response before the start of any form of interaction. When the analyte solution flows close to the surface of the sensor, the analysis enters an association phase which shows the binding of the analyte to the immobilized ligand. This is seen in the steady rise in response on the sensorgram to a point where the complex attains equilibrium and the curve flattens out. Right after the equilibrium phase, a drop in response is observed, which indicates the dissociation phase. This is the stage where the ligand-analyte complex separates, as the SPR system stops the flow of the analyte solution and switches to the flow of a running buffer. More often than not, the complex does not dissociate completely and a regeneration solution, which is usually a mild alkaline or acidic solution, is used to regenerate the sensor surface for subsequent analysis [57,66,67]. Most of the SPR equipment available, such as the Biacore equipment, is able to generate a table of the raw data, the association constant (ka), the dissociation constant (kd), and, most importantly, the equilibrium dissociation constant (Kd), together with some statistical models to fit the data. The equilibrium dissociation constant is, however, the star of the show, because it gives a clear picture of the kinetics of the interaction and binding affinity of the bioaffinity nanoprobe used as the ligand [68]. A very small value of Kd, recorded in the nanomolar to picomolar range, shows that the ligand has a high affinity towards its target and hence will be an excellent bioaffinity probe when used as a biosensor for the detection of pathogens. It shows how easily the ligand detects the analyte in low concentrations, how tightly the ligand binds to the analyte, and how difficult it is to separate the complex [69]. These are the qualities required in a great bioaffinity nanoprobe. The SPR has the advantages of providing an automated and rapid alternative to cell-based assays, the lack of a need for reporter molecules such as fluorochromes or radioisotopes for a binding signal to be recorded, hence the biomolecular interaction is evaluated in real-time, and the ligand and analyte involved in the interaction do not lose their conformational integrity [70]. There are also a few limitations which include degrading the sensor surface due to harsh conditions of regeneration, and the immobilization of a sufficient amount of ligand on the sensor surface must also be successful [71].
FRET
Förster resonance energy transfer (FRET), which is also widely known as fluorescence resonance energy transfer, is a technique whereby excited state fluorophores non-radiatively transfer electromagnetic energy to other fluorophores which are about 1-10 nm away. The fluorophore involved in the transfer is termed the donor, and the one receiving the energy (often ground state level), is termed the acceptor [72][73][74][75]. Energy transfer is facilitated through the energetic coupling of transition dipoles between the two fluorophores and can only occur when there is a spectral overlap between the emission spectrum of the donor and the excitation spectrum of the acceptor [74,76,77]. The transfer eventually leads to the donating fluorophore entering the ground state level and the fluorophore accepting becomes excited [77]. During FRET, the likelihood of an excited donor fluorophore returning to the ground state is commonly known as the transfer efficiency (E). This efficiency is dependent on the physical distance from the center of the donor to the center of the acceptor of the FRET pair, "r", as well as the characteristic Förster distance, also known as the quenching radius, "R o ". R o is typically in the range of 2-8 nm and it is bound by several factors shown by Equation (2). The connection between the transfer efficiency and the two important distances is shown in Equation (1) [75,78,79].
Looking at the first equation, it can be inferred that at r = R o , E = 1 2 and R o defines the tiny distance (nm) that exists between the two fluorophores at the point where half of the entire donor relaxation processes occur by transferring energy to the acceptor [77,80]. The magnitude of R o depends on the orientation (k 2 ), the medium's refractive index (η), and the donor's quantum yield (φ), in addition to the degree at which the spectra of the donor and acceptor overlap (J(λ)). The number of unquenched donor fluorophores represents the rate at which energy is transferred between the donor and acceptor [75,77]. FRET is preferred over other options of interaction analysis because it has ve imal restrictions and can even be used within a living cell. The major requiremen ability of light to be delivered to and collected from the sample, but generally, only benchtop equipment is needed [74]. It is perfect for biomolecular interaction anal cause the majority of the biomolecules are in the nanoscale range and FRET is als in that same range. The nanometric distance measurements can also serve as a 'm gauge' for biomolecular structure analysis [82]. FRET is versatile enough to be ap diverse areas of biomolecular research, but it is also sensitive to environmental con such as solvent pH, viscosity, polarity, and many others [83]. FRET analysis can be out either with a single FRET or the multiplexed FRET methods. Multiplexed FRE ods offer much more advantages including simultaneous analysis of multiple a analysis of intermolecular and intramolecular interactions, and monitoring of coi biomolecular events [84].
The major disadvantage of FRET is that it does not report directly and specifi the interactions between biomolecules, it only measures the donor-acceptor pr FRET is preferred over other options of interaction analysis because it has very minimal restrictions and can even be used within a living cell. The major requirement is the ability of light to be delivered to and collected from the sample, but generally, only simple benchtop equipment is needed [74]. It is perfect for biomolecular interaction analysis because the majority of the biomolecules are in the nanoscale range and FRET is also viable in that same range. The nanometric distance measurements can also serve as a 'molecular gauge' for biomolecular structure analysis [82]. FRET is versatile enough to be applied in diverse areas of biomolecular research, but it is also sensitive to environmental conditions such as solvent pH, viscosity, polarity, and many others [83]. FRET analysis can be carried out either with a single FRET or the multiplexed FRET methods. Multiplexed FRET methods offer much more advantages including simultaneous analysis of multiple analytes, analysis of intermolecular and intramolecular interactions, and monitoring of coinciding biomolecular events [84].
The major disadvantage of FRET is that it does not report directly and specifically on the interactions between biomolecules, it only measures the donor-acceptor proximity and stoichiometry, hence the conclusions drawn are not strong enough without additional data or information [74].
CD
Circular dichroism (CD) is an absorption spectroscopy used to investigate optical isomerism and secondary structures of molecules by taking the difference in absorptions of the left and right lights that are circularly polarized by chiral molecules [85,86]. Chiral molecules are those that are not superimposable on their mirror images and hence exhibit optical activity as an effect [87]. Upon the passage of light through a chromophore solution, the light may either be absorbed or refracted. Absorption is quantified by the molar extinction coefficient, 'epsilon'. Molecules that are active optically have unique molar extinction coefficients for the two different circularly polarized lights. The deviation between the absorbance of the two different circularly polarized lights can be represented by a constant described by the Lambert-Beer law as delta A. The difference between the delta A of the two components or molecules is the measure of the Circular Dichroism. Numerous articles have extensively explained the calculations involved [88]. Electronic CD is produced mainly by molecules whose chromophores can absorb light in the ultraviolet (UV) and also the visible spectral territories and is used to study charge transfer transitions in metal-protein complexes. Vibrational CD, on the other hand, is generated in the infrared (IR) spectral region and it is useful in the analysis of the structure of organic molecules of relatively small size, such as proteins and DNA [86,87]. Figure 5 is a schematic representation of the arrangement in a CD equipment.
CD
Circular dichroism (CD) is an absorption spectroscopy used to investigate optical isomerism and secondary structures of molecules by taking the difference in absorptions of the left and right lights that are circularly polarized by chiral molecules [85,86]. Chiral molecules are those that are not superimposable on their mirror images and hence exhibit optical activity as an effect [87]. Upon the passage of light through a chromophore solution, the light may either be absorbed or refracted. Absorption is quantified by the molar extinction coefficient, 'epsilon'. Molecules that are active optically have unique molar extinction coefficients for the two different circularly polarized lights. The deviation between the absorbance of the two different circularly polarized lights can be represented by a constant described by the Lambert-Beer law as delta A. The difference between the delta A of the two components or molecules is the measure of the Circular Dichroism. Numerous articles have extensively explained the calculations involved [88]. Electronic CD is produced mainly by molecules whose chromophores can absorb light in the ultraviolet (UV) and also the visible spectral territories and is used to study charge transfer transitions in metal-protein complexes. Vibrational CD, on the other hand, is generated in the infrared (IR) spectral region and it is useful in the analysis of the structure of organic molecules of relatively small size, such as proteins and DNA [86,87]. Figure 5 is a schematic representation of the arrangement in a CD equipment. Applications of CD in Biomolecular studies are vast, but it is mostly used for the comparison and characterization of protein secondary structures. It provides an efficient method to study the effect a mutation or change in environmental conditions of the protein may have on the overall structure [90]. CD can also be applied in the analysis of the interaction between molecules such as DNA and DNA-binding ligands. Many ligands that can bind to DNA are not chiral, and hence, are not active optically. However, when they interact with DNA, an induced CD (ICD) signal can be achieved by the ligand by virtue of the joining of the moments of electric transition of the ligand to that of the bases of the DNA. When ICD signals are observed within the absorption bands of the non-chiral ligand, it is a clear indication of binding between the ligand and the DNA [91]. Even though CD provides much lower resolution than other analysis methods such as X-ray crystallography and nuclear magnetic resonance, it has certain advantages that cannot be overlooked. Analysis can be very rapid and inexpensive, only small amounts of sample are required; CD is not affected by the molecule's molecular weight [91,92].
Beyond the advantages of CD, it has a few limitations. For example, specifying the ideal parameters necessary for great CD results in the instrument or experimental procedure is quite challenging and the data obtained are difficult to interpret or make sense of [87]. Applications of CD in Biomolecular studies are vast, but it is mostly used for the comparison and characterization of protein secondary structures. It provides an efficient method to study the effect a mutation or change in environmental conditions of the protein may have on the overall structure [90]. CD can also be applied in the analysis of the interaction between molecules such as DNA and DNA-binding ligands. Many ligands that can bind to DNA are not chiral, and hence, are not active optically. However, when they interact with DNA, an induced CD (ICD) signal can be achieved by the ligand by virtue of the joining of the moments of electric transition of the ligand to that of the bases of the DNA. When ICD signals are observed within the absorption bands of the non-chiral ligand, it is a clear indication of binding between the ligand and the DNA [91]. Even though CD provides much lower resolution than other analysis methods such as X-ray crystallography and nuclear magnetic resonance, it has certain advantages that cannot be overlooked. Analysis can be very rapid and inexpensive, only small amounts of sample are required; CD is not affected by the molecule's molecular weight [91,92].
Beyond the advantages of CD, it has a few limitations. For example, specifying the ideal parameters necessary for great CD results in the instrument or experimental procedure is quite challenging and the data obtained are difficult to interpret or make sense of [87].
FC
Flow cytometry (FC) is a rapid detection and characterization technique used for biomolecules in a salt-dominated solution as they flow through either single or multiple lasers [93]. The word cytometry in the name literally means "cell measurement", as it was originally designed to measure mammalian cells suspended in a flowing stream [94]. FC is able to provide information on the cell number, type, cell physiology, cell viability, susceptibility, genetic identity, and important metabolic parameters on the level of a single cell, and even whole eukaryotic cells across large populations [95]. A flow cytometer basically comprises a source of light, an optical bench, a fluidic system, electronics, and a computer to control the equipment [96]. The sample flows in single file by the action of isotonic sheath fluid in the fluid system and is exposed to a light source or sources. Light signals generated by the light sources are guided by the optical system towards photodetectors, which then transform the light into electronic signals that are stored for later analysis. Because the fluidic system is in the middle of the cytometer, the cell streams are centrally placed, ensuring that the brightness of all the cells is similar. This way, any variation in the value of signals emitted from the cells will reflect actual biological differences [97]. The illumination process produces both fluorescent and non-fluorescent signals. These signals are analyzed by optically joining the signal to a system of detection, which is made up of filters that are linked to a photodetector. The photodetectors' number and configuration permit the concurrent evaluation of many different parameters for a given cell. The electronics part of the cytometer provides a system that converts the analog light signals coming through the photodetectors to digital signals that can be read and stored in the computer [97]. Most flow cytometers available for commercial use have a principal laser, which is an argon-ion laser set at 488 nm. Modern lasers at different wavelengths comprising ultraviolet (350 nm), red (635 nm), violet (405 nm), blue (488 nm), yellow (560 nm), and green (532 nm) allowing the instantaneous use of several fluorophores, with varying excitation needs, are becoming common as well [93,98,99].
Flow cytometers could be either imaging flow cytometers (IFC), which combine the conventional FC and fluorescence microscopy for sample morphology analysis, along with multi-parameter fluorescence [100], or mass cytometers, which integrate time-of-flight mass spectroscopy with FC [101,102]. The advantages of FC that make them so attractive to biosensing are the facts that they are rapid, they can probe a huge number of cells (up to 10 6 -10 8 cells per sample), they can measure fluorescence intensity quantitatively, they can identify pathogens in complex matrices such as food without target enrichment or isolation [103,104]. There are certain limitations of FC that impede the full-scale use of the technique in biomolecular assays; the samples need to be in a single-cell suspension, it is difficult to find the right combinations of antibodies and fluorophores with minimal spectral overlapping, and extra care is needed in the interpretation of FC data. Figure 6 shows a graphical representation of how flow cytometers function. (1), the fluid containing the cell sam ple is introduced at the center of the sheath fluid stream. To prevent mixing, these two fluids main tain a significant difference in velocity. This setup allows the cells to align in a single line, a proces called hydrodynamic focusing. The aligned cells then pass through a laser and a series of detector (2) that measure cell size using forward scatter (FSC), cell complexity using side scatter (SSC), an fluorescence. Before exiting the flow cell as individual droplets (3), the cells are selectively charged with electricity. Electromagnets (4) divert the droplets containing the targeted cells with a charg away from the main stream, guiding them into collection tubes positioned on the side (5). On th other hand, cells without a charge simply fall directly into a waste collection container [105]. Repro duced with permission from Bleichrodt and Read (2019), © Elsevier, 2019.
Electrochemical Sensors
Electrochemical biosensors are a product of biological and electronic technology whereby a biological recognition element is coupled with conducting and/or semi-con ducting materials known as electrodes. Some of the biological recognition elements hav been discussed extensively in the bioaffinity nanoprobes section, including antibodies aptamers, enzymes, and other peptides. Figure 7 is a schematic representation of an elec trochemical biosensor. (1), the fluid containing the cell sample is introduced at the center of the sheath fluid stream. To prevent mixing, these two fluids maintain a significant difference in velocity. This setup allows the cells to align in a single line, a process called hydrodynamic focusing. The aligned cells then pass through a laser and a series of detectors (2) that measure cell size using forward scatter (FSC), cell complexity using side scatter (SSC), and fluorescence. Before exiting the flow cell as individual droplets (3), the cells are selectively charged with electricity. Electromagnets (4) divert the droplets containing the targeted cells with a charge away from the main stream, guiding them into collection tubes positioned on the side (5). On the other hand, cells without a charge simply fall directly into a waste collection container [105]. Reproduced with permission from Bleichrodt and Read (2019), © Elsevier, 2019.
Electrochemical Sensors
Electrochemical biosensors are a product of biological and electronic technology, whereby a biological recognition element is coupled with conducting and/or semi-conducting materials known as electrodes. Some of the biological recognition elements have been discussed extensively in the bioaffinity nanoprobes section, including antibodies, aptamers, enzymes, and other peptides. Figure 7 is a schematic representation of an electrochemical biosensor. In an electrochemical biosensor, an electrochemical method, usually involving an electrode and an electrolyte solution containing the analyte, transforms the chemical energy corresponding to the binding activity between the target and bioaffinity nanoprobe into electrical energy [107,108]. Electrochemical biosensors utilize different transduction methods which include electrochemical impedance spectroscopy (EIS), amperometry (It), and voltammetry (cyclic, differential pulse, linear sweep, square wave) [109,110]. Voltametric electrochemical biosensors have become one of the most versatile detection methods due to their lower noise tendency. These biosensors measure current in a steady potential, controlled by the working electrode, and the target concentration is obtained by observing the highest current intensity.
EIS is a frequency domain system that can measure a wide range of frequencies, providing more kinetic and structural information about the electrode interface than traditional electrochemical biosensors. In this type of biosensor, the interaction between the target and bioaffinity nanoprobe causes changes in the electric field, affecting the impedance values [110]. Electrochemical biosensors are such an attractive choice of pathogen detection technique, especially in food, because they offer a rapid, accurate, sensitive, inexpensive detection mechanism, requiring very small sample quantities. Nanomaterials and nanocomposites are commonly used to enhance the sensitivity of electrochemical biosensors. Additionally, these biosensors can be integrated with microfluidic systems to create compact and efficient devices with multiple functionalities in a single platform. Electrochemical biosensors have proven to be successful in detecting a wide range of pathogens and disease biomarkers. Their applications span research, diagnostics, therapeutics, food safety, and environmental monitoring [111][112][113].
Bekir et al. introduced a highly sensitive electrochemical immunosensor for detecting stressed and resuscitated pathogenic Staphylococcus aureus. The interaction was described by voltammetry, along with impedance spectroscopy. In the dynamic concentration span of 10 1 to 10 7 CFU/mL, an incredible linear response in addition to a low detection limit was recorded. The results were reproducible, indicating the viability of the system [114].
A label-free EIS was designed by Dong et al. based on gold nanoparticles and a poly(amidoamine)-multiwalled carbon nanotube-chitosan (AuNPs/PAMAM-MWCNT-Chi) nanocomposite film-altered glass carbon electrode for detecting Salmonella typhimurium. Bacteria in the linear range of 10 3 to 10 7 CFU/mL were recognized by the sensor, recording a limit of detection (LOD) of 5.0 × 10 2 CFU/mL [115]. In an electrochemical biosensor, an electrochemical method, usually involving an electrode and an electrolyte solution containing the analyte, transforms the chemical energy corresponding to the binding activity between the target and bioaffinity nanoprobe into electrical energy [107,108]. Electrochemical biosensors utilize different transduction methods which include electrochemical impedance spectroscopy (EIS), amperometry (I-t), and voltammetry (cyclic, differential pulse, linear sweep, square wave) [109,110]. Voltametric electrochemical biosensors have become one of the most versatile detection methods due to their lower noise tendency. These biosensors measure current in a steady potential, controlled by the working electrode, and the target concentration is obtained by observing the highest current intensity.
EIS is a frequency domain system that can measure a wide range of frequencies, providing more kinetic and structural information about the electrode interface than traditional electrochemical biosensors. In this type of biosensor, the interaction between the target and bioaffinity nanoprobe causes changes in the electric field, affecting the impedance values [110]. Electrochemical biosensors are such an attractive choice of pathogen detection technique, especially in food, because they offer a rapid, accurate, sensitive, inexpensive detection mechanism, requiring very small sample quantities. Nanomaterials and nanocomposites are commonly used to enhance the sensitivity of electrochemical biosensors. Additionally, these biosensors can be integrated with microfluidic systems to create compact and efficient devices with multiple functionalities in a single platform. Electrochemical biosensors have proven to be successful in detecting a wide range of pathogens and disease biomarkers. Their applications span research, diagnostics, therapeutics, food safety, and environmental monitoring [111][112][113].
Bekir et al. introduced a highly sensitive electrochemical immunosensor for detecting stressed and resuscitated pathogenic Staphylococcus aureus. The interaction was described by voltammetry, along with impedance spectroscopy. In the dynamic concentration span of 10 1 to 10 7 CFU/mL, an incredible linear response in addition to a low detection limit was recorded. The results were reproducible, indicating the viability of the system [114].
A label-free EIS was designed by Dong et al. based on gold nanoparticles and a poly(amidoamine)-multiwalled carbon nanotube-chitosan (AuNPs/PAMAM-MWCNT-Chi) nanocomposite film-altered glass carbon electrode for detecting Salmonella typhimurium. Bacteria in the linear range of 10 3 to 10 7 CFU/mL were recognized by the sensor, recording a limit of detection (LOD) of 5.0 × 10 2 CFU/mL [115].
In 2018, a technique was presented by Helali et al. for detecting Escherichia coli in chicken by EIS and SPR imaging techniques. The detection limit obtained was 10 3 CFU/mL [116].
Shimaa et al. also documented a new electrochemical biosensor for the concurrent detection of Listeria monocytogenes, as well as Staphylococcus aureus. They recorded outstanding sensitivities with LODs of 9 CFU/mL in the case of Listeria monocytogenes and 3 CFU/mL for Staphylococcus aureus [117].
Colorimetric Sensors
Colorimetry is the quantification of ultraviolet-visible (UV-vis) light that is being absorbed or reflected by a medium [118]. Colorimetric sensors are described as a class of optical sensors (sensors that use light in the infrared, visible, or ultraviolet region to analyze chemical or biological interactions), that show a single, double, or multiple change of color when a target molecule is recognized. They are easy to use, portable, cheap, and offer sensitive and selective on-site or in situ applications [119]. Colorimetric biosensors can be used for the detection of a specific analyte in a liquid sample through color changes that occur as a result of the interactions between the target and the bioaffinity nanoprobe, usually with the assistance of a color reagent, and this change in color is observable with the human eye or with very simple, compact optical detectors for quantitative analysis [120].
A colorimetric sensor consists of a source of light, a device for the selection of wavelengths such as filters or monochromators, a cell in which variations in the light absorbed or emitted in the presence of the target molecule can happen, and a sensitive detector [121]. Different types of colorimetric assays have been developed over the years for the application of pathogen detection and these include loop-mediated isothermal amplification (LAMP), polymerized polydiacetylene, gene expression reaction, and so on [120]. Several colorimetric sensors rely on the traditional three-channel visible range, which corresponds to the wavelength ranges partially overlapping with red, green, and blue. The use of many different channels with a smaller spectral range for every one of them is referred to as hyperspectral imaging. This approach can also be employed in colorimetric sensors. Colorimetric sensors can incorporate a broad range of wavelengths, including non-visible wavelengths, starting from near-infrared to ultraviolet, by using hundreds of color channels. This is known as full spectrophotometry.
In order to make the data analysis and instrumentation easier, the analysis of spectra is performed mostly at only a few discrete wavelengths or just by choosing the maximum points in the UV-vis spectra [121]. Figure 8 is a graphical representation of a colorimetric assay for the detection of staphylococcus aureus.
In 2018, a technique was presented by Helali et al. for detecting Escherichia coli in chicken by EIS and SPR imaging techniques. The detection limit obtained was 10 3 CFU/mL [116].
Shimaa et al. also documented a new electrochemical biosensor for the concurrent detection of Listeria monocytogenes, as well as Staphylococcus aureus. They recorded outstanding sensitivities with LODs of 9 CFU/mL in the case of Listeria monocytogenes and 3 CFU/mL for Staphylococcus aureus [117].
Colorimetric Sensors
Colorimetry is the quantification of ultraviolet-visible (UV-vis) light that is being absorbed or reflected by a medium [118]. Colorimetric sensors are described as a class of optical sensors (sensors that use light in the infrared, visible, or ultraviolet region to analyze chemical or biological interactions), that show a single, double, or multiple change of color when a target molecule is recognized. They are easy to use, portable, cheap, and offer sensitive and selective on-site or in situ applications [119]. Colorimetric biosensors can be used for the detection of a specific analyte in a liquid sample through color changes that occur as a result of the interactions between the target and the bioaffinity nanoprobe, usually with the assistance of a color reagent, and this change in color is observable with the human eye or with very simple, compact optical detectors for quantitative analysis [120].
A colorimetric sensor consists of a source of light, a device for the selection of wavelengths such as filters or monochromators, a cell in which variations in the light absorbed or emitted in the presence of the target molecule can happen, and a sensitive detector [121]. Different types of colorimetric assays have been developed over the years for the application of pathogen detection and these include loop-mediated isothermal amplification (LAMP), polymerized polydiacetylene, gene expression reaction, and so on [120]. Several colorimetric sensors rely on the traditional three-channel visible range, which corresponds to the wavelength ranges partially overlapping with red, green, and blue. The use of many different channels with a smaller spectral range for every one of them is referred to as hyperspectral imaging. This approach can also be employed in colorimetric sensors. Colorimetric sensors can incorporate a broad range of wavelengths, including non-visible wavelengths, starting from near-infrared to ultraviolet, by using hundreds of color channels. This is known as full spectrophotometry.
In order to make the data analysis and instrumentation easier, the analysis of spectra is performed mostly at only a few discrete wavelengths or just by choosing the maximum points in the UV-vis spectra [121]. Figure 8 is a graphical representation of a colorimetric assay for the detection of staphylococcus aureus. . Schematic illustration of the staphylococcus aureus (SA) detection process involving a highthroughput colorimetric biosensor using aptamers and the photocatalytic activity of the dsDNA-SG I complex. This process begins with the coating of a 96-well plate with streptavidin. To prevent non-specific adsorption, bovine serum albumin (BSA) is used. Following this, a biotin-labeled capture probe is anchored to the plate surface via the streptavidin-biotin interaction. The SA-specific aptamer is then immobilized onto the 96-well plate via hybridization with the capture probe. When SA is present, the aptamer disengages from the capture probe-aptamer double strand due to a stronger interaction with SA. The resulting single-strand capture probe can hybridize with a DNA nanostructure, a three-way junction (TWJ), which consists of three detection probes (P1, P2, P3). Upon the addition of SG I, a dsDNA-SG I complex forms and catalyzes the oxidation of 3,3 ,5,5tetramethylbenzidine (TMB) under LED photo-irradiation. The intensity of the resulting catalytic color is directly related to the number of bacteria present [122]. Reproduced with permission from Yu et al. (2020), © Springer Nature, 2020 (Open access).
The major limitation of simple colorimetric sensors is low sensitivity, as it is difficult to transform detectable signals into specific color readouts. To overcome this limitation, a couple of nanomaterials such as graphene oxide (GO), gold nanoparticles (AuNPs), magnetic NPs, carbon nanotubes (CNTs), conjugated polymers, and cerium oxide NPs, have been developed and incorporated into the colorimetric assays [123].
Zhang et al. created a rapid, specific colorimetric biosensor for detecting Listeria monocytogenes, by using a vancomycin-conjugated, Fe 3 O 4 NP cluster-improved aptamer as the bioaffinity nanoprobe. The system was a success, with a wide linear range given as 5.4 × 10 3 -10 8 CFU/mL and a 5.4 × 10 3 CFU/mL visible detection limit [124].
A specific, rapid, colorimetric aptasensing method for detecting Salmonella (S.) typhimurium was designed by Yuan et al. The sensitivity reached 7 CFU/mL using the human eye. The system could be adjusted for the concurrent recognition of S. Typhimurium and other foodborne pathogens [125].
Ren et al. described the development of a lateral flow sensor that utilizes plasmonic enhancement to significantly increase the colorimetric signal. The sensor relies on liposomeencapsulated reagents that induce the aggregation of gold nanoparticles (AuNPs). The procedure optimized the performance of the system for detecting Escherichia coli O157:H7 and made it better by 1000-fold. This led to a sensitivity of 600 CFU/mL with the naked eye in apple juice [126].
Optical Sensors
Optical sensors quantify the interaction between a receptor and an analyte by assessing a specific aspect of the reaction as an observable optical signal [127]. The majority of optical sensors measure changes in the sensor's surface properties when the analyte binds to the sensing layer through adsorption or complex formation [127]. Optical biosensors combine biological selectivity with modern micro-and optoelectronics, finding applications in areas such as food safety, therapeutics, and environmental monitoring [128]. There are various types of optical sensors, including colorimetric, chemiluminescence, Fourier-transform infrared (FTIR) spectroscopy, matrix-assisted laser desorption ionization time-of-flight mass spectroscopy (MALDI-TOF-MS), fluorescence, surface plasmon resonance (SPR), Raman spectroscopy, and evanescent field optical fiber [127,129]. Optical sensors are preferred for foodborne pathogen detection due to their ability to detect targets in complex food matrices with minimal sample treatment. They offer high sensitivity and specificity, ease of use, cost-effectiveness, label-free detection, compactness, and minimal invasiveness [128].
Masdor et al. developed three distinct immunoassays, that is, direct, sandwich with gold nanoparticles (AuNPs) and sandwich for the detection of Campylobacter (C.) jejuni on the SPR equipment. In the direct analysis, the polyclonal antibody against C. jejuni was initially attached to the surface to serve as the capturing antibody. C. jejuni cells in different concentrations were subsequently introduced to the ligand, and the resulting response from the interaction was documented in response units (RU). The maximum response was observed at a concentration of 1 × 10 9 CFU/mL and a response of 144.34 RU. The determined limit of detection (LOD) value was 8 × 10 6 CFU/mL. In the sandwich assay, a capture antibody and a mouse control antibody were used. The greatest response was achieved at a concentration of 1 × 10 9 CFU/mL, with a binding response of 131.5 RU. The calculated limit of detection (LOD) value was 4 × 10 4 CFU/mL, and a strong coefficient of correlation of 0.997 was observed. This represents a notable improvement compared to the previous direct format, which had an LOD of 8 × 10 6 CFU/mL. In the case of the sandwich assay with amplification of signal using AuNPs, the antibody-linked AuNPs were introduced over the detected bacteria, which increased the refractive index and, in turn, enhanced the binding response. The maximum response was observed at a concentration of 1 × 10 9 CFU/mL, with an interaction response of 96.6 RU. The calculated LOD was 8 × 10 5 CFU/mL, and a satisfactory coefficient of correlation of 0.998 was noted. The sandwich assay outperformed the others, while the direct assay was the least effective. Comprehensive cross-reactivity studies against various foodborne pathogens revealed minimal non-specific binding, making this assay even more specific than the other available methods [130].
Sanati and colleagues used an asymmetric, Vernier-type double-stage ring resonator (DSRR) integrated with a plasmonic slot waveguide for the identification of Escherichia (E.) coli K12 bacteria in potable water. The efficiency of the sensor was evaluated across a range of liquid environments, and the capacity of the DSRR sensor for the label-free identification of E. coli K12 at visible wavelengths was established. The suggested sensor delivers a high sensitivity value of 480 nm/RIU and an impressively low detection limit reaching down to 3.33 × 10 −5 RIU. This makes the sensor a strong contender for swift and high-definition identification of E. coli bacteria in food items [131].
Kim and colleagues developed a Salmonella sensing platform utilizing retroreflective Janus microparticles (RJP) along with a simple optical system. In contrast to traditional fluorescence-based Salmonella detection methods, the RJP-based platform does not necessitate intricate optical tools, as RJPs can be visualized using a CMOS camera and a standard white LED. The system allows for highly sensitive and quantifiable detection of Salmonella. Moreover, the system exhibited high selectivity for invA by employing oligonucleotides with mismatched sequences. The invA gene encodes a protein that facilitates Salmonella invasion via a type 3 secretion system. Utilizing this system, concentrations of Salmonella varying from 0 to 100 nM were scrutinized with exceptional selectivity and sensitivity, achieving a detection limit of 2.48 pM [132].
Piezoelectric Sensors
Piezoelectric sensors are mass-sensitive sensors that are able to detect targets or analytes using a transduction mechanism that depends on small changes in mass. The technique employed for pathogen detection in this approach is contingent on mass evaluation via piezoelectric crystals. These crystals have the capacity to vibrate at a specific frequency when subjected to an electrical signal of a corresponding frequency. As a result, the vibration frequency is determined by both the crystal's mass and the frequency of the electrical signal applied [133,134]. In the case of foodborne pathogen detection, when the mass increases due to the interaction with target pathogens, the crystal's oscillation frequency changes, and the resulting shift can be measured electrically. This measurement is then used to calculate the additional crystal mass [135]. The two primary categories of mass-sensitive biosensors include surface acoustic wave devices and quartz crystal microbalance devices, which are also known as bulk wave devices [133]. Piezoelectric biosensors offer advantages such as low cost, simplicity, user-friendliness, and direct label-free analysis while maintaining consistent reliability and improved sensitivity [134]. Piezoelectric biosensors utilizing quartz crystal microbalance being the most common type have been customized with various antibodies and other bioreceptors for the recognition of foodborne and waterborne pathogens. These include Salmonella, Escherichia coli, protozoa, Shigella, influenza A and B viruses, Campylobacter, Yersinia, and Vibrio [136].
In a study by Lian and colleagues, they engineered an innovative sensor that integrates graphene, an aptamer, and interdigitated gold electrode (IDE) for the rapid and targeted recognition of Staphylococcus aureus (S. aureus). The biological recognition element in this process is the S. aureus aptamer. A compound known as 4-Mercaptobenzene-diazonium tetrafluoroborate (MBDT) salt served as the molecular bridge, chemically binding graphene to the IDE. These electrodes were, in turn, linked to a series electrode piezoelectric quartz crystal (SPQC). The S. aureus aptamers were then affixed onto the graphene via π-π stacking of DNA bases. When S. aureus is present, it specifically binds to the aptamer, leading to an interaction of the DNA bases with the aptamer and its subsequent release from the graphene surface. This action modifies the electrical characteristics of the electrode surface, thereby leading to a shift in the SPQC's oscillator frequency. The detection process takes only 60 min to complete. The sensor displayed a proportional correlation between shifts in resonance frequency and the range of bacterial concentrations from 4.1 × 10 1 to 4.1 × 10 5 cfu/mL. Moreover, it demonstrated a sensitivity with a lower detection limit at 41 cfu/mL [137]. Sharma et al. were able to detect Listeria monocytogenes (LM), an infectious bacterium, at the infection dose threshold of 10 3 /mL within an hour in both a buffer solution and milk. This was achieved using a unique asymmetrically anchored cantilever sensor and a commercially procured antibody. To validate the responses of the sensor, a secondary antibody-binding phase, akin to sandwich ELISA tests, was utilized for signal boost and the minimization of false negatives. Through the incorporation of a tertiary antibody-binding phase, the team was successful in detecting LM at concentrations as low as 10 2 /mL, a level significantly below the infection dose (<1000 cells) for LM [138]. Table 2 summarizes more of the applications of nanoprobes, taking into consideration the sensor types, targets, limits of detection, and samples tested.
Newer Technologies-Microfluidic Detection Methods
The cutting-edge approach to pathogen detection utilizes compact, integrated biosensing technologies, delivering dependable, sensitive, economical, and quick detection without the necessity for intricate equipment. Microfluidics is a versatile platform engineered for the streamlining, consolidating, and miniaturizing of devices, making it particularly wellsuited for electrochemical, biomedical, and biochemical applications. It is the basis of point-of-care (POC) detection, of which paper-based and lab-on-chips (LOC) are the most outstanding technologies. Several applications of LOCs or microfluidics in foodborne pathogen detection have been covered extensively in the literature. Sun et al. developed a micro-spot paper-based analytical device (µPADs) by the combination of a PVC pad and filter paper. The detection method relied on the observation of a color shift (from colorless to indigo) upon the interaction of a unique enzyme linked to the Cronobacter spp. under examination with a chromogenic substrate. By fine-tuning the enrichment steps, the technique permits an analysis duration of 10 h or fewer and is able to identify living bacteria on the injected sample surface in concentrations as low as 10 1 CFU/cm 2 . This work showed that the production technique is innovative, straightforward, highly reproducible (having an RSD below 5%), and inexpensive (below $0.15 for every micro-spot) [149].
A colorimetric paper-based analytical device (PAD) combined with immunomagnetic separation (IMS) was developed for recognizing Salmonella (S.) typhimurium by Srisa-Art et al. IMS utilized coated anti-Salmonella magnetic beads to detect, separate, and preconcentrate bacteria from samples before testing on paper. A sandwich antibody-based assay was integrated into the process, employing β-galactosidase (β-gal) to be the enzyme of detection for direct S. Typhimurium detection after IMS. The antibody and enzyme complex enabled a colorimetric assay using chlorophenol red-β-d-galactopyranoside (CPRG) for bacteria detection. The procedure showed high specificity to S. Typhimurium with no interference from other pathogens such as E. coli. Without pre-enrichment, the detection limit of S. Typhimurium in culture solution was found to be 10 2 CFU/mL. The developed system was put to use to identify S. Typhimurium in fecal samples from starlings that had been inoculated, as well as in whole milk. The system showed detection thresholds of 10 5 CFU/g in the bird feces and 10 3 CFU/mL in the milk. Notably, this represents the first documented use of a paper-based technique for detecting S. Typhimurium in such samples [150].
Smartphones have become valuable and readily available tools for diagnostics, removing the need for costly signal readers. Combining biosensing technology and digital communication systems, these devices offer immense potential for detecting pathogens in various areas, such as water, food, plant nurseries, medical, environmental, and wastewater. The data collected from the analysis can be easily stored, compared, and transferred between systems, making smartphones an efficient and affordable solution for diagnostics [151].
Cheng et al. reported a nanosensor that employed platinum-palladium (Pt-Pd) nanoparticles as signal boosters in a dual lateral flow immunoassay (LFIA) system, which was combined with a device based on smartphones, for the concurrent detection of Salmonella Enteritidis and Escherichia coli O157:H7. Following optimization, the detection limits were found to be around 20 CFU/mL for Salmonella Enteritidis and roughly 34 CFU/mL for E. coli O157:H7. The recovery rates for the dual LFIA method ranged from 91.44% to 117.00%, indicating its effectiveness in identifying live bacteria present in food samples [152].
Jung and colleagues also utilized the high-resolution camera, steady source of light, and computational aptitude of a smartphone to devise a method that objectively and accurately determines bacterial cell concentrations in food samples, using a regression model based on the intensity of the color of the test lines. They designed a 3D-printed sample container compatible with standard lateral flow assays and developed a custom Android app to extract cell concentration data from color intensity measurements. Tests using Escherichia coli O157:H7 as a representative organism showed that the smartphonebased procedure could detect concentrations between 10 4 and 10 5 CFU/mL in both spinach and ground beef samples [153]. Many examples of the applications of these newer detection techniques have been mentioned extensively in various articles [151,[154][155][156].
Future Perspective
It is a fact that a broad range of bioaffinity nanoprobes have been selected or produced for foodborne pathogen detection and with that, many modes of analysis for the success or progress of these nanoprobes have been reported or enhanced. These bioaffinity nanoprobes, especially aptamers, have made great strides in research into biosensing, diagnostics, and therapeutics but have gained very little success in commercialization. The future for bioaffinity nanoprobes, in general, is focusing on modifying them to meet these criteria: cost-effectiveness, accuracy and precision, sensitivity and selectivity, as well as operation [157].
In the context of developing biosensors for the recognition of foodborne pathogens, it is essential to ensure that biosensors have the capability to specifically detect the target pathogen and provide explicit results that give the analyst or user certainty or confidence in the results, as pathogens in food are usually in trace amounts. This depends largely on the selectivity, sensitivity, and specificity of the bioaffinity nanoprobes. Many nanomaterials have been incorporated into biosensing to assist with this aspect, and many more will be developed in the future. Given the complexity of food structures, there is a need to look into the development of novel bioaffinity probes and the enhancement of the existing options to achieve the ultimate goal of high sensitivity and efficiency in the detection (LOD < 10 2 CFU/mL) of pathogens even with the ever-changing trends in food processing, distribution, and consumption.
One of the key challenges of foodborne pathogen detection methods that currently exist is the need for specific sample preparation protocols which require sample purification and enrichment prior to the analysis. To be able to use biosensors effectively for rapid, pointof-care, or in situ applications, there is a need to develop analysis methods that can function with extremely small sample quantities and minimal sample preparation. The microfluidic chip technology has been a great innovation in miniaturized detection systems, offering the advantages of consumption of minimal samples and reagents, simultaneous analysis, controllable liquid flow, and an incredibly decreased analysis time. This technology, when improved and incorporated into detection techniques, will be helpful in the future [157]. Another reason why the commercialization of systems of analysis utilizing bioaffinity nanoprobes is lagging is the cost associated with the development of the sensors and the display platforms.
Paper-based biosensors have been introduced as the alternative to traditional biosensors because they are cheap, portable, and simple to use. This is very good for in situ pathogen detection even in developing countries that have limited resources. Smartphones, as display platforms for the analysis of detection results, have also been suggested by researchers for the sensing of foodborne pathogens. Given the portability, high camera quality, and availability of smartphones, they will be an effective tool if incorporated into biosensing on a larger scale together with cheaper sensing techniques such as paper-based biosensors.
Multimodal detection, offering a promising strategy for the comprehensive and reliable identification of foodborne pathogens, is a vital future research endeavor. By combining multiple sensing modalities, such as optical, electrochemical, and molecular techniques, a synergistic effect can be achieved, leading to enhanced sensitivity and specificity. For instance, a multimodal biosensor can integrate optical detection methods, such as surface plasmon resonance or fluorescence, with electrochemical transduction for simultaneous measurement of multiple target analytes. This multimodal approach enables the detection of pathogens through different recognition mechanisms, increasing the likelihood of accurate identification even in complex food matrices. Moreover, the combination of different techniques can provide complementary information, allowing for improved discrimination between specific pathogens and reducing false positives. Multimodal detection systems hold great potential in advancing food safety measures by offering robust, rapid, and accurate pathogen detection in a single integrated platform.
Conclusions
Ensuring the safety of food is of utmost importance, considering its significance to human existence and quality of life. Over time, pathogen detection in food has evolved from conventional methods such as culture-based techniques to more advanced approaches including PCR, ELISA, and antibody-based biosensors. The introduction of bioaffinity nanoprobes, particularly aptamers, has revolutionized biosensing due to their high sensitivity. The development of bioaffinity nanoprobes has also led to the emergence of technologies aimed at evaluating and analyzing their efficiency. Techniques such as SPR, FRET, and CD have proven valuable in assessing the performance of bioaffinity nanoprobes. Various biosensing technologies such as optical, colorimetric, and MIPs have emerged to detect foodborne pathogens and ensure food safety. Ongoing research and development in these fields aims to enhance their performance, sensitivity, and specificity, enabling more effective monitoring and ensuring the safety of our food supply. Advancements in biosensor technologies continue to play a crucial role in addressing the challenges associated with food safety and pathogen detection. | 2023-05-29T15:03:36.226Z | 2023-05-26T00:00:00.000 | {
"year": 2023,
"sha1": "d61806077cb32fa684399d4396855c50bc9598e1",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2072-666X/14/6/1122/pdf?version=1685110359",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "c1ca2cd13a99ddc5e14c2efb24f005cab012a72c",
"s2fieldsofstudy": [
"Biology",
"Chemistry"
],
"extfieldsofstudy": []
} |
35333511 | pes2o/s2orc | v3-fos-license | Development of the Safety Attitudes Questionnaire-Korean Version (SAQ-K) and its Novel Analysis Methods for Safety Managers
As the importance of patient safety has been widely acknowledged, numerous improvement programs have been designed and implemented internationally. 1 These programs aimed not only to solve specific problems such as central line-associated bloodstream infection and patient falls, but also to nurture organizational and individual health care workers’ (HCWs’) safety culture.2 Although safety culture has been defined in various ways, the core idea involves “the collective beliefs and perceptions of workers regarding the organization and safety of their workplace operations”.3 Safety culture has served as the foundation of safety improvement endeavors.4 Thus, there is high demand for instruments to measure safety culture among HCWs in hospitals.
Introduction
As the importance of patient safety has been widely acknowledged, numerous improvement programs have been designed and implemented internationally. 1 These programs aimed not only to solve specific problems such as central line-associated bloodstream infection and patient falls, but also to nurture organizational and individual health care workers' (HCWs') safety culture. 2 Although safety culture has been defined in various ways, the core idea involves "the collective beliefs and perceptions of workers regarding the organization and safety of their workplace operations". 3 Safety culture has served as the foundation of safety improvement endeavors. 4 Thus, there is high demand for instruments to measure safety culture among HCWs in hospitals.
In response to the demand, various tools have been developed and tested, [5][6][7] and one of the most embraced and thoroughly validated instruments is the Safety Attitudes Questionnaire (SAQ), which is composed of six domains 8 (Table 1).
Since its inception, SAQ and its variants have been adopted in diverse health care settings, including intensive care units, ambulatory care, emergency departments, operating rooms, and pharmacies. 3,[8][9][10] Also, SAQ has been translated into various languages and validated; therefore, it has provided benchmarks not only for healthcare organizations but also for countries. 10,[11][12][13][14][15] In addition, SAQ has been used to examine the effectiveness of patient safety programs. 2,16,17 Unfortunately, however, despite huge improvements in patient safety, Korea does not have a validated safety culture measurement tool that can be applied nationally, although some healthcare organizations have developed their own instrument. The lack of such tools might have slowed down improvement in patient safety in hospitals across the country. In addition to the dire need for a validated nationwide tool to measure safety culture, more advanced analytical methods of the collected safety culture data are also required. To date, most analyses of SAQ surveys have been conducted by simply calculating the mean and standard deviation of SAQ domain scores 14,18,19 or the percentage of respondents reporting positive perceptions of SAQ domains for each cluster, most frequently a clinical area, 15,20 and then examining patterns across clinical areas. 19 From a statistical standpoint, this traditional approach leaves huge room for improvement, although its simplicity has attracted hospital safety managers with limited knowledge of statistics.
The most frequently omitted step is testing whether SAQ scores are statistically different across clusters (i.e., testing the effect of clusters in explaining the variance in SAQ before calculating summary statistics like the mean for each cluster). Only if there is a difference in SAQ scores across clusters should cluster-specific statistics be estimated; otherwise, the overall SAQ score (pooled estimate) for the entire hospital is sufficient. For ordinary cases, analysis of variance (ANOVA) can be used, but for a highly unbalanced dataset with unequal variances, such as SAQ scores over clinical areas, the assumptions of ANOVA are easily violated, 21 which requires seeking a different approach.
Another issue from the above-mentioned traditional approach arises from many work areas having too small a number of HCWs to provide meaningful information about the precision of clinical area-specific statistics. Compared to hospitals in the US, hospitals in Korea are understaffed, and a lot of clinical areas even have fewer than five HCWs (e.g., outpatient clinics in a hospital). For these small areas, most dispersion measures, such as standard deviation, inter quartile range, or median absolute deviation, cannot be easily calculated, and even if calculated they do not provide much information.
Therefore, we examined methodologies that can be applied to evade the aforementioned issues. First, to test for differences in SAQ scores by cluster, we used the variance components model, which is designed to estimate such between-cluster variances (i.e., within-cluster correlations). 22 Meanwhile, the empirical Bayesian (EB) method was used to obtain a more accurate estimate for each cluster by allowing small clusters to borrow information from large clusters. [22][23][24][25] This study was conducted with three goals: a) To develop and validate the Korean version of SAQ (SAQ-K).
b) To apply the variance components model to test whether SAQ scores vary significantly over clinical areas (i.e., whether SAQ scores differ by clinical area), which is one of the most commonly used clustering categories. 20 c) To obtain precise estimates of SAQ scores of each cluster with the empirical Bayes method.
Development and validation of SAQ-K
With permission of the original developer of SAQ, a team composed of physicians, nurses, patient safety managers, and communication experts translated SAQ into the Korean language. Not only the language difference but also structural and cultural differences in the health care sector were addressed based on the cross-cultural survey adaptation guidelines from. 26 Another researcher who did not participate in the translation, backtranslated the Korean version into English to check for unintended distortions in meaning. The Korean version was then pilot tested with various health care professionals, including but not limited to physicians, nurses, pharmacists, paramedics, and managerial personnel. Two negatively worded items (item 2: "In your clinical area, it is difficult to speak up if you perceive a problem with patient care" and item 11: "In this clinical area, it is difficult to discuss errors") of the original survey were pointed out as difficult to understand and answer properly. This might be attributed to the different language structures of English and Korean; Koreans are not accustomed to answering to negatively worded questionnaire items. Since there was no other way to properly translate these items, they were excluded, and a questionnaire composed of 34 items was developed (TC: 5 items, SC: 6 items, JS: 5 items, SR: 4 items, PM, 10 items, and WC: 4 items).
The questionnaire was administered anonymously to health care professionals working at a large metropolitan hospital in Seoul from October 2013 through November 2013. Participants were allowed to choose the modality of survey between the intranet or paper version. SAQ-K responses were measured on a 5-point Likert scale (1 = Disagree Strongly, 2 = Disagree Slightly, 3 = Neutral, 4 = Agree Slightly, 5 = Agree Strongly) and then converted into a scale of 0 to 100 with the interval of 25 as the original SAQ rubric. For each respondent, six domain scores were obtained by calculating the mean of the items in each domain. Cronbach's alpha was calculated to examine internal consistency, and confirmatory factor analysis (CFA) was conducted to test construct validity. These analyses were performed with statistical software packages, IBM SPSS Statistics 21 and AMOS 21 (SPSS Inc., Chicago Illinois).
Testing differences in SAQ-K domain scores over clinical areas
The following steps apply to each of the six SAQ-K domains. We began by modeling the SAQ-K score of a certain domain for a person (model 0). This model did not include random effects and is denoted as
Obtaining more accurate estimates of cluster-level SAQ-K scores with the empirical bayes method
The EB method was used in this study to obtain more precise estimates of the SAQ-K domain score for each clinical area. Because this method calculates the cluster estimates by pulling the ordinary mean of unit scores to the overall mean, it is also called empirical Bayes shrinkage. The amount of shrinkage for clinical area is determined by the shrinkage factor i B .
Where 2 i σ is the within-cluster variance of clinical area i . Then, the EB mean of clinical area i ,ˆi θ is denoted as Where i y is the mean of clinical area i andμ is the overall mean score of a certain domain of SAQ-K. Therefore, if clinical area i has a wide variance and the shrinkage factor is large, then the EB estimate of the area moves closer to the overall mean accordingly. This shrinkage is prominent usually for the clusters with a small sample size.
Characteristics of respondents
A total of 1,381 questionnaires were returned. After excluding those with too many missing values, especially in clinical area variables, the main interest of this study, 1,142 questionnaires were analyzed. As depicted in (Table 2), most of the survey respondents (73.7%) were female, which is understandable in that most nurses in Korea are female and the number of nurses is much larger than the number of physicians. Health care professionals with 5 to 10 years of experience (25.4%) formed the largest group and were followed by those with 3 to 4 years (21.8%). The tails on both sides of work years accounted for the smallest proportion at 6.7% for respondents with less than 6 months of experience and 5.3% for those with 21 years or more. For job type, nurses constituted the majority at 53.3%, followed by physicians at 33.1%. Supporting staff was 11.6% of the respondents and there was a small number of pharmacists, administrators, and uncategorized HCWs.
5* 3
To examine the dispersion of HCWs by clinical area, we summarized the data in a stem-and-leaf plot, as seen in Figure 1. The numbers of HCWs were heavily dispersed over 72 clinical areas, ranging from 2 to 53 HCWs; many of the areas had fewer than 10 HCWs working in them. Specific names of clinical areas were not of interest in this study and are therefore not listed in this article.
Psychometric properties of SAQ-K
Internal consistency was evaluated with Cronbach's alpha, 27 a coefficient of equivalence among multiple items measuring the same construct. The alphas for SAQ-K were 0.836 for TC, 0.841 for SC, 0.907 for JS, 0.734 for SR, 0.928 for PM, and 0.758 for WC. Since alphas greater than .70 is generally considered to be large enough, 27,28 this suggests that SAQ-K shows high reliability.
The results and fit indices of CFA are provided in (Table 3 & 4). To help readers better understand the CFA results, items in (Table 3) are shown in English as the original SAQ, although actual survey administration and analysis used SAQ-K. GFI, NFI, and CFI were higher than 0.9, AGFI was 0.894, and RMSEA was less than 0.05, indicating a good model fit based on commonly used criteria. 29,30 Effects of clinical area on the score of SAQ-K domains (Table 5) and SR showed the smallest , τ at 3.29. Though , τ appeared smaller than the standard deviation σ of the residuals within clinical areas for all six domains, the LR test results showed that random effects were statistically significant for all six domains, suggesting that SAQ-K scores differed by clinical area and thus clinical areaspecific SAQ-K scores are worth estimating.
Empirical bayesian estimation of SAQ-K domain scores by clinical area
To obtain more accurate estimates of clinical area-specific means, we applied the empirical Bayesian method. (Figure 2) shows the means estimated for the TC domain by ordinary least square (OLS) regression (upper panel) and the EB method (lower panel). Each bar stands for the mean and 95% confidence interval (CI) of a clinical area. The numbers over the CI bars are the clinical area ID numbered by the rank of OLS means; ten clinical areas with the widest CIs by OLS were marked, and these area IDs were marked on the EB means and CIs to clearly show the changes.
For those ten clinical areas, CIs hugely shrunk and their ranks changed, indicating that area mean estimates from the EB method differed from OLS results. Note that any changes in rank were toward the center, the overall TC score mean that was around 65, as denoted in (Table 5). These findings exemplify the shrinkage idea of the EB method, where clusters with large variance borrow information from other clusters to yield more accurate estimates; in other words, CIs become narrower. In return, the means from the EB method are pulled toward the overall mean even to the level at which the rank orders change significantly. Figure 3 shows the estimates from OLS and the EB method side by side for all six domains, helping the reader understand the effects of the EB method in obtaining more accurate estimates. Due to space limits, IDs of only seven clinical areas are displayed. Although the TC domain was shown in (Figure 2), we show it again to better compare the results.
Discussion
The primary aim of this study was to develop and validate SAQ-K. The results showed that SAQ-K had acceptable internal consistency and construct validity, suggesting that it is ready for use in health care organizations across Korea. Note that two items from the original SAQ, one in the TC domain and the other in PM, were excluded in the final version of SAQ-K. These two items were negatively phrased and their Korean translation was hard for respondents to answer, which emphasizes the difficulty of crosscultural adaptation of an instrument and the need for exceptional caution. 26 One reason we want to measure safety culture is to predict future safety incidents. Indeed, many studies have shown that safety improvement programs with a component that nurtures safety culture have significantly reduced the incidence rates of harmful events. 2,31 Therefore, hospital safety managers naturally ask, "Which areas of my hospital have a low safety culture?" The answer to this question guides resource allocation for safety improvement efforts across various areas of the hospital. In this sense, the more accurate the SAQ scores we obtain, the better we can allocate resources, which will eventually enable a hospital to get the most out of its safety programs. Traditional approaches have simply calculated the mean and standard deviation of SAQ domain scores by clinical area, and more likely they only calculated the percentage of respondents with positive perceptions. Both of these approaches are limited in use by overly small sample sizes in many clinical areas. As shown in Figure 1, several clinical areas have only a few HCWs working in them and, therefore, the clinical area mean of such units cannot be estimated accurately due to large standard deviations, as shown in (Figures 2 & 3). However, in the most cases, those simple traditional approaches have prevailed, probably because safety managers more often than not have limited statistical background, especially in advanced statistics like the Bayesian approach.
This study, therefore, applied the empirical Bayes method to provide more accurate mean SAQ domain scores for each clinical area. The empirical Bayes predictor is the best linear unbiased predictor (BLUP), and in this study the EB method improves the accuracy of area-specific estimates by borrowing strengths from other clinical areas. 22 In this process, the mean of each area shrinks toward the overall mean, yielding results as seen in the lower panel of (Figure 2). The advantage of the EB method is that it can be applied even when there are clusters contain very small sample sizes, like one or two, [22][23][24]32 which is frequently the case in hospitals in Korea.
Despite the above-mentioned statistical strengths of the EB method, one might point out overlaps in the CIs of clinical areas and therefore this EB approach might not be necessary. Though seemingly logical, this doubt is easily resolved if one understands how SAQ scores are used in hospitals. Due to the scarcity of resources, a hospital can run safety improvement programs like teamwork training only for a certain number of clinical areas (e.g., ten units in a given year). Selection of those ten units is based on SAQ scores (e.g., bottom ten areas), regardless of whether CIs overlap or not. Therefore, estimating mean SAQ scores of clinical areas as accurately as possible is essential to guaranteeing the just allocation of resources. In addition, SAQ is usually measured annually, leading safety managers to examine changes in SAQ scores by clinical area over time. One might consider building a longitudinal model with the SAQ score of each HCW as the unit of analysis. However, in most cases, this can be challenging because new HCWs are joining and also many HCWs are leaving the area. This naturally leads us to use the mean SAQ score of a clinical area as the unit of analysis, where a sound longitudinal model can only be built with accurate mean SAQ scores of clinical areas. Before we applied the EB method, we tested whether SAQ scores were significantly different among clinical areas. This step was particularly important in this study in that it guided us in judging whether EB estimates were worth applying. For a dataset with not many clusters, one-way ANOVA can be used, but this study covered 72 clinical areas which were heavily unbalanced in sample sizes and had unequal variance, as seen in (Figures 1 & 2). Therefore, we built variance components models with clinical arealevel random effect and tested the difference by observing whether the random effect was statistically significant. In this study, we used the likelihood-ratio test because we were only interested in whether the random effect model was necessary, as seen in (Table 5). If the estimates of CIs for random effect are of interest, score tests (Lagrange multiplier tests) can be used instead. 22,33 Employing variance components models is useful, especially when multiple clustering factors are of interest, such as clinical areas and job type of HCW. In an earlier section, how resources can be allocated over clinical areas was introduced, and the same logic can be applied to job type instead of clinical area. The problem arises when a safety manager needs to decide which clustering factors between clinical area and job type should be used for resource allocation. To illustrate, if variance in TC scores over job type is much bigger than that over clinical area, then the manager can develop a program for the job type with the lowest TC score to match other job types with higher TC scores. 34 The variance components model allows us to perform such comparison between variances. For example, by expanding the above variance component model, a crossed random effects model can be built on the assumption that the SAQ score of an HCW is correlated with both clinical area and job type, without assuming hierarchy between clinical area and job type. We can test whether each of the clustering criteria is significant and compare their variances. Also, we can test whether the effect of job type is nested within a clinical area, or the other way around, with hierarchical models. If there are only a few different job types in a hospital and just marginal differences in SAQ scores among job types are of interest, a generalized estimating equation (GEE) with clinical area as the clustering variable 35 can be used. Though building a GEE model is relatively simple, it is not discussed in this study because it is not within the scope of the study.
Although we developed and validated SAQ-K successfully and provided methodology for its analysis, some issues remain to be addressed. First, many HCWs work in multiple areas of the hospital. For example, a physician in the Department of Anesthesiology and Clinical Care Medicine (ACCM) works in both operating rooms and intensive care units. Which clinical area should the physician choose? A more fundamental question might be which area one perceives as the culture in which one is embedded. SAQ defines one's clinical area as the area where one 'typically' spends time, but this is not clear for many HCWs. Also, sufficient time is required for an HCW to be assimilated into an area's culture. If the period from when an HCW begins to work in an area to SAQ administration is only one week, it might not be wise to think that the HCW's responses accurately reflect the area. In hospitals where HCWs frequently change their working areas or where new trainees join regularly, this issue is amplified. These issues probably cannot be solved perfectly. Rather, what we can do is maintain consistency across the hospital. For example, a SAQ participant inclusion criterion in the Johns Hopkins Hospital is "HWCs with a 50% commitment to one clinical area for at least 4 consecutive weeks before the survey administration". 17 Though still arbitrary, this offers better internal validity when a comparison of SAQ scores within a hospital is of interest.
Conclusion
Culture is complex issue. Measuring culture, therefore, is rightly difficult, and safety culture in hospitals is no exception. However, such difficulty cannot justify our negligence in measuring it. Indeed, without measuring safety culture, we will never be able to improve it. Thus, instruments to measure safety culture are crucial. Through this study, we developed the Korean version of such measurement tool, and we also developed methodologies to better analyze it; these methodologies can be used in other countries as well. We hope that this study helps Korean hospitals provide patients in Korea with safer care, the care that patients well deserve.
Contact the corresponding author at hjeong1@jhu.edu for SAQ-K and Stata commands for the analyses in this article. | 2019-02-15T14:18:07.185Z | 2015-02-06T00:00:00.000 | {
"year": 2015,
"sha1": "b82d6dc3e02448c7fe216b25683732c36576894f",
"oa_license": "CCBYNC",
"oa_url": "https://medcraveonline.com/BBIJ/BBIJ-02-00020.pdf",
"oa_status": "HYBRID",
"pdf_src": "MergedPDFExtraction",
"pdf_hash": "bf7eb276e731861be55f701660913fd222821953",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Computer Science"
]
} |
253447285 | pes2o/s2orc | v3-fos-license | Accounting For Quantum Phenomena In Classical Molecular Dynamics Simulations At Cryogenic Temperatures: Diffraction And The Feynman-Hibbs Correction
The transport properties of 4 He are determined in the temperature range of 10 - 150 K, in classical and quantum frameworks, employing numerical and molecular dynamics simulation (MDS) techniques. Although classical MDS are efficient and comprehensive for understanding the transport properties, the increasing effect of quantum diffraction in the temperature range of 10 - 40 K is not accounted for in the MDS. The diffraction effect causes significant deviations in the simulated results from quantum mechanically calculated theoretical values. In this study, the missing quantum diffraction effect has been accounted for in the MDS through an appropriate reduction in the radius of the 4 He atoms in the simulation system. The obtained results are compared with those obtained from the existing quantum corrections in classical simulations through Feynman-Hibbs (FH) variational principle. It is found that FH corrections fail in the temperature regime where the quantum diffraction effects are large. At 10 K, the MDS result for thermal conductivity, with no quantum corrections, deviates from the quantum mechanical literature values by 12.45%. With FH correction, the same has a deviation of 18.73% and the correction proposed in this article, henceforth referred to as the diffraction correction, yields result with a deviation of only 2.75%. The present article is expected to present a detailed analysis of the benefits and limitations of various existing methods to study the transport phenomena of 4 He at low temperatures. Finally, the proposed method of radius reduction is expected to be applied to complex systems to study fundamental physics and its applications.
I. INTRODUCTION
The knowledge of transport properties of simple gases to complex fluidic systems is important in various interdisciplinary fields such as fuel combustion technologies 1 , nuclear reactor engineering 2 , extraterrestrial explorations 3 , fusion 4 and plasmas technologies 5 . Transport coefficients such as diffusion, thermal conductivity and viscosity that quantify the transfer of energy, momentum and mass in a fluid, act as a bridge connecting the microscopic molecules in a fluid with the macroscopic system as a whole. The interaction and the properties of the individual molecules in a fluid determine the transport coefficients, which can be measured at the macroscopic level. The modelling of atmospheres of other planets 6 , design of thermal protection systems for planetary exploration probes 3,7 , designing inertial confinement fusion reactors 4,5 and the modelling of plasma processes used in gasification of biomass 8 require extensive data on the transport properties of various species. There have been several studies on transport properties of gases in the high temperature regime [9][10][11] , however, the transport properties of neutral gases become very important for the understanding of dynamics of charged species in a weakly ionised plasma medium, at cryogenic temperatures (10 K to 150 K) 12,13 . These systems, referred to as cryoplasmas, can be studied at their molecular level efficiently using classical MDS. However, at low temperatures the quantum effects tend to become more prominent and the classical MDS are unable to provide accurate results, leading to the a) Electronic mail: ramkumar@iitk.ac.in b) Electronic mail: swatis@iitk.ac.in c) Electronic mail: sudeepb@iitk.ac.in divergence in the classically calculated transport coefficients from their experimental values.
There have been studies on the effect of quantum mechanical properties of the gas particles on the transport coefficients in different temperature regimes [14][15][16] . To include the quantum effects, first principle quantum calculations have been carried out which provide accurate results at lower temperatures 17 , where classical values deviate from the experimental results (< 40 K for the case of 4 He gas), but the calculations are lengthy and difficult to perform each time. De Boer used Lennard Jones potential to calculate the transport properties of 4 He gas at very low temperatures (< 4.1 K) 18 and Uehling studied the 4 He gas in the temperature range of 3 K to 300 K using both quantum and classical theories 19 . Owing to the complexity of the quantum calculations, correction procedures to classical methods, to account for quantum effects, have been worked out 20,21 . A semi-classical approach using WKB approximation was proposed by De Boer and Bird 20 which was limited to the temperature regime where the quantum effects are small (> 50 K). Furthermore, the well known Feynman-Hibbs (FH) variational approach, which accounts for the positional uncertainty of quantum particles 21 , has been used successfully for many studies such as adsorption of hydrogen and deuterium 22 , mie fluids 23 , quantum effects in light and heavy water 24 including the transport properties of 4 He at 80 K, neon at 45 K and methane at 110 K 25 , where the quantum effects are small. At cryogenic temperatures, when the wave nature of the particles are considerable, diffraction occurs when particles scatter from each other (quantum diffraction effect). However, later in this article, we shall see that the FH approach fails when the quantum diffraction effects are large.
The purpose of this article is to understand the limits of arXiv:2211.05154v1 [physics.chem-ph] 9 Nov 2022 quantum and classical theories by identifying and quantifying the quantum mechanical (QM) effects that affect the transport properties such as diffusion (D), thermal conductivity (κ) and viscosity (η) of 4 He gas using Lennard-Jones (LJ) interaction potential at cryogenic temperatures. With the known potentials between the particles, MDS is easier to work with than the numerical methods, when the complexity of the system increases. Hence, we focus on correcting MDS results for the aforementioned properties in the temperature range where the QM effects are considerably large (10 K to 40 K), at atmospheric pressure. We show how quantum diffraction effects can be considered classically to correct for the deviations in MDS from the quantum mechanical literature values, which are typically obtained either from experiments or calculated numerically using ab-initio potentials, as reported by Kestin et al. 26 , henceforth referred to as the literature values. The proposed correction reduces the percentage deviation in the results obtained from classical MDS, of the considered transport properties, from ∼20% to ∼ 5% in the temperature range of 10 K to 40 K.
This article is divided into six sections. Section II briefly explains the quantum effects causing the classical results to deviate from the literature values. Section III explores the different methods and frameworks using which the transport coefficients are calculated. Section IV explains the FH and the diffraction corrections to the MDS. Section V discusses the results of the simulation and numerical calculations, and the article is concluded in section VI.
II. QUANTUM EFFECTS
Due to the wave-particle duality of quantum particles, at low temperatures, when the thermal de Broglie wavelength (λ th ), given by λ th = 2πh 2 /(k B mT ), where k B is the Boltzmann's constant, m is the mass of the molecule, T is the temperature, andh is the reduced Planck's constant, is comparable to the particle's dimensions, diffraction effects become dominant. Furthermore, at cryogenic temperatures, the particles with spin half follow Fermi-Dirac statistics and particles with integer spin follow Bose-Einstein statistics and not the classical Boltzmann distribution. When λ th is comparable to the inter-particle distances (d), the effects due to the quantum statistics (symmetry effects) become dominant. For 4 He atoms, the Van der Waals radius (r v ) is 1.4 Å, which corresponds to λ th of 4 He at ∼ 45 K. In Fig. 1 we can see that the d in 4 He gas, obtained from the number density (n) data at atmospheric pressure 27 , from the relation d = n −1/3 , approaches λ th at a much lower temperature, less than 4.222 K, which is the boiling point of 4 He gas at atmospheric pressure 27 . Hence, while taking quantum effects into consideration, quantum statistics effects may be neglected in the temperature range of our study.
It should be noted that under the quantum framework, the calculation of diffusion coefficient needs special considerations. Classically, one can track the motion of individual particles to find the self-diffusion coefficient but quantum mechan- ically, when identical particles interact, they lose their identity after the interaction (indistinguishability) 28,29 . This causes the quantum mechanically obtained values for self-diffusion scattering cross-section to be almost twice that of classically obtained cross-section and surprisingly, the classical values are closer to the experimental values than the quantum mechanical values, even at cryogenic temperatures. This is because experimentally it becomes impossible to track the motion of individual identical particles and hence, some property of the particles, which do not affect the collision process, is used to differentiate the particles in the system 30 . For instance, in the case of 3 He self-diffusion experiments, the molecules are categorised based on their spin state (+½ and -½) 31 as spin does not affect collision interactions. Similarly, different isotopes with mass ratios very close to unity are also used in calculating self-diffusion coefficients experimentally. By doing this, the interacting particles are made distinguishable and the experimentally measured self-diffusion coefficients are not that of true self-diffusion of identical particles. Theoretically, we can calculate these values by assuming a distinguishable scattering process where the self-diffusion coefficient is calculated as a limiting case of binary diffusion 28 .
III. METHODS
In all the calculations done in this article, LJ potential is considered to depict the interaction between the 4 He atoms, which is given by tial. The LJ potential used is truncated at a cut-off distance of 6.6 Å for MDS with no quantum corrections and with diffraction correction, and at 8 Å for the MDS with the FH correction. Beyond this cut-off length the interaction potential is small enough (< 0.02 times ε) to be neglected. The flowchart in Fig. 2 shows the different frameworks under which the transport properties are calculated and the corresponding methods and correction procedures used. These methods are explained in detail in the forthcoming sections. The right half of the chart (cf. Fig. 2) depicts the numerical methods used in different frameworks, where Boltzmann's equations are solved to obtain the transport coefficients. The collision integral is obtained from the scattering angle in classical framework and from the phase shift of the scattering state wave function in the quantum framework. The left half of the chart (cf. Fig. 2) shows the MDS and different correction procedures used in MDS 21 . Under the classical framework, two independent theories, namely the Green-Kubo approximation 32,33 and the Boltzmann's transport equations 34 are used to evaluate the transport properties.
A. Classical Framework
Within the classical framework, the transport properties of 4 He gas can be calculated through the use of MDS or by numerically solving the Boltzmann's transport equations. In the latter method, the scattering angles for binary collisions between the gas molecules are used to evaluate the collision cross-sections 34 and the collision cross-section thus evaluated are used to calculate the collision integrals (Ω) 34 . Using the Chapman-Enskog theory 29 , the Boltzmann's transport equations are solved and the transport coefficients are obtained using the Chapman-Cowling approximation, to the first order, as 29 , Here m is the mass of the interacting particles, P corresponds to the pressure of the system, at which the transport coefficients are calculated and Ω (r,s) are the collision integrals, where r = s = 1 corresponds to the diffusion collision integral and r = s = 2 corresponds to the viscosity / thermal conductivity collision integral. The above formulas are obtained as a first order approximation in solving the Boltzmann's transport equation and collision integrals of other values of r and s are used to calculate the higher order correction terms to the transport properties. MDS although being a classical method, doesn't use Boltzmann's transport equations to obtain the transport properties (cf. Fig. 2). Instead, the Green-Kubo relations 32,33 , which uses the autocorrelation function of the flux corresponding to the transport properties, are used to evaluate the transport coefficients. Viscosity is calculated from the autocorrelation function of the off-diagonal pressure tensor (P i j (t)), given by 35 , the thermal conductivity from the heat flux (J(t)) correlation function, given by 36 , and the diffusion coefficient from the velocity (v(t)) correlation function, given by 37 , where t 0 and t are any arbitrary initial and instantaneous times respectively. Alternatively, the diffusion coefficient can also be calculated from the asymptotic time slope of mean square displacement (M ) of the particles in the system, using the Einstein's relation given by, D = lim t→∞ M /6t 37 . The Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) 38 is used to perform the MDS. While calculating the transport coefficients from the Green-Kubo relations, the autocorrelation functions are evaluated ∼400 times and a cumulative average is taken to cancel out any white noise obtained in the simulation results. The autocorrelation length for heat flux (thermal conductivity calculation) is around 0.5 ns to 0.8 n.s and the same for pressure-momentum flux (viscosity calculation) is ∼ 0.1 ns to 0.4 ns.
B. Quantum Framework
One of the disadvantages of the purely classical MDS is that at low temperatures, the QM effects are not taken into account. In the case of 4 He gas, the QM effects start to dominate at relatively higher temperatures (∼ 40 K) than for any other gases. Using the quantum mechanical theory, the transport properties can be obtained by solving for the phase shifts in the asymptotic region of the scattering state wave function, which then can be used to calculate the collision cross-section in the Chapman-Cowling relations 29 (cf. Fig. 2).
In quantum mechanics, due to uncertainty principle, the relation between scattering angle (χ) and the impact parameter (b) can't be defined to calculate the collision cross-sections. Instead, the phase shifts δ l in the asymptotic region of scattering state wave functions are used to obtain the quantum mechanical collision cross-sections (cf. Fig. 2) 18,28,34 which can be used in the collision integral equation to evaluate the transport coefficients from the Chapman-Cowling expressions (Eqs. 2). To include the effects of quantum statistics, the summations over angular momentum states in the collision crosssections are carried out only over the even l values for Bose-Einstein case and over the odd l values for Fermi-Dirac case due to symmetry, and in the case of collision between indistinguishable particles, a factor of 2 is multiplied to them 18,34 . To obtain the phase shifts, the Schrödinger's equation is solved numerically using the Numerov's integration method 39 for the case of LJ potential. In our calculations, the wave functions are discretised with a step size of 0.025 a.u. and are evaluated up to a distance of 1500 a.u.
The phase shift (δ l ) can be evaluated using, Γ = r 2 U (r 1 ) r 1 U (r 2 ) = j l (kr 1 ) cos δ l − n l (kr 1 ) sin δ l j l (kr 2 ) cos δ l − n l (kr 2 ) sin δ l , and δ l = j l (kr 1 ) − Γ j l (kr 2 ) n l (kr 1 ) − Γ n l (kr 2 ) , where j l and n l are the spherical Bessel and Neumann functions, r 1 , r 2 are two arbitrary positions in the asymptotic region, and U (r) = R(r) r, with R(r) being the radial part of the scattering wave function and r being the radial position. To obtain the quantum scattering cross-section, the phase shift values must be calculated for all even angular momentum states at different energies from 0 to ∞, as the quantum crosssections have summation up to infinite angular momentum state and the collision integral has integral over wave number from zero to infinity. We approximate this by carrying out the summation over angular momentum states and integration over the wave vector until the change in the crosssection and collision integral is not more than a factor of 10 −8 after inclusion of the phase shifts for higher angular momentum states and energy values. The integration in collision integral is performed using the trapezoidal method as it varies smoothly with wave number. For the temperature region that we are interested in (10 K to 150 K), a maximum angular momentum state of l = 22 and a maximum wave vector of 9 a.u. is considered.
IV. QUANTUM CORRECTIONS TO MOLECULAR DYNAMICS SIMULATION
Solving the Schrödinger equation for complex systems to find the collision cross-section becomes challenging and hence, correction terms that account for the missing quantum effects in the classical framework are introduced in MDS.
A. Feynman-Hibbs Correction
One such correction is given by the Feynman-Hibbs variational principle 21 , where the interaction potential (U(r)) is modulated over a Gaussian space with width corresponding to the λ th at a given temperature, which accounts for the spread in the position of the particle due to its quantum nature. The FH potential (U FH ) is given by, Expanding U(r + R) using Taylor expansion and evaluating the integral in Eq. 8 up to second order inh 2 , the FH potential and (c) particle-particle scattering.
is given by, From Eq. 9 it can be seen that in the FH formalism, a correction term, which accounts for the quantum effects, is added to the classical interaction potential (U(r)).
B. Correction for diffraction effects
When a wave is diffracted from a particle, for example diffraction of light from a particle, there is a central maxima of intensity right behind the particle that is obstructing the wave (cf. Fig. 3(a)). The intensity of the diffracted wave is given by 40,41 , where I 0 is the intensity of the incident wave, ξ and η represents coordinates on the obstructing particle and are integrated over the surface of the obstruction and, x and y coordinates represent the point on the surface at which the diffraction intensity is calculated (Fig. 4).
In Eq. 10, the second term represents the Fresnel diffraction amplitude for a circular aperture 40 and the first term is the incident plane wave included as a result of the Babinet principle 41 . In the case of particles scattering from other particles, for an incident particle flux of f 0 , the number of incident particles intercepted by the obstacle per unit time is given by area × f 0 . This creates a shadow region, for the incident particle flux right behind the obstacle, of area π(r 1 + r 2 ) 2 where r 1 and r 2 are the radii of the interacting particles. For particles of identical size, this shadow area is π(2r 0 ) 2 = 4πr 2 0 (cf. Fig. 3(b and c)).
Classical MDS considers only the case of particle scattering and not the intensity allowed in front of the obstacle due to quantum diffraction effect. To account for the wave nature (i.e. diffraction effect) of 4 He particles in the cryogenic temperature limit, we take into consideration the effects of diffraction of waves, of wavelength λ th , with the classical particle scattering by considering an equivalence between incident particle flux and wave intensity ( f 0 ≡ I 0 ). Such comparisons between matter waves (wave nature of quantum particles) and classical optics has been studied extensively under the name of atom optics 42 . The Fraunhofer and Fresnel diffraction theory of classical optics has been used in the study of atomic and molecular scattering experiments [43][44][45] and the validity of classical theories, used in wave optics, such as babinet principle has been studied experimentally for matter waves 46 .
Considering a hard sphere model for the particles, the intensity of waves allowed due to the diffraction is compensated by reducing the radius of the particles which reduces the number of particles intercepted by the obstructing particle. The radius is reduced to the extent that the increase in the allowed flux ( f 0 × ∆area) is equal to the intensity of the diffracted wave over the area in front of the obstacle. This can be mathematically expressed as, where r 0 is the radius of the particle, λ is the wavelength of the diffracted wave, z is the distance at which diffraction intensity is measured and I(r) is given by Eq. 10 with change of coordinates from cartesian to polar. The area of the shadow region is considered to have a radius of 2r 0 as the impact parameter for hard sphere collisions is sum of the radii of the colliding particles. Fig. 3 illustrates the correction considered. The right side of equality in Eq. 11 is the shaded region of the diffraction effect shown in Fig. 3 and the left side of Eq. 11 is the hatched region of the particle scattering shown in the figure (cf. Fig. 3(c)) FIG. 5: Temperature dependence of radius reduction factor (χ) compensating for the quantum diffraction and diffraction effect quantifier (1 − χ) from 10 K to 150 K.
For the case of 4 He collisions, the incident particle is considered to have a wave character with wavelength (λ ) equal to λ th and the particle radius to be equal to σ /2, which is the characteristic length of LJ potential. We calculate the diffraction intensity close to the particle (z = r 0 ) over an area of radius 2r 0 .
Expressing the reduced radius as the original radius multiplied with a reduction factor (χ), the reduction factor's (χ) dependence on temperature is shown in Fig. 5. It is evident that at high temperatures the reductive factor approaches unity and at the lower temperatures it deviates from unity. Physically, 1 − χ(T ) can be interpreted as a factor quantifying the quantum diffraction effect present in the system at a particular temperature as it approaches zero at high temperatures and increases as the temperature decreases (cf. Fig. 5). This correction procedure is implemented in the MDS by replacing the length parameter, σ, in the LJ interaction potential (Eq. 1) with χσ. After making these changes in the simulation parameter, it has been verified that the state variables (temperature, pressure and internal energy) converge to the appropriate values as in the case without the correction. The transport coefficients, namely diffusivity (D), thermal conductivity (κ), and viscosity (η) calculated using different methods and correction terms, in the temperature range 10 K to 150 K, are provided in Table. I. It can be observed in the classical framework that the results from the MDS, without any correction terms, and that obtained by numerically solving the Boltzmann's equations agree with one another, showing that the error in the calculated coefficients in the low temperature regime is due to the missing quantum effects in the classical framework and not due to some artefacts in the calculation methods. The numerical method of solving Boltzmann's equations has been worked out under the assumption that only binary interactions take place during a collision, i.e. only two particle collisions are considered 29,34 . The numerically calculated quantum mechanical values match well with the quantum mechanical literature values, with deviations less than 5% throughout the temperature range considered. Fig. 6 shows the plot of transport coefficients obtained through different corrections and calculation methods. It can be observed that the transport coefficients obtained from various methods closely follow the literature values trend. Fig. 7 shows the percentage deviation in the values calculated using different methods and correction terms, from the literature value. The trends in the percentage deviation of the calculated transport coefficients at different temperatures are similar in the case of all three transport coefficients.
The purely classical values (without any correction term) diverge from the literature values below 40 K as the temperature decreases, which can be attributed to the missing quantum diffraction effects. With the introduction of FH correction in the LJ interaction potential, the error in the diffraction regime (< 40 K) increases, making the calculated coefficients worse. With the increased delocalisation at lower temperatures, the FH approximation doesn't yield the correct kinetic energy of the particles 47 and works well only when the diffraction effect is moderate 23,47,48 . 4 He being a light atom, the quantum effects are larger than that for any other particle, and hence, FH potential can't be used. In the case of the high temperature regime (>100 K) of the temperature range considered, the FH values converge towards the classical values as expected due to the reduced diffraction contribution.
The correction implemented considering diffraction effects works well in the temperature range of our study. In the high temperature regime, the correction due to diffraction becomes negligible, and the values converge towards the classically computed values, and in the low temperature regime, the deviation from the literature value is reduced to ∼ 5% from the 20% deviation in classical results. temperatures < 150 K. Hence, quantum calculations must be performed in this temperature regime or corrections must be included to the classical methods to account for the quantum effects. At temperatures > 150 K, classical methods sufficiently describe the system. Higher the temperature, more accurate the classical framework is, as shown by the color contrast in the classical regime (red) in Fig. 8. As depicted by the color contrast of quantum regime (blue) in Fig. 8, the deviations in the classically calculated transport coefficients increases with decrease in temperature, due to the increased quantum nature of the particles at low temperatures. At temperatures < 5 K the effects of quantum statistics (symmetry) is considerable. In the temperature range of 10 K to 100 K, the quantum diffraction effect is the major contributor to deviations in the classically calculated transport coefficients.
At temperatures < 80 K, where the quantum diffraction effect is large, the correction introduced in classical simulation due to FH variational principle fails though it has been proven to be successful at temperatures where the quantum diffraction effects are small. This is attributed to its inability to correctly yield the kinetic energy of the particles in the system due to the increased delocalisation at lower temperatures. In this temperature regime, by reducing the radius of the particles in the MDS system, we show that the missing quantum diffraction effects can be accounted for, as the reduced radius decreases the collision cross-section to compensate for the reduction in collision cross-section caused due to the diffraction effects.
VII. ACKNOWLEDGEMENT
The authors sincerely thank the high performance computing facility (HPC) at IIT Kanpur, because of which this work has been made possible.
A. Conflict Of Interest
The authors declare no conflict of interest.
B. Data Availability
The data that supports the findings of this study are available within the article. I: Transport coefficients data obtained using numerical quantum calculations (QC), numerical classical calculations (Cls), MDS with no corrections (MDS), FH correction to MDS (MDS+FH) and diffraction correction to MDS (MDS+DC). Here diffusion coefficients (D) is in 10 −6 m 2 /s, thermal conductivity (κ) is in 10 −2 W /(K − m) and viscosity (ν) is in µPa/s. | 2022-11-11T10:17:26.049Z | 2022-01-01T00:00:00.000 | {
"year": 2022,
"sha1": "3dd59df96c82dd90508e2f5213729f9722da9b78",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "3dd59df96c82dd90508e2f5213729f9722da9b78",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Physics"
]
} |
53601028 | pes2o/s2orc | v3-fos-license | Blockchain Practices, Potentials, and Perspectives in Greening Supply Chains
: Blockchain technology is an inchoate technology whose current popularity is peaking. Some of the most pervasive blockchain technology use cases exist for supply chains. Sustainable, and especially green, supply chains can benefit from blockchain technology, but there are also caveats. The sustainability and environmental management research and academic literature is only starting to investigate this emergent field. This paper seeks to help advance the discussion and motivate additional practice and research related to green supply chains and blockchain technology. This viewpoint paper provides insight into some of the main dimensions of blockchain technology, an overview of the use cases and issues, and some general research areas for further investigation.
Introduction
Technological advancements have caused a revisiting of sustainability practices.According to ecological modernization theory, technology can help decouple environmental degradation from economic growth [1].In some cases, technology can benefit both dimensions.As the triple-bottom-line sustainability definition includes social dimensions, whether technology can contribute to all dimensions of sustainability is unclear.
Advances in technology are broad-based and include a variety of production, information, and social technologies.These technologies include current and future developments in such disparate, but possibly interrelated, areas such as additive manufacturing, micro-factories, nanotechnology, Internet of Things (IoT), self-driving vehicles, sharing economies, and blockchain technology [2].Each of these technologies has implications for the sustainability of organizations and especially their supply chains.
Supply chain management is critical for managing sustainability at global and local levels.Whether the focus is on environmental and green initiatives or social responsibility, the largest and deepest influences are supply chain activities.Of all technological developments, blockchain technology can have profound implications for supply chain sustainability, also known as distributed ledger technology.Although we devote a whole section to the definition of blockchain technology and general characteristics, we define it as decentralized databases or ledgers of records that are shared among networks and supply chain participants.In blockchains, records and data are secure, traceable, and auditable, and maintained on a peer-to-peer network [3].The contribution of this paper is providing insights into the potential application of this nascent technology to facilitate green practices in supply chains.Our discussion is grounded on the current understanding of blockchain technology and green supply chain management literature.The evaluation framework used in this study was proposed by Hervani et al. [4].This study would help managers, researchers, and practitioners to further evaluate the potential usage of blockchain technology to improve sustainability, especially along the supply chain.
To further clarify, we provide some insights into the various sustainability-oriented opportunities associated with blockchain technology use cases that occur across and within the supply chain.The supply chain activities include those occurring in upstream, internal organizational, downstream, and loop-closing functions [4].There are similar relationships and implications for each of these activity groups, and there are also unique activity specific cases.After examples are provided, some general research questions are posited.We think this discussion furthers the need to carefully study how blockchain technology specifically, and disruptive technology in general, require more nuanced investigation in sustainable supply chain practice and research.
Blockchain Technology
Blockchain technology became popular through the advancement of cryptocurrency and bitcoin after the 2008 financial crisis [5].Although the primary focus had been on financial applications, the unique characteristics of blockchain technology inspired broader use of this technology in different markets and even for non-financial business purposes.Supply chains [6], real estate [7,8], government [9], healthcare [10], and energy sector [11] use cases have been some effective applications.
Blockchain technology has a number of general characteristics.The integration of these characteristics differentiate blockchain from other similar information technologies.Unlike other business information technologies, blockchain technology uses a unique data structure that stores data as a chain of blocks.Once a new transaction is recorded on the system, it builds a block that is linked to the previous blocks, creating a chain [5].
In terms of openness and access to data, two popular types of blockchain exist: public and private.In the public blockchain, which is generally permissionless, ledgers are publicly available and anyone can record transactions and track the historical transaction on the ledgers.Popular cryptocurrencies, such as bitcoin and ether from Ethereum, were developed on public blockchains.Public blockchains require a high level of security and reliability due to the existence of anonymous users and the lack of trust among them [12].
In a private blockchain, users are known and ledgers are shared among a private group of participants.In a private or permissioned blockchain, access is restricted to a defined group of participants.A validator allows participants to join the system, provides permission to the ledgers, and maintains the privacy needs of the network [9,13].Depending on the type of blockchain, the characteristics slightly change.
Although the main features of both blockchains may overlap and vary in some of the literature, we discuss some of the more popular characteristics.Included amongst these characteristics are decentralized databases, data security, information transparency, information immutability, and smart contracts.
Decentralized Database
Decentralization is an essential characteristic of blockchain technology.In blockchains, no central database, organization, or authority is typically involved in transactions.Decentralized databases of records allow participants in the network to directly interact via a peer-to-peer network.Every participant in the network has the same copy of the ledgers, which are updated with new information or changes in the recorded information in a decentralized manner [3].
Every update in a ledger requires consensus among the network partners.Decentralized consensus is the core of blockchain, which utilizes various algorithms such as proof of work and proof of stake to confirm the reliability of a recorded transaction.Generally, decentralized consensus includes votes or validation of the majority of participants of a network for ensuring the credibility of transactions.Public blockchains require heavy use of consensus algorithms that consume a great amount of power and energy.This characteristic contributes to environmental degradation and negatively affects sustainability values [14].
In a private or permissioned blockchain, the consensus requirement is a set of rules that is defined by the network participants for adding and updating transactions to ledgers.Consensus rules in a private network provide flexibility and ease the use of cumbersome consensus algorithms.
Security
Information is maintained as blocks within blockchain technology.Each block has a timestamp and a hash value that refers to previous blocks on the chain.Hash values have unique cryptographic structures that prevent tampering and altering the information in the blockchain [15].Cryptography logic facilitates authentication and trading for anonymous parties, which is a necessity in public, permissionless blockchains, improving the trust and security of the system.
In a private/permissioned blockchain, the trust in the validator, who gives permission to the parties to record and trace information, plays an important role [16].Security is improved by the decentralized structure of the blockchain.As a result of decentralization, the validity of information is examined by network members based on the consensus rules.This characteristic confines the data misuse and network manipulation.Decentralization also ensures the network is less vulnerable to hacking or crashing.The single point of failure is a common security problem of centralized databases, which has been alleviated by the use of blockchain technology [9].
The timestamp plays a critical role in the supply chain given various time-based competitive issues, such as lead time, delivery, and perishability concerns.The timestamp is also critical to traceability and information transparency.
Information Transparency
Authorized blockchain network participants maintain the same copy of a ledger, which contains a list of transactions.The ledgers are updated with the most recent approved transactions.A complete history of transactions are visible to the network members, allowing for auditability and traceability [17].The level of transparency provided by the blockchain enhances fairness and ease of access to data within a network [3].Transparent information removes intermediaries involved in the processes, increases efficiency, and reduces risks [15].
Growing customer demands for supply chain transparency motivate the application of blockchain technology for supply chain processes.A high degree of transparency provides fundamentals for tracking the origin and flow of products and processes, the parties involved in transactions, and transportation information.Supply chain partners from upstream to end customers can follow and audit the history of records.Since records on the blockchain are time-stamped and secure, data manipulation and fraud are detectable and traceable on the ledgers.This provides trust and reliability for supply chain partners [18].Tracking technologies, such as radio frequency identification (RFID), the IoT, and smart devices link the physical product to the respective electronic records, creating inputs for blockchain technology that are maintained on transparent ledgers [19].
Data and Information Immutability
Blockchain data and information are immutable.Immutability means that records cannot be changed or modified without network member consensus.Participants can be confident that the history of records are reliable and unaltered.Theoretically, this feature comes from the append-only concept of the blockchain, which means records can only be added to ledgers and cannot be modified or removed.However, on a public or permissionless blockchain, where miners vote for transactions and control the system, collusion is possible if the majority of miners decide to alter or remove a transaction.Alternatively, change and removing information on a private or permissioned blockchain requires notifying the network members and follows certain agreements and approval requirements [9,16].
Smart Contracts
Smart contracts are computer codes and scripts that contain terms of contracts and business rules.Smart contracts automatically execute the terms of agreements.Smart contracts check the pre-determined conditions including rules and penalties that are agreed to by parties and trigger the related action to those conditions.
The conditions and terms of contracts are validated by network members [20].These computer codes are self-executives seeking to eliminate human intervention in contracts.Unlike traditional contracts where trust between parties plays an important role, smart contracts remove the need for trust.Terms of contracts and the related legal actions are digitally written as computer programs and stored on the blockchain platform.The digital contracts remove human judgement from transactions.The role of intermediaries, such as financial professionals and legal people that are involved in traditional contracts, can be minimized through smart contract use.The resulting disintermediation improves efficiency and reduces the costs of business activities.
An example of smart contracts is an automatic payment that is performed when a certain regulation is met or a particular value is added to a product [3,21].Transactions need to be verified to be added to the ledgers in blockchain technology.The process of transaction validation by network participants can be facilitated through smart contracts.The validation requirements and consensus rules can be regulated by network members and maintained as digital contracts.Smart contracts can check the pre-defined conditions for approving transactions and add them to the ledgers.Similarly, a change in the approved transactions can follow particular regulations that are stored on smart contracts.This digital transaction approval can simplify the use of blockchain technology in complex and private business networks.
Green and Sustainable Supply Chains: Blockchain Use Cases and Potentialities
The supply chain has numerous intra-and inter-organizational activities.Figure 1 provides a supply chain activities diagram that incorporates a closed loop perspective with some environmental sustainability dimensions [4].The activities generally include: (1) upstream vendor/supplier management concerns, such as supplier selection and development; (2) upstream purchasing, inbound logistics, and inventory management activities; (3) internal operations and productions activities; (4) downstream activities including distribution, green marketing, and consumerism; and (5) closing-the-loop activities such as reverse logistics and the various "Re's", such as recycle, reuse, remanufacture and reclaim.Additional activities and resources needed for the supply chain included in Figure 1 are aspects of waste management, energy utilization, and design concerns.
Based on our observations of a wide range of industries and arguments advanced by practitioners and the more popular business literature, we compiled a number of practical business use cases for greening supply chains.These supply chain activities and exemplary blockchain use cases and possibilities are detailed below.A summary of these use cases and possibilities appears in Figure 1 in the dashed circles.
Vendor Selection
Supplier and vendor selection is viewed by industry and academia as a critical issue for the long-term success of supply chains.Careful selection and evaluation of suppliers are necessary initial steps to ensure the sustainability of supply chains [22].It is this upstream portion of the supply chain that has the most profound influence on overall supply chain sustainability.Issues in the upstream supply chain can be easily hidden from buying organizations, furthering their exposure to risk and affecting supply chain resilience [23].
Supplier selection and evaluation in a sustainability context is a multi-dimensional and complex problem [24].Usually, supplier selection and evaluation is dependent on information.This information is not easily accessible, certifiable, and audited, especially in non-economic, social and environmental, sustainability dimensions [25].It is this major sustainable information limitation barrier that can be effectively alleviated using blockchain technology.
Vendor historical performance and sustainability data can be made available on the blockchain.This accurate and secure data about vendors' environmental performance help companies to improve their vendor selection processes based on green performance values.Using blockchain not only facilitates the vendor selection processes, but provides information regarding the whole supply chain across multiple tiers and sub-suppliers [26].The shared information on the blockchain provides companies the opportunity to help their suppliers in selecting their vendors in different tiers of the supply chain.This would help reduce the risk for the focal companies.Removing intermediaries is also an important outcome that enhances the vendor selection process in the supply chain and reduces costs.
Current supply chain sustainability database systems exist such as the Business Social Compliance Initiative (BSCI) database for voluntary supplier social and environmental auditing in the textile supply chain [27].These databases are available, in some form, to BSCI participants.One of the limitations of BSCI, as with other voluntary databases, is the validity and credibility of their data and audits.Blockchain technology and processes can help address some of these credibility and validity concerns and potentially satisfy third-party and non-governmental organizations (NGOs).These databases may be used for supplier monitoring, development, and selection; their credibility and accessibility can only further support these initiatives.
Supplier Development
Blockchain technology can help improve supplier development programs.The amount of investment in a supplier development program can be recorded on the blockchain platform.The type of knowledge that is exchanged and the type of organizational support has been given to the suppliers are traceable through the smart contract.The recorded information provides the basis for performance measurement of supplier development programs.Comparison between the performance before and after implementing a training program is possible on the blockchain.By means of smart contract, companies can ensure that they trade with suppliers that are involved in supplier development programs.This can also be a principle for selecting suppliers.
Environmental performance measurement and benchmarking systems will be valuable in determining potentially problematic suppliers within the supply chain.There will be issues of direct green supplier and sub-supplier development concerns.It is these traditionally invisible entities in the supply chain that can be the most environmentally risky and poorly performing members of the supply chain.Visibility further down the supply chain can more effectively identify potential sub-suppliers that may require green development and support [26].
Organizations such as Dell, IBM, Lucent, and Pepsico have extensive environmental supplier development and training programs [28].These organizations need to record and monitor suppliers included in their programs, which can number in the thousands of suppliers.Documenting and monitoring these suppliers also allows them to build their supplier capabilities and share them with a broader set of customers.That is, not only will the direct suppliers such as Dell, IBM and Lucent have knowledge of their green development, but other customers can potentially have access through industry associations such as the Electronic Industry Citizenship Coalition (EICC).
Purchasing
Instead of supplier data, product and material data and movement can be maintained on the blockchain.Every product can have several transactional characteristics that are recorded on the blockchain, along with the historical data of a product.These transactions may declare the origin of the product, the quality, quantity, owners, and time.These data provide the ability to trace green quality, recyclability, and carbon footprints.The environmental information ensures customers are aware of safe and sustainable production and transportation of goods.Therefore, customers, with the ability to access this information, would have the opportunity to select sustainable products [29].
To ensure sustainable purchasing, companies can track the journey of resources for rare and high value products.The ability to track the source of products to address biodiversity concerns and contribution of products to resource depletion are two cases that demonstrate the role of blockchains in ensuring the sustainability of products.Using blockchain technology, life-cycle analysis of products can be completed using actual product data, rather than by estimating the values, such as in current life cycle analysis methods, as demonstrated by Favi et al. [30].This accurate and actual information is a revolutionary contribution of blockchain technology in the life-cycle analysis domain.
Materials Management and Inbound Logistics
The location and type of facilities, and design of logistics networks to ensure sustainability can be supported with blockchain data.One particular issue is inventory management in a supply chain through warehousing.A significant amount of warehousing is outsourced to third-party logistics providers.Currently, disparate information systems are used to manage these warehousing relationships.Reduction of auditing and compliance for 'bonded warehouses', as well as tracing products and materials can all be supported through blockchains.Cross-border trade will be influenced from a tracking, finance, and scheduling perspective.
In these cases, traceability and auditing increase the sustainability of the warehouse operations by lessening waste due to product and materials loss.Also, scheduling and planning can be more effective by having utilization information for a network of warehousing choices.Alternatively, the increased use of blockchain in these settings, as in all settings, requires additional energy usage.
Another emergent warehousing and logistics issue is crowdsourcing.Crowdsourcing is an outsourcing strategy that places an open invitation to a broad group of participants to perform a task.This approach is similar to the sharing economy situation someone who has available capacity for storage or delivery can respond to these requests.FLEXE is a company that allows anyone with temporary warehouse capacity to sell it to those that need the space.Companies such as Rideship, Zipmets, and Deliv all provide services for crowdshipping.This crowdshipping takes advantage of nearby delivery services with the ability to service local needs.It also reduces the need to build additional warehousing and vehicles, and increases efficiencies associated with consolidation of materials.All these are win-win, joint environmental and economic benefits, providing opportunities for logistics providers.
A difficulty of these current sharing systems are transaction costs, with most of the benefits accrued by the service providers.The more democratic blockchain systems allow for a broader set of participants, potentially aiding, from a social sustainability perspective, lower income regions and individuals.There are current concerns with the use of blockchain crowdsourcing-related malicious agents in all areas of blockchain processing and activities (see [31]).In addition, there are more secured payment possibilities through the use of cryptocurrencies and tokens, which are pervasive due to blockchain technology.
Transportation between and among facilities is central to both outbound and inbound logistics.When contracting with a third-party transportation company, tracing and monitoring transportation will benefit from blockchain technology [29,32].Transportation causes significant environmental damage and is one of the highest emitters of greenhouse gas emissions, local air pollutants causing smog, and contributes to depletion of energy resources.On the in-bound side, tracing the performance of transport vehicles, as with truckers for example, uses electronic logging devices.Fraudulent actions can occur in the truckers' logs that owners may ignore for purposes of expediency.Some of these behaviors may cause environmental damage, such as driving faster increases emissions and fuel usage.
Changing driver behavior is an important way to save energy resources and improve safe driving.Utilizing cryptocurrency tokens could effectively reward drivers for safe and green practices; these practices may be monitored using blockchain technology with mobile technology.Most current incentive systems are tasked with delivering products quickly and only being rewarded when driving certain distances.These current incentives cause dangerous and unsustainable practices, such as drivers speeding more often and driving longer than allocated hours, creating dangerous conditions.
Building trust in the technology, its broad adoption, and agreed upon industry standards are all issues facing the adoption of blockchain in transportation.
Production and Internal Operations
Production and operations are internal activities within an organization.Whether the production is based on manufacturing goods or delivering services, the transformation of inputs into outputs are central activities of the production stage.Traditional goods manufacturing includes fabricating or assembly activities.Internal production and supply chain activities require environmental management practices, including production management, environmental management systems, eco-design, performance measurement, environmental accounting, reporting, life cycle analysis, source reduction, closed loop internal systems, and a variety of similar greening practices, that fall within the purview of the focal organization.
A linkage of these green practices to external blockchain activities resulting from upstream, downstream, and closed-loop activities needs investigation and determination.Each of the practices and systems can be profoundly influenced from resources and inventory management, flow of materials across the shop floor, to eco-design of products.
The ISO 14001 standards are a popular global environmental management system (EMS) certification.EMSs are critical to internal operations environmental management.Blockchain implications relate to acquiring and maintaining certification.The use of audit teams to certify ISO 14001 organizations may be influenced by the technology.ISO 14001 is dependent on documentation for full certification.This documentation is then audited.Additionally, ISO 14001 certification can occur simultaneously for all sites of a corporation.For multinational corporations that are distributed broadly, distributed ledger and blockchain systems can prove a valuable resource for accumulating, aggregating, and certifying dispersed documentation.Auditing for initial or recertification may become more efficient, and may even not be needed, as documentation can be evaluated and updated continuously.
Within environmental management systems, there are numerous sub-systems, especially with respect to ISO 14000 certification family modules.These subsystem standards include performance measurements, life cycle analysis, climate change, eco-design, and communications.Monitoring environmental performance measurements throughout an organization and its supply chain through a distributed verifiable system provides more accurate data for environmental management purposes.Central to EMS and production systems is the concept of continuous improvement.Continuous improvement requires performance evaluation to determine if goals are being met and if improvements are occurring.Permanent, transparent, and verified performance provides a true measure of improvement.Linking performance measurement and environmental systems globally across an organization's sites helps build broader environmental continuous improvement measures.
Many other internal activities related to production and operations potentially influenced by blockchain technology relate to other supply chain activities, including eco-design, material handling and flow, and supplier management, as examples.We delve further into eco-design and LCA initiatives.
Eco-Design and LCA
Eco-design is a particularly interesting blockchain use case that can be discussed as part of the production and operations or marketing stages of the supply chain.It involves multiple supply chain partners and functions within an organization.Eco-design, with a focus on new product development with environmental criteria playing a prominent role, is influenced by a blockchain in numerous ways.The blockchain helps with easily disseminating information to multiple parties involved, gathering and verifying information, controlling the environmental quality of materials, time management for new product development projects, and coordinating participants.
In some eco-design systems, the environmental impact of materials used requires validation.In some of these cases, specific tests need to be completed.For example, in the cradle-to-cradle design model, hazardous material weighting schemes are used for various materials.This information can be easily stored and accessed by multiple partners after a verification process.This practice is quite suitable for blockchain technology, where scalability concerns related to high volumes of transactions would not be characteristic for materials verification.That is, this process requires time and a limited number of verifications will be required.Once these materials are verified, they would be available for trade and marketing.
Materials will require verification and processing.Green process design is as important in eco-design as in green materials.These green processes require verification and improvement.Internal processes, through ISO 14001 aspects, EMS management can be improved, evaluated, and validated.External processes, potentially through a supply chain blockchain linkage of environmental management systems, will also need verification.These environmentally sound manufacturing process designs and improvements can be used with sourcing, supplier development, supplier selection, and operations management activities.
Eco-design systems integrated with LCA benefit from information accuracy.LCA materials inventory and impact constantly change with many uncertainties [33,34].Significant environmental information uncertainty exists for these systems.LCA tools may have different foundational information, different levels of granularity, missing data, and even inaccurate information.To address some of these uncertainties, various simulation tools have been proposed to complete a sensitivity analysis, as demonstrated by Mueller et al. [34].Blockchain validity, reliability, and transparency can reduce information uncertainty, providing better modeling inputs and outputs for eco-design and LCA tools.
Information standards based on product data technology standards for LCA data have been proposed [35].Expanding these arguments to blockchain, as an information delivery vehicle with appropriate models developed, is natural.The same benefits derived from using these industry standards and protocols occur through blockchain technology and systems, especially with an overall goal of reducing LCA information uncertainty.Benefits for blockchain adoption for these design systems include: less time for LCA data collection; improved data quality; traceability of the data source; using actual data from suppliers, not from a generic source; and storing environmental information of a product through the end of life to better manage its recycling and disposal [18,36].
Outbound Logistics and Marketing
Downstream green supply chain activities include distribution and various customer management activities such as green marketing and packaging.Transportation, similar to inbound logistics, is a large concern for distribution channels.Distribution transportation planning is typically planned by the organization's outbound logistics systems.This may or may not include third-party logistics providers.As mentioned in inbound logistics planning, certification and verification of environmental and social performance concerns exist in this activity; blockchain processes can address these issues.Information sharing on blockchain technology reduces the required paperwork, supporting validation requirements, and prevents data manipulation and counterfeiting within logistics and transportation processes [29].
Similar to the blockchain contributing to the sharing economy associated with crowdsourcing warehousing, which applies to outbound logistics, there are transportation ride sharing activities for commercial transportation.Whether it is rail, trucking, or even light vehicles, the "Uberization" of commercial transportation is also occurring.In these cases, excess freight vehicle capacity can be managed through blockchain activities including authentication of drivers and vehicles to payment through tokens.Ridesharing achieves efficiencies in utilization of vehicles, lessening waste.Authentication of green practices and vehicles increase transparency to customers of transportation sharing.
Packaging can be reused and traced; in this example, blockchain traceability can extend the packaging material life through more efficient management.Recyclable packaging can be monitored and managed more effectively as well.With this monitoring, further confirmation of socially responsible packaging can occur.In an application released by the U.S. Patent and Trademark Office (USPTO), Walmart describes a "smart package" that would include a device that would record information on a blockchain regarding the contents of the package, its environmental conditions, its location, and more.Additionally, multinational supermarket chain Carrefour is already using a similar system where customers can scan packaging for detailed information on a product's source, production processes, and environmental characteristics.
Building packaging with blockchain information transparency can improve green marketing efforts.According to green marketing theory, consumers are more likely to purchase greener products if they are confident that the product is actually green [37,38].This confidence increases with the transparent, verified, and immutable information from blockchains.Overall, substantial green consumer implications exist due to blockchain technology.Two examples of these blockchain activities are consumer token incentive systems to purchase green and product tracing for returning of end-of-life products by consumers.Substantial green consumer theories, including social confirmation theories to perceived behavioral control, can be used to explain the blockchain benefits for green consumer behavior and action [38].
Waste Management
Organizational waste management along the supply chain is critical to many sustainable supply chain activities.Waste minimization is the ultimate goal for organizations and supply chains.However, if waste is generated, then tracking is critical for reasons related to the circular economy and industrial symbiosis.It may also be critical from the perspective of waste disposal and potential liabilities associated with disposal.
For waste minimization purposes, smart contracts can be used to ensure waste is minimized across the supply chain.Performance criteria for suppliers for waste reduction metrics can be included in smart contract execution agreements.Metrics and management around hazardous wastes, such as those identified by the toxics releases inventory (TRI) [39], can be tracked.Specific levels may be dictated in smart contracts for acceptable performance.The waste minimization angle may be to adjust and update smart contracts as part of a continuous improvement process for supply chains.Similar to carbon trading, waste trading can also be managed.
When minimization of waste is not possible, then there are opportunities to environmentally and sustainably manage this waste.One method of accomplishing this goal is to identify how and where the waste can be used to make it a by-product or to minimize its environmental impact.Waste exchanges have been utilized for effective industrial symbiosis realization, expanding the scope from local to national levels [40].One especially cogent application, not typically considered, is the exchange of construction waste from the construction supply chain.In the construction case, there is a strong argument for 'buildings as material banks', which can be effectively managed through blockchains and the Internet [41,42].More on this issue in terms of traceability and verification are described in the reverse logistics discussion.
Many times, eliminating waste completely from the supply chain is impossible.When this occurs, sustainably managing the waste is required.In this situation, the waste management supply chain processes need to be managed and risk plays a significant role.Risk is especially pertinent when managing hazardous wastes, which is also an expensive undertaking.In the United States, the tracking of hazardous wastes is critical due to the long term possibility of becoming a potentially responsible party to superfund sites.This means that companies or even supply chain partners may be responsible for significant multi-million dollar cleanup costs associated with poorly managed landfills and company sites.Having a permanent record and tracking waste disposition can help manage these liability concerns.It may be valuable for government agencies for tracking responsibility of waste as well.There are a number of dimensions of waste management blockchain capabilities and limitations that have been reviewed [43].Fraud and manipulation, wrong or loss of information, manual processing, lack of knowledge about technology, and lack of control are all concerns for waste management in this environment.
Reverse Logistics
Reverse logistics are necessary for a number of take back regulations and building remanufacturing capabilities.One of the major concerns with remanufacturing and reverse logistics planning is the uncertainty in the location and supply of material at their end-of-life.Knowing the location of a material (i.e., traceability) to be taken back or remanufactured can help reduce uncertainty in the materials.Regulatory policies, such as the waste electrical and electronic equipment (WEEE) requirements, state that original equipment manufacturers (OEMs) are responsible for their goods.Thus, traceability of materials in the supply chain, as well as authentication that the material belongs to a particular OEM, improves the efficiency of the process for managing return flows.Mandated producer responsibility through regulations is one aspect; voluntary extended producer responsibility and product takeback can benefit from transparency, traceability, and authentication.
Similarly, circular economy practices have at least four levels of value recovery including product-life extension, reuse, remanufacture, and recycling [44].Other than tracing materials, as identified above, one aspect of the blockchain that may be important is the terms of exchange.Smart contracts may be set up where the financing of returns can be completed electronically.In this case, instead of transferring products and materials, some form of payment is required it is not possible to manage the actual finances.Payment may be based on the quality of the material, which can be traced by data, but also the history of the cost of the product or material.Together, these items can be evaluated and payment can be completed through blockchain payment systems.
The payment scheme is critical to attracting enough product or material to drive the product through the system.Some circular economy principles do not necessarily require that a product or material be at the end of its life, but that it is returned in some condition level.Having this information, such as the number of recycling cycles, or purchase date of a product, assist in determining values.Once the value is determined, the payment can be completed.The payment location may be critical as well.Since globalization of supply chains will continue, paying for the product or material, no matter where it exists, can be more easily completed using cryptocurrencies whose values can be based on local currencies.This supply chain finance application of blockchains along with transparency can enable circular economy practices.
Energy
Energy is an important resource for all supply chain activities and managing energy is central to greening a supply chain.Sustainable energy management typically has environmental relationships associated with air emissions, fuel resource usage, and issues such as biodiversity and hazardous materials emissions.The amount of energy required to run blockchain technology can become overwhelming, especially if there is a need to solve algorithms for solving hashes as part of smart contracts.It is not clear if mining will be necessary for supply chain activities and blockchains.
Distributed storage and operations will require significant energy requirements for electronic databases; as redundancies in data storage will potentially cause exponentially greater energy needs.
Energy-related blockchain activities may support supply chain sustainability.Some organizations and supply chains, in order to achieve zero greenhouse gas emissions, use carbon credit markets for carbon offsets.These markets have been controversial given the difficulties in tracing the location and validity of the offsets.Calculation and the additionality requirements have made them controversial [45].Improving transparency and clarity of a carbon credit can be effective using blockchain technology trust mechanisms.Also, not all carbon credits are created equal.For example, a carbon offset credit generated from a solar farm in a developed country may not have the same total environmental and sustainability of a carbon offset credit from an environmentally sensitive and poorer nation.In the latter situation, there may be more environmental co-benefits such as biodiversity management and offering poverty alleviation opportunities.
Internal emissions trading mechanisms for supply chains can also be better supported through transparency and information sharing from blockchain technology.This type of trading can provide financial benefits for energy use reduction by trading credits.Part of the trading and incentive mechanisms can be financially supported through cryptocurrency exchange [46].
Another example of energy-related blockchain improvement for supply chains is related to decentralized energy management within and between supply chain partners or communities.Rooftop solar power can be more accessible and economically feasible, further supporting adoption of renewable energy.Digital wallets as rewards may be one avenue for incentivizing employees and organizations to adopt more renewable energy along the blockchain.Expanding neighborhood blockchain-enabled micro-grid trading of solar energy to the supply chain, such as that supported by LO3 Energy, can provide certifiable and greener energy.
Research Concerns
Blockchain is a revolutionary technology with the potential to challenge supply chain processes and thought.Some research is required to understand the barriers, enablers, and diffusion of this technology [47].Additionally, research related to the influence of blockchain on sustainable supply chains at organizational strategic and operational levels, its supply chain, broader industry networks, and the macro-economy are all needed.
The research domain is quite broad for a technology that may prove disruptive to the status quo of practice.We will not delve deeply into each of these research questions, but only provide a general set of issues that require investigation.The research on blockchains, sustainability, and supply chains is in its infancy with the academic field fertile for sowing ideas, theories, and analysis; many of which will grow [48].For example, Francisco and Swanson [18] developed a conceptual model that incorporates the theory of acceptance and use of technology.This framework addresses the intentions in the use of blockchain technology for supply chain transparency.
As we are in the early phases of blockchain, adoption and diffusion of the technology is a general concern.Diffusion theory and technology acceptance models may require direct and explicit accounting for multiple stakeholders in acceptance of the diffusion.Multiple agents are involved in the adoption and agreement to be involved in the technology.Sustainability has heterogeneous meanings to participants in a supply chain.This heterogeneity and the need for blockchain may either hinder or aid diffusion.It can hinder diffusion because not everyone will agree that investment in such a technology would help with sustainability.It may be an enabler, since supply chains and partners may seek greater homogeneity and standardization of sustainability.Thus, the role of sustainable supply chains and philosophies and practices can play differentiating roles.Studies on what factors and constructs play a role in barriers or enablers are required.
Company, industry, product, and competitive environment characteristics may each influence the adoption of blockchain technology for sustainable supply chains.For example, in industries with a poor reputation, there might be greater adoption of transparency-based blockchains to support sustainability in supply chains.In this situation, legitimacy building and theory help supply chain participants address reputational issues.Similar issues based on other traditional supply chain characteristics, including building trust, opportunism, and relationship management, can play a role based on the context.Revisiting the various organizational theories for green and sustainable supply chains [49] will be necessary.Competing theories or joint theoretical perspectives are required.For example, will information technology theories such as structuration and internal organizational adoption be more important than technology acceptance and diffusion theory be better predictors?
The theory and research necessarily need to be interdisciplinary and multilevel.For example, the issues of various boundaries and boundary spanning aspects of the research (see [50]) in supply chains are vague.The boundaries and constraints of green and sustainable supply chains include economic, organizational, cultural, technological and proximal boundaries.Whether these and/or other boundaries play a role and where we draw the boundary for blockchain technology and sustainable supply chains becomes a research question related to the impact, effectiveness, performance, and general capabilities.
An overarching question in each of these general research areas is whether blockchain technology is idiosyncratic, unique, disruptive, or just another incremental technology following the status quo rules.Do blockchain technology and the blockchain environment follow similar rules as other supply chain technological and process innovations?In other words, how can adopting blockchains be compared to current information technologies such as enterprise and supply chain wide resource planning systems?Additionally, how can blockchain technology be integrated into the current legacy information systems in the supply chain?The applicability of the current supply chain theories in the blockchain domain is a concern.From an epistemological perspective we are arguably at the level of idealism, where the empirics of blockchain technology have yet to become reality.There is currently much hype and hope associated with blockchain technology where its justification is based primarily on faith and it is uncertain whether this hype and hope result in true verifiable and empirical outcomes.What we have posited in this section is a rationalistic argument that theory and empirics need to be applied collaboratively to evaluate the idealistic notions of blockchain technology and sustainable supply chains conjectured in this manuscript.
Limitations and Conclusions
Actual and potential blockchain sustainable supply chain use cases and applications are extensive.We have only skimmed the possibilities of blockchain application depths; as new technology, knowledge, and needs arise, more use cases will follow.Significant additional possibilities exist.The hype and potential profits associated with blockchain technology provide substantial creative motivation to identify numerous future applications.
Disruptive technologies tend to follow the 'technology mudslide hypothesis' [51].That is, coping with relentless technological change is analogous to climbing a mudslide raging down a mountain.Practitioners and researchers are scrambling to make sense of blockchain technology, where even stopping to take a breath can bury an individual or organization.Incorporating sustainability and supply chains is like adding boulders of different shapes and sizes to this proverbial technological mudslide.In this paper, we attempted provide an overview of the potential of blockchain technology in the sustainable supply chain context.
Admittedly, we have only scratched the surface of the roles that blockchain can play in sustainable supply chain management.Our primary focus was identifying potential uses across the spectrum of green supply chain management functions and activities, specifically on environmental sustainability in the supply chain.Our examination has significant extensibility to social sustainability, and some of this was made explicit in our discussion.
Given the more pragmatic perspective of this manuscript, we only briefly touched upon broader theoretical and philosophical concerns of blockchain technology in sustainable supply chains.A complete and detailed theoretical research evaluation of sustainable supply chain blockchain technology is still required, but true empirical and theoretical evaluation will mature as adoption matures.Moving beyond the hype and hope is necessary for rational determination of effectiveness.
The final thought we present is whether blockchain technology is a true disruptive social innovation, or is another affectation of incremental technology with limited strategic significance for sustainable supply chains.This question remains to be answered.
Figure 1 .
Figure 1.Blockchain application in green supply chain activities.Figure 1. Blockchain application in green supply chain activities.
Figure 1 .
Figure 1.Blockchain application in green supply chain activities.Figure 1. Blockchain application in green supply chain activities. | 2018-11-09T14:41:35.199Z | 2018-10-12T00:00:00.000 | {
"year": 2018,
"sha1": "d0cf0cf487ea5197ecfabfabed2e2f978a49dd96",
"oa_license": "CCBY",
"oa_url": "https://www.mdpi.com/2071-1050/10/10/3652/pdf?version=1539331276",
"oa_status": "GOLD",
"pdf_src": "ScienceParseMerged",
"pdf_hash": "d0cf0cf487ea5197ecfabfabed2e2f978a49dd96",
"s2fieldsofstudy": [
"Environmental Science",
"Business"
],
"extfieldsofstudy": [
"Economics"
]
} |
220981844 | pes2o/s2orc | v3-fos-license | Chloroplast SRP54s are Essential for Chloroplast Development in Rice
Background The chloroplast signal recognition particle 54 (cpSRP54) is known for targeting the light-harvesting complex proteins to thylakoids and plays a critical role for chloroplast development in Arabidopsis, but little is known in rice. Here, we reported two homologous cpSRP54s that affect chloroplast development and plant survival in rice. Results Two rice cpSRP54 homologues, OscpSRP54a and OscpSRP54b, were identified in present study. The defective OscpSRP54a (LOC_Os11g05552) was responsible for the pale green leaf phenotype of the viable pale green leaf 14 (pgl14) mutant. A single nucleotide substitution from G to A at the position 278, the first intron splicing site, was detected in LOC_Os11g05552 in pgl14. The wild type allele could rescue the mutant phenotype. Knockout lines of OscpSRP54b (LOC_Os11g05556) exhibited similar pale green phenotype to pgl14 with reduced chlorophyll contents and impaired chloroplast development, but showed apparently arrested-growth and died within 3 weeks. Both OscpSRP54a and OscpSRP54b were constitutively expressed mainly in shoots and leaves at the vegetative growth stage. Subcellular location indicated that both OscpSRP54a and OscpSRP54b were chloroplast-localized. Both OscpSRP54a and OscpSRP54b were able to interact with OscpSRP43, respectively. The transcript level of OscpSRP43 was significantly reduced while the transcript level of OscpSRP54b was apparently increased in pgl14. In contrast, the transcript levels of OscpSRP54a, OscpSRP43 and OscpSRP54b were all significantly decreased in OscpSRP54b knockout lines. Conclusion Our study demonstrated that both OscpSRP54a and OscpSRP54b were essential for normal chloroplast development by interacting with OscpSRP43 in rice. OscpSRP54a and OscpSRP54b might play distinct roles in transporting different chloroplast proteins into thylakoids through cpSRP-mediated pathway.
Background
Chloroplasts are the site for photosynthesis and other important metabolic processes such as fatty acid and amino acid biosynthesis (Nelson and Ben-Shem 2004;Lopez-Juez and Pyke 2005). The organelle contains up to several thousands of proteins, the majority of which are encoded in the nucleus with only a small fraction encoded in the plastid genome both in Arabidopsis and rice (Abdallah et al. 2000;Richly and Leister 2004). The nucleus-encoded chloroplast proteins are thus required to be transferred into the chloroplasts usually depending on the Toc/Tic complexes on the inner and outer chloroplast membranes (Jarvis and Robinson 2004;Schwenkert et al. 2011). Once these proteins enter into the chloroplast stroma, they may take on the final form, or further transport into the thylakoids through four distinct transport pathways namely, cpSec, ΔpH/Tat, cpSRP and spontaneous pathways (Schünemann 2007;Jarvis 2008). The chloroplast signal recognition particle (cpSRP) and its receptor (cpFtsY) are involved in transporting of chloroplast proteins such as mature light-harvesting chlorophyll a/b binding proteins (LHCPs) to thylakoid membranes (Akopian et al. 2013). Defective or deficient in cpSRP and cpFtsY would impair the biogenesis of chloroplasts, leading to chlorophyll degeneration and chlorotic leaves in plants (Amin et al. 1999;Klimyuk et al. 1999;Asakura et al. 2004;Rutschow et al. 2008;Lv et al. 2015). In Arabidopsis, two null mutants of cpSRP54, ffc1-1 and ffc1-2, produce yellow first true leaves that become green subsequently. The levels of reaction center proteins are significantly lower in the young ffc1-2 leaves but recover to the normal protein level in adult plants grown on agar plates (Amin et al. 1999). However, the ffc1-2 plants are pale green with a fewer number of leaves and reduced rosette diameter grown under soil conditions (Rutschow et al. 2008). The null mutant of Arabidopsis cpSRP43, chaos, exhibits pale green leaves including the first true leaves in the whole lifecycle, with elevated chlorophyll a/b ratio, but a normal level of reaction center proteins (Amin et al. 1999;Klimyuk et al. 1999). The double mutant ffc/chaos exhibits pale yellow leaves at all growth stages with drastically reduced levels of LHCPs except Lhcb4 (Hutin et al. 2002). Similarly, the chlorophyll synthesis and chloroplast development are impaired in company with altered expression of chlorophyll synthesis-associated genes in rice OscpSRP43 mutants, w67 and pgl3, both of them show pale green leaves at all growth stages (Lv et al. 2015;Ye et al. 2018). Furthermore, high temperatures inhibit plant growth and facilitate the progression of leaf senescence in pgl3 (Ye et al. 2018). The maize cpFtsY mutant csr1-1 exhibits a pale yellow-green phenotype while csr1-3 shows a slight pale green phenotype (Asakura et al. 2004). Interestingly, both csr1-1 and csr1-2 are seedling lethal, similar to the Arabidopsis mutants cpFtsY-1 and cpFtsY-2, with completely arrestedgrowth at the cotyledon or the first true leaf stage under photoautotrophic conditions (Asakura et al. 2004;Asakura et al. 2008). It has been shown that the mutation lines of cpSRP pathway genes are able to accumulate truncated light-harvesting chlorophyll antenna (TLA) and enhance energy conversion efficiency in highdensity cultures under bright sunlight conditions in Chlamydomonas reinhardtii (Kirst and Melis 2014;Jeong et al. 2017). Similar phenomena have been observed in tobacco, the cpSRP43 knockdown plants show lower chlorophyll contents in the RNAi canopy leaves with increased leaf-to-stem ratio, improved photosynthetic productivity and canopy biomass accumulation under high-density cultivation conditions (Kirst et al. 2018).
We previously identified a stable-inherited rice pale green leaf 14 (pgl14) mutant (originally termed HM14) at the vegetative stage (Shi et al. 2013). In this study, we isolated PGL14 that encoded for cpSRP54 (hereafter cpSRP54a). A single base substitution in the mutant allele resulting in an altered mRNA splicing is responsible for the pale green phenotype confirmed by genetic complementation. We also isolated a homologue of PGL14, LOC_Os11g05556 (hereafter cpSRP54b). Although the knockout lines of cpSRP54b displayed a similar pale green leaf phenotype to pgl14 with reduced chlorophyll levels, impaired chloroplast structures and down-regulated expression of chlorophyll synthesis/development related genes, they were seedling lethal. Both cpSRP54a and cpSRP54b were chloroplast-localized and could interact with OscpSRP43, respectively. Our results indicated that both of them are required for chloroplast development by interacting with cpSRP43 to potentially participate in protein transport into thylakoids in rice.
Map-Based Isolation of PGL14
We previously identified a chlorophyll-deficient mutant pgl14 exhibiting pale green leaf phenotype from the first leaf to the flag leaf under natural conditions and mapped the recessive mutation to a 299 kb region in chromosome 11 (Shi et al. 2013). To fine map the mutation, a total of 1008 mutant-type F 2 individuals derived from the cross pgl14/Moroberekan were used for genotyping. The PGL14 locus was further narrowed down to a 39.5 kb genomic region between RM26076 and RM26079, covering the BAC clones AC116949 and AC138169 (Fig. 1a). Seven open reading frames (ORFs) were annotated within this region in the database of Rice Genome annotation Project (http://rice.plantbiology.msu.edu/cgibin/gbrowse/rice/), two of which (LOC_Os11g05552 and LOC_Os11g05556) were both annotated as hypothetical loci encoding for cpSRP54. Sequence analysis showed that a single nucleotide substitution from G to A at position 278 was detected in LOC_Os11g05552 in pgl14, and the mutation localized to the predicted splicing site on the last nucleotide of the first intron. RT-PCR analysis showed that the transcript in pgl14 was longer than that of WT, confirming the presence of the altered splicing transcript in the mutant (Fig. 1b). Sequence analysis also showed that the first intron of 119 bp was maintained in the mutant transcript which had a frame shift starting from valine at position 54 and terminating prematurely at position 131 (Fig. 2, Genbank accession MN105082). Therefore, LOC_Os11g05552 was most likely the candidate gene of PGL14.
OscpSRP54a Rescues the Pale Green Phenotype
To verify the function of PGL14, the complementation vector cPGL carrying the entire coding region of PGL14, a 3.6 kb of upstream sequence and a 1.5 kb of downstream sequence was transformed into the rice calli induced from pgl14 mature embryo through Agrobacterium tumefaciensmediated transformation. Five independent transgenic lines were obtained and showed the normal green phenotype similar to WT (Fig. 3a). Sequence characterization of these transgenic plants indicated that the complementary plants displayed a double peak (A and G) while pgl14 and WT presented a single peak, respectively (Fig. 3b), indicating that the wild-type allele has been incorporated into the mutant genome. Furthermore, RT-PCR analysis confirmed that both transcripts were presented in the transgenic plants (Fig. 3c). For the pigment levels, pgl14 showed apparently lowered Chl a, Chl b, carotenoid (Caro) and total chlorophyll (Chls) contents with significantly elevated Chl a/b ratio compared with WT, and all these parameters recovered to WT levels in complementary lines (Fig. 3d).. In addition, the chloroplast ultrastructure of complementary line C-pgl14 was similar to that of WT, displaying normal thylakoid membranes and stacked grana ( Fig. 3e-g). Taken together, PGL14 was indeed the target gene responsible for the pale green leaf phenotype in pgl14, hereafter PGL14 is termed OscpSRP54a.
cpSRP54s are Conserved in Plants
The cpSRP54 is widely present in photoautotrophic organisms. To simplify the phylogenetic analysis, we chose Fig. 1 Map-based cloning of PGL14. a PGL14 localizes to the short arm on chromosome 11 between RM26076 and RM26079, and is narrowed down to a 39.5 kb region covering the bacterial artificial chromosome clones AC116949 and AC138196. The 39.5 kb region contains 7 putative ORFs, the black box indicates LOC_Os11g05552, gray boxes indicate the other ORFs. LOC_Os11g05552 consists of 15 exons and 14 introns indicated as blank boxes and lines, respectively. The black arrow indicates the point mutation (G278A) at 1st intron splicing site. PTsF and PTsR are the forward and reverse primers for PCR analysis in (b). F and R are forward and reverse primers for qRT-PCR analysis; b RT-PCR shows different transcripts from WT and pgl14. The genomic DNA was used as a control 11 OscpSRP54a homologues representing 10 species including the monocot, dicot and algae. Blastp analysis showed that OscpSRP54a shared 47-86% identity at the amino acid level with various cpSRP54 from Chlamydomonas reinhardtii (47%), Arabidopsis thaliana (70%), Glycine max (75%), Nicotiana tobacum (79%), Triticum urartu (81%), Zea mays (85%), Sorghum bicolor (86%), Chlorella sorokiniana (57%) and Spirulina subsalsa (55%) respectively (Fig. 4a). In addition, OscpSRP54a has 78% amino acid identity to OscpSRP54b which possess 47-78% amino acid identity with cpSRP54 from the other species. Apparently, the more closer relationship, the higher the amino acid identity shows in various cpSRP54s in which the function required for chloroplast development has been extensively elaborated in Arabidopsis and C. reinhardtii (Li et al. 1995;Yu et al. 2012;Jeong et al. 2017). The results indicated that cpSRPs were highly conserved in plant species.
To reveal the evolutionary relationship of the cpSRP54 homologues, a phylogenetic tree was constructed. The results showed that cpSRP54 homologues in the higher plants could be classified into two groups as the monocot and dicot (Fig. 4b). Both OscpSRP54a and OscpSRP54b are clustered in the monocot group, however, OscpSRP54a is more closely related to TucpSRP54, ZmcpSRP54 and SbcpSRP54 than OscpSRP54b, probably indicating that OscpSRP54a and OscpSRP54b were likely originated differently.
OscpSRP54b is Essential for Chloroplast Development
To verify whether OscpSRP54b functions similar to OscpSRP54a, we introduced the CRISPR/Cas9 vectors cr1 and cr2 into Kitaake-derived embryogenic calli to edit the two specific target sites in exon 2 and exon 5, respectively (Fig. 5a). A total of 7 and 5 independent T 0 transgenic lines were obtained from construct cr1 and cr2, respectively. Two (cr1-2 and cr1-5) out 7 T 0 lines from construct cr1 were homozygous and exhibited pale green phenotype similar to pgl14 (Fig. 5b). Sequence analysis confirmed that there was one base deletion at the target site in cr1-2 and cr1-5, respectively (Fig. 5a). One (cr2-3) out of 5 T 0 lines from construct cr2 was homozygous and showed pale green leaf phenotype similar to pgl14 (Fig. 5b), and sequence analysis showed that cr2-3 had a single nucleotide deletion at the target site ( Fig. 5a). We then further characterized the performance of cr1-2 and cr2-3. Firstly, we determined the levels of photosynthetic pigments, and found that the contents of Chl a, Chl b, total Chls and carotenoid in cr1-2 and cr2-3 were significantly lower than those of Kitaake, and the Chl a/b ratio of both knockout mutants were much lower than in the WT (Fig. 5c). Then we observed the chloroplast ultrastructure by transmission electron microscopic analysis. The results showed that both cr1-2 and cr2-3 possessed a large number of hollow vesicles, reduced number of grana, irregular grana thylakoids, and destroyed stromal lamella compared with Kitaake ( Fig. 5d-f). Moreover, cr1-2, cr2-3 as well as cr1-5 showed apparently arrested-growth and died within 3 weeks at the seedling stage. These results suggested that OscpSRP54b was essential for the chloroplast development and plant survival in rice.
Rice cpSRPs are Mainly Expressed in Shoots and Leaves and cpSRPs Localize to Chloroplasts
To examine the expression pattern of OscpSRP54a and OscpSRP54b, qRT-PCR was carried out using samples from various tissues at the germination, tillering and heading stages. The results showed that the expressions of OscpSRP54a and OscpSRP54b were detected in all tissues tested with the highest expression levels of OscpSRP54a and OscpSRP54b in the leaves at the tillering stage, followed by the shoots at the germination stage (Fig. 6). Our results indicated that the expression patterns were similar between the two genes which were mainly expressed in above ground parts of the plants.
To determine the subcellular location, we first predicted their physical locations using the ChloroP program (http://www.cbs.dtu.dk/services/ChloroP) and the results showed that both OscpSRP54a and OscpSRP54b were located in chloroplasts ( Supplementary Fig. 1). The subcellular localization of OscpSRP54a and OscpSRP54b were then confirmed by expressing the constructs PAN580-OscpSRP54a and PAN580-OscpSRP54b in protoplasts. The green fluorescence signals of OscpSRP54a: GFP fusion protein and OscpSRP54b:GFP fusion protein overlapped with the chlorophyll autofluorescence signal whereas the free GFP signal was observed in the cytoplasm and nucleus (Fig. 7). These results demonstrated that both OscpSRP54a and OscpSRP54b were chloroplast-targeted proteins.
Down-Regulated Expression of OscpSRP43 in pgl14 and cr1-2 It has been shown that cpSRP43 interacts with cpSRP54 and is critical for chloroplast development by transporting proteins to thylakoids (Schünemann 2007;Akopian et al. 2013). To investigate whether the expression of cpSRP43 was affected in pgl14 and cr1-2, we measured the transcript levels of OscpSRP43, OscpSRP54a and OscpSRP54b b Sequence analysis of the mutation site (red arrow) in WT, pgl14 and C-pgl14; c OscpSRP54a transcripts in WT, pgl14 and C-pgl14; d Pigment contents in WT, pgl14 and C-pgl14 in 8-week-old leaves. Data are means ± SD (n = 3). Means with different letters indicate significant differences according to One-way ANOVA and Duncan's test (p ≤ 0.01). Chloroplast ultrastructure of WT (e), pgl14 (f) and C-pgl14 (g) at the tillering stage. G, grana thylakoid; S, starch granule by qRT-PCR. The results showed that the transcript level of OscpSRP43 was significantly decreased in pgl14 compared with WT, in contrast, the expression level of OscpSRP54b was significantly increased in pgl14 compared with WT, whereas the expression level of OscpSRP54a was similar between pgl14 and WT (Fig. 8a). In addition, the transcript levels of OscpSRP43, OscpS RP54a and OscpSRP54b were all notably reduced in cr1-2 compared with Kitaake (Fig. 8b). The results suggested that both mutations of cpSRP54a and cpSRP54b resulted in down-regulated expression of cpSRP43 in rice.
The expression of genes associated with chlorophyll biosynthesis and chloroplast development was examined in pgl14 and cr1-2. The results showed that expression profile of most genes (such as PsbA, YGL1 and ChlD) was altered between the mutant and its wild type (Supplementary Fig. 2). These results further indicated that both OscpSRP54a and OscpSRP54b Fig. 5 Functional verification of OscpSRP54b for chloroplast development. a Deletion mutation at the target site in three representative knockout lines generated by the CRISPR/Cas9-mediated editing. cr1-2, cr1-5 and cr2-3 are homozygous mutants carrying 1-bp deletion on both homochromosomes. Black boxes indicate exons and lines indicate introns of OscpSRP54b. The sgRNA target sequence is underlined in blue and the PAM motif is highlighted in red letters, F and R are the forward and reverse primers for qRT-PCR analysis; b Phenotype of Kitaake and OscpSRP54b knockout mutants. Bar = 2 cm; c Pigment contents in 1 week-old leaves of Kitaake, cr1-2 and cr2-3. Different letters indicate significant differences according to One-way ANOVA and Duncan's test (p ≤ 0.01); Chloroplast ultrastructure of Kitaake (d), cr1-2 (e) and cr2-3 (f). G, grana thylakoid; S, starch granule; OG, osmiophilic plastoglobuli; SL, stroma lamellae; HV, hollow vesicle It has been shown that AtcpSRP54 interacts with AtcpSRP43 to form the heterodimer which binds to the L18 sequence of LHCPs to form the cpSRP-LHCPs complex for transporting LHCPs to thylakoids (Groves et al. 2001;Goforth et al. 2004). To verify whether OscpSRP 54a and OscpSRP54b interact with OscpSRP43, the full length CDS of OscpSRP54a and OscpSRP54b were fused to C-terminal CFP respectively, and the full length CDS of OscpSRP43 was fused to N-terminal Venus. Coexpression of the OscpSRP54a-cCFP and OscpSRP43-nVenus fusion proteins in rice green tissue protoplasts produced obvious YFP signals overlapped with the auto fluorescence of chloroplasts (Fig. 9). The similar result was obtained by co-expression of OscpSRP54b-cCFP and OscpSRP43-nVenus fusion proteins. In contrast, coexpression of OscpSRP54a-cCFP and OsCSP41b-nVenus, or OscpSRP54b-cCFP and OsCSP41b-nVenus did not show the BiFC fluorescence (Fig. 9). Similarly, coexpression of OscpSRP54a-nVenus and OsCSP41b-cCFP, or OscpSRP54b-nVenus and OsCSP41b-cCFP did not show the BiFC fluorescence (Fig. 9). These results clearly demonstrated that both OscpSRP54a and OscpSRP54b interacted with OscpSRP43, respectively.
Discussion
Deficient chlorophyll contents shown by Arabidopsis cpSRP mutants indicate that cpSRP subunits play important roles for chloroplast development (Pilgrim et al. 1998;Amin et al. 1999;Hutin et al. 2002; Walter et al. . Further studies on cpSRP43 and cpSRP54 defective mutants in Chlamydomonas and Arabidopsis suggest that the cpSRP43/SRP54 complex are able to recognize and bind to the hydrophobic LHCPs during passing through the stroma, thus, the dysfunctional cpSRP43 and/or cpSRP54 disrupt their assembly with LHCPs in thylakoids, leading to impaired chloroplast development (Amin et al. 1999;Hutin et al. 2002;Jeong et al. 2017). In the present study, we identified a chlorophylldeficient mutant pgl14, which possessed a single nucleotide substitution at the splicing site of OscpSRP54a, leading to altered splicing transcripts and terminated prematurely. Complementation by the wild type allele could restore the pgl14 phenotype. OscpSRP54a is a homologue of Arabidopsis cpSRP54 with 70% identity at the amino acid level. The ffc mutants defective in AtcpSRP54 show severely yellow true leaves that subsequently become green (Amin et al. 1999). Unlike the ffc mutants, pgl14 exhibited pale green leaf phenotype in the whole life period. It is noticed that the rice mutant ygl138 has an 18 bp deletion in OscpSRP54 and shows a similar phenotype to pgl14, suggesting that ygl138(t) is allelic to PGL14 . Unlike Arabidopsis which possesses only one copy of cpSRP54, the rice genome possesses two copies of cpSRP54s, OscpSRP54a and OscpSRP54b, which are physically adjacent and share 78% identity at the protein level. Independent knockout plants of OscpSRP54b were seedling lethal but showed pale green leaf phenotype and severely decreased chlorophyll content, resembling to pgl14. These results suggest that both OscpSRP54a and OscpSRP54b are indirectly associated with chlorophyll metabolism probably resulting from the impaired chloroplast development in the mutants.
It has been shown that AtcpSRP54 proteins are equally distributed between thylakoids and stroma by immunolocalization (Hutin et al. 2002). In the present study, we demonstrated that both OscpSRP54a and OscpSRP54b localized to chloroplasts, similar to OscpSRP43 (Lv et al. 2015). The thylakoid membranes in pgl14 seem to be disrupted, and the grana stacks were thinner than those of the wild-type at the tillering stage ( Fig. 3e-g). Nevertheless, pgl14 is viable and capable of seeding at maturity (Shi et al. 2013). However, more severely destroyed stroma and a large number of hollow vesicles were found in the chloroplasts of the OscpSRP54b-knockout lines. The severe and irreversible chloroplast destruction could be the reason for the lethality of OscpSRP54bknockout mutants. These results implicate that functional OscpSRP54b, but not OscpSRP54a, is necessary for rice survival.
Biochemical and genetic studies have indicated that cpSRP54 and cpSRP43 are able to form the heterodimeric complex to participate in localizing LHCPs to thylakoid membranes post-translationally (Hutin et al. 2002;Goforth et al. 2004;Dünschede et al. 2015). Our results suggested that both OscpSRP54a and OscpSRP 54b could interact with OscpSRP43, respectively; indicating that OscpSRP54a and OscpSRP54b both functioned similarly as AtcpSRP54 for cpSRP-mediated protein transportation (Amin et al. 1999;Yu et al. 2012). Nevertheless, the existence of an alternative pathway has been confirmed in Arabidopsis that the targeting of LHCPs to thylakoids can be done when cpSRP54 is absent and unable to form the cpSRP-LHCP transit complex (Tzvetkova-Chevolleau et al. 2007). It has been shown that Glutamyl-tRNA reductase (GluTR) is the initial and rate-limiting enzyme for 5-aminolevolinic acid synthesis. Recent studies have demonstrated that cpSRP43 directly binds to GluTR and prevents aggregation of GluTR, thereby enhancing the stability of active GluTR . In pgl14 and cr1-2, the transcription levels of OscpSRP43 were significantly lower than those of the wild types, we speculated that downregulation of cpSRP43 in the mutants could lead to the decrease of stability and catalytic activity of GluTR, thus inhibiting the chlorophyll biosynthesis and ultimately leading to a decreased level of chlorophyll in the mutants. Interestingly, OscpSRP54b was significantly upregulated in pgl14, whereas OscpSRP54a was notably downregulated in cr1-2 mutant (Fig. 8). We speculate that OscpSRP54b might partially compensate for the defect of OscpSRP54a in pgl14, however, the severe destruction of chloroplasts in cr1-2 inhibited the transportation of LHCPs to thylakoids, resulting in the significantly down-regulation of both OscpSRP54a and OscpSRP54b.
It has been shown that the downregulation of genes involving in chlorophyll biosynthesis, photosynthesis and chloroplast development could be an indirect response to chlorophyll-deficient mutants (Yu et al. 2012;Lv et al. 2015;Qiu et al. 2018). For example, AtcpSRP54 mutations lead to decreased expression of AtGLK1, AtGLK2 and GUN4 which are related to plastid-to nucleus retrograde signaling (Lopez-Juez and Pyke 2005). In our study, the expression level of CHLH was significantly decreased in pgl14 and cr1-2. CHLH, a Mg-chelatase H subunit, is a multi-functional protein involved in plastidto nucleus retrograde signaling and chlorophyll synthesis (Mochizuki et al. 2001;Jung et al. 2003;Wu et al. 2009;Tsuzuki et al. 2011). AtCpSRP54 is found to be associated with chloroplast ribosomes in the stroma, interacts with chloroplast synthesized thylakoid membrane proteins D1 and cytochrome b 6 to perform its conserved role in co-translational targeting (Nilsson et al. 1999;Nilsson and vanWijk 2002;Piskozub et al. 2015). Similarly, PsbA encoding chloroplast D1 protein is significantly downregulated in pgl14 compared with the wild type. In contrast, the knockout mutant of OscpSRP54b induced a notable upregulation in PsbA transcription level. Studies in Synechocystis sp PCC 6803 have demonstrated that chlorophyll synthase/HliD complex binding with the Ycf39 protein and YidC/Alb3 insertase is involved in the photosystem II assembly, suggesting a link between chlorophyll biosynthesis and the Sec/YidCdependent cotranslational insertion of nascent photosystem polypeptides into membranes (Chidgey et al. 2014;Knoppová et al. 2014). In our study, the expression level of YGL1, the homologue of Synechocystis sp PCC 6803 chlorophyll synthase in rice, was significantly downregulated in pgl14, but was significantly upregulated in cr1-2, these results implied that OscpSRP54b and OscpSRP54a might play distinct roles in transporting different chloroplast proteins into thylakoids through cpSRP-mediated pathway although the mechanism requires to be furthered studied.
Plant Materials
The pale green leaf 14 (pgl14) mutant was obtained from ethane methyl sulfonate (EMS) mutagenesis of the wildtype (WT) cultivar IR64 (Wu et al. 2005). The pale green phenotype is controlled by a single recessive nuclear gene (Shi et al. 2013). Normal green leaf cultivar Moroberekan was used as the male parent to cross with pgl14 for construction of an F 2 fine mapping population. The parents and the population were grown under natural summer conditions in the paddy field at the China
Measurement of Pigment Content
The total chlorophylls (Chl) were extracted from 10 mg fresh leaves with 95% alcohol in darkness for 48 h. The extracts were measured spectrophotometrically at 470 nm, 645 nm and 663 nm with a SpectraMax i3x Multi-Mode Microplate Reader (Molecular Devices, USA). Total Chl contents were determined according to the method of Arnon (Arnon 1949), and total carotenoid contents were determined as described by Wellburn (Wellburn 1994). All experiments were carried out with three biological replicates. Student's t-test was conducted using EXCEL2013 and Duncan's test was conducted by SAS 9.0. Means from three replicates were used for analysis.
Transmission Electron Microscopy Analysis
Full expanded leaves of pgl14, WT and C-pgl14 were collected at the seedling stage, while full expanded leaves of Kitaake, cpSRP54b knockout plants cr1-2 and cr2-3 were collected 1 week after transplanting. Leaf sections were fixed with 2.5% glutaraldehyde in phosphate buffer (pH 7.2) for 16 h at 4°C, followed by rinsing, dewatering, embedding, and staining according to the method described by Lv et al. (2015). The chloroplast ultrastructure was observed by a Tecnai G 2 F20 S-TWIN transmission
Map-Based Cloning of PGL14 and Complementation Assay
The mutation was previously mapped to a 299 kb region in chromosome 11 (Shi et al. 2013). A total of 1008 mutant type F 2 individuals derived from the cross pgl14/ Moroberekan were used for fine mapping using simple sequence repeat (SSR) markers (Supplementary Table S1).
The genomic DNA was extracted following the minipreparation method (Lu and Zheng 1992). The genomic DNA fragments of the candidate gene were amplified from WT and pgl14, then sequenced and compared using DNASTAR software. RT-PCR analysis was used to confirm the splicing site in WT and pgl14 using the primers PTs (Supplementary Table S2). The sequences of the genomic DNA fragments and the transcripts were determined at Shanghai Invitrogen Inc. (Shanghai, China).
For functional complementation, a 9.6 kb WT genomic fragment containing a 4.5 kb entire open reading frame (ORF) of PGL14, a 3.6 kb upstream region, and a 1.5 kb downstream region was amplified using the specific primers PQf (Supplementary Table S2). The PCR products were double-digested with BamH I and Kpn I, and the fragments were recovered using the Axygen DNA gel extraction kit (Axygen scientific, USA). Then, the fragments were cloned into the binary vector pCAMBIA1300 to form a new transformation construct, cPGL. The new construct was introduced into the embryogenic calli generated from the mature seed embryos of pgl14 using Agrobacterium-mediated transformation method (Hiei and Komari 2008).
Sequence Alignment and Phylogenetic Analysis
BlastP (https://www.ncbi.nlm.nih.gov/) was used to search homologous protein sequences of OscpSRP54a. The homologous sequences were aligned using the BioEdit software. The neighbor-joining phylogenetic tree was constructed using MEGA 5.1. 1000 bootstrap replicates were used for statistical support for the node values.
Quantitative Reverse Transcription PCR
Total RNA was extracted using the TRIzol method following the manufacture's instruction (Invitrogen, USA). For RNA isolation, the roots and shoots of WT and Kitaake were collected from 10 day-old seedlings, the top full expanded leaves, leaf sheaths, roots, root stems, nodes and internodes of WT and Kitaake were collected from 10 week-old plants, roots, basals, nodes, internodes, flag leaves, flag leaf sheaths and panicles of WT and Kitaake were collected at the grain filling stage. For quantitative reverse transcription (qRT-PCR) analysis of genes associated with chlorophyll biosynthesis and chloroplast development, total RNA was extracted from the top full expanded leaves of pgl14 and WT at 3 weeks after sowing. Total RNA was extracted from 1 week-old leaves of cr1-2 and Kitaake. The first-strand cDNA was synthesized using the First Strand cDNA synthesis kit following the manufacturer's protocol (TOYOBO Biotech, Japan). qRT-PCR was performed in a total volume of 20 μL qRT-PCR reaction buffer containing 2 μL reverse-transcribed product, 0.2 μM of each primer, and 2 × PowerUp SYBR Green PCR Master Mix (ThermoFisher Scientific, USA), on a Thermal Cycle Dice TM Real Time System II (Takara Biotech, Japan) with a cycling program of 2 m at 50°C, 2 m at 95°C, followed by 40 cycles of 15 s at 95°C, 15 s at 55°C, and 60 s at 72°C. The ubiquitin gene (LOC_Os03g13170, Ubq) was used as an internal control. Primers used for qRT-PCR are listed in Supplementary Table S3. The means from three biological replicates were used for analysis by Student's t-test and Duncan's test by EXCEL2013 and SAS 9.0, respectively. The 2 -ΔΔCT method was used to determine the relative transcript levels in gene expression.
Subcellular Localization and Bimolecular Fluorescence Complementation Assay
To determine the subcellular localization of OscpSRP54a and OscpSRP54b, their full length CDSs were amplified using the specific primers SLPGL14 and SLCr1, respectively (Supplement Table S2). The PCR products were double-digested with Xba I and BamH I, and the fragments were inserted into the 5′-terminal of GFP driven by the CaMV 35S promoter in the transient expression vector PAN580 to form the new constructs, PAN580-OscpSRP54a and PAN580-OscpSRP54b, respectively. For BiFC assay, the full length CDSs of OscpSRP54a and OscpSRP54b were amplified using the specific primers BiPGL14, and BiCr1. The full length CDS of OsCSP41b, which encodes for a chloroplast-localized protein (Mei et al. 2017), was amplified using the primer BiCSP41b and fused with cCFP and nVenus fragment as a control. The PCR products were double-digested with Kpn I and BamH I, and the fragments were inserted to the 5′-terminal of cCFP driven by the CaMV 35S promoter in the expression vector pE3449 to form three new constructs, OscpSRP54a-cCFP, OscpSRP54b-cCFP and OsCSP41b-cCFP, respectively. These fragments were inserted to the 5′-terminal of nVenus in pE3308 to form OscpSRP54a-nVenus, OscpSRP54b-nVenus and OsCSP41b-nVenus constructs. The full length CDS of OscpSRP43 was amplified using the primers BiW67 (Supplemental Table S2), and double-digested with Kpn I and Sma I, then the fragments were inserted to the 5′terminal of nVenus driven by the 35S promoter in pE3308 to generate a new construct OscpSRP43-nVenus. The constructs were transformed into rice protoplasts according to the protocol described previously (Zhang et al. 2011).
Conclusions
OscpSRP54a and OscpSRP54b encode two homologous chloroplast signal recognition particles and their loss of function led to pale green leaves. Both OscpSRP54a and OscpSRP54b localize to the chloroplast and are able to interact with OscpSRP43, respectively. These results will facilitate efforts to further uncover the molecular mechanism of chloroplast protein transporting in monocots.
Additional file 1 Supplementary Fig. 1 Subcellular localization prediction of OscpSRP54a and OscpSRP54b using ChloroP program. Supplementary Fig. 2 Expression of genes associated with chlorophyll biosynthesis and chloroplast development. Supplementary Fig. 3 Schematic structure of cr1 and cr2. Supplementary Table S1 Details of SSR markers for fine mapping of pgl14. Supplementary Table S2 Primer sequences for vector construction and reverse transcription-PCR. Supplementary Table S3 Primer sequences for RT-PCR. | 2020-08-06T14:35:57.529Z | 2020-08-06T00:00:00.000 | {
"year": 2020,
"sha1": "457a0583f443c19ec71831aa3f98a676b548aba3",
"oa_license": "CCBY",
"oa_url": "https://thericejournal.springeropen.com/track/pdf/10.1186/s12284-020-00415-2",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "0973a46106c93c91ce71c93d08f2c772000cdfce",
"s2fieldsofstudy": [
"Biology",
"Environmental Science"
],
"extfieldsofstudy": [
"Medicine",
"Biology"
]
} |
50782760 | pes2o/s2orc | v3-fos-license | Epidemiological and clinical features of the emergency visits in a rural hospital in Cubal, Angola
Introduction There is scarce information on the profiles of patients attended in the Emergency Departments (ED) in rural Angola. Methods Retrospective descriptive study including all the patients treated in the ED in Hospital Nossa Senhora da Paz (Cubal) during 6 months (December 2014- May 2015). The epidemiological and clinical data collected were: age, sex, shift, service assignment, reason for consultation and outcome (discharge, admission, referral or death). Results A total of 2384 patients (53.4% women) were attended. The median age was 10 years (range: 0 - 96 years); 57.9% and 40.2% of them were under 17 and 5 years, respectively. No differences were observed regarding the assistance per shift, weekdays, weekends, or mean age per shift. The reason for consultation was registered in 69.9% of the patients; the most common were respiratory tract infections (20.5%), fever (14%), digestive diseases (13.6%) and malaria (10.4%). Up to 47.2% of the patients required in-hospital treatment and 1.3% were transferred to other hospitals. The patients admitted were significantly younger than the patients discharged (median age of 4 vs.16 years, p < 0.01). The mortality rate within the ED was 0.5%. Conclusion Young patients were those who mostly required assistance in the ED. Infectious diseases were the most frequent reason for consultation. Pulmonary tuberculosis was suspected in one third of respiratory infections. The admission rate was high, especially in children under 5 years and in cases of malaria and malnutrition. Low referral rate and low mortality within the ED were observed.
Introduction
Angola was stuck in a thirty-year civil war until 2002 after achieving its independence from Portugal in 1975. It has an estimated population of 25 million inhabitants, 48% of them under-15 years of age. Half of the population lives in non-urban areas, over one third is below the poverty threshold, and only one half has access to drinking water sources and waste treatment systems. The rate of physicians and nurses per 1000 inhabitants are 0.077 and 1.7, respectively. Life expectancy is currently about 51 years in men and 54 years in women. Despite a decreasing trend in the previous years, Angola is still one of the African countries with the highest maternal and under-5 year mortality rates [1][2][3]. Communicable diseases are the responsible of more than 50% of mortality in Angola. Measles, rabies, cholera, leprosy and yellow fever outbreaks are periodically reported [4]. This is considered an area with high transmission rate of Plasmodium falciparum malaria, with more than 2 million confirmed cases in 2014, although its related mortality has recently decreased [2,5]. Despite being considered underestimated, the prevalence of HIV infection among adults is 2.4%, with an antiretroviral treatment coverage of 42% of the population infected. Incidence of tuberculosis (TB) keeps increasing, with no conclusive data on multidrug-resistance related. The main causes of mortality are diarrheic diseases, respiratory infections and malaria. The main causes of child mortality are malaria (up to 33%), prematurity, asphyxia and neonatal tetanus. While mortality from caloric-protein malnutrition, respiratory infections, diarrheic syndromes and meningitis has decreased, it has increased mortality derived from complications of pre-term delivery and neonatal causes. It has also been reported an increase in other causes of mortality such as HIV, and cardiovascular causes (stroke or ischemic heart disease) [1,6].
The Hospital Nossa Senhora da Paz is in a peri-urban area of Cubal Sede, in Cubal municipality (province of Benguela) with an estimated population of 322000 inhabitants, where nearly half of them are under 15 years of age [7]. It has 294 beds (including 30 in the Malnutrition Unit and 150 in the TB Unit), an outpatient clinic, basic laboratory, and radiology and ultrasound service. The Emergency Department (ED) has four beds and is attended by a nurse with training in diagnosis and treatment on a 24-hour work shift with an on-call doctor if needed. Emergency care and treatment is accessible for all the population free of charge, whereas x-rays and lab tests are paid on an affordable tax by the patient. According to the hospital´s Annual Report, 17823 outpatients, 5098 ED visits and 657 natural deliveries were attended in 2014. Almost 5000 patients were admitted and 29 caesarean procedures were performed during this period. No specific emergency medical training or pre-hospital medical care services exist in most African countries, despite positive experiences that have been reported in Ghana and Tanzania [8,9]. A major pitfall is the great variability in infrastructure and human resources of EDs where many health centres are ran by personnel with basic medical training [10]. In consequence, there is scarce information in the literature focusing on pathologies and the clinical profiles of patients urgently attended in this geographical area. The present study aimed to describe the profile of pathologies and patients attended in the ED of a rural hospital in a low-income African country, outcome and mortality within the department. In addition, this study may give indirect information on the community health and resources that should be provided to health professionals.
Methods
In this retrospective descriptive study we analysed all the patients treated in the ED of Hospital Nossa Senhora da Paz from December 2014 to May 2015. The information was obtained from the ED register where the nurse collects age, gender, shift (8am-3pm, 3pm-10pm, and 10pm-8am), destination (discharge, admission, referral or death), service assignment (internal medicine, paediatrics, orthopaedics, obstetrics) and reason for consultation (basic syndromic orientation based on clinical judgement). For further analysis, patients were grouped in age ranges: under 5 years, from 5 to 16 years, from 17 to 60 years and over 60 years. The reason for consultation was then grouped according to body systems and medical specialties. Permission for compilation of the data was obtained from the hospital direction office. All data were analysed with IBM-SPSS. Quantitative variables were expressed as mean ± standard deviation (SD) and median for non-normal distribution variables and qualitative variables as frequencies and percentages. Qualitative variables were compared using Chi-square test and quantitative variables using Student's T-test. Non-normal distribution variables were compared using non-parametric tests (U Mann-Whitney or Kruskall-Wallis). For analytical statistics, 95% Confidence Intervals (CI) were calculated, and a value of p < 0.05 was considered statistically significant.
Results
During the six months of the study 2384 patients were attended in the ED, with a daily mean (±SD) of 13 ± 5 patients, and a median of 13 (range 1-33). The mean number of visits per shift was 4.6, 4.0 and 4.4 in the morning, afternoon/evening and night, respectively. We observed a trend towards fewer visits in the afternoon (p=0.05) with no statistical differences in the attendance between weekdays and weekends (p=0.4). The median age of the patients was 10 years (range: 0-96) (women 16 years, men 6 years) whereas 53.4 % were female. Age and/or gender were not registered in 23 patients. There were no differences in the mean age per shift (p = 0.2). The distribution of patients according to the age range was: under 1 month: 21 (0.9%), 1-12 months: 239 (10%), 1-5 years: 698 (29.3%), 5-16 years: 411 (17.2%), 17-60 years: 894 (37.5%) and older than 60 years: 103 (4.3%). In overall, a 40.2% of patients were younger than 5 years, and a total of 57.9% were younger than 17 years. The distribution by service assignment was: Paediatrics (1308 patients; 55%), Internal Medicine (967 patients; 40.6%), Orthopaedics (62 patients; 2.6%) and Gynaecology/Obstetrics (41 patients; 1.7%). Reason for consultation according to the age group is shown in Table 1. In 701 patients (29.4%) no data was registered regarding the reason for consultation. Considering those patients with all data registered (69.6%), the most frequent syndromic diagnoses were respiratory infections (20.5%, CI 18.6-22.5), fever (14%, 95% CI 12.4-15.8), digestive syndromes (13.5%, 95% CI 12-15.3), malaria (10.4%, 95% CI 9-11.9), trauma/injuries and malnutrition. In case of malaria suspicion, a rapid diagnostic test was performed, and a thick blood smear and peripheral blood extension were performed later for confirmation. Table 2 shows the distribution of age groups for each syndromic diagnosis. Outcomes according to the age group are shown in Figure 1. A total of 1076 patients (45.1%) treated in the ED were discharged, 1126 (47.2%) required in-hospital treatment, 31 (1.3%) were referred to other centres and 12 (0.5%) died within the ED (data on the death causes were not available). The patients admitted (median age of 4 years) were usually younger than those discharged (median age of 16 years) (p < 0.0001) or transferred (median age of 15 years) (p = 0.02). Patients diagnosed of malnutrition had a median age of 1 year (range: 0-23 years), and had a higher rate of hospital admission (95.4% malnutrition vs. 44.9% other consultations, p < 0.001). Patients diagnosed of malaria had a median age of 5 years (range: 1-55 years), and also had a significantly higher rate of admission (77.9% malaria vs. 44.8% other consultations, p < 0.001).
Discussion
The population attended in the ED was very young, in concordance with the hospital´s reference population [7]. Children under five years of age accounted for 40% of the patients, in contrast with other studies in Tanzania and Ethiopia where this age range was four times smaller [8,11]. Consequently, half of the hospital admissions in 2014 were in Paediatrics department. This fact stresses the need for specific resources and paediatric training for the health professionals in the ED. While in Nigeria almost a third of the patients treated in the ED were over 60 years old, in Angola this age group accounted only for the 3.9% of the country's population, thus in our study, this older population represented only a small percentage [12]. Overall, infectious causes were the predominant reason for consultation. They accounted for over 40% of the cases including fever of unknown origin. Our data were congruent with other studies where respiratory infections, gastrointestinal syndromes and malaria were the most frequent reason to seek urgent medical aid in this area [1, 6,13]. Respiratory infections in particular have been reported to be the second cause of mortality in the country [2]. In our study, pulmonary TB was suspected in 30% of the patients that consulted because of respiratory infections, mostly in adults. As a national reference centre for the diagnosis and treatment of TB, our hospital usually receives referrals from all over the country. In 2014, 470 new cases of pulmonary TB were diagnosed and 8% received second-line treatment for multidrug resistant TB (data from the Hospital Annual Report). In our study diarrheic syndromes were the third cause of consultation, especially in children. In Angola, these syndromes are the second most reported infection after malaria. This may be due to the high prevalence of intestinal parasitosis, reported in half of school-aged Angolan children, mainly ascariasis, giardiasis, hymenolepiasis and trichuriasis [2,[14][15][16]. Specifically in the area of Cubal, intestinal parasitosis was reported to be present in 16% of children [15]. Respiratory and digestive syndromes, fever, malnutrition and malaria accounted for 83% of the consultations in children under 5 years, in agreement with previous reports highlighting these as the main causes of infant morbidity and mortality [1,6]. Suspicion of malaria motivated 15% of consultations in patients under 17 years old and a very high admission rate. A mortality rate related to malaria of 8.3% has been described in Cubal [17]. Despite the decrease in the incidence observed in this area, malaria still accounts for 35% of healthcare demand and 20% of hospital admissions [2]. Considering this high impact on the most vulnerable age groups, maintenance of the pharmacological supply and diagnostic tests represents a challenge and an urgent sanitary priority. The number of visits in the ED due to severe malnutrition was high. Most cases required in-hospital treatment, mainly children under 5 years of age. As previously reported, prevalence of acute and chronic malnutrition is 8.2 and 29.7%, respectively, and almost half of Angolan children under 5 years have stunted growth [1,18]. Considering these high rates, it is essential to promote nutrition protocols and strengthen specialised centres in its treatment. Gynaecological and obstetric consultations were fewer than reported in other studies, as these patients were usually attended (and registered) in the maternity centre instead [8,11]. According to data from the Hospital Annual Report 2014, these visits accounted for 28% of the emergencies attended. Angola is considered an endemic area of Schistosoma haematobium. A recent study reported a 61% prevalence of urinary schistosomiasis in children from Cubal, a higher rate than the national average 28% [15,16]. In our study visits due to urologic complaints were significantly higher in adults than in children. This may be due to the fact that younger patients were visited more frequently in the outpatient clinic while adult patients with more advanced clinical symptoms went directly to the ED.
A progressive increase in the incidence of chronic noncommunicable diseases (NCD), classic cardiovascular risk factors and their related complications has been reported in most sub-Saharan Africa countries [3]. Ischemic heart disease is a frequent reason for consultation at ED [13]. More than a third of the Angolan population has hypertension, 8% has hyperglycaemia and 9% has a cardiovascular disease [1][2][3]. It has been reported a prevalence of peripheral arterial disease of 42.6% in this area [19]. Moreover, mortality attributable to these NCD has been estimated to be 24% in Angola [2,3]. In our study, the low rate of visits due to cardiovascular symptoms was similar to other series in the literature and it was related to the low mean age of the population attended [11]. However, when considering the group over 60 years, one every four patients consulted for cardiovascular related symptoms and the number of patients complaining of related neurological symptoms was also higher than average. These data revealed the growing relevance of these chronic NCD in the community and the need to implement screening, early detection and treatment protocols in the community health centres. Finally, the frequency of injuries was lower than in other studies in African countries where they have increased in recent years [2]. The lack of an orthopaedic surgeon in our hospital may be the cause of patients seeking medical assistance in the other municipal hospital with a permanent trauma service. Mortality within ED was low (0.5%). However, due to the short time spent by patients in the service prior to admission or discharge, a selection bias by failing to register some patients cannot be excluded. According to the Hospital Annual Report, the crude hospital mortality rate in 2014 was 7%, similar to other studies in comparable settings [12,20]. Regarding the destination of patients, our results supported those previously published in Ethiopia with admission and referral rates of 47% and 0.3% respectively [11]. The present study has several limitations. It was performed in a single centre. The reason for consultation was not recorded in almost one-third of the patients. In moments of great attendance, probably not all patients were registered by the nurse. Training and clinical capacity of the nursing staff were variable and limited. Together with the few available complementary explorations to support clinical suspicion, this fact may have contributed to the lack of accuracy in some syndromic orientation.
Conclusion
The ED of our institution, located in a rural area of central Angola, attended a very young population. Infectious diseases were the most frequent causes for consultation, mainly respiratory infections, diarrheic syndromes and malaria. Approximately half of the patients required in-hospital treatment, especially in cases of malaria and malnutrition. The mortality and referral rates were low. Priority should be given to measures that guarantee permanently available treatment for malaria and specific paediatric and nutritional care programs for the healthcare providers.
What is known about this topic There is great variability in infrastructure and human resources of Emergency Departments in African countries and scarce scientific information focusing on the pathologies and the clinical profiles of patients treated; Communicable diseases are the responsible of more than 50% of mortality in Angola, although non-communicable causes of mortality are rising; There is still a high incidence of malaria, acute and chronic malnutrition especially in Angolan children under 5 years of age.
What this study adds
In rural Angola, the Emergency Department attended predominantly a very young population with infectious diseases (mainly respiratory infections, diarrheic syndromes and malaria); High hospitalization rates were observed among children under 5 years, especially in cases of malaria and malnutrition; Pulmonary tuberculosis was suspected in one every three patients that consulted because of respiratory infections, mostly in adults. expenses from Soluciones de Gestión y Apoyo a Empresas, S.L. These collaborators had no role in study design, data collection, analysis or interpretation of data or decision to publish the manuscript. For the remaining authors no competing interests were declared. 1659(100) a It includes 101 cases of suspected pulmonary tuberculosis(8 in <5 years, 6 in 5-16 years, 71 in 17-60 years and 16 in older than 60 years) b Others: ophthalmology(5), otorhinolaryngology(36), intoxications/poisoning(8), psychiatry(4), tumours(20), metabolic disorders(13), human immunodeficiency virus(6), measles, dental pathology, sexual assault, pneumothorax | 2018-08-06T13:47:41.926Z | 2018-03-05T00:00:00.000 | {
"year": 2018,
"sha1": "8fe2ca2c457cc34fb2f0bfc41f572733c8d1df22",
"oa_license": "CCBY",
"oa_url": "https://doi.org/10.11604/pamj.2018.29.143.13780",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "8fe2ca2c457cc34fb2f0bfc41f572733c8d1df22",
"s2fieldsofstudy": [
"Medicine",
"Political Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
268065157 | pes2o/s2orc | v3-fos-license | The origin, dissemination, and molecular networks of HIV-1 CRF65_cpx strain in Hainan Island, China
Background HIV-1 CRF65_cpx strain carries drug-resistant mutations, which raises concerns about its potential for causing virologic failure. The CRF65_cpx ranks as the fourth most prevalent on Hainan Island, China. However, the origin and molecular epidemiology of CRF65_cpx strains in this area remain unclear. This study aims to estimate the spatial origins and dissemination patterns of HIV-1 CRF65_cpx in this specific region. Methods Between 2018 and 2021, a total of 58 pol sequences of the CRF65_cpx were collected from HIV-positive patients on Hainan Island. The available CRF65_cpx pol sequences from public databases were compiled. The HIV-TRACE tool was used to construct transmission networks. The evolutionary history of the introduction and dissemination of HIV-1 CRF65_cpx on Hainan Island were analyzed using phylogenetic analysis and the Bayesian coalescent-based approach. Results Among the 58 participants, 89.66% were men who have sex with men (MSM). The median age was 25 years, and 43.10% of the individuals had a college degree or above. The results indicated that 39 (67.24%) sequences were interconnected within a single transmission network. A consistent expansion was evident from 2019 to 2021, with an incremental annual addition of four sequences into the networks. Phylodynamic analyses showed that the CRF65_cpx on Hainan Island originated from Beijing (Bayes factor, BF = 17.4), with transmission among MSM on Hainan Island in 2013.2 (95%HPD: 2012.4, 2019.5), subsequently leading to an outbreak. Haikou was the local center of the CRF65_cpx epidemic. This strain propagated from Haikou to other locations, including Sanya (BF > 1000), Danzhou (BF = 299.3), Chengmai (BF = 27.0) and Tunchang (BF = 16.3). The analyses of the viral migration patterns between age subgroups and risk subgroups revealed that the viral migration directions were from "25–40 years old" to "17–24 years old" (BF = 14.6) and to "over 40 years old" (BF = 17.6), and from MSM to heterosexuals (BF > 1000) on Hainan Island. Conclusion Our analyses elucidate the transmission dynamics of CRF65_cpx strain on Hainan Island. Haikou is identified as the potential hotspot for CRF65_cpx transmission, with middle-aged MSM identified as the key population. These findings suggest that targeted interventions in hotspots and key populations may be more effective in controlling the HIV epidemic. Supplementary Information The online version contains supplementary material available at 10.1186/s12879-024-09101-w.
Introduction
Human immunodeficiency virus type 1 (HIV-1) is classified into four distinctive phylogenetic groups, namely M (major), N (new), O (outlier), and P [1].Of these, HIV-1 group M is particularly significant and includes a broad range of distinct subtypes (A-D, F-H, J, and K), along with circulating recombinant forms (CRFs) and unique recombinant forms (URFs), ultimately initiating the ongoing pandemic [2].The uneven global distribution of various HIV-1 subtypes and CRFs can be attributed to varying founder effects, followed by localized dissemination driven by socioeconomic and behavioral factors [3], sometimes influenced by continuous influxes of new infections from neighboring regions [4].The quest for a globally effective vaccine is in progress [5], but is challenged by the rapid genetic evolution and recombination of HIV, which are impacted by genetic, social, and epidemiological variables.
As of October 30, 2023, the Los Alamos HIV database has documented a total of 140 CRFs for HIV-1.In China, CRF01_AE, CRF07_BC, and CRF08_BC represent the predominant CRFs [6].Most of the newly reported HIV patients (69.2%) on Hainan Island were needle-sharing drug users, with CRF01_AE (84.3%) as the dominant genetic form in 2009 [7].In 2023, CRF01_AE (68.9%) also represented the main genetic form among patients with virologic failure in antiretroviral therapy (ART) on Hainan Island, followed by CRF07_BC.However, no comprehensive studies have been conducted in the past ten years regarding HIV-1 epidemic genetic forms in HIV-positive patients on Hainan Island [8].
CRF65_cpx emerged as a distinct CRF, first identified by Feng et al. in 2013 in western Yunnan Province, China.This CRF consists of genetic components from three genetic forms: B' , C, and CRF01_AE [9].A previous study demonstrated that CRF65_cpx likely originated around the year 2000 among heterosexuals (HETs) in Yunnan Province [10].Over the subsequent years, transmission extended to men who have sex with men (MSM) in Beijing and Anhui, occurring approximately 3 to 7 years later.Subsequently, the CRF65_cpx strains propagated into other areas, including Hebei [11], Jiangsu, Heilongjiang, Jilin [12], and Guangxi [13].In 2019, Ran Zhang et al. confirmed the presence of CRF65_cpx in approximately 0.8% of HIV-positive MSM across 19 cities located in six provinces of China [14].Our previous study indicated that CRF65_cpx ranked fourth among patients with virologic failure in antiretroviral therapy on Hainan Island [8].These findings underscore the national expansion of CRF65_cpx [15].
During its transmission, the CRF65_cpx strain has undergone changes in certain amino acid sites and cytotoxic T lymphocytes (CTL), potentially accelerating the progression of HIV-related diseases [16].Genotypic resistance analyses revealed the presence of natural mutations, such as V179D and K103R/V179D, which are associated with CRF65_cpx resistance to nonnucleoside reverse transcriptase inhibitors (NNRTIs) [15].While individually, V179D and K103R are relatively common polymorphic mutations with limited impact on NNR-TIs susceptibility [17], their combination results in an approximately 10-to 15-fold reduction in susceptibility to efavirenz (EFV) and nevirapine (NVP) [18].Previous studies have demonstrated a positive correlation between the rate of HIV-1 virus evolution and the advancement of the disease [19,20].Furthermore, a significant heterogeneity rate is visible both between and among various subtypes [21].However, no previous reports are available on the evolutionary rate of CRF65_cpx.
Hainan Island, the southernmost province of China, has garnered popularity as a tourist destination due to its pleasant tropical climate.It has historically been regarded as an area in China with relatively low HIV prevalence.However, the dynamics of immigration and tourism have created various challenges in this area.Our study, encompassing the analysis of 58 HIV-1 CRF65_cpx polymerase (pol) sequences on Hainan Island, revealed that the CRF65_cpx strain played a significant role in driving the local HIV-1 prevalence.Here, we present a comprehensive account of the origin and dissemination pattern of the CRF65_cpx strain on Hainan Island using HIV-TRACE and Bayesian analyses.The findings can potentially inform the development of effective HIV surveillance strategies and public health interventions, specifically focusing on key populations.Furthermore, we offer valuable insights for enhancing HIV testing initiatives, monitoring drug resistance, and designing vaccines to control the spread of CRF65_cpx or other CRFs in the Hainan region and across China.
Study design and specimen preparation
An HIV molecular epidemiology and drug resistance monitoring study was conducted at the Fifth People's Hospital of Hainan Province from January 2018 to November 2022.The study recruited 1742 HIV-positive individuals who were either ART-naive or ART-experienced.In accordance with national standards, patient self-assessment interviews were used to gather demographic and epidemiological information.To adhere to ethical standards in China, the blood samples were linked to demographic and clinical data via a unique numerical code.
The viral RNA was extracted, and the HIV-1 pol gene was amplified at the Guangxi Key Laboratory of AIDS Prevention and Treatment (Guangxi Medical University, Guangxi, China).Subsequently, the HIV-1 pol, encompassing the entire protease (PR) and a segment of the reverse transcriptase (RT) (nucleotides 2253 to 3334 in reference strain HXB2; 1060 bp long), were sequenced by Sangon Biotech Company.Previously described primers were used to amplify the HIV pol gene using nested RT-PCR [22].Finally, 1410 available pol sequences were obtained, including 58 (4.11%, 58/1410) CRF65_cpx sequences.
Sequence alignment and subtype assignment
Sequencher v5.1.4.6 was used to assemble the HIV-1 pol sequences, while the MAFFT program was employed for alignment with the HXB2 reference sequence via the online HIV align tool (http:// www.hiv.lanl.gov/ conte nt/ seque nce/ viral ign.html) [23].The HIV-1 subtypes were determined using the online subtyping tool COMET (https:// comet.lih.lu/) for preliminary classification [24], and identified by a phylogenetic maximum likelihood tree (ML tree).The general time-reversible (GTR) substitution with invariant sites (I) and gamma distributed (G) model in FastTree v2.2.10 was used to construct the ML tree (711 sequences) [25], which contains all available pol sequences of subtype C from China and of CRF65_cpx, as well as reference sequences of other genetic forms (A1, A2, B, B' , D, K, F1, F2, H, N, CRF01_AE, CRF07_BC, CRF08_BC and other CRFs) from the Los Alamos HIV Sequence Database (http:// www.hiv.lanl.gov/, accessed on October 30, 2023).Shimodaira-Hasegawa-like values were used to assess the reliability of the tree [26].Sequences clustering with the reference sequences and bootstrap values ≥ 80% were identified as belonging to the same genetic form as the reference sequences.Group N (AJ006022, AJ271370, and AY532635) was set as the outgroup.The results were visualized using FigTree v1.4.3.9 (Table S1 and Fig. S1).
Molecular transmission network analysis
To observe the network growth trend from 2018 to 2021, we constructed a molecular transmission network with a genetic distance threshold of 0.5% substitutions per site, corresponding to a maximum of around 2-3 years of viral evolution separating these strains [27].Briefly, the pol sequences were aligned to the HXB2 sequence using online HIV align tool.The aligned sequences were then uploaded into HIV Transmission Cluster Engine (HIV-TRACE) to calculate the pairwise Tamura-Nei93 (TN93) genetic distance among all sequence pairs [28].Finally, the network was visualized using Cytoscape v3.6.1.In the network, a node represents an individual.Nodes were linked to each other to construct a cluster (consisting of ≥ 2 sequences) if their pairwise genetic distance was below the threshold.To observe the growth of the clusters, the 2012-2018 sequences (year of diagnosis) were designated as baseline, after which the 2019, 2020 and 2021 sequences were introduced to construct the molecular networks.
Sequence dataset and bayesian phylodynamic inference
Given the absence of breakpoints in CRF65_cpx pol, the possibility that some CRF65_cpx sequences were misclassified cannot be excluded.In a previous study [29], 15 CRF65_cpx sequences from the BLAST matches in the HIV Sequence Database (http:// www.hiv.lanl.gov/ conte nt/ seque nce/ HIV/ mainp age.html) were misclassified as subtype C in 2019.Therefore, to identify as many CRF65_cpx pol sequences as possible from the HIV Sequence Database, all the subtype C and CRF65_cpx pol sequences (HXB2:2253-3334) from China and pol sequences of partly other subtypes (http:// www.hiv.lanl.gov/, accessed on October 30, 2023), were downloaded.The downloaded sequences were selected based on the following inclusion criteria: (1) sequences already published in peer-reviewed journals, (2) no uncertainty about subtype assignment, (3) sampling time and city/ province of origin were clearly established in the original publication, (4) the length of nucleotides ≥ 1000 bp, and (5) the proportion of nucleotides with ambiguity ≤ 0.5%.Finally, the global search yielded 83 available CRF65_cpx pol sequences, including 53 misclassified as subtype C (Table S1).The codons with drug resistance mutations (detected via the Stanford University HIV Drug Resistance HIVdb program at https:// hivdb.stanf ord.edu) were manually removed from the alignment to avoid the confounding factor of convergent evolution in the phylogeny inference [30].
To infer the potential origins of the CRF65_cpx on Hainan Island, Bayesian Evolutionary Analysis by Sampling Trees (BEAST) was conducted using 141 available CRF65_cpx pol sequences (Dataset-1), which included 83 sequences from different geographic locations in the HIV Sequence Database (Table S1) and 58 sequences from our study on Hainan Island [30].To improve the temporal signal of the likelihood of Markov chain Monte Carlo (MCMC) convergence of BEAST analysis, five sequences were eliminated based on R squared (R 2 ) in TempEst v1.5.3 [31].Finally, the Dataset-1 R 2 value was 0.618.The maximum clade credibility (MCC) tree generated from Dataset-1 revealed that the Hainan strains formed a monophyletic clade, consisting of 53 Hainan sequences and one Beijing sequence.Dataset-2 was utilized (R 2 = 0.310) to infer viral migration events among the age groups and cities on Hainan Island, which included the 53 Hainan sequences identified in the MCC analysis of Dataset-1.
We implanted Bayesian SkyGrid model in BEAST v1.10.5 to estimate the most recent common ancestor (tMRCA) and the evolutionary rate of CRF65_cpx [32].The Hasegawa Kishino Yano plus Gamma plus Invariant Sites (HKY + G + I) nucleotide substitution model determined by Bayesian Information Criterion (BIC) using jModelTest v2.1.10 [33], along with a relaxed uncorrelated lognormal molecular clock model, was utilized for this analysis.The analysis for each dataset was 100 million generations in triplicate runs and sampling every 10,000 states.The triplicate BEAST runs were combined with LogCombiner v1.10.5pre, and the first 10% was discarded as burn-in.The estimated effective sample size (ESS) was over 200.The MCC trees were generated using Tree Annotator v1.10.5 and visualized in FigTree v1.4.3.A Bayesian stochastic search variable selection (BSSVS) procedure was used to calculate the Bayes factor (BF) to accurately describe the viral dissemination process [34].This study only discussed the results with BF ≥ 3 and with posterior probability (PP) ≥ 0.80.
The demographics of participants on hainan island
Table 1 provides an overview of the descriptive statistics from the study samples.The initial case of CRF65_cpx infection on Hainan Island was diagnosed in 2012, followed by 57 additional cases of the same genetic form in the subsequent nine years.The predominant proportion (89.66%) of these cases engaged in MSM, with the significant majority (96.55%) being male and demonstrating genotypic resistance (100%).The median age of the 58 participants was 25 years, ranging from 17 to 50 years.Of the 58 participants, 82.76% were single, 84.48% were from Haikou, 86.21% were recruited in 2021, 79.31% were on ART treatment, and 43.10% held educational qualifications of college graduate or above.The median of
The expansion of HIV-1 CRF65_cpx strain on hainan island
Of the 58 sequences, 39 (67.24%, 39/58) fell into the molecular network, forming a single cluster at a maximum pairwise genetic distance of 0.5%.Within this network, the nodes exhibited a median degree of 13, ranging from 1 to 25. Notably, 66.67% (26/39) of these nodes displayed more than four links, with all nodes exhibiting genotypic resistance.Within the established cluster, the median number of links was higher for individuals aged 25 to 41 years (n = 14) than for those aged 17 to 24 years (n = 8) and individuals above 40 years of age (n = 4).
From 2019 to 2021, four individuals were added to the cluster annually, with those added in 2019 originating from three cities on Hainan Island.Of these four individuals, all of them were MSM aged 18-24 with a high school education or above, and three of them were directly genetic link to the same HIV infected people (M15) (Fig. 1A).The M15 exhibited a low CD4 + T-cell count of 269 cells/mm 3 in December 2019, followed by a high viral load of 59,572 copies/mm 3 in January 2020.In-depth tracking and investigation revealed that M15 engaged in multiple instances of both commercial and casual sex spanning nearly three months prior to diagnosis in 2019.Clinical surveillance subsequently revealed that medical treatment in this patient was ineffective in 2020.In the subsequent 2020 and 2021 networks, M15 connected directly with half of the new cases: M51 and M50 in 2020, and M31 and M52 in 2021 (Figs.1B and C).Noticeably, the M15 most likely represented the key node driving the local CRF65_cpx transmission.
In 2020, four new nodes, one corresponding to a female individual and two to heterosexually infected individuals, were scattered throughout the network (Fig. 1B).In the 2020 network, a significant detail was that one new node (M31) connected with three others (M57, M15, and M02), which experienced a high viral load (Fig. 1C) and contributed to the HIV-1 CRF65_cpx epidemic complexity.
The origin of HIV-1 CRF65_cpx in China and Hainan island
The BEAST analysis revealed that the mean evolutionary rate of the CRF65_cpx strain was 2.115 × 10 -3 subs/ site/year [95%HPD: (1.691-2.591)× 10 ) were observed in this study (Fig. 3A).
In the MCC tree (Fig. 2), the monophyletic lineage with a high PP (PP = 1.00) at the top consisted of 54 CRF65_ cpx sequences, including 53 sequences from Hainan and one sequence from Beijing.The tMRCA of this lineage was estimated as 2013.2.Further analysis indicated that Haikou might be the epicenter of the CRF65_cpx strain on Hainan Island, which subsequently spread to Sanya
The transmission dynamics between the age and risk subgroups
To reveal the historical migration patterns of CRF65_cpx strain in the age and risk subgroups, Bayesian phylogenetic analysis was conducted using Dataset-1 and Dataset-2, respectively.For the age subgroups, the results showed that the viral migration directions moved from middle-aged people (MP, aged 25 to 40 years) towards the older people (OP, aged over 40 years, BF = 17.6) and young people (YP, aged 17 to 24 years, BF = 14.6) groups (Fig. 3C).Additionally, well-supported transmission between the risk subgroups was inferred from MSM in other provinces to Hainan MSM (BF = 29.1)(Fig. 3D).Migration events with strong BF support (BF > 1000) were observed on Hainan Island, specifically from MSM to HETs (Fig. 3D).
Discussion
This study conducted a comprehensive analysis to elucidate the origin and dissemination patterns of HIV-1 CRF65_cpx on Hainan Island, China.The introduction of This study indicated that the majority of individuals infected with the CRF65_cpx strain were MSM, with 43.10% possessing a college degree or higher.One plausible explanation was the fact that MSM were the predominant population affected by HIV on Hainan Island.Furthermore, individuals identifying as MSM represented a substantial 72.72% of those afflicted with HIV who possessed a college education or higher [35].Consequently, targeted interventions for higher education MSM should be prioritized to effectively reduce HIV transmission.Additionally, the network analysis showed that over two-thirds of HIV-positive individuals were organized into only one cluster, including 39 members, while 66.7% of the nodes had more than four links.A higher clustering rate was observed for CRF65_cpx with a lower genetic cutoff than CRF01_AE, CRF07_BC, and CRF08_BC in other regions of China [36,37].This finding demonstrates the close interconnection within the population affected by this viral strain.It suggests that targeted interventions can be employed for individuals infected with CRF65_cpx.Increasing testing efforts among individuals with high-risk contact with those infected by CRF65_cpx, proactively identifying infected cases and early treatment are important strategies to mitigate secondary transmission.Furthermore, continued monitoring is essential for CRF65_cpx cluster expansion.
From 2019 to 2021, the cluster exhibited a consistent expansion pattern.M15 may hold a pivotal position as a central node or potentially act as a super-spreader in the transmission dynamics of CRF65_cpx on Hainan Island.The Chinese CDC's guidelines for HIV Transmission Network recommend that large clusters (≥ 10 nodes) and nodes displaying high viral loads or drug resistance should be considered as key clusters and nodes for surveillance and intervention [38].Therefore, it is essential to identify key clusters and individuals among HIV-positive patients.The results of this study showed consistent network expansion.Additionally, due to commercial or casual sex and a high viral load, the M15 patient possibly played a key role in local CRF65_cpx transmission.The previous research demonstrated that nodes with unusual characteristics likely represented opportunities for breaking important transmission chains [39].Our findings suggest that the patients in the large cluster, especially the M15 case, should be the primary intervention target.For example, drug adherence monitoring and follow-up should be enhanced for key populations, and viral load assessments should be performed for those living with HIV.
Our findings highlight the therapeutic difficulties faced by patients infected with CRF65_cpx strains.All the CRF65_cpx sequences on Hainan Island carried the V179D mutation, which was in line with findings from an earlier CRF65_cpx investigation [15].However, the prevalence of this mutation was significantly higher in CRF65_cpx than in other subtypes, such as CRF01_AE (7.1%) [40], and subtype F (4.1%) [14].The V17D mutation is a polymorphic accessory mutation, leading to a twofold reduction in NVP, EFV, ETR, and RPV susceptibility [41].Additionally, the V179D and K103R mutation co-occurrence caused a 15-fold decrease in NVP and EFV susceptibility [18].The observed synergistic effect of the V179D and V106I combination indicated a reduction in NVP and EFV susceptibility [42].Based on the findings, implementing intervention measures is recommended for all HIV-positive patients harboring V179D mutations.
Similar to previous analyses, this study confirmed that CRF65_cpx originated from individuals who acquired HIV via heterosexual transmission in Yunnan Province around 2001.5 [16].Approximately four years later, the CRF65_cpx strain disseminated to the MSM population in Beijing.This study established Beijing as a significant epicenter in the spread of the CRF65_cpx strain to other regions, as evidenced by a high BF.It was hypothesized that the CRF65_cpx strain most likely arrived in Hainan via MSM in Beijing in September 2013.Beijing, the nation's capital, is home to a sizable MSM population, which accounted for 73.9% of newly diagnosed HIV/AIDS infections in 2016 [43].The high mobility of the MSM population has contributed to the rapid dissemination of diverse HIV-1 strains throughout China.The phenomenon of novel strains circulating within the MSM population and subsequently leading to crossregional transmission, such as CRF01_AE [44], CRF07_ BC [45], and CRF55_01B [46], is evident.In recent years, MSM has been a major factor in the disease prevalence in Hainan [35].As a tropical tourist destination, Hainan Island attracts individuals from across the country, including MSM.Consequently, they have played a role in the emergence of new HIV CRFs and URFs, heightening the intricacy of the HIV epidemic among MSM populations in Hainan Island.It is worth noting that, compared to HETs, the CRF65_cpx strains in MSM population lack the protective epitopes of HLA-B*2702 and HLA-B*5103.This observation implies the potential adaptability and accelerated disease progression of this strain in MSM population [16].Therefore, a comprehensive investigation into the disease progression and drug resistance aspects of the CRF65_cpx strain are warranted.
On Hainan Island, Haikou represents the epicenter of the spread of the CRF65_cpx strains to other cities via homosexual contact in different years, including Sanya, Danzhou and Chengmai.Given that Haikou serves as the provincial capital of Hainan Province and harbors the largest number of HIV-1 cases, especially among the MSM, its role as a hub is paramount.Sanya follows closely as another significant city in this context.This pattern could likely be attributed to the spillover effect from Haikou to neighboring region on Hainan Island.Additionally, CRF65_cpx has been detected across diverse age groups and within heterosexual populations, underscoring the complexity of the CRF65_cpx epidemic.Consequently, it is evident that the landscape of the CRF65_cpx strains is multifaceted.Therefore, maintaining a robust molecular surveillance system for CRF65_cpx strains, with a particular emphasis on the MSM population, is necessary.
The CRF65_cpx epidemic in Hainan is highly complex.This study revealed the presence of CRF65_cpx strains in both HETs and a broad MSM age group.The Bayesian analysis inferred that viral migration predominantly occurred from MP towards EP and YP, suggesting that MP may potentially play a pivotal role in CRF65_cpx dissemination.Among young MSM, factors such as limited HIV awareness, underestimation of personal risk, peer influence, interaction with older MSM, sexual preferences, and engagement in transactional sex have been identified as drivers of HIV-1 transmission [47,48].Young MSM might face additional vulnerabilities, such as sexual coercion, stigma, and social exclusion, further underscoring the significant influence exerted by older MSM on their younger counterparts.
The study presented several potential limitations.First, sampling bias was introduced because the inclusion criteria relied on patients attending the sampling hospital for the first time or keeping appointments for follow-ups within the recruitment period.Second, the relative sample size was a constraint due to the emerging nature of CRF65_cpx as an HIV-1 strain and the limited sequence availability.Third, there was temporal bias during the sampling period, since a significant proportion of the Hainan sequences were collected in 2021.However, we included all CRF65_cpx sequences downloaded from the HIV Sequence Database to minimize sampling year bias.
In conclusion, this study provides a thorough comprehension of the origin and transmission dynamics of CRF65_cpx on Hainan Island.The results indicate that the CRF65_cpx strain was introduced to Hainan Island by MSM from Beijing around 2013, followed by local dissemination.Notably, MSM aged 25 to 40 years play a pivotal role in bridging the gap between the younger and older MSM population.Given the implications for public health, it is essential to give immediately prioritize to molecular surveillance of the CRF65_cpx strain.This will make it easier to create sensible guidelines and recommendations for the early intervention and containment of the CRF65_cpx pandemic, especially for a population as migratory as the MSM on Hainan Island.
Abbreviation: IQR inter-quartile range, MSM, men who have sex with men
Fig. 1
Fig. 1 The dynamic transmission network of HIV-1 CRF65_cpx from 2019 to 2021 in Hainan, China.The edges (lines connecting the nodes) represent genetic relatedness.The nodes indicate HIV-1 patients or sequences.The characters in the node indicate gender: M denotes male, and F denotes female, while the number in the node represents the patient code.The colors indicate different diagnosis years: purple, 2012-2018; yellow, 2019; green, 2019; and red, 2021.The shapes indicate different transmission routes in (A), (B), and (C).The sizes of the geometrical figures represent the degree (lines connecting the nodes)
Fig. 2
Fig. 2 The Bayesian maximum clade credibility (MCC) tree of HIV-1 CRF65_cpx sequences.The MCC tree was constructed using Dataset-1, consisting of 136 sequences, of which 54 were from Hainan Province and 82 were downloaded from the HIV Sequence Database (Table S1).The colors of the rectangles indicate different provinces.The dotted rectangle represents the monophyletic clade, consisting of 53 Hainan sequences (Dataset-2) and one Beijing sequence.The colors of the triangles in the tip represent the risk factors.The values next to the black dots indicate the time of the most recent common ancestor (95%HPD intervals).Scale years are presented at the bottom of the figure.Note.HETs: Heterosexuals.MSM: Men who have sex with men
the
CRF65_cpx strain occurred via the MSM population in Beijing around February 2013.Subsequently, the infection propagated within the local MSM, leading to the emergence of a substantial and stable transmission cluster.Notably, Haikou city became the hub for the rapid spread of CRF65_cpx on Hainan Island.The transmission dynamics indicated a pattern from MSM to HETs.Particularly, MSM aged 25 to 40 years played a crucial role
Fig. 3
Fig. 3 The inferred well-supported virus migration events of CRF65_cpx between the subgroups.Only results with Bayes factor (BF) ≥ 3 are presented.Arrows indicate the direction of the HIV-1 migration events.The colors were chosen to visually distinguish the different BF level values.A and D containing 136 sequences (Dataset-1) from different provinces, demonstrate viral migration events between provinces in China (3A) and the transmission pattern of HIV-1 CRF65_cpx (3D).B and C including 53 sequences only from Hainan (Dataset-2), present migration events among the age groups and cities in Hainan Province.Note.YP: Young people (aged 17 to 24 years old), MP: Middle-aged people (aged 25 to 40 years old), OP: Older people (OP, aged over 40 years old).MSM: Men who have sex with men.HETs: Heterosexuals.IDUs: Intravenous drug users.NA: Transmission route not available
Table 1
The characteristics of 58 patients infected with HIV-1 CRF65_cpx strain from Hainan, China | 2024-03-01T14:18:07.157Z | 2024-03-01T00:00:00.000 | {
"year": 2024,
"sha1": "a47c9d169c13b414b0eb508175ca70dd30cd0679",
"oa_license": "CCBY",
"oa_url": null,
"oa_status": null,
"pdf_src": "PubMedCentral",
"pdf_hash": "e38e2784502c319bdcc8db40c822f64861917239",
"s2fieldsofstudy": [
"Medicine",
"Environmental Science"
],
"extfieldsofstudy": [
"Medicine"
]
} |
197895691 | pes2o/s2orc | v3-fos-license | International tourist’s expenditure on souvenirs in Ghana: Do their socio-demographics have any influence?
Shopping for souvenir is an important aspect of the tourism activity which almost all tourists participate in. The purpose of this paper is to examine the expenditure pattern of international tourists on souvenirs in Accra. The study adopted the descriptive design and questionnaire was administered to gather data. Data collected from 196 international tourists who were sampled from hotspots in Accra, using convenient sampling method, revealed that there was a significant relationship between socio-demographic characteristics of tourists and expenditure on souvenirs. Therefore, socio-demographic characteristics of an individual was found to have likely influence on the individual’s expenditure on souvenirs. Also, a majority of the respondents, who were first time visitors, spent less on souvenirs. A significant average expenditure on souvenirs signifies that the souvenirs business is a profitable business that needs to be maintained and enhanced. Majority of the respondents had souvenirs as part of their purchase plans. Leisure travellers were noted to spend more on souvenirs than other travellers. Souvenir shopping generates a good amount of money and as such the Centre for National Culture and Ghana Tourism Authority should embark on activities aimed at packaging and promoting the arts and crafts of the country so as to arouse sales of souvenirs.
Introduction
Souvenirs form a major component of the tourism retailing system, employing a significant number of people in its production, distribution and sales (Hashimoto & Telfer, 2007). Souvenirs are the material culture of a destination and being custodians of such cultures, the locals are the majority of employees in this sector (Xie, Wu & Hsieh, 2012;UNWTO, 2012). Production and sale of souvenirs have become an easy step for indigenes and ethnic minority to generate income from tourism even in developed economies like USA (Swanson, 2004). An objective of the United Nations World Tourism Organization's (UNWTO) sustainable tourism development plan is to alleviate poverty, using tourism as a tool to diversify the economy (UNWTO, 2011). Tourism has thus become an alternative source of livelihood (Ashley, Boyd & Goodwin, 2000) and income to host communities, destinations and countries at large. Souvenir making supports the traditional agriculture communities where farming and rearing of animals is the major occupation. The production and sale of arts and crafts is a major source of livelihood, especially, for Ghanaians living in rural areas and for Africans at large (Arko-Achemfuor, 2012). The souvenirs industry employs people with lower level education or no formal education. Moreover, this form of employment is not gender-sensitive; it employs both males and females with skills and technical knowledge to produce and sell arts and crafts as souvenirs. In Tanzania for example, women are more involved in the making and sale of souvenirs (Muganda, Sirima, Moshy & Mkumbo, 2012). Also in Miao, China, women produce handicrafts and embroidery (Henderson, Teck, Ng & Si-Rong, 2009) In Ghana, however, men make the drums, wood and metal works whilst women are into garments, clay works, tie-dye, batik and the gathering of raw materials (The Ford Foundation, 2005). Souvenirs liven up communities; money from the sale of souvenirs creates linkages within small communities as the manufacturers and raw materials are sourced from the local communities (Mshenga & Owuor, 2009). The money generated from souvenirs, therefore, stays in the community (Ayodele, 2002).
Souvenir production also creates an opportunity for medium to small-scale enterprises to participate in tourism. This encourages entrepreneurship and its accompanying benefits of jobs creation and income generation. UNWTO (2012) asserts that the production and sale of handicraft is an attractive and economically viable way through which communities participate in tourism. Souvenirs, therefore, provide a sure way of community participation in tourism through employment. According to United Nations World Tourism Organization [UNWTO], 2012), handicrafts form the majority of souvenirs purchased by travellers, especially in new destinations. In relation to items bought during shopping, Xie and Bao (2006) notice that females buy souvenirs more than males. According to Turner and Reisinger (2001), males prefer action-oriented souvenirs which are of instrumental value that would enable them to perform an important task. On the other hand, females prefer souvenir with sentimental value focusing more on the emotional satisfaction it would offer them. Hugh (2009) concludes that males are likely to buy discounted, low priced souvenirs than females who are less sensitive to prices. Souvenir making has become an economic entity, attracting tourists' direct spending for almost all tourist destinations (Harper, 1981). Shopping is an essential part of the tourism and leisure experience (Sheena, 2006) as every tourist wants to bring home an object from the destination visited. Travelling without shopping for souvenir is regarded a partial activity as the traveller has nothing to prove his/her experience. Shopping for completes the travel experience (Turner & Reisingner, 2001) and complement the role of attractions by preserving tourism experience. Shopping for a souvenir has therefore become a common tourism practice (Nissa, 2009) and almost every tourist shops for souvenirs.
However, it is worth noting that Souvenir shopping is a subsection of tourists shopping; tourists shop for other goods as well. The distinguishing factor between souvenirs and other goods is that souvenirs are related to tourism and travel experience whilst the other goods are not related to tourism. The role of souvenirs is summarized as "look where I have been or look what I have done" (Wicks, 2004:3). As far back as the late 1980s, Gordon (1986), asserted that very few people will take a trip without buying souvenirs to represent a visible summary of the experience and it is still the trend in the 2000s. Souvenirs form a large component of all tourists shopping products (Weng & Tung-Zong, 2012). Whatever the type of tourism (leisure, business), the tourist is likely to take home a souvenir (Follad, 2006). Ming (2011) has the view that shopping for souvenir is an interesting activity which almost all tourists undertake.
Motivations for buying souvenirs include souvenirs as gifts (Kim & Littrell 2001;Gordon 1986) and as a reminder of travel experience (Swanson, 2004). Ward and Tran (2007) posit that souvenirs as gifts are of two categories: self-gifting and gift-giving. Tuomisto (2012) discovered that 71% of tourists to Tampere, Finland bought souvenirs for friends and relatives. Self-gifting involves tourists buying souvenirs for personal use whereas, in gift-giving, souvenirs are used by others. Japanese and Korean tourists deem it a moral right to buy gifts for family, friends and colleagues, especially those who are aware of their travel and/or have made some financial contribution towards the trip (Park, 2000). Souvenirs as gift-giving strengthen the relationship with others and also, serve as a means to appreciate love ones. Souvenirs are mainly for selfuse or for use by other people, especially friends, relatives and colleagues.
Whatever the motivation of buying souvenirs, tourists allocate a significant proportion of their travel expenditure to shopping for gifts and souvenirs (Heung & Cheng, 2000;Ming, 2011). Shopping, of which buying of souvenir is paramount, constitutes over half of the overall travel spending for tourists visiting Hong Kong (Law & Au, 2000). Cai, Lehto and O'leary (2001) reveal that on a scale of preference, Chinese leisure travellers to the USA assign more money to souvenirs than lodging, food and entertainment. Taiwanese tourists in 1999, ranked expenditure on souvenirs second to tobacco and wine (Yoon-Jung, Chia-Kuen, Letho, & O'Leary, 2004). Apichoke (2006) affirms that handicraft products accounted for 30% of tourists' shopping expenses in Thailand.
A study by Malta Tourism Authority (2011) reveals that besides accommodation, tourists spend money mainly on food and shopping (souvenirs and clothing). Shopping constituted 15.1% of tourists' total expenditure of which 6.3% was on souvenirs. Georgian National Tourism Agency, (2011) also found that 17% of tourists' spending is on souvenirs and gifts. A study conducted by Yoon-Jung, (2007) in the USA also revealed that leisure travellers spent the highest amount of money on shopping, followed by those visiting friends and relatives and business travellers. A contrary finding was revealed by Luo and Lu (2011) that business travellers to the Canton Fair in China, spend more on souvenirs than leisure travellers. This contradiction is basically due to geographic differences and both destinations may have different target markets. Furthermore, tourists on honeymoon, also, spend more on souvenirs than other motivation types like sun-sea-sand and hiking among others (Wang & Davidson, 2010). Hugh (2009) is of the view that males are likely to buy discounted, low priced souvenirs than females who are less sensitive to prices. This implies that female spend more on buying souvenirs than males. Tourists with higher level of education spend more on shopping (Yu & Littrell, 2005;Wang & Davidson, 2010) as higher levels of education do commensurate with well-paid jobs.
The spending of consumers, especially tourists, serves as a source of income to producers and vendors of arts and crafts. Research has shown that amount of money spent by tourist has implications on the success and continuity of this sector as it could be an incentive or disincentive to people employed in the art and crafts (Muganda, Sirima, Moshy & Mkumbo, 2012;Henderson, Teck, Ng & Si-Rong, 2009;Mshenga & Owuor, 2009;Ayodele, 2002). However, the pattern of tourists' expenditure on souvenirs in relation to their sociodemographic characteristics has been left out of the picture. Again, research on souvenirs and its' importance to national economies and the tourists is paramount to tourism development (Muganda, Sirima, Moshy & Mkumbo, 2012;Bao, 2006) but the issue in Ghana has not received much attention from tourism academic and tourism regulatory bodies in Ghana.
This study therefore sought to generally examine the expenditure of tourists on arts and crafts in Accra, Ghana. Specifically, the study sought to analyse the socio-demographic characteristics of international tourist to Accra, Ghana; determine the level of tourist expenditure on souvenirs and analyse the relationship between socio-demographic characteristics and expenditure on souvenirs.
Methodology
Inbound tourists were the target population who according to Annku and Lodonu (2012) form majority of consumers of arts and crafts in Ghana; 36% Ghanaians as against 64% non-Ghanaians patronize arts and crafts. With the aid of the Fisher, Laing, Stoeckel and Townsend's (1998) formula, a sample size of 196 tourists was selected. The formula used when the target population is more than 10,000 was adopted since the population of tourists exceeds 10,000. The formula used in calculating the sample size is: = 2 2 Where: n = the desired sample size when the population is more than 10,000 z = the normal standard deviation, usually set at 1.96 which corresponds to 95% confidence level; p = the proportion of the target population that has similar characteristics; q = 1.0 minus 'p' and d = the margin of error which is equal to 0.05 If the z-statistic is equal to 1.96, margin of error (d) equals 0.05% and the proportion of the target population with similar characteristic (p) equals 85% (0.85), then (n) would be: n = (1.96) 2 (0.85) (0.15) 0.05 2 n = 196 The sample size for this study is 196. Data obtained from the GTA (2010) in Accra on tourist arrivals to the metropolis in the month of February was 43,967 (above 10,000). Respondents were sampled using convenience sampling at Arts And Crafts Centre, Kwame Nkrumah Mausoleum and the La Beach all in Accra.
The research adopted the descriptive design and a selfadministered questionnaire was used to collect primary data from respondents. Respondents were approached individually to avoid their responses being influenced by others. Respondents were assured of anonymity and confidentiality after asking for their consent to respond to the research instrument. The data collected were processed using the Statistical Product for Service Solution (SPSS) version 16.0. It was therefore cleaned and analysed using descriptive statistics such as frequencies and percentages. Chi-square test of independence was employed to analyse relationships between socio-demographic characteristics and expenditure pattern a well as travel characteristics and their expenditure pattern.
Socio-demographic characteristics
The study revealed that 61.2% of respondents were females, while 62.8% were between the age range of 21 and 30 years. More than half of the respondents were high school leavers (63.3%) whilst 21.9% had post graduate certificate. It was also noted that half of the respondents (50.0%) earned less than 10,000 dollars annually while 23% of respondents earned ≥ 35, 000 US dollars. With regards to the travel characteristics, more than a quarter of the respondents were in Ghana for volunteerism (28%) and educational purposes (34.2%).
Souvenirs in travel budget
In the Souvenir in travel budget Fig, it shows that more than two-thirds of respondents (84%) had planned to purchase souvenirs during and after their tour in Ghana. Only about 16% did not plan to purchase souvenirs during or after their tour in the country.
Fig 1: Souvenirs in travel budget
The finding is in line with the assertions of Ming (2011), Cai, Letho and O'leary (2001), Heung and Cheng (2000) and Law and Au (2000) that expenditure on souvenirs forms a significant portion of a traveller's budget and that most travellers plan to purchase souvenirs at their destinations.
Expenditure on souvenirs
The expenditure of tourists in this study is a composition of amounts that the international tourists spend on their needs as well as their wants. As souvenirs are used as objects of remembrance, various amounts are allocated to the purchase of variety of souvenirs. It is therefore noted in the Expenditure on souvenirs table (Table 1) that slightly above half (51.0%) of the respondents spent less than a GH¢100 (less than US$ 22.2) on souvenirs whereas 20.4% spent between GH¢101 (US$ 22.4) and GH¢200 (US$ 44.4) on souvenirs. However, less than a quarter (15.3%, 13.3%) of respondent spent GH¢201 (US$ 44.6) to GH¢300 (US$ 66.6) and GH¢ ≥ 301 (US$ 66.8) respectively. On average however, international tourists in Accra spend about GH¢207 (US$ 46) on souvenirs To determine expenditure on souvenir by social groupings, a cross tabulation was used. In addition, a chi square test was performed to determine whether there is a significant relationship between various social groupings and the amount spent on souvenirs and also, to find out whether there exist any relationship between travel characteristics and the expenditure on souvenirs. It was realised from the Expenditure on souvenirs by socio-demographic/travel characteristics table, (Table 2), that slightly above half (53.1%) of male tourists spent less than GH¢100 (US$ 22.2) on souvenirs while 20.3% spent more than GH¢300 on souvenirs.
On the other hand, more than a third of female tourists spent less than GH¢ 100 on souvenirs whereas about 11 per cent spent more than GH¢300 on souvenirs. More male tourists (20.3%) spent more on souvenirs than their female counterparts (11%).
With reference to age, tourists aged 30 and above spent more on souvenirs. A third of this category spent more than GH¢300 on souvenirs. Less than a quarter of the other age categories spent GH¢300 and above on souvenirs. In terms of education, less than a quarter of respondents within each category; high school (20.8%), degree (15.7%) and postgraduate (6.2%) qualification spent more than GH¢300 on souvenirs. It was also revealed in the table that almost a quarter (21.6%) of the respondents earning above US$35,000 spent more than GH¢300 on souvenirs.
Given the p-values at 0.05 significance levels, there was a statistically significant relationship determined between the expenditure of tourists on souvenirs and the sociodemographic characteristics of respondents, which are sex (p=0.006), age (p=0.028), education (p=0.005) and annual income (p=0.006) and their expenditure on souvenirs.
Hypothesis 1: The chi-square test revealed that there was a statistically significant relationship between the sociodemographic characteristics (sex, age, education, income) and expenditure on souvenirs as shown in Table 2.
3.5 Expenditure on souvenirs by travel characteristics As noted in Expenditure on souvenirs by travel characteristics (Table 3), a quarter of the leisure travellers spent more than GH¢300 on souvenirs whilst 50% spent less than 100 on souvenirs. Less than quarter of tourist visiting friends and relatives (13.3%), volunteers (13.6%), business travellers (7.1%) and those for educational purposes (15.8%) spent more than 300 on souvenirs. In terms of frequency of visit 21.6% of repeat visitors spent more than GH¢300 on souvenirs and more than half of first time visitors spent less than GH¢100 on souvenirs. Tourists were interested in souvenirs. Less than a quarter of tourists who were involved in this study had not planned to buy souvenirs and consequently did not make any allocation for it in their travel budget. It can then be, assumed that these respondents were impulse buyers since the purchase of souvenirs was impulsive. On average, souvenirs constituted about 15.3% of the tourists' expenditure and this is less than the allocation by tourists as revealed by the Georgian National Tourism Agency (2011) but more than expenditure as reported by the Malta Tourism Authority (2011).
About half (51.0%) of the tourists spent less than a GH¢100 on souvenirs and this may be attributed to the occupation distribution of respondents since students formed more than half of the respondents (60.7%). Students rely mainly on part time jobs whilst some are even not working and as such, earn a meagre income. This restrict their expenditure especially on luxury items as souvenirs. In spite of this, the average spending of respondents on souvenirs was relatively high GH¢ 207.
It is interesting to note that few male tourists (20.3%) spent more money in purchasing souvenirs. In the same way, a smaller percentage of their female counterpart (11.0%) spent more money on souvenirs. This finding contradicts the assertion by Bao (2005) and Turner and Reisinger (2001) that female tourists spend more money on shopping for souvenirs than male tourists. With reference to age, respondents aged 30 and above spent more on souvenirs; a third of this category (33.3%) spent more than GH¢300 on souvenirs. This may be due to the fact that respondents above 30 years were likely to be gainfully employed and thus are more likely to have higher discretionary money at their disposal since income earned is likely to have a relationship with spending on souvenirs.
A higher proportion of the tourists who had high school qualification (20.8%) spent more than GH¢300 on souvenirs than tourists with degree (15.7%) and post graduate level (6.2%). This is contrary to the finding of Wang and Davidson (2010) and Yu and Littrell (2005) that tourists with higher levels of education spend more on shopping as the study reflects otherwise, with the least educated tourists, spending more money on the purchase of souvenirs.
Respondents with higher income were expected to spend more on souvenirs and this is reflected by the current study as almost a quarter (21.6%) of the respondents earning above US$35,000 per year spent more than GH¢300 on souvenirs. The findings also revealed a significant relationship between socio-demographic characteristics (sex, age, education and annual income) of tourists and their expenditure on souvenirs.
With reference to the purpose of visit, leisure travellers spent more on souvenirs than business travellers and this is in line with the study by Yoon-Jung (2007) that leisure travellers spend the highest amount of money on shopping. These findings, however, contradict the findings of Luo and Lu (2011) that business travellers spend more on souvenirs than leisure travellers. This contradiction may be affected by the attraction within the area that attraction is found since that of Luo and Lu was a fair and business travellers would have been the main market for this product.
Conclusion
Shopping for souvenirs is an activity that tourist in Accra, Ghana do undertake. Souvenirs purchased translates into income for Ghanaians as well as a source of employment. The expenditure of tourist on souvenirs in Ghana implies that the souvenir business is a venture worth investing into as it has a capacity to attract relatively high tourists spending. Therefore institutions that are responsible for promoting arts and crafts that are usually used as souvenirs, such as Centre for National Culture and Ghana Tourism Authority should actively aim at packaging and promoting the Ghanaian arts and crafts to increase interest in arts and craft and also increase sales of souvenirs. Leisure travellers spent more on souvenirs than other travellers such as business travellers though business travellers are described in the tourism literature as price insensitive. This suggests the need to establish a connection between business travellers and the souvenir items either by establishing outlets in business facilities or directing promotional activities towards business travellers.
More also, first time visitors spent less on souvenirs and may be due to inadequate information on arts and crafts in Ghana or difficulty in assessing arts and crafts outlets/centres. Steps need to be taken to promote this business and make it more viable and accessible to all. In addition sex, age, level of education and annual income have a significant relationship with expenditure on souvenirs. This shows that the social grouping in which a tourist is, be it, age, sex among others may have an influence on the likelihood of purchasing souvenirs during and after touring destination. It can be established that the higher the education level of tourists to Accra, the less likely it is to spend high on souvenirs. Further studies can, however, be conducted to determine the magnitude and direction of the relationship in order to provide a deeper meaning of the relationship that exist between sociodemographic characteristics and expenditure on souvenirs. | 2019-07-22T06:02:23.091Z | 2020-03-12T00:00:00.000 | {
"year": 2020,
"sha1": "a63b8ca2c1f2a877de6a3273b75630e8964912f9",
"oa_license": "CCBY",
"oa_url": "https://ijtmr.ktu.edu.gh/index.php/ijtmr/article/download/60/58",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "8b0bf03f88cc9e1de8bf9131b89adf1a9f9a7864",
"s2fieldsofstudy": [
"Business"
],
"extfieldsofstudy": [
"Geography"
]
} |
118556051 | pes2o/s2orc | v3-fos-license | The Series Product for Gaussian Quantum Input Processes
We present the theory for connecting quantum Markov components into a network with quantum input processes in a Gaussian state (including thermal and squeezed), not necessarily vacuum fields.One would expect on physical grounds that the connection rules should be independent of the state of the input to the network. To compute statistical properties, we use a version of Wicks' Theorem involving fictitious vacuum fields (Fock space based representation of the fields) and while this aids computation, and gives a rigorous formulation, the various representations need not be unitarily equivalent. In particular, a naive application of the connection rules would lead to the wrong answer. We establish the correct interconnection rules, and show that while the quantum stochastic differential equations of motion display explicitly the covariances (thermal and squeezing parameters) of the Gaussian input fields We introduce the Wick-Stratonovich form which leads to a way of writing these equations that does not depend on these covariances and so corresponds to the universal equations written in terms of formal quantum input processes. We show that a wholly consistent theory of quantum open systems in series can be developed in this way, and as required physically, is universal and in particular representation-free.
Introduction
The quantum input-output theory has had an immense impact on quantum optics, and in recent years has extended to opto-mechanical systems and beyond. The prospect of routing the inputs through a network, or indeed using feedback has lead to a burgeoning field of quantum feedback control [1]- [5]. The development of a systems engineering approach to quantum technology has benefited from having a systematic framework in which traditional open quantum systems models can be combined according to physical connection architectures.
The initial work on how to cascade two quantum input-output systems can be traced back to Gardiner [6] and Carmichael [7]. More generally, the authors have introduced the theory of Quantum Feedback Networks (QFN) which generalizes this to include cascading, feedback, beam-splitting and general scattering of inputs, etc., [8], [9]. One of the basic constructs is the series product which gives the instantaneous feedforward limit of two components connected in series via quantum input processes: in fact, the systems need not necessarily be distinct and the series product generalizes cascading by allowing for feedback. The original work was done for input processes where the input fields where in the Fock vacuum field state. A generalization to squeezed fields and squeezing components has been given [10], however this was restricted to the case of linear coupling and dynamics: there it was shown that the resulting transform analysis could be applied in a completely consistent manner. More recent work has shown that non-classical states for the input fields, such as shaped single-photon or multi-photon states, or cat states of coherent fields, may in principle be generated from signal models [11], [12] -that is, where a field in the Fock vacuum state was passed through an ancillary dynamical system (the signal generator) which is then to be cascaded to the desired system. Quantum feedback network (QFN) theory concerns the interconnection of open quantum systems. The interconnections are mediated by quantum fields in the input-output theory, [13,8,9]. The idea is that an output from one node is fed back in as input to another (not necessarily distinct) node, the simplest case being the cascade connection (e.g., light exiting one cavity being directed into another). The components are specified by Markovian models determined by SLH parameters which describe the self-energy of the system and how the system interacts with the fields (via idealized Jaynes-Cummings type interactions and scattering).
Here we turn to the problem of the general class of Gaussian states for quantum fields. This includes thermal fields, and of course squeezed fields. In principle, these may be approximated as the output of a degenerate parametric amplifier (DPA) driven by vacuum input, see [13]. In a sense, we have that a singular DPA may serve is the appropriate signal generator to modify a vacuum field into a squeezed field before passing into a given network. We will exploit this in the paper, however, we will have to pay attention to the operator ordering problem when inserting these approximations into quantum dynamical equations of motion and input-output relations.
The programme turns out to be rather more involved than one might expect at first glance. It is always possible to represent a collection of d Gaussian fields using 2d vacuum fields (a Bogoliubov transformation!) and one might hope that the corresponding connection rules applied to the representation in terms of vacuum fields would agree with the intuitive rules one would desire. This turns out not to be the case, and the various feedback constraints cannot be naively applied to the representing fields: the reason is that the represen-tations are a linear combination of creation and annihilation operators for the representing vacuum fields, and we have broken the Wick ordered form of the original equations.
If applied naively, the series product would predict a contribution to the global network model that depended on the covariance parameters of the state. From the physical point of view, this ought to be spurious. In comparison with classical analog linear electronics, we see that the components (e.g. resistors, capacitors, inductors) are described by impedances. When components are interconnected to form a network, the network may be described by an equivalent impedance, derived through an application of Kirchhoff laws. Impedances do not depend on the applied currents or voltages, and are therefore intrinsic to the device or network. Similarly the rules for connecting a quantum feedback network should be intrinsic, and not depend on the state of the noise fields.
Background and Problem statement
Let us begin in the concrete setting of the quantum stochastic calculus of Hudson and Parthasarathy [15] with a fixed initial space h 0 and a noise space that is the (Bose) Fock space over C d -valued L 2 -functions on the time interval [0, ∞). In the language of Hudson and Parthasarathy, we have a multiplicity space of dimension d and we select an orthonormal basis which determines d channels. We denote by A k (t), A k (t) * , and Λ jk (t) the processes of annihilation, creation (for channel j) and scattering (from channel k to channel j). In the following, we shall introduce an Einstein summation convention for repeated channel indices. We will deal with the class of quantum stochastic integrals processes satisfying the appropriate conditions of local integrability, square-integrability [15] without explicit reference. We have for instance the QSDE where the coefficients are adapted and the increments are (quantum) Itō. We have the quantum Itō product formula where the Itō correction comes from the quantum Itō table [15] dΛ with all other products of the fundamental increments vanishing.
Definition 1 Definition The Stratonovich integral is defined algebraically via
This turns out to be equivalent to a mid-point rule [17]. If we consider the QSDE dU (t) = −idE (t) • U (t), with U (0) the identity and E (t) = E jk Λ jk (t) + E j0 B j (t) * + E 0k B k (t) + E 00 a self-adjoint quantum stochastic integral process, then we may convert to the Itō form to get where (setting E ℓℓ to be the d × d matrix with entries E jk) is called the matrix of scattering coefficients unitary (that is, which is the column vector of coupling operators, and where H is the Hamiltonian (H * = H = E 00 + 1 2 E 0j Im For simplicity we will assume that the terms S jk , L j and H are bounded operators on the system Hilbert space h 0 . We generally refer to the triple G ∼ (S, L, H) as the Hudson-Parthasarathy parameters, or informally the "SLH" parameters specifying the model. The unitary process they generate may be denoted as U G (t) if we wish to emphasize the dependence on these parameters.
For X an operator of the initial space, we introduce j t (X) = U (t) * X U (t) and from the quantum Itō rule obtain the Heisenberg-Langevin equation (11) and the Lindblad generator L 00 ≡ L is The maps L αβ are known as the Evans-Hudson super-operators. We shall occasionally write j G t (X) for the dynamical flow of X when we wish to emphasis the dependence on the SLH parameters G.
Let us now write the input processes as A in,j (t) = A j (t) and introduce the output processes as A out,j (t) = U (t) * A in,j (t) U (t) then from the quantum Itō rule we see that
Thermal Fields
Considering the single channel (d = 1) case for the moment, we may introduce non-Fock quantum stochastic processes as follows [18]. For n > 0, we set which are canonical fields on the Fock space with a pair of channels labeled as k = ±. In fact, the map ( This is of course based on an Araki-Woods representation of the fiels [19]. As is well known, these transformation cannot be implemented unitarily. However, from a quantum optics point of view, devices transforming or even squeezing fields in this manner are frequently considered, and it is useful to imagine a hypothetical device -a Bogoliubov box -performing such a canonical transformation on our idealized fields.
Ignoring the second processB, we obtain the non-Fock quantum Itō table It problematic (read impossible) to incorporate a scattering process Λ into this table. We refer to B as non-Fock quantum noise. We need to drop the scattering term from the unitary evolution equation, i.e. set S ≡ I, and with we have where For the flow, we need that the Hudson-Evans super-operator associated with the scattering terms are trivial. This is the case when the entries of the scattering matrix S are (e.g. scalars) commuting with operators of the initial space, but we can get away without assuming that S ++ S +− S −+ S −− is the identity. By inspection we find that flow equation will take the form if and only if we take S ++ S +− S −+ S −− ≡ S 0 0 S * -otherwise we obtain the other noiseB -and in which case the Lindbladian is
The Series Product -Vacuum Inputs
In [9] the authors introduce a rule for combining SLH models in series. For instance, we have the output of the G A ∼ (S A , L A , H A ) fed instantaneously as input to G B ∼ (S B , L B , H B ) and it is shown that this is equivalent to the model generated by Here Im {C} means 1 2i (C − C * ).We note that every model may be written as a purely scattering component and a non-scattering component in series, since we have the law (S, L, H) = (I, L, H) ⊳ (S, 0, 0).
We should remark that it is not necessary to view the two systems A and B as separate systems -specifically, in the derivation of the series product [9] it is not assumed that the A and B operators need commute!
Statement of the Problem
If we wish to have a pair of systems A and B (both accepting d inputs) in series, then we obtain an equivalent Markov model in the limit where the intervening connection is instantaneous. Let L A be the column of the d operators L A ,k , k = 1, · · · , d, and similar for system B. The series product says that the equivalent model has coupling L A + L B and Hamiltonian Suppose we were to apply the series product to two systems with the same single thermal input B, and try and describe this as a series connection using the two vacuum inputs A + and A − . Naively applying the series product to the construction in the A ± format leads to the correct rule L A + L B for the coupling terms, but We have picked up an n-dependent term. For pure cascading, the systems A and B are distinct and so [L * B , L A ] = 0. However, the series product should also apply to the situation where the systems share degrees of freedom. In such cases the additional term is physically unreasonable as it depends on the state of the noise.
It is not immediately obvious what is wrong with the construction. Going to the double Fock vacuum representations and then using the vacuum version of the series product would seem a reasonable thing to do. However, a fully quantum description would involve theB fields as well, and at a schematic level this would involve one or more Bogoliubov boxes -something conspicuously. We will give the correct procedure in this paper.
Notation
We will use the symbol to signify a defining equation. We will denote the operations of complex conjugation, hermitean conjugation, and more generally adjoint by *. For X = [x ij ] an n× m array with complex-valued entries, or more generally operator-valued entries, we write X * for the m × n array obtained by transposition of the array and conjugation of the entries: that is the ij entry is x * ji . The transpose alone will be denotes as X ⊤ , that is the m × n array with ij entry x ji . We will also use the notation X # = (X ⊤ ) * which is the n × m array with ij entry x * ij .
Finite Dimensional Gaussian States
Let a 1 , · · · , a d be the annihilation operators for d independent oscillators. We consider a mean zero Gaussian state with second moments which we assemble into a hermitean d × d matrix, N , with entries n * ji = n ij , and a symmetric matrix M is the d × d matrix with entries m ij = m ji . The covariance matrix is In order to yield mathematically correct variances, we must have both F and N positive. The vacuum state is characterized by having N = M = 0, that is The covariance matrix F defined by (27) must be positive semi-definite, as will be the matrices N and I + N ⊤ . We must also have ran(M ) ⊆ran A linear transformation of the form is called a Bogoliubov transformation if we have again the canonical commutation relations for the primed operators.
The transformationã = U a + V a # is Bogoliubov if and only if the following identities hold This is easily established by inspection, as are the following.
where W = ∆ (U, V ). In particular, the new matrices are Lemma 3 Lemma Given a vac with the choice of the vacuum state, the Bogoliubov transformation a = U a vac + V a # vac leads to operators with the covariance matrix where W = ∆ (U, V ) and We note that the determinant of the covariance matrix is preserved under Bogoliubov transformations. In particular, if we have F = W F vac W * , as in the last Proposition, then F must also be singular. This means that if we wish to obtain a given covariance matrix F for d modes by a Bogoliubov transformation of vacuum modes, we will typically need a larger number D of these modes with F being a sub-block of a transformed matrix W F vac W * . The example in the Theorem shows that in order to obtain the d = 1 covariance we need a Bogoliubov transformation of D = 2 modes. We remark that we may obtain the covariance with the constraint |m| 2 ≤ n (n + 1) ensuring positivity, from 2 vacuum modes via [21,20] The maximal case |m| 2 = n (n + 1) may be obtained from a single mode a 1 via
Quantum Ito Calculus: Gaussian Noise
One would like to extend this to non-vacuum inputs, in particular, those with general flat power Gaussian states for the noise. (We restrict to a single noise channel for transparency but the generalization is straightforward enough.) It is possible to construct noises having the following quantum Itō table where N = [n ij ] and M = [m ij ] have the same properties and constraints as introduced above.
In reality, we are assuming that the fields B j (t) correspond to a representation on a double Fock space, say, are copies of the Fock fields encountered above, The underlying mathematical problem is that we are trying to implement a canonical transformation that is not inner [22,23,24]-specifically the various representations for different pairs (N, M ) are not unitarily equivalent.
Instead we must restrict to QSDE models in the general Gaussian case which are driven by B and B * only. We in fact find the class of QSDEs generating unitaries and we now require that with H again self-adjoint. Let us denote the conditional expectation from the algebra of operators on the system-tensor-Fock Hilbert space down to the system operators (i.e., the partial trace over the Gaussian state) as E (N,M) [·|sys]. As the differentials dB k (t) and dB k (t) * are Itō (future pointing) their products with adapted operators will have conditional expectation zero. Therefore and we deduce that The corresponding Heisenberg-Langevin equations are of the form where the new Lindbladian is Likewise, we find that A little algebra allows us to relate these to the vacuum expressions:
Representation-Free Form
Returning to the problem stated in the Introduction, we have that all the U respectively, where K and L are the Fock representation expressions (9) and (12).
Proof. We first observe that and substituting the QSDE (38) for dU and using the quantum Itō table (36) gives and similarly (51) Combining these terms and using the identity (45) shows that (47) is equivalent to (38).
For the Heisenberg equation, we first note that and similarly (53) Combining these terms and using the identity (46) shows that (48) is equivalent to (43).
Note that in both equations (47) and (48) the Stratonovich differentials occur in Wick order relative to the integrand terms. What is remarkable about these relations is that they are structurally the same as the Fock vacuum form of the QSDEs with S = I. We say that the equations (47) and (48) are representationfree in the sense that they do not depend on the parameters N and M determining the state of the noise.
White Noise Description
We now present a more formal, but insightful account of quantum stochastic processes. Consider a collection of quantum noise input processes {b k (t) : t ∈ R, k = 1, · · · , d} obeying the commutation relations We wish to model the interaction of a quantum mechanical system driven by these processes, and to this end introduce a unitary dynamics given by where (with an implied summation convention with range 1,· · · , d) Here L k and H = H * are system operators. The time ordering T is understood in the usual sense of a Dyson series expansion. From this we may arrive aṫ We claim that U (t) should correspond to the evolution operator for G ∼ (S = I, L, H) without due reference to a particular state for the noise. If we fix the state, say the vacuum, then we use Wick ordering to compute the partial expectations with respect to that state.
To see how to proceed, let us consider a general quantum stochastic integral X (t) described by a formal equatioṅ where the terms x αβ (t) are "adapted" in the formal sense that they do not depend on the noises b k (s) for s > t. As we are talking about the vacuum representation for the time being, we can bootstrap from the vacuum |Ω to construct the Fock space as the completion of the span of all vectors of the type f k(1) (t 1 ) b k(1) (t 1 ) * · · · f k(n) (t n ) b k(n) (t m ) * |Ω , and moreover we can build up the domain of exponential vectors. We quickly see that (58), with Wick ordered right hand side, corresponds to the QSDE Our issue however is how do we put to Wick order a given expression, for instance, the right hand side of (57).
Proposition 5 Proposition For the process X (t) described by (58), we have We may justify this as follows: with the factor of 1 2 coming from the half-contribution of the δ-function. Evidently what the equations in (60) correspond to is our definition of a Stratonovich differential -at least for the Fock vacuum representation. While we can make a connection between (58) and the rigorously defined Hudson-Parthasarathy processes, it should be appreciated at the very least that (60) is the correct mnemonic for doing the Wick ordering -an attempt to convert into a Dysontype series expansion and Wick ordering under the iterated integral signs to get a Maassen-Meyer kernel expansion shows this. At work here is an old principle that "Itôs formula is the chain rule with Wick ordering" [16]. Let us now examine (57) and put it to Wick ordered form. By a similar argument, we have By means of this we may place (57) into the Wick-ordered form and picking up the correct vacuum damping (9), K, as a result.
Setting X t = U (t) (X ⊗ I)U (t), the same Wick ordering rule can be applied to the Heisenberg equations to obtaiṅ (64) Here we use the notation L k,t = U (t) (L k ⊗ I)U (t), etc.
We also remark that we may define the corresponding output fields by where T > t. One may show that the input-output relations are If, on the other hand, we want the state of the noise to be a mean-zero Gaussian with correlations, say then we represent the noise as employing a suitable Bogoliubov transformation. Here we now have double the number of quantum white noises a +,k and a −,k but these are represented as Fock processes.
If we now substitute (68) into (57) we see explicitly that the a ±,k are out Wick order, but this can be rectified by the same sort of manipulation as above.
Once the a ±,k (t) are Wick ordered, we have a equation which we can interpret as the Itō non-Fock QSDE, and this leads to the correct expressions K (N,M) and L (N,M) in the unitary and flow equations respectively.
Given a Gaussian state · on the noise, we may introduce a conditional expectation according to E [·|sys] : A ⊗ B → B A. For instance, E [U (t) |sys] then defines a contraction on the system Hilbert space and we have Now the expression E [Υ sn · · · Υ s1 |sys] will be a sum of products of the operators L, −L * and H times a n-point function in the fields. Similarly, we obtain a reduced Heisenberg equation. To compute these averages we need to be able to calculate n-point functions of chronologically ordered Gaussian fields -this is the realm of Wick's Theorem, so what we have presented may be interpreted as a Gaussian Wick's Theorem [26]. We of course recover the partial traces appearing in the previous section.
Approximate Signal Generator for Thermal States
In this section we show how to go from a general SLH model driven by the output of a Degenerate Parametric Amplifier (DPA) to the limit where the same SLH model is driven by a thermal white noise. We start with the single channel for simplicity.
The Thermal White Noise as Idealization of the Output of a Degenerate Parametric Amplifier
We now show that in the strong coupling limit the output of a degenerate parametric amplifier approximates a thermal white noise. the model consists of a system of two cavities modes c + and c − coupled to input processes A + (t) and A − (t) respectively. Both inputs are taken to be in the vacuum state and the Schrödinger equation iṡ with initial condition U 0 = I and Here ε > κ and k > 0 is a scaling parameter which we eventually model to be large. It is more convenient to work with the white noises a ± (t). The model is linear and we obtain the input-output relations in the Laplace domain to be [10] b s 2 +2κ+κ 2 −ε 2 . In the limit k → ∞ we find the static (s-independent) coefficients and returning to the time domain, the limit output fields are just a Bogoliubov transform of the inputs Here the parameter n corresponds is n = 2εκ It is instructive to look closely at the finite k equations. We have the Heisenberg equationsċ and for k large we may ignore theċ + (t) andċ − (t) terms leaving a pair of simultaneous equations which we may solve to get The output is then and likewisẽ It is relatively straightforward to find a multi-dimensional version of this for a general Bogoliubov transformation
Cascade Approximation
The DPA which is described by driven by the (vacuum) input pair a + (t) a − (t) . It is then put in series with which means that the output a + (t) is fed in as input to the system G ∼ (S, L, H) and a − (t) is left to go away unhindered, G trivial ∼ (1, 0, 0). According to the series product rule, we get DPA and system in series is described by, From this we obtain the Heisenberg equationṡ We now make the approximation √ kc Here we have n = 2εκ , as before. We now make a key assumption: the scattering term S corresponds to a static element. In this case S ≡ e iθ for some real θ. The limit Heisenberg equation therefore simplifies tȯ We are not quite finished as the operators a − (t) and a − (t) are out of Wick order. However, this is easily remedied. For instance, we easily deduce that so that we arrive at Similarly [a − (t) , Y t ] = 1 2 √ n [L * , Y ] t and therefore we get the Wick reordering This leads to the form of the quantum white noise equation with both a + and a − Wick ordered aṡ At this stage we recognize (89) as the equivalent form of the Heisenberg quantum stochastic differential equation for thermal noise.
We also remark that the output process determined by systems in series is B out (t) = U * t A + (t) U t , and from the quantum stochastic calculus we have Using (76) we approximate this as and we have the constraints B dt, consistent with B in driving system A which in turn drives B, corresponds to the dynamics given by the intrinsic series product (22). Proof. We have to show consistency of the quantum stochastic Heisenberg evolution j t (·). To this end we take the open loop equations and impose the constraint dB which we may rearrange as However, the dt term can be recast using the identity The resulting Heisenberg dynamics is therefore the same as for the model (I, L, H) with L = L A + L B , and H = H A + H B + Im{L * B L A }. This is, of course, the form predicted by the series product in the Fock case (22).
Including Scattering
As mentioned above, it is not possible to construct a well defined scattering processes Λ jk in the non-Fock theory. Nevertheless, the effects of static beamsplitter scattering S may be included in a straightforward manner without directly considering unitary QSDE models involving the scattering processes. A clue on how to proceed is given by our earlier observation that if the scattering matrix S entries commute with systems operators -physically, a static beamsplitter -the scattering processes disappears.
In the Fock representation, we could always take the input field A in and apply a unitary rotation A = SA in before passing it though as drive for component. As we have seen, this will require a compensating rotation of the coupling operators, but no change to the Lindbladian. There is also a rotation of the output, however, anticipating this we make the following definition.
Definition 6 Definition Let G andG be SLH model parameters which, for given input noise A in =à in lead to output noises A out andà out respectively. We say that the models' input-output relations are related by a static beamsplitter matrix S if we have The following result shows that for the Fock representation, if the scattering is just a static beam-splitter, then we can produce a related model which avoids the use of the scattering processes. Proof. The Heisenberg dynamics generated by G is (the scattering terms vanish for a static beam-splitter) where and the Lindblad generator is The Heisenberg dynamics forG similarly has no scattering terms in its QSDE, and we see that From the unitarity and scalar nature of S we have that Therefore the QSDEs corresponding to the Heisenberg dynamics for G andG are identical. The input-output relations for G are while forG we have If we require the inputs to be the same (A in = B in ) then we have A out = S B out .
Our strategy for introducing static beam-splitter scattering into the situation where we have non-Fock noise input fields is to say that the initial input A in be replaced by the rotated input SA in , and exploit the fact that the Heisenberg dynamics no longer involves the scattering processes Λ jk explicitly.
Lemma 8 Lemma (The Universal Heisenberg QSDE Description) The Heisenberg dynamics for a general (S, L, H) model with a static beam-splitter matrix S are given by the QSDE for all mean-zero Gaussian input fields A in .
This is of course just the equation (89) written in the Wick-Stratonovich form so as to be representation free! Now let us try and repeat or analysis from Section 7.1. Let us now consider the situation where a Gaussian input A in = A (1) in is driving a system with SLH parameters (S A , L A , H A ) and that its output A where A in +j t (L A ) dt, and the Lindbladians L S are as before.
Proof. Substituting the processes into the QSDEs yields A similar calculation to before shows that The resulting Heisenberg dynamics is therefore same as for the modelG ∼ The correct output for this should however be A out = S B S A B out so that
Conclusions
We have shown that there is a consistent theory for quantum input-output models in series when the driving input processes are in general Gaussian states with a flat power spectrum. This emerges fairly explicitly at the level of the singular input processes b k (t) themselves, but to have a working theory we need to make the connection to the Hudson-Parthasarathy quantum stochastic calculus. This involves quantum stochastic differential equations on the Fock spaces used to represent the noise (which are a mathematical convenience and not physical objects) with the result that the associated dynamical equations appear to depend on the choice of Gaussian state of the noise. In reality this is a mathematical artifact and we show that even here there is a way of expressing the quantum stochastic differential equations (the Wick-Stratonovich form introduced in this paper) which removes these terms. In effect, it is the Wick-Stratonovich form that translates in the physically relevant dynamical equations written in terms of the quantum input processes b k (t). The connection rules are then shown to be genuinely independent of the choice of state. We were also able to include the effects of a static beamsplitter component. At first sight this would seem problematic as the scattering terms Λ jk (t) are not well-defined for non-vacuum states, however, it is possible to ignore them from the model: in fact we need to work at the level of the Heisenberg flow and the input-output relations, neither of which involve the scattering terms. The result is that we may account for static scattering and we find that the series product of [9] again gives the correct rule. In this way we extend the series product to deal with quantum feedback networks driven by general Gaussian input processes. | 2016-08-17T12:00:21.000Z | 2016-02-05T00:00:00.000 | {
"year": 2016,
"sha1": "af9e42306fce3f62e4a2ba5bbca676ecf7588a06",
"oa_license": null,
"oa_url": "http://arxiv.org/pdf/1602.01991",
"oa_status": "GREEN",
"pdf_src": "Arxiv",
"pdf_hash": "af9e42306fce3f62e4a2ba5bbca676ecf7588a06",
"s2fieldsofstudy": [
"Physics"
],
"extfieldsofstudy": [
"Mathematics",
"Physics"
]
} |
5714057 | pes2o/s2orc | v3-fos-license | Orbital Emphysema Following Ocular Trauma and Sneezing
Orbital emphysema is typically a benign condition that occurs following forceful injection of air into the orbital soft tissue spaces. In many cases there is a history of trauma and fracture of an orbital bone, which permits air entry. However, other mechanisms of orbital emphysema have been reported including infection, pulmonary barotrauma, injury from compressed-air hoses, and complications from surgery including dental procedures. Here, we describe a report of a teenager who suffered an isolated medial orbital wall fracture while playing basketball, and several hours later developed orbital emphysema acutely after sneezing. We will review the radiological evaluation of orbital fractures and emphysema.
Orbital emphysema is typically a benign condition that occurs following forceful injection of air into the orbital soft tissue spaces. In many cases there is a history of trauma and fracture of an orbital bone, which permits air entry. However, other mechanisms of orbital emphysema have been reported including infection, pulmonary barotrauma, injury from compressed-air hoses, and complications from surgery including dental procedures. Here, we describe a report of a teenager who suffered an isolated medial orbital wall fracture while playing basketball, and several hours later developed orbital emphysema acutely after sneezing. We will review the radiological evaluation of orbital fractures and emphysema.
Orbital emphysema is typically a benign condition that occurs following forceful injection of air into the orbital soft tissue spaces. In many cases there is a history of trauma and fracture of an orbital bone, which permits air entry. However, other mechanisms of orbital emphysema have been reported including infection, pulmonary barotrauma, injury from compressed-air hoses, and complications from surgery including dental procedures. Here, we describe a report of a teenager who suffered an isolated medial orbital wall fracture while playing basketball, and several hours later developed orbital emphysema acutely after sneezing. We will review the radiological evaluation of orbital fractures and emphysema.
Case Report
A 15-year-old male presented to our emergency room 24 hours after he was playing basketball and was hit in his right eye by a fellow player's knee. Following the impact the patient said that his eye was sore, but he did not experience blurry or double vision. However, several hours after the game, the patient sneezed and felt like the skin around his eye "puffed out". Following this swelling, there was no deterioration or change in his vision; he chose to present to the emergency department as there had not been resolution of the swelling over a 24-hour period. On examination, the patient denied any visual disturbance. Visual acuity was 20/60 on the left and 20/80 on the right; the patient reported this was his baseline as he normally wore corrective lenses, but did not have them with him at the time of exam. He denied headache, nausea, vomiting, dizziness, nasal congestion or bloody discharge from his nose or eye. There was mild swelling of the right upper eyelid, but no crepitus or palpable bony step-off around the orbital rim. There was no proptosis and the remainder of his optical exam was normal including extraocular motion and pupillary reflex.
In light of the mechanism of injury and the abrupt onset of orbital swelling after the patient sneezed, there was a high suspicion for an orbital fracture with concomitant orbital emphysema. Computed tomography (CT) of the facial bones including the orbits was performed with contiguous axial imaging ( Figure 1); coronal and sagittal reformatted images were generated (Figures 2,3). There was a right medial orbital wall (lamina papyracea) fracture, which was depressed by approximately 3 mm. A small amount of retro-orbital fat herniated into the adjacent right ethmoid air cells without entrapment of the medial rectus. There was gas between the right medial rectus and the medial orbital wall extending anteriorly into the periorbital soft tissues and superiorly into the supraorbital fat. While the patient's symptom of orbital swelling after sneezing suggest that orbital emphysema occurred after the sneeze, as no imaging was obtained prior to the patient's sneeze we are unable to formally exclude the possibility that there was orbital emphysema before the sneeze.
The patient was discharged from the emergency department in stable condition with a follow-up appointment with an ophthalmologist, a course of oral antibiotics, and instructions to use nasal decongestants and avoid blowing his nose. Surgical repair and decompression were not performed, and several weeks later the patient was doing well without sequela.
Discussion
Orbital emphysema typically results from forceful entry of air into the orbital soft tissue spaces following an orbital fracture; however, other mechanisms including infection, pulmonary barotrauma, injury from compressed-air hoses, complications from surgery, sneezing, airplane travel, and Boerhaave's syndrome (esophageal rupture) have been reported [1,2]. Fractures typically occur at the thinnest portions of the orbital wall including the medial or inferior orbital wall permitting air entry from the ethmoid and maxillary sinuses, respectively. Although the medial wall (lamina papyracea) is thinner (approximately 0.25 mm thick) than the orbital floor (approximately 0.5 mm thick), fractures of the orbital floor are most common, while isolated medial wall fractures occur in approximately 10-30% of cases of orbital trauma [3][4][5]. In children, the more flexible bones of the orbit are less prone to fracture and shattering, but rather they fracture and function as a trap door which results in a higher incidence of muscle entrapment following orbital wall fractures [6,7]. Air is trapped in the periorbital spaces when the orbital soft tissue acts as a ball valve and presses back the fracture fragment or herniates into the sinus cavity [3,8]. With orbital emphysema there may be complications with a range of severity including proptosis, loss of vision, increased intraocular pressure, and central retinal artery occlusion, with the more severe complications caused by orbital compartment syndrome [3]. In two dramatic cases, one patient developed subcutaneous emphysema and pneuomediastinum following a medial orbital wall fracture, while another developed pneumomediastinum after blowing her nose without prior evidence of ocular trauma [9,10].
CT is effective in identifying the presence and anatomical location of air when orbital emphysema is suspected [11]. Subcutaneous air restricted to the eyelid is classified as palpebral emphysema, and is a rare event arising from disruption of the lacrimal sac and bone [1]. In true orbital emphysema, air is located behind an intact orbital septum, which occurs following fracture of an orbital wall and tearing of the sinus mucosa. A further distinction is made between intraconal (bounded by the medial and lateral rectus muscles) and extraconal orbital emphysema. Air enters into the orbital soft tissue spaces when intranasal pressure increases during coughing, sneezing, or nose blowing [1]. Orbitalpalpebral emphysema occurs when orbital pressure increases beyond the mechanical strength of the orbital septum and air passes freely from the orbit into the eyelid. The presence of orbitalpalpebral emphysema effectively rules out any risk of orbital compartment syndrome, which arises when the pressure of orbital air is sufficient to cause vascular compromise to orbital structures including the retina and the optic nerve. One pitfall in using CT to evaluate the presence of air after orbital trauma is that a wooden foreign body can mimic the presence of orbital emphysema, and the management of these two scenarios is entirely different [12]. There is a wide range of attenuation of wood on CT ranging from air density (-984 HU) to soft tissue/fluid density (+23 HU), which depends upon the type of wood and how much fluid it has absorbed after entering the tissue [5]. In such cases, a careful history is helpful in making the correct diagnosis.
Although radiography has long been the standard for imaging orbital trauma and detecting fractures, it has a high false negative rate (50%) and non-diagnostic rate (30%), which greatly limits it utility [13]. Rather, initial radiological evaluation of orbital fractures is best performed with CT as it can effectively identify bony defects [4,6]. Images in the axial plane permit sensitive examination of medial and lateral orbital walls, while coronal images, which can be obtained directly or though reformatting axial images, are useful in evaluating fractures of the orbital floor and roof. Helical scanning is superior to conventional scanning in the detection of metallic foreign bodies, and when coronal reformats are generated from axial helical scans, imaging time and radiation dose are significantly reduced compared to obtaining direct scans in the axial and coronal planes [4,5]. The reduced scan time and radiation dose are particularly important when evaluating the pediatric population, who are more radiosensitive and may be less cooperative during a scan. In addition to identifying foreign bodies and fractures, it is important for the radiologist to recognize entrapment of extraocular muscles or herniation of periorbital fat into the sinus cavities. Indirect signs of muscle entrapment include kinking or rounding of the rectus [5, 14,15]. The presence of orbital emphysema in the absence of an apparent orbital wall fracture suggests an occult fracture of the orbit [16], and a study by Hunts demonstrated that the location of the orbital emphysema correlates with the fracture location [3]. There is a limited role for MRI in the evaluation of orbital trauma as it has limited sensitivity for detection of fractures and is contraindicated until the absence of a metallic foreign body in the eye has been documented by CT. However, MRI may be useful in further evaluating soft tissue pathology of the rectus muscles, the optic nerve and the brain, and in evaluating vascular damage [5]. Ultrasound is helpful in diagnosing globe defects including foreign bodies, vitreous hemorrhage, lens dislocation, retinal detachment, or globe rupture [6]. Care must be taken; however, to prevent infection or cause unnecessary pain and discomfort when applying the transducer to an orbital wound.
The management of orbital fractures and orbital emphysema depends largely on the clinical presentation and imaging findings. Clinical findings associated with orbital fractures and orbital emphysema include orbital pain, hypoesthesia, restriction of ocular motion, proptosis or enopthalmos (posterior displacement of the globe into the orbit), diplopia, and vision loss. If there is suspicion for orbital compartment syndrome, emergent decompression is necessary and is typically performed by either canthotomy/ cantholysis or needle aspiration of trapped air, which can lead to rapid decompression and resolution of symptoms [8]. Otherwise, orbital emphysema will resolve on its own as the air is absorbed. Surgical repair of orbital fractures within two weeks is indicated in patients with diplopia and CT evidence of entrapped muscle or periorbital tissue, large fractures (>50% of the wall), and enopthalmos that does not resolve. Entrapment of soft tissue may stimulate the oculocardiac reflex, which may stimulates strong vagal responses including bradycardia, nausea, vomiting, syncope and heart block. In these scenarios, rapid repair and release of entrapped soft tissues is recommended [17]. Conservative management is generally reserved for patients with minimal diplopia, preserved ocular motility and the absence of marked enopthalmos [5].
There have not been clinical trials to adequately assess the importance of antibiotics in the management of orbital fractures and this remains an area of controversy. A survey of maxillofacial surgeons reported that 47% prescribed antibiotics at the time of orbital fracture diagnosis; however, another report suggested that antibiotics are only indicated in the setting of an open fracture, concomitant or recent sinus infection, or an immunocompromised state [5, 18].
Our case is a presentation of orbitalpalpebral emphysema caused by sneezing and preceded by fracture of the medial orbital wall. Similar cases have been reported in the adult population, and there has been one other report in the pediatric population of trauma and orbital emphysema resulting from a snowboarding accident followed by nose blowing [19,20]. Interestingly, there have been several reports of orbital emphysema after sneezing or nose blowing without history of trauma, or many months after orbital trauma [1,[21][22][23][24]. Direct trauma to the orbit; however, followed by forceful air entry into the soft tissues spaces during nose blowing, sneezing, or coughing, is the most common cause of orbital emphysema.
In our patient, other than orbital swelling and tenderness, there were no complications from the isolated medial wall orbital fracture and orbitalpalpebral emphysema, which was diagnosed by CT. In the absence of any ocular symptoms and no CT evidence of muscle entrapment, there was no indication for surgical repair. He was prescribed a course of antibiotics and the patient recovered well with conservative management. | 2018-04-03T01:51:32.962Z | 2015-11-06T00:00:00.000 | {
"year": 2008,
"sha1": "6c9731d4e00dc1d49703a586c9b02c937ab61ec2",
"oa_license": "CCBYNCND",
"oa_url": "https://doi.org/10.2484/rcr.v3i1.124",
"oa_status": "GOLD",
"pdf_src": "PubMedCentral",
"pdf_hash": "6c9731d4e00dc1d49703a586c9b02c937ab61ec2",
"s2fieldsofstudy": [
"Medicine"
],
"extfieldsofstudy": [
"Medicine"
]
} |
246822764 | pes2o/s2orc | v3-fos-license | Detailed chemistry modelling of rotating detonations with dilute n-heptane sprays and preheated air
Utilization of liquid fuels is crucial to enabling commercialization of rotating detonation engines in the near future. In this study, Eulerian-Lagrangian simulations are conducted for rotating detonative combustion with dilute n-heptane sprays and preheated air. Two-dimensional flattened configuration is used and a skeletal chemical mechanism with 44 species and 112 elementary reactions for n-heptane combustion is adopted. The flow structure, droplet distribution, and thermochemical parameters in the refill zone are first analyzed. It is shown that the mixture in the refill zone is heterogeneous, including evaporating droplets, vapor, and air. When the total temperature is below 950 K, the average equivalence ratio increases with the total temperature. When it is higher than 950 K, the average equivalence ratio is almost constant. Subsequently, the chemical explosive mode analysis is applied to identify the controlling reactions and dominant combustion modes in the fuel refill zone and reaction fronts. Results demonstrate that the initiation reaction and low-temperature reaction are dominant in the upstream and downstream of the refill zone, respectively. The intermediate species from low-temperature chemistry is found to be important for the chemical explosive mode in the undetonated mixture. The influence of species diffusion and dispersed droplets is further analyzed. Results show that vapor autoignition facilitated by droplet evaporation occurs in the refill zone. Finally, the effects of the air total temperature on the detonation propagation speed and RDE propulsion performance are investigated. It is found that the detonation propagation speed and specific impulse increase with air total temperature.
Introduction
Rotating detonation engine (RDE) is deemed a promising pressure-gain combustion technology due to high thermodynamic cycle efficiency [1]. Fuel flexibility is crucial to materializing RDE towards a practical propulsion system. Typically, liquid fuels have high energy density and are convenient to be stored and transported. In recent years, the interests in liquid fuel RDE have revived. For instance, Bykovskii et al. successfully achieved two-phase rotating detonation waves (RDW) using kerosene sprays with oxygen-enriched air (O2/N2 = 1:1 by vol.) [2] or small addition of hydrogen or syngas [3]. They also found that hydrogen addition enables a more compact RDE combustor [3]. Kindracki [4] also performed liquid kerosene with hydrogen addition in RDE experiments. Rotating detonations were successfully achieved, with a velocity deficit of 20% − 25%, relative to Chapman-Jouguet (C-J) value. More recently, Wolański et al. [5] partially mixed preheated liquid Jet-A or gasoline with hot air, and the resultant reactant composition is higher than the rich flammability limit. They realized a rotating detonation without hydrogen addition. The RDW speed deficit is up to 35%, due to possible losses arising from chamber wall heat transfer, pre-injection deflagration, or reactant nonuniformity [5].
To shed further light on RDE with liquid fuels, numerical studies have also been available. For instance, Hayashi et al. [6] found that a steady JP-10/air RDW can be achieved within wide windows of equivalence ratio and pre-vaporization degree. They also reported that for certain JP-10 droplet concentration (e.g., 0.08-0.163 kg/m 3 for 3 µm droplets), detonation failure occurs. Moreover, Ren and Zheng [7] observed that under ramjet-like conditions rotating detonations with kerosene sprays (no pre-vaporization) can be achieved in a limited range of total pressure (5-7 atm) and increased total temperature is conducive to RDW stability. Besides, a bifurcated wave structure is observed near the spray injector [7].
Meng et al. [8] focused on a more volatile liquid fuel, n-heptane, and considered partially prevaporized n-C7H16 and air mixture to systematically evaluate the influences of droplet diameter (5-50 µm) and pre-vaporization degree on detonation speed and evaporating fuel droplet distribution. They found that the detonation speed decreases with decreased prevaporization degree or increased droplet diameter. They also analyzed the detailed gas-liquid two-phase flow structure and found that a layer with high vapor concentration exists between the droplet-laden mixture/combustion product contact surface [9]. Besides, considering the same fuel, Zhao and Zhang [10] found that the detonation propagation speed increases as the total equivalence ratio increases for the same droplet diameter. When the droplet diameter is less than 5 µm, thrust force from kinetic energy and pressure gain decreases with droplet size. Beyond 5 µm, the former first increases and then decreases with the droplet diameter, while the latter has limited change [10].
The key feature of liquid fuel RDE is that the combustion proceeds in vapor-droplet two-phase mixtures. Since preheated air is widely adopted in practical tests [5], this renders the static temperature and pressure of the undetonated mixture sufficiently high, probably inducing unexpected pre-RDW autoignition. It is shown that low-temperature chemistry (LTC) plays a significant role in autoignition process of liquid sprays in a hot atmosphere, featured by pronounced negative temperature coefficient (NTC) or zero temperature coefficient (ZTC) phenomenon [11]. However, due to the limitations of measurements and modelling approaches (simple chemistry, e.g., in [7][10]), chemical structure and reaction progress in the fuel refill zone of a liquid fuel RDE with hot air have not been studied yet. In this work, we will conduct Eulerian-Lagrangian modelling of rotating detonations with dilute liquid n-heptane sprays. Different from our previous work [8][9][10], a detailed mechanism with 44 species and 112 elementary reactions [12] will be employed for n-heptane combustion. The objectives of our study are to clarify: (1) the thermochemical conditions in the heterogeneous fuel refill zone of a liquid fuel RDE; (2) the chemical structures in n-heptane rotating detonations; (3) the effects of preheated air total temperature and LTC on detonation speed and propulsion indices. Figure 1 shows the two-dimensional computational domain of a flattened model RDE chamber. The flow structure is composed of a RDW, deflagration surface, and oblique shock. The length (x-direction) and width (y) of the domain are 81 mm and 50 mm, respectively. As annotated in Fig. 1, the outlet is non-reflective due to the local supersonic flows. Periodic conditions are enforced at the left and right sides, such that the RDW can propagate across the domain with continuous cycles. The computational domain in Fig. 1 Two-phase heterogeneous mixture is injected from a continuous inlet of the domain (at y = 0), including carrier gas and n-C7H16 droplets. The carrier gas is oxygen-enriched air ( 2 : 2 = 0.432:0.568 by mass), which is deemed conducive for RDW stabilization by Bykovskii et al. [2]. Different from our previous studies [8][9][10], liquid fuel pre-vaporization prior to injection is not considered here and therefore n-C7H16 vapor mass fraction in the carrier gas is zero. The total pressure of the carrier gas is 0 = 20 atm in all simulations. We run a series of spray RDE cases and find that when the total temperature, 0 , is less than 700 K, the RDW cannot stabilize. A slightly higher critical temperature (1,000 K) is reported in Ref. [7], which may be because less volatile fuel (i.e., kerosene) is considered in their simulations. In the following part of this paper, 0 from 700 to 1,400 K will be studied. The gas injection follows the isentropic relations, considering the carrier gas total pressure ( 0 = 20 atm) and predicted gas pressure inside the chamber, and the droplet injection is synchronized with the carrier gas for both injection timing and velocity [9,10].
Physical model for spray RDE
The droplet initial temperature is assumed to be 323 K, to mimic pre-heating by the hot carrier gas in the upstream plenum, as implemented in the RDE tests by Kindracki [4] and Wolański et al. [5]. This enables efficient gasification of fuel droplets inside the combustion chamber and therefore promote detonative combustion efficiency. The initial material density and heat capacity of liquid n-heptane are 680 kg/m 3 and 2,952 J/kg/K, respectively. Well sprayed (hence fine-grained) droplets are favorable for rotating detonations [4]. In this study, mono-sized droplets are considered, with the initial diameter 0 = 5 μm. Droplet aerodynamic breakup is modelled following Ref. [13]. The liquid fuel equivalence ratio is used to parameterize the droplet loading, defined as the mass ratio of the droplets to the oxidant. In our simulations, = 1 is used for all cases.
Numerical method
The Eulerian-Lagrangian method is used to simulate rotating detonations with sprayed liquid fuel droplets. The gas phase is described with the Eulerian method, whilst the individual fuel droplets are tracked with a Lagrangian fashion. Two-way coupling between the gas and liquid phases are implemented, considering the exchanges of mass, momentum, and energy between them. For the gas phase, the Navier-Stokes equations of mass, momentum, total nonchemical energy, and species mass fraction are solved.
For the liquid phase, the droplets are spherical and point-force approximation is adopted. The temperature gradient inside the droplets is neglected, considering their small Biot numbers (~0.057 when the droplet temperature and diameter are 323 K and 5 μm). Droplet interactions (e.g., collisions) are not considered, which is not important for dilute sprays with initial droplet volume fraction less than 0.1%. Besides, since micro-sized droplets are considered, the Basset force, lift force, and body force are not included. With these considerations, the evolutions of mass, momentum, and energy for a single droplet follow where is time and m d = 3 6 ⁄ is the mass of a single droplet, with and being the droplet material density and diameter, respectively.
is the droplet velocity vector, , is the droplet heat capacity, and is the droplet temperature. The droplet evaporation rate ̇ in Eq. (1) is modelled as ̇= hln(1 + ) [14], where and respectively are the density and mass diffusivity at the film over the droplet surface. The Spalding mass transfer number is and ∞ are the fuel vapor mass fractions at the droplet surface and in the gas phase, respectively. The modified Sherwood number h is h = 2 + [(1 + ) 1/3 max (1, ) 0.077 − 1]/ ( ) , with the Schmidt number = 1.0. The function ( ) = (1 + ) 0.7 ln(1 + ) / considers the variation of the film thickness due to Stefan flow effects [14]. In Eq. (2), the Stokes drag is modelled as [15], where and are the gas dynamic viscosity and velocity vector, respectively. The drag coefficient, , is estimated using Schiller and Naumann model [15].
≡ | − |⁄ is the droplet Reynolds number, and is the gas density. Besides, = − ∇ is the pressure gradient force and is the droplet volume.
In Eq. (3), the convective heat transfer rate ̇ is ̇= ℎ ( − ). Here is gas temperature, and is the surface area of a single droplet. ℎ is the convective heat transfer coefficient, following Ranz and Marshall [16]. Furthermore, ̇ accounts for the heat exchange rate associated with the latent heat of evaporation of liquid n-heptane.
The equations for the gas and liquid phases are solved using a customized OpenFOAM code, RYrhoCentralFoam. The solver is carefully validated and verified for shock capturing, molecular diffusion, flame-chemistry interactions and gas-liquid two-phase problems [17][18][19]. For gas phase, second-order backward scheme is used for time marching, and the time step is about 2 × 10 -9 s. Second-order Godunovtype upwind-central scheme is employed to calculate the convection fluxes in the momentum equations. The total variation diminishing scheme is applied for the convection terms in energy and species equations. A detailed mechanism (44 species and 112 reactions) [12] is used. Low-temperature chemistry is included in this mechanism, and hence the LTC effects and two-stage ignition in n-heptane rotating detonations can be studied. One simulation with 88 species and 387 reactions (see details in supplementary document) is also run and the results demonstrate that differences between these two mechanisms are negligible.
For the liquid phase, fuel droplets are tracked from the barycentric coordinates. Equations (1) − (3) are integrated with first-order Euler method and the righthand-side terms are treated in a semi-implicit approach. Details of the numerical method in RYrhoCentralFoam are available in [17][18][19].
The simulations are run on the ASPIRE 1 Cluster from National Supercomputing Centre in Singapore and 360 processors are used for each case. The simulated physical time is about 0.62 − 0.71 ms, corresponding to 15 cycles. Figure 2 shows the distributions of pressure, gas temperature, Heat Release Rate (HRR), and pressure gradient magnitude after the single-waved rotating detonation stabilizes (over ten cycles). The carrier gas total temperature is 0 = 1,000 K. The basic structure of a RDE flow field is well captured. As seen from Fig. 2(b), the temperature in the fuel refill zone (enclosed by the RDW, deflagration surface, and inlet) varies between 510 and 950 K, and the mean is about 822 K. Besides, the pressure in the refill area is 1.78 atm -11.5 atm with a mean of 10.68 atm. A notable feature in Fig. 2(c) is that although HRR is high along the RDW, detonative combustion does not maintain near the injector and only a shock wave can be found (marked as SW in Fig. 2d). A multi-wave structure can also be seen, connecting the induced shock wave (ISW), oblique shock wave (OSW) and the SW, as shown in Fig. 2(d). The ISW is induced by the high-speed injection of the air, which is also found in Ref. [20]. The gas between the ISW and the injector is featured by high speed, low temperature, and pressure. This therefore leads to slower propagation of the SW compared to the RDW. This also affects the heating, evaporation and movement of the liquid droplets in this area, which will be discussed in Fig. 3.
Thermochemical condition
Plotted in Fig. 3 are the distributions of dispersed fuel droplets in the refill zone, and they are colored by their y-component velocity, temperature, evaporation rate, and diameter. After being injected into the combustor, the velocities of n-heptane droplets near the injector are relatively high before the ISW, up to 1,000 m/s, and they gradually relax towards about 375 m/s. Meanwhile, as seen from Fig. 3(b), the droplets are quickly heated to their saturation temperature and start to vaporize. The droplet evaporation rate from Fig. 3(c) gradually increases (due to increased droplet temperature), peaks at 2 mm, and then decreases (due to decreased droplet mass) along the y-direction inside the refill zone. , when the detonation wave propagates steadily at 0 = 700 K and 1,000 K. In this work, is calculated from the ratio of required stoichiometric oxygen atoms to the available oxygen atoms [21], i.e., = 2( + /4)/ , where , , and denote the number of available carbon, hydrogen and oxygen atoms, respectively. Be reminded that since it is based on element conservation, is also well defined in the product gas. However, the ones in the undetonated mixtures (e.g., fuel refill zone) are relevant for our analysis. One can see from Fig. 4 that, very limited vapor (blue areas) exists near the injectors into the RDE chamber, and the starvation of fuel vapour leads to local detonation decoupling near the injector, as shown in Fig. 2(d).
As the air total temperature increases, more fuel vapor is present near the injector, through comparing Fig. 4(c). The average static temperature in the fuel refill zone at different 0 is also given below the x-axis in Fig. 4(c). Here we average respectively based on: (1) the refill zone (based on the criterion of local temperature lower than the corresponding 0 ) and (2) detonation wave front (HRR > 10 13 J/m 3 /s [22]). As 0 increases from 700 K to 950 K, the average static temperature of the gas in the refill zone increases from 581 K to 787 K, the average increases from 0.668 to 0.69. However, when 0 further rises from 950 K to 1,300 K, the average in the refill zone is almost constant, about 0.69, which is slightly higher than the lean flammability limit of n-heptane, 0.56 [23].
For the average at the RDW, when 0 < 950 K, a portion of liquid droplets cannot fully vaporize ahead of RDW, resulting in a relatively low at the RDW in Fig. 4(c). As 0 further increases, the remaining droplets after the RDW gradually decreases, and hence the equivalence ratio at the RDW is close to unity, which is the ER of the injected mixture in our simulations. Therefore, 0 = 950 K is a critical total temperature for the ERs for the fuel vapor availability at the detonation wave. If we linearly extrapolate the three points of (RDW) to 600 K, the corresponding at the RDW is about 0.62, which is near the lean flammability limit. This also justifies why a stable RDW cannot be achieved with 0 < 700 K under the simulated conditions. : average static temperature in the refill zone.
Chemical structure
Here the chemical reaction characteristics in the spray RDE will be extracted with the chemical explosive mode analysis (CEMA) [24]. The equation for a gas reaction system reads: where (•)⁄ is the material derivative, is the vector of temperature and species concentrations, is the chemical reaction, is the diffusion term, is the droplet evaporation term and is the Jacobian of the chemical system. A chemical explosive mode (CEM) is defined as the eigen-mode associated with a positive (real part) eigenvalue which indicates that the local mixture tends to ignite under lossless conditions [24]. Distributions of the CEM eigenvalue are shown in Fig. 5 for 0 = 1,000 K. For better illustration, the logarithmic expression of the eigenvalue is plotted, i.e., ≡ sign[Re( )] • log 10 [1 + |Re( )|]. Details of the CEMA can be found in [24]. Fig. 5(a) is that the heterogeneous two-phase mixture in the refill zone features large positive eigenvalues, indicating that fast droplet evaporation and vapor/air mixing turn the gas into chemically explosive state. Zero-crossing of the eigenvalue occurs at the detonation and deflagration surfaces. As shown in Fig. 5(b), ahead of the RDW, as the droplets are just sprayed into the chamber, the evaporation rate is low and the n-C7H16 vapor cannot mix effectively with the air, which results in overall fuel-lean composition (local ERs: 0-0.03) and the eigenvalue here is around zero. Near the deflagration surface shown in Fig. 5(c), some striped burned zones appear, with small negative eigenvalues (marked as dashed box). The reason for it will be further discussed in Fig. 6. For comparison, one additional simulation is run, in which the low-temperature elementary reactions, i.e., R105-R112 (Appendix A of [12]), are deactivated. Distribution of the CEM eigenvalue from this test is shown in Fig. 5(d). In the refill zone, is around 2, much lower than that of Fig. 5(a), i.e., about 4. This implies that the chemical timescale predicted with LTC included is approximately two orders of magnitude shorter than that without LTC. This clearly shows the LTC promotes the overall reactivity of the undetonated gaseous mixtures. The chemical reactions in liquid n-heptane RDE will be further analyzed through the profiles of the participation index (PI) and explosion index (EI) [24] across the refill zone and the deflagration surface at x = 45 mm (annotated as line 1 in Fig. 5a). High PI (EI) signifies the dominance of the corresponding elementary reaction (species) in the explosive mode. The results are shown in Fig. 6. The curve of along line 1 is also plotted in Fig. 6(a). As seen from Fig. 6(a), the gas reactivity is weak ( being very low, but still positive) near the injector, at y < 0.5 mm. This is because very limited droplet evaporation and the low gas temperature (about 520 K) in the previously mentioned small region before the ISW. Beyond y = 0.5 mm, rises quickly and then levels around 4.0. Besides, at y = 0-5.5 mm (i.e., the shaded area, with heterogeneous mixtures of fuel droplets, vapor, and air), the initiation reaction R104 (n-C7H16 + O2 → 2-C7H15 + HO2) is significant with the highest PI. After that, R107 (RO2 → R'O2H), R111 (OR''O2H → OR''O + OH) and R91 (1-C7H15 → 1-C5H11 + C2H4) become dominant. The first two reactions are low-temperature reactions and last one is a cracking reaction. For these locations, high EI of R'O2H is also observed, which is an intermediate species generated and consumed by the lowtemperature reactions, R107 and R108 (R'O2H + O2 → O2R'O2H), respectively.
Evident from
Beyond the droplet-laden area (i.e., y > 5.5 mm), since the droplet evaporation is completed, the local mixture is gaseous (air and n-heptane vapor). The PIs for the LTC (e.g., R107) decrease. Instead, the PIs of the following elementary reactions become comparatively high: R42 (CH3 + HO2 → CH3O + OH), R100 (n-C7H16 + OH → 2-C7H15 + H2O) and R108 become dominant. At y = 6.35 mm, first zerocrossing of the curve can be found, corresponding to a high temperature EI (indicative of thermal runaway). Burned mixture next to it features < 0. This corresponds to the first-stage ignition (i.e., green areas in Fig. 5c), which produces small radicals such as C3H6 or C2H2. Further downstream, the mixture becomes explosive again before it gets burned in the second-stage ignition near the reactantproduct contact surface, with the being much higher than that before the first-stage ignition. This can also be corroborated from the EI of temperature in Fig. 6(b), and is consistent with the findings for nheptane autoignition in Ref. [25]. Nonetheless, differently, our results indicate that thermal runaway are significant for both stages, which is because of sufficient radical runaway in both two-phase and gasonly mixtures in the refill zone. The n-heptane vapor ignition modes in the refill zone are further analyzed in Fig. 7. Here the ignition mode is identified by projecting ( )⁄ to the CEM using the left eigenvector [26]: where = · , = · and = · respectively represent the projected chemical, diffusion and evaporation terms. Note that here include both the thermal affect for the temperature and the kinetic effect for the vapor. The last term in Eq. (5) can be neglected, following Refs. [26]. The effects of species diffusion and droplet evaporation on gas reactions can be indicated respectively by the ratios of = / and = / . If > 1, ignition is facilitated by diffusion or droplet evaporation; if | | < 1, chemistry is dominant (hence autoignition); if < −1 , diffusion or evaporation dominates chemistry and inhibits ignition [26].
The distributions of and in the refill zone are shown in Fig. 7. It can be seen from Fig. 7(a) that is high (in red) near the spray injector in the entire refill zone, which can be clearly seen from the closeup view of zone A in Fig. 7(A). This indicates that the ignition of the local vapor is promoted by the species diffusion. This is understandable since in this region the equivalence ratio is still very low (see Fig. 4b), and efficient species diffusion (hence reactant mixing) is favorable for vapor ignition. Further downstream, e.g., y > 5 mm, the ignition mode changes to autoignition ( ≈ 0) and fuel diffusion plays a limited role. In these areas, the reactant composition is overall uniform, although reactant stratification still exists due to discrete distributions of the evaporating droplets. Near the deflagration surface (zone B), although diffusion is dominant, mixed local combustion modes are present in Fig. 7(B); the reactions at some pockets are inhibited by the species diffusion. Likewise, how droplet evaporation affects vapor reaction ahead of the rotating detonation wave can be examined through the distributions of . At positions near the injector, the droplets are just sprayed into the chamber and cannot evaporate quickly, which results in small , as shown in Figs. 7(b) and 7(C). As the droplet evaporation accelerates, its contribution towards the CEM becomes high (red spots in Fig. 7C) and is important for the entire droplet-laden refill zone. Nonetheless, some sparse blue dots are observed, which is probably due to the heat absorption by the evaporating droplets. Moving further downstream towards the deflagration front (Fig. 7D), the evaporation affect becomes negligible as the droplets are fully evaporated. Figure 8 shows the change of average detonation wave speed with different air total temperature 0 . Here is calculated from the pressure history at a probe near the head end of the domain. The Chapman-Jouguet speeds of gaseous n-heptane/air mixtures with = 1.0 are added. The liquid fuel RDW speeds are 3%−15% lower than the C-J values. This may be caused by, e.g., imperfect reactant mixing and droplet evaporation [5] [10]. It is also found that the RDW speed with liquid fuel increases when 0 is increased, which is also observed by Meng et al. [9]. This is because the increase of 0 raises the temperature in the fuel refill zone and increases the rate of mixing between reactants which and ultimately causes an increase in . This trend is different from the result of Ref. [7], because the droplets in [7] are finer (2 μm), which enables complete evaporation before the RDW arrives and therefore the equivalence ratio of the undetonated gas has weak dependence on 0 . In addition, the average speeds predicted without LTC are higher than those with LTC. This is because the low-temperature chemical reactions in the fuel refill zone consumes part of the gaseous n-C7H16. Fig. 9. Specific impulse and total pressure ratio versus total temperature. Figure 9 shows the effects of air total temperature on the total pressure ratio and specific impulse . The pressure ratio is = 3 / 0 , where 3 is the total pressure from near the exit (y = 49 mm) and 0 is air injection total pressure, i.e., 20 atm in this study. The specific impulse is calculated following Ref. [10]. One can see that the pressure ratio first increases and then decreases with increasing 0 . From 0 = 700 K to 950 K, an increase in 0 changes the effective equivalence ratio at the RDW (see Fig. 4), which causes a gradual increase in 3 and ultimately an increase in . From 1,000 K to 1,400 K, the average equivalence ratio at the RDW front no longer changes with increasing 0 . However, as 0 increases, chemical reactions accelerate, the deflagrative combustion intensifies and thus lowers the detonated fuel fraction and 3 , which eventually leads to a decrease in the pressure ratio (the detonated fuel fraction versus total temperature is shown in supplementary document for interested readers).
Carrier gas total temperature effects
Nonetheless, the specific impulse monotonically increases with 0 . This is because injection with higher air total temperature expands the area of the low-temperature region before the ISW mentioned in Section 3.1. The gas velocity in the mentioned region is high (more than 1,000 m/s). The increase of the area increases the gas velocity and eventually increases the specific impulse.
The results predicted without LTC are also shown. One can find that the without LTC is consistently underpredicted. This is because the deactivation of the LTC changes CEM eigenvalue (lower reactivity, see Fig. 5) in the refill zone. The change in the CEM eigenvalue results in a lower detonated fuel fraction in the detonation combustion (see supplementary document), thereby a lower pressure of the detonation wave and ultimately to a lower specific impulse and pressure ratio. When 0 > 1,300K, the propulsion indices are almost not affected by the LTC, because the LTC is inhibited for this temperature range.
Conclusion
Two-dimensional rotating detonations with liquid n-heptane sprays and preheated air are simulated with a Eulerian-Lagrangian method and a skeletal chemical mechanism. The results show that the mixture in the refill zone is heterogeneous, including evaporating droplets, vapor, and air. When the total temperature is below 950 K, the average equivalence ratio increases with the total temperature. The chemical structures in the refill zone and reaction front are studied with the chemical explosive mode analysis. It is seen that with fuel vapor addition and efficient mixing, the mixture becomes explosive in most of the refill zone. The initiation reaction (R104) and low-temperature reaction (R107) are dominant in the upstream and downstream of the refill zone, respectively. The LTC intermediate species, R'O2H, is found to be important for chemical explosive mode in the undetonated mixture. The influence of species diffusion and dispersed droplets on fuel vapour ignition is further analyzed. The detonation propagation speed and specific impulse increase with air total temperature. The total pressure ratio firstly increases and then slightly decreases. Inclusion of the LTC in the chemical mechanism would affect the predictions of these parameters, but the difference is minimized when the air total temperature is above 1,300K. | 2022-02-15T06:47:40.242Z | 2022-02-12T00:00:00.000 | {
"year": 2022,
"sha1": "36ce3bd49129ec2eb1ca218d5694b05346e1c055",
"oa_license": null,
"oa_url": null,
"oa_status": null,
"pdf_src": "Arxiv",
"pdf_hash": "36ce3bd49129ec2eb1ca218d5694b05346e1c055",
"s2fieldsofstudy": [
"Chemistry",
"Engineering"
],
"extfieldsofstudy": [
"Physics"
]
} |
86707754 | pes2o/s2orc | v3-fos-license | Specific boundaries between the causal agents of the soybean stem canker
Pathogens within the Diaporthe complex cause seed decay, stem blight and stem canker on soybean, representing a serious threat for this crop species. We herein utilize worldwide sequence data retrieved from Genbank in order to assess the species boundaries between the soybean stem canker causal agents, and define whether or not they should be regarded as members of the same biological species. These studies were complemented with compatibility tests, in order to validate our findings from a biological standpoint. Species delimitation assays supported the occurrence of a speciation event between D. caulivora and D. phaseolourm var. meridionalis. A speciation hypothesis between D. aspalathi and D. phaseolourm var. meridionalis was also supported, based on three reciprocally monophyletic substitutions at locus EF1-α. Compatibility tests further validated species delimitation assays indicating that D. caulivora has developed barriers to gene exchange with D. phaseolorum var. meridionalis. Clarification of the specific boundaries of the SSC pathogens and related entities will be an important asset to future research in soybean pathology, epidemiology and breeding.
INTRODUCTION
Diaporthe Nitschke, with over 800 specific names, constitutes the teleomorphic state of Phomopsis (Sacc.)Bubák, an anamorphic genus with more than 900 specific names recorded.An important number of species within this group has been reported as destructive pathogens causing cankers, diebacks, root rots, fruit rots, leaf spots, blights, decay and wilts on a wide range of plant hosts worldwide, including strategic crop species (Udayanga et al., 2011;2012).Fungi in the Diaporthe species complex constitute an economically relevant threat for the soybean production chain worldwide, with five taxa traditionally recognized: Diaporthe phaseolorum (Cooke & Ellis) Sacc., D. phaseolorum var.sojae (Lehman) Wehm., Phomopsis longicolla Hobbs, D. phaseolorum var.meridionalis F. A. Fernández, D. phaseolorum var.caulivora Athow & Caldwell.The latter two have been reported as the causal agents of the soybean stem canker (SSC).Santos et al. (2011) recently described Diaporthe novem Santos, Vrandecic & Phillips, as a sixth soybean pathogen.Before the arrival of soybean rust to the Americas Diaporthe pathogens were cited as causing more economic losses in soybean production than any other single fungal pathogen, and had been a major concern in South America since 1989 (Sinclair & Backman, 1989).During the 1994/1995 growing season, yield losses due to SSC reached US$ 170 million in Brazil (Yorinori, 1996).SSC was first detected in Argentina in 1996/97, and has since then caused up to 100% yield loss in some instances (Grijalba et al., 2011).
Asexual and sexual names of fungi have recently been granted equal status in the International Code of Nomenclature for algae, fungi, and plants.Therefore, the name Diaporthe has been adopted for this group of fungi, regardless of the spore stage involved (Santos et al., 2011;Crous et al., 2011;Udayanga et al., 2012;Gomes et al., 2013) since Diaporthe (1870) predates Phomopsis (1905).
The taxonomic history of Diaporthe species as soybean pathogens starts early in the 20th century when Diaporthe spp.isolates were obtained from a group of unrelated hosts, including Ipomoea batata L., Phaseolus lunatus L. and Glycine max (L.) Merr.Following the hostspecific hypothesis, those isolates were recognized as independent species and identified as Diaporthe batatas Harter & E. C. Field;D. phaseolorum (Cooke & Ellis) Sacc., and D. sojae Lehman, respectively (Morgan-Jones, 1985;1989;Backman et al., 1985).In contrast, Harter & Field (1912) and Harter (1917) proposed that these three pathogens constitute a single species, reassigning them as three varieties: D. phaseolorum var.phaseolorum, D. phaseolorum var.batatas, and D. phaseolorum var.sojae.In the early 1950's, D. phaseolorum var.caulivora was first described as the causal agent of the SSC and it was considered a perithecial variant from D. phaseolorum var.batatas (Crall, 1950), the causal agent of the dry root in sweet potato (Ipomea batata L.).Hobbs & Phillips (1985) proposed the differentiation of the US Northern and Southern stem cankers.Morgan Jones (1989) further split them into formae speciales, based on morphological and physiological differences, designating D. phaseolorum f. sp.meridionalis for the southern US teleomorphic isolates, and D. phaseolorum f. sp.caulivora for northern isolates.Fernández & Hanlin (1996), based on differences in the number and type of lesions shown by field-grown plants, readopted the concept of "variety".Since then, the accepted denomination has been D. phaseolorum var.caulivora and D. phaseolorum var.meridionalis.Based on nucleotide sequence data, cultural, phytopathological and morphological evidence, Rensburg et al. (2006) proposed that D. phaseolorum var.meridionalis should be treated at the species level along with the red bush die-back causal agent, Diaporthe aspalthi.Species identification in Diaporthe has been traditionally based on host specificity (Udayanga et al., 2011;2012).Few morphological characters can undoubtedly differentiate among taxa (Uecker, 1988).Identification of the SSC pathogens has relied on colony appearance, growth rate, size of stromata, arrangement and morphology of perithecia, presence of α and β conida, and detection of the anamorph phase (Morgan Jones, 1985;1989;Sinclair & Backman, 1989;Fernandez & Hanlin, 1996).The overlap shown by some of these quantitative features has led to several ambiguous identifications.Indeed, it is a well-known general fact, that morphological and phytopathological characters are affected by environmental factors and sampling, often leading to inaccurate species classification (Davis & Nixon, 1992;Padial et al., 2010;Grijalba & Ridao, 2012).In order to circumvent such limitations, the classification of Diaporthe species is presently being redefined to include DNA sequence data (Rehner & Uecker, 1994;Zhang et al., 1998;Mostert et al., 2001;Farr et al., 2002;Santos et al., 2010).
Nevertheless, methods for species delimitation using genealogical data typically rely upon genetic distances or gene trees (Sites & Marshall, 2003;2004).This analytical approach requires arbitrary decisions regarding the thresholds of the species boundary (Hey, 2009).In order to circumvent this problem Yang & Rannala (2010) developed a coalescent-based approach to delimitate closely related species using DNA sequence data.This methodology includes both intra and interspecific variation.This approach to species boundary delimitation has been validated with simulated datasets (Yang & Rannala, 2010;Zhang et al., 2011) and applied to empirical datasets of rotifers, lizards (Yang & Rannala, 2010), forest geckos (Leache & Fujita, 2010), butterflies (Zhang et al., 2011) and rice (Zang et al., 2011).
The study of somatic incompatibility reactions provides a useful criterion for spatial delimitation of fungal individuals, or at least for delimitation of genetically distinct mycelia.This criterion has been applied in several fungal groups including important plant-pathogenic fungi (Pál et al., 2007).It has been proposed that the incompatibility reaction may limit the spread of harmful cytoplasmic or nuclear elements (Caten, 1972), and prevent resource plundering (Debets & Griffiths, 1998).It has also been suggested that vegetative incompatibility may promote the initiation of sexual reproduction in some species as a result of non-self recognition (Dyer et al., 1992).Individuals that share the same heterokaryon or vegetative incompatibility loci can fuse to form a heterokaryon and are then considered to belong to the same vegetative compatible group (Glass et al., 2000).In contrast, fungal isolates that differ at one or more of these loci will not fuse.Instead, programmed cell death or apoptosis occurs in the mycelial cells that are in contact with an isolate representing different vegetative compatible groups (Leslie, 1993).Previous vegetative compatibility assays have been performed for the SSC pathogens, but involving solely two Brazilian isolates (Costamilan et al., 2008).
As previously mentioned, the taxonomic rank for both SSC causal agents has been upraised to the specific level.Nevertheless, genetic and biological boundaries between them have not been addressed so far; indeed, the causal agents of tSSC are in practice still treated as part of the same biological species by soybean breeders and pathologists.As a consequence, much of the research carried out elsewhere treats these pathogens as a single biological entity.We used molecular data retrieved from many different geographic origins in order to clearly assess whether the gene pools of D. aspalathi, D. phaseolorum var.meridionalis and D. caulivora are in fact isolated or not.To further biologically validate the molecular evidence, we implemented vegetative compatibility assays between distinct soybean isolates of D. phaseolorum var.meridionalis and D. caulivora.
Preliminary nucleotide sequence analyses and dataset assembly
Sequence data for soybean pathogen isolates originally identified as Diaporthe phaseolorum var.caulivora, Diaporthe caulivora, Diaporthe phaseolroum var.meridionalis and Diaporthe aspalathi were retrieved form Genebank.Sequence data publicly available for seven loci was included for molecular species delimitation assays (Table 1).Multiple sequence alignment for each locus was attempted using Clustal W (Thompson et al., 1994) and Muscle (Edgar, 2004), as implemented in Mega 5.0 (Tamura et al., 2007), with different parameter settings, and slight manual modifications when necessary.
Bayesian species delimitation assays
Species delimitation assays were performed using the program Bayesian Phylogenetics and Phylogeography (BPP) v. 2.0 (Rannala & Yang, 2003;Yang & Rannala, 2010).This program requires three input files, namely the sequence file (including multiple alignments for every loci under consideration), a species map file (indicating the putative species for each sequence) and a file including specific evolutionary parameters.This latter file is amenable to alternative tailoring in order to account for different evolutionary scenarios.Evolutionary parameters include a guide tree, as well as specification of prior distributions for the scaled ancestral population size (θ 0 ), and root age (τ 0 ).Priors are assigned a Gamma G (α, β) distribution, with a prior mean = α/β and prior variance = α/β 2 .This information is user-provided, and constitutes the starting point (priors) for the program.Prior distributions can affect the posterior probabilities for the different speciation models (topologies).According to coalescence theory, large values for θ 0 (big population numbers) and small values for τ 0 (shallow divergence times) favor conservative models containing fewer species (Leache & Fujita 2010; Yang & Rannala, 2010).In species delimitation, the guide phylogeny is also a most important prior affecting posterior probabilities for the speciation hypotheses (Leache & Fujita, 2010; Yang & Rannala, 2010;Zang et al., 2011).
BPP v. 2.0 uses a reversible-jump Markov chain Monte Carlo (rjMCMC) algorithm to jump back and forward over different topologies and estimate the posterior distributions of species delimitation models, starting from the guide tree.Every model should be compatible with the starting priors and the sequence alignment introduced in the input files.By default, BPP assumes no admixture following a speciogenic event.The JC69 mutation model (Jukes & Cantor, 1969) is assumed to accommodate multiple hits.The sequences are supposed to be close, so that JC69 is deemed adequate.Leache & Fujita (2010) proposed posterior probability values > 0.95 as strong support for a speciation event.
The guide tree herein proposed considers D. aspalathi, D. phaseolorum var.meridionalis and D. caulivora as three separate species (completely resolved tree).Considering the huge population numbers of fungal organisms, a gamma prior distribution G(1, 10) for the root population size (θ 0 ) was set for every assay.Provided that no information (eg.fossil record) is available indicating species history, three different gamma priors, namely G(1, 10), G(2, 200) and G(2, 2000), were attempted for τ 0 .These priors account for different divergence times from the root population.Each analysis was run at least twice, to confirm consistency between runs.Running the rjMCMC analyses for 500,000 generations (sampling interval of five) with a burn-in period of 10,000 produced consistent results across separate analyses initiated with different starting seeds.Convergence was considered as adequate only after the Estimated Sample Size (ESS) was above 300 for every node.
Somatic compatibilty tests
Vegetative compatibility was tested based on the formation of a barrage-zone.Six soybean fungal isolates were tested against each other.All isolates are housed at the Phytopathology Lab, School of Agronomy, University of Buenos Aires.Diaporthe phaseolorum var.meridionalis isolates were obtained at Asunción (Paraguay; Genbank accession number HQ130438, Dm1); Venado Tuerto (Santa Fe, Argentina; HQ130439, Dm2) and Pergamino (Buenos Aires, Argentina; HQ130440, Dm3).Diaporthe caulivora isolates were obtained at Trenque Lauquen (Southern Buenos Aires, Argentina, HM625758, Dc1), Urdampilleta (Western Buenos Aires, Argentina, HM625770, Dc2) and General Pirán (Buenos Aires, Argentina, HM625760, Dc7).Isolates were paired 2-3 cm apart on PDA (potato dextrose agar) in Petri dishes and incubated in darkness for a week at 20ºC and another week at 25ºC (Costamilan et al., 2008).Self-crosses were utilized as negative controls, representing no barrage formation.Each pairing was repeated twice.Hyphal interactions were recorded two weeks after the fungi were plated.The interaction zone and their boundaries were further observed under the microscope and photographed.
Sequence analysis
In the present study we included 162 sequences from 7 distinct nuclear loci, comprising a total of 76240 bp.ITS and EF1-α were the only genomic locations with sequences available for all taxa under study.
Bayesian species delimitation assays
Multilocus bayesian species delimitation assays, irrespective of time divergence assumptions, yielded posterior probabilities (pp) between 99 -100% for the completely resolved tree in every evolutionary scenario (Figure 1).In comparison, the two-species (considering D. aspalathi and D. ph.var.meridionalis as a single species) model displayed extremely low posterior probabilities under all prior combinations (pp <0.02 in all cases).
A speciation hypothesis between D. aspalathi and D. phaseolorum var.meridionalis was also strongly supported (pp= 0.99 -1.0, Figure 1) using the original three species guide tree, and under every time divergence assumption.In order to further explore this finding, a series of assays aimed at assessing the species boundaries between D. aspalathi and D. phaseolorum var.meridionalis were carried out, using solely those loci for which information was available for both taxa (ITS and EF1-α).A speciation hypothesis was once again favored (pp>0.99) in every evolutionary scenario.This speciation hypothesis is sustained by three reciprocally monophyletic substitutions between D. aspalathi and D. phaseolorum var.meridionalis at positions g.99G>T, g.161C>T and g.236 C>T of the EF1-α locus (Figure 2).The ITS region, on the other hand, was identical between D. aspalathi and D. phaseolorum var.meridionalis, as previously stated (Rensburg et al., 2006).
Somatic compatibility tests
The vegetative compatibility tests were performed between D. caulivora and D. phaseolorum var.meridionalis isolates (Figure 3).The presence of a distinctive barrage, or pigmented zone and a lytic gap along the contact zone was detected in every D. caulivora -D.phaseolorum var.meridionalis confrontation assayed, seven days after contact.Microscopically, this pigmented zone comprised of a combination of compartimentalized hyphal segments, vacuolated brown hyphae and empty cells, not observed in unpaired growing mycelia.Conversely, D. phaseolorum var.meridionalis -D.phaseolorum var.meridionalis and D. caulivora -D.caulivora confrontations merged uniformly with no dark line in the contact zone.
DISCUSSION
The multilocus species delimitation test herein assayed clearly indicates that no gene exchange occurs between D. caulivora and the D. aspalathi -D.phaseolorum var.meridionalis cluster.Incompatibility reactions in every D. caulivora -D.phaseolorum var.meridionalis confrontation further strongly validate and confirm the genetic isolation between both groups.These findings supports previous results from Santos et al. (2011) who raised D. caulivora to the specific level using isolates from Croatia, and Grijalba et al. (2011) and Guillin et al. (2011) who reached a similar conclusion for Argentinean isolates.
Our species delimitation assays also supported a speciation hypothesis between D. aspalathi and D. phaseolorum var.meridionalis.This result is somehow unexpected, since it contradicts previous claims by Smit & Knox-Davies (1989a;1989b) and Rensburg et al. (2006), who concluded that both taxa should be considered as part of the same species.These authors have mainly based their proposal on comparative morphology between the two taxa, and an ITS-based phylogenetic reconstruction including other Diaporthe species as well; no EF1-α sequences from D. phaseolorum var.meridionalis was available to them, and therefore they were not included in their combined ITS and EF1-α phylogenetic reconstruction.Therefore, although it is evident that these two taxa are very closely related, it is still not clear whether they are reproductively compatible.
In this regard, it should be emphasized that D. aspalathi has solely been obtained from the red bush, Aspalathus linearis (Burm.f.) R. Dahlgren in South Africa, whereas isolates identified as D. phaseolorum var.meridionalis have been obtained from soybean fields worldwide.The occurrence of three reciprocally monophyletic substitutions at EF1-α suggests that both taxa have been somehow isolated for a considerable period of time, relative to the population size at the founder event (most likely a host jump from soybean to red bush).The present results suggest that genetic divergence between these two groups might be currently taking place, based on ecological grounds (host specialization).Therefore, it cannot be excluded that D. aspalathi and D. phaseolorum var.meridionalis constitute cryptic species at present.Further analyses with a larger number of loci are warranted, in order to assess whether a host-jump based speciation event between these two taxa has already been accomplished.No compatibility or cross infection tests involving D. phaseolorum var.meridionalis and D. aspalathi isolates have been attempted so far to our knowledge in order to biologically validate contrasting hypotheses.Although D. phaseolorum var.meridionalis has been formally accepted as a synonym of D. aspalathi, in view of the present results it is likely that this will need to be further clarified in the future.
It has been stated that traditional morphological characters no longer clarify the taxonomy of Diaporthe at the specific level (Brayford, 1990;Rehner & Uecker, 1994;Crous, 2005).In this regard, Udayanga et al. (2011;2012) and Gomes (2013) proposed phylogenetic trees as platforms for future taxonomic classification within this species complex.Nevertheless, speciation is a continuous process (De Queiroz, 1998;2007) and this implies that delimiting species using genealogical data will necessarily be accompanied by some degree of uncertainty (Leache & Fujita, 2010).This is particularly so when dealing with closely related species.Very importantly also, multiple sequence alignment for a great number of species would most likely bring about ambiguously aligned regions that could greatly skew subsequent phylogenetic analyses (Morrison, 2009), since the characters (nucleotide positions) within will most likely be homoplasic.The number of taxa included not only affects multiple alignment, but also support (or probability) values, and eventually cluster resolution within the topology.This is why multi-species phylogenetic reconstructions shall only be considered as preliminary backbones for further fine-scale analysis such as species boundary delimitations within a particular group of organisms.
Our species delimitation study for the SSC causal agents reveals the potential of the coalescent-based approach for recognizing speciation events for problematic taxa, or groups for which traditional methodologies are not clear-cut due to experimental or historical reasons.This is, to our knowledge, the first attempt to using both infra and supra-specific data for species boundary assessment in plant pathogenic fungi.We propose that the inclusion of a coalescent-based methodology for species delimitation will greatly contribute to the resolution of Diaporthe species complex taxonomy.Additionally, this approach might be a great asset at establishing anamorph-teleomorph connections, an issue greatly lagging in Diaporthe, where only 20% of such links have been resolved so far (Udayanga et al., 2011).
Precise resolution of species boundaries will greatly contribute to optimizing downstream academic and applied studies.It is important to note that we herein adopt the traditional biological species concept (reproductive isolation amongst taxa) based on purely practical grounds: elucidation of the biological relationships amongst the SSC pathogens has implications for agricultural research.Should, for instance, D. aspalathi and D. phaseolorum var.meridionalis still share their gene pools, different hosts (soybean, red bush) might act as alternative sources of inocula; this should not be disregarded by producers and sanitary authorities.This could in turn contribute to the dissemination of a particular disease into new crops species and geographic areas.
Clarification of whether or not a given group of pathogens are reproductively isolated might be an indication of substantially different epidemiological conditions required by the individual taxa, as well as differential preconditions for breeding activities and strategies.In this regard, five loci have been so far described in soybean as conferring vertical resistance against D. phaseolorum var.meridionalis, whereas no major gene conferring resistance against D. caulivora has been described.Because of the cumbersome taxonomic history of the group, these five loci had paradoxically been named as "Rdc" (resistance against "D.caulivora").Pioli et al. (2003) suggested that these loci should be renamed as "Rdm" (Resistance against "D.phaseolorum var.meridionalis").According to the present results, Rdm gene stacking aimed at increasing resistance against D. caulivora, for instance, should not be considered as an appropriate breeding strategy, and this approach should not be favored within corporate or public breeding programs in the future.
FIGURE 1 -
FIGURE 1 -Multilocus Bayesian species delimitation results assuming a 3-species guide tree.The speciation probabilities are provided for each node and each combination of priors.Prior mean θ = 0.1 in all cases (big population numbers); this assumption results in lower speciation probabilities.Left, prior mean τ 0 = 0.1; middle, prior mean τ 0 = 0.01; right, prior mean τ 0 = 0.001.
FIGURE 2 -
FIGURE 2 -Sequence alignment of the variable sites for the EF1-α region defining 5 haplotypes out of 34 sequences originally assigned to Diaporthe caulivora, D. aspalathi and D. phaseolorum var.meridionalis.Each fragment spans 306 bases.Dots indicate similarity, while hyphens indicate gaps.Numbers above indicate the nucleotide position having haplotype AY339353 as a reference sequence during alignment.Nucleotide substitutions g.99G>T, g.161C>T and g.236 C>T define D. aspalathi and D. phaseolorum var.meridionalis as reciprocally monophyletic groups.
TABLE 1 -
Locus name, original GenBank denomination, host, country of origin and GenBank accession numbers of the sequences included in this study. | 2019-03-28T13:41:29.876Z | 2014-08-01T00:00:00.000 | {
"year": 2014,
"sha1": "cd84ee07975b59a972ecc019a7eee8ce5f281d12",
"oa_license": "CCBYNC",
"oa_url": "http://www.scielo.br/pdf/tpp/v39n4/v39n4a06.pdf",
"oa_status": "GOLD",
"pdf_src": "Anansi",
"pdf_hash": "cd84ee07975b59a972ecc019a7eee8ce5f281d12",
"s2fieldsofstudy": [
"Biology"
],
"extfieldsofstudy": [
"Biology"
]
} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.