Split (1)

train · 100k rows

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
102354878	pes2o/s2orc	v3-fos-license	Unsupervised Indoor Positioning System Based on Environmental Signatures Mobile sensors are widely used in indoor positioning in recent years, but most methods require cumbersome calibration for precise positioning results, thus the paper proposes a new unsupervised indoor positioning (UIP) without cumbersome calibration. UIP takes advantage of environment features in indoor environments, as some indoor locations have their signatures. UIP considers these signatures as the landmarks, and combines dead reckoning with them in a simultaneous localization and mapping (SLAM) frame to reduce positioning errors and convergence time. The test results prove that the system can achieve accurate indoor positioning, which highlights its prospect as an unconventional method of indoor positioning. Introduction In outdoor environments, the Global Navigation Satellite System (GNSS) can provide satisfactory positioning, but satellite signals are blocked by walls, causing no signal in the rooms. At the same time, as the indoor environments become more complex, and indoor activities become more frequent, the demand for indoor positioning is increasingly strong. In order to achieve better indoor positioning results, a lot of research has been carried out. Many papers focus on WiFi indoor positioning based on received signal strength, a common local positioning approach with a number of prominent advantages such as low cost and ease of deployment. Ref. [1] proposed an indoor positioning method based on a Monte Carlo algorithm. Ref. [2] proposed the improved indoor position estimation algorithm based on geo-magnetism intensity, which compensates for the flaw of particle filter. Ref. [3] also presented a novel method, and it proposed a low-cost and easy-to-realize positioning system, which uses FM Radio Signal Strength fingerprinting. Many of today's localization systems for indoor and outdoor positioning are based on propagation time measurements of radio signals. Ref. [4] proposed a new WiFi-based positioning method to make the access point (AP) signal features more complete. Ref. [5] proposed two indoor positioning algorithms based on channel state information (CSI). Ref. [6] proposed a new fingerprint-based WiFi indoor positioning system, which is realized by extracting and analyzing the individual multipath propagation delay. However, in order to achieve high positioning accuracy in the presence of non line of sight (NLOS) propagation, these systems require either an expensive manual calibration or additional information. Many paper introduced unsupervised positioning without these costs. Ref. [7] proposed a Wi-Fi radio maps (WRMs) calibration system that automates the initial construction and maintenance of radio maps using crowdsourced fingerprints collected from numerous smartphones without location information. The system incorporates an unsupervised learning algorithm into an incremental and adaptive calibration process. Ref. [8] proposed an unsupervised indoor localization scheme that uses the combination of smartphone sensors, iBeacons and Wi-Fi fingerprints for reliable and accurate indoor localization with zero labor cost. Ref. [9] proposed a novel approach for a channel impulse response (CIR) based fingerprint system, which reduces the calibration and measurement effort and simultaneously improves localization results. Inspired by this, the paper designed an unsupervised indoor positioning system to reduce development costs and improve positioning accuracy. Although there are many works about indoor positioning, there are still some problems: some indoor positioning technologies rely on customized installations, which reduces their scalability [10,11]. The positioning systems based on WiFi [12,13], while providing ubiquitous positioning, require cumbersome calibration works. Some systems can reduce the works [14][15][16], but most need to reduce positioning accuracy. The dead reckoning approach based on sensors of smartphones has begun to receive attention [17,18] in recent years, which uses an accelerometer to calculate the displacement of the user and a compass to learn the orientation of the user. Ref. [19] proposed a method of indoor navigation using a microelectromechanical systems (MEMS)-based strapdown inertial navigation system (INS) aided by Wi-Fi signal strength measurements. Ref. [20] proposed an algorithm to detect the stride using acceleration spectrogram feature by utilizing the accelerometer in a smartphone. The dead reckoning error is quickly accumulated, so the user position needs to be recalibrated. The paper proposes an unsupervised indoor positioning (UIP) system, using sensors on smartphones to detect unique indoor points, such a place where a unique group WiFi APs can be received, which can reset the errors in dead reckoning. Starting with the building layout generated manually or automatically [21][22][23], the system discovers the landmarks through the crowd-sensing method. The landmarks are then used to reset the errors in the dead reckoning, resulting in higher positioning accuracy. The test results prove that the system can achieve accurate indoor positioning, showing the prospect of UIP as an unconventional method of indoor positioning. The rest of this paper is organized as follows: Section 1 introduces the shortcomings of existing indoor positioning research and the research done in this paper. Section 2 introduces how the system uses sensor data to get the displacement and orientation of the user, and proposes a novel method by using landmark in the positioning process. Section 3 describes how to define landmarks by sensor data features and how to identify different landmarks in the experiment, and it is an important innovation in the paper. Section 4 introduces simultaneous localization and mapping and how it works in positioning systems. Section 5 tests the system UIP from different aspects, and the results prove that it has excellent positioning performance. Section 6 provides the conclusions. Related Work There are also two representative indoor positioning systems: the Horus system and the MaLoc system, and they are compared with the UIP system in the test. The Horus system identifies different causes for the wireless channel variations and addresses them to achieve its high accuracy. It uses location-clustering techniques to reduce the computational requirements of the algorithm. The Horus system works in two phases. The offline phase builds the radio map, clusters radio map locations, and conducts other preprocessing of the signal strength models. The online phase estimates the user location based on the received signal strength from each access point and the radio map prepared in the offline phase. The MaLoc is built on a proposed augmented particle filter. To minimize errors in motion estimation and improve the robustness of particle filter, it augments the particle filter with a dynamic step length estimation algorithm and a heuristic particle resampling algorithm. It uses a hybrid measurement model which combines a new magnetic fingerprinting model and the existing magnitude fingerprinting model to improve the system performance and avoid calibrating different smartphone magnetometers. This paper is inspired by these two methods, using a probability calculation method similar to the Horus system, and using the magnetic signal as the positioning technology, which is the method of the MaLoc system. At the same time, this paper has made improvements to the two systems. The main improvements are as follows: 1. The paper uses multi-domain sensors to make up for the single signal to be easily affected by the environment, and the positioning accuracy is insufficient. For example, Horus system is susceptible to noise in the channel. 2. The paper proposes global landmarks, which has not been seen in past positioning systems. Global landmarks enable the system to obtain certain environmental information in advance, which helps to improve positioning accuracy. 3. This paper makes use of the simultaneous localization and mapping (SLAM) architecture and framework that leverages smart phone sensors to both dead-reckon the user location and identify semantic landmarks. These landmarks are used in a SLAM probabilistic framework to reset the accumulated localization error. It makes the UIP system have high positioning accuracy without initial manual calibration. Dead Reckoning Dead reckoning is the method to estimate the track of the users and it needs to sample the user's motion data (e.g., acceleration). The algorithm can calculate the possible displacement and direction of the user from these data, and finally calculate the possible position of the user at the next sampling. In the mobile device positioning system, accelerometer and compass are used to get acceleration and orientation, then process and calculate these data to get the possible position of the user at the next time. Starting from a point which is known, the user's position at time t is updated by the control variable u t = {l t , φ t }, where l t is the displacement and φ t is the orientation change. (1) Displacement: It can be calculated simply by the direct integral of the accelerometer reading which can be obtained by acceleratometer. However, it has a bad positioning error. As shown in Figure 1, after the actual displacement of 30 m, the deviation of the estimated value exceeds 100 m; this is due to the noise and low sampling rate of the accelerometer and the jitter of the phone when the user is walking. The paper uses another effective method. It is identifying the feature of the user's walking [18,24]. The feature comes from the natural upward/downward rebound of the body at every step. To capture it, the accelerometer signal is processed by a low-pass filter, and the system identifies two consecutive local minimum, if the difference value between maximum and minimum is larger than the given threshold, it will be considered as a step of the user, and the displacement is calculated by multiplying the steps by the step size. (2) Orientation: Traditional approaches rely on the compass, which can indicate the orientation by the magnetic field. However, it is noisy and seriously impacts positioning results. Gyroscope is also used in the paper, and it can measure relative palstance, which is a three-dimensional rotation matrix, and the relative angular displacement (RAD) can be obtained when the matrix is multiplied by a time interval. The path structure of the user can be tracked by the gyroscope; however, the estimated path has an error in the initial orientation. As shown in Figure 2a, all of these paths are rotated versions of the real path. Landmarks can correct the deviation. As shown in Figure 2b, the position X 1 of the user at time t 1 is known, dead reckoning with initial bias θ estimates that the position is X 3 at time t 2 , and the real position is X 2 . Assuming that the user encounters the landmark L 2 at time t 2 and the location of L 2 is X 2 , which is known, the algorithm thus considers that the the location of the user is X 2 , which is an accuracy position. Landmark Density Indoors According to observations, signals in indoor environments are abundant, such as sound, light, and magnetic field. In addition, some building structures force the user to move in specific ways, which might create some unique patterns on the sensors. Some signatures may appear after analyzing different sensor signals. WiFi landmark: The indoor environments have many WiFi areas where all locations can receive a different set of WiFi APs, and they are in different sizes. Figure 3 shows the size distribution of WiFi AP coverage areas. If the phone can hear a group of WiFi APs in a small area, it can be considered to be in the area. The figure measures the CDF of WiFi coverage area, for example, if x is 5, y is 0.2, which means there are 20% WiFi APs, and every one of them covers an area not more than 5 m 2 . From the figure, we can know the approximate proportion of WiFi APs that are suitable as landmarks. If the area is small, the positioning error is small. There are eight and five WiFi landmarks in small areas found on the two floors of the laboratory building, and each size is less than 4 m 2 . Magnetic/accelerometer landmark: In order to search for signatures in the magnetic sensor/accelerometer, the paper performs K-means clustering on their measurements. The members of every cluster are mapped to the corresponding physical location. The member locations in most clusters are widely dispersed in space. Fortunately, the members of some clusters are spatially tightly coupled and can be used as landmarks. Different landmarks will be combined to improve positioning accuracy and the specific method will be explained in the next section. Location Estimate Suppose the system combines the different sensor data and uses k-means clustering, and discovers three sensor signatures that can be used as landmarks, as shown in Figure 4a. Considering that dead reckoning has random error, the paper calculates position of the landmark by combining multiple dead reckoning estimates of a landmark. Intuitively, estimation errors are random and independent due to noises on the sensor and human step size. By combining these estimations with errors, it is expected that the mean will converge to a more accurate location. Figure 4b illustrates the feasibility of this method by simple calculation of centroid, where the circle represents the center of dead reckoning estimate-each one has a large error but multiple estimates with errors yield a better estimate. In order to avoid positioning errors due to approximation landmark patterns, landmarks should be in small areas and they will be identified combining with the WiFi signal. The specific method is described in the next section. The Effect of Regular Error Compensation This section is added to verify the effect of landmarks in actual positioning, and the landmarks that are here are not the ones used in the system, but some simple alternatives. The way to get them is also very simple. The user walks naturally in the building with a smartphone from the entrance of the floor, and the system records readings in the inertial sensors and extracts the tuple <time, displacement, direction>. The doors and windows are marked with different numbers, and the user enters them into the system when passing through these places. These numbers are the alternatives of actual landmarks. Since the mapping is known, the actual position of the user can be known, and the effect of the landmark for the estimation error is simulated in this way. The experiment collected 10 traces starting from the entrance. As shown in Figure 5, when pure dead-reckoning without landmark is used (black curve), the error is accumulated rapidly. The error can be corrected periodically by using the landmarks (blue curve). Although the performance is improved, the average positioning error is still about 11 m. Thus, the experiment processes the gyroscope readings to calculate the angular change during walking. Note that gyroscope readings are relative and need to be combined with the compass to infer the absolute orientation. Once the gyroscope and compass are combined (red curve), the error is significantly reduced. Adding landmark calibration on the basis (pink curve), the average error is further reduced. Figure 5 shows the effect of the landmarks, and it is necessary to further design the landmark system. Landmark Design This section details the landmarks of the UIP system. Landmarks refer to signals or combinations of signals in an indoor environment that can be captured by sensors and whose sensor patterns are unique in the environment so that the system can determine the location of the sensor based on landmarks. According to whether the locations can be known in advance, the landmarks are divided into global landmarks and local landmarks. Local Landmark Local landmarks (LLMs) are similar to designs in most unsupervised positioning systems. We obtain the floor graph at first and the system can learn the structure of the building and the coordinates of every point in the building. It is necessary for the radio map. In the offline phase, the user walks naturally along the corridor in the building. Since there are various signals in the indoor environment (here mainly WiFi and magnetic signals), the sensors of the mobile phone capture these signals during walking. The application on the phone can grab the sensor data and transfer it to the server for processing. Defining the signals with high levels of recognition on sensor pattern. The sensor patterns of LLMs cannot be obtained in prior and they are local features of the indoor environment, e.g., WiFi distribution in a certain area, and the phone cannot hear the same WiFi set when it leaves the area. Due to the complexity and diversity of indoor signals, it is important to select an effective signal from a number of signals as a local landmark. In the paper, we choose WiFi and magnetic signal as landmarks because they are ubiquitous in indoor environments and the sensors of the mobile phone can easily obtain these signals. The tasks of finding local landmarks need to (1) distinguish different patterns from the sensed signals, and (2) test if the pattern is limited to a small area spatially. Figure 6 shows the operational flow, and the sensor data are collected by crowd sensing. The process can be shown in matrix form: the element <m,n> is sensor readings of the phone m at time n. The features of magnetic/inertial sensors include average, maximum, minimum, variance and average crossover, and they are media access control (MAC) ID and received signal strength indicator (RSSI) for WiFi. The collected features are normalized, and then the clustering algorithm is performed for every sensing dimension and the combination of different signals (e.g., an accelerometer and a compass) to find the clusters having low similarity to other clusters. Given a cluster, calculate the correlation between it and other clusters. If every value is not larger than the threshold, the cluster will be a candidate. Then, test the area of the cluster. Test if all members of the cluster are in the same WiFi coverage area at first. If they do, calculate the location of each cluster member. If all members are in a small area (about 4 m 2 , the same below), the cluster is considered as a landmark. (1) WiFi landmark: The MAC ID and RSSI are used as features. K-means clustering is applied to identify small areas with low similarity to all locations outside the regions. Assuming that a phone can receive WiFi AP sets A 1 and A 2 at locations l 1 and l 2 , respectively, and the similarity S between l 1 and l 2 is where is the RSSI of AP received by the phone at l i , a ∈ A. The basic principle of the above equation is that, when the AP signal is equally strong at two locations, the similarity of the two locations is proportionally increased, and vice versa. S ∈ [0, 1]. The paper chooses 0.4 as the threshold. Figure 7 shows the results of the test in a building in the university, and 0.4 is the point that balances the quality and quantity of WiFi landmarks. (2) Magnetic/inertial landmarks: There are some turns in the indoor environment. Since gyroscopes provide reliable angular displacement, they have the opportunity to be used as landmarks. The feature called the bending factor is designed in the paper, which captures the concept of path curvature. Similar turns in the same WiFi area will be in the same cluster, and the algorithm will also check whether the cluster is limited to a small area when this coefficient and WiFi are integrated as a feature. Magnetic landmarks are defined by the similar process, and the magnetic field somewhere needs to be significantly different from other places and within a small area. Then, it can be considered as a landmark. Global Landmark Global landmarks (GLMs) is a novel concept, and the idea comes from thinking about the user's own movement state: since the data from the surrounding environment can be used to calculate the coordinates of the user, the user's own motion characteristics may also reflect the characteristics of the environment, and this data can be used as a new landmark. We naturally divide the user's motion state into elevators, stairs, escalators, walking and static state, and these motion data can be obtained by inertial sensors. Their signal characteristics are usually unaffected by the environment, such as the user's movement in the elevator in any building is almost the same. On the other hand, these movements often correspond to certain objects, such as elevators, and we can immediately know where they are through floor plans. These two features make them well suited for indoor positioning. The collection process of GLMs data is the same as LLMs, and it is necessary to identify different kinds of GLMs by sensor patterns. Every one has its own pattern on magnetic sensor or inertial sensor. The following content will analyze the patterns of different GLMs. Elevator: The typical process includes normal walking time, waiting for a period of time, walking into the elevator, overweight/weightlessness, then a stationary time period, the other overweight/weightlessness and finally leaving. As shown in Figure 8, the accelerometer exhibits distinct features. To identify elevator motion patterns, a finite state machine (FSM) is used, which relies on observed state transitions, and different thresholds are set for transition. Escalator: After the elevator is identified, it is effortless to distinguish the escalator and the static from the walking and the stairs by the variance of the acceleration. The next step is to further distinguish the escalator from the static state, and the magnetic field's variance can be used to identify them since the escalator is driven by the motor and the motor affects the surrounding magnetic field. As shown in Figure 9, the waveform of the static state is more stable. Stairs: Finally, the stairs and walking need to be distinguished. When using stairs, gravity will have an obvious effect on the speed increase or decrease of the user. Compared to walking, there is higher correlation between the acceleration in the orientations of gravity and motion when users use stairs. Landmark database is built in the offline phase, and each one has its coordinates in the radio map. The dataset of GLMs is easy to build because their location is known in advance. If we want to add a GLM (e.g., elevator) to the map, we only need to collect the required signal near each elevator and then mark it in the radio map according to the location of the elevator in the building. In contrast, the collection of landmark information requires a man to continuously walk in the experimental environment, sample various types of signals (such as user's acceleration, environmental WiFi signal) at a fixed frequency, and upload them to the server to choose suitable ones as landmarks according to the above requirements. Simultaneous Localization and Mapping In an unknown environment, the user collects information from the surrounding environment (landmarks) while walking, determines the position of the surrounding by the landmark data, and corrects the coordinates of the landmark by its own position. This mode is consistent with SLAM, so the previous dead reckon and landmark system are integrated into the unified SLAM framework. This section will introduce corresponding background and the SLAM algorithm in the paper. EKF-SLAM SLAM is a concept for mobile robots [25] and the idea is used for human positioning in the paper. The system estimates the map (Θ) as well as the user's pose (position (x t , y t ) and orientation (o t )) at time t on a probabilistic framework: where u t = u 1 , . . . , u t is the control variable update history, u t is the control variable (displacement and orientation change) obtained by the sensors at time t − 1. z t = z 1 , . . . , z t is the user's observation history of the environment. n t = n 1 , . . . , n t are data association variables, and n t is the landmark's identity observed at time t. Extended Kalman filter (EKF) is a common method [26,27] which uses a high dimensional Gaussian density distribution. It decomposes P(s t , θ\|u t , z t , n t ) into two independent models as Equations (3) and (4), where θ n t is the location of the landmark n t , p(z t \|s t , Θ, n t ) = g(s t , θ n t ) + ε t , where h and g are nonlinear functions, δ t and ε t are Gaussian noise variables whose covariances are R t and P t , respectively. The method has two limitations, the computational complexity, which is the quadratic of the landmark number, and the data association problem, i.e., how to recognize the detected landmarks when signatures of two or more landmarks are similar. FastSLAM FastSLAM combines particle filters (PF) [28,29] and EKF, and it takes advantage of the structural properties of the SLAM problem: given the motion path, the landmark estimations are independent conditionally, i.e., the correlation of the uncertainty of the landmark estimation depends only on the uncertainty of the path. When the path is known, the estimation errors of different landmarks are independent of each other. Given the number N of landmarks, the path s t = (s 1 , . . . , s t ) is estimated as Equation (5): ap(s t , Θ\|z t , u t , n t ) = p(s t \|z t , u t , n t ) N ∏ n=1 p(θ n \|s t , z t , n t ). (5) Since the path of the user is unknown, FastSLAM estimates the path s t through the particle filter, and it uses particles to represent possible paths where each probability density function p(θ n \|s t , z t , n t ) is estimated using EKF. Based on the particles, each map errors are independent conditionally. The posterior (S [m] t ) of the mth particle contains the path s t, [m] and N landmark estimates, which is represented by the observed landmark typef n,t , mean µ n,t , as shown in Equation (6), Compared to EKF-SLAM, the computational complexity of FastSLAM is logarithmically related to the number of landmarks. In addition, data association decisions can be performed on each particle, thus maintaining multiple posteriors of data associations, which is more robust [30]. FastSLAM can also handle nonlinear models, and it has been shown to converge under certain assumptions [31]. The UIP system will use the SLAM algorithm based on FastSLAM. SLAM Algorithm Like other SLAM algorithms, the algorithm involves sampling, map update, resampling and assuming there is only one landmark observed at a time to keep the algorithm generic. Pose Estimation Given the control variableû t from the dead reckoning, the measure z t at time t and the pose s t−1 , the pose of the particle [m] at time t is sampled as shown in Equation (7). All data can be obtained from the previous section, dead reckoning provides the user's motion status, while sensor data also provides landmark observations. s [m] t ∼ P(s t \|s t−1,[m] , u t , z t , n t ). Equation (7) can be expressed as a product of the pose distribution and the probability of measure z t , where P t is the covariance matrix of the control data at time t and R t is the measure covariance at time t. Given the last pose s t−1 and the control data u t = {l t ,φ t }, including estimated displacementl t and orientation changeφ t , the pose at time t can be predicted aŝ t−1 , u t ). (9) Assuming that the errors satisfy Gaussian distribution, l [m] t and φ [m] t satisfy following probability distributions: where σ l and σ φ are the variances of the estimation errors. Equation (9) can be rewritten as follows: Equations (10)- (14) are concrete implementations of the nonlinear function h in Equation (3). The probability of measure z t involves the integration over all possible landmark locations θ n t , and it is impossible generally. To solve this problem, the algorithm approximates g in Equation (3) to a linear function, resulting in the function in Equation (15). where µ [m] n t ,t−1 is the mean of the estimated position of landmark n t at time t, and θ [m] n t ,t−1 is the estimated landmark location. G θ and G s are the Jacobian of g about θ and s. Therefore, Equation (7) obeys the Gaussian distribution with the following parameters: where Q n t ,t−1 G T θ + R t is the covariance matrix of landmark observation, and z t is the actual observation of landmark location. It is known from Equation (18) that the algorithm measures the distance between the user and the landmark. Map Update The map update is updating the location of the current detected landmark. The algorithm updates the estimate of the observed landmark according to Equation (19), Recall that g is linearized to retain Gaussian distribution of the posterior. This leads to the following update Equations (20)- (22). The derivation of them is equal to that of the EKF measure update: n t ,t−1 . The algorithm inherits the properties of FastSLAM and there are many possible landmark hypotheses, each of which corresponds to an individual particle which has its own judgment about the landmark, and this kind of multi-estimation attribute offers SLAM robustness to errors. Resampling As the particles generated may not match the desired posterior, it is necessary to resample to remove the particles with great errors and put more in the right area. The reason that some particles generated do not yet match the desired posterior is η in Equation (19), which may be different for different particle particle. The normalizer is the inverse of the probability of the measurement: To account for this mismatch, the algorithm resamples in proportion to the following importance factor: n t ,t−1 ) The distribution in Equation (23) can be approximated as a Gaussian distribution by linearizing g with the meanẑ t and covariance G s P t G T s n t ,t−1 G T θ + R t . In addition, the algorithm selects the data association n t satisfying Equation (24): The distribution in Equation (24) is calculated as follows: It can be seen that linearization of g generate a Gaussian distribution over z t with mean t , both of which are functions of the data assocaition variable n t . Experiment Introduction This section evaluates the performance of UIP. Installing a UIP application on the phone and granting the required permissions so that it can grab sensor data from the phone, then send them to the server over the network. The system is implemented on an Android phone. When users walk, the sensors on the phone will get his pose information. The data is sampled (gyroscope at highest permissible rate; other sensors at 24 Hz) and then they are sent to the server for processing. The server side code is written using C and MATLAB (MATLAB2017a, MathWorks, Natick, MA, USA), and all algorithms run on the server. When a new landmark is detected, the server will update the landmark list. Physics laboratory building is selected as the experimental platform, and all landmarks are decomposed into: eight turns, nine magnetic, 15 WiFi landmarks and three global landmarks. In the experiment, three users move around the building to collect data, and each user can use the landmarks detected by the previous users. Figure 10 shows the plan of the experimental area and some landmarks within it. Everyone walks along the corridor from the entrance to the floor. Figure 11 shows the false positive and false negative rates of different GLMs, and it can be seen that some landmarks are easier to detect than other landmarks as the unique patterns. Most values are zero in Figure 11, and the overall false positive rate and false negative rate are about 0.002 and 0.011, respectively, which means that the detection of GLMs is accurate. Figure 12a,b show the location accuracy of the calculated landmark and the number of the observed landmarks change over time respectively. As more users explore the environment, the number of landmarks will increase over time, and the accuracy of the landmarks will also increase as different paths lead to different independent estimates. In about 2 h, the average positional accuracy can converge to less than 1 m. Local Landmark Factually, data in the experiment is limited in the diversity of paths for the limitation of motion areas, and positioning accuracy may increase due to increased path diversity in practical applications. It can be seen from Figure 12c that the false positive rate detected by the three types of landmarks in the experiment is less than 1%, indicating that the matching accuracies are quite high. Similar to GLM, different LLMs have different detection accuracy due to different patterns, but each landmark has an accuracy of more than 60%. Landmark Quantity Although there are a large number of landmarks in the experimental platform, it is difficult to learn how many landmarks exist in other environments, and whether there will be a large impact on positioning if there are fewer landmarks in some buildings. Figure 13 shows the positioning errors of different numbers of landmarks, and even though there are only 10 landmarks, the average positioning error is less than 2 m. In addition, landmarks based on more dimensions (e.g., acoustic signal) can be selected if needed. U s i n g 2 l a n d m a r k s U s i n g 1 0 l a n d m a r k s U s i n g 1 8 l a n d m a r k s U s i n g 2 5 l a n d m a r k s (1) Particle number: The comparison result is shown in Figure 14. It can be seen that the positioning accuracy of 50 particles is greatly improved compared with that of 25 particles; however, if the number of particles continues to increase, the positioning accuracy will not be significantly improved, and the amount of calculation will increase greatly. Therefore, the performance is saturated with about 50 particles where the median accuracy is about 0.53 m, and the number is used in subsequent experiments. (2) Global Landmark: Figure 15 compares the CDF of positioning error with/without GLMs. In Figure 15, the average error still can be less than 1.5 m without GLMs, and in the case of using global landmarks, the performance is greatly improved, CDF within one meter increases by about two times, which is 0.8; the value is over 0.9 within 2 m, which increased by 0.1, indicating that the GLM is a successful idea. (3) Error over time: The Euclidean distance error of the system is calculated to evaluate the positioning accuracy, and Figure 16 shows the variation of the error over time, which is undulating. Due to noise in the sensor readings, the error increases as the user walks, and it will decline when a landmark is observed. As more landmarks are observed, map accuracy increases and user positioning errors decrease, hence the overall trend of the curve is decreasing. The average positioning error does not exceed one meter, which is enough to satisfy most indoor positioning tasks. (4) Online and offline: Figure 17 shows the error distributions in two patterns. In offline patterns, the system learns its error when a user encounters a landmark, so it can track and partially correct past traces, which is not available in online mode. The figure shows that offline system can observably improve the smearing of distribution in online positioning, and the median distance error is slightly enhanced too. Moreover, although the positioning accuracy of the online mode is not as good as offline mode, its error is basically controlled within 1 m. SLAM Framework In order to evaluate the SLAM framework in Chapter 4, the systems with/without the SLAM framework are compared in convergence time and positioning accuracy. Uncertainty of landmarks or user locations is considered in the SLAM system, which is not available in the non-SLAM system. (1) Convergence time: It is the time from the beginning to the point where the positioning accuracy tends to be stable. The length of convergence time affects the positioning efficiency of the system. Figure 18 shows that the accuracy of the SLAM system converges quickly within 1 h and eventually achieves an average accuracy of approximately 0.83 m, and it is better than the non-SLAM system, which converges to an average accuracy of about 1 meter after 2 h. (2) Error distribution: Figure 19 compares the error distribution of the systems with/without SLAM. It can be seen that, after using the SLAM framework, the positioning accuracy is improved significantly, and the median positioning error goes down by about one meter. In the SLAM system, CDF of the positioning error within one meter is about 0.8, while it is about 0.3 in the non-SLAM system. The error of system with SLAM is basically controlled within two m, it can significantly improve the accuracy. As can been seen from Figures 18 and 19, the SLAM system can obtain a high positioning accuracy in a short period of time. Other Systems The UIP system is compared with two classical systems: the WiFi fingerprint identification system [32] and the magnetic fingerprint recognition system [33]. Figure 20 shows that the UIP has a smaller median positioning error which is about 0.5 m, while it is about 0.7 m in the WiFi system and it exceeds one meter in the magnetic system. In addition, CDF of the localization error in UIP exceeds 0.8 within one meter and 0.9 within 1.5 m, which has an obvious advantage over other systems. However, they have better worst-case errors due to the initial manual calibration, and the distribution has a shorter tail. Conclusions In order to avoid the complicated calibration of indoor positioning, an unsupervised indoor positioning (UIP) system is proposed in the paper, which uses landmarks to compensate for the error of dead reckoning. The dead reckoning and landmark observation data are input into the SLAM algorithm to realize online positioning of users.This paper validates the feasibility of the idea of landmarks and provides detailed descriptions of the landmarks and the SLAM framework. The experiments are carried out in many aspects, showing that UIP can find different landmarks accurately, and the probability of error identification is less than 1%. In addition, the positioning error can converge quickly, eventually achieving a median positioning error of 0.53 m, which is better in the offline mode. Compared with the two classical algorithms, UIP also has obvious advantages in positioning accuracy. Future works will be expanding the system, including coordinated positioning of multiple people, using other sensors such as acoustic sensors and optical sensors.	2019-04-09T13:04:14.236Z	2019-03-01T00:00:00.000	{ "year": 2019, "sha1": "a3a6b0d74de2ae8e128e4f3ce69103a5dc9f817f", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/1099-4300/21/3/327/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "bab3cae8b98284390024c381bd7d93507be2b4e5", "s2fieldsofstudy": [ "Environmental Science", "Engineering", "Computer Science" ], "extfieldsofstudy": [ "Computer Science", "Mathematics", "Medicine" ] }
252231290	pes2o/s2orc	v3-fos-license	Jurnal Pertahanan As Russia launched an invasion in every domain of warfare, including cyberspace, Ukraine employed an Information Technology (IT) army equipped with software called db1000n on March 2022. This article seeks to explain how the use of db1000n helps Ukraine against the 2022 Russian invasion. Using a case study as the research method and Susan Martin’s military utility concept as the analytical framework, this article found that db1000n helps to support Ukraine’s cyber warfare effort by being Distributed Denial of Service (DDoS) attack software that possesses three types of military utility, namely technical utility, tactical utility, and strategic utility. Nevertheless, db1000n still has some limitations that cannot secure an easy victory in cyber warfare INTRODUCTION On February 24, 2022, Russia launched an invasion of Ukraine in every domain of warfare including cyberspace. Russian hackers had initiated attacks on many Ukrainian websites since the start of the invasion, making governments' web pages and online banking services could not be accessed by the public (Tidy, 2022). Ukraine's vice prime minister and minister for digital transformation responded to Russian attacks by creating an information technology (IT) army that comprised cyber security specialists on February 22, 2022(Fedorov, 2022. This army's main task is to conduct cyberspace military operations against Russia's information infrastructure. On March 19, 2022, the Ukrainian IT army adopted a software named "death by 1000 needles" (db1000n) as a tool to support Ukraine's cyber warfare effort (IT Army of Ukraine, 2022c). Db1000n is a recent case on military technology. Military technology has been an important part of warfare and winning it. Although there were scholars who argue that the morale of the troops is more important, the contribution of military technology to achieving victory in the battle cannot be simply dismissed (Cohen, 2018). Many military technologies have transformed the domain of warfare as well, for instance the invention of tanks for land warfare, the development of submarines for naval warfare, and the innovation of planes for air warfare (Roland, 2016). Recent development in the information and communication sector opens up the possibility to conduct cyber warfare. Db1000n is also a recent case of cyber warfare. There is no single definition to describe cyber warfare. Choucri (2012) defined cyber warfare as a range of offensive and defensive military activities in cyberspace to protect the states and its interest. On the other hand, Green (2015) stated that cyber warfare is an extension of policy taken by state or state-sponsored actors' in cyberspace as a response to foreign threats or to pose threats to another state. This article will refer to the cyber warfare definition by Lucas (2016) which states that cyber warfare is an attack carried out by state or non-state actors in cyberspace that causes substantial devastation to people and objects in the physical domain. Because db1000n software is a recent case on military technology and cyber warfare, this article proposes one research question, how does the use of db1000n help Ukraine against the Russian invasion? With this question, this article seeks to explain the use of db1000n by Ukraine in terms of military utility. Several studies have been conducted on military technology, cyber warfare, and military utility in the last five years. Siroli (2018) discussed the dual use of hardware and software in the market for cyber warfare at the tactical, operational, and strategic levels. Meanwhile, Mori (2018) Thornton & Miron (2020) examined the Russian military's use of artificial intelligence for autonomous propaganda, espionage, and destruction during cyber warfare. Last, Shandler, Gross, & Canetti (2021) surveyed the public perceptions in the United States, United Kingdom, and Israel towards the use of digital technology in cyber warfare. This article will proceed as follows to answer the research question. First, this article presents the case study method and military utility concept used to examine db1000n in the 2022 Russian invasion. Then, this article analyzes data about the use of Ukraine's db1000n by operationalizing the military utility concept from Susan (Martin, 2016). Finally, this article concludes the findings and discussions on db1000n's military utility. METHODS The methods used in this article are qualitative in the form of a case study. This article fulfills three requirements to use a case study by Robert, namely research question in form of "how" or "why", the research does not require control over behavioral events, and the research focuses on contemporary events (Yin, 2017). The data collection techniques used in this article are internet-based studies and literature-based studies. With that data collection techniques, this study used the following data sources: 1. IT Army of Ukraine's Telegram channel 2. Db1000n GitHub repository 3. Office of the President of Ukraine website 4. Books or book chapters 5. Journal articles This article operationalizes the following steps to analyze the collected data: 1. Setting an analytical framework from the concept 2. Categorizing the collected data based on the analytical framework 3. Creating a description or visualization of the collected data 4. Reflecting the coherency of the description or visualization with the analytical framework 5. Repeating step two, three, and four until getting the answer for the research question. Those five steps were adapted from the data analysis technique for the case study suggested by (Yin, 2017). RESULT AND DISCUSSION Military Utility According to Chapman, Elbahtimy, & Martin (2018), military utility is an important concept in the study of international relations but is still rarely discussed. One of the scholars who keenly discusses the military utility concept is Susan Martin. Martin has been examining the military utility in various publications, namely the utility of biological weapons for deterrence (Martin, 2002); the utility of chemical, biological, and nuclear weapons in war (Martin, 2004); the constraint on the utility of nuclear weapons in international politics (Martin, 2013); and the military utility of chemical agents in Vietnam War (Martin, 2016). Martin (2013) drew upon the structural realism approach to conceptualize military utility. Structural realism has three main ideas as follows (1) the structure of the international system is anarchy; (2) anarchy leads states to the self-help logic; and (3) self-help logic compels states to maximize their security or power (Dunne & Schmidt, 2019). Those three main ideas differentiate structural realism from the main ideas of classical realism which are the egoistic nature of humans determining international politics and national interest defined in terms of power (Dunne & Schmidt, 2019). According to Dunne & Schmidt (2019), both structural and classical realism still shared the same assumptions regarding statism, self-help, and survival Martin (2016) argued that an anarchical understanding of the international system makes structural realism expect the military utility to become an essential factor in the state's decision in times of war. This is because the military utility has material characteristics (e.g. immense destructive power, ability to overcome defenses, and invulnerability to pre-emption) which cannot be taken away from how states will act in the international system (Martin, 2013). In addition, material consequences of a state's actions in international affairs have always become a structural realism priority (Brilmayer, 1999;Martin, 2004). Those logics serve as a foundation for Martin (2016) to conceptualize three types of military utility, which are technical utility, tactical utility, and strategic utility. The technical utility is defined by Martin (2016) as to whether it works to have the desired result on thse target. The term 'works' in the definition is in a narrow sense, meaning the function or operation certainty. Martin (2016) gave three examples to illustrate the technical utility, such as a soldier's rifle may fail to discharge bullets, chemical agent spray may fail to have an effect, or aerial bombs may fail to detonate. It is relatively clear to asses technical utility compared with tactical utility and strategic utility for the reason that only focuses on mechanical aspects (Martin, 2016). Chapman et al. (2018) argued technical utility concerns about the general use of weapons. It is all about whether weapons can kill, injure, or destroy large numbers of troops or affect large areas assuming there is no influence from the target and environmental factors. The technical utility can also be understood as basic use, meaning ignoring the battlefield condition in the use of the weapon calculation (Chapman et al., 2018). Moving forward to the second military utility, the tactical utility has relation to the context on the battlefield. Martin (2016) stated that something has tactical utility when it helps to achieve the user's aims within the theatre of wars. Tactical utility concern whether it supports or not accomplishes objectives in the campaign (Martin, 2016). Hughes (1994) and Martin (2016) also argued that tactical utility can become more difficult to exert because it is time-dependent. In that case, the tactical utility can be said to rest upon the actor's interest in the operation and the phase of the operation. Further elaboration on tactical utility has been done by Chapman et al. (2018). They defined tactical utility as the performance of the weapon on the battlefield. Chapman et al. (2018) gave five examples to illustrate the tactical utility, namely preventing the opponent to capture land, reaching an invulnerable target, undermining the morale of opponent troops, imposing logistical challenges for enemy in the battlefield, and maintaining battle objective in the course of operation. The strategic utility is the last military utility identified by Martin (2016) which refers to whether it maintains the actor's broader goals. While tactical utility concentrates on the scope of the battlefield, the scope of strategic utility comprises political goals in international affairs (Martin, 2016), for instance: 1. sovereignty preservation of a nation; 2. military threats containment; and 3. prevention of further escalation against others. In practice, decision-makers also take into consideration certain strategic utilities that will help further foreign policy in the long term. It is because the discussion of strategic utility has an association with the potential undesired costs (Martin, 2016). While the technical utility is about the weapon and tactical utility is about the battlefield, strategic utility is about the state. Chapman et al. (2018) interpreted strategic utility as how the use of weapons is in line with objectives, principles, or plans from the national strategy. Besides that, Chapman et al. (2018) also include the rhetoric and actions of individual states as one of the frameworks to examine the strategic utility. Chapman et al. (2018) who used Martin's military utility concept to analyze chemical weapons in the Syrian Civil War suggested that the concept is helpful to provide descriptions of what states have more to gain from acquiring or using weapons. This makes the use of the military utility concept as an analytical framework applicable to the research question of this article. Besides that, the qualitative conceptualization of the military utility is also relevant to the method used in this article. Chapman et al. (2018) demonstrated it by operationalizing literature-based studies as the main data collection technique and using contemporary scholarships as the primary data sources in their article. Technical Utility of Db1000n Db1000n is a software to launch a Distributed Denial of Service (DDoS) attack against Russia's information infrastructure. DDoS is defined by Keromytis (2017) as an attack where a huge number of computers are simultaneously seeking access to an online service. It is one of the most common types of attack that can make the target run out of operation and exhaust the intermediate networking path around the target (Gupta & Dahiya, 2021). Many DDoS attack software is already available in the market, such as Mstream, Trinoo, and HOIC (Gupta & Dahiya, 2021). But, db1000n is a DDoS attack software created for Ukraine's cyber warfare effort specifically. The creation of db1000n was first initiated by a Ukrainian cyber security specialist named Ivashko (2022c) on February 26, 2022. Since then, a total of 50 contributors have been helping in the development of db1000n up to now (Ivashko, 2022f). Db1000n is developed and maintained by the contributors in the GitHub (2022) repository, one of the leading online platforms that facilitate software project collaboration and management. The software itself has been through several iterations since its creation. Version 0.8.0 was the first to be adopted officially by the Ukrainian IT army, while the newest as of August 9, 2022, is version 0.9.17 (Ivashko, 2022g). Ivashko (2022d) and contributors used "Go" as the programming language to develop db1000n. Go language was first created at Google by Robert Griesemer, Rob Pike, and Ken Thompson in March 2012(McGrath, 2020. According to Bodner (2021), the Go language is intended for creating software that will last for a long time and can be modified by huge numbers of computer programmers over many years. With that intention, programming in the Go language lean to be more straightforward and sometimes quite repetitive compared to most programming languages nowadays (Bodner, 2021). But in return, the Go language is fast and simple to write (McGrath, 2020). Db1000n can be installed and run by anyone who wants to providse their computer as a DDoS attack launch point. The software is supported in several operating systems such as Windows, Linux, and Macintosh (IT Army of Ukraine, 2022b). Based on the instruction manual, someone who wants to join the attack on Russia's information infrastructure needs to do two things, namely download the latest version of db1000n from GitHub and start the software (IT Army of Ukraine, 2022b). The Ukrainian IT army also gave an optional instruction which is to use virtual private networks (VPN) before launching the software. VPN allows a more secure and reliable connection over the internet (Ashraf, 2018). The following paragraph will describe the way db1000n works. The software regularly fetches files that contain a list of Russia's information infrastructure targets from the Ukrainian IT army's server to the local computer (Ivashko, 2022a). The files arrange what target should be attacked in parallel with other computers that also have db1000n installed on them (Ivashko, 2022a). Db1000n will keep launching DDoS attacks on the arranged target as long as the software and the internet connection run (IT Army of Ukraine, 2022b). In short, db1000n works by managing target change automatically from the Ukrainian IT army's computer to many computers during DDoS attacks against Russia. The first use of db1000n by the Ukrainian IT army was when attacking four Russian government websites on March 24, 2022. These four websites are the portal of foreign economic information, the ministry of foreign affairs, the ministry of economic development, and the ministry of education and science (IT Army of Ukraine, 2022a). The Ukrainian IT army claimed that the attack was successful to make the websites run out of operation (IT Army of Ukraine, 2022e). As of April 23, 2022, only the ministry of economic development website and the ministry of education and science website had come back into operation (Ministry of Economic Development of Russia, 2022; Ministry of Education and Science of Russia, 2022). Db1000n was also used by the Ukrainian IT army to attack Russian company websites, for example, an attack was launched against an express delivery service company in Russia named CDEK to disrupt and wipe out its courier system on March 25, 2022 (IT Army of Ukraine, 2022f). The Ukrainian IT army claimed they managed to achieve the goal within seven hours (IT Army of Ukraine, 2022g). Russian media started to pay attention to Ukraine's db1000n after this attack. One of them was the Russian state-owned news agency Novosti (2022) which claimed that the attack caused national-scale failures in CDEK operations for several hours. As of August 9, 2022, the use of db1000n has been taken down as many as 186 Russian websites (IT Army of Ukraine, 2022j (Ivashko, 2022h) created the software. The most prominent issue is that db10000n sometimes failed to fetch target list files from the Ukrainian IT army's server which caused the computer unable to configure the attack properly (Bondarenko, 2022;Pashagolub, 2022;Chmil, 2022). So far, 189 issues have been solved by db1000n contributors by frequently releasing a new version of the software (Ivashko, 2022e). The remaining 37 issues are still being discussed by the contributors on GitHub up to now (Ivashko, 2022i). Tactical Utility of Db1000n Cyber warfare has different characteristics compared to warfare on land, sea, or air. Cyberspace does not have a definable expanse because it exists in multiple locations of many computer systems and networks (Sloan, 2017). But, this does not mean cyberspace as a new domain of warfare has different nature than warfare in other domains because it is still the application of force against the enemy (Sambaluk & Spafford, 2020). This also implies that cyber warfare tactics have the same basic principles as other domains which is the use of forces to win engagements (Echevarria, 2017). The following section will present a further description of the identified tactical utility of db1000n. The first tactical utility of db1000n it allows an easier way to amass many computers during the attack against the Russian government or company websites. It could be said that DDoS attacks are one of the easiest attacks to launch in cyberspace because these attacks do not need someone to have great cyber security knowledge to inflict damage (Sambaluk & Spafford, 2020). However, one major drawback of DDoS attacks is acquiring a large number of computers as hosts to launch it in the first place (Keromytis, 2017). DDoS attacks according to Gupta & Dahiya (2021) rely heavily on the dispersion of multiple distributed launch points against the target. In other words, the more computers launch DDoS attacks, the more damages are inflicted. In this matter, db1000n supports Ukraine's cyber warfare effort against Russia by being available to be downloaded openly on the web. All versions of db1000n have been downloaded from GitHub more than 345 thousand times since its creation on the last of February 2022 (Shehryar, 2022). The latest version of the software itself, which was released in the first week of August 2022, has been downloaded around two hundred times now (Shehryar, 2022). Meanwhile, the most downloaded version of db1000n is version 0.7.12 with a total number of downloads as many as 35 thousand times (Shehryar, 2022). This version was the predecessor of version 0.8.0, the first version that was officially adopted by the Ukrainian IT army (Ivashko, 2022b). Further details on the last 18 versions of db1000n can be seen in Table 1. Shehryar (2022) The Ukrainian IT army also has acknowledged that the more db1000n downloaded, the more optimal db1000n's tactical utility for cyber warfare against Russia. Some efforts have been done by the Ukrainian IT army to encourage internet users to download the db1000n and keep up to date with the new version of the software. One of the efforts is using their official Telegram channel to remind it online frequently when announcing the result of the DDoS attacks against Russia. For instance, the Ukrainian IT army did it when announcing the result of the attack against the Russian CDEK express delivery service website on March 26, 2022 (IT Army of Ukraine, 2022g). The second and the last tactical utility of db1000n is helped to counter a new Russia's tactics to ward off Ukraine's DDoS attacks against them. The Ukrainian IT army claimed that the Russian websites started to employ two DDoS attack defense mechanisms on March 23, 2022, called Internet Protocol (IP) address change and IP address block (IT Army of Ukraine, 2022d; IT Army of Ukraine, 2022h). An IP address is a unique identification number that is assigned and configured to each device either manually or automatically to be able to communicate on the internet (Rooney & Dooley, 2021). It serves as the beginning and endpoint of message delivery in two ways communication between many computers on the internet. The IP address change tactic makes Russian government or company websites become hidden among many computers' IP addresses that are not Ukraine's DDoS attacks targets (IT Army of Ukraine, 2022d). Whereas the IP address block tactic makes Ukraine's DDoS attacks against Russia's information infrastructure would fail even before these attacks are launched because the target self-isolates from the internet for several moments (Gupta & Dahiya, 2021; IT Army of Ukraine, 2022h). The Ukrainian IT army claimed that the implementation of the IP address change tactic and IP address block tactic caused the success rate of DDoS attacks without using db1000n to only 40% as of April 1, 2022 (IT Army of Ukraine, 2022i). Most of the Russian websites that used both tactics are Russian banks, for instance Cetelem Bank, NK Bank, Kremlin Bank, Tender Bank, and Alternative Bank (IT Army of Ukraine, 2022j). Meanwhile, the Russian government website has been using both tactics for the invasion only by the Ministry of Information (IT Army of Ukraine, 2022j). IT Army of Ukraine (2022j) claimed the IP address change tactic and the IP address block tactic make those five Russian bank websites and the ministry of information website still can be accessed as of August 9, 2022. In that respect, db1000n helps Ukraine's cyber warfare effort against Russia by target switching automatization. As mentioned in the technical utility section, db1000n has a fetch files mechanism to manage DDoS attack targets from the Ukrainian IT army's server to the local computer. The mechanism allows the Ukrainian IT army to change the target of the DDoS attack instantly for every computer that has db1000n installed and turned on when encountering the IP address change tactic or IP address block tactic by Russia. This is also in line with the target adjustment as an important key to maintaining tactical advantages in cyberspace for the long term. According to Libicki (2021), target adjustment ensures the concentration of force against those that respond slowly regardless of how attacks in cyberspace are employed. Nevertheless, a DDoS attack using designated computer software like db1000n is not the ultimate tactic in cyber warfare. There are still many types of software that can be operationalized to win military engagement in cyberspace. One of them is called malicious software (malware). This software works primarily by seeking to duplicate themselves from one computer to many computers and then cause some substantial form of damage such as corrupting data and disabling networks (Cares, 2017). According to Cares (2017), malware can be divided into seven main categories, which are computer viruses, worms, trojan horses, spyware, adware, ransomware, and scareware. Despite that, the tactical utility of DDoS attack software like db1000n is still sufficient compared to malware in the Russian invasion of Ukraine case. Db1000n already overcome the problem of acquiring a large number to launch DDoS attacks based on the download statistics. The automated fetch files mechanism of db1000n also proved relevant to counter the IP address change tactic and IP address block tactic used by the Russian websites. Strategic Utility of db1000n Ukraine officially put into force the Strategy of National Security on February 16, 2022. Just around a week before the 2022 Russian Invasion of Ukraine happened. This strategy which is planned to implement by 2025 serves as the foundation for determining the real and potential threats to Ukraine's security,) directions and objectives of Ukraine's security, and planning and implementing Ukraine's security policy (Office of the President of Ukraine, 2022). In general, the Strategy of National Security itself consists of four sections, namely 1) general provisions; 2) security environment and threats to national security; 3) goals, directions, and objectives of national policy on security issue; and 4) organizational and financial support for the strategy implementation (Office of the President of Ukraine, 2022). The goal of Ukraine's security based on the strategy is the prevention, detection, and elimination of internal and external threats to national security and the conditions that facilitate the occurrence of those threats (Office of the President of Ukraine, 2022). The strategy stated clearly that the main threat to Ukraine's security for the next few years is Russia because of their systematic use of political, economic, informational, psychological, and cyberspace means in the ongoing conflict (Office of the President of Ukraine, 2022). Furthermore, the strategy also stated that something is a threat to Ukraine's security when it is made difficult for Ukraine to protect the national sovereignty, territorial integrity and democratic constitutional order, and other vital national interests (Office of the President of Ukraine, 2022). The goal stated in Ukraine's Strategy of National Security reflects the survival assumption of the structural realism approach. Pashakhanlou (2016) affirmed structural realism and explicitly said that survival is important for the state since it is the prerequisite for the accomplishment of all other goals in international politics. It bases on the main argument by well-known structural realism thinker Kenneth that states seek to ensure their survival in the anarchical structure of the international system (Waltz, 1979). Additionally, Russia as the main threat to Ukraine's security for the next few years further illustrates how structural realism sees the relations among sovereign states. According to Baylis (2020), structural realism described the durability of anarchical structure in the international system implying that interstate relations in the future are more likely to be violent. The Strategy of National Security also gave general orientations as the foundation on how Ukraine should deal with the threats. The general orientations consist of twenty-one points that cover counterintelligence to international cooperation (Office of the President of Ukraine, 2022). Three of them gave concerns on military technology and cyber warfare, which are 1) development of national security capabilities system about timely prevention, detection, and countermeassurment of Ukraine's external and internal threats; 2) completion of establishment, development, and reinforcement of the cyber security system capabilities to effectively combat cyber threats in the modern security environment; and 3) increasing the technological capabilities of the national security with the adoption of the latest systems of hardware and tools (Office of the President of Ukraine, 2022). Refers to Ukraine's security goal and three general orientations that have been mentioned above, the usage of db1000n against Russia is in line with the Strategy of National Security. It can be argued that overall db1000n has strategically relevant utility as a tool to countermeasure and combat Russian threats in cyberspace by boosting Ukraine's technological capability in cyberspace for current circumstances. This strategic utility not only came out of necessity due to the 2022 Russia Invasion of Ukraine happened but also because the db1000n itself has been proven to possess the technical utility and the tactical utility based on the findings and discussion in the two previous sub-sections of this article. In short, the way db1000n works and supports cyber warfare objectives against Russia is well relatable to Ukraine's seeks for survival in the international system. The following section will present further elaboration on this matter. DDoS software like db1000n is a recent expansion on what tools can be strategically wielded when dealing with other states. According to Kello (2017), this creates two implications that states must consider for their foreign affairs, first, there will be problems with the technology's sheer speed of action that can make states not fully comprehend and manage them and, second, there will be revolutionary empowerment that can help states to achieve their goals and preserve their national interests in the anarchical structure of the international system. Additionally, Kello (2017) also argued both of the implications will create shock in international order by making the conflict in cyberspace not easy to model, regulate, and terminate among sovereign states. This further complicates the conflictual and uncertain tendency of international politics for states that have been interconnected in cyberspace. DDoS attacks using software like db1000n make a strategic advantage in current international politics. Buchanan (2020) argued many states have relied on cyber warfare in recent years as a playbook of statecraft to gain advantages over other states by conducting attacks, espionage, or destabilization in cyberspace. Cyber warfare has an accumulation of strategic effects for states because it is one form of covert action with high force, precision targeting, and unexpected timing (Buchanan, 2020). Moreover, the rapid progression of technology also opens up a clear possibility to inflict more damage on targets that cannot be anticipated flawlessly by states over time. According to Buchanan (2020), this pattern has strategic implications for many states because they will embrace it to attack one another as the key part of the struggle for power in cyberspace. Even though there is a utility to ensuring states' survival in a broad sense, the strategic expectations realization of software like db1000n is still challenging for now. Maschmeyer (2021) found there is a trilemma that limits the strategic utility of many tools for cyber warfare, namely slow operational speed to start to produce meaningful effects, constraint on scope and scale of the effects, and efforts to maintain control over a targeted system. Each part of the trilemma is negatively correlated, which means the state's gains in one part of the trilemma tend to produce losses across the other two parts (Maschmeyer, 2021). This is why Maschmeyer (2021) concluded that the existence of those trilemma makes strategic expectations of attacks in cyberspace can fall short and only deliver limited utility for the state. Based on Maschmeyer's trilemma, it could be hypothesized that the use of db1000n faces two problems, namely constraint on the scope and scale of the effects and efforts to maintain control over a targeted system. One piece of evidence that leads to this hypothesis is adopt a new software called "mhDdos" (IT Army of Ukraine, 2022k). This software was created by a cyber security specialist named Oleksandr on June 24, 2022 (Black, 2022). CONCLUSIONS, RECOMMENDATION, AND LIMITATION Therefore, how does the use of db1000n help Ukraine against the Russian invasion? This article concludes with the following answer based on the results and discussion in the previous section. The use of db1000n helps Ukraine to gain advantages against the Russian invasion by being technical, tactical, and strategic DDoS attack software. Technical means straightforward cyber warfare software, tactical stands for effectively countering the Russian defense mechanism in cyberspace, and strategic implies ensuring Ukraine's survival in the current conflictual condition. Despite that, db1000n is not the pinnacle of cyber warfare software because it still needs to be maintained properly and does not guarantee easy success. This article proposes three recommendations based on limitations to fully contribute to the discussion of military technology and cyber warfare. First, further evidence for the military utility of db1000n is needed because the 2022 Russian invasion of Ukraine is still ongoing. Second, a comparative case study is necessary to verify the military utility of db1000n among other DDoS attack software. Last, further research is required to investigate Maschmeyer's trilemma in the strategic utility of db1000n.	2022-09-15T15:57:34.464Z	2022-08-31T00:00:00.000	{ "year": 2022, "sha1": "b0077fb7f33b1c3216c8b15433e710c7765d4099", "oa_license": "CCBYNC", "oa_url": "https://jurnal.idu.ac.id/index.php/DefenseJournal/article/download/1683/ans", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "366fb272ade60ffe8d332ef14bebba6426a8a7e3", "s2fieldsofstudy": [ "Computer Science", "Political Science" ], "extfieldsofstudy": [] }
219792786	pes2o/s2orc	v3-fos-license	Flexible Mixture Priors for Time-varying Parameter Models Time-varying parameter (TVP) models often assume that the TVPs evolve according to a random walk. This assumption, however, might be questionable since it implies that coefficients change smoothly and in an unbounded manner. In this paper, we relax this assumption by proposing a flexible law of motion for the TVPs in large-scale vector autoregressions (VARs). Instead of imposing a restrictive random walk evolution of the latent states, we carefully design hierarchical mixture priors on the coefficients in the state equation. These priors effectively allow for discriminating between periods where coefficients evolve according to a random walk and times where the TVPs are better characterized by a stationary stochastic process. Moreover, this approach is capable of introducing dynamic sparsity by pushing small parameter changes towards zero if necessary. The merits of the model are illustrated by means of two applications. Using synthetic data we show that our approach yields precise parameter estimates. When applied to US data, the model reveals interesting patterns of low-frequency dynamics in coefficients and forecasts well relative to a wide range of competing models. In this paper, we focus on estimating TVP vector autoregressive (VAR) models with a large number of endogenous variables. Due to severe overfitting issues in large TVP-VARs, special emphasis is paid to important modeling decisions, such as whether coefficients evolve gradually, change abruptly or remain constant for subset of periods. In macroeconomic applications, it is common to assume that coefficients evolve according to a random walk, implying that parameter change smoothly over time. As noted by the recent literature (see, for example, Lopes et al., 2016;Hauzenberger et al., 2019), however, this assumption might be overly simplistic and lead to model misspecification. In large TVP-VARs it is often reasonable to assume that most parameters remain constant over time, while only few vary. To capture this behaviour, the Bayesian literature frequently uses shrinkage priors on the state innovation variances to sufficiently push them towards zero (Frühwirth-Schnatter and Wagner, 2010;Belmonte et al., 2014). A severe drawback of this strategy is that it only accounts for the case that a given coefficient is constant for all points in time (labeled static sparsity). Another common situation faced by researchers is that coefficients change only at certain points in time (this is referred to as dynamic sparsity). Using a mixture distribution on the innovation variances, for example, allows to push small parameter changes towards zero (see, inter alia, Gerlach et al., 2000;Giordani and Kohn, 2008;Koop et al., 2009;. 1 Alternatively, Hauzenberger et al. (2019) introduce a more flexible law of motion by assuming a conjugate hierarchical location mixture prior directly on the time-varying part of the coefficients. This location mixture allows for a dynamically adjusting the prior mean on the TVPs to capture situations with a low, moderate or even large number of structural breaks in the coefficients. However, both techniques come with drawbacks. For instance, the mixture innovation model of , equipped with a latent threshold mechanism, discriminates between a high and a low innovation variance state. However, the authors do not discard the random walk law of motion, which might be too restrictive. Hauzenberger et al. (2019) use either a conjugate g-prior (Zellner, 1986) or a conjugate Minnesota prior (Doan et al., 1984;Litterman, 1986), potentially lacking flexibility to disentangle abrupt from gradual changes. In this paper, we carefully design suitable mixture priors for the state equation. In a first variant, a mixture prior is not only introduced on the state innovations, but also on the autoregressive coefficients in the state equation to obtain sufficient flexibility. To achieve parsimony in large models, a latent binary indicator determines the law of motion for the TVPs and detect periods where coefficients evolve according to a random walk and times where the TVPs are better characterized by a stationary stochastic process. Combined with a mixture on the innovation volatilites and suitable shrinkage priors, this approach is capable of automatically capturing a wide range of typical parameter changes. In a second variant, the sparse finite location mixture model of Hauzenberger et al. (2019) is extended by considering non-conjugate shrinkage priors and by replacing the location mixture with a location-scale mixture. Here, an additional mixture on the state variances captures the notion that structural breaks in coefficients happen infrequently (with potentially large TVP innovations), while most of the time coefficients are constant (with TVP innovations pushed towards zero), similar to mixture innovation models. In the previous paragraphs we repeatedly stated that our techniques are well suited to handle overfitting issues in large TVP-VARs. But large TVP models also raise the question of computational feasibility. In this contribution, computational complexity is reduced by using recent advances in estimating large-scale TVP regression (see Chan and Jeliazkov, 2009;McCausland et al., 2011;Hauzenberger et al., 2020). These are based on rewriting the TVP model in its static regression form. In this representation, the TVP model is treated as a very big regression model and the techniques proposed in Bhattacharya et al. (2016) can be used. Since these algorithms are designed for single equation models, we estimate the VAR model using its structural representation and thus estimate a set of unrelated TVP regressions (see Carriero et al., 2019). Based on two applications we investigate the merits of the techniques developed in the paper. First, in an application using synthetic data we illustrate that the proposed methods work well in detecting small and large structural breaks in coefficients. Second, we employ a large US macroeconomic dataset for an empirical application. Our proposed methods reveal interesting patterns in the low-frequency relationship between unemployment and inflation. Moreover, to evaluate predictive performance of our approach, we perform a comprehensive forecasting exercise. This forecasting horse race shows that the proposed framework works well relative to a wide range of competing models. The remainder of the paper is structured as follows. Section 2 introduces a TVP regression model with flexible mixture priors and sketches the main contributions of the paper. Section 3 generally outlines inference in these models, while Section 4 discusses the posterior sampling algorithm of Bhattacharya et al. (2016), when applied to non-centerred TVP regressions. Section 5 and Section 6 show the results for artificial data and US data, respectively. Finally, Section 7 summarizes and concludes. A TVP Regression Let y t denote a scalar time series and x t refer to a K-dimensional vector of predictors, then the observation equation for a TVP regression can be written as: Here, α t is a K-dimensional vector of TVPs that relates x t to the quantity of interest and ε t denotes the measurement error with mean zero and time-varying variance σ 2 t . For the state equation of σ 2 t , we assume a stochastic volatility (SV) specification and refer to Appendix A.1 and Kastner and Frühwirth-Schnatter (2014) for details. Typically, α t is assumed to evolve according to a random walk (RW). In this paper, interest centers on relaxing this assumption. In the following, to achieve both sufficient flexibility and model parsimony, we use two different mixture specifications for α t . In the first variant, we assume that coefficients evolve according to a mixture of a random walk and a white noise process. In the second variant, interest centers on further relaxing the law of motion proposed in Hauzenberger A Flexible State Equation For a mixture between a random walk and white noise process we assume that the evolution of α t is given by: with α 0 denoting a K-dimensional intercept vector, φ t being a K-dimensional diagonal autoregressive coefficient matrix and ς t denoting a K-dimensional vector of state innovations, which are centered on zero and feature a K × K-dimensional variance-covariance matrix Ψ t . Moreover, we assume φ t and Ψ t to evolve according to a regime-switching process: and Here, S t = diag (s 1t , . . . , s Kt ) denotes a binary indicator matrix with {s it } K i=1 being either zero or one.Ψ 1 = diag (ψ 11 , . . . ,ψ K1 ), andΨ 0 = diag (ψ 10 , . . . ,ψ K0 ) refer to K-dimensional diagonal matrices. Equation 3 assumes that coefficients evolve according to a mixture of a random walk and a white noise process, while Equation 4 ensures sufficient flexibility of the state innovations, respectively. For example, if the covariate-specific indicator s it = 1 in the t th period, the i th covariate follows a random walk with state innovation variance ψ it =ψ i1 , while if s it = 0 in the t th period, it follows a stochastic process with variance ψ it =ψ i0 . This specification (henceforth labeled as TVP-MIX) nests a wide variety of popular TVP models, such as standard RW state equations and mixture innovation models. 2 A standard random walk evolution is trivially obtained by setting S t = I K . A so-called mixture innovation model assumes φ t = I K and specifies Ψ t similar to Equation 4 (Gerlach et al., 2000;Giordani and Kohn, 2008;Koop et al., 2009;. Additionally, mixture innovation specifications restrictΨ 0 = κΨ 0 with κ being a small value close to zero andΨ 0 being a diagonal matrix collecting variable specific scaling parameters. 3 Apart from discussing the relation to other popular TVP models, it is also worth highlighting additional features of the model proposed in (2) to (4). If a parameter is almost constant, but also features larger abrupt changes for some periods, we would expect thatψ i0 >ψ i1 . This case is of particular interest, when compared to a standard mixture innovation model with random walk state equation. Conversely, if a coefficient features large, more persistent swings, but also some periods of parameter stability, we would expectψ i0 <ψ i1 . Intuitively, the relative proportions of ψ i0 andψ i1 depend mainly on the nature of coefficient changes. Alternatively, if the i th coefficient is constant or negligible (static sparsity), this can be achieved withψ i1 and/orψ i0 close to zero (Lopes et al., 2016). Note that in the special case of constant coefficients, the proposed specification is not identified. We address this issue in the context of interpreting the state indicators S t . 2 In the empirical application, these are considered as important benchmarks. 3 Related to literature on variable selection (George and McCulloch, 1993;1997) A Hierarchical Pooling Specification For a hierarchical pooling specification, we follow Hauzenberger et al. (2019) and assume that the time-varying part of α t follows a sparse finite mixture in the spirit of Malsiner-Walli et al. (2016). The specification of the state equation α t (labeled as TVP-POOL) reads as: Here, α 0 denotes a K-dimensional constant coefficient vector and γ t is assumed to be a Kdimensional vector of random coefficients featuring a specific structure. That is, conditional on latent group indicators θ t that takes a value n ∈ {1, . . . , N }, γ t follows a multivariate Gaussian distribution: where µ n refers to the group-specific mean and Ψ t denotes the variance-covariance matrix. It is also worth noting that θ t serves as group indicator for γ t . The probability that γ t is assigned to cluster n is defined as P (θ t = n) = ω n . This structure is closely related to the setup of Hauzenberger et al. (2019). In the following, we extend their location mixture prior to a location-scale mixture prior by introducing a regimeswitching specification on Ψ t similar to Equation 4. That is, with bothΨ 0 andΨ 1 being diagonal matrices and S t denoting a binary indicator matrix. Similar to standard mixture innovation models one component serves to detect larger breaks, while a second component handles dynamic sparsity. We therefore discard the conjugate prior assumption of Hauzenberger et al. (2019) and instead assume non-conjugate shrinkage priors on both state variances (described in more detail in Subsection 3.1). Before proceeding, it is also worth sketching the general idea of this random coefficient specification. This model can be seen as a stochastic variant of multiple break point specifications (Koop and Potter, 2007), which is capable of capturing situations with a low, moderate or even large number of structural breaks. To estimate the number of regimes, we follow Malsiner-Walli et al. ) is designed to support inference about the locations γ n , for n = {1, . . . , N }. We expect that many elements in {γ t } T t=1 cluster around zero (i.e. coefficients are constant with Ψ t close to zero), while occasionally there are structural breaks in some coefficients (requires relatively large values in Ψ t ). Especially we aim to detect these two extremes (changes/no changes in α t ) with γ t . The Latent State Indicator Matrix Sofar we remained silent on the evolution of S t . There are many different possibilities how the binary indicators s it , for i = {1, . . . , K} evolve over time. In the following, we assume two laws of motion: 1. Pooled Markov-switching process: When assuming a first-order Markov process for each s it independently, sampling the state indicators can be computationally cumbersome, especially if K is large. Since one has to rely on forward filtering backward sampling algorithms, computation time quickly adds up. Therefore, we replace S t with s t I K . In the following, s t is assumed to be common to all K covariates in period t and governed by a joint Markov process. 4 This process is driven by a transition probability matrix given by: with transition probabilities from state k to l denoted by p kl and following a Beta distribution p kk ∼ B(c 0k , c 1k ), for k = {0, 1} (see Uribe and Lopes, 2017). Independent over time and covariate-specific indicators: The assumption that a joint indicator governs the evolution of large number of coefficients might be too inflexible in certain cases. For this reason, we also specify covariate-specific indicators, coupled with independent mixture priors (see Lopes et al., 2016). In contrast to covariate-specific Markov processes, mixture priors are assumed to be independent over time and thus do not involve computationally demanding forward filtering backward sampling algorithms. In the following, s it is assumed to follow an independent Bernoulli distribution with P (s it = 1) = p i and p i being Moreover, it should be noted that the prior choice on the binary indicators is quite influential. For the random walk/white noise mixture (TVP-MIX), the hyperparameters are chosen in such a way that it is more likely that gradual changes have a higher (unconditional) expected duration (with s t = 1) than abrupt changes (with s t = 0). In the empirical application we therefore set BAYESIAN INFERENCE To discuss inference for both variants outlined in Section 2, we introduce a very general state equation for α t : Equation 8 nests both approaches with the first variant (TVP-MIX) being obtained by setting γ t = 0 K×1 , while the second approach (TVP-POOL) is given by defining φ t = 0 K×K and γ t = µ n , if θ t = n. The Non-Centered Parameterization In this subsection we exploit the non-centered parameterization to writeΨ 0 andΨ 1 as part of the observation equation, enabling shrinkage on the regime-switching state innovation volatilities (Frühwirth-Schnatter and Wagner, 2010). We therefore recast the model as follows: Here,α t is a K-dimensional vector of normalized states, defined asα t =Ψ −1 t (α t − α 0 ) and and, more compactly, as a standard regression model: ) denoting a 3K-dimensional covariate vector and α = (α 0 , ψ 11 , . . . , ψ K1 , ψ 10 , . . . , ψ K0 ) being a 3K-dimensional coefficient vector. On the time-invariantα we use a hierarchical global-local shrinkage prior (see Polson and Scott, 2010) whereα j refers to the j th element inα, λ denotes a global shrinkage parameter and τ j induces local shrinkage. In the empirical application, we focus on the Normal-Gamma (Griffin and Brown, 2010) shrinkage prior. This shrinkage prior has been proven to be successful in macroeconomic and financial application (see, for example, and is quite common in the literature. 5 The exact prior specification is outlined in Appendix A.2. The Static Representation If interest centers on estimating the latent states {α t } T t=1 , we can straightforwardly recast Equation 9 in a static regression form by conditioning on α 0 , the state innovation volatilities {Ψ t } T t=1 and the stochastic volatilities in Σ = diag (σ 1 , . . . , σ T ). We define y as a T -dimensional vector, X as a T × K-dimensional matrix and as a T -dimensional vector with y t , x t and t on the t th position, respectively. Then, the static form of Equation 9 is: Φα =γ + η, η ∼ N (0, I ν ). Moreover, the prior mean a 0 also depends on the structure of Φ andγ. The simplest thing is to set a 0 to a zero vector, which we implicitly assume for the TVP-MIX variants. For the TVP-POOL approach we use a hierarchical mixture prior on a 0 , described in detail next. A Hierarchical Prior Mean The model outlined in Equation 5 to Equation 7 denotes a sparse finite location-scale mixture. After recasting the model in the non-centered parameterization, we are able to replace the location-scale mixture prior on α t (outlined in Equation 6) with a location mixture prior on the normalized latent statesα t , since the scales of Equation 7 (Ψ 0 andΨ 1 ) are now part of the observation equation. That is: with group-specific meanμ n , for n = {1, . . . , N } and variance-covariance matrix I K . In the following, the prior mean is defined as a 0 (=γ) = (a 01 , . . . , a 0T ) , with a 0t =μ n if θ t = n. ω\|ξ ∼ Dir(ξ, . . . , ξ), with ξ referring to an intensity parameter. The prior on the intensity parameter is specified as: with d 0 = 10 in the empirical application. Here, we closely follow Malsiner-Walli et al. (2016), who show that this prior choice is successful in detecting superfluous components and obtaining a parsimonious mixture representation. POSTERIOR COMPUTATION In this subsection, we outline the MCMC sampling step forα. We stress that drawingα is computationally fast, also for relatively large K, but sampling the ν-dimensional vectorα is computationally demanding (Hauzenberger et al., 2020). Thus forα (and the remaining parameters) we use standard MCMC techniques with sampling steps and conditional posteriors outlined in Appendix A.3. Forα, irrespectively of the structure of a 0 and Ω 0 , we obtain standard conditional Gaussian The main issue, however, is that the inversion of Ω −1 1 is computationally costly, since it is ν × νdimensional matrix with ν = T K and {T, K} being potentially large integers. Thus, to avoid high-dimensional full matrix inversions and Cholesky decompositions for drawing the normalized latent statesα, we rely on the algorithms proposed in Bhattacharya et al. (2016), applied to TVP models in Hauzenberger et al. (2020). This method involves the following steps: 3. Define q = a 0 + Φ −1 u, with Φ −1 denoting the lower Cholesky factor of Ω 0 , and r =W q + v. Moreover, using the static representation for a TVP regression the involved matrices are sparse, which can be exploited to achieve additional computational gains (see Chan and Jeliazkov, 2009;Hauzenberger et al., 2019;2020). Depending on the structure of Φ there a two extreme cases as briefly discussed in Subsection 3.2. Computationally the most expensive case is a random walk state equation (φ t = I K , ∀ t), while having no autoregressive structure in the state equation it is computationally less demanding. 7 Recall, the former Φ has a specific lower triangular structure (rendering Ω 0 block diagonal) and in the latter both Φ and Ω 0 are diagonal. Thus, even for a random walk state equation (the most dense case), using sparse algorithms pays off in terms of computation. Moreover, if φ t = S t , for some t, we have to account for an intermediate computational burden lying between the two extremes that eventually depend on the exact structure of Φ (Ω 0 ). Equation-wise estimation for a TVP-VAR The methods outlined in the previous subsection are designed for single equation models. To use these algorithms also for posterior inference in TVP-VARs, we rewrite the multivariate model as a set of unrelated TVP regressions (see Carriero et al., 2019). This can be done by using the structural form of the TVP-VAR: Here, Y t = (Y 1t , . . . , Y mt ) denotes an m-dimensional vector of endogenous variables with B 0t being an m × m-dimensional strictly lower-triangular matrix (with zero main diagonal) defining contemporaneous relationships between the elements of Y t . Moreover, B it , for i = 1, . . . , p, denotes an m × m-dimensional time-varying coefficient matrix, C t is an m-dimensional intercept vector and t refers to an m-dimensional Gaussian distributed error vector, centered on zero and with time-varying In the following, for i = 2, . . . , m, the i th equation of Y t is given by: Moreover, for the first equation (i = 1) we have x 1t = (Y t−1 , . . . , Y t−p , 1) and α 1t = (B 1•,t , c 1t ) . SIMULATION STUDY In this section we use synthetic data to illustrate the features of the proposed mixture variants. For the data generating process (DGP) we assume that the number of observations is T = 100 and the number of covariates is given by K = 5. The covariates are simulated with X j ∼ N (0, I T ) for j = 1, . . . , (K − 1) and X K = ι T with ι T being a T -dimensional vector of ones. For the error variance σ 2 t , we assume an SV specification with log(σ 2 t ) = h t following a random walk process. That is, h t = h t−1 + ϑ t with ϑ t ∼ N (0, 0.1) and h 0 = log(0.1). For the time-varying parameters α t we assume quite specific laws of motion. We define α 0 = (−4, 3, −2, 2, 0) as initial level and assume that both regime-switching autoregressive parameters and regime-switching variances in the state equation are governed by a joint Markov process s t . Here, we let S t = s t I K , φ t = S t , Ψ t = S tΨ1 + (I K − S t )Ψ 0 withΨ 0 = diag (10 −10 , 0.5, 0.1, 10 −10 , 1) andΨ 1 = diag (1, 0.1, 0.5, 10 −10 , 10 −10 ). The joint indicator s t is simulated with transition probabilities p 00 = 0.6 and p 11 = 0.95, effectively leading to a higher unconditional probability that α t follows a random walk evolution. In particular, the first coefficient features larger, more persistent parameter changes (with ψ 11 = 1), while for a small number of periods α 1t is basically constant (achieved through the white noise state equation with ψ 10 close to zero). The second parameter features small gradual changes over time (ψ 21 = 0.1), but also larger abrupt breaks (ψ 20 = 0.5). The third parameter is similar to the first coefficient, but assumes ψ 30 > ψ 10 . The fourth coefficient is assumed to be constant over time (ψ 04 and ψ 14 are both close to zero) and, finally, the fifth parameter features some extremely large breaks (with ψ 05 = 1), while it is otherwise assumed to be constant (here achieved through the random walk state equation with ψ 51 close to zero). To assess the flexibility of our approaches we compare them to models assuming a standard random walk evolution of coefficients and to those assuming constant coefficients. Moreover, we The results with artificial data reveal at least three important features. First, all TVP models yield reasonable estimates for constant coefficients, which is most important for forecasting applications. Focussing on the the fourth parameter, considering a more flexible model pays off to produce less biased and more precise estimates, especially when compared to the constant coefficient model. Second, the TVP-MIX specifications are capable in capturing both rapid shifts and smooth adjustments in the regression coefficients. Our methods in panel (a) and (b) tend to quickly adjust when facing high frequency changes, rendering the methods even more flexible when compared to a typical mixture innovation specification in panel (d). Third, the TVP-POOL model in panel (c) tends to detect sudden changes in the parameters quite well, but is less capable in capturing low frequency movements. This feature differs form a standard random walk evolution assumption on the TVPs. Assuming a standard random walk implies smoothly evolving coefficients makes capturing high frequency changes difficult. Interestingly, the time-varying intercept (the fifth coefficients) tends to soak up movements of other parameters. Models that do not truly detect the large breaks of the third coefficient are particularly prone to this issue (TVP-POOL but also TVP-RW specifications). The gray shaded areas refer to the 68% credible sets of a standard TVP regression with random walk state equation. The black dotted lines indicate the 16 th /50 th /84 th percentiles of a constant coefficient model. Moreover, the red lines denote the true coefficients of α t . EMPIRICAL APPLICATION Structural analysis and forecasting key macroeconomic indicators is of great relevance for policy makers. In the empirical work, we focus on output growth, inflation, unemployment, and/or the interest rate. Focussing on these variables we investigate the merits of our approach by using the popular quarterly US data described in McCracken and Ng (2016). The data set includes 165 macroeconomic and financial variables and ranges from 1959:Q1 to 2019:Q4. 8 In Subsection 6.1 we show some stylized in-sample features of our methods for a small-scale model. By including the four target variables in a small-scale VAR (henceforth S-VAR) we present posterior probabilities of the state indicator matrix S t and estimate the low-frequency relationship between unemployment and inflation. Moreover, in Subsection 6.2, this variable set forms the basis for evaluating the predictive performance of our methods in a comprehensive forecast exercise. For the forecasting exercise, we consider two additional information sets. In our largest specification (L-VAR) we pick 20 macroeconomic indicators, which are commonly considered by the recent literature for forecasting (see, for example, Huber et al., 2020a;Pfarrhofer, 2020). 9 In particular, we include financial market indicators that carry important information about the future stance of the economy (see Bańbura et al., 2010). Moreover, we consider a factor-augmented VAR (FA-VAR). Here, we augment the target variables with six principal components compromising information of the remaining variables in the data set, effectively leading to VAR with ten endogenous variables. 10 In such larger scale-models our methods are capable of handling less frequent (but important) parameter instabilities in a genuine way. Especially forecasting these important macroeconomic aggregates remains a challenging task, since (at least) two issues arise. First, we have to decide on a set of variables, which we want to include in our econometric model. The recent literature on constant parameter VARs highlights that exploiting large information sets yields forecast gains (see, for example, Bańbura et al., 2010;Koop, 2013). Second, it is well documented that important economic indicators feature instabilities in structural parameters and innovation volatilities. 11 In the literature there is strong agreement that SV is important in macroeconomic applications (see Clark, 2011). There is also strong empirical support for shifting parameters in small-scale models (see D' Agostino et al., 2013). However, there is less consensus for time-varying parameters in larger-scale models. With increasing amount of information overall time-variation in parameters tends to reduce. Recent contributions dealing with 8 In the empirical application we start with 1962:Q1 and use the first observations for transformations. 9 In Appendix C we provide further details on the specific variable set, included in the largest specification, and the transformation applied. 10 The number of principal components is motivated by the specification in Stock and Watson (2012), who also consider six factors. 11 See, for example Stock and Watson (2012), Ng and Wright (2013) and Aastveit et al. (2017), which put special emphasis on the recent financial crisis. large-scale TVP-VARs argue that in smaller models the TVP part controls for an omitted variable bias (see Feldkircher et al., 2017;Huber et al., 2020b). 12 In the following empirical application, note thate we consider two lags for every model and allow for SV. In-sample evidence Before proceeding, we briefly elaborate on a potential identification problem when interpreting the state indicators S t (see Frühwirth-Schnatter, 2001). For the TVP-MIX models, identification is ensured by construction (if coefficients indeed feature time variation). Assuming φ t = S t (see Equation 3) automatically imposes inequality constraints on the autoregressive coefficients in the state equation. However, non-identifiability can occur when coefficients are constant. In such a case, elements in S t are hard to interpret, since a no change evolution is supported by both a random walk and a white noise process. Interpreting S t for the TVP-POOL specification is an even more challenging task, since in these models S t solely controls the evolution of state innovations. Here, inference about the state indicator matrix is only useful in combination with inference about the size of state innovation variancesΨ 0 andΨ 1 and with imposing an inequality restriction ex-post (for example,ψ i0 <ψ i1 ). Therefore, we solely focus on two variants of a TVP-MIX model to illustrate the switching behaviour. Figure depicts a specification with elements in S t following an independent mixture distribution (MIX). A comparison between both approaches highlights that a joint indicator evidently leads to a different posterior median of S t than covariate-specific indicators. By restricting S t = s t I K , all covariates are driven solely by a single indicator that pushes all covariates towards either a random walk or white noise state equation in period t. Conversely, with covariate-specific indicators, we see more dispersion across covariates. However, both approaches agree on a white noise state equation in times of turmoil, suggesting a need for abruptly adjusting parameters in these periods. This model feature is in line with the discussion in Primiceri (2005), who suggests that an economically stable period favours more gradual changes (which are more consistent with a random walk state equation) in the coefficients, while shifts in policy rules require quickly adjusting coefficients (which is better captured by using a white noise state equation). 12 Since estimating TVP models with typical MCMC methods remains computationally demanding, several studies take this argument as a reason to opt for approximating the TVP part or rely on dimension reduction techniques, yielding fast inference while accepting a certain risk of misspecification (see, inter alia Eisenstat et al., 2019;Korobilis, 2019;Hauzenberger et al., 2020;Huber et al., 2020b;Korobilis and Koop, 2020). To further illustrate the proposed methods, we estimate the low-frequency relationship between unemployment and inflation. This low-frequency measure corresponds to a long-run coefficient of distributed-lag regression models (Whiteman, 1984) and disentangles systematic co-movements from short-run fluctuations. 13 Panel (a) to (c) in Figure 3 Overall, considering TVP-MIX methods seem to improve the economic interpretability of the lowfrequency component, while a TVP-POOL model aggressively pushes coefficients towards a constant evolution, which could pay off for forecasting. Forecasting evidence In the forecast exercise we consider a wide range of models varying along the evolution assumption of parameters and the information set considered. With respect to the evolution of parameters, it proves convenient to summarize the different specifications (see Table 1). The models differ along three dimensions: the autoregressive parameters φ t , the innovation variances Ψ t and the state indicator matrix S t . First, our main specifications vary between a model that assumes a binary indicator matrix on the autoregressive parameter with φ t = S t (labeled as TVP-MIX) and a model that introduces a hierarchical prior on the TVP-part (a) S t = s t I K with s t following an MS process: (b) Elements in S t follow an independent mixture specification: . . , s Kt ) e.g. Huber et al. (2019) All these TVP models feature a Normal-Gamma (Griffin and Brown, 2010) onα. 14 We compare our methods to two constant parameter models. 15 One variant features a Normal-Gamma (const. (NG)) prior, while the second variant assumes a Minnesota (const. (MIN)) prior. We consider a non-conjugate Minnesota prior, capturing the notion that own lags are more important than lags from other variables (Doan et al., 1984;Litterman, 1986). We estimate this set of models for three information sets (FA-VAR, L-VAR and S-VAR) with each featuring a different number of endogenous variables. Every considered specification features two lags and SV. To asses one-quarter-, one-year-and two-year-ahead predictions, we treat observations ranging from 1962:Q1 to 1999:Q4 as an initial sample and the periods from 2000:Q1 to 2019:Q4 as a hold-out sample. The initial sample is then recursively expanded until the penultimate quarter (2019:Q3) is reached. For each forecast comparison, a small-scale Minnesota VAR with constant parameters (S-VAR const. (MIN)) serves as our benchmark. In the following, Table 2 shows the best performing models for point and density forecasts, being a tractable summary of Table 3 and Table 4. Table 3 depicts root-mean squared error ratios (RMSEs) as point forecast measures and Table 4 the log predictive Bayes factors (LPBFs) as density forecast metrics. The best performing models within each column are indicated by bold numbers. In Table B.1 we provide additional results on continuous rank probability score (CRPS) ratios. This alternative density forecast measure is more robust to outliers than log predictive scores (Gneiting and Raftery, 2007 at three different horizons we obtain a comprehensive picture to evaluate our methods jointly and marginally along the four target variables. Table 2: Overview of the best performing models, indicated by bold numbers in Table 3 and Table 4. indicating that an increasing amount of information pays off for forecasting (see Bańbura et al., 2010). One exception is inflation. For inflation, flexible S-VARs yield more accurate forecasts than FA-VARs and L-VARs for one-quarter-and one-year-ahead point forecasts and one-quarter-ahead density forecasts. Comparing FA-VARs with L-VARs, the results are mixed. One pattern worth noting is that L-VARs tend to outperform FA-VARs for the one-quarter-ahead horizon while the picture reverses for higher-order forecasts. Second, with respect to parameter changes we see that the TVP-POOL specifications forecast particularly well across all horizons and target variables. These models substantially improve upon a wide range of benchmarks. Overall, Table 2 shows that all TVP classes that provide accurate point predictions generally also perform well in terms of density forecasts. Table 4: Density forecast performance (LPBFs) relative to the benchmark (const (Min.)). The red shaded row denotes the benchmark (and its LPS values). Asterisks indicate statistical significance for each model relative to const (Min.) at the 1 ( * * * ), 5 ( * * ) and 10 ( * ) percent significance levels. When examining Table 3 and Table 4 in greater detail, note that a large number of models shown in the tables outperform the Minnesota benchmark in terms of RMSEs (indicated by ratios below one) and in terms of LPBFs (indicated by values above zero). However, the benchmark is a tough competitor when predicting inflation and for higher-order point forecasts. When focussing on the differences occuring through the varying treatment of parameter evolutions, our proposed methods, the TVP-POOL and TVP-MIX specifications, show that their good performance is mainly driven by improved forecast accuracy for output growth and unemployment. In terms of the innovation variance assumption for these specifications, we observe that additional flexibility tends to improve density forecasts performance and yields accurate point forecasts. For the TVP-POOL models this higher degree of flexibility generally pays off across variables and model sizes. For flexible TVP-MIX specifications, forecast ability tends to improve for S-VARs and FA-VARs and is competitive for the L-VARs. Especially the TVP-MIX SSVS MIX and TVP-MIX FLEX MIX models using a small information set yield quite accurate inflation forecasts, being the best performing models for the one-quarter-ahead horizon. Across variables, a notable exceptions is the interest rate for TVP-MIX models. Here, a TVP-MIX SINGLE specification is superior to models assuming a mixture on innovation volatilities. When assessing random walk state equation (TVP-RW) specifications across the information sets, two things are worth noting. First, a standard TVP model with random walk assumption TVP-RW SINGLE is only competitive for one-year-and two-year-ahead forecasts and otherwise forecasts poorly. Second, more flexible TVP-RW variants produce quite accurate forecasts for FA-VARs. Constant parameter models with a Normal-Gamma prior show reasonable forecasts for L-VARs (especially for inflation), but lack flexibility in smaller-scale models. This observation is in line with the fact that in larger-scale model, time-variation in coefficients vanishes (see Huber et al., 2020b). However, few parameter instabilities might still be present since, apart from some exceptions, our methods provide improvements when compared to constant coefficient models. To illustrate the forecast performance over time, Figure 4 depicts the evolution of cumulated joint LPBFs relative to our benchmark. Overall we find that for all four target variable jointly our already leads to a quite diverse model performance across information sets. During this episode and its consecutive three years, L-VARs strictly dominate the other two information sets (FA-VARs and S-VARs). This implies, that for any TVP evolution assumption, the large-scale model outperforms its smaller-scale counterparts. Moreover, during the financial crisis, we observe a substantial increase in LPBFs for a wide range of FA-VARs and L-VARs, while for S-VARs we see similar improvements solely for some TVP-VARs. This feature might indicate that TVPs are capable of mitigating a potential omitted variable bias (Huber et al., 2020b). Second, within each information set, performance across parameter evolution assumptions is A.1. Stochastic volatility specification A stochastic volatility specifications assumes that h t = log(σ 2 t ) follows an AR(1)-process: Following Kastner and Frühwirth-Schnatter (2014), we assume Gaussian priors on the initial state h and the unconditional mean µ h ∼ N (0, 100), a Beta prior on the autoregressive parameter ψ h +1 2 ∼ B(25, 1.5) and a Gamma prior on the state variance ψ h ∼ G(1/2, 1/2). This quite informative prior on ψ h pushes the specification towards a random walk. A.2. The Normal-Gamma prior (Griffin and Brown, 2010) Similar to Bitto and Frühwirth-Schnatter (2019), we introduce class-specific global shrinkage parameters, differentiating between the constant part of the coefficients (labeled λ a ) and regime-switching variances (labeled λ ψ 0 and λ ψ 1 , respectively). In the following, specify τ j \|λ j ∼ G( j , j λ j /2) and λ j ∼ G(ζ, ζ) with λ j = λ k and j = k if j ∈ P k for k = {a, ψ 0 , ψ 1 }. P k denotes a classifier (i.e. defines the set of coefficients belonging to the k th group). In the following, Moreover, we learn the hyperparameter k in a fully Bayesian fashion and specify ζ = 0.01. A.3. Detailed MCMC algorithm In this section, we provide details on each sampling step of the MCMC algorithm and on the full conditional posterior distributions. After defining appropriate starting values, we iterate through the following steps 20, 000 times and discard the first 10, 000 draws as burn-in: 1. The sampling steps (and conditional posteriors) forα t , λ k , τ j , for k = {a, ψ 0 , ψ 1 } and j = 1, . . . , 3K and k are of standard form (Griffin and Brown, 2010): (a) Drawα from a multivariate Gaussian distribution: Here,X is a T × 3K-dimensional matrix withx t on the t th position and: (b) Sample the local shrinkage scalings {τ j } 3K j=1 from a generalized inverse Gaussian (GIG) distribution (Griffin and Brown, 2010): Here, λ j = λ k and j = k if j ∈ P k with k = {a, ψ 0 , ψ 1 }. (c) Sample the associated global shrinkage parameter λ k , for k = {a, ψ 0 , ψ 1 }, from a Gamma distribution distribution: with p k denoting the cardinality of the set P k (see Appendix A.2). 2. Draw the normalized latent statesα from a ν-dimensional Gaussian distribution by exploiting the static representation (see Section 4). 4. Update binary indicators in S t , depending on its law of motion. We recast state equation back in the centered parameterization and evaluate the following regime-switching specification: (a) Conditional on the other parameters in Equation A.2, we follow Kim and Nelson (1999) and sample {s t } T t=1 using standard algorithms. A.4. The spectral decomposition To obtain a time-varying low-frequency measure between two endogenous variable, we follow Sargent and Surico (2011) and Kliem et al. (2016). We therfore recast a TVP-VAR model in its companion form: In the following, the spectral density of Y t at the very low frequency ρ = 0 is given by: For ρ = 0 the low-frequency relationship π ij,t between two variables (Y it , Y jt ) ∈ Y t can be derived with: with Π ij,t denoting the (i, j) th element in Π t . C. DATA In this section we provide further details on the variable used for the large-scale VAR (L-VAR). For the factor-augmented VAR (FA-VAR) we consider the full data set, compromising 165 variables. For brevity we refer to McCracken and Ng (2016) for a detailed description and transformation codes. All variables, serving as a basis for the principal components, are transformed to stationarity as suggested in McCracken and Ng (2016). Finally, we standardise the data by demeaning each variable and dividing through the standard deviation. Especially for principal components standardising is important due to the scale variance of the components.	2020-06-19T01:00:41.118Z	2020-06-17T00:00:00.000	{ "year": 2020, "sha1": "d04333dffbe8fe331c58e6d5a99de868527a88ca", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "d04333dffbe8fe331c58e6d5a99de868527a88ca", "s2fieldsofstudy": [ "Mathematics" ], "extfieldsofstudy": [ "Economics", "Mathematics" ] }
119306646	pes2o/s2orc	v3-fos-license	The color gauge invariance and a possible origin of the Jaffe-Witten mass gap in QCD The physical meaning of a mass gap introduced by Jaffe and Witten is to be responsible for the large-scale (low-energy/momentum), i.e., the non-perturbative structure of the true QCD vacuum. In order to make the existence of a mass gap pefrectly clear it is defined as the difference between the regularized full gluon self-energy and its subtracted (also regularized) counterpart. The mass gap is mainly generated by the nonlinear interaction of massless gluon modes. A self-consistent violation of SU(3) color gauge invariance/symmetry is duscussed in order to realize a mass gap in QCD. For this purpose, we propose not to impose the transversality condition on the full gluon self-energy, while restoring the transversality of the full gluon propagtor relevant for the non-perturbative QCD at the final stage. At the same time, the Slavnov-Taylor identity for the full gluon propagator is always preserved. All this allows one to establish the general structure of the full gluon propagator in the presence of a mass gap. In this case, two independent types of formal solutions for the full gluon propagator have been established. The nonlinear iteration solution at which the gluons remain massless is explicitly present. The existence of the solution with an effective gluon mass is also demonstrated. I. INTRODUCTION Quantum Chromodynamics (QCD) [1,2] is widely accepted as a realistic quantum field gauge theory of strong interactions not only at the fundamental (microscopic) quark-gluon level but at the hadronic (macroscopic) level as well. This means that in principle it should describe the properties of experimentally observed hadrons in terms of experimentally never seen quarks and gluons, i.e., to describe the hadronic world from first principles -an ultimate goal of any fundamental theory. But this is a formidable task because of the color confinement phenomenon, the dynamical mechanism of which is not yet understood, and therefore the confinement problem remains unsolved up to the present days. It prevents colored quarks and gluons to be experimentally detected as physical ("in" and "out" asymptotic) states which are colorless (i.e., color-singlets), by definition, so color confinement is permanent and absolute [1]. Today there is no doubt left that color confinement and other dynamical effects, such as spontaneous breakdown of chiral symmetry, bound-state problems, etc., being essentially nonperturbative (NP) effects, are closely related to the large-scale (low-energy/momentum) structure of the true QCD ground state and vice-versa ( [3,4] and references therein). The perturbation theory (PT) methods in general fail to investigate them. If QCD itself is a confining theory then a characteristic scale has to exist. It should be directly responsible for the above-mentioned structure of the true QCD vacuum in the same way as Λ QCD is responsible for the nontrivial perturbative dynamics there (scale violation, asymptotic freedom (AF) [1]). However, the Lagrangian of QCD [1,2] does not contain explicitly any of the mass scale parameters which could have a physical meaning even after the corresponding renormalization program is performed. The main goal of this paper is to show how the characteristic scale (the mass gap, for simplicity) responsible for the NP dynamics may explicitly appear in QCD. This becomes an imperative especially after Jaffe and Witten have formulated their theorem "Yang-Mills Existence And Mass Gap" [5]. In order to make the existence of a mass gap perfectly clear it is defined as the difference between the regularized full gluon self-energy and its subtracted (also regularized) counterpart. The mass gap is mainly generated by the nonlinear (NL) interaction of massless gluon modes. Our proposal to realize a mass gap in QCD is not to impose the transversality condition on the full gluon self-energy, while restoring the transversality of the full gluon propagator relevant for NP QCD at the final stage. At the same time, the Slavnov-Taylor (ST) identity for the full gluon propagator (important for the renormalization) is always preserved within our approach. As mentioned above, there is no place for the mass gap in the QCD Lagrangian, so the only place when the mass gap may appear is the corresponding system of dynamical equations of motion, the so-called Schwinger-Dyson (SD) equations, which should be complemented by the ST identities ( [1] and references therein). The propagation of gluons is one of the main dynamical effects in the true QCD vacuum. It is described by the above-mentioned corresponding SD quantum equation of motion for the full gluon propagator. The importance of this equation is due to the fact that its solutions reflect the quantum-dynamical structure of the true QCD ground state. The color gauge structure of this equation is also one of the main subjects of our investigation in order to find a way how to realize a mass gap in QCD. In the presence of a mass gap two different and independent types of formal solutions for the regularized full gluon propagator have been established. The general nonlinear iteration solution is always severely singular at small gluon momentum, i.e., the gluons remain massless, and this does not depend on the gauge choice. The massive solution leads to an effective gluon mass, which depends on the gauge choice. No truncations/approximations/assumptions and no special gauge choice are made for the skeleton loop integrals (contributing to the full gluon self-energy) in order to show the existence of these solutions. Both solutions are valid in the whole energy/momentum range, and thus they should preserve AF when the gluon momentum goes to infinity. II. QED It is instructive to begin with a brief explanation why a mass gap does not occur in quantum electrodynamics (QED). The photon SD equation can be symbolically written down as follows: where we omit, for convenience, the dependence on the Dirac indices, and D 0 (q) is the free photon propagator. Π(q) describes the electron skeleton loop contribution to the photon self-energy (the so-called vacuum polarization tensor). Analytically it looks Π(q) ≡ Π µν (q) = −g 2 id 4 p (2π) 4 T r[γ µ S(p − q)Γ ν (p − q, q)S(p)], (2.2) where S(p) and Γ µ (p − q, q) represent the full electron propagator and the full electron-photon vertex, respectively. Here and everywhere below the signature is Euclidean, since it implies q i → 0 when q 2 → 0 and vice-versa. This tensor has the dimensions of a mass squared, and it is quadratically divergent at least in the PT. To make the formal existence of a mass gap perfectly clear, let us now, for simplicity, subtract its value at zero. One obtains Π s (q) ≡ Π s µν (q) = Π µν (q) − Π µν (0) = Π µν (q) − δ µν ∆ 2 (λ). (2.3) The explicit dependence on the dimensionless ultraviolet (UV) regulating parameter λ has been introduced into the mass gap ∆ 2 (λ), given by the integral (2.2) at q 2 = 0, in order to assign a mathematical meaning to it. In this connection a few remarks are in order in advance. The dependence on λ (when it is not shown explicitly) is assumed in all divergent integrals here and below in the case of the gluon self-energy as well (see next section). This means that all the expressions are regularized (including photon/gluon propagator), and we can operate with them as with finite quantities. λ should be removed on the final stage only after performing the corresponding renormalization program. So through this paper the mass gap is a "bare" one, i.e., it is only regularized. Whether the regulating parameter λ has been introduced in a gauge-invariant way (though this always can be achieved) or not, and how it should be removed is not important for the problem if a mass gap can be "released/liberated" from the corresponding vacuum. We will show in the most general way (not using the PT and not choosing any special gauge) that this is impossible in QED and might be possible in QCD. The decomposition of the subtracted vacuum polarization tensor (2.3) into the independent tensor structures is where both invariant functions Π s (q 2 ) andΠ s (q 2 ) are, by definition, dimensionless and regular at small q 2 , since Π s µν (0) = 0 identically due to the subtraction (2.3); otherwise they remain arbitrary. From this relation it follows that Π s µν (q) = O(q 2 ), i.e., it is always of the order q 2 . In this connection a few remarks are in order. The subtraction (2.3) at zero point in QED is justified, since it is abelian gauge theory, and therefore there is no the self-interaction of massless photons, which can be source of the singularities in the q 2 → 0 limit. The vacuum polarization tensor (2.2) has no infrared (IR) singularities in this limit, at least in the PT. So in what follows we will consider the above-mentioned invariant functions as regular at small q 2 , indeed. Also, here and everywhere below (2.5) In the same way, the photon self-energy (2.2) in terms of the independent tensor structures is where both invariant functions Π(q 2 ) andΠ(q 2 ) are dimensionless functions; otherwise they remain arbitrary. Due to the transversality of the photon self-energy which comes from the current conservation condition in QED, one hasΠ(q 2 ) = 0, i.e., it should be the purely transversal On the other hand, from the subtraction (2.3), on account of the relation (2.4), and the transversality condition (2.7) it follows thatΠ which, however, is impossible sinceΠ s (q 2 ) is a regular function of q 2 , by definition. So the mass gap should be discarded, i.e., put formally to zero and, consequently,Π s (q 2 ) as well, i.e., Thus the subtracted photon self-energy is also transversal, i.e., it satisfies the transversality condition q µ Π µν (q) = q µ Π s µν (q) = 0, (2.11) and coincides with the photon self-energy (see Eq. (2.3) at zero mass gap). Moreover, this means that the photon self-energy does not have a pole in the q 2 → 0 limit in its invariant function Π(q 2 ) = Π s (q 2 ). As mentioned above, in obtaining these results neither the PT has been used nor a special gauge has been chosen. So there is no place for quadratically divergent constants in QED, while logarithmic divergence still can be present in the invariant function Π(q 2 ) = Π s (q 2 ). It is to be included into the electric charge through the corresponding renormalization program (for these detailed gauge-invariant derivations explicitly done in lower order of the PT see Refs. [2,6,7,8,9]). Taking into account the subtraction (2.3), on account of the relations (2.10), the photon SD equation (2.1) becomes equivalent to (2.12) It can be summed up into geometric series, so one obtains Since Π s (q) = O(q 2 ) and D 0 (q) ∼ (q 2 ) −1 , the IR singularity of the full photon propagator is determined by the IR singularity of the free photon propagator, i.e., D(q) = O(D 0 (q)) with respect to the behavior at small photon momentum. In fact, the current conservation condition (2.7), i.e., the transversality of the photon self-energy lowers the quadratic divergence of the corresponding integral (2.2) to a logarithmic one. That is the reason why in QED only logarithmic divergences survive. Thus in QED there is no mass gap and the relevant photon SD equation is shown in Eq. (2.13). From it follows that the behavior of the full gluon propagator at small gluon momentum is determined by the behavior of its free PT counterpart. In other words, in QED we can replace Π(q) by its subtracted counterpart Π s (q) from the very beginning (Π(q) → Π s (q)), totally discarding the quadratically divergent constant ∆ 2 (λ) from all the equations and relations. The current conservation condition for the photon self-energy (2.7), i.e., its transversality, and the condition q µ q ν D µν (q) = iξ (here and everywhere below ξ is the gauge-fixing parameter) imposed on the full gluon propagator are consequences of gauge invariance. They should be maintained at every stage of the calculations, since the photon is a physical state. In other words, at all stages the current conservation plays a crucial role in extracting physical information from the S-matrix elements in QED, which are usually proportional to the combination j µ 1 (q)D µν (q)j ν 2 (q). The current conservation condition j µ 1 (q)q µ = j ν 2 (q)q ν = 0 implies that the unphysical (longitudinal) component of the full photon propagator does not change the physics of QED, i.e., only its physical (transversal) component is important. In its turn this means that the transversality condition imposed on the photon self-energy is also important, because Π µν (q) itself is a correction to the amplitude of the physical process, for example such as electron-electron scattering. Concluding, let us emphasize that the photon is always a massless state, since in QED (unlike QCD, see below) the mass gap cannot be realized. III. QCD For our purposes just like in QED it is convenient to begin with the general description of the SD equation for the full gluon propagator, and not for its inverse. Symbolically it can be written down as follows: where D 0 µν (q) is the free gluon propagator. Π ρσ (q; D) is the gluon self-energy, and in general it depends on the full gluon propagator due to non-abelian character of QCD (see below as well). Thus the gluon SD equation is highly NL, while the photon SD equation (2.1) is a linear one. In what follows we omit the color group indices, since for the gluon propagator (and hence for its self-energy) they are reduced to the trivial δ-function, for example D ab µν (q) = D µν (q)δ ab . Also, for convenience, we introduce i into the gluon SD equation (3.1). In comparison with the photon self-energy (2.2), the gluon self-energy Π ρσ (q; D) is the sum of a few terms, where Π q ρσ (q) describes the skeleton loop contribution due to quark degrees of freedom (it is an analog of the vacuum polarization tensor (2.2) in QED), while Π gh ρσ (q) describes the skeleton loop contribution due to ghost degrees of freedom. Both skeleton loop integrals do not depend on the full gluon propagator D, so they represent the linear contribution to the gluon SD equation. Π t ρσ (D) represents the so-called constant skeleton tadpole term. Π (1) ρσ (q; D 2 ) represents the skeleton loop contribution, which contains the triple gluon vertices only. Π (2) ρσ (q; D 4 ) and Π (2 ′ ) ρσ (q; D 3 ) describe topologically independent skeleton two-loop contributions, which combine the triple and quartic gluon vertices. The last four terms explicitly contain the full gluon propagators in the corresponding powers symbolically shown above, that is why they form the NL part of the gluon SD equation. The analytical expressions for the corresponding skeleton loop integrals [10] (in which the symmetry coefficients can be included) are of no importance here, since we are not going to introduce into them any truncations/approximations as well as to choose some special gauge. Let us note that like in QED these skeleton loop integrals are quadratically divergent at least in the PT, and therefore they are assumed to be regularized (see remarks above and below). A. Subtractions Quite similar to the subtraction (2.3), let us formally subtract from the full gluon self-energy (3.2) its value at zero point (see, however, remarks below). Thus, one obtains In this connection let us make in advance a few general remarks. Contrary to QED, QCD being non-abelian gauge theory can suffer from the IR singularities in the q 2 → 0 limit due to the self-interaction of massless gluon modes. Thus the initial subtraction at zero point in the definition (3.3) may be dangerous [1], indeed. That is why in all quantities below the dependence on the finite (slightly different from zero) dimensionless subtraction point α is to be understood. In other words, all the subtractions at zero and the Taylor expansions around zero should be understood as the subtractions at α and the Taylor expansions near α, where they are justified to use. From the technical point of view it is convenient to put formally α = 0 in all derivations below, and to restore the explicit dependence on non-zero α in all quantities at the final stage only. However, we will restore the explicit dependence on α, when it will be necessary for better understanding of the corresponding derivations (see section IV below). Let us remind once more that by mass gap we understand the difference between the regularized gluon self-energy and its subtracted (also regularized) counterpart. To demonstrate a possible existence of a mass gap ∆ 2 (λ; D) in QCD, it is not important how λ has been introduced and how it should be removed at the final stage. The mass gap itself is mainly generated by the nonlinear interaction of massless gluon modes, slightly corrected by the linear contributions coming from the quark and ghost degrees of freedom, namely where index "a" runs as follows: a = q, gh, 1, 2, 2 ′ , and here, obviously, the tensor indices are omitted. In these relation all the quadratically divergent constants Π t (D) and Π a (0; D), having the dimensions of a mass squared, are given by the corresponding skeleton loop integrals at q 2 = 0, which appear in Eq. (3.2). In this connection, let us remind that by the quadratic divergences we conventionally understand the divergent constants having the dimensions of a mass squared and summed up into the mass gap (3.4). Then not losing generality, we can put ∆ 2 (λ) = m 2 f (λ), where m 2 is some fixed mass squared, and f (λ) is some dimensionless function. Its dependence on λ is determined by the divergences of the above-mentioned skeleton loop integrals. However, due to AF the dependence is linear one (up to AF logarithm), so the divergence becomes the quadratic one ∆ 2 (λ) ∼ m 2 λ ∼ Λ 2 , indeed, like in the PT. The subtracted gluon self-energy (3.3) is free from the tadpole contribution, because Π s t (D) = Π t (D) − Π t (D) = 0, by definition, at any D, while in the gluon self-energy (3.2) it is explicitly present The general decomposition of the subtracted gluon self-energy into the independent tensor structures can be written down as follows: where both invariant functions Π s (q 2 ; D) andΠ s (q 2 ; D) are dimensionless functions of their argument q 2 . The subtracted gluon self-energy does not contain the tadpole contribution, see Eq. (3.5). Let us note in advance (see subsection B below) that in this case we can impose the color current conservation condition on it, i.e., to put which impliesΠ s (q 2 ; D) = 0. So the subtracted gluon self-energy finally becomes the purely transversal Let us remind once more that we can expand Π s (q 2 ; D) in a Taylor series near the subtraction point α at any D. Thus the subtracted quantities are free from the quadratic divergences, but the logarithmic ones at large q 2 can be still present in Π s (q 2 ; D), like in QED. Concluding, let us note that we are not going to impose the transversality condition on the gluon self-energy (3.6) itself (see below). That is why we need no its decomposition into the independent tensor structures. B. A self-consistent violation of color gauge invariance/symmetry (SCVCGI/S) QCD is SU (3) color gauge invariant theory, but however: (i). Due to color confinement, the gluon (unlike the photon) is not a physical state. Moreover, there is no such physical amplitude to which the gluon self-energy (like the photon self-energy) may directly contribute. For example, quark/quark and quark/antiquark scattering are not a physical processes. (ii). Contrary to the conserved currents in QED, the color conserved currents do not play any role in the extraction of physical information from the S-matrix elements for the corresponding physical processes and quantities in QCD. In other words, not the conserved color currents, but only their color-singlet counterparts, which can even be partially conserved, contribute directly to the S-matrix elements describing this or that physical process/quantity. For example, such an important physical QCD parameter as the pion decay constant is given by the following S-matrix element: is just the axial-vector current, while \|π j (q) > describes the pion boundstate amplitude, and i, j are flavor indices. (iii). Moreover, in QCD (contrary to QED) exists a direct evidence/indication that the transversality of the full gluon self-energy may be violated beyond the PT, indeed. The color gauge invariance condition for the full gluon self-energy (3.2) can be reduced to the three independent transversality conditions imposed on it. It is well known that the quark contribution can be made transversal independently of the pure gluon contributions within any regularization scheme which preserves gauge invariance, for example such as the dimensional regularization method (DRM) [11] (see Refs. [1,2,8,9] as well). So, we can put Explicitly it can be shown in lower order of the PT (see, for example Refs. [2,6,7,8,9]). It is assumed, however, that in principle it should be valid in every order of the PT, thus going beyond the PT. In the same way the sum of the gluon contributions can be made transversal by taking into account the ghost contribution, so again one can put The role of ghost degrees of freedom is to cancel the unphysical (longitudinal) component of gauge boson (gluon) propagator in every order of the PT, i.e., going beyond the PT and thus being the general one. The previous general condition of cancellation (3.11) just demonstrates this, since it contains the corresponding skeleton loop integrals. As in a quark case, the explicit cancellation can be shown, nevertheless, only in lower order of the PT. For this we should put D = D 0 ≡ D 0 and approximate all other quantities, entering the corresponding skeleton loop integrals in the relation (3.11), by their free PT counterparts (see, for example Refs. [2,8,9]). However, there is no such regularization scheme (preserving or not gauge invariance) in which the transversality condition for the full gluon self-energy could be satisfied unless the constant skeleton tadpole term is disregarded from the very beginning (here T 0 4 is the four-gluon point-like vertex and we omit the tensor and color indices, as unimportant for further purpose). It is nothing else but the quadratically divergent in the PT constant. It explicitly violets the transversality condition for the full gluon self-energy, since formally q ρ Π t ρσ (D) = q ρ δ ρσ ∆ 2 t (D) = q σ ∆ 2 t (D) = 0. In the PT, when the full gluon propagator is always approximated by the free one, the constant tadpole term is set to be zero within the DRM [2,8,9], i.e., Π t ρσ (D 0 ) = 0. So in the PT the transversality condition for the full gluon self-energy is always satisfied. However, even in the DRM this is not an exact result, but rather an embarrassing prescription, as pointed out in Ref. [8]. To show explicitly that even in the PT there are still problems, it is instructive to substitute the first iteration of the gluon SD equation (3.1) into the previous expression. Symbolically it looks like where we omit all the indices. Doing so, one obtains Here we introduce the subtraction as follows: where all other quadratically divergent constant terms are omitted, for simplicity. In the second line of Eq. (3.13) the first integral is not only UV divergent but it is IR singular as well. If now we omit the first term in accordance with the above-mentioned prescription, the product of this integral and the tadpole term Π t (D 0 ) remains, nevertheless, undetermined. Moreover, the structure of the second integral in this line is completely different from the divergent constant integral Π t (D 0 ). This constant term is also not determined, since in general we do not know the behavior of Π s (q 2 1 ; D 0 ) at small and large q 2 1 . All this reflects the general problem that all such kind of massless integrals (d d q/(2π) d )(q µ1 ...q µp /(q 2 ) n ) are ill defined, since there is no dimension where they are meaningful; they are either IR singular or UV divergent [8]. This prescription clearly shows that the DRM, preserving gauge invariance, nevertheless, does not alone provide us insights into the correct treatment of the power-like IR singularities (we will address this problem in the second (II) part of our investigation). However, in the PT we can adhere to the prescription that such massless tadpole integrals can be discarded in the DRM [8]. This is the only way for ghosts to validate the transversality condition for the full gluon self-energy in PT QCD. It makes the full gluon propagator the purely transversal. Then the S-matrix elements for physical quantities and processes in PT QCD become free from unphysical degrees of freedom of gauge bosons, maintaining thus the unitarity of S-matrix in this theory. So, we conclude that beyond the PT, i.e., in the general case the transversality of the full gluon self-energy may be violated. In other words, in the general case, i.e., beyond the PT, we cannot discard the tadpole term (3.12) from the very beginning. If we do not know how to treat it, this does not mean that we should neglect it at all. Its regularized version should be explicitly taken into account. Moreover, all other quadratically divergent constants, which have been summed up into the mass gap (3.4), cannot be discarded like in QED, since the transversality condition for the gluon self-energy is not going to be imposed (see below). In other words, in QCD the quadratic divergences of the corresponding skeleton loop integrals cannot be lowered to the logarithmic ones, and therefore the mass gap (3.4) should be explicitly taken into account in this theory. Thus in order to realize the mass gap (3.4) our proposal is not to impose the transversality condition on the gluon self-energy (3.2), i.e., to admit that in general case indeed. At the same time, we would like to preserve the color gauge invariance condition for the full gluon propagator, i.e., within our approach the relation which is nothing else but the ST identity, always holds. This is important for the renormalization. The lesson which comes from QED is that if one preserves the transversality of the photon self-energy at every stage, then there is no mass gap. Thus in order to realize a mass gap in QCD, our proposal is not to impose the transversality condition on the gluon self-energy, Eq. (3.14), but preserving the ST identity for the full gluon propagator, Eq. (3.15). Concluding, a few general remarks are in order: 1. We would like to emphasize the special role of the constant skeleton tadpole term (3.12) in the NP QCD dynamics. Its existence is a direct evidence that the transversality of the gluon self-energy may be violated beyond the PT theory. 2. The second important observation is that the ghosts themselves cannot now automatically provide the transversality of the gluon propagator in NP QCD. Thus, we sacrifice the general role of ghosts in order to realize a mass gap. To realize a mass gap is much more necessary than to maintain the general role of ghosts. At long last, the role of ghosts is mainly kinematical, while the mass gap dominates the dynamics of QCD at large distance (see below). This is important for understanding of the confinement mechanism. 3. However, let us note in advance that how to restore the transversality of the full gluon propagator relevant for NP QCD will be explained in part II of our investigation. In other words, at the initial stage we violate the transversality of the gluon self-energy in order to realize a mass gap, while restoring the transversality of the full gluon propagator relevant for NP QCD at the final stage. At the same time, the ST identity for the gluon propagator is always valid within our approach. All this will make it possible to maintain the unitarity of the S-matrix in NP QCD. 4. The discussion above does not mean that we need no ghosts at all. We need them in other sectors of QCD, for example in the quark-gluon ST identity, which contains the so-called ghost-quark scattering kernel explicitly [1]. It provides an important piece of information on quark degrees of freedom themselves. If one omits the ghosts, then it will be totally lost (for details see Ref. [12,13,14], a recent publication [15] and references therein). 5. The transversality condition for the gluon self-energy can be satisfied partially, i.e., if one imposes it on quark and gluon (along with ghost) degrees of freedom, as it follows from the relations (3.10) and (3.11). Then the mass gap is to be reduced to Π t (D), since all other constants Π a (0; D) can be discarded in this case, see Eq. (3.4). However, we will stick to our proposal not to impose the transversality condition on the gluon self-energy at all, and thus to deal with the mass gap on account of all the possible contributions. C. General structure of the gluon SD equation Our strategy is not to impose the transversality condition on the gluon self-energy in order to realize a mass gap despite whether or not the tadpole term is explicitly present. At the same time, we would like to preserve the ST identity (3.15), as underlined above. It implies that the general tensor decomposition of the full gluon propagator becomes the standard one, namely is the full gluon invariant function (the full gluon form factor or equivalently the full effective charge ("running")). To show that our strategy works, let us substitute the subtraction (3.3), on account of the relation (3.9), into the initial gluon SD equation (3.1). Then one obtains In the presence of a mass gap, it is instructive to introduce the general tensor decomposition of the auxiliary free gluon propagator as follows: and . (3.20) However, we need the standard free gluon propagator rather than its auxiliary counterpart, despite the latter one being reduced to the former one in the formal PT ∆ 2 (λ; D) = 0 limit. To achieve this goal, the auxiliary free gluon propagator defined in Eqs. (3.18) and (3.20) is to be equivalently replaced as follows: where D 0 µν (q) in the right-hand-side is the standard free gluon propagator now, i.e., Then the gluon SD equation in the presence of a mass gap (3.17), after the replacement (3.21) and doing some algebra, on account of the explicit expression for the auxiliary free gluon form factor (3.20), becomes Here and from now on D 0 µν (q) is the standard free gluon propagator (3.22). Using it explicitly, and on account of the decomposition (3.16), this equation can be further simplified to The only price we have paid so far is the gluon self-energy, while its subtracted counterpart is always transversal. At the same time, the full gluon propagator (3.16) and the free PT gluon propagator (3.22) automatically satisfy the ST identity (3.15). So, one can conclude that our mechanism for the realization of the mass gap is rather self-consistent. Let us emphasize that the expression for the full gluon form factor shown in the relation (3.19) cannot be considered as the formal solution for the full gluon propagator D (see Eq. (3.16)), since both the mass gap ∆ 2 (λ; D) and the invariant function Π s (q 2 ; D) depend on D themselves. It clearly follows from the relation (3.19) that the effect of the mass gap dominates the IR region when the gluon momentum goes to zero, and this effect vanishes when the gluon momentum goes to infinity. This once more underlines a close intrinsic link between the NP dynamics governed by the mass gap and the structure of the true QCD vacuum at large distances (q 2 → 0). It is worth recalling once more that in the opposite limit, i.e., at large q 2 , the subtracted gluon self-energy Π s (q 2 ; D) may still suffer from the logarithmic divergences, like in QED. In the formal PT ∆ 2 (λ; D) = 0 limit, from the gluon SD equation (3.23) one recovers the standard gluon SD equation (3.1), and the gluon self-energy coincides with its subtracted counterpart like in QED, see Eq. (3.3). Then the formal "solution" (3.19) will not depend on the mass gap. The above-mentioned general role of ghosts is to be automatically restored. They will again provide the cancellation of the longitudinal component of the full gluon propagator in this limit. Concluding, we have established the general structure of the full gluon propagator in the presence of a mass gap. Moreover, we have explicitly shown that the initial gluon SD equation (3.17) has the same "solution" as the final gluon SD equation (3.23), that is Eq. (3.16) along with the relation (3.19). IV. NONLINEAR ITERATION SOLUTION In order to find a formal solution for the regularized full gluon propagator (3.16), on account of its effective charge (3.19) (or equivalently the full gluon form factor), let us rewrite the latter one in the form of the corresponding transcendental (i.e., not algebraic) equation, namely suitable for the formal nonlinear iteration procedure. Here we replace the dependence on D by the equivalent dependence on d. Also, for simplicity, we replaced Π s (q 2 ; d) → Π(q 2 ; d). For future purposes, it is convenient to introduce short-hand notations as follows: and In these relations ∆ 2 m are the auxiliary mass squared parameters, while ∆ 2 ≡ ∆ 2 (λ; d) is the mass gap itself. Via the corresponding subscripts the dimensionless constants c m depend on which iteration for the gluon form factor d is actually done. They may depend on the dimensionless coupling constant squared g 2 , as well as on the gaugefixing parameter ξ. We also introduce in advance the explicit dependence on the finite (slightly different from zero) dimensionless subtraction point α, as pointed out above. The dependence of ∆ 2 on all these parameters, as well as on the number of different flavors N f and colors N c , is not shown explicitly, and if necessary it can be restored any time. Let us also recall that all the invariant functions Π m (q 2 ) can be expand in a formal Taylor series near the finite subtraction point α. If it were possible to express the full gluon form factor d(q 2 ) in terms of these quantities then it would be the formal solution for the full gluon propagator. In fact, this is nothing but the skeleton loops expansion, since the regularized skeleton loop integrals, contributing to the gluon self-energy, have to be iterated. This is the so-called general nonlinear iteration solution. As mentioned above, no truncations/approximations/assumptions and no special gauge choice have been made. This formal expansion is not a PT series. The magnitude of the coupling constant squared and the dependence of the regularized skeleton loop integrals on it is completely arbitrary. Let us emphasize once more that through this paper the mass gap ∆ 2 ≡ ∆ 2 (λ, α, ξ, g 2 ) is a "bare" one, i.e., it is only regularized (with the help of λ and α) in order to assign a mathematical meaning to all derivations involving it. It is instructive to describe the general iteration procedure in some details. Evidently, d (0) = 1, and this corresponds to the approximation of the full gluon propagator by its free counterpart. Doing the first iteration in Eq. (4.1), one thus obtains where obviously d (1) (q 2 ) = −P 0 (q 2 ). (4.5) Carrying out the second iteration, one gets where Doing the third iteration, one further obtains where and so on for the next iterations. Thus up to the third iteration, one finally arrives at We restrict ourselves by the iterated gluon form factor up to the third term, since this already allows to show explicitly some general features of the nonlinear iteration solution. A. Splitting/shifting procedure Doing some tedious algebra, the previous expression (4.10) can be rewritten as follows: This formal expansion contains three different types of terms. The first type are the terms which contain only different combinations of Π m (q 2 ) (they are not multiplied by inverse powers of q 2 ); the third type of terms contains only different combinations of (∆ 2 m /q 2 ). The second type of terms contains the so-called mixed terms, containing the first and third types of terms in different combinations. The two last types of terms are multiplied by the corresponding powers of 1/q 2 . Such structure of terms will be present in each iteration term for the full gluon form factor. However, any of the mixed terms can be split exactly into the first and third types of terms. For this purpose the formal Taylor expansions for Π m (q 2 ) around the finite subtraction point α should be used. Thus an exact IR structure of the full gluon form factor (which just is our primary goal to establish) is determined not only by the third type of terms. It gains contributions from the mixed terms as well, but without changing its functional dependence (see remarks below). To demonstrate this in some detail, it is convenient to express the previous expansion (4.11) in terms of dimensionless variable and parameters, namely where M 2 is some auxiliary fixed mass squared, and µ 2 is the point close to q 2 = 0 (to be not mixed up with the tensor index). Also, in the formal PT ∆ 2 = 0 limit c ≡ c(λ, α, ξ, g 2 ) = 0 (unlike c m in the relations (4.2)), since M 2 is fixed. Using further relations (4.2), and on account of the relations (4.12), the expansion (4.11) becomes Taking into account the above-mentioned formal Taylor expansions for example, the mixed term (c 1 c/x)Π 2 (x) can be then exactly split/decomposed as follows: Here and below the dependence on all other possible parameters is not shown, for simplicity. The dimensionless function O 2 (x) is of the order x at small x; otherwise it remains arbitrary. The first term now is to be shifted to the third type of terms, while the remaining terms are to be shifted to the first type of terms. All other mixed terms of similar structure should be treated absolutely in the same way. The mixed term (c 1 c 2 c 2 /x 2 )Π 0 (x) can be split as where the dimensionless function O 0 (x) is of the order x at small x; otherwise it remains arbitrary. Again the first two terms should be shifted to the third type of terms, while the last two terms should be shifted to the first type of terms. Similarly to the formal Taylor expansion (4.14), we can write Then, for example the mixed term (c 2 c/x)Π 0 (x)Π 1 (qx) can be split as where the dimensionless function O 01 (x) is of the order x at small x; otherwise it remains arbitrary. Again the first term should be shifted to the third type of terms, while other two terms are to be shifted to the first type of terms. Completing this exact splitting/shifting procedure in the expansion (4.13), and restoring the explicit dependence on the dimensional variable and parameters (4.12), one can equivalently present the initial expansion (4.11) as follows: where we use notations (4.2) explicitly now, since the coefficients of the above-used expansions depend in general on the same set of parameters: λ, α, ξ, g 2 . The invariant function d 3 (q 2 ; ∆ 2 ) is dimensionless and it is free from the power-type IR singularities; otherwise it remains arbitrary. We have restored the dependence on the mass gap ∆ 2 instead the dependence on the parameter c. In the formal PT ∆ 2 = 0 limit it survives, and is to be reduced to the sum of the first type of terms in the expansion (4.11). The generalization on the next iterations is almost obvious. Concluding, let us underline that the splitting/shifting procedure does not change the structure of the nonlinear iteration solution at small q 2 . It only changes the coefficients at inverse powers of q 2 in the corresponding expansion. In other words, it makes it possible to rearrange the terms in the initial expansion (4.11) in order to get it in the final form (4.19). Also, in the q 2 → 0 limit, it is legitimated to suppress the subtracted gluon self-energy in comparison with the mass gap term in the initial Eq. (4.1). Nevertheless, as a result of the splitting/shifting procedure, which becomes almost trivial in this case, one will obtain the same expansion (4.19) with only different residues, as just mentioned above, and with d 3 (q 2 ; ∆ 2 ) = 1 in this case. B. The exact structure of the general nonlinear iteration solution Substituting the generalization of the expansion (4.19) on all iterations and doing some algebra, the general nonlinear iteration solution for the regularized full gluon propagator (3.16) can be exactly decomposed as the sum of the two principally different terms as follows: (4.20) so that D µν (q) ≡ D µν (q; ∆ 2 ). Here In Eqs. (4.20) and (4.21) the superscript "INP" stands for the intrinsically NP part of the full gluon propagator. Let us emphasize that the general problem of convergence of formal (but regularized) series in Eq.(4.20) is irrelevant here. In other words, it does not make any sense to discuss the convergence of such kind of series before the renormalization program is performed (which will allow to see whether the mass gap survives it at all or not). Anyway, the problem how to remove the UV overlapping divergences (Ref. [16] and for a more detail remarks see appendix C) and usual overall ones [1,2,7,8,9] is a standard one. Our problem will be how to deal with severe IR singularities due to their novelty and genuine NP character. Fortunately, there already exists a well-elaborated mathematical formalism for this purpose, namely the distribution theory (DT) [17] into which the DRM [11] should be correctly implemented (see also Refs. [18,19]). It will be explicitly used in part II of our investigation. We distinguish between the two terms in Eq. (4.20) first of all by the character of the corresponding IR singularities, when the gluon momentum goes to zero. Secondly, by the explicit presence of the mass gap, when it formally goes to zero then the INP term vanishes, while the PT term survives. The third point is that the INP part depends only on the transversal degrees of freedom of gauge bosons, while the PT part has the longitudinal component as well. The INP part of the full gluon propagator is characterized by the presence of severe power-type (or equivalently NP) IR singularities (q 2 ) −2−k , k = 0, 1, 2, 3, .... So these IR singularities are defined as more singular than the power-type IR singularity of the free gluon propagator (q 2 ) −1 , which thus can be defined as the PT IR singularity. The INP part of the full gluon propagator (4.21), apart from the structure (∆ 2 /q 4 ), is nothing but the corresponding Laurent expansion (explicitly shown in Eq. (4.22)) in integer powers of q 2 accompanied by the corresponding powers of the mass gap squared and multiplied by the q 2 -independent factors, the so-called residues Φ k (λ, α, ξ, g 2 ) = ∞ m=0 Φ km (λ, α, ξ, g 2 ). The sum over m indicates that an infinite number of iterations (all iterations) of the corresponding regularized skeleton loop integrals invokes each severe IR singularity labelled by k. It is worth emphasizing that now this Laurent expansion cannot be summed up into anything similar to the initial Eq. (3.19), since its residues at poles gain additional contributions due to the splitting/shifting procedure, i.e., they become arbitrary. However, this arbitrariness is not a problem. The functional dependence, which has been established exactly, is all that matters (a correct treatment of severe IR singularities will be given in part II of our investigation, as mentioned above). It is worth emphasizing once more that just the mass gap term in Eq. (4.1) determines the functional structure of the INP part (4.21)-(4.22) of the full gluon propagator. Thus the true QCD vacuum is really beset with severe IR singularities. Within the general nonlinear iteration solution they should be summarized (accumulated) into the full gluon propagator and effectively correctly described by its structure in the deep IR domain, exactly represented by its INP part (4.21). The second step is to assign a mathematical meaning to the integrals, where such kind of severe IR singularities will explicitly appear, i.e., to define them correctly in the IR region [10,17,18]. Just this violent IR behavior makes QCD as a whole an IR unstable theory, and therefore it may have no IR stable fixed point [1]. This means that QCD itself might be a confining theory without involving some extra degrees of freedom [12,20,21,22,23,24,25,26]. In part II we will show that this is so, indeed. It is instructive to discuss the principal difference between QED and QCD from the point of view of the structures of the full photon and gluon propagators, respectively. In section II it has been explained that the mass gap cannot occur in QED because the photon is a physical state. As a result, the full photon propagator can have only the PT-like IR singularity, i.e., 1/q 2 , see Eq. (2.13). This is in agreement with the cluster property of the Wightman functions, that is, correlation functions of observables in QED [27]. It forbids a more singular behavior of the full photon propagator in the IR than the behavior of its free photon counterpart. In QCD, due to color confinement the gluon is not a physical state. From our investigation then it follows that the mass gap can be realized, and the general nonlinear iteration solution for the full gluon propagator (4.20) becomes inevitably severely IR singular. This is in agreement with the so-called Strocchi theorem [28], which validates such singular structure of the full gluon propagator in the IR. So the existence of a mass gap is a primary reason for a possible violation of the cluster property of the Wightman functions in QCD (to our best knowledge there is no rigorous proof of this property in this theory, see discussion in Ref. [5] as well). At the fundamental quark-gluon level only these remarks about the structure of the Wightman functions in QCD make sense before the solution of the color confinement problem. Concluding, let us note in advance that in the presence of a mass gap there also exists the so-called massive solution. It becomes explicitly smooth at small q 2 in the Landau gauge only (see appendix A). The PT part of the full gluon propagator (4.23), which has the power-type PT IR singularity only, remains undetermined. Its functional structure is due to the unknown in general Π(q 2 ; D) term in Eq. (4.1). This is the price we have paid to fix exactly the functional dependence of the INP part of the full gluon propagator. What we know about the PT effective charge d P T (q 2 ; ∆ 2 ) is that it cannot have the power-type IR singularities; otherwise it remain arbitrary. Also, in the formal PT ∆ 2 = 0 limit it survives, i.e., d P T (q 2 ; ∆ 2 = 0) = d P T (q 2 ). This can be shown directly by restoring the explicit dependence on the mass gap instead the dependence on the coefficient c = (∆ 2 /M 2 ), which appears in the splitting/shifting procedure. According to this procedure, the PT effective charge can be explicitly written down as follows: where A km (q 2 ) are dimensionless functions, which cannot have the power-type IR singularities; otherwise they remain arbitrary (their dependence on the parameters are not shown, for simplicity). The explicit presence of the mass gap (or equivalently the above-mentioned coefficient c) in this expansion just prevents the ghosts to cancel the longitudinal component in the full gluon propagator (4.20). However, in the formal PT ∆ 2 = 0 limit the role of ghosts will be automatically restored, as pointed out above in section III. At the same time, it is worth emphasizing that due to the character of the IR singularity the longitudinal component of the full gluon propagator should be included into its PT part. That is the reason why its INP part becomes automatically transversal, as emphasized above. This also means that the PT part of the full gluon propagator (4.23) contains the free PT gluon propagator, so that we can put d P T (q 2 ) = 1 + d AF (q 2 ), where d AF (q 2 ) should satisfy AF in the q 2 → ∞ limit (see appendix B). Both terms in Eq. (4.20) are valid in the whole energy/momentum range, i.e., they are not asymptotics. At the same time, we have achieved the exact separation between the two terms responsible for the NP (dominating in the IR (q 2 → 0)) and the nontrivial PT (dominating in the UV (q 2 → ∞)) dynamics in the true QCD vacuum. The structure of this solution clearly confirms our conclusion driven above that the deep IR region interesting for confinement and other NP effects is dominated by the presence of a mass gap. In the formal PT ∆ 2 = 0 limit, the nontrivial PT dynamics is all that matters. Let us note in advance that the separation is not only exact but it is unique as well. There exists a special regularization expansion for severe (i.e., NP) IR singularities, while for the PT IR singularity, which is only one present in the PT part of the full gluon propagator (4.23), such kind of expansion does not exist [17,19]. We came to the same structure (4.20)-(4.23) but in a rather different way in Refs. [10,18,19]. In summary, the separation between the INP and PT terms in the full gluon propagator (4.20) is exact and unique. Despite the explicit dependence on the mass gap the gluons remain massless. Moreover, the general nonlinear iteration solution (4.20) is inevitably severely singular in the IR limit (q 2 → 0), and this does not depend on the special gauge choice. V. GENERAL DISCUSSION The mass gap ∆ 2 ≡ ∆ 2 (λ, α, ξ, g 2 ) has not been introduced by hand. It is hidden in the skeleton loop integrals, contributing to the full gluon self-energy. No truncations/approximations/assumptions and no special gauge choice are made for these integrals. An appropriate subtraction scheme has been applied to make the existence of a mass gap perfectly clear. It is dynamically generated mainly by the NL interaction of massless gluon modes. The Lagrangian of QCD does not contain a mass gap, while it explicitly appears in the gluon SD equation of motion. This once more underlines the importance of the investigation of the SD system of equations and the corresponding ST identities ( [1,8,10,29] and references therein) for understanding of the true structure of the QCD ground state. From the relation (3.19) it follows that the mass gap shows up explicitly when the gluon momentum goes to zero. Especially this is clearly seen in the general nonlinear iteration solution (4.20), where it indeed explicitly determines its IR structure. There are no doubts that the mass gap plays a dominant role in the dynamics of QCD at large distances. So the problem how to "liberate/release" it from the QCD vacuum becomes vitally important to understand correctly the mechanism of color confinement. It turned out that for this aim we need to sacrifice the role of ghosts at the initial stage, while restoring the transversality of the full gluon propagator relevant for NP QCD at the final stage. In order to realize a mass gap (more precisely its regularized version), we propose not to impose the transversality condition on the full gluon self-energy, see Eq. (3.14), while always preserving the ST identity for the full gluon propagator, see Eq. (3.15). Such a self-consistent violation of color gauge invariance/symmetry (SCVCGI/S) is completely NP effect, since in the formal PT ∆ 2 = 0 limit this effect vanishes. The first necessary condition for the SCVCGI/S is color confinement, due to which the gluon is not a physical state. The second sufficient condition for the SCVCGI/S is the explicit presence of the skeleton tadpole term in the full gluon self-energy. In other words, not to request the transversality of the gluon self-energy is the first necessary condition to realize a mass gap. In this case whether the tadpole term is explicitly present or not in the gluon self-energy becomes not important. It plays only secondary role, indicating that the above-mentioned transversality may be violated at the initial stage, indeed. All this makes it possible to establish the structure of the regularized full gluon propagator and the corresponding gluon SD equation in the presence of a mass gap. After the restoration of the transversality of the full gluon propagator relevant for NP QCD in part II of our investigation, the SCVCGI/S in QCD will have no direct physical consequences. None of physical quantites/processes in low-energy QCD will be affected by this proposal, i.e., the unitarity of S-matrix in NP QCD will not suffer, as emphasized above. In QED a mass gap is always in the "gauge prison". It cannot be realized even temporarily, since the photon is a physical state. However, in QCD a door of the "color gauge prison" can be opened for a moment in order to realize a mass gap, because the gluon is not a physical state. A key to this "door" is the constant skeleton tadpole term, which explicitly violates the transversality of the full gluon self-energy. On the other hand, this "door" can be opened without a key (as any door) by not imposing the transversality condition on the full gluon self-energy. So in QED a mass gap cannot be "liberated/released" from the vacuum, while photons and electrons can be liberated/released from the vacuum in order to be physical states. In QCD a mass gap can be "liberated/released" from the vacuum, while gluons and quarks cannot be liberated/released from the vacuum in order to be physical states. In other words, there is no breakdown of U (1) gauge symmetry in QED because the photon is a physical state. At the same time, a temporary breakdown of SU (3) color gauge symmetry in QCD is possible because the gluon is not a physical state. In summary, QCD as a theory of quark-gluon interactions may have a mass gap ∆ 2 , possibly realized in accordance with our proposal. The dynamically generated mass is usually related to breakdown of some symmetry (for example, the dynamically generated quark mass is an evidence of chiral symmetry breakdown). Here a mass gap is an evidence of the SCVCGI/S. In the presence of a mass gap the coupling constant plays no role. Thus the SCVCGI/S is also a direct evidence of the "dimensional transmutation", g 2 → ∆ 2 (λ, α, ξ, g 2 ) [1,30,31], which occurs whenever a massless theory acquires masses dynamically. It is a general feature of spontaneous symmetry breaking in field theories. VI. CONCLUSIONS A self-consistent violation of SU (3) color gauge invariance/symmetry for the realization of a mass gap as it has been described in this investigation (section III) and briefly discussed in the previous section is our first main result. The structures of the regularized full gluon propagator and the corresponding gluon SD equation in the presence of a mass gap have been firmly established. This is our second main result (subsection C in section III). The general nonlinear iteration solution (4.20) for the full gluon propagator explicitly depends on the mass gap. However, it is always severely singular in the q 2 → 0 limit. So the gluons remain massless, and this does not depend on the gauge choice. This solution is our third main result (section IV). In the presence of a mass gap gluon may acquire an effective gluon mass, depending on the gauge choice (the so-called massive solution). This is our fourth main result. Nevertheless, we put it into appendix A, since its relation to the solution of the color confinement problem is unclear, even after the renormalization program is performed. Thus, we have shown explicitly that in the presence of a mass gap at least two independent and different types of formal solutions for the regularized full gluon propagator exist (let us remind that the gluon SD equation is highly NL one, so the number of independent solutions is not fixed). No truncations/approximations/assumptions and no special gauge choice are made in order to show the existence of these general types of solutions. The nonlinear iteration solution (4.20) is interesting for confinement. However, three important problems remain to solve: A. How to make this solution to be the purely transversal. B. To perform the renormalization program for the mass gap, and to see whether the mass gap survives it or not. C. How to treat correctly severe IR singularities inevitably present in this solution. Let us now explain each of these points in more detail. A. As we already know, in the presence of a mass gap the ghosts alone cannot provide the cancellation of the longitudinal component of the full gluon propagator. So a universal method of the restoration of the transversality of the full gluon propagator relevant for NP QCD should be formulated. This will make it possible to eliminate the explicit dependence on the gauge-fixing parameter from all the corresponding expressions in NP QCD. It is worth emphasizing in advance that due to the above-mentioned universal method, the PT part of the general nonlinear iteration solution (4.23), which remains undetermined, will be of no importance for us. Only its INP counterpart (4.21)-(4.22), which functional dependence is exactly determined up to the above-mentioned residues, will matter. B. As repeatedly mentioned above, in this paper the mass gap has been only regularized, i.e., ∆ 2 ≡ ∆ 2 (λ, α, ξ, g 2 ). To perform the renormalization program means to remove the dependence on the above shown parameters in a selfconsistent way. We should prove that the product ∆ 2 JW = Z(λ, α, ξ, g 2 )∆ 2 (λ, α, ξ, g 2 ) exists in the λ → ∞ and α → 0 limits. The mass gap's renormalization constant Z(λ, α, ξ, g 2 ) should appear naturally, i.e., it should not be introduced by hand in order not to compromise the general renormalizability of QCD. In other words, the renormalized mass gap should not depend on the gauge-fixing parameter, should be finite, positive definite, etc. Only after performing this program, we can assign to the Jaffe-Witten (JW) mass gap ∆ 2 JW a physical meaning to be responsible for the NP dynamics in QCD. C. This solution is characterized by the explicit presence of severe IR singularities (q 2 ) −2−k , k = 0, 1, 2, 3..., which, in principle, are independent from each other. The only method to treat them in a self-consistent way is the DRM [11], correctly implemented into the DT [16], as emphasized above. All these problems will be addressed and solved in our subsequent paper (part II of our investigation). Concluding, a few preliminary remarks are in order. We have already emphasized that the most appropriate place where the mass gap may appear is the system of the SD equations (complemented by the corresponding ST identities) for the QCD Green's functions in momentum space. The gauge dependence of the gluon Green's function (4.20) in this space is in general twofold: the explicit dependence on the gauge-fixing parameter via the longitudinal component and the implicit gauge dependence in the Lorentz structure, affiliated with its transversal component. Within our approach the former will be solved if we will be able to formulate a general method how to restore the transversality of the full gluon propagator relevant for NP QCD (the problem A). The latter one will be solved if we will be able to find how to renormalize the mass gap in a gauge-invariant way, i.e., the mass gap should survive the renormalization program (the problem B). Solving the above-mentioned problems, we will achieve an explicit gauge independence at the fundamental quark-gluon level, while it is too earlier to discuss the gauge independence of S-matrix elements in NP QCD within our approach. Working always in momentum space, we avoid thus the problem of gauge ambiguity (uncertainty) [32], which has been discovered in much more complicated functional space (see Ref. [33] as well). Substituting this expansion into the previous relation and after doing some tedious algebra, one obtains where Π R (q 2 ; ξ) = 0 at q 2 = −m 2 g ; otherwise it remains arbitrary. Thus the full gluon propagator (3.16) now looks where, for future purpose, in the invariant function Π R (q 2 ; m 2 g ) instead of ξ we introduced the dependence on the gluon effective mass squared m 2 g which depends on ξ itself. The gluon renormalization constant is . (A7) In the formal PT limit ∆ 2 (λ, ξ) = 0, an effective gluon mass is also zero, m 2 g (λ, ξ) = 0, as it follows from Eq. (A2). So an effective gluon mass is the NP effect. At the same time, it cannot be interpreted as the "physical" gluon mass, since it remains explicitly gauge-dependent quantity (at least at this stage). In other words, we were unable to renormalize it along with the gluon propagator (A6). In the formal PT ∆ 2 (λ, ξ) = m 2 g (λ, ξ) = 0 limit the gluon renormalization constant (A7) becomes the standard one [2,8,9], namely and the role of ghosts will be automatically restored, as pointed out above in section III. Concluding, it is interesting to note that Eq. (A2) has a second solution in the formal PT ∆ 2 (λ, ξ) = 0 limit. In this case an effective gluon mass remains finite, but 1 + Π(−m 2 g ; ξ) = 0. So a scale responsible for the NP dynamics is not determined by an effective gluon mass itself, but by this condition. Its interpretation from the physical point of view is not clear. Evidently, the massive solution (A6) is completely independent from the general nonlinear iteration solution (4.20), and it is difficult to relate it to confinement. However, its existence shows the general possibility for a vector particles to acquire masses dynamically, i.e., without so-called Higgs mechanism (which in its turn requires the existence of not yet discovered Higgs particle). The above-mentioned possibility is due only to the internal dynamics and symmetries of the corresponding gauge theory. APPENDIX B: REMARKS ON ASYMPTOTIC FREEDOM As mentioned above, our general solution for the "running" effective charge (3.19) should satisfy AF at large q 2 . We already know that in the q 2 → ∞ limit the subtracted gluon self-energy Π s (q 2 ; D) can be only logarithmically divergent at any D. So neglecting the mass gap term in the relation (3.19), to leading order one obtains g 2 (q 2 ; Λ 2 ) = g 2 (λ) 1 + bg 2 (λ) ln(q 2 /Λ 2 ) , where Λ 2 is the UV cutoff squared, and b > 0 is the standard color group factor [1,2]. We introduce into the numerator g 2 (λ), so that when it formally zero (no interaction at all) then the full gluon propagator should be reduced to its free PT counterpart. For convenience, we also denote d(q 2 ) as g 2 (q 2 ; Λ 2 ), i.e., put d(q 2 ) ≡ g 2 (q 2 ; Λ 2 ). This expression represents the summation of the so-called main PT logarithms. However, nothing should depend on Λ (and hence on λ) when they go to infinity in order to recover the finite effective charge in this limit. To show explicitly that this finite limit exists, let us rewrite the previous expression in the symmetric form [34] g 2 (λ 1 ) 1 + bg 2 (λ 1 ) ln(q 2 /Λ 2 1 ) = g 2 (λ 2 ) 1 + bg 2 (λ 2 ) ln(q 2 /Λ 2 2 ) , since g 2 (q 2 ; Λ 2 1 ) = g 2 (q 2 ; Λ 2 2 ), i.e., nothing should depend on how we denote the UV cutoff, indeed. In the Λ 1,2 → ∞ (and hence λ 1,2 → ∞) limits neglecting the dependence on ln q 2 , from the relation (B2) one obtains ln Λ 2 − 1 2bg 2 (λ 2 ) = ln Λ 1 − 1 2bg 2 (λ 1 ) , and this relation becomes more and more exact with all the UV cutoffs becoming bigger and bigger (and thus the suppression of ln q 2 becoming more and more justified). Evidently, this relation is equivalent to . Thus there exists indeed the limit lim (Λ,λ)→∞ at which it is finite and does not depend on the UV cutoff or the renormalization point (evidently, not losing generality, we can estimate that 2bg 2 (λ) ∼ 1/ ln λ in the λ → ∞ limit). This finite limit is nothing but Λ QCD = Λ P T , which governs the nontrivial dynamics of PT QCD in asymptotic regime at large q 2 . Thus, using the limit (B5), we can rewrite the initial expression (B1) in terms of the finite quantities. It reproduces the well known AF behavior of the effective charge in QCD ar large q 2 , namely where the standard notation α s (q 2 ) = g 2 (q 2 )/4π has been used. It can be generalized like Eq. (B1) as follows: α s (q 2 ) = α s 1 + bα s ln(q 2 /Λ 2 QCD ) , where α s = g 2 /4π is the fine-structure constant of strong interactions, calculated at some fixed scale, for example at Z boson mass. At a very large q 2 one recovers the previous expression. Concluding, within our approach we have shown explicitly the AF behavior of QCD at short distances (q 2 → ∞) not using the renormalization group equations [1,2,8,9,34]. There is no relation between the mass gap (even renormalized) and the asymptotic scale parameter, at least to leading order. APPENDIX C: REMARKS ON OVERLAPPING DIVERGENCES In order to unravel overlapping UV divergences problem in QCD, in each of the standard SD equations the necessary number of the differentiation with respect to the external momentum should be done first (in order to lower divergences). Then the point-like vertices, which are present in the corresponding skeleton loop integrals should be replaced by their full counterparts via the corresponding integral equations. Finally, one obtains the corresponding SD equations which are much more complicated than the previous (standards) ones, containing different scattering amplitudes. These skeleton expansions are, however, free of the above-mentioned overlapping divergences. Of course, the real procedure ( [16] and references therein) is much more tedious than briefly described above. However, even at this level it is clear that by taking derivatives with respect to the external momentum q in the gluon SD equation (3.17), which is convenient to rewrite as follows: the main initial information on the mass gap will be totally lost (we omit the tensor indices, for simplicity). Just the mass gap which appears first in this equation is the main object we were worried about to demonstrate explicitly its crucial role within our approach. Whether the above-mentioned information will be somehow restored or not at the later stages of the renormalization program is not clear at all. Thus in order to remove overlapping UV divergences ("the water") from the SD equations and skeleton expansions, we are in danger to completely lose the information on the dynamical source of the mass gap ("the baby") within our approach. In order to avoid this danger and to be guaranteed that no dynamical information are lost, we are using the standard gluon SD equation (3.17) in the presence of a mass gap. The existence of any kind of UV divergences (overlapping and usual (overall)) in the skeleton expansions will not cause any problems in order to detect the mass gap responsible for the IR structure of the true QCD vacuum. As emphasized above, the problem of convergence of the regularized skeleton loop series which appear in Eq. (4.20) is completely irrelevant in the context of the present investigation. Anyway, we keep any kind of UV divergences under control within our method, since we are working with the regularized quantities. At the same time, the existence of a mass gap responsible for the IR structure of the full gluon propagator does not depend on whether overlapping divergences are present or not in the SD equations and corresponding skeleton expansions. As argued above, the existence of a mass gap is only due to the SCVCGI/S. All this is the main reason why our starting point is the standard gluon SD equation (3.1) for the unrenormalized (but necessarily regularized) Green's functions (this also simplifies notations).	2008-02-28T14:55:25.000Z	2008-02-28T00:00:00.000	{ "year": 2008, "sha1": "b4550e87823c6deccf00fc73d3b1068ae1a6f79e", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "b4550e87823c6deccf00fc73d3b1068ae1a6f79e", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics" ] }
246562039	pes2o/s2orc	v3-fos-license	Research on Air-Flow-Field Characteristics and Structural Optimization of the Guide Channels of the Autoclave : In order to solve the problem of uneven air inlet ﬂow in the guide channel of autoclave, three optimization schemes for the guide plate were proposed. The ﬂow ﬁeld in the guide channels of autoclave was simulated and analyzed by the realizable k-epsilon turbulence model and dynamic boundary method. The results show that the three optimization schemes have a certain effect on solving the problem of uneven air inlet ﬂow in each guide channel, but some of them also have some negative effects. The integration of the three optimization schemes effectively increases the total air inlet ﬂow of the guide channels; the increase ratio reaches 4.15%, improves the ﬂow-ﬁeld characteristics at the outlet of the guide channels and greatly reduces the difference in the air inlet ﬂow in the left and right guide channels; the difference ratio drops from 62.61% to 5.37%. Introduction Autoclave is a large horizontal pressure vessel, which is widely used in industries that require high temperature or pressure production process, such as heavy metal smelting, refractory brick immersion coal infiltration, heating or cooling of composite glass, cable vulcanization, medicine, aerospace industry, insulation materials and military industries, etc. Due to the particularity of the process conditions, the flow-field characteristics in the autoclave and the uniformity of temperature-field changes have a very important influence on its working performance. As the application scenarios and actual working conditions of the autoclave have become more and more complicated, its internal flow-field characteristics and temperaturefield characteristics have become the focus of research by experts and scholars. Li et al. [1,2] simulated the three-dimensional heat transfer and fluid flow in an autoclave with a lengthto-diameter ratio of 10 under nonuniform heating conditions and analyzed the influence of baffles on the fluid flow characteristics in the autoclave. Bohne et al. [3] studied the heat transfer in a small experimental autoclave through calorimeter measurements and fluid dynamics models and observed a complex turbulent flow pattern with a locally varying heat transfer coefficient. Schimmel et al. [4] conducted machine-learning-related research on the fluid-flow and temperature-field simulation of autoclave. Zhang et al. [5] established a simulation method of the flow field and temperature field in the working process of the autoclave through Xflow software and analyzed the influence of the process parameters of the autoclave on its internal flow field and temperature field. Appa et al. [6,7] studied the mass transfer process of the fluid in the autoclave by numerical simulation and verified it based on experimental data. They found that the mass transfer process of the fluid in the autoclave was poor and proposed an estimate empirical formula for fluid mass transfer based on the experimental data. Li et al. [8] studied the turbulent motion characteristics of the fluid in the semicircular tube jacket on the inner wall of the autoclave by numerical simulation and studied the effect of the average Reynolds number Re, δ and τ on the velocity field and flow resistance of the fluid in the jacket. At the same time, the average axial velocity and the secondary flow function distribution of the fluid in the jacket were obtained based on the orthogonal spiral coordinate system. Antonucci et al. [9,10] analyzed the heat transfer phenomenon that occurred during the operation of the autoclave and compared it with experimental data from industrial autoclaves; they also conducted extensive tests on the semi-theoretical method developed. Gao et al. [11] analyzed the flow-field characteristics and temperature rise process in the effective area of the autoclave through numerical simulation methods and verified the accuracy of the simulation through physical experiments. Li et al. [12] studied the flow field and temperature field of the fluid in the autoclave based on the autoclave molding simulation software and optimized the position of the components in the autoclave, thereby improving the efficiency of process design and improving the quality of components operation. Kluge et al. [13] predicted the temperature distribution of various parts in the autoclave by using computational fluid dynamics methods and compared the consistency of the prediction results with experimental data. It was found that more detailed inlet velocity curves and more advanced turbulence models may produce better consistency with experimental data. Jimmy et al. [14] used the particle image velocimetry method to qualitatively measure the flow characteristics of the fluid in the autoclave. At the same time, they studied the heat transfer of the fluid in the autoclave by measuring the temperature at multiple locations during the heating process. Hassim [15] studied the flow pattern of the autoclave through the Fluent software, and the results showed that the air flow velocity in the rear area of the autoclave was relatively uniform. Booth et al. [16] studied the Reynolds number flow in the pipeline in the autoclave and studied the corresponding turbulent energy dissipation rate. Meng [17] analyzed the influence of the length of the specimen in the autoclave on the influence of the fluid turbulence intensity and turbulent kinetic energy on its side by numerical methods and obtained that the optimal length of the specimen was 126 mm. Antonucci et al. [18,19] proposed a new method to analyze the heat transfer in the autoclave. The method has been applied to patch wing panels and was validated by comparison with the experimental data. Ghamlouch et al. [20] designed and manufactured an autoclave model based on the law of similarity, which allows the use of PIV technology to measure the flow field around a representative real industrial molding and to characterize the heat transfer with the help of a thermal imager. Bhatti et al. [21] recounted the latest trends in computational fluid dynamics in an editorial. In addition, some scholars have studied the characteristics of other fluid peristaltic flows [22][23][24] and temperature fields [25,26], which has certain reference significance for the selection and setting of simulation models and boundaries [27,28]. The flow-field characteristics and temperature-field characteristics in the autoclave have been studied above, but the research on the flow-field characteristics in the guide channels in the autoclave is rarely involved. The autoclave studied in this paper is mainly used for heating and cooling composite glass. The uniformity of the air inlet flow of each guide channel in the autoclave and the velocity characteristics of the air at the outlet of the guide channels have a very important influence on the overall flow-field characteristics and temperature-field distribution characteristics in the autoclave. Therefore, in order to solve the problem of uneven air inlet flow in each guide channel in the autoclave, three optimization schemes for the guide plate are proposed. Quantitative and qualitative analysis is performed on the air inlet flow characteristics of the guide channels in the autoclave and the velocity characteristics at the outlet of guide channels under each optimization scheme to measure the optimization effect. The research results provide a certain reference for the design and optimization of the guide-plate structure in the relevant autoclave. Figure 1 shows the overall structure of the autoclave for heating and cooling composite glasses. The autoclave is mainly composed of autoclave body, fans, guide plates, baffles, a heating row and a cooling row and other components. During the working process, the air in the autoclave enters the area between the two baffles and the two sidewalls inside the autoclave through the guide channels on both sides of the autoclave under the action of the fans and diffuses into the area between the two baffles in the autoclave under the action of the guide plates and flows upward into the fans to complete the entire cycle. The heating row and the cooling row located near the two sidewalls in the autoclave heat or cool the air to realize the control of the temperature in the autoclave. It can be seen from the above analysis of the air distribution and flow directions that the flow-field characteristics in the autoclave, especially the air inlet flow characteristics in the guide channels and the flow characteristics of the air at the outlet of the guide channels, have a very important influence on the change efficiency and uniformity of the temperature field in the autoclave. Research Object and Model Theory The air in the autoclave has a high flow velocity driven by the fans, so the realizable k-ε turbulence model is selected when simulating the air flow characteristics in the autoclave. Based on the standard k-ε turbulence model, a new dissipation rate equation is added to the realizable k-ε turbulence model, which improves the vortex viscosity. The realizable k-ε turbulence model has higher accuracy and better calculation performance in rotating uniform shear flow, free flow (jet and mixing layer), channel flow, etc. The realizable k-ε turbulence model is a typical two-equation model, which is the turbulent kinetic energy equation and the dissipation rate equation. The turbulent kinetic energy equation k is shown in Equation (1). and the dissipation rate equation ε is shown in Equation (2). where G k is the turbulent energy generated by the laminar velocity gradient; σ k is the turbulent Prandtl number of the turbulent flow energy equation; σ is the turbulent Prandtl number of the dissipation rate equation; µ t is the turbulent viscosity; S ε is the user defined primary phase; C 1ε , C 2 , C 3ε and C µ are constants which equal to 1.44, 1.90, 1.90 and 0.99, respectively. The dynamic boundary method is used to simulate the rotation of the fans in actual work of the autoclave and the air suction and discharge process caused by the rotation with the fans. Analysis of Simulation Results This article mainly focuses on the research and analysis of the air inlet flow characteristics in the guide channels and the flow-field characteristics at the outlet of the guide channels in the autoclave. Therefore, in the simulation analysis, the heating row and cooling row are ignored, and the corresponding flow-field model is established and meshed. The realizable k-ε turbulence model and dynamic boundary method introduced above are used to simulate and analyze the air inlet flow characteristics of the guide channels and the air-flow-field characteristics at the outlet of the guide channels. The given motor rotation speed is 1500 r/min. Figure 2 shows the characteristics of the fluid flow field at the outlet of the guide channels of the autoclave. Figure 2a is the air velocity contour. It can be seen from the figure that the air velocity at the outlet of the left guide channel is the largest, the maximum speed is greater than 6 m/s, and the air velocity at the outlet of the middle guide channel is the smallest. Figure 2b is the air turbulent kinetic energy contour, which has the same variation law as the velocity cloud map. Figure 2c is the velocity vector of the air, where the arrow direction indicates the air flow direction and the arrow color indicates the air flow velocity. It can be seen from the figure that the air-flow direction at the outlet of the guide channels is tangent to the guide plates and the angle between the tangent at the end of the guide plates and the vertical direction is larger. Therefore, the air flowing through here diffuses and flows to the left and right ends of the autoclave along the tangential direction of the guide plates instead of flowing down in the vertical direction into the area between the two baffles in the autoclave, which has a negative effect on the efficient circulation of the flow field and the uniform change of the temperature field in the autoclave. At the same time, it can be seen from Figure 2c that under the effect of the guide plate structure, the air flow velocity at the outlet of the left and right guide channels is quite different. The reason for the above phenomenon is that, on the one hand, the flow velocity of the air entering each guide channel is different, and on the other hand, the structure of the guide plates makes the amount of air flowing into each guide channel very uneven. This is also not conducive to the efficient circulation of the flow field and the uniform change of the temperature field in the autoclave. In order to quantitatively analyze the air inlet flow of each guide channel, three monitoring surfaces shown in Figure 3 are established at the outlet of the guide channels to monitor the mass flow of the air passing through each monitoring surface. At the same time, it also provides an intuitive and quantitative analysis and explanation for the optimization effect of subsequent guide plates structure optimization. Through analysis, the mass flow of air flowing into each guide channel under the action of the guide plates is shown in Table 1. It can be seen from Table 1 that under the effect of the existing guide plates, the air inlet flow of each guide channel decreases from left to right, and there is a big difference in the air inlet flow between the left and right guide channels, which is 0.098139 kg/s (the difference ratio is 62.61%), which seriously affects the efficiency of flow-field circulation and the uniformity of temperature-field changes in the autoclave. Structural Optimization Analysis Aiming at the problem of uneven air inlet flow between the left and right guide channels and poor flow-field characteristics at the outlet of guide channels under the existing guide plates structure, three optimization schemes for the guide plates and baffles on both sides of the fans are proposed. The optimization effect of each optimization scheme is evaluated through quantitatively analyzing the air inlet flow of each guide channel and the total air inlet flow of the guide channels before and after optimization. Optimized Scheme 1: The Top Part of Guide Plates In order to solve the problem of the large difference in the air inlet flow of the left and right guide channels in the autoclave, optimization scheme 1, as shown in Figure 4, is proposed. It can be seen from Figure 4 that, under the guide plates before optimization shown in Figure 4a, the air flows directly into each guide channel when it flows out of the fans. Optimization scheme 1 deletes the top part of guide plate 1 and guide plate 2 to form the optimized guide plates shown in Figure 4b, which allows the air to have more time and space for diffusion and distribution before entering the guide channels, thereby reducing the unevenness of the air inlet flow of the left and right guide channels. Taking the two guide plates shown in Figure 4 as examples, respectively, the simulation analysis of the air inlet flow characteristics of the guide channels is carried out so as to obtain the results of the air inlet flow characteristics of the guide channels in the autoclave before and after optimization. Through analysis, the mass flow of air entering each guide channel before and after optimization is shown in Table 2. It can be seen from Table 2 that the air inlet flow of the left guide channel in the autoclave after optimization has decreased to a certain extent, while the air inlet flow of the right guide channel has increased significantly. The difference of the air inlet flow between the left and right guide channels is obviously reduced to 0.023733 kg/s (the difference ratio is 16.80%). However, at the same time, it can be seen from the figure that optimization scheme 1 also causes a significant decrease in the air inlet flow of the middle guide channel in the autoclave. The total air inlet flow of the guide channels in the autoclave is also an important index to measure its flow-field characteristics. It can be seen from Table 2 that the air inlet flow of the right guide channel increases, while that of the left and middle guide channels decreases. The total mass flow of air entering the guide channels before and after optimization is expressed by adding up the air inlet flow of the left, middle and right guide channels. Through simulation analysis, the total mass flow of air entering the guide channels before and after optimization is shown in Table 3. It can be seen from Table 3 that while reducing the difference of the air inlet flow between the left and right guide channels in the autoclave, the total air mass flow of the guide channels in the optimized autoclave has a certain increase, and the increase ratio is 0.60%. Table 3. Total air mass flow of guide channels before and after optimization. Before Optimization After Optimization Mass flow (kg/s) 0.337396 0.339422 Optimized Scheme 2: The Shape of Guide Plates For the problem of poor air flow characteristics at the outlet of the guide channels in the autoclave, optimization scheme 2, as shown in Figure 5, is proposed. It can be seen from Figure 5 that, under the guide plates before optimization shown in Figure 5a, when the air flows to the outlet of the guide channels, it diffuses and flows along the tangential direction of the guide plates to the left and right ends of the autoclave. In optimization scheme 2, the end of the guide plates is tangent to the vertical direction so as to form the optimized guide plates structure shown in Figure 5b, which makes the air to flow downwards in the vertical direction into the area between the two baffles in the autoclave when it flows to the outlet of the guide channels, thereby improving the circulation characteristics of the air in the autoclave. Taking the two guide plates shown in Figure 5 as examples, respectively, the Energies 2022, 15, 1119 7 of 12 simulation analysis of the air inlet flow characteristics of the guide channels is carried out so as to obtain the results of the air inlet flow characteristics of the guide channels in the autoclave before and after optimization. The velocity vector diagram of air at the outlet of the guide channels under the guide plates obtained through analysis is shown in Figure 6b. It can be seen from the figure that the air at the outlet of the guide channels in the autoclave after optimization flows vertically downwards along the end of the guide plates, and its flow characteristics have been well improved. Through analysis, the mass flow of air entering each guide channel before and after optimization is shown in Table 4. It can be seen from the table that the difference in the airflow velocity at the outlet of the left and right guide channels in the autoclave after optimization is still large. Through simulation analysis, the total mass flow of air entering the guide channels before and after optimization is shown in Table 5. It can be seen from Table 5 that the total air mass flow of the guide channels in the autoclave after optimization has a certain decrease, and the decrease ratio is 2.44%. Table 5. Total air mass flow of guide channels before and after optimization. Before Optimization After Optimization Mass flow (kg/s) 0.337396 0.329153 Optimized Scheme 3: The Shape of Baffles Aiming at the problem of the large difference in air inlet flow between the left and right guide channels during the working process of the autoclave, optimization scheme 3, shown in Figure 7, is proposed. It can be seen from Figure 7 that, under guide plates before optimization shown in Figure 7a, when the air flows out from the fans, it enters each guide channel along the tangential direction of the fans' rotation. Optimization scheme 3 optimizes the baffles on both sides of the fans into arc-shaped guide plates, thus forming the optimized guide plates on both sides of the fans as shown in Figure 7b. The optimized guide plates make a part of the air flow tangent to the fans rotation direction before entering the guide channels, while the other part of air is tangent to the end of the optimized guide plates as shown in Figure 7c. It can be seen from Figure 7c that, compared with the baffles on both sides of the fans before the optimization, the direction of the combined velocity of the air shift to the middle guide channels when air passes through the optimized guide plates. As a result, more air flows into the remaining two guide channels, reducing the unevenness of the air inlet flow of the left and right guide channels. Taking the two guide plates shown in Figure 7 as examples, respectively, the simulation analysis of the air inlet flow characteristics of the guide channels is carried out so as to obtain the results of the air inlet flow characteristics of the guide channels in the autoclave before and after optimization. crease, and the decrease ratio is 2.44%. Table 5. Total air mass flow of guide channels before and after optimization. Before Optimization After Optimization Mass flow (kg/s) 0.337396 0.329153 Optimized Scheme 3: The Shape of Baffles Aiming at the problem of the large difference in air inlet flow between the left and right guide channels during the working process of the autoclave, optimization scheme 3, shown in Figure 7, is proposed. It can be seen from Figure 7 that, under guide plates before optimization shown in Figure 7a, when the air flows out from the fans, it enters each guide channel along the tangential direction of the fans' rotation. Optimization scheme 3 optimizes the baffles on both sides of the fans into arc-shaped guide plates, thus forming the optimized guide plates on both sides of the fans as shown in Figure 7b. The optimized guide plates make a part of the air flow tangent to the fans rotation direction before entering the guide channels, while the other part of air is tangent to the end of the optimized guide plates as shown in Figure 7c. It can be seen from Figure 7c that, compared with the baffles on both sides of the fans before the optimization, the direction of the combined velocity of the air shift to the middle guide channels when air passes through the optimized guide plates. As a result, more air flows into the remaining two guide channels, reducing the unevenness of the air inlet flow of the left and right guide channels. Taking the two guide plates shown in Figure 7 as examples, respectively, the simulation analysis of the air inlet flow characteristics of the guide channels is carried out so as to obtain the results of the air inlet flow characteristics of the guide channels in the autoclave before and after optimization. Through simulation analysis, the mass flow of air entering each guide channel before and after optimization is shown in Table 6. It can be seen from Table 6 that after optimization, the air inlet flow of the left guide channel in the autoclave has decreased to a certain extent and the air inlet flow of the middle and right guide channel has increased to a certain extent. The above analysis shows that optimization scheme 3 effectively introduces part of the air in the left guide channel into the middle guide channel, thereby reducing the difference in air inlet flow between the left and right guide channels, which is 0.057811 kg/s (the difference ratio is 42.77%). Table 6. Air mass flow of each guide channel before and after optimization. Before Optimization After Through simulation analysis, the mass flow of air entering each guide channel before and after optimization is shown in Table 6. It can be seen from Table 6 that after optimization, the air inlet flow of the left guide channel in the autoclave has decreased to a certain extent and the air inlet flow of the middle and right guide channel has increased to a certain extent. The above analysis shows that optimization scheme 3 effectively introduces part of the air in the left guide channel into the middle guide channel, thereby reducing the difference in air inlet flow between the left and right guide channels, which is 0.057811 kg/s (the difference ratio is 42.77%). Through simulation analysis, the total mass flow of air entering the guide channels before and after optimization is shown in Table 7. It can be seen from Table 7 that while reducing the difference of the air inlet flow between the left and right guide channels in the autoclave, the total air mass flow of the guide channels in the optimized autoclave has a certain increase, and the increase ratio is 3.55%. Table 7. Total air mass flow of guide channels before and after optimization. Before Optimization After Optimization Mass flow (kg/s) 0.337396 0.349380 Overall Optimization The quantitative analysis of the air inlet flow of each guide channel and the total air inlet flow of the guide channels and the qualitative analysis of the flow-field characteristics at the outlet of the guide channels in the autoclave under three optimization schemes are performed. It can be seen from analysis that while the three optimization schemes have a certain effect on solving the problem of uneven air inlet flow in each guide channel, some of them also have some negative effects. In this section, the above three optimization schemes are integrated to form an overall optimized autoclave model based on the autoclave model before optimization. Taking the two autoclaves before and after optimization as an example, the air inlet flow of each guide channel and the total air inlet flow of the guide channels and air flow characteristics at the outlet of the guide channels are simulated and analyzed so as to obtain the results of the flow-field characteristics of guide channels in the autoclave before and after optimization. The velocity vector diagram of air at the outlet of the guide channels under the guide plates after optimization obtained through analysis is shown in Figure 8b. It can be seen from the figure that the air at the outlet of the guide channels in the autoclave after optimization flows vertically downwards along the end of the guide plates, and its flow characteristics have been well improved. At the same time, it can be seen from the figure that the difference in air flow velocity at the outlet of the left and right guide channels in the autoclave is significantly reduced after optimization. Through analysis, the mass flow of air entering each guide channel before and after optimization is shown in Table 8. It can be seen from Table 8 that after optimization, the air inlet flow of the left guide channel in the autoclave has decreased significantly and the air inlet flow of the right guide channel has increased significantly, while the air inlet flow of the middle guide channel has basically not changed. At the same time, it can be seen from the figure that the difference in the air inlet flow of the left and right guide channels in the autoclave after optimization is greatly reduced to 0.005906 kg/s (the difference ratio is 5.37%). That is, through overall optimization, while maintaining the same amount of air inlet flow in the middle guide channel, the problem of uneven airflow in the left and right guide channels in the autoclave has been well solved. Through simulation analysis, the total mass flow of air entering the guide channels before and after optimization is shown in Table 9. It can be seen from Table 9 that the total air mass flow of the guide channels in the optimized autoclave has a significant increase, and the increase ratio is 4.15%. Table 9. Total air mass flow of guide channels before and after optimization. Before Optimization After Optimization Mass flow (kg/s) 0.337396 0.351386 Discussion The model graphs of autoclave are derived from the software Solidworks and the software Visio, the graphs of air velocity vector are derived from the software Ansys Fluent and the data in the table is analyzed by software Ansys Fluent. Conclusions Aiming at the problem of uneven air inlet flow in the guide channels in the autoclave and poor air flow characteristics at the outlet of the guide channels, three optimization schemes for guide channels and baffles on both sides of the fans are proposed. Through the simulation analysis of the air inlet flow characteristics of the guide channels and the air flow characteristics at the outlet of the guide channels in the autoclave under each scheme, the following conclusions are drawn: (1) Optimization scheme 1 solves the problem of the large difference in the air inlet flow of the left and right guide channels in the autoclave by reducing the air inlet flow of the left guide channel and increasing the air inlet flow of the right guide channel. Optimized scheme 1 causes a slight increase in the total air inlet flow of the guide channels in the autoclave, and at the same time causes a certain decrease in the air inlet flow of the middle guide channel. (2) Optimization scheme 2 effectively improves the air flow characteristics at the outlet of the guide channels, but the difference in airflow velocity at the out of the left and right guide channels is still large. At the same time, optimization scheme 2 resulted in a certain decrease in the total amount of air entering the guide channels in the autoclave. (3) Optimization scheme 3 reduces the difference in the air inlet flow of the left and right guide channels in the autoclave by reducing the air inlet flow of the left guide channel while maintaining the same air inlet flow in the right guide channel. At the same time, optimization scheme 3 increases the air inlet flow of the middle guide channel and the total air inlet flow of the guide channels in the autoclave. (4) Integrating the above three optimization schemes effectively solves the problem that the air flow direction at the outlet of the guide channel points to the two ends of the autoclave along the tangential direction of the guide plates and the large difference in airflow velocity between the left and right guide channels. (5) Integrating the above three optimization schemes greatly reduces the difference in the air inlet flow in the left and right guide channels by reducing the air inlet flow in the left guide channel and increasing the air inlet flow in the right guide channel while keeping the air inlet flow in the middle guide channel basically unchanged. The difference of the air inlet flow between the left and right guide channels is reduced from 0.098139 kg/s (the difference ratio is 62.61%) to 0.005906 kg/s (the difference ratio is 5.37%). The overall optimization scheme makes the total air inlet flow of the guide channels in the autoclave increase substantially, and the increase ratio reaches 4.15%.	2022-02-06T16:52:37.038Z	2022-02-02T00:00:00.000	{ "year": 2022, "sha1": "cad32dd13e7edb8aaa022056f26269fd4a87244e", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/1996-1073/15/3/1119/pdf?version=1644312128", "oa_status": "GOLD", "pdf_src": "ScienceParsePlus", "pdf_hash": "3bc4b3788a0dd34d30d219ac7a1c615ebecf9ee8", "s2fieldsofstudy": [ "Engineering", "Environmental Science" ], "extfieldsofstudy": [] }
212923487	pes2o/s2orc	v3-fos-license	Metabolomic biomarkers in midtrimester maternal plasma can accurately predict the development of preeclampsia Early identification of patients at risk of developing preeclampsia (PE) would allow providers to tailor their prenatal management and adopt preventive strategies, such as low-dose aspirin. Nevertheless, no mid-trimester biomarkers have as yet been proven useful for prediction of PE. This study investigates the ability of metabolomic biomarkers in mid-trimester maternal plasma to predict PE. A case–control study was conducted including 33 pregnant women with mid-trimester maternal plasma (gestational age [GA], 16–24 weeks) who subsequently developed PE and 66 GA-matched controls with normal outcomes (mid-trimester cohort). Plasma samples were comprehensively profiled for primary metabolic and lipidomic signatures based on gas chromatography time-of-flight mass spectrometry (GC-TOF MS) and liquid chromatography Orbitrap mass spectrometry (LC-Orbitrap MS). A potential biomarker panel was computed based on binary logistic regression and evaluated using receiver operating characteristic (ROC) analysis. To evaluate whether this panel can be also used in late pregnancy, a retrospective cohort study was conducted using plasma collected from women who delivered in the late preterm period because of PE (n = 13) or other causes (n = 21) (at-delivery cohort). Metabolomic biomarkers were compared according to the indication for delivery. Performance of the metabolomic panel to identify patients with PE was compared also to a commonly used standard, the plasma soluble fms-like tyrosine kinase-1/placental growth factor (sFlt-1/PlGF) ratio. In the mid-trimester cohort, a total of 329 metabolites were identified and semi-quantified in maternal plasma using GC-TOF MS and LC-Orbitrap-MS. Binary logistic regression analysis proposed a mid-trimester biomarker panel for the prediction of PE with five metabolites (SM C28:1, SM C30:1, LysoPC C19:0, LysoPE C20:0, propane-1,3-diol). This metabolomic model predicted PE better than PlGF (AUC [95% CI]: 0.868 [0.844–0.891] vs 0.604 [0.485–0.723]) and sFlt-1/PlGF ratio. Analysis of plasma from the at-delivery cohort confirmed the ability of this biomarker panel to distinguish PE from non-PE, with comparable discrimination power to that of the sFlt-1/PlGF ratio. In conclusion, an integrative metabolomic biomarker panel in mid-trimester maternal plasma can accurately predict the development of PE and showed good discriminatory power in patients with PE at delivery. Scientific RepoRtS \| (2020) 10:16142 \| https://doi.org/10.1038/s41598-020-72852-4 www.nature.com/scientificreports/ have been inconsistent [3][4][5][6][7][8][9][10] . Such biomarkers included those related to placental perfusion/vascular resistance, endocrine dysfunction, renal dysfunction, endothelial dysfunction, and oxidative stress 11 . Overall, these biomarkers have shown poor sensitivity and poor positive-predictive values for the prediction of PE. A test using plasma biomarkers of endothelial dysfunction-specifically the ratio of the anti-and pro-angiogenic factors, soluble fms-like tyrosine kinase-1 (sFlt-1) and placental growth factor (PlGF), respectively-is now commercially available. Although the sFlt-1/PlGF ratio has shown a high degree of accuracy for predicting PE in late pregnancy 12 , the clinical utility of this test in the mid-trimester remains questionable. There are currently no widely accepted biological markers that have been shown to be reliable, reproducible, and cost-effective in the prediction of PE in early pregnancy, which would be critical if a preventative strategy is to be effective. The metabolome refers to the complete profile of metabolites within the maternal plasma that reflect ongoing biochemical process at a single point in time, including metabolic substrates, intermediates, and final products. Metabolomic analysis of samples collected from study subjects with/without underlying diseases is thus ideally suited for developing diagnostic tests, prognostic tools, and personalized therapeutic interventions [13][14][15] . Metabolomic studies have been conducted for the purpose of biomarker discovery for PE in a range of biological specimens (including serum, urine, and placenta) and different analytical platforms have been used to measure a number of putative biomarker molecule species including amino acid, fatty acid, organic acid, and acylcarnitine 16,17 , which have been summarized in detail elsewhere 16 . In brief, several studies with maternal blood in late pregnancy reported metabolic biomarkers for discrimination of preeclampsia [18][19][20][21] , and other previous studies suggested predictive metabolic biomarkers in early pregnancy with the use of metabolomics tools such as nuclear magnetic resonance spectroscopy (NMR), gas-/ liquid chromatography-mass spectrometry (GC-/ LC-MS), with variable sensitivities and specificities (30-100%) [22][23][24][25][26][27][28][29] . Other biologic samples such as urine or placenta in early or late pregnancy has been also explored to examine the underlying pathophysiology of preeclampsia [30][31][32][33][34] . However, few studies have explored comprehensive phospholipid profiles in maternal blood. Phospholipids (PLs) are a major constituent of cell membranes and function also as second messengers in intracellular signal transduction cascades. The different composition of fatty acyl chains within PL subtypes determines the biophysical traits of the individual cell membrane and influences a range of biological processes 35 . Considering that endothelial injury/dysfunction is a key pathogenic feature of PE 36 , it is biologically plausible that changes in circulating PL profiles might be useful in predicting and/or diagnosing the disease. In addition, abnormal circulating PL profiles have been reported in many diseases, including obesity, hypertension, seizure disorders, and gestational diabetes, all of which are risks factors for PE [37][38][39] . In this study, we analyzed comprehensive plasma PL profiles and primary metabolomic signatures from women who did and did not develop PE. We identified unique metabolomic features in women with PE, which was consistent in both early and late gestational ages, and describe a biomarker panel based on binary logistic regression model that showed robust performance in differentiating women with PE from those without. Results Metabolomic biomarkers in the Mid-trimester Cohort. Integrative metabolomic analysis identified and semi-quantified 116 primary metabolites and 213 lipids based on GC-TOF MS and LC-Orbitrap MS, respectively. The profiles were not clearly separated by principal component analysis (PCA) ( Figure S1). Supervised multivariate statistics, PLS-DA modeling demonstrated that the primary discriminatory factor was GA at sampling based on the first two latent variables (Fig. 1A,B). Variable importance projection (VIP) analysis determined that the metabolites that contributed most to the model were Fatty Acid Esters of Hydroxy Fatty Acid (FAHFA) C18:0, LysoPC C19:0, and LysoPS C19:0 ( Fig. 1C,D). Sixty-one metabolites (18.5%) achieved statistical significance based on one-way ANOVA (adjusted by Benjamini-Hochberg analysis, FDR < 0.05) with post-hoc test (Fisher's LSD) of which 26 metabolites were significantly different in all 4 comparisons (Table S1) We first interrogated the clinical characteristics and the metabolic compositional differences in the midtrimester cohort. Table 1 shows the clinical characteristics of the mid-trimester cohort. Cases and controls were matched for maternal age and GA at blood sampling. As expected, cases (women who developed PE) had an earlier GA at delivery and lower birthweight that those who did not (controls). Neither circulating sFlt-1 concentrations nor sFlt-1/PlGF ratios were different between the two groups, but PlGF levels were significantly lower in women who subsequently developed PE as compared with those that did not (median 43 The compositional changes in the metabolic profiles were interrogated based on Student's t-test in which 6 primary metabolites (1.8%) and 17 lipid molecules (5.2%) were present in plasma at significantly different levels in PE cases compared to healthy controls (p < 0.05) (listed in Table 2). The primary metabolites included glycolysis intermediates (3-phosphoglycerate), nitrogenous compounds (xanthine, glutamate), and palmitoleic acid, all of which were elevated in women who developed PE. Lyxose and propane-1,3-diol were present at significantly lower levels in women who developed PE. Metabolite enrichment analysis suggested that arachidonic acid metabolism, malate-aspartate shuttle, β-alanine metabolism were significantly altered (p value < 0.05) ( Figure S3). Controls (did not develop preeclampsia) (n = 66) Cases (developed preeclampsia) (n = 33) p Value A comparative analysis of the metabolomic profiles revealed 17 lipids (5.2%) and 7 primary metabolites (2.1%) that were significantly different in women delivered with PE versus no PE (Student's t-test, listed in Table 4). Of these, the major lipid classes included sphingomyelins (C30:1, C32:1, C33:2, C34:2, and C38:1), that were present at higher levels in the PE versus non-PE group, as well as FAHFA (C18:0), PCs, PE, LysoPC, and LysoPEs, most of which were also significantly higher in the PE group. Similarly, all identified primary metabolites were upregulated in the PE group. Remarkably, the disaccharide, isomaltose, showed a more than two-fold increase in PE. Other primary metabolites of note included 2-deoxytetronic acid, 3,6-anhydro-d-galactose, sugar alcohols (erythritol, xylitol, and myo-inositol), and thymine (Table 4). Isomaltose and erythritol that were originated from dietary resource, have been suggested for the association with disease biomarker, particularly in diabetes 41,42 . Next, we set out to confirm whether the panel of 5 metabolites identified in the mid-trimester cohort could discriminate between cases with and without PE in the at-delivery cohort. The same panel of 5 compounds (SM Table 2. Metabolites that were significantly different in the preeclampsia group in the mid-trimester cohort. * Statistical significance is presented by p value and false discovery rate (FDR) against control. www.nature.com/scientificreports/ (Fig. 3, Figure S5 B). Then, we further explored if there is a combination of metabolomic markers that could more accurately distinguish between PE and non-PE in the at-delivery cohort. All metabolites that were significantly different in the PE group were analyzed (Table 4) and prioritized using binary logistic regression. The analysis suggested three consecutive models, which consisted of SM C30:1 alone, SM C30:1 + oxidized PE (OxPE) C38:3, and SM C30:1 + oxidized PE (OxPE) C38:3 + isomaltose. The calculated AUC (95% CI) values were as follows: 0.844 (0.802-0.886), 0.904 (0.872-0.937), and 0.972 (0.956-0.988), respectively. The model with the three metabolites (SM C30:1 + oxidized PE (OxPE) C38:3 + isomaltose) showed the marginal difference (p = 0.08) compared to the sFlt-1/PlGF ratio (0.812 [0.760-0.864]) (Fig. 3). Ten-fold cross validation was performed for each model and provided in Figure S6. Discussion In the current study, we comprehensively characterized the metabolome in mid-trimester maternal plasma and identified a panel composed of 5 metabolites that can accurately predict the subsequent development of PE. We then tested this same biomarker signature in a second cohort of women delivered in the late preterm period, and were able to accurately discriminate women who were delivered with PE from those delivered with other indications. In addition, we first reported the significant differences in metabolic phenotype according to gestational age at sampling (mid-trimester and at-delivery), suggesting that these biomarkers may be used for both prediction and diagnosis along with pregnancy period. Particularly, lipid molecules, FAHFA C18:0 and LysoPC C19:0 were highly abundant at the time of delivery. In mid-trimester population, we obtained predictive model for PE with 5 plasma metabolites based on binary logistic regression following the variable prioritization based on Pattern Searching algorithm 43 . The regression model presented good predictive performance in the mid-trimester period (16-24 weeks) for the subsequent development of PE. The biomarker signature included SM C28:1, SM C30:1, LysoPC C19:0, LysoPE C20:0, and propane-1,3-diol. Aberrant levels of SMs, which serve as important structural components of cell membranes and as bioactive signal transduction molecules in various cellular response 44 , have been measured in fetal (umbilical artery) blood 45 , placentas 46 , and maternal blood 46 collected from pregnancies complicated by PE. A number of intervention strategies have been proposed to prevent PE including dietary manipulation, exercise, cardiovascular drugs, and antioxidants, but none has been found to be effective. Low-dose aspirin (LDA), www.nature.com/scientificreports/ which acts in part by inhibiting platelet thromboxane A 2 biosynthesis, is the one intervention that has been shown to reduce the development of PE in some high-risk patients [47][48][49] . As such, recent guidelines support the use of LDA starting from mid-pregnancy period to prevent PE in high-risk populations, including women with a history of PE in a prior pregnancy, and women with multiple gestations, chronic hypertension, pregestational diabetes, renal disease, or autoimmune diseases 1,50 . However, the use of LDA in low-risk populations is not currently recommended even though substantial number of the women without these risk factors can develop PE during pregnancy 51 . For the reasons outlined above, the early risk assessment of PE development is highly desirable for low-risk population. Indeed, the application of complex algorithms (e.g. maternal historical and demographic risk factors, blood pressure, uterine artery pulsatility index on Doppler velocimetry, and maternal serum biomarkers such as pregnancy-associated plasma protein A [PAPP-A] and PlGF) to identify women at risk for preeclampsia and use of LDA might also lead to a reduction of PE development 52 . The current study suggests that a new test using maternal plasma biomarkers alone can more accurately identify women at high-risk of PE development and would allow for more timely and appropriate utilization of preventive medications for low-risk pregnant women (those without a history of prior PE and those with no underlying medical conditions). Most guidelines recommend the LDA therapy starting at the late first or early second trimester of pregnancy 1,50 . Therefore, early and accurate identification of women with high risk for PE is critical to allow for the timely adoption of LDA prophylaxis. In the current study, we developed simple binary logistic regression model composed of 5 plasma metabolites, which accurately predicted the development of PE superior to PIGF. The simpler model using only SMs (SM C28:1 and SM C30:1) also showed good performance for the PE prediction. Moreover, our highly quantitative and comprehensive approach suggested the plausible association between the metabolic features and potential patho-mechanism. For instance, the increases in circulating SMs has been described in a lipidomic analysis of placental microvesicles from abnormal pregnancies and appears to correlate with an increase in inflammation and oxidative stress 53 . The additional examination of plasma metabolites from the at-delivery cohort implied an underlying aberrant redox regulation given the significant increase in oxidized PLs (OxPC, www.nature.com/scientificreports/ OxPI, and OxPE) in the PE group. A recent report suggested that this may be due to the enhanced release of the anti-angiogenic factor, endoglin, by circulating exosome in response to exposure to high levels of SMs 53 . Indeed, in our model, the combination of SM C30:1 and OxPE C38:3 was best able to discriminate between patients with and without PE at delivery. The pathophysiologic mechanism regarding these biomarkers needs to be further investigated for exploring novel therapeutic targets. The appropriate development of the prediction model would be completed through the validation of the developed model in independent study population, but we could not split the study population into the development and validation set because of small number of cases. Instead we validated the model in the same study population and tested if the developed model can be also used in the identification of preeclampsia patient at delivery. Further studies with large number of cases with separate development and validation population should be performed. For the clinical application, the biomarker also needs to be validated in other cohorts including the group in the first trimester. As guidelines recommend the start of LDA ideally in the late first trimester 49,50 , the first trimester biomarkers would be clinically more applicable. In addition, the comparative evaluation of prediction accuracy between the metabolic biomarkers and clinical parameters, or the combination of both needs to be evaluated in future studies. In addition, we evaluated the usefulness of metabolomic panel in at-delivery cohort. The identical biomarker composite developed from the mid-trimester cohort also showed good discrimination between PE and non-PE in the late preterm study population at delivery (at-delivery cohort). Furthermore, the metabolic panel derived from at-delivery cohort showed the outperformance in the discriminant power. For at-delivery cohort, we evaluated the maternal plasma at delivery from patients who delivered in the late preterm period (34-0/7-36-6/7 weeks) as the following reasons: (1) Guidelines recommend that women with severe preeclampsia should be delivered at or after 34 weeks of gestation 54 . Therefore most patients with severe preeclampsia are intended to deliver at late preterm period; (2) In addition, to compare the pathophysiologic changes in biomaterials of preeclampsia with those of control pregnancy, proper comparison should be performed between pregnant women with preeclampsia and those who delivered at term without any complications, but this comparison can be biased by the difference of gestational age at delivery. To adjust gestational age at delivery, most studies included pregnant women who delivered at the similar gestational age (preterm) after spontaneous preterm labor or rupture of membranes. However, the result from this comparison can be difficult in interpretation, because the changes in biomaterials could be originate from the parturition mechanism in preterm labor/rupture of membranes, as well as from the pathophysiologic mechanism of preeclampsia. To overcome this problem, we included as controls not only women who delivered preterm after spontaneous parturition but also those who delivered preterm because of maternal medical indication in late preterm period. In conclusion, we demonstrated that a combination of 5 blood metabolites in mid-trimester maternal plasma accurately predicted the development of PE. This unique metabolomic panel was further validated by its accurate predictability that discriminated the patients delivered in the late preterm period for PE from those delivered for other indications. Methods Study design. (1) Mid-trimester cohort. The population for this case-control study consisted of 33 healthy pregnant women whose plasma was taken in the mid-trimester period (gestational age [GA] range 16-24 weeks, median 17.4 weeks) and who subsequently developed PE [cases]. Control subjects included 66 pregnant women with mid-trimester plasma samples who did not develop PE and were matched for maternal age and GA at maternal plasma sampling (ratio 1:2) [controls]. Patients with major fetal malformations and multifetal pregnancies were excluded. The maternal plasma metabolite and PL profiles were compared between cases and controls. PE was defined as the development of hypertension with evidences of end-organ involvement such as proteinuria, low platelet count, renal or liver involvement, cerebral symptoms, or pulmonary edema 1 . (2) At-delivery cohort. After characterization of the metabolomic biomarker panel, we evaluated the discriminatory power of this panel of biomarkers to identify women with PE at the time of delivery. The latter study population consisted of singleton pregnant women who delivered in the late preterm period (34-0/7 to 36-6/7 weeks), had maternal plasma collected within 3 days of delivery, and had no evidence of chorioamionitis. Study subjects were divided into 2 groups according to the indication for preterm birth (i.e., PE or other causes such as spontaneous preterm labor and maternal and/or fetal indication). The plasma metabolomic biomarker measurements were compared between subjects who did and did not have PE at delivery. Metabolomic profiling of maternal plasma. Maternal plasma samples were centrifuged at 2,000 rpm for 10 min and the supernatant stored at -70 °C until assayed. Samples were comprehensively analyzed for primary metabolomic and lipidomic signatures using gas chromatography time-of-flight mass spectrometry (GC-TOF MS) and liquid chromatography Orbitrap mass spectrometry (LC-Orbitrap MS) [See Appendix S1 for further details of methodology] [55][56][57] . In addition, all analytical procedures, including extraction, reconstitution, and MS analysis were performed in random order to minimize potential systematic errors. The study was approved by the Institutional Review Board of Seoul National University Hospital. All patients provided written informed consent for sample collection and the use of biologic materials for research purposes. All methods were performed in accordance with the relevant guidelines and regulations including Declaration of Helsinki. www.nature.com/scientificreports/ Statistical analysis. For clinical information, continuous data were analyzed using the Mann-Whitney U test and categorical data using the Fisher's exact test or Chi-square test, as appropriate. Statistical analyses were conducted using the IBM SPSS version 20 (IBM Corp., Armonk, NY, USA). The integrative metabolomic profiles were pre-processed for data normalization using the MSTUS methodology 58 implemented in NOREVA (https ://idrb.zju.edu.cn/norev a/) 59 . MSTUS (MS Total Useful Signal) is a method of summing and adjusting the total intensities of identified metabolites across all samples). Student's t-test was performed for univariate statistics using EXCEL (Microsoft Office 2010). One-way ANOVA was conducted with post-hoc testing (Fisher's LSD) as indicated. For adjusting multiple comparisons, the false discovery rate (FDR) was computed (adjusted by Benjamini-Hochberg). Partial least squares-discriminant analysis (PLS-DA) and permutation test were carried out for multivariate statistics using SIMCA 15 (Umetrics AB, Umea, Sweden). Binary logistic regression was performed with forward or enter selection, as appropriate, for predictive model construction using IMB SPSS Statistics for Windows, version 25.0 (IBM Corp.) following auto-scaling procedure. Receiver operating characteristic (ROC) analysis, tenfold cross validation, permutation test, and enrichment analysis were performed using Metaboanalyst 4.0 (https ://www.metab oanal yst.ca/). Data availability The datasets analyzed during the current study are not publicly available, but are available from the corresponding author on reasonable request. www.nature.com/scientificreports/	2020-01-02T21:45:12.196Z	2020-01-01T00:00:00.000	{ "year": 2020, "sha1": "839aaa846713120eb763d834addbc5f050ddfc4e", "oa_license": "CCBY", "oa_url": "https://www.nature.com/articles/s41598-020-72852-4.pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "bd1871f3abf91f546d40937dc43c49ec7dc2f5c4", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
254986255	pes2o/s2orc	v3-fos-license	Early literacy and comprehension skills in children learning English as an additional language and monolingual children with language weaknesses Many children learning English as an additional language (EAL) show reading comprehension difficulties despite adequate decoding. However, the relationship between early language and reading comprehension in this group is not fully understood. The language and literacy skills of 80 children learning English from diverse language backgrounds and 80 monolingual English-speaking peers with language weaknesses were assessed at school entry (mean age = 4 years, 7 months) and after 2 years of schooling in the UK (mean age = 6 years, 3 months). The EAL group showed weaker language skills and stronger word reading than the monolingual group but no difference in reading comprehension. Individual differences in reading comprehension were predicted by variations in decoding and language comprehension in both groups to a similar degree. Introduction Reading development depends upon both decoding and oral language skills, as summarised by the Simple View of Reading (Gough & Tunmer, 1986). Moreover, different cognitive skills have been identified that support the development of different aspects of reading: letter knowledge and phonological processing appear to (Babayigit, 2014(Babayigit, , 2015Burgoyne, Kelly, Whiteley, & Spooner, 2009;Burgoyne, Whiteley, & Hutchinson, 2011, 2013Hutchinson, Whiteley, Smith, & Connors, 2003;Lesaux, Crosson, Kieffer, & Pierce, 2010) although not consistently (e.g. Lesaux, Rupp, & Siegel, 2007;Lesaux & Siegel, 2003;Limbird et al., 2014). Moreover, Manis, Lindsey, and Bailey (2004) suggested that differences in reading comprehension may not be apparent at the early stages of learning to read in a second language (L2). Their sample of young Spanish-speaking bilinguals showed similar levels of performance on measures of Spanish and English reading comprehension at US Grade 2 and only slight differences in performance at Grade 1. Research exploring reading development in monolingual and bilingual children has found similar predictors of decoding in both groups (e.g. Chiappe & Siegel, 2006;McBride-Chang & Kail, 2002;Muter & Diethelm, 2001). Similarly, as with monolingual children, research looking at predictors of reading comprehension in bilingual children has found oral language to be a significant predictor of reading comprehension in most studies (e.g. Babayigit, 2014Babayigit, , 2015Lesaux, Crosson, Kieffer, & Pierce, 2010;Lesaux et al., 2007;Proctor, Carlo, August, & Snow, 2005). Indeed, Kieffer (2012) carried out an analysis of longitudinal data collected from Spanish-speaking English language learners in the US and found that English productive vocabulary was the strongest predictor of English reading comprehension when compared to other measures of oral language. However, not all of these studies directly compared second language learners with their monolingual peers (e.g. Kieffer, 2012;Lesaux et al., 2010;Proctor et al., 2005) and as such have not explored the relative importance of oral language to reading comprehension in children from different language backgrounds. Some studies have found that oral language is a stronger predictor of reading comprehension in second language learners compared to monolingual children although the majority of these have not been carried out in the UK (e.g. Droop & Verhoeven, 2003;Lervåg & Aukrust, 2010;Limbird et al., 2014). In one recent UK study, Babayigit (2014) found that oral language was a significant predictor of reading comprehension in both monolingual and bilingual English-speaking children. However, she also found a marginally stronger association between oral language and reading comprehension in the bilingual children in her sample. Conversely, in a later study, Babayigit (2015) found no significant difference in the strength of association between oral language and reading comprehension, although there was a tendency for this association to be stronger in the EAL group. More studies are needed that directly compare the development of EAL and monolingual children in a UK context. Similarly the majority of these studies include children aged 5 years and over at the first point of testing, with many studies in the UK focusing on children in Year 2 and above (Babayigit, 2014(Babayigit, , 2015Burgoyne et al., 2009Burgoyne et al., , 2011Burgoyne et al., , 2013Hutchinson et al., 2003). Children begin reading instruction in the UK at approximately 4 years of age. It is important to note that in the UK there is considerable emphasis on systematic phonics instruction at the early stages of learning to read, precipitated by the publication of the Independent Review of the Teaching of Early Reading (Rose, 2006) and reinforced by the introduction in 2012 of a statutory check of decoding skills for all 6-year olds in UK primary classrooms (Department for Education, 2012a). Research suggests that instructional practices directly influence the cognitive skills children utilise during reading (McGeown, Johnston & Medford, 2012). More research is therefore needed investigating early reading and comprehension skills in this cohort within the context of a phonics led curriculum of early reading instruction in UK classrooms. Recent reports suggest that a large proportion of monolingual English-speaking children are also starting school in the UK with poor oral language (e.g. Bercow, 2008;Law, Todd, Clark, Mroz, & Carr, 2013;Lee, 2013). These children are at risk of difficulties with literacy development (see Pennington & Bishop, 2009 for a review) and educational underachievement much like their EAL peers. However, the nature and aetiology of their language difficulties may be different. While some children learning EAL will have existing language impairments, for many their weaknesses in English may be largely attributed to the challenge of learning a second language. Conversely, the problems facing monolingual children with language weaknesses may be attributed to other factors such as a language delay, a language impairment and/or low socio-economic status (SES; e.g. Clegg, Law, Rush, Peters, & Roulstone, 2015). These children form an important comparison group since EAL children are often put on the Special Educational Needs register (Stow & Dodd, 2003;Sullivan, 2011) and may receive similar instruction and intervention to their monolingual peers with language weaknesses whether or not they have underlying language problems. Given the difference in aetiology of the language difficulties of many EAL children, this may not be appropriate, and it is important to tease these groups apart in order to ensure children receive the most appropriate support. Research has shown that instructional approaches that combine letter knowledge and phoneme awareness are effective for promoting the development of word level reading (Bowyer-Crane et al., 2008;Lundberg, 1994;Torgesen et al., 1999) whereas teaching to promote broader oral language skills is effective in improving reading comprehension skills (Bianco et al., 2010;Borstrom & Elbro, 1997;Clarke, Snowling, Truelove, & Hulme, 2010;Fricke, Bowyer-Crane, Haley, Hulme, & Snowling, 2013). However, this research typically targets native speakers and only a small number of studies focus on ameliorating the language difficulties of bilingual children, particularly in the UK (see Murphy, 2015 for a review). Understanding the similarities and differences in literacy development between EAL children and their monolingual peers with language weaknesses will help to inform the development of effective literacy instruction and intervention approaches for all children. In this paper we report data collected in a randomised controlled trial that aimed to evaluate the effectiveness of an oral language intervention programme for EAL children (Schaefer, Fricke, Bowyer-Crane, Millard & Hulme, 2015). Here, we present longitudinal analyses examining the predictors of reading comprehension skills at the end of Year 1 in UK primary school, from measures taken at school entry (UK Reception). Half of the children were learning to read in EAL and half were monolingual peers with language weaknesses (ML). Our first aim was to look at the similarities and differences in the early reading and language skills of these two groups. Our second aim was to examine the similarities and differences in predictors of reading comprehension and word reading skills across the two groups. Method Participants We recruited participants from 10 primary schools in the South Yorkshire area. Of the 10 schools selected, 8 had a higher percentage of children eligible for free school meals than the UK national average in the year they were recruited (Department for Education, 2012b). We screened all of the children in the Reception class of these schools at school entry and selected 160 children to participate in a randomised controlled trial evaluating the effectiveness of an oral language intervention. We selected children showing the weakest language skills in relation to their classroom peers based on their performance at the beginning of the project on two subtests from the Clinical Evaluation of Language Fundamentals Preschool II UK (CELF-Preschool II UK; Semel, Wiig, & Secord, 2006a); Expressive Vocabulary and Sentence Structure, and the Non Word Repetition subtest from the Early Repetition Battery (ERB; Seeff-Gabriel, Chiat, & Roy, 2008). 80 children (44 male) were monolingual English-speaking (ML) and 80 children (50 male) were learning EAL. We randomly allocated half of the children to receive oral language intervention (n = 40 ML/40 EAL) and half to a waiting control group (n = 40 ML/40 EAL). We excluded children with no functional English skills from the sample. In the EAL group, 13.75 % of children were reported to have Urdu as their first language (n = 22), and 10 % were reported to be Punjabi speakers (n = 16). The remaining children spoke Arabic (n = 3), Bengali (n = 3), Chinese (n = 3), Czech (n = 6), French (n = 1), Karen (n = 2), Kurdish (n = 3), Lingala (n = 1), Malayalam (n = 2), Mandarin (n = 2), Pashto (n = 6), Polish (n = 3), Portuguese (n = 2), Romany (n = 2), Tamil (n = 1), and Vietnamese (n = 1), while the official home language of one child was unknown. At the initial test phase, the two groups did not differ in age, t(158) = 0.63, p = .53, or nonverbal IQ, t(158) = -1.03, p = .305; see Table 1. Tests and procedures Full details of the screening and testing procedures for the intervention study can be found in Schaefer et al. (2015). We gave children a large battery of tests at each time point but only include the measures appropriate to the current research questions in this paper. We collected the data analysed in this paper at two time points; Time 1 (t1) data in the Autumn term of the children's first year of formal schooling (UK Reception), and Time 2 (t2) data in the Summer term of Year 1, at the end of the children's second year of formal schooling. We saw children individually on each occasion and trained research assistants administered the measures. We used raw scores in all analyses. Language skills Listening comprehension (t1, t2)-measured at t1 using a short story from the York Assessment of Reading for Comprehension (YARC; Snowling et al., 2009) which children listened to via headphones and then answered a set of comprehension Early literacy and comprehension skills in children… 775 questions. The maximum score on this task was 8. We used the Understanding Spoken Paragraphs subtest from the Clinical Evaluation of Language Fundamentals 4th Edition (CELF-4 UK; Semel, Wiig, & Secord, 2006b) at t2. This measure has a maximum score of 15. Expressive Grammar (t1, t2)-measured using the Action Picture Test (APT; Renfrew, 2003). Children are asked questions about a series of pictures that elicit different grammatical constructs in response (e.g. ''What is the girl doing?''; ''What is the mother going to do?''; ''What has the cat just done?''). Children receive a score for grammatical complexity with a maximum possible score of 40. We also gave children the Sentence Structure subtest from the Clinical Evaluation of Language Fundamentals in which they are asked to point to the picture that matches a spoken sentence. We used the subtest from the CELF Preschool II UK (Semel et al., 2006a) scale at t1 which has a maximum score of 22. At t2 we used the CELF-4 UK (Semel et al., 2006b) which has a maximum score of 26. Expressive Vocabulary (t1, t2)-measured using the CELF Expressive Vocabulary subtest in which children are asked to name a series of pictures. At t1 we used the subtest from the CELF Preschool II UK (Semel et al., 2006a) scale which has a maximum score of 40. At t2 we used the CELF-4 UK (Semel et al., 2006b) which has a maximum score of 54. We included the APT Information score as a measure of vocabulary with children's responses scored according to the vocabulary used to describe the pictures (max. raw score 40). Literacy skills Letter Sound Knowledge (t1)-measured using the York Assessment of Reading for Comprehension (YARC)-Early Reading (Hulme et al., 2009) Letter Sound Knowledge subscale in which children are asked to provide the sounds of a series of single letters and digraphs. At t1 we used the core subscale which has a maximum score of 17. Invented Spelling (t1, t2)-measured by giving children pictures to name and spell. At t1 we showed children five pictures, while at t2 we extended this to 10 pictures. We scored responses for number of consonants correct. Word Reading (t1, t2)-At t1 we used the Early Word Recognition subtest from the YARC-Early Reading (Hulme et al., 2009) to assess word reading. This measure included 15 regular and 15 irregular words, and had a maximum score of 30. At t2 we measured word reading using the Diagnostic Test of Word Reading Processes (DTWRP; Forum for Research into Language and Literacy, 2012). Three subscales are given on this test-a measure of non-word reading, a measure of exception word reading and a measure of regular word reading. A maximum score of 30 is available for each subscale. We created a composite score of all three measures with a maximum score of 90. Reading Comprehension (t2)-measured using the York Assessment of Reading for Comprehension (YARC)-Passage Reading (Snowling et al., 2009). We administered the Beginner passage and Level 1 from Form A of the YARC. Beginner passages employ a shared reading paradigm where the experimenter and the child take turns in reading sentences while Level 1 passages are read solely by the child. Following the reading of each passage, children are asked a set of comprehension questions. We created a composite score by combining the scores across the two passages with a maximum score of 16. This measure also provides a Text Reading Accuracy score in terms of the number of errors each child makes during reading. Phonological skills Phonological Processing (t1)-measured using the Non-Word Repetition task from the ERB (Seeff-Gabriel, Chiat, & Roy, 2008). The maximum score on this task is 18. Sound Isolation (t1)-measured using the YARC-Early Reading (Hulme et al., 2009) Sound Isolation subscale in which children are asked to repeat a word and then isolate a sound from the beginning, or end of the word. Children are awarded one point for each correct response with a maximum score of 12. General cognitive abilities Non-Verbal IQ (t1)-measured using the Block Design task from the Wechsler Preschool and Primary Scale of Intelligence (WPPSI; Wechsler, 2003). In this test children are given a set of blocks which they have to arrange to match a series of geometric patterns. Statutory measure of achievement Phonics Screening Check (Department for Education, 2012a)-this is a statutory measure of phonic decoding administered annually when children are in Year 1. Children are asked to read a list of real words and non-words (maximum score = 40). Results We provide means and standard deviations for measures of language, literacy, and phonological skills at t1 and t2 in Table 1 along with Cohen's d calculated using the pooled standard deviation to show the effect size of the difference between groups at each time point. We used raw scores in all analyses since norms are not available for children learning EAL on the majority of these measures. However, the YARC Reading Comprehension normative sample did include approximately 14 % of children learning EAL, and we provide standard scores for this measure in Table 2. We first looked at the similarities and differences between groups at t1 when children entered UK Reception and before they had received formal reading instruction. Finally, we looked at longitudinal predictors of reading comprehension and word reading in the ML and EAL groups. Initial Pearson correlations for all t1 measures with t2 Reading Comprehension and t2 Word Reading are shown in Table 3. In the ML group, t2 Word Reading processes correlate with t1 early literacy and phonological awareness measures but not with t1 language, while in the EAL group t2 Word Reading correlates with all t1 language and literacy measures. T2 Reading Comprehension correlated with all t1 language and literacy measures in both groups with the exception of Letter Sound Knowledge and Early Word Reading in the ML group. Our principal interest was to trace possible causal influences from early variations in language and emergent literacy skills to variations in later word reading and reading comprehension skills. For this purpose, a two-group structural . 1) using Mplus 7.4 (Muthén & Muthén, 1998 with missing data being handled with full information maximum likelihood estimation. Before creating the two-group structural equation model we established that strong (scalar) measurement invariance was present for the two latent variables in the model since constraining the unstandardized factor loadings and intercepts to be equal across groups resulted in no significant change in fit against the unconstrained model, Dv 2 (12) = 17.484, p = .132. There was also invariance between the residual covariances that we estimated to take account of common test-specific variance in the CELF and APT tests, Dv 2 (2) = 4.584, p = .101. Further, since Reading Comprehension and Word Reading were measured as single observed variables, we adjusted for the measurement error (Cole & Preacher, 2014). We wished to assess the possibly separable influences of language and emergent literacy skills on later word reading and reading comprehension skills. In the model in Fig. 1, we used five language measures to define a Language factor (CELF Preschool Sentence Structure, CELF Preschool Expressive Vocabulary, APT information, APT grammar and Listening Comprehension), and three measures of early literacy skills (YARC Letter Sound Knowledge, YARC Early Word Reading and YARC Sound Isolation) to define a separate Early Literacy factor at t1. We regressed Word Reading scores (DTWRP) and YARC Reading Comprehension scores at t2 on both these two factors measured at t1. We also regressed the t2 measures (DTWRP scores and YARC Reading Comprehension scores) on a dummy coded variable representing the difference between the intervention and control groups. In addition, we regressed YARC Reading Comprehension on Word Reading .67 (.52) Fig. 1 Model showing the prediction of language and early literacy at time 1 on reading comprehension and word reading at time 2 in monolingual children and EAL children (in parentheses). The model shows the standardized solution with the exception of the paths from Intervention or Waiting Control Group which show the y-standardized solution as dashed lines Figure 1 shows standardized path weights for both the EAL and ML groups (coefficients for the ML group outside the parentheses and for the EAL group inside the parentheses; these standardized coefficients differ slightly between groups due to differences in variance between the groups). In addition, the y-standardized regression coefficients for the dummy coded intervention group variable on t2 Reading Comprehension and t2 Word Reading are shown in order to identify whether changes in these skills occurred as a result of treatment. These y-standardized coefficients are equivalent to Cohen's d (the difference in mean scores between groups in z-score units). These coefficients are small in magnitude and nonsignificant, confirming that the intervention did not produce any reliable changes in Reading Comprehension or Word Reading skill at t2. At t1, the structural model consists of two correlated latent variables (Early Literacy and Language). Early Literacy at t1 predicts Word Reading at t2. In addition, as expected, Language at t1 and Word Reading ability at t2 are both strong predictors of Reading Comprehension at t2. The model accounts for 95 % of the variance in Reading Comprehension skills at t2 for the EAL group and 93 % of the variance in the ML group (the variance explained in reading comprehension in a comparable model where measurement errors were not corrected was 72 and 67 % for the EAL and ML group respectively). The standardized indirect path from Early Literacy at t1 to Reading Comprehension at t2 through Word Reading at t2 is 0.33 and 0.34 for the EAL and ML groups respectively (p \ .001 for both groups). Discussion In this paper we investigate the early language and literacy skills of a group of 160 children; 80 monolingual English-speaking children with language weaknesses and 80 children learning to read in EAL. We report data collected at school entry and again after roughly 2 years of formal literacy instruction in the UK. The aim of the paper was first to compare the early reading and language skills of these two groups of children at school entry and after 2 years of formal reading instruction. A second aim was to examine the predictors of Word Reading and Reading Comprehension from school entry to the end of Year 1 across groups. Comparing their literacy, phonological, and language skills, at both t1 and t2, the monolingual children had significantly better expressive language skills than their EAL counterparts. This is unsurprising at t1 given that the level of exposure to English will vary considerably in the EAL cohort prior to school entry. However, the consistent gap evident at t2 demonstrates that EAL children continue to lag behind in their English oral language skills even after approximately 2 years of formal schooling and when compared to a group of monolingual peers selected as having poor language skills. Conversely, the EAL cohort outperformed the monolingual children on measures of word reading and spelling at both time points, a finding that is not unusual in the literature. For example, Burgoyne et al., (2009) found stronger word reading and text reading accuracy in their EAL participants compared to the monolingual peers. In terms of phonological skills, a phonological processing advantage was found for EAL children on a measure of non-word repetition in this study, which may partly explain the superior word level reading skills of the EAL group. In contrast, there was no significant difference in phonological awareness skills between the EAL and ML groups, which is inconsistent with previous research (e.g. Campbell & Sais, 1995;Kang, 2012;Marinova-Todd et al., 2010;McBride-Chang & Kail, 2002). However, a number of factors are important to bear in mind when interpreting these findings. First, not all studies have found an advantage for second language learners in phonological skills. Lipka and Siegel (2007) found an advantage for L1 learners on measures of non-word repetition and phonological awareness at kindergarten, and no difference between groups at Grade 3, while Geva and Zadeh (2006) found no difference between L1 and L2 learners on measures of phonological awareness although L2 learners were faster at a rapid automatized naming task. In fact, Melby- Lervåg and Lervåg (2014) only found small and nonsignificant differences between first and second language learners in phonological awareness, which were unaffected by correcting for publication bias. Moreover, the EAL sample in this study was heterogeneous in terms of the languages spoken and therefore the degree of similarity between children's L1 and English varied greatly across the sample. Bialystock et al. (2003) found that Spanish-English bilinguals performed better than monolingual English speakers and Chinese-English bilinguals on a measure of phoneme segmentation. One interpretation of these findings was that the similarity in phonological structure between Spanish and English facilitated phonological awareness skills in that group; a facilitatory effect not available to the Chinese-English bilinguals due to differences in the structures of the two languages. Similarly, Loizou and Stuart (2003) found that while English-Greek bilingual children outperformed their English monolingual peers on measures of phonological awareness, no such advantage was found for Greek-English bilinguals compared to their Greek monolingual peers. They explained this finding in terms of the relative complexity of the L2 compared to the L1 (i.e. English-Greek children are learning a second language (Greek) with a far simpler phonological structure than their L1 (English), which facilitates their phonological awareness in their L2). In contrast, Greek-English children are learning a second language (English) with a far more complex phonological structure than their L1 (Greek) and as such, gain no advantage in their phonological development from exposure to two languages. Given the heterogeneity in our sample, the relative phonological complexity of the first language of our participants will vary, and this may have a masking effect on any phonological awareness advantage to be found at the group level. Finally, all of the children in the study received systematic phonics instruction in class, including work on phonological awareness, which may have attenuated the expected differences in phonological awareness between the two groups. An unexpected finding was the lack of any difference in reading comprehension between groups. Many studies have found that children learning EAL demonstrate lower levels of reading comprehension than their monolingual peers (Babayigit, 2014(Babayigit, , 2015Burgoyne et al., 2009Burgoyne et al., , 2011Hutchinson et al., 2003;Lesaux et al., 2010), a finding supported by a recent meta-analytic review (Melby- Lervåg & Lervåg, 2014). However, it is important to note that we selected the monolingual group in this study as having weak oral language in relation to their peers and they were performing in the low average range on the reading comprehension measure as seen in Table 2. Standard scores for the EAL sample also indicate that they show poor performance on this measure of reading comprehension despite having adequate word reading skills. It is also of note that we measured reading comprehension in this study using a single comprehension measure. Studies have shown that reading comprehension assessments can tap different skills (e.g. Bowyer-Crane & Snowling, 2005;Keenan, Betjemann, & Olson, 2008;Kendeou, Papadopolous, & Spandouis, 2012;Spooner, Baddeley, & Gathercole, 2004;Melby-Lervåg & Lervåg, 2014) and as such, a more reliable index of reading comprehension should be obtained by using more than one measure (e.g. Droop & Verhoeven, 2003). However, due to the extensive battery of measures given as part of the main RCT, we were only able to include one measure of reading comprehension in this study. Moreover, the children in this study are still at a very early stage in terms of reading comprehension development, a fact reflected in the use of a shared reading paradigm for half of the task to make it more accessible. As mentioned earlier, research shows that differences in the relative value of decoding and linguistic comprehension to the prediction of reading comprehension appear over time (e.g. Tilstra et al., 2009;Vellutino et al., 2007). As such, different findings may occur when looking at reading comprehension at a later point. The results of our latent variable models did not support the findings from previous studies that oral language is a stronger predictor of reading comprehension in bilingual children (e.g. Babayigit, 2014;Lervåg & Aukrust, 2010;Limbird et al., 2014). In our model, Language at t1 predicted Reading Comprehension at t2 in both ML and EAL learners, and there was no difference in the strength of association between the latent and observed variables across groups. Similarly, Early Literacy at t1 was a significant predictor of Word Reading at t2 in both groups. These findings may be partly attributable to the developmental stage of the children and the fact that the monolingual children in this study had language weaknesses. According to the Simple View of Reading (Gough & Tunmer, 1986), at the early stages of learning to read, reading comprehension will largely be driven by decoding skills with language comprehension becoming more important over time. The children in this study were still at a relatively early stage in their reading development and therefore, their comprehension skills will be largely dependent on their decoding skills. Moreover, given their language weaknesses, these children may be restricted in using their language skills to support their reading comprehension skills. The children in the Babayigit (2014) study in contrast were approximately 9 years of age and therefore at a stage in development where oral language would be expected to play a more significant role in reading comprehension and as such, the predictive strength of these skills may vary. More longitudinal research is needed tracking the reading and language development of children learning EAL and their monolingual peers from school entry to later primary school in order to further investigate these models. These findings have implications for supporting the reading development of children learning EAL and monolingual children with language weaknesses. Children learning EAL seem to acquire word level reading skills more readily than their monolingual peers with language weaknesses as evidenced by performance on both a standardized measure of word reading and the national phonics screening check. Both groups of children appeared to struggle with reading comprehension. However, the mediation analyses indicated that there was no difference in the strength of association between early oral language and emergent literacy skills, and either word reading or reading comprehension between these two groups. As such, when considering interventions to support reading comprehension for both children learning EAL and monolingual children with language weaknesses, it is important to focus on supporting their oral language skills. However, particular attention needs to be paid to the reading comprehension skills of EAL children from the early stages of learning to read as their word reading skills may mask the identification of their reading comprehension difficulties until much later in their school career.	2022-12-23T14:54:02.614Z	2016-10-08T00:00:00.000	{ "year": 2016, "sha1": "1a318f08a1323dd90d5d85b2f32448dc018a86d3", "oa_license": "CCBY", "oa_url": "https://link.springer.com/content/pdf/10.1007/s11145-016-9699-8.pdf", "oa_status": "HYBRID", "pdf_src": "SpringerNature", "pdf_hash": "1a318f08a1323dd90d5d85b2f32448dc018a86d3", "s2fieldsofstudy": [ "Education", "Linguistics" ], "extfieldsofstudy": [] }
222133846	pes2o/s2orc	v3-fos-license	Quadratic approximate dynamic programming for scheduling water resources: a case study We address the problem of scheduling water resources in a power system via approximate dynamic programming.To this goal, we model a finite horizon economic dispatch problemwith convex stage cost and affine dynamics, and consider aquadratic approximation of the value functions. Evaluating theachieved policy entails solving a quadratic program at each timestep, while value function fitting can be cast as a semidefiniteprogram. We test our proposed algorithm on a simplified versionof the Uruguayan power system, achieving a four percent costreduction with respect to the myopic policy I. INTRODUCTION Optimal operation of multi-reservoir systems for economic dispatch is a topic that has been extensively studied [1]- [3]. Succintly, the goal is to obtain a sequence of release decisions that achieve system operation with minimal cost over a planned horizon, while also meeting operational constraints. In systems involving large reservoirs decisions become coupled across time, while also being dependant on the availability of water -which is typically stochastic. The usual framework for solving these kinds of problems is (Stochastic) Dynamic Programming, where the state of the system typically includes the storage level in each reservoir. Standard practice involves discretizing the state variable and computing the value function at each point. However, the number of needed evaluations grows exponentially with the number of states, a phenomenon known as the Curse of Dimensionality [4]. In order to circumvent this issue, several (approximate) techniques have risen which allow for the problem to be solved in continuous spaces. One of such celebrated algorithms is SDDP which seeks to approximate the value function by a set of lower bounding affine functions [5]. However, getting a rich enough approximation might entail the use of too many hyperplanes [6]. Moreover, under quadratic stage cost and affine dynamics the resulting value functions are provably convex quadratic [7]. Given this, we sought to explore an alternative simpler parametric model. Specifically, we aim to tackle this problem by approximating each value function with a suitable convex quadratic function. This simplified model allows us to formulate the scheduling problem as a special case of convex approximate dynamic programming, therefore making the problem tractable on a continuous state manifold while also relaxing the need of computing exact averages, something typical of SDDP [5]. There exist a vast literature on approximate convex dynamic programming. For a certain class of scalar storage problems, the value functions can be proven to be convex piecewise linear, and algorithms with proven convergence guarantees have been developed [8], [9]. Quadratic approximate dynamic programming has been used before (see e.g. [7]), especially for systems with quadratic cost and transition dynamics that are affine in the control (see [10] for further examples including trajectory tracking and portfolio optimization). We build on these contributions for modelling the water scheduling problem as a quadratic approximate dynamic program. The paper is outlined as follows. Section II introduces our dynamic programming model with inflow evolutions. Section III presents the proposed algorithm, which involves sequentially solving several quadratic programs [11, p.152] and one semidefinite program [12]. In section IV we present our numerical results applied on the Uruguayan power system, while also detailing how to incorporate hydrologic uncertainty in our model. Conclusions are given in Section V. II. HYDROELECTRIC SYSTEM MODELLING Consider a model of operation of a hydroelectric system over a horizon K, with time indexed as k = 0, 1, . . . , K − 1. A state vector x k ∈ R n represents current storage level at n reservoirs; a control vector u k ∈ R m models the actions taken by the system operator, including the release and spill term on each hydroelectric plant. Water inflows w k ∈ R p at a subset of the reservoirs are modeled as correlated noise. Notice that we do not enforce p = n since there might not be significant inflows at some of the reservoirs. The cost of operation of the system is modeled through a function g k (x k , u k , w k ), which may include the cost of thermal generation and a penalty for deviating from economic dispatch. Our goal is to obtain a sequence of control actions u = {u 0 , . . . , u K−1 } such that, for a given starting state x, the expected cost of running the system is minimized: (1) Notice that this formulation -which first computes a minimum and then an expected value-differs from typical dynamic programming approaches [13], where the order is inverted. Implicitly, we are assuming at the k−th stage that the disturbance w k is known. This means full knowledge of total inflows at the start of each time interval. Dynamic Programming allows for decoupling of the optimization problem (1) across stages. For this purpose, let us define the cost-to-go function from stage k onwards: As usual, the main idea behind this decoupling is to compute the cost-to-go for stage k + 1 and, in a recursive manner, use this solution to compute the cost-to-go for stage k by using Bellman Equation [13]: (3) where we explicited box constrains U k (x k , w k ) depending on the current state and inflow (e.g.: release and spill terms must be non-negative and bounded), power balance, etc. It can be shown that the value functions are convex, given that the stage cost is convex and the transition dynamics are affine in both the state and the control [10]. A. Hydrologic state space model To capture correlations in water inflows across stages we expand the state variable to include a discrete Markov state e k = e that summarizes the current hydrological environment. Its dynamics are governed by an homogeneous Markov chain, with transition probabilities: This probabilities may be estimated from historical data. One possibility is letting e k take two values (corresponding to dry and wet) as introduced in [14]. Local practice in Uruguay is to use a 5-level model which spans from very-dry to very-wet [15], with transitions given by a non-homogeneous Markov chain. We propose keeping this 5-level discretization while modelling the hydrologic state evolution as time invariant. This entails procuring a single transition matrix P ∈ R 5×5 from the available data, which will be accomplished using Principal Component Analysis [16], [17]. A more thorough description of our proposed model is presented later in Section IV-B. We separate the hydrologic state e from the reservoir levels x and solve the expected value in Bellman Equation in two steps. Since this hydrologic state can only take discrete values, we can compute a different value function V k,e (x) for each possible value of e. Then, for given e k = e, we estimate the future cost-to-go by an expected value over the next hydrologic state e , computed according to the finite probabilistic model given in (4). The generalized Bellman iteration thus becomes: where the outmost expectation is taken over inflows w conditioned to e k = e. The rightmost sum in (5) can be interpreted as an estimate of the future cost-to-go given the current hydrologic state. If the costs g k and the dynamics f k are affine (5) is a linearly constrained quadratic program [11, p.152] and can be efficiently solved using standard techniques. A. Backward pass As has been argued before, our goal is to compute approximate value functionsṼ k,e (x) quadratic in x, for every stage k and hydrologic state e. Each iteration of the backward dynamic programming algorithm is subdivided into two parts: a sampling stage and a fitting stage. The sampling stage consists of obtaining state-cost pairs (x, β) by solving an approximate Montecarlo-based version of (5): Upon obtaining N pairs x s k , β s k,e , we fit the quadratic value function by solving: The computational complexity of our proposed method resides in solving N ×M linearly constrained quadratic programs (as in (6)) and one semidefinite program (as in (8)) for each stage and hydrologic state. B. Forward pass Once all the value functions are approximated, the expected cost of running the system from a certain initial state x and certain hydrologic state e could be obtained by evaluating the fitted functionṼ 0,e (x). However, each stage of the backwards phase introduces errors on the approximations, and therefore the predicted costṼ 0,e might differ from the true cost substantially. In order to gauge the actual cost obtained by our methodology, a forward phase is carried out. This phase implements a Montecarlo simulation scheme which sequentially solves the one-stage optimization problem: starting at k = 0 with initial storage level x 0 = x and hydrological state e 0 = e. The incurred cost of operation over the planned horizon is the expected sum of the running cost per stages: where the expectation is taken over all possible sequences , and the control laws u k are derived from (9). This simulated cost corresponds to deploying our policy, and is therefore a better figure of merit for evaluating performance than the predictionsṼ 0,e (x). Moreover, the obtained policy's performance can be contrasted with the performance of the myopic policy, which at time k seeks to minimize the current stage cost: x k+1 = f k (x k , u, w k ) Intuitively, at each step the myopic policy will use up (possibly all) the available water, minimizing the current cost and disregarding the utility of water in the future. While at first glance a reasonable thing to do, this behavior is generally suboptimal due to the expected inflows over the next steps and the spatial interconnection of the dams. For example, it could be better suited to store water now (at the expense of a higher cost) for use later, when a drought is expected. We expect our methodology to outperform the myopic policy. But how good can our policy really be? Although this question remains unanswered, we can construct a lower bound on the optimal performance. For a given inflow sequence w = {w 0 , . . . , w K−1 } the optimal decisions u = {u 0 , . . . , u K−1 } and the optimal cost can be obtained by solving the K−stage problem: x k+1 = f k (x k , u k , w k ) ∀k = 0, . . . , K − 1 (3). Problem (12) solves for the whole decision sequence u = {u 0 , . . . , u K−1 } at once, by being given full knowledge of all the noise realizations w at the start of the planning horizon. This is in sharp contrast with our proposed algorithm, where at each stage k the controller only has access to the current noise w k . The expected cost of running (12) over all the possible inflow sequences w is indeed a lower bound on (10) since the expectation of the minimum is lower than the minimum of the expectation. Intuitively, (12) ahcieves a lower value because more information about the future inflows is available for planning. A. The Uruguayan system Uruguay is a small country with a demand profile that seldom surpasses 2000M W . It is comprised of 4 hydroelectric plants: 3 of which are located in a cascade-like fashion along the Río Negro basin; the fourth one is located in the Río Uruguay, and is shared with neighbouring Argentina. The combined installed power in said facilities is roughly 1500M W . There are a number of wind farms in Uruguay, with a total installed power amounting to more than 75% of the country's peak load. In recent years, there has been a surge in the installation of solar farms as well [18]. We will employ a one-year horizon with weekly decisions (K = 52 weeks in a year, k = 0, . . . , K − 1). In that regard, non-dispatchable renewables (wind and solar) will be left out of our model since they typically vary on a much faster timescale. Generation will be provided by the four hydroelectric plants and by a single thermal generator representing the aggregate thermal generation of the whole system. The state vector x k ∈ R 4 represents the current volume at each of the four reservoirs. The control u k = r k , s k , t k ∈ R 9 + consists of the release (r k ∈ R 4 + ) and spill vectors (s k ∈ R 4 + ) and the total thermal generation (t k ∈ R + ). The state dynamics are described by where B is the coupling matrix that captures the interconnection between hydro plants: and the vector w k gathers the weekly inflows at each reservoir, as detailed in the next Section. Finally, the cost function g(t k ) is the cost incurred by thermal generation, modeled as linear and time-invariant. B. Markov Model estimation The series used in this case study consist of the weekly measured inflows from the three main reservoirs in Uruguay collected over 105 years . As a first step, we clean up the negative values which correspond mainly to measurement errors, and for this study are considered as Not Aviailable (NA) data in the model estimation phase. We then proceed through several steps: 1) Normalization: Each one of the three series of hydraulic inflow is divided by its weekly median across the time period to remove the seasonal variations along the year. The second step is to apply a logarithm transformation to the normalized series (Box-Cox transformation with λ = 1 [19]). After these two normalization steps, it can be observed that the new series present an approximately Gaussian distribution. In Figure 1 we plot the estimated median inflow and the resulting distribution after transformation. 2) Model estimation: As mentioned before, in this phase entire rows with NA data are removed. To adjust a Markov model to the historic inflows, we considered two clustering techniques in order to group similar inflows in a fixed amount of categories. Our first approach was to perform a K-means algorithm [19] applied directly in the three dimensional space of log inflows. This algorithm consists in separating the data into K clusters in a way that the euclidean distance between each point to the centroid of the assigned group is minimized. Given an initial but not optimal clustering, the algorithm relocates each point to its new nearest center, update the clustering centers by calculating the mean of the updated members, and repeat the relocating-and-updating process until convergence criteria (such as predefined number of iterations, difference on the value of the distortion function) are satisfied. In our study we considered a K = 5 clusters, obtaining the clusterization depicted in Figure 2. As a final step, the rows not assigned to any cluster due to NA values (which came from zeros in the original dataset) are replaced by a small value (in order for the logarithmic transformation to work) and labelled accordingly. That is, they are not used to fit the clusters, but are labelled using the clusters obtained with the previous data. This ensures a more robust estimation of the clusters. The problem with this approach is that the cluster code 1, . . . , 5 is not related to the hydraulicity of the group because the centroids are assigned randomly by the algorithm. This is why we considered a second approach where the clustering is applied to a suitable feature of the inflow vectors. We chose to construct this feature using principal component analysis (PCA) by projecting the points along the principal component axis of the inflow vector. This ensures that the feature variability is maximized. To find the principal direction, the covariance matrix of the data set is calculated, as well as its eigenvectors and eigenvalues. The principal direction is the one associated with the largest eigenvalue, and before doing the projection into this direction, the eigenvector is normalized so the projection is just a convex combination of inflows (i.e. the sum of the components of the PCA eigenvector are normalized to sum 1). After projection, we chose K = 5 clusters containing 20% of the observations each, with increasing inflows in each category. The resulting clusters now have physical meaning, with larger inflows being integrated in the same cluster, and they are depicted in Figure 2. a) Markov process estimation: For both clustering methods explained before, in order to estimate the Markov process, the transitions between different clusters are counted. The parameters p ij of the markovian matrix P represent the probability of making a transition from cluster i to cluster j. This parameters are calculated as follows: where e t represents the state in time t, and the sum is computed along the complete state sequence. C. Simulation using Markov transitions Trials begin at an initial state x 0 and initial hydrologic state e 0 . At each time step the hydrologic sequence is updated with the markovian matrix P derived in (15), and a disturbance vector w k corresponding to said hydrologic state is sampled. In order to approximate the total cost of running the system, we substitute the expected value in (10) with a sample mean carried out over T = 105 forward passes. 1) Performance for varying training points: State-cost pairs are sampled by partinioning the state space in a grid-like fashion. Each of the four reservoirs i = 0, . . . , 3 is uniformly partitioned in N i steps, yielding a total number of N = N 0 × N 1 × N 2 × N 3 state points. The cost at each point is obtained by averaging over M = 10 different noise realizations. It is worth emphasizing that the state variables are not discretized, but these grid points are knots where we anchor our quadratic model to find the specified parameters using (8). The results shown herafter are for varying N 0 , which corresponds to the discretization of the largest reservoir Bonete. For the other reservoirs we fix N i = 3. As an illustrating example, Fig. 3 shows a cut of the quadratic obtained for the fourtieth week of the year with N 0 = 10. 7)). In blue: fitted quadratic function (using (8)). Note that the cost-to-go seems to primarily depend on the state of Bonete dam. Sampling more state-cost pairs at every stage naturally increases the computational effort required to perform the backward pass. Nonetheless, our experiments show that there is no significant performance gain in the obtained policy if more points are used in the training phase (Fig. 4). 2) Bounds on performance and comparison with myopic policy: We can compare the predicted cost-to-go at the start of the year V 0,e0 (x 0 ) with the simulated total cost K−1 k=0 g k achieved by running the system forward starting from e 0 and x 0 , following the learned policy (see (9)-(10)). Fig. 5 shows a comparison between the predicted and simulated cost as a function of the level of the largest reservoir, while starting from a neither-dry-nor-wet hydrologic state (e 0 = 2). A lower bound is constructed by solving the K−stages problem (12) given full knowledge of the noise realizations. Our experiments show that the predictionsṼ 0,e are typically optimistic. The policy achieved by our proposed algorithm typically outperforms the so-called myopic policy (11), in particular for non-empty initial reservoir levels, as portrayed in Fig. 6. D. Simulation using historical series We also perform simulations using the historical series of inflows that were used for fitting our markov model. We Fig. 6. Comparison between the myopic (blue) and learned policy (orange) for varying initial storage levels, along with a performance bound. Our policy achieves a 4% reduction on cost w.r.t. the myopic policy when storage levels are half-full, and performs at most 9% worse than the lower bound policy. compare the cost attained by our policy with the cost attained by a policy that was trained with the Markov model currently in use in Uruguay, and obtain better performance (see Fig. 7). V. CONCLUSIONS We proposed the use of convex quadratic functions to approximate the cost-to-go of a simple economic dispatch problem. We showed that training our method involves solving a sequence of quadratic and semidefinite programs, which can be done with standard convex suites. We benchmarked our algorithm on the Uruguayan power system, obtaining performance that surpasses that of a myopic policy by four percent, and comparable to the theoretical lower bound derived in Section III-B.	2020-10-06T01:01:04.616Z	2020-09-28T00:00:00.000	{ "year": 2020, "sha1": "b779d895478e9041671185d515877dee3c9b2ce4", "oa_license": null, "oa_url": "http://arxiv.org/pdf/2010.02122", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "b779d895478e9041671185d515877dee3c9b2ce4", "s2fieldsofstudy": [ "Mathematics" ], "extfieldsofstudy": [ "Computer Science", "Engineering", "Mathematics" ] }
17710919	pes2o/s2orc	v3-fos-license	Automorphisms and forms of simple infinite-dimensional linearly compact Lie superalgebras We describe the group of continuous automorphisms of all simple infinite-dimensional linearly compact Lie superalgebras and use it in order to classify F-forms of these superalgebras over any field F of characteristic zero. Introduction In our paper [2] we classified all maximal open subalgebras of all simple infinitedimensional linearly compact Lie superalgebras S over an algebraically closed fieldF of characteristic zero, up to conjugation by the group G of inner automorphisms of the Lie superalgebra DerS of continuous derivations of S. An immediate corollary of this result is Theorem 11.1 of [2], which describes, up to conjugation by G, all maximal open subalgebras of S, which are invariant with respect to all inner automorphisms of S. Using this result and an explicit description of DerS (see [10,Proposition 6.1] and its corrected version [2, Proposition 1.8]), we derive the classification of all maximal among the open subalgebras of S, which are AutS-invariant, where AutS is the group of all continuous automorphisms of S (Theorem 3.4). Such a subalgebra S 0 always exists, and in most of the cases it is unique (also, in most of the cases it is a subalgebra of minimal codimension). Picking a subspace S −1 of S, which is minimal among AutS-invariant subspaces, properly containing S 0 , we can construct the Weisfeiler filtration (see e.g. [2] or [10]). Then it is easy to see that where U is a normal prounipotent subgroup consisting of automorphisms of S inducing an identity automorphism of GrS, and AutgrS is a subgroup of a (finite-dimensional) algebraic group of automorphisms of GrS, preserving the grading. We list all the groups AutgrS, along with their (faithful) action on Gr −1 S, in Table 1. This leads to the following description of the group AutS: where InautS is the subgroup of all inner automorphisms of S and A is a closed subgroup of AutgrS, listed in Corollary 4.3. Let F be a subfield ofF, whose algebraic closure isF, and fix an F-form S F of S, i.e., a Lie superalgebra over F, such that S F ⊗ FF ∼ = S. Then all F-forms of S, up to isomorphism, are in a bijective correspondence with H 1 (Gal, AutS), where Gal = Gal(F/F) (see e.g. [13]). Since the first Galois cohomology of a prounipotent algebraic group is trivial (see e.g. [13]), we conclude, using the cohomology long exact sequence, that (3) H 1 (Gal, AutS) ֒→ H 1 (Gal, AutgrS). Since the first Galois cohomology with coefficients in the groups GL n (F ) and Sp n (F) is trivial (see, e.g., [14], [15]), we conclude from (3) and Table 1, that H 1 (Gal, AutS) is trivial in all cases except for four: S = H(m, n), K(m, n), S (1,2), and E (1,6). Thus, in all cases, except for these four, S has a unique F-form (in the SKO(n, n + 1; β) case we have to assume that β ∈ F in order for such a form to exist). Since H 1 (Gal, O n (F)) is in canonical bijective correspondence with classes of non-degenerate bilinear forms in n variables over F (see, e.g., [15]), we find that all F-forms of H(m, n) and K(m, n) are defined by the action on supersymplectic and supercontact forms over F, respectively. In the cases S = S(1, 2) and E (1,6), the answer is more interesting. We construct all F-forms of these Lie superalgebras, using the theory of Lie conformal superalgebras. The present paper is a continuation of [2], which we refer to for terminology not explained here. The base field, unlike in [2], is an arbitrary field F of characteristic 0, and we denote byF its algebraic closure. In the Lie algebra case the problems considered in the present paper were solved by Rudakov [13], whose methods we use. Z-Gradings In papers [2] and [10] the base field is C. However, it is not difficult to extend all the results there to the case of an arbitrary algebraically closed fieldF of characteristic zero. In order to do this one has to replace exponentiable derivations of a linearly compact algebra S in the sense of [2], [10], by exponentiable derivations in the sense of [8] (a derivation d of a Lie superalgebra S over a field F is called exponentiable in the sense of [8] if d(H) ⊂ H for any closed AutSinvariant subspace H of S). Also, we define the group of inner automorphisms of S to be the group generated by all elements exp(ad a), where exp(ad a) converges in linearly compact topology. Then Theorem 1.7 of [2] on conjugacy of maximal tori in an artinian semisimple linearly compact superalgebra still holds overF. Consequently, the classification given in [2] of primitive pairs (L, L 0 ) up to conjugacy by inner automorphisms of DerL stands as well overF. We first recall from [2] and [10] the necessary information on Z-gradings of Lie superalgebras in question over the fieldF. For information on finitedimensional Lie superalgebras we refer to [9] or [10]. Recall that W (m, n) is the Lie superalgebra of all continuous derivations of the commutative associative superalgebra Λ(m, n) = Λ(n)[[x 1 , . . . , x m ]], where Λ(n) is the Grassmann superalgebra in n odd indeterminates ξ 1 , . . . , ξ n , and x 1 , . . . , x m are even indeterminates. Recall that a Z-grading of the Lie superalgebra W (m, n) is called the grading of type (a 1 , . . . , a m \|b 1 , . ). Every such a grading always induces a grading on the Lie superalgebra S(m, n) and it induces a grading on S = H(m, n), K(m, n), HO(n, n), SHO(n, n), KO(n, n + 1), or SKO(n, n + 1; β) if the defining differential form of S is homogeneous with respect to this grading. The induced grading on S is also called a grading of type (a 1 , . . . , a m \|b 1 , . . . , b n ). The Z-grading of type (1, . . . , 1\|1, . . . , 1) is an irreducible grading of W (m, n) called its principal grading. In this grading W (m, n) = j≥−1 g j has 0-th graded component isomorphic to the Lie superalgebra gl(m, n) and −1-st graded component isomorphic to the standard gl(m, n)-moduleF m\|n . The even part of g 0 is isomorphic to the Lie algebra gl m ⊕ gl n where gl m (resp. gl n ) acts trivially onF n (resp.F m ) and acts as the standard representation onF m (resp.F n ). The principal grading of W (m, n) induces on S(m, n), H(m, n), HO(n, n) and SHO(n, n), irreducible gradings also called principal. The 0-th graded component of S(m, n) in its principal grading is isomorphic to the Lie superalgebra sl(m, n) and its −1-st graded component is isomorphic to the standard sl(m, n)-moduleF m\|n . The even part of g 0 is isomorphic to the Lie algebra sl m ⊕ sl n ⊕Fc where sl m (resp. sl n ) acts trivially onF n (resp. F m ) and acts as the standard representation onF m (resp.F n ). Here c acts by multiplication by −n (resp. −m) onF m (resp.F n ). The Z-grading of W (1, 2) of type (2\|1, 1) induces a grading on S(1, 2) = j≥−2 g j , which is not irreducible. Then g 0 ∼ = sl 2 ⊕Fc where c acts on S as the grading operator, and g −1 =F 2 ⊕F 2 , whereF 2 is the standard sl 2 -module. The two copies of the standard sl 2 -module in g −1 are both odd. Finally, the principal grading of the Lie superalgebra E(5, 10) is irreducible of depth 2, with 0-th graded component isomorphic to sl 5 and −1-st graded component isomorphic ro Λ 2F5 , whereF 5 is the standard sl 5 -module. Given a simple infinite-dimensional linearly compact Lie superalgebra S = j≥−d g j with its principal or subprincipal grading, we will call S 0 = j≥0 g j the principal or subprincipal subalgebra of S, respectively. Likewise, if S = j≥−d g j with a grading of a given type, we will call S 0 = j≥0 g j the subalgebra of S of this type. , is the intersection of S with the subprincipal subalgebra of W (m, 1). Since the supersymplectic form is not homogeneous with respect to the subprincipal grading of W (m, 1), L 0 (0) is not graded. We shall call this subalgebra the subprincipal subalgebra of H(m, 1). If g is a Lie algebra acting linearly on a vector space V overF, we denote by exp(g) the linear algebraic subgroup of GL(V ), generated by all exp a, where a is a (locally) nilpotent endomorphism of V , and by t a , where a is a diagonalizable endomorphism of V with integer eigenvalues and t ∈F × . If a group G is an almost direct product of two subgroups G 1 and G 2 (i.e., both G 1 and G 2 are normal subgroups and G 1 ∩ G 2 is a finite central subgroup of G) we will denote it by G = G 1 · G 2 . We will often make use of the following simple result: Proposition 1.2 Suppose we have a representation of a Lie superalgebra g over F in a vector superspace V , and a faithful representation of a group G in V , containing exp(g0), preserving parity and such that conjugation by elements of G induces automorphisms of g. Then the maximal possible G are as follows in the following cases: (a) if g = sl n and V =F n ⊕F n with the same parity, then G is an almost direct product of GL 2 and SL n ; in particular if n = 2 then (f ) if g = P n and V =F n +F n * is the standard P n -module, then G =F × ·GL n ; In all cases when G =F × · G 1 , the groupF × acts on V by scalar multiplication. Proof. Consider the map f : G −→ Aut(g) that associates to every element of G the induced automorphism by conjugation of g. Then the kernel of f consists of the elements of G commuting with g. Suppose, as in (a), that g = sl n and V =F n ⊕F n , where the two copies of the standard sl n -module have the same parity. Then Imf consists of inner automorphisms of sl n , i.e., Imf = P GL n , and ker f = GL 2 . We have therefore the following exact sequence: Since there is in G a complementary to GL 2 subgroup, which is SL n , we conclude that G is an almost direct product of GL 2 and SL n . It follows that if n is odd, then G =F × ·(SL 2 ×SL n ), and if n is even then where C 2 is the cyclic subgroup of order two of SL 2 × SL n generated by (−I 2 , −I n ), proving (a). The same argument proves (b). By the same argument, in case (c) we get the exact sequence Since G contains a complementary toF × subgroup, which is SL 5 , we conclude that G = GL 5 . If g = sl(m, n) and V =F m\|n is its standard representation, then gl m acts irreducibly onF m and gl n acts irreducibly onF n , hence G = GL m × GL n , proving (d). Suppose that g = spo(2k, n) and V =F 2k\|n is the standard spo(2k, n)module. Define on g −1 = p i , q i , ξ j \| i = 1, . . . , k, j = 1, . . . , n the following symmetric bilinear form: . Then G consists of the automorphisms ofF 2k ⊕F n preserving the bilinear form (·, ·) up to multiplication by a scalar, hence G = Finally, let g = P n and V =F n +F n * be the standard P n -module. Then gl n acts irreducibly onF n andF n * which have different parities. It follows that G =F × · GL n . Likewise, if g =P (4), then G =F × · SL 4 since the group of automorphisms of g is SL 4 . On AutS Let S be a linearly compact infinite-dimensional Lie superalgebra overF and let AutS denote the group of all continous automorphisms of S. Let S = S −d ⊃ · · · ⊃ S 0 ⊃ . . . be a filtration of S by open subalgebras such that all S j are AutS-invariant and GrS = ⊕ j≥−d g j is a transitive graded Lie superalgebra. Denote by Aut(GrS) the group of automorphisms of GrS preserving the grading. Denote by Autf S the subgroup consisting of g ∈ AutS which induce an identity automorphism of GrS, and let AutgrS be the subgroup of Aut(GrS) consisting of automorphisms induced by g ∈ AutS. We have an exact sequence: Proof. By transitivity, g −n = g n −1 for n ≥ 1, and, in addition, we have the wellknown injective AutgrS-equivariant map g n → Hom(g If σ ∈ Autf S, then σ = 1 + σ 1 , where σ 1 (L j ) ⊂ L j+1 . Hence log σ = n≥1 (−1) n+1 σ n 1 n converges and e t log σ converges to a one-parameter subgroup of Autf S. Hence σ is an inner automorphism of DerS, proving (b). If S 0 is a graded subalgebra, we have an obvious inclusion Aut(GrS) ⊂ AutS and exact sequence (4), proving (c). Remark 2.2 By Proposition 2.1(a), AutgrS is a subgroup of GL(g −1 ) whose Lie algebra is Gr 0 DerS acting on g −1 . By Proposition 2.1(b), Autf S is a prounipotent group. Invariant Subalgebras Given a linearly compact Lie superalgebra L, we call invariant a subalgebra of L which is invariant with respect to all its inner automorphisms, or, equivalently, which contains all elements a of L such that exp(ad(a)) converges in the linearly compact topology. It turns out that an open subalgebra of minimal codimension in a linearly compact infinite-dimensional simple Lie superalgebra S overF is always invariant under all inner automorphisms of S (see [2]). Example 3.1 We recall that the Lie superalgebra S = SHO ∼ (n, n) is the subalgebra of HO(n, n) defined as follows: where v is the volume form associated to the usual divergence and F = 1 − 2ξ 1 . . . ξ n (cf. [2, §5]). Let S 0 be the intersection of S with the principal subalgebra of W (n, n). Then the Weisfeiler filtration associated to S 0 has depth one and GrS ∼ = SHO ′ (n, n) with the Z-grading of type (1, . . . , 1\|1, . . . , 1) (cf. [2,Example 5.2]). Here and further by GrS we denote the completion of the graded Lie superalgebra associated to the above filtration. By [2, Proposition 1.11], S 0 is a maximal open subalgebra of S. It is easy to see that it is also an invariant subalgebra. This subalgebra is called the principal subalgebra of S. Example 3.2 We recall that the Lie superalgebra S = SKO ∼ (n, n + 1) is the subalgebra of KO(n, n + 1) defined as follows: where v β is the volume form attached to the divergence div β for β = (n + 2)/n and F = 1 + ξ 1 . . . ξ n+1 (cf. [2, §5]). Let S 0 be the intersection of S with the subalgebra of W (n, n + 1) of type (1, . . . , 1\|1, . . . , 1, 2). Then the Weisfeiler filtration associated to S 0 has depth 2 and GrS ∼ = SKO(n, n+1; (n+2)/n) with its principal grading. By [2, Proposition 1.11], S 0 is a maximal open subalgebra of S. It is easy to see that it is also an invariant subalgebra. This subalgebra is called the principal subalgebra of S. If S = SKO(3, 4; 1/3), then, according to Theorem 3.3, S has infinitely many invariant subalgebras which are conjugate to the subprincipal subalgebra. Besides, the principal subalgebra of S is an invariant subalgebra. Note that the principal grading of S has depth 2 and the subprincipal grading of S has depth 1, therefore the principal and subprincipal subalgebras are not conjugate. It follows that the principal subalgebra is invariant with respect to all automorphisms of S. In fact, it is the unique maximal among AutS-invariant subalgebras of S, since the intersection of all the subalgebras of S listed in Theorem 3.3(e) is the subalgebra of S of type (2, 2, 2\|1, 1, 1, 3), which is contained in the principal subalgebra. If S = SKO(2, 3; 1), then S has infinitely many invariant subalgebras which are conjugate to the principal subalgebra, besides, the subprincipal subalgebra is also an invariant subalgebra of S. The principal and subprincipal subalgebras have codimension (2\|3) and (2\|2), respectively, hence they cannot be conjugate. It follows that the subprincipal subalgebra is invariant with respect to all automorphisms of S. In fact, it is the unique maximal among AutS-invariant subalgebras of S, since it contains the intersection of all subalgebras which are conjugate to the principal subalgebra (cf. The Group AutgrS In this section, for every simple infinite-dimensional linearly compact Lie superalgebra S overF, we fix the following maximal among AutS-invariant open subalgebras of S, which we shall denote by Let S −1 be a minimal subspace of S, properly containing the subalgebra S 0 and invariant with respect to the group AutS, and let S = S −d S −d+1 ⊃ · · · ⊃ S −1 ⊃ S 0 ⊃ · · · be the associated Weisfeiler filtration of S. All members of the Weisfeiler filtration associated to S 0 are invariant with respect to the group AutS. Let GrS = ⊕ j≥−d g j be the associated Z-graded Lie superalgebra. In this section we will describe the group AutgrS introduced in Section 2, for every S. The results are summarized in Table 1 (where by ΠV we denote V with reversed parity). Theorem 4.2 Let S be a simple infinite-dimensional linearly compact Lie superalgebra overF. Then AutgrS is the algebraic group listed in the last column of Table 1. In all cases when AutgrS ∼ =F × · G 1 , the groupF × acts on g −1 by scalar multiplication. Proof. Let S = W (m, n) with (m, n) = (1, 1) or S = S(m, n) with (m, n) = (1, 2), with the principal grading. By Proposition 1.2(d), AutgrS ⊂ GL m ×GL n . But the group on the right is contained in AutgrS since it acts by automorphisms of S via linear changes of indeterminates. It follows that AutgrS ∼ = GL m ×GL n . Let S = K(2k + 1, n) = j≥−2 g j with its principal grading. By Proposition 1.2(e), AutgrS ⊂F × · (Sp 2k × O n ). But the group on the right is contained in AutgrS since it acts by automorphisms of S via linear changes of indeterminates, preserving the supercontact differential form dx 2k+1 + k i=1 (x i dx k+i − x k+i dx i ) + n j=1 ξ j dξ n−j+1 up to multiplication by a non-zero number. It follows that AutgrS ∼ =F × · (Sp 2k × O n ). Likewise, if S = H(2k, n) = j≥−1 g j with the principal grading, the groupF × ·(Sp 2k × O n ) acts by automorphisms of S via linear changes of indeterminates, preserving the supersymplectic differential form k i=1 dp i ∧ dq i + n j=1 dξ j dξ n−j+1 up to multiplication by a non-zero number. Hence again, AutgrS ∼ =F × · (Sp 2k × O n ). Consider the Lie superalgebras S = HO(n, n), SHO(n, n) with n > 3, KO(n, n + 1), or SKO(n, n + 1; β) with n > 2 or with n = 2 and β = 0, 1, with their principal gradings. By Proposition 1.2(f ), AutgrS ⊂F × · GL n . On the other hand, the group on the right is contained in AutgrS since it acts by automorphisms of S via linear changes of indeterminates, preserving the odd supersymplectic form n i=1 dx i dξ i and the volume form v attached to the usual divergence up to multiplication by a non-zero number, if S = HO(n, n) or S = SHO(n, n) with n > 3, and preserving the odd supercontact form dξ n+1 + n i=1 (ξ i dx i + x i dξ i ) up to multiplication by an invertible function, and the volume form v β attached to the β-divergence up to multiplication by a nonzero number, if S = KO(n, n + 1) or S = SKO(n, n + 1; β) with β = 0, 1 if n = 2. Therefore AutgrS ∼ =F × · GL n . Let S = SHO ∼ (n, n), with n > 2 even, and let S 0 be the principal subalgebra of S. The group Aut(GrS) consists of the automorphisms of SHO ′ (n, n) preserving its principal grading. By the same argument as for SHO(n, n), Aut(GrS) ∼ =F × · GL n . The subgroup AutgrS consists of the elements in Aut(GrS) which can be lifted to automorphisms of S. Every element in SL n can be lifted to an automorphism of S, since it preserves the form F v defining the Lie superalgebra S. Besides, such automorphisms are inner and act on S via linear changes of variables. On the contrary, the outer automorphisms of SHO ′ (n, n) do not preserve the form F v for any t ∈F, hence they cannot be lifted to any automorphism of S. It follows that AutgrS ∼ = SL n . The argument for S = SKO ∼ (n, n + 1) is similar. Consider S = S(1, 2) = j≥−2 g j with the grading of type (2\|1, 1). By Proposition 1.2(a), AutgrS ⊂F × ·SO 4 . Notice that exp(ad(g 0 )) ∼ =F × ·SL 2 , acting by automorphisms of S via linear changes of indeterminates which preserve the standard volume form v up to multiplication by a non-zero number. Since the algebra of outer derivations of S is isomorphic to sl 2 , AutgrS ∼ =F × · SO 4 . Then exp tD, when applied to x i and ξ j , gives convergent series S i (t) and R j (t), respectively (in the linearly compact topology), hence the change of variables x i → S i (t), ξ j → R j (t) is a one-parameter group of automorphisms of W (m, n) which preserves L. If S = S(1, 2), then, by the same argument as above, the inner automorphisms of S and its automorphism t c , where c is the grading operator of S with respect to its grading of type (2\|1, 1), are induced by changes of variables which preserve the volume form up to multiplication by a non-zero number. We recall that the algebra a of outer derivations of S is isomorphic to sl 2 , with standard generators e, f , h, where e = ad(ξ 1 ξ 2 ∂ ∂x ) and h = ad(ξ 1 ∂ ∂ξ1 + ξ 2 ∂ ∂ξ2 ) (cf. [2, Remark 2.12]). As for S(m, n), t h , where t ∈F, is obtained by a linear change of variables preserving the volume form up to multiplication by a non-zero number. Besides, the element ξ 1 ξ 2 ∂ ∂x is contained in the first member of the principal filtration of S ′ (1, 2), thus it is obtained by a change of variables preserving the volume form up to multiplication by a non-zero number, by Lemma 4.4. On the other hand, the automorphism exp(f ) cannot be induced by any change of variables, since it does not preserve the principal filtration of S. The argument for all other non-exceptional Lie superalgebras is similar. F-Forms Let F be a field of characteristic zero and letF be its algebraic closure. Denote by Gal the Galois group of F ⊂F. Then Gal acts on AutL as follows: α.ϕ := ϕ α = αϕα −1 , α ∈ Gal, ϕ ∈ AutL. We recall the following standard result (cf. [14, §VII.2]): K is a group and A, B, C are groups with an action of K by automorphisms, related by an exact sequence: then there is a cohomology long exact sequence: where the first three maps are group homomorphisms, and the last three are maps of pointed sets. Proof. The same arguments as in [13,Proposition 4.2] show that H 1 (Gal, Autf S) = 0. Then the statement follows from exact sequence (4) in Section 2 and Proposition 5.3. We recall the following well known results on Galois cohomology. All details can be found in [14, § X] and [15, § III Annexe]. (e) if q is a quadratic form over F, then there exists a bijection between H 1 (Gal, O n (q,F)) and the set of classes of F-quadratic forms which arē F-isomorphic to q; (f ) if q is a quadratic form over F, then there exists a bijection between H 1 (Gal, SO n (q,F)) and the set of classes of F-quadratic forms q ′ which areF-isomorphic to q and such that det(q ′ )/ det(q) ∈ (F × ) 2 . Lemma 5.6 Let G be an almost direct product over F ofF × and an algebraic group G 1 , and let C =F × ∩ G 1 (F) be a cyclic group of order k. Then we have the following exact sequence: In particular, if H 1 (Gal, G 1 ) = 1, then H 1 (Gal, G) = 1. Proof. We have the following exact sequence: where π : G → G/G 1 ∼ =F × /C →F × is the composition of the canonical map of G to G/G 1 and the map x → x k fromF × /C toF × . By Proposition 5.3, we get the following exact sequence: This implies exact sequence (5). We fix the F-form S F of each simple infinite-dimensional linearly compact Lie superalgebra S overF, defined by the same conditions as in [2], but over F (in the case of SKO(n, n + 1; β) we need to assume that β ∈ F). This is called the split F-form of S. In more invariant terms, this F-form is characterized by the condition that it contains a split maximal torus T (i.e. T is ad-diagonalizable over F and T ⊗ FF is a maximal torus of S). Theorem 5.7 Let S be a simple infinite-dimensional linearly compact Lie superalgebra overF not isomorphic to H(m, n), K(m, n), E(1, 6), or S(1, 2). Then any F-form of S is isomorphic to the split F-form. Proof. It follows from Propositions 5.2, 5.4 and the description of the group AutgrS given in Theorem 4.2, using Theorem 5.5 and Lemma 5.6. Remark 5.8 Let S = H(2k, n) or S = K(2k + 1, n). Then, according to Table 1 and Lemma 5.6, we have the exact sequence Here and further, Sp 2k = Sp 2k (F) and O n ⊂ GL n (F) is the orthogonal group overF which leaves invariant the quadratic form Given a non-degenerate quadratic form q over F in n indeterminates, associated to a symmetric matrix c = (c ij ), introduce the following supersymplectic and supercontact differential forms σ q and Σ q : Two such F-forms S q and S q ′ of S are isomorphic if and only if q and q ′ are equivalent non-degenerate quadratic forms over F, up to multiplication by a non-zero scalar in F. Proof. It is easy to see that every non-degenerate quadratic form q over F, with matrix c = (c ij ), gives rise to the F-forms H q (2k, n) and K q (2k + 1, n) of the Lie superalgebras S = H(2k, n) and S = K(2k + 1, n), respectively, attached to the corresponding cocycles. By construction, equivalent quadratic forms give rise to isomorphic F-forms of S. Besides, if λ ∈ F × and q ′ is the quadratic form associated to the matrix λc, then S q ∼ = S q ′ , and the isomorphism is given by the following change of variables: The F-forms S q exhaust all F-forms of the Lie superalgebra S, due to Proposition 5.2, Theorem 4.2, Remark 5.8 and Theorem 5.5(e). Example 5.10 Consider the F-form K q (1, 6) of K(1, 6) corresponding to the supercontact form Σ q = dt + 6 i=1 c ij ξ i dξ j . Then the principal grading of K(1, 6) induces an irreducible grading on K q (1, 6): K q (1, 6) = j≥−2 g j , where g 0 = h ⊕ F, h is an F-form of so 6 (F), g −1 ∼ = F 6 , and g 1 = g * −1 ⊕ Λ 3 (F 6 ). Let d be the discriminant of the quadratic form q. If −d ∈ (F × ) 2 , then the g 0 -module Λ 3 (F 6 ) is not irreducible, and decomposes over F into the direct sum of two g 0 -submodules g + 1 and g − 1 , which are the eigenspaces of the Hodge operator * , see Example 6.2 below (they are obtained from one another by an automorphism of g 0 ). It follows that we can define an F-form E q (1, 6) of the Lie superalgebra E(1, 6) by repeating the same construction as the one described in Section 1, namely, E q (1, 6) will be the graded subalgebra of K q (1, 6) generated by g −1 + g 0 + (g * −1 + g + 1 ). Theorem 5.11 Any F-form of the Lie superalgebra S = E(1, 6) is isomorphic to one of the Lie superalgebras E q (1, 6) constructed in Example 5.10, where q is a non-degenerate quadratic form over F in six indeterminates, with discriminant d ∈ −(F × ) 2 . Two such F-forms E q (1, 6) and E q ′ (1, 6) of E(1, 6) are isomorphic if and only if the quadratic forms q and q ′ are equivalent, up to multiplication by a non-zero scalar in F. Proof. By Lemma 5.6 and Theorem 4.2(h), we have the exact sequence The statement follows, due to Proposition 5.2, Theorem 5.5(f ) and the proof of Theorem 5.9. Example 5.13 Consider a Lie superalgebra g − over F with consistent Z-grading g − = g −2 + g −1 + g 0 , where g 0 = h ⊕ F, h is an F-form of sl 2 , g −2 = Fz, and g −1 is a four-dimensional [g 0 , g 0 ]-module such that g −1 ⊗ FF is the direct sum of two copies of the standard sl 2 -module, and where the bracket in g −1 is defined as follows: where q is a non-degenerate bilinear form on g −1 over F, which is symmetric, i.e. q(a, b) = q(b, a). Such a superalgebra g − exists if and only if the discriminant d of the quadratic form q lies in (F × ) 2 . Let S q (1, 2) denote the full prolongation of g − over F (see [4, §1.6]). Then S q (1, 2) is an F-form of S(1, 2). Theorem 5.14 Any F-form of the Lie superalgebra S = S(1, 2) is isomorphic to one of the Lie superalgebras S q (1, 2) constructed in Example 5.13, where q is a non-degenerate quadratic form over F in four indeterminates, with discriminant d ∈ (F × ) 2 . Two such F-forms S q (1, 2) and S q ′ (1, 2) of S(1, 2) are isomorphic if and only if the quadratic forms q and q ′ are equivalent, up to multiplication by a non-zero scalar in F. Proof. By Lemma 5.6 and Theorem 4.2, we have the exact sequence The statement follows, due to Proposition 5.2, Theorem 5.5(f ) and the proof of Theorem 5.9. We summarize the results of this section in the following theorem: Theorem 5.15 Let S be a simple infinite-dimensional linearly compact Lie superalgebra overF. If S is not isomorphic to H(m, n), K(m, n), E(1, 6), or S(1, 2), then the split F-form S F is, up to isomorphism, the unique F-form of S. In the remaining four cases, all F-forms of S are, up to isomorphism, as follows: (a) the Lie superalgebras H q (m, n) := {X ∈ W (m, n) F \| Xσ q = 0} where σ q is a supersymplectic differential form over F, if S = H(m, n); (c) the Lie superalgebras E q (1, 6) constructed in Example 5.10, where q is a non-degenerate quadratic form over F in six indeterminates with discriminant d ∈ −(F × ) 2 , if S = E(1, 6); (d) the Lie superalgebras S q (1, 2) constructed in Example 5.13, where q is a non-degenerate quadratic form over F in four indeterminates with discrim- The isomorphisms between these F-forms are described in Theorems 5.9, 5.11, 5.14. Remark 5.16 It follows immediately from Theorem 5.15 that a simple infinitedimensional linearly compact Lie superalgebra S over C has, up to isomorphism, one real form if S is not isomorphic to H(m, n), K(m, n), E(1, 6), or S(1, 2), two real forms if S is isomorphic to E(1, 6) or S(1, 2), and [n/2] + 1 real forms if S is isomorphic to H(m, n) or K(m, n). Finite Simple Lie Conformal Superalgebras In this section we use the theory of Lie conformal superalgebras in order to give an explicit construction of all non-split forms of all simple infinite-dimensional linearly compact Lie superalgebras. In conclusion of the section, we give the related classification of all F-forms of all simple finite Lie conformal superalgebras. We briefly recall the definition of a Lie conformal superalgebra and of its annihilation algebra. For notation, definitions and results on Lie conformal superalgebras we refer to [5], [6] and [11]. A Lie conformal superalgebra R over F is a left Z/2Z-graded F[∂]-module endowed with an F-linear map, called the λ-bracket, satisfying the axioms of sesquilinearity, skew-commutativity, and the Jacobi identity. One writes [a λ b] = n∈Z+ λ n n! (a (n) b); the coefficient (a (n) b) is called the n-th product of a and b. A Lie conformal superalgebra R is called finite if it is finitely generated as an F[∂]-module. Given a finite Lie conformal superalgebra R, we can associate to it a linearly compact Lie superalgebra L(R) as follows. Consider the Lie conformal superalgebra R[[t]], where t is an even indeterminate, the ∂-action is defined by ∂ + ∂ t , and the n-th products are defined by: For the classification of finite simple Lie conformal superalgebras over an algebraically closed fieldF of characteristic zero we refer to [6]. The list consists of four series (N ∈ Z + ): W N , S N +2,a ,S N +2 , K N (N = 4), K ′ 4 , the exceptional Lie conformal superalgebra CK 6 of rank 32, and Curs, where s is a simple finite-dimensional Lie superalgebra. Example 6.1 Let V be an N -dimensional vector space over F with a nondegenerate symmetric bilinear form q. The Lie conformal superalgebra K N,q associated to V is F[∂]Λ(V ) with λ-bracket: . The annihilation algebra of K N,q is isomorphic to the Lie superalgebra K q (1, N ) defined in Theorem 5.9(b). The Lie conformal superalgebra K N,q is an F-form of the finite simple Lie conformal superalgebra K N . It is easy to check that the F[∂]-span of these elements is closed under λ-bracket (6), hence they form an F-form CK 6,q of the Lie conformal subalgebra CK 6 of K 6 (cf [3, Theorem 3.1]). Likewise, if N = 4 and β 2 = 1/det(q), the F[∂]-span of the elements: is closed under λ-bracket (6). It follows that these elements form a subalgebra S 2,q of K 4,q , which is an F-form of the Lie conformal superalgebra S 2,0 (cf. [3,Remark p. 225]). Remark 6.3 The annihilation algebras of the Lie conformal superalgebras CK 6,q and S 2,q , constructed in Examples 6.1 and 6.2, are the Lie superalgebras E q (1, 6) and S q (1, 2), constructed in Examples 5.10 and 5.13, respectively. Due to Theorems 5.11 and 5.14, Example 6.2 provides an explicit construction of all F-forms of the Lie superalgebras E(1, 6) and S(1, 2). We conclude by classifying all F-forms of all simple finite Lie conformal superalgebras overF . The following theorem can be derived from [7, Remark 3.1]. Theorem 6.4 Let R be a simple finite Lie conformal superalgebra overF. If R is not isomorphic to S 2,0 , K N , K ′ 4 , CK 6 or Curs, then there exists, up to isomorphism, a unique F-form of R (in the case R = S N,a , we have to assume that a ∈ F for such a form to exist). In the remaining cases, all F-forms of R are as follows: • the Lie conformal superalgebras K N,q if R is isomorphic to K N ; • the Lie conformal superalgebras CK 6,q if R is isomorphic to CK 6 ; • the Lie conformal superalgebras S 2,q if R is isomorphic to S 2,0 ; • the derived algebras of the Lie conformal superalgebras K 4,q if R is isomorphic to K ′ 4 ; • the Lie conformal superalgebras Curs F , where s F is an F-form of the Lie superalgebra s, if R is isomorphic to Curs.	2014-10-01T00:00:00.000Z	2006-01-12T00:00:00.000	{ "year": 2006, "sha1": "5656eef84bd936b8c4139d97010d2cd1596f7d54", "oa_license": null, "oa_url": "http://arxiv.org/pdf/math/0601292", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "5656eef84bd936b8c4139d97010d2cd1596f7d54", "s2fieldsofstudy": [ "Mathematics" ], "extfieldsofstudy": [] }
53583630	pes2o/s2orc	v3-fos-license	Enhanced Transmission of Acoustic Waves Through Subwavelength Holes in Hard Plates In 1944, Bethe found the transmittance of electromagnetic (EM) waves through a tiny hole in a perfectly conducting screen varies as being proportional to (kr)4, where k=2π/λ, λ is the wavelength and r is the hole radius [1]. This result shows that a small hole has extremely low transmission or negligible cross section for EM waves of very long wavelength. For a hole of finite thickness, the transmittance is found to be reduced further, because no propagating mode exists inside the hole [2]. Introduction In 1944, Bethe found the transmittance of electromagnetic (EM) waves through a tiny hole in a perfectly conducting screen varies as being proportional to (kr) 4 , where k=2π/λ, λ is the wavelength and r is the hole radius [1]. This result shows that a small hole has extremely low transmission or negligible cross section for EM waves of very long wavelength. For a hole of finite thickness, the transmittance is found to be reduced further, because no propagating mode exists inside the hole [2]. However, in 1998, Ebbesen et al. has observed enhanced transmission of light through either a lattice of subwavelength holes or a single hole surrounded by surface periodical patterns on thin metallic films, where the optical transmission can be much larger than the area fraction of the holes at specific frequencies [3][4][5][6][7][8][9][10][11][12][13]. The holes, once organized or decorated, have the transmission cross section larger than themselves' area, which is different radically from the theory by Bethe. Since then, the remarkable phenomenon has inspired a tremendous amount of attention and works on resonant transmissions of EM waves through various apertures on either metallic or dielectric structure [14][15][16][17][18][19][20][21][22][23][24][25]. Phenomenologically, various observed transmission resonances are associated with two geometrical factors: structural factor (SF) emerging globally from the lattice periodicity and aperture factor owned locally by the individual unit [26][27][28][29]. Structural-factor-related resonances typically have the transmission wavelength comparable to the lattice constant and are dependent strongly on the incidence angle. In sharp contrast, aperture-factor-related resonances have the wavelength determined mainly by the transversal/longitudinal dimensions of the aperture and are not sensitive to the incidence angle. It is well known that acoustic and EM waves share a lot of wave phenomena, but they have something in difference. In nature, acoustic wave is a scalar longitudinal wave in inviscid fluids, while EM wave is a vector transverse wave. Consequently, a subwavelength hole has no cutoff for acoustic wave, but does for EM wave, which underlies the distinct transmissions of acoustic/EM waves through a hole in an ideally rigid/conducting screen. The acoustic transmission of a single hole approaches a constant, 8/ 2 , dislike the EM case, with decreasing the ratio r/ [30]. Transmission/diffraction by an acoustical grating is an old problem, and the previous investigations addressed some cases: one-dimensional (1D) periodic slits in a rigid screen [31,32], a single hole in a thick wall [33,34], and a 1D grating composed of parallel steel rods with finite grating thickness [35,36]. Here we studied the acoustic transmissions through two structures: (1) a two-dimensional array (square lattice) of subwavelength hole and (2) a single hole surrounded by the surface periodic grooves. It is found that the acoustic transmission phenomenon for the structured thin plates is analogous completely to the case of EM wave, except for the transmission phase. For the hole array in thick plates, the transmission peaks are related to the Fabry-Perot-like (FP-like) resonances inside the holes and can occur to the frequencies well below Wood's anomalies. Ultrasonic measurements In our experiments, the measurements of far field transmissions of acoustic waves in the ultrasonic frequency regime (0.2-2.0 MHz) were performed in a large water tank. Two immersion transducers were employed as ultrasonic generator and receiver, and the sample was placed at a rotation stage located between the two transducers at an appropriate distance. The sample could be rotated, so that the oblique incidences were measured. The ultrasonic pulse was incident upon the sample and the transmitted signal was collected by the receiver, collinear with the incident wave. Transmission magnitude, T, and transmission phase, , of the sample were obtained by normalizing the Fourier transformed spectra of the signal through the sample, Apart from measuring the transmission spectrum, we also implemented point-by-point scanning to detect the pressure field distribution in the transmission process. A pinducer (1.5 mm in diameter) replaced the receiving transducer and was located at a distance, z, from the rear surface of the sample to detect the pressure field distribution there. The pinducer was mounted on a two-dimensional translation stage. The scanning was done along the x-y plane parallel to the sample surface, with a spatial step of 0.1mm×0.1mm. Enhanced transmission of acoustic waves through hole array structure First we measured the acoustic transmission of a hole array with the hole diameter d = 0.5 mm, the lattice constant a = 1.5 mm, and the plate thickness t = 0.5 mm. Figure 1 shows the transmittance of the hole array at normal incidence, compared to the transmittance of a smooth brass plate with identical thickness. For the smooth brass plate, very low transmittance is seen because of the acoustic impedance mismatch (ηbrass / ηwater ≈ 25). It is noticed that the transmittance rises at lower frequencies, which indicates a thin brass plate can not block acoustic waves of very long wavelength or very low frequency. This fact is different from the EM case where a sheet of metal as thin as skin depth works well. For the hole array, a pronounced peak is seen at 0.85 MHz and followed by a transmittance zero close to 1.0 MHz which is just Wood's anomaly λ=a. The peak has the transmittance (68%), much lager than the area fraction (8.7%) of holes occupation in the array structure, and shows an acoustic transmission enhancement through the hole array, similar to the EM case. We also investigated the dependence of the transmission peak on the lattice constant. Figure 2 shows the normal transmissions of the hole arrays with identical lattice constant a = 2.0 mm and different hole diameters. The transmission peak and two Wood's anomalies (pointed by arrows) are identified at ~0.75 MHz and ~1.1MHz. With the larger diameter holes (d = 1.2 mm), the peak becomes more pronounced. Comparing with the array of a = 1.5 mm in Figure 1, it is clear to show that the peaks and Wood's anomalies downshift to lower frequencies as the lattice constant increases. In Figure 2, we also plotted the measured transmission phase  for the hole array of d = 1.2, a = 2.0, and t = 0.5 mm, and found 0.98 at the peak frequency. The approximate   phase change reveals the oscillations of the acoustic field on the front and rear surfaces of the plate are out-of-phase, which is distinct from the corresponding characteristic in the EM case. For EM wave transmitted through a hole array, the hole acts as barrier due to the transmission frequency much lower than the cutoff frequency of the hole, and the wave has to tunnel through the hole in a form of evanescent field. So the phase change of the EM wave across the holey film/plate assumes nearly zero [37]. This difference in transmission phase bares the distinct behaviors of a hole to acoustic and EM waves, again. The transmission map is plotted as a function of both the frequency and the incidence angle. The predicted variation of Wood's anomalies versus angle is plotted as solid lines and is superposed on the map. Derived from the conservation of momentum, the variation relation reads: for the Wood's anomaly frequency f (l,m) of order (l, m). It is seen from the map that the measured shifting of Wood's anomalies with the incidence angle agrees well with the solid lines. On the other hand, the peaks exhibit a strong angle-dependent behavior in the same way as Wood's anomalies. In recent investigations, it is demonstrated that the SF resonance can be responsible for enhanced transmissions of EM waves through subwavelength hole arrays [26,27]. We have considered that the acoustic surface wave at the brass-water interface might play no role in the present transmission phenomenon, and shown that the SF resonance holds for acoustic waves by generalizing the proof of EM waves [38,39]. The SF resonance has some spectral features: the resonant wavelength is determined essentially by the lattice constant and is very sensitive to the incidence angle with accompanied by Wood's anomalies. Here, the experimental results for the enhanced acoustic transmission through the hole array in the 0.5 mm thick plate manifests the features of SF resonance. When the plate thickness becomes larger, the situations begin to divide for two types of waves. For EM wave, the transmission peak will diminish after the metallic film/plate becomes thick enough, because the holes have the cutoff. In sharp contrast, there is no cutoff for acoustic waves to propagate through the holes. When the thickness is large enough, for instance t = 2.3 mm, there can be multiple transmission peaks well below the Wood's anomaly, as shown in Figures 4(a) and 4(b). The measured spectra show the typical characteristics of FP resonance in terms of the phase values at the transmission maxima and minima. From Figure 4(c), the peaks are not sensitive to the incidence angle. In fact, these transmission peaks are caused by standing-wave-formed resonances of the acoustic wave establishing inside the hole channel. However, these resonances undergo a tuning, to some degree, by diffraction evanescent waves parasitical to a grating, and consequently deviate from the ordinary FP conditions while the plate thickness becomes comparable to the lattice constant, which will be further discussed later. Enhanced transmission of acoustic waves through bull's eye structure Soon after the discovery of extraordinary optical transmission through a metallic film with two-dimensional array of sub-wavelength holes, it was found that there can be enhanced and collimated transmission through a single sub-micron hole surrounded by finite periodic rings of indentations (denoted as bull's eye) [4]. We also examined the transmission of a bull's eye structure for acoustic waves. The bull's eye structure, shown in the inset of Figure 5, was fabricated by patterning both sides of a thin brass plate with concentric periodic grooves around a single cylindrical hole. The thickness of the brass plate is 1.6 mm, and the diameter of the central hole is 0.5 mm. The groove period is 2.0 mm, and there are a total of 15 grooves. The width and depth of each groove are 0.5 mm and 0.3 mm, respectively. In Figure 5, we showed the measured transmittances as a function of frequency for both bull's eye structure and the reference sample (a smooth brass plate of the same thickness). It can be seen that there is a transmission peak at 0.71 MHz for bull's eye structure, while such peak is missing for the reference sample. In Figure 5, we also plotted the power transmittance calculated by using COMSOL MULTIPHYSICS, a commercial finite-element solver. It can be seen that the predicted peak position agrees well with the experimental data. However, the measured transmittance is much lower than that predicted and the precise reason for this disagreement is yet to be uncovered. For ultrasonic waves in water, wavelength corresponding to 0.71 MHz is 2.1 mm, which is slightly larger than the groove period of bull's eye, 2.0 mm. This close correspondence is a strong clue indicating that the enhanced transmittance is due to the diffraction effect. It has been shown that enhanced acoustic wave transmission through hole arrays in perfectly rigid thin plate, where there can be no surface wave, may be related (and understood via Babinet's principle) to "resonant" reflection by its complementary structure, i.e., planar arrays of perfectly rigid disks [26,39]. In fact, both were found to be associated with the divergence in the scattering structure factor, owing to the coherent addition of the Bragg scattering amplitudes from the periodic array of holes or disks. As a result, a quasi surface mode with frequency close to the onset of the first diffraction order (wavelength λ slightly larger than the lattice constant a) always exists. Such modes are denoted "structure-factor-induced surface modes," or SF resonances. Since diffraction is the ultimate mechanism for the SF resonances, we expect the same to also apply to bull's eye structure, which can be viewed as having 1D periodicity along the radial direction. Besides the transmission enhancement, the collimation effect of the bull's eye structure is very striking [4,40]. As shown in Figure 6(a), the far-field acoustic wave on the transmission side is also in the form of a tight beam with a lateral dimension not exceeding the groove periodicity. The full width at half maximum (FWHM) divergence is ±2 o . As analyzed above, it is the coherent scattering which leads to the emergence of a strongly collimated beam in the far-field region. In Figures 6(b) and 6(c) we also plotted the scanned results at a distance of about 15 wavelengths from the transmission side of the surface, for both the bull's eye structure and the reference sample. Compared with the reference sample, the collimation effect for bull's eye structure is very evident. In addition, it is found by simulation that both the intensity of the acoustic wave field around the central hole region, as well as the collimation effect, would increase with the number of concentric grooves. This is reasonable, since the coherent scattering effect becomes stronger if more concentric grooves are involved. Fabry-Perot resonances tuned via diffraction evanescent waves For the hole arrays, we measured the acoustic samples with various plate thicknesses ranging from 0.5 mm to 3.1 mm, as plotted in the inset of Figure 7. In theory, we employed the mode expansion method to calculate analytically the transmission [39]. We found that the observed transmission peaks are the manifestation of a type of resonance mode that has FP and SF resonances as the two limits. The diffraction evanescent modes play an important role in interpolating between the two limits. To make explicit the role of diffraction evanescent waves, we retained the lowest cylindrical mode inside the holes and 5 lowest plane wave modes outside the holes, and obtained the resonant mode equation as is the diffraction evanescent wavevector, and J1 is the first order Bessel function. Equation (3) is instructive, since a vanishing right-hand side would directly yield the FP resonance condition t =nλ/2, λ being the wavelength. A combination of hole and periodic diffraction evanescent wave effects constitute the correction to the usual FP condition in the form of a non-zero right hand side, implying that the FP resonance can be tuned by varying the periodicity and area fraction of holes. We denote such resonances the FPEV resonances. In Figure 7 we show the measured and calculated FPEV resonance frequencies plotted as a function of inverse plate thickness. The FP condition is indicated by the black dashed straight lines, for n=1, 2, 3, 4, with slopes of 0.5, 1, 1.5 and 2, respectively. The FPEV frequencies are shown as solid black (normal incidence), and red dashed (20 o oblique incidence) lines. Compared with the FP resonances, it is seen that the FPEV resonances always occur at lower frequencies, as though the effective plate thickness is greater than t. The prediction of Equation (3), for the n=1 FPEV resonance, is shown as the blue line. Except in the region of very small a/t values, the blue line has a slope of 0.42. Thus the effect of the diffraction evanescent waves is to shift the resonance condition by ~16%, in the direction of smaller channel length. The difference between the prediction of Equation (3) and the black lines appears at the small t limit, where the transmission peak frequency shows a clear dependence on the incidence angle. This is characteristic of the surface-wave-like mode induced by the SF resonance. In fact, these transmission peak frequencies all occur at close to the Wood's anomaly, as required by the SF resonance condition. Thus the lowest frequency FPEV resonance, which shows little or no dependence on the incidence angle, is smoothly converted to the structure-factor-induced surface mode in the thin-plate limit. The diffraction evanescent wave contributions are dominant at the intermediate values of a/t. To a lesser degree, similar behavior can be observed for the higher order FPEV's. It is seen that as the ratio a/t increases, the lowest order evanescent waves (Equation (3)) can no longer account for the resonant frequency trajectory. Also, in the large a/t limit the curves also display pronounced incident angle dependence, in contrast to FP resonances which are nearly independent of the incidence angle. These are the signals for (1) the lateral scattering interaction is contributing much more to the resonant modes, hence the lowest order evanescent modes are no longer sufficient to account for such strong lateral interactions, and (2) with the increased lateral interaction, SF effect becomes more pronounced, implying incidence angle dependence. These spectral features also correspond to the different field distributions, as shown in Figure 8, where the surface field is localized on the holes for FPlike resonances and the interference pattern is seen in the region between the holes for SF resonances. Max It is interesting to note that the transmission of microwave through a metallic grating of 1D slits has the similar FPEV resonances for the incident polarization with E-field perpendicular to the silts, see open symbols in Figure 7 [37]. This is because the slits have no cutoff to the perpendicular polarization of EM waves, in the same physics as the holes to acoustic waves. The effective fluid model for thick plates For a very small a/t ratio, these resonance wavelengths are much larger than the lattice constant, allowing us to take a view of effective media. Here we employ a simple argument in the same fashion as the EM case [41] with the assumption of brass plate being rigid, and find that the hole array structure fabricated in a rigid plate and filled with a fluid (mass density 0  and bulk modulus  ) may be viewed as an effective fluid with the same thickness, effective mass density 0  and bulk modulus  . It is known that the acoustic wave is characterized by the pressure field, p , and velocity field, u . Averaging the pressure field in the holes, we get the effective pressure field in the effective fluid p p   . Requiring the acoustic energy flow across the surface to be the same for the hole array and the effective fluid, i.e. , we obtain the effective velocity u = u . Also the total acoustic energy for both systems are required to be the same, In Figure 4(c), we see two flat bands appear below 0.75 MHz, and they are the first and second FP resonances, see Figure 4 The effective fluid model allows us to use the holey or slotted hard plate to realize an acoustic medium, and provides some freedom to design acoustic materials, because some material parameters, difficult to be tuned, are related simply to the structural factors of the hole array or the slits. In Figure 9, we illustrated conceptually an acoustic prism made of such slotted hard plate. A detailed discussion is seen in reference [42]. Conclusion We investigated experimentally the acoustic transmission through subwavelength holes fabricated on brass plates at normal and oblique incidence within ultrasonic frequencies regime. The transmission phenomena for both hole array and bull's eye structure in thin brass plates, analogous to the observed enhanced transmission of EM waves through subwavelength hole arrays in a metallic film, exhibit the transmission enhancement because of the SF resonance. At the peak frequency, the transmission phase is nearly , indicating the out-of-phase oscillations of the acoustic field at two surfaces of the plate. For the hole array in thick brass plates, the transmission peaks of acoustic waves are related to the FPlike resonances inside the holes and therefore occur well below Wood's anomaly, since a hole has no cutoff frequency for acoustic propagation. By varying the plate thickness or channel length, one makes the transition from the FP resonance (thick plate limit) to the SF resonance (thin plate limit). Between the two limits there can be interesting deviation from FP resonance conditions, owing to the interaction of the diffraction evanescent waves. In the case of thick plates, the structure can be viewed as a new fluid with effective mass density and bulk modulus scaled, under long wavelength limit, by a factor of area fraction of the holes. The effective medium model describes well the transmission properties of the hole array within a range of incidence angle. Our discussion assumed the approximation of hard plates and did not take acoustic surface waves into account. With acoustic surface waves being involved, transmissions of acoustic waves through structured plates have found far richer and more complicated physical phenomena in the past few years and will attract more attentions in the future [43][44][45]. Although this subject is an old problem, its new phenomena may appear from time to time and the underlying mechanism waits to be unlocked.	2018-11-05T13:37:02.664Z	2013-08-28T00:00:00.000	{ "year": 2013, "sha1": "0a459d9fc11254316df72e1c1f188b0ac7c99e8f", "oa_license": "CCBY", "oa_url": "https://www.intechopen.com/citation-pdf-url/45572", "oa_status": "HYBRID", "pdf_src": "ScienceParsePlus", "pdf_hash": "d07d61af3fed56f96387d4234ba2e70318200465", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics" ] }
219286274	pes2o/s2orc	v3-fos-license	Polaronic Conductivity in Iron Phosphate Glasses Containing B2O3 We report on the electrical properties of glasses with nominal composition xB2O3–(100 − x)[40Fe2O3–60P2O5],x = 2–20, mol.%. The conduction transport in these glasses is polaronic and shows a strong dependence on Fe2O3 content and polaron number density. The changes in DC conductivity are found not to be directly related to B2O3, however structural changes induced by its addition impact frequency-dependent conductivity. All glasses obey Summerfield and Sidebottom procedures of scaling conductivity spectra indicating that the polaronic mechanism does not change with temperature. An attempt to produce a super-master curve revealed that shape of the conductivity dispersion is the same for glasses with up to 15.0 mol.% B2O3 but differs for glass with the highest B2O3 content. This result could be related to the presence of borate units in the glass network. Moreover, the spatial extent of localized polaron motions increases with the decrease of polaron number density, however, this increase shows a larger slope than for previously reported iron phosphate glasses most probably due to the influence of B2O3 on glass structure and formation of polarons. While Summerfield scaling procedure fails, Sidebottom scaling yields a super-master curve, which indicates that polaronic hopping lengths also change with changing polaron number density in these glasses. Introduction Iron phosphate-based glasses belong to a family of electronically conducting materials in which the conduction mechanism follows the small polaron hopping theory [1][2][3][4][5][6][7]. The polaronic conductivity in these glasses originates from the electron transfer between ferric and ferrous ions and, consequently, shows a strong dependence on the total amount of iron oxide, fraction of ferrous/ferric ions and average distance between them. While literature reports numerous studies on the influence of these parameters on the polaronic conductivity [8][9][10], as well as the influence of various modifier oxides, mainly transition metal or alkali oxides which themselves are potential active electrical contributors [6,[11][12][13][14][15][16][17][18], investigations of the impact of the additional glass-forming oxide on the polaronic transport in iron phosphate glasses are sparse. In ionically conducting glasses containing a fixed amount of alkali oxide, a gradual substitution of one glass-forming oxide by another produces a pronounced conductivity maximum at the glass composition with the highest fraction of structural units from both glass-forming oxides [19][20][21][22][23]. This effect, called mixed glass-forming effect, has been identified in various combinations of glass-forming oxides, including those based on phosphorous pentoxide and clearly reflects the importance of the glass structure in ionic transport. Bearing in mind that the mechanism of polaronic transport is very different in nature to the ionic one, it is unrealistic to expect such a straightforward relationship between polaronic conductivity and changes in the mixed glass-forming network. However, polaronic conductivity of iron phosphate glasses does depend on the local structural surroundings of iron ions and their long-range connectivity in the glass network which is inherently related to the basic features of the glass network. In this study, we report a detailed analysis of electrical properties of iron phosphate (IP) glasses of the molar composition 40Fe2O3-60P2O5 doped with boron oxide (IBP glasses). The structural properties of this glass system are relatively well examined [24][25][26]. It was shown that B2O3 enters the glass network as BO4 tetrahedral units and increases the chemical and thermal stability of the glass [24,25]. Moreover, in our previous study, we showed that the replacement of a small amount of Fe2O3 by B2O3 in glasses with fixed phosphate content (60 mol.%) induces local structural disorder which influences polaronic transport giving rise to a deviation from the Summerfield scaling of conductivity spectra [27]. The present work aims to investigate this interesting behavior further and to examine the structural-electrical properties relationship in glasses in which up to 17.7 mol.% of B2O3 is systematically added to 40Fe2O3-60P2O5 glasses at the expense of both Fe2O3 and P2O5. Such variation in composition in which Fe/P ratio is kept constant (0.67), involves smaller changes in overall Fe2O3 amount in comparison to the previously studied system [27], but different structural properties. Indeed, the structural study of both glass systems showed that the addition of B2O3 with simultaneous decrease of contents of both Fe2O3 and P2O5 decreases the thermal stability of glasses, whereas an opposite trend is observed for an exchange of just Fe2O3 by B2O3 [24]. Therefore, in this study, we present a detailed analysis of the electrical properties of these glasses in a wide range of temperatures and frequencies in order to address the role of glass structure in the processes of polaronic transport and to determine parameters that govern them. Materials and Methods The batch composition of the glasses selected for this investigation is xB2O3-(100−x)[40Fe2O3-60P2O5], x = 2-20, mol.%. Glasses were prepared by melting homogenous mixtures of reagent grade chemicals (B2O3-Fe2O3-P2O5) (Sigma Aldrich, Darmstadt, Germany) in appropriate quantities in highdensity alumina crucibles at 1150-1200 °C in the air for 1-2 h. The details of the preparation procedure are given in Ref [24]. Melt is quenched by pouring into a 1 cm × 1 cm × 5 cm steel mold. The resulting glass bars are then moved quickly to an annealing furnace, annealed at 475 °C for 3 h and slowly cooled down to room temperature. The glass samples are designated in accordance with the amount of boron oxide in batch composition. For instance, B-2 glass contains 2 mol.% of B2O3. The powder Xray diffraction (XRD-6000 diffractometer, Shimadzu, Kyoto, Japan) studies are performed to confirm the amorphous nature of the samples [24]. The exact composition of glass samples was determined by particle induced gamma-ray emission (PIGE) technique [28] using 3 MeV protons from the Ruđer Bošković Institute (RBI) Tandetron accelerator (High Voltage Engineering Europa B.V., Amersfoort, The Netherlands) and high-resolution high purity germanium (HPGe) detector (Canberra Industries, Meriden, CT, USA) to detect gamma rays from the measured samples. Gamma rays of boron (429 and 718 keV), phosphorus (1266 keV), iron (847 keV) and oxygen (871 keV) were used for quantitative analysis of spectra and determination of the elemental composition of samples. Proton elastic backscattered spectra (EBS) and collected proton charges were recorded simultaneously to help in the analysis. Details about experimental setup and data analysis procedure can be found elsewhere [28,29]. Before performing electrical/dielectric measurements annealed bars were cut into disks. Thin gold electrodes were sputtered as contacts onto both sides of 1 mm thick sample disks using sputter coater SC7620 (Quorum Technologies Ltd., Laughton, East Sussex, UK). Electrical and dielectric properties were attained by measuring complex impedance using an impedance analyzer (Novocontrol Alpha-AN dielectric spectrometer, Novocontrol Technologies GmbH & Co. KG, Montabaur, Germany) over a wide frequency and temperature range, from 0.01 Hz to 1 MHz at temperatures between 303 K and 523 K. The temperature was controlled to ±0.5 K. Results and Discussion The batch and measured compositions of all IBP glasses are shown in Table 1. It can be seen that compositions match reasonably well, with a clear increasing trend for B2O3, but slightly fluctuating changes for Fe2O3 and P2O5. For a meaningful analysis of the transport properties of glasses, it is important to consider accurate glass composition especially in the case like this, where the glass composition changes in steps of just a few mol.%. An exact glass composition along with the density of glasses (also shown in Table 1) enables precise calculation of charge carrier (polaron) number density which is an important parameter for polaronic transport. Therefore, in our analysis and discussion of electrical parameters we use the exact composition of glasses throughout the entire manuscript. Conductivity isotherms of B-4 glass are shown in Figure 1 and are typical spectra for all IBP glasses studied in this work. Normally, each isotherm exhibits two features, a plateau at lower frequencies that corresponds to the DC conductivity and dispersion at higher ones. The dispersive behavior is more evident at lower temperatures and frequencies, and shifts to higher frequencies with temperature. In Figure 1, the conductivity dispersion for B-4 glass is visible for temperatures up to 483 K. Further, it can be seen that DC conductivity is thermally activated whereas conductivity dispersion shows a weak temperature dependence. In the following, we present in detail the analysis of both quantities in terms of their temperature and compositional dependence. DC Conductivity It is well-known that above room temperature conduction in iron phosphate-based glasses is due to phonon-assisted hopping of polarons between nearest neighboring sites [1][2][3][4][5][6][7] and the DC conductivity is thermally activated with characteristic activation energy. The temperature dependence of DC conductivity, σDC, in this temperature range is usually expressed by Austin-Mott`s relation [1,2,5]: where σ0 is the pre-exponential factor, EDC is the activation energy for the DC conductivity, kB is the Boltzmann constant and T is the temperature (K). In this relation, the pre-exponential factor, σ0, contains important parameters for polaronic transport according to the relation: where C is the fraction of transition metal (TM) ions in lower oxidation state to total TM ion content and R is the average spacing between transition metal ions (R = N^(−1/3)), νph is the phonon frequency (≈10 12 -10 13 Hz), α is the rate of wave function decay, e is the electronic charge. The DC-conductivity variation with 1/T for selected glasses is shown in Figure 2a. The activation energy for DC conductivity, EDC, for each glass is calculated from the slope of log(σDCT) vs. 1000/T and listed in Table 2 along with the values of pre-exponential factor and DC conductivity. The values for DC conductivity at 303 K for all glasses are between 3.08 × 10 −10 (Ω cm) −1 and 8.97 × 10 −12 (Ω cm) −1 and activation energy changes from 0.63 to 0.69 eV which is in good agreement with the values of various iron phosphate-based glasses [6]. The dependence of DC conductivity of glasses upon B2O3 and Fe2O3 content is shown in Figure 2b,c. The DC conductivity decreases perfectly linearly for almost two orders of magnitude with decreasing Fe2O3 content, whereas it shows a scattered decrease with a simultaneous increase in B2O3. Observed trends suggest that the dominant parameter for the polaronic transport in these glasses is Fe2O3 content and not the addition of B2O3. Interestingly, the DC conductivity of binary 40Fe2O3-60P2O5 mol.% glass without B2O3 which we previously reported [30] lies precisely on the linear dependence of DC conductivity versus Fe2O3 content of the present glass system, see the black symbol in Figure 2c. This further supports our conclusion that the polaronic transport in these glasses is controlled solely by the iron oxide content and hence, polaron number density. Here, it should be noted that the polaron number density depends not only on the total amount of Fe2O3, but also on the fraction of Fe 2+ and Fe 3+ ions. For the Fe 2+ /Fetot ratio ≤ 0.5 the polaron number density is determined by the product of number density of the total iron ions and fraction of ferrous ions, whereas for Fe 2+ /Fetot ratio > 0.5 it is determined as a product of the number density of the total iron ions and the fraction of ferric ions [3,4,30]. In the previous structural study of this glass system [24], it was shown that there is no significant variation in Fe 2+ /Fetot ratio since the fraction of ferrous ions for selected glasses changes from 0.16 to 0.22. Therefore, we can approximate Fe 2+ /Fetot ratio to be constant throughout the entire series, which allows us to assume that the polaron number density is proportional to the overall Fe2O3 content. The values of polaron number density calculated using the average value of Fe 2+ /Fetot = 0.19 for all glasses are given in Table 3. For example, the polaron number density, Nv (polarons), for B-2 glass is calculated by multiplying number density of the total iron ions, N (Fe ions), see Table 3, and 0.19 (10.7 × 10 21 cm −3 × 0.19 = 2.03 × 10 21 cm −3 ). Indeed, the DC conductivity exhibits a similarly linear increase with increasing polaron number density, see Table 3. Summary of quantities and parameters of polaronic transport for glasses from xB2O3- Going further in the analysis of the polaronic transport in these glasses we turn our attention to the changes in the pre-exponential factor and parameters it contains, see Table 2. According to Austin-Mott`s theory of small polaron hopping, the conduction process can be characterized by either adiabatic or non-adiabatic hopping. In adiabatic hopping, the electron is at all times relaxed in the potential well of its lattice distortion and hence can respond rapidly to the displacement of the lattice due to the polarization field, while in non-adiabatic hopping the chance of the electron tunneling is rather small [1,2,5]. Hence, in adiabatic hopping conduction, the tunneling term exp(−2αR) in Equation (2) is approximately 1, thus αR becomes negligible. Although in literature there has been much discussion about whether adiabatic or non-adiabatic hopping picture is valid for oxide glasses at high temperatures [3,4], it was shown from DC conductivity data that nonadiabatic hopping model is more suited to describe polaron transport in iron phosphate-based glasses [27]. In order to check the adiabatic or non-adiabatic nature of the hopping conduction in glasses from this study, we have analyzed the dependence of log(σDCT) vs. activation energy, EDC at a fixed experimental temperature T = 423 K, see Figure 4. In such a plot, the slope equals to 1⁄kBT which enables us to determine the mechanism of the polaron hopping in the following way. If the experimental temperature is approximately equal to the temperature obtained from the slope, the process is adiabatic. On the other hand, if the temperatures differ, the process is non-adiabatic. The value of the temperature obtained from the slope for glasses in this study is T = 253 K which is very different from the chosen experimental temperature, T = 423 K This indicates the non-adiabatic hopping of small polarons in these glasses. This result is in accordance with the non-adiabatic hopping mechanism found in similar iron phosphate-based glasses: xB2O3- (40− [30], CaO-BaO-Fe2O3-P2O5 [31], Fe2O3-Bi2O3-P2O5 [32] and Fe2O3-CaO-P2O5 [33]. Going back to the influence of B2O3 on the polaronic transport in these glasses, it seems that the DC conductivity is strongly controlled by the content of Fe2O3 and independent of the B2O3 addition and the structural changes it introduces. The IR and Mössbauer spectroscopy study [24] showed that boron enters phosphate network as BO4 tetrahedral units forming B-O-B and/or B-O-P type bonds rather than B-O-Fe bonds. Consequently, the iron environment in these glasses is not significantly affected by the addition of B2O3. Both Fe 2+ and Fe 3+ ions have distorted octahedral coordination in all glasses similarly as in many other iron phosphate systems [9,17]. In addition, the O/P ratio for these glasses changes from 3.65 to 4.01, see Table 1, which suggests the presence of pyrophosphate (Q 1 ) and ortophosphate (Q 0 ) units in phosphate network with a slight tendency for depolymerization with increasing B2O3 content. The fact that the structural environment of iron ions remains unaffected by the introduction of borate units in the phosphate network can explain why the DC conductivity is not directly dependent on B2O3 addition. Frequency-Dependent Conductivity and Scaling Properties While the analysis of DC conductivity and its temperature dependence gives important information on the long-range transport of charge carriers in glasses, features of the frequencydependent conductivity provide insights into the processes of their localized motions. A relatively simple, but very helpful means of an analysis of the conductivity spectra over a wide range of frequency and temperatures is based on the application of various scaling procedures. In this respect, one of the simplest and widely used scaling procedures is Summerfield scaling [34,35] which uses experimentally easily assessable parameter-DC conductivity as a scaling factor. The Summerfield scaling is expressed by the relation: (σ(ν,T)/σDC(T)) = F(ν/TσDC(T)) and it can be understood as a mobility scaling; it is indicative that the role of temperature is to only speed up the charge carrier dynamics with increasing temperature without changing the conduction mechanism. For all glasses in this study, Summerfield scaling procedure yields a perfect master-curve demonstrating that the time-temperature superposition (TTS) is valid and that the conductivity mechanism is temperature-invariant, as shown for B-2, B-8, B-14 and B-20 glasses in Figure 5. This is in line with the scaling properties of various oxide glasses that exhibit ionic [36,37] or polaronic conductivity [30,37]. Interestingly, our previous study on xB2O3-(40-x)-Fe2O3-60P2O5 (x = 0-20, mol.%) system showed that glass with 2 mol.% of B2O3 fails to produce a master-curve when Summerfield scaling procedure is applied [27]. The cause of this deviation was found to be related to the influence of structure on the polaronic conductivity, i.e., local structural disorder induced by the addition of a small amount of B2O3 which allows changes of the number density of polarons and/or their pathways with temperature. In the glass system studied in this work, we have not observed the same effect and glasses with small amounts of B2O3 obey Summerfield scaling (cf. with Figure 5 for B-2 glass). This result can be related to the fact that within this glass series, B2O3 is added at the expense of both Fe2O3 and P2O5 which has a different influence on the cross-linking of the iron phosphate network. Previous structural study of these two glass systems [24] shows that a gradual substitution of Fe2O3 and P2O5 by B2O3 weakens the network structure as evidenced by the decrease in Tg, whereas an opposite trend is observed for glasses where only Fe2O3 was replaced by B2O3. For this reason, it seems likely that at low amounts of B2O3, small changes in the strength and rigidity of the iron phosphate network affect the mechanism of polaronic transport, i.e., by enabling or disabling the temperature dependence of polaron number density and/or conduction pathways. Furthermore, it is interesting to explore the influence of the glass composition and structure on the conductivity dispersion by applying superscaling in which all individual master-curves within the glass series are superimposed onto each other. Figure 6a,b shows the result of such super-scaling for all investigated glasses. As can be seen from the Figure 6a individual master-curves do not accurately overlap and a super-master curve could not be obtained. There are two possible reasons for such a result: either the shape of the conductivity dispersion changes or/and individual master curves get shifted along the x-axis with the changes in the glass composition. Intending to detect possible changes in the shape of the conductivity master curves, we have shifted the individual master curves along the x-axis in an attempt to produce a super-master curve, see Figure 6b. In the shifting procedure, the master-curve of B-2 glass was taken as a reference curve. All master-curves except that for B-20 glass perfectly overlap when shifted which suggests that the shape of their conductivity dispersion is the same. On the other hand, the master-curve for B-20 glass exhibits evidently different shape of the frequency-dependent conductivity, thus failing to superimpose onto the super-master curve of glasses from B-2 to B-18. Although it is hard to argue about the origin(s) of different shape of the conductivity dispersion for B-20 glass without more detailed structural characterization of the whole glass series and investigation of the glasses containing higher B2O3 amounts, it is likely related to the structural features of the glass at highest B2O3 content within this series. Possibly, the extensive inclusion of BO4 tetrahedra and formation of B-O-P and B-O-B bonds in this glass influences the local dynamics of polarons. The values of the shift needed to produce a super-master curve for B-2 to B-18 glasses are given in the legend of Figure 6b. Considering the magnitude and direction of the shift, it can be seen that values are very small and they do not imply any trend with the change of glass composition. The shift in conductivity super-master plot has been reported for various series of ionically conducting oxide glasses where it has been correlated with the alkali oxide content and hence number density of mobile ions [38,39] as well as changes in the typical length of the hop of the ions with their number density [40]. Recently, we have shown that the Summerfield master-curves of polaronically conducting iron phosphate glasses of similar composition and structure, but widely ranging fraction of ferrous ions (0.23 ≤ Fe 2+ /Fetot ≤ 0.58) also exhibit a shift on the x-axis upon super-scaling [30]. Our analysis revealed that the observed shift can be correlated to the product of polaron number density, Nv and the spatial extent of localized motions of polarons, <r 2 LOC(∞)> 1/2 . Therefore, it is further of interest to consider the characteristic spatial extent of localized hopping of polarons which can be assessed from the scaling property of the permittivity spectra. Scaling of Permittivity Spectra and Length Scales of Polaronic Transport The spectra of the real part of permittivity, ε′ (ν), at different temperatures for B-10 glass is shown in Figure 7a as a representative for all other glasses in this study. At each temperature, permittivity spectrum exhibits a typical frequency-dependence; at higher frequencies, the real part of the permittivity tends to a constant value, ε′∞, which results from rapid polarization processes occurring in the glasses under an applied field whereas, with decreasing frequency, it increases and approaches a limiting low frequency plateau denoted as the low-frequency static permittivity, ε′s. The static permittivity is associated with the polarization effects of the polarons with respect to the immobile glass matrix and determines the difference Δε = ε′s−ε′∞ called dielectric strength of relaxation [40,41]. Here it should be noted that permittivity spectra of polaronic glasses exhibit a well-defined low frequency plateau, as shown in Figure 7a, which is not the case for most ionically conducting glasses since in their spectra effects of electrode polarization dominate at low frequency, hence, masking the permittivity plateau. In contrast, polaronic glasses exhibit a small increase of the real part of permittivity below the point corresponding to static permittivity due to a negligible electrode polarization and surface effects. The imaginary part of permittivity, ε″, increases linearly with decreasing frequency in a log-log plot, see Figure 7c, due to the contribution of DC conduction [42]. Since permittivity and conductivity are directly related quantities, the scaling properties of conductivities are mirrored in the scaling properties of permittivity. The Summerfield scaling procedure for the permittivity spectra includes rescaling of the real part of permittivity by subtracting ε′∞ and multiplying by T, while the x-axis is scaled by the product (TσDC(T)). Expectedly, from the results of conductivity scaling, the Summerfield scaling procedure yields a perfect permittivity master curve for all glasses in this study. The representative plot of permittivity master-curve is shown for B-10 glass in Figure 7b. For each glass, the scaling of the permittivity spectra yields a parameter Δε T which can be related to the typical spatial extent of localized motions of polarons, <r 2 LOC (∞)> 1/2 . via relation [30,43]: where kB is Boltzmann's constant, ε0 is the permittivity of free space and Nv is the number density of polarons. For each glass, the parameter ΔεT is determined from the onset of the static permittivity plateau in permittivity master-curve as shown graphically for glass B-10 in Figure 7b. The obtained values of the extent of the localized motions of polarons are given in Table 3 and shown in Figure 8. As can be seen from Table 3 and Figure 8, the value of < rLOC 2 (∞)> 1/2 increases from 2.53 Å to 3.03 Å as the polaron number density, Nv, decreases from 2.03 × 10 21 cm −3 to 1.58 × 10 21 cm −3 for B-20 glass. Such a trend corroborates our previous findings [30] that the spatial extent of the localized polaron hopping decreases with increasing polaron number density and represents a realistic estimate of the extent of polaron cavity. Indeed, from Table 3 it is visible that the magnitude of the < rLOC 2 (∞)> 1/2 values is close to that of polaron radius, rp, calculated from the equation proposed by Bogomolov and Mirilin: rp = (1/2) (π/6N) 1/3 where N is the total number of iron ions [44,45]. Also, it should be noted that the values of < rLOC 2 (∞)> 1/2 for these glasses are higher than those obtained for glasses from xHfO2- [30]. This result is in line with their lower values of polaron number density, compare Table 3 and Table 3 in [30]. However, it is interesting to see that the decrease of < rLOC 2 (∞)> 1/2 with Nv for the present glass system is characterized by a larger slope than that of HfO2-Fe2O3-P2O5, CeO2-Fe2O3-P2O5 and HfO2-Fe2O3-B2O3-P2O5 glasses [30], see Figure 8. This observation can be related either to the influence of borate units on the formation of polarons or to the inherent property of the polaronic transport in iron phosphate glasses with low Nv. In order to ascertain the actual cause of this behavior, more investigations on the iron phosphate glasses with a similar range of polaron number density but free of B2O3 are required. These studies are underway. Further, stimulated by the fact that all glasses in our study exhibit a well-defined static permittivity plateau which enables us to precisely determine dielectric strength, Δε, we apply an alternative scaling procedure for conductivity spectra proposed by Sidebottom which is expressed by the form: (σ(ν,T)/σDC(T)) = F(νε0Δε/σDC(T)) [40]. The Sidebottom scaling procedure considers the simultaneous change in the typical hopping distance of the charge carriers with changes of their number density and it is considered as truly universal, since it applies whenever scaling is possible at all, i.e., when the shape of the conductivity dispersion does not change with temperature [45]. Figure 9. shows that Sidebottom scaling procedure produces conductivity master-curve for B-2, B-8, B-14 and B-20 glasses and the same result was obtained for all other glasses in this study. Finally, we show that Sidebottom scaling yields a perfect conductivity super-master curve for glasses from B-2 to B-18, see Figure 10. Having in mind that the Summerfield scaling failed to do so, this result implies that the lengths of a polaron hop change with composition in these glasses. Conclusions The analysis of electrical properties of xB2O3-(100−x)[40Fe2O3-60P2O5], x = 2-20 mol.% nominal composition glasses in a wide range of temperatures and frequencies reveal that the DC conductivity depends strongly on the polaron number density determined by the overall amount of Fe2O3. The polaronic transport in all glasses is non-adiabatic. The addition of B2O3 is not found to be the factor that influences long-range DC conductivity in this glass system, however, it seems that it affects the short-range dynamics of polarons as evidenced in the features of the frequency-dependent conductivity. The consideration of the scaling features of conductivity spectra shows that all glasses obey Summerfield and Sidebottom scaling procedures indicating the validity of the timetemperature superposition principle. Further, the superposition of individual master curves onto each other in an attempt to produce a super-master curve revealed two important results. First, the shape of the conductivity dispersion is the same for glasses with up to 15.0 mol.% B2O3 but differs for glass with 17.7 mol.% B2O3. A different shape of the frequency-dependent conductivity of the latter glass is most probably related to the structural features of the glass at high B2O3 content. Second, while the Summerfield scaling procedure appears to fail, Sidebottom scaling yields a super-master curve for glasses with up to 15.0 mol.% B2O3, which indicates that while the time-temperature superposition principle is fulfilled, polaronic hopping lengths also change with changing polaron number density in these glasses. The calculated spatial extent of localized motions of polarons, assessed from scaled permittivity spectra, for all glasses are close in values to their polaronic radii. Interestingly, the dependence of the values of the spatial extent of localized polaron motions on the polaron number density complements well the dependence previously found for iron phosphate glasses containing HfO2 and CeO2 with larger polaron number density. However, this dependence for glasses from this study is characterized by a larger slope which could also be related to the influence of B2O3 on glass structure and formation of polarons. Conflicts of Interest: The authors declare no conflict of interest.	2020-06-04T09:12:22.887Z	2020-05-30T00:00:00.000	{ "year": 2020, "sha1": "adb0e7383031084894a727af7e33acaf8b038239", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/1996-1944/13/11/2505/pdf", "oa_status": "GOLD", "pdf_src": "ScienceParsePlus", "pdf_hash": "c9964a77ad21f448b343e94aed57e84b676fa0b3", "s2fieldsofstudy": [ "Materials Science", "Physics" ], "extfieldsofstudy": [ "Materials Science", "Medicine" ] }
40745433	pes2o/s2orc	v3-fos-license	STUDIES ON STINGLESS BEES 827 ELECTROANTENNOGRAPHY IN THE STUDY OF TWO STINGLESS BEE SPECIES ( HYMENOPTERA : MELIPONINI ) The first recorded electroantennographic preliminary studies on stingless bees have been performed using two species of Frieseomelitta from Brazil. Experiments with F. silvestrii and F. varia showed that antennae respond to hexane extracts of heads and abdomens of both species and posterior tibia of F. silvestrii (which carry plant resin), as well as to the pure compounds 2-heptanol and 2-nonanol, which occur in the mandibular glands of both species, and to the terpenes α-cubebene, humulene, and β-caryophyllene found on their tibia and in the cerumen of their nests. INTRODUCTION Stingless bees of the hymenopteran group Meliponini display complex communication patterns, and much of that communication must stem from glandular secretions.The study of pheromones in stingless bees is still in its infancy.The most important exocrine glands in stingless bees that might be implicated in pheromone production include, in the head, the mandibular glands; in the thorax, the thoracic salivary glands; and in the abdomen, the Dufour gland and epidermal glands. Mandibular gland secretion of stingless bees has been more extensively studied because, among them, it appears to play an important part in chemical communication (Kerr et al., 1963).Engels and Francke have together analyzed the cephalic secretions of a large number of species and made relations to age and development (Engels et al., 1993(Engels et al., , 1997;;Francke et al., 2000;and earlier papers).We have only recently made the first chemical studies of the Dufour glands of stingless bees (Cruz-López et al., 2001;Patricio et al., 2003).Some early studies with pure chemicals were made to study their effects on the behaviour of stingless bees (Weaver et al., 1975;Smith & Roubik, 1983), but the subject has progressed slowly. The electroantennograph technique (EAG) measures the summation of receptor potentials from responding cells in insect antennae (Kaissling, 1971).It has been much used to identify active pheromone substances in a mixture of volatile compounds extracted from insects, and has played an important part in the identification of lepidopteran pheromones.As for honeybees, it is known that they detect odours by using olfactory receptor cells in their antennae (Lacher, 1964;Vareschi, 1971;Williams et al., 1982), and no doubt stingless bees do too. The recognition of active compounds in EAG studies can be a short-cut in identifying compounds recognized by antennal receptors that may be used in pheromone communication.Until our present work there have been no EAG studies on any stingless bees.We have analyzed the mandibular glands (Cruz-López et al., 2002), the hind tibia that carry plant resin and nest propolis (Patricio et al., 2002), and the Dufour glands (Patricio et al., 2003) of three species of Frieseomelitta.They therefore provided a good starting point from which to initiate EAG studies.In our experience, while EAG with Lepidoptera is relatively easy and yields large potential responses, with ants it is very difficult and gives much lower potentials (Wadhams & Morgan, unpublished).We give here a full account of the first studies by EAG with stingless bees, using two species of Frieseomelitta from Brazil, that show that this approach to studying communication chemicals is practical and promising.A brief report on our study of the chemicals from the heads of Frieseomelitta has already been published (Cruz-López et al., 2002). METHODS AND MATERIALS Insects.Workers of F. varia Lepeletier 1836 and F. silvestrii Ihering 1912 were provided from the collection of Professor Paulo Nogueira Neto and from the Laboratório de Abelhas, Instituto de Biociências, Universidade de São Paulo, Brazil.F. varia were collected at Ribeirão Preto in northern São Paulo State.F. silvestrii came originally from Luiziania in Goiás State, near Brasília.Groups of about 50 worker bees with brood cells, honey, and pollen pots were used, and supplied with water. Extracts were prepared by placing the body parts of worker bees (heads, abdomens, and hind legs), and of pieces of cerumen taken from their nest, in small glass tubes containing hexane (0.5 ml), and allowing them to stand for 30 min.Pure compounds (geranylfarnesol, α-cubebene, humulene, and βcaryophyllene, 10 mg each) were dissolved in hexane (10 ml). Electroantennography was performed using excised antennae of F. silvestrii and F. varia workers.Electrodes were drawn from 1 mm i.d.borosilicate tubes to produced capillaries with a taper length of 0.5 cm with the tips filled with saline solution prepared according to Maddrell (1969) but omitting the sucrose.An antenna was placed between two electrodes, one inserted at the base of the antenna and serving as the reference electrode, and another, the recording electrode, placed at the tip of the antenna.In all, 20 antennae were used for F. silvestrii and five for F. varia. For the electroantennogram, hexane solution (10 ml) was impregnated onto a strip of filter paper, 15 x 1.5 mm, which was then placed in a disposable glass pipette.While air was continuously blown over the antenna, the pipette tip containing the strip of filter paper was inserted in a hole at the end of the tube carrying the air stream, thus delivering vapor from an unknown quantity (equivalent to one worker head, abdomen, or tibia) of extract or from 10 mg of one of the pure compounds.The vapor was puffed from the filter paper with a solenoid-controlled nitrogen stream (3 ml/sec).The signals, amplified with an AC coupled amplifier, were displayed on an oscilloscope (Devices 3131, Dynamic Electronics Limited).A permanent trace was obtained with a Devices M2 pen recorder.Only two replicates of the experiment were made with F. varia because only a limited number of live bees were available.Results of F. silvestrii were analyzed using student's t test.No statistical test was performed with the F. varia results because of the small number of replicates. Satisfactory electroantennography experiments were conducted with F. silvestrii, but by the time the experiments were performed with F. varia, only two good replicates were obtained.Table 1 shows the EAG responses of F. silvestrii to abdominal, head, and hind leg extracts in hexane.A typical electroantennogram is shown in Fig. 1.Also tested were two important components of the mandibular secretion of F. silvestrii: 2-nonanol and 2-undecanol; three volatile terpenes from the posterior tibia: α-cubebene, humulene, and β-caryophyllene; and geranylfarnesol, a component of the Dufour gland (Patricio et al., 2003).All the extracts and synthetic compounds except geranylfarnesol gave significant EAG responses.One of the sesquiterpenes from the tibia, α-Cubebene, (Patricio et al., 2002) gave a particularly strong response (Table 1), although there was less of this compound found on the tibia of F. silvestrii (Patricio et al., 2002) than β-caryophyllene (which gave a weaker response). However, from the analytical conditions used in the analysis of resin on the tibia (Patricio et al., 2002) it was not possible to know which enantiomer of βcaryophyllene was present on the tibia, since commercially available (-)-β-caryophyllene was used.The limited experiment on F. varia nevertheless gave strong responses to solvent extracts of abdomen, head, and cerumen (Table 1).Nonanal and 2-undecanol were the important mandibular compounds to test with F. varia.Shortage of live bees limited the experiment on F. varia; nevertheless, responses to solvent extracts of head, abdomen, cerumen, and the pure compounds nonanal and 2-nonanol were all strong, and only that to 2-undecanol was comparatively weak (Table 1). DISCUSSION The electroantennographic (EAG) technique has been used to determine which of the compounds present in a mixture of volatiles can stimulate olfactory receptors.No physiological information of this type was available for stingless bees, therefore preliminary EAG studies were undertaken using the antennae of F. varia and F. silvestrii.Head, abdominal, and tibial extracts of F. silvestrii all gave equally strong antennal responses.The responses of F. varia antennae to body extracts were very similar to those obtained with F. silvestrii.Strongest responses of the pure compounds were given by α-Cubebene and 2-nonanol.It is noteworthy that in both species 2-nonanol gave stronger responses than 2-undecanol.Further planned experiments that included extracts of individual dissected glands were cut short by the decline of the colonies.Geranylfarnesol is only a trace component in the Dufour glands of F. silvestrii but is the major substance in the closely related F. silvestrii languida (Patricio et al., 2003).It was available for tests, but its poor response, not significantly different from that of the hexane control, could be due to low volatility because of its relatively high molecular mass (358 mass units).We have reported earlier on the plant resin substances found on the hind tibia of these species (Patricio et al., 2002) and we have tested three of the substances found there both on F. silvestrii and an extract of the cerumen from the nest of F. varia.The results show that individual substances and mixtures of compounds present in the abdomen, cerumen, and on the tibia of Frieseomelitta bees can elicit olfactory responses in bee antennae.This is consistent with the hypothesis that such volatiles can elicit behavioral responses.What kind of behavior these materials induce has yet to be discovered. Our brief study shows that EAG can provide a route to learning more about chemical communication in stingless bees.From the mixtures of chemicals found in glands, the EAG can select substances that show a positive response and deserve further study in behavioral experiments.The relatively simple mixture of substances in F. silvestrii makes it a suitable candidate for behavioral studies with synthetic compounds.**Appling 10 µl of the hexane extract (see Methods and Materials) and 10 µl of a solution (1 g L -1 ) of pure compounds.† The absolute configuration of the β-caryophyllene on the tibia (Patricio et al., 2002) has not been determined. TABLE 1 Mean and standard deviation of the EAG responses of the Frieseomelitta silvestrii and F. varia antennae to some body extracts and synthetic samples of compounds identified in their exocrine glands.	2017-06-17T05:00:16.515Z	2004-11-01T00:00:00.000	{ "year": 2004, "sha1": "2e5e8d9d28980c5751ffda87e82155fa71e62501", "oa_license": "CCBY", "oa_url": "https://www.scielo.br/j/bjb/a/WHWdq7x8JqZtCKjGhn9jJBJ/?format=pdf&lang=en", "oa_status": "GOLD", "pdf_src": "ScienceParseMerged", "pdf_hash": "2e5e8d9d28980c5751ffda87e82155fa71e62501", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [ "Biology", "Medicine" ] }
125856427	pes2o/s2orc	v3-fos-license	A Randomized Algorithm for Robust BMI Optimization Robust bilinear matrix inequality (BMI) optimization, which is to minimize an objective function subject to a parameter dependent BMI constraint, is considered. A recursive algorithm employing a branch-and-bound technique and randomization of a parameter is provided for solving the problem. When the algorithm ﬁnds a solution, this so-lution satisﬁes the parameter dependent constraint with a prescribed accuracy in a probabilistic sense. Furthermore, the objective function value at that solution ensures that the feasible set whose objective function value is less than this value is too small to be found. Introduction Decision and control under uncertain environment can be recast as the robust optimization [1] which is to minimize an objective function subject to a parameter dependent constraint. For example, robust output feedback controller design [2], switching rule design of a linear switched system [3], operation planning of uncertain power systems [4] are rewritten as the robust optimization. However, the optimization is difficult due to its parameter dependency [5]. A randomization technique is one of the solutions for its difficulty [5]. For example, scenario approach [6] gives a standard convex optimization with many randomly sampled constraints. They show that an appropriate selection of the number of random samples leads to an approximate solution in a probabilistic sense. In addition to this, necessary number of random samples is of polynomial order of given accuracy parameters and the number of decision variables. This approach is extended to a robust nonconvex optimization in [7]. These researches clarify how many random samples is needed for finding an approximate solution with prescribed theoretical guarantee. However, these algorithms require us to solve a standard nonconvex optimization with many randomly sampled constraints. Furthermore, they do not provide any algorithm for solving this problem with some guarantee. Another approach is a recursive algorithm employing randomly sampled constraints one by one. For a robust convex problem, there are some recursive type randomized algorithms [8] for robust convex optimization. It has also been extended to robust nonconvex feasibility problem [3,9]. However, these algorithms are specialized for switching rule design of linear switched systems. Then, these do not deal with optimization problem. In this paper, we consider a robust bilinear matrix inequality (BMI) optimization whose constraint is parameter dependent and bilinear with respect to decision variables. The BMI optimization is a standard form of nonconvex optimization [10][11][12] and this problem is its generalization. Our aim is to characterize difficulty of the problem from a viewpoint of randomized algorithms. To achieve this goal, we develop a randomized algorithm with guarantee on feasibility in a probabilistic sense. We employ branch-and-bound techniques [13]. Then, we show that this solution satisfies the constraint with a prescribed accuracy in a probabilistic sense. Furthermore, we clarify the volume of feasible set whose objective function value is less than the obtained objective function value is enough small with high probability. We hasten to note that these techniques do not lead to a solution within a reasonable computational time due to NPhardness of the BMI optimization [14]. We have already developed a randomized algorithm for robust nonconvex problem with parameter dependent decision variable [4]. This algorithm utilizes a full randomization technique, that is, it employs random sampled constraints and random sampled decision variables. It stops within finite number of iterations and these numbers are of polynomial of the problem size. Furthermore, we can provide guarantee on feasibility and optimality in a probabilistic sense. However, in general, the feasible set whose objective function value is enough small to find its element by using the full randomization technique. In this case, our randomized branch-and-bound algorithm could find a solution. This paper is organized as follows. In Section 2, we formulate the robust BMI optimization. In Section 3, we show a randomized branch-and-bound algorithm. A numerical example is shown in Section 4. Finally, we provide concluding remarks in Section 5. Problem Formulation Let us consider a robust bilinear matrix inequality (BMI) optimization: where x ∈ X ⊆ R n x and y ∈ Y ⊆ R n y are decision variables, c x ∈ R n x and c y ∈ R n y are coefficients of the objective function, and δ ∈ Δ ⊆ R d is an uncertain parameter. The BMI Proceedings of the 48th ISCIE International Symposium on Stochastic Systems Theory and Its Applications Fukuoka, Nov. [4][5]2016 constraint F(x, y, δ) ≤ 0 is defined by where F i j (δ) = F T i j (δ) for i = 0, 1, 2, . . . , n x and j = 0, 1, 2, . . . , n y . Furthermore, F i j (δ) could be nonlinear in δ ∈ Δ. In this paper, we assume that the sets X and Y are hyper cubes which are defined by where x i ≤ x i and y j ≤ y j . We can also deal with other general sets such as ellipsoids, polytopes, and so on, if we add appropriate constraints to F(x, y, δ) ≤ 0. Even if there is no uncertainty, it is difficult to find a solution within a reasonable computational time. Furthermore, the problem is also difficult from a viewpoint of its parameter dependency. This is because we cannot verify whether given candidate solution (x, y) is feasible, or not, within finite computational time. A Randomized Branch-and-Bound Algorithm for Robust BMI Optimization Now, we consider a randomized branch-and-bound algorithm for robust BMI optimization. To employ a branchand-bound technique, we have to prepare for algorithms for finding upper and lower bounds of the optimal value of the problem (1) within a subset of X × Y. Lower bounds It is well known that a linear matrix inequality (LMI) relaxation approach for the standard BMI optimization leads to a lower bound of the optimal value of the original problem. Now, we apply this technique for the robust BMI optimization. An LMI relaxation of a parameter dependent BMI constraint is given by If F(x, y, δ) ≤ 0 for a fixed δ ∈ Δ, then F L (x, y, xy T , δ) ≤ 0 holds. Since x and y belong to hyper cubes X and Y respectively, (x, y, W) has to satisfy for any i = 1, 2, . . . , n x and j = 1, 2, . . . n y . Thus, the LMI relaxation of the problem (1) is Please notice that the relaxation problem (2) is still difficult due to its parameter dependency. Now, we introduce the probability measure P δ on the set Δ. Then, we employ ranodmization techniques for solfing the problem (2) in a probabilistic sense. We summarize an algorithm in Appendix. Upper bounds In this section, we introduce a full randomization algorithm for finding an upper bound of the robust BMI optimization within a subset X s × Y s of X × Y. We assume that X s and Y s are also hyper cubes. We employ randomization of all decision variables x and y and an uncertain parameter δ from the sets X s , Y s , and Δ, respectively. To generate random samples, we introduce a probability measure P δ on Δ and the uniform probability measure P xy over X s × Y s . An algorithm is shown in Algorithm 1. The algorithm stops within a finite number of random samples from the sets X s , Y s , and Δ. In fact, the maximum number of random samples from the set X s × Y s is¯ and that from the set Δ isk¯ . When we select appropriate numbersk and¯ , we can guarantee quality of an obtained solution. Draw (x ( ) , y ( ) ) ∈ X s × Y s according to P xy ; 3: for k := 1 tok do 4: Updatê 10: end for 11: return (γ,x,ŷ) Then, the following statements hold. The probability that (x,ŷ) satisfies is less than or equal to β a . 2. The probability that this problem satisfies That is, obtained solution (x,ŷ) is a feasible solution in a probabilistic sense with high probability 1 − β a . This implies that (x,ŷ) is a feasible solution in a probabilistic sense with the same probability level. We can also see that objective function value at (x,ŷ) is an upper bound of the optimal value of the problem (1) with this probability. Furthermore, the objective function valueγ at (x,ŷ) is an approximate minimum in a probabilistic sense with probability 1 − β b . Implementation of the algorithm is easy because of simpleness of the algorithm This is one of the advantages of the algorithm. However, since the feasible set of the problem is generally too small, the algorithm cannot find a solution in many cases. To solve this issue, we employ a branch-andbound technique. A randomized branch-and-bound algorithm Now, we show a randomized branch-and-bound algorithm for the problem (1). We utilize Algorithms 1 and 3 for finding upper and lower bounds. In Algorithm 2, we prepare for a provisional optimal valueγ and a provisional optimal solution (x,ŷ). Furthermore, we construct a subproblem set S. Then, we iterate the following procedures. We first extract a subproblem (C,γ L (C)) from S such that the lower bound of the optimal value of the problem (1) within C attains minimum among all subproblems in S 1 . Then, we split the set C into two distinct sets C 1 and C 2 such that C = C 1 ∪ C 2 . We solve the original problems (1) and the relaxation problems (2) in C 1 and C 2 , respectively Based on the results, upper and lower bounds of the optimal value are updated. If obtained lower bound is greater than the current upper bound, we prune this region. This is because there is no optimal value in this region. Otherwise, we insert this subregion into the problem set S. In this loop, if the upper boundγ and the minimum lower bound of the optimal value in the solution set S become close enough, the algorithm stops withγ and (x,ŷ) as output. Properties of the algorithm are follows. 3. The probability thatγ L ≤ γ * + μ holds is greater than 1 − β, where γ * is the optimal solution of the problem (1) and μ is a tolelance which is defined by Proof: 1. Let us consider two events, F κ and B κ : F κ : At the κ-th executions of Algorithm 1, the -th random sampled pair ofx (κ ) andŷ (κ ) satisfiesk 1 (κ) random sampled constraints, that is, Note that this event implies the event {x Null}. B κ : A pair ofx (κ ) andŷ (κ ) satisfies That is, a pair ofx (κ ) andŷ (κ ) is not a probabilistic solution with a given accuracy α. Our aim is to show the probability that any pair ofx (κ ) and y (κ ) , κ = 1, 2, . . ., which is output of the algorithm is not a probabilistic solution is less than or equal to β. That is, we try to show holds. 2. We can see that this statement holds due to a property of branch and bound techniques. 3. When we selectk 2 (κ) and¯ 2 according to (5) and (6), Lemma 2 in Appendix implies thatγ L i satisfieŝ with probability 1 − 6β/(π 2 k 2 ), where γ * L (C i ) is the optimal value of the problem (2) over C i ,γ L i is its lower bound given at Step 5 of Algorithm 2 and r L (C i ) is the largest radius of a ball within the set That is, Since eachγ L i satisfies the above inequality with probability 1 − 6β/(π 2 k 2 ) the probability that all ofγ L i satisfy the above inequality is greater than or equal to Notice thatγ L is the minimum in S. That is, holds with probability 1 − β. Then, when we define the set we cannot find a feasible solution in S c . Thus, we can interpret γ(S c ) = ∞. Replacing min (C,γ L )∈S γ * L (C) with γ * , we therefore see that holds with probability 1 − β. Select a subproblem (C,γ L ) in the subproblem set S such thatγ L is the minimum in S; 3: Split the set C into two hyper cubes C 1 and C 2 such that C = C 1 ∪ C 2 and C 1 ∩ C 2 = ∅; 4: for i := 1, 2 do 5: Solve the robust convex optimization (2) within C i by executing Algorithm 3 withk 2 (κ) and¯ 2 . Then, obtain the objective function valueγ Li at the obtained solution; 6: ifγ Li >γ orγ Li = Null then Solve the problem (1) within C i by executing Algorithm 1 withk 1 (κ) and¯ 1 (κ). Then, obtain a triplet (γ Ui ,x Ui ,ŷ Ui ) as output of the algorithm; 10: ifγ Ui <γ then This theorem says that obtained solution (x,ŷ) satisfies a parameter dependent constraint with prescribed probability. That is, γ * C ≤γ, where γ * C is the optimal solution of a chance constrained version of the original problem 1: The statement 2 and 3 implies thatγ ≤ γ * if we select enough small ε and r m . We therefore see that γ C ≤γ ≤ γ * if we select enough small ε and r m . That is, x x = 1 and n y = 1. This model is taken from [10]. The nominal values of the coefficient matrices are All parameters are allowed to vary 10% around these nominal values. We have set the set X and Y according to the paper [10]: We have selected parameters as α = 0.01, β = 0.01, ε = 0.01, and r m = 1.0 × 10 −4 . We have obtained the following solution (γ,x,ŷ) = (−0.1913, 1.0476, 1.4639). In this example, we have generated about 800,000(≈¯ 1 (κ)) random samples from the set (x, y) at Step 9 in Algorithm 2. On the other hand,k 1 (κ) was about 2,500. By employing posteriori analysis, we have examined this solution satisfies the statements of the theorem with probability 1 − 0.001. Furthermore, when we have employed Algorithm 1 for solving the problem (1), we have found 0.1725 which is clearly infeasible. In this case, selecting α a = α b = β a = β b = 0.01, the maximum number of random samples is about 500,000. Even if the full randomization technique cannot find a solution, our algorithm could find a solution. Concluding Remarks We have proposed a branch-and-bound randomized algorithm for robust BMI optimization. When the algorithm stops with output, the obtained solution satisfies a parameter dependent constraint with high probability. Furthermore, it is shown that the obtained objective function value is close to optimal value in a probabilistic sense. In fact, the volume of feasible set whose objective function value is less than the obtained suboptimal value is enough small with high probability. In addition to this, we have shown that erroneous pruning in the branch-and-bound algorithm occurs with small probability. where [·] + is the projection onto the cone of positive semidefinite matrices [5] and p, q, r, and s are defined by p(x, y, W, δ) = diag(p 11 , p 12 , . . . , p 1n y , p 21 , p 22 , . . . , p 2n y , p n x 1 , p n x 2 , . . . , p n x n y ), q(x, y, W, δ) = diag(q 11 , q 12 , . . . , q 1n y , q 21 , q 22 , . . . , q 2n y , q n x 1 , q n x 2 , . . . , q n x n y ), r(x, y, W, δ) = diag(r 11 , r 12 , . . . , r 1n y , r 21 , r 22 , . . . , r 2n y , r n x 1 , r n x 2 , . . . , r n x n y ), s(x, y, W, δ) = diag(s 11 , s 12 , . . . , s 1n y , s 21 , s 22 , . . . , s 2n y , s n x 1 , s n x 2 , . . . , s n x n y ), p i j , q i j , r i j , and s i j are given by For simplicity, we introduce a vector z ∈ R n which is composed of x, y, and W. To solve this problem, we introduce a randomized algorithm (Algorithm 3) based on the cutting plane technique based on the analytic center [8]. The analytic center is the minimum point of the logarithmic barrier function B(z) − N i=1 ln b i − a T i z determined by a convex polytope Z {z ∈ R n : a T i z ≤ b i , i = 1, 2, . . . , N}. Note that it is easy to find the analytic center due to convexity of the barrier function. Algorithm 3 is a randomized algorithm based on the analytic center. When we introduce the feasible set of the problem (2) and the largest radius r S L of a ball which is contained by the set S L as S L = {z ∈ R n : f (z, δ) ≤ 0, ∀δ ∈ Δ}, r S L = max r,z c {r : {z ∈ R n : (z − z c ) T (z − z c ) ≤ r 2 } ⊆ S L }, we can provide properties of the algorithm. Then, ifγ Null, the probability thatγ,x,ŷ andŴ satisfies is greater than or equal to 1 − β, whereγ * α is the optimal value of the chance constrained problem for k := 1 tok do 3: Draw δ k from the set Δ according to the probability measure P δ ; 4: if f L (z −1 , δ k ) 0 then 5: Compute a subgradient g := ∂ z f L (z −1 , δ k ) and goto Update; where a 2(n x +n y +n x n y )+ := g and b 2(n x +n y +n x n y )+ := g T z −1 ; 12: end for 13: return (γ,ẑ).	2019-04-22T13:07:51.584Z	2017-01-01T00:00:00.000	{ "year": 2017, "sha1": "212a271051df915e05984a918cd0a2665f6a078f", "oa_license": null, "oa_url": "https://www.jstage.jst.go.jp/article/sss/2017/0/2017_111/_pdf", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "8fef5f01d9b5ef34218cc4160112485206f16604", "s2fieldsofstudy": [ "Mathematics", "Engineering", "Computer Science" ], "extfieldsofstudy": [ "Computer Science" ] }
263796822	pes2o/s2orc	v3-fos-license	Quantum storage of 1650 modes of single photons at telecom wavelength To advance the full potential of quantum networks one should be able to distribute quantum resources over long distances at appreciable rates. As a consequence, all components in the networks need to have large multimode capacity to manipulate photonic quantum states. Towards this end, a multimode photonic quantum memory, especially one operating at telecom wavelength, remains a key challenge. Here we demonstrate a spectro-temporally multiplexed quantum memory at 1532 nm. Multimode quantum storage of telecom-band heralded single photons is realized by employing the atomic frequency comb protocol in a 10-m-long cryogenically cooled erbium doped silica fibre. The multiplexing encompasses five spectral channels - each 10 GHz wide - and in each of these up to 330 temporal modes, resulting in the simultaneous storage of 1650 modes of single photons. Our demonstrations open doors for high-rate quantum networks, which are essential for future quantum internet. On the quest to reach large multimode quantum storage capacity, the use of multiple spatial channels has yielded good results, with a channel number up to 665 realized in gaseous atomic ensembles [37] based on Duan-Lukin-Cirac-Zoller (DLCZ) protocol [41].Further improvements to the multimode capacity of such memory could be achieved via increasing the number of stored temporal modes in each of these channels.However, the simultaneous storage of multiple temporal modes remains challenging in atomic gas ensembles due to limitations of the applied storage protocols [41][42][43].Specifically, the number of temporal modes that can be stored is related to the optical depth (OD) of the storage media [44].Fortunately, the atomic frequency comb (AFC) quantum memory with rare-earth ion-doped (REID) materials does not feature any constraint in terms of OD for simultaneous temporal storage, and, thus, is promising for developing highly multimode photonic quantum memory using several photonic degrees of freedom [45].Multiple storage channels have been demonstrated with the AFC protocol in the spatial domain [34] and in the spectral domain [27][28][29], which is facilitated by the inhomogeneous broadening in REID materials.Storage of multiple temporal modes in one spectral channel has also been achieved, for instance the storage of 1250 temporal modes in Yb 3+ :Y 2 SiO 5 at 979 nm [46].Furthermore, the storage of multiple temporal modes in multiple channels has been reported, including storage of 12 spectro-spatial-temporal modes [39] and 130 spectro-temporal modes [40] in Pr 3+ :Y 2 SiO 5 at 606 nm.Despite these important advances, a quantum memory with large multimode capacity at telecom wavelength has yet to be demonstrated, as an essential step towards a future quantum network compatible with existing telecom infrastructure. In this paper, we present the multimode storage of 1650 modes of single photons at telecom wavelength with a 10-m-long cryogenically cooled erbium doped silica fibre (EDF). In our demonstration, five individual AFC spectral channels -each with 10 GHz bandwidth and separated by 5 GHz isolation -are prepared by using an optical frequency comb [47] combined with frequency chirping [28].Up to 330 temporal modes of heralded single photons generated from cascaded second-order nonlinear process [48][49][50] are stored in each spectral channel.Our achievements pave the way towards developing future quantum internet by utilizing multimode quantum memory compatible with the infrastructures of fibre-based communication. Results Multimode quantum storage scheme.Based on the inhomogeneous broadening in REID materials, an N × M multimode quantum memory can be realized by using N spectral channels and M temporal modes in each spectral channel (see Fig. 1(a)).Through spectral tailoring of the inhomogeneous broadened absorption line of REID materials, AFC channels with a number of N in the spectral domain are prepared.In each spectral channel, a train of single photons in M temporal modes are stored, with the maximum value of M determined by the time-bandwidth product of the AFC memory.By doing so, a total number of N × M modes of single photons are stored, simultaneously.To develop such a quantum memory with multimode capacity at telecom wavelength, a promising REID material is low doping concentration EDF, which has a telecom-band transition wavelength and THz-wide inhomogeneous broadening (see Figs. 1(b), (c)). Experimental setup.The experimental setup is composed of an EDF based AFC quantum memory with five spectral channels, a heralded single photon source at telecom wavelength, and a coincidence detection system, as illustrated in Fig. 2 We prepare an AFC quantum memory with five spectral channels in a 10-m-long EDF cooled to a temperature of 10 mK and exposed to a magnetic field of 2000 Gauss (see Fig. 2(a)(I) and Methods for details).More precisely, we propose and apply an approach -optical frequency comb combined with frequency chirping -to prepare five individual AFCs, each with a bandwidth of 10 GHz, at central wavelengths of 1532.11nm, 1532.00 nm, 1531.88 nm, 1531.76 nm and 1531.65 nm, labeled as channels of 1, 2, 3, 4, and 5 (see Supplementary Information Note 1 for more details on preparing five AFCs by using optical frequency comb combined with frequency chirping method).Figure 1(d) depicts a typical AFC (channel 2) in our experiment, and all the five AFCs are shown in Supplementary Information Note 1.Note that different from using multiple individual lasers in parallel to prepare spectrally multiplexed AFC quantum memory [28], our method not only prepares such an AFC quantum memory by using a single laser, but also guarantees each channel with large bandwidth.Moreover, the scheme applies spectral isolation among channels, thus making the crosstalk among different storage channels to be negligible.When five spectral distinct photons -spectra matching with the five AFCs -are absorbed by the five AFCs respectively, five collective atomic excitation states can be generated.The state in one of the channels, e.g., in channel 1, can be expressed as: where W 1 is the total atom number constituting the AFC in channel 1, the amplitude c j depends on both the resonance frequency and position of the j th atom, δ j is the detuning of the j th atom with respect to the photon carrier frequency, k is the wave number of the input photon, z j is the position of the j th atom, \|g j and \|e j represent ground and excited states of the j th atom, respectively.Following the creation of the collective atomic excitation, the different terms in the collective excitation state \|Ψ 1 , having different detunings, begin to accumulate different phases.Due to the periodic nature of AFC, i.e., δ j = m j ∆ (m j = 0, ±1, ±2, ... and ∆ is the teeth spacing of AFC), at the time t = 1/∆ all terms in the state acquire phases equal to an integer multiple of 2π, which are all equivalent to 0. This process of rephasing leads to the re-emission of input photon in its original quantum state. To characterize the five spectral channels of quantum memory, we generate heralded single photons at telecom wavelength in a fibre pig-tailed periodically poled lithium niobate (PPLN) module [48][49][50], as shown in Fig. 2(a)(II).The light from a continuous-wave (CW) laser (PPCL300, PURE Photonics) operating at 1540.60 nm is modulated to light pulses and subsequently sent into the PPLN module.The cascaded second-harmonic generation (SHG) and spontaneous parametric down conversion (SPDC) processes result in correlated photon-pairs centred at 1540.60 nm with a bandwidth of ∼60 nm.After efficient filtering, the central wavelengths of signal and idler photons are 1531.88nm and 1549.32 nm with a bandwidth of ∼100 GHz.The signal photons are sent into the five AFC spectral channels. It is worth noting that AFC itself is both a memory and a filter [51].In our experiment, the spectral modes that do not match the storage bandwidth will be either absorbed and re-emitted spontaneously (with a decay time of ∼10 ms) or pass through the memory, in which case they can be discriminated by their arrival time.Finally, we establish a coincidence detection system consisting of superconducting nanowire single photon detectors (SNSPDs, P-CS-6, PHOTEC Corp.) and a time-to-digital converter (TDC, quTAG, qutools), as shown in Fig. 2(a)(III) (see Methods for details of the SNSPDs).All SNSPD detection signals are delivered to the TDC to perform coincidence measurements.The system storage efficiency can be calculated by the ratio of the counts of recalled heralded photons to that of the input heralded photons.The second-order cross-correlation function (g s,i (0)) between signal and idler photons after/before storage is calculated as: where p si is the probability of 3-fold coincidence detections of trigger signal, idler photons, and signal photons, p s (p i ) is the probability of coincidence detections of trigger signal and signal (idler) photons, respectively.Note that the coincidence detection in our experiment with trigger signal enables us to select the recalled photons by their arrival time, thus reducing the noise photon from the residual spontaneous emission.According to the Cauchy-Schwarz inequality [52], a non-classical field satisfies g s,i (0) > 2.More details of measuring g s,i (0) through coincidence counts are given in Methods.Measurement.First, we investigate the performance of the five AFC spectral channels. To this end, we vary the storage time from 5 ns to 230 ns by programming the tooth spacing from 200 MHz to 4.35 MHz.For each case, we map heralded single photons onto each of the five spectral channels, and subsequently collect coincidence statistics of the recalled photons to gauge storage performance in terms of storage time, efficiency, and preservation of quantum properties manifested in the g (2) s,i (0) values.According to the measured results, the system storage efficiency of each channel with different storage times ranges from 0.1% to 1%.The g s,i (0) still remains above 25 for the maximum storage time of 230 ns (see Table I), demonstrating strong non-classical properties for all storage times (see Supplementary Information Note 3 and 4 for more analysis of the storage efficiency and g s,i (0) for five individual AFCs).Furthermore, these measurements demonstrate that our memory, in principle, is able to perform quantum storage of 2300 (10 GHz × 230 ns) temporal modes of telecom-band single photons in each channel, for five parallel channels amounting to a total of 11500 modes. Second, to examine the effect of crosstalk between different spectral channels, we measure the g s,i (0) between recalled signal photons and heralding idler photons corresponding to different spectral channels [36].If there is no crosstalk between different spectral channels, the measured g s,i (0) between uncorrelated channels should be around 1. On the other hand, with increasing crosstalk, the value of g s,i (0) would also increase (the relationship between crosstalk and g s,i (0) is shown in Supplementary Information Note 5).To eliminate the crosstalk between spectral channels, a sufficiently large separation between adjacent AFCs is required [27] -in our case we choose 5 GHz separation.We first assess the crosstalk resulting from the source and detection system without the quantum memory.For the crosstalk characterization, we label the signal photons corresponding to channels 1 to 5 as S 1 , S 2 , ..., S 5 , and similarly, the idler photons as I 1 , I 2 , ..., I 5 .The choice of which signal and idler channels to detect is determined by the tunable FBG filter settings.Measuring coincidences between S m and I n , where m, n ∈{1, 2, 3, 4, 5}, we calculate a 5 × 5 matrix of g s,i (0).The measured average g s,i (0) between recalled photons and idler photons is 25.48 ± 1.05 for five correlated spectral modes and 1.01 ± 0.10 for 20 pairs of uncorrelated spectral modes.The results confirm that the crosstalk between different spectral channels is negligible in our quantum memory. Third, to quantify the multimode capacity of our quantum memory, we create a train of heralded single photons -with pulse durations of and spaced by 300 ps -for simultaneous storage in all five spectral channels.With the storage time of 200 ns, 330 temporal modes are simultaneously stored in each spectral channel (see Fig. 3(c)).The total stored mode number is, thus, up to 1650.Furthermore, to assess the crosstalk between different temporal modes, we obtain 1650 × 1650 values of g s,i (0) through measuring cross-correlation function between different idler photons and recalled signal photons.Figure 3(d) presents s,i (0) corresponding to different temporal modes in channel 2 (the full 1650 × 1650 array of g s,i (0) is shown in Supplementary Information Note 7).The average g s,i (0) between correlated signal and idler photonic modes (1650 pairs) is 22.92 ± 0.07, verifying that the non-classical correlations are intact.For uncorrelated modes the g s,i (0) is 1.01 ± 0.01 on average with error bar from Monte Carlo simulation, confirming the negligible crosstalk between different temporal modes.In addition, we conclude the key metrics of our multimode quantum memory in Table II, which also includes the state-of-the-art of multimode quantum storage of non-classical light based on REID materials. Discussion An important feature of our demonstration is the large multimode capacity of quantum memory for storage of single photons at telecom wavelength.Yet there are clear avenues to further advance the multimode capacity.First, by making use of the entire THz-wide inhomogeneous broadening of the EDF, at least seventy individual 10-GHz-wide AFCs could be generated with a separation of 5 GHz.Generating more large-bandwidth channels requires a broader optical frequency comb, which can be achieved by a nonlinear broadened comb [53], Kerr soliton microcomb [54,55], and mode-locked fibre laser comb [56] etc.Second, by increasing the temporal mode number in each channel to the upper bound determined by the time-bandwidth product of the AFC, a total number of spectro-temporal modes of 10 GHz × 230 ns = 2300 can be realized in each channel.Combining with these improvements, the multimode capacity would exceed 70 × 10 GHz × 230 ns ≈ 160000.In addition, we note that the multimode capacity of our quantum memory could be further increased via adapting multiplexing in the spatial degree [30][31][32][33][34][35][36][37][38]57]. Several upgrades could be applied to develop quantum memories available for future quantum networks, including storage time and storage efficiency.The storage time of AFC quantum memory is determined by the teeth spacing (∆) of AFC, i.e., T storage =1/∆.In our demonstration, the frequency stability of laser system for AFC preparation limits the ∆ to several MHz, resulting in a storage time of hundreds of nanoseconds.A straightforward improvement is to employ a laser system with Hertz linewidth to prepare AFC.By doing so, the storage time would be further limited by the homogeneous linewidth of Er 3+ ions in fibre, i.e., inversely proportional to its optical coherence time.In our experiment, the optical coherence time reaches tens of microsecond with a temperature of 10 mK and a magnetic field of 0.2 T. The optical coherence property can be further improved, for instance by reducing the spin flip flops and superhyperfine broadening.The interaction of Er 3+ -Er 3+ spin flip flops can be reduced by decreasing the doping concentration of Er 3+ ions [58].And the superhyperfine broadening due to the interaction between Er 3+ ions and neighboring Al 3+ ions could be modified by removing the co-dopants of EDF [59].In addition, it is desirable to develop on-demand quantum memories.A promising option to enable this feature is spinwave storage [60,61], which allows optical coherence to be mapped onto a long-lived spin state, such as the hyperfine level.The Λ-like structure has been harnessed in erbium ions [62], which demonstrated erbium ions doped solids has potential for realizing on-demand quantum memory.With regards to the storage efficiency, the non-zero background absorption (d 0 ) of the prepared AFC is one of main limits.In our demonstration, d 0 is caused by the limited ratio of the lifetime of Zeeman sublevels (∼300 ms) to the waiting time following the AFC preparation (200 ms), in which the waiting time is to ensure that recalled photons are not masked by spontaneously emitted photons and guarantee the memory with high signal-tonoise ratio (SNR > 1000).Along with the waiting time, the population of Zeeman sublevels decays accordingly, which inevitably fills the AFC troughs [6,28].To further increase the efficiency of the memory, on one hand, the lifetime of Zeeman sublevels could be increased by decreasing the doping concentration of Er 3+ ions [6].On the other hand, the trade-off between the waiting time and the SNR of memory should be explored, i.e., appropriately sacrificing the SNR of memory by reducing the waiting time to improve the storage efficiency.Furthermore, utilizing an impedance-matched optical cavity could also improve the storage efficiency [63,64].Inspired by these methods, towards a high-performance EDF based AFC quantum memory, a favorable avenue is to create an impedance-matched in-fibre cavity in a piece of EDF with low doping concentration (more details see Supplementary Information Note 8). In summary, we have demonstrated a multimode quantum memory that is suitable for storage of spectro-temporal modes of single photons at telecom wavelength.A quantum memory with five 10-GHz-wide AFC channels in a 10-m-long cryogenically cooled EDF is prepared, and 330 temporal modes of heralded single photons are stored in each channel leading to the multimode capacity up to 1650.The key method introduced here is the combination of an optical frequency comb with frequency chirping to prepare large-bandwidth multi-channel AFCs, thus, enabling the large multimode capacity and high spectral isolation of our memory.With the improvements of storage time and efficiency, EDF based AFC quantum memory would pave the way for constructing future quantum internet compatible with current telecom infrastructure. Methods Erbium doped silica fibre.The experiment utilizes a 10-m-long, single-mode, commercial EDF with Er 3+ ions doping concentration of 200 ppm, co-doped with Al, Ge, and P. The fibre is spooled to a home-made copper cylinder with a diameter of 4 cm and fixed in the dilution refrigerator (LD400, Bluefors), in which the temperature can be cooled below 10 mK.The measured absorption at 1532 nm is 0.35 dB/m at T=10 mK.The EDF is fused with single-mode fibres for each end, and is exposed to 2000 Gauss magnetic field, the strength of which is optimized by observing the lifetime of Zeeman sublevels of 4 I 15/2 level.With a magnetic field of 0.2 T, the lifetime reaches 0.278 ± 0.035 s, which ensures the persistent time of AFC (see Supplementary Information Note 9 for details).On the other hand, the applied magnetic field can eliminate a part of magnetic two-level systems, thus increasing the optical coherence time of erbium ions [65].The loss of the whole fibre sample is about 2.5 dB, including bending, splicing, and transmission loss. It is a strong indication of quantum correlations for photons-pairs if g s,i (0) 2. According to Cauchy-Schwarz inequality, it also indicates that the auto-correlation function of heralded signal photons is 1, i.e., these signal photons are denoted as single photons [66]. For all the coincidence measurements in this paper, the width of coincidence window is 600 ps.In the section of inhomogeneous broadening absorption profile (70 GHz wide), we prepare five 10-GHz-wide AFCs with a separation of 5 GHz between the edges of adjacent AFCs.In principle, over seventy such AFCs can be prepared if we make use of the entire inhomogeneous broadening.(c) Simplified energy level scheme of Er 3+ ions in erbium doped silica fibre.AFCs are prepared through frequency-selective optical pumping that transfers atomic states from the ground state (\|g ) to the auxiliary state (\|aux ) via the excited state (\|e ), forming AFCs.(d) A typical trace of 10-GHz-wide AFC measured in our experiment (the central wavelength of the AFC is 1532.00nm, and comb spacing ∆ is 100 MHz, leading to T storage =10 ns).s,i (0) between correlated recalled signal and idler photon modes (1650 pairs) is 22.92 ± 0.07.For uncorrelated modes, the average g The bandwidth of each AFC is 10 GHz, and the spacing of adjacent AFCs is 5 GHz.In total, the available bandwidth for quantum memory is 50 GHz.Inset: Details for a 2-GHz-wide section of the AFC is shown, where the comb spacing ∆ is 100 MHz.OD: Optical Depth. Note 2: Heralded single photon source In this section, we present properties of the heralded single photon source at telecom wavelength corresponding to different AFC spectral channels.The experimental setup is shown in Fig.S. 3. The scheme for generating correlated photon-pairs is cascaded second-harmonic generation (SHG) and spontaneous parametric down conversion (SPDC) processes in a single periodically poled lithium niobate (PPLN) waveguide [49,50,[75][76][77] (the main parameters of the PPLN see Table S. I).The light from a narrow linewidth continuous-wave We measure the photons counting rate of correlated photon-pairs corresponding to different spectral channels under different power of pump light.quadratic polynomial curves.measured counting rate of idler photons is slightly higher than signal photons.This is mainly owning to the difference in transmission efficiency and bandwidth of the filters.The total transmission loss of idler photons is 3.9 dB, including the loss of DWDM, PC, FBG, etc., while the loss of signal photons is 5.5 dB.The bandwidth of the FBG with the central wavelength of 1549.32 nm (1531.88nm) is ∼0.06 nm (0.05 nm). s,i (0) for typical correlated photon-pairs corresponding to channel 2 of the quantum memory with different generation probabilities of idler photons (P idler ).The measured g s,i (0) fits well with an inversely proportional curve.Considering that the output coupling efficiency of the PPLN module is 71.65%, the total transmission efficiency is 40.74%, and the detection efficiency of SNSPD is 60%, the measured efficiency of idler photons is 17.51%.For a measured P idler of 0.62%, the intrinsic generation probability of idler photons in the waveguide is 3.54%, which results in a measured g s,i (0) for correlated photon-pairs corresponding to different channels. Channel a Signal photon (nm) b Idler photon (nm) c g s,i (0) Note 3: Performance of the five-channel quantum memories In this section, we first take channel 2 as an example to illustrate the storage performance.slight oscillation of the storage efficiency is visible.This may be due to a quantum beat generated by the interaction of the Er 3+ with other surrounding ions (such as the co-doped Al 3+ ) [12,59].For longer storage time, the decay is exponential with decay constant of 48 ns.We attribute this to the decoherence effects of Er 3+ and the imperfect preparation of AFCs [6]. s,i (0) after storage as a function of storage time for channel 2. Unlike the efficiency, the cross-correlation function g (2) s,i (0) remains almost constant at a level above 25 even at the maximum storage time of 230 ns.First and foremost, this shows that non-classical properties of the correlated photon-pairs are preserved in the memory and, in addition, that even at the largest storage time the signal-to-noise ratio (SNR) of the system is high enough (> 730) that the noise does not degrade the cross-correlation function.(see Note 4 for more details on the role of SNR in determining g Channel 5 0 ns correlated photon-pairs, where the SNR is defined as the ratio of coincidence of idler-recalled photons to idler-noise photons.The measured 3-fold coincidence detecting probability p si of trigger signal, signal and idler photons, and the 2-fold coincidence detecting probability p s (p i ) of trigger signal and signal (idler) photons are expressed as where η c is the internal generation probability of correlated photon-pairs in the waveguide, η s is the overall measured efficiency of recalled signal photons including transmission, storage, collection and detection efficiency, η i is the overall measured efficiency of idler photons, and η n is the detected probability of noise photons in each experimental trial, respectively.The p s •p i gives a good estimation for the accidental coincidences of uncorrelated background [78]. Combining Eq. ( 1) with Eq. ( 2) in main text, g s,i (0) can be calculated as During the experiment, η s is the only considered efficiency that will decrease with increasing storage time, and detection of idler and noise photons satisfies η c η i η n .The SNR of quantum memory system is defined as SNR= η c η i η s /(η c η i η n ) = η s /η n , where η c η i η n means the detected probability of coincidences between idler photons and noise photons.To analyze the contribution of SNR to g s,i (0), SNR is substituted into Eq.( 2), then the g s,i (0) can be re-expressed as If SNR is large enough, g s,i (0)=1/η c + 1 will be independent with η s , which means g s,i (0) will be unaffected by decreasing system storage efficiency once the overall measured efficiency (η s ) of recalled photons is still much larger than that of noise photons (η n ). Considering the real experimental parameters, the internal generation probability of correlated photon-pairs is 3.54%, and the measured coincidence probability of noise photons and idler photons (i.e., η c η i η n ) is 0.86 × 10 s,i (0) keeps unchanged even with longer storage time and decreasing efficiency.According to the measured efficiency, we estimate the g s,i (0) with different storage time of quantum memory. Note 5: Relationship of crosstalk and cross-correlation function In this section, we describe the relationship of the crosstalk between different channels and the corresponding cross-correlation function g As shown in Fig. S. 7(a), here we consider the situation of our experiment.η i is the overall measured efficiency of idler photons, η 1 is the transmission and collection efficiency of signal photons before quantum memory and η 2 is the measured efficiency of signal photons after the quantum memory, p 1 and p 2 are intrinsic generation efficiency of correlated photon-pairs.According to Eq. ( 2), the cross-correlation function corresponding to different channels can be expressed as It can be clearly seen that the cross-correlation function between uncorrelated channels (a).The experimental time sequences are also shown in Fig. 2(b) (see Methods). ( 2 ) s,i (0) is 27.90 ± 0.18 for 5 pairs of correlated spectral modes (m=n) and 1.07 ± 0.02 for the 20 pairs of uncorrelated spectral modes (m =n), where error bars are acquired through Monte Carlo simulation.The detailed values are shown in Supplementary Information Note 6.The results indicate that the crosstalk between different spectral channels is negligible.Next, we evaluate whether the multimode memory induces additional crosstalk.To that end, we send the signal photons into the five AFC spectral channels with the storage time of 200 ns.The recalled signal photons from different channels are indexed as R 1 , R 2 , ..., R 5 .Measuring coincidences between R m and I n , we obtain the recalled signal photons from five channels (see Fig. 3(a)).The transmitted photons output from the EDF indicate the optical delay of the fibre, and the time delay between the transmitted photons and the recalled photons shows the storage time of the AFC memory.We again calculate the 5 × 5 array of g (2) s,i (0) values as shown in Fig. 3(b).The measured average g (2) photons at 1549.32 nm with a bandwidth of 100 GHz are filtered out.The signal photons are sent into five-channel quantum memory to filter and to store.After the storage time of T storage , the signal photons are recalled from AFCs.Finally, the idler photons and recalled signal photons corresponding to different channels are selected by FBGs and detected by SNSPDs, subsequently performed coincidence analysis in the TDC.Superconducting nanowire single photon detectors.All detections of single photons are carried out by a set of SNSPDs system provided by PHOTEC Corp.The system consists of SNSPD devices, cryostat system, and electronic control system.The niobium nitride (NbN) SNSPDs, manufactured by Shanghai Institute of Microsystem and Information Technology (SIMIT), operate at ∼2.2 K in the cryostat system with a dark counting rate of ∼100 Hz and a time jitter of ∼100 ps.All detectors have a dead time of less than 50 ns and a detection efficiency of ∼60%.Calculation of cross-correlation function g(2)s,i (0).Considering the counting effective events in m experimental trials, we record the total 3-fold coincidence counts C si of the trigger signal, idler and signal photons, the coincidence counts C s (C i ) between signal (idler) photons and trigger signals.With the photon detected probability in each trial calculated as p si = C si /m, p s = C s /m, p i = C i /m, the second-order cross-correlation function g(2) FIG. 1 : FIG.1: Spectro-temporal multimode quantum storage of single photons at telecom wavelength.(a) General scheme for storage of N × M modes of single photons.A train of M temporal modes are stored into the AFCs with N spectral channels, resulting in storage of N × M modes of single photons.(b) The absorption profile of 4 I 15/2 to 4 I 13/2 transitions of Er 3+ ions in the EDF at 10 mK.The inhomogeneous broadening is up to 2 THz (only 1 THz are shown here).In the section of inhomogeneous broadening absorption profile (70 GHz wide), we prepare five 10-GHz-wide AFCs with a separation of 5 GHz between the edges of adjacent AFCs.In principle, over seventy such AFCs can be prepared if we make use of the entire inhomogeneous broadening.(c) Simplified energy level scheme of Er 3+ ions in erbium doped silica fibre.AFCs are prepared through frequency-selective optical pumping that transfers atomic states from the ground state (\|g ) to the auxiliary state (\|aux ) via the excited state (\|e ), forming AFCs.(d) A typical trace of 10-GHz-wide AFC measured in our experiment (the central wavelength of the AFC is 1532.00nm, and comb spacing ∆ is 100 MHz, leading to T storage =10 ns). FIG. 2 : FIG. 2: Experimental setup and time sequences.(a) Experimental setup.(I) Preparation of AFC memory with five spectral channels.The erbium doped silica fibre (EDF) is cooled to a temperature of 10 mK by a dilution refrigerator and exposed to a magnetic field of 2000 Gauss provided by a superconducting magnet.Light from a continuous-wave laser (CW-laser) is firstly modulated to an optical frequency comb (OFC) by an intensity modulator (IM) and a phase modulator (PM).Then another PM is utilized to generate chirped light on each comb of the OFC.The modulated pump light is sent to the EDF via an optical circulator (Cir).Optical switch (OS) and variable optical attenuator (VOA) are used to control the pump time and intensity.Beam splitter (BS) and optical power meter (OPM) are utilized to monitor the pump light power.Fibre Bragg grating (FBG) is used to select signal photons recalled from different AFCs.(II) Preparation of the heralded single photon source (HSPS).Light from a CW-laser is modulated to 300 ps pulses by an IM and sent to the periodically poled lithium niobate (PPLN) module for generating correlated photon-pairs.By utilizing two dense wavelength division multiplexers (DWDMs), signal photons at 1531.88 nm and idler photons at 1549.32 nm with a bandwidth of 100 GHz are filtered out.FBG is used to select the idler photons corresponding to different spectral channels.Subsequently idler photons are directly sent to the detection system.The signal photons are sent into the EDF for storage.Erbium doped fibre amplifier (EDFA) and VOA are used to adjust the pump power.Polarization controller (PC) and polarizing beam splitter (PBS) are utilized to control the polarization of pump light.(III) Detection system.The idler and recalled signal photons are detected by two superconducting nanowire single photon detectors (SNSPDs).Coincidence measurements are performed by a time-to-digital converter (TDC).PCs are used to control the polarizations of signal and idler photons.(b) Time sequences.The cycle time of the experiment is one second, including 300 ms for AFCs preparation, 200 ms for waiting spontaneously emitted photons from excited states, and 500 ms for storage.The AFC pump laser is modulated to OFC and frequency chirped light within the preparation time of 300 ms.The SHG & SPDC pump laser with the repetition of 1 µs is modulated into 330 light pulses per cycle with pulse durations of and spaced by 300 ps (more details see Methods). FIG. 3 : FIG. 3: Characterization for the multimode quantum memory.(a) Results for spectral-multiplexed storage in five AFC spectral channels with a storage time of 200 ns.(b) Crosstalk between different spectral modes of idler photons and recalled photons.The average measured g (2) s,i (0) between recalled photons and idler photons is 25.48 ± 1.05 for 5 correlated spectral modes and 1.01 ± 0.10 for 20 uncorrelated modes.(c) Results for storage of 1650 spectro-temporal modes of single photons at telecom wavelength.Heralded single photons with 330 temporal modes are generated and stored into five spectral channels with a storage time of 200 ns, totally, 1650 modes are stored into the five AFCs.(d) Crosstalk between different temporal modes of idler photons and signal photons recalled from channel 2 (mode numbers: 1, 80, 160, 240, 330).The average g (2) fiveFig.S. 2 : Fig.S. 2:Typical AFCs with five spectral channels.A typical example of AFCs with five spectral channels prepared by utilizing optical frequency comb and frequency chirping method.The bandwidth of each AFC is 10 GHz, and the spacing of adjacent AFCs is 5 GHz.In total, the available bandwidth for quantum memory is 50 GHz.Inset: Details for a 2-GHz-wide section of the AFC is shown, where the comb spacing ∆ is 100 MHz.OD: Optical Depth. Fig.S. 4 : Fig.S. 4: Measured results of correlated photon-pairs related to channel 2. (a) Photons counting rate of signal photons and idler photons versus pump power.Inset: Idler photon-counting rate (blue circle) versus pump power.The blue solid line is the quadratic polynomial fitting curve of the photon-counting rate with the quadratic and linear parts shown as the red and green curves, respectively.The quadratic and linear components are corresponding to the contributions of generated photon-pairs and noise photons.(b) Coincidences counting rate between signal photons and idler photons versus pump power.(c) Cross-correlation function (g (2) s,i (0)) of correlated photonpairs versus P idler .All error bars are calculated by standard deviations of counts which obey Poisson distribution. Fig.S. 4 Fig.S. 4(b) shows the coincidence counting of idler photons and signal photons with different pump powers.Then we measure the cross-correlation function (g (2) s,i (0)) of photon-pairs corresponding to different channels of quantum memory under different generation probabilities of idler photons.Fig.S. 4(c) shows g Fig.S. 5 ( Fig.S. 5(a) shows the system storage efficiency, defined as the ratio of the count rate of recalled heralded photons to that of the input heralded photons, for a typical channel of quantum memory (channel 2) versus storage time.At small storage time (< 100 ns), a Fig.S. 5(b) shows g ( 2 ) s,i (0) in our quantum memory).The cross-correlation function after storage versus storage time for all five-channel quantum memories are shown in TableS.III. Fig.S. 5 : Fig.S. 5: Performance of multi-channel quantum memory.(a) A typical example (channel 2) for system storage efficiency as a function of storage time.The comb tooth spacing are prepared ranging from 200 to 4.35 MHz, which determines the storage time from 5 to 230 ns.(b) A typical example (channel 2) for cross-correlation function after storage as a function of storage time.The measurement time for each point is 1000 s.All error bars are calculated by standard deviations of counts which obey Poisson distribution. −9 in a 600-ps time window.According to the measured storage efficiency with different storage time, we calculate the SNR with different storage time, as shown in Fig.S. 6(a).The SNR is 730 with storage time up to 230 ns.Besides, we calculate the g (2) s,i (0) with different SNR as shown in the inset of Fig.S. 6(a).It indicates that once the SNR is larger than 20, the g (2)s,i (0) is nearly unaffected.It explains why g(2) ( 2 )Fig.S. 6 : Fig.S. 6: Cross-correlation function with different SNR.(a) The SNR of quantum memory versus storage time.Inset: The estimated g (2) s,i (0) versus SNR of quantum memory.(b) The results for the calculation of g ( 2 ) s,i (0).For simplicity, we consider a twochannel quantum memory, shown in Fig.S. 7(a), where 1 and 2 represent the input channels of the quantum memory, 1 and 2 represent the corresponding outputs of the quantum memory, respectively.Here we simply consider the crosstalk as the leakage efficiency of photons into the mismatched channels.The crosstalk coefficients of different channels can be expressed as C 11 , C 12 , C 21 , C 22 with C ij (i, j = 1, 2) representing the efficiency of photons from channel i passing through channel j and satisfying C 11 + C 12 = 1, C 12 + C 22 = 1. ( 1 and 2 , 2 and 1 ) will be 1 -caused by accidental coincidences of uncorrelated backgroundonly if no crosstalk exists (C 12 =0, C 21 =0), otherwise it will be larger than 1.Here, according to the experimental parameters, we assume g 12 =1 (i.e., C 12 =0) and p 1 =p 2 =p=0.0354,then calculate the crosstalk C 21 between different channels with different g 21 .The calculated result is shown in Fig.S. 7(b).It indicates that the crosstalk between different channels increases with the increasing of cross-correlation function, thus cross-correlation function is a good indicator to characterize the crosstalk. Fig.S. 7 : Fig.S. 7: Relationship of crosstalk and cross-correlation function.(a) A typical two-channel quantum memory with storage of two pairs of correlated photon-pairs.(b) Crosstalk between two different channels versus cross-correlation function of different channels. Fig.S. 10 :Note 9 :Fig.S. 11 : Fig.S. 10: Impedance-matched in-fibre cavity quantum storage.(a) Conception of impedancematched cavity quantum storage.(b) Scheme of EDF based impedance-matched quantum memory.Two DBR mirrors are fabricated at both ends of the EDF.(c) Our recent progress on the fabrication of EDB mirror in fibre. Fig.S. 12 : Fig.S. 12: Counts of spontaneously emitted photons in 1-ns-wide time window versus waiting time.All error bars are calculated by standard deviations of counts which obey Poisson distribution. TABLE II : Comparison on the time-bandwidth product and multimode capacity of quantum memories for non-classical light based on different REID materials. a λ is the wavelength of the stored single photons.b T storage is the storage time of AFC quantum memory.c BW is the available storage bandwidth of quantum memory.d TBP is the time-bandwidth product of quantum memory.e Number of stored modes (i.e., multimode capacity). TABLE S . I: Main parameters of PPLN module. laser operating at 1540.60 nm is modulated into light pulses with a repetition rate of 1 TABLE S . II: Values of g Correlated photon-pairs corresponding to different channels of quantum memory.b Central wavelengths of signal photons corresponding to different channels.c Central wavelengths of idler photons corresponding to different channels. a TABLE S . III: Performance of each channel of quantum memory.	2022-09-05T06:44:00.660Z	2022-09-02T00:00:00.000	{ "year": 2022, "sha1": "77dbcd8efd0767cdbe58b728dd28e7d28a72e7d1", "oa_license": "CCBY", "oa_url": "https://www.nature.com/articles/s41534-024-00812-1.pdf", "oa_status": "GOLD", "pdf_src": "ArXiv", "pdf_hash": "77dbcd8efd0767cdbe58b728dd28e7d28a72e7d1", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics" ] }
205962841	pes2o/s2orc	v3-fos-license	Identification of autoantibodies to ECH1 and HNRNPA2B1 as potential biomarkers in the early detection of lung cancer ABSTRACT Identification of biomarkers for early detection of lung cancer (LC) is important, in turn leading to more effective treatment and reduction of mortality. Serological proteome analysis (SERPA) was used to identify proteins around 34 kD as ECH1 and HNRNPA2B1, which had been recognized by serum autoantibody from 25 LC patients. In the validation study, including 90 sera from LC patients and 89 sera from normal individuals, autoantibody to ECH1 achieved an area under the curve (AUC) of 0.799 with sensitivity of 62.2% and specificity of 95.5% in discriminating LC from normal individuals, and showed negative correlation with tumor size (rs = −0.256, p = 0.023). Autoantibody to HNRNPA2B1 performed an AUC of 0.874 with sensitivity of 72.2% and specificity of 95.5%, and showed negative correlation with lymph node metastasis (rs = −0.279, p = 0.012). By using longitudinal preclinical samples, autoantibody to ECH1 showed an AUC of 0.763 with sensitivity of 60.0% and specificity of 89.3% in distinguishing early stage LC from matched normal controls, and elevated autoantibody levels could be detected greater than 2 y before LC diagnosis. ECH1 and HNRNPA2B1 are autoantigens that elicit autoimmune responses in LC and their autoantibody can be the potential biomarkers for the early detection of LC. Introduction Lung cancer (LC) is the leading cause of cancer death worldwide. In 2015, LC is estimated to account for more than 220,000 new cancer cases and 158,000 cancer deaths in the United States. 1 Presently, early detection of LC is limited to low-dose spiral computed tomography (LDCT) to screen LC in at-risk individuals. [2][3][4] While LDCT offer mortality benefit in high-risk individuals, 5 its application as a screening procedure have been restricted by the plentiful costs 6 and high false positives (50%), requiring individuals to have unnecessary follow-up examinations and unnecessary surgery therapy. 7 Hence, the search for improving biomarkers that can complement with imagining for LC screening remains an important goal. Serum biomarkers would enhance diagnostic capabilities to complement the currently available diagnostic tests, since they are noninvasive and reproducible screening tools. Many studies have demonstrated that cancer sera contain autoantibodies that react with a unique group of autologous cellular antigens called tumor-associated antigens (TAAs). [8][9][10] The immune response appear months to years before the clinical diagnosis of a tumor, which is triggered by the release of TAAs from tumors in cancer patients, suggesting serum autoantibody detection is highly suitable for early cancer diagnosis. 11 Early detection is essential for the optimal management of cancer. Thus, extensive studies are being conducted to identify and validate new biomarkers that would add to current markers and increase the sensitivity and specificity of cancer detection. Some studies have been developing autoantibody assays that could complement CT scanning in LC diagnosis and management. [12][13][14][15][16][17][18] The only one commercial test currently is a blood test that measures autoantibodies to LC-associated antigens called EarlyCDT-Lung. It was developed to assist physicians in the early detection of LC in a high-risk population with pulmonary nodules detected by CT. 16 Due to the low sensitivity of this test, additional antigens need to be explored, and alternative ways to improve the assay performance must also be pursued. In the present study, we identified ECH1 and HNRNPA2B1 as new autoantigens for LC using immunoproteomic approach (serological proteome analysis, SERPA), and found their autoantibodies could improve the performance characteristics for distinguish those with early-stage LC patients without clinical symptoms from the high-risk population, who are heavy smokers. Serum sample collection Five groups in two independent sample sets (discovery and validation) were used for this study. All of the serum samples were collected from the New York University (NYU) Lung Cancer Biomarker Center, a member of the National Cancer Institute-sponsored Early Detection Research Network (NCI-EDRN). The discovery set (set 1) included two groups (group 1 and group 2). Group 1 consisted of 28 LC patients with 219 serial serum samples, including samples from prior, at-diagnosis and post-treatment, in which, 94 serum samples from 25 patients included at least 2 to 11 serial samples, which have been collected on average 32 mo (range, 0.5-120 mo) before the diagnosis of LC, and 125 serum samples from 19 patients included at least 2 to 13 serial samples collected on average 29.9 mo (range, 3-110 mo) post-treatment of LC. All of the patients are heavy smokers with average 54.4 pack years of smoking. When every patient in group 1 was diagnosed as LC, two normal individuals were collected by matching for age, gender and smoking pack-year in the same hospital. Therefore, total 56 normal individuals were included in group 2. A separate set of samples (set 2: validation set), including three groups (group 3, group 4 and group 5) were used for validation of the autoantibody. Group 3 (n D 90) consisted of LC serum samples collected within eight weeks of biopsy-proven LC diagnosis. LC cases were staged as pathological stage I-III by the thoracic surgeons. Group 4 was the controls to group 3, in which there were 89 normal serum samples from healthy people who are never smokers or light smokers (<20 pack year). Group 5 consisted of 90 COPD (Chronic obstructive pulmonary disease) patients who were high-risk smokers and exsmokers and participated in CT scan LC screening. The detailed characterizes of five groups are shown in Table 1. The utilization of human samples was approved by the Institutional Review Board (IRB) at the Scripps Research Institute and at NYU. All clinical information were collected under Health Insurance Portability and Accountability Act (HIPAA) compliance from study participants with written informed consent under clinical research protocols approved by appropriate IRBs. One-and two-dimensional Western blotting and proteomic analysis To screen the autoantibody-positive sera, H1299 cells were lysed directly in Laemmli's sample buffer (Bio-Rad, Hercules, CA) and loaded onto 10% SDS-PAGE gel followed by running the gel at 80 V for 2 h, which is then transferred onto nitrocellulose membrane (Osmonics Inc., MA) for Western blotting. The membrane was then cut into 0.5-cm wide stripes. After blocking with 5% nonfat milk prepared in Tris-buffered saline (TBS), containing 0.05% Tween-20 (TBST), for 1 h at RT, the nitrocellulose membrane strips were incubated with sera at a dilution of 1:200. Horseradish peroxidase-conjugated goat anti-human IgG (Caltag Laboratories, San Francisco, CA) was used as secondary antibody with a dilution of 1:10,000 for 1 h at RT. The positive bands were detected with Enhanced Chemiluminescence (ECL) kit (Amersham, Arlington Heights, IL). For 2-DE Western blotting, the proteins on 2-DE gel are directly transferred onto nitrocellulose membrane and incubated with two pools of five sera from patients with LC in group 1 and from normal individuals in group 2 at a dilution of 1:500. All pools of sera were assayed in duplicated. After identifying the interesting protein spots, gel spots in 2-DE gel were excised and digested to perform liquid chromatography tandem mass spectrometry (LC-MS/MS) analysis. MS/MS spectra derived from peptides were submitted for database search using TurboSequest (available in Bioworks version 3.3.1) against the human IPI database (v3.48), in both correct and reverse orientations to enable false-discovery rate (FDR) calculation. The following filters were applied in Bioworks: DCn 0.85; consensus score 10.0; protein probability 1 £ 10 ¡3 and X corr 1.5, 2.0 and 2.5, for singly, doubly and triply charged peptides, respectively. The commercial recombinant proteins of ECH1, GAPDH and HNRNPA2B1 were electrophoresed by SDS-PAGE and transferred onto a nitrocellulose membrane that was then cut into strips. The strips were incubated with sera diluted 1:200 and subsequently with HRP-conjugated goat anti-human IgG diluted 1:10,000. Statistical analysis Data regarding the different seroreactivity of the autoantibodies were summarized by mean of OD value. Due to the sera antibodies against three TAAs were not normally distributed (Shapiro Wilk's test), nonparametric Mann-Whitney U-tests were used to compare differences of antibody levels between two groups, and nonparametric Kruscal-Wallis test were used to compare differences of antibody levels among multiple groups. x 2 tests were used to compare the differences of frequency between two groups and among multiple groups. Spearman's test was used to evaluate the correlation between autoantibody level and tumor size or lymph nodes number. The receiver operating characteristic (ROC) analysis of single-variable was conducted for each autoantibody for the distinguishing of LC from controls, leading to estimates of area under the curve (AUC) with 95% confidence interval (CI). The optimal cutoff thresholds for designating positive reaction were determined at the point on the ROC curve at which Youden's index (sensitivity C specificity ¡1) was maximal. Differences were considered statistically significant when p < 0.05. Statistical analyses were performed using SPSS software (version 18.0). Results A novel autoantibody against 34 kD autoantigen was found in serial serum samples from patients with early stage of lung cancer (discovery set 1) Ninety-four serial sera collected before the diagnosis from 25 patients with early stage of LC (group 1) and 56 sera from normal human individuals (group 2) matched to group 1 were examined for autoantibodies using Western blotting with LC cell line H1299. Interestingly, there were two patients who appeared to have an increase in autoantibody for antigen/antigens with molecular weight 34 kD before the diagnosis of LC ( Fig. 1A and B). Of 25 patients who had sequential serum samples in group 1, 10/25 patients (40.0%) were identified by Western blotting analysis to contain antibodies against 34 kD cellular protein, while no reactivity with the 34 kD protein was detected in 56 matched normal human sera. Western blotting analysis of five representative sera from LC (group 1) and normal individuals (group 2) are shown in Fig. 1C. 34kD proteins were identified to be ECH1, GAPDH and HNRNPA2B1 by SERPA approach Cytosolic, membrane and nuclear fractions of H1299 cell line were obtained using differential extraction of proteins. Proteins were separated by 2-DE and TAAs were screened with two pools of sera, which consisted of patients with early-stage LC in group 1 (Fig. 1C: lane 1-5) and matched healthy individuals in group 2 (Fig. 1C: lane 6-10). A total of six spots with 34 kD molecular weight exhibited reactivity with early-stage LC sera pool but not with controls (Fig. 2), in which one spot was from cytosolic ( Fig. 2A and B), four from membrane ( Fig. 2D and E) and one from nucleus ( Fig. 2G and H). These spots were excised from the gel, trypsin-digested and subsequently analyzed by LC-MS/MS. Ultimately, GAPDH (glyceraldehyde-3-phosphate dehydrogenase), ECH1 (Enoyl Coenzyme A hydratase 1) and HNRNPA2B1 (heterogeneous nuclear ribonucleoproteins A2/B1 isoform B1) was successfully identified by MS in the cytosolic, membrane and nuclear fractions, respectively (Table 2). Potential of autoantibody to ECH1 and HNRNPA2B1 as biomarkers in lung cancer diagnosis: Evaluation in an independent validation set 2 Table 3 and Fig. 3A and B show the individual and average OD values of autoantibodies against GAPDH, ECH1 and HNRNPA2B1 in the three groups. Autoantibodies to ECH1 (Fig. 3A) and HNRNPA2B1 (Fig. 3C) were revealed significant higher OD value in LC (group 3) than that in normal individuals (group 4) and COPD patients (group 5) (p < 0.01), while not for GAPDH (Fig. 3B). To evaluate the value of anti-ECH1 and anti-HNRNPA2B1 autoantibodies in LC diagnosis, we performed ROC analysis further. When the discrimination between the LC patients and normal individuals was performed, the overall accuracy of autoantibodies against ECH1 and HNRNPA2B1 was 78.8% and 83.8% (Table 4), with AUC of 0.799 (Fig. 3D) and 0.874 (Fig. 3G), respectively. The similar results were observed when LC was compared with COPD patients ( Fig. 3E and H, Table 4), as well as normal controls combined with COPD patients (Fig. 3F and I, Table 4). Commercial recombinant protein ECH1 and HNRNPA2B1 were further used for Western blotting analysis to verify the sera with positive reactivity in the discovery study ( Fig. 1D-F). Additionally, we evaluated the correlation of these two autoantibodies with histological and clinical parameters in LC patients (group 3) (Table 5 and Fig. 4A-H). The frequency of anti-ECH1 autoantibody in LC patients is statistically higher than controls who are non-smokers or light smokers (p < 0.001, Table 5), but the frequency does not have relationship with smoking pack years (p D 0.811, Table 5). In addition, the patients with tumor size <3 cm were observed to have significantly higher serum level of autoantibodies against ECH1 (OD median: 0.206) than those with tumor size 3 cm (OD median: 0.163) (Fig. 4A), and there is a negative correlation between anti-ECH1 antibody level and tumor size (r s D ¡0.256, P D 0.023) (Fig. 4E). For the anti-HNRNPA2B1, the patients with stage I have the highest autoantibody level and frequency among all of the stages (p D 0.012 and p D 0.015, Table 5). The patients who have not developed lymph node metastasis showed higher autoantibody level to HNRNPA2B1 than those who have lymph node (OD value: 0.261 vs. 0.213, Fig. 4D), and the negative correlation between autoantibody level and lymph node number was observed (r s D ¡0.279, p D 0.012, Fig. 4H). Autoantibodies against ECH1 can be detected in the prediagnostic sera of lung cancer patients but not HNRNPA2B1: Evidence from study set 1 We next questioned whether these two serum autoantibodies can be detected in sera drawn before time of diagnosis, and whether anti-ECH1 and anti-HNRNPA2B1 antibodies might develop over time. Ninety-four pre-diagnostic serum samples drawn at a various range of time (ranging 0.5-120 mo before diagnosis) from 25 patients in group 1 were used to validate these 2 biomarkers discovered. When there was more than one sample in a calendar year in a LC patient, we plotted the average OD for that year. We selected the OD values from the samples with collection date closest to diagnosis for each autoantibody in 25 LC individuals and compared it with the 56 matched-controls. This analysis showed an AUC of 0.763 (95% CI: 0.641-0.884, p < 0.001) with sensitivity of 60.0% and specificity of 89.3% for anti-ECH1 (Fig. 5E), however, AUC for anti-HNRNPA2B1 was only 0.400 with p value of 0.154 (Fig. 5F). For both of these two autoantibodies, no significant fluctuation of serum autoantibody occurred during the years before LC diagnosis, as well as post treatment (all of p > 0.05) (Fig. 5A-D), however, the difference was observed among the sera groups drown before diagnosis and 56 sera from matched normal controls for anti-ECH1 (p D 0.006) (Fig. 5A). In the further pairwise analysis, it was found that sera from less than 2 y before diagnosis showed an increased level for anti-ECH1 compared with normal controls (all of p < 0.05) (Fig. 5A). Discussion Serum autoantibodies detection has been actively explored as a means to provide novel biomarkers to aid in the early clinical diagnosis of cancer, since autoantibodies are typically produced and secreted in the serum before symptoms manifest and highly stable in the serum samples. 2,19 Autoantibody signatures might be useful in cancers in which there are high-risk populations and where existing detection methods lack sensitivity and specificity, particularly in LC. 2 To identify specific LC biomarker, an immunoproteomic approach combined with MS was used to identify interesting proteins and found that three 2DE-Western blotting-positive spots corresponded to ECH1, HNRNPA2B1 and GAPDH. ECH1 and HNRNPA2B1 were verified in the subsequent validation study. ECH1 is a member of the hydratase/isomerase superfamily and functions in the auxiliary step of the fatty acid b-oxidation pathway. A variety of studies demonstrated that ECH1 may be associated with tumor progression: abnormal expression has association with hepatocellular carcinoma secondary to hepatitis C virus infection, 20 and the pathogenesis of gastric cancer. 21 Additionally, the downregulation of ECH1 has been shown to be associated with the DNA damage-induced apoptosis resistance of B cell chronic lymphoid leukemia. 22 A higher expression level of ECH1 was confirmed in tissue from patients with gastric carcinoma with lymph node metastases indicating that ECH1 is a critical factor in the development of lymphatic metastasis in gastric cancer. 23 However, no study has yet addressed the relevance anti-ECH1 autoantibody to the development of cancers. HNRNPA2B1, one of the most abundant and important nuclear RNA-binding proteins involved in packaging nascent mRNA, alternative splicing, 24,25 cytoplasmic RNA trafficking, 26 and translation. 27 Numerous of researches indicate that HNRNPA2B1was involved in the tumorigenesis of pancreatic cancer, 28,29 prostate cancer, 30 hepatocellular carcinoma, 31 and LC. 32 Higher expression level of HNRNPA2B1 in tissues from patients with gastric adenocarcinoma, 33 non-small cell LC 34 and hepatocellular carcinoma 35 has been reported. The autoimmune responses to HNRNPA2B1 have been described previously in patients with hepatitis, 36 but not reported for cancer. This study indicated that both autoantibodies could distinguish LC from normal individuals with the accuracy of 78.8% for anti-ECH1 and 83.8% for anti-HRNPNA2B1, as well as COPD patients with accuracy of 80.0% and 87.2%, respectively. The earlyCDT-lung assay test, a currently available commercial kit, included a panel of six autoantibodies against p53, NY-ESO-1, CAGE, GBU4-5, Annexin 1 and SOX2 was reported to have sensitivity of 38.0% and specificity of 88.0% in LC patients, was also shown to have no significant difference based on LC stages. 16 In the present study, the frequency of autoantibody to HNRNPA2B1 is 90% in stage I patients, which is significantly higher than that in stage II (56.7%) and stage III (70.0%). The results indicate that anti-HNRNPA2B1 might be potentially taken as a biomarker to detect the early stage of LC. Although the mechanisms leading to autoantibody production in cancer patients are not completely understood, emerging evidence indicates that most TAAs are cellular proteins whose aberrant regulation of function could be linked to malignancy. 8 Further analysis of autoantibodies in the clinical characterizations of LC disclosed that anti-ECH1 level showed negative correlation with tumor size, anti-HRNPNA2B1was inversely correlated with lymph node metastasis. It may be suggested that the presences of these two autoantibodies are associated with aggressiveness of LC. Advanced cancer may have different molecular characteristics compared with preclinical disease; therefore, it is necessary to determine whether biomarkers discovered from patients with established disease can also be applied to samples from earlier time points before diagnosis, or even at a preclinical stage. In our recent study, we demonstrated that autoantibodies discoverable in sera from established LC may also be detectable in pre-diagnostic sera. 37 In this study, we found anti-ECH1 level in the serum samples drawn as early as 2 y before diagnosis increase compared with the matched normal individuals (group 2), along with high sensitivity (60.0%) and specificity (89.3%) with an observed AUC of 0.763 (95%CI: 0.641-0.884), while anti-HNRNPA2B1 has the similar antibody level among serial serum samples from LC patients and normal controls. This suggested that autoantibody to ECH1 may also be used as potential biomarkers to help identify those who may be at risk for developing LC later on in their life. However, more works are needed to further evaluate how to apply this clinically as there is a frequent temporal change in positive reactivity through the pre-diagnostic phase. 37 In recent decades, the potential utility of autoantibodies as cancer biomarkers to monitor therapeutic outcomes, or as indicators of cancer prognosis post-therapy, has been explored. [38][39][40] It was reported the increased anti-NPM1 autoantibody level in the sera from prostate cancer patients after surgery treatment. 41 Using 125 serial serum samples drawn after treatment from 19 LC patients in this study, autoantibody level to both ECH1 and HNRNPA2B1 have not shown variation in the follow-up samples post treatment. It suggested that there was no association between these two antibodies and clinical outcome, and it might not show value as a potential prognostic biomarker after treatment. Conclusions Early detection of LC will allow the patients to receive treatment while still in early stage, even in pre-clinical stage, where it is still possible to prevent the progression to LC. The combinatorial utilization of serum biomarkers with low dose CT examination is one of the promising approaches, which is noninvasive and is of high sensitivity and specificity. This study suggest that ECH1 and HNRNPA2B1 can elicit humoral immune response in LC, and their autoantibodies can be taken as potential biomarkers in the early detection of LC, and even the predictor in the pre-clinical stage of this disease. The subsequent works are needed to furtherly evaluate how to apply these clinically to aid physician to make decision in the early detection of LC. Disclosure of potential conflicts of interest No potential conflicts of interest were disclosed.	2018-04-03T04:41:40.683Z	2017-03-31T00:00:00.000	{ "year": 2017, "sha1": "a7cff49da82bf95b8824336281d8531620ade459", "oa_license": null, "oa_url": "https://www.tandfonline.com/doi/pdf/10.1080/2162402X.2017.1310359?needAccess=true", "oa_status": "GOLD", "pdf_src": "TaylorAndFrancis", "pdf_hash": "0213edbe8d9183051d0244d814e03c1f50046f10", "s2fieldsofstudy": [ "Biology", "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
270839215	pes2o/s2orc	v3-fos-license	Mapping deterioration in electrospun zein nonwoven nanostructures encapsulating corn oil Electrospun nonwovens of biopolymers are gaining popularity in filtration, coatings, encapsulation, and packaging materials. However, their applications are hindered by limited stability, particularly when loaded with lipids. This research aimed to apply a multiscale approach to gain insights into deteriorative processes, e.g., oxidation, limiting the shelf life of these complex materials, using corn oil-loaded electrospun zein nonwovens as a model system. Oil-doped zein electrospun nonwovens were stored in the dark at 23 °C and 33% relative humidity for 28 days and tested at selected intervals to monitor their morphology and mechanical properties. Lipid oxidation was assessed using the thiobarbituric acid reactive species (TBARS) assay. The photophysical properties of intrinsic, i.e., tyrosine (Tyr), and extrinsic, i.e., boron-dipyrromethene undecanoic acid 581/591 (BODIPY C11), lumiphores were also monitored to evaluate changes in local molecular rigidity, and oxidation, respectively. The protein secondary structure was determined with Fourier transform infrared spectroscopy (FTIR). Scanning electron microscopy (SEM) analysis of the oil-loaded electrospun nonwovens revealed that the diameter of the ribbon-like fiber significantly decreased during storage from 701 ± 23 nm to 620 ± 44 nm. Breakage of the electrospun fibers was observed and correlated with increased brittleness and molecular rigidity of the nonwoven material, reflected by an increase in Tyr emission intensity and phosphorescence lifetime. Changes in tensile strength, brittleness and matrix rigidity also correlated with a zein secondary structure transition from unordered to ordered β-sheets. Raman and luminescence micrographs showed oil migration during storage, thereby increasing lipid oxidation. The correlation between local rigidity and lipid distribution/oxidation suggests that reorganizing protein structures increased material brittleness and displaced encapsulated oils within the electrospun fiber. Understanding deteriorative mechanisms aids in developing innovative strategies to improve the stability of these novel food-grade materials. Introduction Advances in fabrication technology allow the production of model foods with designed structural elements to, for example, tailor sensory properties and extend shelf-life.Among them, electrospinning of foodgrade materials (e.g., edible proteins and carbohydrates) is gaining traction, leading to the manufacture of various structures from fibers, ribbons, and beads to their mixture and diameters between 100 and 1500 nm (Aceituno-Medina et al., 2013;Tang et al., 2019).During electrospinning, a high electrostatic force draws the polymer solutions from a charged wire or spinneret, stretching the solution into ultrafine structures (Li et al., 2009).Electrospun nonwovens contain fibers with submicron diameters that entangle during fabrication to produce a nonwoven sheet.These materials are gaining popularity in filtration, coatings, encapsulation, and packaging applications (Lim et al., 2019). Nonwovens of food-grade biopolymers allow the inclusion of bioactive compounds within the fibers to ensure compatibility and improve their functional properties (e.g., nutraceutical, antimicrobial, or antioxidant activities).For example, gelatin nanofibers with encapsulated essential oils have been produced as edible packaging with antimicrobial properties, while a whey protein isolate-pullulan blend was explored for the encapsulation of resveratrol in product fortification (Seethu et al., 2020;Tang et al., 2019).Despite promising applications, the stability of nonwovens is limited due to deteriorative phenomena affecting both the biopolymer carrier and the encapsulated lipids (Moomand and Lim, 2014;Xu et al., 2008).For example, García Moreno et al. (2016) reported the limited oxidative stability of electrospun fibers containing fish oil and its effects on their physical stability.Similarly, Drosou et al. (2022) observed that the structural integrity of the electrospun encapsulant diminished during storage. Additives, such as plasticizers and antioxidants, are useful for stabilizing nonwoven structures.However, growing consumer concerns regarding food sustainability and clean food labels require alternative methods besides the incorporation of additives to increase stability (Cavaliere and Ventura, 2018;Grant et al., 2021).Thus, understanding the limiting factors influencing the stability of food-grade materials and their products must remain a priority, as the list of acceptable additives that extend shelf life is continually decreasing.Identifying the underlying mechanisms of deterioration is crucial in informing handling best practices, storage conditions, packaging material selection, and other mitigation strategies. Since composite nonwovens contain multiple phases (e.g., encapsulant and encapsulate), concurrent deteriorative phenomena may occur.Thus, a multiscale approach that considers phenomena at the macro-, micro-, nano-and molecular levels will facilitate an improved understanding of deterioration in these matrices.Moreover, monitoring deterioration at multiple length scales provides distinct perspectives on how physical and chemical changes evolve in a matrix on their own and concerning each other.For example, previous work using a multiscale approach to understand ageing in biodegradable gelatin films concluded that protein stiffening is responsible for adverse changes in tensile stress and brittleness during storage (Colaruotolo, 2019;Colaruotolo et al., 2021).For nonwoven composites, the deteriorative mechanisms will differ based on specific constituting biopolymers and encapsulates and must be assessed without altering existing structures.To this end, optical spectroscopic tools are versatile due to their non-destructive and minimally invasive testing procedures, allowing sample analysis in situ with minimal sample preparation.These techniques provide complementary assessments of the nonwoven state and its existing structure. Zein was chosen as the biopolymer in this study due to its popularity as an encapsulant of lipid-soluble bioactive compounds, while corn oil was the encapsulate.Corn oil was selected due to its common use as a hydrophobic bioactive carrier (Liu et al., 2019;Moomand and Lim, 2015;Wang et al., 2019); however, no bioactives were added to facilitate data interpretation.In zein fibers with encapsulated oil, solvent (alcohol) evaporation from the polymer-oil solution during spinning decreases the solvent quality near the jet surface, causing the oil to migrate to the center of the structure during formation (Moomand and Lim, 2014), effectively encapsulating it.This study aimed to quantify and better understand the relationships between deteriorative processes in zein nonwovens with encapsulated corn oil and inform best practices to extend shelf life and mitigate deleterious effects, such as oxidation, matrix stiffness or increased brittleness. Zein nonwoven fabrication and storage Spin dope solutions, i.e., homogeneous and stable solutions to be spun, were prepared as detailed in Moomand and Lim (2015) with modifications.Zein at 20% (w/w) was dissolved in 80% (w/w) 7:3 2-propanol:water solution by stirring at 1200 rpm for 15 min.Corn oil was added during mixing in the dark at 1200 rpm at room temperature (21 ± 2 • C) for 15 min to achieve an oil:protein ratio of 3:10 (w/w).Electrospinning was conducted using an N.S. lab Nanospider (Elmarco s. r.o, Liberec, Czech Republic) at 21 ± 2 • C and 35 ± 5% relative humidity (RH).The polymer solution was electrospun through a wire and carriage spinneret (0.7 mm diameter orifice) at 150 mm/s.This setup, i. e., the wire and carriage spinneret, was selected since it allows for higher throughputs than needle configurations, better mimicking industrial procedures and conditions.Additionally, this setup avoids clogging and reduces the probability of developing complex fiber structures that would unnecessarily complicate analysis and monitoring deterioration effects (Prahasti et al., 2020;Menaka and Srinivasan, 2023).The spinning voltage was set at 50 kV to attain the higher levels required by a wire carriage configuration.The nonwoven samples were stored in the dark to reduce photodegradation, within sealed containers holding a saturated solution of magnesium chloride that allowed maintaining a 33% RH within the chamber.The samples were placed on a mesh affixed 3 m above the bottom of the container that separated them from the solution and stored at 23 ± 2 • C. Mechanical characterization of zein nonwovens The mechanical properties of the electrospun nonwovens during storage were assessed using a Modular Compact Rheometer (MRC 302, Anton Paar, Graz, Austria).Strips (20 × 50 mm) of electrospun material were affixed between the testing platform and a 20 mm plate (PP20/S), separated by a 10 mm gap.The samples deformed in tensile mode at a constant velocity of 12 mm/min until mechanical failure.Triplicate samples were tested after 0, 7, 14, 21, and 28 days of storage with a sample size of n = 5.Acquisition time (s), extension (mm), and force (mN) were recorded and used to calculate the engineering stress (σ E ): where F is the normal force in mN, A 0 corresponds to the original crosssectional area of the nonwoven sheet in m 2 , and the Hencky strain (ε H ) as: where L is the length of material in mm at time t (s), and L 0 is its initial length (mm).Young's modulus (Y) assessed the material extensibility (Steffe, 1996) and was calculated as: where the subscripts i and f correspond to the initial and final measures, respectively. The jaggedness of the force vs. deformation curve described the brittleness of the materials (Calzada and Peleg, 1978).The percent of direction reversals in this relationship, which corresponds to the percentage of the data points identified as local extrema, was calculated by counting the events where the force decreased from its previous value, after which it increased again in the subsequent point, as described in Corradini and Peleg (2006). Morphology of zein nonwovens Scanning electron microscopy (Model S-570) (Hitachi High Technologies Corp., Tokyo, Japan) with an accelerating voltage of 20 kV and a 30,000× magnification evaluated the electrospun nonwoven morphology after the samples were prepared with a 20 nm gold deposit layer using a vacuum-assisted sputter coater (Model K550, Emitech, Ashford, Kent, England).Image analysis was done using Pro-Plus 5.1 (Media Cybernetics Inc., Rockville, MD, USA) and ImageJ (U.S. National Institutes of Health, Bethesda, MD, USA). Thiobarbituric acid reactive species (TBARS) assay This assay followed Hadad and Goli's (2019) protocol with modifications.Electrospun material (5 mg) was dissolved in 4 mL of trichloroacetic acid (15% w/v) and 0.375% (w/v) thiobarbituric acid (TBA) in 0.25 M HCl in a glass test tube.Samples were placed in a water bath at 100 • C for 30 min and then submerged in cold water until they reached room temperature.The absorbance spectra of the samples were recorded from 300 to 800 nm at 21 • C with a 5 nm slit using a Duetta-Fluorescence and Absorbance Spectrometer (Horiba Scientific Inc., Edison, NJ, USA).The absorbance spectrum of the TBA reagent was recorded as a control and subtracted from the sample spectra before normalization to the initial absorbance (day 0) at 300 nm.The absorbance intensity at 530 nm was reported in triplicate samples (n = 6) tested at 0, 7, 14, 21, and 28 days of storage. Characterization of deterioration using luminescence spectroscopy techniques 2.6.1. Steady-state measurements Luminescence spectroscopy provides information on the molecular mobility and lipid oxidation of the matrix by recording the fluorescence emission spectra of intrinsic and extrinsic probes.Tyrosine (Tyr) is zein's predominant aromatic amino acid and an intrinsic reporter of molecular mobility.BODIPY 581/591 undecanoic acid (BODIPY C11) has dual emission bands at 515 and 595 nm, corresponding to oxidized and unaltered compounds, respectively.This oxidation-sensitive probe, dispersed in corn oil at 16.66 μM and mixed into the zein in 2-propanol dispersion, was electrospun directly onto strips (1 cm × 2 cm) of black paper (Astrobrights #22321, Neenah Inc, Alpharetta, GA, USA), using the settings described above.The strips were mounted on quartz slides (13.5 mm × 30 mm × 0.5 mm) (FireflySci, Staten Island, NY, USA) and placed at a 45 • angle in 1 cm light path quartz cuvettes (FireflySci) before equilibrating them to 21 • C. The fluorescence emission spectra of the probes were recorded using a Fluoromax-4 spectrophotometer (Horiba Scientific Inc., Edison, NJ, USA) using long-pass filters (Thorlabs, Newton, NJ, USA) to decrease recording light scattering and the different excitation and emission wavelengths and slits for each probe, as summarized in Table 1.The emission spectra without added probes (i.e., controls) were collected using the same experimental conditions, and subtracted from those of the samples to eliminate the background signal of the matrix.Triplicates (sample size of n = 6) were tested at 0, 2, 5, 7, 10, 14, 17, 21, 25, and 28 days of storage. Time-resolved measurements Tyr phosphorescence lifetimes were measured using a Cary Eclipse spectrophotometer (Agilent, Santa Clara, CA, USA) for the nonwovens attached to a 13.5 mm × 30 mm × 0.5 mm quartz slide (FireflySci) and placed at a 45 • angle in 1 cm-light path cuvettes.Samples were equilibrated at 5 • C using a Peltier system (Quantum Northwest, Liberty Lake, WA, USA) under a constant stream of nitrogen to displace the oxygen.Table 1 includes the experimental conditions for recording Tyr phosphorescence.The phosphorescence lifetime decays were characterized using a multi-exponential function (3 components) using OriginPro 2021 (OriginLab Corp., Northampton, MA, USA), and the average lifetime reported. Fluorescence microscopy Luminescent probes with an affinity for lipids (i.e., LipidTOX™ deep red and BODIPY C11) were loaded into the nonwovens.The neutral lipid stain LipidTOX™ deep red (10 μL) was added to 0.3 g of corn oil before including the oil in the spin dope solution.The LipidTOX™ allows visualization of the location and migration of oil in the nonwoven fibers throughout storage.BODIPY C11 in corn oil at 100 μM monitored and mapped lipid oxidation within the matrices.Table 1 summarizes the excitation and emission wavelengths for both probes.Micrographs of the nonwovens were obtained using a DM IRE2 luminescence microscope (Leica Microsystems, Wetlzar, Germany) by overlying the signals at each excitation wavelength. Confocal Raman Microspectroscopy (CRM) Raman mapping of samples on a WITec alpha300 R microscope (Witech, Ulm, Germany) equipped with an ultra-sensitive EMCCD detector and 532 nm solid-state diode laser excitation source collected the Raman scattered light in the backscattering geometry using a groove/mm grating after focusing on an area consisting of ribbon structures with a 100× magnification lens.A Raman scatter map of the selected area was obtained from 100 scans per line and 100 lines per image using 20 mW laser power and 0.05 s integration time.The Raman image was generated using the filter manager and true component analysis function of the WITec software using different colors and gradients to differentiate peaks and intensities.Based on the sample characteristics and purpose of the measurement, the amide I region (~1610 -1690 cm − 1 ) for protein and the C-H stretching region (~2830-3090 cm − 1 ) for lipids were monitored, as well as the disulfide region (490-550 cm − 1 ) and Tyr ring vibration/doublet band (I 850 /I 830 ). Fourier transform infrared (FTIR) spectroscopy The secondary structure of the protein in the nonwovens was analyzed using an FTIR spectrophotometer (Prestige21, Shimadzu Corporation, Kyoto, Japan) equipped with an attenuated total reflectance (ATR) attachment (MIRacle, Pike, Madison, WI).Reflectance spectra of the nonwoven matrices were collected at frequencies from 800 to cm − 1 .The Happ-Genzel apodization function was applied at a 4 cm − resolution and 6 mm aperture, averaging 32 scans.Spectra collected using Lab Solutions software (Shimadzu Corporation, Kyoto, Japan) were processed and analyzed using OriginPro 2021 (OriginLab Corp.).Vibrations of the C=O and C-N bonds in the 1600-1700 cm − 1 range, i.e., amide I region, monitored the predominant protein secondary structure during storage.Fourier self-deconvolution was used to identify the hidden peaks in the amide I region using a smoothing window of 5 and a smoothing factor of 0.15.The peaks fitted with Voigt functions allowed determining each peak contribution to the overall signal of the amide I region. Modeling degradation kinetics The kinetics of the main quality attributes studied, Y(t), specifically, matrix rigidity (expressed as an increment in the intensity of the intrinsic fluorophores), lipid oxidation, and formation of disulfide bonds were described using a logistic model (Aragao et al., 2008): where Y max corresponds to the maximum value of the attribute of interest, k is the reaction rate, and t c is a critical time when change becomes prevalent.The experimental data was adjusted to Eq. ( 4), using the nonlinear regression routine in Mathematica 13.2 (Wolfram Research, Inc. Champaign, IL, USA). Statistical analysis Statistically significant differences in all reported data were evaluated using one-way ANOVA and Tukey tests at α = 0.05 for mean comparisons in OriginPro (2021) (OriginLab Corp.).The goodness of fit of the mathematical characterization of the kinetic reactions was evaluated based on the adjusted R 2 and the mean squared error. Morphology and mechanical properties of nonwovens Scanning electron micrographs of fresh nonwovens showed ribbons with a width of 701 ± 23 nm, which decreased to 649 ± 58 nm and 620 ± 44 nm after 14 and 28 days, respectively (Fig. 1).The decrease in fiber width was attributed to the evaporation of the solvents (i.e., water and alcohol) initially bound or entrapped in the zein matrix and polymer rearrangements during storage.SEM is extensively used to visualize and assess the morphology of electrospun matrices to identify characteristic lengths and numbers of fibers, ribbons, beads, or their combination, after production (Pedram Rad et al., 2018).However, SEM has seldom been used to assess the deterioration of electrospun matrices during long-term storage.Besides the decrease in fiber width, physical breakage of the fibers was apparent through storage and previously reported (Bajsic et al. (2016)) for repeatedly UV-exposed polycaprolactone nonwovens with encapsulated titanium dioxide, causing extensive production of hydroperoxides and resulting in polymer chain scission and breakage.Herein, reactive oxygen species (ROS) form during the oxidation of the encapsulated oil, which alters the zein network integrity, increasing its brittleness.Therefore, ROS led to protein oxidation and solvent loss, reducing the amount of plasticizer, which could have contributed to shorter and thinner ribbons.The loss of plasticizer can also drive an increase in polymer-polymer interactions, which concurs with extensive reports that oxidation is a significant factor in protein network brittleness and increased breakage propensity (Plackett, 2011;Shah et al., 2023;Ullsten et al., 2016). The Young's modulus of the nonwovens was estimated from their tensile stress vs. deformation curves using eq.( 3).The Young's modulus of the nonwovens as a function of storage time is summarized in Table 2.After 28 days of storage, the Young's moduli of the nonwovens increased significantly by about 26%, similar to previous reports on biopolymerbased nonwovens, which exhibited a reduction in extensibility (i.e., higher rigidity) with increasing storage time (Taherimehr et al., 2021).Increased rigidity was at least partially attributed to the physical ageing of the biomaterial, a process where polymer chains rearrange within the matrix to achieve a minimum thermodynamic state, and the annealing rate depends on storage conditions, e.g., RH and temperature (Struik, 1977).Polymer chain rearrangement during physical ageing decreases the free volume between and within polymer chains, which modulates its mechanical properties, i.e., increasing material rigidity (Merrick et al., 2020).A reduction in material extensibility is reflected in the jaggedness of the tensile stress vs. deformation curves, which measure the brittleness of the material (Corradini and Peleg, 2006).The direction reversals increased from 31 ± 4 to 4.0 ± 3 % (Table 2), indicating higher brittleness, possibly due to a loss in nonwoven plasticization by the residual solvents as they evaporated during storage.This mechanical parameter provides valuable information on the increased propensity of the material to exhibit minor fractures as a function of time (Laurindo and Peleg, 2008).The increased brittleness aligns with the observed ribbon breakage in scanning electron micrographs upon storage of the nonwoven material. Lipid oxidation within nonwovens Lipid oxidation progressively increased by 78%, as assessed using TBARS, with the highest rate occurring in the first two weeks (Table 3).These results are comparable to those of Hadad and Goli (2019) and Aytac et al. (2017) for electrospun mucilage and poly-lactic acid (PLA) nanofibers carrying flaxseed oil and thyme essential oil, respectively.The minor discrepancies observed herein may be attributed to the incorporation of different biopolymers, lipids, and antioxidants (Aytac *Different letters per column denote significant differences (p < 0.05). L.A. Colaruotolo et al. et al., 2017;Hadad and Goli, 2019).The effect of fiber fabrication, e.g., exposure to a 50 kV voltage, on lipid oxidation was evaluated by comparing the TBARS of the oil before and after electrospinning, revealing no significant differences.Considering the TBARS assay results, it is clear that lipid oxidation is an important deteriorative effect in nonwovens during storage.The TBARS method, however, has low selectivity, e.g., besides detecting products from lipid oxidation, TBARS also reacts with protein oxidation products.Additionally, corn oil contains both mono-and polyunsaturated fatty acids and malonaldehyde, the primary compound measured by TBARS, is only formed during the oxidation of polyunsaturated lipids and not monounsaturated fatty acids (Barrera-Arellano et al., 2019).Thus, this technique does not fully characterize lipid oxidation in multi-component materials (Fantini and Yoshioka, 1992).Incorporating liposoluble fluorophores LipidTOX™ deep red and BODIPY C11 in the nonwovens allows lipids and the oxidation state to be visualized within a system, respectively.Micrographs of the nonwovens obtained immediately after production showed a relatively even distribution of oil throughout the core of the fiber (Fig. 2A i).The displacement and uneven distribution of the oil during storage are likely associated with solvent loss and increased protein-protein interactions, which appeared to displace the oil out of the core of the zein fiber encapsulant.The red emission of the BODIPY C11 (Fig. 2A ii) corresponds to the unoxidized probe and is evidence that the oil was not initially oxidized, as confirmed by a faint green emission of the oxidized BODIPY C11 (Fig. 2A iii).After 28 days of storage, the lipid was no longer evenly distributed (Fig. 2B i), with a concomitant increase in the ratio of the emission intensities of the oxidized to unoxidized BODIPY C11 (Fig. 2B ii and iii).Oxidation within the observed area using the emission intensities of the unoxidized and oxidized BODIPY C11 could be quantified.However, since the local oil concentration did not remain constant during storage, changes in emission could be confounded by changes in probe concentration.In comparison, steady-state luminescence spectroscopy provided a more reliable quantitative measure of oxidation.The fluorescence emission intensity at 595 nm, representing the unoxidized BODIPY C11, decreased slightly during the first five days of storage, consistent with a lag or delay period in the kinetics of lipid oxidation corresponding to the initiation stage.After the lag phase, the unoxidized BODIPY C11 emission progressively decreased to 44% of its initial value (Fig. 3).Unlike the TBARS assay, BODIPY C11 does not report on secondary oxidative products, thereby reducing the risk of overestimating the extent of lipid oxidation (Domínguez-Rebolledo et al., 2010).The primary advantage of BODIPY C11 in assessing lipid oxidation is that it accounts for the depletion and formation of two forms of a single compound (BODIPY C11), whose concentration remains constant throughout the entire storage time and does not rely on measuring diverse oxidized products (e.g., hydroperoxides) with ephemeral lifetimes. Zein structure in the nonwovens The luminescence emission of the aromatic amino acids (AAA), i.e., phenylalanine (Phe), Tyr, and tryptophan (Trp), in a protein, is sensitive to changes in the microenvironment where the AAAs are embedded.In the case of zein, the lack of Trp in its composition requires focusing on Tyr emission intensity and lifetimes, using steady-state and timeresolved measurements, respectively.Tyr is a useful probe to characterize the molecular mobility of protein confined in a matrix (e.g., Tyr buried in the hydrophobic core of the protein increases fluorescence emission, and phosphorescence lifetime) (Corradini et al., 2017;Draganski et al., 2015).In the present study, Tyr fluorescence emission from the nonwovens, monitored during storage, revealed an initial decrease in signal, probably due to material equilibration after electrospinning, followed by a monotonic increase until it plateaued (Fig. 4B).An increase in Tyr fluorescence emission intensity can stem from a combination of several effects, such as less solvation of the residue, solvent polarity affecting protein conformation, medium viscosity restricting the movement of the residue, and low temperature.(Lakowicz, 2006).Tyr tends to preferentially bury in the protein interior via hydrophobic forces to reduce exposure to the hydrophilic solvent.Observed changes in the fluorescence intensity corresponded to protein structure reorganization correlated to the physical ageing of a polymer with a glass transition temperature above the storage temperature (Monnier et al., 2021).Polymer chain rearrangement effectively reduces the free volume in the matrix, facilitating the formation of inter/intra molecular bonds, which increases the brittleness of a material (Low et al., 2018).Although the Tyr fluorescence intensity signaled that protein rearrangements may be responsible for the increased rigidity and brittleness of the electrospun samples, the dependence of this measurement on probe concentration complicated its interpretation.Therefore, to verify the aforementioned trend, Tyr phosphorescence emission lifetimes were collected as this photophysical property is concentration-independent. Lifetimes exhibited a similar increase as intensity, with initial average lifetimes of 3.59 ± 0.14 and 5.38 ± 0.37 ms after 28 days, corresponding to a 50% increase compared to the recently produced material (Fig. 4A). Besides monitoring Tyr emission, when excited at 270 nm, an additional band appeared at ~433 nm, which has been attributed to structural fluorescence arising from charge-transfer along a protein backbone with extensive β-sheet structures (del Mercato et al., 2007;Pinotsi et al., 2016).This emission band, tested under similar conditions, was also reported in zein/gluten doughs, particularly those with a high zein/gluten ratio due to the high β-sheets proportion in zein (Sadat et al., 2022).The emission intensity at 433 nm increased to a level similar to that of Tyr through time.This trend supports the hypothesis of protein rearrangements during storage. Additionally, in the nonwovens, a decrease in the unordered random coil structures from 25 to 0% occurred within 10 days, while the ordered structures, particularly β-sheet, increased from 2 to 19% by the end of storage time as assessed with FTIR analysis (Fig. 5).These observations are in accordance with the results obtained using luminescence spectroscopy techniques.Less drastic effects observed over a shorter time were reported in a study on electrospun zein nonwovens that encapsulated epigallocatechin gallate, showing a notable increase in β-sheet structures with increasing storage times (Li et al., 2009).Ordered structures have been characterized as more rigid than unordered structures due to increased inter/intra protein bonds (Perticaroli et al., 2013;Skrbic et al., 2016).The increased proportion of β-sheet structure during storage aligns with previous reports on increased material rigidity and brittleness. Chemical mapping of nonwovens To better understand the changing protein-lipid distribution regarding a change in chemical bonds, Raman micro-spectroscopy was employed.Micrographs of the nonwovens immediately after manufacturing presented a relatively even distribution of protein (red) L.A. Colaruotolo et al. and lipid (blue) (Fig. 6A).Conversely, after 28 days of storage, distinct regions of higher lipid content appeared closer to the ribbon surface (Fig. 6B).After 28 days, the decrease in the intensity of the Raman band located at 3008 cm − 1 , which corresponds to the unsaturated bonds in the vegetable oil, decreased by 47% suggesting lipid oxidation.The intensity ratio of the 1305 and 1267 cm − 1 bands originated from the rocking of the -C-H bonds and the in-phase twisting of = C-H bonds, respectively and increased by 35%, indicating oxidation of the encapsulated corn oil (Muik et al., 2007). Bands correlating to protein characteristics include the ratio of the peak intensities at 850 and 830 cm − 1 and the intensity of the scatter band at 520 cm − 1 , representing the Tyr doublet and disulfide bonds, respectively.In the electrospun fiber matrix, the ratio of the Tyr doublet signal decreased from 1.90 to 1.33, suggesting an increase in hydrogen bonding that correlated with a higher tensile strength in zein matrices (Turasan and Kokini, 2017).A 8% increase in the disulfide bond band was also observed, and while further investigation into the conformations of disulfide links is possible via deconvolution of the band at 520 cm − 1 , due to the low signal-to-noise ratio obtained from extensive diffraction in the tested matrices, the deconvolution required excessive smoothing, potentially resulting in loss of signal (Wang et al., 2022).Therefore, the intensity at 520 cm − 1 was directly used to benchmark the formation of disulfide bonds often formed due to the oxidation of cysteine amino acids, allowing inter/intra-molecular crosslinking (Joye et al., 2009).Oxidation can be speculated to be driven by secondary corn oil oxidation products or from the continual interaction of protein with molecular oxygen during storage.Chemical analysis of oxidative agents in the system throughout storage is needed to confirm this hypothesis. Correlation and kinetics of the observed changes A comparison of the measurements used to characterize physical (i.e., increased rigidity) and chemical (i.e., lipid oxidation) changes based on the fluorescence emission intensity of Tyr and BODIPY C11, respectively, contributes to elucidate the connection between these two phenomena.Plotting the normalized fluorescence intensity of Tyr vs. the oxidized BODIPY C11 revealed three phases that corresponded to the equilibration of the material with its environment, observable degradation, and finally, plateauing for both deteriorative phenomena (Fig. 7).This correlation highlights that rigidity and lipid oxidation changes follow a similar pattern and occur simultaneously.To better understand the timeline of the most relevant deteriorative effects in the nonwovens, the kinetics of the increase in matrix rigidity (i.e., the fluorescence intensity of the band at 430 nm), lipid oxidation (i.e., the fluorescence intensity of the oxidized BODIPY C11) and protein oxidation (i.e., the Raman band of the disulfide bonds), were characterized using Eq. ( 4).The experimental data adjusted with the logistic equation (Fig. 8), allow insights into the extent (Y max ), rate (k), and critical time (t c ) of each reaction (Corradini, 2018).Table 4 presents the parameters of Eq. ( 4) and the measures of goodness of fit. As revealed in Fig. 7, lipid oxidation and increase in matrix rigidity exhibited similar kinetics and are well described by the selected Fig. 6.Raman micrographs of electrospun zein with encapsulated corn oil after A) 0 and B) 28 days of storage.Red and blue spectra correspond to the protein and lipid, respectively.L.A. Colaruotolo et al. empirical model.Despite their similarity, scrutiny of the kinetic parameters revealed differences in the onset and rate of the respective phenomena, where an increase in rigidity preceded the onset of lipid oxidation, which, once initiated, accelerated.Conversely, disulfide bond formation proceeded slower than the other two phenomena, as evidenced by a rate an order of magnitude lower than the others. From the data, it is inferred that electrospun zein nonwovens with encapsulated corn oil first underwent an equilibration stage that involved evaporation of the residual solvent from the zein matrix and redistribution in which the glassy protein polymer rearranges into more ordered secondary conformations with the formation of additional inter/intramolecular bonds.This rearrangement consequently displaced lipids from the core to the surface of the electrospun fibers, evidenced by the Raman mapping.Lipid oxidation proceeded at an increased rate due to the localized clusters of lipids near the surface of the fibers, wherein a thinner protein layer conferred less protection from oxygen.Protein rearrangement increased rigidity and brittleness as more disulfide bonds were formed, thereby resulting in fiber breakage.To conclude, the limiting stability of the corn-oil loaded electrospun zein nonwoven can be attributed to protein rearrangements that lead to changes in the distribution of oil across the ribbon structures, thereby altering the nonwovens mechanical properties, and modulating lipid/protein oxidation.Strategies to extend the stability of these materials should focus on slowing protein reorganization to deter the onset of physical ageing, which may decrease the rate of lipid oxidation and increase overall stability. Conclusions This research explored the efficacy of spectroscopic techniques to assess food quality attributes of novel materials, i.e., electrospun nonwovens, as a function of time.Investigating deterioration at multiple length scales uncovered the main phenomena and respective interrelated progression that limit their stability.SEM showed ribbon-like fibers where the diameter significantly decreased during storage (701 ± 23 nm vs. 620 ± 44 nm).Extensive fiber breakage correlated with increased brittleness, as reported from bulk and local measurements, e. g., decreased extensibility by 26%, increased Tyr emission intensity by 40% and phosphorescence lifetime by 50%.Changes in mechanical properties and matrix rigidity also correlated with a transition in zein secondary structures from unordered (25% vs. 0%) to ordered β-sheets (2% vs. 19%) during storage.Raman and luminescence micrographs illustrate oil migration during storage, increasing lipid oxidation.The correlation between local rigidity and lipid distribution/oxidation suggests that protein structural reorganization increases material brittleness or propensity to physical breakage and displaces encapsulated oil into localized sites.Understanding the causes of deterioration aids in developing innovative strategies to improve the stability of these novel food-grade materials.It can also inform formulation and production to directly address deteriorative phenomena.The methods proposed herein to characterize deterioration at multiple length scales can track deteriorative mechanisms in other novel materials, expediting product development and extending product shelf-life. Fig. 4 . Fig. 4. A) Normalized fluorescence emission intensity of Tyr (λ exc = 280 nm, purple squares) and 440 nm band (λ exc = 350 nm, light blue circles) in zein nonwovens as a function of storage time.B) Tyr phosphorescence lifetimes in zein nonwovens during storage. Fig. 5 . Fig. 5. Protein secondary structure distribution as a function of storage time, as calculated from the amide I region of the FTIR spectra. Fig. 7 . Fig. 7. Correlation between normalized fluorescence emission intensity of Tyr and BODIPY C11 in electrospun zein matrices during storage. Fig. 8 . Fig. 8. A) Increased rigidity, represented by the intensity of the band at 440 nm, B) lipid oxidation, and C) disulfide bond formation as a function of storage time, fitted with Eq. 4. Filled circles: experimental data, dashed lines: adjusted fit. Table 1 Luminescent probes and testing conditions used for each analytical method. Table 2 Mechanical properties of zein nonwovens with encapsulated corn oil during storage. Table 3 Lipid oxidation, assessed using TBARS, of zein nonwovens with encapsulated corn oil during storage.	2024-06-30T15:21:22.642Z	2024-06-28T00:00:00.000	{ "year": 2024, "sha1": "410e3c339f353d05af2a1b6e7dc0d434fac15650", "oa_license": "CCBYNCND", "oa_url": "https://doi.org/10.1016/j.crfs.2024.100801", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "ae11d457b47b174dac2bf6b949286a0e1552231b", "s2fieldsofstudy": [ "Materials Science", "Environmental Science" ], "extfieldsofstudy": [ "Medicine" ] }
265353235	pes2o/s2orc	v3-fos-license	Risk factors associated with preterm birth among mothers delivered at Lira Regional Referral Hospital Background The World Health Organization (WHO) defines Preterm Birth (PTB) as “a live birth taking place before the expected 37 weeks of gestation”. Annually, approximately 15 million infants are born prematurely, constituting significantly to infant mortality during the initial four weeks of life, responsible for 40% of deaths among children under the age of five. Evidently, preterm deliveries have contributed to 46% of admissions to the neonatal intensive care unit (NICU) at Lira Regional Referral Hospital (LRRH) over the past three years. Paradoxically, while the prevalence of preterm births remains high, there is a lack of documented information regarding the underlying risk factors. Consequently, the primary objective of this study was to assess the potential risk factors associated with preterm birth at LRRH. Methods An analytical cross-sectional research was undertaken at LRRH, employing a quantitative methodology. The study utilized secondary data obtained from a total of 590 comprehensive maternal medical records, of deliveries that occurred at the facility between April 2020 and July 2021. The collected data underwent analysis using STATA version 17 software. To identify predictors of preterm birth, a Logistic regression model was applied, yielding adjusted odds ratios (AOR) alongside 95% confidence intervals (CI). The significance level was set at p < 0.05 to establish statistical significance. Furthermore, assessments for multicollinearity and model fitness were conducted using the Variance Inflation Factor (VIF) and linktest, respectively. Results The prevalence of preterm delivery among mothers who gave birth at LRRH stood at 35.8%. The outcomes of logistic regression analysis revealed that maternal employment status had a statistically significant association with preterm birth (AOR = 0.657, p = 0.037, 95%CI: 0.443–0.975); having a baby with low birth weight (AOR = 0.228, p < 0.001, 95% CI: 0.099–0.527) and experiencing preeclampsia (AOR = 0.142, p < 0.001, 95% CI: 0.088–0.229) were also identified as significant predictors of preterm birth in the study. Conclusions and recommendations The occurrence of preterm delivery is significantly higher (35.8%) among mothers who gave birth at LRRH when compared to the national average (13.6%). The prevalence of preterm birth among mothers was linked to factors such as employment status, delivery of low birth weight infants, and the presence of preeclampsia. Consequently, the research proposes a set of recommendations. Firstly, the Ministry of Health (MoH) should evaluate the present state of readiness within the healthcare system to effectively handle cases of preterm birth both within medical facilities and the community. Secondly, the Ministry of Gender, Labour, and Social Development should leverage Labor Officers to implement and uphold the regulations stipulated in the Employment Act and Labor Laws. Supplementary Information The online version contains supplementary material available at 10.1186/s12884-023-06120-4. Introduction Preterm birth (PTB), as defined by the World Health Organization (WHO), refers to "a live birth taking place prior to the completion of 37 weeks of gestation" [1].Annually, approximately 15 million infants are delivered prematurely, and this global pattern is experiencing an upward trajectory [1].Notably, PTB emerged as the primary cause of neonatal mortality within the initial four weeks of life, ranking as the second leading factor behind under-five deaths on a global scale; prematurity is responsible for 40% of under-five fatalities [2].Geographically, preterm birth prevails across 184 nations, displaying incidence rates ranging from 5 to 18% among all newborns.Specifically, Africa and South Asia jointly contribute to 60 to 85% of all instances of preterm births [3].Comparing demographics, approximately 12% of newborns in less affluent nations are born prematurely, in contrast to 9% in more economically prosperous countries [1]. Preterm birth rates in Sub-Saharan Africa have been investigated, revealing estimates of 10.9% in Gambia and 12% in Tanzania [4].Within global rankings, Uganda holds the 28th position for preterm births, standing at 13.6 per 1000 live births.Tragically, 25% of the 27 deaths per 1000 live births of newborns can be directly attributed to preterm births.In light of the new global infant mortality goals aiming for fewer than 10 deaths per 1000 live births by 2035, Uganda must intensify its efforts to reduce preterm birth rates [4].Achieving this objective would demand a targeted intervention on addressing preventable causes of preterm birth in low-income nations [4].While factors such as demographics, social circumstances, obstetric care, and pregnancy-related conditions have been implicated in preterm births worldwide, it's important to note that these risks differ according to geographical regions, leading to marked disparities between affluent and low-income countries [1,4,5]. To address the issue of premature births, the Ministry of Health, operating under the Government's initiative, has established a nationwide framework for pregnancy care and assistance.This includes bolstering emergency medical services and referral systems, enhancing maternal, neonatal, and child health provisions across all tiers of healthcare, expanding access to sexual and reproductive health amenities with a particular emphasis on family planning and age-appropriate information, and elevating the quality of national laboratory and diagnostic services [6].Additionally, the United States Agency for International Development (USAID) through its Regional Health Integration to Enhance Services-North, Lango Activity (RHITES-N, Lango), provides support to hospitals like the LRRH in terms of maternal healthcare services.Despite the implementation of these measures, data shows that over the past three years, 46% of admissions to the Neonatal Intensive Care Unit (NICU) at LRRH were attributed to preterm deliveries [7].While, prior researches focusing on various contexts have produced inconclusive findings regarding the predictors of preterm birth, there is a notable absence of documented evidence concerning predictors of premature birth specifically within Lira district.Given this research gap, the current study aims to evaluate the risk factors linked to preterm birth at LRRH, with the ultimate goal of pinpointing potential interventions capable of mitigating the occurrence of preterm births. Data source and study design The collection of secondary data involved the retrieval of information from the medical documents (specifically, prenatal cards and registers) of mothers who gave birth at LRRH during the period from April 2020 to July 2021.A cross-sectional sample of 590 comprehensive records of mothers who gave birth at LRRH was examined. Study variables and measurement The dependent variable in this study was preterm birth, dichotomized as birth occurring before 37 weeks (preterm birth) and at 37 weeks or later (term birth).Accordingly, a mother who gave birth before 37 weeks was classified as having experienced preterm birth and assigned a code of 1, while a mother who delivered at 37 weeks or beyond was categorized as having undergone term birth and assigned a code of 0. The independent variables included various aspects of family socio-demographics.Age was measured in completed years and subsequently grouped into three categories during the analysis: <18 years, 18-30 years, and 31 years or older.Marital status was stratified into four categories: married, single, divorced, and widow/ widower.Education level was divided into four categories: no education, primary education, secondary education, and tertiary education.Employment status was dichotomized as employed or not employed.Location was classified as rural or urban.Other independent variables included Body Mass Index (BMI) measured in kg/ m², which was segmented into four categories during cases of preterm birth both within medical facilities and the community.Secondly, the Ministry of Gender, Labour, and Social Development should leverage Labor Officers to implement and uphold the regulations stipulated in the Employment Act and Labor Laws.Keywords Factors associated, Preterm birth, Lira Regional Referral Hospital analysis: underweight, normal weight, overweight, and obese.Substance use was categorized as active drug user or not an active drug user.Maternal factors were also considered, with mode of delivery categorized as either normal delivery or caesarean section.Antenatal Care (ANC) attendance was categorized as either less than 3 times or 3 times or more.Anemia was divided into two categories: <10 g/dl and ≥ 10 g/dl.History of abortion was classified as either having experienced abortion or never having experienced abortion.The birth weight of the mother's baby was grouped into "yes" (< 2,500 g) or "no" (≥ 2,500 g).Comorbidities were classified as, whether present or absent.HIV status was categorized as sero-positive or sero-negative.Parity was stratified as having fewer than 4 children or having 4 or more children.Inter-pregnancy interval was segmented as less than 24 months or 24 months or more.Previous preterm birth was dichotomized as "yes" or "no, " and preeclampsia was similarly classified as present or absent.Furthermore, fetal factors were considered, including the sex of the child, categorized as male, female, or both (in the case of twins of different sexes).Pregnancy outcome was classified as singleton or twins.Congenital abnormalities were recorded as "yes" or "no." Data analysis The data analysis was conducted using STATA version 17.0 software.In the initial univariate analysis, frequencies and percentages were computed to describe the variables considered in the study.At bivariate analysis, associations were examined using the Pearson Chisquare test with a significance level set at p < 0.05.Factors that demonstrated significance in this analysis were selected for inclusion in the subsequent multivariate analysis.The purpose of the multivariate analysis was to estimate the individual net effects of each independent factor on the dependent variable.Notably, due to the low prevalence of preterm births (ranging from 5 to 18%) as indicated in the literature, which categorizes it as a rare event, the binary complementary log-log model was a potential choice for isolating net effects, as opposed to the logistic and probit models.However, an evaluation of three Link functions (Logistic, Complementary log-log, and Probit) was performed to determine their suitability for fitting the data.Based on the Akaike's Information Criteria (AIC) and Bayesian Information Criteria (BIC), the logistic model exhibited the lowest AIC and BIC values, rendering it the most plausible model for identifying predictors of preterm birth.Subsequently, two diagnostic tests were executed: a multicollinearity test and a model fitness test employing the Variance Inflation Factor (VIF) and link test, respectively. Ethics and consent to participate This study employed the Research Ethics Committee of Makerere University's School of Statistics and Planning to obtain approval and oversee the research process, ensuring adherence to both the regulations set forth by the Uganda National Council for Science and Technology (UNCST) and other international guidelines concerning the involvement of human subjects.The study made use of aggregated data sourced from a hospital, for which institutional authorization had been obtained to carry out the research.Informed consent was not pursued, as the data extracted did not contain any elements that could potentially reveal the identities of the patients.Physical copies of the data were securely stored and accessible solely to the research team.Similarly, electronic databases were safeguarded with passwords, with access restricted exclusively to the research team.Lastly, a request for a waiver of consent was submitted to Lira Regional Referral Hospital, an entity established to ensure the protection of study participants. Limitations The research employed pre-existing data gathered at the establishment.The occurrence of the Covid-19 pandemic influenced the quantity of mothers who gave birth at the facility during the specified time frame (April 2020 to May 2021).To tackle this issue, the investigation period was prolonged until July 2021.Certain instances of insufficient or absent information, notably within the records, were faced.Nevertheless, this challenge was managed by incorporating information from antenatal cards to supplement the absent data. Background characteristics and differentials in preterm births by Socio-demographic, maternal and fetal characteristics of the mothers Table 1 shows the distribution of participants based on socio-demographic, maternal, and fetal attributes.Some entries exhibited a lack of approximately 10% information in specific variables, causing a deviation from the total count of 590 respondents.These incompletely recorded instances were omitted from the analysis.The tabulated data indicates that out of the total 590 female participants, roughly 36% experienced preterm deliveries.The largest segment (82%) of the respondents fell within the age range of 18 to 30, approximately 56% were not engaged in employment, and nearly 48% resided in rural areas.In relation to marital status, a significant majority of the mothers (87%) were married, while an estimated 6% possessed limited or no literacy.Concerning maternal characteristics, nearly 9% of the mothers attended less than three antenatal care sessions, and around 40% underwent Caesarean section deliveries.Moreover, about An analysis of maternal characteristics demonstrates significant associations with preterm birth.These include parity (chi-square = 4.37, p = 0.037), low birth weight baby (chi-square = 20.46,p < 0.001), preeclampsia (chisquare = 81.34,p < 0.001), and previous preterm births (chi-square = 8.66, p = 0.003). Furthermore, an analysis of fetal characteristics indicates that the sex of the child (chi-square = 6.98, p = 0.031) and pregnancy outcome (chi-square = 10.56,p < 0.001) are also significantly associated with preterm birth. Multivariate analysis The factors that were found to be statistically and significantly associated with preterm birth during the bivariate analysis were further examined in multivariate logistic regression.These factors were; employment, parity, low birth weight baby, preeclampsia, previous preterm births, pregnancy outcomes, and the sex of the child.In Table 2, it is evident that the factors indicating a significant association with preterm birth were: maternal employment, the presence of a low birth weight baby, and the occurrence of preeclampsia.The results from the logistic regression analysis indicated the predictive potential of certain socio-demographic factors in relation to preterm birth.Specifically, the analysis indicated that maternal employment played a role in predicting preterm birth.Accordingly, mothers who were not employed exhibited a roughly 34% reduced odds of giving birth to a preterm baby when compared to those who were employed (AOR = 0.657, p = 0.037, 95% CI: 0.443-0.975).NB: All variables do not sum to 590 due to missing data; * Significance at 5%; X 2 = Chi-square 1 Formal employment (salary and wage earners) 2 In case of twins of different sexes Table 1 (continued) Examining maternal factors, the results revealed significant associations between preterm birth and having a baby with low birth weight and experiencing preeclampsia.In this regard, mothers who gave birth to babies weighing 2500 g or more displayed a nearly 77% reduced odds of having a preterm birth, in comparison to those who gave birth to babies weighing less than 2500 g (AOR = 0.228, p < 0.001, 95% CI: 0.099-0.527).Furthermore, mothers without preeclampsia exhibited an approximately 86% reduced odds of experiencing preterm birth when contrasted with mothers who had preeclampsia (AOR = 0.142, p < 0.001, 95% CI: 0.088-0.229). However, fetal factors such as the sex of the child, pregnancy outcome, and the presence of a congenitally abnormal baby were not found to have a statistically significant association with preterm birth. Prevalence of preterm birth The prevalence of preterm birth among mothers who delivered in LRRH was 35.8%.This indicated high prevalence compared to 13.6% in Mulago National Referral Hospital, Uganda [4], 18.3% in Nairobi, Kenya [8], 17.5% in a rural district hospital, Rwanda [9], 24.4% in Kilimanjaro Christian Medical Centre, Northern Tanzania [10], 25.9% in Jimma University Specialized Teaching & Referral Hospital, South-West Ethiopia [11], 24.3% in Mansoura University Hospital, Egypt [12], and 12.3% in Fafen Zone, Somalia area, Eastern Ethiopia [13].The high prevalence of preterm birth in LRRH could be due to poor quality of antenatal care.This is a result of inadequate equipment such as antenatal ultrasound machines to identify fetal antenatal conditions and skills to manage these diagnoses, inadequate skills for the early identification of mothers at risk of preterm birth for timely management, inadequate multiple micronutrient supplementation, poor referral systems, inadequate and poor health infrastructures, inadequate health supplies, and health system structural factors.These technical, interpersonal, resource, and infrastructural factors impede the provision and experience of good quality maternity care at health facilities [6]. Socio-demographic factors The study anticipated that socio-demographic characteristics, such as age, BMI, marital status, education level, location, employment, and drug use (smoking/alcohol), would not exhibit a significant association with preterm birth among mothers delivered at LRRH.However, the findings revealed that employment emerged as a noteworthy predictor of preterm birth.According to the results, mothers who were unemployed demonstrated a significant difference (AOR = 0.657, p = 0.037, 95% CI: 0.443-0.975),being 0.657 times less likely to deliver preterm babies compared to those who were employed.This outcome aligns with studies conducted in Indonesia, which identified that working mothers faced a 16.2 times higher risk of delivering late preterm infants (LPI) in comparison to housewives (OR = 16.2;95% CI: 2.315-123.444)[14]; in Mulago Hospital, Uganda, being unemployed (AOR = 0.36, 95%CI: 0.15-0.86,p = 0.021) was associated with a 64% reduction in the likelihood of experiencing preterm birth [4].Similarly, in Cyprus, long working hours (OR: 3.77, 95% CI: 2.08-6.84)were about 4 times linked to preterm birth [15].At Mansoura University Hospital, Egypt, increased risks were also observed between long working hours and temporary contracts, and the risk of preterm delivery (AOR = 2.36, CI: 1.18-7.78)and (AOR = 1.98,CI: 1.72-8.74)respectively [12].However, the findings of this study are not corroborated by research conducted in Nigeria, which revealed that maternal occupation did not significantly affect gestational age (χ²=10.143,p = 0.428) and birth weight (χ²=16.807,p = 0.079) at delivery.Nevertheless, it did significantly affect stillbirth rates (χ²=28.134,p = 0.002) [16].Furthermore, the nature of employment might impact a pregnant woman differently, depending on whether it involves manual or labor-intensive work.Similarly, a heavy workload could subject an expectant mother to stress, potentially leading to pregnancy complications and resulting in preterm delivery. Maternal factors The study hypothesized that antenatal care (ANC) visits, mode of delivery, parity, history of abortion, inter-pregnancy interval, mothers with low birth weight babies, HIV status, comorbidity, anemia, preeclampsia, and previous preterm births are not significantly associated with preterm birth.However, this study found that mothers with low birth weight babies and preeclampsia were statistically significant predictors of preterm birth.The statistical analysis indicated that a mother with a low birth weight baby was a statistically significant predictor of preterm birth (AOR = 0.228, p < 0.001, 95% CI: 0.099-0.527).Accordingly, mothers who gave birth to babies with a weight of ≥2500 g were 0.228 times less likely to experience preterm birth than those who gave birth to babies weighing < 2500 g.This finding aligns with studies conducted at Shire Suhul General Hospital in Northern Ethiopia.Those studies discovered that mothers with a history of bearing neonates weighing less than 2500 g, including the most recent birth (AOR: 2.78, 95% CI: 1.39-5.55),were 2.8 times more likely to have preterm deliveries compared to their counterparts [17].Similarly, a study at Jimma University Specialized Teaching and Referral Hospital in South West Ethiopia demonstrated that a history of low birth weight (OR = 0.085, CI: 0.04-0.18,p < 0.001) increased the likelihood of preterm delivery by a factor of 0.085 compared to those without a history of poor birth outcomes [11].Further studies conducted in Abu Dhabi, United Arab Emirates, revealed that low birth weight babies (AOR = 17.62,CI: 11.05-28.10,p < 0.001) were nearly 18 times more likely to experience preterm births than their counterparts [5].Additionally, in Public Hospitals of Fafen Zone, Somali Region, Eastern Ethiopia, newborns with birth weights less than 2500 g (AOR = 3.78, 95% CI: 1.55-9.84,p < 0.001) had a 3.78 times higher likelihood of being delivered preterm than babies with birth weights of ≥2500 g [13].These findings could be attributed to various factors, including intrauterine growth restriction resulting from genetic factors and uterine infections, inadequate prenatal nutrition, chronic health conditions like diabetes, heart problems, and high blood pressure, placental issues, or maternal infections preventing proper oxygen and nutrient delivery to the fetus.Such conditions may contribute to the occurrence of preterm births and low birth weight babies [13,17]. Additionally, preeclampsia was a statistically significant predictor of preterm birth (AOR = 0.254, p < 0.001, 95% CI: 0.185-0.348).This implies that mothers without preeclampsia were 0.254 times less likely to experience preterm birth than those with preeclampsia.This result is consistent with studies conducted at Mukalla Maternity and Childhood (MCH) Hospital in Yemen, which revealed that mothers with pre-eclampsia (AOR = 4.120; CI: 1.818-9.340,p < 0.001) were 4.12 times more likely to deliver preterm babies than those without [18].Similarly, at Mulago Hospital in Uganda, mothers with preeclampsia (AOR = 16.24,95% CI: 3.11-84.70,p < 0.001) were 16 times more likely to have a preterm birth than mothers without preeclampsia [4].In Nanjing Maternity and Child Health Care Hospital, China, the odds of preterm birth among mothers with preeclampsia were 2.46 times higher (AOR = 2.46, 95% CI: 1.78-3.40,p < 0.001) than for mothers without the condition [19].Likewise, in a regional hospital in Accra, Ghana, mothers with pre-eclampsia/eclampsia (AOR = 3.4, 95% CI: 1.0-11.9,p < 0.001) were 3.4 times more likely to experience preterm delivery than those without preeclampsia [20].Preeclampsia may result from pregnancy complications characterized by high blood pressure, poor nutrition, high body fat, insufficient blood flow to the uterus, genetic factors, and signs of damage to other organ systems, most frequently affecting the liver and kidneys. Fetal factors The study also hypothesized that fetal factors (such as the sex of the child, pregnancy outcome, and congenital abnormalities status) are not associated with preterm birth.However, this study revealed that no fetal factors were associated with preterm birth, which contradicts studies conducted in Public Hospitals of the central zone, Tigray, Ethiopia.These previous studies showed that neonates with congenital/birth defects (AOR = 3.19, 95% CI: 1.22, 8.34, p < 0.05) were three times more likely than those without any birth defects to experience preterm birth [21].Similarly, research conducted at Shire Suhul General Hospital in Northern Ethiopia found that visible physical neonatal congenital anomalies in the most recent baby (AOR = 10.4;95% CI: 1.66-65.2,p < 0.05) increased the odds of preterm birth occurrence compared to normal babies [17].This discrepancy could be attributed to the interaction of genetic and environmental risk factors contributing to preterm delivery in Ethiopia, which may not be a prevalent issue in Uganda, especially in Lira District [21]. Conclusions This study examined the occurrence of preterm birth (PTB) and its related determinants within the context of deliveries at LRRH.The prevalence of preterm birth among mothers who gave birth at LRRH was documented at 35.8%.Specifically, the likelihood of preterm delivery was found to be lower among unemployed mothers in comparison to their employed counterparts.Similarly, mothers whose babies had a birth weight of less than 2500 g were observed to have a higher probability of delivering preterm compared to those whose babies had a normal birth weight.Conversely, the presence of preeclampsia was associated with a greater probability of preterm birth, whereas mothers without preeclampsia exhibited a reduced likelihood of delivering preterm babies. Recommendations Basing on the key findings from the study, the followings recommendations were made: 1. To address the elevated occurrence of preterm birth within the LRRH and Lango sub-region, it is imperative for the Ministry of Health to evaluate the existing state of the healthcare system's readiness to handle preterm births.This involves devising comprehensive plans and making necessary preparations, particularly in terms of procuring essential equipment.Furthermore, conducting training and mentorship programs aimed at imparting vital managerial and clinical proficiencies is essential.These programs are designed to empower frontline healthcare providers in promptly identifying and effectively managing expecting mothers who face a heightened risk of preterm birth. A crucial aspect of this initiative involves bolstering the referral system, ensuring a seamless pathway for timely and appropriate care.Concurrently, elevating the quality of antenatal care necessitates Table 1 Differentials in preterm births by Socio-demographic, maternal and fetal characteristics of the mothers (N = 590) 14% of the mothers tested positive for HIV, 6% gave birth to low weight infants, and almost 23% experienced preeclampsia.In regard to fetal attributes, a considerable proportion of the mothers (97%) had single babies, 54% of whom were female, and roughly 3% exhibited congenital abnormalities.Table1also provides an overview of variations in preterm birth across different socio-demographic, maternal, and fetal characteristics.The table demonstrates that several factors are significantly linked to preterm birth, including maternal employment, parity, low birth weight babies, preeclampsia, previous preterm births, the sex of the child, and pregnancy outcome.Differences in preterm birth based on socio-demographic characteristics reveal that maternal employment (chi-square = 4.90, p = 0.027) is notably associated with preterm birth. Table 2 Multivariate Analysis of factors associated with preterm birth	2023-11-23T14:51:38.102Z	2023-11-23T00:00:00.000	{ "year": 2023, "sha1": "29256de393e5cb0c01c57f53df9f393b78601dc4", "oa_license": "CCBY", "oa_url": "https://bmcpregnancychildbirth.biomedcentral.com/counter/pdf/10.1186/s12884-023-06120-4", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "57ea52c10f0e8a9ad5b16a5907033209461679e1", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
7075948	pes2o/s2orc	v3-fos-license	The C Ring of the F1Fo ATP Synthase Forms the Mitochondrial Permeability Transition Pore: A Critical Appraisal The mitochondrial permeability transition pore (MPTP) is a non-specific pore in the inner mitochondrial membrane (IMM) whose opening is triggered by high matrix [Ca2+] to which it is sensitized by [Pi] and oxidative stress. MPTP opening plays a critical role in necrotic cell death such as in cardiac ischemia/reperfusion (I/R) injury and the action of cytotoxic drugs. Indeed MPTP inhibition with cyclosporine A (CsA) protects tissues from I/R injury (1). Matrix cyclophilin D (CyP-D), the target of CsA, facilitates MPTP opening but the identity of the pore-forming proteins (to which CyP-D binds) remains unresolved (2–4). Extensive evidence supports the adenine nucleotide translocase (ANT) being the site of inhibition by adenine nucleotides and bongkrekic acid (BKA – an ANT ligand) and activation by carboxyatractyloside (CAT – another ANT ligand) and oxidative stress (5). Furthermore, the ANT binds CyP-D and, when reconstituted into proteoliposomes, it produces Ca2+-activated pores similar to the MPTP (6, 7). In addition, liver mitochondria lacking ANT1 and ANT2 exhibit MPTP opening that is insensitive to adenine nucleotides, CAT and BKA, and requires higher [Ca2+] than control mitochondria (8). However, since pore opening can still be observed, other IMM proteins must be able to form the MPTP. One candidate is the mitochondrial phosphate carrier (PiC), which binds CyP-D, more so following oxidative stress, and may be the locus of MPTP activation by Pi (9, 10). However, neither partial knockdown nor over-expression of PiC in cell lines affected MPTP opening (10, 11), although MPTP opening in heart mitochondria from mice with cardiac-specific PiC knockout were less calcium sensitive (12). Overall, the available data suggest that the ANT and PiC play roles in MPTP opening but that another IMM protein must also be involved. Several recent papers suggest that this may be the F1Fo ATP synthase. CyP-D was shown to bind to the F1Fo ATP synthase and modulate its hydrolytic activity (13, 14) and in 2013, the laboratories of both Bernardi (15) and Pinton (16) presented data that implicated the ATP synthase in MPTP formation. Bernardi and colleagues (15) detected Ca2+-activated channels, similar to the MPTP, in phospholipid bilayers containing reconstituted dimers of mammalian F1Fo ATP synthase. Similar channel activity was demonstrated in yeast mitochondria and this was strongly attenuated in mutants lacking the e and γ subunits needed for ATP synthase dimer formation (17). However, high levels (0.3 mM) of Ca2+ were required for channel opening which, unlike MPTP opening, also required Bz-423. No data were presented on the effects of oxidative stress, CsA, or recombinant CyP-D. Furthermore, Pinton and colleagues (18) pointed out that Bernardi’s laboratory had previously demonstrated MPTP opening in ρ0 cells, which lack the mitochondrial DNA encoding the α and A6L subunits of the ATP synthase. In addition, the ATPase inhibitor protein F1, which promotes ATP synthase dimerization, attenuates rather than promotes MPTP opening, and enhances cell survival under ischemic conditions (19). Rather, Pinton and colleagues (16) implicated the c-subunits of the Fo ATPase in MPTP formation, showing that their knockdown reduced MPTP opening in response to ionomycin or hydrogen peroxide and their over-expression enhanced opening. The c-subunits form a ring structure in the IMM, and so represent an attractive candidate for forming the MPTP, but direct evidence for this was not provided. However, the paper of Alavian et al. (20) claims to do this. Alavian et al. (20) confirmed the observations of Bonora et al. (16), but they also reconstituted the purified c-subunit into proteoliposomes and demonstrated channel activity. Most channels conducted at ~100-pS but a few did so at 1.5–2 nS, similar to the MPTP conductance (21). However, the channels were insensitive to Ca2+ and CsA and were only inhibited by much higher concentrations of ATP and ADP than required to inhibit MPTP opening. The authors proposed that other F1Fo ATP synthase components are needed for MPTP regulation, which they investigated using purified monomeric F1Fo ATP synthase reconstituted into proteoliposomes. Some infrequent channel activity was observed, which was increased by addition of CyP-D and further by 100 μM Ca2+. These effects were prevented by 5 μM CsA, a concentration much higher than the Ki for CyP-D (2 nM). Channels sensitive to both Ca2+ and CsA could also be detected in patches from sub-mitochondrial vesicles (SMVs) enriched in F1Fo ATP synthase or IMMs and these were absent when OSCP and β subunits of the ATP synthase and bound CyP-D were removed by urea treatment, while adding back purified β-subunit to reconstituted c-subunits largely abolished channel activity. However, the authors did not address whether the IMM and SMV preparations also contained ANT and PiC, which is very likely. Rather, they concluded that the sites through which Ca2+, ADP, and CyP-D (and thus CsA) modulate channel activity are on the F1 domain of the ATP synthase whose association with the c-subunit ring may loosen upon Ca2+ and CyP-D binding. This might cause expansion of the ring converting it into a high conductance channel and, using fluorescent probes, they presented evidence consistent with the c-subunits moving apart during MPTP opening. Furthermore, channel activity was greatly enhanced when glycine residues in the c-subunit transmembrane domains were replaced with valines, thus moving the packed helices further apart (20). However, interpretation of these data is difficult because the expressed c-subunits ran at 15 kDa on SDS-PAGE and not 7.6 kDa, the size of the mature c subunit, suggesting that the mitochondrial targeting sequence had not been removed. Indeed several studies in this paper and that of Bonora et al. (16) showed expression of 15 kDa unprocessed protein rather than the true c subunit. In summary, when the evidence for an involvement of the F1Fo ATP synthase and more specifically its c-subunit in MPTP formation is reviewed critically, it is legitimate to conclude that it is no better than that for the involvement of the ANT and PiC. Perhaps the truth lies in a synthesis, and that an interaction between the ANT, PiC, and F1Fo ATP synthase in the ATP synthasome (22) is critical for MPTP formation, as we (2) and subsequently others (4, 18) have concluded. A scheme illustrating how the different components may interact is presented in Figure Figure11. Figure 1 A hypothetical model of the MPTP that proposes an interaction between the ANT, PiC, and F1Fo ATP synthase in the ATP synthasome. It is proposed that the MPTP forms at the interface between interacting domains of the PiC, ANT, and Fo ATP synthase following ... The mitochondrial permeability transition pore (MPTP) is a non-specific pore in the inner mitochondrial membrane (IMM) whose opening is triggered by high matrix [Ca 2+ ] to which it is sensitized by [Pi] and oxidative stress. MPTP opening plays a critical role in necrotic cell death such as in cardiac ischemia/reperfusion (I/R) injury and the action of cytotoxic drugs. Indeed MPTP inhibition with cyclosporine A (CsA) protects tissues from I/R injury (1). Matrix cyclophilin D (CyP-D), the target of CsA, facilitates MPTP opening but the identity of the pore-forming proteins (to which CyP-D binds) remains unresolved (2)(3)(4). Extensive evidence supports the adenine nucleotide translocase (ANT) being the site of inhibition by adenine nucleotides and bongkrekic acid (BKA -an ANT ligand) and activation by carboxyatractyloside (CAT -another ANT ligand) and oxidative stress (5). Furthermore, the ANT binds CyP-D and, when reconstituted into proteoliposomes, it produces Ca 2+activated pores similar to the MPTP (6,7). In addition, liver mitochondria lacking ANT1 and ANT2 exhibit MPTP opening that is insensitive to adenine nucleotides, CAT and BKA, and requires higher [Ca 2+ ] than control mitochondria (8). However, since pore opening can still be observed, other IMM proteins must be able to form the MPTP. One candidate is the mitochondrial phosphate carrier (PiC), which binds CyP-D, more so following oxidative stress, and may be the locus of MPTP activation by Pi (9, 10). However, neither partial knockdown nor over-expression of PiC in cell lines affected MPTP opening (10,11), although MPTP opening in heart mitochondria from mice with cardiac-specific PiC knockout were less calcium sensitive (12). Overall, the available data suggest that the ANT and PiC play roles in MPTP opening but that another IMM protein must also be involved. Several recent papers suggest that this may be the F1Fo ATP synthase. CyP-D was shown to bind to the F1Fo ATP synthase and modulate its hydrolytic activity (13,14) and in 2013, the laboratories of both Bernardi (15) and Pinton (16) presented data that implicated the ATP synthase in MPTP formation. Bernardi and colleagues (15) detected Ca 2+ -activated channels, similar to the MPTP, in phospholipid bilayers containing reconstituted dimers of mammalian F1Fo ATP synthase. Similar channel activity was demonstrated in yeast mitochondria and this was strongly attenuated in mutants lacking the ε and γ subunits needed for ATP synthase dimer formation (17). However, high levels (0.3 mM) of Ca 2+ were required for channel opening which, unlike MPTP opening, also required Bz-423. No data were presented on the effects of oxidative stress, CsA, or recombinant CyP-D. Furthermore, Pinton and colleagues (18) pointed out that Bernardi's laboratory had previously demonstrated MPTP opening in ρ 0 cells, which lack the mitochondrial DNA encoding the α and A6L subunits of the ATP synthase. In addition, the ATPase inhibitor protein F1, which promotes ATP synthase dimerization, attenuates rather than promotes MPTP opening, and enhances cell survival under ischemic conditions (19). Rather, Pinton and colleagues (16) implicated the c-subunits of the Fo ATPase in MPTP formation, showing that their knockdown reduced MPTP opening in response to ionomycin or hydrogen peroxide and their over-expression enhanced opening. The c-subunits form a ring structure in the IMM, and so represent an attractive candidate for forming the MPTP, but direct evidence for this was not provided. However, the paper of Alavian et al. (20) claims to do this. Alavian et al. (20) confirmed the observations of Bonora et al. (16), but they also reconstituted the purified c-subunit into proteoliposomes and demonstrated channel activity. Most channels conducted at 100-pS but a few did so at 1.5-2 nS, similar to the MPTP conductance (21). However, the channels were insensitive to Ca 2+ and CsA and were only inhibited by much higher concentrations of ATP and ADP than required to inhibit MPTP opening. The authors proposed that other F1Fo ATP synthase components are needed for MPTP regulation, which they investigated using purified monomeric F1Fo ATP synthase reconstituted into proteoliposomes. Some infrequent channel activity was observed, which was increased by addition of CyP-D and further by 100 µM Ca 2+ . These effects were prevented by 5 µM CsA, a concentration much higher than the K i for CyP-D (2 nM). Channels sensitive to both Ca 2+ and CsA could also be detected in patches from sub-mitochondrial vesicles (SMVs) enriched in F1Fo ATP synthase or IMMs and these were absent when OSCP and β subunits of the ATP synthase and bound CyP-D were removed by urea treatment, while adding back purified βsubunit to reconstituted c-subunits largely abolished channel activity. However, the authors did not address whether the IMM and SMV preparations also contained ANT and PiC, which is very likely. Rather, they concluded that the sites through which Ca 2+ , ADP, and CyP-D (and thus CsA) modulate channel activity are on the F1 domain of the ATP synthase whose association with the c-subunit ring may loosen upon Ca 2+ and CyP-D binding. This might cause expansion of the ring converting it into a high conductance channel and, using fluorescent probes, they presented evidence consistent with the c-subunits moving apart during MPTP opening. Furthermore, channel activity was greatly enhanced when glycine residues in the c-subunit transmembrane domains were replaced with valines, thus moving the packed helices further apart (20). However, interpretation of these data is difficult because the expressed c-subunits ran at 15 kDa on SDS-PAGE and not 7.6 kDa, the size of the mature c subunit, suggesting that the mitochondrial targeting sequence had not been removed. Indeed several studies in this paper and that of Bonora et al. (16) showed expression of 15 kDa unprocessed protein rather than the true c subunit. In summary, when the evidence for an involvement of the F1Fo ATP synthase and more specifically its c-subunit in MPTP formation is reviewed critically, it is legitimate to conclude that it is no better than that for the involvement of the ANT and PiC. Perhaps the truth lies in a synthesis, and that an interaction between the ANT, PiC, and F1Fo ATP synthase in the ATP synthasome (22) is critical for MPTP formation, as we (2) and subsequently others (4, 18) have concluded. A scheme illustrating how the different components may interact is presented in Figure 1.	2016-06-17T21:24:54.048Z	2014-08-07T00:00:00.000	{ "year": 2014, "sha1": "aa8903acc96e37c80f0eb4d669873ad766e81765", "oa_license": "CCBY", "oa_url": "https://www.frontiersin.org/articles/10.3389/fonc.2014.00234/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "aa8903acc96e37c80f0eb4d669873ad766e81765", "s2fieldsofstudy": [ "Medicine", "Chemistry" ], "extfieldsofstudy": [ "Biology", "Medicine" ] }
54631284	pes2o/s2orc	v3-fos-license	Alpha-Lipoic Acid Functionalized Silver Nanoparticles for Colorimetric Detection of Copper Ion. A facile, low cost, rapid and easy-to-use colorimetric sensor for the detection of Cu2+ ion is described using alpha-lipoic acid (ALA) functionalized AgNPs. The prepared ALA-AgNPs are yellow, and spherical with an average particle size of 9.23 ± 3.36 nm. The addition of Cu2+ ion resulted in the aggregation of ALA-AgNP solution and caused a visible color change from yellow to orange-brown, which could be detected by the naked eye and UV-vis spectrophotometer. Parameters affecting the sensitivity of the sensor were investigated, including ALA concentrations, reaction times, reaction volumes and pH. Under optimal conditions, the colorimetric sensor could detect Cu2+ ion in the range of 0.625 × 10-5 to 10 × 10-5 M with a linear coefficient (r) of 0.9978. The limit of detection (LOD) and limit of quantification (LOQ) were 0.43 × 10-5 M (4.3 μM) and 1.45 × 10-5 M (14.5 μM), respectively, which is lower than the limit set by the U.S. Environmental Protection Agency (EPA). This proposed method showed good selectivity and could detect Cu2+ within 5 min. Finally, the proposed method could be successfully applied for the analysis of Cu2+ ion in tap water samples. Introduction was observed in the spectrum at wavelength 394 nm (Fig. 2 a), and the color of the solution is yellow (Fig. 2, inset a). For ALA-AgNPs, the SPR band was slightly shifted and the intense absorption peak was observed at 399 nm (Fig. 2 b). The color of ALA-AgNPs is lighter than AgNPs as illustrated in Fig. 2, inset b. The yellow and transparent color of the two nanoparticles indicates the good dispersity in water. 28 FT-IR spectroscopy was used to analyze spectral signal to characterize and confirm the existence of ALA on the surface of AgNPs. The FT-IR spectra of standard ALA and ALA-AgNPs are shown in Fig. 3 for comparison. The FT-IR spectrum of standard ALA (Fig. 3 a) shows a characteristic band at 1694 cm -1 due to the stretching of carbonyl group (C=O) in the carboxylic acid. 29 The band at 734 cm -1 corresponds to weak S-S stretching vibration. 22 This weak S-S band is absent for ALA-AgNPs which is attributed to the deprotonating and formation of Ag-S bond (Fig. 3 b). 22 Thus, the FT-IR spectra indicated the successful functionalization of ALA to AgNPs surface. The mechanism of the sensing system The principle of Cu 2+ colorimetric detection is depicted in Fig. 1. The AgNPs functionalized with ALA exhibit high surface charges of carboxylic group and the electrostatic repulsion in between enable them to disperse uniformly in aqueous solution. The ALA-AgNPs exhibit a unique SPR property with maximum band at 399 nm ( Fig. 4 a). The addition of Cu 2+ ion (1.0×10 -4 M) led a red-shift of the peak at 399 nm where its intensity was reduced and a new peak was emerged at about 550 nm (Fig. 4 b). The SPR of 399 and 550 nm are related to the quantities of dispersed and aggregated ALA-AgNPs, respectively. Thus, the absorbance ratio at 550 and 399 nm (A550/A399) express the molar ration of aggregated and dispersed ALA-AgNPs and this was used as a sensing signal. The carboxylic groups of ALA functionalized to the surface of AgNPs interact with Cu 2+ ion to cause an aggregation and hence, a change of solution color from yellow to orange-brown (Fig. 4, inset a and b). In contrast, the addition of Cu 2+ ion Analytical Sciences Advance Publication by J-STAGE The A550/A399 signal reached maximum within 5 min after addition of Cu 2+ ion (1.0 × 10 -4 M) and gradually decreased to the minimum and then kept constant in 30 min. As the criteria for colorimetric sensor is fast detection, therefore, the reaction time of 5 min was selected as ideal reaction time for UV-vis spectra measurement and photograph since this time was enough to record digital images of color change and register UV-vis spectra. The reaction time of 5 min was then used for further experiments. In general, the pH value plays an important role for the interaction of metal ions and the nanoparticles. Consequently, the pH was investigated in the pH range from 4-10. As illustrated in Figure 6 d, the maximum A550/A399 signal was obtained at pH 5.3. At higher pH value, the A550/A399 signals were decreased. The possible reason is that under high alkaline conditions, stable dispersion of AgNPs could be formed through interaction with OHion. 30-31 According to the results above, the optimal conditions (1.0 × 10 -4 M ALA, reaction time 5 min, volume ratio 1,000 : 1,000 and pH 5.3) were used for Cu 2+ detection. concentration from 0.625 × 10 -5 -10 × 10 -5 M with a linear coefficient (r) of 0.9978. The limit of detection (LOD) and limit of quantification (LOQ) were calculated from 7 linear calibration curves using Eq. 1 and Eq. 2, respectively. 32 where Sa is the standard deviation of the response (y-intercept) and b is the slope (sensitivity) of the calibration curves. The LOD was 0.43 × 10 -5 M (4.3 µM) which is below the safe limit for drinking water (20 µM) set by US Environmental Protection Agency (EPA). ion. Real sample analysis In order to evaluate the feasibility of the developed method in detecting Cu 2+ ion in samples, analytical recoveries were studied. When mixing with ALA-AgNPs, the samples did not show any color change. The tab waters samples were then spiked with different concentration of Cu 2+ ion. Samples were analyzed in triplicate, and the average signal response was used to calculate the Cu 2+ concentration using standard curve obtained prior to analysis. The percentage recoveries were calculated by Eq. 3. where C1 = concentration determined in fortified sample, C2 = concentration determined in unfortified sample and C3 concentration of fortification. 33 Recoveries of four tap water samples were summarized in Table 2. The recoveries were in the range of 84.04 -96.40 % with RSD 2.02 -3.82 %. Since the acceptable recovery in analytical analysis in 1.0 × 10 -5 M level is 80-110 % and RSD is 7.3-11.3 %, 35 therefore, recoveries for all samples are acceptable. To validate the ALA-AgNPs assay, Cu 2+ ion spiked water samples were analyzed with the developed method and ICP-OES. These results are shown in Table 2. Comparison between the two methods was done by a paired sample t -test. There was no significant difference between the results obtained from these two method at a 95 % confidence level (t = -1.841, degree of freedom = 3 and Sig (2-tailed = .163)). This result indicates that the developed method is reliable for detecting of Cu 2+ ion in tap water samples. Thus, the colorimetric sensor using ALA functionalized AgNPs is a promising analytical tool for detection of Cu 2+ in water sample. Analytical Sciences Advance	2018-12-12T19:53:54.703Z	2019-04-01T00:00:00.000	{ "year": 2019, "sha1": "a77a3a5e82a7bcbcfdecf7c994c8def2c0070c43", "oa_license": null, "oa_url": "https://www.jstage.jst.go.jp/article/analsci/35/4/35_18P442/_pdf", "oa_status": "GOLD", "pdf_src": "Adhoc", "pdf_hash": "f63be7206a1a37a32eceaea01c47960737f96a02", "s2fieldsofstudy": [ "Chemistry", "Materials Science" ], "extfieldsofstudy": [ "Chemistry", "Medicine" ] }
271278953	pes2o/s2orc	v3-fos-license	Community Point of Care Testing in Diagnosing and Managing Chronic Kidney Disease Chronic kidney disease (CKD) poses a significant global health challenge with increasing prevalence and associated morbidity. Point-of-care testing (POCT) provides an opportunity to improve CKD management and outcomes through early detection and targeted interventions, particularly in underserved communities. This review evaluates the roles of POCT in CKD, focusing on utility (through screening programs, monitoring of kidney function, and assessing participants on renally excreted medications), accuracy, and acceptability. Screening programs employing POCT have demonstrated promising outcomes, with improved rates of CKD diagnosis in groups with disparate health outcomes, offering a vital avenue for early intervention in high-risk populations. These have been conducted in rural and urban community or pharmacy settings, highlighting convenience and accessibility as important facilitators for participants. In addition, POCT holds significant promise in the monitoring of CKD, particularly in groups requiring frequent testing, such as kidney transplant recipients and patients on renin-angiotensin-aldosterone inhibitors. The consideration of the variable analytical performance of different devices remains crucial in assessing the utility of a POCT intervention for CKD. While the convenience and improved accessibility of home self-testing versus healthcare professional management is important, it must be balanced with acceptable levels of accuracy and precision to maintain patient and clinical confidence. Despite challenges including variability in accuracy and the user-friendliness of devices, patient feedback has generally remained positive, with studies reporting increased patient satisfaction and engagement. However, challenges regarding wider uptake are limited by healthcare professional confidence (in test reliability), the potential for increased workload, and early prohibitive costs. In conclusion, POCT represents a growing and valuable tool in enhancing CKD care, particularly in resource-limited settings, but careful consideration of device selection and implementation strategies is essential to achieve desired outcomes. Introduction Chronic kidney disease (CKD) is a growing global public health issue that currently affects between 9 and 13% of the population worldwide.In combination with acute kidney injury (AKI), it is predicted to be a top-five cause of mortality by 2040 [1][2][3].The prevalence, incidence, and progression of kidney disease varies between and within countries, influenced by access to healthcare, political systems, available technology, and socioeconomic status, with people in the lowest socioeconomic quartile having up to 60% increased risk of progressive CKD compared to those in the highest quartile [1,[4][5][6]. The most common causes of CKD worldwide are diabetes mellitus, accounting for 30-50% of people with CKD (80% in low-and middle-income countries) and hypertension [7,8], both of which are also disproportionately overrepresented in young ethnic minority populations in high-income countries (e.g., in Black and South Asian populations in the United Kingdom), and are associated with an increased risk of accelerated decline in kidney function [9][10][11]. Diagnosis of CKD is made using estimated glomerular filtration rate (eGFR), calculated primarily using isotope dilution mass spectrometry (IDMS)-traceable creatinine, age, biological sex, and the urinary albumin: creatinine ratio (uACR) [12].An alternative or adjunct to creatinine is cystatin-C.This has been shown to improve eGFR estimation, particularly when used in conjunction with creatinine.It is therefore recommended by the KDIGO CKD 2023 guidelines for confirmatory CKD testing and is utilised in paediatrics and parts of mainland Europe [12].However, it has still not been widely adopted (e.g., it is not recommended by the National Institute for Health and Care Excellence in the UK, due to the risk of false positives) and is not readily available in a number of laboratories [13,14].Furthermore, there are no studies to our knowledge using POCT for cystatin-C; therefore, it will not be discussed in this review. CKD is associated with an increased risk of end-stage renal failure (ESRF), cardiovascular events, and premature death.A health economic analysis in the UK published in 2011 reported that CKD was responsible for an excess of 40,000 premature deaths, 12,000 excess myocardial infarctions, and 7000 excess cerebrovascular accidents each year in the United Kingdom alone [15]. As CKD is irreversible and progressive, organisations such as the National Institute for Health and Care Excellence (NICE) in the United Kingdom have placed emphasis on the early detection of the disease to facilitate intervention and slow the progression of the disease and its associated comorbidities [13]. Rates, detection, and management of kidney disease tend to be better defined in high-income countries; however, emerging evidence suggests that low-and middle-income countries are likely to have similar if not greater burdens of disease [16,17].Furthermore, detection may be limited by healthcare infrastructures such as proximity to routine testing, costs, and available workforce. Point-of-care testing (POCT), also referred to as 'rapid tests', is defined as in vitro procedures that are performed in close proximity to patients and are characterised by a rapid turnaround time [18,19].Aligned with growing expectations for timely investigations and management amongst patients, the market for POCT has been rapidly increasing.It was worth approximately USD 44 billion in 2022 and is expected to rise to USD 78 billion by 2030 [20].Within this market, kidney function testing has also been advancing with a range of POCT devices available for most routine renal biochemistry tests and urinalysis. The purpose of this review is as follows: (1) summarise the available devices, with a focus on accuracy and precision; (2) discuss the roles of POCT in CKD; and (3) describe current healthcare professional and patient/service user perspectives as a potential facilitator or barrier to the wider use of POCT in daily clinical practice. POCT Devices for Kidney Disease: Accuracy, Precision and Acceptability The diagnosis and classification of CKD is based on the estimated glomerular filtration rate (eGFR) and urinary albumin: creatinine ratio (UACR), as defined by the Kidney Diseases: Improving Global Outcomes (KDIGO) CKD 2023 Guidelines [12].eGFR is calculated using an equation derived from age, biological sex, and endogenous filtration marker concentration, of which creatinine, a waste product from predominantly skeletal muscle, is most commonly used worldwide.It is measured using either the modified Jaffe or enzymatic assays with isotope dilution mass spectrometry (IDMS) as a reference standard and quantified in units of mg/dL or µmol/L.Turnaround time for the assay is 3 of 13 usually a few hours; however, accounting for sample delivery and communicating results to the requesting clinician can add lengthy delays which may impact care. POCT is a rapid alternative to laboratory testing.There are a number of devices available, but due to variability in design (Figure 1) and sample analysis methodology for both POCT and laboratory assays, the accuracy and precision of POCT (intra-individual and inter-device) compared to laboratory testing is inconsistent. Diagnostics 2024, 14, x FOR PEER REVIEW 3 of 14 or enzymatic assays with isotope dilution mass spectrometry (IDMS) as a reference standard and quantified in units of mg/dL or µmol/L.Turnaround time for the assay is usually a few hours; however, accounting for sample delivery and communicating results to the requesting clinician can add lengthy delays which may impact care. POCT is a rapid alternative to laboratory testing.There are a number of devices available, but due to variability in design (Figure 1) and sample analysis methodology for both POCT and laboratory assays, the accuracy and precision of POCT (intra-individual and inter-device) compared to laboratory testing is inconsistent.Accuracy typically is expressed as absolute or percentage bias.The coefficient of variation (CV = standard deviation divided by mean) is a marker of precision.Analytical total error (ATE = bias + 1.96 × standard deviation) takes into account systematic bias (inaccuracy) and random bias (imprecision) to provide a more comprehensive assessment.Standardised reference ranges have been developed by various groups, including the Clinical Laboratory Improvement Amendments (CLIA), the Laboratory Working Group for the National Kidney Disease Education Program (NKDEP), and the Clinical & Laboratory Standards Institute (CLSI) [21][22][23].A summary of ATE and acceptable percentage bias is described in Table 1.Accuracy typically is expressed as absolute or percentage bias.The coefficient of variation (CV = standard deviation divided by mean) is a marker of precision.Analytical total error (ATE = bias + 1.96 × standard deviation) takes into account systematic bias (inaccuracy) and random bias (imprecision) to provide a more comprehensive assessment.Standardised reference ranges have been developed by various groups, including the Clinical Laboratory Improvement Amendments (CLIA), the Laboratory Working Group for the National Kidney Disease Education Program (NKDEP), and the Clinical & Laboratory Standards Institute (CLSI) [21][22][23].A summary of ATE and acceptable percentage bias is described in Table 1.The Statsensor (Nova Biomedical, Waltham, MA, USA) has been widely assessed for creatinine (and eGFR).It is small, portable, requires a small sample volume (1.2 µL), and is user-friendly.Imprecision (CV) has been reported as ranging from 3 to 13% over a wide range of creatinine concentrations [24][25][26][27][28][29][30][31].However, it tends to be less accurate with older ages and greater creatinine concentrations.In vitro studies have suggested this is linked to high creatine and urea concentrations falsely elevating creatinine results [27].A new model called the Nova Max Creatinine device has since been released by Nova Biomedical, with one study (N = 517) reporting 98.9% sensitivity and 85.0% specificity in detecting eGFR < 60 mL/min/1.73m 2 ) compared to laboratory reference standards [32]. The i-STAT Alinity (Abbott, Abbott Park, IL, USA) measures a wider array of biochemical markers, including creatinine and eGFR.It is also a light (660 g) hand-held device requiring 65 µL of blood and has a turnaround time of 2 min.The Epoc Blood Analysis System (Siemens Healthineers, Erlangan, Germany) is a slightly larger device but remains hand-held and portable.It requires a larger volume of blood (92 µL) and therefore would be less suitable for home self-testing. Currin et al. compared both the i-STAT and Statsensor to the laboratory Jaffe method IDMS-traceable creatinine assays and iohexol-derived measured GFR in a rural South African population (N = 674).Both laboratory and POC tests overestimated Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) eGFR compared to measured GFR, though, the positive bias was less for i-STAT and Statsensor (8.4 vs. 19.9 vs. 28.6 mL/min/1.73m 2 , respectively).However, the POCT devices showed wider imprecision (4.6% and 10.2% vs. 3.5%, respectively) [33].Van der Heijden et al. compared both devices alongside the Epoc Blood Analysis System.I-Stat had the best accuracy (ATE = 6%), followed by Epoc, then Statsensor.All devices showed considerable variability around the creatinine reference standard, although Epoc performed best (95% limits of agreement −0.49 to 0.49 mg/dL). A list of devices with a summary of operational and analytical characteristics is described in Table 2.These devices can be used across all age ranges, although there are fewer validation data in children, particularly with the handheld devices. The Role of Point of Care Testing (POCT) in Kidney Disease For kidney disease, POCT involves a blood test (for creatinine, eGFR, and potassium) and/or a urine test (UACR or urinary protein: creatinine ratio).POCT provide an opportunity to target early interventions in settings more convenient for service users, to improve health outcomes and patient quality of life [34]. The National Institute for Health Research's Horizon Report in 2014 identified four areas where POCT would offer additional benefits to routine care: (1) screening for CKD by detection of elevated creatinine levels; (2) dose adjustment of prescribed medications in patients with renal impairment; (3) monitoring of CKD and (4) detection of AKI (including AKI on a background of CKD) [35]. Screening for CKD Studies describing POCT CKD screening programs include FINISHED (First Nations Community Based Screening to Improve Kidney Health and Prevent Dialysis), Kidney Evaluation for You (KEY), and a pharmacist-led study in the United Arab Emirates [34,36,37].FINISHED was a comprehensive screening, triage and treatment program for indigenous populations in rural and remote communities across Canada, to identify individuals with diabetes, hypertension and CKD.Out of 1700 people screened, 25.5% of adults (N = 343) were found to have CKD and 15% of children were identified to have early kidney disease (either through POC blood testing or urinalysis) [36,38].This facilitated earlier treatment for CKD in a cohort with a higher prevalence of kidney disease but reduced access to healthcare, improved nephrology follow-up (8.4% versus 2.5% prior to screening) and was cost-effective with a gain of CAD 23,700 per quality-adjusted life year [39]. KEY was a community-based pilot study in diverse Australian regions (a rural mining town, regional centre, and metropolitan city) to screen high-risk individuals for conditions, including CKD.Similar to FINISHED, 20.5% of undiagnosed CKD cases were identified out of 402 participants enrolled in the study, using the i-STAT POC device for creatinine/eGFR or urine dipstick and subsequent UACR testing with the expectation of delaying future morbidity and mortality through early identification [34]. Pharmacy-led initiatives in community settings have also been performed in Canada and the United Arab Emirates.Donovan et al. piloted a study using the Statsensor device, identifying 10 CKD cases from 89 participants [40].In the United Arab Emirates, patients with risk factors for CKD (diabetes mellitus, hypertension, or a positive family history of CKD) were targeted and POCT was performed on 400 participants at pharmacies using the PICCOLO (Abaxis) blood chemistry analyser (see Table 2), identifying undiagnosed CKD in 38.8% (N = 155) of participants [37]. Limitations included a lack of confirmatory testing for diagnosis (i.e., diagnosis was made with a single eGFR < 60 mL/min/1.73m 2 ), selection bias, and lack of data on sampling success and patient experience. Furthermore, in three of the four studies, the POCT devices were operated by a healthcare professional, in most cases due to sample volume required ± device operation (the exception being Donovan et al. which used Statsensor), which precludes the option for home testing and potentially increases staff workload.This adds further workload and does not allow the option for home testing. Despite these potential barriers, the studies discussed all highlight the potential benefits of using POCT in rural, remote, or urban community settings: (1) screening for the presence of CKD, (2) determining the severity and associated risk, and (3) initiating and optimising management through implementation of focused strategies to delay disease progression and manage risk factors, with the potential benefits of reduction in long-term morbidity and mortality. Presently, screening for CKD using capillary POCT remains limited in the literature.This is likely related to the increased costs required to support this process and the potential increased workload that would result from greater identification.However, given the rapid advances in technology, the growing demand for more accessible care, and increasing awareness of the associated morbidity and premature mortality associated with CKD, it is likely that this is a field which will continue to grow as resources and financial support are redirected towards it. Dose Adjustment of Prescribed Medications in Patients with Renal Impairment The KDIGO 2023 Clinical Practice Guideline for the Evaluation and Management of CKD are for sodium-glucose co-transporter 2 inhibitors (SGLT2i) to be prescribed along with a maximum tolerated dose of angiotensin-converting enzyme inhibitors (ACEi) or angiotensin receptor blockers (ARB; see Figure 2) [12].The combination of these medications for people with diabetic and non-diabetic proteinuric kidney disease has demonstrated a risk reduction in cardiovascular events, reduced proteinuria, and slowed the progression of kidney disease [41,42], adding up to 20 additional years before reaching ESRF for a middle-aged person optimised with an eGFR of 60 mL/min/1.73m 2 [43].In addition, fineronone has recently been shown to also deliver similar outcomes in addition to ACEi or ARB [44].However, these medication groups require renal function monitoring after initiation and incremental dose changes (see Figure 3), which provides a potential target for POCT to (1) guide rapid up-titration and (2) utilise patient-led home monitoring. Presently, screening for CKD using capillary POCT remains limited in the literature.This is likely related to the increased costs required to support this process and the potential increased workload that would result from greater identification.However, given the rapid advances in technology, the growing demand for more accessible care, and increasing awareness of the associated morbidity and premature mortality associated with CKD, it is likely that this is a field which will continue to grow as resources and financial support are redirected towards it. Dose Adjustment of Prescribed Medications in Patients with Renal Impairment The KDIGO 2023 Clinical Practice Guideline for the Evaluation and Management of CKD are for sodium-glucose co-transporter 2 inhibitors (SGLT2i) to be prescribed along with a maximum tolerated dose of angiotensin-converting enzyme inhibitors (ACEi) or angiotensin receptor blockers (ARB; see Figure 2) [12].The combination of these medications for people with diabetic and non-diabetic proteinuric kidney disease has demonstrated a risk reduction in cardiovascular events, reduced proteinuria, and slowed the progression of kidney disease [41,42], adding up to 20 additional years before reaching ESRF for a middle-aged person optimised with an eGFR of 60 mL/min/1.73m 2 [43].In addition, fineronone has recently been shown to also deliver similar outcomes in addition to ACEi or ARB [44].However, these medication groups require renal function monitoring after initiation and incremental dose changes (see Figure 3), which provides a potential target for POCT to (1) guide rapid up-titration and (2) utilise patient-led home monitoring.The screening programs discussed included a triage and treatment follow-up after identification [34,36,37,40,46].Two were performed in a pharmacy setting, which has proven to be a popular location in community settings.This is logical, as pharmacies specialise in medication management and are often a trusted part of diverse local communities [47]. This is highlighted a Dutch study, where 46 elderly patients on renally excreted medications had POCT, identifying over half with mild-to-moderate renal impairment, leading to medication adjustments and positive experiences reported by participants (4/5 using a Likert scale; 1 = bad, 5 = excellent) [48].Ramos et al. used the NovaMax Statsensor device on 552 patients across 15 pharmacies in Spain.If dose adjustment or withdrawal was recommended, due to a decline in function or an eGFR below the recommended threshold, patients were referred to their physician, leading to modification of medications in 39 (7.1%) patients [49]. These community programs were able to successfully identify patients at risk on renally excreted medications in local community healthcare settings and established a strong foundation for future work to design and evaluate complex interventions using POCT to start renin-angiotensin-aldosterone inhibitors (RAASi) and monitor them through the up-titration process.However, additional POCT to monitor potassium concentrations is important, particularly for patients with advanced CKD and/or diabetes mellitus. Monitoring of CKD There is an opportunity for POCT in specific CKD cohorts requiring frequent surveillance, such as renal transplant recipients, patients on RAASi or diuretic therapies, or patients requiring potentially nephrotoxic treatments (e.g., intravenous injection of contrast The screening programs discussed included a triage and treatment follow-up after identification [34,36,37,40,46].Two were performed in a pharmacy setting, which has proven to be a popular location in community settings.This is logical, as pharmacies specialise in medication management and are often a trusted part of diverse local communities [47].This is highlighted in a Dutch study, where 46 elderly patients on renally excreted medications had POCT, identifying over half with mild-to-moderate renal impairment, leading to medication adjustments and positive experiences reported by participants (4/5 using a Likert scale; 1 = bad, 5 = excellent) [48].Ramos et al. used the NovaMax Statsensor device on 552 patients across 15 pharmacies in Spain.If dose adjustment or withdrawal was recommended, due to a decline in function or an eGFR below the recommended threshold, patients were referred to their physician, leading to modification of medications in 39 (7.1%) patients [49]. These community programs were able to successfully identify patients at risk on renally excreted medications in local community healthcare settings and established a strong foundation for future work to design and evaluate complex interventions using POCT to start renin-angiotensin-aldosterone inhibitors (RAASi) and monitor them through the up-titration process.However, additional POCT to monitor potassium concentrations is important, particularly for patients with advanced CKD and/or diabetes mellitus. Monitoring of CKD There is an opportunity for POCT in specific CKD cohorts requiring frequent surveillance, such as renal transplant recipients, patients on RAASi or diuretic therapies, or patients requiring potentially nephrotoxic treatments (e.g., intravenous injection of contrast dye).There is now an established and growing utility for POCT in each of these subgroups. Kidney transplant recipients often have to travel greater distances than most patients and have to test more frequently, making this an ideal option for POCT.Key factors to consider in choosing a device are its accessibility, user-friendliness (in home settings), accuracy, and precision. Murray et al. conducted a small observational feasibility study (N = 15) in stable kidney transplant recipients in the UK, using the Abbott i-STAT Alinity analyser at home for both creatinine and potassium concentrations compared to their standard laboratory testing [31].The hand-held device was deemed easy to use by patients; however, the test success rate was 70% (42/60), with two-thirds of unsuccessful tests being due to device-related reasons.Within-patient bias for creatinine was 2.25 umol/L (95% CI: −12.13, 16.81 µmol/L); however, potassium was poorer at 0.66 mmol/L (95% CI: −0.147, 2.79 mmol/L) [31]. Lint et al. also reported that post-transplant patients (N = 30) were highly motivated and satisfied with home-testing using the Nova Biomedical Statsensor Xpress device, despite trust in the accuracy of creatinine being low, likely due to variation in subsequent measurements [50].Lint et al. also performed a larger observational study with 138 kidney transplant recipients, comparing creatinine capillary POCT vs. venous POCT vs. IDMStraceable enzymatic assays [30].The coefficient of variation (CV) was 10.4% for capillary sampling and 5.2% for capillary sampling, which was above the authors' defined allowable total error of 6.9% and the NKDEP's desirable goal of 7.6%.These findings were similar to results published by Nataatmadja et al. in a study of 60 participants with CKD (20 with kidney transplants) who reported CV ranging from 5.8% to 11.3% [29].In addition, there was no statistically significant difference between trends of capillary POCT samples compared to venous samples, suggesting that monitoring trends rather than detecting single sample changes in kidney function may be a more effective use of POCT testing [30]. Finger-prick capillary sampling using hand-held devices at home provides improved accessibility for patients, improving patient engagement and satisfaction and preserving venous access.Accuracy and precision in this cohort are vital, as there is a lower threshold for inaccurate results to cause anxiety, reduce patient and clinician confidence, and trigger further investigations.It is likely that a compromise between patient accessibility and user-friendliness versus accuracy and precision is required, and interpretation according to known limitations in test performance.Therefore, choosing the most appropriate device to meet these demands is key to a successful intervention. Patient and Healthcare Professional Perspectives on POCT for Kidney Disease To ensure the successful uptake and expansion of POCT in healthcare settings, confidence in the tests, through weighing up the benefits and limitations for both healthcare professionals and service users, is vital.Factors which will contribute to this decision include accuracy and precision, cost (to the service user and the healthcare system), accessibility (e.g., travel time, transport, and/or organising childcare), invasiveness of the test (e.g., finger-prick versus venepuncture), and previous experience. Studies of patient and healthcare professional perspectives for kidney function assessment using POCT are limited; however, the perspectives from the wider use of noncreatinine POCT are valuable, as the themes regarding facilitators and barriers for implementation are likely to be similar. In rural regions or areas where there are limited healthcare resources, one qualitative study based on 101 semi-structured interviews from patients, healthcare professionals, and policymakers in South Africa identified themes surrounding protracted turnaround times for 'routine testing', due to reduced accessibility and longer travel times (for patients and samples) [51]. However, participants from a mixed-methods service evaluation involving participants receiving home visits in the UK (N = 47) felt that POCT had benefited their care, with improved accessibility, reassurance with instant results, and reduced travel time as important factors [52].This has been previously described in the Australian KEY screening program (N = 402), which reported improved satisfaction with POCT in 99% of partici-pants, with increased convenience and better understanding of their condition (96% of participants) [34]. From a clinical perspective, alongside patient satisfaction, additional useful features for POCT include user-friendliness (linked with the amount of training required) and simple interpretation of results.Furthermore, the technical performance of devices, proven effect on clinical outcomes, and reliability of testing were all potential facilitators to the uptake of POCT [53]. Test reliability is key to providing user confidence, as both patients and healthcare professionals have expressed concerns regarding the risk of inaccurate results, which can lead to false reassurance, unnecessary anxiety, additional confirmatory testing, and extra workload [53,54]. Barriers to consider include increased costs (at least in the short term), legislative roadblocks, accessibility, extra workload, and access to maintenance and repairs for POCT equipment, particularly in rural settings [51]. Conclusions POCT provides an accessible tool for achieving rapid results, enabling improved detection of CKD, monitoring of kidney function in high-risk groups, adjustment of renally excreted medications, and instant screening prior to interventions such as contrastenhancing agent administration.It has been demonstrated to improve patient satisfaction, experience, and engagement in clinical services, and provides an attractive implementation tool for rural and remote regions where conventional testing is more challenging and health inequalities are compounded.POCT provides an exciting alternative to delivering healthcare within local communities, which may have the added benefit of reducing health inequalities that currently exist in nephrology. Due to the range of devices and differences in their technical specifications, the choice of device will often be suited to the remit of the intervention in question, and these factors need to be considered by the care team to optimise effectiveness. The scope for further expansion of kidney function testing includes optimisation of CKD medications and monitoring of renal function in patients on RAASi or diuretic therapies, initially targeting high-risk groups.Barriers included reduced accuracy and precision compared to laboratory testing, and increased short-term costs. Figure 1 . Figure 1.Illustrates the operational, analytical and human factors, which should be considered for each device in a given clinical context. Figure 1 . Figure 1.Illustrates the operational, analytical and human factors, which should be considered for each device in a given clinical context. Figure 3 . Figure 3.An illustration of the KDIGO treatment algorithm for the initiation and follow-up management of ACEi and ARB.This figure has been reproduced from the published KDIGO 2024 CKD Clinical Practice Guidelines for the Evaluation and Management of CKD [45].Abbreviations: ACEi = Angiotensin-converting enzyme inhibitors; ARB = angiotensin receptor blocker; eGFR = estimated glomerular filtration rate. Figure 3 . Figure 3.An illustration of the KDIGO treatment algorithm for the initiation and follow-up management of ACEi and ARB.This figure has been reproduced from the published KDIGO 2024 CKD Clinical Practice Guidelines for the Evaluation and Management of CKD [45].Abbreviations: ACEi = Angiotensin-converting enzyme inhibitors; ARB = angiotensin receptor blocker; eGFR = estimated glomerular filtration rate. Table 1 . List of analytical total error (ATE) and coefficients of variation (CV) for creatinine and potassium, and estimated glomerular filtration rate (eGFR). Organisation Limits of Acceptable Performance/Analytical Total Error (ATE) National Kidney Disease Education Programme; BV = Spanish Society of Clinical Chemistry and Molecular Pathology (SEQC) Table of Desirable Quality Specifications based on Biological Variation; RCPA = The Royal College of Pathologists of Australasia and the Australasian Clinical Biochemist Association Quality Assurance Program. Table 2 . List of devices discussed with summary of operational and analytical product characteristics. * Manufacturer locations are as follows: Nova Biomedical, Waltham, MA, USA; Abbott, Abbott Park, IL, USA; Zoetis, Parsippany, NJ, USA; Siemens Healthineers, Erlangan, Germany; Abaxis, Union City, CA, USA; Radiometer, Copenhagen, Denmark. Range of costs is indicative and may be subject to change depending on the specific supplier, promotions, and bulk purchasing agreements.For exact up-to-date pricing, the supplier should still be contacted directly.* Unable to ascertain costs of sensor cassette.	2024-07-19T15:04:57.891Z	2024-07-01T00:00:00.000	{ "year": 2024, "sha1": "97fe674e2d06418446756a5db69d4f67d6b3dbaf", "oa_license": "CCBY", "oa_url": "https://doi.org/10.3390/diagnostics14141542", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "59e1d07668daad0a17cd17e409c56e56c1919cf6", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
11761864	pes2o/s2orc	v3-fos-license	Identifying a Window of Vulnerability during Fetal Development in a Maternal Iron Restriction Model It is well acknowledged from observations in humans that iron deficiency during pregnancy can be associated with a number of developmental problems in the newborn and developing child. Due to the obvious limitations of human studies, the stage during gestation at which maternal iron deficiency causes an apparent impairment in the offspring remains elusive. In order to begin to understand the time window(s) during pregnancy that is/are especially susceptible to suboptimal iron levels, which may result in negative effects on the development of the fetus, we developed a rat model in which we were able to manipulate and monitor the dietary iron intake during specific stages of pregnancy and analyzed the developing fetuses. We established four different dietary-feeding protocols that were designed to render the fetuses iron deficient at different gestational stages. Based on a functional analysis that employed Auditory Brainstem Response measurements, we found that maternal iron restriction initiated prior to conception and during the first trimester were associated with profound changes in the developing fetus compared to iron restriction initiated later in pregnancy. We also showed that the presence of iron deficiency anemia, low body weight, and changes in core body temperature were not defining factors in the establishment of neural impairment in the rodent offspring. Our data may have significant relevance for understanding the impact of suboptimal iron levels during pregnancy not only on the mother but also on the developing fetus and hence might lead to a more informed timing of iron supplementation during pregnancy. Introduction The clinical importance and prevalence of iron deficiency (ID) make the understanding of this micronutrient deficiency an important challenge for both the scientific and medical communities. Iron is an essential micronutrient and ID affects more than 2 billion people around the world. ID occurs in many forms ranging from marginal tissue iron depletion to the most severe form of iron deficiency anemia (IDA). It has been estimated that, globally, 50% of anemia can be attributed to ID. IDA ranks at number 9 among 26 mortality risk factors and accounts for over 800,000 deaths and 35 million disability-adjusted lost life years [1]. North America alone bears 1.4% of the global burden of ID and IDA and it has been estimated that 35-58% of healthy women show some degree of ID, with a higher prevalence during pregnancy [1,2,3,4,5]. This high prevalence of ID during pregnancy seems at odds with the practice of routine iron supplementation as part of the prenatal care provided in most developed countries. Factors that contribute to the still high prevalence are complex and include concerns of early iron supplementation generating oxidative stress [6], a low compliance rate (50%) of taking iron supplements even with optimal motivation and guidance due to the undesirable adverse effects [7,8,9], and a rise in risk factors like type-2 diabetes and obesity that can cause ID and IDA despite adequate food intake [10,11,12,13]. ID and untreated IDA during pregnancy have many negative consequences for the offspring and have been shown to be associated with a higher incidence of low birth weight and prematurity [14,15,16,17,18], long-term cognitive abnormalities such as language learning impairments and behavioral changes [19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35], alteration in thermoregulation [36], changes in lipid metabolism [37], stroke and seizures [38,39], altered motor function and coordination [40,41,42], and in many cases alteration of Auditory Brainstem Responses (ABRs, [35,43,44,45,46], a measure of nerve impulse conduction in the auditory system). Gestational ID also has been shown to change iron homeostasis in the offspring resulting in increased risk of developing ID later in life despite adequate nutrition [47,48,49]. It is alarming that the reported prevalence of ID for US children under 2 years of age is estimated to be 25% [23,50,51,52], which might well be an underestimation due to the difficulty in diagnosing ID in the absence of anemia. Despite the recognition that ID during pregnancy can have multiple adverse effects on the developing fetus, it remains elusive to what degree and during which gestational time window maternal ID has to occur to affect fetal development to a degree that leads to functionally relevant long-term impairments. We addressed this question using a highly controlled animal model system that allowed us to stage the initiation of ID during pregnancy using a defined feeding regimen and to analyze the fetal development as well as the neural function in the young adult offspring. Our prior studies on the effects of IDA during pregnancy suggested that IDA affects a very early arising precursor cell pool. The impairment of this cellular population is likely to contribute to an ultimate disruption of proper CNS development during postnatal development [53]. Based on those studies, we now address the fundamental questions of whether (i) the generation of neural impairment in the offspring is limited to cases of maternal IDA as opposed to ID without anemia and whether (ii) the time window during which the iron depletion occurs in the dam determines the degree of functional impairments in the offspring. Results Maternal iron restriction that does not cause severe maternal anemia renders the embryo iron depleted with postnatal onset of anemia One of the major challenges in understanding the impact of iron on development is the difficulty of measuring iron concentrations in the developing embryo. Animal models allow us to gain insight into the relationship between maternal iron intake and fetal iron levels. We established a feeding protocol in which rat dams were provided a customized iron-deficient diet that led to maternal iron deficiency (ID) but not to severe iron deficiency anemia (IDA). As shown in Figure 1A, this diet did not generate a discernable difference in hematocrit values in the pregnant dams compared to control dams, although hemoglobin (Hb) and red blood cell counts (RBC) were slightly below the control range by 21 days of gestation. Even though severe anemia was absent in the dams, we found progressively reduced serum iron levels, ranging from a .40% reduction at 15 days of gestation up to an 82% decrease at 21 days of gestation, relative to controls (CTLs; Figure 1B). To determine the impact of the decrease in serum iron concentration in the dam on the iron status in the developing embryo, we measured the iron concentration of whole embryos using atomic absorption spectroscopy (AAS; Figure 2A). While we did not see a significant reduction in tissue iron concentrations at E13, at E15 embryos from iron deficient dams had only 44% of the levels of iron seen in control embryos. Despite a developmental increase in total iron concentration in the iron deficient embryos from an average of 8 ng Fe/mg tissue at E15 to approximately 14 ng Fe/mg at E19, these values were far below the normal Fe concentrations that increased from 19 ng Fe/mg at E15 up to 63 ng Fe/ml at E19. This lack of iron accumulation in the embryos was correlated with the respective total body weights. At E13, when the embryos did not yet show a significant difference in Fe concentration compared to controls, there also was not a difference in the weight of the embryos. At E15, E17, and E19, however, there was a persistent lack of weight gain in the iron deficient embryos compared to controls, rendering ID fetuses nearly 40% lighter than control fetuses at E15. By E19, the weight increased although ID embryos never reached the same weight as controls ( Figure 2B). In order to determine whether a continuation of the maternal feeding protocol postpartum has a persistent impact on the developing pups, we maintained the dams on the diet they received throughout gestation and tested the iron status of the pups one week after birth. As expected, the iron status of the pups continued to deteriorate and after one week serum iron concentrations were reduced by 63% and hematocrit levels showed a 40% decrease ( Figure 2C, D). While the reduction of iron concentrations in whole embryos and in the serum of Figure 1. Iron restriction two weeks before conception induces maternal iron deficiency without severe anemia. A. Hematocrit levels of pregnant rats exposed to an iron-deficient diet two weeks prior to conception (ID -2wks) (red bars) were evaluated at gestational day E15, E17, E19, and E21 and compared to time matched pregnant rats that received a control diet (black bars). CTL pregnant dams: n = 6 for each gestational age; ID -2wks pregnant dams: n = 6 for each gestational age; expressed values are mean 6 SEM. B. The serum iron concentrations in iron-deficient diet treated animals (ID -2wks) and control animals (CTL) were determined via atomic absorption spectroscopy (AAS) at E15, E17, E19, and E21 (CTL pregnant rats: n = 3 for each time point; ID -2wks pregnant rats: n = 3 for each time point; Mean 6 SEM, * p,0.001). While there were no significant changes in the hematocrit values between the two groups, iron restricted dams showed significantly lower serum iron concentrations compared to control dams, confirming their iron deficiency (ID) status. doi:10.1371/journal.pone.0017483.g001 postnatal animals suggested a change in overall iron homeostasis, our functional readout was focused on the impact of iron deficiency on brain development. We were, therefore, particularly interested in the status of brain iron levels in postnatal animals. AAS analysis of two central nervous system (CNS) regions that are particularly easy to harvest and allow for a good volumetric normalization (spinal cord and cerebellum) confirmed that there was a significant decrease in the concentration of iron in these tissues ( Figure 2E). The decrease in iron concentration was more pronounced in the cerebellum, which had a much higher base-line level than the spinal cord. Embryonic iron deficiency leads to changes in auditory nerve conduction velocity Having shown that the early maternal ID severely impairs the iron homeostasis and weight of the offspring, we next asked whether this gestational insult also results in impaired CNS development in the offspring. Studies in humans, monkeys, and rodent animal models have suggested functional impairments in neural signaling as measured by impaired conduction velocity in the auditory system. This measurement of nerve conductivity is particularly useful as it allows for a unified and non-invasive analysis of a stereotypical neural response profile that results in a wave diagram with each wave representing the neural activity along the ascending pathway (see for review [54,55,56,57]). Specific ABR wave parameters like interpeak latencies between P2 and P1 are thought to be a reflection of the myelination status [58,59], while changes in P3 and P4 interpeak latencies are also suggestive of impairments in synapse maturation [60,61,62,63]. Measurements of Distortion-Product of Otoacoustic Emissions (DPOAE) can be used to exclude impairments that are due to a peripheral loss of hair cell Figure 2. Maternal iron restriction prior to conception is associated with embryonic iron deficiency, postnatal anemia, and reduced postnatal central nervous system iron levels. The ID -2wks and CTL pregnant dams were sacrificed at different time points during gestation (E13, E15, E17, and E19) and embryos were characterized in respect to total iron content by AAS analysis. A. The nutritional ID -2wks regimen generated embryonic ID as early as E15 (Mean 6 SEM, * p,0.001). B. The embryos of ID -2wks dams were significantly smaller than CTL embryos at all gestational ages after E13. (For A and B: at E13, n = 12 CTL embryos and n = 10 ID-2wks embryos; at E15, n = 6 CTL embryos and n = 6 ID -2wks embryos; at E17, n = 11 CTL embryos and n = 13 ID -2wks embryos; at E19, n = 6 CTL embryos and n = 6 ID -2wks embryos; for all gestational ages, each CTL and ID -2wks group was composed of embryos from 3 separate litters; Mean 6 SEM, * p,0.05). C. Hematocrit measurements in ID -2wks pups (red bars) revealed values significantly lower than in CTL pups (black bars) at postnatal day 7 (P7) (Mean 6 SEM, * p,0.0001). D. Serum iron concentrations measured at P7 were significantly decreased in blood samples collected from ID -2wks rat pups (red bars) compared to CTL pups (black bars)(For C and D: CTL and ID-2wks each had n = 8 pups from 3 separate litters, Mean 6 SEM, * p,0.0001). E. AAS at postnatal day 45 showed reduced iron level in ID spinal cord and cerebellum compared to CTL samples (spinal cord: n = 8 CTL and n = 5 ID -2wks samples; cerebellum: n = 5 CTL and n = 5 ID -2wks samples; each CTL and ID -2wks group contained tissue samples harvested from 3 separate litters, Mean +/2 SEM, * p,0.05). doi:10.1371/journal.pone.0017483.g002 function that would also affect the general neural output [64,65,66,67]. To determine the impact of the early maternal ID on the CNS development of the offspring, we conducted an ABR analysis of the ID and normal offspring at P40-45 days, a time point at which the auditory system is fully developed. As shown in Figure 3A, ABR latency analysis revealed significantly increased P2-P1 interpeak latencies in ID offspring compared to CTLs at all frequencies examined, ranging from 0.2560.18 ms to 0.496 0.034 ms (* p,0.0001). When we compared the interpeak latency values of the individual animals in the ID group to control animals, it was apparent that iron deficient animals showed a more pronounced variability across all animals compared to controls ( Figure 3B). We also analyzed the interpeak latency of P4-P1 to determine whether the impairment affects other aspects of auditory information processes detectable by ABR analysis. As shown in Figure 3C, interpeak latencies P4-P1 were also significantly increased in the ID group. In order to exclude possible defects of inner and outer hair cells that would influence the ABR measurements, we also conducted a DPOAE assessment. As shown in Figure 3D, DPOAE amplitude values of ID group were indistinguishable from the control group at all frequencies which strongly suggested normal hair cell function. Wave morphology was not different compared to CTL and hearing thresholds were not elevated in ID animals (data not shown). Taken together, our data show that animals exposed to ID during gestation develop an IDA that is associated with a significant defect in neural function as determined by ABR latency analysis. Delayed onset of gestational iron restriction does not prevent the development of IDA in the offspring The dietary regimen used so far in our study (ID -2wks) was selected to model a pre-pregnancy state that is defined by Figure 3. Auditory Brainstem Responses (ABR) measurement revealed decreased conduction velocities in the offspring born to iron deficient dams. Female Sprague-Dawley rats were provided with the iron-deficient diet two weeks before conception and remained on this diet throughout pregnancy and lactation. Rat pups born to these dams were maintained on iron-deficient diets until testing at postnatal day 45 (P45). Cochlear function was also assessed for all animals (CTL and ID -2wks groups) at ,P45. A. The ABR (P2-P1) interpeak latencies recorded from ID -2wks (red bars) animals showed a highly significant increase compared to CTL (black bars) at all frequencies tested. Auditory nerve function was analyzed by measuring the Auditory Brainstem Responses (ABR) to tone pips at 5 log-spaced frequencies from 5.6 to 24.4 kHz (at 70 dB sound pressure level-SPL). ABR peaks and troughs were identified using a semi-automated tool and verified by a trained observer. Latencies were calculated from onset of stimulus and the (P2-P1) interpeak latency was computed; Mean 6 SEM, p,0.0001). B. The ABR interpeak latencies displayed increased variability in the ID -2wks animals (red circles) compared to CTL (open circles) and the mean value was increased in iron deficient animals. C. The ABR (P4-P1) interpeak latencies recorded from ID -2wks (red hatched bars) animals were significantly prolonged compared to CTL (black bars) at all frequencies tested (Mean 6 SEM, p,0.0001). D. The ID dietary regimen was not associated with outer hair cell damage or cochlear dysfunction as revealed by Distortion-Product of Otoacoustic Emissions (DPOAE) analysis. There was no significant difference between control (black bars) and iron-deficient rats (red bars) in DPOAE amplitudes. (For A-D: n = 35 CTL rats from 5 separate litters and n = 15 ID -2wks rats from 3 separate litters; Mean 6 SEM). doi:10.1371/journal.pone.0017483.g003 suboptimal tissue iron concentrations of the dam prior to conception that does not result in a stage of severe maternal IDA by the time the pups are born. While this dietary regimen led to neural impairment in the offspring, the development of IDA in the offspring, lack of weight gain, and lower core body temperature were all factors that could have contributed to the ABR latency defect. In order to begin to better define which of these factors might play a defining role to the development of neural impairment in the offspring, we changed the onset of the timing of the iron restriction along specific gestational windows. As outlined in Figure 4, we generated three additional groups of rats in which the iron-deficient diet (2-6 mg Fe/g) was introduced at the beginning of pregnancy (ID E0), and at the beginning of the 2nd (ID E7) or the 3rd (ID E14) trimester. All experimental groups remained on the irondeficient diet during lactation and after weaning until the pups reached postnatal day 45, when the animals were subjected to hematological analysis along with weight and temperature measurements followed by an ABR measurement as outlined before. Control group (CTL) consisted of age-matched rats whose dams were maintained on a control diet which contains 240 mg Fe/g and is identical to the iron deficient diet in all other ingredients. As shown in Figure 5A, the hematological analysis revealed that the offspring from all the dietary groups had established severe anemia by the time of the ABR measurement at P45. All groups showed a significant reduction in hematocrit values, hemoglobin concentrations, red blood cell counts (RBC), mean corpuscular volumes (MCV) and serum iron concentrations relative to the CTL group. Other parameters that could impact neural function were also affected but to various degrees depending on the time when the ID diet was introduced to the pregnant dam. As shown in Figure 5B, all iron deficient groups showed a significant decrease in body weight, while only the ID-2wks offspring presented with a significant decrease in core body temperature compared to controls. Core body temperatures were not affected when the iron restriction started at embryonic day 0 (ID E0), day 7 (ID E7) or day 14 (ID E14) and were comparable to control values ( Figure 5C). The establishment of a neural impairment in the offspring is dependent on the onset of the iron restriction during pregnancy defining a window of vulnerability To monitor whether the delayed onset of the maternal iron restriction changes the functional neural impairments in the offspring we had detected in the ID-2wks dietary group, we performed ABR testing of the P45 offspring as outlined before. In all three groups DPOAE analysis revealed again that outer hair cell function was comparable to control animals (data not shown). ABR tests in the ID E0 and ID E7 offspring showed prolonged P2-P1 interpeak latencies ranging from 0.1760.06 ms up to 0.3960.05 ms compared to CTL values. ABR latency responses at the highest frequency (24 kHz) did not reach statistically significant levels in the ID E7 group ( Figure 6A and 6B). There was also a decreased severity of the ABR defects with a later occurring insult. Surprisingly, when pregnant females were exposed to iron restriction at the beginning of the third trimester (ID E14), and the offspring was subjected to ABR testing, we did not find any statistically significant interpeak latency differences in the ID E14 group compared to the control group across all frequencies ( Figure 6C). We also analyzed the P4-P1 interpeak latency differences to determine whether the impairment affects the entire brainstem auditory response in such a manner that changes observed at later peaks are even greater in magnitude than those detected through analysis of the P2-P1 interpeak latencies. As shown in Figures 6 and 7, ID animals showed a relatively larger increase than CTL when comparing the P4-P1 latencies (0.25-0.9 ms) to the P2-P1 latencies (0.5 ms -.1.5 ms). While all P4-P1 latencies, except one, were increased by over 1 ms in the ID E0 group ( Figure 7A), the increase in latency was less severe in the ID E7 group. Nonetheless, even in this group, the alterations seen in the P4-P1 analysis were greater than in the P2-P1 analysis and also were significantly increased compared to control animals at all frequencies tested ( Figure 7B). In contrast, and consistent with the measurements of P2-P1 changes, the ID E14 animals did not show a significant difference in P4-P1 interpeak latencies compared to the control group ( Figure 7C). Taken together, our data show that dietary regimens that are not severe enough to cause a transition of maternal ID to severe IDA To determine the restricted developmental window in which maternal iron restriction impacts fetal development, we established one control (CTL) group and four experimental groups of iron deficient animals: CTL group: animals received control iron diet (240 mg Fe/g food) throughout the experiment; ID-2wks group: animals began the iron-deficient diet (2-6 mg Fe/g food) two weeks before conception; ID E0 group: the iron-deficient diet was introduced the day of plug detection (E0); ID E7 group: animals were placed on the iron-deficient diet on embryonic day 7 (E7); ID E14 group: animals were fed the iron-deficient diet starting on embryonic day 14 (E14); All pregnant Sprague-Dawley female rats in different diet regimens groups were maintained on the respective diet from the assigned start point throughout pregnancy and lactation. After weaning, the pups were maintained on iron-deficient diet until the day of auditory testing (postnatal day 45). doi:10.1371/journal.pone.0017483.g004 still disrupt fetal iron homeostasis as early as E15 and result in severe IDA in the offspring. We provide evidence showing that maternal iron restriction that occurs prior to the onset of pregnancy and during the first trimester has the most severe impact on the neural development of the offspring. Maternal ID during the second trimester still negatively affects the neural conduction velocity (measured by ABR) of the offspring but to a lesser extent. Our data further demonstrate that the development of neural impairments in the offspring is not dictated by the presence of IDA, low serum iron concentration or low birth weight, but is determined by the time window during which the dam received the iron-deficient diet. Discussion The purpose of this study was to determine the impact of dietary maternal iron deficiency (ID) on fetal development, with the main focus on the identification of critical periods of gestation when the developing CNS is most vulnerable to maternal ID. The animal model we established was intended to mimic the extremely prevalent human condition of marginal maternal ID that occurs in an estimated 30 to 50 percent of pregnancies worldwide. After careful titration of the iron concentration in the diet, we were able to establish a model where the pregnant dam was rendered iron deficient but did not develop severe anemia during pregnancy. The analysis of the most prolonged form of diet restriction, in which iron levels were already reduced prior to the onset of pregnancy led to some unexpected findings. For example, we found that despite the fact that the dams received the irondeficient diet not only prior to conception but during the entire pregnancy, the dams did not develop severe anemia that had an impact on hematocrit levels, which remained in the normal range. However, maternal serum iron concentrations decreased contin- Figure 4). A. The table summarizes the results of the hematological tests. All groups of experimental iron deficient animals (ID -2wks, ID E0, ID E7, and ID E14) presented with significantly decreased hematocrit (Ht) values, significantly lower hemoglobin (Hb) levels, profoundly decreased red blood cells (RBC) counts and markedly reduced mean corpuscular volume (MVC) compared to controls, demonstrating the presence of anemia in dietary iron restricted rats. Serum iron concentrations in iron-deficient diet treated animals were significantly lower compared to control animals, confirming the anemic condition of experimental animals. (Ht measurement: n = 11 CTL rats, n = 9 ID -2wks rats, n = 11 ID E0 rats, n = 9 ID E7 rats, and n = 9 ID E14 rats; Hb, RBC, and MCV determination: n = 6 CTL rats, n = 4 ID -2wks rats, n = 4 ID E0 rats, n = 4 ID E7 rats, and n = 6 ID E14 rats; Mean 6 SEM, * p,0.0001; serum iron analysis: n = 5 rats for all groups, Mean 6 SEM, * p,0.01; each CTL and experimental group included rats from 3 separate litters). B. Experimental iron deficient animals exhibited marked growth retardation compared with controls. (n = 35 CTL rats, n = 15 ID -2wks rats, n = 14 ID E0 rats, n = 27 ID E7 rats, and n = 29 ID E14 rats; all CTL and experimental groups contained rat pups from at least 3 separate litters; Mean 6 SEM, * p,0.0001). C. Experimental group ID -2wks (n = 18 rats from 3 separate litters) offspring exhibit significantly reduced core body temperature in comparison to CTL animals (n = 17 rats from 3 different litters; * p,0.05). These changes are not seen in the other three experimental groups (n = 13 ID E0 rats, n = 13 ID E7 rats, and n = 14 ID E14 rats). All groups are composed of animals from 3 separate litters. doi:10.1371/journal.pone.0017483.g005 uously throughout pregnancy, confirming iron depletion. Upon more detailed examination of the maternal blood via HESKA Hema True Veterinary Hematology System, we observed minimal reductions in the measurement of mean corpuscular volume and hemoglobin compared to control values. Taken together, these results are consistent with ID and the absence of severe anemia. This dietary model thus mimics a situation that would not cause any level of clinical concern in the human population, as the absence of severe anemia would not prompt any immediate intervention and is likely to remain unnoticed. We believe that this is an important aspect of our study as it underscores the need for monitoring the status of iron beyond the level of anemia. Iron restriction to the dam two weeks prior to conception was associated with lower iron concentrations in embryonic tissues as early as embryonic day 15 (E15). The embryos maintained a very similar low iron level throughout gestation without an appreciable increase in serum iron concentrations suggesting that maternal iron stores, that apparently were not severely depleted, did not compensate for the continuously increased demand for iron by the embryo. Our observation seems to argue against the often proposed notion stating that the developing embryo is partly protected from the maternal ID through compensatory changes in the iron transport mechanisms of the placenta, which minimize the impact of ID in the fetus [68,69,70,71]. However, our data is consistent with (i) recently published data, suggesting that gestational ID might have a restrictive effect on the constitution of adequate fetal iron stores, [72,73] and with (ii) the observation by Naghii et al., indicating that fetal brain tissue can be iron depleted even in the absence of anemia in the pregnant mother [74]. Furthermore, studies from the Georgieff laboratory have shown that a transgenic model of hippocampus-specific iron deficiency without any signs of anemia also resulted in altered neuronal development as well as significant cognitive dysfunctions (reduced spatial recognition memory performance and procedural memory) [75,76,77]. Wu et al. have also shown that a short period of perinatal iron deficiency results in altered associative behavior . ABR peaks and troughs were identified using a semi-automated tool and verified by a trained observer. Latencies were calculated from onset of stimulus and the (P2-P1) interpeak latency was computed. The rats were sedated with an intraperitoneal injection of ketamine (80 mg/kg) and acepromazine (1 mg/kg). A. The offspring born to dams exposed to ID diet from the time of mating, ID E0 group (yellow bars), displayed significantly prolonged ABR (P2-P1) interwave latencies at all frequencies examined compared to controls (CTL, black bars) (n = 35 CTL rats, n = 14 ID E0 rats; Mean 6 SEM, p,0.05). B. The offspring of dams provided with the iron-deficient diet in the second trimester, ID E7 (blue bars), exhibited significantly prolonged (P2-P1) interpeak latencies at a subset of frequencies tested compared to CTL (black bars) (n = 35 CTL rats; n = 27 ID E7 rats, p,0.05). C. The pups born to dams provided with the iron-deficient diet in the third trimester of pregnancy, ID E14 (green bars, n = 29), had normal ABR (P2-P1) interpeak latencies compared to CTL (black bars, n = 35). The control and experimental dietary groups were composed of rat pups from 3 separate litters. doi:10.1371/journal.pone.0017483.g006 in adult rats without anemia [78]. These studies support our notion that anemia is not necessarily a defining predictor of neural impairments that are associated with iron deficiency. Our model of gestational ID was not only associated with persistently lower body weights as early as E15 but was also associated with functional impairments as determined by the Auditory Brainstem Responses (ABR) analysis of neural conduction velocity in the offspring. The increase in P2-P1 interpeak latencies without hearing threshold changes suggested that, at least in part, impaired myelination could be one explanation for our findings. While disruptions in myelination have been demonstrated in rodents with prenatal and lactational ID [43,45,79,80,81,82,83,84] and hypomyelination is also recognized as a defining feature associated to perinatal ID in prematurely born infants [44], the greater increase in latency changes seen also by analyzing P4-P1 interpeak latencies (which reflects the entire brainstem) is consistent with findings in the literature suggesting that neuronal components could also be affected by iron deficiency. As we established a functional criterion for the ''diagnosis'' of neural impairment that could clearly be associated with the maternal dietary intake, it was possible to design a timed dietary restriction study that cannot be performed in humans but is critical for understanding the window of vulnerability. We generated maternal dietary groups that were defined by the onset of iron restriction solely based on food source. Using the ABR test as a functional readout of the neural CNS integrity of the offspring, we determined that the maternal exposure to an iron-deficient diet either prior to conception, at the start of first trimester, or at the onset of second trimester had a significant negative impact on the ABR latency response in the offspring, placing the window of vulnerability for the fetus in the first two trimesters of gestation. Interestingly, the negative CNS effects on the fetus were more severe when the ID was initiated at the beginning of pregnancy compared to ID initiated prior to the onset of pregnancy. An important result of our research was that the offspring from pregnant rats receiving the ID diet at the beginning of the third Figure 7. Conduction velocity is decreased throughout the brainstem along the auditory pathway in the offspring of the various dietary groups, as reflected by ABR (P4-P1) interpeak latency. Auditory nerve function was analyzed by measuring the auditory brainstem responses (ABR) to tone pips at 5 log-spaced frequencies from 5.6 to 24.4 kHz (at 70 dB SPL). ABR peaks and troughs were identified using a semiautomated tool and verified by a trained observer. Latencies were calculated from onset of stimulus and the (P4-P1) interpeak latency was computed. The rats were sedated with an intraperitoneal injection of ketamine (80 mg/kg) and acepromazine (1 mg/kg). A. The offspring born to dams provided with ID diet from the time of mating, ID E0 group (yellow hatched bars), displayed significantly prolonged ABR (P4-P1) interpeak latencies at all frequencies examined compared to controls (CTL, black bars) (n = 35 CTL rats, n = 14 ID E0 rats,p,0.05). B. The offspring from experimental group ID E7 (blue hatched bars) exhibited significantly prolonged (P4-P1) interpeak latencies at all frequencies tested compared to CTL (black bars) (n = 35 CTL rats; n = 27 ID E7 rats, p,0.05). C. The offspring born to experimental group ID E14 (green hatched bars) had normal ABR (P4-P1) interpeak latencies compared to CTL (black bars) (n = 35 CTL rats; n = 29 ID E14 rats). The control and experimental dietary groups were composed of rat pups from 3 separate litters. doi:10.1371/journal.pone.0017483.g007 trimester exhibited normal auditory nerve conduction velocities as reflected by ABR recordings despite IDA being present in all experimental rats at the time of testing. This result might shed some light on seemingly contradictory results found in humans; some investigators reported no alterations of ABR latencies in ID children with or without anemia [85,86], while others found that IDA in 6 month old babies produced abnormally prolonged ABR latencies that persisted years after the children's iron status was corrected with iron therapy [43,81]. Our study indicates that there is a precise window of vulnerability during which ID can affect the fetal CNS leading to postnatal neural impairments. The lack of defect seen in some studies could be a consequence of the insult occurring outside the window of vulnerability, rather than a perceived general insensitivity of newborns to iron restriction. Without the establishment of the maternal iron status during pregnancy or at least cord blood iron levels, it cannot be resolved whether the initiation of ID and the possible onset of a later apparent pathology happened in utero. Therefore, any conclusions about the importance and/or need of pre-and postnatal iron supplementation have to be made with caution. The window of vulnerability, which we have defined in our studies, also offers an important clue for the underlying cellular impairment and seems to be consistent with our previous work on the study of cellular targets of ID [53]. As illustrated in Figure 8, ID during embryonic CNS development seems to affect neural precursor cell populations leading to an imbalance of specific precursor populations, which might be ultimately responsible for reduced oligodendrocyte numbers, abnormal myelination, and neural pathologies. We have shown previously that the differentiation and proliferation of embryonic glial precursor cells can be modulated by exogenous iron concentrations [87]. Interestingly, the timing of the generation, expansion and differentiation of these precursor cells coincides with the window of vulnerability we have defined in our diet regimens. In the rat embryo, gliogenesis starts around day E13.5 with the generation of embryonic glial restricted precursor cells (GRPs), which shortly before birth give rise to oligodendrocyte precursors (OPCs) that persist postnatally [88]. Even slight alterations of proliferation and/or of differentiation abilities of precursor cells (GRPs and OPCs) occurring before the time of oligodendrocyte generation could translate into a myelination defect. The notion of early embryonic cells being especially responsive to suboptimal iron levels is also supported by studies from Badarocco et al., whose interesting experiments using intracranial apotransferrin injections suggest that iron may affect oligodendrocyte development at early stages of embryogenesis rather than during late development [89]. The defects that are associated with iron deficiency are, however, unlikely to be restricted to the glial population. As shown in Figure 8, neuronal progenitor cells develop in a similar time window and could equally well be targets of iron deficiency. The cognitive impairment seen in the offspring is most likely a combination of an impact on both neuronal and glial embryonic progenitor cell pools. The absence of a functional defect in late onset ID suggests that, if the diet is introduced at embryonic day 14 (red arrow) and maintained, the time point of iron depletion is past the period when the most vulnerable embryonic populations expand and generate their progeny. Under these conditions, the impact of ID on glial cell generation and/or neuronal maturation does not seem to be severe enough to result in detectable auditory dysfunction. It should be pointed out that despite the lack of an ABR latency defect in our study at the late gestational time point (ID E14), we cannot exclude the presence of other abnormalities. For instance, a number of studies in rodent and primate models of late gestational and postnatal ID reported significant and persistent behavioral abnormalities and neurochemical disturbances [19,90,91,92,93,94,95,96,97]. As development occurs along a temporal continuum, it would be expected that deficiencies occurring at different times cause different types of outcomes. The timed maternal iron restriction coupled with a persistently deficient diet used in our model, led to another surprising finding that sheds some new light on the role of anemia in iron deficiency. Due to the associated fatigue, muscle weakness, and dyspnea during exertion, anemia is considered a confounding factor in various behavioral tests and raises criticism on the accuracy of the results and their interpretation. The ABR electrophysiological test we employed in the evaluation of neurodevelopmental impairments induced by ID seems not to be influenced by anemia. As shown in Figure 5A, all offspring of our different feeding groups developed clinically relevant anemia, yet only three out of four groups showed functional impairment in ABR patterns. This observation strongly suggests that anemia is not a determinant factor of pathological ABR patterns recorded in iron deficient animals and allows dissociation of the impairment caused by ID from effects that might be primarily due to anemia. As ABR analysis can be conducted in multiple species, including humans, this type of analysis may provide a particularly useful noninvasive tool for examining the effects of micronutrient deficiencies on CNS development. While IDA was not required to cause ABR defects, the occurrence of IDA did, however, severely impact the survival rate of the offspring. The litter survival rate was approximately 75% in the four experimental groups compared to the control groups. This decrease in survival rate may lead to an unintentional sample population bias of ''survivors'' at P45 when the functional measurements are conducted. While we cannot completely rule out that the surviving animals are on some level more robust that their littermates, it is important to note that we cannot correlate survival with absence or presence of ABR defects. The ID -2wk, ID E0 and ID E7 groups had as many animals that died during the experiments as the ID E14 group yet none of the surviving animals in the last group showed an ABR defect, while all the animals in the other three groups showed an ABR defect of various severities. Another novel observation in our study is the time-specific presence of thermoregulatory deficits in the offspring from mothers who were iron restricted two weeks prior to conception. The three other feeding groups did not develop such defects. As ID has sometimes been linked to poor thyroid function [98], which in turn can result in a thermoregulatory dysfunction [99,100,101,102,103], the decreased core body temperature in the most severe feeding regimen (ID -2wks) might involve a thyroid component. However, based on our data, we do not think that thyroid involvement is critical for the development of the ABR dysfunction, as the ID E0 and the ID E7 groups have significant ABR dysfunctions but normal core body temperatures. In summary, the analysis of developmental defects that are associated with exposure to iron restriction during different phases of fetal development emphasizes the influence of the precise window of vulnerability and the associated duration and magnitude of ID that is exerted on the developing fetal CNS. The prolonged absolute and interpeak latencies revealed by the comprehensive ABR analysis in ID animals could have important clinical implications. If ID at an early age causes hypomyelination or delayed neural maturation along the auditory pathway altering the properties of the auditory message and its transmission through the brainstem structures, it is possible that such a dysfunction could have functional significance for language acquisition and the development of other higher cognitive and emotional functions. In lieu of this information, our data not only reiterates the importance of iron supplementation as part of prenatal and gestational IDA therapeutic regimen but also as a preventive strategy. Identifying the windows of vulnerability when specifically vulnerable cell populations are generated during embryonic development will be critical in delineating the restricted window of therapeutic opportunity when iron supplementation could correct or prevent the functional deficit. Animals and dietary treatment Adult female and male Sprague-Dawley rats (10-12 weeks of age) were purchased from Charles River Laboratories (Wilmington, MA). All animals were housed individually under controlled environmental conditions (0600-1800 h light cycle, 23uC and 32% humidity) and were provided free access to food and water. For the time mated pregnancies, the female/male pair was housed in the same cage for 24-48 hours or until the detection of vaginal plug. There are no data to support a notion of decreased fertility in any of the iron deficient (ID) diet treated groups. The litter size was comparable to controls and the newborn pups appeared healthy although their overall size is smaller in the ID -2wk and ID E0 groups. We did, however, observe a decreased survival rate in all iron deficiency groups. By postnatal day 45, on average 10-20% of the offspring died with the most severe losses in the ID -2wk group. The lower survival rate was most likely a result of the severe anemia experienced by the offspring. Determination of gestational age For the time-mated pregnancies, the first day of gestation was determined by the appearance of vaginal plug. For exact timing of the gestational age, the crown-to-rump length (CRL) of the embryo was determined. The measurement of CRL, using a standard millimeter scale, was taken from the vertex to the rump with care taken to avoid pressure that would distort the natural curvature of the embryos. Embryonic size based on CRL in addition to vaginal plug detection allowed for a more accurate evaluation of gestational age. Experimental design (dietary regimens) The feeding protocols used in our study are depicted in Figure 4. The experimental animals were divided in four groups of specific dietary regimens, using the same iron-deficient diet containing 2-6 mg Fe/g of food (rodent TD. 80396 ID Diet -Harlan Teklad Custom Research Diets). The first diet regimen was designed to generate embryonic ID in early pregnancy. The dams were started on the iron-deficient diet two weeks before conception and the pups born to these dams represented the experimental group ''ID -2wks''. The next group of experimental animals was fed the irondeficient diet on the day of conception or appearance of vaginal plug, designated ''ID E0''. The third experimental group of animals, designated ''ID E7'', comprised of rat pups born to dams exposed to the iron-deficient diet from the seventh gestational day onwards. The rat pups born to dams feeding on the iron-deficient diet from gestational day 14 onwards were assigned to the fourth dietary group of rats, labeled ''ID E14''. As our control (CTL), we used age-matched rats born to dams that were maintained on control iron diet throughout the duration of the experiment. The control diet contained about 240 mg Fe/g of food (Adjusted Iron Diet (240) -TD.05656 from Harlan Teklad Custom Research Diets). All other components of the diet, composition of nutrients and caloric values were identical to the ID diet except the concentration of iron. All control and experimental animals were given deionized-distilled water through glass sipper tubes. All groups received the assigned dietary treatment (irondeficient and control diets) during lactation and after weaning until postnatal day 45, when the animals were subjected to ABR and DPOAE testing as well as evaluation of body weight, core body temperature, and iron status. Animal procedures were approved by the University Committee of Animal Resources at the University of Rochester, New York. Hematocrit and serum iron level measurement The animals were anesthetized with ketamine (100 mg/kg i.p.) and blood samples for hematocrit measurement were collected in heparinized microcapillary tubes (VWR, West Chester, PA) from the right cardiac atrium before the rats were perfused. Tubes were centrifuged for 15 minutes at room temperature in a high speed hematocrit centrifuge (IEC MB Microhematocrit Centrifuge). The Hematocrit value was calculated as the length of red blood cell column per total blood sample column length and expressed as a percentage. For a more detailed hematological analysis including hematocrit (Ht) values, hemoglobin (Hb) concentrations, red blood cell (RBC) counts, and mean corpuscular volumes (MCV), we analyzed the fresh blood samples, collected in EDTA-coated tubes, using HESKA Hema True Veterinary Hematology System. Whole blood aliquots for serum iron determination were allowed to clot, and then were centrifuged at 10,000 x g at 4uC for 15 minutes to separate cells from sera. Serum samples were isolated then frozen and stored at 280uC until analysis for iron concentrations by atomic absorption spectrophotometry (AAS). Tissue iron measurement using Atomic Absorption Spectroscopy (AAS) After anesthesia with ketamine (100 mg/kg i.p.) and blood collection for hematocrit and serum iron determination, the animals were transcardially perfused with11.2 mg/ml heparin in phosphate-buffered saline via the left ventricle to remove blood from organs. The embryos or CNS tissues were immediately dissected, washed in distilled deionized water, vacuum dried, weighed, and stored in iron-free teflon vials at 280uC until analyzed as previously described. Frozen embryonic and postnatal CNS tissues were wet-digested in ultra-pure nitric acid and analyzed for iron concentration using the graphite furnace atomic absorption spectrophotometer. Standards, blanks, and standard curves were used as previously specified [104,105,106]. Core body temperature measurement All experimental ID and control rats were subjected to core body temperature measurement by using a RET-2 rectal probe and TH-5 Thermalert Monitoring Thermometer (Physitemp). The core body temperature was consistently recorded for all groups of animals at six o'clock in the evening of the day of auditory testing. Auditory Brainstem Responses (ABR) recording and analysis Control (CTL) and iron deficient (ID) dietary groups of rats (ID -2wks, ID E0, ID E7, and ID E14 groups) were ABR tested on postnatal day 40-45. For the test, rats were anesthetized with acepromazine (1 mg/kg i.p.) and ketamine (80-100 mg/kg i.p.). Needle electrodes were inserted at vertex and pinna, with a ground near the tail. ABR potentials were evoked with click tone pips at 5 log-spaced frequencies from 5.6 to 24.4 kHz at 70 dB SPL. The response was amplified (10,000x) and 2,048 responses were averaged with using software and high-frequency speakers from Intelligent Hearing Systems (Smart EP, Miami, FL). To compare the change of latencies, the ABR peaks and troughs were identified by a semi-automated tool and verified by a trained observer by visual inspection of recorded wave forms. Latencies were calculated from the onset of the stimulus, and the interpeak latencies of wave II-I (P2-P1), wave III-I (P3-P1), and wave IV-I (P4-P1) were computed. The change of interpeak latencies was then calculated. Distortion Product of Otoacoustic Emissions (DPOAE) acquisition The function of the outer hair cells in the ear was analyzed by measuring DPOAEs with secondary tones (f2) at 5 log-spaced frequencies from 5.6 to 24.4 kHz. DPOAE testing was performed with f2/f1 = 1.2 and unequal primaries (L1 = 10dB+L2) at 70 dB and 60 dB SPL using high-frequency speakers and software from Intelligent Hearing Systems (SmartOAE) and high-frequency microphones (ER10B+) from Etymotic Research (Elk Grove Village, IL). Cochlear function was assessed for all animals at approximately 40-45 days of age and rats were sedated with ketamine (80-100 mg/kg i.p.) and acepromazine (1 mg/kg i.p.). Statistical analyses ''n'' represents the number of animals in any group analyzed and has a value $ 3. Regardless the sample size (n value), each control and experimental group was composed of animals from at least 3 separate litters. Statistical analyses were performed with the two-tailed Student's t-test and conducted comparing nutritionally ID to control rats, under the same conditions. A p-value less than 0.05 was considered statistically significant. Comparison of mean values of ABR interpeak latencies between different dietary groups and controls was performed with analysis of variance (ANOVA) using a Bonferroni posttest analysis and significance was defined as a p value less than 0.05.	2014-10-01T00:00:00.000Z	2011-03-15T00:00:00.000	{ "year": 2011, "sha1": "0997e91820ac1c13df6d435400915b03971df7ba", "oa_license": "CCBY", "oa_url": "https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0017483&type=printable", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "0997e91820ac1c13df6d435400915b03971df7ba", "s2fieldsofstudy": [ "Biology", "Medicine" ], "extfieldsofstudy": [ "Biology", "Medicine" ] }
237393180	pes2o/s2orc	v3-fos-license	Infants Younger Than 90 Days Admitted for Late-Onset Sepsis Display a Reduced Abundance of Regulatory T Cells Objective To provide epidemiological data of infants < 90 days of age with suspected late-onset sepsis (LOS) and evaluate distinct immunological specificities. We hypothesized that previously healthy infants < 3 months of age with sepsis have a yet undefined immunological predisposition; e.g. differences in lymphocyte subsets including regulatory T cells. Methods We performed an exploratory, single center study between January 1st, 2019 and June 1st, 2021. Routine diagnostics included conventional culture (blood, cerebrospinal fluid, urine), PCR and inflammatory markers in infants < 90 days of age with suspected sepsis. We additionally analyzed lymphocyte subsets and CD4+ CD25+ forkhead box protein (FoxP3)+ Tregs at admission for sepsis workup as compared to age-matched controls. Results A convenience sample cohort of n= 51 infants with sepsis workup was enrolled. Invasive bacterial infection (IBI) was diagnosed in 25 (49.0%) patients including two infants with a rhinovirus co-infection and viral infection in 14 (27.5%) neonates. No infectious cause was found in 12 cases. Infants with suspected LOS displayed a decreased abundance of CD4+ FoxP3+ T cells as compared to controls, which was most pronounced in the subgroup of infants with IBI. We also noticed elevated HLA-DR-positive CD3+ cells in infants with LOS and a higher CD4/CD8-ratio in infants with viral infection as compared to healthy controls. Infants with viral infections had a higher number of natural killer cells as compared to infants with IBI. Conclusion Our exploratory data support the concept of a potential immaturity state and failed immune tolerance development for young infants with LOS. Future large-scale studies are needed to elucidate pre-sepsis conditions and to target the microbiome-immunity interplay as a potential risk pattern. INTRODUCTION Episodes with suspected infections are highly relevant in early infancy, particularly in the first three months of life (1). Lateonset sepsis (LOS) in infants < 90 days of age presents with nonspecific clinical signs such as fever. Diagnostic approaches are imprecise due to low sensitivity of cultures and limited specificity for laboratory markers, e.g. C-reactive protein or procalcitonin (2). As infections in young infants often have a fulminant course, antibiotic treatment is consequently initiated as soon as clinical suspicion is evident (3)(4)(5). However, diagnostic markers or signatures to clearly differentiate between invasive bacterial infection (IBI), viral infection or no infection are lacking. Molecular biology tools (e.g. multiplex pathogen-PCR; nextgeneration sequencing) are cost-intensive and not available as specific bedside tests to guide a medical decision for antibiotic treatment (5). There are previous studies indicating that 7-13% of infants < 90 days of age have invasive bacterial infection (IBI; including urinary tract infection, UTI; bacteremia and/or meningitis) (3,6) but often causes of fever remain unknown. The diagnosis of a viral infection may help to discontinue antibiotic therapies and herewith reduce the total exposure in these infants (7). From an immunological perspective, the group of infants admitted with suspected LOS < 90 days of age has largely been neglected for comprehensive research approaches. We herein address the hypothesis of a potential immaturity state and failed immune tolerance development for young infants with suspected LOS. Historically, neonatal T cellswhen compared to those from adults -were considered immature, less functional or unresponsive against specific antigens. This idea has been recently replaced by a concept including the important role of regulatory T cells in the immunological adaptation to the rapidly changing environment in the first months of life (8,9). Neonatal T cells, in particular CD4+ and CD8+ cells, combine functionality in protecting from foreign pathogen as well as sustaining tolerance to self-antigens. Regulatory T cells (Tregs) represent a subpopulation of CD4+ T cells with an ability to limit the immune response against self-and non-selfantigens. This temporary feature seems to be crucial for the ontogenetic control of immune activation and feto-maternal tolerance. An increased abundance of Tregs in preterm infants, however, correlates with sepsis risk, possibly through their immunosuppressive capacity (10). In term infants, Tregs have been shown to be less abundant in peripheral blood during RSV infection (11,12). Further, Tregs are reduced in infants prone to allergies (13) whereas in adults Tregs are upregulated in the course of sepsis which correlates to mortality risk (14). B cells inherit an important role in adult sepsis -patients' immunological responses in an antibody-dependent and antibody-independent manner (15). In neonates, B cells lack antigenic exposure and have limited expression of B cell receptors resulting in low levels of primary IgG responses to immunizations and infections (16). However, the role of B cells and NK cells in the neonatal response to sepsis is incompletely understood and mainly studied in preterm infant cohorts (17,18). In animal models, activation of NK cells has been shown to alter neonatal sepsis outcomes, hence they are proposed as candidate for immunomodulatory prophylactic treatment (19). Reduced abundances of NK cells at birth have proved to be associated with increased risk for LOS in human term and preterm neonates (18). The distinct immune responses of the newborn mediate the transition from an intrauterine environment to an extrauterine life. Age-dependent susceptibility to infection remains an important threat in young infancy and reflects developmental trajectories in cell autonomous immunity (CAI), nutritional immunity and metabolic pathways such as serum-iron concentrations, postnatal development of physiological barriers and production of antimicrobial proteins and peptides (APP) (9). The scope of the present study was to delineate whether the group of infants < 90 days with an uneventful postnatal period and suspected LOS carries certain immunological characteristics that are related to lymphocyte subsets or Treg frequencies. MATERIALS AND METHODS Between January 1 st , 2019 and June 1 st , 2021, we performed a convenience sample, single-centre observational study (Fever Without Source study; FWS study) with infants that were admitted from the community for suspected sepsis with an uneventful previous history for infections. The inclusion criteria were age < 90 days and > 72 hours and suspected lateonset sepsis with sepsis workup and empirical antibiotic treatment. Afebrile infants admitted to hospital for noninfectious reasons (i.e. with elective surgery) in the first 3 months of life served as controls. After written informed consent was given by the parents, infants with suspected LOS and controls were enrolled by the attending physicians. Besides routine sepsis diagnostics ("sepsis work-up": biomarkers of inflammation, conventional cultures of blood, urine and cerebrospinal fluid), we performed extended viral diagnostics including an in-house multiplex PCR from nasopharyngeal aspirates (NPA) and enterovirus in-house PCR from stool samples at the local Institute for Clinical Chemistry. In infants with suspected meningitis the enterovirus PCR from stool samples was performed by the institute for hygiene and environment in Hamburg (surveillance program for enterovirus meningitis/ encephalitis). The multiplex-assay from NPA (RP2Plus Biofire ® respiratory 2.1 plus panel) included adenovirus, human rhino-/ enterovirus, respiratory syncytial virus (RSV), influenza virus A and B, parainfluenza virus 1-4, metapneumovirus, coronavirus V (4 subtypes: HKU1, NL63, 229E, OC43), MERS CoV, Bordetella pertussis and parapertussis, chlamydia pneumoniae, mycoplasma pneumoniae. Stool samples were tested for rotavirus, adenovirus, and norovirus using commercially available antigen tests (Ridascreen, r-Biopharm, Darmstadt, Germany), when infants developed symptoms of gastroenteritis. The infants' clinical and demographic characteristics, laboratory markers, and clinical presentation were immediately documented in a written case report form. Following pseudonymization, study personnel transferred the data into a prespecified Microsoft Excel database (Microsoft Office 2010, Versions 14.0). Data of all infants were monitored by two researchers (IF and MTD) using the original patient files. The diagnostic results were not blinded to the clinicians in the unit. Ethics All study parts were approved by the local committee on research in human subjects at the University of Lübeck (Reference number: 20-228). Blood samples were obtained exclusively within a medically required blood withdrawal at admission and before any treatment with anti-infective agents. The additional blood volume (< 1% of body blood volume per blood sampling) was in line with current guidelines of the European Medical Agency (EMA) on the investigation of medicinal products in term and preterm infants; Committee for Medicinal Products for Human Use and Pediatric Committee (PDCO, 2009). Definitions Gestational age was calculated from the best obstetric estimate based on early prenatal ultrasound and obstetric examination. Lateonset sepsis (LOS) was defined as sepsis occurring after the first 72 h of life. Clinical LOS was defined as pediatricians decision to treat the infant with antibiotics for at least 5 days due to the following reasons: two clinical signs of systemic inflammatory response: temperature > 38.0°C or < 36.5°C, tachycardia > 200/min, newonset or increased frequency of bradycardias or apneas, hyperglycemia > 140 mg/dl, base excess < 210 mval/l, changed skin color and increased oxygen need; and one laboratory sign: Creactive protein > 10 mg/l, platelet count < 100/nl, immature/total neutrophil ratio>0.2 and white blood cell count<5/nl (20). Cultureconfirmed sepsis was defined as clinical sepsis with proof of causative agent in cultures of blood, urine or cerebrospinal fluid. Invasive bacterial infection (IBI) was defined as culture-confirmed sepsis or clinical sepsis. Viral infection was defined as proof of causative virus via multiplex PCR from nasopharyngeal aspirate, enterovirus PCR from stool samples, RSV rapid test, influenza rapid test or PCR test for adenovirus, norovirus and rotavirus. Fever was defined as central body temperature of > 38.0°C. Analysis of White Blood Cell Counts, Lymphocyte Subsets and Tregs After blood withdrawal, the EDTA whole blood samples were stored for a maximum of 24 hours at room temperature before processing (21). As part of the clinical routine laboratory evaluation, white blood cell counts including numbers of lymphocytes were assessed by the central laboratory of the university hospital from EDTA blood (ethylenediamine tetraacetic acid). Further immunological analyses were performed in two different laboratories: First, the central laboratory of the university hospital performed the determination of lymphocyte subsets (CD3+/CD4 +/CD8+/CD19+/[CD16/CD56]+/activated T cells and the CD4/ CD8-ratio) on weekdays, Monday-Friday from 8a.m. to 4p.m. Counting of lymphocyte subsets and their activation status in whole blood was performed on a BD FACSCanto ™ II system (BD Biosciences, San Jose, CA, USA) with BD FACSCanto clinical software using multitest 6-Color TBNK (T cells, B cells and natural killer cells) kits and multitest CD3/CD8/CD38/HLA-DR kits according to the manufacturer's instructions. Cytometer performance was checked weekly using BD FACS 7-Color Setup Beads, while the quality controls BD Multi-Check Control and BD Multi-Check CD4 Low Control were run alternately twice a day. Gating strategies and representative plots are available in the BD user manual BD Multitest ™ 6-color TBNK (22). Second, we performed immunological analyses in the research laboratory of the pediatric department as previously described (10,23). The analyses included the determination of CD3+ lymphocytes, CD4+ lymphocytes, CD3+ CD4+ lymphocytes, CD3+ CD4+ CD25+ lymphocytes, CD3+ CD4+ Foxp3+ lymphocytes and CD3+ CD4+ Foxp3+ CD25+ lymphocytes. A cell viability test was performed on a regular basis to control for dead cells (eBioscience, San Diego, CA, USA). We used flow cytometry to determine the cell population percentages and absolute cell counts (Supplementary Table 1). In order to characterize T cell populations, we performed a whole blood staining with fluorochrome-labelled antibodies using cell permeabilization and fixation reagents (FoxP3/Transcription Factor Staining Buffer Set; eBio-science). Afterwards, cells were stained with surface antibodies (multicolor flow cytometry) specific for CD3/fluorescein isothiocyanate (FITC) (eBioscience), CD4/phycoerythrin (PE) (Miltenyi Biotec, Bergisch Gladbach, Germany), CD25/brilliant violet (BV421; BioLegend, San Diego, CA, USA) followed by intranuclear staining for FoxP3 (eFluor660; eBioscience). FoxP3 staining was conducted according to the manufacturer's protocol. We used fluorescence activated cell sorter (FACS) staining buffer (eBioscience) to dilute the fixed and stained cells, before storage at 4°C. Flow cytometric analysis was performed within a timeframe of 4 days. We used a BD LSR II cytometer, FACS Diva software (BD Bioscience, San Jose, CA, USA) and FlowJo (Tree Star, Ashland, OR, USA; Version 10.7.1) for Treg analyses. Tregs were determined by their position in the forward-/side-scatter plot (size/granularity) and co-expression of CD3, CD4, CD25 and FoxP3 (Supplementary Figure S1). Fluorescence minus one (FMO) controls were used to establish gating boundaries and to identify any background spread of fluorochromes. Statistical Analysis Data analysis was performed using the SPSS data analysis package (Version 26.0; SPSS Inc., Munich, Germany). Differences between groups were evaluated with Fisher's exact, Mann-Whitney U, Kruskal-Wallis-test and Dunn's multiple comparison test. A Pvalue < 0.05 was considered as statistically significant for single tests. RESULTS During the observational period, we screened all 100 infants with suspected LOS < 90 days of life ( Figure 1). We excluded 49 infants for the following reasons: lack of immunological data at hospital admission (n=13), denial of consent (n=18), not approached by attending physician (n=17) and underlying Shwachman-Diamond syndrome with bone marrow failure (n=1). We included n = 51 infants < 90 days of age with fever and suspected LOS with a median gestational age of 39.9 weeks (IQR 37.8 -40.4) weeks, a median birth weight of 3750 grams (IQR 3300g -4305g) grams and a median age of 45.0 days (IQR 19.0 -60.0 days) at hospital admission ( Table 1). The study cohort included 4 former preterm infants > 30 weeks of gestation with uneventful history and 10 infants with postnatal antibiotic treatment due to suspected but unconfirmed early-onset sepsis. The most frequently documented clinical symptoms were fever (body temperature > 38.0°C 94%; > 37.5°C 100%), worsened general condition (78.4%) and agitation (51.0%). Patients were hospitalized for a median of 7.0 days (IQR 4 -12 days). IBI was Table 2), while no bacterial or viral pathogen was found in 12 infants (23.5%). 85.7% of infants with viral infection had elder siblings, while infants with IBI were often exposed to antenatal antibiotics (40%) and to be delivered by Caesarean section (32%, Table 1). Our control group included n = 30 infants < 90 days of age that were admitted to hospital for non-infectious reasons: elective surgery for hernia or foot deformity (n =15), events that required 24 to 72-hour inpatient monitoring such as apparent life-threatening events (ALTE) or suspected but unconfirmed seizure (n = 12), failure to thrive and hyperbilirubinemia (n = 3). All infants included in the control group did not have any clinical or laboratory sign of infection or underlying diseases. The cohort included 2 preterm infants of 35 + 1 and 34 + 3 weeks of gestation. Microbiological Diagnostics Among infants with IBI, n = 5 (20.0%) had Streptococcus agalactiae (GBS) culture-proven LOS. Two of these cases were complicated by meningitis, or endocarditis. Two mothers were tested positive for GBS during pregnancy, one was tested negative and two mothers had an unknown GBS-status. Three infants with GBS sepsis had previous ante-and/or postnatal antibiotic treatment for positive maternal GBS status and/or suspected early-onset sepsis that was not confirmed retrospectively. Urosepsis caused by Escherichia coli was the most frequent cause of IBI, i.e. 15/18 with urosepsis cases. 3 of these infants had a positive blood culture. Other pathogens causing urosepsis were Klebsiella oxytoca (n = 1) and Enterococcus faecalis (n = 2). Seven infants with urosepsis were diagnosed with malformation of the urogenital tract ( Table 2). Enterovirus Infection Is Frequently Diagnosed in Young Infants With Suspected LOS Infection with enterovirus was the most frequent cause of viral infection resulting in sepsis workups (50% of viral infections). Rhinoviruses were found as causative agent in 3 cases; single cases were caused by rotavirus, respiratory syncytial virus (RSV), coronavirus NL 63 and coronavirus NB infection ( Table 2). Notably, there was a significant decline in viral infections as a cause of suspected LOS <90 days in the COVID-19 pandemic time-frame of our observational study ( Figure 1). Admission for Sepsis Decodes Underlying Diseases Interestingly our cohort included 4 infants (7.8%) with suspected or proven major underlying conditions, that were not known before hospital admission such as Trisomy 21, batteredchild syndrome (BCS) and two infants with more than one dysmorphic feature and suspected syndrome. Two of these cases were diagnosed with urosepsis and one had pneumonia as potential infectious complication of the underlying disease. In the case of BCS no infectious cause for the symptomatic infant was found. Infants <90 Days With IBI Display Lower Frequencies of Regulatory T Cells Than Healthy Controls Infants with suspected LOS displayed a decreased abundance of CD4+ FoxP3+ T cells as compared to controls, which was most pronounced in the subgroup of infants with IBI ( Table 3 and Figures 2-4). In order to account for a potential age-related bias, we divided our cohort in three age-defined subgroups: 3 to 30 days, 31 to 60 days and 61 to 90 days ( Figure 5). The trend of lower regulatory T cells in infants with IBI was observed in all age groups with significant levels in infants > 30 days of age. Differences with regard to frequencies of CD3+ or CD4+ cells were not observed. We also noticed elevated HLA-DR-positive CD3+ cells in infants with LOS and a higher CD4/CD8-ratio in infants with viral infection as compared to healthy controls. Infants with viral infections had a higher number of natural killer cells as compared to infants with IBI. No differences were observed in the B-cell and cytotoxic T cell compartment ( Table 4). DISCUSSION In this single-centre prospective exploratory study we provide epidemiological and immunological data of young infants < 90 days of age with suspected and proven LOS. IBI occurred in 49.0% of all infants that were admitted for sepsis workup, whereas 27.5% had proven viral infection. Notably, in about one quarter of the cases (23.5%) no infectious cause was found despite thorough phenotyping and extensive microbiological diagnostics. As a specific immunological hallmark, infants with suspected LOS displayed a decreased abundance of CD4+ FoxP3+ T cells as compared to controls, in particular infants with IBI. It was the main hypothesis of our study that this immunologically unexplored group of infants with suspected community-acquired LOS in the first 90 days of life is characterized by certain immunological specificities. The comprehensive approach of stratification between IBI and viral infection, in this group of patients is crucial for future development of biomarkers that help clinicians' decision making for initiation and duration of anti-infective treatment. Clinical susceptibility may act in concert with specific, potentially malleable immunological predisposition. Our data suggest that Tregs are diminished in community-acquired IBI in young infants which is a novel observation for sepsis patients. Pagel et al. demonstrated higher Treg frequencies in preterm infants with early-onset sepsis (EOS), independent of gestational age (10). Studies with adult sepsis-patients revealed higher abundances of Tregs during the course of sepsis as well as an association with poor prognosis and increased mortality in patients with septic shock (14). In these studies, the depletion of Tregs was reported to decrease mortality (14,24). However, the scientific question remains unsolved, whether altered Treg frequencies display a risk factor for the onset of sepsis (as it has been suggested for preterm infants with EOS (10),) or whether the course of sepsis correlates with dysregulation/dysfunction of T cell subsets including regulatory T cells (as it has been proposed for adult patients with septic shock (14),). Animal and human cord blood studies consistently reported essential differences in Treg-associated function and capacity between term and preterm neonates (25)(26)(27)(28). Since Tregs play a role for a permissive physiological colonization of young infants (29,30), we propose that reduced Treg quantities may represent a dysregulation pattern of the mutual immune-microbiome interaction as hallmark of susceptibility to IBI. Additional risk factors may include higher exposure rate to antenatal antibiotics and Caesarean section as noted in the IBI subgroup of our cohort. The dynamics of GBS late-onset infections despite positive screening and intrapartum prophylaxis in some IBI infants also suggest fluctuations in host immunity that contribute to failure of natural niche occupation of pathobionts (31). On the other hand, infants with viral disease were most often younger siblings which is a known risk factor for household transmission of viruses. Interestingly, we noted a significant effect of extended hygiene measures during the COVID-19pandemic, i.e. a decline in cases of suspected LOS that were eventually diagnosed with viral infection. Decreased quantities of Tregs were found in young infants with RSV infection (12). While we only present a trend towards lower Treg frequencies, Raiden et al. studied immunological characteristics in 36 infants with RSV infection and described a marked reduction of Treg frequencies from 6.5% to 1% but also less dramatic changes depending on the Treg phenotype (11). An increased recruitment from the infectious focus, i.e. the lungs or lung-draining lymph nodes, higher susceptibility to apoptosis or an increased instability of infant Tregs have been proposed as underlying mechanisms for the Treg reduction in peripheral blood (32,33). While CD56 + T cells (NK cells) play a key role in the promotion of severe systemic inflammation in adults (34), we found no remarkable differences in NK cell numbers between subgroups of suspected LOS and controls. Infants with viral infections had a higher number of NK cells as compared to infants with IBI which suggest a specific role of NK cells in antiviral immunity even in young infancy (35). We also noticed elevated HLA-DR-positive CD3+ cells (activated T lymphocytes) in infants with LOS and a higher CD4/CD8-ratio in infants with viral infection as compared to healthy controls. Term neonates are known to have higher counts of CD4+ T cells in peripheral blood than adults (36) which is why CD4/CD8 ratio is usually 3:1, and declines to adult values of 2:1 until the age of four years (37). Higher CD4/CD8 ratios as seen in infants with viral infection in our cohort may also be linked with a potential immaturity state carrying increased infection risk. Strengths and Limitations The major strength of our study is the novel approach of providing detailed microbiological, virological and immunological data of a well phenotyped cohort of infants < 90 days of age with suspected sepsis as compared to healthy controls. The topic is of high clinical relevance given the potential long-term consequences of both, failure to treat IBI appropriately and unnecessary exposure to antibiotics early in life. Our study cohort is unique as previous studies on cellular immunity in early life sepsis investigated neonates who had never been discharged home, in particular preterm infants, or assessed healthy babies with regard to developmental immune trajectories (9,10). There are limitations to our study, i.e. the single center approach, small sample size and lack of pre-sepsis data which would require a large-scale birth cohort study. Our study design was exploratory and based on a convenience sample including differences in available blood volumes for all immunological tests and times from sampling to blood processing. Our study describes quantitative data, whereas functional capacities or single cell signatures are not yet studied in detail and subject to future studies. CONCLUSION Infants admitted with suspected LOS are characterized by lower quantities of regulatory T cells. Our exploratory data support the concept of a potential immaturity state and failed immune tolerance development for young infants with LOS. Future large-scale studies are needed to elucidate pre-sepsis conditions and to target the microbiome-immunity interplay as a potential risk pattern. DATA AVAILABILITY STATEMENT The data analyzed in this study is subject to the following licenses/ restrictions: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. Requests to access these datasets should be directed to matsingmar.fortmann@uksh.de. ETHICS STATEMENT The studies involving human participants were reviewed and approved by the ethics committee "Ethikkommission der Universitä t zu Lü beck". Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin. FUNDING The FWS study is part of a clinician scientist grant provided by the German Center for Infection Research (DZIF) which is funded by the German Ministry for Education and Research. There has been no involvement in study design, collection of analysis, interpretation of data, writing of the report and decision to submit the manuscript for publication by the German Ministry for Education and Research. The first version of the manuscript was written by IF and CH. No payment, honorarium, grant or other form of payment has been given to the authors. The study was also supported by the University of Lübeck junior research grant (IF), Lübeck-Hilfe für krebskranke Kinder (https://luebeck-hilfe-fuerkrebskranke-kinder.de) and Annemarie-König-Stiftung.	2021-09-03T13:12:00.689Z	2021-08-27T00:00:00.000	{ "year": 2021, "sha1": "5c08b371f6fcc18612b73c86601237818e75f3c8", "oa_license": "CCBY", "oa_url": "https://www.frontiersin.org/articles/10.3389/fimmu.2021.666447/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "5c08b371f6fcc18612b73c86601237818e75f3c8", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
71716088	pes2o/s2orc	v3-fos-license	Verbalizing phylogenomic conflict: Representation of node congruence across competing reconstructions of the neoavian explosion Phylogenomic research is accelerating the publication of landmark studies that aim to resolve deep divergences of major organismal groups. Meanwhile, systems for identifying and integrating the products of phylogenomic inference–such as newly supported clade concepts–have not kept pace. However, the ability to verbalize node concept congruence and conflict across multiple, in effect simultaneously endorsed phylogenomic hypotheses, is a prerequisite for building synthetic data environments for biological systematics and other domains impacted by these conflicting inferences. Here we develop a novel solution to the conflict verbalization challenge, based on a logic representation and reasoning approach that utilizes the language of Region Connection Calculus (RCC–5) to produce consistent alignments of node concepts endorsed by incongruent phylogenomic studies. The approach employs clade concept labels to individuate concepts used by each source, even if these carry identical names. Indirect RCC–5 modeling of intensional (property-based) node concept definitions, facilitated by the local relaxation of coverage constraints, allows parent concepts to attain congruence in spite of their differentially sampled children. To demonstrate the feasibility of this approach, we align two recent phylogenomic reconstructions of higher-level avian groups that entail strong conflict in the "neoavian explosion" region. According to our representations, this conflict is constituted by 26 instances of input "whole concept" overlap. These instances are further resolvable in the output labeling schemes and visualizations as "split concepts", which provide the labels and relations needed to build truly synthetic phylogenomic data environments. Because the RCC–5 alignments fundamentally reflect the trained, logic-enabled judgments of systematic experts, future designs for such environments need to promote a culture where experts routinely assess the intensionalities of node concepts published by our peers–even and especially when we are not in agreement with each other. failed to support several of the deep divergences recovered in the preceding study, particularly within the Neoaves sec. (secundum = according to) Sibley et al. (1988) [3]. Thomas (2015) [4] used the term "neoavian explosion" to characterize the lack of congruence between inferences of early-diverging lineages (see also [5]). Similarly, after reviewing six phylogenomic studies, Suh [6] concluded that the root region of the Neoaves constitutes a "hard polytomy". Multiple analyses have dissected the impact of differential biases in terminal and genome sampling, as well as evolutionary modeling and analysis constraints, on resolving this complex radiation [7,8,9]. Suh [6] argues that a well resolved consensus is not imminent (though see [10]). Brown et al. (2017) [11] analyzed nearly 300 avian phylogenies, finding that the most recent studies "continue to contribute new edges". These recent advancements provide an opportunity to reflect on how synthesis should be realized in the age of phylogenomics [11,12,13]. The neoavian explosion can be considered a use case where multiple studies provide strong signals for conflicting hierarchies. Resolution towards a single, universally adopted tree is unlikely in the short term. Rather than focusing on the analytical challenges along the path towards unitary resolution [9], we turn to the issue of how the persistence of conflict affects the design of synthetic data infrastructures. In other words, how do we build a data service for phylogenomic knowledge in the face of persistent conflict? This question is of broad relevance to systematists, comparative evolutionary biologists, and designers of biological information services interested in robust, reproducible, and reusable phylogenomic data. And it turns on the issue of improving identifiers and identifier-to-identifier relationships for this domain. Particularly verbal representations of the neoavian explosion are not well designed for conflict representation and synthesis [14]. To alleviate this, some authors use tree alignment graphs in combination with color and width variations to identify regions (edges) of phylogenomic congruence and conflict [15]. Other authors may show multiple incongruent trees sideby-side, using color schemes for congruent clade sections [9]. Yet others may use tanglegrams a novel answer to our central question: "how to build a synthetic knowledge environment in the face of persistent phylogenomic conflict?" The discussion focuses on the feasibility and desirability of creating such an integration service, emphasizing the role of trained expert judgment in providing them [30]. Methods Syntactic and semantic conventions 1. Taxa are models, concepts are mimics. We typically refrain from using the terms "taxon", "taxa", or "clade(s)". We take taxa to constitute evolutionary, causally sustained entities whose members are manifested in the natural realm. The task for systematics is to successively approximate the identities and limits of these entities. Thus, we assign the status of 'models' to taxa, which systematists aim to 'mimic' through empirical theory making. This perspective allows for realism about taxa, and also for the possibility to let our representations stand for taxa [31], at any given time and however imperfectly, to support evolutionary inferences. In reserving a model status for taxa, we can create a separate design space for the human theory-and language-making domain. In the latter, we speak only of taxonomic or phylogenomic concepts-the products of inference making [21]. 2. Sameness is limited to the same source. Therefore, for the purpose of aligning the neoavian explosion use case, we need not speak of the "same taxa" or "same clades" at all. Similarly, we need not judge whether one reconstruction or the other more closely aligns with deepbranching avian taxa, i.e., which is (more) 'right'? Instead, our alignment is only concerned with modeling congruence and conflict across two sets of concept hierarchies. The concepts are labeled with the "sec." convention to maintain a one-to-one modeling relationship between concept labels and concepts (clade identity theories). Accordingly, there is also no need to say that, in recognizing each a concept with the taxonomic name Neornithes, the two author teams are authoring "the same concept". Instead, we model the two labels 2015.Neornithes and 2014.Neornithes, each of which symbolizes an individually generated phylogenomic theory region. As an outcome of our alignment, we may say that these two concepts are congruent, or not, reflecting the intensional alignment (to be specified below) of two phylogenomic theories. But, by virtue of their differential sources (authorship provenance), the two concepts 2015. Neornithes and 2014.Neornithes are never "the same". "Sameness" is limited in our approach to concepts whose labels contain an identical taxonomic name and which originate from a single phylogenomic hierarchy and source. That is, 2015.Neornithes and 2015.Neornithes are (labels for) the same concept. The alignments are generated with the open source Euler/X software toolkit [28]. The toolkit ingests multiple trees (T 1 , T 2 , T 3 , etc.) and articulation sets (A 1-2 , A 2-3 , etc.), converting them into a set of logic constraints. Together with other default or facultative constraints (C) needed for modeling tree hierarchies, these input constraints are then submitted to a logic reasoner that provide two main services. First, the reasoner infers whether all input constraints are jointly logically consistent, i.e., whether they permit at least one "possible world". Second, if consistency is attained, the reasoner infers the set of Maximally Informative Relations (MIR). The MIR constitute that unique set of RCC-5 articulations for every possible concept pair across the input sources from which the truth or falseness of any relationship in the R 32 lattice can be deduced [14,26,33]. Many toolkit options and functions are designed to encode variable alignment input and output conditions, and to interactively obtain adequately constrained alignments. The toolkit also features a stylesheet-driven alignment input/output visualization service that utilizes directed acyclical graphs [28]. A step-wise account of the user/toolkit workflow interaction is provided in [26]. Special challenges for multi-phylogeny alignments Aligning phylogenomic trees entails several special representation and reasoning challenges. We address three aspects here that have not been dealt with extensively in previous publications. 1. Representing intensional parent concept congruence via locally relaxed coverage. The first challenge relates directly to the issue of parent node identity. Unlike comprehensive classifications or revisions [14,26,34], phylogenomic reconstructions typically do not aspire to sample low-level entities exhaustively. Instead, select exemplars are sampled among all possible low-level entities. The aim is to represent lower-lever diversity sufficiently well to infer reliable higher-level relationships. Often, terminal sampling is not only incomplete for any single reconstruction, but purposefully complementary to that of other analyses. Generating informative genome-level data remains resource-intensive [10]. This makes it prudent to coordinate terminal sampling globally, by prioritizing the reduction of gaps over redundant terminal sampling. In the case of 2015.PEA (198 terminals) versus 2014.JEA (48 terminals), only 12 species-level concept pairs have labels with identical taxonomic names. By default, the logic toolkit applies a coverage constraint to every input concept region. Coverage means that the region of a parent is strictly circumscribed by the union of its children [35]. However, this constraint is relaxable, either globally for all concepts, or locally for select concepts. To relax coverage locally, the prefix "nc_" (no coverage) is used in the input, as in 2014.nc_Psittacidae. This means: either a parent concept's referential extension is circumscribed by the union of its explicitly included children, or there is a possibility of additional children being subsumed under that parent but not mentioned in the source phylogeny. Either scenario can yield consistent alignments. In other words, if a parent concept has relaxed coverage, it can attain congruence with another parent concept in spite of each parent having incongruent sets of child concepts. Managing coverage in the toolkit input is not trivial. Relaxing coverage globally is akin to saying "anything goes", i.e., any parent could potentially include any child. This would yield innumerable possible worlds, and therefore has no value for our purpose. On the other hand, applying coverage globally means-counter-intuitively in the case of phylogenomic trees-that only parents with completely congruent sets of children can themselves attain congruence. The challenge for experts providing the input is thus to relax coverage locally, and strictly in the service of 'neutralizing' lower-level sampling differences between trees that should not yield conflict at higher levels. The effect of locally relaxed coverage is illustrated in Figs 1-4, using the example of parrots-2015./2014.Psittaciformes. At the species level, the author teams sampled wholly exclusive sets of concepts for this alignment region (Figs 1 and 3 (Fig 2). The absence of even partial concept region overlap at the terminal level 'propagates up' to the highest-level parent concepts, which are therefore also exclusive of each other. Asserting higher-level node congruence in light of lower-level node incongruence requires a conception of node identity that affirms counter-factual statements of the following type: if 2014.JEA had sampled 2014.Psittacus_erithacus, then the authors would have included this species-level concept as a child of 2014.Psittacidae. This is to say that 2015./2014.Psittacidae, and hence their respective parents, are intensionally defined [25,36,37]. In all toolkit visualizations, the input and aligned, noncongruent concepts sec. 2015.PEA are shown as green rectangles (T 2 −18 concepts). Input and aligned, non-congruent concepts sec. 2014.JEA are shown as yellow octagons (T 1 −6 concepts). Congruent sets of aligned, multi-sourced concepts (first shown in Fig 4) are rendered in gray rectangles with rounded corners. In this input visualization, each phylogenomic tree is separately assembled via parent/child (is_a) relationships (solid black arrows). All species-level concepts sec. 2015.PEA and 2014.JEA are exclusive of each other. Under strict application of the coverage constraint, this is represented by asserting eight articulations (dashed magenta arrows) of disjointness (!) of each species-level concept from the other-sourced order-level concept. The legend indicates the numbers of nodes and edges for each input tree, parent/child relationships, and expert-asserted input articulations. See also S1 File. https://doi.org/10.1371/journal.pcbi.1006493.g001 shows reasoner-inferred non-/congruent concepts and articulations (see legend)-i.e., none in this particular case. The reasoner infers 108 logically implied articulations that constitute the set of MIR. See also S2 File. Although the input and alignment of Figs 1 and 2 are empirically defensible, they fail to capture certain intuitions we have regarding the higher-level 2015./2014.Psittaciformes relationship. For instance, we may wish to say: "Sure, the author teams sampled complementary species-level concepts. Yet these trees are not actually in conflict. At higher levels, there likely is agreement that parrots are parrots, and non-parrots are non-parrots". That is: 2015. Psittaciformes = = 2014.Psittaciformes. To obtain this intuitive alignment, we have to locally relax coverage at select lower levels (Fig 3). In particular, 2015 .nc_Nestor, 2014, in spite of the mutually exclusive species-level concepts sampled. Jointly, these four instances of relaxing coverage render the articulation 2015.Psittacidae = = 2014.Psittacidae consistent, and hence also 2015.Psittaciformes = = 2014.Psittaciformes (Fig 4). https://doi.org/10.1371/journal.pcbi.1006493.g002 Verbalizing phylogenomic conflict Using a combination of published topological information (and support), more or less direct reiterations of phenotypic traits (cf. discussions and supplementary data of 2015.PEA and 2014.JEA), and trained judgment [30], we align these concept regions as if there are congruent property criteria that each region entails, i.e., something akin to an implicit set of synapomorphies or uniquely diagnostic features. Of course, the phylogenomic data provided by 2015.PEA and 2014.JEA do not signal intensional definitions directly. But neither do their genome-based topologies for parrots provide evidence to challenge the status of such definitions as previously proposed [38]. In addition, particularly 2015.PEA (supplementary information; sections on "detailed justification for fossil calibrations" and "detailed phylogenetic discussion; pp. 3-21) provide a provide an in-depth account of how their preferred topology relates to published, property-centered circumscriptions of dozens of higher-level clade Verbalizing phylogenomic conflict concepts. We have to assume, fallibly and non-trivially, that such topology-to-synapomorphy relations are also implied by JEA.2014, as reflected (inter alia) in their discussion. Three clarifications are in order. First, Region Connection Calculus is at best a means of translating the signal of an intensional definition. The congruent (= =) symbol means, only: two regions are congruent in their extension. The RCC-5 vocabulary is obviously not appropriate for reasoning directly over genomic or phenomic property statements. The reasoner does not assess whether 2015.Psittacidae, or any included child or aligned concept, has 'the relevant synapomorphies'. Doing so would not be trivial even if property-based definitions were provided for all higher-level node concepts, because we would still have to make theory-laden assumptions about their congruent phylogenomic scopes [26,39,40]. Second, we are not providing detailed textual narratives that would justify each assertion of higher-level congruence. Such narratives are possible, and even needed to understand disagreements, because they explain the reasoning process behind an expert-made assertion. However, our main objective here is to focus on the issue of RCC-5 translation of systematic signals; not on a character-bycharacter dissection of each congruent articulation. Third, a sensible intensional alignment strategy uses a minimal number of instances of locally relaxed coverage in order to compensate for differential child sampling at lower levels, so that parent coverage can remain in place at higher levels to expose incongruent node concepts. The benefits of this strategy will be shown below. 2. Representing clade concept labels. Our modeling approach requires that every region in each source tree receives a taxonomic or clade concept label. However, the source publications only provide such labels for a subset of the inferred nodes. In particular, 2015.PEA (p. 570: Fig 1) obtained 41 nodes above the ordinal level. Of these, 17 nodes (41.5%) were explicitly labeled in either the published figure or supplement (pp. [9][10][11][12]. The authors also cite [20] as the primary source for valid name usages, yet that list is not concerned with supra-ordinal names. Similarly, 2014.JEA (p. 1322: Fig 1) inferred 37 nodes above the ordinal level, of which 23 nodes (62.2%) were given an explicit label. They provide an account (cf. supplementary materials SM6: 22-24) of their preferred name usages, sourced mainly to [20] and [41]. In assigning clade concept labels at the supra-ordinal level when the authors may have failed to do so (consistently), we nevertheless made a good faith effort-through examination of the supplementary information and additional sources [1,3,42,43,44,45,46,47]-to represent the authors' preferred name usages. Where usages were not explicit, we selected the only or most commonly applied clade concept name at the time of publication. This effort yielded 13 additional labels for 2015.PEA (Table 1), and 7 such labels for 2014.JEA (Table 2). If no suitable label was available, we chose a simple naming convention of adding "_Clade1", "_Clade2", etc., to the available and immediately higher-level node label, e.g. 2014. Passerea_Clade1. The numbering of such labels along the tree topology starts with the most immediate child of a properly named parent, and typically follows down one section of the source tree entirely ("depth-first"), before continuing with the higher-level sister section. Using this approach, we added 11 labels for 2015.PEA (Table 1) and 7 labels for JEA.2014 (Table 2). If greater numbers of labels need to be generated, including siblings, then it is sensible to have a rule for ordering sibling nodes, e.g. by assigning the next-lowest number to the sibling whose child's name appears first in the alphabet. Our numbering of the labels 2014.Passerea_Clade2 Verbalizing phylogenomic conflict aligned with higher-level concepts of [20]. We highlight these instances here because they represent a widespread phenomenon in phylogenomics. It is useful to understand how such discrepancies can be modeled with RCC-5 alignments (Figs 5 and 6). Table 1, p. 1): "Taxonomy follows Gill and Donsker (2015; fifth ed)". As shown in Fig 5, their phylogeny accommodates four sampled genus-level concepts that would correspond to children of the family-level concept Eurylaimidae sec. Gill & Donsker (2015) [20]. However, these concepts are arranged paraphyletically in relation to the In summary, our approach represents non-monophyly as an incongruent alignment of the phylogenomic tree and the source classification used to provide labels for that tree's monophyletic clade concepts. There are four distinct regions in the phylogeny of 2015.PEA where such alignments are needed: {Caprimulgiformes, Eurylaimidae, Hydrobatidae, Procellariidae, Tityr-idae} sec. Gill & Donsker (2015) [20]. Each of these is provided in the S7-S9 Files. Configuration of input constraints and alignment partitioning The source phylogenies specify 703 and 216 clade or taxonomic concepts, respectively. The frequent instances of locally relaxed coverage increase the reasoning complexity in relation to multi-classification alignments [14], making specialized RCC-5 reasoning useful [48]. The reasoning and visualization challenges commend a partitioned alignment approach. To keep the Results concise, we show visualizations of the larger input and alignment partitions only in the Supporting Information. A detailed account of the input configuration and partitioning workflow is given below. Underlying all alignments is the presumption that at the terminal (species) level, the taxonomic concept labels of 2015.PEA and 2014.JEA are reliable indicators of either pairwise congruence or exclusion [14,26,32]. That is, e.g., 2015.Cariama_cristata = = 2014.Cariama_ cristata, or 2015.Charadrius_hiaticula ! 2014.Charadrius_vociferus. Because the time interval Verbalizing phylogenomic conflict separating the two publications is short in comparison to the time needed for taxonomic revisions to effect changes in classificatory practice, the genus-or species-level taxonomic concepts are unlikely to show much incongruence; though see [49] or [50]. We note that 2015.PEA (p. 571) use the label 2015.Urocolius(_indicus) in their phylogenomic tree, which also corresponds to the genus-level name endorsed in [20] Gill & Donsker (2015). However, in their Supplementary Table 1 the authors use 2015.Colius_indicus. We chose 2015.Urocolius and 2015.Urocolius_indicus as the labels to apply in the alignments. The toolkit workflow favors a partitioned, bottom-up approach [29]. The process of generating, checking, and regenerating input files must be handled 'manually' on the desktop (note: improved workflow documentation and semi-automation of input-output-input changes are highly desirable). The performance of different toolkit reasoners was benchmarked in [28]. To work efficiently, the large problem of aligning all concepts at once is broken down into multiple smaller alignment problems, e.g. 2015./2014.Psittaciformes (Figs 3 and 4). To manage one particular order-level alignment, we start with assembling each input phylogeny separately, with relaxed coverage applied as needed (Fig 3). The RCC-5 articulations for low-level concept pairs are provided incrementally, e.g., in sets of 1-5 articulations at a time. Following such an increment, the toolkit reasoning process is re-/deployed to validate input consistency and infer the number of possible worlds. There is an option to specify that only one possible world is sought as output, which is equivalent to just checking for input consistency, as opposed to inferring all possible worlds. Doing so saves time as long as the input remains (vastly) under-specified. The stepwise approach of adding a small number of articulations at a time leads to increasingly constrained alignments, while minimizing the risk of introducing many new. difficult-to-diagnose inconsistencies. Once a set of small, topographically adjacent alignment partitions is well specified, these can serve as building blocks for the next, larger partition. Hence, the basic sequence of building up larger alignments is: (1) obtain a well-specified low-(order-or family-) level alignment; (2) record the inferred parent-level articulations from this alignment; (3) propagate the latternow as low-level input articulations-for the next, more inclusive alignment; (4) as needed, prune the lowest-level (sub-ordinal) input concepts and articulations of (1) from this alignment; (5) repeat (1) to (4) for another paired region; (6) assemble the more inclusive alignment by (manually) connecting the pruned, propagated concepts and articulations from two or more lower-level alignments, by adding to them the higher-level concepts from each input phylogeny. Depending on the interplay between (ranked) higher-level names recognized in each phylogeny and the number of terminal concepts sampled, steps (1) to (6) Telluraves, which are therein only represented with two concepts labels and one congruent articulation. These two complements are the core partitions that inform our use case alignment, globally. The corresponding S10 and S11 Files include the input constraint (.txt) and visualization (.pdf) files, along with the alignment visualization (.pdf) and MIR (.csv). The two large partitions yield unambiguous RCC-5 articulations from the species concept level to that of 2015./2014.Neornithes. They can be aggregated into a synthetic, root-to-order level alignment, where all subordinal concepts and articulations are secondarily pruned away (see above). Such an alignment retains the logic signal derived from the bottom-up approach, but represents only congruent order-level concept labels as terminal regions, except in cases where there is incongruence. We present this alignment as an analogue to Fig 1 in [4] (p. 515), and compare how each conveys information about congruent and conflicting higher-level clade concepts. Lastly, we further reduce the root-to-order alignment to display only 5-6 clade concept levels below the congruent 2015./2014.Neoaves. This region of the alignment is the most conflicting, and therefore forms the basis for our Discussion. Higher-level congruence Our alignments show widespread higher-level congruence across the neoavian explosion use case; along with several minor regions of conflict and one strongly conflicting region between concepts placed immediately below the 2015. /2014.Neornithes partition shows 305 aligned regions-247 without the "no coverage" regions-of which 60 congruently carry at least one concept label from each source phylogeny. This alignment also shows eight congruent species-level concept regions. These would be the only instances of congruence if coverage were globally applied (Figs 1 and 2). Therefore, relaxing the coverage constraint yields 52 additional instances of higher-level node congruence. Similarly, the 2015./2014.Telluraves partition has 231 aligned regions-194 without the "no coverage" regions-of which 38 are congruent. This corresponds to an increase of 34 regions, compared to four congruent species-level concept regions present under strict coverage. Correcting for the redundant 2015./2014.Telluraves region, we 'gain' 85 congruent parent node regions across the two phylogenies if node identity is encoded intensionally (Figs 3 and 4). Indeed, this approach yields the intuitive articulation 2015. Neornithes = = 2014.Neornithes at the highest level. The first of these is less problematic from a standpoint of achieving integration: for a given alignment subregion, the more densely sampled phylogeny will entail additional, more finely resolved clade concepts in comparison to its counterpart. Typically, this distinction belongs to the phylogeny of 2015.PEA, due to the 4:1 ratio of terminals sampled. There are 83 above species-level clade concepts sec. 2015.PEA that can be interpreted as congruent refinements of the 2014.JEA topology (see S10 and S11 Files). Conversely, only two such instances of added reso ; Figs 9 and 10, and S14 and S15 Files); and finally, 2015./2014.Neoaves (26 overlaps; Figs 11-13, and S16-S18 Files). We will examine each of these in sequence. 1. 2015./2014.Pelecanimorphae. The two author teams sampled four family-level concepts congruently for this alignment region (Fig 7). However Within 2015.Pelecaniformes, we obtain five additional overlapping articulations between five concepts that make up the 2015/2014 supra-familial topologies in this alignment (Fig 7). This conflict is due to the differential assignment of 2015. Australaves as sister. This is the first occurrence of conflict that cannot justifiably be resolved by relaxing parent coverage, but instead reflects divergent phylogenomic signals. Whole-concept and split-concept resolution How to speak of such overlap? In Fig 9, we utilize clade concept labels that pertain to each input phylogeny. In the resulting alignment, the articulation 2015.Eutelluraves >< 2014. Afroaves is visualized as a dashed blue line between these regions. Yet Fig 9 also specifies the extent of regional overlap at the next lower level. Accordingly, only the region 2015./2014.Coracornithia is subsumed under each of the overlapping parents. This is indicated by the two inclusion arrows that extend 'upward' from this region. The other two paired child regions are respectively members of one parent region. If we call the input regions 2015.Eutelluraves "A" and 2014.Afroaves "B", we can use the following syntax to identify output regions that result from overlapping input concepts [26]: A � B (read: "A and B") constitutes the output region shared by two parents, whereas A\b ("A, not b") and B\a ("B, not a") are output regions with only one parent. We call this more granular syntax split-concept resolution ("merge concepts" in [26]), as opposed to whole-concept resolution which preserves the syntax and granularity provided by the input concept labels. (Fig 9). Moreover, in this case the split-concept resolution syntax is redundant or unnecessary, because each of the three resolved regions under "A" (2015.Eutelluraves) and "B" (2014.Afroaves) is congruent with two regions already labeled in the corresponding input phylogenies. We will see, however, that this granular syntax is essential for verbalizing the outcomes of more complex alignments that contain many overlapping regions. Split-concept resolution for the neoavian explosion In Fig 13, the same 'zoomed-in' alignment is shown under split-concept resolution. This permits identifying all output regions created by the 26 overlaps of the neoavian explosion (see Table 4). The entire set consists of 78 labels; i.e., 26 labels for each split-resolution product {A � B, A\b, B\a} for one instance of input region overlap. Not all of these split-concept resolution labels are semantically redundant with those provided in the input. Specifically, 51 labels are generated 'in addition' for the 12 terminal congruent regions (compare with Fig 12). These are indeed unnecessary synonyms for regions already identified in the input. However, the relative number of additional labels generated per input region is telling. This number will be highest for regions whose differential placements are the primary drivers of incongruence. As explained above, these are: 2015./2014.{Phoenicopterimorphae, Charadriiformes, Columbi-morphae}, respectively with 14, 8, and 7 additional labels. Six redundant split-concept resolution labels are further produced for input regions that are unique to one phylogeny; e.g., 2014. Columbea is also labeled 2015.Neoaves_Clade1 \ 2014.Passerea (where the "\" means: not). The remaining 21 split-concept resolution labels identify 15 salmon-colored alignment regions-11 uniquely and 4 redundantly with 2-3 labels each-for which there are no suitable labels in either of the phylogenomic input trees (Table 4). Forty-six additional articulations are inferred to align these regions to those displayed in Fig 12. Although these novel regions are not congruent with any clade concepts recognized by the source phylogenies, they are needed to express how exactly the authors' respective clade concepts overlap. Three distinct reference services are gained by generating the split-concept resolution labels. First, in cases where no whole-concept resolution labels are available, we obtain appropriately short and consistent labels to identify the split regions caused by overlapping clade concepts. Second, the {A � B, A\b, B\a} triplets have an explanatory function, by using the same syntactic set of input labels (A, B) to divide complementary alignment subregions of an overlap. If we focus on one label of a triplet, we can find the two complements, and thereby systematically explore the 'reach' of each split in the alignment. Third, the clade concept labels (A, B) used in the split-concept resolution labels will be exactly those that identify overlapping regions across the source phylogenies. Analysis of clade name performance We can now also ask to what extent the clade names (syntax) used by the two author teams succeed or fail to identify congruent and incongruent concept regions (semantics). Such name:meaning (read: "name-to-meaning") analyses were carried out in three previous alignment use cases, with rather unfavorable outcomes for the respective names in use [14,32,51]. Here, based on the alignment of Fig 11, the 97 x 83 input concepts yield a set of 8,051 MIR (S16 File). Of these, 384 MIR involve one of four "no coverage" regions added to 2014.JEA concepts. We therefore restrict the name:meaning analysis to the remaining 7,667 MIR (Table 5). Interestingly, the clades names used by the respective author teams fare rather well. Only nine of 7,667 pairings in the MIR (0.12%) are unreliable as identifiers of in-/congruence of the respective RCC-5 articulation. In seven instances, two congruent concepts have different Table 4 Fig 13. For each split-concept resolution label (or label cluster), we provide the two immediate children or constituent concepts 1 and 2 -i.e., what is jointly subsumed 'underneath' the split-as well as the set of lower-level concept regions (using whole-concept resolution labels) that are differentially distributed by the split between the two source phylogenies. � = Two children listed. Pelecaniformes is the single instance in which the meaning of the same name is less inclusive in one source (Fig 7). Lastly, the overlapping relationship 2015.Otidimorphae_Clade1 >< 2014.Otidimorphae_Clade1 involves the same name (Figs 12 and 13), though it is not actually used by the author teams (see Methods). # Split-concept label(s) Constituent clade concept 1 Constituent clade concept 2 Lower-level concept regions differently In summary, the clade concept names used by 2015.PEA and 2014.JEA rarely provide an incorrect signal regarding in-/congruence. This desirable outcome seems to reflect their recognition that newly inferred clade concepts merit the use of unique names. Comparison with other conflict visualizations We now compare these results with conflict analysis and visualization tools created for the Open Tree of Life project (OToL)-a community-curated tree synthesis platform [13,22,23,24]. The OToL approach is explained in [11,15,23,52,53]. The method starts off with 'normalizing' all terminal names in the source trees to a common taxonomy [24]. Having the same terminal name means taxonomic concept congruence (= =). To assess conflict from the perspective of one rooted input tree (A), a source edge j of that tree is taken to define a rooted bipartition S(j) = S in \| S out , where S in and S out are the tip sets of the ingroup and outgroup, respectively. The algorithm progresses sectionally from the leaves to the root. Concordance or conflict for a given edge j in tree A with that of tree B is a function of the relative overlap of the corresponding tip sets, as follows [23]. Concordance between two edges in the input trees A and B is obtained when B in is a proper subset (�) of A in and B out � A out . On the other hand, two edges in trees A and B are conflicting if none of these sets are empty: A in intersects ( T ) with B in , A in T B out , or B in T A out . In other words, conflict means that there is reciprocal overlap in the ingroup and outgroup bipartitions across the two trees. We applied this approach in both directions, i.e. starting with 2014.JEA as primary source and identifying edges therein that conflict with those of 2015.PEA, and vice-versa. The visualizations are shown in Figs 14 and 15, respectively. Most of the red edges in Fig 15, which is based on the more densely sampled tree sec. 2015. PEA, are consistent with the overlapping RCC-5 relationships shown in Figs 7 to 13. However, within the 2015.Pelicanimorphae, certain RCC-5 overlaps (Fig 7) are not recovered ("false positives"). In addition, numerous edges within the 2015.Passeriformes are shown as conflicting ("false negatives") but are congruent refinements based on the RCC-5 alignment (Fig 8). Using the less densely sampled tree sec. 2014.JEA as the base topology creates is instructive (Fig 14). Here, a much larger subset of the topology 'backbone' is inferred by the OToL algorithm as conflicting-an outcome that would appear inconsistent. Table 5. Clade name-to-RCC-5 relationship reliability analysis for the higher-level neoavian explosion alignment. Relationship data are derived from the set of MIR corresponding to Fig 11 and Verbalizing phylogenomic conflict therefore attach as children to a higher-level parent in the OToL taxonomy. The latter is used to place terminals that are differentially sampled between sources. For instance, 2015.Ciconiiformes-which has no close match in 2014.JEA-may end up attaching as a child of 2014.Neognathae instead of 2014.Pelecanimorphae (Fig 7). Hence the OToL taxonomy is used to represent concept intensionality, but it cannot do so reliably if it lacks relevant input concepts. Black edges indicate concordance, whereas red edges signal conflict. Clade and terminal concept labels are added in accordance with the present study. Moreover, consistency or inconsistency of the edge concordance/conflict analysis with the RCC-5 alignments (Figs 7 to 13) are signaled via a green "+" circle and a black "-" circle, respectively. See also S19 File. Verbalizing phylogenomic conflict Key phylogenomic conflict representation conventions We review the key conventions of our approach before discussing services that can be derived from our alignments. 1. Using the taxonomic concept label convention of [14] allows us to individuate each concept entailed in 2014.JEA and 2015.PEA, even if the taxonomic or clade concept name components are identical, as in 2015.Pelicaniformes < 2014.Pelicaniformes. 2. Because our main intention is to represent phylogenomic congruence and conflict across these inferred phylogenies, there is no need to speak of sameness in any profound sense, such as referring to the "same {clades, nodes, species, taxa}". Such language is best used once we shift from modeling similarities and differences between human-made phylogenomic theories, to hopefully (but not necessarily) robust evolutionary inferences. We thereby avoid blurring the lines between two important communication goals best met by maintaining complementary manners of speaking [21]. 3. Linking concepts via is_a (parent/child) relationships permits the assembly of single-source hierarchies, whereas RCC-5 articulations express the relative congruence of concept regions across multi-source hierarchies. Uncertainty can be accommodated via disjunctions of the base five relations [33]. 4. Under parent coverage, differential child-level sampling will propagate up to yield incongruent relationships among parent-level clade concepts [14,26,29]. Local relaxation of the coverage constraint can mitigate this effect. However, this requires expert judgment [30], reflected in input articulations that stipulate counter-factual circumstances. We can thereby indirectly model intensional (property-based) node concept definitions in RCC-5, and obtain instances of clade concept congruence in spite of incongruent terminal sampling (Figs 1-4). 5. Because every clade concept region to be aligned requires a label suited for human communication, we need to supply such labels when the sources fail to do so. A pragmatic solution is to utilize the next available higher-level name and add the suffix "_Clade#", as in 2015. Neoaves_Clade1 or 2014.Passerea_Clade3 (Tables 1 and 2). This may involve deciding on a breadth-vs. depth-first approach, and having a rule to prioritize between sibling nodes. 6. In some instances, the source may provide a clade concept label for a non-monophyletic tree region. Representing such mismatches is achieved by providing an(other) alignment between (1) the reference classification and (2) the phylogeny to which the labels are incongruently applied (Figs 5 and 6). 7. Multi-rooted, bottom-up, and incremental partitions may be required to manage the logic reasoning complexity of large or even global alignments [28]. Consistent alignments of higher-level concept hierarchies (Fig 11), can be derived from this bottom-up approach by propagating the inferred parent-level articulations while pruning the children used for aligning lower-level partitions [14,28,51]. 8. Overlapping relationships among higher-level clade concepts can be represented using either whole-concept or split concept resolution (compare Figs 9 and 10; Figs 12 and 13). The latter option provides a uniquely powerful syntax to partition and label the alignment regions created by concept overlap (Tables 3 and 4). 9. The reasoner-inferred MIR are useful for quantifying all pairwise instances where the names used by each source succeed, or fail, in matching the signal of the RCC-5 relationships (Table 5). 10. The alignments can be compared with other conflict representation methods, such as the OToL concordance/conflict visualizations [11,15,23]. This is particularly illustrative in cases where differential sampling of low-level concepts generates unequal assessments between the OToL and RCC-5 approaches (Figs 14 and 15). The latter is more reliable in cases where expert judgment is needed to represent higher-level concept intensionality under strongly divergent sampling schemes. Knowledge representation services What can we gain from this approach, both narrowly for this use case and for future data integration in systematics? Data representation designs have inherent trade-offs. Unlike other semi-/automated phylogenomic conflict visualization methods [13,23,24], the above approach requires extensive upfront application of human expertise to obtain the intended outcomes. In return, the RCC-5 alignments deliver a level of explicitness and verbal precision exceeding that of published alternatives [4,5,6,9,16,17]. We can not just verbalize all instances of congruence and conflict, but transparently document and therefore understand their provenance in a global alignment (Figs 11 and 13). In other words, the RCC-5 alignments provide a logically tractable means to identify and also explain the extent of conflict. We can derive novel data services from the alignment products. (Note that these services are envisioned but not yet implemented in a web-based platform.) Example queries include the following. 5) Highlight sets of concepts where incongruence is due to differential granularity (sampling), versus actual overlap. (6) Identify and rank concepts that participate in the greatest number of overlapping relationships (Table 3). (7) Identify and rank the longest chains of nested, overlapping concept sets (Fig 12). (8) Highlight the congruent, lowest-level concept pairs whose incongruent placement into higher-level regions causes the chains of overlap. (9) List all split-concept resolution labels in complementary triplets {A � B, A\b, B\a}, and provide for each the two immediate children and (again) the set of lower-level, whole-concept resolution regions that are differentially distributed by the split (Fig 13 and Table 4). (10) Identify clade names that are unreliable across the source phylogenies; including identical clade name pairs that participate in concept labels with an incongruent relationship, or different clade names whose concept labels have a congruent relationship ( Table 5). All of the above queries, and many others we could propose, are enabled by our RCC-5 representation and reasoning conventions, which therefore present a new foundation for building logic-based, machine-scalable data integration services for the age of phylogenomics. Conceptualizing node identity and congruence this way addresses a gap in current systematic theory that is not adequately filled by other syntactic solutions. Linnaean naming. We have shown elsewhere that homonymy and synonymy relationships are unreliable indicators of congruence [14,26,32]. Code-enforced Linnaean naming is designed to fixate the meaning of names by ostension, while allowing the intensional components to remain ambiguous [21,54,55,56,57]. This trade-off effectively shifts the burden of disambiguating varying intensionalities associated with Linnaean names onto an additional, interpreting agent-typically human experts. Our RCC-5 alignment approach can be viewed as a way to formalize the disambiguation effort, so that it can attain machine-interpretability. Phyloreferencing. Similarly, node-based phyloreferences [58,59,60] are not well suited to reconstruct an alignment such as that of 2015./2014.Pelecanimorphae (Fig 7). This would require: (1) an elaborate notion of phyloreference homonymy and synonymy (e.g., 2015.Pelecanifores versus 2014.Pelecaniformes, or 2015.Strisores versus 2014.Caprimulgimorphae); (2) node-based definitions with inclusion/exclusion constraints that cover all terminals in the phylogeny; and (3) synapomorphy-based definitions at higher levels to model the local relaxation of coverage constraints. All of these functions may be feasible in principle with phyloreferences, provided that human experts are permitted to enact them. However, it may be fair to say that phyloreferences were not mainly designed to bring out fine differences between node concepts across multiple phylogenies. They are best utilized when concept evolution and conflict are not the main drivers of an information system design. Role of trained judgment The two largest alignments of 2015./2014.Neornithes (without) / 2015./2014.Telluraves jointly entail 895 concepts and 95 instances of relaxed parent coverage. They provide us with 97 congruent regions in the global alignment, of which 85 regions are obtained only because of the indirect modeling of intensional node definitions. The contingency of the alignment outcome on expert intentions is neither surprising nor trivial. We should therefore explore this dependency more deeply. Redelings and Holder [23: pp. 5-6] comment on the OToL synthesis method: "Any approach to supertree construction must deal with the need to adjudicate between conflicting input trees. We choose to deal with conflict by ranking the input trees, and preferring to include edges from higher-ranked trees. The merits of using tree ranking are questionable because the system does not mediate conflicts based on the relative amount of evidence for each alternative. [. . .] In order to produce a comprehensive supertree, we also require a rooted taxonomy tree in addition to the ranked list of rooted input trees. Unlike other input trees, the taxonomy tree is required to contain all taxa, and thus has the maximal leaf set. We make the taxonomy tree the lowest ranked tree. [. . .] Our method must resolve conflicts in order to construct a single supertree. However, the rank information used to resolve conflicts is an input to the method, not an output from the method. We thus perform curation-based conflict resolution, not inference-based conflict resolution." Clearly, the outcomes of the OToL synthesis method are also deeply dependent on expert input regarding the relative ranking of input phylogenies and of the OToL taxonomy [24]. We have shown (Figs 14 and 15) that these choices can lead to inconsistent outcomes whenever the sequence of input trees determines how concordance and conflict are negotiated by the algorithms. If the less densely sampled tree is prioritized, and the taxonomy cannot accommodate all components of a lower-ranked tree, then the method will show more conflict in comparison to an inverse input sequence. Any global rule of priority among trees is a poor proxy for modeling individual node concept intensionality, which requires making reliable, local decisions between (1) conflict due to differential granularity versus (2) conflict due to overlap. We can now return to the challenge posed in the Introduction. How do we build a data service for phylogenomic knowledge in the face of persistent conflict? Our answer is novel in the following sense. Assuming that such a service is desirable, we show that achieving it fundamentally depends on making and expressing upfront empirical commitments about the intensionalities of clade concepts whose children are incongruently sampled. Without embedding these judgments into the alignment input, we lose the 85 congruent parent regions recovered under relaxed parent coverage. We furthermore lose the ability to distinguish the former from more than 340 alignment regions that are not congruent. And we lose the power to express the nature of this residual conflict-granularity versus overlaps-and how to resolve it. In other words, the first step for building the phylogenomic data knowledge service will be to recognize that conceptualizations of node identity within such a system just cannot be provided through some mechanical, 'objective' criterion. Instead, we need an inclusive standard of objectivity that embraces trained judgment as an integral part of identifying and linking node concepts [30]. In that sense, phylogenomic syntheses are inference-based (contra [23]) and also driven by a specific purpose. As integrative biologists, our goal in providing RCC-5 alignments is to maximize intensional node congruence. There may not be a more reliable criterion for achieving this than expert judgment, which draws on complex and context-specific theoretical knowledge [40,43,61]. Logic representation and reasoning can help render these constraints explicit and consistent, and expose implicit articulations through the MIR which encompass all node concepts in an alignment. But logic cannot substitute the expert aligners' intensional aims and definitions. Building a phylogenomic data knowledge service forces us to become experts about externally generated results that conflict with those which we may (currently) publish or endorse. We need to become experts of another author team's node concepts, to the point where we are comfortable with expressing counter-factual statements regarding their intensionalities, in spite of incongruent child sampling. This will require a profound but necessary adjustment in achieving a culture of synthesis in systematics that no longer manages conflict this way: "If we do not agree, then it is either our view over yours, or we just collapse all conflicting node concepts into polytomies". In contrast, we need to develop the following culture of synthesis: "We may not agree with you, but we understand your phylogenomic inference well enough to express our dis-/agreements in a logic-compatible syntax. Therefore, we are prepared to assert and refine articulations from our concepts to yours for the purpose of maximizing intensional node congruence". Only then can we expect to also maximize the empirical translatability of biological data linked to diverging phylogenomic hypotheses. Shifting towards the latter attitude will be more challenging than providing the operational logic to enable scalable alignments. Automation of certain workflow components is certainly possible. Ultimately, the logic or technical issues are not the hardest bottlenecks to overcome. Designers of future data environments capable of verbalizing phylogenomic conflict and synthesis need to reflect on how to promote a culture where experts routinely re-/assess the intensionalities of node concepts published by peers. If we wish to track progress and conflict across phylogenomic inferences, we first need to design a value system that better enables and motivates experts to do so. Response to reviewers He we discuss various reviewer comments that merit a response but would break up the main flow of the narrative if inserted earlier. We take liberty to assign a header to each comment. Phylogenetic clade definitions and taxonomic concepts are fundamentally mismatched. One reviewer pointed out that clade hypotheses are about branching patterns and relationships of descent, and therefore are mismatched with our notion of node intensionality. We disagree in the following sense. We believe that we are not conflating two fundamentally different kinds of clade conceptualizations, as much as bringing out with the RCC-5 alignments one aspect in the dual, or hybrid nature of clade concepts. The latter are not either this or that-with parallels to the taxa as classes-versus-individuals literature-both can be both, depending on the pragmatic interest [36,37,62]. For the purpose of synthesis and integration, modeling the intensional aspect of clade concepts is critical. We see this purpose reflected (e.g.) in the matching of high-level terminals in [3]. No mechanism for quantitatively expressing uncertainty about tree topology. The same reviewer pointed out that we select single point estimate topologies for each author team, thereby not accounting for the complex likelihood surfaces of the reconstructions and the relative uncertainty of each topology. Applied to what we show here, this criticism is valid. However, it would be feasible perform RCC-5 alignments on a cluster of paired topology alternatives with similar likelihood values. The products can be compared in order to manage uncertainty, through identification of stable versus variable regions across multiple alignments. If most of the variation occurs at higher levels, this would mean that the vast majority of our low-level RCC-5 input articulations could be reused. Phylogenetic conflict is not limited to two trees. Another reviewed pointed out the need to align more than two phylogenies in situations where many recent reconstructions are available to inform a synthesis [5,6,11]. While the current logic toolkit handles three or more input trees in principle, there certainly are unrealized opportunities to model transitive relationships (example: for concepts A, B, C in the input trees T1, T2, T3: if A T1 = = B T2 and B T2 = = C T3 then A T1 = = C T3 ). 'Smartly' breaking down alignments of three or more trees while exploiting transitive relationships, as well as visualizing the outcomes accessible ways, are important future improvements for this approach. "Not every clade [concept] is worth labeling and discussing". We can agree with that assessment. But, having a framework to do so is critical to evaluating the feasibility of a phylogenomic data knowledge service, and should not trail behind discussions regarding its desirability. If we have no formalized means of translating Fig 1 of [3] into a machine-accessible language (Fig 11), then we cannot fully understand the costs and benefits of building the service. Incentivizing alignment production. One reviewer pointed out that efforts to align multiple trees are costly, and inquired about our suggestions for incentivizing such expert contributions. An initial answer would point to the creation of an e-journal, where multi-phylogeny and -taxonomy alignments can be published either as stand-alone articles or in association with separate publications of new tree reconstructions. The platform of a formal journal best responds to expert needs to receive academic credit [63]. Knowledge systems such as [64] could represent the information input and output. The most valuable product of such an ejournal are the expert-vetted sets of RCC-5 articulations, which represent a new kind of "systematic intelligence". Scientists and commercial publishers may utilize this intelligence to improve the precision and recall of systematically structured data [54], where business models would focus on the latter clients for revenue. Needless to say, these are ideas that will take time to concretize and test.	2019-03-08T14:19:04.927Z	2017-12-14T00:00:00.000	{ "year": 2019, "sha1": "51d0aa36b481c3e91b8858ec342674cc6c1b1c68", "oa_license": "CCBY", "oa_url": "https://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1006493&type=printable", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "51d0aa36b481c3e91b8858ec342674cc6c1b1c68", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [ "Computer Science", "Biology", "Medicine" ] }
197615437	pes2o/s2orc	v3-fos-license	Electromechanical characterization of didactical piezoelectric sensors based on crystalline grade PET A mechanical shock or impact sensor system for teaching purposes was developed and characterized. The device consists of Mylar layers (3×3 to 21×21 cm2) confined by aluminum sheets. The metallic surfaces are wired in order to send their signals to a digital oscilloscope. The instrument receives voltage variations after the surface of the transducer is stroked by a plastic projectile (capsule mass: 50 g) released in free fall (heights: 0.15 - 3 m). The respective piezoelectric constant was determined to be d = 42.02±10−12 m/V. This proposed experiment can be attractive for a basic instrumentation course which combines both theoretical and experimental concepts of mechanics and materials sciences. Introduction The trend of technological improvement of sensors leads to deeper and more detailed scientific investigations [1]. That is a reason for institutions to promote both research and development of sensor engineering. Besides, this topic is a key element of the curricula for numerous scientific profiles in Physics and Engineering. For instance, since collisions are an important concept in the field of mechanics; it can be used to link other technological concepts such as electromechanical transducers; this scope could be very attractive for academic courses oriented to applications in several fields such as sports, liberal arts or forensic science. In fact, many applications arise from the force evaluation of a left-jab boxing movement or the estimation of the initial speed of vehicles after collisions. Accordingly, if the academic interest is focused on Physics, teachers center the lesson on the mechanical phenomena, paying no attention to the data acquisition details. Such didactic proposals use the sensors as black-boxes to only correlate the observable inputs vs. the output parameters [2]. On the other hand, if the interest is centered in the instrument, the engineering aspects take priority over the physical analysis. Concerning the subject of collisions, we found only few didactic proposals based on inexpensive components to manufacture suitable transducers. In this work we suggest the construction of a cost-effective system based on an array of aluminum-Mylar-aluminum layers used as transducer and a didactic oscilloscope to monitor the electrical signals produced by an impact. Such development and the corresponding characterization are also attractive and useful for several biomechanical applications. Piezoelectricity in polymers Under mechanical stress, some ceramic and polymeric materials exhibit a change in their electric charge density; the inverse effect leads to mechanical deformations induced by external electric fields; this material property is known as piezoelectricity [3]. A theoretical toy-model could be the following: the mechanical stress modifies the electronic orbitals around the molecules; in this action, some regions acquire positive or negative charge, this reconfiguration promotes the formation of local dipole moments [3]. The aggregate effect of the dipole moments produces the macroscopic and observable electric field. Piezoelectric materials have a molecular configuration suitable for electrical polarization: a non-centrosymmetric structure; this morphological asymmetry permits the migration of electric charges in defined molecular regions [1]. Particularly, in a polymer composed by bulky monomers and stable macromolecules, the electric polarization is easily promoted within the system by different kind of stimuli; where the type of atomic bonds and their location is crucial for increasing the local dipole moment. Thus, structures with adequate molecular architectures and regular arrangements are suitable to exhibit a polar order [3,4]. For instance, organic materials such as vinylidene polyfluoride / PVDF and polyethylene terephthalate / PET, display such structural characteristics. In fact, PVDF is an efficient piezoelectric-polymer utilized in several instruments [4][5][6], but it can be expensive for some schools. In contrast, PET is less efficient but almost costless. A highly crystalline PET (within in the polymers context) is Mylar, a common use product in many home appliances and also in some piezoelectric applications [3,4]. Fig. 1 shows the molecular structure of a Mylar monomer. In its manufacture, Mylar is slowly cooled, favoring the formation of spherulites and lamellae arrangements which exhibit crystalline structures at a local level, being these structures responsible for the whitish color of the material due to scattering effects. Context of the characterization equation In general, the constituent equations of a piezoelectric material can be described by four fundamental parameters [1,4] The fundamental equations that describe the piezoelectric effect are complex: they interconnect electrical and mechanical variables in all directions; so their description requires tensor theory, a subject that students should learn in advanced physics-math courses. The scope of this document is a didactic approach for first-year university students; thus, for simplicity, we use a uniaxial model where the effects of mixed components of voltage, load, and electric field, are considered irrelevant [8]. In this way, the one-dimensional equations for our piezoelectric material are: where d* represents the mechanical deformation produced by an electromagnetic stimulus. However, our interest is centered on the electrical response to a mechanical stress, and then this parameter is neglected. In addition, if the material is dielectric, the global electric displacement is zero, i.e. there are no free electric charges. Hence, Eq. 1 is simplified as follows: . According to the characteristics of the system, we can approximate this electric field to that of a parallel plate capacitor (gap: h), so the voltage is V = Eh. In addition, if an average tension T is applied on the area A of the capacitor during a short period, it can be associated to a constant force of value F = TA. Thus, we can rewrite Eq. 2 as follows: . As the voltage generated by the piezoelectric system is proportional to the applied force, we can consider F as the average force of the corresponding momentum (the impact on the piezoelectric), thus we can approximate it by the average linear momentum change Δp in the time interval Δt: . In our experimental case, the impetus corresponds to the impact of a body of mass m in free fall from a height x. In fact, the momentum change is: Δp = mΔv = m�2 , (g = 9.78 m/s 2 ). Finally, Eq. 2 becomes: Eq. 3 represents the theoretical reference used to evaluate our experimental characterization. Indeed, in order to obtain the experimental value of d, a voltage study was performed as a function of: 1) the distance between the Al-sheets, 2) the condenser area A, and 3) the initial height of the projectile in free fall. Although Δt is neither a constant nor a controllable parameter for each impact, we calculate the average value for each case. Materials and methods. The experimental device was made-up from several simple and cheap elements, as it is shown in Fig. 2. Plastic capsules (parts of candies and toys by Ferrero and Kinder-Surprise) used as projectiles were filled with lead filings and drooped on the center of the sensor. The piezoelectric transducer was composed by a square core of Mylar (dimensions: 3x3 to 21x21 cm 2 ) covered on both sides with aluminum foil sheets (commonly used in food packaging) and adhered with double-face tape to ensure a steady and homogeneous blending; the tape also prevents electrical contact between Al-sheets. The metallic surfaces were connected to thin Cu wires of negligible impedance and to a BNC cable connected the transductor to send the signals to a digital oscilloscope (Tektronix TDS-210). Furthermore, the number of Mylar layers was varied in order to register any variation on the voltage responses. The aluminum foils provided an easy handling of the condenser, as well as the possibility of building sensors of different areas and geometries. However, this material also leads to some disadvantages since it is prone to tearing after some impacts; thus each event was carefully carried out. The experimental implementation of our device intends to determine the scope and validity of Eq. 3, as well as the accurate evaluation of the piezoelectric deformation coefficient d. In this way, by modifying a single variable and controlling the others, it is possible to obtain voltage trends in terms of: 1) the height of the free fall, 2) the number of Mylar plates, and 3) the area of the capacitor. In order to reduce the stochastic error; for each case, we obtained the mean of ten measurements. Finally, in order to ensure that the impact of the projectile happened at nearly the same point, a thread was vertically suspended as a track. Results and Discussion. The acquired data was adjusted via linear functions and the d coefficient was obtained from the slope value analyses. It is also important to mention that the time interval Δt of the impact was variable, within an average time range of 0.5-15 ms. Fig. 3 exhibits a typical electric signal recorded from the impact of a projectile on the transducer. The oscilloscope shows a variable voltage or pulse profile which develops over a period of 15 ms. By means of the automatic average command of the digital oscilloscope we obtained a voltage value which is representative and suitable to use in the Eq. 3; as explained before. Nevertheless, future work should take advantage of the dynamic character of these signals to deepen into the mechanics and physical meaning of the impact. Fig. 4 shows the obtained experimental data (voltage as a function of the square root of the height x) for three different sensor areas: 5x5, 7x7 and 10x10 cm 2 . It was observed that stronger signals (with respect to the background noise) were produced by implementing two plates. The worst theoretical fitting gave rise to a correlation coefficients of V5×5=0.460√ +0.058, R 2 =0.933. Fig. 5 shows the dependence of voltage with the number of Mylar plates for sensors of 5x5, 7x7 and 10x10 cm 2 . In these experiments, the projectile was dropped from a constant height of 1 m in free fall (the average thickness of the transducers was 43 μm). In this case, the voltage signal increases almost linearly with the number of plates. At first, we expected to observe a saturation effect as a result of attenuation occurring in the inner layers of the device due to Mylar deformations. A possible IOP Conf. Series: Journal of Physics: Conf. Series 1221 (2019) 012059 IOP Publishing doi:10.1088/1742-6596/1221/1/012059 5 explanation for the observed trend is that in a parallel plate capacitor, the accumulated voltage is directly proportional to the separation between plates, as described Eq. 3. This suggests that the thicker the Mylar stacks, the stronger the output signal. In this case, the worst theoretical fitting gave rise to a correlation coefficients of V10×10=0.382N+0.175, R 2 =0.973. Fig. 6 shows the sensor area dependent voltage signals obtained for free fall experiments performed at a fixed height. It is observed that the experimental data nearly satisfies the theoretical description given by Eq. 3. Here, the linear fitting gave rise to the following results: V = 0.002/A -0.013, R 2 = 0.997. In general, we can observe that Eq. 3 is adequate to reasonably describe the proposed physical sensor-system. Average piezoelectric constant Finally, from the obtained experimental results, an average value for the piezoelectric constant was obtained, namely d = 42.02 ± 5 x 10 -12 m/V (estimated error: 12%). In the case of PVDF, some reports claim experimental values ranging from 23 to 30×10 -12 m/V [4]. Therefore, our result is congruent in the order of magnitude and very acceptable, at least for didactic scientific applications.	2019-07-20T02:03:43.953Z	2019-06-01T00:00:00.000	{ "year": 2019, "sha1": "ec21be94aa52c294ad878e50485ed96261aa1733", "oa_license": null, "oa_url": "https://doi.org/10.1088/1742-6596/1221/1/012059", "oa_status": "GOLD", "pdf_src": "IOP", "pdf_hash": "2cec655c3431e461068a61fd7553a27b1db5dbc5", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Materials Science" ] }
213629014	pes2o/s2orc	v3-fos-license	Solution-Based Learning Implementation in Design Course Student participation is a key success in almost every learning process. It is believed that concept-level of understanding could be achieved only if students are actively involved in the learning activities. The concept of student centered and active learning is usually adopted in order to improve student participation. Problem Based Learning is one of popular methods used in this paradigm, in which students learn both thinking strategies and domain knowledge of a subject through the experience of problem solving. In addition to this current well-known learning method, this paper introduces a new approach so called Solution Based Learning in which students learn about a subject by analyzing solutions that has been given by some experts. Basically, the concept urges students to analyze the way of thinking of the experts on why they could give the solution like this and that. Students then may adopt the thought or criticize it in a positive manner. In order to see its viability and applicability, this new concept has been implemented in Product Design and Development Course as a case study. Based on the process performance analysis, it is shown that the proposed approach is able to improve student participation, to make class more active and to improve the overall student understanding of the subject of study Introduction Product Design and Development (PDD) Course plays a very important role in enhancing student competency in design and manufacturing. The basic competencies every student should fulfill after completing the course include a comprehensive understanding of industrial product design and a widerange ability to bring the design theory into real implementation. Hence, the success in delivering the course will give a significant contribution to the student's standard competences. Good understanding of PDD concept is a foundation for students to do a final project, as it is a combination of many important preceding subjects (figure 1). Therefore, the ability to identify market opportunity and customer need, to define product specification, to generate product concept, to evaluate and select optimum concept, to create prototype and to do final product evaluation are some skills students should have in doing a project, especially that relates to PDD topics. One of important indicators in evaluating learning process is the level of students' active participation in the class and their average final grade. From previous experience, it is observed that Teacher-Centered Learning (TCL) could not give optimum result. Student participation is very less due to monotonous environment as teaching method is relied only upon the slides in a lecture-based environment. Fundamental principles that significant to teaching and learning include: (1) knowledge is constructed from the experience of the learner; (2) knowledge resides in the mind rather than (3) learning is a personal interpretation of the world in that the learner's beliefs and values are used in interpreting objects and events; (4) learning is an active process of making meaning from experience; (5) learning takes place in contexts relevant to the learner; (6) reflection is an essential part of learning; and (7) learning is a collaborative process in which multiple perspectives are considered [1]. Figure 1. Relationship between Product Design Course and other subjects It is reasonable to say that there is no best method that suitable for all learning situations. A proven good teaching method for a subject is not always good when applied for another subject. Hence, teaching method should be build based on the characteristics of the course, students' motivation and facilities availability. The introduction of new approach of Student Centered Learning (SCL) gives different paradigm and diverse point of view in order to optimize its goal achievement. Combination of many methods in teaching should be a good alternative to a single method. Student Centered Learning The term Student Centered Learning (SCL) is widely used in literature. Many terms have been linked with SCL and therefore the term can mean different things to different people. The notion that unifies them together is that the student is the central entity who must be actively engaged in learning things. Rogers [2] describes SCL as the shift in power from the expert teacher to the student learner, driven by a need for a change in the traditional environment. The power shift paradigm from teacher to the students has emphasized focus on learning rather than on teaching [3,4]. The use of Student Centered Learning concept appears to be thoughtful of today's society. Lea et al. [5] reviewed several studies and found that overall it was an effective approach. In this approach, students felt that there was more respect for the students, it was more interesting and it boosted their confidence. Correspondingly, Hall and Saunders [6] found that students had increased participation, motivation and grades in an Information Technology course where 94% of the students would recommend it to others over the more conventional approach. The implementation SCL is believed to provide numerous benefits in the classroom including a greater depth of understanding of concepts, broader knowledge base, improved communication and interpersonal/social skills, enhanced leadership skills, increased creativity, and improved writing skills. There are many types of SCL, including project-based learning and problem-based learning. Project Based Learning is a comprehensive viewpoint focused on teaching by involving students in investigation. Students pursue solutions to nontrivial problems by asking and refining questions, debating ideas, making predictions, designing plans and/or experiments, collecting and analyzing data, drawing conclusions, communicating their ideas and findings to others, asking new questions, and creating artifacts [7]. Teacher as facilitator gives a "driving question" to students working as a team to respond to, and then direct them to create an artifact (or artifacts) to present their gained knowledge. Artifacts may include a variety of media such as writings, art, drawings, three-dimensional representations, videos, photography, or technology-based presentations. Problem-based learning assigns students learn about a subject through the experience of problem solving. Students learn both domain knowledge and thinking strategies that can help the students develop flexible knowledge, effective problem solving skills, self-directed learning, effective collaboration skills and intrinsic motivation [8]. Students, working in groups, identify what they already know, what they need to know, and how and where to access new information that may lead to resolution of the problem. Teacher, known as the instructor or tutor plays a significant role to facilitate learning by supporting, guiding, and monitoring the learning process [9]. The tutor needs to build students' confidence, encourage them, while also stretch their understanding. Solution Based Learning: A New Approach Problem-based learning has been implemented widely in many disciplines, mainly in medicine, pharmacy, biology, social, economy, law, as well as engineering. In engineering, however, projectbased learning is more popular due to its unique characteristic (i.e. the creation of artifacts at the final phase). Based on the characteristic of PDD course, project-based learning is considered as a good candidate to be adopted in learning process. However, in order to improve its effectiveness, some modifications might be needed. Presented in this paper is a research on project-based learning modification so-called "solution-based learning" as a contrary to "problem based learning". While problem-based learning aims on student activeness through dealing with a "problem" to be solved, solution-based learning provides students a "solution" to be analyzed. Solution-based learning (SBL) is a new term coined for learning method that encourage students to learn from proven solutions given by the experts (learning from experts). SBL has inherent capacity towards increased emphasis on preset values and measures, clear outcome indicators, and ability to generate meaningful feedback [10]. Therefore, when the outcome of problem-based learning is a proposed solution from student discussion to a problem defined by facilitator, the outcome of solutionbased learning is students' comprehensive understanding about "how and why the solution is like this". Students have also opportunity to examine pros and cons of the solution (that might be outdated already) that can be starting point to make improvements. This method enables students to know and see how the experts found solution of the problems under investigation, and then adopt them to solve similar cases. The example of implementation of the proposed method is shown in figure 2. In order to implement SBL, supporting media should be provided while combination of lecture, group discussion, question-answer, and student presentation as well as information technology practice should be used. The media used in this research includes mechanical products, industrial products and art-jewelry products. The examples of real products as learning media can be found in figure 3. A total of 12 products are available to be chosen by students that have been randomly grouped. Implementation and Discussion The new approach of solution-based learning has been implemented in Product Design and Development course with 60 students registered. The students were grouped into 10 groups with 6 students in each group. Every group took one product as object of investigation. Presentations of all groups were made in order for students share their thought and improve their presentation skill. In addition, discussion and question-answer session were employed so that multi-direction interaction could be made. Learning atmosphere was very joyful and student participation was very high. The classroom scenery can be depicted in figure 4. Student perception on SBL method has been evaluated by using a questionnaire. Two other courses delivered in conventional method were used as comparators in order to see the objectivity. Twenty-two items were questioned by using rating scale (1)(2)(3)(4). In general, it is shown that SBL method perceived better than traditional method. This result is in line with the finding of Chandra [11] and Rowe [12]. Outstanding improvements include (figure 6): -Learning environment is more joyful -The assignment is very helpful -New knowledge is engaged more -Feedback for assignment is managed better -Classroom become more alive In addition to questionnaire-based analysis, Focus Group Discussion (FGD) was also conducted, incorporating 10 representative students from all groups. Some advantages of SBL identified by students include improvements in: -Computer skills related to CAD system and software -Teamwork skills -Time management -Self-evaluation and objectivity It is important to note that many students were identified uploading their project to social media like Facebook and Instagram, and used the media to show their project to public. This indicates that they were happy with the project. Conclusion Solution-based learning, a new approach of Student Centered Learning has been introduced. The concept places students in the center of activity in analyzing a proven solution provided by experts. The outcome of solution-based learning is students' comprehensive understanding about how and why the experts gave that solution of the problem. Opportunity to examine pros and cons of the solutions can be used as starting point to adopt them to solve similar cases or to make new improvements. The concept of solution-based learning has been implemented in Product Design and Development course and it is found that in general, the proposed approach is able to improve student participation, to make class more active and to improve the overall student understanding of the subject.	2020-01-23T09:08:34.844Z	2020-01-21T00:00:00.000	{ "year": 2020, "sha1": "9f83244be78025c472a868c1d59a0b6bdae9ad81", "oa_license": null, "oa_url": "https://doi.org/10.1088/1757-899x/722/1/012067", "oa_status": "GOLD", "pdf_src": "IOP", "pdf_hash": "7766bdafb7c8828f5a3aa990a17a22f1d472e750", "s2fieldsofstudy": [ "Education" ], "extfieldsofstudy": [ "Computer Science" ] }
215975149	pes2o/s2orc	v3-fos-license	Widespread Difficulties and Applications in the Monitoring of Historical Buildings: The Case of the Realm of Venaria Reale : Environmental monitoring represents a key step for restaurateurs to follow who strive to ensure the preservation of buildings and artifacts while allowing for people’s thermal comfort. This paper describes the possibilities and main issues arising from the study of indoor microclimates. The presented case study focuses on the monitoring data analysis for two rooms of the Realm of Venaria Reale, in Turin. The adopted methodology provides for the gathering of knowledge about the history, the geometry, and the change of use in the course of the lifetime of the building. This information allows us to construct a virtual model of the building, through which it is possible to evaluate the past and present and to hypothesize future scenarios regarding the indoor environmental conditions. Moreover, this paper presents a specific index, namely the Heritage Microclimate Risk (HMR), which enables us to evaluate the risk level to which the artifacts kept within historic buildings are exposed. With that database of information, we can preemptively define which actions (managerial and structural) would need to be taken for the preservation of the artifacts and the building itself, avoiding the possible risk component taken by working on the real building. Introduction The focus of this research concerns the indoor microclimates of historical buildings, their modification, and their control through the identification and management of the parameters that characterize them. Specifically, this paper analyzes the opportunities and benefits resulting from the correct management of indoor microclimates, aimed to guarantee a preventive conservation of cultural heritage. Moreover, we will deepen our understanding of the principal issues that are typically faced during this kind of research. Very frequently, studied artifacts and architecture need microclimatic conditions that are not standardized, and most are distinguished by particular system requirements. This underlines how the requirements of preserving cultural heritage are highly dependent on the object's or building's own history: use, as well as architectural, structural, and managerial changes over time. From this point of view, the approach proposed by Marco Pretelli and Kristian Fabbri is effective: they define the concept of Historical Indoor Microclimate (HIM) [1], which takes into account the modifications of the microclimatic characteristics of an indoor area over the years. Indeed, HIM develops a historical vision of architecture from construction to the present day. Moreover, HIM considers the relationship between those changes and the degradation of the artifacts and buildings. Once this relationship is understood, it is possible to define specific strategies aimed at preventive conservation. Case Study The Realm of Venaria Reale is situated in Italy in the province of Turin. It is one of the Sabuda Residences, part of the serial sites that UNESCO has included on the Heritage of Humanity list since 1997. The Realm is marked by the intermingling of building phases, styles, and techniques. It is the result of the intervention of six different architects: Amedeo Castellamonte, Michelangelo Garove, Filippo Juvarra, Benedetto Alfieri, Giuseppe Battista Piacenza, and Carlo Randoni. Not one of them completely realized their own vision of the building, adding to the unfinished character that distinguishes it. The Realm of Venaria Reale and all the artifacts kept within it have undergone changes over the years, such as demolitions, architectural additions, changes in types of use, etc. These stratifications represent the keys to understanding the building and allow for the study of the evolution of different artistic and architectural techniques, as tastes and needs have changed over the years. Today, after the biggest cultural heritage valorization and restoration project undertaken to date to ensure that "the giant became [the] Realm" [2], Venaria Reale became an important cultural and touristic attraction: exhibitions, concerts, shows, and international events are presented there. Currently, the compound is one of the most visited sites in Italy thanks to the beauty of its architecture and of the artifacts presented within it. This confirms the belief that investing in culture is as fundamental as it is strategic. Scientific Literature In the 1970s, scientist and researcher Garry Thompson (1925Thompson ( -2007 became interested in the study of indoor microclimates and preventive conservation, focusing some of his research on the conservation of movable assets hosted in museums [3]. In 1967, he took part in a conference [4] about climatology in museums, during which several studies concerning the analysis of microclimatic parameters (air temperature, relative humidity, lux, etc.) and their impact on heritage were presented. In 1972, UNESCO, with the aim to ensure the best conservation of heritage and to support international cooperation, subdivided the concept of heritage into three groups: monuments, groups of buildings, and sites. For each typology, UNESCO recommends integrated protection, balancing control, and conservation. The gradual affirmation of awareness of the significant impact that environmental conditions have on heritage conservation provoked various debates, not only about the concept of restoration and conservation but also about standards and indexes. As a result, the American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) and other institutions defined and readjusted many times their guidelines for managing the microclimate of museums, libraries, archives, etc. [5][6][7][8]. Moreover, institutions such as the International Council of Museums (ICOM) and the International Centre for the Study of the Preservation and Restoration of Cultural Property (ICCROM) historically operate in this field, in terms of awareness and research. Different studies conducted by researchers who are affiliated with these institutions underline the importance of environmental monitoring with the aim to reduce material degradation in favor of preventive conservation. Gael De Guichez-a chemical engineer and member of ICCROM, who in the second half of the twentieth century proposed specific methodologies for museums-exemplifies this trend. In the 1980s, indoor monitoring was expanded to the outdoors, which hosts important architectonical monuments. Other publications by Dario Camuffo [9][10][11] offer relevant scientific contributions to this field. Some authors also wrote about Indoor Microclimate Quality (IMQ) [12], which is not defined by specific standards and is easily confused with Indoor Environmental Quality (IEQ) [13,14], which is defined by standards for new buildings (EN 15251). Finally, recent research, deepening our understanding of the field of microclimates, mostly focuses on museums [15][16][17][18][19][20][21][22][23][24][25][26], evaluating models and strategies able to establish the conditions of artifacts (e.g., evaluation protocols). Other researchers use building simulations, as we did for the study of Venaria Reale, to study the future impact of climate change [27] or to analyze hygrothermal conditions of museum storage [28]. Nevertheless, the scientific literature offers numerous case studies where several different approaches have been adopted. The standards considered are not always the same and neither are the methodologies, the monitoring instruments, software, etc. This makes the possibility of reaching a unique and common vision of the action necessary to achieve preventive conservation difficult, as well as making the replicability of a shared strategy to apply to multiple similar case studies challenging. As mentioned in the introduction, the approach of this research follows the concept of HIM [29][30][31] and highlights the difficulties and the usefulness of studying the indoor microclimate for preventive conservation. Study Objective The maintenance of a building largely depends on its microclimatic conditions. We propose in this study a preventive conservation of architecture-in this case the Realm of Venaria Reale-and of artifacts having historic and artistic value through the understanding and management of the indoor microclimate. It is clearly fundamental to identify choices and paths that allow us to preserve, conserve, and make available the cultural heritage. Unfortunately, combining the two antithetical purposes of enjoying and conserving is complicated. To do this, an approach that is both transversal and interdisciplinary is needed. The science of prevention and conservation is in fact a subject that is particularly complex, because it includes different actors from several areas, for example, from technical physics to architecture, chemistry, and environmental engineering. The aforementioned multidisciplinarity might be one of the reasons why today it is not yet completely clear what a "suitable environment" means (let alone an "environment"). Therefore, to establish the line of action necessary for a preventive conservation of cultural heritage, there is no simple or immediate solution. The study of the indoor microclimate in the Realm of Venaria Reale presented in this paper is an opportunity to explain many reasons why there is a strong need today for a conscientious management of cultural heritage and how imperative it is considering all the difficulties in this field. Indeed, to manage a historic building, for example, we have to guarantee its sustainability, preservation, and energy efficiency, making both its preventive conservation and accessibility possible. To do this successfully, we need to know both the problems and strengths of the building and of the indoor microclimate control process. Materials and Methods The research methodology is structured has follows: It is not always necessary (or possible) to complete all these steps, and it is not even mandatory to do them in this order, but one of the strong points of the study of indoor microclimate is that all the steps of this method give us good information, allowing us to further expand our knowledge of the environmental conditions in the building and thus of the artifacts' conservation and human comfort. Archive Search and On-Site Visits This step is part of the acknowledgement of the case study and consists in finding useful material which primarily allows gathering information about the building's geometry. During this step, it is possible to find bibliographical and archival material about the phases of construction, for example, but also of its structure, and information about the use of spaces, visitors' paths, and opening hours, etc. In the case of the Realm of Venaria Reale, we collected much information by talking with some architects who worked on the restoration of the complex and the Centre of Conservation and Restoration (CCR) as part of the Venaria Reale team. This way, it was possible to obtain some sections, layouts, bibliographic advice, etc., for the Realm. Monitoring Campaign Currently, since the end of the restoration in 2007, there is a monitoring system inside the Realm consisting of repeaters and probes that send data wirelessly to any device. In this case, the monitoring campaign was not performed, but the result was positive. We acquired ten years of monitoring data for the two rooms studied from the CCR, that is, from 7 August 2007 to 3 August 2011 for Room 38, and to 16 March 2017 for Room 33. Construction of a Virtual Building Model and Its Validation Having obtained the information about the geometry of the model, the outdoor climate data, the building's materials, the Heating, Ventilation and Air-Conditioning (HVAC) system operations, the opening hours, etc., we were able to produce a 3D model of the case study. For the case presented in this paper ,we made the first 3D model using the SketchUp program (version 20.0 Trimble Inc., Sunnyvale, CA, USA) and, using a plug-in, we uploaded that model to IES.VE (Virtual Environment by Integrated Environmental Solutions), a dynamic simulation software that makes it possible to elaborate information about energy use, light levels, occupant comfort, CO2 emissions, airflow, etc., generating data, images, and videos. To be validated, data produced by the virtual model in the current scenario must be compared to the real monitored data, and these have to respect the parameters reported by guideline 14 of ASHRAE. The validation parameters are as follows: Validated if R 2 > 0.5. Building Simulation The building simulation allows seeing the current conditions of the indoor microclimate of the building, but it is possible to hypothesize different scenarios, such as changing the building's material, or the outdoor climate, or also the doors' and windows' opening or closing. This allows verifying the consequences of several possible modifications to the building using a virtual model without the risk of damaging the actual site. HMR Calculation The Heritage Microclimate Risk is a specific index with which you can estimate the aggressiveness of the microclimate against a specific material or artifact. It is calculated as follows: and regarding the single variable (x): where mrh is the hourly microclimate risk of the reference period; h represents the total hours of the reference period. In the case presented in this paper, the reference period corresponds to the monitoring campaign. The hourly microclimate risk (mrh) is determined by the following: where hr(x) is the heritage risk of the microclimatic variable (x); hr(x,set),min is the heritage risk of the microclimatic variable (x) with the minimum set-point, defined as the lower range established by the Standard (UNI 10829, 1999) or other guidelines; hr(x,set),max is the heritage risk of the microclimatic variable (x) with the maximum set-point defined as the lower range established by the Standard (UNI 10829, 1999) or other guidelines; n is the number of hours of the reference period. Results The two rooms studied in the Venaria Reale are in a partition of the building, next to the "Galleria Grande" on the east side. The images below (Figure 1) show on the left (a) a rendering of the Realm realized by 3D Studio Max; on the right (b) the part of the building where the two rooms are located. This last image was designed in SketchUp. The blue lines show the partition of the building, which is shown in the image on the left (a), and the red rectangles show the rooms analyzed, which is shown in the image on the right (b). As already mentioned, it was possible to acquire much monitoring data-Temperature (T) and Relative Humidity (RH)-from the CCR, dating from 2007. The HVAC system has been operational the whole time. Below are some illustrative data resulting from an analysis of the probes' indoor microclimatic data. The characteristics of the current microclimate, the interpretation of the data, their relevance, and the difficulties encountered are discussed here. The T and RH ranges in both rooms are between 15 °C and 28 °C, and 20% and 80%, respectively. The following chart ( Figure 2) shows that the values of T have been recorded more frequently in Room 38. Then, if and how these values could be dangerous for the materials conservation is discussed below. The frequency distribution and cumulated frequency of the indoor temperature in Room 38 provide important information, that is, the temperatures higher than 25 °C indicate a risk of overheating the materials in the room. Moreover, considering the artifacts hosted in this room ( Figure 3), it is especially important to monitor the wood's conditions because it dries out more easily than other materials. Data from both rooms include many gaps. Some are due to conscious choices made by the staff (e.g., breaks for annual periods such as from 30 August 2011 to 24 April 2013), while other shorter gaps might be caused by technical malfunctions of the probes (e.g., dead batteries or problems with the Wi-Fi connection). As mentioned above, we acquired ten years of monitoring data for the two rooms studied from the CCR, that is, from 7 August 2007 to 3 August 2011 for Room 38, and to 16 March 2017 for Room 33. We standardized all these data and, by taking the average year by year and hour by hour, it was possible to obtain a "year-type". These data (Figure 3) are fundamental for the validation (Table 1) of the virtual environmental simulation. Regarding the possibility of simulating future scenarios using the virtual environmental simulation, we assumed that in 2100 the outdoor climate will rise by 1.5 °C. We simulated this hypothetical scenario to evaluate what consequences it would have on the indoor temperature and relative humidity, and therefore on the conservation of artifacts and the building. The results show that, if the HVAC is operational, no serious modifications or damage will occur due to the indoor microclimate ( Figure 5). Indeed, the mean difference between the indoor temperatures calculated on a whole year is 0.28 °C, with peaks of 0.8 °C. Considering an operational HVAC system, if there was a 1.5 °C increase in temperature, we could assume a reduction in heating costs. On the contrary, if we consider that the HVAC system is switched off, we would have an annual mean difference of the indoor temperature of 1.45 °C, with a minimum peak of 0.69 °C and a maximum peak of 1.58 °C ( Figure 6). We must view these data as a wake-up call-without an operational HVAC system, the effects of climate change could expose the conservation of the artifacts hosted in Venaria Reale to some risk. As shown in Table 2, the maximum gap between the relative humidity recorded in 2017 and the simulated humidity in a hypothetical 2100 occurs when the HVAC system is active. If we consider the HVAC system being switched off, we see the biggest gap in the temperature between these two scenarios. To evaluate how the increase in T could affect the materials hosted inside the room, we calculated and compared the HMR for different materials in both scenarios, namely 2017 and 2100 (Table 3). We can see that the HMR in 2100 will be more aggressive by +1% and +3%. Table 3 can be summarized as follows: HVAC on: HMRT 2100 < 0.01 HMRT 2017; HMRRH2100 < 0.02 HMRRH 2017. HVAC off: HMRT 2100 > 0.002 HMRT 2017; HMRRH2100 > 0.03 HMRRH 2017. At this point, it would be useful to be able to evaluate the aggressiveness of the microclimate on the artifacts and the related variations due to climate change. To do that, we defined a proposal of a specific index which allows estimating a percentage level of "aggressiveness" that microclimate factors can have on materials, namely the Heritage Microclimate Risk (HMR). It is not a global or multicriterial index; indeed, it evaluates each parameter separately, taking into account specific standardized ranges per physical variable, such as material, fresco, painting, etc. HMR is useful for making decisions about the placement or borrowing of artifacts because it allows us to verify if the conditions of a specific place-for a temporary period or definitively-are suitable or not for its conservation. In this context, and in accordance with the aim of this paper, we must also highlight a difficulty, namely the heterogeneity of standards and indexes. Considering what is reported in UNI 10829 and MIBACT for the categories all materials, wood, oil on canvas, and fabric (silk), we can see that the range established by the standards are rather stable for the temperature parameter, whereas they vary more for the relative humidity. Indeed, the HMR value calculated considering the category "all materials" is very low. Calculating the same microclimate risk for the individual materials hosted and exhibited in Room 33, we obtain higher values. Figure 7 shows four risk envelopes. The first comprises "all materials", with a low risk. The risk then rises, varying according to the specific material ranges. Moreover, there are other problems with this kind of survey in addition to the ones already listed, including the retrieval of information about the building's materials, the history of the building, and the interactions with curators and stakeholders. Discussion The possibility of understanding a building through the study of its indoor microclimatic conditions makes it possible to define how to avoid or minimize possible risk situations in a timely manner. This is certainly a desirable preventive approach to conservation problems, as recommendable as it is rarely practiced. When the aim is just to obtain better conditions for the thermal comfort of visitors, making practical choices is simple, even if the concept of comfort has changed over the years. Today we do not accept that it is too cold or hot in an indoor space, and much confidence is placed in the HVAC systems. Moreover, in the last 30 to 40 years, the role of many heritage buildings has changed significantly. Often, the original use of a heritage building can change and a private home can be converted into a museum, for example. Instead of being used to host exhibitions, research, and cataloguing, a building is today increasingly used as a space that is more flexible and dynamic, keeping up with current economic, social, and tourism needs. In fact, it is not surprising today to find cafés, restaurants, and workshops inside a historical museum, and this makes the management of such historical architecture more complicated. Therefore, when we talk about preventive conservation. we must consider three cardinal elements, establishing degrees of priority depending on the case-the building, the artifacts hosted within it, and the people who use that space. As seen in this paper, the study of the indoor microclimate associated with the virtual building simulation and the calculation of HMR makes it possible to know, in advance, the consequences that many decisions and actions could have on conservation and heritage buildings, as well as on new construction projects. Conclusions This paper proposes seriously considering the importance of studying indoor microclimates. It provides key information about the conditions of preserving cultural heritage. Indeed, it makes it possible to evaluate the quality of the environment for people, artifacts, and the building itself, proving to be a valid reference for restoration projects, management interventions, and prevention in historic buildings and for choices to be made about the availability to visitors of the cultural heritage. In this study, the intangible-the air-becomes tangible, because it allows making decisions about use and conservation. Moreover, the comparison between the indoor conditions of Rooms 33 and 38 in 2017 and 2100, assuming that climate change will cause a rise in outdoor air temperatures of 1.5 °C, shows how a building simulation associated with the study of the indoor microclimate can be used to understand hypothetical future scenarios, thereby allowing interventions in time, with the goal of putting preventive conservation in place. Therefore, understanding and managing the indoor microclimate represent a fundamental tool for architects, helping them to make timely preventive decisions. Author Contributions: All the authors contributed equally to this work with regard to all statements. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding.	2020-03-12T10:31:47.971Z	2020-03-10T00:00:00.000	{ "year": 2020, "sha1": "0dcd3181a3c26af3d8e4ca3426a288656788ef39", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2571-9408/3/1/8/pdf", "oa_status": "GOLD", "pdf_src": "ScienceParsePlus", "pdf_hash": "45cafbe9a5b2259c1f3730d5088d8319fd721aec", "s2fieldsofstudy": [ "Environmental Science", "History" ], "extfieldsofstudy": [ "Computer Science" ] }
237452478	pes2o/s2orc	v3-fos-license	Rainbow-link: Beam-Alignment-Free and Grant-Free mmW Multiple Access using True-Time-Delay Array The millimeter-wave (mmW) communications is a key enabling technology in 5G to provide ultra-high throughput. Current mmW technologies rely on analog phased arrays to realize beamforming gain and overcome high path loss. However, due to a limited number of simultaneous beams that can be created with analog/hybrid phased antenna arrays, the overheads of beam training and beam scheduling become a bottleneck for emerging networks that need to support a large number of users and low latency applications. This paper introduces rainbow-link, a novel multiple access protocol, that can achieve low latency and massive connectivity by exploiting wide bandwidth at mmW frequencies and novel analog true-time-delay array architecture with frequency dependent beamforming capability. In the proposed design, the network infrastructure is equipped with the true-time-delay array to simultaneously steer different frequency resource blocks towards distinct directions covering the entire cell sector. Users or devices, equipped with a narrowband receiver and either a single antenna or small phased antenna array, connect to the network based on their angular positions by selecting frequency resources within their rainbow beam allocation. Rainbow-link is combined with a contention-based grant-free access to eliminate the explicit beam training and user scheduling. The proposed design and analysis show that rainbow-link grant-free access is a potential candidate for latency-critical use cases within massive connectivity. Our results show that, given less than 1e-5 probability of packet loss, a rainbow-link cell, over 1 GHz bandwidth using 64 element antenna array, attains sub-millisecond user-plane latency and Mbps user rates with an approximate 400m line-of-sight coverage and a density of up to 5 active single antenna users per second per meter square. power consumption and cost [2], [3]. These stringent requirements coincide with all three use categories in 5G specified as enhanced mobile broadband (eMBB), ultra-reliable lowlatency communication (URLLC) and massive machine-type communication (mMTC). Specifically, the eMBB is tailored for high peak rate and throughput. General requirements for URLLC are sub 1ms user plane latency and packet error rate as low as 10 −5 [4]. For mMTC, the general requirement, as suggested by its name, is to provide massive wireless connectivity (e.g. beyond 10 5 links) to machine-type devices in a given area [5]. The millimeter-wave (mmW) communications is a key technology in 5G. Due to abundant bandwidth and antenna array beamforming, mmW communications is the enabler of eMBB. It is also envisioned as a promising candidate for the future high-end IWSN [6]. However, the current mmW solutions have many disadvantages in terms of latency, connection density, power and cost for the industry internet of things (IoT) applications. Firstly, mmW systems rely on beamforming to overcome severe propagation loss. To keep reasonable cost and power consumption, current systems utilize phased antenna arrays at both base station (BS) and user equipment (UE) for beamforming. The overhead associated with beam alignment and beam scheduling using analog arrays is a latency bottleneck. Secondly, the analog phased antenna arrays can connect only a limited number of devices based on the number of beams and the number of radio frequency (RF) chains. Each beamformed transmission occupies the entire bandwidth, therefore frequency domain multiple access cannot be supported. As a result, phased antenna array based mmW networks cannot support a high number of connected devices. Lastly, cost and power consumption of wide-band mmW devices are high which presents a major limitation for IoT applications. Meeting diversified service requirements with reasonable cost and power consumption is inherently hard. Specifically, to tackle URLLC requirement, key 5G NR features such as grantfree multiple access must be employed [7]. In this regime, the network capacity is often limited by insufficient radio resource elements. [8] evaluates URLLC use cases with 10MHz band at 4GHz carrier frequency where the capacity attains 10 users per cell. Alternatively, in mmW band large unlicensed band resources are available. However, due to the need of beamforming, these resources cannot be shared between a large number of spatially separated users [9]. To Fig. 1: Due to narrow angular coverage of analog beams, BS with phased array can only serve a limited number of scheduled users. On the other hand, TTD array can connect with users in the entire spatial sector without any scheduling. With phased array, the entire band is pointed to a certain direction, while with TTD array, each beam is pointed to a specific frequency subband (represented by colors). reducing the initial access and beam alignment latency [10]. Latency due to resource scheduling among multiple users is highlighted by work [11], where optimized frame structure and fully digital arrays were proposed as a solution. The ability of a digital array to connect with users in all directions is useful, however its power consumption and cost is overwhelmingly high. Therefore, to address aforementioned weaknesses of existing solutions for mMTC and URLLC, rethinking of array architectures and multiple access schemes is required. In this work, we leverage a new beamforming technique referred to as rainbow beam [12], that is enabled by a novel True Time Delay (TTD) array architecture [13]. Rainbow beams exhibit beam patterns that cover the entire angular space by uniquely mapping different frequencies, e.g. orthogonal frequency-division multiplexing (OFDM) subcarrier (SC), onto specific directions. As indicated in Figure 1, rainbow beams are particularly attractive in wideband systems as different SCs and corresponding beamforming directions can be allocated to a large number of users. As long as the users select their appropriate frequency resources, no beam training or scheduling is needed. Based on this feature, in our design a grant free mmW multiple access protocol is combined with rainbow beamforming. It has a potential to provide low user-plane latency for a massive amount of users in IoT applications. The idea of utilizing frequency dependent beamforming to enhance coverage of directional beams (therefore provide multiple access) has been studied in sub-terahertz communication system [14], [15], [16]. However, previous studies largely considered broadband, high-rate communications, as opposed to IoT use cases. Differing from most mmW networks, in our design the multiple access based on rainbow beams is regarded as an enabling solution for latency-critical mMTC [17]. On one hand it provides wide angular beam coverage in a power efficient manner by spatially spreading SCs; on the other hand it operates in mmW band where sufficient bandwidth resources are available to reduce system overhead. Our main contributions are the following: • We propose a rainbow-beam based frequency domain multiple access which does not require explicit beam-training and supports grant-free random access. • We propose a detailed design for radio interface, downlink (DL) synchronization procedure, and random access protocol. • We analyze synchronization performance and probability of collision (packet loss) and their impact on latency for given network design parameters. 1 • Our analytical study is supported by simulations that evaluate the proposed multiple access scheme in terms of latency, reliability and effective rates. The rest of the paper is organized as follows. In Section II, we introduce the system model. The proposed rainbow link protocol and radio interface design are included in Section III. The analysis of DL synchronization and uplink (UL) grant-free transmission are presented in Section IV. Section V presents simulation results. Discussion and future work suggestions are in Section VI. The paper is concluded in Section VII. Notations: Scalars, vectors, and matrices are denoted by non-bold, bold lower-case, and bold upper-case letters, respectively, e.g. h, h and H. The element in i-th row and jth column in matrix H is denoted by [H] i,j . Transpose and Hermitian transpose are denoted by (.) T and (.) H , respectively. The l 2 -norm of a vector h is denoted by \|\|h\|\|. diag(A) aligns diagonal elements of A into a vector, and diag(a) aligns vector a into a diagonal matrix. The i-th element in set S is denoted as [S] i . II. SYSTEM MODEL In this section, we introduce the system model of mmW communication with TTD beamforming. All important notations are summarized in Table I. UL#2 ... The network consists of a massive number of quasi-static mmW UEs that are served by a BS. We assume each UE, a machine type device, is equipped with a phased array with N U antenna. Furthermore, we assume each UE has much narrower bandwidth than the BS to further reduce power and cost. As a result, each user can only access a subset of SCs through orthogonal frequency-division multiple access (OFDMA). The indices of those SCs accessible by user u are denoted by set B u whose cardinality satisfies \|B u \| B, i.e., the SCs that a user can access is much smaller than the total number of SCs in the broadband. In practical OFDM based system, adjacent SCs are typically bundled into blocks for frequency resource utilization. This mechanism is controlled by g group which denotes how many SCs are consolidated into a resource block (RB). To be clear on notations, we use \|B u \| for the number of SCs and K for the number of RBs in B u , i.e., K = \|B u \| /g group . Proposed frame structure in rainbow link We assume the BS has a linear array with N B elements. The antennas are critically spaced, i.e., half of wavelength that associates with f c . In this work we focus on a sparse geometric channel of M multipath components 2 , where the angle of arrival (AoA), angle of departure (AoD), and complex path gain of the u-th UE and m-th multipath component are denoted as θ u,m , φ u,m , and g u,m , respectively. θ u,m and φ u,m are assumed to be uniform randomly distributed in region −π/2 to π/2. The channel H u,b , for u-th user and b-th SC is where α u,b,m = g u,m exp(j2πbτ u,m BW/B) is the SC-wise complex gain, with g u,m and τ u,m representing the multipath gain and delay with respect to the first antenna element as a reference. We denote a B (φ) and a U (θ) as the narrowband 3 array response vectors, i.e.,[a U (θ)] n = exp [jπ(n − 1) sin(θ)] and [a B (φ)] n = exp [jπ(n − 1) sin(φ)]. Amplitude of g u,m is characterized by the free space path loss model [18]. In general, channel with multiple taps would be frequency selective. However, since all UEs operate over a small portion 2 For mmW communication, typically M ≤ 4 3 The narrowband array response model holds true when the propagation delay across the array aperture is less than the sampling duration [12]. For typical mmW BS array aperture ≤ 0.32m, the propagation delay across the aperture is up to 1 ns, which is less than 2.5ns, the sampling duration of a 400MHz system. of broadband bandwidth, we assume the channel is approximately flat. We also assume the BS is equipped with a reconfigurable TTD array with a single RF chain for its transceiver as shown in Figure 2. In our previous work, we have shown that TTD array can realize frequency dependent beams so that each OFDM SC is mapped to a particular beamforming direction [12]. Delay taps in the array are set to be uniformly spaced with inter-element delay spacing of ∆τ , regardless of whether the delays are introduced in baseband or RF. The full-range rainbow beam 4 can be achieved with tap ∆τ = 1/BW such that the equivalent analog combiner for the b-th SC is given by Inspecting (2), it is clear that beamformer w b steers SC b onto a unique directions with respect to the rest of SCs. On the other hand, any steering vector, i.e., a B (φ) for an arbitrary φ, is also synthesized at a certain SC. Hence it is referred to as full coverage rainbow beamforming. For a certain user, we define b u as the index of SC whose encoded spatial direction is the closest to the AoA θ u of a certain path. This particular SC is also denoted as anchor SC. In other words, the beamforming gain at anchor SC is the highest among all subcarriers. To better present our proposed network protocol, we briefly introduce the signal model in the DL and UL. In the DL, we denote the transmit OFDM symbol at the b-th SC as S (DL) [b]. Given an ideal bandlimited filter at the UE defined by its narrowband B u , the received frequency domain signal is In the above equation, v u represents the precoder of the u-th user (which is not frequency dependent) and scalar β u,b characterizes the post-beamforming gain of the b-th SC. z (DL) [b] refers to the noise at the b-th SC. . The received signal at the BS in the UL is expressed as where the post-beam channel gain in the UL β * u,b is the conjugate of the one in the DL as long as the same SC and UE-side beamformer v u is used. Since each SC is mapped to a spatial direction, UE can measure the receive power on SCs during DL boardcasting and identify the segment of SCs with the highest received power as the segment aligned with its AoA. Note that although the channel gain is assumed to be flat across the narrowband B u , post-beamforming gain β u,b is still frequency dependent (b dependent) due to TTD beamforming precoder w b at the BS side. This unique feature of rainbow-beam with TTD array causes beamforming gain loss on different SCs. Even if a UE has broadband sampling capability, there will only be certain SCs that it can use. This design consideration is addressed in Section IV . III. RAINBOW LINK MEDIUM ACCESS CONTROL In a conventional mmW UL access, a 4-step random access procedure is employed. The BS broadcasts synchronization and beam training pilots during DL. The UEs independently conduct synchronization, measure the received signal strength of BS's beams, and send feedback through the random access channel. After the BS receives feedback, it applies user scheduling and resource allocation scheme by reserving dedicated time-frequency resources for UEs before sending them access grant. UEs then use the scheduled resources to complete UL transmission. In this scheme a non-trivial latency is expected not only in the grant request, but also in the resource scheduling because the served UEs in each time slot are limited to a small angular region covered by a narrow analog beam from the BS. The rainbow link on the other hand, leverages the frequencydependent beam steering capability of TTD array which can simultaneously connect multiple UEs. We propose to use a fixed beam configuration given by beamforming precoder/combiner in (2) for both DL synchronization and UL access, i.e., on the BS's side no beam switching is required. Due to large coverage of rainbow beam in angular domain, UEs always have beamforming gain to and from BS without explicit beamforming training or feedback. Elimination of beamforming feedback significantly cuts the overhead on beam maintenance and resource scheduling which then makes the 2-step contention based random access possible. This section presents the proposed protocol for multiple access with rainbow link. We start by discussing details of DL synchronization and UL grant-free transmission. Based on the discussion, we then provide the design of radio interface as well as relevant performance metrics that are analyzed in Sections IV and V. A. Rainbow beam DL synchronization Given a wideband OFDM transmission from the BS, the synchronization of the narrowband UEs require estimation of the correct timing offset and fast Fourier transform (FFT) window for decoding. Effectively, a narrowband receiver at the UE side needs to select a set of SCs B u that are then used for UL transmission. Since rainbow beam maps each SC to a specific spatial direction, UEs should select SCs that are mapped to its AoD so that they leverage the maximum beamforming gain from BS and improve DL for the detection of the synchronization signal. Due to narrowband receiver processing at the UE, the synchronization involves following challenges: • UEs should synchronize using a received signal that contains only a fraction of the entire OFDM preamble. • UEs needs to locate (in the frequency domain) a segment of SCs with the minimum beamforming gain loss as its operating band B u . To tackle the first challenge, we propose to load synchronization sequences on all SCs. Since the AoD of a UE in DL channel might land on any SC across the wideband, we employ pseudorandom noise (PN) sequences instead of Zadoff-Chu sequences that are used in 4G Narrowband Internet of Things (NB-IoT). PN sequences have a low peak to average power ratio, zero auto-correlation with its time-shifted version, and do not require fixed narrowband reception for completeness [20]. The preamble sequence is assumed to be known at UEs. The second challenge is essentially dependent on the UE's ability to maintain beam alignment with the BS. With no prior information about which segment it should select, a UE would have to traverse the entire broadband signal and repeatedly search for b u , the anchor SC with the highest beamforming gain. Undoubtedly, traversing the entire broadband frequencies (with low sampling rate) would cause prohibitive overhead. We assume that UEs are quasi-static and thus the AoDs change slowly. In this case, using the previously selected segment as a prior, UEs can keep track of changes in AoDs. This simplification eliminates the extra overhead of UEs searching for b u . As illustrated in Figure 2, synchronization sequences are loaded onto different SCs in the DL broadcast. The DL synchronization signal contains frequency multiplexed narrowband beacons 5 . In order to participate in contention UEs are required to successfully synchronize. The synchronization algorithm at the UE involves the following steps: • Perform down conversion by multiplying the broadband signal with exp (−1j2πtb u /B), filtering, and sampling to obtain a narrowband baseband signal. b u here refers to the anchor SC from the previous connection. • Locate the correct FFT window by estimating the integer and fractional timing offset. • Identify the new anchor SC b u with the highest signal strength. The updated B u for UL transmission is determined accordingly. B. Rainbow Beam UL Grant-free Transmission The proposed work cycles and the corresponding timing diagram of the UL grant-free protocol is shown in Figure 3. When their packets are generated, each UE activates and synchronizes in the nearest DL time frame where BS broadcasts syn-chronization signal, after which UE transmits in an arrive-andgo manner. Namely, UE does not send a scheduling request for access grant as in grant-based access. For the UL transmission it utilizes multiple RBs in the narrow band centered 6 at the detected anchor SC, i.e., UEs are sending packets that contain preamble and data so that BS can extract UE's identity and other necessary control information. Naturally in a grant-free access, multiple packets transmitted in the same time-frequency resource block from different UEs would cause a collision and packet loss. For the sake of tractable analysis, we assume decoding on SCs without collision always succeeds, and decoding on SCs with collision always fails. In order to improve the packet loss rate, we propose a simple repetition coding. With abundant time-frequency resource blocks in mmW spectrum, UEs can perform repetitive transmissions on multiple resource blocks [21]. In our proposed scheme, each UE repeats its transmission on n RBs in its narrow band of K RBs. The n replicas offer repetition coding and diversity gain to reduce packet loss. Making a large number of repetitions alleviates the loss of packets for a certain user. But the contention also becomes harsh when other UEs do so. In this work we assume a fixed number of repetition n for all users and study the optimal design of n. Notice that the optimal n will also be influenced by the notion of grouping. When multiple SCs are consolidated as a RB, UEs load more useful symbols into a packet but have less resources to avoid collisions. We will analyze these trade offs in Section IV. C. Radio Interface Design In order to analyze rates, coverage, and latency of the proposed system, this section briefly discusses radio interface design. Due to its popularity, reliability, and flexibility for contention based grant-free random access [23], we adopt a frame slotted Aloha. Both UL and DL transmissions are slotted in terms of frames 7 and users can transmit without grant request or allocated resources from BS. Critical specifications in radio interface for the proposed system include [25]: • Subcarrier spacing and OFDM symbol duration. • Cyclic prefix to accommodate combined channel delay spread and the largest delay tap of TTD array In URLLC and mMTC use cases, short packets are commonly used [26], [27]. In that sense, large spacing between SCs shall be considered. A natural choice is then to adopt a numerology scaled from the current 5G and 4G networks. As pointed out by [28], in current standards, SC separation (SCS) is of the form: 2 k × 15kHz for integer k. Specifications with k > 2 cannot be implemented in sub-6GHz spectrum. To get 6 From a geometric argument, putting the anchor SC in the center of Bu intuitively ensures that there is good channel gain. 7 The propose protocol utilizes subset of time-frequency resource, similar to the interleaved NB-IoT resource units with LTE [24], and we focus on relevant UL/DL only. The uplink grant-free access with proactive repetition scheme similar to [22], [23]. The abbreviation P, Cont, ACK, and Sync denote packet arrival, contention attempts, acknowledgement (ACK) and synchronization attempts, respectively. A short guard period counts for propagation and processing delay. In the diagram, UE succeeded in the first synchronization attempt and the second contention attempt a rough understanding of the system, we then use the k = 5 case as listed in Table II. Here the length of CP is about 250ns which is sufficient to accommodate the combination of largest delay of TTD element 8 and the root mean square delay spread of mmW frequencies in most outdoor environment [18]. Each frame would contain multiple transmission time interval (TTI)s. Frame length is a critical design to accommodate various control data and payloads. With the grant-free multiple access, the structure of DL TTI in the proposed scheme is relatively simple. The detailed frame structure is specified in Table III. Although BS transmits wideband signals, only a portion of the band is accessible to each narrowband UE. Therefore both UL and DL reference signals (preambles) in our design are much longer than the ones in [25] due to lack of broadband receiving capabilities. We also assume open-loop power con-trol with a maximum UE transmit power of 23 dBm [29] which allows for shorter UL reference symbols than that of DL frame. The proposed numerology and frame design are used throughout the rest of the paper. It is worth mentioning that this numerology is tailored to achieve URLLC with large SC spacing and mini-slots. In general, depending on the application, there are designs with more efficient control fields. However, these investigations are beyond the scope of this paper. D. Performance Metrics To simplify the discussion, we assume that a successful DL synchronization and UL transmission without collision together lead to a successful packet delivery. During each DL broadcasting, BS encodes feedback per SC to indicate whether previous UL payload on that SC is successfully decoded. Synchronized UEs would persistently transmit until their packets are delivered, while UEs with unsuccessful synchronizations do not participate in contention. The probability of successful synchronization is dependent on the receiver DL preamble. For users that are within the rainbow link coverage, we analyze the following performance metrics: • Access latency: defined as the amount of time between packet arrival and its successful delivery. • Effective rate: defined as the ratio of the number of data symbols (payload) in a packet and the time required to successfully delivery through contention. Both metrics depend on several factors including SNR, user density, the number of repetitions, and grouping of subcarriers into RBs. The proposed system is expected to serve a large number of spatially separated users with low latency. In the next section, we analyze DL synchronization and UL contention, and mathematically formulate performance metrics. IV. PERFORMANCE ANALYSIS OF RAINBOW LINK In this section, we provide analysis of the proposed rainbow link access. For the sake of tractable analysis and concise notation, we make two additional assumptions. Firstly, we mainly focus on a single path channel, i.e., the summation and the mdependency in (1) are dropped. The underlying rationale is that each UE can only access a narrow bandwidth B u , therefore it is very likely that only a single path aligns in the spatial direction encoded on SCs within B u as illustrated in Figure 4. Secondly, we omit the discussion of beamforming in the UE's side. Specifically, N U = 1 is used, and we drop beamformer v and array response a U (φ u ) of UE in the analysis. The rationale behind is that in the term β u,b = v H u H u,b w b , UE's beamformer serves merely as another factor that has no dependency on b. In other words, a multi-antenna UE results merely in a higher beamforming gain and is not related to any unique feature of rainbow beamforming 9 . Fig. 4: The narrowband operation of UE and rainbow beams naturally reject multipath components. For signal at a given subcarrier, its multi-path components propagate via a different AoD and hence are filtered out by UE due to its limited bandwidth. A. Analysis of DL synchronization Synchronization based on OFDM waveform usually involves estimation of timing offset and carrier frequency offset [30]. With TTD array, in addition to timing offset UEs need to locate their anchor SCs. Here we employ a similar analysis as in [30] of NB-IoT without addressing the carrier frequency offset. In order to locate the anchor SC b u with the highest gain during DL synchronization, a user needs to find: b u is then the anchor SC whose steering angle is the most aligned with θ u . Plugging (1), (2) into (3) and simplifying: The expression above leads to Fejér kernel F NB (·) which is defined as F N (x) = \|sin (N x/2) / sin (x/2)\| 2 . b u should then satisfy the following criterion: From (6), one can see that as the total number of SCs in UE's band increases, the edge SC b u ± \|B u \| /2 have an increasing gain loss compared with the anchor SC b u . This has two implications. Firstly, to ensure a relatively flat beamforming gain across the UE narrowband bandwidth, the total number of SCs in the broadband should be much larger than number of antennas at the BS N B , i.e., N B /B 1. This is because the gain difference in narrow band is only relevant to the ratio N B /B as shown in (6). Secondly, due to the gain loss, there are only limited number of SCs that can be included in B u . For instance, a generic threshold \|B u \| ≤ B/N B ensures that the SC at the edge of B u has less than 3dB gain loss compared with the anchor SC . Increasing number of antennas N B at the BS might result in large gain difference which then limits SCs that a user can access. Notice that these phase and gain differences do not affect UL data transmission since they can be treated as part of the channel. With UL preamble sequence, BS estimates the channel per SC and thus naturally compensates it. The rest of the subsection then addresses the estimation of timing offset. Let L denote the number of symbols in DL preamble sequence. Let l be the index of samples in a OFDM symbol such that −N CP ≤ l ≤ \|B u \|. Then l-th sample of the m-th symbol in preamble sequence can be expressed as: where D u amounts to normalized timing offset that includes an integer and a fractional part due to down sampling. In the equation above, only SCs within the narrow band B u are taken into account. This assumes that UEs sample the wideband signal without aliasing. In practice, such assumption can be achieved with a high quality low pass filter. With channel gain g u remaining the same for B u during L symbols, the received frequency domain symbol is then: In (8),z is the per SC noise in frequency domain. Synchronization is essentially an estimation of D u from observed samplesS However, since θ u is unknown, the gain loss cannot be accurately modeled and causes degradation in synchronization performances. We evaluate the following estimator for timing offset and its synchronization performance: In (9) UEs merely perform correlation betweenS As \|B u \| increases, the summation in (9) involves more symbols, but the correlation between the two sequences is also more distorted. To conclude, in DL synchronization UEs suffer from gain loss due to the rainbow beam pattern and can therefore only operate on a limited number of SCs. This effect will be analyzed via numerical simulations. We note that as a unique feature of frequency dependent rainbow beam pattern, the synchronization does benefit from beamforming gain contributed by BS antenna. The length of preamble L should be designed accordingly to fulfill high probability of detection and low synchronization error. B. Analysis of packet loss rate Next we analyze collisions and packet loss rate (PLR) for given radio resources and user density. The purpose of the analysis is to mathematically guide the design of various system parameters. Since UEs are narrowband, there is no collision between UEs that contend in non-overlapped bands. Given that all users choose a consecutive K RBs, a subband would suffer from collisions from users landing on a total number of 2(K − 1) + 1 = 2K − 1 RBs. To tackle all the possibilities for a given number of repetitions, a combinatorial approach is needed. In the analysis, we focus on a specific user occupying bandwidth B u and calculate the density function of the number of occupied SCs in the band before this user adds its replicas. From there, packet loss rate is simply the probability that all replicas from this user are covered by already occupied SCs. Based on our model, the packet loss rate P PLR (B, K, U, n, g group ) is given in the following proposition as a function of the number of SCs B, the number of active users U , the number of RBs that a user can access K, the grouping factor g group , and the number of repetitions n for each user. Proposition 1: With a uniform spatial distribution of users, an approximated expression of PLR of the proposed rainbow link is where f (x, p, i) = C i x (1 − p) x−i p i is the binomial probability distribution function, C i x is the operator that evaluates x choose i, p 0 = [1, 0, · · · , 0] T ∈ R K+1 is the unit vector, and the entry in m 1 -th row and m 2 -th column of the matrix T ∈ R (K+1)×(K+1) is given by Proof: See Appendix A. We note here that with all other parameters fixed, P PLR as a function of n has a minimum. However, since n is an integer and the formula in (10) is complicated, we discard discussions on conditions of optimal n. In practice, we can simply calculate P PLR numerically for purposes of system design. In general, increasing the number of SCs that a UE accesses can drastically reduce packet loss rate. However, from a geometric perspective, the generic argument is that \|B u \| is limited by B/N B , the number of spatial directions in the main lobe of an analog beam. This can also be seen from (8) that SCs outside of the main lobe of b u have large gain loss. C. Latency and Rate Calculation With both major causes of protocol overhead addressed, in this subsection we discuss the performance metrics used in the evaluation of the system. Since the system is designed to fulfill stringent latency requirements, we mainly focus on the overall latency. Based on latency analysis, we then derive the formulas for effective data rate of UEs. As illustrated by Figure 3, the overall latency T tot can be computed as follows: Specifically, T activation accounts for the random offset to the nearest broadcasting when a UE activates. T failure accounts for unsuccessful synchronization attempts which might include multiple frames. The extra T DL is the overhead for successful synchronization. In T cont , T packetloss stands for packet loss in contention and the extra T DL stands for the overhead receiving ACK from DL broadcasting. The equation to calculate the effective rate is: (13) where N payloads is the total payload and N packet is the payload per packet in terms of OFDM symbols. The most non-trivial trade-off in our proposed design is controlled by g group . On one hand, as indicated by the numerator in (13), using a large number of SCs per RB (large g group ) boosts the data rate of UEs. On the other hand, large g group also enlarges the denominator in (13) through (12), (10). The interpretation is that given a fixed \|B u \|, increasing g group enhances useful payloads but also reduces the number of RBs which means UEs become more vulnerable to collisions. Since the contention overhead depends on the number of users participating in contention at that time, it is hard to evaluate it analytically. Simulation results will analyze these performance metrics in the next section. V. NUMERICAL AND SIMULATION RESULTS In this section, we support our analysis in the previous sections with numerical results. We start with a brief illustration on how certain system parameters are chosen in the simulations. Table II. The UE's baseband bandwidth is then approximately 16MHz. • Grouping factor g group : To keep the discussion simple, we assume that for each user, the number of resource blocks K in B u is an integer. This means g group would be a factor of \|B u \|. In simulations, we consider cases where g group = 1, 2, 4. Using g group = 8 can greatly enhance the rate of UEs but there will be only 4 RBs in B u to avoid collisions. • User active rates p: For the ease of implementation in our simulations, p exclusively controls how many active users are added per frame. Specifically, in each frame we consider 1000 new links uniformly distributed in a semicircle covered by the rainbow beam pattern. Activations of these new UEs are treated as independent events, i.e., U as a random variable follows binomial distribution U ∼ B (1000, p). These new users will contend with persistent users added in a previous frame. Given N B , specifications of link budget at different distances are given in Table IV. The network is assumed to operate at 60 GHz carrier frequency at which free space path loss model in [18] is adopted with line-of-sight (LoS) pathloss exponent equal to 2. Specifically, DL noise power is calculated based on BS's active bandwidth of 1 GHz, i.e., a UE's receiving signal-to-noise ratio (SNR) is independent of its active bandwidth. With the radius of coverage, the activation rate p, and the current frame length (specified in Table III), one can compute user capacity of the system. For instance, a 400 m radius of coverage and p = 0.03 gives the following density on average: (14) in unit of UE activation per m 2 per second. This density ρ serves as a generic metric that evaluates network capacity. A perceivable drawback of the proposed system is the low spectral efficiency. The drawback comes from the following facts: • UE packets are short such that good coding schemes are not applicable. Non-payload portions cannot be ignored. • UEs trade its bandwidth for low rate of collisions. Even if g group = 4 is employed along with 16-quadrature amplitude modulation (QAM) and 1/2 coding scheme, maximum R eff 10 for a single UE is merely 4×18 125×10 −6 × 4 × 2 3 = 1.536 Mbps. Thus the proposed systme is not intended for high rate applications. In the evaluation we use the following parameters unless otherwise specified. The BS has an array size of N B = 64 elements. The total number of SC is B = 2048 and UEs operate on \|B u \| = B/N B = 32 SCs. A maximum UE transmit power is set to 23 dBm which provides a coverage of 400m to achieve about 10 dB SNR for UL transmission. For users near the cell edge, due to their very low SNR, multiple synchronization attempts might be needed to ensure a satisfying detection performance. We evaluate the probability of successful synchronization as a function of SNR and its impact on access latency. An overall histogram of DL SNR values for UEs randomly distributed in a semi-circle of 400 m radius is given in Figure 5. This range of SNR values are used throughout simulations. A. Synchronization Performance Through Monte Carlo simulations, we explore probability of correct estimation of D * u as a function of broadband SNR. Figure 6 gives detailed comparison of synchronization performance with and without including impacts of the aforementioned gain loss and a multiple-path channel. As specified in Table III, a PN sequence with length L = 22 is employed as preamble. The results in Figure 6 clearly show that the gain loss has a non-negligible impact. The more SCs are used for narrowband transmission, the larger the impact of beamforming gain loss is. As \|B u \| approaches B/N B , the synchronization performance starts to saturate. 10 the maximum rate is when a UE always succeeds in its first attempt of contention With only \|B u \| = 8 SCs in band, probability of successful synchronization is not visibly impacted by the rainbow beam pattern. This means that with \|B u \| ≤ B/4N B , i.e., only a small number of SCs included in B u , the impact of rainbow beamforming is trivial. With 16 or 32 SCs in band, synchronization performance loss due to distortion becomes noticeable. There is a 4 dB equivalent SNR loss due to distortion when system uses \|B u \| = 32 SCs, which is effectively the threshold for \|B u \| as discussed previously. In such case, the detection performance is merely comparable with the case of \|B u \| = 16 when there is no distortion. The implications of numerical results are the following. Firstly, the proposed narrowband synchronization beacon is robust to both thermal noise and frequency selectivity introduced by the rainbow beam. In a typical use case of mmW IoT with large number of devices, high probability of DL synchronization can be achieved. Secondly, synchronization performance would benefit greatly from an accurate estimation of b u where (9) can then be altered to count for the gain loss coherently. Otherwise, our analysis suggests that \|B u \| ≤ B/4N B can be a generic threshold for achieving negligible beamforming gainloss. Figure 7 presents both the simulated and theoretical packet loss rate given by (10). In the evaluation, we focus on the two cases g group = 1, 2. B. Collision and Packet Loss Rate We have the following observation from the results. Firstly, the theoretical results agree well with simulations for small and moderate n values. This confirms the correctness of the Proposition 1 where we derived PLR. Admittedly, discrepancy occurs for large n values because our derivation omitted the cases when each user uses repetition transmission that occupies the entire band. Secondly, the results indicate that for a given number of active users, there is an optimal value of repetitions that minimizes P PLR . The optimal value of n decreases with an increasing number of active users in contention. This makes sense since intuitively users should avoid making too many repetitions when the network is crowded, so as to leave resources for others. Lastly, compared to probability of failed synchronizations, packet losses due to collisions among users have much smaller probabilities in the given network settings. Results from Figure 7 show that even with g group = 2 and with 20 active users transmitting simultaneously, the packet loss rate is still below 10 −3 . With multiple retransmissions, the system can then provide high reliability. C. Latency and Rates Next we evaluate the total latency T tot due to DL synchronization and UL contention. Here we focus on the trade-off based on g group as specified in IV-C. We note here again that users are generated with uniform distribution in a 2D-sector of 400 m radius and activation probability of 0.03 per frame over 1000. Simulation results for user-plane latency are presented in Figure 8. In Figure 8a, a comparison on synchronization over- head between the proposed TTD array and phased arrays with multiple RF chains is presented. In both cases, UEs operate on only 32 SCs and BS has a 64-element array. To make a fair comparison, in the case of phased array BS employs random beamforming strategy. In each frame, it steers analog beams randomly to serve users. Thus in both cases, there is no overhead due to control signaling or scheduling 11 . From the result, the performance of the proposed system approaches that of a fully-digital array. As discussed in the previous section, the degradation comes from beamforming gainloss due to frequency dependent beam pattern. For conventional phased arrays, as the number of RF chains decrease, the overall beam coverage is limited as depicted in Figure 1. UEs start to experience prohibitively long synchronization overhead that cannot satisfy requirements of latency-critical uses cases. The Figure 8b shows that with 2-grouping and active rate p = 0.03, T sync dominates the overall latency while with 4-grouping, T cont dominates the overall latency. The tail in contention latency is due to persistent, spatially clustered users in certain narrow band segment. Although in both cases the general URLLC requirement for reliability greater than 1 − 10 −5 at 1 ms latency is satisfied, in 4-grouping case the margin is small. This means that 2-grouping can still support more active users per frame (higher active probability p) while in 4-grouping case p = 0.03 is nearly the maximum activity rate it can support for URLLC. In Figure 9, we show the effective rates for different grouping strategies. For instance, subcarrier grouping with g group = 4 achieves higher rates than g group = 2 yet users are more likely to collide. As stated before, the proposed design is tailored for URLLC and for sporadic, latency critical transmissions. Higher grouping ratio g group might bring higher rates, but is very unlikely to satisfy URLLC requirement. In Figure 10, we demonstrate trade-off of g group with respect to network capacity. Specifically, we gradually increase p to see what's the approximate maximum traffic that the network can support with g group = 1 and g group = 4, respectively. The saturation for g group = 1 appears at about p = 0.165 and as mentioned earlier the saturation for g group = 4 appears at about p = 0.03. Using Equation (10), we numerically calculate P PLR and find that the optimal number of repetitions n are 5 and 3, respectively. Although with g group = 1, each user has only one forth of the data rate, BS can serve 5 times as many users as that in the case of g group = 4. In both cases, the achievable latency for a reliability of 1 − 10 5 is very close to 1 ms, as can be seen from Figure 10. Although in the first case there is no grouping and rate of each user is small, the system can still achieve a higher sum The narrow bandwidth and number of SCs used by the UE are critical parameters for system performance. With the increased number of SCs that UE can access, there is improvement in both synchronizations and contentions. On the other hand, the physical nature of rainbow beamforming limits the number of SCs (approximately B/N B ) that a user can access: only a fraction of radio resources are aligned with the AoA of a UE and SCs at edge of the segment B u suffer severely from beamforming mismatch. Based on the presented analysis, the proposed system has the following performance: • Power and Coverage: The proposed rainbow link based network can achieve a coverage up to 400 m when BS with 64 TTD analog antenna array serves single antenna UEs with the transmit power of 23 dBm assuming lineof-sight condition. • Rates and user capacity: The BS can serve up to 5 active UEs per second per m 2 with Mbps data rates depending on the grouping of SCs per RB. • Latency and reliability: The UEs experience 1 ms latency and reliability higher than 1 − 10 −5 when the optimal number of repetitions n is used. The greatest strength of the proposed system is its flexibility to support combined URLLC and massive multiple access. Based on the proposed design of numerology and multiple access protocol, the system is better suited to serve a massive number of users than to provide high data rates. When conservative grouping is used (in our case no grouping or 2grouping), the proposed system can eventually achieve a sum rate at BS higher than that of the case with larger grouping. The open research questions include but are not limited to the following: Efficient Frame Design: If no grouping is used, one frame has only 18 symbols in its payload based on the proposed frame design. Typical URLLC packets have 32 bytes, so either dense modulation scheme or more aggressive grouping strategies would have to be used. In this work we assume 1 : 1 UL DL ratio. Alternative frame design, for instance, a frame length of 250 µs with 1 : 3 UL and DL ratio could support higher data rate for each user with extra DL TTIs all contributing to the payload. Robustness for Non-LoS channels and user mobility: In the current discussion, UEs are assumed to be in LoS channels. But the same discussion might not be applicable for non-LoS channel conditions. For instance, when a UE suffers from blockage, it loses track of its anchor SC and needs to switch to a different path. As pointed out in III-A, the switching of B u without any prior requires prohibitively long overheads. Thus the current design is not robust to mobility of UEs and non-LoS channel conditions. These questions need to be addressed for practical implementations of rainbow-link. Dynamic Control of Random Access: Increasing bandwidth of the narrowband UE, at the expense of the hardware cost, can greatly boost performance of the proposed system. While we analyzed the system performance under uniform user distribution, distributions of users based on clustered model could further intensify contentions. In this case there might be no optimal number of repetitions as the UEs would persistently contend with other UEs in the same spatial sector. To overcome this effect, the BS can control the number of repetitions that each user can make. If there are too many collisions, BS would tell each UE to reduce its repetitions so as to make sure that at least some UEs can transmit. Alternative Multiple Access Protocol: The proposed system can benefit from non deterministic repetition coding strategy. As it has been proposed in [31], each user can use a random number of repetitions n for each contention period. This might increase system robustness to clustered scenario. Unfortunately, the analysis used in [31] is not directly applicable due to its assumption of sufficient length of code words. Alternatively, instead of repetition coding, a non-orthogonal coding can be applied. Each user can transmit data on all SCs in its band and multiply data symbols with a unique code in frequency domain. Though collisions will happen on SCs, BS can still decode with these unique code words. This is similar to [32] where NOMA is employed for multiple access. VII. CONCLUSION In this work we proposed a novel frequency domain multiple access, referred as rainbow link, that exploits wideband spectrum at mmW to serve a large number of narrowband users. By exploiting TTD array architecture at the BS, frequency resources are mapped to specific spatial directions so that users can be assigned a subset of SCs in OFDMA and leverage beamforming gain of the entire array. Rainbow link enables a grant free multiple access and can support a very large number of users with stringent latency requirements. With a single RF chain and 64-element array, the BS can provide reliable access with 400 m LoS coverage and sub-millisecond overhead for up to 5 UE activations per second per square meter. We believe the proposed rainbow-link can be a candidate for future critical mMTC use cases. A. Generic Formula for Set Covering Distribution For the simplicity of notation, in the appendix we use B for the total number of RBs so that it integrates g group . Before the detailed of derivation of (10), we first rewrite it in a more intuitive manner: P K,n (i) is probability of packet loss of a user given i other users are transmitting on nearby SCs in its narrow baseband B u . Specifically, nearby SCs actually refer to the (2K − 1) SCs whose narrow band segments have intersection with B u . Since the user's choice of n SCs are completely random, P K,n (i) can be decomposed in a straight forward manner: P K,n (i) = K j=0 P (j among K SCs occupied by i users) Essentially, packet loss results in the case when all n choices of the user fall on occupied SCs. Therefore, once we know the probability distribution of the number of occupied SCs, packet loss rate can be thoroughly calculated. The strategy here is to characterize occupancy of the selected narrow band as a Markov chain where each competing user causes transition of states in terms of number of occupied SCs. Let p i ∈ R K+1 be the vector characterizing state of the narrow band with i competing users, i.e., Since any segment of joint K SCs can be chosen as a narrow band, competing users can have different numbers of overlapping SCs with the user of interest. As a consequence, for a competing user, its n repetitions might fall only partially on those overlapped SCs. Then, those repetitions (on overlapped SCs) contributed by a competing user may or may not land on occupied SCs in B u . All these probabilities are given by: This formula is essentially a generic approximation. To our best knowledge, the exact formula here can not be tracked analytically. The main reason is that probabilities for different configurations of j repetitions in B u are not even. Competing users with a highly overlapped band might generate some configurations that other users cannot. This means that state of the band can not be fully characterized by the number of occupied SCs alone. Since a complete treatment is computationally infeasible, we here assume that all configurations for j repetitions are equivalent. This assumption works well for small and medium value of n and introduces deviation for n approaching K. With the transition matrix, calculation of p i is relatively simple: And then: Ruifu Li received the B.Sc. degree from the University of Wisconsin-Madison in 2020. He is currently working towards his M.S. degree in electrical and computer engineering at University of California, Los Angeles. His research interests include signal processing and wireless communications.	2021-09-10T01:16:26.826Z	2021-08-21T00:00:00.000	{ "year": 2021, "sha1": "a032dc8fba841bd4707f8006e87d79154c813a40", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "a032dc8fba841bd4707f8006e87d79154c813a40", "s2fieldsofstudy": [ "Engineering", "Computer Science" ], "extfieldsofstudy": [ "Engineering", "Computer Science" ] }
251356860	pes2o/s2orc	v3-fos-license	How Do Male Football Players Meet Dietary Recommendations? A Systematic Literature Review The aim of this review was to determine whether male football players meet dietary recommendations according to a UEFA expert group statement and to identify priority areas for dietetic intervention, including training periodization and field position. A database search of PubMed, Web of Science, EBSCO and Scopus was performed. To be included within the final review, articles were required to provide a dietary intake assessment of professional and semi-professional football players. A total of 17 studies met the full eligibility criteria. Several studies showed insufficient energy and carbohydrate intake compared to the recommendations. A majority of athletes consume adequate protein and fat intakes compared to the recommendations. In addition, several studies showed the insufficient intake of vitamins and minerals. This systematic review showed that football players do not meet the nutritional recommendations according to the UEFA expert group statement. Future research should be focused on how to apply nutritional recommendations specific for athletes in accordance with training periodization and positions on the field. Introduction Football is a sport discipline characterized by high-intensity efforts followed by periods of active recovery or passive rest. Football players complete a weekly microcycle of 8-10 training units during the preparation period. The competitive period is characterized by the highest-intensity efforts, which increase energy expenditure [1]. The match effort, on the other hand, is characterized by a variable tempo, during which the athlete jogs, sprints and dynamically changes the direction of running, performing about 726 different actions of movements and turns [2][3][4]. The number of matches in a season at an elite level number up to 60 [5], and players run about 9-12 km in each of them [6]. This distance depends mainly on the players' position on the field [3,7]. It is also worth noting that energy expenditure during a match is dependent on the position of the player [8] and that energy and nutrient requirements need to be individualized [9]. Properly balanced nutrition is one of the factors that support performance enhancement and post-exercise recovery [10,11]. An adequate nutrient supply contributes to training adaptation [12,13] and can have an impact on reducing the risk of injury and recovery time [14]. Training periods vary in terms of physiological demands and differences in energy expenditure generated. Planning the nutrition of athletes with adjustments to the periods of training and individualization requires special attention [15]. Improvement in an athlete's exercise capacity is influenced not only by the appropriate choice of exercise load but also by the use of appropriate nutritional strategies to maximize adaptive changes [16,17]. Recently, it has been observed that physiological demands in football have increased [2,18]. Evidence suggests that high-intensity and sprint activities have increased by 30-80% [2]. Hence, nutritional strategies are rated by elite teams as one of the most important ways to accelerate this process. In 2017, Collins et al. [19] indicated that specific nutrition guidelines for football had not been updated for over a decade [20]. Three years later, an UEFA expert group statement on nutrition in elite football has been published, highlighting that increased interest in and focus on nutrition brings with it greater challenges in football society [21]. The authors indicated that this nutritional guideline represented a vital first step, bringing together the best scientific and on-field practitioners from around the world to add much-needed continuity to recommendations for players and teams. The continued development of good nutritional practices within the game will also have an impact far beyond clubs' training grounds, helping to promote good health across society as a whole [21]. Therefore, the aim of this review was to determine whether male football players meet the dietary recommendations according to the UEFA expert group statement, and to identify priority areas for dietetic intervention, including training periodization and field position. Materials and Methods The review is reported using the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines [22]. Population, intervention, comparison, outcome and study design (PICOS) criteria are defined in Table 1. All keywords used in the literature search are listed in Table 2. Due to the updated nutritional recommendations for football players, the data from all studies were compared to the UEFA expert statement [21] in the discussion section. Concept Key Words Football (soccer) players "football" OR "soccer" OR "players" OR "soccer players" Dietary intake "nutrient requirement" OR "nutritional supplement" OR "dietary supplement" OR "dietary intake" OR "daily food" OR "food intake" OR "dietary assessment" OR "dietary requirement" OR "sports nutrition" OR "food diary" OR "macronutrient" OR "nutrient needs" OR "dietary needs" OR "nutrient intake" OR "RDA" Search Strategy One author (KD) searched PubMed (MEDLINE), Web of Science, Scopus and EBSCO (including SPORTDiscus) databases for literature relevant to the aims of the systematic review. Eligibility Criteria All original studies (including cross-sectional studies, longitudinal studies, observational studies and randomized controlled trials) were included in the study. Studies not written in English were excluded from the analysis. Studies published prior to 2011 were excluded from the analysis. Only studies with adult male professional or semi-professional footballers as participants were eligible for the review. "Professional" athletes are defined as individuals who exercise > 10 h/week and whose athletic performance has achieved the highest level of competition, and "semi-professional" athletes exercise > 6 h/week with an emphasis on improving performance. Only studies reporting typical consumption values (e.g., energy as kcal/day, protein as g/day or calcium as mg/day) were included. Studies that qualitatively assessed diet were not included in the review. All inclusion and exclusion criteria are specified in Table 3. Study Selection Process Publications were initially reviewed based on the title and abstract by two authors (K.D. and A.K.). Duplicates and articles deviating from the topic of the literature review were removed. All articles that passed title and abstract review were subjected to full-text review. Each full-text article was checked against the inclusion and exclusion criteria presented in Table 3 by two authors (K.D. and A.K.). Differences of opinion regarding the inclusion of articles in the literature review were verified by the third author (A.Z.). The stages of article selection and reasons for excluding papers are shown in Figure 1. Study Quality: Risk of Bias The quality of individual studies included in the review was assessed for bias (quality) using the Academy of Nutrition and Dietetics Evidence Analysis Manual [23], which permits the assessment of relevance and validity, with the allocation of either a positive, neutral or negative quality ranking. All studies were compared against this checklist by two different reviewers (K.D., A.K.). Only studies that accurately described study selection, including inclusion/exclusion criteria, were included in the review. At end stage, a third reviewer (A.Z.) reviewed any discrepancies. In quality assessment, no studies had negative rating; therefore, all of the obtained articles were included. Protein The average protein intake in the analyzed papers was 1.9 ± 0.3 g/kg BM/day (range 1.3 ± 0.4 g/kg BM/day [26]-2.5 g/kg BM/day [38]). Three studies [26,29,33] showed lower protein intake, and in one paper [39], protein intake exceeded the protein recommendations advocated by the UEFA expert group statement (1.6-2.2 g/kg BM/daily) [21]. The adequate intake of protein was observed in the other papers. The average proportion of fats in the total daily ratio during the preparation period was 30.5 ± 4.0% and during the in-season period 31 ± 2%. The proportion of energy from fat in the diet in the pre-season period ranged from 25 ± 3% [29] to 34 ± 12% [32]. During the in-season period, the proportion of energy from fat in the papers analyzed was 29 ± 4% [35] to 37 ± 5% [30]. Five studies evaluated fat intake dependent on the position occupied on the field, of which two [24,40] concerned the preparation period and three the competitive period [30,34,39]. The lowest fat intake during the preparation period was recorded in center defenders-fullbacks-(1.0 ± 0.4 g/kg BM/day) [24] and the highest in midfielders (1.4 ± 0.6 g/kg BM/day) [24]. During the season, the lowest proportion of dietary fat was recorded in goalkeepers (0.9 ± 0.2 g/kg BM/day) [34]. In contrast, the Anderson et al. [39] study found that goalkeepers consumed the highest amount of fat (1.9 ± 0.4 g/kg BM/day). Vitamins and Minerals The average intake of vitamins and minerals in the diets of athletes was analyzed in four papers [24,26,29,40]. All studies included in the review evaluating vitamin and mineral intake were from the pre-season period [24,26,29,40]. In the analyzed studies, the insufficient intake of magnesium, calcium, zinc, folic acid and vitamins such as B 1 , B 2 , B 12 , A, C, E and D was compared to the recommended standards [42][43][44]. Książek et al. [29] analyzed the average intake of vitamins and minerals with respect to the supplementation used. It was shown that the mean intake of folic acid, vitamin D, vitamin E, iron and zinc in the diet that included supplementation was consistent with the standard intake. Four studies assessing iron and calcium intake and three articles assessing vitamin D intake were included in the systematic review. The average reported daily intake of iron was 14.3 ± 3.8 mg/day. Two out of four studies [29,40] reported intake that met the iron intake requirement, which is 15-18 mg/day. The average daily calcium intake from studies was 1027 ± 323 mg/day. In three of the four papers, insufficient calcium intake was observed in relation to recommendations [24,26,29]. Three studies reported vitamin D intakes of 12.2 ± 8.7 µg/day [40], 3.1 µg/day [24] and 4.9 µg/day (56.5 µg/day after accounting for supplementation) [29]. With regard to the Institute of Medicine (US) Committee to Review Dietary Reference Intakes for Vitamin D and Calcium (15 µg/day, 1300 mg/day) [44], only 1 paper [29] reported the recommended intake. Discussion This review aimed to assess the adequacy of dietary intake in male professional and semi-professional football players when compared to the dietary recommendations advocated by UEFA expert group statement and to define dietary areas demanding improvement. This systematic review found that football players' total energy and carbohydrate intake did not meet sports-nutrition recommendations; on the other hand, the majority of athletes had adequate protein and fat intake. Energy Intake in Football Players Adequate energy supply is a crucial factor for enhancing physiological adaptations during the pre-season and for optimizing performance in-season. According to the UEFA expert group statement [21], the mean daily energy expenditure of goalkeepers and outfield players has been estimated at~2900 kcal/day and~3500 kcal/day, respectively, with matchday energy expenditure estimated at~3500 kcal. Several studies (n = 7) showed that football players' energy intake does not meet the demands of the pre-season [29,31,32] or the inseason [30,[32][33][34]36]. Inadequate energy supply in the diets of the athletes studied may have been due to inadequate reporting related to a lack of recording and undernutrition during the study [45]. Methods used to measure dietary intake are hampered due to errors in precision (repeatability, reproducibility and reliability) and validity (accuracy). Moreover, existing literature has shown significant variability between methods used to evaluate dietary intake, with the frequent underreporting and misreporting of consumption [46]. Another reason for insufficient energy supply may be a real problem associated with low energy availability (LEA). Inadequate energy supply in the diets of athletes may contribute to relative energy deficiency in sport (RED-S). RED-S is a common and contemporary topic in sports nutrition and is defined as a range of signs and symptoms that may negatively affect the health and performance of physically active individuals (both female and male). The effects include endocrine, physiological, metabolic and psychological dysregulation, which may ultimately affect physical capacity and performance [47]. There is little research concerning RED-S in football players, and this has so far focused only on female football players [48]. Excessive energy intake relative to the daily requirements may contribute to weight gain by increasing the proportion of body fat. Excess body fat negatively affects acceleration capacity, overall energy expenditure and injury risk. It is important that sports nutritionists and performance teams provide players with body-composition targets [21]. Such targets can be achieved using periodized nutrition. Periodizing nutrition focuses on the different energy needs and varying intake during different phases of the year [16]. In football, energy demands are lower in the in-season and higher in the pre-season [1], when the nutritional strategy should be focused also on obtaining optimal body composition [49]. Carbohydrate Intake in Football Players Due to football involving activities of varying intensities, including walking, jogging, sprinting, changing direction, jumping, striking the ball and contact with opposition players, CHO are the primary fuel during matches [21]. During the pre-season, the main task of a football player is to improve their athletic performance, to avoid injury and illness and to prepare to play throughout the season. The maintenance of football players' athletic performance largely depends on the availability of CHO as energy substrates [50]. The suggested CHO intake is 4-8 g/kg BM/day and depends on variations in loads and individual training goals (such as maintaining or changing body composition). If there were a greater intensity and volume of training, higher CHO intakes (6-8 g/kg BM/day) would likely be required. Only in two of the six analyzed papers [31,32] did not players consume CHO according to the recommendations (4-8 g/kg BM/day) during the pre-season. In season, the maintenance of maximal performance in football players is highly related to the replenishment of glycogen stores via proper CHO intake. Currell et al. [51] showed that CHO intake during matches improved running performance (agility, dribbling and kicking accuracy). In-season CHO recommendations suggest 6-8 g/kg BM/day prior to a match and when there is fixture congestion (a 2-3-day period between games) [21]. One [37] of two papers evaluating match days showed an optimal CHO supply (6-8 g/kg BM/day). The optimum physique varies according to a player's physiology, their field position and playing style [21]. Although goalkeepers are usually taller and have higher body mass (higher fat mass) than field players [52], they still have the optimum physique, in terms of position-specific and seasonal trends. Moreover, athletes have different energy requirements depending on their position on the field [53]; therefore, the amount of CHO should be adjusted individually for each athlete. Players whose positions on the field are characterized by a higher energy expenditure should increase their daily intake of CHO. Protein Intake in Football Players According to the UEFA expert group statement [21], to enhance training adaptation, support recovery and stimulate muscle protein synthesis, the recommended protein intake is 1.6-2.2 g/kg BM/day. In this review, the majority of studies reported that protein intake matched the recommendations for the pre-season [24,31,32,40] and for the competitive period [30,32,[34][35][36]. In one study [39], players exceeded protein recommendations while not consuming enough CHO. Daily football training leads to musculoskeletal and tendinous tissue damage. Due to the important role of protein in muscle development and maintenance, football players should consume higher quantities of protein than the general population. According to the "food-first" philosophy, it is easy to achieve the recommended level of protein intake (1.6-2.2 g/kg BM/day) with a mixed diet. According to Burke et al. [54], the food-first philosophy states that nutrient delivery should come from whole foods and drinks, and there are situations wherein a "food-only" approach may not always be optimal for athletes. In special situations (e.g., dietary energy restriction, rehabilitation after injury), athletes require a higher proportion of protein in the diet [55][56][57]. To achieve the recommended amounts of nutrients in the diet of an athlete, and due to the difficulty in consuming large meals immediately after exercise, periodic supplementation with supplements, especially from group A according to the Australian Institute of Sport (AIS), may be considered. According to recent reports, food-first but not always food-only [58] football players in special cases should develop a personalized nutrition and supplementation plan via collaboration with their dietitian to optimize their performance. Fat Intake in Football Players Most of the studies included in this review indicated that the athletes studied consumed fat according to the recommended standard (20-35% of total energy) [21]. In contrast, in the study by Iglesias-Gutiérrez et al. [30], the athletes studied exceeded the recommended intake standard for this macronutrient (37 ± 5%). The study authors suggest that the high fat supply in football players' diets may be dictated by high animal-protein intake [30]. The average proportion of fats in football players' diets during the starting period was 31 ± 2%. A diet high in saturated fatty acids can negatively affect an athlete's health [59]. Football players, when planning their individual diets, should pay attention to the ratio of total fat (<35% of total energy requirements), especially taking into account the optimal supply of saturated fatty acids (<10% of total energy requirements) [60]. Micronutrient Intake in Football Players Due to the increased energy expenditure associated with intensive training and participation in matches, football players have a greater need for vitamins and minerals compared to the general population [42]. Regulatory ingredients play an important role in athlete nutrition by participating as precursors to metabolic pathways and physiological processes, which can help maximize adaptive return in athletes [61]. Studies included in the review showed the suboptimal supply of vitamin B 1 and B 2 [26], folic acid [24,40], vitamin A [24,26], vitamin D [24,40], vitamin C [26,40], calcium [24,26], magnesium [24,26], iron [24] and iodine [40]. In line with the food-first philosophy [54], nutrients (including vitamins and minerals) should come from standard foods and beverages, rather than from isolated ingredients in foods, dietary supplements or sports foods. All vitamins and minerals are important for health and performance. However, according to the UEFA expert group statement [21], of particular note are vitamin D, iron and calcium. Vitamin D deficiency can lead to several health issues, including an increased risk of bone injuries, chronic musculoskeletal pain and viral respiratory tract infections [41,62,63]. Athletes with low 25-(OH)D levels (<30 ng/mL or <75 nmol/L) have also been shown to be at a higher risk of injury, including musculoskeletal injuries [64]. Close et al. [58] proposed recommendations for using dietary supplements in sport foodfirst but not always food-only. Authors have suggested several reasons why a food-only approach may not always be optimal for athletes, one of which concerns vitamin D [65]. Iron deficiency in the diet of athletes can have a negative impact on aerobic capacity [61,66], which may indirectly translate into the decreased effectiveness of training units and reduced adaptive capacity. Two studies included in this review showed an insufficient supply of this micronutrient in football players during the preparation period [24,26]. Football players, especially in-season, are at risk of iron deficiency [67,68]. Therefore, its serum concentration should be monitored, and dietary supply should be controlled. Proper doses of iron may be required to improve health through supplementation to correct any deficiency [58]. An optimal supply of calcium is necessary to maintain bone health, skeletal muscle function, cardiac contraction and nerve conduction. It has been shown that athletes can lose calcium with sweat [69]. Four papers included in the review showed an inadequate supply of calcium in the diets of the athletes studied [24,26,29,40]. To achieve the recommended dietary amount, football players should follow a mixed diet that includes calcium-rich foods (mainly dairy products, with smaller amounts from bony fish, legumes, certain nuts, plus fortified soy beverages and breakfast cereals). Consideration of the need for calcium supplementation may be required in cases of RED-S or in situations of increased calcium requirements (vegetarian/vegan diets) [58]. The reason for the non-adherence to dietary recommendations may be due to athletes' knowledge on how to choose the right food for the main, pre-and post-workout meals. Physiological factors responsible for appetite can also contribute to an inadequate supply of energy, especially during long/high-intensity training periods [60]. Moreover, logistical issues, such as time constraints (short breaks between training sessions), can have an additional influence on insufficient nutrients intake. Using Research Evidence in Practice It is common knowledge that translating theoretical assumptions into practice is difficult to implement [70,71]. Translation of science depends on technical competency, personal attributes and practical skills (e.g., ability to build rapport with athletes, coaches, support staff). The delivery of nutritional knowledge should be supported by practical skills (e.g., ability to promote behavior change, cooking skills), resources and technological tools (e.g., relevant software, mobile applications) contained within the practitioner's toolkit [72]. Therefore, the main goal of sports organizations/clubs should be hiring a qualified dietitian who should critically assess the translational potential of research and the applicability of these results in practice. Translating innovation in scientific research into practical applications for football players should be in the most accessible way, e.g., creating infographics, menu and recipe guides, video-based content, lectures or mobile applications. Nutritional education can be also be implemented by a special tool, such as the Athlete's Plate (AP). The AP was developed to teach athletes how to design their plates depending on training load [73]. Furthermore, a necessary next step for knowledge translation are personal connections between sports organizations or clubs/staff members, and sports researchers/scientists may be a way to facilitate the use of nutrition research in sports. Developing a network of researchers to generate "credible knowledge" and facilitate the implementation of research in sports can help increase nutritional awareness among athletes [74]. The aforementioned activities can facilitate the implementation of research evidence in practice. Limitations This systematic review included only studies that assessed the intake of nutrients using validated methods. The underreporting of the quantitative estimates of dietary intake is one of the most common obstacles preventing the collection of accurate habitual dietary intake data [75,76]. In addition, data on dietary intake or anthropometric measurements were not provided by the authors of several studies included in this review. The missing data were calculated based on the available information in these studies, which may have reduced the accuracy of the results. It should not be underestimated that the delivery of and compliance with recommendations in the context of football is complex, because sports nutrition often has to deal with variable training loads, breaks and stops, injuries, returnto-play and complementary pharmacological management [77]. Moreover, the systematic review registration code is missing. Conclusions This review found that football players at a medium to high sporting level are not implementing the nutritional recommendations identified by the UEFA expert group. Among other things, insufficient supply of energy, CHO, vitamins and minerals (vitamin B 1 , B 2 , folic acid, vitamin A, C, D, E, calcium, magnesium, iron, iodine) were shown. Athletes do not periodize their CHO supply based on their training period. Moreover, the majority of athletes consume adequate protein and fat intake compared to recommendations. In order to monitor athletes' implementation of dietary recommendations and provide education on the subject, it is important for sports clubs to employ nutritionists. Nutritional periodization for macronutrient modification in the diet should take into account the nutritional goal, training period, position on the field and tactical tasks assigned to the player. This review highlights that further research should be focused on how to apply nutritional recommendations specific to athletes in specific positions on the field and be appropriate to the physical demands.	2022-08-06T15:13:38.655Z	2022-08-01T00:00:00.000	{ "year": 2022, "sha1": "e697260d32e0d528cb3dfcbe7bcfeeefe9ec2e84", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/1660-4601/19/15/9561/pdf?version=1659532266", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "5f9803021c0749919ad98c785e26670ae598586b", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [] }
82297021	pes2o/s2orc	v3-fos-license	An easy, cheap computerized method to assess two-dimensional trajectory parameters Short notes don’t have an abstract. Movement is essential for the survival of mobile organisms.Its study can help to determine taxonomic status (1), to isolate pheromones (2) and to understand biological mechanisms (3).It can also provide information on the health, physiological state and motivation of animals.However, it has rarely been rigorously quantified.We devised a manual method in 1973 (4) and computerized it in 1991 (5) but, despite its continuing use (e.g.6), this processing became obsolete due to the evolution of computers.Plenty of modern programs exist (7 and references therein, 8, 9 and references therein) but require expensive equipment, take a long time and are generally appropriate for only one kind of assessment.We developed a user-friendly, cheap method that allows simultaneous assessment of orientation, linear speed and angular speed of any moving agent.This software was tested on the ant Myrmica rubra, in a colony being maintained in the laboratory (Fig. 1A).Stimuli presented to the foragers were pieces (1 cm 2 ) of white paper and isolated heads of congeners, which emit the species' alarm pheromone. Ant trajectories were manually recorded, using a water-proof marker pen, on a glass slide set over the ants' foraging area.They were then traced onto transparent polyvinyl sheets, which stuck to the screen of any PC (Fig. 1B).The trajectories could then be analyzed using the newly elaborated software installed on the PC: 1.The distance between two points on the screen, initially assessed in pixels, is converted into a metric unit using a dialog box, for both the Xand Y-axes.2. The successive points of the trajectory are entered by clicking with the mouse, which visualizes, in red, the trajectory on the screen (Fig. 1C).The point towards which the moving agent was expected to go is then located, in green, on the screen (Fig. 1C).3. The user then states that the trajectory entering is finished and, after that, he/she enters, in a window, the total time spent by the moving agent to move along its trajectory.4. Validating the last operation starts the calculation, by the newly-elaborated software, of the three following variables (Fig. 2).Comparison of the manual (M) and the computerized (L) method.Ten ant trajectories obtained in the presence of a blank paper (control) and of an isolated congener's head (test) were analyzed and the difference between the two methods was evaluated.Differences are less than the experimental errors.O=orientation (angular degrees), V=linear speed (mm/sec), S=angular speed (angular degrees/cm).Assessment of the linear (V) and angular speed (S) of Tribolium castaneum and of Paramecium caudatum under control and experimental conditions.T. castaneum was observed directly, like the ants, while P. caudatum was observed under a stereomicroscope (Mag.= 23 X), this requiring a unit adaptation.N = number of individuals observed; results of non-parametric χ 2 tests between control and experiments: P= level of probability; NS = difference not significant at P = 0.05.An activated GSM had an impact on the observed animals. T. castaneum An easy method to assess two-dimensional trajectory parameters The manual and the computerized methods give identical results (Table 1), but the computerized one is 30 times faster and therefore allows analysis of many more trajectories, and is more precise, human errors being avoided. The newly-computerized method was then used to make five assessments, and was thus tested. Trajectories of the beetle Tribolium castaneum were successfully analyzed under normal conditions, near a switched-on mobile phone (GSM) and near a switched-off GSM (Table 2).The new method is particularly applicable to small moving animals.Note the effect of an activated GSM on the insects' movement.2. Trajectories of the protozoan Paramecium caudatum were analyzed under normal conditions, near a switched-on GSM and near a switched-off GSM (Table 2).A camera lucida was applied to the stereomicroscope under which P. caudatum were set.The new method allows analysis of the movement of any microscopic agent in this manner.Note, once more, the effect of an electromagnetic field on living organisms.3. Pieces of white paper (1 cm 2 ) were deposited for 8 days on ant cemeteries (Fig. 3A) and were Fig. 3. -Three illustrated uses of the method.A: pieces of paper were deposited on ant cemeteries and then presented to foragers.They were not attractive to the ants but decreased their angular speed.They thus may be impregnated with trail pheromone deposited by ants leaving the cemeteries sites.B: isolated heads of three ant species were presented to foragers of these species to see if such cross tests can help recognizing unknown species.Here, the head of an individual of M. sabuleti (pointed by an arrow) is presented to workers of M. rubra, which are not attracted by the non-specific stimulus.Cross tests and assessments using our method can thus help discriminating between species.C: trajectories of ants moving near a small amount of ethanol or chloroform.Ethanol increased the ants' linear and angular speed while chloroform decreased their linear speed.Simple ethological tests together with our software-based method can help detect minute amounts of drugs in samples. then presented to foragers whose movement was analyzed using the described method (Table 3).The foragers were not attracted by the papers but their angular speed considerably increased.Ants transporting corpses move thus randomly away from their nest and in a sinuous increasingly slowing-down pattern as they come nearer to a cemetery.They finally stop there and drop the corpses.While returning then to their nest, they deposit their trail pheromone along a short distance (personal observation), which explains the ethological effect of cemetery sites on the ants.The new computerized method thus provided, in a few minutes, an explanation for the presence of ant cemeteries, on given places, far from the nests.4. Myrmica ants are attracted by their specific alarm pheromone contained in the head of workers (Table 4).Cross-tests using isolated heads of known and unknown ants (f.i.newly collected) (Fig. 3B) followed by analysis of the numerous recorded ant trajectories enable recognition of an unknown (f.i.collected) species.Such a long process can be efficiently performed only by using this rapid computerized method.Such taxonomic recognition of closely related species can be Locomotion of Myrmica sabuleti foragers in front of their cemeteries.Blank pieces of paper or paper deposited for 8 days at cemeteries were presented to foragers.The orientation towards the paper (angular degrees), the linear speed (mm/sec) and the angular speed (angular degrees/cm) of 60 or 30 (= N) foragers were assessed using our software.The distributions of the values obtained for each two stimuli were compared using the nonparametric χ 2 test.P = level of probability; NS = difference not significant at P = 0.05.Cross-tests between three Myrmica species, using isolated worker heads presented to foragers.The orientation (O; angular degrees) towards the head, the linear speed (V; mm/sec) and the angular speed (S; angular degrees/ cm) of 10 foragers were assessed each time, using our software.Ants clearly oriented themselves only towards isolated heads of their own species.Cross-tests, together with our computerized method, are thus helpful for taxonomic purposes. An easy method to assess two-dimensional trajectory parameters extended to any animals that have specific pheromonal secretions.It can be used as an aid to morphological or genetic determination.This technique should be applied, for instance, to related bumblebee species (10), virgin females responding only to the pheromonal secretion of conspecific males. 5. Myrmica ants react to ethanol by increasing their linear and angular speed (Fig. 3C), and do so down to 0.0001 µl of ethanol, which corresponds to an aqueous solution of 0.001% (Table 5).These ants also react to chloroform, but by decreasing their linear speed (Fig. 3C), this occurring down to a presentation of 0.00001 µl of chloroform, e.g. an aqueous solution of 0.0001%.For revealing these kinetic reactions, many trajectories must be analyzed, and this can be done, in a short time, only by using this rapid, simple method.So, using this method, Myrmica ants can be used to detect small amount of any given drugs in collected material.Response of Myrmica sabuleti workers to ethanol and chloroform.10 µl of differently-concentrated solutions of these substances were presented to foragers and the locomotion of 10 of them was assessed using the here related software.The concentration (%) is given in the first column; the quantity (µl) presented, in the second one.O = orientation towards the stimulus, angular degrees; V = linear speed, mm/sec; S = angular speed, angular degrees/cm.χ² tests between results for 'pure water' and 'substances': P = level of probability, * = P < 0.05 or 0.02, ** = P < 0.001, otherwise = result non significant at P = 0.05. In conclusion, Roduit (11) wrote: 'no universal solution exists for the analysis of trajectories'.This is true when the solution requires highly technical equipment, sophisticated software and many conditions for being used.On the contrary, a simple method -requiring cheap material, easy-to-use software and having no conditions for being used -may be universal or, at least, used in a first step to check if it may be promising to use more onerous methods.The user-friendly system we have here related is such a simple method.It requires no program license and can be used by many persons at the same time.It is thus competitive with other more sophisticated methods.The software, labeled OVS, will be available on the website of the journal as soon as the present paper is published. Fig. 1 . Fig. 1. -Three steps in the computerized analysis of trajectories.A: ants are kept in the laboratory in artificial nests.Trajectories are recorded on a glass slide set above the ants' tray and are then traced on a polyvinyl sheet.B: this sheet is stuck to the screen of a PC.Each trajectory is entered using a mouse.C: the updated software visualizes each trajectory and quantifies its orientation, linear and angular speed. M Fig. 2. -Mathematical reasoning underlying the quantification of the orientation (O), linear speed (V) and angular speed (S) of a trajectory.The three variables are defined in the text.	2019-03-19T13:06:44.505Z	2012-07-01T00:00:00.000	{ "year": 2024, "sha1": "e6c19a627009cdd1dca190ea6f9cfc198b43a757", "oa_license": "CCBY", "oa_url": "https://belgianjournalofzoology.eu/index.php/BJZ/article/download/158/239", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "7e9ba9cc21dffff77ae459503ddb8a0dbe22ed46", "s2fieldsofstudy": [], "extfieldsofstudy": [ "Biology" ] }
256183495	pes2o/s2orc	v3-fos-license	Mitochondrial Transplantation in Mitochondrial Medicine: Current Challenges and Future Perspectives Mitochondrial diseases (MDs) are inherited genetic conditions characterized by pathogenic mutations in nuclear DNA (nDNA) or mitochondrial DNA (mtDNA). Current therapies are still far from being fully effective and from covering the broad spectrum of mutations in mtDNA. For example, unlike heteroplasmic conditions, MDs caused by homoplasmic mtDNA mutations do not yet benefit from advances in molecular approaches. An attractive method of providing dysfunctional cells and/or tissues with healthy mitochondria is mitochondrial transplantation. In this review, we discuss what is known about intercellular transfer of mitochondria and the methods used to transfer mitochondria both in vitro and in vivo, and we provide an outlook on future therapeutic applications. Overall, the transfer of healthy mitochondria containing wild-type mtDNA copies could induce a heteroplasmic shift even when homoplasmic mtDNA variants are present, with the aim of attenuating or preventing the progression of pathological clinical phenotypes. In summary, mitochondrial transplantation is a challenging but potentially ground-breaking option for the treatment of various mitochondrial pathologies, although several questions remain to be addressed before its application in mitochondrial medicine. Introduction Mitochondria are cytoplasmic double-membrane organelles defined as eukaryotic cells' powerhouses due to their involvement in the cellular bioenergetics. In particular, mitochondrial synthesis of adenosine triphosphate (ATP) is associated with the functionality of aerobic oxidative phosphorylation (OXPHOS) [1]. In addition to energy production, mitochondria play many critical roles in cellular function and signalling, including fatty acid biosynthesis, calcium homeostasis, reactive oxygen species production, cell survival, proliferation, apoptosis, autophagy, stem cell differentiation, and regulation of the immune response [2,3]. The complex functionality of mitochondria in cell biology is supported by their morphological and ultrastructural plasticity. Under physiological conditions, they form a dynamic and highly interconnected tubular network [4]. Due to their endosymbiotic origin [5], mitochondria have their own circular doublestranded DNA (mtDNA) that is transcribed and replicated following principles that are different from those regulating nuclear DNA (nDNA) [6]. Furthermore, unlike nDNA, mtDNA is characterised by maternal uniparental inheritance [7]. Human mtDNA is 16.6 kb long and comprises 37 genes, of which only 13 code for subunits of complexes I, III, IV, and ATP synthase. The remaining genes code for 2 ribosomal RNAs and 22 transfer RNAs, both of which are necessary for the protein synthesis machinery of mitochondria [8]. Nuclear genes encode the remaining proteins involved in OXPHOS and other enzymes required for mtDNA replication, repair, transcription, and translation [9]. Each individual mammalian cell contains hundreds or thousands Figure 1. Schematic representation of the treatments for mitochondrial diseases. Illustrated are the current treatment strategies for mitochondrial diseases, which can be classified into pharmacological and metabolic (red) and molecular (blue) approaches. Pharmacological and Metabolic Approaches Most of the pharmacological agents aim to reduce symptoms, particularly exercise intolerance and fatigue, and slow down the progression of the disease. Traditional treatments attempt to improve the function of the electron transport chain (ETC). In particular, several vitamins and cofactors are used to: (1) increase electron flux through the respiratory chain (CoQ10, riboflavin); (2) act as antioxidants (idebenone, alpha-lipoic acid, vitamins C and E), and/or increase ETC substrate availability (dichloroacetate and thiamine); and (3) feed the mitochondria (L-carnitine) [27]. However, in recent years, emerging experimental strategies have been proposed which are focused on: (1) stimulating mitochondrial biogenesis, (2) modulating mitophagy and mitochondrial dynamics, (3) bypassing OXPHOS deficits, (4) using mitochondrial replacement therapy (MRT), and (5) treating chronic hypoxia [24]. Schematic representation of the treatments for mitochondrial diseases. Illustrated are the current treatment strategies for mitochondrial diseases, which can be classified into pharmacological and metabolic (red) and molecular (blue) approaches. Pharmacological and Metabolic Approaches Most of the pharmacological agents aim to reduce symptoms, particularly exercise intolerance and fatigue, and slow down the progression of the disease. Traditional treatments attempt to improve the function of the electron transport chain (ETC). In particular, several vitamins and cofactors are used to: (1) increase electron flux through the respiratory chain (CoQ10, riboflavin); (2) act as antioxidants (idebenone, alpha-lipoic acid, vitamins C and E), and/or increase ETC substrate availability (dichloroacetate and thiamine); and (3) feed the mitochondria (L-carnitine) [27]. However, in recent years, emerging experimental strategies have been proposed which are focused on: (1) stimulating mitochondrial biogenesis, (2) modulating mitophagy and mitochondrial dynamics, (3) bypassing OXPHOS deficits, (4) using mitochondrial replacement therapy (MRT), and (5) treating chronic hypoxia [24]. Molecular Approaches Molecular strategies aim to override or correct genetic alteration to mitigate the consequent mitochondrial deficits. The nDNA-associated MDs can benefit from genetic approaches based on the targeted re-expression of the wild-type gene and from the genome editing techniques based on clustered regularly interspaced palindromic repeat/Cas9 (CRISPR/Cas9) technology, zinc finger nucleases (ZFNs), or transcription-activator-like effector nucleases (TALENs) [28]. However, mtDNA-related MDs must be treated with alternative methods. Indeed, unlike nDNA, not only is mtDNA not easily accessible due to the mitochondrial double membrane structure, but it is also present in multiple copies within the organelle [29]. Thus, the delivery of nucleic acids to mitochondria is challenging. Three main experimental approaches have been developed to overcome this issue: (1) allotopic gene therapy, (2) mtDNA-selective endonuclease-based strategy, and (3) mitochondrial genome editing [28]. The first approach aims to re-express the mitochondrial gene at the nuclear level-followed by the import of the gene-encoded protein into the organelle, where it is supposed to function properly-by adding a mitochondrial-targeting sequence (MTS) at the N-terminus. Although clinical trials began in LHON patients, several challenges must be overcome to increase the successful application of this treatment [30]. The second approach exploits the observation that a break in the DNA double-strand leads to rapid degradation of the mtDNA molecule. Consequently, the selective elimination of mutated mtDNA via specific endonucleases might promote a heteroplasmic shift towards wild-type mtDNA copies within the organelle [31][32][33]. However, the strategy can only allow the potential treatment of heteroplasmic conditions. The third approach is based on two recent mitochondrial base editor technologies that are potentially able to correct homoplasmic single-nucleotide changes [33]. On the one hand, the DddA-derived cytosine base editor (DdCBE), which catalyses the conversions of C-to-T (or G-to-A) in human mtDNA [34]. On the other hand, transcription-activator-like effector (TALE)-linked deaminases (TALEDs) introduce targeted A-to-G editing in human mtDNA [35]. However, these approaches are still under development, and further studies are needed to verify their efficiency and safety in vivo. Intercellular Transfer of Mitochondria Cell-to-cell mitochondria transfer is considered a type of intercellular interaction that occurs physiologically in organisms [36]. Our knowledge of mitochondrial cell communication was enhanced by two experiments performed in the early 2000s: (1) Rustom et al. discovered cell-to-cell migration of organelles via nanotubular structures [37]; (2) Spees et al. demonstrated the natural transfer of wild-type mitochondria from mesenchymal stem cells (MSCs) into parenchymal cells displaying mitochondrial dysfunction [38]. Since then, a considerable body of data suggests that mitochondria and their components can be actively released into the extracellular space and transferred between cells under both healthy and pathological conditions [39,40]. Accordingly, knowledge of mitochondrial function is shifting from a microscale (single cell) to a macroscale cellular perspective, drawing attention to communication between cells through exchange of mitochondria. The biological role of this interaction is still controversial and depends on (among other things) the metabolic state of the cells. In fact, mitochondria released from a donor cell might be degraded or integrated by a recipient cell [41]. According to some studies, transmitophagy processes are responsible for exogenous mitochondria degradation [42]: (1) Davis et al. show that the mitopsosis [43] of dysfunctional mitochondria from optic nerve neurons occurs in adjacent glial cells [44]; (2) Phinney et al. suggest the selective discharge of depolarized mitochondria from MSCs to macrophages. Importantly, the authors unexpectedly noted that the recipient macrophages integrate exogenous organelles in order to enhance their own bioenergetics [45]. In line with this, other research papers point out that damaged cells can also take up exogenous functional mitochondria from a donor cell and integrate them into their endogenous mitochondrial networks to improve biological processes [46]. This seems to be the case for: (1) astrocytes transferring mitochondria to neurons during focal cerebral ischaemia [47], (2) endothelial progenitor cells protecting the brain endothelium during stroke and brain injury [48], and (3) bone-marrow-derived mesenchymal stromal cells (BM-MSCs) providing mitochondria to diverse cell types in various conditions [49,50]. In short, intercellular communication between recipient and donor cells can generate what Liu et al. refer to as the "find me" and "save me" connection signal, a specific mechanism that occurs in damaged tissues [51]. In addition to the well-characterized apoptotic processes, other biochemical events have been observed in mitochondria under stress conditions. It could be that mtDNA escapes into the cytosol via the Bak/Bax permeabilization pathway, where it is identified by various pattern recognition receptors (PRRs) as a "foreign" damage-associated molecular pattern (DAMP). This process leads to the activation of the cGAS-STING1-TBK1-IRF3 cascade, which stimulates the expression of type I interferon and proinflammatory cytokine (IL-1, IL,4, IL-10 and TNFα) genes [52]. The mtDNA can also bind endosomal toll-like receptor 9 (TLR9) to boost a cellular proinflammatory response [53]. Given the cellular connections, one could boldly hypothesize that the proinflammatory response may also trigger a chain reaction involving the neighbouring cells. According to this view, a cell might both help the dysfunctional neighbouring cell to eliminate dysfunctional organelles and restore the pool of healthy mitochondria [54,55]. In view of the above, cell-to-cell mitochondrial transfer seems to be essential for the maintenance and restoration of the organism's homeostasis [56]. Mechanisms of Mitochondria Intercellular Transfer The molecular mechanisms responsible for mitochondrial transfer are far from being clearly and completely understood. The best-known molecular mechanisms for intercellular transfer of mitochondria are: (1) tunnelling nanotubes (TNTs) and (2) extracellular vesicles (EVs) (Figure 2). be the case for: (1) astrocytes transferring mitochondria to neurons during focal cerebral ischaemia [47], (2) endothelial progenitor cells protecting the brain endothelium during stroke and brain injury [48], and (3) bone-marrow-derived mesenchymal stromal cells (BM-MSCs) providing mitochondria to diverse cell types in various conditions [49,50]. In short, intercellular communication between recipient and donor cells can generate what Liu et al. refer to as the "find me" and "save me" connection signal, a specific mechanism that occurs in damaged tissues [51]. In addition to the well-characterized apoptotic processes, other biochemical events have been observed in mitochondria under stress conditions. It could be that mtDNA escapes into the cytosol via the Bak/Bax permeabilization pathway, where it is identified by various pattern recognition receptors (PRRs) as a "foreign" damage-associated molecular pattern (DAMP). This process leads to the activation of the cGAS-STING1-TBK1-IRF3 cascade, which stimulates the expression of type I interferon and proinflammatory cytokine (IL-1, IL,4, IL-10 and TNFα) genes [52]. The mtDNA can also bind endosomal toll-like receptor 9 (TLR9) to boost a cellular proinflammatory response [53]. Given the cellular connections, one could boldly hypothesize that the proinflammatory response may also trigger a chain reaction involving the neighbouring cells. According to this view, a cell might both help the dysfunctional neighbouring cell to eliminate dysfunctional organelles and restore the pool of healthy mitochondria [54,55]. In view of the above, cell-to-cell mitochondrial transfer seems to be essential for the maintenance and restoration of the organism's homeostasis [56]. Mechanisms of Mitochondria Intercellular Transfer The molecular mechanisms responsible for mitochondrial transfer are far from being clearly and completely understood. The best-known molecular mechanisms for intercellular transfer of mitochondria are: (1) tunnelling nanotubes (TNTs) and (2) extracellular vesicles (EVs) ( Figure 2). Under the stimulation of stress conditions, mitochondria can be transported bidirectionally between cell A and B via the TNT structure and mitochondria-containing EVs. The formation of TNT is driven by the actin, and mitochondrial transfer between TNT-linked cells is regulated by Miro1. EV endocytosis is mediated by the NAD + /CD38/cADPR/Ca 2+ pathway: Under stress, intracellular NAD+ increases and diffuses to the extracellular environment. CD38 then catalyses NAD + to produce cADPR, a second messenger that acts on Ryanodine Receptor (RyR) on the Endoplasmic reticulum to induce the release of the intracellular Ca 2+ . Following the increase of cytoplasmic Ca 2+ mediates the activation of exocyst complex, leading to the formation and release of vesicles. Transfer via TNTs In 2004, Rustom and collaborators identified TNTs as distinct structures for intercellular communication while working on PC12 rat pheochromocytoma cell culture [37]. TNTs are plasma membrane-derived tubular cytoplasmic extensions of 50-1500 nm in width and 5-120 in length [57,58]. These dimensions facilitate unidirectional or reciprocal movement of signalling molecules and cellular components, promoting long-distance cell-to-cell interconnection [59]. Different cell types have been reported to form TNTs with each other in both physiological and pathological states [60]. TNTs have an F-actin-based structure and rely on adaptor-motor proteins to mediate cargo transfer [61]. It is possible to speculate that mitochondrial transfer between TNT-linked cells shares similar mechanisms with Miro1-mediated axonal transport of mitochondria into neurons. Indeed, Miro 1 (the outer mitochondrial membrane (OMM) Rho-GTPase Ca2+-dependent protein) appears to play a central role in regulating mitochondrial mobility along the TNTs [62]. In neurons, Miro1 acts as a mitochondria-loaded vehicle that interacts with OMM mitofusins and the molecular motor kinesin-1 via the Milton adapter protein (TRAK1/2). In this way, mitochondria move along the tracks of parallel polarized microtubule array [63]. The formation of TNTs is driven by the F-actin cytoskeleton-remodelling-based projection system. Three steps are involved in this mechanism: the development of a membrane protrusion, its expansion, and its fusion with the target cell membrane [64]. Currently, many proteins and pathways related to nanotubes formation have been identified [60,65]. Transfer via EVs Another system for cell-to-cell transfer of mitochondria is mediated by EVs. EVs were identified in 1971 by Aaronson S. et al., who discovered that the freshwater phytoflagellate Ochromonas danica produces membranous extracellular macromolecules [66]. Since then, a large number of reports in the last three decades, have suggested EVs' involvement in intercellular signalling [67,68]. EVs are nano-sized circular phospholipid structures that are released into the extracellular space by almost all the cell types under both physiological and pathological conditions [69]. EVs are designated as biological cargoes because they carry and deliver a variety of bioactive molecules such as proteins, mRNAs, miRNAs, mtDNA, lipids (including cholesterol and cytokines), mitochondria, and their components [70,71]. They are classified in three general classes according to their nano-diameter and biogenesis: (1) exosomes (30-150 nm), which are of endocytic origin through inward budding of endosomal membrane; (2) microvesicles (150-1000 nm), which are formed by outward budding of the plasma membrane; and (3) apoptotic bodies (>1 µm), which are released by apoptotic cells [69]. Due to exosomes' limited size, only microvesicles (eMVs) might contain intact mitochondria and contribute to mitochondrial intercellular transfer [72,73]. Indeed, recent papers have indicated horizontal mitochondrial transfer from donor cells to host cells through eMVs: (1) Islam et al. showed that bone-marrow-derived stromal cells (BMSC) supply alveolar epithelial cells with functional mitochondria via eMVs in a Cx43-dependent process [74]; (2) Phinney et al. reported that MSCs deliver eMVs with depolarized mitochondria to macrophages under pathological conditions [45]; (3) Hough et al. showed that airway myeloid-derived regulatory cells (MDRCs) secrete eMVs, which transfer mitochondria to T cells in a model of allergic airway disease [75]; (4) Davis et al. discovered that retinal ganglion cells can release mitochondria-rich EVs, which are captured by astrocytes in the optic nerve papilla [44]; and (5) Hayakawa et al. found that astrocytes release mitochondrial EVs that are captured by damaged neurons to support neuronal survival [76]. On the one hand, the eMV-mediated mitochondrial intercellular transfer mechanism is regulated by the NAD+/CD38/cADPR/Ca2+ pathway that activate the exocyst complex, leading to the formation and release of vesicles [36]. In addition, the loading of mitochondria into EVs may depend on optic atrophy 1 (OPA1) and sorting nexin 9 (Snx9) proteins, although the precise process is unknown [77]. On the other hand, anchorage on recipient cells is mediated by integrins localized on the eMVs surface [78]. After attachment to recipient cells, EVs can either immediately fuse with the membrane of the host cell or be internalized by a variety of mechanisms, such as phagocytosis, micropinocytosis, and clathrin-dependent endocytosis [79,80]. Overall, these finding suggest that cell-to-cell mitochondrial transfer pathway paves the way for a novel approach to mitochondrial transplantation for the treatment of different disorders. Mitochondrial Transplantation (MT) As further expansion in the field of innovative therapeutic strategies for mitochondrial disorders, two scientific evidences supported the idea of testing mitotherapy based on mitochondria delivery methods [81]. On one hand, according to the endosymbiotic theory, α-proteobacteria-derived mitochondria could be integrated into the host cell's mitochondrial network [82]. On the other hand, evidence for cell-to-cell mitochondrial transfer suggested that mitochondria retain their bacterial ability to enter cells [83]. Many recent procedures of mitochondria transplantation rely on the transfer of exogenous functional mitochondria to injured cells to finally restore mitochondrial deficiency in several diseases [84]. In Vitro Methods for MT Mitochondrial transplantation (MT) is an innovative technique based on the possibility of introducing healthy exogenous mitochondria into dysfunctional cells or tissues to modulate mitochondrial function [85]. The concept of "mitochondrial transplantation" was developed in 1982 by the pioneering study of Clark and Shay. These researchers purified mitochondria from donor human fibroblasts affected by mtDNA mutations, which induced chloramphenicol (CAP) and efrapeptin (EF) resistance ( Figure 3A). Then, they transferred the organelles to antibiotics sensitive dysfunctional human recipient cells via simple coincubation method, and they were able to produce antibiotics-resistant vital mammalian cells, denominated "mitochondrial transformants" [86]. They also showed that antibiotics resistance could not be transferred from mouse to human cells. This mitochondrial transformation experiment led to some basic conclusions: (1) isolated mitochondria retain the ability to invade host cells through endocytosis mechanisms; (2) mtDNA from donor cells could be integrated into recipient cells, allowing them to transfer genetic material and induce functional changes; (3) integration of exogenous mitochondria from different species into the recipient mitochondrial network could potentially be limited by species barrier; and (4) cells might have variable mitochondrial uptake capacity according to their physiological state [87]. In 1988, King and Attardi confirmed that mitochondrial transformation is possible. They proposed the first precise MT technology for the transfer of healthy mitochondria in vitro. The researchers demonstrated that antibiotic resistance could be successfully generated in sensitive, partially mtDNA-depleted 143B human osteosarcoma cells through direct microinjection of human mitochondria isolated from CAP-resistant cell lines. They proved the retention of exogenous mtDNA copies in host cells from six to ten weeks after transplantation via the analysis of multiple mtDNA polymorphisms [88]. Although very accurate, the injection method was less efficient than Clark and Shay's coincubation protocol because of the limited number of target cells in each transplantation experiment and the damage to the recipient cell. These preliminary data suggested that the uptake of a few healthy mitochondria could very rapidly repopulate a mtDNA depleted cells. In 1997, Pinkert et al. successfully microinjected mice liver mitochondria into mice zygotes, which incorporated a high rate of foreign mitochondria [89]. These results represented an initial step in the development of mitochondrial replacement therapy (MRT) techniques, which in MDs could prevent the transmission of mutant mtDNA to the foetus [90] ( Figure 3B). l. Sci. 2023, 24, x FOR PEER REVIEW 9 of 23 cells becomes significant. As a matter of fact, stem cells from patients might be transplanted ex vivo and introduced back in an organism's injured area to reprogram and/or repair its metabolism [55]. Such an attractive approach pushed researchers to explore novel physical-principlebased MT techniques [101]. In 2016, Wu et al. utilised a photothermal nanoblade to deliver mitochondria isolated from MDA-MB-453 cells into 143B-Rho0 cells [102] (Figure 3E). In detail, the photothermal nanoblade techniques causes the opening of the plasma membrane via laser pulses and the delivery of mitochondria using a titanium-coated borosilicate glass micropipette ~5 mm in length and with a tip ~3 μm in diameter using a fluid pump. Although the respiration of 143B-Rho 0 cells seemed restored, only a few cells per hour could be transplanted in each experiment. Thus, the researchers adjusted the protocol with a biophotonic-laser-assisted surgery tool (BLAST) for the rapid massively parallel delivery of micron-sized large cargo into recipient cells [103]. In parallel, Macheiner's groups proposed a novel mitochondrial transfer method, called magneto-mitotransfer, which employs an anti-TOM22 magnetic bead to labelled mitochondria and carries them into host cells with the support of a magnetic plate ( Figure 3F). They reported the rapid autologous MT of human fibroblasts MRC-5 and the related functional changes. Moreover, in 2021, Sarcel et al. developed "MitoPunch", a pressuredriven MT device to transfer mitochondria to recipient cells seeded on a porous polyester membrane ( Figure 3G). They show the transfer of mitochondria isolated from HEK293T cells to 143B-Rho 0 cells and the host retention of exogenous mtDNA copies [104]. Overall, the success of MT seems to be influenced by several factors: (1) the method of isolating mitochondria, which stresses organelles and affects their viability; (2) the selection of the donor cell line, with stem cells being preferred; (3) the amount of mitochondria to be transplanted; and (4) the delivery method [105]. The need to deliver intact and functional mitochondria has therefore led researchers to develop new tools. In this scenario, Gäbelein's group recently developed a single-cell technology called FluidFM to transplant mitochondria between living cells without compromising mitochondrial and cellular integrity or viability. FluidFM combines atomic force microscopy, optical microscopy, and nanofluidics. This procedure uses specific probes that permit However, only in early 2000s, with the discovery of cell-to-cell transfer of mitochondria, did the real therapeutic potential of MT come to light. Since then, researchers' interest triggered several MT studies on coincubation methods to investigate mitochondrial uptake in vitro [91,92]. The development of two strategies proved essential to assess the validity of MT: (1) advances in confocal microscopy technology [93] and (2) availability of Rho 0 cells, completely deprived of mtDNA by ethidium bromide treatment [94], which are suitable models for proving mtDNA transfer [95]. Several MT approaches have been developed to improve in vitro transfer using chemical mediators. In 2017, Chang and his research group implemented Clark and Shay's technique by using a penetrating peptide (Pep-1) conjugated with mitochondria to induce pores in the host cell's plasma membrane to facilitate mitochondrial transfer. As a result, the conjugation of mitochondria isolated from human osteosarcoma 143B cybrid donor cells with Pep-1 promoted the mitochondrial internalization by a MELAS cybrid model ( Figure 3C). The confocal microscopy analysis showed that the exogenous mitochondria colocalized with host cells mitochondria and favoured the recovery of the mitochondrial membrane potential [96]. In neonatal rat cardiomyocytes (NRCMs), Maeda et al. used transactivators of transcription dextran complexes (TATdextran) to increase cellular internalization of exogenous mitochondria and improve mitochondrial recipient function [97]. Remarkably, Caicedo et al. designed a protocol called MitoCeption based on the application of centripetal force to directly transfer mitochondria, isolated from MSCs, to human breast cancer MDA-MB-231 cells. In this protocol, mitochondria suspension is slowly added to the recipient cells in a Petri dish, which is then centrifuged and placed in the cell incubator facilitating the transplantation. The confocal imaging, fluorescenceactivated cell sorting (FACS), and mitochondrial DNA analysis demonstrated the successful dose-dependent transfer of exogenous mitochondria to cancer cells [98]. Later, Cabrera et al. proposed a new transplantation protocol to simplify the MitoCeption procedure. Specifically, mitochondria are directly added into a microcentrifuge tube, where the recipient cells are in suspension. Following this, the tubes are centrifuged ( Figure 3D). The novel protocol appears to be quicker and successful [99]. This procedure allowed for an allogenic transplant in vitro, repairing the metabolic activity, mitochondrial mass, and mtDNA sequence stability in UV-damaged peripheral blood mononuclear cells (PBMCs). In 2018, Kim's research lab proposed an easy and quick centrifugation method without additional incubation steps. They first isolated mitochondria from human-umbilical-cord-derived MSCs via differential centrifugation, then transferred them into both L6 cells and human umbilical cord Rho 0 cells via centrifugation. At the end, in the recipient cells, they observed an increase in ATP production, mitochondrial membrane potential, and oxygen consumption levels and a decrease in ROS [100]. According to Kim's results, the application of an outer physical stimulus in MT experiments, such as centripetal force, appears to be successful regardless of cells' types and physiological states. However, two critical issues must be considered. On one hand, the possible mechanical damage to the recipient cells' plasma membrane. On the other hand, such a strategy cannot be performed in vivo. Here, the considerable therapeutic potential of stem cells becomes significant. As a matter of fact, stem cells from patients might be transplanted ex vivo and introduced back in an organism's injured area to reprogram and/or repair its metabolism [55]. Such an attractive approach pushed researchers to explore novel physical-principlebased MT techniques [101]. In 2016, Wu et al. utilised a photothermal nanoblade to deliver mitochondria isolated from MDA-MB-453 cells into 143B-Rho0 cells [102] ( Figure 3E). In detail, the photothermal nanoblade techniques causes the opening of the plasma membrane via laser pulses and the delivery of mitochondria using a titanium-coated borosilicate glass micropipette~5 mm in length and with a tip~3 µm in diameter using a fluid pump. Although the respiration of 143B-Rho 0 cells seemed restored, only a few cells per hour could be transplanted in each experiment. Thus, the researchers adjusted the protocol with a biophotonic-laser-assisted surgery tool (BLAST) for the rapid massively parallel delivery of micron-sized large cargo into recipient cells [103]. In parallel, Macheiner's groups proposed a novel mitochondrial transfer method, called magneto-mitotransfer, which employs an anti-TOM22 magnetic bead to labelled mitochondria and carries them into host cells with the support of a magnetic plate ( Figure 3F). They reported the rapid autologous MT of human fibroblasts MRC-5 and the related functional changes. Moreover, in 2021, Sarcel et al. developed "MitoPunch", a pressure-driven MT device to transfer mitochondria to recipient cells seeded on a porous polyester membrane ( Figure 3G). They show the transfer of mitochondria isolated from HEK293T cells to 143B-Rho 0 cells and the host retention of exogenous mtDNA copies [104]. Overall, the success of MT seems to be influenced by several factors: (1) the method of isolating mitochondria, which stresses organelles and affects their viability; (2) the selection of the donor cell line, with stem cells being preferred; (3) the amount of mitochondria to be transplanted; and (4) the delivery method [105]. The need to deliver intact and functional mitochondria has therefore led researchers to develop new tools. In this scenario, Gäbelein's group recently developed a single-cell technology called FluidFM to transplant mitochondria between living cells without compromising mitochondrial and cellular integrity or viability. FluidFM combines atomic force microscopy, optical microscopy, and nanofluidics. This procedure uses specific probes that permit minimally invasive access to cells and fluid flow to extract and inject mitochondria ( Figure 3H). However, the technique is still not appropriate for use in vivo [106]. For in vitro and in vivo applications, the use of specific physiological vehicles with low immunogenicity and toxicity, such as extracellular vesicles (EVs), improved the preservation of mitochondria integrity when delivered (Table 1). Two main studies explored the novel EV-mediated MT method ( Figure 3I). In 2021, Ikeda et al. isolated EVs from humaninduced pluripotent stem cell (iPSC)-derived cardiomyocytes (iCMs) and coincubated them with recipient iCMs damaged by hypoxia. The confocal microscopy analysis indicated that exogenous mitochondria were transferred into host cells and fused with their endogenous mitochondrial networks, leading to a significant increase in ATP generation and an improvement in contractility [107]. In the same year, Peruzzotti-Jametti et al. investigated the hypothesis that neural stem cells (NSCs) release and transport functioning mitochondria through EVs into host cells to modulate mitochondrial activity. First, they verified whether the EVs were loaded with mitochondria via microscopic and functional analyses. Then, they incubated them with mtDNA-depleted L929 Rho 0 cells. Mitochondrial function was then assessed in the recipient cells [71]. In Vivo Methods for MT In addition to the reported in vitro MT studies, mitochondria can also be transplanted directly into animal models. In 2009, McCully et al. [108] presented the first in vivo MT study in an ischaemia-reperfusion heart model using an allogenic transplant in a rabbit. They isolated functional mitochondria from the left ventricular tissue of a healthy animal and injected them into the cardiac ischemic area of a different rabbit just before reperfusion. The study highlights the potential of MT to improve functional recovery of the heart and survival of cardiomyocytes after ischaemia. In addition, cardiomyocytes could uptake exogenous mitochondria just 2 h after injection. Further studies demonstrated that MT significantly reduces cardiac ischaemia markers (creatine kinase-MB and cardiac Troponin I) and the apoptosis protein Caspase-3, as well as infarct size [109]. Next, Cowan et al. performed two in vivo MT procedures: (1) direct injection of mitochondria into rabbit myocardial ischemic zone and (2) vascular perfusion of mitochondria through the coronary artery. Analysis of the results demonstrated that direct mitochondria injection proved to be more efficient in cardiac protection [110]. To summarize, direct mitochondrial transplantation appears to be the most promising way to modulate cardiomyocytes metabolism. Since then, several studies in vivo have been performed in different animal tissues, such as liver, lung, and brain [111] (Table 2). Worth to mention, two studies demonstrate the feasibility of MT on patients with ischaemia-reperfusion injury and single large-scale mtDNA deletion syndromes (SLSMDs). Intravenous reinfusion of CD34 + ex vivo transplanted cells. [160] In the first case, autologous mitochondria were isolated from a piece of healthy rectus abdominis muscle and direct injected to the myocardium of five paediatric patients who required central extracorporeal membrane oxygenation (ECMO) support for ischaemiareperfusion dysfunction. The data show improvement in ventricular function within several days after treatment, but future studies are necessary to demonstrate the efficiency of the strategy [112]. Recently, in a compassionate use study, a mitochondrial augmentation therapy (MAT) approach has been applied on six patients with SLSMDs [160]. MAT is a cell technology platform in which autologous hematopoietic stem and progenitor cells (HSPCs) are augmented ex vivo with mitochondria obtained from donor cells or tissue. HSPCs have been shown to migrate to distal tissues and to abrogate disease-related deterioration of mitochondrial dysfunctions or metabolic disease. In this case, CD34+ cells were collected from a patient with a non-inherited primary mitochondrial disease, and PBMCs were collected from the maternal donor. Using a coincubation method, healthy mitochondria isolated from maternal PBMCs were transplanted into the patients' CD34+ cells, which in turn were intravenously reinfused into the patients. MAT resulted in decreased heteroplasmy of mtDNA deletion, increased mtDNA levels, and improved ATP content in peripheral blood mononuclear cells in four out of six patients, as well as improved muscle strength and endurance in two individuals. Despite these encouraging results in peripheral tissues, this strategy does not reach the brain, which is also affected in SLSMDs (Table 2). However, this approach is not applicable in the case of homoplasmic mutations or in the presence of high heteroplasmy levels. Several data show that mitochondrial dysfunction plays a pivotal role in neurodegeneration. Indeed, cells of the nervous system require a considerable amount of energy to carry out synaptic transmission, neurogenesis, and neuronal differentiation [161]. Therefore, MT strategy has also been experimented upon in animal models to treat neurodegenerative diseases and to investigate mitochondria supply in the brain [162]. For example, Shi et al. reported that systemic injection of mitochondria isolated from human hepatoma cells (HepG2 cells) into the brain of PD mice mitigates the progression of PD by improving ETC function and decreasing ROS production [143]. To facilitate mitochondrial transplantation, in another paper, the authors injected PeP-1-conjugated mitochondria into the medial forebrain bundle (MFB) of 6-OHDA PD rats. As a result, neuronal mitochondrial function and dynamics improved in the substantia nigra (SN) [142]. Additionally, encouraging data for MT in the CNS came from experimental models of schizophrenia (SZ) and Alzheimer's disease [162]. Dos-Santos et al. demonstrated the therapeutic potential of MT in the crush-based animal model of optic-nerve glaucoma. The group performed an intravitreal injection of liver-isolated mitochondria and showed that MT transiently protects RGCs and reduces oxidative stress [156]. It is worth noting that the route of administration is crucial to the effect of MT in vivo. Currently, four routes of administration are used: (1) intracerebral injection, an invasive and painful procedure; (2) intrathecal injection, an alternative to bypassing the blood-brain barrier (BBB); (3) systemic or intracarotid administration, a less invasive and safer route; and (4) intranasal administration, which very recently has been shown as a simple and effective delivery system [140,144,163]. There are two main limitations to consider when performing MT in the CNS. First, the BBB which protects the brain from compounds that are toxic to brain neurons [164]. Second, the immune response in the brain is particularly alarming, and mitochondria are strongly immunogenetic organelles [165,166]. Indeed, some studies have investigated the immune reaction after mitochondrial transplantation in different tissue in vivo, confirming that there is an activation of the immune system following injection of mitochondria [167,168]. To reduce the immunological risk associated with MT, it would be very useful to untwist the mechanism underlying an immune response during this procedure [111]. In summary, future research on in vivo MT should focus on bypassing the BBB and preventing harmful immune responses in the brain. The study by Pluchino's group mentioned above has provided a strategic method in this direction using NSC-derived EVs [71]. Remarkably, the delivery through vesicles not only preserves mitochondrial integrity but also reduces the damage associated with the immune response. Indeed, EVs hiding mitochondria from microglia and macrophages would act as a Trojan Horse, thus reducing immune reactivity. Challenges and Future Prospective Although mitochondrial transplantation has emerged as a possible therapeutic strategy for MDs, there are still several challenges to master before it can be used in clinical applications. While much progress has been made concerning the delivery, the remaining and still unsolved issues include: (1) mitochondrial yield and purity, (2) mitochondrial long-term storage, and (3) transplant rejection [169]. The isolation/preservation of undamaged and coupled mitochondria is an essential point. Indeed, it is now well known that dysfunctional mitochondria trigger a systemic or tissue-specific inflammatory response through various mechanisms, such as activation of inflammasomes and release of DAMPs [170]. Several methods of isolating mitochondria rely on differential centrifugation, which allows a rapid preparation characterised by high yield and low purity [171,172]. Techniques such as density gradient centrifugation and affinity purification by magnetic beads were used to further purify mitochondria or to separate distinct populations of mitochondria [172,173]. However, unlike differential centrifugation, yield is lower and purity is higher in these procedures [174]. Furthermore, isolation methods are time-consuming, and this could reduce mitochondrial viability and transplantation success [175]. Therefore, it is very important to define a suitable and rapid isolation method for mitochondrial transplantation. According to McCully, isolated mitochondria can remain active and coupled for approximately 1 h on ice, and storage after this time significantly affects transplantation efficiency [175]. This storage range limits the clinical applications of MT. Thus, the therapeutic uses of MT could be substantially broadened if isolated mitochondria could be employed as a storable preparation [169]. Currently, when isolated mitochondria are cryopreserved at −80 • C, impairment of OMM integrity occurs [111]. In order to overcome this critical aspect, several studies of mitochondrial long-term storage have been made using cryoprotectors, such as DMSO [176] and trehalose [177]. However, despite the preservation of the integrity, these methods cause a reduction in mitochondrial functionality. Therefore, it is crucial to develop a method of mitochondrial cryopreservation to maintain both their stability and bioenergetic capacity. Lastly, transplant rejection might occur in vivo. As mentioned above, some studies report the increase of autoimmune and inflammation markers after allogenic mitochondrial injection in different tissue in vivo [168]. This response could lead to the elimination of externally provided mitochondria from the organism to maintain a homeostatic state [167]. Two approaches could bypass this problem: (1) autologous mitochondrial transplantation, which caused no significant increase in immunity markers in animal models [109]; (2) mitochondrial delivery via EVs, which have lower immunogenicity compared to isolated mitochondria [178]. Interestingly, the extrusion of mitochondrial components via the generation and release of mitochondrial-derived vesicles (MDVs) from both the outer and inner membranes of mitochondria has been recently described. MDVs are mainly engaged for mitochondrial quality control and transporting selected mitochondrial cargoes [179]. Depending on the mitochondrial stress level, the cargoes can have a different fate. It has been demonstrated that they can be delivered to the late endosome/multivesicular bodies for packaging into EVs, to lysosomes for degradation, and to peroxisomes [180]. MDVs from damaged mitochondria might carry mtDNA, which-once inside EVs-could be released to the extracellular space. Neighbouring cells might recognise the genetic material as DAMPs and activate proinflammatory pathways [181]. These aspects suggest that MDVs generation and trafficking is regulated in a different way under pathological conditions. Future research focussing on MDVs as possible therapeutic tools to deliver mtDNA will have to take into account all the above-mentioned crucial issues. Moreover, it has recently been shown that the retention time of exogenous mtDNA in recipient cells does not last long [91,182]. Indeed, once in the recipient cell, exogenous mitochondria could be integrated into the endogenous mitochondrial network via the fusion pathway (MFN1/2 and OPA1) and/or undergo mitophagy [183]. Chang and colleagues reported the increase in mitochondrial-fusion-related protein expression and inhibition of fission after transplantation [184]. Fusion and fission are two conserved processes which control mitochondrial dynamics, playing a key role in the regulation of mitochondrial remodelling and cellular metabolism [185,186]. Fission is mediated by the interaction of a cytosolic GTPase, dynamin-related protein 1 (DRP1) with four OMM proteins (Fis1, MiD49, MiD51, Mff), resulting in the division of the mitochondrion into two similar organelles [187]. Fusion of the outer and inner mitochondrial membranes, on the other hand, requires the action of two distinct classes of integral DRPs, such as the mitofusins (MFN1/MFN2) located at the OMM level and optic atrophy protein 1 (OPA1), located in the intermembrane space [3]. Defects in both mitochondrial fission and fusion impair mitochondrial function and contribute to numerous pathological conditions. Collectively, mitochondrial dynamics evolve as a compensation mechanism to maintain heteroplasmic condition in cells. Indeed, when mutated mtDNA accumulates, fusion compensates mitochondrial function by mixing wild-type and defective mtDNA copies [188]. On the other side, fission antagonizes fusion processes and triggers segregation of damaged mitochondria for subsequent elimination by mitophagy [81]. Amongst several other challenges when transplanting mitochondria, it is crucial to better understand how mitochondrial fission and fusion are regulated in order to promote the integration of exogenous mtDNA copies into the host endogenous mitochondrial network. On the other hand, the selective elimination of external mitochondria represents a barrier to organelles' retention. Two reasons could lead to transplant rejection: (1) genome balance; and (2) haplotype matching between exogenous and endogenous mtDNA (for a comprehensive review, see [189]). The concept of "genomic balance" refers to the relationship between mitochondrial and nuclear genomes to maintain cellular destiny and homeostasis [190][191][192][193]. Thus, the administration of exogenous mtDNA copies with different genetic profiles could disrupt the existing balance between mtDNA and nDNA in recipient cells. These mtDNA-nuclear mismatches may lead to complications in male mice [194]. At the same time, mtDNA-mtDNA interactions could be problematic. For example, in mice, the combination of two mtDNA haplotypes from the same subspecies resulted in physiological alterations [195]. Collectively, it could be that to restore homeostasis, either the host cell is reprogrammed or exogenous mitochondria are eliminated as a compensatory mechanism. Consequently, future investigations should focus on unravelling the mechanisms behind the mitonuclear and/or mtDNA-mtDNA interactions to better understand how these relations could affect mitochondria transplantation. The transplantation of autologous mitochondria would prevent this mismatch as they are derived from the same patient. Conversely, this would not be appropriate for individuals with congenital mitochondrial disorders. To minimize mtDNA difference during allotransfer, it has been proposed that mitochondria derived from genetically close family members from the maternal lineage without the mutation or with very low heteroplasmy levels should be used [169]. A future solution may be the application of the in vitro mtDNA editing approaches combined with the possibility to generate induced pluripotent stem cells (iPSC) from patient somatic cells. In this scenario, patient-derived iPSC could be used tout court or differentiated in specific cell types, such as mesenchymal or neural precursor cells, in order to induce the reversion of the pathogenetic mutation outside of the body. Subsequently, after appropriate quality control and selection steps, these modified cells may be used as ex vivo donors for mitochondrial transplantation. Conclusions The main aim of this review is to investigate mitochondrial transplantation as a potential therapeutic strategy for MDs. Diagnosis, treatment, and the development of effective therapies are severely limited by the variability in MDs' clinical expression. Unlike heteroplasmic conditions, MDs caused by homoplasmic mtDNA mutations do not benefit yet from advances in molecular approaches. Therefore, providing healthy mitochondria in the presence of dysfunctional mitochondria appears to be an attractive option, paving the way to novel, intriguing alternatives. Interest in mitochondrial transplantation was sparked by the discoveries of intercellular transfer of mitochondria in the early 2000s. Since then, several studies have been carried out to improve MT techniques and evaluate their validity in different disease models both in vitro and in vivo. The rationale of MT is to induce a heteroplasmic shift in a homoplasmic MDs condition to attenuate or prevent the progression of pathological phenotypes. Three aspects must be considered for a successful transplantation: (1) maintenance of mitochondrial function and integrity during transfer, (2) mitochondrial immunogenicity, and (3) metabolic/genetic interaction of exogenous mitochondria with mitochondria of the recipient cells. Once these conditions are met, mitochondrial transplantation could become a reality to overcome the limitations of currently available approaches, designed to correct specific mutations with the unprecedented advantage of having a broad applicability in various mitochondrial diseases.	2023-01-24T16:35:44.013Z	2023-01-19T00:00:00.000	{ "year": 2023, "sha1": "ec4ecf87140827a2987613236e94ec59af64d37e", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/1422-0067/24/3/1969/pdf?version=1674121733", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "973780cd0d00228fc7b5c8533b70ac954c6c048a", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [] }
18690577	pes2o/s2orc	v3-fos-license	High waste-to-biomass conversion and efficient Salmonella spp. reduction using black soldier fly for waste recycling The recycling of organic waste worldwide is not effective, which leads to water pollution and loss of potential crop fertilizers. Available resources have to be used more efficiently as the world population increases. An innovative solution is to use insects for the management of organic waste. Here, we used black soldier fly to convert organic waste into animal feed protein, as fly larvae, and plant fertilizer, as compost residue. A continuous fly reactor was monitored for 9 weeks. We analyzed physicochemical and microbial parameters, and we evaluated the sanitary risk. Results show 55.1 % of material degradation and 11.8 % of biomass conversion based upon total solids. We observed higher levels of N and P in the treatment residue than in the inflow material. Results also show a lower concentration of Salmonella spp. and viruses. Compost treatment with black soldier fly is therefore an efficient system for nutrient recycling. Introduction The volume of solid waste generated every year is increasing worldwide. Although municipalities strive to provide solid waste management, much of the generated waste is left untreated (Hoornweg and Bhada-Tata 2012). A large fraction of solid waste in low-and middle-income countries consists of organic material, e.g., food waste, animal manure, and market waste. If left untreated, the valuable energy and nutrients contained in the waste can contribute to an increase in annual greenhouse gas emissions (UNEP 2010) and a leaching of nutrients into water bodies, causing eutrophication (Sharpley et al. 1994). Over and above this, there is a risk of spreading the pathogens contained in the organic waste. The waste management coverage in high-income countries is considerably higher, but at the same time the waste generated per person is up to three times greater and still increasing (Hoornweg and Bhada-Tata 2012). The proportion of organic waste is around 30 % and only a small proportion of the nutrients returns into the food cycle (Cordell et al. 2009). This is a growing concern, since the nitrogen and phosphate cycle has been identified as key factors that has to be maintained within certain levels for the planet to be able to support human existence in the future (Rockström et al. 2009). One way of securing human well-being is, according to Griggs et al. (2013), sustainable production systems. In a report by the United Nations Food and Agricultural Organization, it was stated that the consumption of insects in the food sector, as animal feed and directly for human consumption, is one such solution (van Huis et al. 2013). The black soldier fly, encountered between 46°N and 42°S (Martínez-Sánchez et al. 2011), is of particular interest. Their ability to consume prodigious amounts of organic waste has long been known. Black soldier fly larvae have been reported to consume and degrade a number of organic materials with material degradation up to 70 % (Diener et al. 2011). The larvae in the final larval stage, the prepupae, crawl out of the feeding material to find a dry and dark place to pupate (Sheppard et al. 2002) and are thus self-harvesting. The prepupae contain around 40 % protein and 30 % fat and have been demonstrated to be a good alternative feed in fish (St-Hilaire et al. 2007) and pig (Newton et al. 1977) production, with the potential to replace fish meal and fish oil as source of animal feed protein. Replacing fish meal would alleviate the pressure on the wild fish population (Tacon and Metian 2008). With the increasing consumption of fish and meat, the demand for animal feed protein has increased greatly, while its availability is limited. As a consequence, the price of fish meal has reached new heightsvalued USD 1,820 per tons (World Bank, May 2014)making the production of compatible alternative animal feed protein a lucrative business. By producing an economically valuable product, organic waste management could become a financially sustainable system that would increase resource reuse efficiency and reduce negative environmental impacts (Fig. 1). Hygiene aspects play an important role when dealing with organic waste and animal feed (Sidhu and Toze 2009). In several studies, black solider fly larvae waste treatment systems have been shown to remove bacteria in the Enterobacteriaceae family: the concentration of Salmonella spp. was reduced by 7 log 10 in 8 days in a small batchoperated fly larvae system treating fecal sludge (Lalander et al. 2013); Erickson et al. (2004) observed the reduction of Salmonella enterica Serovar Enteritidis and Escherichia coli O157:H7 in cattle manure, while Liu et al. (2008) have reported on the reduction of E. coli in poultry manure. However, a l t h o u g h a h i g h r e d u c t i o n i n b a c t e r i a o f t h e Enterobacteriaceae family occur, reduction of other organisms has not been observed. All these systems were operated in batch mode. The question is whether the same results would be obtained in a continuous flow reactor. In a large-scale fly larvae waste management system, it is likely that a continuous flow reactor would be implemented, as it would be too timeconsuming to operate in batch mode. One hypothesis is that the operation over a long period would support the buildup of a microbial community, helping to break down the waste material and thus improve the hygienic quality of the treatment residue. The objective of this study was to establish a continuous flow fly larvae reactor and to analyze important process and hygiene parameters in order to evaluate material degradation, waste-to-biomass conversion rate, nutrient flow, and average weekly reduction of evaluated pathogens and indicator organisms. Materials Pig manure (21.1±1.7 % total solids) was collected on a pig farm in Uppsala municipality, Sweden. Dog food (Puppy Original, Pro Plan, Purina) mixed with water (21.9±0.2 % total solids) was used as the model substrate for organic waste (Vinnerås et al. 2003). Human feces were collected fresh in plastic bags and stored at −20°C immediately upon collection. A mixture of pig manure, dog food, and human feces was prepared (4:4:2; 28.7±1.2 % total solids), divided into feeding portions, and kept at −20°C until use. Fig. 1 A graphical representation of the concept of fly larvae composting: organic waste is consumed by fly larvae, in the sixths and final larval stage, the prepupae migrate out of the compost. The prepupae can be used as animal feed and the treatment residue used as organic fertilizer. The loop is closed when the animal manure and food waste is diverted as substrate into the fly larvae compost Ascaris suum, which infects pigs, is often used as model for Ascaris lumbricoides, which infects humans (Johnson et al. 1998). A. suum extracted from the feces of infected pigs were purchased from Excelsior Entinel, Inc. and received in aqueous solution at a concentration of 100,000 eggs mL −1 . Propagation of bacteriophage фX174 to a concentration of 10 9 plaque forming units (PFU) mL −1 was performed in unselective bacterial nutrient medium (NB, Oxoid AB, Sweden) using the host strain Escherichia coli (ATCC 13706). The phage was collected by centrifuging the solution at 2,000×g for 10 min followed by sterile filtration and kept at 4-6°C until use. Black soldier fly eggs were received twice a week from the Research Institute of Organic Agriculture, Frick, Switzerland. Experimental set-up and sampling Three fly larvae reactors consistent of white 37 L Sortera bins (37×55×28 cm), with a lid and slanting wall for prepupal migration, that were placed inside 45 L black Samla bins (39× 57×28 cm) with a lid, both from IKEA (Uppsala, Sweden). The Samla bins were used to keep light out of the reactor and as collection bins for migrating prepupae. The reactors were kept at 25°C throughout the experiment. Black soldier fly egg packages were received in cardboard strips. Individual egg packages were divided into small (around 150-250 larvae), medium (around 400-500 larvae), and large (around 800-1,000 larvae) egg packages and hatched individually. Twice a week, a more or less equal number (based on the number and size of egg packages) of 5-10-day-old larvae were added to each reactor. The eggs were hatched in the same feed mixture used in the reactors. The total number of larvae at steady-state was estimated to be 7,200. The reactors were fed three times a week. The preprepared feed mixture was thawed overnight and added in the mornings. The feed portions were adjusted to the estimated number of larvae present in the reactors, to a feeding rate of 100 mg per larvae per day (Diener et al. 2009); at steady-state each unit was fed 5.4 kg per week. The reactors were operated as plug-flow: old material in the back of the reactor was pushed 5 cm to the front prior to fresh feed mixture being added. All samples were collected from the front 10 cm of the reactor. On the first feeding occasion in week 4, the feed mixture was inoculated with 1 mL A. suum of concentration 100,000 egg mL −1 . On the first feeding occasion in week 7, the feed mixture was inoculated with Reovirus type 3, Canine Adenovirus I, and Porcine enterovirus I. The reactors were weighed weekly throughout the course of the experiment. All migrating prepupae were counted/weighed. Once a week, triplicate samples of 5 g were collected in each reactor and analyzed separately to determine the concentration of bacteria and bacteriophages in the influent and effluent material. The concentration and viability of A. suum eggs was determined on days 7, 14, and 28 after inoculation. The concentration of viruses in the material was determined 2, 4, 7, 9, 11, and 14 days after inoculation. Samples for physicochemical analysis were collected at the start and end of the experiment. Physicochemical analyses The material was dried at 105°C for 14 h to determine total solids and at 550°C for 6 h to determine volatile solids, i.e., the organic content. A radiometer electrode was used to measure pH. All analyses were conducted at room temperature on 10 g of sample diluted with 50 mL deionized water and were left to settle for 1 h prior to analysis. In a closed container, 1 g of sample was mixed with 20 mL deionized water. The material was thoroughly mixed and diluted 1:10 in deionized water. Spectroquant © test kit number 114544 was used for total ammonium nitrogen (NH 4 -N) analyses and kit number 114764 for nitrate (NO 3 -N) analyses. In a 50-mL volumetric flask, 0.5-0.6 g samples were dissolved into 10 mL concentrated H 2 SO 4 and brought to "rolling boil" on a hotplate. When the sample had visually dissolved, an additional 5 mL H 2 SO 4 was added. The sample was allowed to cool and was thereby diluted-1:100 for total nitrogen analysis and 1:50 for total phosphorous analysis-in deionized water. Spectroquant © test kit number 1.00613.0001 was used for total nitrogen analysis and number 1.00673.0001 was used for total phosphorous analysis. Bacterial and bacteriophage analyses The buffer used for all experiments was buffered NaCl peptone water with Tween 80 at pH 7, hereafter referred to as Tween buffer. Sample (5 g) was extracted and further diluted in the Tween buffer. A 100 µL volume of selected dilution was spread on Slanetz-Bartley agar (Oxoid AB, Sweden) and incubated at 44°C for 48 h for Ent. spp. enumeration and on xylose lysine desoxycholate agar (XLD) (Oxoid AB, Sweden) containing 0.15 % sodium-novobiocin and incubated at 37°C for 12 h for Salmonella spp. enumeration. The plates were enumerated with a detection limit of 100 CFU mL −1 . To lower the detection limit of Salmonella spp. concentration, 200 μL per plate of sample of dilution 10 −1 was spread on five plates per sample and enumerated with a detection limit of 10 CFU mL −1 . To lower the detection limit further, the sample was enriched as follows: 5 g per unit was immersed in 45 mL buffered peptone water and incubated at 37°C for 24 h, upon which 100 μL was immersed into Rappaport-Vassiliadis salmonella enrichment broth (RVB) and incubated at 41.5°C for 24 h. Positive results were evaluated by spreading 10 μL on a XLD plate and incubated at 37°C for 12 h, lowering the detection limit to 1 CFU g −1 . Total thermotolerant coliforms were enumerated in double layer agar using violet red bile agar (VRB) (Oxoid AB, Sweden); 1 mL of sample was mixed with 7-8 mL of agar, and upon solidification of the first layer, an additional 7-8 mL agar was added. The plates were incubated at 44°C for 24 h and counted with a detection limit of 10 CFU g −1 . For sampling of coliphage фX174, the host E. coli (ATCC 13706) was cultured in unselective microbial medium (NB, Oxoid AB, Sweden) at 37°C for up to 4 h. A 1 mL sample volume of suitable dilution was mixed with 2 mL soft agar and 1 mL host solution, poured onto blood agar base (BAB) plates (Oxoid AB, Sweden) and incubated at 37°C for 16±2 h. A. suum ova extraction A. suum ova extraction was conducted according to the procedure of the United States Environmental Protection Agency (EPA/625/R-92/013), using 30 g of material. Prior to incubation, 15-20 of the extracted ascaris eggs were verified. The tubes were incubated at 28°C for 30 days in 0.1 N H 2 SO 4 . After approximately 30 days, the incubated material was analyzed under the microscope (~100 ova per sample). A 1 mL Sedgwick Rafter Counting Cell was used to count eggs; pre-larvae and larvae were counted as viable. Virus analyses Samples for virus analysis were diluted 1:5 in NaCl peptone water with Tween buffer, further diluted 1:2 in beef extract with glycine, and kept at −70°C until analysis. After thawing, the samples were vortexed for 15 min at 60 rpm, centrifuged at 3,000×g for 15 min at 5°C, and the supernatant filtered through a 0.45 μm filter. The samples were then filtered through a sterile gel filter (PD10, GE Healthcare) to remove toxicity. The samples were analyzed by endpoint titration using the following cell cultures: MDCK (ATCC CCL-34) for adenovirus, BHK (ATCC CCL-10) for reovirus, and swine kidney cell line (SK-6) for enterovirus, all cells were prepared 24 h prior to infection. To assess the virus concentration in the ingoing material, the samples were diluted tenfold and eight replicates of 50 μL were assayed for each dilution, while 48 replicates of the lowest dilutions were used to lower the detection limit. Material conversion The waste-to-biomass conversion rate (BCR) on a total solids basis was calculated by: where Mig.PP Tot.TS and Waste Tot.TS were the total solids in the migrating prepupae and waste, respectively. The calculations for the nutrient flow balance and material degradation were based on the assumption that: where Waste Tot.Ash , Residue Tot.Ash , and Mig.PP Tot.Ash were the total amount of ash in the inflow waste, the residue, and the migrated prepupae, respectively. The amount of ash in the residue was estimated based on the measured values in the inflow material and migrated prepupae. The material reduction (Mat TS.Red ) on a total solids basis was calculated as: where Residue Tot.TS and Waste Tot.TS were the total amount of total solids in the treatment residue and the inflow waste, respectively. Reduction of microorganisms The log 10 reduction of evaluated microorganisms that occurred during 1 week in the treatment (Δ log 10 Red) was calculated as: where C Mat i was the estimated concentration of the material in the treatment unit after substrate inoculation week i and C Mat:out iþ1 the theoretical concentration of the material in the treatment unit prior to substrate inoculation 1 week later (i+1). The theoretical concentration of the material in the treatment unit week i (C Mat i ) was calculated as: where C Sub:in i and m Sub:in i were the concentration and mass, respectively, of the inflow substrate week i and C Mat i and m Mat i the concentration and the mass, respectively, of the treatment unit week i. Mass balance The material in the fly larvae composts was assumed to be mainly polysaccharides (cellulose and starch); thus, the chemical formula C 6 H 10 O 5 was used when calculating the amount of oxygen needed for the respiration. For each polysaccharide, six oxygen molecules were required for full degradation into six carbon dioxide and five water molecules. For the mass balance, it was assumed that the total mass of material reduced on organic content basis had been completely respired into carbon dioxide and water. The nutrient mass balance was conducted by multiplying the measured concentration of the evaluated nutrients with the mass of the total solids, for each fraction. Statistical analyses Two-tailed, paired Student's t test (95 % confidence interval) was used to evaluate whether the reduction of microorganisms was significant and if a statistically significant difference occurred between the total solids, organic content and the concentrations of nutrients in the inflow and outflow materials. Linear regression with t test was used to evaluate if the pH and the reduction of microorganisms changed over time. All analyses and graphical presentations were conducted in R (R Core Team 2012). Results and discussions Three black soldier fly compost reactors were continuously fed a mixture of human and pig manure and organic waste during 9 weeks. Once weekly, samples of the outflow material were collected and physicochemical and microbial parameters analyzed. The process efficiency in the fly larvae reactor was verified by determining the material degradation, waste-tobiomass conversion rate, mass balance of the organic content and nitrogen and phosphorous, the average weekly reduction of the evaluated microorganisms, and the survival of Adeno-, Reo-and Enterovirus and A. suum eggs in the process. Change in physicochemical properties The organic content significantly (p<0.05) decreased in the process, from 86.6 % in the inflow material to 72.9 % in the outflow material (Table 1). The pH was found to significantly (p<0.05) increase with time; the average pH in the inflow material was 5.6, while the average pH in the outflow material was 8.3. There was no significant (p>0.05) difference in the change in pH over time between the three reactors. The prepupae were 12.3 % ash and 59.9 % moisture. The concentration of total phosphorous significantly (p<0.05) increased per gram total solids in the process, from around 33 mg g −1 to almost 48 mg g −1 ( Table 2). The concentration of total nitrogen, organic nitrogen and nitrate was not significantly different (p<0.05) in the inflow and outflow material, while the concentration of total ammonium nitrogen significantly increased (p<0.05). Some of the nitrogen was incorporated into the prepupal biomass in the form of proteins. The nitrogen profile changed significantly in the process: the organic nitrogen present in the inflow material was converted into ammonia, changing the organic nitrogen to total nitrogen ratio from 0.91 to 0.78 and the C/N ratio decreased from 16.4 to 14.3. Material conversion and nutrient flows The total flow of nutrients was estimated for the fly larvae composting process (Fig. 2). It was found that close to half of the total nitrogen in the inflow material was degassed as ammonia, while just over 10 % was incorporated into the larval biomass. Approximately 30 % of the total phosphorous and 10 % of the ash was not accounted for in the system, as these two components should not disappear it is likely that it is due to experimental and/or analytical factors. The flow balance was estimated by multiplying the concentration of total phosphorous for the inflow and outflow with the corresponding mass. It could be that the total phosphorous concentration was not constant for all outflow material. Less than 3.5 % of the total phosphorous was incorporated into the larval biomass. The total material reduction was 55.1 % on a total solids basis. Of the total 32 kg flowing into the fly larvae compost, Table 2 Concentrations (mg g −1 total solids (TS)) of nutrients in inflow and outflow materials and the prepupae, and the ratio of Org-N/ Tot-N and C/N in the inflow and outflow materials; average values and standard deviation (SD) of the three fly larvae reactors presented a Significant difference (p<0.05) in concentration per gram total solids between in-and outflow material . 2 The chart show the total mass balance of material and nutrients in the 9 week continuous flow fly larvae treatment, calculated on total solids basis; averaged values from the three replicate units are presented in the scheme. The arrows for total nitrogen and phosphorous are amplified ten times around 20 kg turned into carbon dioxide and water and 2.4 kg (11.8 %) was converted into larval biomass (Table 3). The concentration of total ammonium nitrogen increased, while the concentration of organic nitrogen decreased by almost 63 %. The waste-to-biomass conversion rate of 11.8 % on a total solids basis presented in this paper only included the migrating prepupae. Sheppard et al. (1994) reported an optimal conversion rate of 8 % on a total solids basis in a lowmaintenance system operating in ambient temperature, converting hen manure into prepupal biomass. The optimal conversion rate was found for the May to August period, while a considerably lower conversion rate was found for the period May to December. In good accordance with this study, Diener et al. (2011) reported a conversion rate of 11.8 % on a total solid basis in mixed organic waste. The highest waste-to-biomass conversion rate for black soldier fly larvae reported hitherto is 22.9 %, in which human feces was converted to prepupal biomass in a small-scale system feeding ten larvae one lump sum of material (Banks et al. 2014). The conversion rate drastically decreased to 1.6 % in a treatment with 100 larvae with lump sum feeding, while it was 15.9 % with incremental feeding with the same number of larvae. It appears that several factors impact the conversion rate: the waste source is a major contributor, but also the temperature at which the system operates, the larval density and the feeding regime appears to have an impact. In this study, it was observed that some prepupae never left the reactor, but pupated in the outflow region. A higher wasteto-biomass conversion rate could therefore be expected in a more optimized system. One way of forcing the prepupae to leave the system could be by increasing the moisture percentage in the outflow region. Moreover, the pig manure fraction contained a considerable proportion of straw, which cannot be consumed by the larvae. A substrate with lower proportion of straw would also yield a higher conversion rate. The material degradation was over 50 %, in accordance with previous findings for degradation of different animal manures (Myers et al. 2008;Sheppard et al. 1994). For other feed stuff, higher degradation has been achieved. A 70 % degradation on a total solids basis was demonstrated for mixed organic waste (Diener et al. 2011) and hen feed (Gobbi et al. 2013), while a 75 % reduction on a total solids basis was demonstrated in fecal sludge (Lalander et al. 2013). The higher degradations observed in those materials is probably due to a higher percentage of easily available carbon. The straw in the pig manure made the waste substrate less degradable to the larvae. The total phosphorous was found to be concentrated in the outflow material, which decreases the amount of field work during application. Around 44 % of the total nitrogen evaporated, mainly as ammonia. Ammonia evaporation is governed by the ammonium-ammonia equilibrium, which in turn depend on the pH and temperature (Emerson et al. 1975). This system was run at 25°C and the pH increased from pH 5.6 at the start to pH 8 after 3 weeks of treatment (Table 1). As a consequence, the evaporation of ammonia can become considerable in this type of wellventilated system. In a large-scale system with controlled ventilation, emissions of ammonia can be kept low by condensation of the outgoing air. Preliminary studies on batch fly larvae systems demonstrated that the emissions of nitrous oxide (N 2 O) were so small as to be negligible (unpublished data). Of the total amount of nitrogen in the system, loss through denitrification is likely to be small since the amount of nitrate in the system was not great and the process was relatively well-aerated through larval movement. Although a considerable proportion of the total nitrogen was lost, the concentration of total ammonium nitrogen was higher in the outflow than in the inflow, thereby improving the fertilizer value of the treatment residual in comparison to untreated material. The outflow material that may be used as organic fertilizer comprised on a total solids basis 4.8±0.10 % total phosphorous and 5.1 ± 0.2 % total nitrogen, of which around 77.9 ± 3.0 % was organic nitrogen and 19.4±2.3 % total ammonium nitrogen -0.98±0.15 % of total residue (Table 2). Table 3 Process efficiency of the fly larvae continuous reactor calculated on total solids basis in the material and in the total system including the prepupae (PP); average values and standard deviation (SD) of the three fly larvae reactors presented In Hygienic quality of treatment residue The concentration of Salmonella spp. reduced drastically in the process, from an average inflow concentration of around 10 7 CFU g −1 to below the detection limit of 1 CFU g −1 . The concentration in the inflow was not measured in weeks 5-7, but is assumed to be similar to the concentrations measured in weeks 3-5 (Figs. 2 and 3a). The concentration of total thermotolerant coliforms appeared to decrease in the material (Figs. 2 and 3b), while the concentration of Enterococcus spp. did not appear to be reduced in the process (Figs. 2 and 3c). The inflow concentration of bacteriophage фX174 varied somewhat, but the concentration in the outflow was quite stable from week 5 onwards (Figs. 2 and 3d). In order to get an estimation of the reduction of the evaluated organisms in the treatment, the average log 10 reduction that occurred during one week in the treatment was calculated (5). In average, total thermotolarent coliforms and фX174 was reduced 0.6 and 1.7 log 10 , respectively, in one week (Table 4). The weekly reduction of thermotolerant coliforms was found to increase with time. No statistically significant improvement in weekly reduction of the фX174 was found over time. No reduction of Enterococcus spp. occurred in the treatment. There were no significant difference in the reduction of microorganisms between the reactors (p>0.05). One of the hypotheses in this study was that a continuous system could improve the hygienic quality of the fly larvaetreated material compared to a batch system, due to interactions with an improved microbial and fungal community. For the bacteria and bacteriophages, the results agreed with the findings for a batch system, with no improved reduction (Lalander et al. 2013). In the previous study, the reduction of bacteriophages was found to be equal in the larvae treated and control material and was thus concluded not to be affected by the larval activity. The weekly log 10 reduction of thermotolerant coliforms was found to be 0.6 log 10 in this experiment, but was found to improve with time. No identification of the specific coliform species was performed; however, it is likely that a large part of the coliforms included in the analyses were not E. coli. Erickson et al. (2004) demonstrated a greater reduction in E. coli O157:H7 at 27-31°C in fly larvae-treated chicken manure compared to the reduction found in cattle manure. The reduction was increased with increasing temperature and similar reduction as observed in this study was observed in chicken manure at 23°C. Chicken manure is considerably higher in ammonia than cattle manure (Paul and Beauchamp 1989); uncharged ammonia is well known to be toxic to microorganisms and as the concentration of uncharged ammonia increases with temperature (Nordin et al. 2009). The reduction Week Log 10 conc. (CFU g −1 ); d фX174 (PFU g −1 ) in the inflow (square) and residue (triangle) material over the course of the 9-week experiment of E. coli observed by Erickson et al. (2004) in chicken manure and the increased reduction observed at elevated temperatures could therefore have been caused by increased concentration of uncharged ammonia and be unrelated to larval activity. The mechanism behind the reduction of salmonella observed in this study is not known. Choi et al. (2012) found that methanol extracts of the black soldier larvae had antibacterial effects on the Gram-negative bacteria studied, but no such effect was seen on the Grampositive ones. The authors speculated that the antibacterial effect could be due to interaction between the bacterial ribosome or bacterial cell wall and the active substance. In accordance with Choi et al. (2012), a reduction was seen for the Gram-negative salmonella and some reduction for thermotolerant coliforms in this study, while no reduction of the Grampositive enterococcus occurred. The concentration of viable viruses decreased in the material, and was below the detection limit in all samples after 14 days (Table 5). Adenovirus was the most persistent of the viruses studied, while enterovirus was less persistent than reovirus. The maximum virus concentrations in the material, which were recorded at day 2 for all viruses, were 2.5-4 log 10 units below inflow concentration. This is more than the approximate 1 log 10 dilution of the ingoing material-assuming that the inflow material was mixed evenly in the compost-and is probably due to inactivation of the viruses, possibly in combination with the non-uniform distribution of viruses in the material. As the virus concentration in the outgoing material decrease with time, an increased time interval between feeding and harvesting will yield a lower virus concentration in the outgoing material. There are several factors that affect the survival of viruses: larval activity can be one; a natural timedependent die-off in ambient temperature another; adsorption onto particles is a third; and ammonia inactivation can be a fourth effect. Ammonia has been proven to have a good reduction effect on single stranded viruses (Emmoth et al. 2011). A study on the inactivation of adenovirus (ad41) during storage reported a decimal reduction time of 12.5 days at 20°C in biosolids which have similar dry matter content (Wei et al. 2009). This is considerably longer than the findings in this study. Further investigations are required for a better understanding of the mechanisms involved. The concentration and viability of ascaris eggs did not change significantly (p>0.05) in the course of the treatment, nor was there any significant difference in concentration and viability of ascaris eggs between the three units over the course of the experiment (p>0.05). The viability of the eggs was 90.3 % when they were inoculated into the treatment, 82.9±4.3 after 1 week and 78.6±6.7 after one month in the system. The concentration was 72.6±11.7 eggs per gram after one week and 84.6±68.8 after one month. As for the bacteria and bacteriophages, the reduction of Ascaris eggs did not appear to change in the continuous system compared to what was found for the batch system (Lalander et al. 2013). It was not possible to determine the total reduction/destruction of ascaris eggs. When attempting to estimate the flow of the eggs in the system, based on the Table 4 The average log 10 reduction that occur in one week in the treatment (Δ log10 Red) for the evaluated microorganisms, the p-value for this reduction and the p-value for the time effect, i.e., whether a change in Δ log10 Red occurred over time The concentration in the outflow material was below the detection limit during most of the experiment, statistical analyses was thus not possible concentrations found in the outflow material, the total number of eggs exceeded the total number added, although no eggs were found in the fresh inflow material. Due to the adhesive nature of the ascaris eggs, it is likely that the concentration of eggs was not constant in the system. In this system, not much material was removed throughout the experiment due to the slow accumulation of solids during the 10 weeks of the study. Thus, it is possible that the eggs accumulated in the outflow region while the concentration of eggs could have been lower in the inflow region. However, as the concentration of eggs did not decrease with time in the outflow region, it does not appear that the eggs were destroyed. The viability of the eggs was not significantly (p>0.05) reduced with time. However, it was observed that the eggs extracted after one month in the system had not started to develop into prelarvae after 30 days in the fly larvae reactor at 25°C, which could otherwise be expected. After 30 days at 28°C after the eggs were separated from the material, 79 % of the eggs had developed into prelarvae. It appeared that the development of the eggs had been inhibited during the treatment, despite the temperature being in a range that supports development. One reason could be the increased total ammonium nitrogen concentration, as uncharged ammonia is known to inactivate the development of ascaris eggs (Fidjeland et al. 2013). Additionally, Fidjeland et al. (2013) found that the main effect of ammonia was to shorten the lag phase for the bilinear inactivation of ascaris. It is therefore possible that a longer retention time in the fly larvae system or during post-storage could lead to inactivation of the ascaris eggs present. Sanitary risks of the fly larvae system The minimum retention time of material in the fly larvae compost was set at two days, while the hydraulic retention time was over 30 days. This was enough for inactivation of the zoonotic salmonella bacteria, while the other investigated organisms survived the treatment. Viruses had an indicated reduction of 2 log 10 whereas ascaris eggs were not inactivated/destroyed. System management is therefore required that results in additional security in terms of breaking disease transmission, e.g., the selection of crops to be fertilized and incorporation of the fertilizer into the soil. In a previous study, it was shown that the content of the zoonotic bacteria such as salmonella was reduced in the material to such a degree that it was not detectable in the harvested prepupae (Lalander et al. 2013). Other zoonotic bacteria were not tested, but in other studies a, similar reduction of another Gram-positive bacteria, E. coli O157, has been observed (Erickson et al. 2004). Due to the reduction of zoonotic bacteria, the risk of disease transmissionwhen feeding animals other than the ones providing the feed to the larvaeis a low-risk management system. For preservation and greater hygienic security, it may be required to process the prepupae further. With drying for example, the low water content and the temperature required for drying improve the hygienic status of the material, while preserving the larval meal at the same time. Conclusions The material degradation was 55.1 % and the waste-tobiomass conversion rate 11.8 % on a total solids basis in the continuous fly larvae composting system. Optimizing prepupal migration efficiency could increase the waste-to-biomass conversion rate. A large proportion of the nitrogen evaporated as ammonia: these losses can be kept low by condensing the outgoing air. Both the phosphorous and total ammonium nitrogen concentrations increased in the outflow material. The hygienic quality of the material was found to increase in the fly larvae treatment: the pathogen Salmonella spp. and viruses were reduced in the system, while the reduction of the other organisms studied was small or insignificant. Minimal retention time was found to be a key factor for pathogen reduction.	2017-08-03T02:51:04.538Z	2014-01-01T00:00:00.000	{ "year": 2014, "sha1": "ae1fdbc3394755b68e0ce50d574efe6eb5434bcc", "oa_license": "CCBY", "oa_url": "https://link.springer.com/content/pdf/10.1007/s13593-014-0235-4.pdf", "oa_status": "HYBRID", "pdf_src": "Springer", "pdf_hash": "bd72c79898928a77b2a56cb5813a6ce639369bf7", "s2fieldsofstudy": [ "Environmental Science" ], "extfieldsofstudy": [ "Environmental Science" ] }
72029201	pes2o/s2orc	v3-fos-license	Surgical management of congenital idiopathic clubfoot by twin incisions Background: Idiopathic clubfoot surgery is known for a number of complications, that is, skin slough and wound dehiscence medially, infection and limitation of full correction at the time of surgery which leads to excessive scar hypertrophy, recurrence of deformity, and a poor functional outcome. Many procedures done in the past have reported complications ranging from 3 to 30%. We are doing surgical correction of congenital idiopathic clubfoot by two incisions avoiding dissection of the medial skin to overcome these problems and are reporting our results. Materials and Methods: We used two incisions (a medial and a posterior incision) in cases with moderate to severe congenital idiopathic clubfoot. These were those cases who did not want to continue with Ponseti’s technique so as to avoid repeated follow-ups. Assessment of results was done by comparing preoperative and postoperative Pirani scores. Results: Our results were excellent in 7.9% (6 feet), good in 72.3% (55 feet), and fair in 15.8% (12 feet) cases. All of our cases had primary wound healing and no delayed scar related complication. Conclusions: Two incisions avoid stress on the medial skin. They permit release of all contracted structures posteromedially. We are describing this technique for its simplicity. Introduction Surgery for congenital idiopathic clubfoot has a number of complications in the postoperative period such as delayed wound healing, gaping, infection, and recurrence of the deformity. Multiple types of surgical approaches have been described in literature with the aim to reduce the complications and give adequate tissue exposure so as to release all contractures and allow the incision to heal primarily without wound dehiscence or scar contracture. The two most commonly used incisions: Turco and Cincinnati have been reported to have wound complications like skin slough, dehiscence, infection, and limitation of full correction at the time of surgery in 3-30% of cases leading to hypertrophic scar and recurrence of the deformity. [1][2][3][4] In India, a subtropical country, healing gets markedly delayed if the surgical wound gets infection or there is wound dehiscence. Maintaining the foot in correction in a plaster of Paris cast along with a window for dressing in a hot and humid climate, makes the task more difficult due to soiling of the cast leading to breaking and further infection. With Ponseti's technique coming in vogue surgeries have become less and even if surgery is required, it need not be as extensive as with the Cincinati [5] complete subtalar release or Turco's [6] posteromedial release. We have used twin incisions, a modification of NC Carroll's technique making it simpler for the correction of the idiopathic clubfoot to avoid the above mentioned complications. In this technique two incisions are given, sparing the midincision zone of Turco's incision, reducing the chances of gaping or infection. Dual incisions have been also described by Carroll. [1,2] They differ from our incisions as the medial incision on the foot in the Carroll's technique is used to raise a flap for complete subtalar release which at times may undergo marginal flap necrosis. Materials and Methods Seventy-six idiopathic clubfoot in 56 children aged 3-18 months (average 9 months), presenting to us from January 1997 to June 2007, were treated by us by a medial and a posterior incision, sparing the intervening skin and dissection for posteromedial release following percutaneous plantar fasciotomy. Forty-two of these children were male and 14 were female. There was bilateral involvement in 20 children. All these feet had moderate to severe deformity as assessed by Pirani score. A foot with a Pirani score as modified by Flynn et al., [7] less than 2.5 and which got corrected with Ponseti's weekly casting was considered as mild. A foot was considered to be having moderate deformity when the child had a Pirani score between 2.5 and 5 (21 feet). These feet did not get corrected fully with weekly Ponseti's casting or recurred in a few weeks' time due to poor parent compliance, ignorance, illiteracy, or low socioeconomic status. [3] A foot was said to be having severe deformity if the Pirani score was more than 5 (55 feet). Very rigid feet, with a Pirani score above 6, neglected CTEV with bony deformity and candidates with secondary clubfoot were not considered to be fit for this procedure. The surgery was done with patient in semisupine position and a sandbag put below the opposite iliac crest and lateral border of the operative foot touching the couch. The twin incisions used were: The technique is described below: Medial incision The incision is started 1 cm behind and above the medial malleolus along its posterior border and is curved down to middle of the first metatarsal. After identifying upper border of abducter hallucis muscle (AH), it is retracted down. The dissection is done straight down to the deep fascia, so as to avoid superficial dissection and jeopardize the vascularity of flaps raised. Behind the medial malleolus, the sheath of tibialis posterior (TP) and adjoining flexor digitorum longus (FDL) tendon is identified. This is made easy by the fact that the TP tendon is three times thicker than the FDL tendon. The superficial part of deltoid ligament and the spring ligament are cut. The TP tendon is cut and followed down to the navicular tuberosity and all its insertions are excised. The proximal cut end of TP is sutured to the tendon of FDL, and can be used in future as a motor unit for tendon transfer. Capsulotomy of talonavicular, naviculocuneiform, and cuneiform-first metatarsal joints is done. If necessary, Z-plasty of FDL and flexor hallucis longus FHL) tendons is done to avoid clawed toes. The anteromedial subtalar joint between navicular and sustentaculam tali is opened. The tendon of AH is cut and reduction of navicular confirmed. Posterior incision The incision is made approximately 3" vertically along the lateral border of tendo Achilles. Z-plasty is done with medial division in the distal segment. The distal segment is reflected and space between FHL and peroneus longus (PL) is developed. By forcefully dorsiflexing the foot, posterior capsule of the ankle and subtalar joints was stretched and the posterior capsule of the ankle joint, subtalar joint, and interroseus talocalcaneal ligament all were cut from here. Posterior talofibular and calcaneofibular ligaments are cut. Z-plasty of tendo Achilles is done with foot in neutral position. Closure is done in two layers. Deep layer is repaired by 3.0 vicryl and skin by 3.0 prolene, ethilon, or silk [ Figures 1 and 2]. A well-padded cast above the knee without stress on suture line is applied for 3 weeks [ Figure 3] followed by removal of sutures and keeping foot in maximum possible abduction and dorsiflexion without pain for a further duration of 3 weeks in the AK cast. Further maintenance of correction is achieved by keeping the foot in surgical shoes maintaining 150 dorsiflexion of ankle and eversion of foot for further 6 weeks duration. The children were advised stretching, consisting of manipulation of the foot by mother to full or slight overcorrection of equinus, and eversion for 5 min at least three times a day for the next 3-6 months. Figures 4-5 shows the postoperative appearance and 6 year follow-up of the same child. Final results were assessed at two years by an improvement in the Pirani score to less than 0.5 as good, less than 1.5 as fair, and less than 2.5 as poor. Results Average age at surgery was 13 months (9-20 months). Follow-up evaluation was done at 2 years. The maximum long follow-up is of 7 years. The average surgical time was 45 min. The average wound healing time was 14 days ranging from 10 to 30 days. On comparison of the preoperative and postoperative Pirani score with number of patients, it was found that six patients improved to 0-0.5, 55 improved to 0.5-1.5, 12 improved to 1.5-2.5. 3 patients still had a Pirani score above 2.5 and were categorized as poor results. Number of patients with a preoperative Pirani score of 5-6 was 55 which came to 0 postoperatively [ Table 1]. Only 3 feet (3.9%) had a residual forefoot adduction at 1-year. We used the Chi-square test to assess our results as follows: Preoperative group:  2 trend = 128.719 P < 0.000001 Postoperative group:  2 = 141.50 at 4 degrees of freedom P < 0.000001 Upon comparison of Pirani score between preoperative and postoperative groups, we observed a positive linear trend and significant reduction of the cases towards the poorer scores. The finding was statistically significant;  2 trend = 128.719, P < 0.000001 and  2 = 141.50; P < 0.000001; respectively. This suggests that the technique of simple twin incisions significantly increases the probability of better results (P < 0.000001). We had excellent results in six (as seen by postoperative Pirani score below 0.5), good in 55 (as seen by postoperative Pirani score <1.5) [Figures 5] and fair results in 12 feet (as seen by postoperative Pirani score below 2.5). A foot resisting treatment, not acceptable socially was taken as a failure of treatment as seen in 3 feet with increases the financial cost as well as agony to the patient. At times patients prefer to accept deformity rather than come repeatedly for follow-ups due to poverty and illiteracy. A wound Discussion Postoperative wound complications are significant in India due to its subtropical location with hot and humid climate leading to increased chances of skin and soft tissue infection resulting in recurrence of the deformity. Characteristics of these complications can be associated with the surgical approach used. [8] Wound complications following posteromedial release are inherent to the technique only as the contracted skin posteromedially will never suffice as a cover once the foot is brought to the neutral position. Turco's technique [6] advocates casting in under correction followed by gradual correction in serial casting in weekly casts. This is a difficult situation as the parents expect the child to have normal foot once surgery is over and repeated casting at times under anesthesia with infection, gaping, and persistence of deformity is taken as a sign of failed surgery. Two incisions for complete subtalar release have also been described by Carroll NC [1,2] where he uses a posterior incision similar to our incision and a medial, distally based flap. This flap is reported to get necrosed. Our incisions are different. Our medial incision follows the distal part of Mackay [9] incision and our posterior incision is lateral to the tendo Achilles. Our incisions for posteromedial release overcome these problems of wound healing as they neither put tension on the medial incision nor on the posterior incision. The medial skin where maximum tension is expected is not violated at all and even on stretching it does not give way. The intact bridge of skin between the two incisions never showed signs of vascular compromise or necrosis. A reduction below 0.5-1 of Pirani score was achieved in 80.2% feet which is found to be acceptable in moderate to severe cases. In cases with residual forefoot adduction with rigid deformity or an initial severe deformity, cuboid decancellation is done. This may be done at the time of first surgery only or at a later date for forefoot adduction. Thus, it has the advantage to maintain a normal looking foot with preserved joint functions. In our series we did cuboid decancellation in 3 feet only at repeat surgery of poor results. We compared our results with those of other published studies and we have found a bilateral verus unilateral ratio of 51.56:49.44 comparable to that of Turco et al., [10] 55:45. Skin necrosis rate as described by Turco in 1971 [10] was 6%, and Aggarwal et al., [11] was 3%. However, we found no skin necrosis following our technique of two incisions. We had a recurrence rate of 3.9%, maybe due to starting the treatment late in some patients and failure to carry out postoperative regimen of gentle stretching. However, surgically treated clubfeet, in which at least part of the retracted ligaments are removed and the tendons are lengthened, frequently relapse as well, presumably due to surgical scar and muscle fibrosis. [12] The overall results as reported by Turco [10] for his procedure on 31 cases in 1971 was 90% success and 10% failure. In our series, we had 7.9% excellent, 72.3% good, 15.8% fair, and 3.9% poor results. We recommend these two incisions in the operative correction of moderate to severe idiopathic clubfoot because we did not observe any wound related complications and achieved full correction of the foot.	2019-03-09T14:04:23.385Z	2013-01-01T00:00:00.000	{ "year": 2013, "sha1": "1d2eb2845b56562169a2148e5da1fab36a9585bb", "oa_license": "CCBYNCSA", "oa_url": "https://doi.org/10.4103/2319-2585.117383", "oa_status": "GOLD", "pdf_src": "MergedPDFExtraction", "pdf_hash": "933c5e161d4ce64a5867b497f4b4804fae1e37c6", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
251329105	pes2o/s2orc	v3-fos-license	Emission reduction targets and outcomes of the Clean Development Mechanism (2005–2020) The Clean Development Mechanism (CDM) allows developing countries to earn carbon credit units by reducing greenhouse gas emissions. Here we assess the emission reduction outcomes of the CDM between 2005 and 2020. The analysis covers 3,311 CDM projects hosted by 79 countries and over 10,000 Monitoring Reports. We identify which host countries and project types departed from original forecasts more. Overall, the total amount of actual emission reductions was 16% below the targets envisaged by project proponents. Emission reduction projects consistently under-performed over the year, but performance varied between and within regions. Industrial HFCs and N 2 O projects exceeded their targets, whereas landfill gas and methane avoidance projects under-performed by larger mar-gins. Economic gains were unevenly distributed. Estimated revenues relative to GDP were higher for larger emerging economies, and disproportionately smaller for the deprived members of the Global South. Four host countries (China, India, South Korea and Brazil) not only dominated the market, but also gained an advantage from the higher carbon prices before 2012. Least Developed Countries had their carbon credits issued in more recent years when prices were much lower. The results show an imbalance in economic outcomes and raise questions about the effectiveness and equity of this Kyoto mechanism. Weak targets under Paris Agreement could intensify these challenges. Introduction Article 6 of the Paris Agreement provides a basis for establishing new mechanisms for mitigating greenhouse gas (GHG) emissions after 2020. The new mechanisms are likely to involve the use of internationally transferred mitigation outcomes to achieve nationally determined contributions (NDCs). These mechanisms are expected to replace the Clean Development Mechanism (CDM), a multi-billion euro carbon finance mechanism established under the Kyoto Protocol [1]. The CDM enables Annex I Parties (industrialized countries) to produce and acquire carbon offset units through investments in GHG mitigation projects in non-Annex I Parties, predominantly in the Global South. These offset units, known as Certified Emission Reductions (CER), represent a reduction, avoidance or sequestration of one metric ton of CO 2 that would otherwise be emitted into the atmosphere. Annex I Parties can use CERs to meet their emission reduction targets under the Kyoto Protocol [2,3]. The CDM issued the first CERs in October 2005. Since then, it has mobilized more than US$162 billion of financing to non-Annex I Parties. As of September 2021. US$162 billion is the total capital investment by CDM projects, including Programmes of Activities (PoAs), that have been issued CERs. This is a conservative estimate, because not all projects reported their financial details. The US$162 billion is estimated from a fraction of projects with CERs issued (N = 2,786) that collectively accounts for 69.5% of estimated (ex ante) annual emissions [5]. This article assesses the emission reductions outcomes of 3,311 CDM projects that have been issued CERs. The main objective is to measure project performance by comparing actual emission reductions (ex post) with estimated emission reductions (ex ante). There were signs of under-performance as early as 2006 [4], but evidence is conclusive only when most projects are given time to operate to their proposed timeframe. Most projects have a crediting period of either 7 or 10 years, and most of them were registered on or before 2011. We identify the project types and host countries that did not meet the emission reduction targets stated in CDM project proposals, known as Product Design Documents (PDDs). To determine how many CERs to issue, the CDM Executive Board (EB) requires the project proponent to submit a Monitoring Report (MR), which specifies the amount of estimated and actual emission reductions in a given monitoring period. The data we used are derived from these MRs published between 2005 and 2021, and an official CDM project database. The CDM was also designed to channel carbon finance to the Global South. The size of direct economic contributions to non-Annex I Parties depends on the quantity and price of CER units. Our analysis compares the price expectations of project proponents and market CER prices to understand the discrepancies in economic contributions. We show the extent to which direct economic contributions relative to GDP, are unevenly distributed between higher-and lower-income developing countries. We explain this in terms of the timing of CER issuance and price volatility. Our study covers the active lifetime of the CDM to provide conclusive evidence and insights into the ways forward. The findings raise questions about the effectiveness of the CDM in meeting its targets and delivering equitable outcomes. CDM-style Paris mechanisms may result in similar patterns of uneven distribution of benefits, if their scope and designs largely replicate those of Kyoto mechanisms. These results are particularly important for re-thinking the limits of these mechanisms in delivering development benefits and how they should be designed in order to benefit the economically deprived members of the Global South in more direct and effective ways. Methods The timeframe of our analysis, which runs from 2005 to 2020, is important for understanding our findings. The first commitment period of the Kyoto Protocol commenced in 2005. The first few years after 2005 witnessed a market boom with higher CER prices. Sharp changes occurred towards 2012, when the first commitment period completed. The end of 2020 is another watershed. The second commitment period concluded on 31 December 2020. The European Union no longer accepts the use of international units, including CERs, for compliance under the EU ETS after 2020. While the CDM continues to operate, no new projects are expected to seek registration. Emission reduction data were collected from MRs. Other project information was gathered from the official database of CDM projects (cdm.unfccc.int/Projects/projsearch.html), including project type, host Party (country), duration of monitoring period, date of CER issuance, and start date of validation. The database was accessed in March 2021 and updated in September 2021 [5]. CDM projects included in the analysis The full CDM database has 8,206 registered CDM projects. Our project-level analysis excluded 352 Programmes of Activities (PoAs), because new projects can be added to the programme without undergoing the complete CDM project cycle, and therefore it is hard to attribute emission reduction estimates to their component Project Activities. PoAs account for 2.2% of all CERs issued [5]. A total of 3,312 projects (out of 7,854), formally known as CDM Project Activities, have been issued CERs. The first set of carbon units was issued in October 2005, and the latest date of issuance was 30 August 2021. We excluded one project (CDM reference #1) and the first 19 Monitoring Reports (MRs) of another project (CDM reference #3), because their MRs did not clearly specify the amount of estimated emission reductions for the relevant monitoring period and their PDDs did not include a breakdown of yearly estimates. The project activities of these two projects (first 19 MRs of CDM Project #3) generated a total of 66,958,005 CER units. Our analysis on emission reductions is based on the remaining 3,311 CDM projects, which were hosted by 79 non-Annex I Parties. Estimating emission reductions and measuring performance The amount of emission reduction is stated in the MRs of these 3,311 projects. Each project has at least one MR published, and the average is four MRs. There are over 10,000 MRs. Typically, each Monitoring Report has only one issuance record, but those that transcend the first and second Kyoto commitment periods have two records. We reviewed each MR and recorded the amount of actual emission reductions reported by the project proponent, which was verified by an independent auditor and accepted by the CDM EB before CERs were issued. The majority of MRs were revised after reviews, and some of them were withdrawn. We adopted the revised and final versions of MRs and excluded withdrawn MRs from the analysis. Each MR indicates the amount of emissions actually reduced during a given monitoring period, against a ex ante estimate for the same period. We identified 13 MRs that indicate their actual emission reductions exceeding the amount of CERs issued. These 13 MRs reported more actual emission reductions than CERs issued. A common explanation is that the CDM EB requires the amount of CERs issued be capped at the level of average annual emissions estimated in the registered PDD. In analyzing emission reductions, we adopted the value of emissions actually reduced, as reported in the MR. As a result, the aggregate values do not perfectly match the total amount of CERs issued. Duration of monitoring period ranges from 1 days to 4,107 days and averages 419 days (N = 10,999). The start date of the earliest monitoring period was 1 April 2000, and the latest end date was 31 December 2020. Ex ante estimates are also extracted from MRs. However, some of the pre-2010 MRs do not indicate the amount of estimated emission reductions for the corresponding monitoring period. To document the original expectations of project proponents, we retrieved these emission reduction targets from the first PDD that was approved for registration, rather than the post-registration, revised PDD. Moreover, we did not take the self-reported estimates at face value, but cross-checked the calculation presented in the MR against the approved PDD to ensure accuracy and consistency. We believe that there are typos, miscalculations, and tendencies for strategically choosing variables (e.g., number of actual operating days) to present a lower value of estimated emission reductions in MRs. For some projects, the amount of estimated emission reductions varies every year. We therefore adopted the estimates specific to the monitoring period concerned, rather than the yearly average of the entire crediting period, where applicable. For this reason, we did not use the estimated emission reduction data from the official CDM Database. This database also does not provide data on actual emission reduction, which are gathered from individual MRs. Performance rate is calculated to indicate the extent in which a project met its pre-determined emission reduction targets within the crediting period(s) in which CERs were issued. It is the ratio of actual emission reductions to estimated emission reductions. Weighted performance rates are used to describe projects (N = 3,311). Each project resulted in a different amount of emission reductions, and therefore its impact on the overall performance rate of a host country or project type is weighted by the total amount of emission reductions generated from this project during the crediting period(s). Average performance rates (not weighted) are used to describe CER issuance records (N = 10,999). Comparing carbon prices We compared the price expectations of project proponents with the market CER prices. A sub-sample of the 3,311 projects was used. An expected price is the level of market CER price predicted by the project proponent. It is the offer price at which they expected to sell their CER units. Price expectations are a key driver of CDM investment decision-making. The expected level of CER price is often stated in the PDD or its attachments for demonstrating the project's expected higher financial performance over an economic baseline (i.e., return on investment without CER incomes) in order to justify that the investment would create 'additional' emission reductions, especially for renewable energy projects [6,7]. We identified expected prices from 2,113 CDM projects. Where multiple expected prices were presented in a PDD, the lowest one was adopted as a conservative estimate (i.e., the closest to market prices, in most cases). Weighted reference prices are calculated on the basis of market CER prices. It is used in this study to approximate the market value of a CER unit at the time when it was issued, and to compare with the target (i.e., expected price). Our calculation was based on an average value of the daily CER closing prices in the 30 trading days prior to the date of CER issuance. We adopted the prices of monthly CER futures contracts traded in the Intercontinental Exchange, because there was very limited spot CER trading from the end of 2012. Monthly CER futures prices are the closest alternative to spot prices. Spot price is the current price at which a CER unit can be traded for immediate delivery. The price set in a monthly futures contract is usually very close to its spot price, due to the short timeframe. Yet, we were unable to retrieve this price information for CER units issued before March 2008, which account for 5.7% of total CER units issued. We used annual average secondary CER prices obtained from the World Bank's reports (2007,2009) to substitute for these missing prices. Weighting is based on emission reductions. A CDM project might have its CERs issued at different times during or after the crediting period. Accordingly, there were multiple reference prices for the different monitoring periods of the same project. The reference prices of each project were weighted by the amount of actual emission reductions the project generated in the corresponding monitoring period. As a result, these prices reflect the timing of CER issuance. Because CER prices were on a downward trend after 2011, the more the CER units issued in later years, the lower the weighted prices. These prices are plotted in a graph and ordered by the start date of project validation, which marks the time when the project proposal was submitted for third-party validation [8]. Same as the 'Start of Project's Public Comment Period' specified in [8]. The official database of CDM projects includes complete information on the validation start date of each project, which is therefore reliable and consistent for the purpose of research [5]. Estimating economic contributions The volume of CER units generated relative to the GDP of a given host country is used by the UNEP DTU Partnership as "an immediate expression of the importance of CDM to the economy" [9]. The ratio of these two variables provides an indication of the direct economic benefits of producing and exporting these units, given the economic scale of the host country. We used the amount of actual emission reductions for estimating economic contributions, but also included revenue estimates to account for price variations. We estimated revenues (ex post) from sale of CER units for each of the 3,311 projects, which are the product of actual emission reductions in a given monitoring period and the reference CER price (without weighting). Because we used secondary market prices as reference prices, the value of CER units generated from 'bi/multilateral' CDM projects might have been underestimated. Bi/multilateral projects have a formal arrangement at the time of project development with an Annex I Party that intends to buy and use the credits. They operate under a forward purchase agreement, which typically specifies a mutually agreed price level or a price range at which the CER units generated from the project would be sold to a buyer from an Annex I Party. Purchase agreements with set prices could protect project proponents and host countries from volatile prices. On the other hand, 'unilateral' CDM projects do not have a letter of approval from an Annex I Party at the time of registration and do not have a prior contractual arrangement with an international buyer [2]. We do not have access to all forward transaction data and therefore cannot include agreement prices in our analysis. Despite this limitation, our revenue estimates are useful for comparing economic contributions between countries. Many emission reduction purchase agreements are guaranteed only up to 2012 [4], but over 45% of CER units were issued after 2012. Agreement on a transaction price was typically reached long before the project started and therefore likely to be higher than market prices, which began to fall sharply from 2011. The large differences between the two prices since 2011 challenge the assumption that all purchase agreements would be exercised at the indicative prices and volumes [10]. CER units might eventually be sold outside the range of these agreement prices-likely at a lower level, given the availability of significantly cheaper CER units in the market. Renegotiations and cancellations did occur. Official data suggest that 19% of buyers have withdrawn from the bi/multilateral projects to which they were committed [9]. Moreover, 69.6% of all CDM projects were bi/multilaterally established (Table 1), but China alone accounts for most of these projects and has an exceptionally high percentage of bi/multilateral projects. Excluding China, less than half of the CDM projects were bi/multilaterally established. Although market prices do not perfectly reflect actual transaction prices, the changes in the latter track market trends. This is important for the purpose of this analysis, which aims to highlight the time-sensitive economic contributions of the CDM in a period of over 15 years. The revenues analysis adopted the amount of CERs issued, rather than actual emission reduction, because only CERs are traded in the market. All revenue estimates were discounted by 2% to account for the transfer of proceeds from CER sales to the Adaptation Fund established under Kyoto Protocol. To address the different price levels at different times of CER issuance, the aggregate revenue estimates were converted to 2010 prices using the International Monetary Fund (IMF) Euro Area GDP deflators for the time (year) of CER issuance. For estimating economic contributions to individual host countries, the IMF GDP estimates were adopted as denominators and expressed in Euro (converted from USD using year-end exchange rates). The ratio between revenues (current prices) and GDP was calculated for each country and each year between 2005 and 2021, using the nominal GDP estimate for that year. The values were then aggregated for each country. Project type, geographical region, Commitment Period, host country and whether it is a 'Least Developed Country' (LDC) were specified in the CDM database. Summary The key measures and indicators used in the analysis are listed below: Actual emission reductions: as stated in CDM Monitoring Reports. They indicate the amount of net GHG emissions reduced, avoided or sequestered in the combined monitoring periods of a project and determine the amount of CERs issued. Combined monitoring periods are the fraction of crediting period(s) in which CER units were issued. Estimated emission reductions: as stated in CDM Monitoring Reports and PDDs. These estimates are a fraction of the total estimated emission reductions of a CDM project and represent what was expected to achieve in the combined monitoring periods. The remaining emission reductions to be generated in the rest of the crediting period(s) (i.e., CERs yet to be issued) was excluded. Performance rate: ratio of actual emission reductions to estimated emission reductions. Expected CER price: as stated in PDDs. Representing the offer price expectation of the CDM project proponent and the expected revenues per CER credit sold. Non-Euro currencies were converted to Euro, at the exchange rate at the start date of validation. Weighted reference price: average value of the daily prices of monthly CER futures contracts over the 30 trading days prior to the date of CER issuance. The average values calculated for each project were weighted by the amount of actual emission reductions the project generated in its corresponding monitoring period. Estimated revenues: product of the CER units issued in a given monitoring period and the reference price at the date of CER issuance. Discounted by 2% and converted to 2010 prices. Economic contributions: ratio of 1) total amount of actual emission reductions, and 2) estimated revenues, to the GDP of a given host country. Emission reductions The 3,311 CDM projects reduced 2,043 million tonnes of emissions, which is 16.4% below the aggregate PDD forecasts (2,444 million). Fig 1 shows , most of their projects achieved 80% of their target or more. African LDCs generally performed better than the rest of the world, but the rest of Africa was not. It should be noted that these discrepancies can be attributed to inflated emission baselines or poor project performance, or both. Project types explain some of these results. These projects involve destruction of HFCs and N 2 O gases from industrial processes, which are of high global warming potential and generate a disproportionately large amount of emission reductions. There are only 79 HFCs and N 2 O projects, but they produce 10.6 million tonnes of CO 2 e each on average and account for 41% of total actual emission reductions from all projects. South Korea and China generated 85.0% and 44.6% of their emission reductions from these projects respectively, much higher than India (24.9%). These three countries were responsible for over 80% of all HFCs and N 2 O-based emission reductions. Excluding these 79 projects from the sample (N = 3,311) reduced the overall performance rate from 84% to 74%. Despite their effectiveness in meeting emission reduction targets, HFCs projects are highly controversial and have been called into question for creating perverse incentives for manufacturing more of these gases to increase baseline emissions, by which to maximize CER revenues. It is likely that part of the claimed or reported emission reductions are not real and additional, raising question about the environmental integrity of these projects [11,12]. These projects are no longer accepted for CDM registration since 2007. The majority of CDM projects (1,950) produce hydro and wind power. Few of these projects achieved their emission reduction target. Their MRs have suggested several reasons. Emission factors were declining in some countries, meaning that the burning of fossil fuels becomes less polluting over time and therefore renewable energy displaced fewer emissions than proposed. Weather is a more common explanation. Seasonal variations are foreseeable, but the availability of water (rains) and wind is more often overestimated than less. Another reason, which is not always explicitly stated in MRs but widely discussed elsewhere, is a mismatch between policy intervention and actual need. China, for example, produced 67.4% of all emission reductions based on wind, hydro, and solar power. The Chinese government has considerably increased the incentives for installing renewable energy capacities, mobilizing a massive amount of private and public capital into these industries. However, the installed capacity booms did not meet with actual energy demand and improvements in the capacity of supporting infrastructure, resulting in excess supply and eventually curtailment of wind, hydro, solar power across the country [13][14][15]. The reduced operation of power generating facilities inevitably left more emissions unabated. Landfill gas and methane avoidance projects recorded the lowest performance rates among project types with more than 100 projects issued CERs. Both of them involve the handling and disposal of waste, but landfills reduced more emissions. Landfill gas projects involve the capture and degradation of the methane and CO 2 produced during landfill waste decomposition by flaring or use for power generation. Mexico achieved only 67% of its ex ante emission reductions. Fifty-one Mexican projects, out of 72, involved either landfill gases or methane avoidance. Project performance was remarkably low. One of these projects (CDM reference #1371) commenced in 2008, but achieved only 3.5% of the pre-project estimate during a fiveyear monitoring period from 2015, citing non-operation in 1,531 days (out of 1,824 days). On average, these 51 projects departed from their targets by over 50%. Their MRs attribute poor performance to factors such as low efficiency of the gas capturing systems, gas release resulting from exposure of historical waste during infrastructure installation, and lower baseline emissions due to a change in local regulations causing decreased waste disposal (CDM reference #1123, #2992, and #4211). Some landfill sites begin as unmanaged dumping grounds which are not run by the government and therefore historical records of landfill activity can be unreliable or inaccessible, making it difficult to estimate the historical waste composition which is a key determinant of landfill gas production [16]. New legislations and policies that promoted waste minimization also resulted in unexpectedly smaller volumes of disposed waste delivered to the landfill sites, directly reducing the baseline GHG emissions that would have been released into the atmosphere (CDM reference #4211). Other reasons listed in MRs (all project types) include poor estimation before project implementation, accidents, maintenance and repairs, natural hazards, and changes in economic conditions. CDM projects consistently under-performed over time. This is clearer for countries and regions with more issuance records available, including China, India, and the rest of Asia and Latin America, which account for 81.1% of all CER issuance records (Fig 3). Similarly, the average performance rates of wind power, hydropower, methane avoidance, biomass energy, and landfill gas projects varied before 2010 but became steady afterwards (Fig 4). These five project types account for 74.3% of all issuance records. The fewer emissions actually reduced directly lead to a smaller amount of carbon finance. This is exacerbated by the shrinkage of the CDM market from 2011. Carbon prices Price expectations and market prices moved in the same direction up to a point. Fig 5 displays the movement of expected prices and weighted reference prices between 2003 and 2020, along with the daily closing prices of CER units. The discrepancies between expected and market prices increased since the middle of 2008. CDM projects that began validation at a later time were less likely to reach the forecasted price when their CER units were available to the market. In 2011, it became clear that CER units could only be used to a limited extent in the phase 3 (2013-2020) of the EU ETS. This triggered a collapse of CER prices, as the EU ETS was the largest buyer of CER units. While the CER prices struggled to return to higher levels, a small number of projects proposed around this time continued to show confidence in the price. There were tendencies for overestimating CER prices, or being too optimistic. The decrease in price disadvantaged countries that issued CERs in more recent years, notably the LDCs. This can be illustrated by mapping estimated CER revenues. Economic contributions The direct economic contributions of the CDM are unevenly distributed. Nonetheless, the differences are relatively small, indicated by the low contrasts in filled colours (Fig 6). About 53.5% of CER units were issued on or before 2012. Emerging economies, particularly China, earned most of the CER units, but some LDCs held a fair share of the market. The main driver was the EU ETS, which would accept new project credits/CERs after 2012 only if the project is registered in one of the LDCs. As a consequence, the demand for CER units from non-LDCs diminished, and investment in CDM projects hosted by a LDC increased. The biggest beneficiaries were Asian LDCs. Bhutan, Cambodia, and Lao PDR hosted 19 CDM projects and earned 903, 339, and 286 CER units per million GDP (Euro), respectively. Hydropower contributed to 98.3% of the CER units issued to them. Total revenues for the first and second Kyoto commitment periods are estimated as €10,480 million and €173 million (2010 prices) respectively. Host countries do not receive a fair share proportionate to their economic scale, because their CER units were issued in different periods of time. The decrease in CER price intensified the uneven distribution of benefits between large emerging markets and small LDCs. Fig 7 shows the amount of estimated CER revenues relative to nominal GDP. The clearer contrasts in colour reflect the effects of price and timing of CER issuance. China stands farther apart from other non-Annex I Parties by generating a disproportionately large return from the CDM. India, South Korea, and Brazil are three other larger national beneficiaries. Revenues was also high for lower-middle income economies (non-LDCs), such as Nicaragua, Vietnam, and Bolivia. In contrast, these estimates were remarkably lower for Asian LDCs, including Bhutan, Cambodia, and Lao PDR, and African LDCs, notably Uganda and Lesotho. The small pie charts at the bottom of Figs 6 and 7 suggest that China, India, South Korea and Brazil together earned 81.1% of all CER units issued, but accounted for 90.2% of the estimated revenues (or 90.5% for the first Kyoto commitment period). This implies that the average market value of their units is higher than those of the rest of the world. The differences between China and other host countries are likely to be larger than presented in Fig 6, because China has an exceptionally high proportion of bi/multilateral partnerships which could shield revenues from sliding prices-much higher than its 'competitors', namely, India, South Korea, and Brazil (see Table 1 above). Therefore, the economic benefits of the CDM are highly concentrated in a handful of emerging economies, especially China. The timing of project and CER issuance contribute to an uneven distribution. As indicated in Fig 8, emerging economies had a large proportion of their CER units issued on or before 2012, followed by the rest of Latin America (Fig 8B). Benefits to these countries were comparatively attractive, thanks to the higher prices in the first few years of CDM implementation. LDCs were underrepresented in the market until 2012, when they gained momentum from European and international policy interventions that supported LDC participation (Fig 8C). However, revenues from the post-2012 market were lower, even if the sale prices were protected by purchase agreements, as most of which were established in the later years and reflected the declining market conditions. This remains true, but to a lesser extent, for the rest of Africa and Asia (Fig 8D). The direct economic contributions of the CDM were disproportionately small for most African and Asian host countries other than China, India, South Korea, and Vietnam. Discussion and conclusions New carbon crediting mechanisms under the Paris Agreement present an opportunity for redressing some of the previously unforeseen or underestimated problems of Kyoto mechanisms. The CDM is part of the global action to create a cleaner and more equitable future by financing emission reduction activities in developing countries. CDM projects that have been issued CERs are in principle able to produce 4.6 billion CERs to the end of 2020 [17]. This Kyoto mechanism has presented a steep learning curve for its successors [18,19]. A total of 3,311 registered CDM projects have reduced, avoided or sequestered a minimum of 2,043 million tCO 2 e. This is 16% below the original PDD forecasts, or 26% lower if the controversial HFCs and N 2 O gas projects are excluded. Carbon prices moved in the opposite direction as CDM project proponents forecasted. The extended period of low prices could reduce the incentives for operating the project and bring it to a halt. In this situation, the project proponent could manage to support the project using internal resources or through conventional financing [6,20], implying that international finance through the CDM would no longer play a decisive role-even if the project continues to reduce emissions. Furthermore, the economic benefits of the CDM were concentrated in emerging economies. Direct benefits to LDCs were disproportionately smaller, because more of their CER units were subject to the frustrating prices. The results shed light on the financing outcomes of the CDM. The amount of carbon mitigation finance mobilized did not meet prior expectation. For instance, in October 2009, the World Bank anticipated that the CDM could raise €13.5 billion (US$18 billion) in direct carbon revenues for developing countries by the end the first Kyoto commitment period [21]. Our conversative estimate (€10.5 billion) is 22.2% lower. This is conservative because secondary market prices are typically higher than the direct financial benefit to host countries [20]. According to our data, the pre-2012 weighted average price of CERs was EUR 12.49 (constant 2010 prices). Annual reports of the World Bank suggest that the weighted average price in the primary CER market between 2005 and 2012 was EUR 9.37 (constant 2010 prices) [15, 20, 27, 28]. This raises questions about the effectiveness of the CDM in mobilizing finance, considering that the mechanism itself was a significant global investment and involved massive resources and efforts in negotiation, coordination, and operation. The financing outcomes do not appear equitable. The World Bank anticipated that China, India, South Korea and Brazil would receive 78.3% of direct carbon revenues by the end of the first commitment period [21], whereas our study suggests 90.5% for the same period. Large emerging economies have considerable advantages over the poorest ones. For example, the Chinese government required local project proponents to identify an Annex I Party partner before seeking national approval and therefore many Chinese CDM projects are bound to purchase agreements. This could protect the country's interests from the price collapse since 2010, and reflects the stronger institutional support provided by the Chinese government for building international partnership for CDM activities [22,23]. LDCs, on the other hand, did not gain any advantage during the market boom before 2012. The post-2012 rules that prioritized LDCs met with a market downturn. Earlier research has raised PLOS CLIMATE questions to the ability of the CDM to deliver development benefits [24,25]. We echo these previous studies by suggesting that the design of the CDM, market volatility, and differential capacities for project development result in a 'development deficit' that Paris mechanisms must address. The unmet expectations are concerning in the wake of the Paris Agreement. Systematic under-performance can be an indication of certain problems and loopholes in the processes of project registration and auditing gone undetected. There have been reports of inflation in baseline emissions and price expectations, and strategic manipulation of project variables in order to demonstrate feasibility and additionality [6,8,[26][27][28]. Under-performance can also arise from poorly designed or managed projects. If these problems and loopholes are actually more significant and prevalent than previously documented, the performance of the mechanism may be even lower than reported here. A strong rulebook under Article 6 and measures for enhancing regulatory effectiveness would be important for ensuring the reliability of project proposals and audit reports, and the performance of projects. Unlike the Kyoto Protocol, the Paris Agreement requires Parties to formulate their NDCs, which are voluntary emission reduction targets. The bottom-up approach for determining emission reduction targets can lead to a lack of ambition in controlling domestic emissions and consequently weaken the aggregate demand for carbon units [19,29]. Sharp reductions and movements in carbon price as a result of changes in national or international commitments could marginalize economically deprived members of the Global South in the processes of project development, implementation, and transferring carbon units. LDCs arguably encounter more barriers than large emerging economies. As recipients of carbon finance, their economic interests in the international transfer of carbon units risk being compromised by the volatility of the market, which can arise from a unilateral decision to retreat from climate action by dominant GHG emitting countries and a rapid decline in demand from these countries. Ambitions in emission reduction, programme continuity, and additional institutional support to economically deprived Global South countries will be crucial for delivering equitable outcomes under the new crediting mechanisms.	2022-08-05T15:22:56.646Z	2022-08-03T00:00:00.000	{ "year": 2022, "sha1": "31268bdb56ce7d0fc72ba918aa62da6ab0a4ac49", "oa_license": "CCBY", "oa_url": "https://journals.plos.org/climate/article/file?id=10.1371/journal.pclm.0000046&type=printable", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "0b223b70556de3c39658716be02c5f889e443f93", "s2fieldsofstudy": [ "Economics" ], "extfieldsofstudy": [] }
39544708	pes2o/s2orc	v3-fos-license	Ax-Schanuel type theorems and geometry of strongly minimal sets in differentially closed fields Let $(K;+,\cdot, ', 0, 1)$ be a differentially closed field. In this paper we explore the connection between Ax-Schanuel type theorems (predimension inequalities) for a differential equation $E(x,y)$ and the geometry of the set $U:=\{ y:E(t,y) \wedge y' \neq 0 \}$ where $t$ is an element with $t'=1$. We show that certain types of predimension inequalities imply strong minimality and geometric triviality of $U$. Moreover, the induced structure on Cartesian powers of $U$ is given by special subvarieties. If $E$ has some special form then all fibres $U_s:=\{ y:E(s,y) \wedge y' \neq 0 \}$ (with $s$ non-constant) have the same properties. In particular, since the $j$-function satisfies an Ax-Schanuel theorem of the required form (due to Pila and Tsimerman), our results will give another proof for a theorem of Freitag and Scanlon stating that the differential equation of $j$ defines a strongly minimal set with trivial geometry (which is not $\aleph_0$-categorical though). Introduction Throughout the paper we let K = (K; +, ·, ′ , 0, 1) be a differentially closed field with field of constants C. Let E(x, y) be (the set of solutions of) a differential equation f (x, y) = 0 with rational (or, more generally, constant) coefficients. A general question that we are interested in is whether E satisfies an Ax-Schanuel type inequality. A motivating example is the exponential differential equation y ′ = yx ′ . We know that (the original) Ax-Schanuel ( [Ax71]) gives a predimension inequality (in the sense of Hrushovski [Hru93]) which governs the geometry of our equation. In this case the corresponding reduct of a differentially closed field can be reconstructed by a Hrushovski-style amalgamation-with-predimension construction ( [Kir09]). Zilber calls this kind of predimension inequalities adequate (see [Asl17b,Asl18a] for a precise definition). This means that the reduct satisfies a strong existential closedness property which asserts, roughly speaking, that a system of exponential equations which is not overdetermined has a solution. Being overdetermined can be interpreted as contradicting the Ax-Schanuel inequality. Thus, having an adequate Ax-Schanuel inequality for E will give us a complete understanding of its model theory. For more details on this and related problems see [Asl17b,Asl18a,Kir09,Zil04,Zil05]. Note however that adequacy will not play any role in this paper, we just mentioned it here in order to give a basic idea why Ax-Schanuel inequalities are important. Ax-Schanuel type statements can also be applied to diophantine geometry. Indeed, they can be used to prove a weak version of the famous Zilber-Pink conjecture in the appropriate setting (see [Zil02,PT16,Kir09,Asl18b]). Thus, we want to classify differential equations of two variables with respect to the property of satisfying an Ax-Schanuel type inequality. The present work should be seen as a part of that more general project. We explore the connection between Ax-Schanuel type theorems (predimension inequalities) for a differential equation E(x, y) and the geometry of the fibres of E. More precisely, given a predimension inequality (not necessarily adequate) for solutions of E of a certain type (which is of the form "td − dim" where dim is a dimension of trivial type) we show that fibres of E are strongly minimal and geometrically trivial (after removing constant points). Moreover, the induced structure on the Cartesian powers of those fibres is given by special subvarieties. In particular, since an Ax-Schanuel theorem (of the required form) for (the differential equation of) the j-function is known (due to Pila and Tsimerman [PT16]), our results will give a new proof for a theorem of Freitag and Scanlon [FS15] stating that the differential equation of j defines a strongly minimal set with trivial geometry (which is not ℵ 0categorical though). In fact, the Pila-Tsimerman inequality is the main motivation for this paper. Thus we get a necessary condition for E to satisfy an Ax-Schanuel inequality of the given form. This is a step towards the solution of the problem described above. In particular it gives rise to an inverse problem: given a one-variable differential equation which is strongly minimal and geometrically trivial, can we say anything about the Ax-Schanuel properties of its two-variable analogue (see Section 5 for more details)? On the other hand, understanding the structure of strongly minimal sets in a given theory is one of the most important problems in model theory. In DCF 0 there is a nice classfication of strongly minimal sets, namely, they satisfy the Zilber trichotomy (Hrushovski-Sokolović [HuS93]). Hrushovski [Hru95] also gave a full characterisation of strongly minimal sets of order 1 proving that such a set is either non-orthogonal to the constants or it is trivial and ℵ 0 -categorical. However there is no general classification of trivial strongly minimal sets of higher order and therefore we do not fully understand the nature of those sets. From this point of view the set J defined by the differential equation of j is quite intriguing since it is the first example of a trivial strongly minimal set in DCF 0 which is not ℵ 0 -categorical. Before Freitag and Scanlon established those properties of J in [FS15], it was mainly believed that trivial strongly minimal sets in DCF 0 must be ℵ 0 -categorical. The reason for this speculation was Hrushovski's aforementioned theorem on order 1 strongly minimal sets (and the lack of counterexamples). Thus, the classification of strongly minimal sets in DCF 0 can be seen as another source of motivation for the work in this paper, where we show that these two problems (Ax-Schanuel type theorems and geometry of strongly minimal sets) are in fact closely related. Let us give a brief outline of the paper. After defining the appropriate notions we formulate the main results of the paper in Section 2. We prove those results in Section 3. Then we give a brief account of the j-function and apply our results to its differential equation in Section 4. Section 5 is devoted to some concluding remarks and inverse questions. We have gathered definitions of several properties of strongly minimal sets that we need in Appendix A. Setup and main results Recall that K = (K; +, ·, ′ , 0, 1) is a differentially closed field with field of constants C. We may assume (without loss of generality) K is sufficiently saturated if necessary. Fix an element t with t ′ = 1. Let E(x, y) be (the set of solutions of) a differential equation f (x, y) = 0 with rational (or, more generally, constant) coefficients. Now we give several definitions and then state the main results of the paper. Definition 2.1. Let P be a non-empty collection of algebraic polynomials P (X, Y ) ∈ C[X, Y ]. We say two elements a, b ∈ K are P-independent if P (a, b) = 0 and P (b, a) = 0 for all P ∈ P. The P-orbit of an element a ∈ K is the set {b ∈ K : P (a, b) = 0 or P (b, a) = 0 for some P ∈ P} (in analogy with a Hecke orbit, see Section 4). Also, P is said to be trivial if it consists only of the polynomial X − Y . Definition 2.2. For a non-constant x ∈ K the differentiation with respect to x is the derivation ∂ x : K → K defined by ∂ x : y → y ′ x ′ . Recall that f (x, y) = 0 is the differential equation defining E and denote m := ord Y f (X, Y ) (the order of f with respect to Y ). Definition 2.3. We say the differential equation E(x, y) has the P-AS (Ax-Schanuel with respect to P) property if the following condition is satisfied: Let x 1 , . . . , x n , y 1 , . . . , y n be non-constant elements of K with f (x i , y i ) = 0. If the y i 's are pairwise P-independent then td C C x 1 , y 1 , ∂ x 1 y 1 , . . . , ∂ m−1 x 1 y 1 , . . . , x n , y n , ∂ xn y n , . . . , ∂ m−1 xn y n ≥ mn + 1. (2.1) The P-AS property can be reformulated as follows: for any non-constant solutions (x i , y i ) of E the transcendence degree in (2.1) is strictly bigger than m times the number of different P-orbits of y i 's. Note that (2. Remark 2.4. Having the P-AS property for a given equation E may force P to be "closed" in some sense. Firstly, X − Y must be in P. Secondly, if P 1 , P 2 ∈ P then P 1 (y 1 , y 2 ) = 0, P 2 (y 2 , y 3 ) = 0 impose a relation on y 1 and y 3 given by Q(y 1 , y 3 ) = 0 for some polynomial Q. Then the P-AS property may fail if Q ∈ P. In that case one has to add Q to P in order to allow the possibility of an Ax-Shcanuel property with respect to P. Similar conditions on P are required in order for P-independence to define a dimension function (number of distinct P-orbits) of a pregeometry (of trivial type), which would imply that the P-AS property is a predimension inequality. Note that the collection of modular polynomials has all those properties. However, the shape of P is not important for our results since we assume that a given equation E has the P-AS property. Definition 2.5. A P-special variety (in K n for some n) is an irreducible (over C) component of a Zariski closed set in K n defined by a finite collection of equations of the form P ik (y i , y k ) = 0 for some P ik ∈ P. For a differential subfield L ⊆ K a weakly P-special variety over L is an irreducible (over L alg ) component of a Zariski closed set in K n defined by a finite collection of equations of the form P ik (y i , y k ) = 0 and y i = a for some P ik ∈ P and a ∈ L alg . For a definable set V , a (weakly) P-special subvariety (over L) of V is an intersection of V with a (weakly) P-special variety (over L). 1 Remark 2.6. If the polynomials from P have rational coefficients then P-special varieties are defined over Q alg . Furthermore, if E satisfies the P-AS property then for the set U := {y : f (t, y) = 0 ∧ y ′ = 0} we have U ∩ C(t) alg = ∅ and so P-special subvarieties of U over C(t) are merely P-special subvarities. Notation. For differential fields L ⊆ K and a subset A ⊆ K the differential subfield of K generated by L and A will be denoted by L A . Let C 0 ⊆ C be the subfield of C generated by the coefficients of f and let K 0 = C 0 t = C 0 (t) be the (differential) subfield of K generated by C 0 and t (clearly U is defined over K 0 ). We fix K 0 and work over it (in other words we expand our language with new constant symbols for elements (generators) of K 0 ). Now we can formulate one of our main results (see Section A for definitions of geometric triviality and strict disintegratedness). Theorem 2.7. Assume E(x, y) satisfies the P-AS property for some P. Assume further that the differential polynomial g(Y ) := f (t, Y ) is absolutely irreducible. Then is strongly minimal with trivial geometry. • If, in addition, P is trivial then U is strictly disintegrated and hence it has ℵ 0categorical induced structure. • All definable subsets of U n over a differential field L ⊇ K 0 are Boolean combinations of weakly P-special subvarieties over L. As the reader may guess and as we will see in the proof, this theorem holds under weaker assumptions on E. Namely, it is enough to require that (2.1) hold for x 1 = . . . = x n = t (which can be thought of as a weak form of the "Ax-Lindemann-Weierstrass" property 2 ). However, we prefer the given formulation of Theorem 2.7 since the main object of our interest is the Ax-Schanuel inequality (for E). Further, we deduce from Theorem 2.7 that if E has some special form, then all fibres E(s, y) for a non-constant s ∈ K have the above properties (over C 0 s ). Corollary 2.8. Let E(x, y) be defined by P (x, y, ∂ x y, . . . , ∂ m x y) = 0 where P (X,Ȳ ) is an irreducible algebraic polynomials over C. Assume E(x, y) satisfies the P-AS property for some P and let s ∈ K be a non-constant element. Then is strongly minimal with trivial geometry. • If, in addition, P is trivial then any distinct non-algebraic (over C 0 s ) elements are independent and U s is ℵ 0 -categorical. • All definable subsets of U n s over a differential field L ⊇ C 0 s are Boolean combinations of weakly P-special subvarieties over L. Remark 2.9. Since U s ∩ C = ∅, in Theorem 2.7 and Corollary 2.8 the induced structure on U n s is actually given by strongly special subvarieties (over L), which means that we do not allow any equation of the form y i = c for c a constant. In particular we also need to exclude equations of the form P (y i , y i ) = 0 for P ∈ P. We also prove a generalisation of Theorem 2.7. Theorem 2.10. Assume E(x, y) satisfies the P-AS property and let p( is strongly minimal and geometrically trivial. As an application of Theorem 2.7 we obtain a result on the differential equation of the j-function which was established by Freitag and Scanlon in [FS15]. To be more precise, let F (j, j ′ , j ′′ , j ′′′ ) = 0 be the algebraic differential equation satisfied by the modular j-function (see Section 4). Strong minimality and geometric triviality of J follow directly from Theorem 2.7 combined with the Ax-Schanuel theorem for j (see Section 4). Of course the "furthermore" clause does not follow from Theorem 2.7 but it is not difficult to prove. Theorem 2.7 also gives a characterisation of the induced structure on the Cartesian powers of J. Again, that result can be found in [FS15]. The proof of Theorem 2.11 by Freitag and Scanlon is based on Pila's modular Ax-Lindemann-Weierstrass with derivatives theorem along with Seidenberg's embedding theorem and Nishioka's result on differential equations satisfied by automorphic functions ( [Nis89]). They also make use of some tools of stability theory such as the "Shelah reflection principle". However, as one may guess, we cannot use Nishioka's theorem (or some analogue of that) in the proof of 2.7 since we do not know anything about the analytic properties of the solutions of our differential equation. Thus, we show in particular that Theorem 2.11 can be deduced from Pila's result abstractly. The key point that makes this possible is stable embedding, which means that if M is a model of a stable theory and X ⊆ M is a definable set over some A ⊆ M then every definable (with parameters from M) subset of X n can in fact be defined with parameters from X ∪ A (see Section A). Let us stress once more that the set J is notable for being the first example of a strongly minimal set (definable in DCF 0 ) with trivial geometry that is not ℵ 0 -categorical. Indeed the aforementioned result of Hrushovski on strongly minimal sets of order 1 led people to believe that all geometrically trivial strongly minimal sets must be ℵ 0 -categorical. Nevertheless, it is not true as the set J illustrates. Proofs of the main results Proof of Theorem 2.7 Taking x 1 = . . . = x n = t in the P-AS property we get the following weak version of the P-ALW property for U which in fact is enough to prove Theorem 2.7. We show that every definable (possibly with parameters) subset V of U is either finite or co-finite. Since U is defined over K 0 , by stable embedding there is a finite subset A = {a 1 , . . . , a n } ⊆ U such that V is defined over K 0 ∪ A. It suffices to show that U realises a unique non-algebraic type over K 0 ∪ A, i.e. for any u 1 , u 2 ∈ U \ acl(K 0 ∪ A) we have tp(u 1 /K 0 ∪ A) = tp(u 2 /K 0 ∪ A). Let u ∈ U \ acl(K 0 ∪ A). We know that acl(K 0 ∪ A) = (K 0 A ) alg = (K 0 (ā,ā ′ , . . . ,ā (m−1) )) alg . Since u / ∈ (K 0 A ) alg , u is transcendental over K 0 (A) and hence it is P-independent from each a i . We may assume without loss of generality that a i 's are pairwise P-independent (otherwise we could replace A by a maximal pairwise P-independent subset). Applying Lemma 3.1 toā, u, we deduce that u, u ′ , . . . , u (m−1) are algebraically independent over K 0 A . Hence tp(u/K 0 ∪ A) is determined uniquely (axiomatised) by the set of formulae (Recall that g is absolutely irreducible and hence it is irreducible over any field). In other words g(Y ) is the minimal differential polynomial of u over K 0 A . Thus U is strongly minimal. A similar argument shows also that if A ⊆ U is a (finite) subset and u ∈ U ∩ acl(K 0 A) then there is a ∈ A such that u ∈ acl(K 0 a). This proves that U is geometrically trivial. If P is trivial then distinct elements of U are independent, hence U is strictly disintegrated. The last part of Theorem 2.7 follows from the following lemma. Lemma 3.2. Every irreducible (relatively) Kolchin closed (over C(t)) subset of U n is a P-special subvariety of U n . Proof. Let V ⊆ U n be an irreducible relatively closed subset (i.e. it is the intersection of U n with an irreducible Kolchin closed set in K n ). Pick a generic pointv = (v 1 , . . . , v n ) ∈ V and let W ⊆ K n be the Zariski closure ofv over C. Let d := dim W and assume v 1 , . . . , v d are algebraically independent over C. Then v i ∈ (C(v 1 , . . . , v d )) alg for each i = d + 1, . . . , n. By Lemma 3.1 each v i with i > d must be in a P-relation with some The algebraic variety defined by the equations P i (y i , y k i ) = 0, i = d + 1, . . . , n, has dimension d and contains W . Therefore W is a component of that variety and so it is a P-special variety. We claim that W ∩ U n = V . Since v 1 , . . . , v d ∈ U are algebraically independent over C, by Lemma 3.1v,v ′ , . . . ,v (m−1) are algebraically independent over C(t). Moreover, the (differential) type of each v i , i > d, over v 1 , . . . , v d is determined uniquely by an irreducible algebraic equation. Therefore tp(v/C(t)) is axiomatised by formulas stating thatv is Zariski generic in W and belongs to U n . In other wordsv is Kolchin generic in W ∩ U n . Now V and W ∩ U n are both equal to the Kolchin closure ofv inside U n and hence they are equal. Thus definable (over C(t)) subsets of U n are Boolean combinations of special subvarieties. Now let L ⊆ K be an arbitrary differential subfield over which U is defined. Then definable subsets of U n over L can be defined with parameters fromL = K 0 ∪ (U ∩ L alg ) (see Section A). Then Lemma 3.2 implies that irreducible Kolchin closed subsets of U n defined overL are weakly P-special subvarities of U n over L. Finally, note that since U does not contain any algebraic elements over C(t), the type of any element u ∈ U over C(t) is isolated by the formula f (t, y) = 0 ∧ y ′ = 0. Proof of Theorem 2.10 We argue as above and show that for a finite set A = {a 1 , . . . , a n } ⊆ U p,q there is a unique non-algebraic type over K 0 A realised in U p,q . Here we will use full Ax-Lindemann-Weierstrass. If u ∈ U p,q \ (K 0 A ) alg then q(u) is transcendental over K 0 (A) and so q(u) is Pindependent from each q(a i ). Moreover, we may assume {q(a 1 ), . . . , q(a n )} is P-independent. Then by the P-AS property ..,n ≥ m(n + 1) + 1. But then ..,n ≥ m(n + 1) + 1, and hence u, u ′ , . . . , u (m−1) are algebraically independent over K 0 A . This determines the type tp(u/K 0 A) uniquely as required. It also shows triviality of the geometry. Proof of Corollary 2.8 Consider the differentially closed field K s = (K; +, ·, ∂ s , 0, 1). The given form of the differential equation E implies that U s is defined over C 0 (s) in K s . However, in general it may not be defined over C 0 (s) in K, it is defined over C 0 s = C 0 (s, s ′ , s ′′ , . . .). Since ∂ s s = 1, we know by Theorem 2.7 that U s is strongly minimal in K s . On the other hand the derivations ∂ s and ′ are inter-definable (with parameters) and so a set is definable in K if and only if it is definable in K s (possibly with different parameters). This implies that every definable subset of U s in K is either finite or co-finite, hence it is strongly minimal. Further, Theorem 2.7 implies that U s is geometrically trivial over C 0 (s) in K s . By Theorem A.2, U s is also geometrically trivial over C 0 s in K s . On the other hand for any subset A ⊆ U s the algebraic closure of C 0 s ∪ A is the same in K and K s . This implies geometric triviality of U s in K s . The same argument (along with the remark after Theorem A.2) shows that the second and the third parts of Corollary 2.8 hold as well. The modular j-function The modular j-invariant satisfies the following order three algebraic differential equation: where S denotes the Schwarzian derivative defined by Sy = y ′′′ y ′ − 3 2 y ′′ y ′ 2 and R(y) = y 2 −1968y+2654208 2y 2 (y−1728) 2 . Let J be the set defined by (4.2). Note that F is not a polynomial but a rational function. In particular constant elements do not satisfy (4.2) for Sy is not defined for a constant y. We can multiply our equation through by a common denominator and make it into a polynomial equation F * (y, y ′ , y ′′ , y ′′′ ) = q(y)y ′ y ′′′ − 3 2 q(y)(y ′′ ) 2 + p(y)(y ′ ) 4 = 0, (4.3) where p and q are respectively the numerator and the denominator of R. Let J * be the set defined by (4.3). It is not strongly minimal since C is a definable subset. However, as we will see shortly, J = J * \ C is strongly minimal and MR(J * ) = 1, MD(J * ) = 2. Thus whenever we speak of the formula F (y, y ′ , y ′′ , y ′′′ ) = 0 (which, strictly speaking, is not a formula in the language of differential rings) we mean the formula F * (y, y ′ , y ′′ , y ′′′ ) = 0 ∧ y ′ = 0. Definition 4.1. For each N we let Φ N (X, Y ) be the N-th modular polynomial (see, for example, [Lan73]). Denote Φ := {Φ N (X, Y ) : N > 0}. We say two elements are modularly independent if they are Φ-independent. For an element a ∈ K its Hecke orbit is its Φ-orbit. Let us form the two-variable analogue of equation (4.2): Now we formulate the (differential version of the) Ax-Schanuel theorem for j established by Pila and Tsimerman in [PT16]. If j i 's are pairwise modularly independent then td C C(z 1 , j 1 , ∂ z 1 j 1 , ∂ 2 z 1 j 1 , . . . , z n , j n , ∂ zn j n , ∂ 2 zn j n ) ≥ 3n + 1. (4.5) In other words (4.4) has the Φ-AS property. Note that here K does not need to be differentially closed, the result holds for arbitrary differential fields (indeed, the Ax-Schanuel theorem is a universal statement). As a consequence of Theorems 2.7 and 4.2 we get strong minimality and geometric triviality of J (note that F * (Y 0 , Y 1 , Y 2 , Y 3 ) is obviously absolutely irreducible as it depends linearly on Y 3 ). Lemma 3.1 for j is of course a special case of the Ax-Schanuel theorem for j. Nevertheless it can also be deduced from Pila's modular Ax-Lindemann-Weierstrass with derivatives theorem ([Pil13]) by employing Seidenberg's embedding theorem. Therefore only Pila's theorem is enough to prove strong minimality and geometric triviality of J. Moreover, Corollary 2.8 shows that all non-constant fibres of (4.4) are strongly minimal and geometrically trivial (after removing constant points) and the induced structure on the Cartesian powers of those fibres is given by (strongly) special subvarieties. Note that it is proven in [FS15] that the sets F (y, y ′ , y ′′ , y ′′′ ) = a have the same properties for any a. Remark 4.3. To complete the proof of Theorem 2.11, that is, to show that J is not ℵ 0categorical, one argues as follows (see [FS15]). The Hecke orbit of an element j ∈ J is contained in J. Therefore J realises infinitely many types over j and hence is not ℵ 0 -categorical. Concluding remarks The P-AS property states positivity of a predimension of the form "td −m · d", where d is the number of distinct P-orbits. 3 Transcendence degree, being the algebraic dimension, is non-locally modular. On the other hand d is a dimension of trivial type. And it is this fact that is responsible for triviality of the geometry of U. For the exponential differential equation the predimension is of the form "td − ldim" where ldim stands for the linear dimension. The linear dimension is locally modular non-trivial and accordingly the strongly minimal set y ′ = y is not trivial; indeed it is non-orthogonal to C. Here strong minimality is obvious as the equation has order one. However one may ask whether strong minimality can be deduced from this type of predimension inequalities in general. The answer is no. For example, consider a linear differential equation with constant coefficients ∂ 2 x y − y = 0. We showed in [Asl17a] that an Ax-Schanuel statement holds for it. However U is the set defined by y ′′ − y = 0 and the set y ′ = y is a definable infinite and co-infinite subset (the differential polynomial Y ′′ − Y is absolutely irreducible though). An interesting question is whether there are differential equations with the P-AS property with trivial P. As we showed here, if E(x, y) has such a property then the corresponding U must be strongly minimal and strictly disintegrated. There are quite a few examples of this kind of strongly minimal sets in DCF 0 . The two-variable versions of those equations will be natural candidates of equations with the required P-AS property. For example, the geometry of sets of the form y ′ = f (y), where f is a rational function over C, is well understood. The nature of the geometry is determined by the partial fraction decomposition of 1/f . As an example consider the equation One can show that it defines a strictly disintegrated strongly minimal set ( [Mar05]). The two variable analogue of this equation is But this is equivalent to the equation y ′ y = (x − y) ′ . Denoting z = x − y we get the exponential differential equation y ′ = yz ′ . It is easy to deduce from this that (5.7) does not satisfy the P-AS property with any P (it satisfies a version of the original exponential Ax-Schanuel inequality though and therefore has a predimension inequality which is of the form td − ldim). Indeed, the fibre of (5.7) above x = t is of trivial type but the section by x = t + y is non-orthogonal to C. So according to Theorem 2.10 the equation (5.7) does not satisfy any P-AS property. Of course, all the sets y ′ = f (y) can be treated in the same manner and hence they are not appropriate for our purpose. Thus, one needs to look at the behaviour of all the sets E(p(y), q(y)), and if they happen to be trivial strongly minimal sets then one can hope for a P-AS inequality. The classical Painlevé equations define strongly minimal and strictly disintegrated sets as well. For example, let us consider the first Painlevé equation y ′′ = 6y 2 + t. Strong minimality and algebraic independence of solutions of this equation were shown by Nishioka in [Nis04] (note that strong minimality was discovered earlier by Kolchin (see [Mar05])). We consider its two-variable version ∂ 2 x y = 6y 2 + x. (5.8) The goal is to find an Ax-Schanuel inequality for this equation. Note that (5.8) does not satisfy the P-AS property with trivial P. Indeed, if ζ is a fifth root of unity then the transformation x → ζ 2 x, y → ζy sends a solution of (5.8) to another solution. If one believes these are the only relations between solutions of the above equation, then one can conjecture the following. One could in fact replace x's with y's in the condition (x i /x j ) 5 = 1 as those are equivalent. Nagloo and Pillay showed in [NP14] that the other generic Painlevé equations define strictly disintegrated strongly minimal sets as well. So we can analyse relations between solutions of their two-variable analogues and ask similar questions for them too. In general, proving that a certain equation has the required transcendence properties may be much more difficult than proving that there are equations with those properties. In this regard we believe that "generic" (in a suitable sense) equations must satisfy the P-AS property with trivial P. However we do not go into details here and finish with a final remark. Zilber constructed "a theory of a generic function" where the function has transcendence properties analogous to trivial AS property described here ( [Zil05]). He also conjectured that it has an analytic model, i.e. there is an analytic function that satisfies Zilber's axioms and, in particular, the given transcendence properties. Wilkie constructed Liouville functions in [Wil05] and showed that they indeed satisfy the transcendence properties formulated by Zilber. Later Koiran [Koi03] proved that Liouville functions satisfy Zilber's existential closedness statement as well. However, those functions do not satisfy any algebraic differential equation, so we cannot translate the result into a differential algebraic language. If there is a differentially algebraic function with similar properties then it may give rise to a differential equation with the desired properties. Appendix A On strong minimality In this appendix we define some standard properties of strongly minimal sets that are used throughout the paper. We also prove that geometric triviality of a strongly minimal set does not depend on the set of parameters over which our set is defined. It is of course a well known classical result, but we sketch a proof here since we use the result in the proof of Corollary 2.8. For a detailed account of strongly minimal sets and geometric stability theory in general we refer the reader to [Pil96]. Algebraic closure defines a pregeometry on a strongly minimal set. More precisely, if X is a strongly minimal set in a structure M defined over A ⊆ M then the operator cl : Y → acl(AY ) ∩ X, for Y ⊆ X, is a pregeometry. Definition A.1. Let M be a structure and X ⊆ M be a strongly minimal set defined over a finite set A ⊆ M. • We say X is geometrically trivial (over A) if whenever Y ⊆ X and z ∈ acl(AY ) ∩ X then z ∈ acl(Ay) for some y ∈ Y . In other words, geometric triviality means that the closure of a set is equal to the union of closures of singletons. • X is called strictly disintegrated (over A) if any distinct elements x 1 , . . . , x n ∈ X are independent (over A). • X is called ℵ 0 -categorical (over A) if it realises only finitely many 1-types over AY for any finite Y ⊆ X. This is equivalent to saying that acl(AY ) ∩ X is finite for any finite Y ⊆ X. Theorem A.2. Let M be a model of an ω-stable theory and X ⊆ M be as above. If X is geometrically trivial over A then it is geometrically trivial over any superset B ⊇ A. Proof. By expanding the language with constant symbols for elements of A we can assume that X is ∅-definable. Also we can assume B = {b 1 , . . . , b n } is finite. Let z ∈ acl(BY ) for some finite Y ⊆ X. By stability tp(b/X) is definable over a finite C ⊆ X and we may assume that C ⊆ acl(B) ∩ X. Therefore z ∈ acl(CY ). By geometric triviality of X (over ∅) we have z ∈ acl(c) for some c ∈ C or z ∈ acl(y) for some y ∈ Y . This shows geometric triviality of X over B. As we saw in the proof all definable subsets of X m over B are definable over acl(B)∩X (which is the stable embedding property). The same argument shows that ℵ 0 -categoricity does not depend on parameters (see also [NP16]). Of course this is not true for strict disintegratedness but a weaker property is preserved. Namely if X is strictly disintegrated over A then any distinct non-algebraic elements over B are independent over B.	2018-04-13T09:13:49.000Z	2016-06-06T00:00:00.000	{ "year": 2016, "sha1": "919c7bfc16fb702a5f27a2dec80ec912eb42f343", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "dbf7ee7b0d5425efd4b06d54f6633bb4b5ae7351", "s2fieldsofstudy": [ "Mathematics" ], "extfieldsofstudy": [ "Mathematics" ] }
195873415	pes2o/s2orc	v3-fos-license	Spatial variation of vector vortex beams with plasmonic metasurfaces The spatial variation of vector vortex beams with arbitrary polarization states and orbital angular momentum (OAM) values along the beam propagation is demonstrated by using plasmonic metasurfaces with the initial geometric phase profiles determined from the caustic theory. The vector vortex beam is produced by the superposition of deflected right- and left-handed circularly polarized component vortices with different helical phase charges, which are simultaneously generated off-axially by the single metasurface. Besides, the detailed evolution processes of intensity profile, polarization distribution and OAM value along the beam propagation distance is analyzed. The demonstrated arbitrary space-variant vector vortex beam will pave the way to many promising applications related to spin-to-orbital angular momentum conversion, spin-orbit hybrid entanglement, particle manipulation and transportation, and optical communication. Optical vector vortex beams have polarization singularities and also possess phase singularities in both their right-and left-handed circularly polarized (RCP and LCP) component vortices. With particular characteristics in polarization and phase structures, vector vortex beams have made a wide variety of applications in the fields of optical tight focus 1-3 , optical tweezing [4][5][6] , spin-to-orbital angular momentum conversion 7-10 , confocal microscopy 11,12 , and metrology 13,14 . An optical vortex beam with homogeneous polarization state has a helical phase front with azimuthal phase dependency of exp(ilϕ), where ϕ is the azimuthal angle, l is the topological charge (TC) of helical phase with lħ representing the OAM carried by each photon. A vector vortex beam with space-variant polarization state has the singularity of polarization which is coincident with the corresponding singularity of helical phase. As the most common vector vortex beam, radially or azimuthally polarized beam is produced by the equally-weighted superposition of two component vortices with opposite circular polarization states and opposite OAM values of l = +1 and l = −1. While the high-order vector vortex beam is composed by two component vortices with large OAM values and opposite circular polarizations 15 , which can be used for ultrasensitive angular measurement and spin object detection 13,14 . The OAM value and polarization distribution of an optical beam can be changed along the beam propagation in free space. Recently, the topological charge inversion of optical vortex has been demonstrated 16,17 and the Poincaré beams with polarization transformation during the free space propagation are shown 18 . The vector vortex beams with both varying OAM and polarization state has also been studied 19,20 . It is essential to generate vector vortex beams with arbitrarily polarization states and OAM values varying along the beam propagation, which has not been demonstrated yet. Here, an effective approach to realize the spatial variation of vector vortex beams with arbitrary polarization states and OAM values of RCP and LCP component vortices along the beam propagation is demonstrated based on the plasmonic metasurfaces with the initial phase profiles designed from the caustic theory. With only a single metasurface, both the deflected RCP and LCP component vortices with different helical phase charges are Department of Mechanical and Aerospace engineering, Missouri University of Science and technology, Rolla, MO, 65409, USA. Correspondence and requests for materials should be addressed to J.G. (email: gaojie@mst.edu) or X.Y. (email: yangxia@mst.edu) generated simultaneously, and the vector vortex beam is produced by the superposition of these two component vortices. Furthermore, the detailed evolution processes of intensity profile, polarization distribution and OAM value along the beam propagation distance is presented. The spatial variation of vector vortex beams could have miscellaneous applications due to the polarization singularities and the OAMs carried in both RCP and LCP component vortices. The variation of vector vortex provides different OAMs and polarization states at different propagation distances, which give more degrees of freedom for the applications involved with both OAM and polarization, especially in the regimes of particle manipulation, optical communication and quantum information processing. For the particle manipulation, both OAM and polarization of light can be transferred to microparticles with torques, and different OAMs or polarization states will create different mechanical torques to microparticles. Thereby, the space-variant vector vortex beam could generate space-depended torque that has distinct magnitude and orientation at each location, and such feature could be used as particle and DNA investigations 46,47 . For the optical communication and quantum information processing, both OAM and polarization state can represent the bit and qubit of information. The space-variant OAM and polarization state provide extended degrees of freedom for the entanglement based on both spin and orbital angular momentum 48,49 . Results Design of plasmonic metasurface. As shown in Fig. 1(a), the subwavelength nanoslit antennas with different orientation angles are etched in a gold film with thickness of 50 nm on glass substrate using focused ion beam (FIB, FEI Helios Nanolab 600 Dual Beam system). The unit cell of the nanoslit antenna has the period of 330 nm, and the width and length of nanoslit antenna is 60 nm and 200 nm, respectively. As a circularly polarized beam is incident on the metasurface, the introduced geometric phase shift to the converted spin component transmitted from each nanoslit antenna is 2θ which is twice as the orientation angle of nanoslit, while the original spin component has no phase shift. Also for RCP incidence the transmitted LCP component has a phase shift of 2θ, while for LCP incidence the transmitted RCP component has a reversed phase shift of −2θ 15,50 . The simulated relationship between the phase shift of the converted polarization and the orientation angle of nanoslit antenna is shown in Fig. 1(b). Figure 1(c) shows the simulated electric field \|E\| distributions at the wavelength of 633 nm, where the RCP and LCP components have different field profiles, indicating strong polarization anisotropy. Figure 1(d) gives the SEM image of the fabricated homogeneous array of nanoslit antennas. For the transmission under circular polarization basis, the total transmitted beam through the metasurface contains both the original spin component and the converted spin component. The polarization conversion efficiency is defined as the intensity ratio between the converted spin component and the total transmitted beam. Figure 1(e) plots the measured and simulated spectra of original spin transmission (green lines), converted spin transmission (red lines) and conversion efficiency (CE, black lines). The observed maximum conversion efficiency is around 55% near 800 nm, where the plasmonic resonance occurs. Here a HeNe laser at 633 nm is used in experiments due to its high-quality laser beam for the measurements of both beam intensity profiles and interference patterns. The transmission efficiency of plasmonic metasurfaces is relatively low due to the large Ohmic loss of metal. However, the principle of geometric phase-based metasurfaces can be extended to the recently developed dielectric metasurfaces made of silicon or titanium oxide which have very low absorption loss for manipulating light with ultrahigh transmission efficiency and high polarization conversion efficiency [24][25][26] . It has been shown that the vector vortex beam can be generated by the superposition of two component vortices with opposite circular polarizations and different helical phase charges, with the optical field for vector vortex beam expressed as 15,51 : T represents the LCP component with charge of l L , and Θ is a phase constant to represent the phase difference between the RCP and LCP components, so that the total vector field is written as 52 : where m = (l R − l L )/2 represents the TC of polarization state for the vector vortex beam, Θ is the initial polarization orientation for ϕ = 0, and n = (l R + l L )/2 is the average OAM value of RCP and LCP component vortices. The parameter m describes polarization rotations per round trip along the azimuthal direction, which is called as the index of the vector vortex beam 53 . For example, the case for n = 0, m = 1 and Θ = 0 corresponds to the radially polarized beam which is a transverse magnetic (TM) beam, while the case for n = 0, m = 1 and Θ = π/2 gives the azimuthally polarized beam which is a transverse electric (TE) beam. To distinguish the original spin component and the converted spin component, a linear phase gradient is imposed into the geometric phase profile, so that the converted spin component is deflected by an angle with respect to the original spin component. As shown in Fig. 2(a), after passing through the metasurface, the incident LCP or RCP beam is converted into two deflected RCP or LCP vortex beams with the desired helical phase charges. As the incident beam is linearly polarized, the converted beam will be a vector vortex beam with the superposition of both the output RCP and LCP components. The initial geometric phase profile encoded on the metasurface is determined from the caustic theory. It has been shown that the caustic curve can be generated by modulating the phase on the initial input plane, which has been used to produce a series of nondiffracting Bessel-like beams with arbitrary trajectories 54 . Assuming the initial input field is Ψ = = x y z Gr i kQ x y ( , , 0) ( ) exp( ( , )) 0 0 0 0 0 , the optical field at propagation distance z could be calculated by the paraxial Fresnel integral: 54 , the point at the propagation distance z = z′ corresponds to a circle C(z = z′) on the input plane and a conical ray bundle emitted from this circle will intersect at this point, as shown in Fig. 2(b). As the circle C(z) shifts and expands on the input plane, the continuous caustic curve represented by f z g z z [ ( ), ( ), ] is formed by the locus of the apexes of the conical ray bundles emitted from the circle C(z). The function of circle C(z) for generating a conical ray bundle . Due to the large wavenumber k, the terms (l/k)•dϕ/dx 0 and (l/k)•dϕ/dy 0 can be neglected, so that after adding the vortex structure, the optical field at z is still majorly contributed by the circle C(z). In this case, the helical phase on the circle C(z) is exp(jlϕ), and by using polar coordinates of www.nature.com/scientificreports www.nature.com/scientificreports/ (cos , sin ), the complex field distribution at the propagation distance z is calculated by www.nature.com/scientificreports www.nature.com/scientificreports/ In order to generate space-variant vector vortex beam, the initial input plane is divided into four different zones, and different zones have different TCs of helical phase. Since the optical field distribution at the propagation distance z is mainly determined by the circle C(z) on the input plane, vortex fields with helical phase charges varying as the propagation distance can be created by different circular zones with the corresponding charges on the input plane. As the beam propagates, different circular zones will generate optical vortices with varying charges. The yellow dashed circles divide the input plane into different circular zones, and the distance Zt(s) corresponding to each yellow dashed circle is defined as the transition distance. A function L(z) is used to represent the TCs of helical phase at different distances: The vector vortex beam is produced by the superposition of two vortex beams with opposite circular polarizations and different helical phase charges. Since the metasurface formed by rotating the nanoslit antenna will generate geometric phase profiles of ϕ geom (x, y) and −ϕ geom (x, y) for the LCP and RCP components, the phase profiles for the LCP and RCP components need to be designed separately. First, for the LCP component, the caustic phase function of Eq. (6) is 4 . Then a linear phase gradient of π Λ x 2 / (Λ = 1.6 µm) is imposed to the caustic phase function P L , so that the converted LCP beam will propagate with a deflection angle of α λ = Λ arctan( / ) and the total complex amplitude is π A linear phase gradient π Λ x 2 / is also imposed to P R . Since the geometric phase shift is negative for the RCP component, the total complex amplitude for the converted RCP beam is Finally, the total geometric phase profile for producing space-variant vector vortex beam is obtained by superimposing the complex amplitudes of both RCP and LCP components and then calculating the argument of the summation complex function, as shown in Fig. 2(c). The total geometric phase profile from the superposition of the RCP component with helical phase charge of L R and the LCP component with charge of L L is then expressed as 15,55 : where Θ z is the phase difference between the RCP and LCP components as a function of propagation distance, and Θ z has different values on different divided circular zones on the input plane. Θ z determines the initial polarization orientation of vector vortex beam, with Θ z = 0 and Θ z = π/2 correspond to TM and TE polarization mode, respectively. There are two channels of the vector vortex beams generated by the total geometric phase profile from Eq. (7). The first term of Eq. (7) with phase gradient 2πx/Λ will deflect the RCP component at a positive angle, while the second term with phase gradient −2πx/Λ will deflect the RCP component at a negative angle. For the LCP component, the deflection angle is reversed. For the channel at the positive angle, the beam is the superposition of Figure 3(c) gives the calculated polarization distributions. It can be seen from Fig. 3(b) at z = 10 µm, a dark line parallel to the transmission axis of the polarization analyzer is observed with the petal number of 2. The polarization distribution at z = 10 µm in Fig. 3(c) also clearly shows the initial polarization orientation of π/2 and the polarization rotation of 2π per round trip, indicating the azimuthally polarized vector beam with the TC of polarization state of 1 and TE polarization mode. At z = 30 µm, the dark line is perpendicular to the transmission axis of the analyzer, and the polarization distributions show the initial polarization orientation of 0 and the polarization rotation of 2π per round trip, giving the radially polarized vector beam with the TC of polarization state of 1 and TM polarization mode. At z = 50 µm and z = 70 µm, the beam spot becomes larger due to the increased vortex charge of l R L , 3,4 = 2. In these www.nature.com/scientificreports www.nature.com/scientificreports/ two cases, the polarization rotation is 4π per round trip, so that the TC of polarization state equals 2 and the corresponding petal number becomes 4. Besides, the initial polarization orientation is π/2 for z = 50 µm and 0 for z = 70 µm, indicating the polarization mode is TE for z = 50 µm and TM for z = 70 µm, respectively. The evolution of vector vortex beam during the transition process is measured at the wavelength of 633 nm. Figure 3(d) displays the measured beam intensity distributions without polarization analyzer at different propagation distances, and Fig. 3(e) plots the corresponding intensity distributions analyzed with vertical or horizontal linear polarizer, which shows coincidence with the simulation results. Furthermore, the transition process of the first-type vector vortex beam with the spatial variation of polarization mode in the order of (TE → TM → TE → TM) is demonstrated in the supplementary video, which shows the transition processes of both beam intensity and polarization distributions from z = 10 µm to z = 70 µm with the interval of 2 µm. It is shown that for the vector vortex beam, the original linear polarization (LP) state becomes the circular polarization (CP) state, and then changes into another LP state with different polarization orientation compared to the original LP state. The second category of vector vortex beam with the variation of OAM value has L R (z) = {2, 3, 4, 5} and L L (z) = {0, 1, 2, 3}, giving the OAM value varies in the order of n = (L R + L L )/2 = {1, 2, 3, 4} and the TC of polarization m = (L R − L L )/2 = 1 keeps invariant. All these vector vortices are TE polarization modes with the initial polarization orientation Θ = π/2. The simulated beam intensity and polarization distributions are shown in Fig. 4(a-c). It is observed that the beam spot in Fig. 4(a) has the enlarged spot size as the OAM value n increases with respect to the propagation distance z. The petal structures in Fig. 4(b) always display the petal number of 2 at different propagation distances, due to the TC of polarization m = 1. From the polarization distributions in www.nature.com/scientificreports www.nature.com/scientificreports/ Fig. 4(c), it is observed that near the beam center the polarization state is circular polarization because in this case the superimposed Bessel beams have different orders. For the vortex helical phase of exp(ilϕ), the beam field has a Bessel distribution with order l as . For the superposition of two vortex beams with exp(il R ϕ) and exp(il L ϕ) and l R > l L , the Bessel function with order l R exhibits a larger radius than that with order l L , so near the beam center, the intensity of LCP component is stronger than the intensity of RCP component, resulting in almost circular polarization near the beam center. Away from the beam center, the superposition of LCP and RCP components with the similar intensity exhibits nearly linear polarization. The initial polarization orientation is π/2, giving the TE polarization mode. The measured beam intensity distributions are shown in the Fig. 4(d,e). In order to investigate the transition process of OAM value, the vector vortex beam is interfered with a reference spherical wave under RCP, LCP and LP at different propagation distances, as shown in Fig. 5. The vector beam is consisted of RCP and LCP components, and when the reference spherical wave has only RCP or LCP polarization, the interference results will only give the OAM information of the RCP or LCP component. If the reference spherical wave has linear polarization, due to the linear polarization is the superposition of RCP and LCP components, the interference with LP spherical wave will give the information of both RCP and LCP components in the vector beam. Here the RCP interference patterns in Fig. 5(a) only give the information of RCP components, showing a vortex spiral structure with the spiral number equals to l R . The LCP interference patterns in Fig. 5(b) only give the information of LCP components, showing a vortex spiral structure with spiral number equals to l L . The LP interference patterns in Fig. 5(c) contain both the inner part with l L spirals (marked by the white arrow) and the outer part with l R spirals (marked by the yellow arrows), giving the information of both RCP and LCP components in the vector vortex beam. Figure 6(d,e) are the corresponding measured beam intensity distributions, which agree with the simulation prediction. Discussion In summary, the spatial variation of vector vortex beams with arbitrary polarization states and OAM values along the beam propagation has been demonstrated by using single plasmonic metasurfaces with the encoded initial phase profiles determined from the caustic theory. Based on the geometric phase profile induced by nanoslit antenna array, the deflected vector vortex beam is produced by the superposition of RCP and LCP component vortices with different helical phase charges transmitted through the same metasurface. Three different categories of vector vortex beams with the spatial variation of polarization mode, OAM value, and TC of polarization state are studied. Along the beam propagation distance, the detailed beam evolution process is analyzed for intensity profile, polarization distribution and OAM value. The demonstrated arbitrary spatial variation of vector vortex beam will advance many promising applications in spin-to-orbital angular momentum conversion, spin-orbit hybrid entanglement, quantum information processing, particle manipulation and transportation, and optical communication. Methods Simulations. The finite-integration time-domain (FITD) solver of the CST Microwave Studio software is employed to simulate the optical field distribution and the transmission spectrum. In the simulation, periodic boundary conditions are used along both x and y directions in the unit cell. The permittivity of gold is taken from spectroscopic ellipsometry data, and the refractive index of glass substrate is 1.45. The evolution processes of intensity and polarization distributions for vector vortex beams shown in Figs 3, 4 and 6 are calculated by using the Fresnel-Kirchhoff diffraction integral: where Ψ x y ( , ) 0 0 is the complex amplitude distribution located at the z = 0 plane with surface area S and normal direction → n , r′ is the vector between the source point and a point in the z = 0 plane, r is the vector between the point in z = 0 plane and a point in the plane at the propagation distance z, and k = 2π/λ is the wavevector. The polarization distributions are plotted by using the Stokes parameters. Sample fabrication. A 50 nm-thick gold film is deposited on a glass substrate using electron-beam evaporation. Then the nanoslit antenna arrays are milled in the gold film using focused ion beam system (FEI Helios Nanolab 600, 30 kV, 9.7 pA). The metasurface contains 300 × 300 unit cells, and each unit cell has the period of 330 nm and contains a nanoslit with size of 200 nm × 6 nm at a specified orientation angle which is determined by the designed geometric phase profile. Optical characterization. The transmission spectra of the metasurface under circular polarization basis shown in Fig. 1(d) are measured with a collimated broadband Tungsten-Halogen source, where a combination of a linear polarizer and a quarter-wave plate are used to convert the incident light to circularly polarized wave. The light beam is focused normally onto the sample using a 50× objective lens and the transmitted light is collected by another 10× objective lens to a spectrometer (Horiba, iHR 550). The deflected vector vortex beam is measured by a microscope imaging system with a 10× objective lens, a 0.5× tube lens and a CCD camera placed on a translation stage with a tilted angle of α = arctan(λ/Λ) = 21° to the optical axis, as shown in Fig. 2(d).	2019-07-11T14:29:30.477Z	2019-07-10T00:00:00.000	{ "year": 2019, "sha1": "2209cd7482dc0e6e503afa50163464b4e0602ee6", "oa_license": "CCBY", "oa_url": "https://www.nature.com/articles/s41598-019-46433-z.pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "f2b257e1af4a968374b05db75b124a7e074636cf", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics", "Medicine" ] }
215410117	pes2o/s2orc	v3-fos-license	Ecklonia stolonifera Extract Suppresses Lipid Accumulation by Promoting Lipolysis and Adipose Browning in High-Fat Diet-Induced Obese Male Mice. Obesity develops due to an energy imbalance and manifests as the storage of excess triglyceride (TG) in white adipose tissue (WAT). Recent studies have determined that edible natural materials can reduce lipid accumulation and promote browning in WAT. We aimed to determine whether Ecklonia stolonifera extract (ESE) would increase the energy expenditure in high-fat diet (HFD)-induced obese mice and 3T3-L1 cells by upregulating lipolysis and browning. ESE is an edible brown marine alga that belongs to the family Laminariaceae and contains dieckol, a phlorotannin. We report that ESE inhibits body mass gain by regulating the expression of proteins involved in adipogenesis and lipogenesis. In addition, ESE activates protein kinase A (PKA) and increases the expression of lipolytic enzymes including adipose triglyceride lipase (ATGL), phosphorylated hormone-sensitive lipase (p-HSL), and monoacylglycerol lipase (MGL) and also thermogenic genes, such as carnitine palmitoyltransferase 1 (CPT1), PR domain-containing 16 (PRDM16), and uncoupling protein 1 (UCP1). These findings indicate that ESE may represent a promising natural means of preventing obesity and obesity-related metabolic diseases. Introduction Obesity is one of the most prevalent global health problems and predisposes toward metabolic diseases such as type 2 diabetes and hyperlipidemia [1,2]. It is caused by excessive fat accumulation when energy intake exceeds energy expenditure [3]. Therefore, lipid metabolism and energy expenditure may be targeted for the treatment or prevention of obesity-associated diseases. Mammals have two types of adipose tissue that have different functions-white adipose tissue (WAT) and brown adipose tissue (BAT) [4]. WAT stores excess energy as triglyceride (TG) and is largely composed of adipocytes containing large unilocular lipid droplets. By contrast, BAT is morphologically and functionally different from WAT, being largely composed of adipocytes containing small multilocular lipid droplets and large numbers of mitochondria. Brown adipocytes express uncoupling protein 1 (UCP1), which is responsible for uncoupling fatty acid oxidation from ATP synthesis, causing the loss of energy as heat [5]. Thus, BAT is a specialized tissue that may represent a useful target for the prevention of obesity [6]. Furthermore, the possibility of converting WAT adipocytes to BAT-like cells has attracted scientific attention, a process referred to as "browning" [7]. Experimental Animals and Diet The animal studies were in accordance with the criteria outlined in the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Science (NAS) and published by the National Institutes of Health (NIH) and were approved by the Institutional Animal Care and Use Committee of CHA University (IACUC, Approval Number 190193). Male Institute of Cancer Research (ICR) mice (4 weeks of age) purchased from Joong Ah Bio (Suwon, Korea), were housed at 3 mice per cage and maintained in a temperature-and humidity-regulated facility on a 12 h light/dark cycle. After a 1 week period of adaptation, the mice were randomly allocated to four groups (n = 9 per group): a chow diet group (CD, containing 10 kcal% as fat; D12450B, Research Diets, NJ, USA), a high-fat diet group (HFD, containing 60 kcal% as fat; D12492, Research Diets), a low-dose ESE-supplemented group (HFD + ESEL), and a high-dose ESE-supplemented group (HFD + ESEH). The mice were orally administered ESE at doses of 50 mg/kg/day (HFD + ESEL) or 150 mg/kg/day (HFD + ESEH) for 6 weeks. During the experimental period, body mass, dietary intake, fasting blood glucose, and rectal temperature were measured each week. At the end of the experiment, the mice were terminally anesthetized with CO 2 after fasting for 12 h, and then blood and tissue samples were collected. Body Mass and Food and Water Intake Measurements During the experimental period, the mice were weighed weekly at 09:00 am using an analytical balance. The food and water intake were calculated each week as the differences between the initial quantities supplied and the amounts of food and water remaining at the end of the week. Fasting Blood Glucose Measurement Fasting blood glucose was measured in blood obtained from a tail vein after withholding food for 12 h using an Accu-Check Blood Glucose Meter (Roche, Seoul, Korea). Rectal Temperature Measurement The rectal temperatures of the mice were measured weekly using a Testo 925 Type Thermometer (Testo, Lenzkirch, Germany). Biochemical Analysis Blood samples were collected at 09:00 am by cardiac puncture under terminal anesthesia. After the blood clotted for 30 min, the serum was separated by centrifugation at 4 • C and 3000× g for 20 min. The serum concentrations of triglycerides, total cholesterol, low-density lipoprotein (LDL)-cholesterol, and high-density lipoprotein (HDL)-cholesterol, and the activities of aspartate transaminase (AST) and alanine transaminase (ALT), were determined using colorimetric assay kits (Roche, Basel, Switzerland). Histological Analysis Subcutaneous WAT was fixed in 4% paraformaldehyde and embedded in paraffin. Sections were prepared and stained with hematoxylin and eosin (H&E). Photomicrographs were obtained using a Nikon E600 microscope (Nikon, Tokyo, Japan). Cell Viability To determine the appropriate concentrations of ESE and dieckol to be used in further investigations, cell viability testing was performed using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT). 3T3-L1 preadipocytes were seeded (5 × 10 2 cells/well) in 96-well plates and incubated overnight. Stock solutions of ESE and dieckol were prepared in DMSO. Then, the cells were treated with ESE (0, 12, 25, 100, or 200 µg/mL) or dieckol (0, 12, 25, 50, or 100 µM), in comparison to the control cells that had no treatment, for 24 h. After this, 20 µL of a 5 mg/mL MTT solution (yellow color) was added to each well, and the cells were incubated for a further 3 h, and this procedure led to the formation of formazan crystals (purple color). After removing the MTT-containing medium, 100 µL DMSO was added to elute the formazan crystals. Finally, the intensity of the dissolved formazan crystals that are associated with the enzyme activity and the number of viable cells was quantified at 570 nm on a Biotek-enzyme-linked immunosorbent assay (ELISA) reader (BioTek, Winooski, VT, USA). Oil Red O Staining After 8 days of differentiation, fully differentiated 3T3-L1 adipocytes were fixed in 4% formaldehyde for 1 h at room temperature and washed twice with 60% isopropanol. The fixed cells were stained with Oil red O (ORO) solution for 30 min and then washed with distilled water. After drying, the stained cells were imaged and the stain was eluted using 100% isopropanol. The absorbance was measured at 490 nm. To analyze hepatic lipid accumulation, cryostat sections of the liver were stained with ORO solution, and photomicrographs were obtained. Statistical Analysis Data are expressed as means and standard deviations (SDs), and were analyzed by one-way ANOVA, followed by Tukey's post-hoc test (IBM SPSS Statistics Version 20.0, Chicago, IL, USA). Statistical significance was accepted when p < 0.05. ESE Prevents the Development of Obesity in HFD-Induced Obese Mice To determine whether ESE has an anti-obesity effect, we fed mice CD, HFD, or HFD supplemented with 50 mg/kg/day (HFD + ESEL) or 150 mg/kg/day (HFD + ESEH) ESE for 6 weeks. As shown in Cells 2020, 9, 871 5 of 13 Figure 1A,B, the body mass of the mice fed an HFD was much higher than that of the mice fed a CD. However, the body masses of the mice fed the HFD and co-administered ESE were dose-dependently reduced. The mice fed the HFD gained 14.1 ± 1.3 g over 6 weeks, whereas the gains in the ESE-treated HFD-fed mice were 9.1 ± 1.9 g for ESEL and 7.3 ± 1.5 g for ESEH. The masses of the WAT, liver, kidney, lung, and spleen were also measured at the end of the experiment, and this showed that ESE-treated HFD-fed mice had smaller WAT depots than the HFD-fed mice, but there were no differences between the groups in the masses of the other organs ( Figure 1C,D). In addition, food and water intake were not affected by ESE treatment ( Figure 1E). supplemented with 50 mg/kg/day (HFD + ESEL) or 150 mg/kg/day (HFD + ESEH) ESE for 6 weeks. As shown in Figure 1A and B, the body mass of the mice fed an HFD was much higher than that of the mice fed a CD. However, the body masses of the mice fed the HFD and co-administered ESE were dose-dependently reduced. The mice fed the HFD gained 14.1 ± 1.3 g over 6 weeks, whereas the gains in the ESE-treated HFD-fed mice were 9.1 ± 1.9 g for ESEL and 7.3 ± 1.5 g for ESEH. The masses of the WAT, liver, kidney, lung, and spleen were also measured at the end of the experiment, and this showed that ESE-treated HFD-fed mice had smaller WAT depots than the HFD-fed mice, but there were no differences between the groups in the masses of the other organs ( Figure 1C, D). In addition, food and water intake were not affected by ESE treatment ( Figure 1E). As shown in Figure 1F, the fasting blood glucose in the HFD group (133.6 ± 3.7 mg/dL) was higher than that in the CD group (85.8 ± 4.8 mg/dL). However, the fasting glucose was significantly lower in both the ESEL (104.4 ± 5.2 mg/dL) and ESEH (93.4 ± 7.7 mg/dL) groups. In addition, the serum concentrations of triglycerides, total cholesterol, and LDL-cholesterol were lower, and that of HDL-cholesterol was higher, in ESE-treated HFD-fed mice than in mice fed the HFD alone (Table 1). These results indicate that ESE not only suppresses WAT mass gain, but also regulates fasting blood glucose and serum lipid levels. As shown in Figure 1F, the fasting blood glucose in the HFD group (133.6 ± 3.7 mg/dL) was higher than that in the CD group (85.8 ± 4.8 mg/dL). However, the fasting glucose was significantly lower in both the ESEL (104.4 ± 5.2 mg/dL) and ESEH (93.4 ± 7.7 mg/dL) groups. In addition, the serum concentrations of triglycerides, total cholesterol, and LDL-cholesterol were lower, and that of HDL-cholesterol was higher, in ESE-treated HFD-fed mice than in mice fed the HFD alone (Table 1). These results indicate that ESE not only suppresses WAT mass gain, but also regulates fasting blood glucose and serum lipid levels. ESE Inhibits Lipid Accumulation in the WAT of HFD-Induced Obese Mice and 3T3-L1 Cells To evaluate the effects of ESE and dieckol on lipid accumulation, we measured WAT and 3T3-L1 adipocyte size. As shown in Figure 2A, the WAT adipocytes were larger in HFD-fed mice than in mice fed a CD. However, mice administered ESE had dose-dependently smaller adipocytes than those fed HFD only, such that the cells were similar in size to those in the CD group. Before evaluating the effects in 3T3-L1 cells, we first assessed the cytotoxicity of ESE and dieckol in these cells. As shown in Figure 2B, 200 µg/mL of ESE and 100 µM of dieckol were cytotoxic. Therefore, concentrations of ESE (12, 25, 50, or 100 µg/mL) and dieckol (12,25, or 50 µM) below these were used in further experiments. To assess lipid accumulation, we fully differentiated 3T3-L1 cells and stained them using ORO solution. Examination of these cells showed that ESE and dieckol reduced the number of lipid droplets ( Figure 2C) and lipid accumulation in a dose-dependent manner ( Figure 2D). Thus, ESE suppresses the HFD-induced increase in adipocyte size in WAT and lipid droplet accumulation in differentiating 3T3-L1 cells. ESE Inhibits Lipid Accumulation in the WAT of HFD-Induced Obese Mice and 3T3-L1 Cells To evaluate the effects of ESE and dieckol on lipid accumulation, we measured WAT and 3T3-L1 adipocyte size. As shown in Figure 2A, the WAT adipocytes were larger in HFD-fed mice than in mice fed a CD. However, mice administered ESE had dose-dependently smaller adipocytes than those fed HFD only, such that the cells were similar in size to those in the CD group. Before evaluating the effects in 3T3-L1 cells, we first assessed the cytotoxicity of ESE and dieckol in these cells. As shown in Figure 2B, 200 μg/mL of ESE and 100 μM of dieckol were cytotoxic. Therefore, concentrations of ESE (12, 25, 50, or 100 μg/mL) and dieckol (12,25, or 50 μM) below these were used in further experiments. To assess lipid accumulation, we fully differentiated 3T3-L1 cells and stained them using ORO solution. Examination of these cells showed that ESE and dieckol reduced the number of lipid droplets ( Figure 2C) and lipid accumulation in a dose-dependent manner ( Figure 2D). Thus, ESE suppresses the HFD-induced increase in adipocyte size in WAT and lipid droplet accumulation in differentiating 3T3-L1 cells. ESE Reduces Adipogenesis and Lipogenesis in the WAT of HFD-Induced Obese Mice and 3T3-L1 Cells To determine the mechanisms of the effects of ESE on adipocyte size and lipid accumulation, we performed western blot analysis. As shown in Figure 3A, HFD-fed mice had higher expression levels of proteins involved in adipogenesis (C/EBPα, PPARγ, and FABP4) than CD-fed mice. However, ESE-treated HFD-fed mice showed a lower expression of these proteins in WAT. Similarly, in 3T3-L1 cells, ESE and dieckol significantly reduced the expression levels in a dose-dependent manner ( Figure 3B). In addition, ESE reduced the expression of proteins involved in lipogenesis. As shown in Figure 3C, WAT from the HFD group had higher expression levels of lipogenic proteins (LPAATθ, lipin1, and DGAT1) than that from the CD group, but ESE treatment reduced these. Furthermore, ESE and dieckol significantly reduced the expression of lipogenic proteins (SREBP1, LPAATθ, lipin1, and DGAT1) in 3T3-L1 cells ( Figure 3D). Taken together, these results suggest that ESE inhibits adipogenesis and lipogenesis in WAT and 3T3-L1 cells. 3B). In addition, ESE reduced the expression of proteins involved in lipogenesis. As shown in Figure 3C, WAT from the HFD group had higher expression levels of lipogenic proteins (LPAATθ, lipin1, and DGAT1) than that from the CD group, but ESE treatment reduced these. Furthermore, ESE and dieckol significantly reduced the expression of lipogenic proteins (SREBP1, LPAATθ, lipin1, and DGAT1) in 3T3-L1 cells ( Figure 3D). Taken together, these results suggest that ESE inhibits adipogenesis and lipogenesis in WAT and 3T3-L1 cells. ESE Stimulates Lipolysis in the WAT of HFD-Induced Obese Mice and 3T3-L1 Cells Lipolysis involves the sequential activity of lipolytic enzymes (ATGL, p-HSL, and MGL), which is induced following PKA activation [26]. Therefore, we determined whether ESE increases lipolysis by upregulating the expression of lipolytic enzymes (ATGL, p-HSL, and MGL) using western blot analysis. As shown in Figure 4A, ESE administration increased the phosphorylation of PKA and the expression levels of the downstream proteins (ATGL, p-HSL, and MGL) fourfold in WAT. Moreover, ESE significantly increased the immunofluorescence staining intensity of PKA and p-HSL in WAT sections, as shown in Figure 4B. Consistent with this, dose-dependent significant increases in the protein expression of lipolytic enzymes were identified in 3T3-L1 cells treated with ESE or dieckol ( Figure 4C). These results suggest that ESE increases lipolysis in adipocytes by inducing the activation of PKA and increasing the expression of lipolytic enzymes. ESE significantly increased the immunofluorescence staining intensity of PKA and p-HSL in WAT sections, as shown in Figure 4B. Consistent with this, dose-dependent significant increases in the protein expression of lipolytic enzymes were identified in 3T3-L1 cells treated with ESE or dieckol ( Figure 4C). These results suggest that ESE increases lipolysis in adipocytes by inducing the activation of PKA and increasing the expression of lipolytic enzymes. ESE Promotes Browning in the WAT of HFD-Induced Obese Mice and 3T3-L1 Cells BAT uses FFAs released by lipolysis, and CPT1 is essential for the delivery of FFAs into mitochondria [27]. BAT and browned WAT adipocytes are characterized by UCP1 expression and consume energy by oxidizing FFAs inefficiently in mitochondria; therefore, the change from a WATto a BAT-like phenotype is associated with an increase in body temperature [28]. Therefore, we determined the effects of ESE on heat generation by measuring the temperatures of the mice. As shown in Figure 5A, rectal temperature was unaffected by which treatment group the mice were in for the first 2 weeks. However, at the end of the experimental period, the ESE-treated groups had higher rectal temperatures (ESEL: 37.38 ± 0.19°C; ESEH: 37.86 ± 0.11 °C) than the other groups (CD: 36.20 ± 0.07 °C; HFD: 36.32 ± 0.11°C). We then performed western blot analysis to measure the expression of the thermogenic proteins CPT1, PRDM16, and UCP1. As shown in Figure 5B, HFD-fed mice expressed lower levels of these proteins than CD-fed mice, but ESE administration increased ESE Promotes Browning in the WAT of HFD-Induced Obese Mice and 3T3-L1 Cells BAT uses FFAs released by lipolysis, and CPT1 is essential for the delivery of FFAs into mitochondria [27]. BAT and browned WAT adipocytes are characterized by UCP1 expression and consume energy by oxidizing FFAs inefficiently in mitochondria; therefore, the change from a WATto a BAT-like phenotype is associated with an increase in body temperature [28]. Therefore, we determined the effects of ESE on heat generation by measuring the temperatures of the mice. As shown in Figure 5A, rectal temperature was unaffected by which treatment group the mice were in for the first 2 weeks. However, at the end of the experimental period, the ESE-treated groups had higher rectal temperatures (ESEL: 37.38 ± 0.19 • C; ESEH: 37.86 ± 0.11 • C) than the other groups (CD: 36.20 ± 0.07 • C; HFD: 36.32 ± 0.11 • C). We then performed western blot analysis to measure the expression of the thermogenic proteins CPT1, PRDM16, and UCP1. As shown in Figure 5B, HFD-fed mice expressed lower levels of these proteins than CD-fed mice, but ESE administration increased the expression levels of CPT1, PRDM16, and UCP1. Similarly, CPT1 and UCP1 immunoreactivity were higher in the ESE-treated groups, as shown in Figure 5C. In addition, ESE and dieckol dose-dependently increased the expression of the same proteins in 3T3-L1 cells ( Figure 5D). These results suggest that ESE promotes browning by increasing the expression of CPT1 and UCP1 in mitochondria, resulting in a greater loss of energy as heat. Cells 2020, 9, x 9 of 14 the expression levels of CPT1, PRDM16, and UCP1. Similarly, CPT1 and UCP1 immunoreactivity were higher in the ESE-treated groups, as shown in Figure 5C. In addition, ESE and dieckol dosedependently increased the expression of the same proteins in 3T3-L1 cells ( Figure 5D). These results suggest that ESE promotes browning by increasing the expression of CPT1 and UCP1 in mitochondria, resulting in a greater loss of energy as heat. ESE Reduces Hepatic Lipid Accumulation in the Livers of HFD-Induced Obese Mice HFD-induced obesity is a risk factor for non-alcoholic fatty liver disease (NAFLD) because it is associated with lipid accumulation in the liver [29]. Therefore, we next determined whether ESE affects hepatic lipid accumulation in HFD-induced obese mice. As shown in Figure 6A, hepatic lipid staining by ORO was dose-dependently reduced by ESE in HFD-fed mice. In addition, the serum AST and ALT activities, which are widely used indicators of liver damage [30], were higher in the HFD than in the CD group, but reduced by ESE treatment ( Figure 6B). Furthermore, we performed western blot analysis to measure the expression levels of the key hepatic lipogenic proteins SREBP1, lipin1, and DGAT1 [31]. As shown in Figure 6C, HFD-fed mice expressed higher levels of these hepatic lipogenic proteins in the liver, but ESE treatment significantly reduced these. Taken together, these results suggest that ESE limits hepatic lipid accumulation in HFD-fed mice. ESE Reduces Hepatic Lipid Accumulation in the Livers of HFD-Induced Obese Mice HFD-induced obesity is a risk factor for non-alcoholic fatty liver disease (NAFLD) because it is associated with lipid accumulation in the liver [29]. Therefore, we next determined whether ESE affects hepatic lipid accumulation in HFD-induced obese mice. As shown in Figure 6A, hepatic lipid staining by ORO was dose-dependently reduced by ESE in HFD-fed mice. In addition, the serum AST and ALT activities, which are widely used indicators of liver damage [30], were higher in the HFD than in the CD group, but reduced by ESE treatment ( Figure 6B). Furthermore, we performed western blot analysis to measure the expression levels of the key hepatic lipogenic proteins SREBP1, lipin1, and DGAT1 [31]. As shown in Figure 6C, HFD-fed mice expressed higher levels of these hepatic lipogenic proteins in the liver, but ESE treatment significantly reduced these. Taken together, these results suggest that ESE limits hepatic lipid accumulation in HFD-fed mice. Discussion Obesity develops as a result of excessive fat accumulation, which occurs because of an imbalance in energy intake and expenditure. Excess energy is stored as TGs in WAT, which expands through both cellular hyperplasia and hypertrophy [32]. Targeting energy expenditure represents an interesting concept as a means of combating obesity. The discovery of adipose browning and its Discussion Obesity develops as a result of excessive fat accumulation, which occurs because of an imbalance in energy intake and expenditure. Excess energy is stored as TGs in WAT, which expands through both cellular hyperplasia and hypertrophy [32]. Targeting energy expenditure represents an interesting concept as a means of combating obesity. The discovery of adipose browning and its associated lipid metabolism has generated interest in the search for natural dietary compounds that could be a therapeutic strategy for prevent obesity and its associated metabolic syndrome [33,34]. In the present study, we evaluated two mechanisms whereby ESE might prevent obesity, suppress fat accumulation, and increase energy expenditure-the upregulation of lipolysis and the induction of browning, using HFD-induced obese mice and 3T3-L1 cells, respectively. Our data support the effects of ESE in regulating energy expenditure as a potential new anti-obesity agent. ESE, one of the marine algae, is a natural dietary component that has been used as a food additive or flavoring material in many countries [35]. Marine algae are rich in minerals, dietary fibers, and bioactive compounds, such as polyphenols, which have potential health benefits [36]. Recently, brown marine algae have been identified to be a potential source of anti-obesity substances [37][38][39][40]. It has previously been shown that Ecklonia stolonifera has an anti-adipogenic activity in 3T3-L1 adipocytes because it downregulates C/EBPα and PPARγ expression [41,42]. However, the effects of ESE on adipocyte lipolysis and browning has not been investigated. Furthermore, there was a need for more animal studies to assess the efficacy of ESE as an anti-obesity agent. Therefore, in the present study, we assessed the anti-obesity effects of ESE using both in vivo and in vitro models. The overconsumption of HFD induces body mass gain secondary to the accumulation of WAT [43]. As expected, HFD-fed mice became obese and had higher WAT mass than mice fed a CD. However, the administration of ESE to HFD-fed mice for 6 weeks significantly ameliorated the body mass gain and reduced the size of WAT adipocytes. HFD-induced obesity is also associated with higher concentrations of fasting blood glucose, and serum triglycerides and cholesterol [44]. In the present study, ESE ameliorated the increase in fasting blood glucose, serum triglyceride, total cholesterol, and LDL-cholesterol concentrations, and the reduction in HDL-cholesterol concentration. We also measured the expression of adipogenic proteins in the WAT of each group using western blot analysis. We found that ESE reduced the expression of C/EBPα, PPARγ, and FABP4, which are important for the late stages of preadipocyte differentiation, in HFD-fed mice. Furthermore, ESE reduced the expression of the lipogenic proteins SREBP1, LPAATθ, lipin1, and DGAT1, implying that it may reduce lipid synthesis. The expression of these proteins was also reduced to similar extents by treatment with either ESE or dieckol in 3T3-L1 cells. Therefore, both the in vivo and in vitro data suggest that ESE may prevent the development of HFD-induced obesity by regulating the expression of adipogenic and lipogenic factors. HFD-induced obesity is associated with a number of metabolic diseases, including NAFLD, which is characterized by hepatic lipogenesis and progressive steatosis [45]. Hepatic lipid accumulation results from an imbalance between lipid accumulation and disposal, and the key metabolic pathway of lipogenesis are regulated by hepatic lipogenic proteins including SREBP1, lipin1, and DGAT1 [46][47][48]. In the present study, ORO staining showed that the administration of ESE reduces hepatic lipid accumulation and the expression of these lipogenic factors in the liver of HFD-fed mice, which implies that ESE inhibits hepatic lipogenesis. Lipolysis is the hydrolysis of TGs stored in intracellular lipid droplets. In adipocytes, the hydrolysis of TGs provides fuel for FAO, and this catabolic process is activated by PKA and involves several lipolytic enzymes (ATGL, HSL, and MGL) [49]. Specifically, a recent study showed that the phosphorylation of HSL is induced following PKA activation, which releases FFAs that can be oxidized in mitochondria [50,51]. In the present study, we found that ESE treatment increased the expression of p-PKA, ATGL, p-HSL, and MGL in HFD-fed mice. Consistent with this, there was significantly higher PKA and p-HSL immunofluorescence in the WAT of ESE-treated groups than in the HFD-only group. In addition, ESE increased the expression of CPT1, an enzyme that facilitates the transfer of FFAs into mitochondria and is rate limiting for FAO. Likewise, ESE and dieckol increased the expression of these lipolytic enzymes and CPT1 in 3T3-L1 cells. Therefore, ESE may increase energy consumption by increasing the expression of these proteins in WAT. UCP1 is expressed in the mitochondrial inner membrane and is responsible for the energy-wasting phenotype of BAT. In addition, it has been reported that BAT-like trans-differentiated subcutaneous WAT expresses high levels of UCP1 and shows greater mitochondrial activity [52,53]. In the present study, western blotting showed that ESE increased the expression of thermogenic genes (UCP1 and PRDM16) in WAT from HFD-fed mice and 3T3-L1 cells in a dose-dependent manner. Consistent with this, there were higher intensities of UCP1 and CPT1 immunostaining in WAT from HFD-fed mice. Furthermore, the rectal temperatures of ESE-treated mice were higher than those of the other groups. These results suggest that ESE may increase energy expenditure by increasing the expression of thermogenic genes in the WAT of HFD-induced obese mice. The thermogenic effect of ESE in BAT has not yet been reported, thus further studies are required to understand the details of energy expenditure as heat by UCP1 in BAT, which is morphologically and functionally different from WAT. In conclusion, we have shown that ESE inhibits lipid accumulation in WAT by downregulating adipogenesis and lipogenesis. In addition, we have provided evidence that ESE increases lipolysis and FAO, promotes WAT browning, and increases the loss of energy as heat. These findings suggest that ESE has important effects on lipid metabolism and represents a potential dietary means of treating obesity and the related metabolic diseases.	2020-04-08T19:07:49.220Z	2020-04-01T00:00:00.000	{ "year": 2020, "sha1": "9274d2ce140aa9a4ae579d70f94d237c0073f36a", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2073-4409/9/4/871/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "f22eb66e508a4d40aabf4b1630f4fff261b51dc4", "s2fieldsofstudy": [ "Medicine", "Environmental Science", "Biology" ], "extfieldsofstudy": [ "Medicine", "Chemistry" ] }
246679970	pes2o/s2orc	v3-fos-license	Relaxation of the parameter independence assumption in the `bootComb` R package Background. The bootComb R package allows researchers to derive confidence intervals with correct target coverage for arbitrary combinations of arbitrary numbers of independently estimated parameters. Previous versions (<1.1.0) of bootComb used independent bootstrap sampling and required that the parameters themselves are independent - an unrealistic assumption in some real-world applications. Findings. Using Gaussian copulas to define the dependence between parameters, the bootComb package has been extended to allow for dependent parameters. Implications. The updated bootComb package can now handle cases of dependent parameters, with users specifying a correlation matrix defining the dependence structure. While in practice it may be difficult to know the exact dependence structure between parameters, `bootComb` allows running sensitivity analyses to assess the impact of parameter dependence on the resulting confidence interval for the combined parameter. Availability. bootComb is available from the Comprehensive R Archive Network (https://CRAN.R-project.org/package=bootComb). Introduction The bootcomb R package Henrion (2021) was recently published. This package for the statistical computation environment R (R Core Team, 2021) allows researchers to derive confidence intervals with correct coverage for combinations of independently estimated parameters. Important applications include adjusting a prevalence for estimated test sensitivity and specificity (e.g. Mandolo et al. (2021)) or combining conditional prevalence estimates (e.g. Stockdale et al. (2020)). Briefly, for each of the input parameters, bootComb finds a best-fit parametric distribution based on the confidence interval for that parameter estimate. bootComb then uses the parametric bootstrap to sample many sets of parameter estimates from these best-fit distributions and computes the corresponding combined parameter estimate for each set. This builds up an empirical distribution of parameter estimates for the combined parameter. Finally, bootComb uses either the percentile or the highest density interval method to derive a confidence interval for the combined parameter estimate. Full details of the algorithm are given in Henrion (2021). A key point of the algorithm is that the best-fit distributions for the different parameters are sampled from independently. This requires the parameters to be independent. This may not be a realistic assumption in some real-world applications. While for most practical applications the input parameters are typically estimated from independent experiments (otherwise the combined parameter could be directly estimated), the parameters themselves may not be independent. This is for instance the case when adjusting a prevalence for the diagnostic test's sensitivity and specificity. The latter two parameters are not independent: higher sensitivity can be achieved by lowering specificity and vice versa. If the experiments estimating these parameters are sufficiently large, then the violation of the assumption of parameter independence may only have negligible impact on the resulting confidence interval for the combined parameter. However, for the sake of general applicability and to allow running sensitivity analyses, the author felt it was beneficial to extend bootComb to handle dependent parameters. Methods Copulas are multivariate distribution functions where the marginal probability distribution of each variable is the uniform distribution on the interval [0, 1]. Copulas allow to specify the intercorrelation between random variables. An important probability theory result, Sklar's Theorm (Sklar, 1959), states that any multivariate probability distribution can be expressed in terms of its univariate marginal distributions and a copula defining the dependence between the variables. Mathematically, let X 1 , X 2 . . . , X d be d random variables and define U i = F i (X i ), i = 1, . . . , d. Then the copula C of (X 1 , . . . , X d ) is defined as the joint cumulative distribution function of (U 1 , . . . , U d ): . . , d are continuous. Then, via the probability integral transform (Angus, 1994), the random vector (U 1 , U 2 , . . . , U d ) has marginals that are bootComb makes use of the fact that the above can be reversed: given a sample (u 1 , . . . , u d ), a sample for . The inverse functions F −1 i (u) will be defined if the marginals F i (x) are continuous. For the use of bootComb, where users input confidence intervals for an estimated numeric parameter, this will always be the case. bootComb will proceed as follows to generate samples from a multivariate distribution of d dependent variables: • Estimate best-fit distributions F 1 , . . . , F d for each of the d parameters X 1 , . . . , X d given the lower and upper limits of the estimated confidence intervals for each parameter. • Since the marginals of this normal distribution are all N (0, 1), compute u i = Φ(z i ) where Φ is the cumulative distribution function of the standard normal. The resulting vector (x 1 , . . . , x d ) will be a sample from the multivariate distribution of (X 1 , . . . , X d ). Note that the dependence structure was completely specified through the covariance matrix Σ (since the covariances are assumed to be 1, this really is a correlation matrix) and marginal distributions for each parameter were specified by F i , i = 1, . . . , d. Results We repeat the 2 examples from Henrion (2021) here, but look at the effect of specifying a dependence between the input parameters. HDV prevalence in the general population With an application to hepatitis D and B viruses (HDV and HBV respectively) from Stockdale et al. (2020), Henrion (2021) showed how to use bootComb to obtain a valid confidence interval forp aHDV , the prevalence of HDV specific immunoglobulin G antibodies (anti-HDV) in the general population. If, however, the 2 input prevalences are not independent, e.g. if anti-HDV is more common among people with presence of HBsAg the higher the population prevalence of HBsAg is, then that assumption of independence would not hold. We can investigate how strong an effect dependence of the parameters can have on the resulting confidence estimate. For example, let's run the same example using bootComb with specifying the following covariance matrix for the bivariate normal copula: This yields the 95% confidence interval (0.10%, 0.26%), a slightly wider interval -which makes sense, as the positive correlation means it is more likely for pairs of bootstrapped input parameters to be both near the upper (respectively lower) end of their confidence intervals. For this particular application, a dependence between both prevalence parameters,p HBsAg andp aHDV \|HBsAg , is unlikely and we have therefore not considered this example any further. Henrion (2021) gave an example of adjusting an estimated SARS-CoV-2 seroprevalence for the estimated sensitivity and specificity of the test assay. Specifically: However in this case, the assumption of independence is not fully realistic: there is a trade-off between sensitivity and specificity of the test assay, and as such one would expect a negative dependence between the two parameters: sensitivity can be increased at the cost of decreased specificity and vice versa. Assuming that the sensitivity and specificity are negatively correlated with the copula correlation parameter ρ = −0.5 between these two parameters, using the extension of bootComb we can now account for the dependence of the parameters: The reported confidence interval is now (3.8%, 19.4%) -marginally wider than when the dependence was ignored. If we additionally specify returnBootVals=TRUE in the function call, we can extract and plot the sampled pairs of sensitivity and specificity values to check the dependence structure. This is shown on Figure @ref(fig:Fig1): as the correlation parameter ρ in the copula between the sensitivity and specificity is decreased from 0 to -1, the dependence between both parameters becomes more and more pronounced as one would expect. This shows that a simple correlation matrix specified for the Gaussian copula results in this case in a non-trivial dependence structure between two beta-distributed variables, respecting the specified marginal distributions. We can also visualise the effect on the estimated confidence interval, as shown on Figure Figure @ref(fig:Fig2). We can see that in this case, with a negative correlation, the width of the CI increases at the correlation becomes stronger. However, looking at the scale of the y-axis we see that this is just a marginal effect. Conclusions The R package bootComb has been extended and, using Gaussian copulas, it can now handle the case of dependent input parameters. For many applications, the effect of dependence between the parameters will be marginal or even negligible. However, the package now allows users to do sensitivity analyses to assess the effects of a miss-specified dependence structure between the parameters that are being combined. Data Availability Statement All data to support this work are contained within the article. The software package itself is available from https://cran.r-project.org/package=bootComb. Conflicts of interest Author Marc Y. R. Henrion declares none.	2022-02-10T06:47:57.393Z	2022-02-09T00:00:00.000	{ "year": 2022, "sha1": "4f4281b052e5fb77d354d38c55aa43973895d22e", "oa_license": "CCBYNCND", "oa_url": "https://www.cambridge.org/core/services/aop-cambridge-core/content/view/2EE5834C49C7668A3E7C70E2537925A9/S2516712X22000132a.pdf/div-class-title-relaxation-of-the-parameter-independence-assumption-in-the-bootcomb-r-package-div.pdf", "oa_status": "GOLD", "pdf_src": "Arxiv", "pdf_hash": "4f4281b052e5fb77d354d38c55aa43973895d22e", "s2fieldsofstudy": [ "Mathematics" ], "extfieldsofstudy": [ "Mathematics" ] }
229684782	pes2o/s2orc	v3-fos-license	Synthesis and Characterization of Store-Operated Calcium Entry Inhibitors Active in the Submicromolar Range The store-operated calcium entry, better known as SOCE, forms the main Ca2+ influx pathway in non-excitable cells, especially in leukocytes, where it is required for cell activation and the immune response. During the past decades, several inhibitors were developed, but they lack specificity or efficacy. From the non-specific SOCE inhibitor 2-aminoethyl diphenylborinate (2-APB), we synthetized 16 new analogues by replacing/modifying the phenyl groups. Among them, our compound P11 showed the best inhibitory capacity with a Ki ≈ 75 nM. Furthermore, below 1 µM, P11 was devoid of any inhibitory activity on the two other main cellular targets of 2-APB, the IP3 receptors, and the SERCA pumps. Interestingly, Jurkat T cells secrete interleukin-2 under phytohemagglutinin stimulation but undergo cell death and stop IL-2 synthesis when stimulated in the presence of increasing P11 concentrations. Thus, P11 could represent the first member of a new and potent family of immunosuppressors. Introduction The control of Ca 2+ homeostasis is a challenging process for the cell. Indeed, many cellular functions, including contraction, differentiation, and proliferation, are regulated by an increase of the intracellular Ca 2+ ion concentration [1]. These increases are commonly due to the release of Ca 2+ ions from internal stores (mainly the endoplasmic reticulum (ER)) and to an influx of Ca 2+ ions from the extracellular space. The diversity of the Ca 2+ -release and Ca 2+ -influx mechanisms in various cell types is responsible for the variety of Ca 2+ responses occurring after cell stimulation. In leukocytes in general, and in lymphocytes in particular, cell stimulation induces the synthesis of inositol 1,4,5-trisphosphate (IP 3 ) allowing the release of Ca 2+ ions out of the ER through the IP 3 receptor (IP 3 R). This depletion activates a Ca 2+ influx known as store-operated calcium entry (SOCE) [2]. During the last 15 years, the two main proteins responsible for SOCE have been unveiled: STIM1, the Ca 2+ -sensor protein localized in the ER membrane, and Orai1, the pore-forming protein localized in the plasma membrane. The importance of Orai1 for lymphocyte function appears from the fact that the protein was identified after the discovery of a point mutation creating a non-functional channel, inducing a severe combined immunodeficiency [3]. Thus, even if Orai1 channels are present and the leukocyte repertoire unchanged, the immune system remains deficient. For more than 30 years, several compounds able to block SOCE have been characterized and developed [4][5][6]. Indeed, a SOCE inhibitor could represent a new way to impair the activation of immune cells and could be of interest in the treatment of chronic inflammatory diseases and leukemias/lymphomas as well as to impair graft rejection or failure. Among all these molecules, 2-aminoethoxy diphenylborinate (2-APB) has emerged as an interesting compound due to its special properties and the ease to modify its basal structure. Furthermore, 2-APB has intriguing properties: at a concentration between 1 and 5 µM, 2-APB potentiates the leukocyte SOCE but is a full blocker at concentrations above 20-30 µM [7,8]. We previously characterized several 2-APB analogs having either this bimodal function or displaying only inhibiting or only potentiating effects [7,9]. The group of Mikoshiba on the other hand has developed tandem 2-APB molecules with very potent SOCE inhibition capacity, called DPB162-AE and DPB163-AE [6], though DPB163-AE at low concentrations activates SOCE [6] while DPB162-AE also affects ER Ca 2+ -store content in various cell types [10]. Very recently, asymmetrical 2-APB analogues obtained after the exchange of one of the two phenyl groups, or the replacement of the ethanolamine, were tested on MDA-MB-231 cells, which exert an Orai1-dependent SOCE [11]. However, none of these molecules were able to totally block the MDA-MB-231 cells SOCE at 10 µM, and some could induce Ca 2+ release from internal stores, demonstrating a lack of specificity [11]. Importantly, 2-APB also has a similar biphasic effect on the IP 3 R [12,13] but also inhibits the ER Ca 2+ pumps (SERCA [14,15]) and can directly open Orai3 channels, allowing an Orai3-dependent SOCE [16][17][18]. These data all underscore that one limit to the use of 2-APB and its analogs is their lack of specificity, implying that any new interesting molecules derived from them must be tested for its effects on other potential targets like IP 3 R and SERCA. In this new study, we designed and synthesized 2-APB together with 16 new, non-commercially available 2-APB analogs. In an attempt to get more potent SOCE inhibitors, only the phenyl rings were slightly modified or replaced ( Figure 1). Indeed, we previously showed that the presence of the ethanolamine group is associated with the inhibitory capacity of the molecule and that groups larger than phenyl rings impair the potentiation ability [7]. Among these molecules, our interest was particularly drawn to the molecule we called P11 as it is more specific for the SOCE over IP 3 R and SERCA, and inhibits the SOCE with a Ki << 1 µM. This most potent SOCE inhibitor induces apoptosis of phytohemagglutinin (PHA)-stimulated Jurkat cells and impairs the synthesis of interleukin-2, clearly blocking T cell activation. This 2-APB analog P11 could, therefore, represent a new important step in the creation of more potent and specific SOCE inhibitors. Results In our previous work, we used commercially available 2-APB analogs containing various types of modifications of the phenyl groups of the molecule. Replacement of the two phenyl groups by larger groups or the impairment of the free rotation of the phenyl groups by the creation of a carbon bridge between them totally impaired the potentiation ability and gave rise to several interesting, inhibitory molecules [7]. On the contrary, decreasing the size of the two aryl groups only shifted the potentiation/inhibition capacity to higher concentrations [7]. In an attempt to identify and characterize new, stronger SOCE inhibitors, we decided to perform our synthesis of 2-APB analogs by increasing the size of the two aryl groups by adding small groups like methyl or methoxyl groups, or larger ones like other aromatic rings (molecules are depicted in Figure 1). Using a previously described methodology, these compounds were prepared in a one-pot procedure using air-stable amine borane complexes and Grignard reagents formed under Barbier conditions. The amine borane complexes have been shown to provide easy access to borinic acids and their derivatives. Indeed, by direct addition on the diisopropylamine borane, the diarylaminoborane can rapidly be obtained without undergoing a third addition leading to the triarylborane. In all cases, the diarylaminoborane was readily transformed into the corresponding 2-aminoethylborinate, via simple methanolysis followed by transesterification using 2-aminoethanol. To visualize SOCE, we performed a classical protocol: Jurkat cells in Ca 2+ -free HBS medium are stimulated by 1 µM thapsigargin (TG, black arrow, Figure 2A) during 600 s to allow the release of Ca 2+ ions by the ER, and the subsequent opening of the store-operated Ca 2+ channels (SOCC). Addition of 1 mM CaCl 2 allowed for a massive entry of Ca 2+ ions through SOCC leading thus to an increase of the cytosolic calcium concentration ([Ca 2+ ] cyt ). To test 2-APB and its analogues, different concentrations of compounds were added 30 s prior to Ca 2+ addition. ] cyt ) measurement in Jurkat cells using Indo-1 fluorescence. Cells were treated for 10 min with thapsigargin (1 µM, black arrow) to allow Ca 2+ release from endoplasmic reticulum (ER) and opening of the SOCE channels. After 10 min, 1 mM CaCl 2 was added, allowing Ca 2+ entry through the SOCE channels. 2-APB (A) or P11 (B) at the indicated concentrations was applied 30 s prior to CaCl 2 (red arrow). Results are representative for five experiments. (C) Dose-response curves of compounds P1 (= 2-APB), P3, P7, P9, and P11 on Jurkat cell SOCE. Experiments were performed as in Figure 2A, and the peak [Ca 2+ ] cyt was expressed as % of the peak [Ca 2+ ] cyt recorded in the absence of any compounds. Compounds were added 30 s prior to CaCl 2 . Results are expressed as mean ± SEM (n = 3 to 5 according to the tested molecule). Figure 2A, our 2-APB had the typical dual effect on SOCE amplitude: a large potentiation at 5 µM (> 200% of control [Ca 2+ ] cyt increase) and an almost total inhibition at 50 µM. No significant differences were found between commercial and our 2-APB (not shown). Figure 2B depicted the effect of increasing compound P11 concentration: noteworthy the SOCE was totally blocked for concentrations > 100 nM. As shown in We realized the same kind of experiments with all our 2-APB analogues (only four molecules were shown on Figure 2C) and all the molecules have lost the possibility to increase the [Ca 2+ ] cyt , confirming that enlarging the two phenyl groups impairs the potentiation process, resulting in inhibitors with a K i ranging 1000-fold (between 75 nM and 75 µM, Table 1). Noteworthy, we also tested all these molecules on a lymphoblast cell lines devoid of any Orai1 expression due to a double mutations [19] and none was able to induce a [Ca 2+ ] cyt increase due to the activation of Orai3 channels as 2-APB can do. Addition of Methyl and/or Methoxyl Groups The addition of one methyl on both phenyl rings in para and meta positions respectively for compound P3 (and P5) allowed the creation of inhibitors with an inhibition constant of 2-3 µM (Table 1, Figure 2B). Interestingly, the addition of a second methyl to both phenyl rings significantly increased the inhibition efficacy of the molecule (p > 0.05, n = 3, compounds P7 (and P10), Figure 2B). Among these two molecules, the compound P10 was the most active with a K i ≈ 0.5 µM (Table 1), and complete inhibition was reached at 3 µM. We also created 2-APB analogs with one methoxy group on the two aromatic rings (compounds P2 and P4). Even if the presence of a methoxy group on the aromatic rings in para position had the same effect as a methyl group (compound P3 ≈ P2), the replacement in meta drastically impaired the inhibitory capacity of the compound P4 (K i = 75 ± 21 µM, Table 1) with a maximal inhibition of ≈ 80% (not shown). Compound P16 represents a chimera between P3 and P2 ( Figure 1), bearing one methyl and one methoxy group on either aromatic ring. As compounds P3 and P2 had similar properties, unsurprisingly P16 had a similar K i (3.1 ± 0.6 µM, Table 1). From these molecules, it seems that the more the two rings are ramified, the more they inhibit the Jurkat cell SOCE. Fluorine-and Chlorine-Containing Analogs Compounds P13 and P14 are, respectively, analogs of compound P3 where the methyl groups were replaced by chlorine or fluorine. The presence of a halide only weakly modified the K i ≈ 2 µM and 5 µM respectively for compound P13 (Cl) and compound P14 (F) vs. 3.5 µM for compound P3 (Table 1). Interestingly, the compound P15 which has one methyl + one chlorine on both phenyl rings had a large increase in its inhibitory capacity (K i ≈ 0.3 µM, Table 1). These results confirm that the more ramified is the phenyl group, the more efficient the molecule inhibits the SOCE. However, the use of chlorine or fluorine did not hugely improve the inhibition capacity of the molecules. Compounds P8 and P9 These two compounds are analogs of P3 with a bulky alkyl residue (tBu, P8) or a linear alkyl chain (nBu, P9). P8 and P9 are better inhibitors with K i = 350 ± 40 nM and 641 ± 103 nM, respectively (n = 3 for both, Table 1 and Figure 2B). Thus, the addition of larger groups increases the efficacy of the molecules. Noteworthy, compound P7, bearing a 3,4-dimethyl substituent, displayed the same behavior as P9 ( Figure 2B). Fused Ring Analogs The best inhibitor we previously described was a compound in which the two phenyl groups were replaced by 2 benzothienyl groups attached to the central boron atom by the thienyl parts (K i ≈ 0.4 µM, full inhibition at concentration > 3 µM [7]). Our compound P12 is very similar, except that the benzothienyl groups are attached to the boron atom by the benzyl parts. The reversal of this group had no real effect as the K i of compound P12 was very similar at 374 ± 96 nM (n = 3, Table 1). The naphtyl analog P6 had a slightly better K i , 275 ± 17 nM (n = 3, Table 1). However, the best inhibitor was obtained when the group was formed by two phenyls bound by a single C-C linker. This compound P11 had a K i < 100 nM (75 ± 21 nM, n = 3, Table 1) and fully inhibited the SOCE at concentrations below 1 µM ( Figure 2B). Interestingly, when we added a gem-dimethyl between the two phenyl groups (compound P17), no effect on the SOCE amplitude was seen for concentrations >3 µM, and higher concentrations disrupted the fluorescence and impaired the measurement. Thus, the free rotation of the two phenyl groups on the same side seems important for the high efficacy of compound P11. P11 Specifically Inhibits Store-Operated Calcium Entry (SOCE) but Does Not Affect the Ca 2+ -Extrusion Mechanisms As shown in Figure 2A, after TG treatment and addition of extracellular Ca 2+ , the [Ca 2+ ] cyt rapidly increased and reached in control conditions a peak of about 1 µM followed by a decay to a new equilibrium around 500 nM. As [Ca 2+ ] cyt results from a balance between Ca 2+ influx and efflux, this decay was due to the activity of the plasma membrane Ca 2+ ATPases ("PMCA") and the Na + /Ca 2+ exchanger ("NCX") counteracting the SOCE (TG has blocked the SERCA activity, impairing Ca 2+ uptake in the ER). To ensure that compound P11 acted on the SOCE and not on the Ca 2+ -efflux mechanisms, we next performed a Mn 2+ -quenching experiment with Indo-1. Mn 2+ ions enter the cells through SOCC but cannot be pumped outside the cells by PMCA and NCX; once in the cells, Mn 2+ bound to Indo-1 and quenched its fluorescence measured at 430 nm. Thus, an increase of the Mn 2+ influx amplitude was associated with an increase of the Indo-1 quenching speed, and reversely [7,9]. From this result, it is clear that the compound P11 targeted the SOCE. P11 Is Also Able to Block SOCE in Other Cell Types We also tested P11 on the DG75 (Burkitt B lymphoma), U937 (monocytic lymphoma), and MDA-MB231 (breast carcinoma) cell lines. With the SOCE measured in the same conditions, compound P11 was significantly more efficient on these three cell lines than on Jurkat cells with K i = 32 ± 2 nM, 40 ± 5 nM, and 50 ± 5 nM, respectively, for DG75, U937, and MDA-MB231 cells vs. 75 ± 21 nM for Jurkat cells ( Figure 3B). Thus, in several cell lines, full inhibition of the SOCE was already obtained at a concentration of 100 nM ( Figure 3B). We next tested P11 efficacy on two lymphoblastoid cell lines ("LCL") established from an Orai1-deficient with a severe combined immunodeficiency ("LCL Orai1−") and a healthy ("LCL Orai1+") patients ( Figure 3C [19]). As shown in Figure 3C, in absence of extracellular CaCl 2 , TG induced the Ca 2+ release from the ER in the two types of cells. However, CaCl 2 addition after 10 min of TG treatment was only able to induce a massive [Ca 2+ ] cyt rise in LCL expressing Orai1 proteins, confirming the SOCE default of the SCID patient. Noteworthy, 1 µM P11 only induced an immediate decrease of [Ca 2+ ] cyt in Orai1-expressing LCL and was inefficient in Orai1-deficient LCL. These results confirm that Orai1 is the target of P11 in these cells. Selectivity of Compounds P11 and P9 Despite interesting properties, 2-APB is a complex molecule that lacks specificity. Thus, 2-APB was firstly described as an IP 3 R inhibitor [20] but it blocks also SERCA [14] and some TRP channels (TRPC3 for example [21]). Thus, on Jurkat cells, 2-APB has three main targets that play an important role in Ca 2+ homeostasis. As our molecules are analogs of 2-APB, it was obvious to check whether they also affected SERCA and IP 3 R activity. Among our various molecules, we selected to perform experiments on P11 and two structurally close compounds, P8 (Ki ≈ 300 nM) and P9 (Ki ≈ 600 nM). Measurements of SERCA-mediated 45 Ca 2+ store loading and IP 3 -dependent release from the ER Ca 2+ store were performed on permeabilized cells in unidirectional conditions. As this approach required the use of cells forming strongly adherent monolayers, we used L15 fibroblasts stably overexpressing IP 3 R1, similarly to our previous work [9]. Unfortunately, P8 induced already a strong Ca 2+ leak from the ER by an IP 3 -independent pathway, impairing the experiment. As shown in Figure 4, both P11 and P9 displayed some inhibitory effects on the activity of SERCA and/or IP 3 R. Compound P9 did not significantly inhibit IP 3 R activity but was only a partial inhibitor of the SERCA activity: −52 ± 3% (n = 4) at 100 µM with a K i value of 6 ± 0.3 µM (n = 4). Interestingly, below 3 µM, compound P9 thus only inhibited SOCE and quickly lost its efficacy at decreasing concentrations ( Figure 4). In contrast, compound P11 was a weak inhibitor of the IP 3 R (−29 ± 9% at 100 µM, n = 3); calculation of the K i was however impossible because too high P11 concentrations would be needed to determine this value with sufficient precision. This molecule was also a more complete inhibitor of the SERCA activity, with an almost total inhibition at concentrations > 30 µM (−89 ± 1%, n = 3). However, compound P11 was therefore still a hundred times less efficient for inhibiting SERCA activity than for inhibiting SOCE (K i = 7.4 ± 2 µM (n = 3) vs. 75 ± 21 nM (n = 3). Thus, at concentrations < 1 µM, P11 is a specific SOCE inhibitor without any effect on SERCA or IP 3 R in Jurkat cells. . Dose-response curves for compounds P9 (A) and P11 (B) indicating a preferential inhibition of SOCE over the loading of the ER Ca 2+ stores by SERCA pumps ("Loading") and the Ca 2+ release from the ER by IP 3 Rs ("IICR"). The effects of P9 and P11 on the IP 3 -induced Ca 2+ release ("IICR") and the SERCA-mediated Ca 2+ loading ("loading") were investigated in permeabilized L15 fibroblasts. The ER Ca 2+ stores of saponin-permeabilized L15 fibroblasts were loaded until steady-state with 45 Ca 2+ and Ca 2+ release was induced by incubation in a Ca 2+ -free efflux medium in the presence of TG (10 µM). The indicated concentration of P9/P11 were added either during the loading phase for assessing SERCA activity or during the efflux phase before addition of IP 3 (0.7 µM) to assess IP 3 R activity. Activity was expressed as % of the activity measured in the absence of any compounds. Results are expressed as mean ± SEM (n = 3 for "loading, n = 5 for "SOCE" and "IICR"). All our subsequent experiments were therefore performed only with compound P11 as it is a specific inhibitor of SOCE that can be used at concentrations <1 µM. Compound P11 Impairs the Activation of Jurkat Cells Activation of T cells needs a [Ca 2+ ] cyt increase for several minutes to allow the synthesis of interleukin-2 ("IL-2") and the further activation of the immune functions. Therefore, inhibition of the [Ca 2+ ] cyt rise due to SOCE impairs T cell activation and proliferation [7,22]. Having shown that compound P11 was able to block the Jurkat cell SOCE, we next performed experiments using phytohemagglutinin ("PHA"). PHA is commonly used to cross-link the T cell receptor allowing the synthesis of IL-2 and thereafter the T cell activation [22]. As shown in Figure 5A, in presence of 1 mM extracellular CaCl 2 , PHA induced a Ca 2+ mobilization which was largely inhibited by increasing [P11]. Noteworthy, PHA was able to induce a weak Ca 2+ release from internal stores (orange curve). One micromolar of P11 did not affect this Ca 2+ release as the PHA-induced Ca 2+ rise was very similar to the one recorded in absence of extracellular Ca 2+ concentrations not only blocked the SOCE of PHA-stimulated cells, but also emptied the cytosolic Ca 2+ content. In the absence of PHA stimulation, Jurkat cells did not produce any IL-2 ("UT", Figure 5B). On the contrary, under PHA stimulation, approximately 100 pg/mL IL-2 was detected in the medium (103 ± 4 pg/mL, n = 3, Figure 5B) after 24 h. Addition of increasing P11 concentrations at the time of PHA stimulation progressively decreased the IL-2 synthesis to 26 ± 6 pg/mL at 1 µM P11 (−74%, n = 3, Figure 5A). By itself, 1 µM P11 was not able to induce IL-2 synthesis (not shown). As this diminution in IL-2 synthesis can both be explained by the decrease of its synthesis by all cells, or by a decrease in the number of cells that can synthesize IL-2, we next evaluated cell death by trypan blue staining in the presence of PHA stimulation with and without the addition of compound P11. Under resting culture conditions, the percentage of dead cells measured was lower than 5% (3.4 ± 0.4%, n = 5, Figure 5C). After 24h PHA stimulation, the number of dead cells did not significantly increase (4.4 ± 1.0%, n = 4). Furthermore, compound P11 by itself was not toxic as no significant increase in dead cells was noticed at a maximal concentration of the compound (at 1 µM: 4.0 ± 0.4%, n = 5). Addition of increasing concentrations of compound P11 simultaneously with PHA, however strongly increased the proportion of dead cells up to 33.2 ± 1.3% at 1 µM (n = 4, p < 0.05, Figure 5C). By calculating the amount of IL-2 synthetized by the remaining living cells, it was clear that increasing concentrations of compound P11 decreased this synthesis capacity by ≈ 63% (39.2 ± 8.6 pg/mL vs. 107.6 ± 4.1 pg/mL for PHA-stimulated cells in the absence of P11, Figure 5D). These results clearly show that compound P11 has a dual effect on activated Jurkat cell, on the one hand reduction of IL-2 synthesis and on the other an induction of their death. Compound P11 Induces Apoptosis of PHA-Stimulated Jurkat Cells To verify the mechanism of cell death induced by compound P11, we next performed caspase-3 activity assays using the fluorogenic Ac-DEVD-AFC substrate as well as a TUNEL assay. Moreover, similar results were obtained when the percentage of dead cells was measured with the TUNEL assay (Figure 7). In resting conditions, less than 5% of the cells had fragmented DNA (4.2 ± 0.3, n = 3). PHA alone induced a threefold increase in the number of cells with fragmented DNA (13.3 ± 1.6%, n = 3). At concentrations > 10 nM, compound P11 induced a ≈ 4.5-fold increase (17.9 ± 1.2%, n = 3 at 1 µM). Taken together, these results demonstrate that compound P11 is not toxic for the cells when added alone but could profoundly impair Jurkat cell activation by inducing their apoptosis. Discussion Since the discovery of the molecular partners implied in SOCE in the middle of the 2000s, the characterization of their modulators has emerged as attracting enough for pharmaceutical companies like GSK and Hoffmann-La Roche or newcomers like Calcimedica. Indeed, it appeared that aberrant SOCE could be associated with autoimmune disorders, inflammation, or cancers [23]. Inhibiting SOCE appears as a new and promising way to treat this kind of disease or to ease it. SOCE was functionally characterized during the 1980s (formerly named capacitative calcium entry by Jim Putney [24]), but its molecular mechanism was only unveiled in the last decade [3,25]. After store depletion, the ER proteins STIM1 interact with and activate the plasma membrane Ca 2+ channels mainly consisting of Orai1 proteins [3]. Noteworthy, loss-of-function mutations in the Orai1 gene, rendering the Orai1 protein unfunctional or absent in the plasma membrane, via SOCE inhibition, mainly impair the immune system [26]. On the opposite, some gain-of-function mutations in Orai1 or STIM1, allowing a constitutive SOCE or an increased SOCE amplitude are also associated with diseases like the York platelet syndrome or the Stormorken syndrome [27]. From this fact, the control of Orai1 channels (directly or indirectly via STIM1) and the SOCE emerged as a new way to control the immune system, but also the proliferation of cancerous cells or other diseases. Noteworthy, two Orai1 paralogues exist, Orai2 and Orai3 with "less global" roles: thus, Orai2 regulates the SOCE amplitude by forming heteromeric channels with Orai1 with reduced conductance in some T cells [28]. About Orai3; there exists a dichotomy between SOCE observed in breast cancer cell lines and in cells from patient samples. In estrogen-receptor expressing cells, Orai3 channels are responsible for the SOCE, while Orai1 are the ones responsible in estrogen-receptor-negative cells [29]. Thus Orai3 channels could also represent a target of interest for the treatment of the estrogen-receptor-positive breast cancer cells which represent 80% of breast cancers [29]. However, in the vast majority of the cell types, the spatial emplacement of the three Orai proteins in the cells seem to be different: Orai1 mainly is located in the plasma membrane [3,30], Orai2 in the internal compartments, and Orai3 in the ER membrane [31][32][33]. Thus, in T and B cells only Orai1 participates in SOCE [3,9,19]. Despite huge therapeutic potential, the characterization of new SOCE modulators is impaired by the limited knowledge of the Orai1 structure. Indeed, only the crystal structure of the Drosophila melanogaster dOrai, which shares 73% sequence homology with hOrai1, has been published at a 3.35 Å resolution [34]. Furthermore, no hOrai1 structure has yet been established when bound to STIM1 in its active conformation. A controversy also still exists on the number of Orai proteins needed to form a functional channel, tetramer [35] or hexamer [36,37], although the latter seems to be the most accepted currently. Therefore, the development of new SOCE modulators is still not possible by docking work but powered by the characterization of analogs of existing modulators. Among SOCE modulators, we focused our work on 2-APB, based on its following three properties: (i) 2-APB is a small molecule; (ii) 2-APB potentiates SOCE amplitude at low concentrations (≈ 5 µM) and inhibits SOCE at higher concentrations (> 30 µM); (iii) the relative ease to synthesize analogs. In three previous studies, we already showed that the central boron-oxygen-carbon plays a role in the potentiation capacity and the ethanolamine and the two phenyl groups in the inhibition [7,9,38]. Thus, by modifying one or the other part of the 2-APB molecule, it is possible to create analogs with better potentiation or inhibition ability. Furthermore, our 2-APB analogs request Orai1 expression to act on the T cell SOCE [9]. In the present work, we only modified the two phenyl rings by adding simple radicals like methyl or by replacing them with more complex cycles to develop new SOCE inhibitors (Figure 1). Interestingly, all our molecules modified on the two phenyl groups lost their potentiation capacity. In Djillani et al. 2014 [7], we showed that two analogs where one (nicknamed "methoxy-APB") or two methoxyl ("dimethoxy-APB") groups were added on the same phenyl, keeping the second one intact, still had a potentiation capacity (even if it was decreased). On the opposite, our new compounds with added methoxyl groups on the phenyl groups have only inhibiting properties, confirming that increasing the size of both phenyl groups allows the creation of new SOCE inhibitors. Among our new molecules, compound P11 was the most potent SOCE inhibitor. With an inhibitory constant between 32 and 75 nM depending on the cell type used and a complete SOCE inhibition at a concentration well below 1 µM, P11 is one of the best SOCE inhibitors described in the literature. The twofold difference in Ki between cell types could reflect a difference in the Store-Operated Calcium Channels (SOCC) composition. Indeed, it has been shown that the expression ratio of Orai1/STIM1 modulates the sensitivity of the SOCE to 2-APB [39]. Furthermore, Orai1 can be associated with some TRP channel subunits like TRPC1, TRPC3, and TRPC6 [40] which can also interact with 2-APB, and thus potentially with its analogs. The free rotation between the two phenyl groups on the same side seems to play an important role in the inhibition capacity of compound P11. When this rotation is impaired by a second carbon "bridge," like in compound P24, the molecule was not able to inhibit, even partially, the SOCE at concentrations below 1 µM. Thus, we assume that when 2-APB or its analogs enter the SOCC pore, the two "wings" (phenyl or larger/longer aryls) need to tether and reach the hypothetic inhibitory binding site(s). However, the great advantage of compound P11 is that it avoids the major side effects of 2-APB, as it did neither inhibit the IP 3 R nor the SERCA pumps of the ER. Furthermore, it did not induce a Ca 2+ leak from the ER by an unknown pathway, for example, as our compound P8 or as DPB162-AE [10] did. This absence of side effects could be related to the apparent non-toxicity of the molecule on resting cells: compound P11 by itself at 1 µM did neither increase DNA fragmentation nor cell death. Jim Putney showed more than 30 years ago that the Ca 2+ influx was activated by the ER depletion, and that this Ca 2+ entry he called "capacitative calcium entry," eventually named store-operated Ca 2+ entry (SOCE), allowed the refilling of the ER [24]. More recently, Zheng et al. [41] showed that the refilling of ER Ca 2+ is fully impaired in cells that do not express all three Orai isoforms. However, after 24 h in culture, Ca 2+ stores became eventually refilled, meaning that SOCE is critical for the immediate refilling of the ER Ca 2+ stores. It is known that ER Ca 2+ store depletion leads to ER stress induction that, if not counteracted by the unfolded protein response, will provoke apoptosis. In our experiments, the cells were stimulated with PHA and different [P11] for 24 h to study the apoptosis induction and to measure IL-2 synthesis and caspase-3 activity. Remarkably, P11 was not by itself cytotoxic, and it exerted a cytotoxic effect only when the cells were stimulated and the [Ca 2+ ] cyt rise impaired ( Figure 5A). Surprisingly P11 induced a decrease of [Ca 2+ ] cyt beneath pre-stimulation values. As the cells were still under PHA stimulation, we assume that there was a constant ER Ca 2+ release and that the SOCE blockade impaired the refilling of the ER Ca 2+ content and could induce an ER stress leading to the apoptosis. Enlarging the 2-APB molecule on the two phenyl groups allows the development of increasingly efficient inhibitors. We anticipate that our compound P11 could also be further modified and are confident that molecules with an inhibition capacity in the low nanomolar range could be obtained. Thus, from our studies and those of Mikoshiba's group, it appears possible to create 2-APB analogs with higher potency and selectivity. Due to its dual effect on SOCE, the advantage of 2-APB over other SOCE modulators is that, depending on the substitution performed, it is possible to create inhibiting or potentiating agents. Our compound P11 could, therefore, represent a new promising branch of the 2-APB family that could help to cure human disorders in which Ca 2+ homeostasis is modified. Cell Lines Jurkat (acute T cell leukemia from ATCC), DG75 (Burkitt B lymphoma, kind gift of Martin Rowe, University of Birmingham, UK), and U937 (monocytic cell line from ATCC) cell lines were basically maintained in RPMI-1640 medium (Lonza, Levallois-Perret, France) supplemented with 10% heat-inactivated fetal calf serum and 2 mM L-glutamine, at 37 • C in a 5% CO 2 humidified atmosphere. The MDA-MB231 breast cancer cell line (form ATCC) was maintained in same conditions as the leukocyte cell lines except that DMEM (Lonza, Levallois-Perret, France) medium was used instead of RPMI1640. Epstein-Barr virus immortalized B cells from Orai1-deficient and healthy patients were a kind gift of Capucine Picard and Alain Fischer (study center of primary immunodeficiencies, AP-HP, Hopital Necker, Paris France [19]) and maintained as the leukocytes cell lines. The L15 fibroblast cell line stably overexpressing IP 3 R1 [42] was a generous gift of Katsuhiko Mikoshiba (now at Shanghai Tech Univ., China). The cells were maintained in DMEM medium (Gibco ® , Life Technologies, Gent, Belgium) supplemented with 10% heat-inactivated fetal calf serum, 3.8 mM L-glutamine, 1% non-essential amino acids, 85 IU mL −1 penicillin, 85 µg/mL streptomycin, 400 µg/mL geneticin, and 25 mM Na-Hepes (pH 7.4) at 37 • C and 10% CO 2 . Cytosolic Ca 2+ Concentration and SOCE Measurement [Ca 2+ ] cyt was recorded by a fluorimetric ratio technique as previously described [7,43,44]. Leukocytes were spun and resuspended at a density of 10 6 cells/mL in Hepes buffered saline (HBS; 135 mM NaCl, 5.9 mM KCl, 1.2 mM MgCl 2 , 11.6 mM Hepes, 11.5 mM glucose adjusted to pH 7.3 with NaOH). Due to the presence of 0.002% Ca 2+ in the NaCl, this HBS medium contained 0.1 mM of Ca 2+ . Then, cells were incubated in the dark with 4 µM Indo-1-AM for one hour at room temperature under slow agitation. Cells were then centrifuged and resuspended in HBS medium before measurement. To perform the experiments, 0.5 to 1 × 10 6 cells were suspended in 2 mL HBS in a quartz cuvette and inserted into a spectrofluorophotometer (RF-1501 Shimadzu Corporation, Kyoto, Japan) connected to a PC computer (Dell Computer Corp., Montpellier, France). A temperature of 37 • C was maintained by circulating water from a bath. Ultraviolet light of 360 nm was used for excitation of Indo-1, and emissions at 405 and 480 nm were recorded. Background and autofluorescence of the cell suspension were subtracted from the recordings. The maximum Indo-1 fluorescence (R max ) was obtained by adding 1 µM ionomycin to the cell suspension in the presence of 10 mM extracellular CaCl 2 . Minimum fluorescence was determined following the depletion of external Ca 2+ with 5 mM EGTA. [Ca 2+ ] cyt was calculated according to the equation [Ca 2+ ] cyt = K d (R-R min )/(R max -R), where K d is the apparent dissociation constant of Indo-1-calcium complex (230 nM), and R is the ratio of fluorescence values recorded at 380 nm in absence and presence of 10 mM CaCl 2 [43]. To visualize the SOCE, cells were placed into the quartz cuvette and 0.5 EGTA was added just before the beginning of the recordings. To induce the SOCE, the cells were treated with 1 µM thapsigargin (TG) during 10 min in Ca 2+ -free HBS to induce Ca 2+ release from the ER and the opening of SOCE channels [7,45]. Then, 1 mM CaCl 2 was added to measure the change in [Ca 2+ ] cyt subsequently to Ca 2+ influx [45]. In the dose-response experiments, different concentrations of 2-APB and its analogs were added 30 s prior to Ca 2+ readdition [7,9]. In experiments with PHA, cells were placed in the cuvette, and 0.9 mM CaCl 2 was added to reach a 1 mM extracellular concentration. Free Ca 2+ concentrations were calculated using the Bad4 software [46] Measurement of Mn 2+ Influx by Indo-1 Quenching To study directly the divalent ion influx through the SOCC with selected 2-APB analogs, we performed Mn 2+ quenching experiments. Mn 2+ ions bind to Indo-1 and quench the 430 nm emission wavelength fluorescence after excitation at 360 nm. The decrease of fluorescence is directly dependent on Mn 2+ entry through SOCC. For measurements, cells were treated for 10 min with 1 µM TG to open SOCC, whereupon 100 µM MnCl 2 was added instead of CaCl 2 . After 60 s a 2-APB analog was added as indicated and the change in the curve slope was measured. After permeabilization, the non-mitochondrial Ca 2+ stores were loaded for 45 min in 120 mM KCl, 30 mM imidazole-HCl (pH 6.8), 5 mM MgCl 2 , 5 mM ATP, 0.44 mM EGTA, 10 mM NaN 3, and 150 nM free 45 Ca 2+ (28 µCi mL −1 ). Subsequently, efflux was initiated by incubation in efflux medium (120 mM KCl, 1 mM EGTA, 10 µM TG and 30 mM imidazole-HCl pH 6.8). The latter medium was replaced every two minutes for 18 minutes. The exact amount of Ca 2+ at each time point is calculated by summing in retrograde order the amount of radioactivity remaining in the cells at the end of the efflux and the radioactivity collected during the successive time intervals. This allows to calculate the steady-state content of the ER Ca 2+ stores (indicative for the SERCA activity, after subtraction of the loading achieved in the presence of 2 µM TG) the basal leak rate of Ca 2+ out of the stores and the IP 3 -induced Ca 2+ release after addition of IP 3 (0.7 µM) [14]. These various parameters were measured both in control conditions or in the presence of various 2-APB analogs as indicated. To assess the effect of the 2-APB analogs on the SERCA-mediated loading of the ER Ca 2+ stores, the analogs were added during the 45 Ca 2+ -loading phase as described previously [9]. Obviously, compounds that strongly stimulated the basal Ca 2+ leak out of the stores (as P8) could not be investigated for their effect on SERCA activity (as Ca 2+ loading was already compromised by the increased leak) or on IP 3 -induced Ca 2+ release (as the decreased luminal Ca 2+ level impacted both the driving force for Ca 2+ release and the sensitivity of the IP 3 R) [48]. Interleukin-2 (IL-2) Assay Forty-eight-well plates were seeded with 10 6 Jurkat T cells per well. Cells were then incubated for 24 h with or without 10 µg/mL PHA and 0 to 1 µM P11 in complete culture medium. After 24 h, supernatants were collected, and IL-2 amounts were quantified using a Quantikine Human IL-2 Immunoassay (R&D Systems Europe, Lille, France). Optical density (OD) was measured at 450 nm using a Victor3 plate reader (Perkin Elmer, Courtaboeuf, France). The concentration of supernatant IL-2 was determined from the OD curve obtained with the standard. Results were the mean of four repeats. In parallel plates, cell viability was accessed after treatment with trypan blue and counting with a Malassez hematimeter (Sigma-Aldrich, Saint Quentin Fallavier, France). Fluorometric Assay for Caspase-3 Activity Caspase-3 activity was measured using the fluorogenic substrate Ac-DEVD-AFC (Tebu-bio, Le Perray-en-Yvelines, France) according to the manufacturer's instructions. Briefly, 48-well plates were seeded with 10 6 Jurkat T cells per well. Cells were then incubated for 24 h with 0, 1, or 10 µg/mL PHA and 0 to 1 µM P11 in complete culture medium. After 24 h, cells were collected and lysed in RIPA lysis buffer supplemented with protease inhibitor cocktails (Life Science Biorad, Marnes la Coquette, France). The protein samples were quantified using the DC protein assay reagent (Biorad, Les Ulis, France). Cell lysates (50 µg) were diluted with reaction buffer (200 mM Hepes pH 7.4, 1 mM EDTA, 20% sucrose and 20 mM dithiothreitol) and incubated with the fluorogenic substrate (20 µM final concentration) for 45 min up to 7 h at 37 • C. The release of 7-amino-4-trifluoromethylcoumarin (AFC) was measured by fluorescence (excitation at 405 nm, emission at 505 nm) using a Victor3 plate reader (Perkin Elmer, Courtaboeuf, France). Apoptosis Detection by the Terminal Transferase dUTP Nick End Labelling (TUNEL) Method We measured the apoptosis levels in Jurkat cells by the "in situ cell death detection" kit (Roche Applied Science, Meylan, France). Briefly, 50,000 cells/well were seeded in a 96-well plate and treated with or without 10 µg/mL PHA and 0 to 1 µM P11 for 6, 12, or 24 h. Cells were spun, fixed, and permeabilized according to the manufacturer's instructions and the terminal transferase dUTP Nick end labelling reaction was performed in the dark during 60 min at 37 • C. For visualization of the results, an epifluorescence microscope (Axioskop, Karl Zeiss, Le Pecq, France) was used at an excitation wavelength of 488 nm, and an emission wavelength of 545 nm. Cells were also loaded with 4',6' -diamidino-2-phenylindole (DAPI) to visualize the nuclei. Total cell number (>200) and TUNEL-positive cells were counted in several fields (>5) and the ratio of apoptotic cells was calculated. This experiment was repeated three times. Chemicals TG and ionomycin were purchased from Calbiochem-Merck (Nottingham, UK). Synthesis of 2-APB and its analogs were done using previously described methodologies [49]. General procedure for APB analog synthesis: To a solution of arylbromide (3 eq), magnesium (3.5 eq), and diisopropylamine borane complex (DIPAB, 1 eq) in dry THF (1 mL/mmol) was added at room temperature to a solution of PhMgBr (2 M in THF, 0.05 eq). After 10 min, (or the end of gas evolution on large scale), the reaction mixture was heated to 70 • C until no arylbromide remained in the reaction mixture (8 to 16 h) as witnessed using TLC monitoring. At 0 • C, 1 mL/mmol of dry MeOH was added (caution, exothermic reaction). After 1 h, the reaction mixture was concentrated under reduced pressure, and diluted in a mixture of 1N HCl/MeOH (7:3, 10 mL/mmol), after 1 h, the product was extracted with Et 2 O (3 × 10 mL/mmol) and the combined organic phases washed with 1N HCl (10 mL/mmol) and brine (3 × 10 mL/mmol). Organic phases were then dried over anhydrous Na 2 SO 4 and concentrated under reduced pressure. The residue was dissolved in Et 2 O (4 mL/mmol) and 1.2 eq of ethanolamine was added. Depending on substituents, precipitation occurs spontaneously or requires addition of pentane. Crystals were filtered, washed with pentane, and dried under vacuum to yield the pure borinic acid 2-aminoethyl ester. DIPAB Diisopropylamine-borane complex (55124-35-1) A 3000 mL three-necked round-bottomed flask equipped with a mechanical stirrer, a thermometer, and a dropping funnel was charged with THF (1500 mL) and NaBH 4 (56.75 g, 1.5 mol) The heterogeneous mixture was vigorously agitated using a mechanical stirrer and cooled with an ice/salt bath below 0 • C. The dropping funnel was charged with 37.8 mL of H 2 SO 4 (0.71 mol) The H 2 SO 4 was added dropwise maintaining the internal temperature below −5 • C over 3 h. An iPr 2 NH (140 mL, 1 mol) solution in THF (100 mL) was added dropwise maintaining the temperature below 0 • C over 1 h 30 m. The mixture was vigorously agitated during 20 h at room temperature. The mixture was filtrated over a N • 3 fritted funnel and the resulting solid was triturated with THF (3 × 400 mL). THF filtrate was concentrated under reduced pressure the residue was taken with CH 2 Cl 2 , and then filtrated to eliminate all solid residues. The filtrate was washed with water (4 × 200 mL). The organic phase was dried over Na 2 SO 4 and concentrated under reduced pressure to give DIPAB as colorless oil which solidified upon cooling (102 g, 90%). 1 1.02 g of ((diphenylboryl)oxy)ethanamine were obtained following the general procedure using 1.73 g of bromobenzene as a white solid (92% yield). 1 13 C NMR to the cell lines, absence of effect on the two other main targets of 2-APB, SERCA and IP 3 R, absence of cytotoxicity on non-activated cells, but the ability to induce the caspase-3 activity and the apoptosis of PHA-activated cells. Thus, P11 is able to highjack the Ca 2+ signaling of the T Jurkat cells from proliferation to death.	2020-12-24T09:08:23.869Z	2020-12-01T00:00:00.000	{ "year": 2020, "sha1": "43ab4543654b981a88246bb9643835cc48809967", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/1422-0067/21/24/9777/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "77c981ff23a689b7016a5029ce9e3d36a7fd7591", "s2fieldsofstudy": [ "Biology", "Chemistry" ], "extfieldsofstudy": [ "Medicine" ] }
229317868	pes2o/s2orc	v3-fos-license	Abscisic Acid Inhibits Asymbiotic Germination of Immature Seeds of Paphiopedilum armeniacum Paphiopedilum armeniacum is a rare orchid native to China with high ornamental value. The germination of P. armeniacum seeds is difficult, especially for the mature seeds, which is the major limitation for their large-scale reproduction. This study explored the reasons for seed germination inhibition from the aspects of the important plant endogenous hormone—abscisic acid (ABA). The major endogenous hormone contents of seeds were determined at different developmental stages. The ABA content was 5.8 ng/g in 73 days after pollination (DAP) for the immature seeds, peaked at 14.6 ng/g in 129 DAP seeds, and dropped to 2.6 ng/g in the late mature stage of the 150 DAP seeds. The reduction of ABA content in the mature seed suggests a possible contribution to the increased expression of CYP707A, an ABA catabolism gene. The germination rate of the immature seeds was reduced to 9% from 69% when 5 μg/mL ABA was added to the Hyponex N026 germination medium. The result showed that ABA can inhibit the germination of P. armeniacum immature seeds. However, for the heavily lignified mature seeds, reduction in endogenous ABA level does not result in an increase in the germination rate. Lignin accumulation in the seed coat imposes the physical dormancy for P. armeniacum. In summary, the germination of P. armeniacum is regulated by both ABA and lignin accumulation. Introduction The Paphiopedilum genus belongs to the family of Orchidaceae, containing 107 species throughout the world, and are distributed across southern India, Nepal, Bhutan, north-eastern India, Burma, southern China, Hong Kong, South-east Asia, the Malay Archipelago, the Philippines, New Guinea, and the Solomon Islands [1][2][3]. The Paphiopedilums are attractive and popular in the horticulture market due to the unique pouch-shaped lip of the flower and their abundant colors. Unfortunately, the Paphiopedilums are left in a dangerous situation with their supply decreasing sharply because of habitat destruction and excessive collection. In an effort to conserve this endangered species, the Convention on International Trade in Endangered Species of Wild Fauna and Flora placed the entire wild Paphiopedilum species on the list to forbid international transactions [4]. The conventional propagation of Paphiopedilum orchids using axillary buds division from the mother plant is unproductive and time-consuming [5]. Asymbiotic germination is the most common and efficient approach for the large-scale propagation of Paphiopedilum [1]. However, the germination of fully mature Paphiopedilum seeds is often difficult. Studies on the embryo development and germination of Paphiopedilum showed that Paphiopedilum seeds have an optimal germination stage before reaching maturity: The germination of P. armeniacum was about 65% in 120~130 days after pollination (DAP) seeds, while the 180 DAP seeds failed to germinate [6]. In 180 DAP seeds, P. wardii and P. hangianum had the highest germination (65% and 73%, respectively), while the germination was reduced to 20% in 300 DAP seeds [7,8]. The germination of 270 DAP seeds of P. delenatii and P. callosum reached above 90%, while The germination of 300 DAP seeds of P. callosum was only 20% [9,10]. Orchid seeds have a unique characteristic: a capsule containing thousands of tiny seeds of about 0.05-6.00 mm with various shapes [11]. The mature orchid seeds are composed of undifferentiated embryo covered with testa and no morphologically differentiated endosperm and cotyledons, which is different from most flowering plants [12]. The research about the failure of orchid germination focused on some hypotheses: (1) morphological dormancy caused by underdeveloped embryos; (2) physiological dormancy caused by inhibitors accumulated in mature seeds; and (3) physical dormancy resulting from thick impermeable testa [13]. Many studies have reported that abscisic acid (ABA) is an inhibitor of seed germination and plays an important role in inducing seed dormancy [14][15][16]. Early research found that mature terrestrial seeds of Epipactis helleborine, an orchid species with poor germination, contain five times more free ABA than Dactylorhiza maculate, an orchid species with higher germination [17]. The orchid species Calanthe tricarinata, Cypripedium japonicum, and Cypripedium formosanum have relatively high ABA in their seeds, and some of the treatments to reduce the ABA content in Calanthe tricarinata and Cypripedium japonicum mature seeds were shown to improve seed germination [18][19][20]. At present, the research of Paphiopedilum seed germination is still mainly based on the observation of seed morphology. P. armeniacum, a native orchid species to the Yunnan province of China, has been praised for its high ornamental value. In a previous study, key anatomical features in embryo development were shown to be associated with the ability of embryos to germinate in vitro [21]. The fertilized egg is formed around 45 DAP. At 87 DAP, it undergoes longitudinal and lateral division of the cell to form a preliminary globular embryo with the suspensor also beginning to degenerate. After further development around 129 DAP, the final mature embryo is still ellipsoidal without morphological differentiation. In general, the highest germination rate occurs when the globular embryo is formed and the suspensor has not fully degenerated. Similar embryogenesis and development have also been explored in P. barbigerum, P. appletonicnum, P. concolor, P. hirsutissimum, P. delenatii, and P. helenae, although different Paphiopedilum embryos have different developmental and maturity cycles [22][23][24]. Transcriptome and lignin analysis studies further revealed the large amounts of non-methylated lignin that accumulated during the seed maturation of P. armeniacum, which negatively correlates with the germination performance [25]. However, the role of ABA and other important plant hormones are still uncertain. A better understanding of seed physiology will provide valuable information concerning the propagation and help with the conservation of those hard-to-germinate species. The goal of this study is to investigate the changes in endogenous hormones during seed development and the effect of ABA on the asymbiotic germination of P. armeniacum. P. armeniacum Seed Is Composed of Only an Undifferentiated Embryo and One Layer of Heavily Lignified Seed Coat Orchids typically produced capsules with dust-like seeds inside (Figure 1(A1-A3)). As P. armeniacum seeds developed, they gradually changed from a whitish color to brown and black ( Figure 1(A1-A3)). Before 94 DAP, the color of the seeds in the capsules was white, and after 94 DAP, the color began to turn brown. At 150 DAP, the seeds became dark brown and black. Scanning electron microscopy was used to collect information on seed size and seed coat structure. The individual seeds were spindle-shaped, and the surface of the seed coat was composed of numerous testa cells ( Figure 1B). The testa cells were large and rectangular in the middle, and small and round at the ends. This pattern probably resulted from a slow cell division in the integuments during embryo development [26]. During seed development, the seed size increased before maturity at 115 DAP because of cell division and elongation. After maturation, the size began to decrease significantly due to the dehydration-induced shrinkage. To get an overview of the embryo structure inside the seed coats, the seeds were pretreated with sodium hypochlorite to loosen the rigid seed coat structure [27]. The ellipsoid seed embryo was located in the middle of the spindle-shaped seed, and the seed coat was relatively thin in the middle of the seed compared to both ends ( Figure 1C). Unlike the model plants such as Arabidopsis, these embryos consisted of cells with similar size and no structural polarization. The size of the ellipsoidal embryos increased from 73 to 115 DAP. From 115 to 150 DAP, the embryo size remained relatively constant ( Table 1). The embryos occupied a small proportion of the space inside the testa (20 to 30%), leaving a large amount of air space. Previous studies on histochemical analysis showed that the seed coat is a one-layer structure and consisted of the hydrophobic lignin, serving to protect the embryo [21]. The mature seed coat cells are all vacuolated. At maturity, the cells become dehydrated and collapse into a thin layer. Table 1. The size of seed and embryo at different developmental stages of P. armeniacum. The seed dimension was measured from SEM images, and the embryo dimension was measured from optical microscope images. n = 3; ±SD. The size of seed and embryo is calculated from three independently grown biological replicates. In each replicate, approximately 30 seeds were analyzed. P. armeniacum Seed Germination Rate Increased During the Early Stages of the Seed Development and Decrease Dramatically after 94 DAP The asymbiotic seed germination for different stages of seed development is shown in Figure 2. The seeds collected at 73 DAP showed a germination rate of 49%. As the seeds developed, the germination rate increased and reached a maximum of 69% at 94 DAP. As the seeds continued to develop, the germination rate decreased sharply and was reduced to 15% in 115 DAP seeds. After 136 DAP, the seed germination rate was as low as 10%, while in 157 DAP seeds, it dropped to a low of 2%. The germination characteristics of the seeds also appeared to be correlated with the morphological observation of seed color. The period of high seed germination was when the color of the seed was still whitish, while poorer germination appeared to be associated with darker seeds. This observation provides a certain convenience for asymbiotic germination and sowing. The physical appearance of the seeds can be used to judge their maturity. Determination of Major Endogenous Hormones Level in Developing Seeds The coordinated interaction of endogenous hormones such as indole-3-acetic acid (IAA), ABA, and gibberellins (GAs) is known to play certain roles for seed germination. The content of four major endogenous hormones during seed development was measured by high-performance liquid chromatography tandem mass spectrometry (HPLC-MS/MS; Figure 3 and Figure S1). The ABA content was maintained at a relatively low level of around 5.8 ng/g at the early stage of seed development and peaked at 14.6 ng/g in 129 DAP seeds ( Figure 3A). A similar trend was found for Cypripedium formosanum and other orchid species [19,20], which indicated that ABA level remained low during early seed development and rose rapidly mid-way during seed maturation. The increased amount of endogenous ABA may contribute to the germination inhibition before the embryo's maturation. In addition, as an important abiotic stress hormone, ABA may play an important role in the dehydration tolerance of seeds. However, as the seeds matured further, the ABA level dropped to 2.4 ng/g in 150 DAP seeds. At 150 DAP, the role of germination inhibition is possibly served by the accumulation of lignin instead of ABA [25]. ABA level peaked at 129 DAP and decreased after seed coat lignification was accomplished. IAA content increased progressively in the early stages of seed development ( Figure 3A). The IAA level was 17.5 ng/g in 73 DAP seeds and peaked at 19.9 ng/g in 94 DAP seeds. The level was reduced to 17.2 ng/g in 115 DAP seeds and 10.9 ng/g in 129 DAP seeds. In 150 DAP seeds, it reduced to the lowest level at 4.3 ng/g. This result showed that IAA content was higher in the early stages of the seed development compared to mature seeds. Previous research revealed that the globular embryo of P. armeniacum was formed by 94 DAP. The high content of IAA may facilitate the formation of zygotic embryos to globular embryos. Gibberellic acid (GA 3 ) level remained relatively low throughout the whole development of the seeds ( Figure 3B). There are other bioactive gibberellins such as GA 1 or GA 4 besides GA 3 , so the other types of GAs in P. armeniacum may play an important role in seed development. The content of trans-zeatin riboside (TZR) is high in the early stage of seed development and is later decreased and maintained at a low level during the later stages ( Figure 3B). As a cytokinin, TZR plays an important role during the cell division of zygote embryo formation to spherical embryo formation [21,28]. Gene Expression Related to Hormone Biosynthesis The hormonal action of ABA is precisely controlled by the balance between its biosynthesis and catabolism. One differentially expressed gene (DEG) related to ABA biosynthetic pathway and two DEGs in the ABA catabolism pathway were discovered. In the metabolic pathway of ABA (Figure 4), the expression of NCED, encoding an important rate-limiting enzyme for ABA synthesis, increased as the seeds matured and plateaued at 108 DAP. The expression level remained high at the later stages of seed development from 108 to 150 DAP. The expression of the ABA catabolism enzyme ABA 8'-hydroxylase (CYP707A) increased significantly with the maturity of the seeds. Combined with the endogenous ABA level change in the developing seed, there is a possibility that the increased expression of PaNCED contributes to the ABA accumulation in the early stage, while increased expression of the ABA catabolic gene PaCYP707A is responsible for the ABA level reduction in mature seeds of P. armeniacum. In the metabolic pathways of GA, the gene expression levels of GA20OX and GA3OX, the important enzymes for GA synthesis, increased with the development of seeds. However, the gene expression level of catabolism-related gene GA2OX was higher at 66 DAP than later periods, which remained low with little fluctuations. Table S1. Exogenous ABA Treatment Inhibit Immature Seed Germination To investigate the role of exogenous ABA on seed germination, 5 µg/mL ABA was applied to the asymbiotic germination medium, and the germination performance was recorded. The results showed that ABA inhibited the germination performance by reducing embryo expansion and testa rupture of 94 DAP immature seeds. Although the attached seed coat in the control group ( Figure 5A) did not completely fall off, most of the seed embryos swelled and broke through the seed coat, forming an obvious white protocorm. In the treatment group with ABA added (Figure 5B), most of the embryos also swelled but did not break through the seed coat. The seed germination of the ABA-added group was significantly reduced to only 9% ( Figure 5C) compared to the 94 DAP seed germination of 69%. This result showed that exogenous ABA did in fact lead to the inhibition of immature seed germination. The seeds harvested at 94 DAP were used to represent immature seeds' germination performance. Light microscope images of (A) the germinated immature seeds without any treatment and the immature seeds treated with 5 µg/mL ABA (B) in the culture medium. The immature seeds with exogenous ABA treatment demonstrated a reduced ability on both embryo swollen and seed coat rupture, as compared to the one without ABA treatment. Red arrow: Embryo that have breakthrough seed coat; Yellow arrow: Embryo that have not broken through the seed coat. Scale bar = 1 mm. (C) Asymbiotic germination rate of the CK (control) and ABA-treated immature seeds. n = 3; ±SD; Asterisks indicate significant difference from the WT (Student's t test, p < 0.01). Reduction in Endogenous ABA Has no Significant Effect on Mature Seed Gemination To investigate the role of endogenous ABA on mature seed germination, 94 DAP immature seed capsules were treated with fluridone, an inhibitor of ABA biosynthesis, once a week for a month. At 124 DAP, the ABA content was measured, and the results showed that fluridone treatment reduced the ABA level to 2.8 ng/g compared to 9.4 ng/g in the control ( Figure 6). However, there was no significant difference in the germination rate between the seeds that contained low ABA level (2.8 ng/g) and those that contained high ABA level (9.4 ng/g), as shown in Figure 6. This suggests that the asymbiotic seed germination was not improved when the ABA content was reduced in mature seeds. Figure 6. Effect of fluridone (Flu) on ABA content and asymbiotic germination of mature seed in P. armeniacum. The seeds were harvested at 124 DAP and used to represent the germination performance of mature seeds under reduced endogenous ABA level. n = 3; ±SD; an asterisk indicates a significant difference from the WT (Student's t test, p < 0.05). Discussion P. armeniacum is a rare orchid native to China with high ornamental value. To help improve the conservation efforts and commercial production of this endangered species, a fundamental understanding of the reproductive biology of Paphiopedilum is very important. Basic knowledge of embryo and seed development can aid in the design of experiments that can help explain how to improve asymbiotic seed germination in large scale production. The embryo development and germination pattern of P. armeniacum are unique compared to that of other flowering plants (Figure 7). The embryos are poorly developed and consist of cells of similar size and no obvious structural histodifferentiation [29]. Embryo germination gives rise to a tubercle structure called a protocorm, from which the shoot apical meristem subsequently differentiates [30]. There are no necessary signals and nutrients to support histodifferentation before germination occurs because of the lack of an endosperm. This delayed histodifferentiation ensures that only the seeds that arrived at the proper conditions can start the vegetative establishment. The immature seeds at 73 and 94 DAP achieved a high germination rate, estimated from the testa rupture ( Figure 2). However, the protocorms formed from those immature seeds failed to differentiate and develop into plantlets ( Figure 8A,(A1),B,(B1)). Eventually, the growth-arrested protocorms turned brown and withered (Figure 8(A1)). A properly developed and differentiated protocorms can form only from mature seeds ( Figure 8C,(C1)), which will form a shoot at the top of the protocorm and then develop into a plantlet ( Figure 8D). Paphiopedilum seeds require a precise mechanism to determine the timing of the germination. Based on this fact, there are two approaches for improving the current conservation strategies. The first one is to figure out what causes the growth arrest at the early protocorm stages. The second approach is to identify the major germination inhibition factors and then focus on reducing the barrier. Germination gives rise to a tubercle structure called a protocorm, from which the histodifferentiation occurs and the shoot apical meristem subsequently differentiates. The high level of ABA inhibits the germination of immature seeds. Once the lignification rises [25], the ABA level starts to decline. Figure 8. Protocorm and plantlet formation from immature and mature seeds of P. armeniacum. Protocorm was formed from immature seeds harvested at 73 DAP after 60 days cultivation (A), which failed to develop further and turned brownish after 160 days cultivation (A1). Protocorm was formed from immature seeds harvested at 94 DAP after 60 days cultivation (B), which enlarged slowly and failed to develop shoots after 140 days cultivation (B1). Protocorm was formed from mature seeds harvested at 124 DAP after 60 days cultivation (C), which rapidly enlarged and developed a shoot on top after 70 days of cultivation (C1). The plantlet was able to grow properly from mature seeds harvested at 124 DAP (D). The over-mature seeds harvested at 157 DAP failed to germinate due to the physical restriction imposed by lignin accumulation (E). P1: growth-arrested protocorm; P2: properly developed protocorm; S: shoot formed from the top of the protocorm; Scale bar = 0.5 mm. Many studies demonstrated that ABA has the effect of inducing and maintaining seed dormancy, and GAs can stimulate seed dormancy release and germination [14,15,31,32]. In this study, the levels of four plant hormones (ABA, GA 3 , IAA, and TZR) were determined at five different developmental stages of seed development. Our results revealed that the ABA content increased as the seed approach maturation, peaking at 14.6 ng/g. The dynamic change of ABA content in P. armeniacum seed was similar to that of Rosa hybrida, Pinus taeda, and other species [33,34]. There is the possibility that the high level of ABA in the early stage can inhibit the premature germination of seeds [35]. To confirm this, exogenous ABA application was performed, and the results showed that exogenous ABA treatment inhibited immature seed germination ( Figure 5). The ABA level declined after lignification was accomplished at 157 DAP, with a significant reduction in germination rate. Analysis of the relative quantification of genes showed that the content of ABA would be reduced in mature seeds of P. armeniacum due to the increased expression of ABA catabolized genes. This is different from the relatively high ABA content found in other mature orchid seeds, such as Cypripedium formosanum and Cypripedium japonicum [19,20]. The question is whether ABA inhibits the mature seed germination or not. The ABA biosynthesis inhibitor fluoridone was applied to the capsules once a week for a month, and the hormone analysis confirmed that the treatment effectively reduced the ABA contents from 9.4 ng/g to 2.8 ng/g. No significant differences were seen in the germination rate between the mature seeds that contained a low ABA level (2.8 ng/g) and those that contained a high ABA level (9.4 ng/g), as shown in Figure 6. This suggests that ABA does not play a significant role in repressing mature seed germination. Mature orchid seeds generally have a heavily lignified seed coat that provides protection to the embryo. Several studies attributed lignin accumulation to germination inhibition of mature seeds [21,25]. Instead of ABA, the heavily lignified seed coat formed in the mature seeds may serve as the major inhibitor for germination. Plant Material and Seeds Collection P. armeniacum were maintained in the greenhouse of south China botanical garden under controlled conditions with a temperature of 15~35 • C, humidity between 65%~95%, and light intensity less than 400 µmol m −2 s −1 . The flowers were hand-pollinated and labeled from March to April 2019. Capsules were collected for the seed morphological measurement, asymbiotic seed germination, and hormones measurement experiments. Morphological Characterization of Seeds and Embryos For the morphological investigation, seeds were observed under stereomicroscope (Nikon SMZ745T, Tokyo, Japan) and optical microscope (Nikon E200MV, Tokyo, Japan). In order to collect information on the embryo features, the seeds were treated with 1% sodium hypochlorite for 1 h to loosen the rigid seed coat structure. To collect detailed information about seed surface morphology characteristics, the seeds were prepared for scan electron microscopy (SEM). The seeds were mounted on stubs and coated with gold palladium in a sputter-coater JEE-420, and examined with a JSM-6360LV scanning electron microscope (JEOL, Tokyo, Japan) with a filament voltage of 15 kV. The dimensions (length and width) were measured on at least 30 seeds for every sample. Qualitative data such as seed shape, morphology, and orientation of testa cells were analyzed, and representative images were recorded. Seed Asymbiotic Germination Assay Seed capsules collected at 73, 94, 115, 136, and 157 DAP were surface sterilized by dipping into 75% (v/v) ethanol for 2 min, agitated for 15 min in 1 g l −1 mercuric chloride and 0.05% (w/v) Tween 20, and finally rinsed four times with sterile distilled water. The surface-disinfected capsules were cut open longitudinally, and the seeds were scooped out with sterile forceps and placed on germinating medium. Hyponex N026 medium was supplemented with 1.5 g/L activated charcoal, 2 g/L peptone, 15 g/L sucrose, and 5% coconut water were used for P. armeniacum germination as described preciously [21]. Germination was recognized when the embryo swelled and the testa ruptured. The germination rate is calculated from three independently grown biological replicates. Each biological replicate contained seeds collected from 10 capsules collected at the same time interval, germinated in 10 culture flasks. Measurement of the Endogenous Plant Hormones Levels Endogenous levels of ABA, GA 3 , IAA, and TZR were determined as described by [36]. Freshly harvested seeds (100 mg) at 73, 94, 115, 129, and 150 DAP were ground in liquid nitrogen, and the extraction was performed with acetonitrile solution. The QuEChERS method was used to purify the impurities. The nitrogen purge method was used to concentrate the sample. High Performance Liquid Chromatography-Mass Spectrometry (HPLC-MS/MS, SCIEX-6500Qtrap, ABI, USA) analysis was used to detect the hormone levels. The HPLC-MS/MS was equipped with a poroshell 120 SB-C18 column (2.1 mm × 150 mm). The individual hormones and their metabolites are quantified by comparison of the measured response ratio of endogenous hormone to its internal standard and the ratio of hormone of known concentration to internal standard. The hormone content was expressed as ng g −1 fresh weight (FW). Treatment with Exogenous ABA To examine the effect of exogenous ABA on seed germination, 1 mL of 0.5 mg/mL ABA solution were added in the 100 mL germination medium. ABA solution was prepared with sterile water and filtered with 0.22 µm membrane. Treatment with Fluridone Ninety-four DAP seed capsules were treated with fluridone as previously described [20]. Twenty seed capsules were sprayed with 100 mL fluridone 10 µM, once a week for 1 month. After 30 days, the 124 DAP seeds were collected for germination, and ABA was analyzed as described in Sections 4.3 and 4.4. RNA Extraction, De Novo Assembly, Functional Annotation of Unigenes, and Analysis of Differential Genes Expression One-hundred milligrams of seeds were used for RNA-seq analysis. The method for total RNA extraction was described previously [25]. The RNA purity, concentration and integrity were evaluated using agarose gel electrophoresis, Nanodrop One (Nanodrop Technologies Inc., Wilmington, DE, USA), and Agilent 2100 (Agilent Technologies Inc., Palo Alto, CA, USA). See Figure S2 for the quality of the RNA used for the RNA seq analysis. The analysis of expression level of related genes was based on the previous sequencing results of the seed transcriptome of P. armeniacum, under the BioProject PRJNA550294 in the National Center for Biotechnology Information (NCBI). The transcripts were assembled using Trinity v2.4.0 program with default parameters, and gene expression was estimated by applying the fragments per kilobase per million mapped reads (FPKM). The unigenes were searched against public databases, including the non-redundant protein database (NR), Swiss-Prot, Clusters of Orthologous Groups of proteins (COG), Gene Ontology (GO), and the Kyoto Encyclopedia of Genes and Genomes Pathway (KEGG). The libraries were constructed from seeds at five different developmental stages (66 DAP, 87 DAP, 108 DAP, 122 DAP, and 150 DAP). Pairwise comparison was performed, and DESeq2 was used to analyze the differentially expressed genes (DEGs). In the process of detecting DEGs, \|log2(Fold Change)\| ≥ 1 and FDR < 0.05 are used as screening criteria. Expression level from 108 DAP was taken as the middle point, and the DEGs obtained from the two comparisons (66 vs. 108, 108 vs. 150) of high overall expression were selected as a representative in the Figure 4. Statistical Analysis Statistical analysis was performed using Student's t-test as indicated in the figure legends. All experiments were performed on at least three independently grown biological replicates. All values represent the mean ± SD. Conclusions Seed germination is a complex process controlled by multiple factors. Both chemical and physical inhibition play important roles in seed germination. The accumulation of endogenous ABA in immature P. armeniacum seeds may play a critical role in germination inhibition. Once the seed coat begins to lignify, the ABA level declines and the heavily lignified seed coat becomes the major germination inhibitor instead. This information present here will provide theoretical guidance for large scale propagation of P. armeniacum.	2020-12-17T09:11:54.668Z	2020-12-01T00:00:00.000	{ "year": 2020, "sha1": "b800534bc699501441d80bb8d977ae5282c19734", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/1422-0067/21/24/9561/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "da02ee9a02728131001393cfdf66cdbb9fcfd60d", "s2fieldsofstudy": [ "Agricultural And Food Sciences" ], "extfieldsofstudy": [ "Medicine", "Biology" ] }
258461258	pes2o/s2orc	v3-fos-license	A stochastic cellular automaton model of culture formation We introduce a stochastic cellular automaton as a model for culture and border formation. The model can be conceptualized as a game where the expansion rate of cultures is quantified in terms of their area and perimeter in such a way that approximately geometrically round cultures get a competitive advantage. We first analyse the model with periodic boundary conditions, where we study how the model can end up in a fixed state, i.e. freezes. Then we implement the model on the European geography with mountains and rivers. We see how the model reproduces some qualitative features of European culture formation, namely that rivers and mountains are more frequently borders between cultures, mountainous regions tend to have higher cultural diversity and the central European plain has less clear cultural borders. I. INTRODUCTION The topic of border formation between nations or cultures is complex and it has been subject of interdisciplinary discussion for centuries [1,Chapter 2]. The role of natural boundaries, such as rivers and mountains, in the formation of borders was the basis of many early thoughts on borders [2, p.22-23], but has since been under criticism in the academic literature [3]. Simultaneously, the statistical physics of social dynamics aims at making simple models of complex social phenomena to capture some, but not all aspects of the phenomena. This point of view has with some success been applied to areas such as traffic, networks, economics [4,Chapter 1]. In particular, models of culture and language inspired by statistical physics have been intensively studied in recent decades. Some of the most studied examples are the voter model [5], the Axelrod model [6] along with many others (see for example the review [4]). In this paper, we apply the methods of statistical physics to another complex topic of the social sciences: The problem of border formation. We show that a simple model taking only the locations of seas, rivers and mountains as input can reproduce significant features of actual border locations. The model is a stochastic cellular automaton with coarsening dynamics constructed such that approximately round cultures spread faster. It is inspired by yet substantially different from the agent-based model from [7] which took spreading of information as a starting. Instead, the inspiration for our model stems from ideas in popular culture about the role of geography and military power in border formation [8]. For instance, the popular strategic games Risk [9] and Civilization [10] which involve using armies to conquer territory. * klausen@math.ku.dk † alaurits@ist.ac.at Thus, the model becomes a concrete mechanisation of 19th century naturalistic thoughts on borders where power and natural geography played a central role (although a major difference is that we present a probabilistic rather than deterministic point of view). The failure of naturalistic models (such as the one considered in this paper) in describing more than just some overall probabilistic correlations may well be used as yet another argument against the 19th century naturalistic point of view that was implicitly build to fuel contemporary imperialistic agendas [1,Chapter 3]. With these objections in mind we nevertheless construct a simple model for borders that take only natural boundaries and power into account. In particular, we ignore the myriad of other factors such as climate, diseases, individuals, cities, trade routes, taxation, technology, natural disasters, religion, crops, ideologies etc. Adding model features to account for these factors would come at the expense of the simplicity of the model and we believe they may be better dealt with using other approaches. Over the last 25 years, stochastic cellular automata have been used to model similar spacial forms and reproduce dynamic spatial behaviour [11]. Examples of the approach include urban growth [12,13] and forest fires [14]. We study our model both both with periodic boundary conditions (i.e. on a torus) with no geographical features and a map of Europe with the geographical features of rivers and mountains. On the European map we compare our model to historical data of border locations in the years 1200-1790 from [15][16][17]. Although we compare outcomes of our model to historical data, we emphasise that the purpose of the model is not to predict the actual borders of Europe, but rather to demonstrate a (computable, random) process capable of forming reasonable cultural borders. Since the current European borders to some extent reflect the cultural/linguistic boundaries our model is in turn a model of border formation. This is also the reason we focus on the European map as it could be argued that it is the region where cultural identities and political borders are the most interrelated [18]. II. MODEL: THE GAME OF EUROPE We consider a 200×200 pixelated map of either a torus (meaning with periodic boundary conditions) or of Europe, which consists of 40000 cells. A country is a set of cells and we assign each (land) cell to a country, see for example Figure 1. The countries should not necessarily be thought of as real countries, but could just as well be interpreted as cultures, tribes, or even ideas. One can envision our model as a game of Europe where countries with varying power compete against each other using armies. We emphasise again that we only use this terminology to reflect our inspiration and for clarity of presentation. At each time step, every cell is updated based on interactions between neighbouring countries. To define the interactions we first need to define the powers of the different countries. A naive approach is to define the power of a country to be its area divided by its perimeter. This is motivated by the following: 1. The size of the army a country has is roughly proportional to its population, which again is roughly proportional to its area. 2. This army should be positioned in the border regions -having size the perimeter, as this is where the interaction with neighbouring countries takes place. The validity of point 1. above can well be criticised. See for instance the reference [15], where the size of an army is argued to be given by the wealth of the corresponding country, or [19], where fertility is considered. The model presented here, though more naive, is much simpler, since we do not need any a priori information of wealth of different areas or some other descriptor of population, the size of an army, etc. If we define the power of a country to be its area divided by its perimeter the power has dimensions of length, and thus large countries are stronger (more powerful). This leads to an uninteresting dynamics, where one country quickly dominates. For a more interesting dynamics we therefore choose the power of a country to be some dimensionless quantity. Denoting the area of country C by α C and its perimeter by π C we define its power as There are many ways of constructing such a dimensionless power, e.g. α C /π 2 C is another possible choice. Equation (1) is a natural choice and leads to a more local dynamics as will be explained below. More precisely a cell is a pair of integers (i, j) and the area of a country is simply defined as the number of cells the country consists of. The perimeter is defined as the number of cells (i, j) owned by the country such that at least one of the cells in the 3 × 3 square around the cell (i, j) belongs to a different country [20]. For the "battle" of a cell (i, j) we need to define the neighbours of (i, j). All cells (i , j ) such that the centre of (i , j ) is closer to (i, j) than some number R are considered the neighbours of (i, j). See Figure 2 for an illustration. More formally, the relation is that We call R ∈ [1, ∞) the radius of influence. Countries occupying more of the neighbouring cells are stronger, as they have more of their army in the neighbourhood. Denote by N C (i, j) the number of neighbours cell (i, j) of country C. The local power of the country C is its (global) power times its number of neighbourhood cells. Finally, we add some randomness to the dynamics. This is controlled by the fluctuation p ∈ [0, 1]. Each country's power at the cell (i, j) is multiplied by independent fluctuations uniformly in (1 − p, 1 + p). We denote the (random) fluctuation of country C in cell (i, j) by Φ C (i, j). (All Φ C (i, j) are independent.) In total, the local power of country C in the battle of cell (i, j) is given by . The country with the highest local power at cell (i, j) wins the battle and conquers the cell (i, j) (or, defends, if the country already occupied cell (i, j)) [21]. The dynamics is computed at all cells simultaneously. After each time step the map is updated accordingly, and the process repeats. We can now describe how the choice of the power in Equation (1) gives a more local dynamics. The other relevant choice of the power Π C is the square of the choice in Equation (1). If a cell has N A neighbours of country A and N B of country B the local powers (without the fluctuation) are Π A N A and Π B N B , whereas for the other choice of the power it would be Π 2 A N A and Π 2 B N B . For the second choice the global powers Π A and Π B matter more for the battle of some cell, while for the first (Equation (1)) more weight is put on the local effect of how many neighbours N A and N B the cell has of the different competing countries. On the map of Europe, the above-described dynamics is modified by geographic parameters. Further, we have to describe boundary conditions. These geographic parameters and the boundary conditions are explained in Section IV. On the torus (meaning the 200 × 200 cell grid has periodic boundary conditions) there are neither geographic features nor boundary conditions and it is thereby a model with no geography. III. FREEZING TRANSITION ON THE TORUS The model presented above is a Markov chain on the set of configurations. As such, any realisation of the dynamics will for large enough time end up in some irreducible recurrence class [22, Theorem 1.40] [23]. Clearly, any configuration of only one country occupying all cells is an irreducible recurrence class (of just one configuration) and one can ask the question if these are the only ones. To investigate this we define a configuration to be freezing if it is a recurrence class. That is, the probability of staying in a freezing configuration is 1 and as such the dynamics "freezes" if it reaches such a configuration. In particular, the one-country configurations are freezing. For small fluctuations p, there may be other freezing configurations, but for p large there are none. Even, for large enough p (p = 1 is trivially large enough) any other configuration is transient, meaning that, with probability 1, the dynamics eventually ends up in a one-country configuration. Thus there exists a smallest p, the critical p c , such that for any p > p c the only irreducible recurrence classes are the one-country configurations. A necessary condition for the only irreducible recurrence classes being the one-country configurations is that no other configuration is freezing. We calculate here the minimal value p tile for a tiling configuration (meaning all countries are k × k squares for a k > 2R, see Figure 3) to be non-freezing. Trivially then p c ≥ p tile . It is a simple calculation to see that for a tiling configuration the cells that are easiest to conquer are the corners. All countries have the same power (they all have the same shape) and thus we need for such a tiling configuration to be non-freezing. Here N D (N A ) denotes the number of neighbours of the corner · · · · · · · · · · · · · · · · · · . . . of the defending (respectively strongest attacking) country and we suppress in the notation the dependence on the cell (i, j). A counting argument (see Figure 3) finds where N R denotes the number of cells in the neighbourhood corresponding to radius of influence R (see Figure 2). Thus, Equation (2) gives For large R we have N R πR 2 and so p tile 2/πR. For small R the values of p tile are given in Table I. Finally, an empirical observation is that for sufficiently large R the shapes of the countries in the simulations turn out to be approximately hexagonal after many time steps. A configuration of hexagonally shaped countries is however non-freezing for p smaller than p tile , but it does signify a metastable configuration. IV. GEOGRAPHICAL FEATURES We now describe the geographical features of Europe taken into account in the model. This differs from the work [7], where only the coastline is incorporated. The parameters of the model are summarised in Table II. Coasts and boundary conditions: We implement the sea as cells that cannot be conquered. Further, if a cell is on the perimeter of a country only because it borders the sea (meaning its 3 × 3 square [20] contains only sea and its own country) we add to the perimeter the sea-border parameter s ∈ [0, 1] instead of 1. The parameter s models how much easier it is to defend a country which has the sea instead of neighbours. In all of the following we set s = 0.5 we therefore do not include it as a parameter. Similarly, for the boundaries of the map we choose the same parameter s -our main goal with penalising the boundaries (which are beyond the Ural, Kaukasus and Sahara) is to make sure that the boundary conditions do not influence the centre of the map. Rivers are like the sea implemented as cells that cannot be conquered and give the same effect to the perimeter as the sea, through the parameter s (that we set to 0.5). Additionally, if a country occupies both sides of a river, it gets a bonus to its area through the parameter A r ≥ 0 and the river no longer counts as a border. Technically this is implemented as follows. Suppose cell (i, j) is part of the river. If all cells in the 3 × 3 square around cell (i, j) [20] are either part of the river, or belong to the same country, then that country gets a bonus to its area of A r . The parameter A r is supposed to capture the effect that rivers work as trade routes as well as giving benefits to fertility and infrastructure, as many civilisations arose around rivers [24, p.1]. The location of the implemented rivers is seen in Figure 4. Mountains are implemented by a variable m(i, j) ∈ [0, 1] at each cell (i, j) indicating how mountainous the cell is. Mountains have two effects. The first effect is that when calculating the local powers at cell (i, j) the country already owning the cell gets a bonus parametrised by the parameter D m > 1, i.e. for the defending country D its local power is Thus a "fully mountainous" region with m(i, j) = 1 gets the defensive bonus D m . The second effect is that the perimeter contribution of cell (i, j) is weighted by where P m ∈ [0, 1] is a parameter. Here a "fully mountainous" region with m(i, j) = 1 contributes only P m to the perimeter. The parameters D m , P m model how much easier it is to defend mountainous territory. The parameter D m models a local effect -it is easier to defend, while P m models a global effect -the army of the country may be stationed elsewhere. The values of m are plotted in Figure 4. V. METHODS In this section, we briefly describe the sources of data and how we prepare the data. The mountainous parameter is defined via the elevation data from the GMTED2010 dataset [25,26]. Major rivers in the EU are from the WISE Large rivers and large lakes dataset [27] and rivers Don, Volga, and Ural are from [28]. All data is changed to the Lambert Azimutal Equal Area projection using QGIS [29]. From the elevation data, the value m is calculated as follows. For each cell, the mean Figure 4. The code and datasets are available at Github [30] and ISTA [31] [32]. The data of historical borders are from [16,17], which in turn is based on [15]. The data consists of maps of Europe divided into states at 5 year intervals in the period 1200-1790 (i.e. the years 1200, 1205, 1210 and so on). What exactly constitutes a "state" is discussed in [15], where also the dataset is described in more detail. The dataset does not cover the full extent of the map we use, in particular for the earlier years. It does however cover most of Europe (apart from the earlier years, where northern Norway, Sweden and Finland and Russia east and immediately west of the Urals is not covered), see Figure 8. The maps of [15][16][17] are changed to the Lambert Azimuthal Equal Area projection using QGIS [29] and rasterized to the desired resolution using the python package "geocube" [33]. By lowering the resolution of the data to the desired 200 × 200 grid, some of the smaller countries disappear, as their area is smaller than one cell. VI. RESULTS Simulations on the torus: To gauge the effect of the parameters p, R on the torus we perform simulations for various values of p and R. As a starting point for the investigations of the parameters we choose (p, R) = (0.2, 4), which were found to give reasonable dynamics on the European map. Interestingly, the choice (p, R) = (0.2, 4) is just above the parameters of freezing for the torus (see Section III), Fluctuations: In Figures 5(a) and 5(c) we vary the fluctuations and plot the average country size as a function of time respectively the fluctuation. In Figure 5(c) we see that at time step number 1000 the average country size has a maximum (in p) around p = 0.2. In particular, the model is non-monotone in the fluctuation p and as such, p cannot be interpreted as an effective temperature of the model. Radius: For completeness, we study also the effect of the radius. These findings are shown in Figure 10 in the Appendix. We see that the average size of countries is monotone increasing in the radius of influence R. Simulations on the map of Europe: In this section, we vary the parameters to determine their effects in the European geography. As a starting point, we again choose (p, R) = (0.2, 4). Fixing R = 4 the value p = 0.2 is close to the value of fastest growth on the torus. On the European map, however, the choice (p, R) = (0.2, 4) is very slow evolving and barely above frozen (see Figure 5(b)). As for the geographical parameters (A r , D m , P m ) we investigate their effects in the Appendix and find that reasonable choices are (A r , D m , P m ) = (8, 2, 0.5). This leads us to the standard choice of values (p, R, A r , D m , P m ) = (0.2, 4, 8, 2, 0.5) reported in Table II that we use for most of our investigations. Fluctuations: In Figures 5(b) and 5(c) we vary the fluctuation size and plot the average country area as a function of time respectively the fluctuation. As for the torus we see a non-monotone dependence of the average country size in the fluctuation p. For small fluctuation p (i.e. below 0.8) larger fluctuation leads to a larger average size at time step 1000, but for large fluctuation sizes the dynamics reverses. Interestingly, the fluctuation p giving the maximal growth is quite different for the European map compared to that of the torus. This may be understood as follows. On the map of Europe, some small island and peninsula states exist forcing the average country size to be small, even if mainland Europe is split between few large countries. This effect explains why (for large times) the average country size on the map of Europe is much smaller than that on the torus (see Figure 5(c)). This additionally explains the difference in which p's give the largest average sizes. Namely, on the European map, the average country size is essentially given by the reciprocal of the number of such island and peninsula states. The dependence of the number of such states on the fluctuation p is a completely different dynamics than that of the number of countries on the torus. Geographical parameters: In the Appendix, we discuss the effects of the parameters A r , P m and D m and show how the parameters P m and D m contribute in two very different ways. Noticeably the parameter D m is the most influential. The effect of changing the parameter P m in comparison is much smaller. Finally, the average country size is monotone increasing in the parameter A r . VI.1. Findings We finally present the main findings. Namely, that the model, with appropriate parameters, reproduces some of the qualitative features of the cultural borders in Europe. In particular, we find that mountainous regions have a higher frequency of borders, which can be interpreted as higher cultural diversity in mountains. This effect is also present in historical data. Locations of borders: In the following, we say that a land cell corresponding to one country is a border if at least one cell in its 3 × 3 square [20] belongs to a different country. In Figure 6 the frequency for each cell to be a border is shown and in Figure 7 we plot the correlation of the border frequency and the mountainous parameter m. We see that mountain regions are more frequently borders and that larger areas that are flatter tend to have a very low density of borders. In this way the model reproduces the idea of mountains acting as natural borders, which we can also confirm in historical data in Figure 7 (see also the discussion below). Further, this reflects the higher cultural diversity that is often seen in mountainous areas [34]. Inspecting Figure 6 one sees that the frequency of borders close to rivers are significantly increased (compare also with the snapshot Figure 1), an effect observed in current subnational borders in [35]. Historical comparison: We next compare to the historical data from [15][16][17]. In Figure 7 we plot the correlation between border frequencies and the mountainous parameter m also for the historical data from [15][16][17]. The historical borders are of course correlated since the different data points of [15][16][17] are only separated by 5 year intervals. These correlations, in the form of static borders, are the cause of the border frequency being 1 for many cells with low mountain parameters. However, there is no inherent reason to use one year over another. We thus plot in Figure 7 the average border frequency of the historical data and compare the simulated data to these. As with the simulated data, we see a corre- [15][16][17]. lation between border frequencies and how mountainous an area is: More mountainous regions are more often borders. The verifies historically the idea of mountains acting as natural borders as we also see for our simulated data. In Figure 8 we plot the border frequencies of the historical data from [15][16][17]. We see that the area that was then the Holy Roman Empire has a very high density of borders. This is because the Holy Roman Empire was not considered a country in the source of the data [15][16][17]. Instead, all the smaller individual states, usually German, that were part of the Holy Roman Empire were considered (a phenomenon sometimes known as Kleinstaaterei [36]). This digression illustrates the point that our model is not a model for borders between territorial states, nor for the concrete historical borders of Europe, but rather for the cultures that might to some extend predate the emergence of the territorial state. In particular, one should not compare the specific border frequencies in Figure 6 from the model with the historical data in Figure 8. Contested areas: Finally, we consider which cells are most contested. For each square we find how often it was conquered. This is then averaged over 20 simulations. The results are shown in Figure 9. We see that the average number of times a cell changed countries is higher on the central European plain than elsewhere, in particular in the mountainous regions in the Alps and Caucasus. This might be connected to the, from one point of view, less clear cultural borders on the central European plain (as can be seen on old historical maps [37,38], although one should be wary of context of such maps). Together with the above discussion on border location we reach the conclusion that not only are mountain regions more often borders, they are also much more stable borders. that each cell was a border region (colour gradient) using averaged versions of the historic maps from [15][16][17] discussed in Section V. In particular, the dataset from [15][16][17] does not define the Holy Roman Empire as a country. That leads to a high density of historic borders in central Europe. Figure 9. Heatmap of the average number of times each cell has been conquered before time step 1000 (colour gradient). The scale is cut-off at 30 as some squares have been conquered much more than 30 times. The parameters are chosen to be p = 0.2, Ar = 8, Dm = 2, Pm = 0.5 and the result is averaged over 20 simulations. VII. DISCUSSION AND OUTLOOK We have seen that for suitable choices of the parameters, the model efficiently reproduces many of the features of historical political and cultural borders: There is a high density of borders in mountain regions and along rivers and clear/stable borders in mountain regions and unclear/unstable borders on the large central European plain. We underline that this is another way (popularised in [8]) of approaching the problem of culture spreading than what was done in [7]. Although the model is designed in such a way borders in mountainous regions and along rivers should occur more often, the results show a (somewhat realistic) probabilistic model capturing the naturalistic approach to border formation exists. The model presented is an attempt to take the geography into account in a simple way. As discussed one could imagine adding additional geographical features, but this would come at the expense of simplicity of the model. Another point of discussion in regards to the model is that we have left out the possibility of new "countries/cultures" to form. With such a possibility one might be able to get a continual dynamics instead of our coarsening dynamics and then study the steady state as was done in [7]. One potential way to incorporate the possibility of new countries forming could entail a probability for fracturing of large countries every time step. However, this would introduce more parameters and thus also come at the expense of simplicity of the model. Additionally, many models of our inspiration in statistical physics (see the review [4]) do not include the possibility of new countries, parties etc. forming. Thus our model could still function as a starting point for such investigations. ACKNOWLEDGMENTS Thanks to Kim Sneppen, Svend Krøjer, Peter Wildemann, Peter Rasmussen and Kent Baekgaard Lauritsen for discussions and suggestions. FRK acknowledges support from the Villum Foundation for support through the QMATH center of Excellence (Grant No. 10059) and the Villum Young Investigator (Grant No. 25452) programs. Appendix: Additional plots Dependence of country size on radius R In Figure 10 we plot the average size of countries for simulations on the torus for various values of the radius of influence R. As one could expect, the average size of countries increases in as R increases. Effectively countries much smaller than R don't have enough local power to defend themselves and get conquered. This leads to a larger average country size for large R. The results were averaged over 20 iterations and the shaded regions are between the 5% and 95% quantiles. Influence of the parameter Ar To study the effect of the parameter A r we plot in Figure 11 the average country size against time for different values of the parameter A r . We see that countries tend to get larger (and larger quicker) when the parameter A r is large. This is the behaviour one would expect, as increased A r gives countries around rivers more military strength to expand and get larger. Figure 11. Plot of the averaged country size over time for different value of the river area bonus Ar. Here the parameters were p = 0.2, R = 4, Dm = 2, Pm = 0.6 instead of the normal choice. The results were averaged over 20 simulations and the shaded regions are between the 5% and 95% quantiles. Comparison of the two mountain parameters We investigate whether the two mountain parameters have different effects. To do this we plot the frequency of borders compared to the mountainous parameter m. In Figure 12(a) we make a plot as in Figure 7. In Figure 12(b) we bin all cells into 4 bins depending on the value m. Judging from Figure 12(a) it looks like only the mountain defence parameter has a significant effect on the model, but in Figure 12(b) we see that the two mountain parameters have different effects.	2023-05-04T06:28:51.343Z	2023-05-03T00:00:00.000	{ "year": 2023, "sha1": "2561aa6ee6e1f43698dc8731a37e2870240c859e", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "2561aa6ee6e1f43698dc8731a37e2870240c859e", "s2fieldsofstudy": [ "Computer Science" ], "extfieldsofstudy": [ "Physics", "Computer Science" ] }
119085454	pes2o/s2orc	v3-fos-license	PLCK G165.7+67.0: Analysis of a Massive Lensing Cluster in a Hubble Space Telescope Census of Submillimeter Giant Arcs Selected Using Planck/Herschel We present Hubble Space Telescope WFC3-IR imaging in the fields of six apparently bright dusty star-forming galaxies (DSFGs) at $z$ = 2-4 identified by their rest-frame far-infrared colors using the Planck and Herschel space facilities. We detect near-infrared counterparts for all six submillimeter sources, allowing us to undertake strong-lensing analyses. One field in particular stands out for its prominent giant arcs, PLCK G165.7+67.0 (G165). After combining the color and morphological information, we identify 11 sets of image multiplicities in this one field. We construct a strong-lensing model constrained by this lensing evidence, which uncovers a bimodal spatial mass distribution, and from which we measure a mass of $(2.6 \pm 0.11)$ $\times$ $10^{14}$ $M_{\odot}$ within $\sim$250 kpc. The bright ($S_{350}$ $\approx$ 750 mJy) DSFG appears as two images: a giant arc with a spatial extent of 4.5"that is merging with the critical curve, and a lower-magnification counterimage that is detected in our new longer-wavelength ground- and space-based imaging data. Using our ground-based spectroscopy, we calculate a dynamical mass of $1.3^{+0.04}_{-0.70} \times 10^{15}$ $M_{\odot}$ to the same fixed radius, although this value may be inflated relative to the true value if the velocity distribution is enhanced in the line-of-sight direction. We suggest that the bimodal mass taken in combination with the weak X-ray flux and low SZ decrement may be explained as a pre-merger for which the intracluster gas is diluted along the line of sight, while the integrated surface mass density is supercritical to strong-lensing effects. 1. INTRODUCTION Clusters of galaxies with masses ∼10 15 M are extremely useful but rare tracers of the distribution of mass in the universe (Bahcall 1977;Mo & White 1996). Finding galaxy clusters, and establishing cluster properties and cluster scaling relations, are fundamental to cosmology studies (Vikhlinin et al. 2009;Mantz et al. 2010;Rozo et al. 2010;Allen et al. 2011;Benson et al. 2013;Hasselfield et al. 2013;Planck Collaboration et al. 2014). As ensembles of discrete galaxies, they can be discovered in optical and near-infrared wide-area surveys such as the Sloan Digital Sky Survey (SDSS) (i.e., Koester et al. 2007a,b;Rykoff et al. 2014Rykoff et al. , 2016. Although originally discovered at optical wavelengths, galaxy clusters with masses of 1 -10 × 10 15 M will almost always contain a massive component of hot intracluster gas which makes them distinct X-ray sources. This reservoir of hot baryons is a salient feature of massive clusters, as there is no physical mechanism to dissipate it. To take advantage of this requisite feature, the ROSAT archives offer the all-sky advantage to efficiently detect the most extreme sources of X-ray emission produced by thermal bremsstrahlung and line emission in the intra-cluster gas (Rosati et al. 1998;Ebeling et al. 2007Ebeling et al. , 2010. A galaxy cluster bound by gravity also has a distinct signature at radio wavelengths. This is because the same large reservoirs of intra-cluster gas which give rise to the X-ray flux also distort the cosmic microwave background radiation (CMB) by inverse Compton scattering. From the ground, searches for galaxy clusters by detection of this Sunyaev-Zel'dovich (SZ) effect using the South Pole Telescope (SPT; Carlstrom et al. 2011) yield hundreds of candidates (Bleem et al. 2015). Targeted searches using the Atacama Cosmology Telescope (ACT; Fowler et al. 2007) that employ a similar technique are also successful (Sehgal et al. 2011(Sehgal et al. , 2013. From space, Planck High Frequency Imager (HFI) data is used to extend the search for the SZ decrement to all available extragalactic sky (Lamarreet al. 2003;Planck Collaboration et al. 2016a). To complement the cosmological SZ approach, the detection of galaxy over-densities by the astronomical technique of color-selection has recently been explored. Ultra-bright infrared galaxies are targeted, such as the Dusty Star Forming Galaxies (DSFGs) which produce stars at rates of up to ∼1000 M yr −1 and yield prodigious amounts of dust. This warm dust radiates as a modified blackbody spectrum with a prominent peak in the rest-frame far-infrared. Submillimeter data are well-suited to conduct the color search for the DS-FGs because this wavelength range corresponds to the observed-frame thermal dust peak at redshifts typical of DSFGs of z ≈ 2 -4 (Casey et al. 2014;Planck Collaboration 2015, and reference therein). In this regime, there is the unusual advantage that one records the flux density of the DSFGs closer to the peak of their rest-frame SED as their redshift increases. As a result, the benefit of the high flux density of DSFGs largely compensates for the cosmological dimming (Blain 1999;Planck Collaboration 2015), thereby gaining leverage for the detection of high-redshift objects. A Planck/HFI census was undertaken to find DSFGs by color using the cleanest 26% of the data, or ∼10,000 deg 2 . To be selected by Planck/HFI, the DSFGs had to be separately detected in each of the cleaned 857 GHz and 353 GHz maps, be compact at Planck resolution ( > ∼ 5 ), and have flux density ratios in the 353, 545, and 857 GHz maps consistent with being red and dusty sources (Planck Collaboration 2015). These socalled 'cold' sources in the CMB are extremely rare, with number densities of ∼1 per tens of square degrees, requiring the wide-field survey area of Planck/HFI. We flag the brightest 228 of these cold sources as the candidate DSFGs. To classify the sources, follow-up observations are made of this sample higher spatial resolution using the Herschel/SPIRE. Details and initial results are discussed elsewhere (Planck Collaboration 2015. In sum, 15 of the 228 sources are discovered to be individual DSFGs boosted in brightness as a result of gravitational lensing (Planck Collaboration 2015; Cañameras et al. 2015). In this paper, we present new HST imaging and lensing analysis for six of the 15 strongly-lensed Planck/Herschel-selected sources. We focus our study on one particular object in our sample, namely, PLCK G165.7+67.0 (hereafter G165). To better understand the properties of this one field, we acquire multiwavelength imaging and spectroscopic follow-up observations which will be discussed in detail. This paper is organized as follows. In §2 we compare our sample of strongly-lensed DSFGs with others in the literature. In §3 we present new HST imaging data for our sample of six Planck/Herschel-selected strongly-lensed DSFGs. We also present new groundand space-based observations of the high mass cluster, G165. In §4 we describe the data reduction and analysis of the follow-up data obtained for G165. In §5, we construct the strong lensing model for G165. This analysis is followed by a discussion in §6 in which we make independent determinations of its mass, its lensing strength, and the properties of its low inferred cluster gas pressure. In §7 we summarize our results. An Appendix is provided to describe the imaging and lensing analysis of all six fields in our HST sample. We assume throughout a ΛCDM cosmology with H 0 = 67 km s −1 Mpc −1 , Figure 1. STRONGLY-LENSED DSFGS Although the details of the search strategies for strongly-lensed DSFGs differ, most algorithms set a high 350 µm flux density (S 350 ), or a high 500 µm flux density (S 500 ) cut of 100 mJy. To date, dozens of strongly-lensed DSFGs in the redshift range 2 < z < 4 satisfy these criteria. In Figure 1 we assemble the set of lensed DSFGs for the surveys or subsets thereof for which there are Planck S 350 flux densities, spectroscopic redshifts for the lens and the lensed sources, and images of the lensed sources. For each DSFG, the symbol size is proportional to the size of the Einstein radius we estimated from the resolved image of the lensed source. For reference, Einstein radii of 1, 5, and 10 are used, which are typical of individual massive galaxy lenses (M ∼ 10 11 M ), galaxy group lenses (M ∼ 10 13 M ), and galaxy cluster lenses (M ∼ 10 15 M ), respectively. The legend gives the color-coded references. The brightest lensed DSFG, the "Cosmic Eyebrow" (z = 2.0439), stands out for its high submillimeter flux density (S 350 = 1300 mJy). It was found by cross-correlating the sources in the WISE all-sky source catalog "AllWISE," with infrared-bright galaxies in the Planck compact source catalog (brown circle in Figure 1; Díaz-Sánchez et al. 2017). Note that the peak submillimeter flux density of the Cosmic Eyebrow is measured from Planck/HFI data, which has higher uncertainty by a factor of ∼10 than the Herschel/SPIRE photometry used for the other comparison samples in Figure 1. In the redshift range 2 < z < 4, a search for lensed DS-FGs within the Herschel Astrophysical Terahertz Large Area Survey (H-ATLAS) using S 500 as a discriminator yields 22 lensed DSFGs covering a search area of 14.4 deg 2 (Harris et al. 2012;Bussmann et al. 2013;Calanog et al. 2014;Negrello et al. 2017). This same approach applied to the Herschel Multi-tiered Extragalactic Survey (HerMES) field extends the areal coverage by a factor of ∼7, resulting in 13 new lensed sources (Wardlow et al. 2013;Bussmann et al. 2013;Calanog et al. 2014;Nayyeri et al. 2016). By applying similar flux density cuts to the Herschel Stripe 82 Survey (Hers82), an additional three lensed DSFGs are found (Nayyeri et al. 2016). From the ground, South Pole Telescope (SPT) data enables the selection of strongly-lensed DSFGs based on the ratio of flux densities at 1.4 and 2.0 mm S 1.4mm /S 2mm , which are consistent with thermal emission by dusty galaxies (Vieira et al. 2010;Carlstrom et al. 2011). The brightest sources in the sample are then followed-up at higher resolution using most notably the SubMillimeter Array (SMA) and the Atacama Large Millimeter Array (ALMA). A total of 26 stronglylensed DSFGs are identified, which tend toward higher redshifts owing to their selection at longer wavelengths, and whose identifications are typically explained as galaxy-galaxy lensing events (Weiß et al. 2013;Vieira et al. 2013). The Planck collaboration used Planck/HFI data to extend the search for lensed DSFGs to all available sky. This enables the detection of the most rare and extreme sources picked out by the detection of the thermal dust peak. To focus detection on only the most extreme sources, a strict flux density criterion is imposed that exceeds 500 mJy at 545 GHz. The expectation is that the brightest sources that are also compact at Planck resolution will be too faint to be explained by a single field DSFG. These sources are most likely: (1) multiple DSFGs, or (2) a single strongly-lensed DSFG. The latter case would flag the presence of a large foreground mass. On follow-up at higher resolution using Herschel/SPIRE, the vast majority of sources resolve out into clumps of several submillimeter bright objects in Cañameras et al. (2015). The images of the individual DSFGs are strongly-lensed into giant arcs and are multiply-imaged in four fields: G145, G165, G045 and G080. For G092, the two NIR counterparts are not consistent with being counter-images despite having similar colors, and are more likely to be separate, possibly interacting submillimeter bright galaxies at a similar source redshift (labeled as '1a' and '1b'). For G244, the DSFG is a 1. 4 partial Einstein ring that is resolved in the ALMA imaging (Cañameras et al. 2017a,b). This arc appears in the NIR as an extended source, which is blended with the main lensing galaxy such that giant arcs are not detected in these HST data. The cutout boxes in the upper right corners show the smoothed image (for G080), the single band image (for G145), and the galaxy halo subtracted image (for G092), which more easily show the red, low surface brightness arcs. In each panel, north is in the direction of the compass arrow and east is counter-clockwise with respect to the compass arrow. A 5 scale bar is shown in the bottom left corner for reference. close projected proximity (Planck Collaboration 2015. These are candidate galaxy over-dense regions which are potentially the high redshift predecessors of massive lensing clusters at lower redshifts (Planck Collaboration 2015;Flores-Cacho et al. 2016;Martinache et al. 2018;Kneissl et al. 2018). At the same time, a small minority of 15 of 228 sources remained isolated, while also meeting additional flux density thresholds of S 350 > 300 mJy and/or S 500 > 300 mJy. These sources show signatures of individual DS-FGs that are boosted in brightness as a result of strong lensing. Of these, 11 sources could be followed-up at higher resolution using observing facilities from the northern hemisphere. Spectroscopic measurements of the lens and source redshifts, and identification of giant arc structures, strengthen the lensing interpretation (solid red disks in Figure 1, Cañameras et al. 2015;Nesvadba et al. 2016;Cañameras et al. 2017a,b). The Harrington et al. (2016) sample (black-and-reddashed circles) is closely related to the (Cañameras et al. 2015) sample. Their selection also relies on color using a combination of Planck and Herschel, yet the intersection is incomplete owing to the use of different Planck catalogs. Harrington et al. (2016) select sources by cross correlating Herschel/SPIRE with: Planck PCCS (6 candidates), Planck HerMes (1 candidate), and Planck HerS-82 (1 candidate). The selection of the Planck/Herschel sample (Cañameras et al. 2015) was made by applying color criteria to: Planck PCCS and Herschel/SPIRE (6 candidates), and to Planck OT2 and Herschel/SPIRE (five candidates). In sum, 3 of 8 of the Harrington et al. (2016) lensed DSFGs are new. In sum, there is a tendency for Planck/Herschel selected sources to have higher flux densities and larger Einstein radii than those drawn from the literature. The cluster scale of the lens may partially explain this difference, in that a larger magnification factor (µ) can be achieved, especially in the case of an Einstein ring such that µ ∝ √ M lens , where M lens is the mass of the lens. The wider areal coverage of a factor of ∼10 relative to the SPT and a factor of ∼100 or more relative to H-ATLAS, HerMES and Hers82 surveys also helps by allowing to set higher flux density thresholds resulting in the identification of larger lenses in some cases. OBSERVATIONS AND REDUCTION We present new observations using HST/WFC3-IR for the six fields in our sample. The WFC3/IR images are re-drizzled using the software package DrizzlePac (Fruchter & et al. 2010). We adopt values for the photon-sensitive effective size of a pixel to its real size (final pixfrac), and a final pixel scale (final scale), of 0.85 and 0. 06, respectively. The results of re-drizzling the data in each case show improvements in image quality over the pipeline products of up to 10%. The final reduced images reach comparable depths to the CLASH clusters. For one representative case, G165, we compute 10σ limiting magnitudes of F 110W AB < ∼ 26.9 mag and F 160W AB < ∼ 26.2 mag for point sources inside 0. 4 apertures. We find that the image depth and bandwidth are sufficient to make NIR detections of the strongly-lensed DSFG in each of our sample fields. We also identify other examples of giant arcs and/or image multiplicities in some cases. In Figure 2, we present the HST color images of the central regions for each of the six fields. The WFC3 F 160W images are used as detection images for the matched aperture photometry in both HST bands. We custom-built our own code to cope with the unusual morphologies peculiar to arcs in the central regions of massive lensing clusters. We detect sources by applying a σ-clipping algorithm with respect to the local background RMS values. The local background, in turn, is estimated following a similar approach as in SExtractor (Bertin & Arnouts 1996). Briefly, the image is divided into patches of 100 × 100 pixels, with the background in each patch represented by the σ-clipped median. The local background is then estimated by a smooth 3 × 3 spline interpolation over these patches. To ensure robust detection of objects, we smooth the image with a Gaussian filter of FWHM ∼ 0. 2. Objects are de-blended using the watershed PYTHON algorithm in astropy.photutils (The Astropy Collaboration et al. 2018). Artifacts such as diffraction spikes are visually identified and removed. We assign apertures to each galaxy image by measuring the semi-major/minor axis sizes at 3 times the FWHM lengths of the detected objects, typically amounting to 0. 6. Elliptical annuli are used to get the best estimate of the background, with an inner radius equal to the photometric aperture and an outer radius equal to 1.2 times the inner radius. We compute the aperture flux, and then subtract the area-scaled local background level within the annuli. The flux uncertainty is computed as the quadratic sum of the local smooth background RMS value only. We note that aperture corrections are minimal owing to our large extraction aperture of > 0. 4. In three fields, G145, G045, and G080, the DSFG arclet family members are too faint and/or blended with the halos of bright cluster members near in projection to measure a flux. In these cases, an upper limit on their fluxes is reported. In Table 2 we present the photometric catalog of the lensed DSFGs for our sample. The columns are: Arc ID, RA (J2000), DEC (J2000), F 110W AB and F 160W AB mag, the magnification factor µ of the lensed DSFG as measured by our mass-traces-light lens model, the effective Einstein radius in arcseconds as measured from our lens model, the redshift of the lens, and the redshift of the lensed source. LBT Observations We acquired imaging of G165 in K AB -band using the LBT Advanced Rayleigh Ground layer adaptive Optics System (ARGOS) during instrument commissioning time in December 2016 (Rabien et al. 2018). AR-GOS corrects ground-layer distortions in the imaging of the two 8.4-m apertures using two 3-beam constellation lasers as guide stars that are fixed to each aperture. The ARGOS instrumentation operates through the LUCI imager and multi-slit spectrometer. High-quality corrections of up to FWHM ∼ 0. 25 in K AB -band are achievable across a large field of view (4 × 4 ) at a native pixel scale of 0. 12 pix −1 . We acquired LBT/LUCI + ARGOS data in monocular mode on two separate nights: 46 min of observation using LUCI2 on 9 December and 42 min using LUCI1 on 15 December. We note that the LUCI1 set of observations have a slightly shorter exposure time and, in turn, a slightly higher per-pixel RMS uncertainty. However, they yield a fainter point source detection limit, due to the lower FWHM as measured in isolated and unsaturated stars. We choose to analyze the data separately from each night, and only to combine the photometric measurements at the last step. The background subtraction of our LBT/LUCI + ARGOS KAB-band imaging data is accomplished in an iterative process as illustrated above for a small region near the cluster center. The difference in image flatness between the stacked images after the 1st stage, and the final image upon extending the bright object masks (2nd and 3rd stages), is evident. We sample the background by placing test boxes down that are isolated from bright sources. We find the flatness to improve, or alternatively for the RMS level of the background to decrease, by typically 9% between the first and second stages and to converge after the 3rd stage. In each panel north is down and east is to the right. Random dithers of up to 40 are imposed to optimize the sky-subtraction in the crowded cluster regions and to eliminate detector artifacts. Such large dithers require high point source stability across the field. As a check, we estimate the pointing error at each dither position by stacking the object frames, and then measuring the positional centroids of 13 pre-selected cluster members that span the full field-of-view and that are isolated from bright sources. We find that the typical translational shift between images is ∼2 pix, or ∼0. 25, with negligible rotation. The WCS information in each object frame is updated accordingly. At this point we resample the images onto the same pixel grid using the fluxconserved, overlapping pixel-area method in the Python routine astropy.reprojection, as needed. We designed our own reduction pipeline so that will be optimized to ensure high flatness across the chip and to maximize the signal-to-noise of the data. As a first step, we subtract the dark frames from all object frames. We then proceed to find the best esti-mate of the background. Within a single exposure, the sky-background varies by ∼100 ADUs, comparable to the integrated flux of some of the fainter cluster members. Therefore, instead of creating sky-frames by taking the median at multiple pointings directly, we apply a "normalizing-rescaling" approach to construct master sky-frames taken from neighboring exposures, which are then scaled to the background level of each object frame prior to the subtraction. We designate each dither pointing "i" as the coaddition of 24 individual object frames in 5 s exposures, all taken the same position (120 s total science time) plus ∼0 s readout time owing to non-destructive readouts. The 5 s exposure was chosen to be small to avoid persistence and non-linearity effects. We find that a reasonable compromise in image combination is to collect the temporally closest five dither pointings about each ith dither pointing in a running boxcar, equating to a total clock time of ∼14 min including other overheads. The result of including more dither pointings is a slight improvement in the background noise, but a degradation in image flatness. We mask out the bright sources in all the frames of the running boxcar to avoid biasing the result or "master-background" upward of its true value with unwanted cluster halo light. Before dividing this master frame into the ith dither pointing, we divide the dither pointing frame by its 5-pass, 3-σ clipped median value 1 to obtain the mean image of these "normalized" object frames. As a last step, we scale this new running master background frame to the 3-σ clipped median of this ith dither pointing to match the sky-background level at the exact time of the exposure. Our background-subtracted dither pointings yield the "1st stage" result ( Figure 3). Following an iterative approach, we introduce two additional stages to the background subtraction, each time using the previous stage result as a starting point. The main difference is that we continue to extend the bright object mask into the fainter outskirts of the masked sources. We avoid aggressively expanding the bright source mask, as increasing the number of masked pixels improves the flatness but at the expense of the noise level as fewer frames are available from which to estimate the background. The "3rd Stage" result is obtained by performing another iteration of the "2nd Stage" result. To assess image quality after each reduction stage, we compute the background RMS values inside of seven test boxes of size 50 × 200 pixels located in regions isolated from bright sources. We find the background RMS to decrease on average by 9 % following stage one, and to converge to the < ∼ 1% level following stage two (see Figure 3). ∼20 a The Einstein radius in arcseconds is reported at the redshift of the lensed DSFG. b The spectroscopic redshift comes from this paper. c The spectroscopic redshift comes from Cañameras et al. (2015). d The lensed image is only partially resolved in our HST image ( Figure 2). These values for the magnification factor, the size of the Einstein radius and the redshift are drawn from the ALMA data in Cañameras et al. (2017a,b). e The spectroscopic redshift comes from Harrington et al. (2016). h We find these two images are likely to originate from two different sources at a similar redshift. i The Spectroscopic redshift comes from SDSS DR14 data. To make corrections for pixel-to-pixel variations, we first tried applying a flat-field to the data in the usual way prior to subtracting off the background. The result was unsatisfactory because image artifacts remained in the data. On reversing the order of these two operations, we found an improvement in the image flatness and the removal of image artifacts. This improvement arises because our master flat-field is constructed by combining local sky frames generated as a natural part of our background-subtraction algorithm. We then stacked the sky-subtracted and flat-fielded object frames to produce our final data product. We report a mean K-band FWHM of 0. 53 for the 9 Dec 2016 run (LUCI2), and 0. 29 for the 15 Dec 2016 run (LUCI1). In all, for the 9 Dec 2016 and the 15 Dec 2016 runs, respectively, we reach a 10σ limiting magnitudes of K AB < ∼ 22.6 mag and K AB < ∼ 23.5 mag inside apertures of 4 × FWHM. We do not combine the final images from the two different detectors as the 15 Dec 2016 data has higher spatial res-olution ( Figure 4). We emphasize, however, that our photometry is measured using both nights of data and weighted by an inverse variance-weighted mean of the two fluxes, as described in §4.1. Dozens of arcs appear in this high spatial resolution K-band image, which we use in combination with our HST data to make identifications of single galaxy images that appear in multiple positions in the image plane, or "arclet families." The exquisite quality of our LBT/LUCI + ARGOS images can be seen when combined side-by-side with our HST images as in Figure 5. For more details, see Rabien et al. (2018). Spitzer Observations We acquired imaging in the G165 field on 2 February 2016 using the Spitzer InfraRed Camera (IRAC) in the 3.6 µm and 4.5 µm channels as part of a larger program (Cy13, GO-13024, PI: Yan) to image the fields of massive lensing clusters that would make good tar-1a 1b Figure 4. High resolution K-band image (FWHM ≈ 0. 29) for the central region of G165, using LBT/LUCI + ARGOS. Dozens of strongly-lensed galaxy images and arclet families are detected. The lensed galaxy DSFG 1a is spatiallyresolved (labeled as "1a"). Notably, we identify a new image which we designate as its counter-image, DSFG 1b (labeled as "1b"). See Figure 5 for a 3-band color image which includes these data and our HST data set. North is in the direction of the compass arrow and east is to the left. A 10 scale bar is shown in the bottom left corner for reference. gets for JWST. The on-target exposure time was set to 12,000 seconds in each of the two channels. The Spitzer Science Center (SSC) processed these data using the standard SSC pipeline, and we made the final image mosaics based on these products. A detailed analysis of the full data set of this entire program will appear in an upcoming paper (Yan et al. 2018, in preparation). We also refer the reader to the description in Griffiths et al. (2018), where the reduction of their IRAC data from the same Spitzer program is discussed. In Figure 5, we show the 3.6 µm IRAC mosaic of this field (right panel). The doubly-imaged DSFG, G165 DSFG 1a and G165 DSFG 1b, is stronglydetected in both Spitzer/IRAC channels (S 3.6 and S 4.5 ). MMT Observations We obtained spectroscopy in the field of G165 on 14 Feb 2015 using MMT/Hectospec (Fabricant et al. 2013), as a part of a larger program (2015A; PI: Frye). Hectospec is a multi-fiber spectrograph that assigns optical fibers on the sky with minimum allowed separations of > ∼ 20. To maximize the wavelength range for measuring redshifts, we selected the 270 grooves mm −1 grism, which covers a wavelength range of 3700 − 9150Å at a dispersion of 1.21Å pixel −1 . We chose to position 23 fibers (20 galaxy targets plus 3 standard stars) with priorities set to the positions of the brighter examples of prominent giant arcs and cluster members, as selected by their near infrared (NIR) photometric redshift estimates made using Canada France Hawaii Telescope (CFHT) plus Spitzer/IRAC imaging. We refer to Cañameras et al. (2018) for details on the photometric analysis and photometric redshifts. The observations comprised of a single Hectospec pointing with 7 × 1020 s exposures taken under variable seeing conditions of ∼1 -2. This was sufficient for our science goal given the 1. 5 fiber widths and relatively bright magnitudes of the targets of i AB 18 -22 mag. The reductions proceeded in a standard fashion using the IDL/Hectospec Reduction Software package (HSRED) obtained from the Smithsonian Astrophysical Observatory Telescope Data Center: (https://www.mmto.org/node/536). We removed cosmic rays using the code "LA Cosmic" (van Dokkum 2001). Corrections for pixel-to-pixel variations, fringe corrections, and fiber identifications are accomplished using a dome flat. Background subtractions were made after first averaging together the spectra set to blank sky positions, taken under the same conditions as the science data. The wavelength solution is found in two ways, using both a HeNeAr lamp exposure, as well as the positions of prominent night-sky lines. We coadded the individual exposures to yield the final reduced spectra. As the planning for this observing run took place prior to receiving the HST data set, we were not able to fine tune the target list to include new arclet family members. Secure spectroscopic measurements are made for 19 objects, which we define as the high significance detection of two or more spectral features (>2σ level in the continuum). Our catalog results in measurements for five new cluster members with z = 0.326 -0.376, and eleven new sources with redshifts 0.388 < z < 0.622. Three galaxies have MMT/Hectospec redshifts that place them in the foreground. See §4.2 for additional details and the redshift catalog. Gemini Observations We obtained further spectroscopy in the field of G165 using the Gemini-North multi-object spectrograph (GMOS) as a part of a larger program (GN-2016A-Q-30, PI: Frye). The observations took place on 27 April, 2016. We selected the B600 line mm −1 grating, which has a wavelength coverage measured from our data of a total of 2975Å about the central wavelength for each slitlet at a dispersion of 0.92Å pixel −1 . As we did not have the HST images in time to plan this observing run, we populated the slit masks first with prominent arcs selected from the CFHT image from Cañameras et al. (2015), followed by cluster members selected from our Gemini pre-imaging data. We chose 1 slits to match typical seeing on-site. We acquired six science exposures of 1200 s, two each at central wavelengths of 645 nm, 650 nm, and 655 nm to correct for chip gaps. Arc spectra were obtained within ± 1 night of the observations using the CuAr lamps at similar central wavelengths. Dispersed flat-fields were taken at each of the three central wavelength (and hence grating tilt) configurations. Figure 5. Left: Color image of a central region of the G165 field in HST WFC3-NIR F 110W +F 160W . DSFG 1a appears at the expected location based on the submillimeter imaging (Cañameras et al. 2015). Sharing this crowded central region are three other arclet families which all show fold images about an axis of symmetry, as labeled. Middle: Three-band image of the HST and LBT/LUCI+ ARGOS Kband data. DSFG 1a stands out on account of its red color. Our lens model also predicts for there to be a fainter counterimage, DSFG 1b, which is discovered in the K-band data at its model-predicted location. Right: Spitzer/IRAC imaging at 3.6 microns. DSFG 1a and DSFG 1b are both bright, S3.6,AB≈19.1 mag and S3.6,AB∼19.9 mag, respectively. All images are 10 × 30 on a side and have the same orientation as in Figure 2. The initial calibrations of bias-subtraction and flatfielding proceeded in the standard way using the IRAF Gemini reduction package. 2 We removed cosmic rays prior to the background subtraction using the IRAF task GEMCRSPEC. For the wavelength calibration, there is a tendency for the IRAF algorithm to introduce wavelength offsets of the stacked spatial rows, especially for the smaller spectral "boxes." To avoid introducing this undesirable spatial feature into the data, we switched to using a pipeline written in IDL by one of us (BLF). The IDL pipeline includes the tasks mentioned below and is discussed elsewhere (Frye et al. 2002(Frye et al. , 2007(Frye et al. , 2008. Briefly, the IDL pipeline avoids re-pixelization by identifying the flexure-induced instrumental curvature imprinted onto the slitlet edges between the 2D spectra. This curvature amounts to 1-3 pixel shifts from center to edge of the CCD, which are easily fit by low-order polynomials. We then wavelength-calibrate the data in two ways: using the arc lamps and using the night-sky lines. As both outputs had an RMS on the wavelength fit of <0.5Å, we choose to use the sky lines for the potential benefit that the wavelength references are embedded directly onto the data at the time of the observations. Cosmic ray hits on the object were removed in 1D by comparison of the stacked spectra from the six different exposures using our IDL task SPADD (Frye et al. 2002(Frye et al. , 2007(Frye et al. , 2008. Thresholds are set for acceptable number of cosmic ray hits per pixel in the stack to avoid removal of real spectral features. We measured redshifts for the 1D co-added spectra using our IDL task SPEC (Frye et al. 2002(Frye et al. , 2007(Frye et al. , 2008. Our catalog results in spectroscopic measurements for 32 galaxies in the G165 field. Of these, we find nine cluster members that are new with 0.326 < z < 0.376, and 18 new lensed sources with 0.386 < z < 1.065. Five galaxies have Gemini/GMOS redshifts that place them in the foreground of the lens. ANALYSIS AND RESULTS FOR G165 We describe our algorithms for performing the matched photometry for the HST plus LBT imaging. We then analyze the combined results of the MMT, Gemini, and archival ground-based spectroscopy. The Photometry To include the LBT/LUCI + ARGOS data into our catalog alongside the HST photometry, we first translate the central locations of our photometric apertures defined by the F 160W image onto the K-band using the WCS information. Although the FWHM resolution of our LUCI1 K-band data (FWHM ∼ 0. 29) is higher than that of our two HST bands (0. 22 and 0. 18 for F110W and F160W bands, respectively), we do not alter the aperture sizes and ellipticities, as there is adequate matching to detect the vast majority of the sources. The data from LUCI1 and LUCI2 are obtained under different weather conditions, and the field orientation angles and plate scales are slightly different for LUCI1 and LUCI2. As a result, we opted to conduct K-band photometry separately for LUCI1 and LUCI2 images, and only then to compute the aperture fluxes by applying an inverse variance-weighted mean of the two values. As the photometric depth at K is shallower than for HST data, the aperture fluxes for some sources and arclets fall below their 1-σ uncertainties. In such cases, we report the 10σ detection limit of the aperture fluxes. Table 3 gives the complete photometric catalog for all eleven arclet families. The columns are: Arc ID, RA(J2000), DEC(J2000), F 110W AB mag, F 160W AB mag and K AB mag. The lensed DSFG, G165 DSFG 1a, has F 110W AB = 23.0 +0.2 −0.2 , F 160W AB = 22.2 +0.2 −0.2 , and K AB = 18.9 +0.2 −0.2 , bright enough to make ground-based spectroscopic follow-up feasible. Figure 6. Histogram of available spectroscopic redshifts in the field of G165. The redshift distribution combines the results from our MMT-Hectospec and Gemini-GMOS data sets (green), and objects drawn from the literature (SDSS, DR 14; magenta). We measure a value for the redshift of this lens of z = 0.351 based on 18 cluster members. These cluster members range in position from the central region out to a cluster-centric radius of ≈ 0.8 Mpc. The Spectroscopy: G165 Cluster Members The catalog for all 62 objects in the G165 field with measured redshifts is given in Table 4. The columns are: Source ID, RA (J2000), DEC (J2000), redshift (z), Reference (MMT/Hectospec: H; Gemini/GMOS: G, or SDSS: S), SDSS i AB mag, F 110W AB mag, and F 160W AB mag. Only objects with secure redshifts appear in the table, i.e., we require that two or more spectroscopic features be detected at the > ∼ 2σ level relative to the continuum In the case of a single emission line, we require that it be detected blueward of rest-frame Hα, and also show a continuum break. In total, we now have spectroscopic redshifts for 18 galaxies in the cluster. Of these, six cluster members are drawn from our MMT/Hectospec data, and an additional nine cluster members come from our Gemini/GMOS data. The remaining three cluster members are drawn from the literature (SDSS, DR13). We specify cluster membership by requiring the redshifts to be in the range 0.326 < z < 0.376, equating to ±3σ with re-spect to the mean of z = 0.351. Five galaxies with spectroscopic redshifts extending outside of this range (0.388 ≥ z ≥ 0.399) are not included in this set. This group of more distant objects have a mean redshift of z=0.391 and a velocity dispersion of ≈1600 km s −1 , making them members of a potential background galaxy group. The cluster members in common with the smaller HST field, and with F 110W AB < 20 mag, all fall reasonably well onto the red sequence of the color-magnitude diagram (CMD; see §B.1 in the Appendix). The velocity dispersion based on all the cluster members is σ = 2,600 ± 50 km s −1 , equating to a large dynamical mass of M V = 1.0 ± 0.1 × 10 16 M within 1 Mpc. If we restrict the angular extent to match the scale of our HST observations of θ = 50, or ≈ 250 kpc, then 13 cluster members are removed. The velocity dispersion for this smaller galaxy set is σ = 1,600 +200 −300 km s −1 , equating again to a large mass, M V = 8.8 ± 2.3 × 10 14 M . Note that this mass may be biased upward as a result of a line-of-sight arrangement of mass (see §6.1 for a comparison of the various mass estimates). The uncertainties are obtained by a bootstrapping technique, in which we compute the velocity dispersion of all the combinations of data sets formed by extracting one source and then sampling by replacement. We take the bootstrap uncertainty to be the 14% and 86% confidence levels. The cluster separates out naturally into two main mass concentrations, which we refer to as the Northeast (NE) and Southwest (SW) regions. We take the cluster center to be situated at the center of this bimodal mass, with a positional uncertainty that depends on the relative masses. Given that each of the two mass regions produce similar numbers of arcs and arclet families, conservatively we expect the mass ratio to be < ∼ 10. The uncertainty on the cluster center translates into an uncertainty in the virial radius, and none of the above takes into account the potentially large systemic errors due to the unknown radial and velocity structure of the cluster. The value for the dynamical mass within the cluster core is relatively common for massive clusters (Girardi et al. 1993), and at the same time is higher than the mean value for CLASH clusters by a factor of three (Siegel et al. 2016). Our values for σ and hence also for M V are potentially inflated owing to the likely presence of radial substructure in the line-of-sight direction. We return to the discussion of biases in the determination of the dynamical mass in §6.1. STRONG LENS MODELING 5.1. The Approach We perform a strong lensing analysis for the fields in our sample by an approach that relies on the assumption that the light approximately traces the mass, or "LTM," such that the galaxies are biased tracers of the Figure 7. HST color image of the G165 field. We identify eleven new arclet families (labeled and color-coded) in this rich field. The doubly-imaged DSFG notably bisects the critical curve (lavender-colored ellipse "1a"), and appears again just north of the gravitational potential as "1b." The critical curve is obtained from our LTM model (see §5), using the pipeline by (Zitrin et al. 2009(Zitrin et al. , 2015. We make arclet family designations by a combination of their similar colors, morphologies, image symmetries, and model-predicted locations (see Figure 8 for image stamps). dark matter. A similar LTM methodology has been used to constrain the 2D mass distribution for cluster lenses extending back to some of the first examples of image multiplicities in cluster environments such as Cl0024 (Broadhurst et al. 2000). This lensing analysis was subsequently extended to accommodate the properties of the first cluster field to show large numbers of arclet families, the HST Advanced Camera for Surveys (ACS) image of A1689 (Broadhurst et al. 2005). To construct our mass maps, we use the well-tested implementation of the LTM pipeline by Zitrin et al. (2009Zitrin et al. ( , 2015. We also refer to Acebron et al. (2018) and Cibirka et al. (2018) for additional descriptions. In our LTM model, the lensing galaxies are assigned a power law mass density distribution scaled in proportion to their galaxy luminosities. The power law index is left as a free parameter and is the same for all lensing galaxies. The superposition of the mass distributions of the individual lensing galaxies, which makes up the initial 2D mass distribution, is then smoothed by a Gaussian kernel to approximate the dark matter distribution, whose width is the second free parameter of the model. The dominant dark matter and galaxy distributions are, in turn, summed up with a relative weight, which is also a free parameter of the model, and then normalized (to a specific source redshift), which adds a fourth free parameter to the model. Finally, the model accommodates a two-parameter external shear to provide additional flexibility. The values for these six parameters are constrained by the positions, orientations, and relative brightnesses of the arclet families. The exquisite spatial resolution of HST makes feasible the designation of arclet families based on morphology and color. By good fortune, the HST images show obvious axes of symmetry superimposed onto the field (see Figure 5), which allow for the identification of image multiplicities even without the aid of measured redshifts in some cases. Arclet families, in turn, constrain the model by imposing the condition that each family member image originates from the same source. The best-fit model is the one that minimizes the angular separations between the observed and predicted (relensed) image positions in the image plane. Notably, in addition to providing confirmation of the locations of the counter-images, the strong lensing model also has the predictive power to locate new image counterparts which can be searched for in the data to iteratively improve on the model result. As spectroscopic redshifts are not available for every arclet family, some redshifts may be left as free parameters to be optimized in the minimization of the model. In such cases, we typically assume a redshift prior of z=2 -3 to align approximately with the peak in the star-formation rate density (Madau & Dickinson 2014). The best-fit model and errors are optimized through a Monte-Carlo Markov Chain using thousands of steps. The lens model for the G165 cluster field is discussed in detail below, and a lensing analysis for the other five fields appears in §B.1 of the Appendix. We emphasize that all arclet families discovered in this study are supported by our physical LTM model. The lensed DSFG spectroscopic redshifts and other relevant information can be found in Table 2 and references therein. G165 Our HST image is rich with giant arcs and arclet families. The presence of giant arcs, and structures consisting of several giant arcs, has been noted before (Cañameras et al. 2015). A preliminary mass model for G165 was made using ground-based CFHT data available at the time (Cañameras 2016;Cañameras et al. 2018). In total, we present here 11 designations of arclet families, all of which are new to the literature (Table 3). The reference center for our lensing analysis is set to the location of the lensed DSFG at (RA, DEC) = (11:27:14.731,+42:28:22.56). By inputting the positions and brightnesses of the cluster members into our LTM lensing algorithm we construct a mass map that is constrained by the arclet family information. Our model reproduces all lensed galaxy images with respect to their locations (rms ∼ 0. 65). The arclet families are marked on a color image along with the critical curve in Figure 7. Postage stamp images of the arclet family members appear in Figure 8 organized by family name. Below we give a description of each of the 11 unique arclet families. G165 DSFG 1 (Arcs 1a, 1b) − Arc 1a is the NIR counterpart of the lensed DSFG at z = 2.2 detected in the submillimeter data set (Cañameras et al. 2015). This giant arc, which orthogonally bridges the critical curve, has a NIR angular extent of ∼5. Our model estimates for Arc 1a to be a merging image with a high areal magnification factor of > ∼ 30 that varies along the long axis of the arc. A counter-image is predicted which is detected only in our redder LBT/LUCI + ARGOS K-band image at the model-predicted location. A bright image at the exact model-predicted location is also seen in our Spitzer/IRAC imaging data. We designate this arc as the counter-image G165 DSFG 1b (see Figure 5). Interestingly, while the F 160W AB -K AB color is consistent between the two images, the F 160W AB -S 3.6,AB color differs by a large > ∼ 2.6 mag. This color difference is owing in part to contamination. G165 DSFG 1b appears to be situated behind a bluer and lower redshift galaxy which influences the photometry and therefore renders the color unreliable (see Figure 5). G165 DSFG 1a is also an arc that is merging. As such, the background source crosses a cluster caustic such that G165 DSFG 1a represents only a region (and only a portion of the starlight) of that background source, while G165 DSFG 1b unveils the entire source and thus the total integrated galaxy light. It is noteworthy that G165 DSFG 1 is the only arclet family in this field to have a measured spectroscopic redshift. This family is used for the internal minimization or "anchor" of our model. G165 2a, 2b, 2c (Arcs 2a, 2b, 2c) and G165 8a, 8b, 8c (Arcs 8a, 8b, 8c) − The Arc 2 family members are the brightest in the field, with K AB -band magnitudes for each of the three arcs of ≈18.5 mag, making them also excellent sources for follow-up spectroscopy to measure the redshift. Here we leave the redshift to be optimized in the modeling, with a most likely prior in the range of z = 2 -3. The bluer arclet trio that make up Arcs 8, which are situated near in projection, are undetected at K. Arcs 2a & 2b and Arcs 8a & 8b fold about an axis of symmetry, as do Arcs 7a & 7b and Arcs 10a & 10b discussed below. G165 3a, 3b, 3c (Arcs 3a, 3b, 3c); G165 4a, 4b, 4c (Arcs 4a, 4b, 4c) and G165 6a, 6b, 6c (Arcs 6a, 6b, 6c) − For the following description we refer to the close-up image in Figure 5. The family members Arcs 6a & 6b are compact arcs which are consistent with an interpretation as a background spheroidal galaxy. These arcs are situated on opposite sides of an axis of symmetry, as marked. Adjacent in projection on the sky, the slightly redder family members Arcs 4a & 4b present more extended morphologies. Coincident with Arcs 4a & 4b, the bright family members Arcs 3a & 3b describe a fold arc conjoined at the axis point. The third image of each of these families, Arcs 3c, 4c, and 6c, appear at an an angular separation of ≈14. This set of third images for each family retains similar colors as well as image morphologies and relative image placements. G165 5a, 5b (Arcs 5a, 5b) − These faint and blue galaxy images are situated just inside the critical curve, and are the only secure arclet family members to reside on the opposite side of the gravitational potential. Arcs 5a & 5b are two merging images folded about the critical curve. Meanwhile, the dashed circle labeled as "5c?" marks the position of a candidate counter-image that awaits confirmation pending additional model constraints. G165 7a,7b,7c (Arcs 7a, 7b, 7c); G165 10a,10b (Arcs 10a, 10b) − Arcs 7a & 7b and Arcs 10a & 10b project onto an arc-like structure that is parallel to Arcs 2a, 2b, and 2c. Arcs 7a & 7b are especially red and low in surface brightness. The counter-image which we designate as Arc 7c appears southward at the model-predicted lo- Arcs 9c Figure 8. Image stamps showing each of the eleven arclet families in the G165 field, as labeled. The two HST bands are shown in most cases, which are demonstrated here to be valuable in the identification of the arclet families. The K-band image extends the reach of HST at comparable spatial resolution, thereby enabling the discovery of the ultra-red counter-image of the lensed DSFG, Arc 1b, which is also confirmed by our Spitzer image at 3.6 µm ( Figure 5). The image stamps all have a 1 bar provided for reference, and maintain the same orientation as in Figure 2. cation. The candidate counter-image labeled as "10c?" appears near to its expected location but at a different color and so is not included in our lens model. G165 9a, 9b, 9c (Arcs 9a, 9b, 9c) − This arclet family trio is distinctively blue and compact. Arcs 9a & 9b are split by an axis of symmetry. Arc 9c appears at the model-predicted location at an angular separation of 10. Note two other candidate counter-images are marked in Figure 7 on the opposite side of the gravitational potential, which await confirmation as additional model constraints become available. G165 11a,11b,11c (Arcs 11a, 11b, 11c) − The blue Arcs 11a & 11b are images that merge across the critical curve as indicated by the pair of star forming knots within Arc 11a that appears again in Arc 11b with reverse parity. Arc 11c appears at the model predicted location southeast of the other two arclet family images at an angular separation of 18. From our lens model we compute a large effective Einstein radius of 15 at z = 2.2 and 17 at z = 9. By integrating up the mass surface density, we measure a lensing mass of 2.6 ± 0.11 × 10 14 M within a ∼250 kpc radius. We refer to §6 for independent measurements of the mass and estimates of the lensing strength. Our lens model is consistent with that of the nonparametric Weak and Strong Lensing Analysis Package (WSLAP) model of Diego et al. (2007). A detailed comparison of the two models especially in light of additional model constraints will appear in an upcoming paper (Pascale et al. 2018, in preparation). Note given the significant visibility of both G165 DSFG 1a and G165 DSFG 1b in the K-band and Spitzer/IRAC, the James Webb Space Telescope (JWST) resolution and sensitivity will be needed at 1 -4 µm to significantly refine these models. We recount the estimate the lensing mass from our strong lensing model. We then make two independent estimates of the lens mass from our spectroscopy. Lensing mass We measure a lensing mass of 2.6 ± 0.11 × 10 14 M within ≈ 250 kpc by applying the constraints imposed by the eleven arclet families ( §4.2). Of these, we have spectroscopic confirmation only for G165 DSFG 1a of z=2.2357 (Harrington et al. 2016). We choose to allow the redshifts of other arclet families to vary as free parameters with values of z = 2 -3. While the approach works reasonably well in that it yields accurate model predictions of the counter-images, nevertheless, the lack of redshifts is non-ideal. This is because uncertainties in the lensed galaxy redshifts translate into uncertainties on the normalization of the lens model, which in turn lead to changes in the value for the total mass of dark plus visible matter. Additional spectroscopy of arclet families is needed in this field. We find the mass density to fall off rapidly beyond 250 kpc, and to reach ≈ 4×10 14 M within 1 Mpc. Dynamical mass Our value for the dynamical mass is a factor of ∼3.5 higher than that of the lensing mass at a fixed radius of 250 kpc. At the same time, the redshifts of only five cluster members enter into the computation (compared to 18 cluster members extending out to the virial radius). We find that the bias imposed by the small numbers is made evident in the choice of the aperture size when estimating the velocity dispersion, and hence on the enclosed mass. For example, for an aperture of 103 or ∼0.5 Mpc, we compute a velocity dispersion for the six cluster members of σ = 1600 ± 200 km s −1 and a cor-responding dynamical mass of M V = 1.9 ± 0.10 × 10 15 M . This value is unchanged from the dynamical mass contained within ≈250 kpc. If we now push the aperture size out by only another 10 to ≈300 kpc, the addition of the two cluster members skews the value for the velocity dispersion of the cluster members upward to σ = 2600 ± 50 km s −1 , equating to M V = 5.8 ± 2.0 × 10 15 M . This amounts to a factor of three difference in mass from the value measured within 250 kpc. As the true velocity dispersion should not correlate with a change in aperture size in the far-field (Girardi et al. 1993), our mass value is uncertain based on the radial velocities of the cluster members. We infer from this that our information on the cluster is redshift-limited. In general, it is expected that 10 -20 spectroscopic redshifts of cluster members are needed to obtain a robust value for the velocity dispersion (i. e., Girardi et al. 1993). It is natural to ask if the velocity structure may be skewed from a spherical distribution, especially given the obvious bimodal distribution of the cluster members into a northeastern (NE) and a southwestern (SW) region (see Figures 7 and 9). It is useful to introduce a "bifurcation" line drawn normal to the line connecting the NE and the SW regions at its mid-point. We plot the projected distance of the cluster members residing in the NE and SW regions relative to this bifurcation line as a function of the relative radial velocity in Figure 9. We find that the two cluster members with the largest projected distances from the bifurcation line of ∼800 and 1050 km s −1 also have some of the highest negative relative velocities (two green hollow squares). These two cluster members are two of only three objects in our spectroscopic sample that show nebular emission lines indicative of ongoing star formation. It is tempting to ascribe this behavior to that of galaxies situated behind the cluster that are falling into the gravitational potential well. We find the trend that the cluster members situated in the NE half are preferentially redshifted relative to the SW half. This results in an elongation of the velocity structure. If true, then the implication is that the cluster velocity dispersion obtained from this nonspherical mass distribution is inflated relative to its true value. Additional spectroscopy is needed to fill in the sparse redshift sampling of the cluster members to obtain a larger, more representative set of cluster members out to the far-field. This discussion relates also to the question of cluster gas pressure, which is given in §6.3. Caustic mass We have spectroscopy for seventeen galaxies that reside within ± 5000 km s −1 from the mean cluster redshift of z=0.351. This redshift information provides the means to measure the caustic mass in a formalism developed in Diaferio & Geller (1997) and Diaferio (1999) (see also Serra et al. 2011;Alpaslan et al. 2012;Windhorst et al. 2018). The approach is to estimate the mass of a cluster of galaxies out to the virial radius by analyzing the distribution of its constituent galaxies in redshift space (i. e. projected separation from the cluster center R as a function of line-of-sight velocity with respect to the cluster median redshift v los ). On the assumption of a virialized cluster, this distribution resembles the bell of a trumpet (with the spread in v los increasing at low R), whose area can be related to the gravitational potential (and hence mass) of the cluster. It is useful to work in phase-space by depicting v los as a function of their projected distances from the cluster center. We adopt the virialized region from the prescription in Jaffé et al. (2015), such that v los ≤ 1.5 σ is within a projected distance of R 200 , where σ is the velocity dispersion (Jaffé et al. 2015). Indeed, the vast majority of cluster members (black dots in Figure 10), fits well within this radius (grey shaded region, Figure 10). We convert our redshift catalog of cluster members into a continuous density field by using an adaptive density kernel. The contour (black curve) identifies the region in the redshift-space distribution that corresponds to the escape velocity of the cluster (assuming spherical infall), which in turn is related to its gravitational potential as v 2 esc = −2φ(r). In practice, we impose the condition of spherical symmetry by rewriting this density threshold into a symmetric version about the v los = 0 line. To do this, we check the absolute values of v los for this doublevalued function in small increments of radius along the density threshold contour. The caustic equates to the minimum of those two absolute values, and is reflected along the v los = 0 line to construct the "tuning fork" shape (green contour). The amplitude of the caustic A(r) is then related to the cluster mass M such that GM ∝ r 0 A 2 (r)dr. Emission No emission Figure 10. The "trumpet" diagram depicting the caustic mass estimator for G165. We start with the redshift catalog of cluster members (black dots) which we convert into a density field. The contour (black) identifies the region in redshift space at which the galaxy density equates to some threshold value, which in this case corresponds to the escape velocity of the cluster. We rewrite the equation for this threshold density into a form that is symmetric about v los = 0 (green "tuning fork") to satisfy the requirement of spherical symmetry in our model. The mass estimate is then made by integrating the area under the caustic lines. We measure a mass of Mcaustic = 1.9 ± 0.18 × 10 15 M within r ≈ 0.8 Mpc. By applying this estimator, we measure a mass of ≈1.9 ± 0.18 × 10 15 M within 0.8 Mpc. The uncertainty on this value is derived by a "jacknife" resampling approach consisting of making 20 realizations in which two galaxies at a time are removed at random and the mass recomputed. Analysis of this set yields the stated estimate in the uncertainty of the mass. Note the mass has been rescaled to be median-biased with respect to the dynamical mass, which is calibrated as a function of redshift and cluster richness of comparable systems in Alpaslan et al. (2012). This value is a factor of ∼5 higher than the value for the lensing mass extrapolated out to 1 Mpc, and a factor of ∼3 lower than the value for the dynamical mass computed within 1 Mpc. If G165 does have an aspherical mass distribution elongated along the line-of-sight (see §6.1), then this value will be an overestimate. In a recent paper, Hayashi et al. (2017) point out that new cluster members undergoing infall show high line-of-sight velocities at all radii. This is potentially insightful for the G165 field, for which there is a bimodal segregation of cluster members. Most of our spectroscopy of the cluster members show the features of red elliptical galaxies as expected for the central region of a galaxy cluster. At the same time, the two galaxies with the highest velocities of v los ≈ 950 and v los ≈ 1750 km s −1 are also two of only three galaxies showing nebular emission line features (magenta diamond-shaped symbols) indicative of recent star formation. This finding of higher star formation in the high velocity outliers is consistent with the picture that these two objects are infalling members, with star formation potentially induced by interactions with other cluster members. G165 as a Lens We compare the lensing strength of G165 with that of another massive lensing cluster at a similar lens redshift, the Hubble Frontier Fields cluster Abell 2744 (HFFs; PI: J. Lotz, GO-13495). Abell 2744 provides a useful benchmark for its well-constrained lens model and similar size of its effective Einstein radius. Its strong lensing model is well-constrained with 29 arclet families identified from deep HST imaging in seven bands with 5-σ limits in each filter of m AB ∼ 29 mag (Mahler et al. 2018). These limits are 2 and ∼3 AB mag deeper than the 10-σ limiting magnitude for G165 for the F 110W AB and F 160W AB filters, respectively (see §3.1). For consistency, we construct the models for both clusters by our LTM approach, where the lens model for G165 comes from this paper and the one for Abell 2744 is from Zitrin et al. (2014). We show the lens models in two left-most panels in Figure 11. We find in both cases a similar elongated shape and similar effective z = 9 critical curve size of ∼15 for the two fields. To compute the lensing strengths, we assume the same background luminosity function (Finkelstein 2016), and then compare the number distribution of lensed background galaxies in the two fields. Overall, the clusters G165 and A2744 yield significantly brighter objects compared to a blank field at all magnitudes. At high-redshifts, the clusters G165 and A2744 yield on average similar numbers of z ≈ 9 objects (right panel in Figure 11). G165 is an ideal lens through which to investigate high-z objects (z = 9 -12). This owes in part to the relatively low redshift of the lens plane of z = 0.351 for which the level of the intra-cluster light (ICL) contamination at the NIR wavelengths corresponding to the Lymanbreak for z > 9 galaxies is minimized (Windhorst et al. 2018). G165 also has a reasonably high ecliptic latitude of 35 • , reducing its background from the peak with the zodiacal plane. On imaging, the lens size is ideal for JWST/NIRCam observations as the lens fills (but does not overfill) the field of view out to 2 -3 times its Einstein radius. Note that for relatively shallow exposures typical of a JWST short program reaching limiting fluxes of m AB ∼ 27 mag, JWST will not have an advantage over HST at 1 -1.6 µm. For such cases, the improvement of JWST will come from imaging at the longer wavelengths (λ > 1.6µm), which enable robust detections of the stellar continuum of any new high redshift galaxy candidates situated behind lensing clusters. We emphasize that this cluster is an excellent candidate to monitor caustic crossing events. Stars from the ICL (see Mihos 2016, and reference therein) are the primary obstacles to detect bright caustic crossing events, since the ICL limits the maximum magnification factors to ∼10,000. When these microlenses are not present, the maximum magnification can reach factors of > ∼ 10 5 , allowing us to see the effects of the much more numerous (but fainter) and smaller stars. The ICL at the position of G165 DSFG 1a is small but still considerable. Deep imaging at optical wavelengths bluer than the cluster member starlight will yield a constraint on the colors that can be converted into an estimated stellar mass per unit area that can come from the ICL. The G165 Cluster Gas Pressure Given our different search strategy to find the G165, it is natural to ask how this massive lensing cluster compares with others selected by more commonly used methods, such as X-ray brightness or the detection of the SZ decrement. G165 has high mass and high dark matter concentration, as evidenced by the prominent displays of giant arcs and arclet families even in these relatively shallow (single orbit) HST images. As such, we would expect for G165 to be replete with large amounts of cluster gas. Table 4 continued Table 4 continued G165 is in fact undetected in ROSAT imaging (R6+R7 bands, or ∼0.7 -2 keV). Put another way, G165 is at most a low luminosity X-ray source with an upper limit on the flux computed from the RASS diffuse map of 1.12 × 10 −4 counts s −1 arcmin −2 . It is unusual for a truly relaxed cluster to have an X-ray flux so low as to be undetected by ROSAT at this redshift and mass scale. At the same time, at these lower luminosities, the scaling relations correlating the X-ray luminosity to cluster mass are more uncertain owing to a large intrinsic scatter in the data. (Bruch et al. 2010). G165 also misses out on membership in the Planck Sunyaev Zel'dovich (PSZ) cluster catalog as a result of its low SZ signal, which falls below the minimum detection threshold. In the Planck Compton-Y parameter map there is a small fluctuation at the position of the cluster which may represent a weak detection of intercluster gas, or it may be noise given the detection is only at the 1-2σ level. This lack of a significant SZ signal might be a consequence of radio emission washing out a shallow decrement, projection effects, or an overestimation of the cluster mass. Radio sources have an inverted spectrum with respect to IR sources that can counteract the SZ signal. As DSFG G165 1a is the one image in the field with high submillimeter flux arising from high star formation and/or AGN activity, this lensed DSFG is the most likely source to be radio-loud. There is a weak radio emitter detected near to the position of the IR source. From NVSS data we measure a total flux from the cluster including this IR source of <40 mJy at 1.4 GHz (Condon et al. 1995). Although present, this modest radio signal is insufficient to compete with the SZ effect at the relevant frequencies (100-353 GHz), thereby ruling out radio contamination as an explanation for the relative SZ silence. The last conventional explanation is the lensing configuration. The G165 field contains an obvious bimodal substructure. This division of mass is strengthened by our glimpses into the velocity structure, in which the cluster members show a velocity gradient between the SW and NE sides (see Figure 9). There are examples of post-mergers that produce significant enhancements of X-ray flux such as the well-studied Bullet cluster Clowe et al. 2006), and the "El Gordo" cluster (Menanteau et al. 2012). If the field is elongated along the line-of-sight direction as a series of two smaller galaxy structures, then we may be catching G165 during a less well studied evolutionary "pre-merger" phase. In this scenario, the total cluster gas pressure dilutes across the large structure, which reduces the gas pressure as well as the X-ray emission hence reducing the SZ decrementȦt the same time, the surface mass densities integrated along the line-of-sight are still supercritical to strong lensing effects. Additional spectroscopy of cluster members is needed to test the hypothesis that cluster orientation explains the low inferred cluster gas pressure. SUMMARY Searching wide-field imaging data sets for giant arcs is now fairly common, yet conducting searches for unresolved giant arcs at submillimeter wavelengths is still relatively rare. We obtained HST WFC3-NIR imaging of the fields of six lensed DSFGs selected in a novel search by their rest-frame FIR color and compactness using Planck/Herschel data. We conduct a more detailed analysis of the G165 field which shows spectacular examples of giant arcs and arclet families. We find: 1. Each of our six sample fields shows the NIR counterpart of the strongly-lensed DSFG. In four fields, the DSFG image appears in more than one location in the image plane as an arclet family at HST resolution (G165, G045, G145 and G080). 2. We use the LTM approach to construct a mass map in the fields for which there is at least one arclet family seen in our data (G165, G045, G145 and G080). For the cases without arclet families that are resolved in our data set (G092, G244), we still generate a κ-map through the galaxy brightnesses and orientations. For G165, we estimate a lensing mass of 2.6 ± 0.11 × 10 14 M within ∼250 kpc and an effective Einstein radius of ≈ 15 at z = 2.2. 3. For G165 we identify eleven arclet families by their similar colors, morphologies and model predictions. Obvious axes of symmetry lend additional support to our arclet family designations. Figure 11. Comparison of the G165 lens with HFF cluster A2744. The two left panels show magnification maps corresponding to the z = 9 critical curves for G165 (left) and A2744 (center) obtained from our LTM methodology (Zitrin et al. 2015). The right panel shows a z = 9 luminosity function for a blank field (black) placed in the background of our LTM model for G165 (red), our LTM model for A2744 (green), and a fully parametric model (blue). The two clusters yield similar numbers of lensed high-redshift galaxies for similar exposure times. Our lens model suggests that G165 has a similar lensing strength to HFF cluster A2744. A2744 The lensed DSFG (G165 DSFG 1a) appears in the NIR as a red giant arc with an angular extent of 5, and magnification factor of > ∼ 30. Its counterimage, G165 DSFG 1b is fainter, and detected only in our high resolution LBT/LUCI + ARGOS K-band image and in our Spitzer/IRAC images. There is a F 160W -S 3.6 color difference between the two images that arises because G165 DSFG 1a is a merging image and so represents only a portion of that background source, while G165 DSFG 1b uncovers the entire background source. 4. For G165 we present ground-based spectroscopy using MMT/Hectospec and Gemini/GMOS. We measure 51 new redshifts, which augment the spectroscopic catalog of objects in this field by a factor of five. From these data we calculate a dynamical mass of 8.8 ± 2.3 × 10 14 M within 250 kpc. We also estimate a caustic mass for G165, which is 1.9 ± 0.18 × 10 15 M within ∼0.8 Mpc. These masses estimates are high, possibly due to the observed aspherical distribution of mass. 5. The lensing properties for G165 are not far different from those of other well-studied massive lensing clusters. In a counting simulation, for G165 we predict similar numbers of high redshift object detections to z ∼ 9 as A2744, another well-studied lensing cluster with similar lens redshift and dark matter properties. 6. Based on the 18 spectroscopic redshifts of cluster members for G165, we find hints of a velocity gradient across the cluster. Such a line-of-sight orientation will dilute the intercluster gas below the ROSAT and Planck-SZ effect detection limits, while maintaining a high surface mass density integrated over the line-of-sight that amply suffices to explain the observed strong lensing effects. We appreciate helpful discussions with Eiichi Egami, Xiaohui Fan, Dan Marrone, Ann Zabludoff and Scott Tremaine. Support for program HST GO-14223 was provided by NASA through a grant from the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA Contract NAS5-26555. J.M.D. acknowledges the support of project AYA2015-64508-P (MINECO/FEDER, UE). R.A.W. was funded by NASA JWST Interdisciplinary Scientist grants NAG5-12460, NNX14AN10G, and 80GNSSC18K0200 from Goddard Space Flight Center. MP was funded by a UA/NASA Space Grant for Undergraduate Research. This work is based in part on observations made with the Spitzer Space Telescope, which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under a contract with NASA. This work makes use of the Large Binocular Telescope, which is an international collaboration among institutions in the United States, Italy, and Germany. We would like to thank the staff at Gemini Observatory North, and at the MMT, for performing the observations in service mode. A. NIR COUNTERPARTS OF OUR LENSED DSFG SAMPLE We searched for the NIR counterparts of the lensed DSFG submillimeter sources. Using the submillimeter positions as a guide, we detect red and relatively-bright NIR counterparts for all six lensed DSFGs at the expected locations with respect to their positions in the submillimeter data (Figure 2, Cañameras et al. 2015). In all cases, the lensed DSFG images in the HST images stand out as the reddest sources in the field. Note these galaxy images are significantly magnified, even if their size is smaller than or equal to the instrumental resolution of HST. Despite their small angular extents in some cases, these lensed sources are still amongst the brightest DSFGs in the sky in the NIR due to their large estimated magnification factors. Figure 12. Color-magnitude diagrams in the six sample fields computed from our HST/WFC3-NIR photometry (blue open circles). The cluster members used in our lens models are indicated by the red stellar symbols and the objects with measured spectroscopic redshifts in each cluster that are also situated in our HST images are depicted by the gold-filled circles. The G165 field shows a tight red sequence despite the obvious bimodal distribution of the mass, possibly indicating that these early type galaxies share a similar star formation history. In the G145, G165, G045 and G080 fields, we detect multiple images of a single background DSFG. For G145 and G080, we find that two of the images match up with peaks in the submillimeter (the "+" symbols in Figure 2). In another field, G092, the NIR counterparts are also identified, yet show different morphologies despite their similar colors. These images are more likely to be two unrelated and possibly interacting DSFGs at a similar redshift (see §5). For G244, we detect the submillimeter arc, but do not spatially resolve the Einstein ring structure, although two sets of arclet families are identified in this field using high resolution ALMA data (Cañameras et al. 2017a,b). Finally, for G165 we find that G165 DSFG 1a bridges the critical curve. We detect another red source at the model-predicted location of the counter-image, G165 DSFG 1b that is prominent in both Spitzer/IRAC channels (dashed circle in Figure 5). The colors between the two images are different, which was initially unexpected as lensing is achromatic. At the same time, G165 DSFG 1a is an arc that is merging with an image of itself. Here, the background source is crossing a cluster caustic, such that G165 DSFG 1a represents only a portion of that background source, while G165 DSFG 1b shows the entire source (see §5.2 for more details). The estimation of the strong lensing properties appears below. B. LENSING ANALYSIS We apply our well-tested LTM pipeline to the G045, G145, G092, G080 and G244 fields, while the lens model for G165 was already discussed in detail in §5.2. For each field, the red lensing galaxies populate a distinctive region of the CMD. Galaxies on this "red sequence" have similar colors because they have a similar redshift and share a roughly similar star formation history. The red sequence is easily established in each of the six fields (Figure 12), where the red star-shaped symbols denote the cluster members used in our model. To reduce the chances for contamination from foreground/background objects, we impose a conservative magnitude cut on our selection of lensing galaxy members in the range of F 110W AB = 20 -22 mag, depending on the field. We have spectroscopic information on cluster members within the HST field-of-view in four clusters, G165, G045, G145, and G080, which aids further in their identification (Figure 12, gold-filled circles). The positions and brightnesses of the cluster members serve as inputs to the LTM model. We emphasize that all arclet families discovered in this study are supported by our physical LTM model. Note that the spectroscopic redshifts of the lenses, the lensed DSFGs, and other relevant information can be found in Table 2 and references therein. B.1. G045 Four peaks of the lensed DSFG are detected in the submillimeter and ALMA imaging (Cañameras et al. 2015;Nesvadba et al. 2016). Of these, we find NIR counterparts for three images which we designate here as G045 DSFG 1a, 1b and 1c (see Figure 2). We measure a spectroscopic redshift for the lens which is z=0.556, based on seven redshifts in the 3-σ clipped range 0.535 < z < 0.577 drawn from our spectroscopy which will appear in a separate paper (Frye et al. 2018b, in preparation). Of these, the redshift for one cluster member is situated within the field of view of our HST data (gold-filled circles in Figure 12). The reference center for our analysis is the location of the lensed DSFG image at (RA, DEC) = (15:02:36.012, +29:20:50.51). Our lens model recovers both the image positions and angular separations of the counter-images with an rms ∼ 0. 4. In turn, the model yields high magnification factors of > ∼ 9, > ∼ 9, and > ∼ 7 for G045 DSFG 1a, 1b, and 1c, respectively. In an independent analysis, the magnification factors of 10 -22 were measured for smaller emission line regions within each arc (Nesvadba et al. 2016). We compute effective Einstein radii of 8 at the lensed DSFG redshift, and 10 at z = 9, respectively. B.2. G145 The positional centroids from the submillimeter image are indicated by the gold plus symbols in Figure 2. We find NIR counterparts for two of these three peaks, which we designate as G145 DSFG 1a and G145 DSFG 1b. These two small arcs are only marginally-resolved using HST. Initially, only one counterpart image was identified, DSFG G145 1a. A careful search unveiled a second image with a similar color, at the model-predicted location, which we designate as DSFG G145 1b. Using these two arcs as inputs, the model predicts, in turn, a third image that coincides with the image in the submillimeter, but which is not detected by HST. The lack of a detection is not surprising, given the faintness of the other two NIR counterparts, which both hover around the limiting magnitude of our observations. The redshift distribution of galaxies in this field is broad, with a somewhat poorly defined peak at z ≈ 0.837, which we take to define the lens plane. This value is based on four redshifts in the 3-σ clipped range 0.822 < z < 0.852 drawn from our spectroscopy, which all fall within the HST field of view (gold-filled circles in Figure 12). This spectroscopy will appear in a separate paper (Frye et al. 2018b, in preparation). We note that there is no spectroscopic information available from the archives or other sources. The redshifts for the four lensing members are situated within the field-of-view of our HST data (gold-filled circles in Figure 12). The reference center for our analysis is the DSFG located at (RA, DEC) = (10:53:22.249,+60:51:43.93). Our lens model recovers both the image positions and angular separations of the counter-images with an RMS of ∼0. 1. In turn, we estimate magnification factors of 12±0.5 and 5±0.5 for G145 DSFG 1a and G145 DSFG 1b, respectively. We estimate the uncertainty by sampling the values for the magnification in a neighboring annular region of width 2, an approach that works reasonably well for images which are not very near in projection to the critical curve ( > ∼ few arcseconds). Our model yields an effective Einstein radius of 10 at the redshift of the lensed DSFG. B.3. G092 The single "tadpole-shaped" arc detected in the SMA imaging breaks up into two lensed sources, G092 DSFG 1a and G092 DSFG 1b, in our HST images. These arcs are not easily reproduced by our lens model despite their similar colors. A clue to their nature is given by subtracting off the light of the central elliptical galaxy using Galfit. By doing this, we uncover significant differences in the smooth vs. clumpy components of the two images (Figure 2, inset). The measured redshift of z = 3.3 is integrated over both components. Based on the available information, we infer that these two images are two different galaxies at a similar redshift. As such, this may potentially be an example of a pair of interacting galaxies which induces the ultra-high star-formation rates of ∼1000 M yr −1 obtained from correcting Figure 13. Surface mass density distributions of the ratio of the local surface mass density to the critical value, or κ, for each of our sample fields using the LTM methodology (Zitrin et al. 2015). For visualization purposes, κ is scaled to the redshift of the lensed DSFG in each field. The values of κ are overlaid onto the contours, and the cluster orientations are the same as in Figure 2. the value in (Cañameras et al. 2015, their Table 2) by the magnification factor provided from our lens model. There is only a single available redshift in this field from the literature, which is of high value as it corresponds to that of the central lensing galaxy (z = 0.448 from SDSS DR 14). The reference center for our analysis is the location of the DSFG at (RA, DEC) = (16:09:17.842, +60:45:19.41). Even without an established arclet family, we construct a map of the surface mass density through the cluster brightnesses and orientations of the member galaxies (see Figure 13). By adopting our best-fit scenario that these are two singly-imaged lensed sources at a similar redshift, we compute high magnification factors of ∼20 for each image. B.4. G080 The submillimeter imaging shows three bright peaks of this one lensed DSFG. The positional centroids of the peaks are indicated in Figure 2 by the gold "+" symbols and labels. We designate the two NIR counterparts that we detect in our HST imaging as G080 DSFG 1a and G080 DSFG 1b. These images are red, faint and low surface brightness features that are visible only upon smoothing the background image (see inset of Figure 2). Interestingly, there is a shift by up to 0 5 in the positional centroids of G080 DSFG 1a and G080 DSFG 1b between the SMA and HST images, equating to a physical extent in the source plane of ∼4 kpc. We find no good explanation for these positional offsets. We compute a redshift for the lens of z = 0.670 that is based on ten redshifts in the 3-σ clipped range 0.649 < z < 0.691 drawn from our spectroscopy in this field, which will appear in a separate paper (Frye et al. 2018b, in preparation). Of these, the redshifts of four of the cluster members are situated within the field of view of our HST data (gold-filled circles in Figure 12). The reference center for our analysis is the location of the lensed DSFG image at (RA, DEC) = (15:44:33.202, +50:23:43.53). Our lens model recovers both the image positions and angular separations of the counter-images with an rms ∼2.2. From this analysis we estimate high magnification factors of ∼20 for each of the two images. An effective Einstein radius of ∼7 is computed at the redshift of the lensed DSFG. B.5. G244 We confirm the NIR counterpart of the lensed DSFG as a red and spatially-extended image, although the spectacular ring-like structure and two arclet families seen in the ALMA data are blended with the primary lens in our HST image and are thus unresolved (Figure 2). The primary lensing galaxy consists of a single object with a measured redshift of z = 1.5 which is blended with the lensed DSFG. Elsewhere in the field there are two blue arcs in the near projected proximity of the brightest cluster galaxy that appear to be unrelated images, and no other arclet families are identified. The expected location of the DSFG (z = 3.0) is the reference center for our analysis at (RA, DEC) = (10:53:53.107, +05:56:18.44). Without arclet families we cannot construct a lens model for this field. Nevertheless, we are able to approximate the surface mass density relative to the critical value through the galaxy brightness and its orientation to yield a κ-map ( Figure 13). Note that this field already has a published model based on the exquisite ALMA data (Cañameras et al. 2017a,b). C. INTRINSIC PROPERTIES OF THE LENSED DSFGS By our estimates, the image magnification factors of all the fields in our sample range from factors of 5 -30 or more, making these objects not intrinsically luminous. For example, the six sources in our sample have submillimeter flux densities in the range of 330 -1054 mJy (Cañameras et al. 2015), which equate to estimated unlensed flux densities corrected by their magnification factors from our lens models of 25 -100 mJy which are more representative of the field DSFG population. We obtain rare spatially-resolved images of the lensed DSFG in some cases, which offers a unique and high resolution view into the complex morphologies of the star forming component. For the lensed DSFGs in G092 and G045 in particular, distinct stellar clumps are detected with magnification-corrected sizes of ∼300 pc, similar to that expected for giant H II regions and for high redshift intensely star forming galaxies (Förster Schreiber et al. 2011). For G165, the physical and kinematical properties of the clumps within the lensed DSFG are presented in Cañameras et al. (2018). Importantly, G165 DSFG 1a is situated on the critical curve. This offers the opportunity to search for transient lensing events in the form of compact star forming clumps traversing the critical curve, which can yield higher magnifications than cluster lensing by factors of ∼200 or more (Kelly et al. 2018). Such approaches may offer a viable route to access first light sources with JWST (Windhorst et al. 2018), and to constrain the amount of compact dark matter . G165 DSFG 1a has a very high apparent (i.e., without correcting for lensing effects) star-formation rate of 17733 ± 171 M yr −1 (Cañameras et al. 2015, their Table 2). Corrected by the magnification factor from our model, this becomes < ∼ 590 M yr −1 . The corrected star-formation rate is also high, and is exactly the galaxy type in which one may expect to find bright stars that undergo caustic (or micro-caustic) crossings, as they can be magnified by factors of thousands. Given the fortuitous placement of G165 DSFG 1a and the intensity of the ICL, microlensing events should be relatively common in this field. Under these circumstances, G165 is a good candidate for frequent monitoring using JWST to search for caustic crossing events.	2019-01-24T16:49:13.000Z	2018-05-12T00:00:00.000	{ "year": 2019, "sha1": "b41b299a6e7e79ac3bd2203bc93da77eab325ddc", "oa_license": null, "oa_url": "https://repository.arizona.edu/bitstream/10150/632529/1/Frye_2019_ApJ_871_51.pdf", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "b41b299a6e7e79ac3bd2203bc93da77eab325ddc", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics" ] }
216360799	pes2o/s2orc	v3-fos-license	Spontaneous membranoproliferative glomerulonephritis in a young Crl:CD-1(ICR) mouse Here, we reported a spontaneous case of membranoproliferative glomerulonephritis observed in a young ICR mouse. A 5-week-old female mouse was euthanized owing to abdominal swelling and increased body weight. At necropsy, generalized subcutaneous edema, and clear, colorless, non-viscous ascites were observed. Histologically, the kidneys showed diffuse, bilateral glomerular lesions. The lesions were characterized by thickening and double contour of the basement membrane and an increase in mesangial cells and matrix, resulting in the narrowing of the capillary lumen. Additionally, eosinophilic hyaloid material accumulated in the subendothelial areas and Bowman’s space. The material was positive for periodic acid-Schiff, complement component C3, or immunoglobulin G, stained red by Masson’s trichrome, and stained blue by phosphotungstic acid-hematoxylin stain and was considered to be plasma due to glomerular leakage. The glomerular lesion was diagnosed as membranoproliferative glomerulonephritis, and an uncertain endothelial injury was suspected as the cause. Here, we report a spontaneous case of membranoproliferative glomerulonephritis (MPGN) observed in a young Crl:CD-1 (ICR) mouse, a strain frequently used in toxicity studies. A 5-week-old female ICR mouse was received from the Charles River Laboratories Japan, Inc. (Kanagawa, Japan) and placed in quarantine. Two days after receipt, the animal presented with abdominal swelling and a marked increase in body weight (from 21.9 g at receipt to 28.0 g). Therefore, 3 days after the receipt, the animal was euthanized by exsanguination from the axillary artery and vein under isoflurane anesthesia and necropsied. The procedures for animal care and housing were in compliance with the institutional guidelines for the care and use of laboratory animals. Blood biochemical tests were performed using serum sampled at necropsy. For histopathological examination, the kidneys, adrenal glands, heart, liver, lung, skeletal muscle, and spleen were fixed in 10% neutral buffered formalin, paraffin-embedded, sectioned, and stained with hematoxylin and eosin (HE). Additionally, kidneys sections were stained with periodic acid-Schiff (PAS), periodic acid-methenamine-silver (PAM), Masson's trichrome (MTC), phosphotungstic acid-hematoxylin (PTAH), and direct fast scarlet (DFS) methods. For immunohistochemistry, kidneys sections were subjected to a labeled polymer method using the Histofine MOUSESTAIN KIT or Simple Stain Mouse MAX-PO (Nichirei Biosciences Inc., Tokyo, Japan) for antibodies against CD31 (product No.: ab28364; prediluted; Abcam, Cambridge, UK), smooth muscle actin (SMA; clone: 1A4; product No.: M0851; prediluted; Dako, Glostrup, Denmark) and C3 (clone: EPR19394; product No.: ab200999; 1:1000; Abcam), or directly against endogenous mouse immunoglobulin G (IgG), and were counterstained with hematoxylin. The validity of the positive staining was assessed using negative control slides, in which the primary antibodies were substituted with antibody diluents. For electron microscopic examination, small pieces of the kidney fixed in neutral buffered 10% formalin were refixed in 2.5% glutaraldehyde, postfixed in 1% osmium tetroxide, and routinely embedded in Epon resin. Ultrathin sections of the selected areas were prepared, contrasted with hafnium chloride and lead citrate, and examined using an HT7700 transmission electron microscope (Hitachi High-Technologies Corp., Tokyo, Japan). At necropsy, generalized subcutaneous edema, and clear, colorless, non-viscous ascites were observed. Blood biochemical tests revealed low levels of total protein, albumin, and calcium, as well as high levels of blood urea nitrogen, total cholesterol, triglycerides, phospholipids, phosphorus, and potassium (Table 1). Histologically, bilateral and diffuse glomerular lesions were observed. The capillary walls were thickened, and the mesangial area was expanded, resulting in narrowing of the capillary lumen (Fig. 1A). Along the capillary walls or with-in the Bowman's space, eosinophilic hyaloid material was accumulated. Furthermore, in some Bowman's capsules, swelling or basophilia of parietal epithelial cells and thick-ening of the basal membrane were observed. The thickened capillary walls, expanded mesangial areas, and accumulated material along the capillary walls and in the Bowman's a) The results for normal mice are obtained from 8-week-old female Crl:CD-1 (ICR) mice (N=3) of the same lot as the present case that was assigned to a toxicity study and were dosed with vehicle (0.5% methylcellulose water solution) by oral gavage once daily for 2 weeks. SD, standard deviation. space were all positive for PAS (Fig. 1B). The PAM stain revealed collapsed tufts, thickening and double contour of the basement membrane, and an increase in mesangial cells and matrix (Fig. 1C). The hyaloid material was negative for PAM and was observed in subendothelial areas. Red blood cells were sometimes observed in the mesangial areas or between the double-contoured basement membrane. The hyaloid material was stained red by MTC (Fig. 1D), stained blue by PTAH, and was negative for DFS. Immunohistochemistry for CD31 confirmed the narrowing of the capillary lumen ( Fig. 2A). Furthermore, increased expression of SMA was observed in the mesangial and subendothelial areas, suggesting activation, proliferation, and interposition of mesangial cells (Fig. 2B). The hyaloid material was generally positive for IgG (Fig. 2C) and C3 (Fig. 2D); it was negative for IgG in the subendothelial areas. In the kidneys, other than in the glomeruli, tubular basophilia mainly in the cortex, as well as minimal tubular vacuolation and hyaline cast were observed. No remarkable changes were observed in the blood vessels and interstitium. Other than in the kidneys, only minimal arteritis in the heart, and minimal to mild extramedullary hematopoiesis in the adrenal glands and spleen were observed. The glomerular lesion was diagnosed as MPGN based on histological characteristics such as basement membrane thickening and mesangial cell proliferation, according to INHAND 1 . Accumulation of hyaloid material is also a characteristic of hyaline glomerulopathy; however, unlike in the present case, in hyaline glomerulopathy, mesangial proliferation is usually not observed, and hyaloid material is ultrastructurally observed as microtubular or fibrillary structures 2 . MPGN is a light microscopic pattern of injury that can be caused by several diseases, and detailed morphological characteristics such as the presence and location of immune complex deposits are important to investigate the cause of MPGN 3 . In the present case, the accumulation of hyaloid material was characteristically observed in the glomeruli. The special stains and immunohistochemical evaluations suggested that this material included plasma proteins such as fibrin, IgG, and C3. Additionally, this material was observed in the subendothelial areas and Bowman's space, while it was not obvious in the basement membrane or subepithelial areas. Therefore, the material was deemed as plasma due to glomerular leakage, rather than immune complex deposits. Reportedly, similar lesions, called insudative lesions, can be observed in glomerular diseases such as diabetic nephropathy and thrombotic microangiopathy 4,5 , and an association with endothelial injuries has been suggested 6 . Furthermore, chronic endothelial injury can demonstrate the MPGN pattern without immune complex deposits 3,7 . Thus, in the present case, an uncertain endothelial injury was suspected as the cause of MPGN. It has been reported that the loss of vascular endothelial growth factor (VEGF) in podocytes leads to endothelial injury and thrombotic angiopathy in mice 8 . However, in the present case, changes in VEGF expression could not be demonstrated using immunohistochemistry (data not shown), and the possible mechanism of endothelial injury was unknown. Moreover, the relationship between glomerular lesions and minimal arteritis in the heart remained uncertain. In mice, spontaneous MPGN has been reported in mice deficient in factor H (FH), the principal inhibitor of the alternative pathway of the complement system 9 . However, the glomerular lesion in FH-deficient mice is distinct from the present case and is characterized by global deposition of C3 along the capillary wall; and insudative lesions have not been reported in FH-deficient mice. Additionally, a mutant strain ICR-derived glomerulonephritis (ICGN) mouse spontaneously develops nonproliferative glomerular lesions accompanied by hypoproteinemia and hypercholesterolemia as early as 40 days of age 10 . The glomerular lesion in ICGN can be distinguished from the present case by the absence of mesangial cell proliferation even in the terminal stage and presence of spike-like protrusions in the basement membrane, with deposition of immunoglobulin along the capillary wall. Furthermore, spontaneous glomerular lesions have been reported in a 33-day-old female and an 8-week-old male ICR mice 11,12 , and the morphological characteristics are similar to those in the present case, including capillary wall thickening, the double contour of the basement membrane, and subendothelial accumulation of hyaloid material or plasma component negative for Congo red. Although some differences have been observed such as in the presence of hypoproteinemia, mesangial proliferation, or stainability of accumulated material for immunoglobulins and C3, this could be reflective the difference in the disease stages. Here, we presented a spontaneous case of MPGN observed in a young ICR mouse. Though further research is needed to clarify the incidence and pathogenesis of spontaneous MPGN in young ICR mice, we believe that this report could provide relevant information for pathological examinations in toxicity studies. Disclosure of Potential Conflicts of Interest: Authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. (B) Higher magnification of (A). The electron dense material can be observed in the subendothelial area. Relatively electron-lucent, irregularly-shaped material can also be seen (arrow). The basement membrane is indicated by the arrowheads. Bar=5 μm.	2020-04-27T20:38:27.823Z	2020-03-22T00:00:00.000	{ "year": 2020, "sha1": "0f64e04ca7446b437181dce4f5f37f4038e9b944", "oa_license": "CCBYNCND", "oa_url": "https://www.jstage.jst.go.jp/article/tox/33/3/33_2019-0081/_pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "f4e921ef9814460e99e44f57cb64604eafe901af", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
267955186	pes2o/s2orc	v3-fos-license	Husserlian Neurophenomenology: Grounding the Anthropology of Experience in Reality : Anthropology has long resisted becoming a nomothetic science, thus repeatedly missing opportunities to build upon empirical theoretical constructs, choosing instead to back away into a kind of natural history of sociocultural differences. What is required are methods that focus the ethnographic gaze upon the essential structures of perception as well as sociocultural differences. The anthropology of experience and the senses is a recent movement that may be amenable to including a partnership between Husserlian phenomenology and neuroscience to build a framework for evidencing the existence of essential structures of consciousness, and the neurobiological processes that have evolved to present the world to consciousness as adaptively real. The author shows how the amalgamation of essences (sensory objects, relations, horizons, and associated intuitions) and the quest for neural correlates of consciousness can be combined to augment traditional ethnographic research, and thereby nullify the “it’s culture all the way down” bias of constructivism. This is not a "view", an "interpretation" bestowed upon the world.Every view about. .., every opinion about "the" world, has its ground in the pregiven world.It is from this very ground that I have freed myself through the epoché; I stand above the world, which has now become for me, in a quite peculiar sense, a phenomenon.[1] What, then, does the future hold?In total with the postmodernists, I have come to the regrettable but obvious conclusion that there is no easy accommodation of the scientific and hermeneutic intellectual frames.Since the hermeneutic frame is for me fatally damaged by its denial of objective truth and the possibility of scientific anthropology, my solution is to proclaim that it is not anthropology at all in any reasonable sense of the term.The wave of the anthropological future that I hope for is a scientific anthropology taking into full account the human capacity for discriminating among highly complex combinations of circumstances and reacting systematically to their similarities and differences.Scientific archeology will benefit from such an anthropology, and it will contribute to it in turn.More properly, it will be a part of this anthropology because a properly scientific anthropology searches for significant relationships among all possible sets of variables at all times and places.[2] Introduction Anthropology has long resisted becoming a nomothetic science.Every time our beloved discipline gets too close to a potential theoretical paradigm, it recoils as from a plague.Anthropologists repeatedly miss any opportunity to build upon empirically relevant theoretical constructs and back away into a kind of natural history of sociocultural differences [2][3][4][5][6].For a discipline to attain the status of a paradigmatic science in the Kuhnian sense [7]. it must eventually establish its foundations in the structures of reality.Until then, it remains a pre-scientific, naturalistic exploration of the surface of things, an "artificial" science as Herbert A. Simon [8] characterizes it.Contemporary anthropology is like biology before Darwin, chemistry before the periodic table, or astrophysics before Newton.Anthropology has come close to this paradigmatic Rubicon in the past (e.g., "psychic unity" and Levi-Strauss-style structuralism), only to veer away from fully acknowledging the roots of its scope in the structures of reality-that is to say, the evolutionary biology of the human organism, and the neurophysiology of the brain. The time is long past since we should be grounding our theory-construction in methods that reveal and study the structures in the real world that produce the similarities and differences between "cultures"; that is, the neural structures that mediate knowledge and operate to confirm that knowledge in direct interaction with the world.One approach that anthropologists have largely ignored is the phenomenology of Edmund Husserl, an approach that provides a rationale and a skillful method for revealing the essential structures of sensory perception [9].We will see that this method requires the systematic separation of what is given directly by the world to perception and all the cultural information that we automatically layer atop the given.By combining the phenomenology of essences with modern neuroscience, we can then ground anthropological theories-especially those about experience, sensory perception, and embodiment-in research that allows us to discriminate between mental universals and cultural variations.But first let me trace the history of how sociocultural anthropology got to be so methodologically outmoded in the 21st century. How Did We Get This Way? The irony is that the founder of anthropology as a full-on scientific discipline, the great German ethnologist Adolf Bastian , sought an anthropology grounded in the psychophysics of our species see [10][11][12][13][14].He reasoned, and I think correctly, that the variation in ways of life of different peoples was generated by the same hidden "psychic unity" of the species (read: structure of the human brainmind).It was the task of anthropologists qua ethnographers to perform intensive, long-term, Malinowski-like field research among many traditional peoples (as Bastian himself did for many years) before "civilization" had completely corrupted their lifeways.However, it was not his vision to collect all the weird and wonderful ways that humans carried out their affairs, but rather, by collecting the surface manifestations of variance among many peoples, to inductively produce accurate empirical generalizations that in turn might be deductively explained by theories about the underlying structure-the Elementargedanken or "elementary ideas"operating within the neuropsychology of humankind.Alas, British social anthropology failed to apply Bastian's injunctions and preferred to rely upon second-hand reports from others who had dealings with traditional folk "out in the colonies".Only after the example set by Malinowski's famous exile to the Trobriand Islands did British ethnography turn to participant observation (as opposed to the later American holocultural analysis), and by that point the earlier notions of a structural underpinning of "psychic unity" beneath apparent variation of lifeways were repudiated. This repudiation of a structural basis of human mentation and behavior was compounded by one of Bastian's students, Franz Boas, who famously migrated to the United States to escape virulent antisemitism in Germany.Boas established himself at Columbia University in New York, and, from that platform, essentially founded American-style "cultural anthropology".From a historical perspective, it is understandable why American cultural anthropology originated in the anti-racist, anti-social Darwinist polemics of the early 20th century [15].To acknowledge even the slightest biological inheritance of social and cultural traits was to leave open the door to claims of inherent and abhorrent racial characteristics, be they Jewish or African or Aboriginal American.This rather reactionary and scientifically naïve sociopolitical stance, understandable as it might be in hindsight, generates empirically unsupportable claims of the "it's culture all the way down" sort that effectively denies any biogenetic origins of mind, experience, sensation, cognition, evalua-tion, ethics/morals, or sociality.Everything of interest to the anthropologist is supposedly a cultural or historical construct, the sole result of social learning (e.g., [16][17][18][19]). In the quest for knowledge, whether pre-scientific or scientific, the methods used to acquire knowledge are intimately bound up with the assumptions about reality, and how the truth of things is to be ascertained and embedded in the minds of the curious [7,20].As the philosopher of science and evaluation Michael Scriven [21] showed us, any agent (human or non-human) that operates as a finite information storage and retrieval system (i.e., a brain) and that finds itself in an over-rich data environment must theorize.By theorizing, the agent imposes redundancies upon the data field, and thus may adapt to the data field by projecting patterns of redundancy upon the field.All animals with or without brains are just such agents.Among humans, people operate in this way by creating systems of knowledge which, when shared within groups, become belief systems and worldviews.Embedded in such knowledge systems are the means by which the truth of things is ascertained, often by controlling how the quest for truth is accomplished.This relationship between knowledge systems and methods applies as much to scientific as to pre-scientific knowledge systems.Anti-structuralist (or anti-essentialist) social scientists have their views and their methods of confirming their views, as do structuralists and essentialists.But there is a profound difference between the two approaches to knowledge.The antistructuralist/essentialist views and attendant methods can never produce a nomothetic body of theory, whereas the structuralist/essentialist views have a good chance of discovering a theoretical basis for both similarities and differences among human groups.The difference between the approaches is crucial, for one offers no methods by which observations of the surface of things can penetrate into the reality below the surface, whereas the other may well do. In the past-say, back in Bastian's, Boas' or even Levi-Strauss' day-the technologies available for penetrating to reality were very limited, for each of them were dealing with the products of the human brain operating in groups.There were few methods for directly observing what the brains of people were doing as they expressed themselves, communicated with each other, and behaved cooperatively.Levi-Strauss, for instance, developed a very flawed method of deductive reasoning with respect to observed patterns in social structure and mythology, and only once, to my knowledge, acknowledged that his structures must in fact be neural structures (see his 1971 book L'Homme nu, the fourth volume of Mythologiques; for the English translation, see [22]). 2.1.Experience, Embodiment, and the Senses Some anthropologists, perhaps weary of the status quo in mainstream sociocultural anthropology, began to ground their understanding of human lifeways upon subjective and intersubjective experience [23][24][25][26], including extraordinary or transpersonal experiences [27,28].With a significant stimulus from earlier theorists such as Durkheim, Levi-Strauss, and Levy-Bruhl [29,30] and the tantalizing cybernetic theorizing of Gregory Bateson [31,32], Victor Turner [23,33] and others began shifting their focus from institutions and social organization (people viewed from the outside, as it were) to the everyday experience of people going about their daily lives (people viewed from the inside).This shift in focus led to the anthropology of experience approach to doing ethnography, as well as an understanding of the sociality from an intersubjective standpoint see e.g., [34][35][36][37][38].The reorientation of perspective towards the lived experience of people in different societies raised interesting methodological issues about how the ethnographer should or could go about "getting into the heads" of their hosts; i.e., how is one to access the privacy of consciousness [39]?This is not a new challenge by any means, but rather requires an expansion of what it means to place "participant observation" into a first person, intersubjective mode [40,41]. The anthropology of experience approach and its quest for appropriate methods led quite naturally to interest in the ideas and methods of phenomenological philosophy [9,30,[42][43][44][45][46][47][48][49], especially the formative work of Edmund Husserl [50][51][52].The blending of the anthropology of experience with phenomenological methods led quickly to the recognition, either explicitly or implicitly, that experience is always "embodied"-a realization already well established in early 20th century Husserlian phenomenology [53] (pp.[152][153][154].Embodiment became "a thing" among psychological scientists (see e.g., [54][55][56]) and became de rigueur within the anthropology of experience cohort [39,49,[57][58][59][60][61][62][63].It is easy to demonstrate using even casual introspection that everyday experience is had from an embodied standpoint.I am staring at this screen with my eyes open, from a posture of sitting, and typing with my fingers on a keyboard.Embodiment is as simple and yet as profound as that; as the old saw goes, "wherever you go, there you are", taken there in and by your body, even when you may be "out of body" in a vision or dream. Yet with this concern with embodiment, with few exceptions (e.g., [64]), scant attention has been paid to the part of the body that mediates experience, namely the nervous system (see [65][66][67][68] for exceptions).This persistence of cultural constructivism has gradually alienated anthropology from mainstream post-neuro-turn science [69,70], (see especially [71] on this issue).Even among phenomenological anthropologists, researchers tend to base their findings upon those phenomenologies that support the "culture all the way down" bias (see [42,46,72] for summaries). Even a cursory reading of Husserl will show that embodiment of experience is fundamental to his understanding of the sensory: The Body is, in the first place, the medium of all perception: it is the organ of perception and is necessarily involved in all perception.In seeing, the eyes are directed upon the seen and run over its edges, surfaces, etc.When it touches objects, the hand slides over them.Moving myself, I bring my ear closer to hear.Perceptual apprehension presupposes sensation-contents, which play their necessary role for the constitution of the schemata and, so, for the constitution of the appearances of the real things themselves.[53], (p.61) It should be obvious that the anthropology of experience with its understanding of embodiment and its increased reliance upon phenomenological methods would lead to both a sharper focus upon the role of the senses in experience [73][74][75][76][77]-already a primary focus of early 20th century Husserlian methods-and to a closer proximity to psychobiology.Hundreds of books have been written about human social organization, kinship, ritual, behavior, subsistence, and social institutions with no reference at all to people's experiences, their bodies, or their brains.However, it is much harder to avoid human physiology when focusing the ethnographic gaze upon embodied experience and the senses. Introducing Neuroanthropology In order to claim, as many anthropologists do, that there are no structural underpinnings mediating apparent individual and cultural variations, logic requires that initially the organ of learning, the brain, be a tabula rasa, a "blank slate" [78].In other words, the human brain must, by some magic, be at birth like a brand-new, unformatted hard drive.But, in the 21st century, in the age of neuroscience, we know this is not the case [79].Of course, the emergence of anthropology during the first three quarters of the 20th century occurred during a time when science was burgeoning, but neuroscience was still limited to neurology [80].Neurologists at the time generally steered clear of addressing the relations between the brain and mind, and few neuroscientists drew any interdisciplinary connections between neurobiology and the social sciences.But, by the 1980s, interdisciplinary neuroscience had begun what we now refer to as the global neuroscientific turn, or simply the neuro-turn, a neurobiological engagement with research and theoretical issues in the traditional purview of the social sciences and humanities [64,[81][82][83].This movement led to several new sub-fields in neuroscience resulting in new journals such as Culture and Brain, Social Cognitive and Affective Neuroscience, and Social Neuroscience, as well as numerous related books and articles (e.g., [84][85][86][87]). Biocultural perspectives began to emerge mid-century and may be seen as one early hint of a neuro-turn in anthropology (see [88][89][90]).Physical anthropology, of course, has had an abiding interest in the evolution of the primate and hominin brains, especially reflected in the work of Ralph L. Holloway [91,92].Although the roots of anthropological interest in the brains of fossilized and living peoples dates back to the mid-20th century [93][94][95], it may be viewed as a more direct outgrowth of neuroanthropology, a movement my colleagues and I founded in the 1980s and thereafter elucidated [96][97][98][99], in synch with other scientists [100][101][102][103], in an attempt to engage cultural anthropology with the neurosciences. Despite the discipline's aversion to anything suggestive of biological structure, some anthropologists beginning in the 1980s began exploring new approaches to studying human lifeways from an evolutionary biological perspective.The early work leading to the field of evolutionary psychology began to appear, spearheaded by Canadian anthropologist Jerome Barkow and his colleagues [104][105][106].They reasoned that modern peoples are actually operating from a brainmind that evolved during the Upper Paleolithic, and many "cultural" features we encounter among living peoples make better sense when viewed as ancient adaptations to a physical and social world much changed [107]. Today, neuroanthropology is the study of how the brain mediates experiences, social relations, techno-skills, histories, habits and institutions, and social learning among Homo sapiens, and incidentally other large brained social animals, especially other primates, and extinct hominins.We now know that we share many of the neural structures with our fellow creatures with big brains, especially with other primates, and there is much to be learned from cross-species comparisons.As such, neuroanthropology stands as a necessary alternative to the "naïve" culturological position described above-the notion that upon inventing "culture" humanity somehow left their neurobiology behind.Neuroanthropology dispenses with the "it's culture all the way down" fiction and leaves open the possibility of a "cultural" theory grounded upon psychophysical structures (e.g., [68,[108][109][110][111]). Husserlian Phenomenology As I have suggested above, of all the schools of phenomenology that might have methods to offer us in our quest to ground the scientific study of experience and the senses in reality, the school that promises to be the most applicable is that of Edmund Husserl (1859Husserl ( -1938)), considered by many to be the Father of Phenomenology (Those readers wishing to read about Husserl's career, see [112] Chap. 1, [113] Chap. 1, and [51,114]).The reason Husserl's work is advantageous is that, unlike most other phenomenologists such as Heidegger, Merleau-Ponty, and Schutz who have influenced some anthropologists, Husserl was a mature contemplative who showed that to properly reveal and study consciousness via introspection, an extraordinary degree of skill in contemplative methods is required. Husserl's project is clearly set out in his article "Philosophy as Rigorous Science" [115], in his set of lectures from 1902 to 1903 published in a short book entitled The Idea of Phenomenology [116], his book Ideas: General Introduction to Pure Phenomenology (commonly called Ideas I; [117] (pp.41-47), and is described as a "new science" that describes and analyses the essential structures of "pure experience" and consciousness.Husserl considered phenomenology to be an eidetic science-the term "eidetic" deriving from the Greek eidos or "essence" (see [118], Chap.2).Phenomenology is also an a priori science; that is, it is not "empirical" in the sense of collecting facts from observations had here and there, and testing ideas in light of the facts.Rather, phenomenology is an approach that seeks a first-hand introspective exploration and description of the essential structures of experience, leading to an analysis of essences upon the basis of which scientists can build inductive theories about the world inside and outside the body.Phenomenology is designed not only to precede natural and social science, to ground science in the truth accessible via experience, but also to critique the taken-for-granted grounds of sciences that are, after all is said and done, dependent upon perception. Grounding Sensory Experience in Essential Structures Husserl's methods of contemplation are unique.They are also complicated, and there is insufficient space here to do them justice or to show the reader how to perform basic meditations on the senses (see [119][120][121]).What I do wish to do is emphasize that Husserl's intention was to separate those elements of experience given by the world and those contributed by the mind and projected upon the sensory given.People carry on within their lifeways uncritically blending the two sources of information in what Husserl called the natural attitude, the state of consciousness in which we blithely follow our daily habitual engagement with the world of things and people. Let us make this clear to ourselves in detail.At the natural standpoint we simply carry out all the acts through which the world is there for us.We live naively unreflective in our perceiving and experiencing, in those thetic acts in which the unities of things appear to us, and not only appear but are given with the stamp of "presentness" and "reality".When we pursue natural science, we carry out reflexions ordered in accordance with the logic of experience, reflexions in which these realities, given and taken alike, are determined in terms of thought, in which also on the ground of such directly experienced and determined transcendences fresh interfaces are drawn.At the phenomenological standpoinit, acting on lines of general principle, we tie up the performance of all such cogitative theses, i.e., we "place in brackets" what has been carried out, "we do not associate these theses" with our new inquiries; instead of living in them and carrying them out, we carry out acts of reflexion directed towards them, and these we apprehend as the absolute Being which they are.We now live entirely in such acts of the second level, whose datum is the infinite field of absolute experience-the basic field of Phenomenology.[117] p. 155 This is why treating Husserlian phenomenology as just another philosophical phenomenology is wrongheaded (see [122]).The intent of Husserlian phenomenology is to strip away all of the taken-for-granted-ness about the world, bundle it all up (metaphorically speaking), and store it in a mental closet until next it is needed, leaving the "pure ego" or Watcher free of beliefs, ontological assumptions, and other hindrances to studying one's own intentional acts. Husserl variously called his method performing a reduction (from the Latin root meaning to "lead back", "bring back", or "restore or return to a previous state"), to bracket (from mathematics where an expression in brackets is treated as a unit), to return to the things (focus upon what is given in perception and only in perception), to enter the epoché (attitude in which taken for granted assumptions about reality are suspended).Simply put, Husserl's method is one of focusing awareness solely upon what is given a priori ("primordially") in sensory experience and suspending all else that is not given in the act [117] (pp.107-111), [123] Chap.3, [124] p. 39, [125] (pp.206-211).Following his teacher Franz Brentano [126], Husserl considered every act of consciousness, whether from the natural standpoint or from within the epoché, as an intentional one; that is, consciousness is always about something ("consciousness of").No matter the state of consciousness, there is always an object and always a subject within the context of the act. When Husserl performed the reduction upon any sensory experience, he learned that within the field of perception there are objects, relations between objects, a perceptual field, a horizon beyond which no perceptions can occur, and intuitional knowledge.He further discovered, among many other things, that "pure" perception is comprised of sensory qualia (hyle or "stuff"), forms (morphé), ideas (eidos), interrelations, and intuitions that are mediated by essential structures (or essences) of perception, i.e., in short, the sensuous elements of perception paired with intuitive knowledge.With respect to the object (say a coffee mug), both hyle and morphé interact within the perceptive field to instantiate the idea of the object; the object is not just any random thing, it is a "coffee mug".Moreover, the object/idea relationship is reciprocal.I may see an object and experience it as a "coffee mug", or I might go looking for a coffee mug (an idea in my head) and find an object that "fulfills" the idea with appropriate hyle and morphé. Intuition permeates the perceptual act.For instance, we can only perceive the object from a certain perspective, yet we are aware of the object as a whole.We do not perceive a visible half of a coffee mug, but rather the entire coffee cup, visible aspects as well as invisible aspects.Also, we perceive that although the hyle (in this case visual data) can vary with changes in light conditions and changes in perspective, we always intuitively grasp that it is the same object.Husserl himself uses striding towards a tree in his garden as an example: The colour of the tree-trunk, as we are aware of it under the conditions of pure perception, is precisely "the same" as that which before the phenomenological reduction we [. ..] took to be that of the real (wirklichen) tree.Now this colour, as bracketed, belongs to the noema.But it does not belong to the perceptual experience as a real (reelles) integral part of it, although we also find in the experience "a colour-like something", namely, the "sensory colour", the hyletic phase of the concrete experience in which the noematic or "objective" colour "manifests itself in varying perspectives".[. ..]But one and the same noematic colour of which we are thus aware as self-same, in itself unchanged within the unity of a continuously changing perceptual consciousness, runs through its perspective variations in a continuous variety of sensory colors.We see a tree unchanged in colour-its own colour as a tree-whilst the positions of the eyes, the relative orientations, change in many respects, the glance wanders ceaselessly over the trunk and branches, whilst we step nearer at the same time, and thus in different ways excite the flow of perceptual experience.Let us now start sensory reflexion, reflexion upon the perspective variations: we apprehend these as self-evident data, and are also able, shifting the standpoint and the direction of attention, to place them with full evidential certainty in relation with the corresponding objective phases, recognize them as corresponding, and thereby also see without further difficulty that the perspective colour-variations, for instance, which belong to some fixed colour of a thing are related to that fixed colour as continuous "variety" is related to "unity".[117] (pp.283-284) If you can grasp the distinction between the "objective" color of the tree and the sensory variations we experience through time, you will go a long way toward understanding what Husserl is getting at in general (and, by the way, this is supported by vision neuroscience; see [56] p. 165).What he is saying is that the naïve "natural attitude" observer is experiencing an objective tree defined as having a brown trunk and green leaves, while at the same time glossing over as insignificant the variety of colors that are actually occurring in our sensorium as we change our orientation relative to the tree, and as illumination changes.The role of the hyle is to fulfill the empty eidetic intention "tree".The "objective" tree is the same regardless of the variety of sensory colors we perceive changing through time.We do not have to think about this variety/unity correspondence; it is instantaneous, intuitive, and automatic [127] p. 208. An essential structure, or essence, is an attribute of the perceptual act that is obdurate relative to our will.In other words, we cannot by any act of will modify the structures of perception.Essences are as it were "wired-in".In neurobiological terms, essences are how we experience the result of millions of years of evolution behind sensory systems that present the world as it is before our sensorium [128].The essential structures of sensory experience revealed by Husserl's personal meditations are myriad and scattered throughout his writings, most of which have yet to be published (see [129]).Here are several essences that you might uncover by your own efforts as a novice phenomenologist: Pattern.Each and every object you reduce to its primordial given has a form (morphé); i.e., it is ordered, never random, shapeless, or chaotic.Suspending the natural attitude therefore does not produce a "blooming, buzzing confusion" of hyle.It is easy to see how the mind apprehends the sensory and intuitive patterns supplied by the primordial given and project's meaning upon those patterns.You might say that the given is inherently salient within the context of the intentional act. Things.Our experience is full of things, "thingly-real" objects with stable boundaries, and duration.Things are filled with apparent qualia (hyle in Husserl's terms as contrasted with form or morphé: colors, textures, tones, tastes, etc.), relations among objects, and intuitions (class membership, position within the horizon, aesthetics, pragmatics, and so on). Part-whole relationship.If the object is a three-dimensional thing, we never perceive it in its entirety at the same time.We intuitively infer the whole from the perspectives we are afforded upon its parts.We never "see" part of the coffee mug per se; we see the mug as a whole. Eidos ("form", "type", "species").Things before our gaze are intuited within direct experience as exemplars of a class, an idea, an eidos. Entanglement.Objects (including hyletic data) are always given within the context of an environment of other objects with which they relate and with which they interact within the field. Horizon.Objects always present within the context of a horizon (limits to what can be perceived at the moment).We never perceive the whole world, but only those objects and horizon within our perceptual field. Impermanence.No matter the object of our focus, it is perceptually impermanent.It is not there to your perception before you experience it, and it will not be there at some point in the future.Phenomenologically speaking, there is no such thing as a permanent object, including the ego or Watcher. Object-Watcher discrimination.Reduction of the relationship between the object and the Watcher as the subject.The object only arises before a subject, and the subject intends the object.There is always a subject to any intentional act. Intentionality.Every moment of consciousness is constituted as a system of essential structures linking an object and a subject perceiving the object. Attention.You may exercise control over your relationship with the object, or the object may "draw" your attention.This is the function within consciousness that allows for the willful control of focus and modification of the intentional act. Many other essences are detectable and describable within the scope of "pure" phenomenological methods.I have only suggested a few to illustrate the flavor and the technique of reducing aspects of perceptual experience.Indeed, once a contemplative is skilled in entering the phenomenological attitude, he can rely upon it in any state of consciousness, including ASC such as psychotropic drug trips, lucid dreams, hypnagogic and hypnopompic states, meditative states, and absorption states.In my own work and writings, I have relied heavily upon both my own experiences enriched by phenomenological methods, and the reports of other mature contemplatives that have mastered these skills, not just those who have mastered Husserl's methods, but from various other traditions, especially Buddhist mindfulness meditators see [130].Regardless of the approach (as long as it entails intense, skilled, and mature contemplation), reducing the perceptual field always uncovers essences upon which "culture" has no impact save at higher neocortical levels of processing mediating "meaning", which becomes layered on, or "sedimented" upon, the primordial given. This layering is so automatic and occurs so rapidly that people in general, regardless of their cultural background, are unaware of the processes involved.Only through disciplined phenomenology are we able to make and study these discriminations. In the unbroken naïveté in which all psychology, all humanistic disciplines, all human history persists, I, the psychologist, like everyone else, am constantly involved in the performance of self-apperceptions and apperceptions of others.I can of course, in the process thematically reflect upon myself, upon my psychic life and that of others, upon my and others' changing apperceptions; I can also carry out recollections; observingly, with theoretical interest, I can carry out selfperceptions and self-recollections, and through the medium of empathy I can make use of self-apperceptions of others.I can inquire into my development and that of others; I can thematically pursue history, society's memory, so to speakbut all such reflection remains within transcendental naïveté; it is the performance of the transcendental world-apperception which is, so to speak, ready-made, while the transcendental correlate-i.e., the (immediately active or sedimented) functioning intentionality, which is the universal apperception, constitutive of all particular apperceptions, giving them the ontic sense of "psychic experiences (Erlebnisse) of this and that human being"-remains completely hidden.[1] p. 209 As Husserl repeatedly noted, culture (including scientific concepts and theories) drapes over our everyday experience like a "garment of ideas" or "garment of symbols" (e.g., [1] p. 51).This, of course, refers to all your life-long enculturation ("Enculturation" is anthropology-speak for the process by which we learn to be a "culture-bearing" member of our society)-everything you learn from other people while growing up as a member of a group.Husserl [115][116][117] argued that only through the application of the phenomenological method could the sciences and philosophy be grounded in direct knowledge about how the mind constructs the reality we experience. Considering Intersubjectivity Essential structures of sensory perception extend to objects that happen to be other people, or for that matter, other sentient beings.My perception of you is no different in most respects from my perception of a coffee mug.You are before my consciousness as a thinglyreal and enduring object constituted by hyle and morphé, relations to other things in my sensory field, and within my inevitable horizon.All of the essences summarized above apply to people as well as inorganic things.It is in the domain of intuitive knowledge that a crucial difference is discerned.When you appear within my perceptual field, I immediately know you as "like me", with both a body and an inferred stream of consciousness.I was in fact born this way, seeking and finding my mother's face and breasts [79]. My grasp of your consciousness and experience derives from a special type of intentionality Husserl called empathy (Ger: Einfühlung; [131] p. 120 and p. 135; see also [132][133][134], [121] p. 92, [135] Chap.6).This is not to be confused with our everyday, fuzzy notion of empathy that refers to various feelings (compassion, pity, concern, kindness, and the urge to help) or identification with someone's plight [132,136].My empathy in the phenomenological sense is a generalized intuition associated with my experience of your body, behavior, and expressions [117] p. 210, [124] p. 149; see also [137], [133] p. 11, [138].We now know that there are systems of so-called mirror neurons in the mammalian brain that subserve such intuitions [139].In other words, I do not just experience your presence as a physical thing, or even a body behaving, but I intuitively know you as a class of "person" and that you, like me, are experiencing a stream of consciousness that I cannot access."More specifically, what counts in the strict sense as empathy are those experiential acts in which a foreign subject is not merely hypothesized or inferred, but rather given and experienced herself" [132] p. 274.In the same way, when my dog Luke is present to my perception, I experience not only his physical being, but as an exemplar of a class "dog" and that, like myself, he is experiencing a stream of consciousness in which I am an object (see [140]).In neither case can I experience the content of Luke's or your stream of consciousness in the same primordial way I experience your respective bodies, behaviors, and expressions. Empathetic intuition includes the knowledge that the world that I encounter within the epoché is the same knowledge you encounter under the same circumstances.This realization led Husserl to posit one of the most important concepts he contributed to philosophy, as well as to anthropology, social psychology, and other social sciences, the lifeworld (Ger.: Lebenswelt, Those readers wishing to read more about lifeworld phenomenology, see [1]; see also [141][142][143][144]).Husserl borrowed the term from philosopher Richard Avenarius (1843-1896) and others to characterize the intersubjectively shared, pre-scientific, primordial givenness of our experience of the environing world and to emphasize the role of intersubjective experience in the performance of the transcendental epoché-that is, realizing that the environing world I am experiencing within the horizon of the primordial sensory given is the same transcendental world you are experiencing via the same essential structures of primordial perception [117] (pp.[130][131][132][133][134][135][136][137], [145]. Husserlian Neurophenomenology Husserlian essences, whether involved in experiences of things, events, or other sentient beings, are precisely the kind of data that immediately falsify constructivist assumptions, for they not only evidence real structure at the roots of experience, but they are of the kind that can be confirmed by reference to neuroscience, and, in particular, research into the neural correlates of consciousness (NCC; see [146][147][148]).As far as I know, Husserl himself never suggested a merger of his methods with those of neuroscience.Again, his project was the production of a "pure" subjective study upon which to rest the findings of the natural sciences, including psychologies of various sorts [149] (pp.[166][167][168][169][170].However, Husserl was not ignorant of the fact that consciousness and sensory essences were somehow being generated by the brain: [I]s it possible, we are asking, for the matter here at issue to be understood in such a way that the cerebral states (states of the [body]) precede, in an Objectively temporal sense, the corresponding conscious lived experiences, or must not, for reasons of principle, the brain state and its conscious accompaniment be simultaneous, in conformity with the absolute sense of simultaneity?Thereby is not a parallelism given eo ipso?Namely, in this way: to every conscious lived experience in my consciousness C m there corresponds a certain state in my [body], a certain organic state [read: NCC].On the other hand, to everything without exception that comprises the [body] there correspond real events of a certain kind in every subject, and consequently also in me: certain real perceptual possibilities, which, if not corresponding to the state of the brain [. ..], then to another state in connection with it in a natural-scientific nexus.[53] p. 305 Husserl raises the question of whether there is a lawful connection (be it simultaneous or over some short "objective" duration) between brain states and states of consciousness.More specifically, do the essential laws of consciousness and experience correspond with laws of psychophysical functions [150]; see also [151,152]?Neuroscience has long since answered Husserl's question in the affirmative.Thus, it now makes sense to speak of Husserlian neurophenomenology (HNP) as an interdisciplinary approach to the study of experience and the senses applied to ethnological scopes of inquiry (The phrase "Husserlian neurophenomenology" was coined in an article by neuropsychologist Francisco Varela [153] who was himself a mature contemplative.Varela was influenced by Husserl's writings and seems to have agreed to some extent with Husserlian methods (see [154]).But he contrasted Husserlian neurophenomenology with his own approach which he called "experiential neuroscience".Varela was a practitioner of Tibetan Tantric Buddhism, being a follower of the late Chögyam Trungpa Rinpoche, and the methods Varela used in this regard were different than those used by Husserl.To my knowledge, Varela never applied Husserlian methods in any systematic way). Husserlian Neurophenomenology in the Field In the context of ethnography carried out among traditional societies, unless they have been trained in contemplative phenomenology, our hosts are most likely to be as naïve as is the ethnographer about the essential structures of our individual perceptions.And, like most ethnographers, they will probably be relatively uninformed in neuroscience, although psychophysiological knowledge can be found in some cultural traditions (e.g., see [155] on knowledge of neurophysiology among the Desana people of Brazil).Thus far, there are but a handful of ethnologists who have more than a passing familiarity with either contemplative methods or neuroscience evidence pertaining to NCC.This situation places practical limits on the extent to which most ethnographers can avail themselves of an HNP perspective in the field.But, this merely poses a challenge, not an insurmountable wall, for to shift anthropology toward a scientific future will undoubtedly require retraining [156].I would like to conclude this article by exploring some potentially fruitful alternatives that could bring HNP to bear in the field. Take Up the Challenge The easiest way to operationalize HNP for any individual ethnographer is to train themselves to apply the reduction before they enter the field.The method is easily teachable.Although it is too much to expect every ethnographer to train in neuroscience as well, this need not be an obstacle because they can collaborate with a willing neuropsychologist.Once a fieldworker is skilled enough in Husserlian phenomenological methods, there are myriad ways to apply these skills in the field.Indeed, this has already been carried out successfully by anthropologists exploring the more esoteric, transpersonal experiences that are the foundation for many traditional societies' spiritual lives and worldviews (e.g., [27,[157][158][159]).It is important to note that Husserlian methods can be applied in any state of consciousness in which sufficient awareness and lucidity are present.Indeed, there is a causal connection between the intensity of awareness and intensity of lucidity.The phenomenologically trained ethnographer who subjects themself to native spiritual practices such as drug trips, ritual practices, vision quests, and the like can easily expand their analysis of sensory experiences by applying the reduction.In this way, they can discern the difference between alterations in the sensory given and the meaning attributed to the experiences by their hosts. Experimental Methods Experimental ethnography has long been used in the field and in ethnological analysis (see e.g., [160][161][162][163][164][165]).For instance, volunteer hosts have been trained to use cameras and then asked to photograph and film events of meaning to themselves.Fieldwork in applied anthropology occasionally takes on an experimental expression (e.g., [166]).There is thus no good empirical or ethical reason for not training host volunteers to apply Husserlian methods and then describe their revelations in their native tongues, thus blending Boasian and Husserlian methods for getting at claims of universality for essential structures, and simultaneously exploring how each culture expresses these features of sensory and intersubjective experiences. An alternative strategy might involve collaboration among professional anthropologists raised in different cultural backgrounds.Native-born anthropologists (Ph.D. level anthropologists today derive from societies in Africa, South and Central America, Native America, First Nations in Canada, Aboriginal Australia, and some islands of the Pacific) and anthropology students could be trained to carry out Husserlian meditations to verify cross-cultural patterns in essential structures and those encountered in alternative state of consciousness (ritual trance, dreaming and co-dreaming, absorption states, states driven by entheogens, ordeals, vision quests, and so on) relevant to their own spiritual traditions. Training in the Anthropology of the Senses The most obvious cohort of professional anthropologists that might be (should be?) interested in Husserlian methods are those with a focus upon the anthropology of the senses [49,[73][74][75][76][77][167][168][169][170][171].The anthropology of the senses marks a return to "the things" in Husserl-speak, a return to the sensuousness of everyday life [171] (pp.3-10).Yet, even with the attention being given to sensuousness in experience, few anthropologists ground their work on the neurobiology of the senses.Yet, the structure of the sensory systems, as well as the sensorium itself, is as resistant to plasticity as the thyroid, spleen, or intestine; they are some of the most biogenetically ordered neural systems in the body.Offering them a sure-fired method for discriminating the a priori givenness of sense data from cultural overlays of meaning, preferences, and praxis would surely be a welcomed addition to their ethnographic toolkit.There will undoubtedly be resistance from those with an ideologically driven constructivist bias, but enough ethnography could be based upon Husserlian methods to allow for collaboration with neuroscience colleagues to ground neuroanthropological theories in both ethnographic fieldwork and in neuroscience. Collaborative Teamwork As I mentioned, it is rare to find an ethnographer who is also trained in both contemplative phenomenology and neuroscience.Such a broad training would not be strictly necessary, for one of the most powerful applications of Husserlian neurophenomenology in cross-cultural research could be implemented using collaborative teamwork.There is a long history of collaborative ethnography in the discipline.We are all familiar with the expeditions carried out by teams of researchers in the latter 19th and early 20th centuries [172].Since then, most ethnographic research has been carried out by lone researchers (à la Malinowski) or spouses operating as a team of two (à la the Tedlocks).But cross-disciplinary teamwork has also proved productive in a number of ethnographic venues, including healthcare facilities, ethnic schools in urban areas, market research, linguistic research, as well as more generally in traditional societies (e.g., [173][174][175][176]).In fact, collaborative ethnographic approaches and methods are on the rise, and some have suggested that collaborative fieldwork strategies offer distinct advantages, including being conducive to more validity in interpretations, analyses, and theory building [177].There are now even handbooks and guides to collaborative ethnographic research [178,179], as well as a journal, Collaborative Anthropologies. Hence, there is no good reason not to approach the ethnographic research on the embodiment of experience and the senses using collaborative teamwork that might include neuroscientists, cross-cultural psychologists, ethnographers, and contemplative phenomenologists.The inclusion of a neuropsychologist or a neuroanthropologist and a trained phenomenological ethnographer would seem to be the minimum requirement to assure the application of a Husserlian neurophenomenology. Conclusions Summarizing, I have argued that anthropology can only become a normal science of humanity when it develops methods that can link its theories to the structures of reality.Those structures are the physiological structures of the body and its brain.The shift in the discipline to the study of lived experience and the senses invites the inclusion of Edmund Husserl's contemplative phenomenology and its methods.These methods, which do require training in their application, allow the practitioner to isolate and study the essential structures of consciousness while, at the same time, suspending all the cultural sediments naturally layered atop the primordial given.Combining descriptions of essences with neurobiological research on the neural correlates of consciousness (in other words, Husserlian neurophenomenology) may prove a powerful approach to add to the ethnographer's toolkit. Operationalizing an ethnographically useful application of Husserlian neurophenomenology may prove challenging, for it, at minimum, requires training (or retraining) the ethnographer's gaze.Ethnographers with training in both contemplative methods and in neuroscience are thin on the ground.It is perhaps impractical at the present time to require an ethnographer to be proficient in both skills, although the time may come when both are routinely taught in anthropological curriculum.I have suggested several approaches, including (1) an ethnographer skilled in Husserlian methods collaborating with a neuropsychologist familiar with the research on NCC; (2) the use of experimental methods with volunteer hosts willing to learn Husserlian methods and then asked to study their sensory experiences and describe in their native language what they discover and what it means to them; and (3) the use of collaborative teamwork in the field so as to bring a number of skills to play, including ethnographic participant observation bolstered by Husserlian methods and neuropsychology.	2024-02-27T17:30:13.900Z	2024-02-17T00:00:00.000	{ "year": 2024, "sha1": "56c353d971d0248a020641141cd4c8ba94ac6611", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2673-9461/4/1/6/pdf?version=1708337266", "oa_status": "GOLD", "pdf_src": "ScienceParsePlus", "pdf_hash": "c7a5a7a5229ae72b48ac6b41c89d1020b24ad753", "s2fieldsofstudy": [ "Philosophy", "Psychology" ], "extfieldsofstudy": [] }
16906940	pes2o/s2orc	v3-fos-license	Seasonality of Rotavirus in South Asia: A Meta-Analysis Approach Assessing Associations with Temperature, Precipitation, and Vegetation Index Background Rotavirus infection causes a significant proportion of diarrhea in infants and young children worldwide leading to dehydration, hospitalization, and in some cases death. Rotavirus infection represents a significant burden of disease in developing countries, such as those in South Asia. Methods We conducted a meta-analysis to examine how patterns of rotavirus infection relate to temperature and precipitation in South Asia. Monthly rotavirus data were abstracted from 39 published epidemiological studies and related to monthly aggregated ambient temperature and cumulative precipitation for each study location using linear mixed-effects models. We also considered associations with vegetation index, gathered from remote sensing data. Finally, we assessed whether the relationship varied in tropical climates and humid mid-latitude climates. Results Overall, as well as in tropical and humid mid-latitude climates, low temperature and precipitation levels are significant predictors of an increased rate of rotaviral diarrhea. A 1°C decrease in monthly ambient temperature and a decrease of 10 mm in precipitation are associated with 1.3% and 0.3% increase above the annual level in rotavirus infections, respectively. When assessing lagged relationships, temperature and precipitation in the previous month remained significant predictors and the association with temperature was stronger in the tropical climate. The same association was seen for vegetation index; a seasonal decline of 0.1 units results in a 3.8% increase in rate of rotavirus. Conclusions In South Asia the highest rate of rotavirus was seen in the colder, drier months. Meteorological characteristics can be used to better focus and target public health prevention programs. Introduction Rotavirus causes a significant proportion of diarrhea in infants and young children worldwide leading to dehydration, hospitalization and in some cases death. It has been estimated that the proportion of hospitalizations for childhood diarrhea due to rotavirus has increased from 22% to 39% from 1986 to 2004 [1] and for 2004 there were an estimated 527,000 rotavirus-related deaths annually worldwide [2]. Rotavirus is a common infection in children worldwide and it is estimated that by age 5 nearly every child will have an episode of rotavirus infection [3,4]. Children in developing countries, such as those in South Asia, bear the greatest burden of deaths, up to 82%, from rotavirus infection, primarily due to malnutrition or lack of access to rehydration therapy [3]. A recent review on the burden of rotavirus diarrhea estimated that rotavirus is responsible for the death of approximately 122,000-153,000 children annually in India alone [4]. As of 2006, there are two live oral vaccines which have undergone large scale clinical trials and are being introduced globally [5,6]. These vaccines may have a significant impact in developing countries. Increased understanding of the epidemiology and, in particular, seasonal patterns of rotavirus in these countries is needed to ensure effective vaccine intervention programs [6]. Seasonal fluctuations in infectious diseases, and specifically in rotavirus infection, are a well-known and well documented phenomenon [7,8]. Rotavirus manifests at low endemic levels with pronounced seasonal outbreaks with considerable geographic variation. The virus is considered a winter disease with most cases observed in the coldest time of the year [7][8][9]. A review of 34 studies conducted worldwide found that rotavirus is more common in the cooler months in the temperate zones, yet the peak of disease can vary from autumn to spring [7]. The strong winter peak was seen primarily in the Americas, though in cities which cover various climates such as Toronto, Ontario, Canada and Houston, Texas, USA [7]. However in tropics (within 10N or 10S of the Equator) the pattern is less defined and autumn/spring peaks are more common [7]. A recent review considered the tropical region specifically and associated monthly disease incidence with meteorological variables for the same month. The analysis found that a 1uC increase in mean temperature resulted in a 10% decrease in rotavirus incidence and a 1 cm increase in mean monthly rainfall was associated with a 1% decrease in rotavirus incidence [8]. A review of 43 studies conducted in Africa found that rotavirus was detected year-round in nearly every country and generally exhibited distinct seasonal peaks during the dry months [10]. Another study of hospitalization for rotavirus which was conducted in three Australian cities found that admissions for rotavirus diarrhea peaked in winter and spring and was lowest in summer [11]. They also found that higher temperature and humidity in the previous week were associated with a decrease in rotavirus diarrheal admissions in all three cities [11]. The core of seasonality in infectious diseases is thought to be related to temporal oscillations in the governing transmission cycles of pathogenic agents and host susceptibility. Seasonal factors operate at many levels, creating temporal changes in human behavior, including sanitation and hygiene practices, and probability of environmental exposures [7,[12][13][14][15][16]. Despite growing attention to disease seasonality, a solid theoretical underpinning for seasonal peaks is limited. Better insight into the nature of disease seasonality can be gained from understanding the association with meteorological and environmental drivers. With increasing monitoring of rotavirus worldwide [10,17], a more detailed documentation of seasonality will soon be available. In developing countries consistent meteorological monitoring data is limited therefore satellite imagery can be advantageous and used as a proxy for the combined effects of temperature and precipitation. Remote sensing data are primarily used to derive characteristics related to vegetation cover, landscape structure, and water content. The Vegetation Index (VI) is a measure of density of plant growth which is calculated from remote sensing data. Very low values of VI (0.1 and below) correspond to barren areas of rock, sand, or snow. Moderate values represent shrub and grassland (0.2 to 0.3), while high values indicate temperate and tropical rainforests (0.6 to 0.8). Measures of the VI have been used in previous epidemiological studies [18] and has demonstrated predictive properties for onchocerciasis in Ethiopia [19], schistosomiasis in Brazil [20], West Nile Virus in New York City [21], and cryptosporidiosis for specific climates in Africa [22]. The objective of this study is to increase the understanding of seasonality of rotavirus diarrhea and associations with meteorological characteristics in South Asia. We assembled a set of time series representing monthly rate of rotavirus from 39 studies conducted in South Asia and examined the relationships between disease rate and meteorological factors, namely precipitation and ambient temperature. We also consider the association between rate of rotavirus and vegetation index, a remote sensing measure of the combined effects of temperature and precipitation. Outcome Data We conducted a meta-analysis of studies published on rotavirus infection from 1966 to 2010 in South Asia, defined as India, Pakistan, Bangladesh, Sri Lanka, Nepal, and Bhutan. Literature was gathered using the Ovid MEDLINE search engine. The keywords used in the search were rotavirus infections or rotavirus. These keywords were combined with seasonality and seasons, resulting in 567 citations. The search was further narrowed by adding the keywords Asia, Asia Central, Asia Western, Asia Southeastern resulting in a total of 152 citations after duplicate citations were removed. This database of studies was supplemented with references from review articles of rotavirus [7,8,16] and with studies from our own archives that included studies which do not appear in Ovid MEDLINE as they are published in South Asian scientific journals. This resulted in a total of 166 studies for evaluation. Studies were evaluated based on the following four criteria: 1) studies must provide observational data on rotavirus in humans who are not immuno-compromised (i.e., HIV/ AIDS patients); 2) studies must include at least a full year of data to cover all seasons; 3) studies must provide disease incidence or prevalence data on a daily, weekly, or monthly aggregation; 4) studies must provide data for a single location (i.e., hospital, village, etc). Studies which did not meet the criteria were excluded (Figure 1). A total of 40 studies met the study criteria of which 26 studies presented data in the published journal article (2 presented data on the same cohort; therefore the duplicate study was not included). We contacted the authors of 11 studies; 6 were able to provide data in the requested format. The remaining 8 studies provided monthly data aggregated over the entire study period. The final analysis comprised data from 39 studies. Data on rotavirus were aggregated on a monthly basis producing 47 time series, reflecting discrete time periods and locations (Table 1). Four studies provided data for more than one location or period of time. Exposure Data: Temperature and Precipitation Based on the latitude and longitude of the study location for each times series, we supplemented monthly rotavirus data with time specific monthly mean ambient temperature and monthly cumulative precipitation, obtained from the National Climatic Data Center databases (792 of 1046 months = 75.7%). When meteorological data specific to the time of the study were unavailable (254 of 1046 months = 24.3%), normals for monthly average temperature and cumulative precipitation were gathered from Global Historical Climatology Network. Each study location was classified based on the Köppen Climate Classification [23]. Using the study locations' latitude and longitude information each study location was superimposed on a map of Köppen Climate Classification and using a GIS spatial overlay each location was classified into a specific climate category. For the purposes of analysis we used only the four major classifications; moist tropical climates (Climate A), arid and semiarid climates (Climate B), humid mid-latitude areas (Climate C) and colder temperate areas (Climate D) (Figure 2). [81] Colombo, Sri Lanka solar illumination angles and sensor view angles over the period of record, caused by satellite overpass time drift. GIMMS processing also reduces NDVI variations that arise from sensor band calibration, volcanic aerosols, cloud cover and water vapor. We acquired GIMMS data on a bimonthly basis as a normalized difference vegetation index (NDVI) composite product with 1 km spatial resolution [24]. MODIS Terra data are hosted by NASA's Warehouse Inventory Search Tool (WIST) and are available from 2000 to present at temporal resolutions of 16 days and 1 month and spatial resolutions of 250 m, 500 m, 1 km and 0.25 degrees. From the MODIS Terra satellite, we acquired monthly composite Enhanced Vegetation Index (EVI) data at 1 km resolution. We chose 1 km spatial resolution to remain consistent with the GIMMS data. Although MODIS data are available on a bimonthly basis, we acquired monthly composite data to help remove errors caused by clouds, sun glare, and the satellite instrument itself [25] and to match the monthly rotavirus time series. The MODIS sensor has many advantages over the AVHRR satellite series for calculating vegetation indices. MODIS offers a higher spatial resolution and also has narrower bandwidths in the red and near infrared that improve its sensitivity to chlorophyll and produces less distortion from atmospheric water vapor [25]. The Normalized Difference Vegetation Index (NDVI) measures vegetation cover, and over time can be used to monitor plant growth. NDVI is calculated as follows [26]: NDVI~r nir{rred rnirzrred Where r nir is the reflectance in the near-infrared portion of the electromagnetic spectrum and r red is the reflectance in the red region. In the red region of the electromagnetic spectrum, sunlight is absorbed by chlorophyll, causing low red-light reflectance. In the near-infrared portion of the spectrum, the leaf's spongy mesophyll causes high reflectance. NDVI values range from 21.0 to 1.0, where healthy vegetation has increasingly positive values and decreasing negative values indicate rock, soil, snow, ice or clouds [27]. The Enhanced Vegetation Index (EVI) is a modification on NDVI that adds adjustment factors for soil and aerosol scattering. Liu and Huete [28] defined EVI as follows: where L is a soil adjustment factor. C 1 and C 2 are coefficients that correct the red band for aerosol scattering by factoring in the blue band. Reflectance values for the near-infrared, red, and blue wavelengths are designated by r nir , r red , and r blue , respectively. Usually, G = 2.5, C 1 = 6.0, C 2 = 7.5, and L = 1. Although MODIS Terra provides both NDVI and EVI products, we chose to analyze the EVI product for several important reasons. EVI is better at correcting for distortions caused by reflected light from particles in the air and from the ground cover below the vegetation. In areas with heavy vegetation, EVI is more sensitive to small changes and does not become saturated as easily as NDVI. NDVI responds mostly to variations in the red band which occur from chlorophyll, whereas EVI depends more upon the near-infrared band that is responsive to structural variations in the canopy [25]. Vegetation indices were not available for four studies prior to 1981. The only instrument acquiring daily global imagery was the AVHRR sensor on board the NOAA-6 satellite, which was launched in 1979. NOAA-6 has a daylight overpass time of 0730 hours whereas NOAA-7, launched in 1981, has a daylight overpass time of 1430 hours. NDVI products are not created from the NOAA satellites with early morning daylight overpass times (NOAA-6 and NOAA-8), because the lower solar zenith angle produces inconsistent and less intense radiance measurements than the afternoon solar illumination conditions from the other NOAA satellites [29]. Landsat multispectral scanner MSS offers an alternative to the AVHRR sensor for vegetation studies prior to 1981 [30]. However, due to its 18-day repeat cycle Landsat MSS does not produce enough cloud-free images to create a monthly time series for measuring seasonal vegetation dynamics [29]. Each study area was defined by locating the city on aerial imagery within Google Earth and creating a polygon around the city measuring 40 square kilometers that avoided large water bodies. ESRI's ArcMap 9.2 software and the Arc2Earth extension converted the polygon KML files to ArcGIS format. Bimonthly GIMMS data were acquired as geo-referenced TIF images in a geographic coordinate system, which are fully compatible with ArcGIS 9.2. Duke University's Marine Geospatial Ecology Toolset converted the monthly MODIS data from the HDF file format to ArcGIS rasters. For quality control, each study location polygon was overlaid onto the satellite imagery in conjunction with global country data to ensure proper alignment. We used Python scripts and ArcToolbox's Zonal Statistics function to create a time series of the mean vegetation index for each study area. For the GIMMS data, bimonthly statistics were averaged to create one monthly value. Data Standardization The studies selected for this analysis used different measures for rotavirus outcome. Of the 39 studies, 25 studies (64.1%) presented outcome data as number of cases, 13 studies (33.3%) presented outcome as percent positive stools, and one study (2.6%) presented the outcome as incidence. In order to standardize different outcome measures the raw values were normalized into z-scores on a time series-by-time series basis. The z-score was calculated using the mean and standard deviation for the complete duration (all months) of each study as follows: where Z ij is the z-score for the actual value for outcome, u ij (rotavirus cases, incidence or percent positive stools), for time series i in month j and u u i and s i are the mean and standard deviation for each time series, respectively. We also standardized exposure characteristics (temperature, precipitation and vegetation index) to adjust for location specific variations and assess relative associations [22]. Data Analysis: Mixed Effects Models The relationship between the normalized outcome and meteorological parameters was examined using a regression model adapted to time series data. We applied a linear mixed effects model to link the z-score of monthly rotavirus values with z-score of temperature, z-score of precipitation, and z-score of VI individually and control for the length of the time series at each study location. The mixed effects regression model was defined as follows: where Y ij is the rotavirus z-score, and X ij is the exposure of interest (temperature z-score, precipitation z-score, or VI z-score), the fixed effects: b 0 is the population intercept, b 1 is the population slope, and random effects: b 0 is the study intercept and b 1 is the study slope. The mixed effects model accounts for unequal lengths of times series included in the analysis, so studies which contribute less data have a smaller effect on the overall results. This model also allows us to explore the relationship between temporal fluctuations in rotavirus and meteorological conditions of each study as compared to the overall pattern for all studies. We examined the relationships overall for all locations as well as for tropical climates (Climate A) and humid mid-latitude climates (Climate C). As there were only two studies, for the same location, in the arid/semi arid (Climate B) category we did not examine relationships for this climate category. We considered four regression models. Model 1 was defined as the synchronous model and assessed the association between rotavirus z-score and temperature, precipitation, and VI z-score individually. In Model 2, the exposure was lagged by one month to assess whether increased rotavirus rates are associated with temperature, precipitation and VI exposure from the previous month. Model 3 was the synchronous model (Model 1) adjusted for the distance from the equator. In Model 3, we included the square of the study location's latitude to emphasize proximity to the equator and the interaction with the exposure in the regression model as a fixed effect for each individual exposure predictors to control for the location's distance from the equator. By adjusting for the distance from the equator we can account, in part, for heterogeneity in the interaction between annual temperature and precipitation levels and overall climate characteristics. Finally, in Model 4 the exposure was lagged by one month and the model was adjusted for the distance from the equator. Regression parameters were tested at a = 0.05 significance level. All analysis was conducted using R Software (Version 2.9.2). The regression parameters can be interpreted in terms of the zscore; a one unit change in z-score of an exposure variable is associated with an estimated change in z-score of dependant variable. A one unit change in z-score is equivalent to the standard deviation estimate of the given time series, thus to express the estimated effect in the original units of the explanatory variable the regression parameter can be divided by the value of the corresponding standard deviation estimate. Assuming that the majority of values of monthly measures of rotavirus infections are within the range of six standard deviations, we can then express the effect in terms of percent change by dividing the estimated effect in the original units of explanatory variable by 6 and multiplying the result by 100%. Data Analysis: Poisson Harmonic Regression To evaluate varying seasonal patterns based on study location, we conducted a seasonality assessment. Annual seasonal patterns were examined for study locations that included rotavirus data for 3 or more years using Poisson harmonic regression. There were six study locations which included sufficient data for detailed seasonality analysis. Seasonality is characterized as systematic, periodic fluctuations within the course of a year. It is assessed by several parameters: 1) the time when the seasonal curve reaches its maximum, 2) annual maximum value (peak), and 3) annual minimum value (nadir) [31]. These seasonal parameters are calculated based on values predicted by the harmonic regression: ln (E½y t )~b 0 zb 1 sin (2pvt)zb 2 cos (2pvt) zb 3 sin (4pvt)zb 4 cos (4pvt)ze t where y t is a time-series rotavirus z-score, t is time in months, v is frequency (v = 1/12), b 0 is intercept, b 1 , b 2 , b 3 , and b 4 are regression parameters, and e t is the error term. We use two harmonics to account for the double peak seen in several time series. Using the estimate for the regression parameters, we calculated peak timing and intensity based on the d -method [32]. The relative intensity, a measure of the shape of the seasonal pattern from peak to nadir, is calculated by dividing the estimated seasonal maximum value by the estimated seasonal minimum value. Results A total of 47 time series were used in this analysis to assess the relationship between monthly rotavirus rate in South Asia with ambient temperature and precipitation. All study locations provided a total of 1046 months of data; on average, each study location had 22 months of data. The scatterplots of independent variables with rotavirus z-score demonstrate a stronger association in temperature for both climate categories than in precipitation and vegetation index (Figure 3). Descriptive analysis illustrates the average values of temperature, cumulative precipitation, and vegetation index for all study locations and also for each climate subcategory ( Table 2). As all studies are from South Asia, there is little difference in the average temperature range by climate categories. The primary variation by climate category is seen in precipitation. The vegetation index is similar for moist tropical areas (A) and humid mid-latitude areas (C); but averaged much smaller values in the arid and semiarid region (B). The highest levels of rotavirus are seen in the colder, drier months of the year (,December-March). When assessing synchronized relationships (Model 1), all exposure parameters, temperature, precipitation, and VI z-score were significant negative predictors of rotavirus z-score and this relationship held for both the tropical (Climate A) and humid-mid latitude (Climate C) climates (Table 3). For all studies, a one unit increase in temperature z-score, precipitation z-score and VI z-score resulted in a 20. 20.193) unit decrease in the rotavirus z-score, respectively. Therefore, the overall effect of a 1uC decrease in monthly ambient temperature is associated with 0.078 unit increase above the annual rotavirus level, which is proportional to 1.3% increase in rotavirus infections. Similarly, a decrease of 10 mm in precipitation is associated with 0.3% increase in rotavirus infections. A decrease of 0.1 units in vegetation index for a seasonal change -compared to the annual norm of 0.3 -is associated with a 3.8% increase above the annual rotavirus level. In the tropical climates (Climate A) temperature is more strongly associated with a decrease in rotavirus and in the humid mid-latitude climate (Climate C) temperature and precipitation have similar negative association with rotavirus z-score. In the tropical climates a decrease of 1uC from annual average of 26.2uC in monthly ambient temperature is associated with 0.13 unit increase above the rotavirus annual level, or a 2.2% increase in rotavirus infections, while in the humid midlatitude climate (Climate C) a decrease of 1uC in temperature below annual mean of 24.2uC results in lesser but significant increase of 0.7% increase. VI demonstrated a similar negative association with rotavirus z-score in both climates. All three exposures remained significant when lagged associations were considered (Model 2) though the strength of the relationships decreased. This was also seen for both climate categories (Table 3). Overall for all locations, when adjusting for latitude (distance of the study location from the equator (Model 3)) only the relationship with temperature z-score remained significant suggesting that the association is stronger for locations farther from the equator. This is seen clearly when considering the relationship by climate category; in the humid mid-latitude climate (Climate C), which is further from the equator, the relationship with temperature is stronger than in the tropical climate (Climate A), which are locations closer to the equator. Overall for precipitation z-score and VI z-score the relationship becomes insignificant after adjusting for latitude. Similar associations were seen after adjusting for latitude and lagging the exposure variable (Model 4). Poisson harmonic regression was used to analyze the seasonal patterns for six time series which provided monthly rotavirus data for more than 3 years. The seasonal patterns vary by location but typically demonstrate a peak in the winter months. The seasonality assessment demonstrated a winter peak for all the locations ( Table 4). The estimate for peak timing of seasonal peak ranged from late October to late February. Figure 4 shows the annual time series of rotavirus z-score superimposed for the years of the study for 2 locations, Chandigarh, India which is in the humid mid-latitude climate (C) and Matlab, Bangladesh which is in the tropical climate (A). Discussion Our findings provide a quantitative link between the rate of rotavirus and meteorological parameters in South Asia. We found that higher rate of rotavirus is associated with cold, dry months. Our study demonstrates that the strength of these associations vary by climate category. The use of standardized z-scores allows for direct comparisons of detected effects across geographical zones with different weather conditions and across various measures of disease. This study also demonstrates the utility of the remote sensing vegetation index in assessing seasonal associations. The seasonal pattern seen in rotavirus varies by climatic region and is also associated with local weather. Our findings agree with a review of studies in tropical regions which found that a reduction of rotavirus rates was associated with increases in temperature and precipitation [8]. Our analysis focuses on South Asia, which includes both temperate and tropical areas, whereas the Levy et al. study includes just the tropics; only seven out of 39 studies were included in both analyses. The similarity of findings indicates that both studies capture a unique and stable phenomenon suggesting an effect of climate and local weather on the seasonal pattern of rotavirus infection: an increase in rate of rotavirus in cooler and drier times of the year for a given location. The Levy et al. study acknowledged the diversity of climates prevalent even within the defined tropical band [8]. In our study we controlled for various climatic regions by classifying each study using Köppen Climate Classification. Most of the study locations we examined fell into only two broad classifications, tropical climates (Climate A) and humid mid-latitude climates (Climate C). With the limited number of studies we were only able to use the broadest categories though the Köppen Climate Classification system provides more detail. Within the broad tropical climate (A) category there are three subcategories which are distinguished by precipitation patterns and in the mid-latitude climate (C) category there are six subcategories which are distinguished by type of summer as well as precipitation patterns [23]. However, we did demonstrate different degree of association between meteorological characteristics and rotavirus prevalence by climate category. In tropical climates (Climate A) an increase in temperature was associated with a greater decrease in rotavirus than in humid mid-latitude climates (Climate C). We also found that temperature is a stronger predictor of rate of rotavirus after controlling for the study latitude or distance from the equator in the humid mid-latitude climate (C) whereas the relationship decreased for the moist tropical climate (A); this suggests the viral transmission pattern might depend on broad range of environmental conditions, which can also be drivers for temporality in social behaviors and health practices. The primary reasons for seasonal fluctuation in rotavirus prevalence are still unknown. Typically for a virus, it is not expected that the seasonal relationship would be primarily driven by a favorable environment for the pathogen. Temporal and geographic trends have been better recognized in temperate than tropical climates, perhaps because though temperatures vary across seasons, temperatures do not reach the lower levels seen in temperate climates, or patterns of disease prevalence are not as marked. It has been suggested the seasonal pattern seen in rotavirus may be driven by airborne transmission of the disease [16,33,34]. The drop in humidity and rainfall dries the soils which may increase the aerial transport of the contaminated fecal matter. In the U.S. the seasonal pattern in rotavirus shifts by geographic location with peak activity occurring first in the Southwest from October through December and last in the Northeast between March and May [13,34]. Influenza has also demonstrated strong seasonal patterns [15,35,36] and associations with humidity [37]. A seasonal pattern similar to that of rotavirus has been demonstrated for influenza in the U.S., with peak activity beginning in the Southwest in mid-December and shifting to the Northeast towards mid-January [36]. The seasonal pattern seen in rotavirus may be influenced by socio-demographic factors. Pitzer et al. suggest that the seasonal pattern seen in rotavirus in the U.S. is influenced by spatiotemporal patterns in birth rate and that young children experiencing their first infection of rotavirus are the primary drivers of epidemics and therefore of the seasonal patterns [9]. As mentioned by Bobak et al., in populations with reliable data, birth rates vary by season of the year [38]. In the U.S. most births occur in the summer and early fall and the fewest births are seen in the spring which can be explained by temperature and light effects [12,39]. Seasonal patterns in birth rates have been demonstrated in South Asia, though specific regional patterns are not clear [12,39]. In regions with distinct periods of very high temperatures, a drop in births at this time can be noticeable, especially in developing countries, which may, in part, explain our findings in the tropical (A) climates. Seasonality of newborn susceptibility might be insufficient to warrant disease seasonality. It is likely that seasonal patterns can be reinforced by changes in rotavirus genotype. In the temperate climate of England and Wales, where rotavirus infections are highly seasonal, a child's risk of a laboratory-confirmed rotavirus infection in the first year of life was significantly higher for children born in summer compared to those born in the winter [40]. The authors suggested that maternal immunity and age-specific levels of exposure to rotavirus could explain the findings. Our preliminary findings suggest that the seasonal pattern for rotavirus may vary by strain type [41]. A similar meta-analysis approach may be used to assess the differing seasonality by strain type. School schedules and holiday travel may also affect the seasonal pattern of rotavirus infections by changing the probability of exposure to the virus. Other viral diseases, influenza in particular, have demonstrated associations with holiday travel and school schedules [42]. In many countries, vacations and family travels are somewhat synchronized and it may be difficult to separate the independent effects. Given the potential airborne transmission of rotavirus it can be expected that the seasonal pattern may be driven by crowdedness. Rotavirus is most prevalent in children under 5 years of age. Over 70% of the studies considered were Table 4. Estimated peak timing from Poisson harmonic regression for studies with more than 3 years of data including study reference, study location, study duration, estimated peak and 95% confidence interval, p-value, calendar interpretation of estimated peak and relative intensity for both estimated peaks. conducted in children less than 5 years of age and therefore, our results are mostly driven by the seasonal patterns seen for this age group. The degree to which seasonal patterns in children under 5 years of age in South Asia can be affected by school schedules and holiday travel needs to be further investigated. The studies used in our analysis covered a large time span ) and it was difficult to gather temperature and precipitation data specific to the study time for the earlier study periods and studies in more rural areas. For the studies in which averages were used, temperature and precipitation may not represent the weather specific to the year for which rotavirus data is available. This may limit the precision of our estimates of associations with meteorological factors. Humidity has also been shown to be associated with rotavirus seasonality [8,11]. An analysis of birth cohort data from Vellore, India, demonstrated a positive correlation between humidity and the strain-specific incidence of rotavirus diarrhea [41]. However, humidity is another factor for which consistent data was difficult to obtain for all study locations. To overcome this limitation, we consider the use of remote sensing data to provide a proxy for meteorological characteristics for areas in which direct measures of temperature and precipitation are not available. The Vegetation Index (VI) is a measure of the combined effect of temperature and precipitation and is therefore a strong proxy for humidity as well. Study specific temperature and precipitation data was available for only 75% of study months whereas VI measures were available for 93% of study months. A strength of this analysis is the demonstration of the use of VI as a proxy for the combined effect of temperature and precipitation. The precision of our estimates may also be affected by the coarse monthly temporal aggregation. The incubation period for rotavirus is short, approximately 2 days, and the disease is highly infectious which suggests that the seasonal peak of disease may be very short [33]. A finer temporal aggregation, of weeks or days, for rotavirus data would allow us to capture the short peak more accurately using Poisson harmonic regression. Our sub-analysis of seasonality characteristics in locations with sufficiently long time series data demonstrates that the form of the seasonal pattern might be more complex then typically observed in temperate climates, which usually exhibit one well defined winter peak. The ongoing global rotavirus monitoring and surveillance efforts will soon provide valuable information for detailed seasonality characterization. The seasonal pattern seen in rotavirus may be influence by our ability to measure infection in a systematic manner. Due to the long time span covered by studies used in this analysis, 1976-2009, various diagnostic methods were used for rotavirus. Enzyme immunoassay is currently the most common technique while earlier studies were likely to use electron microscopy. It is estimated that a child with rotavirus diarrhea excretes 10 9 to 10 12 viruses per gram of stool during the acute phase of diarrhea [16,33]. The enzyme immunoassay, which is based on VP6 antigen detection, and electron microscopy, by which the 70 nm viral particles are visualized, both have a limit of detection of 10 5 to 10 6 viral particles per gram of stool, and hence are unlikely to miss cases of rotavirus diarrhea in acute illness. Since both methods have similar limits of detection, it is unlikely that differences in methods would affect detection of seasonal patterns of infection. The use of various diagnostic methods required over time required standardization of data to z-score. While necessary for this analysis, this approach reduces the effect of temporal heterogeneity but can't eliminate potential bias, which might lessen the observed effects. To improve our understanding of rotavirus seasonality, better capacity and standardization of global monitoring and surveillance efforts are required [17]. Understanding seasonal patterns of rotavirus allows for testing of efficacy of vaccines currently in use [5,6]. The details of seasonal patterns can be used to target vaccination programs, test efficacy of the vaccine implementation programs and the vaccine itself [6]. In particular, it may be useful to consider appropriately timed . Annual time series of rotavirus z-score superimposed for years of the study for two locations. Panel A is data for a study conducted in Chandigarh, India [44] and Panel B is data for a study conducted in Matlab, Bangladesh [65]. Data are shifted by 6 months in order to center the winter peaks. doi:10.1371/journal.pone.0038168.g004 immunization booster programs in settings which have reported poor efficacy of rotavirus vaccine which wanes further in the second year of life, and have demonstrated strong seasonality [43]. The typical schedule for rotavirus immunization in developing countries includes vaccination at 6, 10, and 14 weeks of age and is constrained by the very young mean age of primary rotavirus infections and the increased background risk of adverse events after six months of age. However, if booster vaccination programs were to be considered for older children lacking immunity, prerotavirus season vaccination may be substantially more effective than post-season vaccination. Proper assessment of seasonality and dependency on climate characteristics also provides insights on effects of vaccination strategies, including potential shifts in seasonal peaks and duration of outbreaks. Conclusion In this study, we found the highest rate of rotavirus was seen in the colder, drier months in South Asia. This study capitalizes on previously published methodology for assessing the relationships between meteorological characteristics and disease outcomes [22,31]. With expansion of remote sensing and improved weather monitoring, global weather forecasting on a very refined spatial scale is feasible. A better understanding of the links between disease ecology and environmental conditions warrants the discovery of underlying governing principles of transmission of rotaviral infection. Based on the findings, higher of rates of rotaviral infection during the relatively cold and dry periods in South Asia, innovative strategies for global disease forecasting, early warning systems, timing of vaccination programs, and other preventive measures can be developed at the regional level.	2016-05-04T20:20:58.661Z	2012-05-31T00:00:00.000	{ "year": 2012, "sha1": "fbd8b481fd85a6d4517e97c86a56eac1b75f7651", "oa_license": "CC0", "oa_url": "https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0038168&type=printable", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "fbd8b481fd85a6d4517e97c86a56eac1b75f7651", "s2fieldsofstudy": [ "Environmental Science" ], "extfieldsofstudy": [ "Biology", "Medicine" ] }
114768770	pes2o/s2orc	v3-fos-license	Transient two-phase CFD simulation of overload operating conditions and load rejection in a prototype sized Francis turbine An increasing shift in operating conditions of hydropower turbines towards peak load operations comes with the necessity for numerical methods to account for such operations. This requires modifications to state-of-the-art CFD simulations. In the first part of this paper a 1D hydroacoustic model to represent the pressure oscillations in the penstock was introduced and coupled with a commercial CFD solver. Based on previous studies, various changes in cavitation and turbulence modeling were done to influence the behavior of a cavitating vortex rope typically occurring at high load conditions of a Francis turbine. In the second part, mesh motion was added to this model to simulate a load rejection starting from full load conditions. It was shown that additional extensions to the 3D CFD model are compulsory to model specific operating conditions as well as transient operations. Thus, accordance with measurement data at overload operation was improved and only small deviations remained. For the load rejection the maximum overspeed was well captured and the comparison of guide vane torques with model test measurements showed a sufficient agreement. With the gained insights, occurring effects which influence the performance and the life-time can be detected and conclusions for the hydraulic design as well as the operating mode can be drawn. Upcoming studies will focus on evaluating the flow field in detail and on reducing the remaining deviations by further extending the mathematical model. Introduction The increasing number of wind turbines and solar power plants require an extended operation range for hydro-power turbines to stabilize the power grid. This leads to additional challenges given that this operating modes need greater care regarding transient flows such as pressure peaks and fluctuations. CFD methods are state of the art to predict the hydraulic performance of a turbine at steady-state operation, but for more demanding cases like transient operations they still have some shortcomings. For instance if it is necessary to consider the interaction with the hydraulic waterway up-/downstream of the turbine or if due to load changes the guide vane opening angle is varying. So far not many investigations have been made on this field. Flemming et al. [1] investigated the influence of two-phase modeling for overload surge trying to extract cavitation parameters like mass flow gain factor and cavitation compliance. Ruprecht et al. [2] considered the response of the waterway passage by means of a Method of Characteristic approach (MoC), where Chirkov et al. [3] used an one-dimensional hyperbolic system and a two-phase approach. The investigations of the latter two research teams were done with inhouse CFD solvers, in model size and without a comparison to measurement data. Transient operations considering moving guide vanes were performed by Nicolle et al. [4] simulating a start-up procedure by using a new mesh for every time step. Li et al. [5] were focusing on a load rejection process, modeling the whole plant including the penstock in a 3D manner with a tetrahedral unstructured mesh and dynamic mesh functionality. Similar investigations were carried out by Lui et al. [6] using measured pressure signals as inlet boundary condition to the spiral case. Zhang et al. [7] coupled a MoC approach to the 3D solver, considering a compressible water model and applying a dynamic mesh approach to simulate moving guide vanes during the load rejection. So far no investigations have been published that describe the coupling of a commercial CFD code to finite difference system, which is showing more flexibility in space and time discretization than the MoC approach. The same holds true for a direct comparison between prototype CFD results and measurement site data for the mentioned turbine operations. The operation conditions discussed in this paper are overload and load rejection of a prototype sized Francis turbine. To accomplish this, a one-dimensional (1D) hydroaccoustic model to represent the waterway upstream of the turbine is developed and coupled to a three-dimensional (3D) commercial CFD solver. In a second step an already available moving node approach of the commercial code is extended by an automated remesh procedure to represent moving guide vanes. Results of both operating conditions are compared with measurement site data. Benefits from this work are the possibility to represent conditions which were so far not able to simulate and therefore give the opportunity to get a full insight into the unsteady 3D flow field. Hydroacoustic Modeling Derived from the mass and momentum equation and after neglecting the convective terms because of high wave speeds at low flow velocities, the dynamic behavior of a pipe element with length dx can be described according to eq (1): As discussed in Nicolet [8] and Moessinger et al. [9] the hyperbolic partial differential equation system is discretized in space using a centered scheme on a staggered grid. Temporal differentiating is done using second order implicit Euler to be consistent with the commercial CFD solver. The resulting implicit matrix equation is solved with a Gauss-Seidel method using an additional under-relaxation factor to ensure stability and convergence. Test-cases to validate the 1D code with one-dimensional commercial code (SIMSEN) and 3D commercial CFD solver (ANSYS CFX) have been applied and described in previous studies [9]. 3. Francis turbine at overload condition 3.1. Case study At high load operation, Francis turbines feature a large axisymmetric cavitating vortex rope beneath the runner. Under certain conditions this cavitation volume can act as an internal energy source leading to instabilities. Serious pressure oscillations may occur due to the interaction of the oscillating vortex rope with the hydraulic system. When simulating this operating point it is crucial to account for the water-dynamic upstream, otherwise constant boundary conditions lead to unstable (as seen in figure 1 for constant discharge at the inlet) or unreliable results. For this investigation a Francis turbine of the upper power range with unbranched bare steel penstock and a draft tube with direct outflow to the tailwater has been chosen. During Table 1 is selected. Numerical setup The previously described and validated coupling is applied to model the above introduced power plant. To avoid scaling effects and take to the buoyancy into account, prototype sized modeling is chosen, which allows a direct comparison with measurement data. Considering the axisymmetric appearance of the vortex rope, it is assumed sufficient to model only one passage of stay vane, wicket gate and runner by applying periodical boundary conditions. The draft tube is modeled completely including elbow and piers. Transfer between stationary and rotating frame is realized by circumferentially averaging the flow variables (stage interface, wicket gate to runner) and by transient rotor-stator coupling (runner to draft tube), accounting for all transient flow characteristics. To capture the dynamic interaction between the cavitation volume and discharge variations together with pressure oscillations, a multiphase simulation is carried out. This comes along with the additional advantage that non-physical vorticity peaks in the vortex center can be avoided. The mass transfer between the phases representing evaporation and condensation is described by the Rayleigh-Plesset equation [10] and the Zwart-Gerber-Belamri cavitation model [11]:ṁ where the model constants are as follows: Turbulent scales are modeled by the SST k-ω-model with automatic wall function in previous studies and a scale-resolving SAS-SST model at current simulations. In general, ANSYS CFX 15.0 uses a finite volume based discretization scheme up to second order accuracy for convective fluxes and truly second order accuracy for diffusive fluxes. Time dependent computations are performed with a second order accurate time scheme. The computational grid consists of three domains (tandem cascade, runner and draft tube) with block-structured hexahedral cells. The mesh size of SVWG (865,000), RU (606,000) and DT (11,027,000) was kept constant for all investigations. For the upstream 1D penstock inlet boundary condition, total energy is prescribed. Based on a measured pressure at the spiral case, pipe losses and the expected discharge according to the hill chart, the fixed total specific energy follows as: At the 1D-3D interface an averaged pressure extracted from the 3D stay vane inlet is set as a 1D pressure boundary condition. A discharge return value of the 1D model in addition with an appropriate inflow angle serves as the 3D domain inlet condition. As buoyancy is considered the hydrostatic pressure profile at the draft tube outlet is set in accordance to the measured tail water level. iso-surface represents the second invariant of the velocity gradient tensor [12] (Q criteria): Flow visualization where the color visualizes the ratio between the turbulent eddy viscosity ν t and the laminar molecular dynamic viscosity ν: As the color range is in a logarithmic scale it is obvious, that for RANS simulations especially in the necking area of the vortex rope large viscosity ratios indicate a strong influence of the turbulence model. For the SAS-SST model, the viscosity ratio was significantly reduced. Additionally, downstream of the cavitation volume a vortex core of helical shape is visible. This phenomenon has been found as well by applying a SST-SAS turbulence model to Kaplan turbines [13]. Transient simulation 3.4.1. Previous results In previous studies [9] the pressure oscillations were damped out when applying the 1D3D coupling (cf. figure 3) as a consequence of the underestimated rope frequency in the 3D-CFD simulation. To investigate the influence of the system eigenfrequency, three According to eq. (8) cavitation volume and the associated pressure level affect the frequency. Various modifications regarding the cavitation and turbulence modeling have been made to influence the vortex rope behavior. Figure 4 shows the FFT of the dynamic pressure at draft tube location for the measurement and 1D3D coupling of previous and current results. Compared to the measurement, the modifications substantially improve the rope frequency. After an Figure 4: FFT for dynamic pressure at the draft tube measurement location initialization phase, the amplitudes reach a stable level with some underlying low frequency. Figure 5 shows a smaller time interval of the simulated and measured pressure oscillations. The slightly underestimated amplitudes could be a consequence of the remaining discrepancies in the frequency range. Further studies will include different cavitation models and take the compressibility of both phases into account. Francis turbine at load rejection 4.1. Case study Load rejection occurs when a turbine in operation is suddenly disconnected from the grid. To limit the resulting overspeed of the runner, guide vanes have to close at the maximum feasible speed and prevent undesired pressure peaks at the same time. As shown in figure 6, during guide vane closure the turbine accelerates to an overspeed of 150% of the nominal runner speed. Starting at overload, the vortex rope in the draft tube collapses, the pressure suddenly drops and rises with further closure of the guide vanes to a stable level as the guide vanes are fully closed. Figure 6: measured values for load rejection process Numerical setup After disconnecting from the grid, runner speed is no longer constant but depends on the runner torque and runner/generator inertia. This can be expressed in an explicit equation which will be solved after every time step. The flow field during the load rejection is highly instationary and asymmetric. Therefore it is necessary to take all stay vanes, guide vanes and runner passages as well as the spiral case into account. This sums up in a total mesh size for the block structured hexahedral grid of 25.2 · 10 6 cells. Transition between the domains are set to general grid interfaces (GGI), where the transfer between stationary and rotating domain is handled by transient rotor stator coupling, accounting for all transient flow characteristics. To ensure stability, turbulence scales are modeled by the SST k-ω-model with automatic wall functions and a time step which corresponds to 1/72 of the initial runner revolution. Adaptive grid method Moving guide vanes during load rejection are realized with a combination of node motion and automated remeshing. ANSYS CFX already comes with the ability to deform meshes [15]. While the boundary movement is specified, all remaining nodes shift according to a displacement diffusion by solving the equation: Mesh stiffness Γ disp exponential increases as the control volume size decreases: For this case, larger control volumes absorb more mesh motion. An increase of the stiffness model exponent C stif f from 2 (default) to 15 yields to a much more abrupt stiffness variation, ensuring a desirable mesh quality along the guide vane closure. However during the load rejection process, starting from overload, total guide opening angle varies over 32 • . This angular rotation is too large to ensure a good mesh quality along the complete guide vane closure using only the mesh deformation method. Therefore an automated remesh process is introduced, allowing a remesh procedure during runtime (cf. figure 7). Depending on an arbitrary logical condition (maximum runtime, minimum cell angle) the simulation is paused and with the aid of an in-house mesher, depending on the current guide vane angle, a new mesh is generated. After loading the mesh into CFX and interpolating the old results, the simulation proceeds. lead to a massive overprediction of the runner speed. By coupling the 1D code to represent the pressure rise due to the closing wicket gates a substantial improvement of the speed prediction is achieved. A pure 1D approach (SIMSEN) is also able to predict the maximum overspeed well, but shows some discrepancies when the runner speed is decreasing. For a final statement on how the predicted runner speed of the CFD simulation will develop, more simulation time is necessary. Figure 9 shows a comparison between model test measurement and prototype CFD of the average torque of four wicket gates. From the beginning the torque is slightly overpredicted Figure 9: comparison of wicket gate torque between model test measurement and prototype CFD with increasing deviations during the closure process. However the development from opening to closing tendency is captured sufficient well. Steady state simulations of different openings will follow and give an indication if these deviations are related to the transient closing procedure. Furthermore, time step size variation and mesh adaption are other starting points to improve the simulation. Thereafter a more detailed comparison of additional measurement points will be possible. Conclusion A previously developed and validated 1D-3D coupling method was further analyzed. When applied to a prototype sized Francis turbine at overload conditions, accordance with measurement data was improved by adjustments regarding cavitation and turbulence modeling. Upcoming studies about different cavitation models and the consideration of compressibility for both phases are aiming at further reducing the remaining deviations. In a second application, full condition load rejection was investigated. The guide vane closure was taken into account by moving mesh motion. The runner speed was modeled as a function of runner torque and runner/generator inertia. Simplified models led to a significant overprediction of maximum runner torque. By considering a full 360 • computational domain and including the penstock by the already introduced 1D code, a better accordance with measurement data and commercial 1D solver (SIMSEN) was achieved. Also a comparison of guide vane torque showed at the beginning good agreement, with increasing deviations during the load rejection process. After a detailed evaluation of the simulation more effort has to be put in the computational model to ensure stable convergence and reasonable results along the whole closing procedure.	2019-04-15T13:08:05.615Z	2016-11-01T00:00:00.000	{ "year": 2016, "sha1": "fd9722946bf24778848ef0e57484266e112089b4", "oa_license": null, "oa_url": "https://doi.org/10.1088/1755-1315/49/9/092003", "oa_status": "GOLD", "pdf_src": "IOP", "pdf_hash": "87ab6541fe5e64644b9489b71c37d7cbf4e18d94", "s2fieldsofstudy": [ "Engineering" ], "extfieldsofstudy": [ "Physics", "Engineering" ] }
81476244	pes2o/s2orc	v3-fos-license	A study on knowledge and practices of over the counter medications among 1 st year medical students There is no regulation for the use of OTC drugs in India. The poor economic status and busy lifestyle of an individual makes him rely on the OTC drugs. In India, it has been shown that literate people were 76% more likely to selfmedicate than illiterate people. Selfmedication is widely practiced worldwide and often considered as a component of self-care. The World Health Organization (WHO) has appropriately pointed out that responsible self-medication can help prevent and treat diseases that do not require medical consultation and provides a cheaper alternative for treating common illnesses. The practice of self-medication must be based on authentic medical information otherwise irrational use of drugs can cause wastage of resources, increased ABSTRACT INTRODUCTION The non-prescriptive drugs or over-the-counter drugs (OTCs) are the drugs that are purchased without prescription. There are currently more than 300000 different OTC drugs available only in US. 1,2 There is no regulation for the use of OTC drugs in India. The poor economic status and busy lifestyle of an individual makes him rely on the OTC drugs. In India, it has been shown that literate people were 76% more likely to self-medicate than illiterate people. 3 Selfmedication is widely practiced worldwide and often considered as a component of self-care. 4 The World Health Organization (WHO) has appropriately pointed out that responsible self-medication can help prevent and treat diseases that do not require medical consultation and provides a cheaper alternative for treating common illnesses. 5 The practice of self-medication must be based on authentic medical information otherwise irrational use of drugs can cause wastage of resources, increased resistance of pathogens, and can lead to serious health hazards such as adverse drug reaction and prolonged morbidity. Self-medication assumes a special significance among the medical students as they are the future medical practitioners and have a potential role in counselling the patients about the advantages and disadvantages of selfmedication. Medical students also differ from the general population because they are well-exposed to the knowledge about diseases and drugs. 6 Therefore, the present study was taken up to analyze the use of OTC drugs among medical students in a tertiary care teaching hospital. The Present study is planned to provide baseline data regarding use and knowledge of over the counter drugs among 1 st year MBBS medical students. Present study is done to find out pattern and extent of use of OTC drugs and other than OTC drugs among the medical students of a tertiary care teaching hospital. Objective of the study was to analyze the use of over the counter drugs among medical students in a tertiary care teaching hospital. METHODS This was descriptive, observational and cross sectional and non-interventional study. Study was carried out at Rajarajeswari Medical College and Hospital. All the First year MBBS students of Rajarajeswari Medical College were chosen for conducting the study. Duration of the study was two months (February-March 2018). It was Semi structured and self-administered questionnaire study. Inclusion criteria All 1 st year MBBS students (246) of Rajarajeswari Medical College were included in this study. Exclusion criteria Those not willing or giving consent to participate in the study. Sampling method A total of 246 students were present in the first year and complete enumeration method was chosen since all of them were being included in the study. Study protocol The study was conducted after getting approval from the Institutional ethical committee. All the students participating in the study were explained clearly about the purpose and nature of the study in the language they could understand. They were included in the study only after obtaining a written Informed Consent Form. This study was used to conduct the study and the questionnaire consisted of questions which included basic details of the student and questions regarding their knowledge and practice of OTC drugs. A total of 250 students were present in the first year but 246 students were included in the study as 4 were absent despite 3 consecutive attempts. Data analysis At the end of the study, all the data was collected and compiled in MS Excel Sheet and results were analyzed using Statistical Package for Social Science (SPSS) version 21.0. All qualitative variables were presented using descriptive statistics with frequency and percentages. DISCUSSION There were 45.6% using OTC as it is time saving which is in accordance with the findings of a study from Boudha, Kathmandu where 48% of respondents used OTC medicines due to ease to get the medicine. 7 Only 9% of the respondents always check the expiry date before using any OTC drugs which differs to the result obtained from a study done in Malaysia. 7 In the present study, 84% were aware about the OTC drugs similar to the study conducted in Lagos State, Ikeja, and West Bengal. 8,9 Around half of the respondents (44%) use OTC when symptoms are minor and manageable comparable to study by Ghosh et al. 10 Many of the respondents (72.8%) used OTC drugs for fever (antipyretics) i.e., paracetamol which is in line with the study from Pune, West bengal, and Lagos Residents. 8,9 The current study showed that more than half of the respondents (54%) had good knowledge on overthecounter drugs which is similar to a study from Nigeria and Taiwan. 10 The present study showed that nearly half of the respondents (46%) had level of knowledge regarding overthe-counter drugs from average to poor. However, study by Kasabe et al, found that majority of the respondents stated that their level of knowledge was moderate to low. 11 The study showed that about half (47%) of the respondents had good practice on over-the-counter drugs and similar results were obtained in a study from Nigeria and Taiwan which is also in line with study conducted by Bhambhani et al. 12 CONCLUSION The results show that students are lacking in the concept of over-the counter drugs, consultation of doctor before using the OTC drugs, its adverse effects, caution to be taken, reading the instruction before use, checking the expiry date etc. In order to control this problem, appropriate measures have to be taken and the prescription system has to standardized and implemented. Conducting awareness programs and restricting drug advertisements for public can be recommended.	2019-03-18T14:04:22.801Z	2018-11-24T00:00:00.000	{ "year": 2018, "sha1": "4d1e55a011b43a1a9545f22a84d569ab10016a03", "oa_license": null, "oa_url": "https://www.ijbcp.com/index.php/ijbcp/article/download/3026/2195", "oa_status": "GOLD", "pdf_src": "MergedPDFExtraction", "pdf_hash": "2512b82f8140e3da16efa15d39ee3df1a611bf96", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
253386030	pes2o/s2orc	v3-fos-license	A genomic approach to analyze the cold adaptation of yeasts isolated from Italian Alps Microorganisms including yeasts are responsible for mineralization of organic matter in cold regions, and their characterization is critical to elucidate the ecology of such environments on Earth. Strategies developed by yeasts to survive in cold environments have been increasingly studied in the last years and applied to different biotechnological applications, but their knowledge is still limited. Microbial adaptations to cold include the synthesis of cryoprotective compounds, as well as the presence of a high number of genes encoding the synthesis of proteins/enzymes characterized by a reduced proline content and highly flexible and large catalytic active sites. This study is a comparative genomic study on the adaptations of yeasts isolated from the Italian Alps, considering their growth kinetics. The optimal temperature for growth (OTG), growth rate (Gr), and draft genome sizes considerably varied (OTG, 10°C–20°C; Gr, 0.071–0.0726; genomes, 20.7–21.5 Mpb; %GC, 50.9–61.5). A direct relationship was observed between calculated protein flexibilities and OTG, but not for Gr. Putative genes encoding for cold stress response were found, as well as high numbers of genes encoding for general, oxidative, and osmotic stresses. The cold response genes found in the studied yeasts play roles in cell membrane adaptation, compatible solute accumulation, RNA structure changes, and protein folding, i.e., dihydrolipoamide dehydrogenase, glycogen synthase, omega-6 fatty acid, stearoyl-CoA desaturase, ATP-dependent RNA helicase, and elongation of very-long-chain fatty acids. A redundancy for several putative genes was found, higher for P-loop containing nucleoside triphosphate hydrolase, alpha/beta hydrolase, armadillo repeat-containing proteins, and the major facilitator superfamily protein. Hundreds of thousands of small open reading frames (SmORFs) were found in all studied yeasts, especially in Phenoliferia glacialis. Gene clusters encoding for the synthesis of secondary metabolites such as terpene, non-ribosomal peptide, and type III polyketide were predicted in four, three, and two studied yeasts, respectively. Introduction Studies on the strategies exhibited by microbial communities inhabiting cold regions worldwide to thrive at low temperatures have increased in recent decades. Cold environments, characterized by temperatures below 5°C, predominate on Earth and represent a harsh habitat for microbial life leading to a strong reduction of growth rates (Feller and Gerday, 2003;D' Amico et al., 2006;Margesin and Feller, 2010;Margesin and Miteva, 2011). Cold-adapted (i.e., psychrophilic) microorganisms have evolved molecular and metabolic adaptations to overcome the adverse effects of low temperatures, including the synthesis of cryoprotectant molecules, cold-active enzymes, and polyunsaturated lipids regulating the degree of membrane fluidity (Margesin et al., 2007;Buzzini and Margesin, 2013;Alcaíno et al., 2015;Baeza et al., 2017). An increase in antioxidant enzymes activities was observed in Antarctic fungi after incubation at low temperatures, suggesting that the antioxidant defense could play a significant role in microbial survival when subjected to cold conditions (Gocheva et al., 2006(Gocheva et al., , 2009Kostadinova et al., 2012). When the Antarctic yeast Mrakia blollopis was submitted to a cold-shock an accumulation of TCA-cycle metabolites, lactic acid, and polyamines was observed (Tsuji, 2016). Polyamines are recognized to regulate important functions in cell growth, development, and cellular protection against stresses (Valdés-Santiago and Ruiz-Herrera, 2013). In the genome of the coldadapted yeast Mrakia psychrophila there is a high numbers of genes encoding for Major Facilitator Superfamily (MFS) proteins (133), which play multiple roles including the incorporation of nutrients from the environment (Perlin et al., 2014;Su et al., 2016), and suggested as an adaptation to survive in its environment. Considering the protein 3D structural level, the highly flexible and larger active site exhibited by cold-active enzymes is considered responsible for their higher specific activity in comparison to their mesophilic counterparts (Collins et al., 2008;Jung et al., 2008;Feller, 2013). The reduced proline content found in bacterial cold-active proteins has been proposed as an adaptation strategy to attenuate the negative effect of proline isomerization on protein folding (Wedemeyer et al., 2002;Feller and Gerday, 2003;Feller, 2010). OMICs technologies have become practical approaches for studying microorganisms, opening a window to find novel and holistic strategies of cold adaptation expressed by microbial cells (Casanueva et al., 2010). A higher number of rRNA and tRNA genes have been found in psychrophilic than mesophilic bacterial genomes suggesting an adaptation strategy to compensate the reduced translation rate under low-temperature conditions (Médigue et al., 2005;Methé et al., 2005;Riley et al., 2008;Choi et al., 2020). However, in the few studies about genomes of psychrophilic yeasts the number of rRNA and tRNA genes was variable and not consistently higher than mesophilic yeasts (Firdaus-Raih et al., 2018;Baeza et al., 2021;Yusof et al., 2021). Genes associated with cold tolerance were found in genomes of the yeast Naganishia vishniacii isolated from McMurdo Dry Valleys, Antarctica; putative genes encoding for solute transfers, chaperones, associated with photoprotection, desaturases, trehalose synthesis, and lipid metabolism were found (Nizovoy et al., 2021). The yeast genus Glaciozyma includes psychrophilic species isolated from Antarctica and European glaciers, comprising some species formerly classified into different genera (e.g., Leucosporidium antarcticum, which has been re-classified as Glaciozyma antarctica; Turchetti et al., 2011). Genes encoding for proteins stimulating cold survival, e.g., antifreeze proteins (AFPs), were identified in the genome of G. antarctica PI12 (Wong et al., 2019). The comparisons with other genomes of two mesophilic yeasts, two thermophilic, and one psychrophilic fungi revealed that a third of the genes identified in G. antarctica were absent in the more closely related yeasts, suggesting that these genes, which are unique for G. antarctica, could be associated to its psychrophilic habitus (Firdaus-Raih et al., 2018). Furthermore, a 4% of the annotated open reading frames (ORFs) found in the G. antarctica genome corresponded to small ORFs (smORFs) not exceeding 100 amino acids, which may have a role in cold adaptation (Mat-Sharani et al., 2015). Thousands to millions of smORFs have been found in other organisms, including humans, insects (i.e., Drosophila melanogaster), metazoans, and yeasts (i.e., Saccharomyces cerevisiae), in which they fulfill key physiological functions (Couso and Patraquim, 2017). The analysis of genomes and transcriptomes of a few Antarctic yeasts revealed the presence of a considerable number of putative genes involved in the synthesis of enzymes regulating the response to stress; most of them were classified as oxidative, general, and cold-associated responses, including desaturases, chaperones and chaperonins, glutathione S-transferase, catalase, and enzymes for the production and regulation of compatible osmolytes. Furthermore, putative genes involved in the synthesis of secondary metabolites, i.e., non-ribosomal peptide synthase (NRPS), terpenes, and type III polyketide synthases (PKSs) were found (Baeza et al., 2021(Baeza et al., , 2022. A global comparative analysis of in silico generated proteomes among Antarctic yeasts in comparison with their different optimal temperature for growth (OTG) and growth rates (Gr), revealed that a general trend of high content of flexible proteins associated with low OTG was found in yeasts characterized by fast growth. When subjected to cold stress, yeasts characterized by similar growth parameters exhibited similar transcriptomic level (Baeza et al., 2022). Interestingly, a significant upregulation of genes encoding for heat-shock proteins and chaperones, fatty acid desaturases, and proteasomal subunits, associated with the downregulation of ribosomal subunits were observed in yeasts exhibiting the higher OTG (e.g., Candida sake and Wickerhamomyces anomalus), but not in those with lower OTG (e.g., Mrakia gelida, and Cryptococcus sp.). In contrast, some yeasts characterized by lower OTG showed an upregulation of ribosomal subunits, suggesting the presence of differences strategies of adaptation to cold depending on their specific OTG (Baeza et al., 2022). In Rhodosporidium kratochvilovae strain YM25235 subjected to low-temperature stress, 1,300 differentially expressed genes (DEGs) were detected by next-generation deep Frontiers in Microbiology 03 frontiersin.org sequencing technology (RNA-seq). From them, the pathways included the MAPK signaling pathway, metabolic pathways, and amino sugar and nucleotide sugar metabolism were overrepresented (Guo et al., 2021). The MAP-Kinase HOG1 would play a role in cold adaptation in R. kratochvilovae by promoting biosynthesis of polyunsaturated fatty acids and glycerol (Chen et al., 2022). In recent years, the genomes of cold adapted yeasts from Antarctica were sequenced and characterized; their diverse responses to cold adaptations were studied, analyzing genomes and transcriptomes (Wong et al., 2019;Baeza et al., 2021Baeza et al., , 2022Nizovoy et al., 2021;Touchette et al., 2022). The extension of these studies to cold-adapted yeasts isolated from other geographical areas, which are characterized by chemical-physical and ecological conditions (including a different annual and seasonal temperature range) very different from those of the Antarctic habitats, is essential to advance toward a global understanding of yeasts' adaptative strategies to live in cold habitat worldwide. In this work, the genomes of four strains of different cold-adapted yeast species isolated from the Italian Alps and Apennines were sequenced, analyzed, and compared in relation to their different optimal growth temperatures. In this framework, a comparative analysis among yeasts was performed considering their growth properties, genome size, protein flexibility, and putative genes encoding for cold stress response and secondary metabolites. Materials and methods Yeast strains, culture conditions, determination of the optimum temperature and growth rates The yeast strains used in the present study were Goffeauzyma gilvescens DBVPG 4707, Glaciozyma martinii DBVPG 4841, Naganishia antarctica DBVPG 5271, and Phenoliferia glacialis DBVPG 5880, all preserved in the Industrial Yeast Collection DBVPG of the University of Perugia, Italy. They were all isolated from soil or sediments of cold habitats associated to high mountain glaciers of Italian Alps and Apennines. More in-depth information on strains used in this study are available on the website of DBVPG Collection (www.dbvpg.unipg.it). All the strains were routinely maintained in physiological inactive state (−80°C). Working cultures were grown on YEPG (yeast extract 10 g L −1 , peptone 10 g L −1 , glucose 20 g L −1 , agar 15 g L −1 ) agar slants at 25°C (G. gilvescens) or 15°C (G. martinii, N. antarctica, and P. glacialis). Calibrated (A 600 = 0.8, which was considered an average cell concentration = 10 7 cells/mL) suspensions of 48 h-old cells of the 4 strains were used to inoculate 200 μl of YEPG into 96 wells microplate to obtain A 600 = 0.1. Microplates were incubated at 4°C, 10°C, 15°C, 20°C, 25°C, and 30°C and A 600 was checked every day for 10 days. Growth curves were determined in triplicate; the growth rates (Gr) of each strain were calculated from the exponential phase of the growth curves while, the temperature in which the highest A 600 during exponential phase was observed was considered the optimum temperature for growth (OTG). Genomic DNA purification DNA Was extracted using standard methods based On The treatment of yeasts cells with CTAB (hexadecyltrimethylammonium bromide) and chloroform/isoamyl alcohol buffers. Briefly, 72 h yeast cells were suspended In pre-warmed CTAB extraction buffer containing CTAB, NaCl, EDTA, Tris HCl, PVP (Polyvinylpyrrolidone), β-mercaptoethanol, and proteinase K (sigma, Aldrich) and incubated at 60°C for 1 h. then chloroform/ isoamylalcohol (24:1) Solution Was added and The Mix Was centrifuged To recover The aqueous phase. The DNA Was then purified using pectinase and RNAase (sigma, Aldrich) incubating The Mix at 37°C, and phenol:Chloroform:Isoamylalcohol (25:24:1; Sigma, Aldrich). After centrifugation, DNA contained In The aqueous phase Was additionally washed with sodium acetate (3 M), isopropanol and ethanol 70% (sigma, Aldrich). After an additional centrifugation, DNA Was suspended In ultrapure water and maintained at −20°C until quantification and sequencing The DNA Was quantified By qubit Fluorometric assay (ThermoFisher scientific) To obtain a concentration higher than 20 ng/μl. Next-generation sequencing NGS by the Omics2view company (Germany, http://www. omics2view.consulting) on MGI DNBSEQ-G400 in 2 × 150 bp mode, including sequence processing as follow. Quality of demultiplexed reads was checked with FastQC v0.11.7 (Andrews, 2018). A summary QC report created with MultiQC v1.10 (Ewels et al., 2016) is available for download. Read quality trimming was performed with the BBTools package v38.45 (Bushnell, 2019). This comprised the removal of optical duplicates, human sequences, adapter sequences, low entropy reads, and trimming of bases with quality scores <20. Reads with invalid or ambiguous bases and reads with a length < 50 base pairs (bp) were discarded. Only reads surviving quality trimming as pairs entered downstream analysis. Read quality recalibration and error correction was performed with the BBTools package v38.45 (Bushnell, 2019). Qualitytrimmed reads were aligned to a preliminary de-novo assembly made with Tadpole from a subset of the quality-trimmed reads. Alignment information was used to recalibrate the base quality of all quality-trimmed reads. Sequencing errors were corrected by consecutively applying BBTools programs BBMerge, Clumpify, and Tadpole in error correction mode on the quality-recalibrated reads. Results are referred to as "filtered reads. " 31-bp kmers of filtered reads were normalized with BBNorm from the BBTools package v38.45 (Bushnell, 2019) to a target kmer depth of 100×, with a minimum kmer depth of 3×. Contiguous sequences (contigs) were assembled from normalized reads and further Prediction of secondary metabolite biosynthetic clusters The draft genomes were submitted to antiSMASH 6 fungal version to predict biosynthesis of secondary metabolites clusters in strict mode using default parameters (Blin et al., 2019). Each predicted core biosynthetic gene (CBG) and additional biosynthetic gene (ABG) of different gene clusters was compared to NCBI nr database by Blastp using default parameters, and the complete sequence of first 10 hits were retrieved. The sequences were analyzed using the maximum likelihood method and JTT matrix-based model (Jones et al., 1992) using the MEGA11 software (Kumar et al., 2018;Stecher et al., 2020). Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using a JTT model and then selecting the topology with a superior log-likelihood value. Prediction, annotation, and comparative analysis of coding sequence Bioinformatic analysis was performed using Geneious Prime 2020.1.2 software (Kearse et al., 2012). The CDSs with lengths ≥210 nt were predicted in draft genomes using the generic training of the Augustus gene prediction program (Stanke and Morgenstern, 2005;Stanke and Tzvetkova, 2006), trough the Geneious Augustus plugin. The CDS were in silico extracted, translated, and compared by Blastp against a local curated fungal protein database (updated in June 2021). The hits having at least 30% similarity and E values ≤10 −10 were considered for annotation. The annotated CDSs were classified according to cellular function predicted using KAAS -KEGG Automatic Annotation Server (Moriya et al., 2007) 1 trough GHOSTX, and using default parameters and the datasets for Genes and Eukaryotes. The translated and annotated CDS were used for comparative analysis of their flexibilities calculated based on their sequences of amino acids. For that, the percentage of amino acids classified according to their flexibility index (Li et al., 2006;Rao et al., 2011) either as very flexible (Vf: E, G, K, N, Q, and S) or moderately flexible 1 https://www.genome.jp/kaas-bin/kaas_main?mode (Mf: A, D, H, I, P, R, T, V) were calculated for each CDS. Furthermore, the global flexibility of each CDS was predicted using MEDUSA, a Deep-Learning based protein flexibility tool that uses as input the information from homologous protein sequences and amino acid physicochemical properties, using default parameters (Vander Meersche et al., 2021). From that, the percentage of residues grouped in three predicted flexibility classes was considered, M0, M1, and M2, from rigid to flexible. For each CDS the percentages of amino acids grouped as Vf, Vf plus Mf (VMf), M2, and M1 + 2, were calculated. The comparisons among yeasts were performed globally, considering the results for all CDS, and using the CDSs classified by cellular function by applying one-way ANOVA followed by a post hoc Tukey analysis. The data distribution was visualized using histograms and quantile-quantile (QQ) plots and did not find any indication that the data did not have a normal distribution. The CDS that displayed significant differences (significance threshold of 0.05) in their calculated flexibilities in the comparisons among yeasts at the global level or grouped by cellular function were selected to evaluate a probable correlation to yeast's growth parameters. For that, the differences among yeasts in calculated flexibilities of CDS were plotted vs. their differences in the optimal temperature for growth or growth rate, and linear regressions were applied. Prediction of small open reading frames The ORFs between 30 and 300 nt were predicted in the draft genomes using the ORF finder included in Geneious software, annotated as SmORFs, and classified as those present in the CDS (CDSSmORFs), into 1,000 nt upstream (5'SmORFs) and into 1,000 nt downstream (3'SmORFs) from the CDS. SmORFs were translated and annotated by Blastp comparison to a local SmOrfs database constructed with the data downloaded from the web pages smORFunction 2 (Ji et al., 2020) and SmProt 3 (Hao et al., 2018;Li et al., 2021). Multigene phylogeny This analysis was conducted to compare the P. glacialis isolates from the Italian Alps and Antarctica and other members belonging to order Kriegeriales for which the genomes are available in the NCBI database were included. The corresponding genomes were downloaded (Supplementary Table S1), and the CDS were predicted and annotated, as mentioned above. The CDS encoding for putative proteins common to at least eight yeasts, including both P. glacialis isolates, were selected for more analysis. Furthermore, the analysis was performed for SmORFs. The translated SmORFs with at least 30% similarity in Blastp comparison Frontiers in Microbiology 05 frontiersin.org against the local SmORF database mentioned above and common to three and all the studied yeasts were considered. The reference SmORF sequences were also included. The Evolutionary distances were computed for the individual proteins and concatenated CDS using the Jukes-Cantor method; the phylogenetic trees were obtained by neighbor-joining with 1,000 bootstrap replications. The analyses were conducted in MEGA11 software (Kumar et al., 2018;Stecher et al., 2020). Yeast growth parameters The complete growth curve was determined for each strain at temperature ranging from 4°C to 30°C, to obtain both OTG and Gr (Figure 1). G. martinii showed the lowest OTG (at 10°C), and the most limited temperature growth range (from 4°C to 15°C). N. antarctica and P. glacialis exhibited similar OTG at 15°C while G. gilvescens showed the highest OTG (20°C). All the strains exhibited growth at 4°C while only G. gilvescens and P. glacialis grew at 25°C. No growth was observed at 30°C (Figure 1). The maximum Gr of G. martinii and N. antarctica corresponds to a value more than twice lower than the Gr of the other two species and were exhibited at 4°C-10°C with no significant differences for G. martinii (0.0726 ± 0.0022 and 0.0710 ± 0.0014, respectively) and at 4°C for N. antarctica (0.0736 ± 0.0023; Figure 1). Draft genomes, prediction, and annotation of CDS, and secondary metabolite biosynthetic clusters The contigs assembled in draft genomes ranged from 401 in N. antarctica to 1,908 in G. gilvescens, the size was from 20.7 Mpb in G. gilvescens to 33.8 Mpb in P. glacialis, and the GC content was from 50.9% in G. gilvescens to 61.2% in P. glacialis (Supplementary Table S2). The predicted CDS varied from 2,221 in G. gilvescens to 10,303 in P. glacialis, while the annotated ones ranged from 11% in P. glacialis to 20% in G. gilvescens. Secondary metabolite biosynthetic gene clusters were predicted in the draft genomes: terpene clusters in the four yeasts, type III polyketide synthase clusters (T3PKS) in G. gilvescens and N. antarctica, and non-ribosomal peptide synthetase clusters (NRSP) in G. gilvescens, G. martinii and P. glacialis (Table 1). The identification of protein encoded by core biosynthetic gene (CBG), additional Growth parameters. Growth curves were determined at temperatures from 4°C to 30°C. The growth rate (discontinuous lanes) was calculated from the exponential phase of the growth curves. Continuous lanes show to the maximum A 600 observed during exponential phase at different temperatures and express the optimal temperature for growth for each strain. Goffeauzyma gilvescens (circle), Glaciozyma martinii (down triangle), Naganishia antarctica (square), and Phenoliferia glacialis (up triangle). Frontiers in Microbiology 06 frontiersin.org biosynthetic gene (ABG), and transport related gene (TRG) in the clusters was inferred by a phylogenetic analysis that included the sequences of the best five Blastp hits for each sequence obtained from the NCBI database (Supplementary Figure S1). In terpene clusters, the CBG was closest to squalene synthase in G. martinii (seq 22,166) and P. glacialis (seq 24,553), to bifunctional farnesyl-diphosphate farnesyl transferase/squalene synthase in G. gilvescens (seq 14,972) and N. antarctica (seq 22,146), and to geranylgeranyl-diphosphate geranylgeranyl transferase, terpenoid cyclase, squalene/phytoene synthase and phytoene desaturase in G. gilvescens (seq 7,405), G. martinii (seq 17,625), N. antarctica (seq 20,744) and P. glacialis (seq 22,851), respectively. The ABG was close to amino acid transporter in N. antarctica and NAD-binding protein and NAD-dependent mannitol dehydrogenase in P. glacialis (seq 22,851 and 24,553, respectively). In T3PKS clusters the CBG found in G. gilvescens and N. antarctica apparently corresponded to chalcone synthase, and in N. antarctica the CBG and TRG were closest to anthranilate synthase and sugar transporter, respectively. The CBG of NRPS clusters predicted would correspond to nonribosomal peptide synthetase in G. martinii and P. glacialis, and to L-aminoadipatesemialdehyde dehydrogenase in G. gilvescens. The ABG of NRPS predicted in P. glacialis was close to an oxidoreductase (Supplementary Figure S1). Modules and metabolic pathways The annotated CDSs were classified according to general and specific completed modules and metabolic pathways. Forty-four modules were found, with the highest number in P. glacialis followed in decreasing order by N. antarctica, G. martinii, and G. gilvescens (Supplementary Table S3), 28 were found in the four studied yeasts and 9 in tree yeasts (i.e., Polyamine biosynthesis, arginine, assimilatory sulfate reduction and beta-oxidation acyl-CoA synthesis). The modules for which more putative genes were found were citrate cycle, beta-oxidation, leucine degradation, citrate cycle second carbon oxidation, and glycolysis. Seven modules associated to metabolism of cofactors and vitamins (i.e., lipoic acid biosynthesis, biotin biosynthesis and ubiquinone biosynthesis) were found only in two yeasts. Regarding metabolic pathways, more putative genes were found for P. glacialis and G. martinii, and the top5 at the middle level were signal transduction, carbohydrate metabolism, amino acid metabolism, transport and catabolism, and cell growth and death (Figure 2). At the specific level the top5 pathways were cell cycle, MAPK signaling pathway, spliceosome, ribosome, and autophagy (Supplementary Table S4). Of 201 specific pathways for which at least one putative gene was found, about the half were common to four yeasts, whiles 32 were found only in one yeast, most in G. martinii such as nitrogen metabolism, ether lipid metabolism, non-homologous end-joining, and atrazine degradation. In all studied yeasts, two or more CDS encoding for the same putative protein was found, especially in G. martinii and P. glacialis. The 20 putative proteins encoded by a higher number of CDS found in all yeasts are shown in Figure 3. Among them, P-loop containing nucleoside triphosphate hydrolase protein, alpha/beta hydrolase protein, armadillo-type protein, major facilitator superfamily domain-containing protein, and general substrate transporter protein can be mentioned. Stress related genes A total of 423 orthologous genes associated to response to different kind of stresses were found, most of them related to oxidative and cold stress, followed by general, osmotic, and heat stress responses, especially in P. glacialis and G. martinii (Supplementary Table S5). The top5 predicted proteins related to stress were mitogen-activated protein kinase 1, catalase, dihydrolipoamide dehydrogenase, glyceraldehyde 3-phosphate dehydrogenase, glycogen synthase kinase 3 beta, ribulosephosphate 3-epimerase. Small open reading frames Hundreds of thousands of SmORFs were found in all yeasts, in decreasing order in P. glacialis (772,037), G. martinii (462,525), N. antarctica (295,488), and G. gilvescens (199,985), and higher numbers of 3'SmOrfs were predicted ( Figure 4A). No identical SmORFs nucleotide sequences were found among the studied yeasts, and very few considering translated sequences, being higher between P. glacialis and G. martinii in CDSSmOrfs and 3'SmOrfs (from 59 to 64 identities; Figure 4B). Compared to the SmORFs database, only 0.1%-0.5% of SmORFs yielded hits (considering at least 50% of coverage and similarity), most of them corresponding to the classifications Homo sapiens and mouse ( Figure 4C). A phylogenetic analysis was performed considering the translated SmORFs with at least 30% similarity by Blastp to the SmORF database and common to three and all yeasts studied. The majority of common putative SmORFs corresponded to Literature mining classification. The yeasts were grouped close in either analysis of Figure S2B), not observing a preference for hierarchical clustering among yeasts in any of the SmORF used for the analysis. Comparison of protein flexibilities based on amino acid composition The analysis was performed considering the predicted flexibilities of translated CDSs annotated and classified in cellular pathways. The M1 + 2 percentage showed the highest degree of significant (p < 0.05) variations among almost all yeasts under study, being higher between G. martinii vs. G. gilvescens and N. antarctica, and P. glacialis vs. G. gilvescens and N. antarctica. The pathways exhibiting more significant differences were carbohydrate metabolism, cell growth and death, and transduction signal. The clustering of species according to their differences revealed that P. glacialis was close to G. martinii, while N. antarctica was close to G. gilvescens ( Figure 5A). When the 4 yeast species were grouped according to their OTG (obtaining 3 groups: OTG = 10°C, OTG = 15°C, and OTG = 20°C), those with OTG equal to 10°C and 15°C clustered separately from yeast exhibiting an OTG of 20°C ( Figure 5B). On the other hand, when Gr was considered (Gr = 0.07, Gr = 0.16, Gr = 0.18), the yeast species showing either highest or lowest Gr clustered together, differently from those with intermediate Gr, which clustered separately ( Figure 5C). The significant differences found among studied yeasts in CDS groups and parameters were analyzed concerning their differences in growth parameters ( Figure 5D). Overall, positive, and negative correlations were obtained between Vf and M1 + 2, respectively, and Gr, while a positive correlation was found between M1 + 2 and OTG. In the analysis by pathways, both positive and correlations between M1 + 2 and Gr in 2 and 3 pathways, respectively, and positive correlation between M1 + 2 and OTG in 6 pathways were observed. These correlations were generally found in curves with only two points, except for M1 + 2 vs. OTG in for global and general metabolism pathway (R 2 ≥ 0.98). Multigene phylogeny of Alpine and Antarctic Phenoliferia glacialis isolates To highlight the differences between strains belonging to the same species but isolated from different geographical areas, the strain DBVPG 5880 of P. glacialis isolated from Italian Alps, was compared to an isolate from Antarctica, and to other members of the order Kriegeriales with available genomes at the NCBI database. The identified CDS encoding for proteins common to at least 8 yeasts (including both P. glacialis isolates) were considered for analysis at individual proteins and concatenated CDSs. In the trees generated from individual analysis of 65 proteins, both isolates of P. glacialis were generally closed related (Supplementary Figure S3). In some cases, alpine P. glacialis was closest to G. antarctica PI12, such as for proteins 2-isopropyl malate synthase, 3-hydroxybutyrate dehydrogenase, and acetyl-CoA carboxylase. Figure 6 summarized the results of trees classified for proteins common to 8-12 yeasts. The two strains of P. glacialis clustered in closed related positions in just 2 of the 6 trees obtained (Figures 6D,E); in the trees with higher number of analyzed proteins ( Figures 6A-C,F) Antarctic P. glacialis was always positioned external to the entire group of strains, differently from Alpine P. glacialis. When the analysis was performed with concatenated CDS, both P. glacialis isolates clustered together and with Rhodotorula species (especially with R. diobovata, R. taiwanensis, and R. mucilaginosa) which vary in dependence of the number of CDS analyzed ( Figures 7A-C), and with G. antarctica (Figures 7D,E). When the analysis was restricted to all CDS common exclusively to P. glacialis isolates and G. antarctica, these three yeasts clustered with R. taiwanensis MD1149, and R. mucilaginosa F6, with a close relation between alpine P. glacialis and G. antarctica PI12 ( Figure 7F). Discussion The four yeasts under study showed variable growth parameters (OTGs ranging from 10°C to 20°C). Interestingly, N. antarctica showed its max Gr at 4°C, but its max OTG (the highest A 600 during the exponential phase) was found at 15°C. The long lag phase may explain the data observed at 4°C, in comparison with those obtained at 10°C and 15°C. The maximum A 600 was reached by G. gilvescens and P. glacialis, which also showed the highest Gr and the ability to grow at 25°C, although the growth of P. glacialis was greatly reduced. All the strains were isolated from cold habitats of Italian Alps and Apennines and grow at 4°C, but only G. martinii and N. antarctica exhibited no growth at temperature equal or above 20°C. The Alpine P. glacialis strain herein studied differed from a previously studied Antarctic isolate of the same species exhibiting higher OTG and temperature range for growth (Baeza et al., 2021). This temperature range is narrower than that exhibited by isolates of the yeast-like fungus Aureobasidium subglaciale, which grew from 0°C to 30°C, and even tolerated incubation at 50°C for 2 h (Zajc et al., 2022). The sizes of draft genomes and GC contents ranged from 20.7 to 33.8 Mpb and 50.9% to 61.2%, respectively, in agreement with other yeasts isolated from cold habitats (Brejová et al., 2017;Coleine et al., 2019Coleine et al., , 2020Cunha et al., 2020;Baeza et al., 2021). Comparing the calculated protein flexibilities among yeasts, significant differences were found mainly for medium plus very flexible (M1 + 2, predicted by MEDUSA), which was higher for A B C FIGURE 4 SmORFs prediction and comparison among yeasts. The total (white columns) SmORFs and those annotated (gray columns) by comparison SmOrfs database, are shown in (A) the number of translated SmORFs identical among yeasts is shown in (B) and those annotated by comparison to a SmORFs database in (C). CDSSmORFs, SmORF located into a CDS; 5'SmORFs, SmORF located until 1,000 nt upstream of a CDS; 3'SmORFs, SmORF located until 1,000 nt downstream of a CDS (3'SmORFs). Frontiers in Microbiology 10 frontiersin.org G. martinii than G. gilvescens and N. antarctica, and for P. glacialis than G. gilvescens and N. antarctica. Considering the growth parameters, the yeasts showing most significant differences in OTG were also discriminated according to their predicted protein flexibilities; similar clustering did not occur for Gr. Overall, a correlation between protein flexibilities calculated as M1 + 2 and yeast growth parameters was directly proportional when considering OTG and variable for Gr. Similarly, in previous genomic and transcriptomic studies of eight Antarctic yeasts, a correlation was observed between calculated protein flexibilities and OTGs, not Gr (Baeza et al., 2021(Baeza et al., , 2022. Interestingly, the correlation between OTG and calculated protein flexibilities was negative in Antarctic yeasts, while it was positive in the yeasts isolated from the Alps and Apennines. It could be an interesting point concerning potential differences in adaptation strategies in yeasts colonizing different cold environments, thus suggesting the need of a more in-depth study of cold-adapted yeasts at the genomic and biochemical levels. A considerable number of putative genes involved in the regulation of response to different kinds of stress were found, including, in addition to cold, general, oxidative, and osmotic stress, thus highlighting the putative adaptation of yeasts to the Comparative analysis of yeast species according to predicted protein flexibilities and growth parameters. The predicted flexibility of CDSs grouped by cellular functions were compared among species individually (A) and grouped according to their optimal temperature for growth (B) and growth rate (C). The heatmap shows the number of cellular pathways for which significant difference (Tukey post hoc tests, p < 0.05) were found between yeasts. In (D) the plot of the difference between yeasts for a pathway in flexibility parameters (P1-2) vs. difference in growth parameters (F1-2) is shown. OTG, optimal temperature for growth; Gr, growth rate; Vf, very flexible; M2, class2 flexibility predicted by MEDUSA; M1 + 2, class 1 + class 2 flexibility predicted by MEDUSA. Frontiers in Microbiology 11 frontiersin.org challenging environmental conditions occurring in Alpine ecosystems. Some genes associated with response to cold were found, among them: dihydrolipoamide dehydrogenase, enoyl-[acyl-carrier protein] reductase II, glycogen synthase, omega-6 fatty acid, desaturase/acyl-lipid omega-6 desaturase, stearoyl-CoA desaturase (Delta-9 desaturase), ATP-dependent RNA helicase DeaD, elongation of very-long-chain fatty acids protein 6, glycogen debranching enzyme, and very-long-chain acyl-CoA dehydrogenase, involved in different cellular functions such as cell membrane adaptation, compatible solute accumulation, regulation of protein folding, and changes in RNA structure. Similar genes patterns were also found in other microorganism (including yeasts), as a common reply to cold stress (Aguilera et al., 2007;Inouye and Phadtare, 2014;Zhang and Gross, 2021). Among the more represented pathways MAPK signaling, spliceosome, ribosome, and autophagy were observed, especially in P. glacialis and G. martinii. In a genomic and comparative transcriptomic study of a cold-adapted oleaginous yeast R. kratochvilovae, cultivated at 15°C and 30°C, the significantly downregulated genes were primarily related to metabolic and cellular processes and cellular organelles, whereas the MAPK signaling pathway was among overrepresented pathways (Guo et al., 2021). A recent study of Chen et al. (2022) demonstrated that the overexpression and knockout of the MAP-Kinase HOG1 played an important role Comparison among yeasts based in phylogenetic analysis of individual proteins. Comparisons were made based on the patristic distances extracted from threes computed for proteins (Supplementary Figure S3) that were common to at least 8 yeasts, including both P. glacialis isolates. Frontiers in Microbiology 12 frontiersin.org in cold adaptation of R. kratochvilovae, mainly by promoting the biosynthesis of polyunsaturated fatty acids, such as linoleic acid and linolenic acid, and glycerol (Chen et al., 2022). Autophagy denotes a variety of mechanisms for the metabolization via vacuole of aged and damaged proteins, lipids, and membranebound organelles and has also been associated play a critical role in aging (Tyler and Johnson, 2018). The role of autophagy in response to abiotic stress, including cold, has been studied mainly in plants and fishes Ren et al., 2021). A considerable number of genes classified in the autophagy pathway were found in Antarctic yeasts, especially in Candida sake (Baeza et al., 2021). Among complete modules according to putative genes in the four studied yeasts, the pathways for polyamine biosynthesis and citrate cycle (TCA cycle Krebs cycle) were found. TCA-cycle metabolites, lactic acid, and polyamines were accumulated in the Antarctic yeast Mrakia blollopis after a cold shock exposure (Tsuji, 2016). Polyamines are aliphatic hydrocarbon chains containing one or more amine groups. They have recognized functions in cell growth, development, and cellular protection against stresses such as oxidative, osmotic and temperature (Valdés-Santiago and Ruiz-Herrera, 2013;Gevrekci, 2017;Murray Stewart et al., 2018). Putative antifreeze glycopeptide AFGP polyprotein and antifreeze protein were identified in N. antarctica and P. glacialis. Similarly, a putative gene encoding for ice-binding protein was found in the whole genome of the endemic Antarctic ascomycetous fungus Antarctomyces pellizariae (Batista et al., 2020). The antifreeze or Ice Binding Proteins (IBPs) can bind to the ice surface and be incorporated into the ice crystal matrix controlling the ice shape and inhibiting re-crystallization, potentially decreasing cell injury (DeVries, 1986;Bouvet and Ben, 2003;Garner and Harding, 2010). IBPs also stimulate microbial biofilm formation on ice and the formation of channels through the ice, helping interchange of nutrients and gasses (Davies et al., 2002;Bar Dolev et al., 2016;Cheung et al., 2017). Most studies revealed the production of IBPs by fish, bacteria, and plants, and in a more limited way by fungi. The IBP AFP1, produced by G. antarctica, was successfully expressed in E. coli (Hashim et al., 2013). Antifreeze proteins from the Antarctic yeast Goffeauzyma gastrica, maintained, after partial purification, a good antifreeze property after several freeze-thaw cycles (Villarreal et al., 2018). The redundancy in putative CDS encoding for the same protein or proteins families was one additional aspect that can Maximum likelihood phylograms based on the concatenated CDSs. The coding sequences for proteins common to at least 8 yeasts, including both P. glacialis isolates, were concatenated and the evolutionary distances computed by Jukes-Cantor method. The phylogenetic trees obtained by neighbor-joining with 1,000 bootstrap replications and statistically supported bootstrap values (≥50%) are shown. (A), 50 proteins common to 8 yeasts; (B), 40 proteins common to 9 yeasts, (C), 34 proteins common to 10 yeasts; (D), 22 proteins common 11 yeasts; (E), 17 proteins common to 12 yeasts. In (F), was considered the proteins common at least for P. glacialis and G. antarctica (31 proteins, common to 7 yeasts). Abbreviation of yeasts as in Figure 6. Frontiers in Microbiology 13 frontiersin.org be putatively considered as an adaptation strategy of yeasts to their extreme environments. Gene duplication has been proposed as a mechanism of genomic adaptation (as a genetic robustness) to a changing environmental conditions, including transport of nutrients, and protection to biotic and abiotic stress (Levasseur and Pontarotti, 2011;Kondrashov, 2012). P-loop containing nucleoside triphosphate hydrolase (P-loop NTPase), Alpha/Beta hydrolase (α/β hydrolase), Armadillo repeat-containing proteins (ARMCs), and the major facilitator superfamily protein (MFS) were found in the studied genomes among the most redundant ones. P-loop NTPase is a superfamily of enzymes crucial for almost all aspects of life, and some members contribute to stress responses in plants (Leipe et al., 2002;Cheung et al., 2013Cheung et al., , 2016. α/β Hydrolase is a superfamily of hydrolytic enzymes that share a common fold having widely catalytic functions and phylogenetic origin (Ollis et al., 1992). Some α/β hydrolases have been involved in salt and cold tolerance in plants and microorganisms, including yeasts (Brody et al., 2001;Takahashi et al., 2006;Liu et al., 2014;Wu et al., 2020). ARMCs form a large family of enzymes primarily characterized in multicellular organisms (e.g., in plants) and exhibiting very versatile and fundamental functions, including response to stress, such as osmotic (Kojima et al., 2013;Sharma and Pandey, 2015;Huang et al., 2021). MFS is a ubiquitous transporter superfamily whose members have a broad spectrum of substrates, such as inorganic and amino acids, nucleosides, lipids, and short peptides (Yan, 2015); some members of MFS are involved in resistance to oxidative stress and fungicides in fungal pathogen Alternaria alternata (Chen et al., 2017). An increase of MFS copy numbers led to an increase of the gene redundancy in Antarctic yeast Naganishia vishniacii (Nizovoy et al., 2021) and Mrakia psychrophila isolated from permafrost on the Qinghai-Tibet Plateau (Su et al., 2016) and suggested that it could be considered a further adaptation strategy contributing to the accumulation of nutrients under cold imposed conditions. Gene clusters for the biosynthesis of secondary metabolites were predicted: terpene clusters were found in all four sequenced yeasts, NRSP in G. gilvescens, G. martinii, and P. glacialis, and T3PKS in G. gilvescens and N. antarctica. NRPS was mainly found in bacteria and fungi, especially Ascomycota, which assemble numerous peptides having functions in primary metabolism, cellular development and morphology, and stress responses such as oxidative in Aspergillus fumigatus (Stack et al., 2007;Süssmuth and Mainz, 2017;Sayari et al., 2019). Terpenoids are structurally the most diverse groups of natural compounds, widely distributed in nature, that play important physiological and metabolic functions, mainly antioxidant activities in microorganisms (Avalos and Carmen Limón, 2015;Jaeger and Cuny, 2016). Type III PKS has widely described in plants and bacteria and recently in Antarctic yeasts (Baeza et al., 2021), and produces a wide array of compounds such as chalcones, pyrones, acridones, phloroglucinols, and stilbenes, with several physiological roles such as pigmentation, salinity and dehydration resistance, stress adaptation, and cell wall remodeling (Yu et al., 2012;Parvez et al., 2018;Navarro-Muñoz and Collemare, 2019). A class of genetic elements with increasing relevancy are putative functional small ORFs (SmORFs) of 10-100 codons (which millions can be found in eukaryotic genomes) generally considered non-coding sequences because of their short length that impair bioinformatic analysis (Albuquerque et al., 2015;Couso and Patraquim, 2017). In G. antarctica SmORFs represented 4% of annotated ORFs, validated as expressed transcripts via cross-referencing against available transcriptome data (Mat-Sharani et al., 2015). High numbers of SmORFs were found in all four Alpine yeasts studied, especially in P. glacialis, of which 19% were at 5′UTR. The SmORFs showed to be "exclusive" for each yeast as no identical SmORFs at nucleotide sequence were found and resulted infimum when translated sequences were compared; only 0.1%-0.5% of them showed similarity to SmORFs of other organisms, mainly from H. sapiens and mouse. The function mainly described for SmORFs, especially those at 5′UTR, is related to the regulation of the expression of the main ORF by mechanisms that include the alteration of ribosome capacity to initiate, terminate and reinitiate translation (Wang and Rothnagel, 2004;Gunišová and Valášek, 2014;Renz et al., 2020). Although some examples of microproteins working independently have been found, such as repressors in S. cerevisiae, Neurospora crassa and H. sapiens (Renz et al., 2020), their appropriate functional study is still complicated. Anyway, it is possible to postulate a role for SmORFs as an additional adaptation strategy of yeasts to cold environments. The results obtained by the comparison of the genomes showed that Alpine P. glacialis herein studied exhibited some differences when compared to another strain of the same species, but isolated from Antarctica (Baeza et al., 2021). The draft genome of Alpine P. glacialis was 33.8 Mbp with 61.2% GC (87% completeness), whiles Antarctic P. glacialis was 15.9 Mb with 50.4% GC (completeness 71%). Two terpene and one NRPS biosynthetic cluster were predicted in Alpine P. glacialis and three NRPS in Antarctic P. glacialis. The ratio sum of CDS / total of different putative proteins was 2.1 for Alpine strain and 1.8 for the Antarctic one. When comparing the top 20 redundant putative proteins, just 9 of them exhibited a coincidence between both yeasts: alcohol dehydrogenase, α/β hydrolase, general substrate transporter, MFS, NADP-dependent oxidoreductase domaincontaining, P-loop NTPase, and pyridoxal phosphate-dependent transferase. In the comparisons by multigene phylogeny computed with protein sequences individually and concatenated CDS of both strains, including other yeasts of the order Kriegeriales, the two P. glacialis did not always cluster together and, in some trees, they showed a close relation with G. antarctica PI12. Conclusion The findings obtained from the genomic approach used in this study to analyze the cold adaptation of the four studied yeasts are consistent with those reported for other yeasts isolated from cold habitats regarding genomic structure. Frontiers in Microbiology 14 frontiersin.org Furthermore, the high numbers of putative genes encoding for cold stress response and gene clusters encoding for the synthesis of secondary metabolites suggest the ability of such yeasts to successfully survive and adapt their physiology to these harsh ecological conditions. The different correlation between the OTG and calculated protein flexibilities found in yeasts isolated from the Italian Alps and Apennines and those from Antarctica suggest the existence of a certain degree of variability in adaptation strategies of yeasts thriving in different cold regions. As expected, Alpine and Antarctic P. glacialis isolates were close in the multigene phylogenetic analysis, including other members of the order Kriegeriales. Interestingly, both strains were clustered close to G. antarctica PI12 suggesting more similitude among yeast from cold environments. On the other hand, as already described for Antarctic yeasts, a considerable number of SmORFs and genetic redundancy were found. These two aspects are less studied and could be proposed to play roles in adaptation to cold, which are attractive to include in future comparative studies of adaptative yeast' strategies from different cold environments worldwide. Data availability statement The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: NCBI BioProject, accession no: PRJNA859078. Author contributions BT and PB carried out DNA purifications and determination of yeast growth parameters. MB performed bioinformatic and statistical analyses. MB, BT, and PB contributed to design of the experiments, discussion of the results, and manuscript writing. All authors have read and agreed to the published version of the manuscript. Funding This research was funded by Agencia Nacional de Investigación y Desarrollo, grant number Fondecyt 1180233. Supplementary material The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmicb. 2022.1026102/full#supplementary-material SUPPLEMENTARY FIGURE S1 Prediction of function for genes in secondary metabolite clusters. The phylogenetic analysis included the best 10 Blastp hits for each sequence obtained from the NCBI database. The evolutionary history was inferred by using the Maximum Likelihood method and JTT matrix-based model. The percentage of trees in which the associated taxa clustered together is displayed next to the branches. SUPPLEMENTARY FIGURE S2 Phylogenetic analysis for translated SmORFs. The analyses were performed with SmORFs common to three (A), and all yeast (B) studied, including the reference Blastp hits SmORFs from Mouse and Literature mining classification. The evolutionary history was inferred by using the Maximum Likelihood method and JTT matrix-based model. The percentage of trees in which the associated taxa clustered together is displayed next to the branches. SUPPLEMENTARY FIGURE S3 Phylogenetic analysis for individual proteins used in multigene phylogeny analysis. The evolutionary history was inferred by using the Maximum Likelihood method and JTT matrix-based model. The percentage of trees in which the associated taxa clustered together is displayed next to the branches.	2022-11-08T14:14:45.717Z	2022-11-08T00:00:00.000	{ "year": 2022, "sha1": "084e6a2bfb27bb2896e614fc577fd4bfa6a1dd3c", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Frontier", "pdf_hash": "084e6a2bfb27bb2896e614fc577fd4bfa6a1dd3c", "s2fieldsofstudy": [ "Biology", "Environmental Science" ], "extfieldsofstudy": [ "Medicine" ] }
211543080	pes2o/s2orc	v3-fos-license	Neuroprotective and Anti-Inflammatory Properties from the Bioactive Constituents of Myracrodruon Urundeuva, a Brazilian Medicinal Species Myracrodruon urundeuva Fr. All. (Anacardiaceae) is largely used, by the Northeast Brazil population, for its antiinflammatory actions, already confirmed by experimental studies. In addition, this medicinal species presents neuroprotective properties, as demonstrated by in vitro and in vivo Parkinson’s disease models. The trunk bark of M. urundeuva is a rich source of bioactive polyphenol compounds, as tannins and chalcones considered precursors of flavonoid biosynthesis. Not only the standardized plant extract, but also its isolated compounds present analgesic and anti-inflammatory effects, observed in acute models of nociception and inflammation in rodents. Besides, due to their anti-inflammatory and antioxidant activities, chalcones drastically inhibit myeloperoxidase (MPO) activity and have been effective in an allergic conjunctivitis model. Possibly, all these effects of M. urundeuva are involved with the neuroprotection offered by the standardized extract, SEMU (rich in chalcones), and isolated compounds, in rat mesencephalic cells exposed to 6-OHDA and in the classical hemiparkinsonian model of Parkinson’s disease induced by 6-OHDA in rats. We showed that the standardized extract reverted behavioral (apomorphine-induced contralateral rotations) and neurochemical alterations (striatal monoamine concentrations, as well as DAT and TH immunostainings) observed in 6-OHDA-lesioned animals and protected glia cells. The extract significantly decreased striatal iNOS, COX-2, TNF-alpha and NF-kB immunostainings and, more importantly, HDAC immunostaining. All together, these data indicate the benefits of M. urundeuva for inflammation related pathologies and may stimulate translational studies focusing on its use for the prevention or treatment of neurodegenerative disorders as Parkinson’s disease. do-Sertão". The ethnopharmacological study has initiated by the observation of the use of its stem bark for the treatment of gynecological diseases, such as cervicitis, due to its antiinflammatory properties. These actions were supported by clinical studies that resulted in a specific pharmaceutical preparation, available, in Brazil, for more than 25 years [21]. The success of this treatment guided preclinical studies, aiming at the knowledge of the medicinal properties, as well as the possible mechanisms of action involved in the pharmacological action of the plant. This initial pharmacological screening resulted in the discovery of peripheral antiallergic, anti-inflammatory, analgesic, antiulcer and antioxidant properties attributed to that medicinal plant [22][23][24][25][26]. Recently, we also observed by the first time the antidiabetic and hipolypidemic effects of the stem bark decoction from a cultivated specimen of M. urundeuva [27]. Furthermore, guided by the knowledge of the pathophysiological processes involved in PD, we decided to investigate whether the effects observed on the periphery would be reproduced at the level of the central nervous system. We started the study by testing the neuroprotective activity of a fraction rich in chalcones, the major component of the plant extract. Chalcones are aromatic ketones and intermediates in flavonoid biosynthesis. They form a central core for a variety of important biological compounds and possess different activities, like antibacterial, antifungal, anti-inflammatory, antioxidant and antitumor, among others [28][29][30]. Chalcones are one of the largest classes of natural products, being studied with enthusiasm and already used in the medical clinic for the treatment of pain, gastric ulcer, liver Slowing down or even disrupting the processes that culminate in neuronal death, presuppose a reasonable understanding of these events. Furthermore, the lack of the necessary knowledge hampers the development of a rational neuroprotective therapy, which could be favored by the necessary comprehension of what triggers neurodegeneration. Nevertheless, the knowledge of some processes involved in neuronal death facilitates the planning of therapeutic strategies capable of delaying the disease evolution [7]. Among the targets for neuroprotective therapy, we can mention oxidative stress, mitochondrial dysfunction, excitotoxicity and neuroinflammation [8,9], the latter being much studied by investigators working on the neuroprotective actions of medicinal plants [10]. Phytochemicals are bioactive components of medicinal plants. They are usually secondary metabolites, known to prevent age associated declines in cognitive and motor functions, among others [11,12] that are, at least partly, due to their antioxidant and anti-inflammatory properties. Although some neuroprotective agents show promising effects in preclinical studies, they should be submitted to well designed clinical trials, in order to provide reliable translational data [13,14]. In this sense, several teams of investigators are working on potential neuroprotective treatments for PD, focusing on novel therapeutic targets, including drugs with antioxidant and anti-inflammatory actions [15]. The majority of neuroprotective agents are antioxidants and, many of them, bioactive components of medicinal plants [16]. Neuroprotective drugs should present the ability to reverse or prevent some of the damage leading to neuronal cell degeneration. Thus, neuroprotection refers to strategies and mechanisms that lead to a relative preservation of the neuronal integrity and to a reduction in the rate of neuronal loss over time. On the contrary, the neurotoxicity leading to neuronal injury may result in motor incoordination and cognitive dysfunction. Investigations of new neuroprotective drug development focus on mechanisms implicated in protein accumulation, oxidative stress, neuroinflammation and dopamine deficits in PD [17]. Acute neuroinflammatory processes, whether innate or adaptive, have a protective action in neuronal homeostasis; however, when becoming chronic, these processes trigger cascades that lead to neurodegeneration. Although it is not known whether neuroinflammation is a cause or a consequence of neuronal death, research, both in vitro and in vivo, has already shown the benefits of anti-inflammatory therapy in neural protection [18,19], including the potential of medicinal plants as neuroprotection agents [20]. This perspective led us to study Myracrodruon urundeuva Fr. All. (Figure 1), a medicinal plant widely used in Northeastern region of Brazil, due its anti-inflammatory and healing actions. Myracrodruon Urundeuva: Neuroprotective Potential on PD The medicinal species M. urundeuva belongs to the Anacardiaceae family and is popularly known as "Aroeira- mesencephalic cells, evidencing the antioxidant effect as one of the neuroprotection mechanisms [37,38]. Three dimeric chalcones, denominated A, B and C urundeuvines, were isolated from the CEF (Figure 3). The arrangement of the hydroxyl groups found in the dimethyl chalcones is also related to their anti-inflammatory action, whose inhibition of 5-lipoxygenase is dependent upon the C-ring hydroxylation [39]. Whereas oxidative stress and neuroinflammation form a vicious cycle that amplifies the neurodegeneration process and adds to the indications of antioxidant and anti-inflammatory effects of the plant, we decided to carry out tests for evaluating whether the neuroprotective potential observed in the mesencephalic cell culture could be reproduced in vivo. Thus, the chalcone-enriched fraction (CEF) of the stem bark from M. urundeuva was tested in vitro in rat mesencephalic cells exposed to 6-hydroxydopamine (6-OHDA). This neurotoxin causes dopaminergic degeneration by mechanisms related to mitochondrial dysfunction, specifically in the I and IV complexes, leading to oxidative stress and consequent neuronal death [36]. At concentrations ranging from 1 to 100 μg/ml, CEF was able to increase the cell viability index, as determined by the 3-(4,5dimethylthiazol-2yl)-2,5-diphenyl tetrazolium test (MTT), in a concentrationdependent manner. The cell death was also evaluated by the fluorescence technique, using acridine orange and ethidium bromide (AO/EB) staining assays, observing the apoptotic pattern reversion in cells treated with CEF ( Figure 2). These results were corroborated by immunostaining for DAT, the dopamine transporter, and TH, the tyrosine hydroxylase, this which is one the key enzymes in the catecholamine biosynthesis. CEF was able yet to decrease nitrite formation and to inhibit lipid peroxidation in been consistently associated with PD and also identified in brain regions most affected by this disease [45,46]. Microglia have an important role in homeostasis and neural protection. When overactivated, as a result of injuries and immunological stimuli [47,48], microglia become neurotoxic through mechanisms involving the excessive formation of free radicals and the increase in inflammatory mediators expression [49]. 6-OHDA is able to induce striatal microgliosis, from the second hour after the exposure to the cells, this effect persisting for weeks [50]. The extract, at the doses used, was able to inhibit microglia activation, relieving the process of neurodegeneration. Taking into account that the neurotoxicity caused by microglia activations is usually progressive, increasing and accelerating neuronal damage [51], agents capable of inhibiting or decreasing this activation are suitable for clinical trials, aimed at the discovery of new neuroprotective agents. At the end of treatment, the apomorphine-induced rotational behavior test was performed. Apomorphine is a dopaminergic agonist that, once administered to animals with unilateral striatal lesions, causes stereotyped rotational movements in the contralateral direction to the lesion. This test is widely used for evaluating the functional efficacy of the PD induction model, as well as to predict the possible neuroprotective action of drugs that can reduce the stereotypy of the animal [41]. The groups lesioned with 6-OHDA and treated with SEMU (20 or 40 mg/kg, p.o.) decreased significantly, in a dosedependent manner, by 79 and 87%, respectively, the number of contralateral rotations/h, relatively to the untreated 6-OHDA-lesioned group. The sham operated group did not present stereotyped rotations. The presumable neuroprotective effect of SEMU was confirmed by the determination of dopamine (DA) levels, through high-performance liquid chromatography (HPLC). When compared to the injured animals, those treated with SEMU presented significantly higher levels of DA in the lesioned striata (Table 1). The maintenance of DA levels is of primary importance for new pharmacological options in antiparkinsonian therapy, considering that the symptoms of the disease, as well as its evolution, are directly related to dopaminergic depletion in the nigrostriatal pathway. Through the histological technique of cresyl violet, the viability of the striatal neurons was possible to observe microscopically, with the finding of a large number of viable and preserved cells, approximately 2.1 times in the animals treated with SEMU 40 mg/kg, when compared to those that did not receive the treatment. Immunostaining for TH was also performed and the results corroborate with those previously found, being possible to observe an increase of almost 13 times in enzyme labeling, in the animals receiving SEMU at its highest dose. The extract standardization evidenced a large amount of phenolic compounds, presenting 13.3% of total polyphenols. These substances have been studied for their neuroprotective activity, after considering the anti-inflammatory and antioxidant actions of curcumin and green tea catechins on the central nervous system [42]. This scenario directed our research to the evaluation of the neuroprotection mechanism of action observed in SEMU, focusing on the nigrostriatal pathway inflammation established in the 6-OHDA model. Considering that the neuroinflammatory process is completely associated with mediators released by glia cells, we performed immunostaining for GFAP and OX-42, in order to verify the presence of astrocytes and activated microglia, respectively. Microglia are more numerous in basal ganglia and substantia nigra, in rats, making these regions more susceptible to inflammatory aggressions, an important factor for success in establishing the model of parkinsonism [43,44]. Activated microglia and subsequent neuroinflammation have The treatment started 30 min before and 1 hour after the unilateral 6-OHDA stereotaxic injection. Two weeks after the stereotaxic procedure, the animals were sacrificed by decaptation to remove both the right and left striata. The results were expressed as mean ± SEM. Ipsilateral is the striatum of the side where 6-OHDA was injected (right lesioned-side) and contralateral is the striatum opposite to the lesion (left side). 6-OHDA group: Lesioned without treatment; SO group: Sham operated animals. this chapter do not violate copyright laws. All original sources have been appropriately acknowledged and/or referenced. Where relevant, appropriate permissions have been obtained from the original copyright holder(s). reduction of the neuroinflammatory process. The immunostainings for mediators, such as TNF-alpha, COX-2 and iNOS, in addition to the NF-kB transcription factor (Figure 4), were shown to be much more tenuous in the striatum of animals treated with the extract of M. urundeuva, proving its central anti-inflammatory action [53]. Nevertheless, NF-κB, known as the major transcriptional factor of a wide range of cytokines, shows its pathway activation much more pronounced in patients with PD. Furthermore, NF-kB is an important target in the reduction of neuroinflammation and oxidative stress. The inhibition of microglia activation associated with the reduction of NF-kB activation are probably the main factors responsible for plant-mediated neuroprotection [54]. In additional tests, we have shown that the same extract is able to inhibit the brain histone deacetylases (HDACs), which are involved with epigenetic mechanisms and activation of neuronal survival genes, promoting the protection of dopaminergic neurons in PD [55]. Chalcones are one of the bioactive components present in M. urundeuva extracts and responsible for some of the biological activities of this plant. In addition, chalcone derivatives were shown to be neuroprotective, through the downregulation of tau protein phosphorylation and insoluble Aβ peptide formation, involved with the pathogenesis of Alzheimer's disease, a neurodegenerative disease as PD [56]. Furthermore, HDACs inhibitors have been shown to be neuroprotective, both in vitro (astrocyte assays) and in vivo models (rat traumatic brain injury and brain ischemia models) [57,58]. Conclusion Taking together, all these findings indicate that SEMU appears to be a promising agent of natural origin for protection against PD and neurodegeneration. Neuroinflammatory events are known to be present in neurodegenerative diseases, including PD. They are involved with the dopaminergic neuronal loss and constitute a potential target for neuroprotection. However, it is not known whether neuroinflammation is the primary cause of neurodegeneration, but certainly its presence may contribute to the continued loss of dopaminergic neurons [59]. There are, however, still unanswered questions related to the concentrations of phytochemicals able to cross the blood brain barrier, as well as their quality, that could interfere with the final results [60]. Despite of all these problems, translational studies dealing with anti-inflammatory medicinal plants as M. urundeuva and its bioactive components should be highly stimulated and could focus on their use for the prevention and treatment of neurodegenerative diseases as PD.	2020-01-02T21:45:36.422Z	2019-03-16T00:00:00.000	{ "year": 2019, "sha1": "67a554b1304e6464b6ba9012fb7dcec16bc051da", "oa_license": "CCBY", "oa_url": "https://scholars.direct/Articles/translational-neuroscience/tnrr-2-003.pdf?jid=translational-neuroscience", "oa_status": "HYBRID", "pdf_src": "Anansi", "pdf_hash": "ed3aba0e4ea793d6c5495082aad69fe77fba6366", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [ "Biology" ] }
55247198	pes2o/s2orc	v3-fos-license	The Measures of the Persian Medicine Vis-à-vis the Treatment of Premature Ejaculation Premature ejaculation is a male sexual dysfunction characterized by ejaculation which occurs within about one minute of vaginal penetration and inability to delay ejaculation on the vaginal penetrations. The individuals suffering from the premature ejaculation will reach orgasm before even the minimal penile or sexual stimulation, prior to or immediately after penetration. Premature ejaculation is the most prevalent male sexual dysfunction. It has detrimental effects on the relationship of the partners; it may cause mental distress, anxiety, and depression. Its prevalence rates run the gamut from 20% to 30%. The etiology of this problem is unknown, yet again there are some hypotheses in this regard including anxiety, penile hypersensitivity, and serotonin receptor dysfunction. The use of the Persian Medicine and herbal remedies for the treatment of diseases has a long historical background. Nutrition is the principal stepping stone of the treatment in the Persian Medicine; it proffers various food and remedy options for treating premature ejaculation. Therefore, the authors of the present paper aimed at taking into account the nutritional advice and suggestions of the Persian Medicine to ameliorate the aforementioned issue. In this review article, the seminal and original books of the Persian Medicine on the male sexual dysfunction have been thoroughly reviewed. A list of available, inexpensive and effective foods and medications which has been experienced for centuries was elicited. Due to the prevalence and the effects of the premature ejaculation on both marital relationships and subsequently fertility, constructing treatments like this list can be efficacious for to solving this problem. Introduction Also called early ejaculation, rapid ejaculation, premature ejaculation is one of the most common sexual disorders in men-with an estimated prevalence of 30% in the world. The individuals suffering from the premature ejaculation will reach orgasm before even the minimal penile or sexual stimulation, prior to or immediately after penetration. In fact, this disorder is referred to the situation in which, first the interval between the moments of penetration to the ejaculation is unusually short and the ejaculation occurs before or shortly after the penetration. Second, the person has no control over his ejaculation. Third, this condition causes mental and psychological pressure or dissatisfaction for the affected person, his partner or both. Men with premature ejaculation tend to report less sexual satisfaction than men who have a normal ejaculation delay. The etiology of this problem is unknown, and there are some hypothesis, including anxiety, penile hypersensitivity, and serotonin receptor dysfunction [1]. Due to its prevalence and its effects on both marital relationships and subsequently fertility, understanding the causes and treatment of premature ejaculation is of paramount importance. The Persian medical scholars deem nutrition and the rectification of the lifestyle to be quintessential for the bodily health and the reproductive system accordingly [2]. In this regard, the authors of the present study aimed at garnering a sundry of nutrition strategies or measures zoomed particularly on treating premature ejaculation through an extensive scrutiny of the literature available on the traditional medicine. The purpose of this study is proffering explanations for the causes of this problem and providing a typology of food items recommended for the improvement or treatment of this irksome disorder. Methodology In this mini review article, the seminal and original books of the Iranian traditional medicine on the male sexual dysfunction have been reviewed. The criteria for selecting the sources are as follows: Their writers, in addition to practicing medicine, were remarkably successful in writing and therefore could pencil their thoughts on the topic vividly. The authors have been elicited from various centuries of medical history. These sources include the Canon of Medicine by Avicenna (Ibn Sina, 11 th century AD), Al-Hawi by Razi (9 th century AD), Zakhireye Khwarazmshahi by Ismail Gorgani (11 th century AD), Ekseer e Azam by Hakeem Muhammad Azam Khan (19 th century AD) and Tibb-i Akbari (an expanded version of the Arabic treatise Sharh al-asbab waal-'alamat ) by Muhammad Muqim Arzani (18 th century AD) and the Summary of Wisdom (Kholāsa Al-Hekma) by Mohammad Hossein Aghili Khorasani(13 th century AD), amongst others. The aforementioned works were browsed thoroughly with germane keywords like premature ejaculation and food and nutrition recommendations and measures for premature ejaculation treatment. Razi and Jorjani and Avicenna believed that patients must be treated with food as far as possible. Abu Zeid Balkhi, who was Razi's master, professed that food is the first thing to be noted and carefully considered in maintaining health. Under the section on Treatment he mentioned that drugs should be avoided unless it is absolutely needed, because unlike food, drugs are in contrast with human nature. Nazem Jahan wrote, "For increasing sexual power, food has more advantages than drugs" [2,3]. In the present research, a list of useful foods and nutrition in the Persian Medicine was gathered which is available to the general public; therefore, their inclusion in the diet of the individuals can be efficacious for the treatment merged with other conventional methods. In various recent studies, the role of many of these substances has been confirmed in ameliorating the premature ejaculation. Findings Perusing the pertinent literature, one cannot help but notice that premature ejaculation has been always the cynosure of attention for the Persian scholars; in many available sources, miscellaneous food items and nutrients have been recommended for its treatment [4]. The pioneers of the Persian medicine, like Avicenna (Ibn Sina), held that simultaneous orgasms of both males and females during copulation would be ideal for a consummate gestation to occur. In this regard, apart from providing some measures and suggestions during coitus, they zoomed on the Couples' orgasm synchronization [5]. Jorjani has written: Rapid ejaculation has four causes: excessive semen volume, thin watery semen, hot temperament of semen, and weakness of retentive power of seminal tracts. He has mentioned lowfrequency intercourse, thin semen, burning sensation on ejaculation, and ejaculation without enough erection as signs of these causes, respectively. He suggested different treatment modalities according to the causes of the disease and temperaments [6,7]. From the perspective of the Persian Medicine, the treatment for all disease or maladies in general and male sexual dysfunction in particular boils down to proper nutrition. In this view, the cause of ejaculation speed is categorized into four main groups and based on the cause, appropriate treatment is provided. The increase of coldness and wetness in the genitals: as an explanation of these patients it can be posited that their semen is too much diluted and the color approaches whiteness. These people usually do not have high libido or sex drive. On the other hand, the color of their urine is yellowish-white. The roster of foods that can help these patients include pistachios, almonds, cinnamon, saffron, hazelnuts, carrot seeds, alfalfa, stinging nettle (Urtica dioica), ginger jam, Abgoosht (or Dizi: an Iranian stew) with warm spices, cooked peas, turnips, coconut, flaxseed, sesame, carrot jam and milk and honey. Furthermore, these people should avoid taking dough (a savory yogurt-based beverage (and yogurt. Oiling the testicles with chamomile oil and exercise are very much recommended for these patients. Moreover, the frequency of intercourse should not exceed once or twice per week. The increase in warmness and wetness: semen has a moderate texture and massive volume; these individuals have high physical strength and have a lot of libido; the color of their urine is strong yellow and the color of the semen tends to be yellowish. Swollen veins/ arteries of testicles are evident in these individuals. These people have the proclivity to quaff cool water. The pungency of semen: the patient feels warm when the semen exits; the color of the fluid in these people is yellow. If the temperature of the semen is high, the semen is concentrated. However, if the temperature of the semen is not reached to the extreme, the liquid is diluted. These people tend to partake of cold foods and drinks. The consumption of lettuce seeds, tiger lotus flower sharbat or syrup, violet sharbat, rosé flower sharbat, chicory, dough and yogurt, tamarind (Tamarindus indica), Barberry juice, barley and jujube can be a great boon to these people. Weakness of the main organs of the body: the main organs of the body are those which are vital to the survival of the person and his offspring. These organs include the brain, the heart, the liver and the testicles (and in women, the ovaries). The strength of these members is absolutely essential for the proper sexual performance. People who have weakness in core organs of the body tend to have low sexual desire; they suffer from erectile dysfunction, ejaculation without proper erection and weakness in pulse. Psychological causes: Other causes of premature ejaculation include depression, anxiety, and severe phobia. The use of brewed lemon balm, chamomile, basil, lavender, orange blossom, apple juice and rose water are useful for these people (Table 1). Conclusion To wrap up, it can be safely construed that the use of nutritional supplements in traditional medicine, along with other modern treatments, can help patients ameliorate premature ejaculation to a considerable extent. It seems that combination of medical and herbal therapy can be beneficial in obtaining better result in treatment. More clinical trials are mandatory.	2019-03-17T13:11:57.224Z	2018-01-01T00:00:00.000	{ "year": 2018, "sha1": "12cd73049911aef7d7615166f08e80c333ac5fe6", "oa_license": "CCBY", "oa_url": "https://doi.org/10.4172/2327-5162.1000258", "oa_status": "HYBRID", "pdf_src": "MergedPDFExtraction", "pdf_hash": "b4618ef6b87013e5f1dce69eeaaf69dedc00fda2", "s2fieldsofstudy": [ "Medicine", "Psychology" ], "extfieldsofstudy": [ "Medicine" ] }
154829765	pes2o/s2orc	v3-fos-license	TRADE AND FINANCIAL LIBERALIZATION AND ITS EFFECTS ON GROWTH, EMPLOYMENT AND INCOME DISTRIBUTION IN LATIN AMERICAN COUNTRIES Strong capital inflows and comprehensive trade and financial liberalization characterized the last decade in the majority of Latin American countries. Despite some modest improvement in poverty incidence, the evolution of employment, wages and income distribution has frustrated even the most “Panglossian” of the Washington Consensus’s policy maker that largely run the continent along the last years. Introduction Strong capital inflows and comprehensive trade and financial liberalization characterized the last decade in the majority of Latin American countries.Despite some modest improvement in poverty incidence, the evolution of employment, wages and income distribution has frustrated even the most "Panglossian" of the Washington Consensus's policy maker that largely run the continent along the last years. Considering the evolution of household income distribution along the last two decades in Latin America countries an comprehensive analysis observed an asymmetrical pattern of growth with a high income concentration during the "lost decade" of 80's and a distributive rigidity during a more expansionist phase observed in average in the region along the nineties (Sáinz, and Fuente (2001).But even this evaluation can not be assured since there is a strong underestimation of the income of the richer strata.Due to a disappearance of regular jobs in the continent a polarization process with a hollowing out of middle class and a top-driven increase in inequality seems to be happening in many countries in recent years as a social consequence of the economic and structural changes led by external opening .But unfortunately this performance is not the bottom line.Nowadays an implosive decline is taking place in Argentina with tragic consequences on poverty incidence. Given the diversity of experiences of liberalization in the continent and the superposition of many economic and social changes to identify and even more to isolate the effects of trade and financial liberalization on income distribution it is not a simple question. In an effort to bridge a classical/sraffian theory of income distribution with a structuralist approach to economic development and a institutionalist approach to labor markets 1 , this paper tries to address to these questions considering the balance of payment constraint through its effect on interest rate, exchange rate, relative prices and in GDP growth as the dominant macroeconomic force shaping income distribution.Some routes can be singularized.From the classical/sraffian surplus approach emerges the proposition that there is an inverse relation between the rate of interest (formed exogenously by monetary forces) and product wage2 .This relationship will be considered as a clue factor connecting financial liberalization and functional income distribution.From this perspective, the level of productivity in wages goods sector is essential for the determination of real wages3 . From the classical and structuralist approach we retain the basic conception that in a surplus labor economy economic growth generates not only a reduction in poverty -an indisputable stylized fact-but trough an increase in formal employment an improvement in the distribution of labor income4 .From both approaches we take that structural heterogeneity between sectors is a primary source of income differentiation.Thus, the impact of external liberalization on income distribution depends on whether financial and trade liberalization contributes to reduce or to increase the real rate of interest, the level of employment, and the structural heterogeneity. The majority of studies on income distribution concentrate exclusively on personal/family income distribution with no concern on functional distribution.In general the nationals survey data only report evidences adequate to this concept.This dominant approach underreports non wage income -barely covered in household surveys-and exclude some important connections between macroeconomic forces and income distribution5 .This data inadequacy and other pitfalls arising from the exclusive focus on labor income will be observed and explored along this position paper. Given the high level of income concentration observed in Latin American countries, this paper will not cover the distributive social policies that in the last decade was enlarged in many countries in order to alleviate the poverty incidence, its focus will be on the forces that shape the primary income distribution6 . Besides this introduction the main argument unfolds in four sections.In the first we briefly examine how the balance of payment constraint has historically influenced the income distribution in Latin America countries.A brief analyses of the first experiment of external liberalization intended in the beginning of eighties in Argentina and Chile was viewed an avant-premiére of what happened in almost all countries in the nineties.In the second section we consider the determinants of the large capital flows and their connections with balance-of payment liberalization in the continent .We stress the major role played by external pull forces and distinguish differences on trade liberalization in Mexico (and some small countries above Panama) and in South American and the differences on financial liberalization in Chile in contrast with other countries.In the third section we analyze some evidences on growth, employment, relative prices, wages and income distribution in the continent considering two stylized situations: an "import-led consumption boon" (Taylor and Vos, 2000) based in fix or semi-fixed exchange rate and the more recent period characterized by floating exchange rate with income concentration and no growth.Some important differences are observed among countries.In the last section we conclude and explore some questions for future research. Growth and Income Distribution in Latin America before Balance of Payment Liberalization Despite remarkable national differences in the continent, the majority of Latin American countries historically exhibited large inequities in income distribution.As stressed by Londono and Székely(1997) there is nowadays an excess inequality between individuals, families or social classes not only in countries like Brazil or Colombia but also in countries like Chile, Mexico or Argentina (even before the recent turmoil) 7 .The unequalizing forces on income distribution are not new but have a long historical trajectory in the continent.Given a high concentration in land property, the historical pattern led by commodity export in a liberal economic order since the beginning of XX century in all countries, brought about a economic dynamic characterized by unstable growth and high unemployment.With exception of Argentina (or Uruguay) where a surplus of high quality land existed for a highly demanded international wage good that formed their main staple, the majority of continent had a pattern of growth where a kind of "commodity lottery"8 created a succession of boom and down cycle with large swings in income favoring the land owners.In this model of growth, a "Dutch disease" was a strong tendency in many successful export economies retarding a export diversification and import substitution and forming a large surplus labor in depressed rural areas. As a response to the international crisis of the thirties and the sharp contraction of international trade the continent was forced to create a inner dynamic based in import substitution in industry and strict control of foreign exchange.In large countries that after the II World War tried to spur an industrialization process, old and new forces contributed for a high level of income concentration.The existence of a large surplus labor in countryside and in big cities and the structural heterogeneity between wage goods sectors, specially agriculture, and industrial sectors seen to have played an important role for this distributive reality 9 .Both aspects were envisaged by structuralists authors as central factors for the income of small rural producers and urban and rural wage earners.Again, the Argentinean and Uruguayan cases differed from the continent.In fact, in these countries the industrialization and urbanization from the thirties through the fifties brought about a real wage rate expansion while the land prices declined improving the income distribution10 .But as soon as the balance-of-payment constraint interrupted the economic growth through increases in exchange and interest rates, inflationary pressures exerted a depressing effect on real wages.The "populist" cycle of go-and-stop in Argentina was an attempt to change income distribution in the presence of strong external constraint11 .But this situation was not peculiar to Argentina but a more generic case in Latin America. The balance of payment constraint that chronically accompanied the industrialization process in Latin American countries was the main macroeconomic force for the persistence of high levels of surplus labor and for structural heterogeneity.In fact, in a situation marked by small financial inflows and a high dependency of export revenues on commodities, a limited capacity to import was the main constraint for a higher rate of growth and employment. In an " industrialization -cum -dollar shortage", pressures for exchange rate devaluation and for systematic increases in food and raw material prices brought about a persistent inflation and, in many countries, a short cycles of economic growth.No other country exceed Brazil in its effort to industrialize and no other large country exceeds Brazil in income concentration.A surplus labor with a initial shortage of "good" lands in a heterogeneous primary sector (very different situation compared to Argentina) shaped a very concentrated wealth and income distribution.A huge surplus rural labor coupled with a significant lag in food productivity for internal markets generated a high poverty level and a permanent pressure on unskilled labor.Different from Mexico, that industrialized without strict exchange control, in Brazil the balance of payment constraint exerted a bigger pressure on real rate of exchange and inflation.With a crawling peg adopted in midi sixties coupled with a rate of interest fully adjusted for inflation, the real wage rate for unskilled labor was repressed widening the income differential between capital and labor and within skilled and non-skilled labor. Although the orthodox rhetoric emphasize the negative aspects of the "import substitution industrialization" particularly along the seventies -is better to call it as state-led industrialization12 -the persistence of high rate of economic growth with structural changes brought about a significant reduction in poverty and in surplus labor in the majority of Latin America countries causing in consequence a decline in inequality in many of them. agricultural employment and the absence of a corresponding increase in food supplies that made continued inflation inevitable.It caused a persistent upward pressure in the prices of food paid by the urban consumer, and it was the rise in food prices which caused the persistent rise in urban wages and salaries...." "...the basic cause of the inflation was the disproportionality in the growth of production in different sectors of he economy, ...between wage-goods and non-wage goods" (Kaldor 1978:128) Due to the extraordinary surge in international credit supply that followed the end of Bretton Woods monetary system, the net transfer of resources grew from zero to almost 3% of real GDP in Latin American countries along the seventies13 .A positive income effect spread all over from this growth-cum -debt pattern.With a negative real interest rate and a high growth of exports the debt accumulation did not arise major concerns 14 .In countries like Brazil or Mexico, where a developmental state were strong, the surge in capital inflow was accompanied by a high growth in exports and in import substitution.The strict control of financial system responded for this strategy.In many countries the dramatic end of "dollar shortage" and the beginning of a era of "dollar abundance" stimulate a surge in imports and income growth. Many factors contributed to decline in poverty and in improved income distribution that characterized this decade in the continent, specially the reduction of rural surplus labor absorbed in modern activities.The real important exception occurred in Chile.The institutional rupture and a monetarist economic policy practiced in the second half of the decade in this country had important distributive consequences on employment, wages and in poverty levels. The shifts in relative prices in favor of oil, raw materials, and foods stuff affected unequally the different countries.In oil exports like Mexico or Venezuela the change in prices favored a boom in exports and in economic growth, in Mexico, thanks to the social democratic policies implemented along the decade the income distribution improved.In Brazil, despite an intense reduction in poverty and a high positive income mobility, the income distribution did not change substantially.The shift in terms of trade with a stable real rate of exchange and positive real rate of interest increased the inflation rate precluding partially the transformation of the positive employment effects on higher real wages.In Brazil as well as Mexico or Argentina the role of sector public not only as investor but as employer was very important in promoting a richer and bigger urban middle class through a growing demand for skilled labor. In fact despite the high difference in wage levels between the rural labor and the urban professionals, the urbanization and economic growth of this period brought about a significant growth of middle class job.It is important to note the dominant role of public employment amongst Latin American professionals15 . Given the high concentration of land property it is very difficult to use the inverted Kusnetz curve in Latin America countries for the long run, but if we compare the seventies years with the two decades before, we can argue that the transformations induced by high growth and structural change with a reduction in surplus labor in countries like Brazil or Mexico were shifting the unequalizing forces toward the descendent part of a Kusnetz curve.This whole situation dramatically changed in the eigthies. During this decades, the Latin America countries experienced a low growth (1.2% in average for 18 countries), a sharp fall in regular employment, an increase of more than 3 points in GINI index ( a growth of 6% in great contrast with the decline of this index during the seventies), a decline in real wages, in wage share and an increase in poverty.(Londono, Székely, 1997).The unilateral decision of American Federal Reserve to increase the interest rate and the sharp valorization of US$ against the yen and the mark brought about a real interest rate on debt of 14,5% between 1981/1985 in acute contrast with the negative rates charged in the seventies (Schulmeister, 2000).With the Mexican default in 1982 new capital ceased to flow to the continent, with partial exception of Chile (for geopolitical reasons) and Colombia.The net resources transfer became negative in more than 4% of GDP in average (ECLAC, 2002).Severed from international finance, the highly indebted countries had to promote an intense export effort.Despite this, the rate of growth exports in US$ was lower than the rate observed in Latin American countries in the seventies.The decline in the terms of trade more than compensated that bigger export effort.(ECLAC, 1995) The high inflation and the distributive conflict that characterized this decade was a consequence of this dramatic increase on external solvency and liquidity problems that spread in the majority of the countries16 .Before we explore with more details the consequences of this change in balance of payment constraints on income distribution , lets consider very briefly the first experience of trade and financial liberalization in the continent practiced in Chile and Argentina from 1979 through 1981. Supported by the IMF and guided by a monetary theory of balance of payment, the Chilean government changed the previous stabilization policy based on monetary control and started a comprehensive program of trade and financial liberalization.( Díaz-Alessandro, 1984, Frenkel, 1988).A fixed rate of exchange and a liberalization of capital account were the crucial policies of this strategy.Based in the supposition that Chile was a small country that could obtain the capital influx it needed and that difference between domestic and foreign debt was not relevant, the local authorities with the IMF approval, allowed for a fast expansion of a current account deficit financed by short term debt.With a strong valorization of the real exchange rate, large spreads in the rate of interest were used to attract the internationally abundant short term capital17 .After few years of rapid income growth and import boom with real wages increases, in few years the local banks accumulated a great stock of short term debt. 18his experience was very similar to that occurred at this same period in Argentina.In 1982 it was over. With a radical contraction in external loans and finance outflow to Latin American countries, the rate of exchange in Chile and Argentina was strongly devaluated, the controls of capital were introduced, the trade barriers were erected and the private debt was nationalized.A huge wealthy transfer from public to private hands and a capital outflow from Latin American residents were a common consequence of this episode 19 . It is important to note that the contraction of GNP in Chile -15% between 1981/83-in Argentina -11%-and Uruguay -14%-exceeded the contraction occurred in the rest of continent in these years, a direct macroeconomic consequence of capital account liberalization.After that avant -à-lettre experience (that was spread in the continent in the nineties), Argentina, Chile and Uruguay followed the same and compulsory macroeconomic policy based on the achievement of large trade surplus in order to finance the international debt service through exchange rate devaluation and contraction on internal absorption.Low and unstable growth, high inflation, high interest rate, huge decline in public investments, decrease in formal employment were common characteristics that explain the general expansion of poverty and income concentration. Although this situation were most alike in the continent, important differences have to be noticed.In the midi of the decade the rate of interest on American Bonds fell down and the US$ devaluation against the yen and the mark yielded an important reduction on real interest rate on international debt.In addition, the world export prices and the real rate of world export growth increased.This created opportunities for indebted countries to re-launch a trajectory of economic growth.Chile and Colombia explored this opportunity.The common aspect of these two countries was the higher rate of export growth in US$ than the rate observed in the majority of Latin American countries.(ECLAC, 1995).Chile was particularly lucky in its terms of trade and more than other countries was politically supported by American hemispheric political interests.Given a well succeeded keynesian and sector developmental policies, these countries could diversify their exports (along traditional and non-traditional manufactures) preserving competitive real rate of exchange through tariffs and capital controls. Despite a high export effort, the Brazilian economy had very low growth far behind Chile, Colombia or even Mexico.With a highly indexed price system, all the efforts for real devaluation and increases in real rate of interest were followed by nominal increases in wages and other contracts enacting a self-sustained inflationary process.In a high inflation environment the wages less protected, the non-asset holders, the less skilled and the small land producers suffered the burden of inflationary crisis.This process was partially suspend during the short periods that followed several "heterodox stabilization plans" that failed for the same reason: the lack of external finance.Some social policies established in the middle of the decade were effective in reducing the social consequences of this crisis but were not enough to interrupt this unequal path. As we have seen the balance of payment constraint trough its impact on the rate of income growth, on the rate of interest, exchange rate and relative prices played historically a dominant role on macro forces that influenced the evolution of poverty levels and income distribution in Latin America countries.With the abundance of capital inflow that came to the continent in the nineties the balance of payment constraint has sharply reduced, but contrary of what happened in the seventies a mild, unstable and unequal growth with a high rate of unemployment predominated in the majority of the countries 20 .Before we examine how the liberalization of balance of 19 According to a World Bank estimate the capital fugue in Latin America reached U$ 90 billions between 1980/1983, mainly from Argentina, Mexico, Venezuela, and Central America. 20For an analysis of how an epoch of dollar abundance was accompanied by mediocre economic growth in Brazilian case see Medeiros and Serrano (2001) payment that spread all over in Latin America in the nineties contributed for this deceptive result lets observe briefly the macro dynamic of capital flows in Latin America countries. Capital Flows and Balance of Payment Liberalization in Latin America: A New Financial Lottery 21 As it happened in the seventies, the large volume of capital flows (debt, foreign direct investment, portfolio equity) that came to Latin America in the nineties, was largely autonomous from the domestic policies practiced in these peripheral countries.Deregulation and innovations in American and international financial markets, reschedule of public debts of high indebted countries along Brady' initiative (as it happened in Mexico, Venezuela, Argentina or Brazil), reduction in the international (American and Libor) interest of rate, continuous devaluation of US$ (up to 1995), and diversification of European banks to new external markets were the external push factors for the surge of capital flows in the continent.(See Table 1 and 2) The sudden dollar abundance was the material basis for the launch of a strategy of balance of payment adjustment that intended in many Latin American countries to reach simultaneously a price stabilization and sustained economic growth.In a rather different approach from the seventies, the trade opening that started in many Latin American countries in the second half of the eighties, in the middle of dollar shortage, and the capital account opening (as it was spread with different intensities in the continent in the beginning of the nineties), were considered in Argentina or Mexico, or Brazil but not in Chile or Colombia as an industrial policy in itself in a macro strategy very similar to that implemented in Argentina and Chile in 1979.This strategy had the same characteristics we already has observed: the use of the exchange rate as nominal anchor (in a rigid fix parity as in Argentina case, or in a semifixed regime as in Mexico before 1994 or Brazil before 1999) in a open and deregulated financial market brought about an appreciation in real exchange and a huge surge in imports 22 .The rate of interest was pushed in order to sterilize capital inflow and attract new capital inflow that came in seeking for high arbitrage opportunities created by the fix (or semi-fixed) exchange regime 23 .The radical shifts in current account-from surplus to deficit-and a fast increase in deficit ratio to exports stimulated speculative movements (led in general by national resident bond holders as it was evident in Mexico in 1994 or in Brazil in 1999) against the exchange regime.Increases in interest rate contributed to alter the composition of capital flux increasing the rate of short term capital to reserves.The Mexican crisis in 1994, or Brazilian in 1999 or Argentinean in 2001 was in this sense, very similar to that happened 12 years before in Southern Cone .24 Despite the similarity of all crisis in the age of liberalization the national meaning of balance of payment liberalization was not all alike.In the case of Mexico there was a big difference.Mexico was highly backed by IMF/US Treasury not only because Mexico was big for international markets, the common argument, but because Mexico started to be, in a context of NAFTA agreement, a huge market for U.S. exports.In fact, during the eighties the large American trade deficit with Asian countries was enlarged by Latin American exports surplus in order to finance the negative transfer of resource.In the nineties, a large American export surplus with the continent was the counterpart for a positive net resource transfer.But as a matter of fact, the contribution of South American to U.S. exports was and still is very low in contrast with the growing importance of Mexico imports. The IMF/American Treasury rescue of Mexican economy gave more time for the survival of this exchange rate regime in the continent, not exactly in Mexico, or in Chile where it did not return since its failure in 1983, but in South American Brazil and Argentina.From 1994 until 1999 Brazil practiced the same model with the same dynamic and results ending with the exchange rate devaluation.During this period the Brazilian imports were essential for the survival of the rigid Argentinean exchange regime. These policies grounded in free trade and liberalized finance has collapsed in the end of the decade when the net resources transfer to Latin America countries became again negative (ECLAC 2002).At the end of the decade the majority of countries adopted a fluctuating exchange regime.25 (See Table 3 and 4 in the Annex) In 1983, the balance of payment liberalization in Southern Cone was defeated by the high American rate of interest and the high valorization of US$.Again, that same factors and reasons were in the roots of this new failure.In fact, with the appreciation of US$ dollar in 1995 and an increase in the international rate of interest, the real interest rate on international debt increased -see Table 1-; after 1998, the rate of import in industrial countries declined and the terms of trade and the Latin American exports revenue dropped.(See Table 5). In times of a surge in foreign direct investment that came in great volumes in Brazil or Argentina in the last years of the nineties, the ratio of current account deficit to exports -a solvency ratio-has grown and so the external fragility.Contrary to the prompt rescue occurred in Mexican case (an north-American integrated country) and in Asia's 24 Despite its persistent failure , in a Report on Capital Account Liberalization the IMF still considers that there is not anything unique in external liberalization that distinguish it from domestic financial liberalization "...the mechanisms through which internal and external financial liberalization can expose threats to financial stability are largely the same.Both internal and external liberalization squeeze margins and leave less leeway for poor loan and management practices.Both give banks and other intermediaries additional access to risky investments.Both give banks gambling for redemtion access to additional sources of expensive funding.There is nothing unique or different about external financial liberalization in this context."(IMF, 1998) bail out in 1997 (an American effort to impulse trade and finance liberalization in the region), the big default and crisis in Russia in 1998, inaugurated a different approach to be followed by IMF/ American Treasury.In consequence the commercial banks assumed a more conservative exposure in emergent countries resulting in a sharp contraction on external loans.The Latin American international bonds issues declined and the Brady bond spread on U.S. bond spread grown substantially ending the liberal expectation for a new survival of this model. As we observed the external forces played a dominant role on outflow of capital from industrialized countries.These flows showed, by its turn, a pro-cyclical pattern contributing for a high volatility in economic growth in LAC.In fact, given the growing dependency on external markets and on the external rate of interest the macroeconomics policies implemented in LAC did not counter balanced the external cycle, on the contrary, it showed a pro-cycle performance, in a succession of bubbles of imports surge and private capital gains in equities followed by depressive policies aimed to "restore confidence" (Ocampo, 2002) when external cycle of investment was reversed.A remarkable aspect of this macroeconomics is a permanent pressure on domestic interest rate.In a fixed exchange regime the rate of interest was used as a major instrument to attract mobile capital, in a float system to control the exchange flotation.As we will observe in the nest section this monetary policy based in a ever-high-interest rate was specially strong in Brazil.(See Table 6 and 7) A comprehensive financial liberalization -a free substitution of dollar and domestic assets-that characterized many Latin America countries during the nineties was not a necessary condition to obtain a positive flux of resources necessary to resume the income growth.The Chilean experience with non remunerated deposits established in 1991 did not sever the country from financial markets but was positive to protect its export competitiveness and to improve the maturity composition of net capital inflows.The same can be said about Colombia that followed a similar policy (IMF, 1998).In this sense the associated effect of capital liberalization seem to be the excess of short term capital and the volatility in income growth.The simultaneous removal of trade barriers and cut in tariffs with a overvalued rate of exchange -as was argued by Ocampo and Taylor(1998), and Taylor and Voz (2000)-amplified the destabilized effects of financial liberalization on income growth . Although the external push factors seen to explain the boost of FDI inflow in Latin American countries, the internal deregulation and privatization in services (finance, insurance and business service) and in public utilities (telecommunications, electricity) was indeed a crucial factor pull factor for the enlargement of international presence of transnational corporations in essential sectors. Through these routes the influence of balance of payment deregulation on income distribution was very significant and as we will argue in next section, the mechanisms of transmission include much more than its effects on wage differentials. Growth and Income Distribution in Latin America in the Age of Balance of Payment Liberalization We have seen from the previous analysis that the macroeconomic dynamics of Latin America in the nineties was shaped by the interactions of the surge of capital flows accompanied by internal pull factors induced by financial deregulation and trade overture.In the case of Mexico, the interactions with the American regional trade policy was dominant.From a macro perspective the main novelty observed in the nineties was the sharp discontinuity with the 80's and the instability observed in growth, inflation, relative prices employment, and wages.Lets observe briefly some data. First of all, it is important to notice that the rate of economic growth for the majority of Latin American countries was low by historically standards.Between 1990 and 1999 the average rate of GDP growth in Latin America was 3.2% in contrast with 1.0% observed in the eighties but far behind the 4.9% observed between 1951/80.The rate of per capita growth was 1.4% far exceeding the 1.0% negative observed in the "lost decade".But this performance was very differentiated with a high dispersion within the continent.Chile reached a GDP a high rate of 6% in sharp contrast with the Brazilian rate of 2.5%.If we exclude Brazil, the average rate of GDP growth in Latin American was 3.6% and 1.7% in per capita GDP (ECLAC, 2001).In comparison with its own history or with others Latin American countries, the Brazilian economy had in the nineties a very poor growth performance. As already observed this low rate of GDP growth was remarkable unstable (see Table ).Large discontinuities occurred in 1994, 1998 and 2001 and the pattern of growth followed the movements of net transfer of resource.Again this characteristic was differentiate between countries.Argentina and Venezuela had a "roller coasted" (O'Connell, 2002) kind of growth in contrast with Chile or Colombia where the GDP volatility was less pronounced. In a huge contrast with the 80s, in the era of dollar abundance, the annual rate of inflation has dramatically contracted from three digits in the beginning of the decade (four digit in case of Brazil) to one digit reached in the in the middle of the nineties. In no less impressive contrast with the precedent decade, despite the high rate of exports achieved in many countries, the deficit in balance of trade and in current account was a general aspect in the region (ECLA, 2001).The rate of import growth far exceeded the rate of exports.This rate was dispersed and important differences between Mexico and the countries below Panama affirmed in more recent years.Again the Chilean economy distinguished by its lower sustainability (deficit in current balance and exports). As we have observed in the last section, a high real interest rate distinguished the Latin America countries along the nineties.In context of a great GDP volatility a pro-cyclical behavior of public revenues was accompanied by a high interest rate resulting in consequence a high contraction on non financial outlays with a depressive effect on growth and employment.Again this feature was very dispersed across countries.Brazil has presented the highest real rate of interest and Chile the lowest.(Table 7) Some remarkable structural changes and transformations in labor market occurred in the nineties induced by these macro forces.In the case of Mexico and others small countries above Panama these transformation were led by the NAFTA arrangements: -save for Chile, in most countries the last decade was accompanied by a low growth in aggregate productivity and low investment ratio but with a very high dispersion between and within sectors.In Mexico, a high growth in manufacturer sector driven to exports was in sharp contrast with the low productivity in agriculture and service sector.According to Hofman (2000) between 1991/98, Mexican's labor productivity did not increase in aggregate terms despite the large increases observed in manufacture.In Brazil, the aggregate labor productivity increased only 1% between 1991/98 (Hofman, 2000).The productivity gap between agriculture and industry has declined, but the productivity differences within the manufacturer sector, and between the whole sector and the services sector has enlarged; -between 1991 and 2000, the rate of unemployment has increased significantly -see Table 3 -in almost all countries principally in Argentina and Brazil and has declined in urban Mexico.The rate of employment followed the rate of economic growth with minor changes in elasticities (Stallings, Weller, 2001); -during the nineties the Latin American countries became a source of mass immigration (legal and illegal) to the United States.This flow, highly concentrated in Mexico, Dominican Republic, Haiti, El Salvador, Jamaica and Cuba responded for 50% of American legal immigration (circa of one million people every year).(Borjas, 1999); -between 1990 and 1998, formal employment in private enterprises and in public sector has declined in Latin America and the informal employment has increased.This downgrade in employment structure was most pronounced in Brazil but affected all countries with the only one exception occurring in Chile (Tokman and Klein (2000) based in ILO data); -despite the general and almost universal fall in Agricultural employment, the industry contribution for employment growth was strongly differentiate among countries.In countries above Panama, the manufacturing employment has expanded in contrast with the pronounced fall occurred in southern countries.In Mexico, in sharp contrast with Brazil, the agriculture sector absorbed labor mainly in low productivity activities.In all countries the growth in employment in finance services, far exceeded the average employment creation (Stallings, Weller, 2001). The principal findings on income distribution can be thus summarized: -the information about functional income distribution is precarious and scanty among the countries.The share of wages in manufacturing valued added fell down in the majority of countries between 1985-1992(UNCTAD, 1996)).In Brazil and Mexico the wage share in national income has strongly contracted; -the Gini index on personal income distribution has increased along the decade in Argentina (Frenkel, and Rozada, 2000), Bolivia, Chile, Colombia, Mexico, Peru, Venezuela and many among others (Behrman, Birdsall, Szekely, 2001) and stayed high in the same level in Brazil26 .The incidence of poverty has increased in Argentina, Mexico and Venezuela an fallen down in the majority of others countries.(Behrman, Birdsall, Szekely, 2001); -in general the employment of low income earners (first and second quintile) and high income earner (fifth quintile) had a faster rate than the employment of middle income earner (third and fourth quintile).(See Table 8).This hollow-out of middle earners was very intense in Brazil and reflected the de-industrialization of labor force.In Mexico, the labor income concentration was essentially caused by a decline in real income of lower and middle earner and the expansion of real income of high income earner (Tookman, Klein, 2000); -what distinguish LAC income concentration is the importance of the concentration at the top of the distribution.In the last years considering only the standard household surveys (that shows a strong underestimation of profits and interest earnings) a top-driven inequality favoring the richest 1% individuals has been into play in LAC reflecting co-movements in the structure of employment and the distribution of wages (Székely, Hilgert, 1999); -the average real wage in formal sector has moderately grown in many countries.It showed a steady growth in Chile, and a decline in Argentina, Mexico, Peru and Venezuela (Stallings, Weller, 2001 based in ECLAC data). -in average in Latin American countries the minimum real wage and average industrial wage have increased slowly in many countries the minimum wage has not accompanied the expansion of average wage (Tookman, Klein, 2000 on ILO data). (See Table 9).In Mexico, the real minimum wage has declined.The real minimum wages paid in 2000 were in average much lower than the wage paid in 1980.Chile and Colombia were important exeption followed by Brazil (ECLAC, 2002) ; -the information on wage structure does not permit a detailed analysis on labor market segmentation.The conventional proxy for skilled labor is based on schooling years and the remuneration associated to it is assumed to be the premium paid to skill.In many countries the average wage of graduate professionals to workers with 7-9 years of schooling sharply increased in the nineties with the exception of Argentina that showed a moderate increase and Brazil that did not show a significant rise.(Starling, Weller, 2001).In both cases a "devaluation of human capital" seen to have appeared (Sainz, and Franco, 2001).This general trend in the wage gap between skilled and non skilled in the majority of Latin American countries is well documented in studies based on nationals data; -even when controlling for personal characteristics the segmentation in labor market, between large and small firms and between long term and short term labor contracts was a decisive factor underling the wages differences.(Tookman, Klein, 2000). The majority of studies on income distribution in Latin America stress the role of trade liberalization only on the distribution of labor income.Based in a "before-and-after" country narratives on balance payment liberalization covering 17 Latin American countries Taylor and Vos (2000) try to distinguish the effects of trade and finance liberalization on income distribution.They conclude that "...the structural changes associated with trade liberalization have generally caused a rise in income inequality, most pronounced in a widening of the income gap between skilled and unskilled workers.Capital account liberalization leading to greater capital inflows and through that aggregate demand expansion, employment growth and/or price stabilization has offset tendencies towards greater inequality and permitted poverty reductions in a number of instances.Yet financial opening also has been associated with greater volatility, impeding sustained improvements in equity or poverty reduction."(pg 3) The mechanism of transmission from trade liberalization to personal inequality considered by the authors in many Latin American national experiences is based on structural changes in demand composition for skilled and unskilled labor in a deregulated labor market.As far as the capital account liberalization is associated with the increase on capital inflow the authors considered that it had mixed effect for income distribution.A positive influence trough exchange rate and income growth on real wage and poverty levels, partially counterbalanced by a high volatility precluding a sustained improvement in income distribution. For Ros (1999), the substantial increase in wage inequity occurred in the last decade in Mexico can also be attributed to shifts in composition of demand favoring the skilled labor.The reduction in the mark-ups in the tradable sector due to trade overture has lead to a high cut in non skilled labor increasing the premium paid to skill. In a very different methodological approach Behrman, Birdsall and Szekely (2001) emphasized the effects of financial liberalization on income distribution.Considering a trade liberalization index (average and dispersion of tariffs), a financial liberalization index (controls on FDI, current balance controls, capital account controls) and inequality index (Gini on household surveys) and poverty incidence they conclude that trade liberalization did not "explain" the changes observed on inequity and poverty in Latin America countries along the eighties and nineties.The financial liberalization seem to them to have played a mayor role for the increase observed in inequity. 27 From our point of view, the connections of financial and trade liberalization on income distribution in the nineties were not qualitatively very different from those we historically observed in the region.Lets consider primarily the "import led consumption boom" based on overvalued real exchange rate.The analysis will be concentrated in Argentina, Brazil, Mexico, where a manufacturer sector has been so far the machine of economic growth with a important export diversification (far more developed in Brazil than Argentina or Mexico previously the NAFTA) and in Chile and Colombia.In countries like Venezuela, Bolivia, Honduras, Nicaragua or Peru, although the financial and trade opening have implied in common aspects observed in large countries, the dominant factor that sets their economic cycle and income distribution is still governed by the external prices of their main commodity.Small countries like Panama, Jamaica or Republic Dominican have their principal income from services and private transfers. Due to high inflation that typically characterized Latin American countries in eighties and in beginning of nineties, the exchange rate based stabilization programs brought a sharp decline in inflation with a positive impact on growth and poverty incidence.This was very important in Brazilian and Argentinean experiences of hyperinflation .In this phase a imports surplus resulted from the exchange appreciation coupled with trade liberalization.The expansion of domestic credit and the increase in real wages boost the internal demand and economic growth.A high consumption of upper classes follows the wealth effects brought about fast growth in asset and equities prices.In the cases in which the cut in inflation comes together with a lower real interest rate, the prices of 27 The mechanisms of transmission from financial liberalization to inequity envisaged by authors in this econometric paper are based in a (neo-classical) supposition that a greater capital influx associated with financial liberalization increases the "capital endowments" yielding a decline in the price of capital.Assuming a complementary relation between capital and skilled labor, the authors conclude that the financial liberalization shifts the labor demand increasing the returns to skill. wages goods and allows increases the minimum real wages, the positive effect on income distribution depends on the length of this phase and the effects on labor market.While the shifts in relative prices against trade goods and the displacement of domestic manufacturer occurs as happened everywhere but specially in Argentina and Brazil, the contraction in manufacturer employment and the expansion of service and non salaried workers push downwards the job structure and average wage.(See Table 10) .The fast increase in import coefficient leaked to outside the stimuli of income growth on employment 28 . The decline in industrial mark-ups in the tradable sector and the increase in dollar labor cost compelled a strong cut in employment -mainly non skilled-and modernization in the manufacture that could resist the strong external competition29 .The capacity of competition contemplates different situations across LAC.In Mexico the winners were the manufactures integrated to U.S., outside Mexico the manufacturers with important absolute advantages based on natural resources.In a country like Brazil, the imports penetration in electronics and capital goods caused a significant import de-substitution.Combined with a huge penetration of imports in traditional sectors as textiles, footwear and garment the trade and financial liberalization brought about an important deindustrialization in labor force. The increase of industrial unemployment in countries like Brazil or Argentina made the positive effect of exchange rate on wages be partially counterbalanced by diminishing bargaining power of industrial wage earners.The results in wage share and in labor income distribution depend of the intensity of this process and the evolution of the interest rate and the minimum wages.In Brazil for instance, the decline in industrial and rural employment and a consequent expansion of urban informal employment was accompanied by an increase in minimum wages.As soon as a high interest rate was in motion in order to sustain an overvalued rate of exchange the pressures on employment and in labor costs levels in traded sector held the average wages.A significant contraction on wage share has occurred.A kind of hollow out of middle wage earners has befallen simultaneously with a small but positive growth in low wages.Measured by conventional concentration index (as Gini index) the income distribution has not changed in Brazil along the nineties, a clear result of changes of opposite directions. In Mexico, despite a major industrial employment, the fiscal contraction and a high interest rate destroyed employment outside the "maquila" sector generating a large surplus labor.Because Mexico did not have a high inflation as Brazil or Argentina the cuts in inflation resulted from the exchange rate anchor was not followed by a such high reduction in poverty incidence and contrary of what happen in Brazil the minimum wage has declined.Contrary what happen in Brazil, in Mexico, the productivity in agriculture has lagged far behind the industrial productivity.Probable, the major income concentration observed in Mexico in the nineties is due to the effects of trade liberalization on Agriculture sector (Stalling and Peres, 2000). As well as in Mexico, in Argentina the wage distribution worsened all the time.The main process was the decline in traded employment and a rising iniquity in non traded sector mainly in services (Frenkel, 98, Taylor and Vos, 2000). The rise of productivity was very unequal among sectors and within sectors.In Brazil a high agricultural productivity growth diminished the agriculture/manufacture gap but the structural change occurred increasing the gap between manufacture/industrial services and personals services.In Mexico the agriculture sector was a sponge for surplus labor.A productivity gulf inter and intra sectors enlarged the heterogeneity of the whole economy acting negatively on distribution of labor income and in capitalist rents 30 . The "import led consumption boom" did not have a long life in Latin America.With the current account deficit and the external financial fragility rising, the authorities has driven first to contractionary policies centered on fiscal cuts and high increases in interest rate.After the "Tequila Crisis" in 1994 and Brazilian exchange rate crisis of 1998, the devaluation on rate of exchange in float exchange rate regimes inaugurate another macroeconomic cycle with different impacts on income distribution 31 . In the previous exchange rate regime, a high interest rate was necessary to attract capital inflow to sustain a rigid exchange rate, in a float exchange rate regime a high interest rate was necessary to prevent sharp devaluation and control the pressures on inflation 30 Ocampo and Taylor (1998) connect trade liberalization and income distribution though its impact on skill demand and on the structural change associated to that policy.If, as a consequence of a liberal trade reform a divergence in productivity among sectors increases in a low grow economy, the income distribution effect will be very different from a more balanced productivity growth. 31 "A surge in capital inflows associated with exchange-based stabilization programs and accompanied by trade liberalization can artificially boost real wages by allowing exchange rates and trade balances to deviate significantly from their sustainable levels.Because of the erosion of competitiveness due to currency appreciation, employment tends to decline in manufacturing, although it may increase in services.While tariff cuts and appreciation lower imports costs for industry, and efforts to rationalize production lead to labor shedding profits in industry may nevertheless be squeezed as increased competition reduces firms' total sales.When the bubble bursts and capital flows are reversed, the exchange rate comes under pressure, and a deflationary adjustments follows, involving cuts in domestic absorption and depreciation of the currency.Labor tends to lose the gains achieved during the boom phase, not only through shrinking employment; while there is a rise in employment in export sectors, it is often more than offset by a decline in the non traded goods sectors..." (Unctad:1997:140) Similarly Taylor and Ganuza (1998) grounded on evidences for 15 countries in Latin America in the last two decades distinguished two stylized scenarios.In the first, the inflation speeds up and balance of payment worsens, devaluation and fiscal contraction generates a go-stop cycle with a average low growth.The effects on poverty as occurred in Latin America during the eighties depend on the extent of output cuts and reduction in real wages related to devaluation and inflation rate."The second scenario is more relevant to the 1990s, when large capital inflows together with liberalization of current and capital accounts of the balance of payments permitted visible output and employment expansions.Inflation rates went down (dramatically, in some cases) in response to exchange rate-based stabilization programs.In the 1970s, favorable external factors...went together with similar phenomena.Poverty typically declines in such episodes because of faster output growth as well as slower inflation and or increases in real minimum wages which price stabilization permits.However, residual domestic price increases often lead to real appreciation (especially if the nominal exchange rate is pegged as an anti-inflation "anchor") with possibly favorable effects on real wages but adverse impacts on output in traded goods sector.Ultimately, the authorities may be driven toward contractionary policies (fiscal cuts, high interest rates exacerbated by attempts at sterilization of capital inflows) to attack widening trade deficits.A return to go-stop cycles is a real and present danger, especially if external conditions deteriorate as during the "Tequila crisis of 1994-5". ( 1998:3) rates.Thus, a high interest rate seem to be a persistent and long lasting effect of trade and financial liberalization 32 .Naturally, this situation is not alike across the countries and depends on how the exports react to devaluation.In this respect there are two different histories.In Mexico and other small countries above Panama, after 1994, a fast rate of exports released the external constraints favoring a higher income growth despite a meagre result in terms of employment.The competitivity in labor intensive sectors as electronics, garment, textiles in a cross border integrate trade network has rapidly increased.For this performance the low wages paid in industrial sector was an important component.In addition, the high growth in Mexican exports was extraordinarily pulled by American growth -that responds for about of 90% of Mexican exports-occurred along the second half of the nineties. In Brazil, the rate of exports in dollars did not respond so fast and so high.The structure of exports -more concentrated in sectors based on natural resources -in a epoch characterized by a huge decline in terms of trade and tariff protection in industrialized countries explain this result.The sharp increase in interest rate and a drop in fiscal real disbursement in order to attract capital influx, constrain the imports rate and controls the internal debt was the basic economic policy of this regime.In Brazil, as well as many other Latin American countries, the excess of interest rate over GDP growth was a basic factor for a depressive macroeconomics.The trade adjustment that succeed from this policy was basically centered on imports reduction. With a devaluation in real exchange rate, the positive effect on poverty incidence and in real wages brought about by the shifts in relative prices during the "import-consumption boom" was partially canceled.As it happened in Brazil or Argentina after the devaluation, the inflation pressures are growing despite the economic recession 33 . The price pressures on real wage comes additionally from other source.As a result of privatization and internationalization, the tariffs of public utilities have increased everywhere affecting most the urban low income earners.The positive impacts on employment that resulted from exports and import substitution did not compensate in any country the negative distributive impact from the shifts in relative prices and the negative impact that a depressive macroeconomics exerted on the general level employment. 32 "...Where do the high interest rates often observed in the wake of market liberalization come from?"Ask Ocampo and Taylor (1998) and they answers, "One possible explanation is that economic actors at home may pull back from the local market in a dynamic process."(pg14) As we have seen, first of all, the domestic rate of interest depends on the American rate of interest over American bonds.Given that rate, the expectation of devaluation and the country risk explain this almost permanent pressure on rate of interest in the continent.Efforts to sterilize capital inflows in anti-inflation programs based on exchange rate used as nominal anchor -as was the case in many Latin American countries in the first half of the nineties-or to control a float exchange rate as it was the case in many countries from the midi nineties do not change substantially this broad pressure that result from financial fragility. 33Considering the social consequences of Argentinean acute crisis O'Connel following the collapse of the exchange rate regime observed that, "One dramatic element of this matter is that food products being the major exportables, their prices have tended to rise in association with the exchange rate, hitting relatively more the lowest paid sectors of society.Not surprisingly, therefore, that as shown above poverty levels have shot up and the proportion of people below the food-poverty line has increased to a socially unacceptable level."(O'Connell, 2002).As we have seen in the first section of this paper this situationapart for its dramatic dimension-has been a permanent tension in Argentinean economy since the Second War Two revealing the modesty of structural change occurred in the last decades. Open Questions for Research and Final Remarks In the last section we presented some evidences and explored some connections that running from financial and trade liberalization affect directly or indirectly the income distribution in Latin America countries.Although its is very difficult and to some extent meaningless to isolate the effects of financial and trade liberalization (in a sense that the real exchange rate that is essential for the balance of trade is strongly influenced by financial liberalization) on income distribution let us summarize some transmission mechanism and explore new aspects. Financial liberalization in Latin America countries has created a high volatility in income growth engendering a short cycle of economic activity following the capital influx movements.Coupled with trade liberalization, it has increased the Latin American external financial fragility and, consequently, has created a strong instability on the rate of exchange.A high interest rate and a pro cyclical fiscal expansion amplified the discontinuities in income growth.As it was considered this cycle has generated a pattern of distribution formed by a sequence of exchange rate evaluationincome growth-expansion of internal credit -expansion of real wage with changes favoring the most qualified workers followed by exchange rate devaluation -credit crunch-increase in interest rate -decline in income growth and in real wages. From this sequence some structural and long lasting effects seem to appear: a "bad specialization" and dependency on commodities and on exports based on natural resources increased ( as it happened in the countries below Panama) or on low wages as in the Mexican case, an increase in structural unemployment and in industrial heterogeneity as it happened in the majority of countries, and a high fragile external finance position. As we considered historically the balance of payment constraint through its influences on the rate of exchange, rate of interest, on employment level and on structural heterogeneity was the principal macroeconomic force that shaped income distribution.After the balance of payment liberalization a renewed external dependency is shaping a high concentrated income distribution.However some new aspects seen to be assuming great importance. So far the impacts of high interest rate on income distribution was considered through its effects on wage share.This ultimate effect depends on how much the financial costs are passed through prices.As we have seen in a regime of fixed or semi fixed exchange rate the new financial pressures can not be translated to prices in tradable sectors resulting in a compression of margins and unemployment.In a float system these pressures are more easily passed through on prices with a stronger effect on labor share. But as was put by UNCTAD 34 the immediate short term effect of a rise in interest rate is a redistribution of property income from profits to interest earners.This movement has to be seen in a economic circumstance where large public debts and extraordinary 34 "The rise in interest rates has also been a key factor in the increase in interest payments as a proportion of value added in the corporate sector.....Thus, the rise in interest rates and the share of interest in national income were initially reflected in a redistribution of property income from profits to rentiers....However, mark-ups in trade and industry typically respond to sustained increases in interest rates in the same way as they respond to changes in other costs.Since, under financial liberalization, interest rates adjust rapidly to changes in the price level, mark-up pricing implies that the greater interest burden tends to be shifted onto labour."(147) capital gains that spread in stock market in most countries created a new class of rentiers 35 .Since the limits of industrial and finance capital were blurred by financial deregulation the expansion of this new rentier class encompass the main interest of big business and wealthy people.As we observed in the last section, this dimension of financial liberalization on income distribution is strongly underreported in standard household surveys in LAC based essentially on labor incomes36 . Income distribution is strongly influenced by the relative political and economic power of different groups and social classes.The fragility of public sector and the enrichment of a class of bond holder rentiers (indexed to the rate of interest or to exchange rate) has been a remarkable feature of the last decade above all in a country like Brazil.As a matter of fact this is not a new situation as we have seen in the South Cone in the beginning of eighties.But it is a new reality in Brazil where so far the developmental state was very strong.The highly subsided process of privatization that took place in the continent and the internalization of large business groups in a deregulated financial market has enlarged the edge of private sector against the public sector.A depressive macroeconomics aimed to protect the private wealthy is the counterpart of his economic transformation. In relation to the distribution of labor income some new process are in action.The high interest rate affected firms and sectors differentially.In large export oriented companies with strong links to foreign capital the access to a lower interest rate allow them to obtain a rent not available to the remaining companies increasing the income differences among companies.In addition, the process of privatization and internationalization of banks, business services and public utilities has created a strong differentiation in income and techniques amplifying the role played by rents captured by big firms.When we consider the low increase in average and minimum wages the differential rents captured by big business can not be exaggerated.But a new dimension, barely explored in the literature, comes from the effects of business income concentration on wages differentiation.As far as the labor market became more flexible and deregulate, a rent sharing in privatized and other winners sectors favoring some lucky professionals seems to be in place increasing the wage dispersion37 . 35 "For a number o reasons, financial liberalization in many middle-income countries gave rise to a massive expansion of public and private debt.....a number of governments in highly indebted countries, particularly in Latin America, faced net financial transfers abroad as net new borrowing fell short of interests payments.Consequently, they had to borrow at very high rates at home in order to service their external debt. The increased public and private debt has witnessed the emergence of a new class of rentiers whose incomes depend as much on capital gains on financial assets as on interest payments.....The attractive terms which have been offered on public debt in order to shift government financing from the central bank to the private sector has meant that low risk, often tax free, government bonds constitute an increasing proportion of total investments of such rentiers.This tendency has gone so far that in some middle-income countries the corporate sector has become a surplus sector, lending directly and indirectly to the public sector."(146). This new process has emerged in economies where the public employment and the influence of unions were important for the middle employment jobs.Somehow the contraction of these jobs reflects this structural change This movement has not to do with a general tendency of wage differentiation based on skill established in literature and observed empirically in many countries.The average wages on skill (proxies by years of schooling) hides a differentiation that depends on the wages paid to special professionals in different sectors and firms.The income differentiation in the high income wage earners has been great and strongly underreported.As we have seen in the case of Brazil, the average wage of graduate professional in comparison with workers with 7-9 years did not improve in the nineties -the stagnation in the wages paid in public sector can respond partially for this performance-but the dispersion of wages between firms and sectors has increased.As in the case of rentier's income, this kind of wage differentiation is barely accessed in national data based on household informed wages. The connections between functional and personal income distribution are not usually explored in the studies on income distribution.Its reasonable to suppose two links.If we consider that the non labor income is far more concentrated than labor income, and that there is a public spending transference based on wage contract, the decline in wage share leads directly and indirectly to a higher personal income concentration.This second link is based in the fact that in Latin America countries the taxes on profits, capital gains and finance rents is not effective as it happens with income wages.Despite this importance, this result is, however, very difficult to be tested because the non wage income is usually underreported. The heterogeneity of techniques and productivity and consequently the dispersion of rents was an important dimension for primary income distribution in real terms in Latin American countries.While different income social groups have different patterns of consumption changes in relative prices alter, as above examined, income distribution in real terms .The low level of productivity in food production -a dominant share of wage goods-was and still is an essential cause for the high incidence of poverty and low real wages.But when we consider the basic needs in housing, infra-structure and collective goods we have new dimensions.The mere income distribution that arisen from an increase in wages or in pensions or other nominal income transference does not translates in a fully access to these services.Their demand can not be supplied by market forces because they depend on public investments in infraestructure. The Latin American countries ceased to invest in infra-structure -save for Chile -since the last decade; the boom in privatization and FDI in energy, telecommunications, roads, has not so far, except for a massive diffusion of a new product such as the cell phone, provided a mass diffusion of public facilities for the majority of people.On the contrary, a huge exclusion from modern and old services characterize the normal life of common people in Latin America.Here we have the old and modern dimensions of structural heterogeneity of internal markets whose reduction is today as it was in the past a necessary base for a less polarized economic growth.1995=100) 19801995=100) 19901995=100) 19921995=100) 19931995=100) 19941995=100) 19951995=100) 19961995=100) 19971995=100) 19981995=100) 19991995=100) 20001995=100) 2001 Argentina (a) Argentina (a) TABLE 1 : Real Interest Rate and Export Growth in World Economy TABLE 3 : Latin America and Caribbean: Principal Economic Indicators Annual growth rate , per cent , * in millions of dollars * TABLE 4 : Latin America: Real Exchange Rate* (1995=100) ECLAC based on IMF data.*corresponds to the average real rate of exchange of each country in relation to the currencies of its main trade partners weighted by their volume of imports according to the average observed in1995-1998. TABLE 5 : Oil and Non-Oil Commodity Prices, 1981-1998(annual percentage change in nominal U.S. dollars, World Bank commodity price indexes) TABLE 7 : Latin America: Real Interest Rates ECLAC 2002, based on IMF data.Nominal rates on short-term loans adjusted by consumer prices TABLE 8 : Latin America: Income Distribution in Urban Households, by Quintile TABLE 9 : Latin America: Average (a) and Minimum (m) Real Wages ( Argentina: manufacturer average wage; Brazil: formal labor; Chile: until 1993 only non agricultural labor; Colombia: manufacturer average wage; Mexico: manufacturer average wage; Paraguay: only capital; Venezuela: urban workers TABLE 10 : Labor Costs and International Competitivity Source: Tokman and Klein, 2000.Data related to manufacture industry in 1997.(1) Variation in local currency in constant prices deflated by consumer index.(2) Difference between productivity and labor cost	2018-12-27T04:41:12.511Z	2005-12-01T00:00:00.000	{ "year": 2005, "sha1": "14aff07734b9f97c74499babc2df31693069fd50", "oa_license": "CCBY", "oa_url": "https://www.scielo.br/j/rec/a/967vqNSB4rpMJjWb6DhdKmt/?format=pdf&lang=pt", "oa_status": "GOLD", "pdf_src": "ScienceParseMerged", "pdf_hash": "14aff07734b9f97c74499babc2df31693069fd50", "s2fieldsofstudy": [ "Economics" ], "extfieldsofstudy": [ "Economics" ] }
252298701	pes2o/s2orc	v3-fos-license	Dietary Supplementation with Eucommia ulmoides Leaf Extract Improved the Intestinal Antioxidant Capacity, Immune Response, and Disease Resistance against Streptococcus agalactiae in Genetically Improved Farmed Tilapia (GIFT; Oreochromis niloticus) A 7-week rearing trial was designed to investigate the effects of Eucommia ulmoides leaf extract (ELE) on growth performance, body composition, antioxidant capacity, immune response, and disease susceptibility of diet-fed GIFT. The results showed that dietary ELE did not affect growth performance or whole-body composition (p > 0.05). Compared with the control group, plasma ALB contents increased in the 0.06% dietary ELE group (p < 0.05), and plasma ALT and AST activities decreased in the 0.08% dietary ELE group (p < 0.05). In terms of antioxidants, compared with GIFT fed the control diet, 0.06% dietary ELE upregulated the mRNA expression levels of Nrf2 pathway-related antioxidant genes, including CAT and SOD (p < 0.05), and 0.06% and 0.08% dietary ELE upregulated the mRNA levels of Hsp70 (p < 0.05). In terms of immunity, 0.06% dietary ELE suppressed intestinal TLR2, MyD88, and NF-κB mRNA levels (p < 0.05). Moreover, the mRNA levels of the anti-inflammatory cytokines TGF-β and IL-10 were upregulated by supplementation with 0.04% and 0.06% dietary ELE (p < 0.05). In terms of apoptosis, 0.06% and 0.08% ELE significantly downregulated the expression levels of FADD mRNA (p < 0.05). Finally, the challenge experiment with S. agalactiae showed that 0.06% dietary ELE could inhibit bacterial infection, and significantly improve the survival rate of GIFT (p < 0.05). This study demonstrated that the supplementation of 0.04–0.06% ELE in diet could promote intestinal antioxidant capacity, enhance the immune response and ultimately improve the disease resistance of GIFT against Streptococcus agalactiae. Introduction Tilapia (Oreochromis niloticus) is the most exported farmed fish in China, with a total production of 1.65 million tons farmed in 2020 [1]. Nevertheless, in recent years, outbreaks of streptococcal disease have caused significant losses in the tilapia industry. Currently, Streptococcus is an important pathogenic bacterium that affects tilapia, and the disease is likely to occur when the water temperature is above 31 • C [2]. Usually, tilapia is more resistant to diseases; however, the seasonal high temperatures in summer are probably the cause of low immunity in tilapia and the increase in the susceptibility of these fish to pathogenic bacteria, resulting in a significant increase in mortality due to streptococcal infection [3,4]. Thus, to improve the ability of tilapia to fight bacterial infection, additives (such as Bacillus pumilus and white button mushrooms) have been applied to enhance immunocompetence and antioxidative status [5,6]. The use of feed additives to improve the immunity and disease resistance of tilapia has become a major trend worthy of further research to provide a reference for the tilapia industry. Eucommia ulmoides is an endemic plant species in China. The leaves and bark of E. ulmoides can be used as a growth promoter for animals, with growth-supporting, lactation, and immune-enhancing effects [7]. As a byproduct of the traditional Chinese herb E. ulmoides, E. ulmoides leaves are quite common in China [8] and show higher antioxidative activity than the cortex, fruits, and flowers [9]. In recent years, experts have studied the composition and efficacy of E. ulmoides leaves, they concluded that Eucommia ulmoides leaf extract (ELE) as a feed additive showed no drug resistance and almost no toxic side effects [10,11], indicating that ELE is a very valuable feed additive for development. Furthermore, ELE is rich in bioactive compounds (e.g., flavonoids, chlorogenic acid, peach leaf coralline, kynurenine) with anti-inflammatory, antioxidant, antiviral, and hepatoprotective properties [12]. In recent years, several researchers have reported that ELE increases the body weight of weaned piglets [13] and broiler chickens [14] and increases the feed intake of piglets [15]. Moreover, in a study on lambs, it was found that the addition of extracts from Eucommia ulmoides leaves in the diet did not affect their average daily weight gain or feed efficiency [16], which may be related to the amount of ELE added and the particular animal species. In aquatic animals, Huang et al. [17] found that 1.0% dietary ELE supplementation could improve the growth performance of large yellow croaker (Larimichthys crocea) and enhance antioxidant capacity and immunity. Zhang et al. [18] studied channel catfish (Ictalurus punctatus) and reported that 0.2% dietary ELE supplementation improved the intestinal microbiota structure and reduced the incidence of disease. The above studies show that ELE is increasingly being used in aquaculture, but its use as an additive for fish is still relatively rare, and the immunomodulatory regulatory mechanisms of ELE deserve further study. Toll-like receptor (TLR) family-mediated innate immunity is the first line of defense against disease [19]. TLRs are the primary receptors for the recognition of pathogenassociated molecular patterns (PAMPs) by the innate immune system that initiate the signaling pathways that regulate the adaptive immune response [20]. In addition, the TLRs can bind myeloid differentiation factor 8 (MyD88) to activate nuclear factor kappa-B (NF-κB) and apoptotic signaling pathways [21,22]. In aquatic animals, many studies have indicated that TLR2 plays a critical role in the innate immune response [23,24]. However, no information regarding the effects of ELE on the immune response and apoptosis related to the TLR2-MyD88 pathway in tilapia has been reported. Additionally, the nuclear factor erythroid 2-related factor 2 (Nrf2) signaling pathway plays a critical role in the resistance to exogenous or endogenous oxidative stress [25]. Likewise, the mechanism by which ELE regulates the antioxidant status of tilapia via the Nrf2 signaling pathway deserves investigation. In this study, the genetically improved farmed tilapia (GIFT), one of the tilapia strains, was chosen as the subject of this experiment. GIFT was developed in response to the growing demand for superior growth rates and increased resistance to emerging diseases among fish in aquaculture through international efforts [26]. The GIFT is now reportedly being cultured in about 87 countries around the world [1] and is one of the most popular aquaculture species in China. Thus, the objectives of our study were to examine the effects of ELE on growth, antioxidant capacity associated with the Nrf2 signaling pathway, immune response and apoptosis induction associated with the TLR2-MyD88 signaling pathway, and the disease resistance of GIFT against Streptococcus agalactiae. Table 1 shows the ingredients and proximate composition of the experimental diets. ELE was purchased from HANOVE Animal Health Products Co., Ltd., Wuxi, China. According to the recommended dosage (0.03-0.06%) of this product in omnivorous fish feed, the ELE was supplemented in the diet at five levels (0% (control), 0.02%, 0.04%, 0.06%, and 0.08%). All of the ingredients used in this experiment were crushed and passed through a 60-mesh sieve, made into pellets (the grain diameter is 1.0 mm), and then dried in an oven at 45 • C for 24 h. The specific steps and instruments used were described in our previous report [27]. After drying, the pellets were put into self-sealing bags and stored at −20 • C until further use. Experimental Fish and Procedures GIFT juveniles were provided by the breeding farm of the Freshwater Fisheries Research Center (FFRC) of the Chinese Academy of Fishery Sciences (Wuxi, China). Before the experiment, all fish were temporarily reared in floating cages for two weeks to adapt to the experimental environment. Afterward, 300 healthy fish (initial body weight was 12.04 ± 0.03 g) were randomly assigned to 15 floating cages (1 m × 1 m × 1 m) (20 fish per cage). Each diet consisted of three replicates. The experiment lasted for 7 weeks, during which the fish were fed twice a day, each time to apparent satiety. Additionally, the water quality indicators were recorded daily (YSI ProDSS Multiparameter Water Quality Meter, Ohio, USA), the water temperature was maintained between 31 and 33 • C, the amount of dissolved oxygen was higher than 6 mg/L, and the pH was kept at 7.0-7.5. Sample Collection After 7 weeks, the experimental fish were fasted for 24 h, after which the number of GIFT per cage was counted, and all fish were weighed. Three fish were randomly taken from each cage. First, blood was drawn from the caudal vein and immediately centrifuged for 10 min (3000 rpm, 4 • C). Then, upper plasma samples were obtained and stored in a −20 • C freezer for plasma biochemical analysis. Intestinal samples were collected by dissection. A portion of the intestinal tissue was stored in 4% paraformaldehyde for Antioxidants 2022, 11, 1800 4 of 17 pathological analysis, and the remaining intestinal samples were stored in a −80 • C freezer for gene and enzymatic activity analysis. Proximate Composition and Chemical Analysis The experimental diets and whole-body composition were analyzed based on the method of AOAC [28]. Plasma total protein (TP), albumin (ALB), alanine aminotransferase (ALT), and aspartate aminotransferase (AST) were determined with an automatic biochemical analyzer. The intestinal activities of antioxidant factors (malondialdehyde (MDA), superoxide dismutase (SOD), catalase (CAT), glutathione (GSH), and glutathione peroxidase (GPx)) were analyzed with the corresponding reagent kits. The major kits, testing equipment, and main methods are presented in Table 2. Histology Hematoxylin and eosin (HE) staining was used to analyze the intestinal histology. First, the intestinal tissue samples were extracted from 4% paraformaldehyde. Then, intact wax blocks were obtained by gradient alcohol dehydration and embedding, followed by serial sectioning with a microtome (Leica Company, Wetzlar, Germany), HE staining and dehydration sealing. Finally, a Zeiss microscope (Axioplan-2, Oberkochen, Germany) was used to observe the intestinal pathological changes, and photographs were collected for analysis. Real-Time PCR Analysis First, the TRIzol method (Vazyme Biotech Co., Ltd., Nanjing, China) was used to extract total RNA from the intestinal tissues. Then, the quality and quantity of the RNA were checked with a NanoDrop 2000 spectrophotometer. Finally, the reaction system was set up according to the instructions of the HiScript ® II One Step qRT-PCR SYBR Green Kit (Q221-01, Vazyme, Nanjing, China) and performed on a CFX96 real-time PCR detection system thermocycler (Bio-Rad). The specific primers for the reference gene (β-actin) and target genes in this experiment are displayed in Table 3. The mRNA expression levels were calculated from the standard curve, normalized to β-actin, and quantified using the relative standard curve method. Streptococcus Agalactiae Challenge Test Ten fish from each cage were challenged with Streptococcus agalactiae (S. agalactiae) in indoor recirculating culture barrels with a controlled water temperature at 32 ± 1 • C, the pH value ranged from 7.6 ± 0.2, and dissolved oxygen levels were maintained at 6-7 mg/L. Before the challenge, a pre-experiment was performed to determine the halflethal concentration (1 × 10 6 CFU/mL) of S. agalactiae using a bacterial turbidimeter (SGZ-6AXJ, Yue Feng Instrument Co., Ltd., Shanghai, China). The specific method is described in our previous study [29]. Then, the fish were challenged by intraperitoneal injection with 1 mL/100 g (1% of body weight). The mortality rate within 144 h was recorded. Statistical Analysis The data were subjected to normality and homogeneity tests. Then, the experimental data (means ± SEMs) were analyzed using SPSS 24.0 statistical software for one-way analysis of variance (ANOVA). When the difference was significant (p < 0.05), Duncan's multiple comparisons tests were performed. Furthermore, orthogonal polynomial contrasts were used to assess the significance of linear or quadratic models to describe the response of the dependent variable to dietary ELE levels. p-values < 0.05 were considered statistically significant. Table 4 shows the GIFT growth performance results. The FBW, FCR, WGR, SGR, and SR were not influenced by dietary ELE levels (p > 0.05). Table 5 presents the whole-body composition of the GIFT, and no significant effect of ELE supplementation was found on the moisture, protein, lipid, and ash content in all diets (p > 0.05). weight (g)-initial body weight (g)). d Weight gain rate (WGR) (%) = 100 × (final body weight (g-initial body weight (g))/initial body weight (g). e Specific growth rate (SGR) (% day −1 ) = 100 × [(In (final body weight (g))-In (initial body weight (g)))/days]. f Survival rate (SR) (%) = 100 × (survival fish number/total fish number). Plasma Parameters The results of the plasma parameter assessment of the GIFT fed different diets are presented in Table 6. Plasma ALB had a positive linear with increasing dietary ELE inclusion levels (p < 0.05). At a dietary ELE level of 0.06%, plasma ALB showed the highest level (p < 0.05). In addition, both plasma ALT and AST had negative linear responses with increasing dietary ELE inclusion levels (p < 0.05), and plasma ALT activity of the fish fed 0.08% dietary ELE were lower than those fed the control diet (p < 0.05). The plasma AST activities of the fish fed 0.06% and 0.08% dietary ELE were lower than those fed the control diet (p < 0.05). Plasma TP levels were not influenced by dietary ELE levels (p > 0.05). Table 7 shows the results of intestinal antioxidant enzyme activities of the GIFT fed different diets. The CAT and SOD had an open upward parabola with increasing dietary ELE inclusion levels (p < 0.05). The highest CAT and SOD activity was observed in the 0.04% and 0.06% ELE groups, respectively, which were notably higher than those in the group administered the control diet (p < 0.05). The GSH-Px had a positive linear response with increasing dietary ELE inclusion levels (p < 0.05), and at a dietary ELE level of 0.06%, GSH-Px showed the highest activity (p < 0.05). In addition, the GSH had an open upward parabola with increasing dietary ELE inclusion levels (p < 0.05), but no significant differences in GSH content were observed among all groups (p > 0.05). Furthermore, dietary ELE levels did not affect GSH and MDA contents (p > 0.05). Figure 1 shows photomicrographs of intestinal sections, and the data on the number of goblet cells and villus length are shown in Table 8. When the ELE inclusion level was 0.06%, the number of goblet cells was significantly larger than that in the control group (p < 0.05). In addition, no significant effect of ELE supplementation was found on villus length compared with the control group (p > 0.05). Figure 1 shows photomicrographs of intestinal sections, and the data on the number of goblet cells and villus length are shown in Table 8. When the ELE inclusion level was 0.06%, the number of goblet cells was significantly larger than that in the control group (p < 0.05). In addition, no significant effect of ELE supplementation was found on villus length compared with the control group (p > 0.05). Figure 2 shows the results of the relative expression of the Nrf2 pathway and Hsp70. The Nrf2, CAT, and SOD had an open upward parabola with increasing dietary ELE inclusion levels (p < 0.05). Moreover, the mRNA levels of Nrf2 in the 0.04% and 0.06% dietary ELE groups were higher than those in the control group (p < 0.05, Figure 2A). The CAT mRNA expression level in the fish fed 0.04% ELE was significantly higher than that in fish fed the control diet (p < 0.05, Figure 2C), and the SOD mRNA expression level was markedly upregulated in the fish fed 0.06% ELE (p < 0.05, Figure 2D). No notable changes were observed in Keap1 mRNA levels among all treatment groups (p > 0.05, Figure 2B). In addition, Hsp70 levels had a positive linear relationship (p < 0.05) with increasing dietary ELE inclusion levels and were higher in the 0.06% and 0.08% ELE diets than the control diet (p < 0.05, Figure 2E). mRNA expression level in the fish fed 0.04% ELE was significantly higher than that in fish fed the control diet (p < 0.05, Figure 2C), and the SOD mRNA expression level was markedly upregulated in the fish fed 0.06% ELE (p < 0.05, Figure 2D). No notable changes were observed in Keap1 mRNA levels among all treatment groups (p > 0.05, Figure 2B). In addition, Hsp70 levels had a positive linear relationship (p < 0.05) with increasing dietary ELE inclusion levels and were higher in the 0.06% and 0.08% ELE diets than the control diet (p < 0.05, Figure 2E). Figure 3A shows that no significant differences were observed in the relative expression levels of TLR2 mRNA between the dietary ELE supplementation groups and the control group (p > 0.05). The TLR2 mRNA levels in the 0.04% and 0.06% dietary ELE groups were remarkably lower than that in the 0.08% dietary ELE group (p < 0.05). Furthermore, the MyD88 had a negative linear response with increasing dietary ELE inclusion levels (p < 0.05). Compared with the control group, 0.06% and 0.08% dietary ELE levels significantly decreased the MyD88 mRNA expression levels (p < 0.05, Figure 3B). Figure 3A shows that no significant differences were observed in the relative expression levels of TLR2 mRNA between the dietary ELE supplementation groups and the control group (p > 0.05). The TLR2 mRNA levels in the 0.04% and 0.06% dietary ELE groups were remarkably lower than that in the 0.08% dietary ELE group (p < 0.05). Furthermore, the MyD88 had a negative linear response with increasing dietary ELE inclusion levels (p < 0.05). Compared with the control group, 0.06% and 0.08% dietary ELE levels significantly decreased the MyD88 mRNA expression levels (p < 0.05, Figure 3B). Relative Expression of the Genes in the NF-κB Signaling Pathway The NF-κB expression levels had a negative linear response with increasing dietary ELE inclusion levels (p < 0.05). Compared with the control group, the NF-κB mRNA expression level was remarkably downregulated with 0.06% dietary ELE (p < 0.05, Figure 4A). Conversely, the TGF-β expression levels had a positive linear response with increasing dietary ELE inclusion levels (p < 0.05), and the mRNA expression levels of TGF-β were higher in the 0.06% dietary ELE group than in the control group (p < 0.05). Similarly, the IL-10 expression levels had a positive linear response with increasing dietary ELE inclusion levels (p < 0.05), the mRNA expression levels of IL-10 increased with increasing dietary ELE from 0% to 0.06%, and the highest levels of both were found in the 0.06% dietary ELE group (p < 0.05, Figure 4B,C). In addition, no clear changes were found in the expression levels of the proinflammatory factors TNF-α and IL-8 among all dietary treatments (p > 0.05, Figure 4D,E). Figure 3A shows that no significant differences were observed in the relative expression levels of TLR2 mRNA between the dietary ELE supplementation groups and the control group (p > 0.05). The TLR2 mRNA levels in the 0.04% and 0.06% dietary ELE groups were remarkably lower than that in the 0.08% dietary ELE group (p < 0.05). Furthermore, the MyD88 had a negative linear response with increasing dietary ELE inclusion levels (p < 0.05). Compared with the control group, 0.06% and 0.08% dietary ELE levels significantly decreased the MyD88 mRNA expression levels (p < 0.05, Figure 3B). Relative Expression of the Genes in the NF-κB Signaling Pathway The NF-κB expression levels had a negative linear response with increasing dietary ELE inclusion levels (p < 0.05). Compared with the control group, the NF-κB mRNA expression level was remarkably downregulated with 0.06% dietary ELE (p < 0.05, Figure 4A). Conversely, the TGF-β expression levels had a positive linear response with increasing dietary ELE inclusion levels (p < 0.05), and the mRNA expression levels of TGF-β were higher in the 0.06% dietary ELE group than in the control group (p < 0.05). Similarly, the IL-10 expression levels had a positive linear response with increasing dietary ELE inclusion levels (p < 0.05), the mRNA expression levels of IL-10 increased with increasing dietary ELE from 0% to 0.06%, and the highest levels of both were found in the 0.06% dietary ELE group (p < 0.05, Figure 4B,C). In addition, no clear changes were found in the expression levels of the proinflammatory factors TNF-α and IL-8 among all dietary treatments (p > 0.05, Figure 4D,E). Relative Expression of the Genes in the Apoptosis Signaling Pathway The FADD expression levels had a negative linear response with increasing dietary ELE inclusion levels (p < 0.05), and at dietary ELE levels of 0.06% and 0.08%, the relative expression of FADD mRNA in the intestine was markedly lower than that in the control Relative Expression of the Genes in the Apoptosis Signaling Pathway The FADD expression levels had a negative linear response with increasing dietary ELE inclusion levels (p < 0.05), and at dietary ELE levels of 0.06% and 0.08%, the relative expression of FADD mRNA in the intestine was markedly lower than that in the control group (p < 0.05, Figure 5A). In addition, the expression levels of Caspase8, Bcl2, Bcl-xl, and AP-1 were not influenced (p > 0.05, Figure 5B-E). Figure 6 shows the mortality rate of the GIFT fed with different dietary ELE levels with the Streptococcus agalactiae challenge after 144 h. The mortality rate had a negative linear response with increasing dietary ELE inclusion levels (p < 0.05), and the lowest mortality rate of GIFT was observed in the fish fed 0.06% ELE (p < 0.05). Figure 6 shows the mortality rate of the GIFT fed with different dietary ELE levels with the Streptococcus agalactiae challenge after 144 h. The mortality rate had a negative linear response with increasing dietary ELE inclusion levels (p < 0.05), and the lowest mortality rate of GIFT was observed in the fish fed 0.06% ELE (p < 0.05). Effects of ELE Supplementation on Growth Performance and Whole-Body Composition In recent years, studies on aquatic animals have confirmed that ELE can promote growth performance, such as in grass carp (Ctenopharyngodon idella) [30], turbot (Scophthalmus maximus L.) [31], and large yellow croaker (Larimichthys crocea) [17]. However, our current results showed that dietary ELE supplementation did not improve the growth performance of GIFT. The differences in fish species and cultural environment could cause a different outcome. Since there are still relatively few studies on ELE in fish compared with mammals, more studies are needed to probe the mechanism of the effects of ELE on growth performance in aquatic animals. Furthermore, our current study showed that dietary ELE supplementation did not affect body composition, which is consistent with the findings in large yellow croaker [17]. Effects of ELE Supplementation on Intestinal Morphology Intestinal morphology has a direct link to intestinal development and health status [32]. The length of the intestinal villus reflects the absorption of nutrients in the intestine, so the morphology of the intestinal villi directly reflects the growth and development of the body [33]. Our current results showed that dietary ELE supplementation did not significantly affect intestinal villus length compared with the control group, which indicated that the addition of ELE did not affect nutrient absorption in the intestine or negatively affect growth. Moreover, goblet cells maintain intestinal homeostasis by secreting mucus in the intestine to help the body absorb nutrients and defend against pathogens [34]. In this study, when the ELE level reached 0.06%, the number of goblet cells increased significantly, indicating that ELE can promote the proliferation of intestinal goblet cells to some extent. This indicates that appropriate ELE supplementation could maintain intestinal structural integrity and improve the immune barrier function of the intestine. Zhang et al. [18] proposed that ELE supplementation can improve intestinal villi structural disorders, which supports our findings. Effects of ELE Supplementation on Antioxidant Status Intestinal health is also closely related to intestinal antioxidant capacity [35,36]. The increase in the levels of relevant intestinal antioxidant enzymes and intestinal antioxidantrelated genes could reflect an improvement in intestinal health [37]. In our experiment, dietary ELE supplementation significantly activated the Nrf2 signaling pathway, which is Effects of ELE Supplementation on Growth Performance and Whole-Body Composition In recent years, studies on aquatic animals have confirmed that ELE can promote growth performance, such as in grass carp (Ctenopharyngodon idella) [30], turbot (Scophthalmus maximus L.) [31], and large yellow croaker (Larimichthys crocea) [17]. However, our current results showed that dietary ELE supplementation did not improve the growth performance of GIFT. The differences in fish species and cultural environment could cause a different outcome. Since there are still relatively few studies on ELE in fish compared with mammals, more studies are needed to probe the mechanism of the effects of ELE on growth performance in aquatic animals. Furthermore, our current study showed that dietary ELE supplementation did not affect body composition, which is consistent with the findings in large yellow croaker [17]. Effects of ELE Supplementation on Intestinal Morphology Intestinal morphology has a direct link to intestinal development and health status [32]. The length of the intestinal villus reflects the absorption of nutrients in the intestine, so the morphology of the intestinal villi directly reflects the growth and development of the body [33]. Our current results showed that dietary ELE supplementation did not significantly affect intestinal villus length compared with the control group, which indicated that the addition of ELE did not affect nutrient absorption in the intestine or negatively affect growth. Moreover, goblet cells maintain intestinal homeostasis by secreting mucus in the intestine to help the body absorb nutrients and defend against pathogens [34]. In this study, when the ELE level reached 0.06%, the number of goblet cells increased significantly, indicating that ELE can promote the proliferation of intestinal goblet cells to some extent. This indicates that appropriate ELE supplementation could maintain intestinal structural integrity and improve the immune barrier function of the intestine. Zhang et al. [18] proposed that ELE supplementation can improve intestinal villi structural disorders, which supports our findings. Effects of ELE Supplementation on Antioxidant Status Intestinal health is also closely related to intestinal antioxidant capacity [35,36]. The increase in the levels of relevant intestinal antioxidant enzymes and intestinal antioxidantrelated genes could reflect an improvement in intestinal health [37]. In our experiment, dietary ELE supplementation significantly activated the Nrf2 signaling pathway, which is the most important antioxidative stress defense mechanism in cells [38]. In this study, 0.04% and 0.06% dietary ELE significantly upregulated Nrf2 mRNA expression levels. Further-more, the downstream factors CAT and SOD were also affected by dietary ELE levels, and the highest CAT and SOD mRNA levels were present in the 0.04% and 0.06% dietary ELE groups, respectively. The results also indicated that dietary ELE supplementation could improve intestinal antioxidant capacity, which is supported by a study on channel catfish [18]. In addition, the activities of antioxidant enzymes in fish are positively correlated with the levels of their associated genes [39]. As found in this study, with the activation of antioxidant defense mechanisms, the highest CAT activity was found in the 0.04% dietary ELE group, and the highest SOD and GSH-Px activities were both present in the 0.06% dietary ELE group. This further demonstrated the efficacy of ELE to enhance antioxidant capacity. The specific reason for this result may be due to the action of the main components of ELE (chlorogenic acid [7], E. ulmoides flavonoids [40], and E. ulmoides polysaccharides [41]), which have a scavenging effect on free radicals. However, the specific mechanism needs further study. In addition, heat shock proteins (HSPs), also known as stress proteins, are preferentially synthesized after stress, among which Hsp70 has important cellular functions, such as cytoprotective and antioxidant effects [42]. Many studies have pointed out that herbs can enhance the expression of Hsp70 in tilapia, a mixture of Chinese herbs and a commercial probiotic Bacillus species could improve the expression of Hsp70 after various stresses [43], and dietary blackberry syrup supplementation could improve the resistance of Nile tilapia to Plesiomonas shigelloides [44]. Likewise, in this experiment, the expression levels of Hsp70 mRNA were elevated with dietary ELE supplementation. It was further shown that appropriate dietary ELE supplementation (0.04-0.06%) could improve the antioxidant capacity of the body. Nevertheless, the highest (0.08%) or lowest (0.02%) levels did not significantly improve the antioxidant capacity. The reason may be that the effective active ingredients of ELE have a suitable range of action, and too high or too low levels may not play their proper role. Effects of ELE Supplementation on Immunocompetence As a member of the TLR family, TLR2 is involved in the induction of innate immune responses. [20]. MyD88 is an important junction protein for TLRs to mediate innate immune responses, which can activate NF-κB in downstream signaling pathways and ultimately cause inflammatory transmitters and the release of cytokines [45]. According to a previous report on Ussuri catfish (Pseudobagrus ussuriensis), downregulating the mRNA expression levels of proinflammatory cytokines via the TLR2-MyD88-NF-κB pathway could contribute to immune competence and disease resistance [46]. In the current study, appropriate dietary ELE supplementation (0.04-0.06%) reduced the relative gene expression of TLR2. It is, therefore, reasonable to assume that pathogen binding to the TLR2 protein is reduced, which in turn reduces the relative gene expression of TLR2 [47]. As the corresponding adaptor molecules of TLR2, the expression levels of MyD88 mRNA were inhibited with the addition of 0.06% ELE. In addition, the 0.06% dietary ELE group had the lowest level of NF-κB mRNA expression, indicating that appropriate dietary ELE supplementation might enhance GIFT immunity. Kim et al. [48] reported that Eucommia extract has high antiinflammatory activity and can inhibit NF-κB expression. Furthermore, NF-κB-regulated downstream cytokines are also involved in the regulation of the immune response [49]. The present study demonstrated that 0.04-0.06% dietary ELE enhanced the mRNA expression levels of the anti-inflammatory factors TGF-β and IL-10 in the GIFT intestine, while the pro-inflammatory factors TNF-α and IL-8 were not affected by dietary ELE levels. The elevation of anti-inflammatory gene transcripts suggested that ELE may have a significant anti-inflammatory effect, which is consistent with a previous study on channel catfish, which showed that ELE could reduce inflammation [18]. Nevertheless, our experimental result showed that a higher level (0.08%) did not tend to improve the immune response, as reported by Huang et al. [17] where in large yellow croaker ELE exerts a suppressive effect on immune competence at high doses. Moreover, considering the economic benefits, a higher addition level (0.08%) is not recommended for GIFT. From the above experimental results, it can be inferred that appropriate dietary ELE supplementation (0.04-0.06%) could enhance GIFT immunity by suppressing the expression of relevant inflammatory factors in the TLR2-MyD88-NF-κB pathway. In addition, plasma ALB, ALT, and AST are important nonspecific immune indicators in fish [50,51]. In our experiment, appropriate dietary ELE supplementation decreased plasma ALT and AST activities, which indicated that the hepatopancreas tissue is protected and that no significant amount of ALT and AST escapes from the cells into the blood [52]. In addition, 0.06-0.08% dietary ELE significantly increased the ALB content compared with the control diet, indicating that ELE can increase the plasma ALB content in tilapia, maintain blood osmolarity balance, promote the exchange of substances between blood and tissues, and thus improve nonspecific immunity. In addition, 0.04-0.08% dietary ELE showed increased tilapia survival rates after challenge with Streptococcus agalactiae. The result of the bacterial challenge test also supported our findings that ELE had positive effects on protecting tilapia from S. agalactiae infection. Effects of ELE Supplementation on Apoptosis The apoptotic signaling pathway is also activated by TLR2 [22]. TLR2 sends apoptotic signals through MyD88 in a pathway involving FADD and Caspase 8, and the binding of MyD88 to FADD is sufficient to induce apoptosis [53]. In this study, 0.06% and 0.08% dietary ELE significantly downregulated the expression levels of FADD mRNA, indicating that TLR2-mediated apoptosis was significantly inhibited by appropriate dietary ELE supplementation. In addition, TLR2 can regulate apoptosis through the NF-κB pathway [54], and NF-κB then acts on a variety of apoptosis-related target genes, including Bcl2, Bcl-xl, and AP-1. In this study, these apoptosis-related genes (Bcl2, Bcl-xl, and AP-1) were not affected by dietary ELE levels. This could be explained by the FADD-mediated apoptotic pathway being the main pathway of TLR2-mediated apoptosis [55] rather than the TLR2-NF-κB pathway. However, the regulatory mechanisms of TLR2-mediated apoptotic pathways remain complex and variable and need to be further explored. Conclusions In general, our current study showed that dietary ELE supplementation had no significant effect on the growth performance of GIFT. However, it was confirmed that supplementation with 0.04-0.06% ELE in the diet could promote intestinal antioxidant capacity by activating the Nrf2 signaling pathway, enhance the immune response by suppressing the TLR2-MyD88-NF-κB signaling pathway, and ultimately improve the disease resistance of GIFT against Streptococcus agalactiae (Figure 7).	2022-09-16T15:13:11.244Z	2022-09-01T00:00:00.000	{ "year": 2022, "sha1": "3642139d143a2fa31983e58fc84fc97fa63ea393", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2076-3921/11/9/1800/pdf?version=1663058450", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "8a8c61cc5b984953266f6774799463f22059bba5", "s2fieldsofstudy": [ "Agricultural And Food Sciences" ], "extfieldsofstudy": [ "Medicine" ] }
22913718	pes2o/s2orc	v3-fos-license	Tetra-μ-acetato-bis[(pyridine N-oxide)copper(II)](Cu—Cu) The molecule of the title binuclear copper(II) complex, [Cu2(CH3COO)4(C5H5NO)2], occupies a special position on a crystallographic inversion centre; the coordination environment of the CuII atom is slightly distorted square-pyramidal and is made up of four O atoms belonging to four acetate groups in the basal plane with the O atom of pyridine N-oxide ligand in the apical position. The Cu—Cu distance is 2.6376 (6) Å. The molecule of the title binuclear copper(II) complex, [Cu 2 (CH 3 COO) 4 (C 5 H 5 NO) 2 ], occupies a special position on a crystallographic inversion centre; the coordination environment of the Cu II atom is slightly distorted square-pyramidal and is made up of four O atoms belonging to four acetate groups in the basal plane with the O atom of pyridine N-oxide ligand in the apical position. The Cu-Cu distance is 2.6376 (6) Å . Tetra-µ-acetato-bis[(pyridine N-oxide)copper(II)](Cu-Cu) Yue Cui, Qian Gao, Chao-Yan Zhang and Ya-Bo Xie S1. Comment The crystal structures of binuclear copper(II) complexes have been extensively studied because of their possible anticarcinogen properties (Li et al., 2007), and numerous papers dealing with binuclear copper complexes have been published (Zhang, 2009). Herein, we report the synthesis and crystal structure of a new binuclear copper complex. The molecule of the title binuclear copper(II) complex, [Cu 2 (C 2 H 3 O 2 ) 4 (C 5 H 5 ON) 2 ], occupies a special position in the crystallographic inversion centre; coordination environment of the Cu II atom represents a slightly distorted tetragonal pyramid and is made up of four oxygen atoms belonging to four acetato-group in the basal plane as well as the oxygen atom of pyridine N-oxide ligand in the apical position. The Cu-O bond distance between Cu II atom and acetato O atoms vary from 1.9605 (18) Å to 1.9710 (18) Å, while the Cu-O bond distance involving Cu II atom and the O atom of the S2. Experimental A solution containing a 1:2:5 molar ratio of picolinic acid N-oxide (0.0139 g, 0.1 mmol), CuCO 3 (0.0247 g, 0.2 mmol) and acetic acid (1 ml, 0.5 mmol/ml) in a mixture of ethanol(5 ml) and water (10 ml) was sealed in a 25 ml teflon reactor and kept at 453 K for 3 days, then slowly cooled to 373 k and kept at this temperature for 24 h more. After cooling to room temperature, the mixture was filtered and the filtrate was allowed to stand at room temperature. Block crystals suitable for the X-ray investigation were collected. S3. Refinement All H atoms were placed geometrically (C-H = 0.93-0.96 Å) and included into refinement in the riding motion approximation with U iso (H) = 1.2U eq (C) [1.5U eq (C) for methyl H atoms]. Molecular structure of the title compound with displacement ellipsoids drawn at the 30% probability level; hydrogen atoms are shown as small circles of arbitrary radius. The unlabelled atoms are derived by the symmetry transformation -x + 1, -y, -z. Tetra-µ-acetato-bis[(pyridine N-oxide)copper(II)](Cu-Cu) Crystal data where P = (F o 2 + 2F c 2 )/3 (Δ/σ) max = 0.001 Δρ max = 0.43 e Å −3 Δρ min = −0.33 e Å −3 Special details Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger. (7) H9A-C9-H9B 109.5	2016-05-12T22:15:10.714Z	2009-07-01T00:00:00.000	{ "year": 2009, "sha1": "c5fea9b966789f5311db3617e5e6872a60ed22b6", "oa_license": "CCBY", "oa_url": "https://journals.iucr.org/e/issues/2009/08/00/ya2100/ya2100.pdf", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "0e9e5da46dd82e3b018da4e04873e1d61a2503ec", "s2fieldsofstudy": [ "Chemistry" ], "extfieldsofstudy": [ "Chemistry", "Medicine" ] }
116667204	pes2o/s2orc	v3-fos-license	Cooperative activities of forearm muscles under loading applied to thumb or each finger Human thumb and fingers are usually subjected to an external loading during daily activity. The information of how muscles in the forearm cooperate with each other in order to response to the external loading is still unknown. Such information may be helpful in understanding muscle function pathology and motor disorder. A novel method called electromyography computed tomography (EMG-CT) was developed to visualize muscle activity within a whole cross-section of the forearm by measuring surface EMG signals around the forearm. The current study aimed to extend the previous work by using the EMG-CT to investigate muscle cooperative activity under loading application to thumb or each finger. Loads of 0.98-9.8 N were applied to the thumb or each finger of four subjects in eight loading directions. The loading directions on thumb and index, middle, and little fingers were inner, outer, and upper directions. EMG signals around the subject’s forearm were recorded during the loading by using EMG band consisting of 40 pairs of bipolar electrodes. The results show different muscle cooperative activity pattern between loading conditions. During load was applied to thumb, muscle in lower region in pronation cross-section were highly active. When load was applied to a finger, muscles in lateral-lower region were highly active. In all subjects, total muscle activity in the whole cross-section and the maximum value of muscle activity increased in proportion to loading. This study demonstrates effectiveness of EMG-CT method by showing that the muscle cooperative activity of an individual is specific to force application conditions. Introduction Human fingers can move either independently or in synchrony with each other to achieve daily tasks such as grasping, gripping, and pinching. Loss of finger and hand function may be caused by neuromuscular diseases, such as Parkinson's and muscular dystrophy. An observation of muscle cooperative activities is important for developing rehabilitation processes or treatments for neuromuscular disorders. Researches attempted to understand muscle cooperative activities using cadaveric models (Garcia-Elias et al., 1991;Haugstvedt et al., 2001) and implanted force transducers in vivo (Schuind et al., 1992;Dennerlein, 2005). Garcia-Elias et al. (1991) used a stereophotogrammetric measurement system to study extensor mechanism of the fingers in human forearm specimens. Changes in length and orientation of different zones of the extensor mechanism at different finger configurations were reported. Haugstvedt et al. (2001) used human cadaveric upper extremity specimens to determine torque generated by the muscles rotating the forearm at various pronation/supination. The relationships between moment arm and angle of the flexor carpi ulnaris, extensor carpi ulnaris, supinator, biceps, pronator teres, and the pronator quadrates were reported. Studying from cadaveric model gains us insights of anatomical aspect of muscles, but the effects of physiological forces such as muscle contraction cannot be considered by this approach. Furthermore, Schuind et al. (1992) applied force transducers to the flexor pollicis longus and flexor digitorum superficialis, and profundus tendons of the index finger of patients. The tendon forces generated during passive and active motion of the wrist and fingers were recorded and reported. Dennerlein (2005) measured the in vivo tendon force of the flexor digitorum superficialis of the long finger during open carpal tunnel release surgery using force transducer. Forces were measured during isometric pinch and dynamic tapping of the finger. The results showed that tendon forces were a complicated function of fingertip force and motion. These studies provide insights of how forearm muscles generate forces during contraction. However, the measurements using force transducers were invasive and cannot provide the distribution of muscle activity within the whole cross-section in detail. The authors have proposed an electromyography computed tomography (EMG-CT) method that provides a visual image of muscle activity within a whole cross section of the forearm by measuring surface EMG signals using an EMG band comprising 40 pairs of bipolar surface electrodes (Nakajima et al., 2014). An EMG conduction model was formulated for reverse-estimation of muscle activities from measured EMG signals. By minimizing the difference between the estimated EMG values from the conduction model and the measured values, the muscle activities were estimated. The EMG-CT method provides a muscle activity distribution in the forearm in a noninvasive manner and will provide new information regarding the muscle cooperative activity within the forearm cross-sectional area. In previous study, a visual image of muscle activity within the forearm during loading applied to the proximal interphalangeal joint of the middle finger in the inner direction was obtained using EMG-CT method as the first trial (Nakajima et al., 2014). The method is expected to contribute to investigate the relationship between muscle cooperative activity pattern and the external loading response of human thumb and finger by visualizing a complex muscle activity within the forearm. This study aimed to investigate muscle activity patterns within a whole cross section of the forearm by EMG-CT under simple loadings in the inner, outer, and upper directions applied to the thumb and index, middle, and little fingers, in which the muscles in the forearm may work together in a complex manner. The EMG tomographic images reveal muscles cooperative activity within the forearm when different load conditions are applied to fingers. Method 2.1 Subjects Four healthy right-hander male subjects participated in the experiment and right forearm was examined. The subjects had no history of neuromuscular disorders of the forearm. Forearm length, circumference, and fat thickness for each subject were measured. Forearm length was defined as the length between the lateral epicondyle of the humerus and the styloid prominence. Circumferential length was measured at the midpoint of the forearm length. Fat and skin thickness were recorded as average values from the medial, lateral, anterior, and posterior midpoint of the forearm measured with a skinfold caliper. Table 1 shows age, height, weight, forearm length, circumferential length, and average fat and skin thickness of the subjects. Experimental procedure Each subject sat comfortably on a chair with his forearm placed on an adjustable stand, shoulder abducted and flexed at approximately 45°, elbow joint flexed at approximately 45°. Forearm and wrist held in pronation on the table, palm held with bars ( Fig. 1). The subject was requested to relax his fingers and maintain this position. Load was applied to the subject's finger with a weight-and-pulley system. Eight loading types were examined in the study as shown in Table 2. Inner, outer, and upper loadings were applied to the thumb at the distal phalanx ( Fig. 2A). The muscle activities to resist applied inner, outer, and upper loadings correspond to thumb extension, flexion, and abduction, respectively. Inner loadings were applied to the index, middle and little fingers at the proximal phalanx (Fig. 2B), corresponding to the muscle activities of finger extension. Outer loadings were applied to the middle finger at the distal and middle phalanges (Fig. 2B), corresponding to the muscle activities of finger flexion. The weights of these loadings were 0.1, 0.3, 0.5, and 1.0 kg. A loading was applied approximately 5 s three times with 5 s rest intervals. The subjects were instructed to maintain the posture during load applied. The procedures were approved by the Ethical Review Board for the Protection of Persons in Biomedical Research, Graduate School of Engineering, Hokkaido University, and all subjects signed an informed consent agreement. EMG setup and data acquisition EMG signals from the forearm were recorded with an EMG band, consisting of 40 pairs of bipolar surface electrodes (3-mm diameter disciform brass electrodes). Each four electrodes were placed on a custom-built electrode plate (Fig. 3). The inter-electrode distances of the bipolar electrodes were 15 and 45 mm. The middle point of the EMG band was positioned at the midpoint of the forearm length. The circumferential position of ulna was find by hand and a specific electrode plate was placed on the skin surface above ulna. The ground electrode was placed at lateral epicondyle. Before attachment of the electrode band, the subject's forearm skin was shaved with a razor and cleaned with an alcohol swab. To keep constant contact pressure of the electrodes to the skin surface in the measurements, a fabric elastic band was wrapped around the electrode band. All EMG signals were amplified by a factor of 367, sampled at 1000 Hz and filtered with a third-order 9-Hz Butterworth high-pass filter followed by a second-order 570-Hz Butterworth low-pass filter. The EMG signals were collected for 30 s per trial using a custom program (LabVIEW 8.5, National Instruments, TX, USA). The signals were filtered using a second-order 10-Hz Butterworth high-pass filter, a second-order 100-Hz Butterworth low-pass filter, and band-stop filters using a custom program (MATLAB, MathWorks, USA). The root mean square (RMS) value and mean power of each channel were calculated from the recorded signals in 500-ms windows. The average data were calculated from three trials for each loading condition. Reverse calculation of muscle activation An EMG conduction model for the reverse calculation of muscle activation was developed (Nakajima et al., 2014). The cross-sectional area of the subject's forearm was modeled as a circular region. The circumferential length of the circular region was set to the measured circumferential length. The muscle region was divided into elements for calculation (Fig. 4). The set of element nodes was distributed across the circular region with an element size of 1 mm for the surface region and 5 mm for the inside region. The number of elements in the model was 916 ± 45 (Table 1). During contraction, a muscle fiber depolarizes and generates current. Muscle action current was used to quantify levels of muscle activity. The strength of the muscle activity of activated virtual fiber k in element j (mk) (mA) was defined as in Eq. (1) (1) where (mV) is the mean square value of muscle action potential from muscle fiber k recorded by bipolar electrode i, di is the distance between the pair of bipolar electrodes i, V0(di) (mV/mA) is a transformation coefficient depending on di, lik (mm) is the distance between muscle fiber k and bipolar electrode i, l0 is the unit length (1 mm), and b(di) is the power exponent of the attenuation as a function of di (Nakajima et al., 2014). In the study, di was 15 and 45 mm. V0 and b were 162 mV/mA and -2.12 at 15 mm and 115 mV/mA and -1.74 at 45 mm, respectively (Nakajima et al., 2009(Nakajima et al., , 2014. The mean square value of muscle action potential from all muscle fiber k detected by bipolar electrode i, was described as (2) The muscle action potential from an element j is a superposition of the contributing action potentials from all the fibers within the element. When mk within the element j is assumed as a constant value of mj, can then be rewritten as (3) Figure 4 Representation of an electromyography conduction model used for calculations. To estimate mj, an objective function described as Eq. (4) was minimized, where was measured and was calculated from the EMG conduction model as in Eq. (3) at bipolar electrode i. (4) The total muscle activity S was defined as the summation of muscle activity mj weighted by the area of element aj within the forearm as in Eq. (5). Results The muscle activity in the forearm of subject A is presented in a tomographic image as a typical example. Figure 5 shows the strength of the muscle activation weighted by the area of each element within the cross-section under each loading. Figure 6 shows a typical pattern of muscle alignment in the middle part of the forearm obtained from a MR image. The tomographic images were shown in the pronation posture as in Fig. 6, by circumferentially arranging the surface EMG signals within the conduction model with reference to the position of the ulna in the calculations. In the study, thirteen muscles that were observed in the cross section were focused on: the extensor carpi ulnaris (ECU), extensor digiti minimi (EDM), extensor digitorum communis (EDC), extensor pollicis longus (EPL), abductor pollicis longus (APL), extensor carpi radialis longus (ECRL), extensor carpi radialis brevis (ECRB), flexor digitorum profundus (FDP), flexor pollicis longus (FPL), brachioradialis (BR), flexor carpi ulnaris (FCU), flexor digitorum superficialis (FDS), and palmaris longus (PL). Patterns of muscle cooperative activity could be observed under all investigated loadings. Differences of patterns with the loadings in an individual were clearly observed. The activated area and the maximum value of muscle activity increased with loads in all subjects. Examples of the muscle activity pattern of all subjects during 4.9 N loading are presented in Fig. 7. The relationships between external loading and total muscle activity in all subjects are presented in Fig. 8. The slopes varied with different subjects and applied loadings. In contrast, the coefficients of determination (R 2 ) was high for all subjects and loadings (R 2 = 0.93 ± 0.10), indicating a strong linear relationship between load application and muscle activation. Discussion In this study, EMG-CT was used to observe muscle cooperative activities in the forearm during loadings applied to thumb and each finger. The intensity level of each element was calculated from the detected EMG signals. EMG-CT method provided a visual image of muscle activity in the whole cross section of the forearm. When loadings were applied to the thumb in different directions, differences of patterns in an individual were clearly observed. For example, during inner loadings in subject A (Fig. 5A), the lateral-lower region was activated. When the load was increased, it could be observed that more muscle elements in the inner upper and medial regions were also activated as synergistic muscles to stabilize the joint. During outer loadings (Fig. 5B), the active regions were the medial and lower regions. More muscle elements in medial region were activated when the loading increased, although the activated area was smaller than that during inner loadings. During upper loadings (Fig 5C), the medial and lateral-lower regions were activated. When the loading increased, muscle elements in the lateral-lower region were more highly activated. Kaufman et al. (1999) measured EMG signals of FPL, EPL, and APL with needle electrodes during thumb extension/flexion and adduction/abduction, and reported that these muscles were activated during the motions. The activities of APL and EPL were large during thumb extension, whereas those of FPL were large during thumb flexion. In general, EPL and APL lie in the inner lateral-lower part and FPL in the central part of the measured cross section. The present results as shown in Figs. 5 and 6 agree with this trend, although it is difficult to recognize exactly the activities of specific muscles in our imaging. When an outer loading was applied to the distal phalanges of the thumb and middle finger, differences in patterns were also clearly observed, as shown in Figs 5B and 5F. In contrast, when an inner loading was applied to the proximal Tadano, Tonsho, Keeratihattayakorn, Yamada, Nakajima, Harada and Iwasaki, Journal of Biomechanical Science and Engineering, Vol.13, No.3 (2018) [DOI: 10.1299/jbse. phalanx of the index (Fig. 5D), middle (Fig. 5E), and little ( Fig. 5H) fingers, the patterns of muscle activity distribution appeared similar. Using needle electrodes, Darling et al. (1994), Maier and Hepp-Reymond (1995), and Valero-Cuevas et al. (1998) found that FDP, FDS, and EDC were activated during finger motions. FDP and FDS lie in the medial part and EDC on the surface of the lateral-lower part. The present results agree with the previously reported trend. In addition, when the point of force acts on the middle phalanx (Fig. 5G), the moment arm of the loading decreases compared with the case when the force is on the distal phalanx. This decrease was evident in the different total muscle activity, although the same muscle activation patterns were observed. Tadano, Tonsho, Keeratihattayakorn, Yamada, Nakajima, Harada and Iwasaki, Journal of Biomechanical Science and Engineering, Vol.13, No.3 (2018) [DOI: 10.1299/jbse.18-00065] , Tonsho, Keeratihattayakorn, Yamada, Nakajima, Harada and Iwasaki, Journal of Biomechanical Science and Engineering, Vol.13, No.3 (2018) [DOI: 10.1299/jbse.18-00065] , Tonsho, Keeratihattayakorn, Yamada, Nakajima, Harada and Iwasaki, Journal of Biomechanical Science and Engineering, Vol.13, No.3 (2018) [DOI: 10.1299/jbse.18-00065] Tadano, Tonsho, Keeratihattayakorn, Yamada, Nakajima, Harada and Iwasaki, Journal of Biomechanical Science and Engineering, Vol.13, No.3 (2018) [DOI: 10.1299/jbse.18-00065] During inner loading applied to little finger (Fig. 7C), all subjects seem to use similar muscle activity distribution pattern. This reflects the validity of the method. During outer loading applied to middle finger (Fig. 7B), subject B, C, and D use largely similar pattern. It is noted that subject A seems to use more muscles in lateral-lower part. During inner loading applied to thumb (Fig. 7A), different patterns were observed. Although main active muscles were in lateral-lower part, subject B seems to use more muscles in upper part and subject D use more muscles in medial part. Difference of muscle activity distribution pattern in thumb may be caused by the fact that many muscles were used to control thumb while other fingers used fewer muscles. Furthermore, there was a strong linear correlation between total muscle activity and applied loading (Fig. 8). The same trend was observed in every subject. When the loading increased, higher muscle forces were required to maintain static equilibrium. The increase in total muscle activity indicates that more muscles are activated to produce force. Differences in total muscle activity may have been caused by differences in coactivation under each loading. High total muscle activity may indicate high coactivation between muscles. This observation indicates that the strategy to activate muscles to resist loadings may vary. Some loading types, such as inner loading on the thumb and little finger ( Fig. 8A and H), require high coactivation levels. Further analysis of the relationship between force and muscle activity will allow the development of a model for estimating muscle force sharing in the forearm. However, there was also large variation in total muscle activity between subjects. This variation between subjects may have been associated with differences in individual muscle structures, including muscle size and muscle position. In this study, palm and forearm were held, but this condition did not guarantee that the subject did not use finger muscles during a thumb loading. During the thumb motions, the surface of the medial part, which may include FCU, was activated. This activation may have resulted from unintentional wrist movements by the subject during the experiments, whereas the other muscles may have operated synergistically to stabilize the joint. In other words, this method may be able to detect such effects in detail. Also, the loading types examined in this study were limited to the forearm in pronation posture, and a change in arm posture may affect muscle cooperative activities. The EMG conduction model did not consider the bone region within the forearm. Because this region does not generate muscle activity, the results may not have been affected by this omission. In further studies, determining how to detect the bone region in the cross section using the EMG band will be an important issue. In addition, the muscle activity was represented as mA dipole/s which passes normally through the cross-sectional area, although muscle fibers are not exactly aligned to the longitudinal direction of the forearm. The subjects in this study were young. It is known that age also affects muscle function. To gain more understanding regarding ageing and pathology in live human, a various range of the subject's age should be considered. This study measured activity only in extrinsic muscles within the forearm. It seems that during finger motion, intrinsic muscles also work during finger motion. Measuring intrinsic muscle activity will also improve our understanding of finger function. This method provides tomographic information of muscle activity within the forearm that cannot be assessed with a needle electrode. This facility will be very useful in clinical application. The impaired muscle region can be identified and properly treated. Furthermore, changes in load sharing between muscles within the forearm can be used to characterize pathological changes that affect muscle function. Muscle dysfunction caused by neuromuscular disease can also be observed by the method, enabling the development of more efficient treatments. In addition, the accessibility of deep muscle activity can provide a more delicate control of a prosthetic arm using the pattern of muscle activity from EMG-CT. In summary, the present study applied various loadings to the thumb and index, middle, and little fingers. The results provide effectiveness of EMG-CT method by showing that the muscle cooperative activity of an individual is specific to force application conditions.	2019-04-16T13:28:21.169Z	2018-01-01T00:00:00.000	{ "year": 2018, "sha1": "1c25542db8550ff3bfb5a5f8fbe7ec24cd8a160f", "oa_license": null, "oa_url": "https://www.jstage.jst.go.jp/article/jbse/13/3/13_18-00065/_pdf", "oa_status": "GOLD", "pdf_src": "Adhoc", "pdf_hash": "fdd6e3eea85312ddac03a660ba561ec72fff4bd2", "s2fieldsofstudy": [], "extfieldsofstudy": [ "Medicine" ] }
222003226	pes2o/s2orc	v3-fos-license	Reflections on the Use of SSCI Papers in Evaluating Social Sciences Research in Chinese Universities Purpose: This review demonstrates how to position Social Sciences Citation Index (SSCI) papers reasonably in order to promote the reform and development of the system for evaluating social sciences research (E-SSR) in China. Design/Approach/Methods: This review examines the contributions made by SSCI papers after such papers became a tool in the E-SSR system in Chinese universities, and the resultant issues. This review analyzes documents pertaining to the E-SSR systems of more than 50 world-class universities with consideration to the inherent characteristics and historical mission of social sciences research in China. The findings serve as the basis from which to examine the future trends in the reform of the E-SSR system in Chinese universities. Findings: The application of SSCI papers as an E-SSR tool is not common in world-class universities. To date, the reform of the E-SSR system in Chinese universities has involved: (i) establishing a pluralistic evaluation mechanism, with equal importance placed on SSCI papers and other research achievements; (ii) emphasizing the need for caution in using SSCI papers as an E-SSR tool and instituting distinct treatments for various disciplines; (iii) reducing the importance attached to journal language and ranking, while emphasizing innovation quality and practical contributions; and (iv) establishing China’s E-SSR standards in order to achieve an equilibrium between internationalization and localization. Originality/Value: This review argues that although SSCI papers constitute an indicator of E-SSR system, their importance must not be overstated. The main purpose of the E-SSR system is to facilitate the development of social sciences with a style and characteristics unique to China. The value of SSCI papers in China's E-SSR system In recent years, Chinese universities have used institutional regulations to guide and encourage the publication of SSCI papers by their faculty members. Universities also use the number of papers as a yardstick to evaluate the academic levels of faculty members and disciplines. In the social sciences, SSCI papers have become the basis of talent recruitment, evaluation of professional titles, and awarding of scientific research grants. Chinese universities have begun releasing the number of SSCI-indexed papers published by their various faculties and departments as well as the frequency with which these papers were cited. In China, the role of SSCI is not just a database for scientific researchers, but an important standard for the E-SSR system and related achievements (Dang, 2005). Moreover, many Chinese universities treat SSCI and CSSCI papers differently when awarding the publication of papers, such that "the award amount for an SSCI paper published in an international journal is usually several times, or even ten to twenty times, more than that for a local CSSCI paper" (Xu & Jiang, 2018, p. 49). This reflects the premium status of SSCI papers within the E-SSR system in Chinese universities. Impact of introducing SSCI papers into China's E-SSR system Although the SSR level of a country (or particular scientific research institution or scholar) cannot be accurately measured using the number of SSCI papers published, this indicator reflects its international status within a research field to a large extent (Kousha & Thelwall, 2007). As such, introducing SSCI into China's E-SSR system and encouraging faculty members to publish papers in international journals have contributed to the internationalization of SSR in the country. On the one hand, the international academic impact of Chinese social sciences has increased, facilitating the transformation of "silent China" into "China with a voice" (Liu, 2019, p. 111). The standardization of the country's academic research paradigm and improvement in the academic rationality of research have also contributed to this shift (Liu & Ding, 2014). On the other hand, the excessive reliance on SSCI papers has created several issues that cannot be ignored, including a lack of in-depth research into Chinese real social problems. As Liu and Ding (2014) note, "When selecting topics to write papers for publication in SSCI journals, scholars from China were more inclined towards popular issues that western countries were concerned about" (p. 88). According to Qin and Zhang (2008), a desire to be published in English journals and to attract the attention of foreign academic journals resulted in three phenomena: First, Chinese scholars began researching topics currently studied by Americans; second, they began studying local issues of interest to Americans; third, they began interpreting local issues through direct recourse to Western theories. Consequently, many local studies have become a database for American research or evidence of the applicability of Western theories in the Chinese context. As a result, Qin and Zhang (2008) note, "the local issues that really need to be studied and solved were frequently ignored" (p. 8). Therefore, although the publication of SSCI papers offers rich rewards for scientific research, research has failed to meet the country's social development needs. In this respect, emphasis on publishing SSCI papers has made "scholars mistakenly equate internationalization with a high quality level, thus weakening the role of scientific research at serving the country's development" (Liu & Ding, 2014, p. 91). Exaggerating the evaluative function of SSCI papers and placing this indicator at such a "supreme" position in the evaluation system have also led to a discrimination against the publication of papers in Chinese (Chou, 2014, p. 12). Discussion The world-class universities of other countries rarely apply SSCI papers as an E-SSR tool. A review of faculty member handbooks of 54 world-class universities, including Harvard University and Stanford University, reveals no mention of SSCI papers in respect to promotion, tenure, or performance assessment. In contrast, the corresponding guidelines for faculty members in the social sciences in Chinese institutions almost always include SSCI papers. Drawing on the experiences of other countries, as well as the developmental characteristics of the social sciences in China and the reality of its E-SSR system, this review proposes the following reform directions for China's E-SSR system. Establish a pluralistic evaluation mechanism, with equal importance placed on SSCI papers and other research achievements In world-class universities, both print and nonprint research achievements are considered for promotion, tenure, and performance assessment. In this respect, published papers are just one form of print achievements. For instance, Texas State University's College of Education considers 10 categories of print materials-including books, articles, conference proceedings, abstracts, reports, and book reviews-when evaluating promotions. Books comprise five subcategories: scholarly monographs, textbooks, edited books, chapters in books, and creative books. Articles published in both refereed and non-refereed journals are also included (Texas State University, 2018). In addition to SSCI papers, the status of other forms of achievement-including academic monographs, consulting reports, textbooks, and translations-should be enhanced in the E-SSR system of Chinese universities. Instead of giving supremacy to papers, especially SSCI papers, a pluralistic evaluation mechanism should be established. As Chen (2011) argues, if Chinese scholars wish to step out onto the world stage without being overbearing or self-effacing, they must first "improve their personal abilities and strive to improve the overall academic level. They should remain patient, work hard, spend a decade boosting their strength, and another decade heeding lessons from the past and training conscientiously. By the time a large number of works embracing an international perspective and local sentiments have accumulated, the goal of internationalizing Chinese academia will be achieved naturally" (p. 66). Apply SSCI papers as an E-SSR tool with caution and institute distinct treatments for various disciplines In their empirical study, Klein and Chiang (2004) confirm that SSCI cannot accurately reflect the quality of published results as well as the existence of obvious ideological biases. As a result of the uneven distribution of disciplines, the use of SSCI has produced numerous biases in overall E-SSR (Chu et al., 2003). The SSCI contains the lowest number of ethnological journals, but the highest number of psychological journals. Indeed, due to its inclination and preference, SSCI cannot provide fair evaluations for academic research in the fields of modern languages and literature or theories of socialism with Chinese characteristics (Zhuo, 2011). It is necessary to recognize that the SSCI's uneven distribution of disciplines prevent its application to all disciplines in the social sciences. Accordingly, SSCI should be used discriminately according to the characteristics and particularities of individual disciplines. Reduce the importance of language used and journal ranking, and emphasize innovation quality and actual contributions This article reviews the recruitment and promotional literature of 54 world-class universities, finding that their evaluation standards primarily emphasizes originality, high accomplishment level and quality, as well as potential. For example, the criteria for lifelong tenure in Harvard University's Faculty of Arts and Sciences include scholarly achievement and impact on the field, evidence of intellectual leadership and creative accomplishment, potential for future accomplishments, and potential contribution to the university and broader scholarly community (Harvard University, 2012). The evaluation standards focused on faculty members' level of achievements in scientific research and the influence of those achievements. Essentially, these criteria have nothing to do with the journals in which faculty members' papers were published or the language in which they were written. According to Li (2020), "The specific journal that published the paper, the language of the paper, the amount of attention garnered after publication and the ways the attention was given-all these have no bearing on the scientific value of knowledge innovation, and should not be misinterpreted as standards for evaluating scientific research" (p. 13). Therefore, China's E-SSR system should downplay the language used and journal ranking, and shift the actual evaluation foci to innovation quality and practical contribution. Establish China's standards of E-SSR with the aim of achieving an equilibrium between internationalization and localization In terms of the nature of the various disciplines, the natural sciences constitute a pure knowledge system that emphasizes value neutrality. In contrast, the social sciences comprise a broad duality: namely, a knowledge system and an ideology (Yang, 2017). In this sense, there are "borders" within the social sciences (Zhuo, 2011, p. 9). The famous sociologist Xiaotong Fei advocated cultural awareness in later life. This idea is particularly important for the development of social sciences in China. Summarizing the course of cultural awareness, Fei (1997) claimed that "Every form of beauty has its uniqueness; it is precious to appreciate other forms of beauty with openness. When beauty is represented by diversity and integrity, the world will be blessed with harmony and unity" (p. 22). In this respect, we need to identify the unique advantages and inherent characteristics of China's social sciences and develop these through self-reflection. Instead of blindly looking up to others, we should "step out" with our own unique stance. We must self-adapt to the global perspective, and engage in equal exchange and mutual growth with other cultures in creating a global civilization where "beauty is represented by diversity and integrity." China's E-SSR system should serve the purpose of "building social sciences with the characteristics and a style and imposing manner that are unique to the country, but are also of universal significance" (Yang, 2017). In the process of internationalizing the social sciences, China's E-SSR standards should be formulated to achieve an equilibrium between internationalization and localization. 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.	2020-07-30T02:07:34.933Z	2020-07-27T00:00:00.000	{ "year": 2020, "sha1": "98982d290ebb8f1da89c69976cc07df2112301ff", "oa_license": "CCBYNC", "oa_url": "https://journals.sagepub.com/doi/pdf/10.1177/2096531120943210", "oa_status": "GOLD", "pdf_src": "Sage", "pdf_hash": "d3e654a9ab2115187e1def284dd3f8a61b634743", "s2fieldsofstudy": [ "Education", "Political Science" ], "extfieldsofstudy": [ "Sociology" ] }
225698858	pes2o/s2orc	v3-fos-license	A quality framework for assessing the designed curriculum – a basis for the Czech PE curriculum revision This paper presents a general curriculum quality framework that has been developed on the basis of research in the Czech Republic and internationally. The cogency of this framework and its relevance to the Czech PE curriculum is demonstrated by reference to curriculum research. Finally its potential as a quality assessment tool for the review of the Czech Physical Education curriculum is discussed. INTRODUCTION Since the fall of communism in the Czech Republic in 1989 significant educational reforms have taken place, including major reform of the Czech school curriculum. The current national curricular documents (the Framework Educational Programmes -FEPs) were issued from 2005 onwards for the different levels of education. The general Czech curriculum, in its concept and design, is perceived as problematic (e.g. Dvořák, Starý, & Urbánek, 2015;Janík, 2013;Janík et al., 2018Janík et al., , 2019Kuhn, 2011, etc.). According to Šíp (2014), this is due to the inadequacies of the reform process and the fact that the strengths and weaknesses of the original curriculum were not clearly established prior to the reforms. We agree with this view (c.f. Kuřina, 2014;Vlček, 2019) and also believe that the changes were not sufficiently discussed beforehand, with academic experts, in schools or with teachers. This might be the reason behind the unsatisfactory outcomes of the curricular reform and the failure to accept some changes. There is also a large body of research, which indicates that the physical education (PE) curriculum in the Czech Republic is not working as it should (Habrdlová, Lupač & Vlček, 2017;Lupač, 2016;Habrdlová & Vlček, 2015;Tupý, 2018;Vlček & Mužík, 2012;Vlček, 2019). From the reform's beginning, a health concept underpinned the Czech PE designed curriculum. Despite the fact that the programmes have been shown to be problematic and have been subject to ongoing content reviews, approved by the Ministry of Education and Sport (MEYS), the health oriented concept of the Czech PE curriculum has not changed significantly and in fact in educational practice different PE concepts still prevail (Fialová, Flemr, Marádová, & Mužík, 2014;Vlček, 2019). The research also shows that there is a lack of internal congruence in the PE curriculum between the health-oriented learning objectives and the more balanced educational content (health/movement) and that the health-oriented PE concept is not accepted by the public and is not being implemented by teachers (cf. Fialová et al., 2014;Vašíčková, 2016;Janík, Vlček, & Mužík, 2016). In 2018 the MEYS has announced a new cycle of revision for the Czech FEPs which will also affect the PE designed curriculum. This project is now referred to as Revision of the FEP and preparation of the Education Policy Strategy of the Czech Republic until 2030+ (Strategy 2030+) 1 . In January 2019 an expert group was established under the leadership of Prof. Arnošt Veselý. Their task is to prepare the initial document Guidelines for Education Policy of the Czech Republic 2030+ to define the vision, priorities and objectives of education policy beyond 2030. It will describe what should be achieved and how these goals can be achieved. For the purposes of the review it is important that the quality of the existing curriculum is assessed but the question is how, and what tools can be used. In 2011 an interdisciplinary team, which included the author, conducted an expert survey of experienced directors (head-teachers) and teachers in grammar schools 2 , to hear direct-1 For more information see http://www.nuv.cz/t/rrvp. 2 A grammar school is a selective high school at the stage of upper secondary education typically beginning at age 15 or 16 years (ISCED 3). ly from experts in the field regarding what makes a quality designed curriculum ( Janík et al., 2011). These results were combined with similar studies conducted in the Czech Republic and abroad (Böttcher, 2006;Dvořák, 2012;Dane & Schneider, 1998;Egger et al., 2002;Gehrmann, Hericks, & Lüders, 2010;Gandal & Vranek, 2001;Halbheer & Reusser, 2008;Havel, 2016;Kurz, 2005;Squires, 2012Squires, , 1998Stake, 1967Stake, , 1972Thijs & Van den Akker, 2009;Zhu, Ennis, & Chen, 2011) and a general quality framework developed ( Janík et al., 2011). Unfortunately, this framework has not been used in discussions about the PE curriculum quality in the Czech Republic. This paper presents this quality framework and discusses its applicability as a potential assessment tool for the planned PE curriculum revision in the Czech Republic. To the Curriculum and its quality It is important to note that the term curriculum is not defined consistently by different researchers. In a narrow definition, a curriculum means a teaching program. In the broadest sense, the curriculum means all the learning that takes place at school or in other institutions, both planned and unscheduled (cf. Lawton & Gordon, 1993, p. 66). In this text we view the 'curriculum' as somewhere between these extremes. We distinguish five forms of the curriculum (Průcha, 2002): • Concept form -vision, rationale or 'basic philosophy' underlying a curriculum. • Designed form -official documents (e.g. syllabi) prescribed at both the government level (the educational framework) and at the school level, but also and associated teaching materials, text books, etc. • Implemented form -curriculum as interpreted and used (especially by teachers). • Results form -outcomes of the actual process of teaching and learning. • Effects form -the impact of the acquired and attained learning outcomes on learners. Frequently in discussion of the curriculum only the designed form is considered. However, in any quality assessment it is not sufficient to consider just the designed form of the curriculum -all five forms must be taken into account. Janík et al. (2013, p. 20, cf. Průcha, 1996Helmke, 2007, p. 40) distinguish two different meanings of the term 'quality' (cf. the Oxford definition of quality 3 ). 1) A descriptive meaning which identifies a desirable characteristic or attribute of a pedagogical phenomenon, for example the practicality of a designed curriculum. 2) An evaluative meaning which describes a desirable standard or optimal level of achievement. This involves setting targets using normative approach as defined by Terhart, (2000, pp. 815-816) and their evaluation and measurement using an empirical approach - (Terhart, 2000, p. 817) as for example used in PISA (Programme for International Student Assessment) and TIMSS (Trends in International Mathematics and Science Study). In this text we use the first definition. This means that quality can be viewed as a complex entity that can be broken down into specific attributes. Through monitoring and evaluation of these attributes, the problematic aspects of a curriculum can be identified ( Janík et al., 2013, p. 21). The Quality Framework The framework presented here, developed by Janík et al. (2011), was developed to define the general characteristics of a quality designed curriculum. It comprises specific quality criteria grouped within four key areas 4 . These key areas provide an overall quality framework and the detailed quality criteria describe these areas more comprehensively. • Area 1: feasibility and practicality Criteria/characteristics: 1.1 respects reality and is manageable, 1.2 respects teaching/learning practice and is practical, 1.3 is instructive, inspirational and motivating for teachers, 1.4 is usable in the school environment by managers and teachers, 1.5 encourages communication and cooperation in school. • Area 2: professional accuracy and congruence Criteria/characteristics: 2.1 is consistent with similar documents such as assessments and inspection criteria, 2.2 is logical and interconnects goals and educational content, 2.3 accurately reflects the relevant discipline and the current state of disciplinary knowledge. • Area 3: clarity and comprehensibility Criteria/characteristics: 3.1 is thoughtfully structured and well-arranged, 3.2 is written so that it is accessible, understandable and accepted, 3.3 it is structurally interconnected, 3.4 it is concise but includes the essentials. • Area 4: flexibility within overall bindingness Criteria/characteristics: 4.1 provides a reasonable space for free decision-making, 4.2 is durable but provides flexibility for updating educational practice, 4.3 provides a desirable degree of uniformity between schools, 4.4 it defines what is important for pupils to acquire (the core curriculum) and is binding. DISCUSSION Area 1 requires a quality curriculum to be feasible and practical. The terminology has been modified in translation to better explain the framework. has confirmed that curriculum reform in the Czech Republic took place largely on paper, at the documentation level; to a much lesser extent in teachers' mind-set, at the cognitive-emotional level, or in their implemented teaching (cf. Janík et al., 2018;Pešková, Spurná, & Knecht, 2019). Hence, many PE teachers accepted curriculum reform formally, in theory, but not in practice (Vlček, 2019;Janík, Vlček, & Mužík, 2016, p. 139). Despite the health-oriented concept clearly expressed in the Framework Educational Programme for Basic Education (FEP BE) learning objectives, most teachers are not motivated by the health-oriented PE concept and prefer a focus on sport (Fialová et al., 2014, pp. 77-83). Furthermore, research on the results and effects forms of the curriculum show that the levels of physical activity and the health status of the Czech population are unsatisfactory (cf. Antošová & Kodl, 2014; OECD/European Observatory on Health Systems and Policies, 2017; Mitáš & Frömel, 2013). It appears, therefore, that the PE curriculum is problematic presumably in part because it is not feasible and practical. Area 2 relates to professional accuracy and congruence. Congruence (or consistency, coherence, fidelity, or alignment) has been consistently emphasized as a critical quality criterion (Stake, 1967(Stake, , 1972Dane & Schneider, 1998;Egger et al., 2002;Thijs & Van den Akker, 2009;Zhu, Ennis, & Chen, 2011). Again, all five forms of the curriculum should be considered. Is the concept underpinning the curriculum congruent with the goals as expressed in the designed curriculum and the subject matter? Do teachers actually teach the designed curriculum? There are many studies which show that this is not the case in the Czech Republic. For example, there is significant incongruence between the exclusively health-oriented learning objectives of the educational area 'Man and Health' and the health-movement oriented PE content. Other research shows that the health-oriented concept form is not accepted by the public (Mužík & Hošková, 2010; and is not what is being implemented (implemented form) by teachers (cf. Fialová et al., 2014;Vašíčková, 2015, Janík, Vlček, & Mužík, 2016. Area 3 requires a quality curriculum to be clear and comprehensible. This has been shown to be one of the most problematic areas of the Czech curriculum. Teachers have expressed the view that the PE learning objectives in the designed form are too abstract and difficult to understand Vlček & Mužík, 2012). Another problem is that the FEP BE is conceptually confusing because of the lack of a clear relationship between the visionary key competences of overall basic education and the PE learning objectives and expected PE outcomes (Vlček, 2019;c.f. Knecht, 2014;Píšová, Kostková, & Vlček, 2011), that is, it is not thoughtfully structured. Area 4 relates to the need for a curriculum to be binding and durable but to allow the flexibility to be updated as educational content and practice changes. The two level Czech educational programmes provide considerable opportunity for free decision making by schools and teachers. However, the PE designed curriculum defined in FEP BE is not binding and does not contain a core curriculum that describes the main learning objectives and educational content of individual disciplines including PE. Therefore, there is a lack in uniformity as far as implemented curriculum is concerned (as described above) and especially in the results curriculum form in terms of learning outcomes evaluation (cf. Fialová, 2015, p. 54). CONCLUSION It is clear that the quality framework presented here is consistent with the numerous research studies on curriculum quality and also with research into the PE curriculum. It is a comprehensive tool that takes account of all the different quality criteria that have been identified. It provides a tool to comprehensively assess quality y of a curriculum. In the introduction we referenced a large body of research that indicates that the designed curriculum in the Czech Republic, including PE, is not working as it should, and that a national review is currently underway. As a part of the review process, studies for the revision of the curriculum for particular school subjects have been prepared, including a study by Tupý (2018) identifying problem areas of the PE curriculum, also in the designed curriculum. However, the issues of curriculum quality are extensive, complex, and often multifactorial (c.f. Fend, 1998Fend, , 2008Hopmann & Riquarts, 1995;Hopmann & Gundem, 1998;Thijs & van den Akker, 2009) and there are issues with all five forms of the Czech PE curriculum which should be addressed (Vlček, 2019). In our view only a comprehensive quality framework such as presented here, while developed in relation to the curriculum generally, can provide a sufficiently robust tool in the upcoming PE curriculum review and redesign in the Czech Republic. Although we believe that a rigorous and comprehensive review of the curriculum, based on all five forms, is an essential prerequisite to a quality redesigned PE curriculum, it is important to note that comparative research (Vlček, 2019) and other texts critiquing the Czech PE curriculum (Habrdlová, Lupač, & Vlček, 2017;Lupač, 2016;Habrdlová & Vlček, 2015;Tupý, 2018;Vlček & Mužík, 2012) have shown that the designed PE curriculum in the Czech Republic is largely consistent with the curriculum of other developed countries in Europe and also USA. This is an important observation; it means that the Czech PE curriculum, while it may have some problems, is not fundamentally broken or significantly different from the curriculum in other countries.	2020-07-02T10:38:29.374Z	2020-06-24T00:00:00.000	{ "year": 2020, "sha1": "d7b2afdfb56c6e25fac9bd22c194b6eaa6c2dd88", "oa_license": "CCBY", "oa_url": "https://karolinum.cz/data/clanek/8066/Kinan_56_1_0047.pdf", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "5b2afdab1a745c17486b0d2c0946fb98158358ed", "s2fieldsofstudy": [ "Education" ], "extfieldsofstudy": [ "Political Science" ] }
3017701	pes2o/s2orc	v3-fos-license	Cardiovascular medication utilization and adherence among adults living in rural and urban areas: a systematic review and meta-analysis Background Rural residents face numerous barriers to healthcare access and studies suggest poorer health outcomes for rural patients. Therefore we undertook a systematic review to determine if cardiovascular medication utilization and adherence patterns differ for rural versus urban patients. Methods A comprehensive search of major electronic datasets was undertaken for controlled clinical trials and observational studies comparing utilization or adherence to cardiovascular medications in rural versus urban adults with cardiovascular disease or diabetes. Two reviewers independently identified citations, extracted data, and evaluated quality using the STROBE checklist. Risk estimates were abstracted and pooled where appropriate using random effects models. Methods and reporting were in accordance with MOOSE guidelines. Results Fifty-one studies were included of fair to good quality (median STROBE score 17.5). Although pooled unadjusted analyses suggested that patients in rural areas were less likely to receive evidence-based cardiovascular medications (23 studies, OR 0.88, 95% CI 0.79, 0.98), pooled data from 21 studies adjusted for potential confounders indicated no rural–urban differences (adjusted OR 1.02, 95% CI 0.91, 1.13). The high heterogeneity observed (I2 = 97%) was partially explained by treatment setting (hospital, ambulatory care, or community-based sample), age, and disease. Adherence did not differ between urban versus rural patients (3 studies, OR 0.94, 95% CI 0.39, 2.27, I2 = 91%). Conclusions We found no consistent differences in rates of cardiovascular medication utilization or adherence among adults with cardiovascular disease or diabetes living in rural versus urban settings. Higher quality evidence is needed to determine if differences truly exist between urban and rural patients in the use of, and adherence to, evidence-based medications. Background Rural and urban communities have distinct characteristics in terms of demographics, social, and physical environments, and may vary in access to healthcare facilities and services. Rural residents tend to be older and are more likely to be obese, have less education, and lower income than urban residents [1][2][3][4][5][6]. Rural populations also have a higher prevalence of chronic conditions such as diabetes and heart disease [1,[7][8][9]. Some studies have shown worse health outcomes among rural populations, including a higher risk of cardiovascular-related morbidity and mortality compared to urban populations [3,4,[10][11][12][13]. For example, in Canada, mortality from circulatory diseases is higher in rural than urban communities, as is the risk of heart failure-related mortality, hospitalization and emergency department visits [4,10,11]. Similarly, in the UK a higher risk of in-hospital death due to ischemic heart disease has been observed among rural residents [12], and in Australia, mortality due to six major chronic diseases consistently increased in areas that were increasingly remote [13]. Collectively, these characteristics suggest increased healthcare needs for those living in rural communities. However, rural residents report several barriers to accessing healthcare including transportation difficulties and distance to care, social isolation, financial constraints, limited health care facilities (hospitals and pharmacies), physician shortages, and lack of access to specialist care [14][15][16][17][18][19]. Indeed, in the US, rural areas contain 19% of the population but only 11% of the physician workforce [15], and the ratio of specialists per population consistently declines as locations become smaller and more remote [15,17]. The lack of access to primary care physicians, specialists, or health care facilities has been postulated to result in decreased prescribing of evidence based medications. However, a previous systematic review found no clear rural-urban difference in the prevalence or intensity of prescription drug use in older adultsalthough that review included a wide variety of health conditions and medications [20]. It is possible that important differences may exist for certain disease states or medication conditions. As a result, we conducted a systematic review that evaluated whether cardiovascularrelated medication utilization and adherence patterns differ for rural versus urban adults with cardiovascular disease or diabetes. These two disease states were selected as they affect a large number of patients, are associated with high morbidity and mortality, often require multiple medications to manage, and outcomes are known to be different between rural and urban patients [21]. Inclusion and exclusion criteria Controlled clinical trials or observational studies were included if they enrolled adults with established cardiovascular disease (atrial fibrillation, hypertension, heart failure, coronary artery disease) or diabetes, and reported cardiovascular medication use or adherence patterns for patients living in rural versus urban communities. Medications of interest included acetylsalicylic acid (ASA), antithrombotic, anticoagulant, antihypertensive (including angiotensin converting enzyme inhibitors (ACEI) and angiotensin receptor blockers (ARB)), or lipid lowering agents. The research question, inclusion and exclusion criteria, and review methods were outlined in a protocol developed a priori according to the PRISMA guidelines [22]. Since the definition of rural and urban varied substantially between studies, we a priori defined populations described as urban, city dwelling, or metropolitan in the primary publication as urban. Conversely, rural descriptors included town, village, country dwelling, nonmetropolitan or remote communities. Any definition of adherence or persistence used in primary studies was accepted. Only full text, peer reviewed articles, were included. Studies evaluating the use of medications for acute management, such as during hospitalization, were excluded, as were studies conducted in developing countries where management approaches may be substantially different. The populations of interest were those with established cardiovascular disease or diabetes, in whom several evidence-based medications are recommended for use. Two researchers (GKM, DLW) independently screened all studies and extracted all data using pre-defined forms and definitions, and disagreements were resolved through discussion, or by a third researcher (DTE). Literature search strategy A comprehensive search strategy implemented by a research librarian was done in April 2012 in the following electronic databases: MEDLINE®, PubMed, Embase, International Pharmaceutical Abstracts, CINAHL, and Web of Science® and reference lists of included articles were also manually searched. Previously identified included studies were searched in Scopus to gather additional subject headings. No language, study design or date restrictions were applied. The MEDLINE® search strategy is listed in (Additional file 1: Table S1). Data extraction and quality assessment Studies were evaluated for bias, and the STROBE checklist was used to assess the quality of reporting [23]. Study authors were contacted for missing information on rural-urban comparisons, and in two cases additional data were provided (unpublished data V. Maio 2012, I. Carey 2011). Both unadjusted and adjusted data were abstracted or calculated where possible [24]. If more than one adjusted analysis was reported, the analysis that adjusted for the most confounders was extracted, and medication use data for patients without contraindications to treatment were preferred over populations that may have included patients who were not eligible for a specific therapy. Where possible, studies reporting multiple rural or urban populations were combined. Data analysis methods To summarize the effects of rural and urban location on medication utilization or adherence both unadjusted and adjusted pooled effects were calculated. As we expected heterogeneity between studies, we pooled effect estimates using a random effects model with inverse variance weighting and Review Manager 5.1 software [25]. Heterogeneity was assessed using the I 2 statistic with an I 2 statistic >50% being considered as moderate heterogeneity. There was no a priori degree of heterogeneity that precluded pooling. For studies reporting multiple outcomes within the same cohort [e.g., % receiving betablockers (BB) and % receiving an ACEI], a pooled estimate of the odds of treatment were calculated using methods recommended by Borgenstein et al. [26] that accounts for the fact that patients within each outcome are not mutually exclusive (i.e., a patient may have received both a BB and an ACEI). Since the correlation of outcomes is unknown, we used a moderate correlation of 0.5 with sensitivity analyses using 0.25, 0.75 or 1 and found there was little impact on the results. Subgroup analyses were further conducted to explore the robustness of our results and potential sources of heterogeneity. Studies reporting data not suitable for meta-analysis (e.g. outcomes other than OR, or with missing data) were summarized narratively. Publication bias was assessed using funnel plots and Egger's test. Results A total of 11092 citations were identified in the literature search and 51 unique studies (described in 52 publications), met the inclusion criteria ( Figure 1) [2,8,9,. Fifteen studies were cohort studies and 36 were cross sectional or repeat cross sectional studies (Additional file 1: Table S2). The included studies were published over a 21-year time span (1990 to 2011) and had quality scores based on the STROBE checklist that ranged from 8.5 to 21 (median 17.5) out of a total of 22 possible items (0.5 points given for partial reporting). Two reports provided data on the same study [44,48], and six studies included data for more than one patient cohort [43,49,54,66,71,72]. Thus, 58 patient populations (or cohorts) were included in our study. Two studies were in a language other than English and were translated using on-line resources and local expertise [37,70]. There was good agreement between reviewers on study selection (kappa 0.82, 95% confidence interval [CI] 0.75 to 0.90). Included studies varied in their characteristics (Additional file 1: Table S2). Studies ranged in sample size from 32 to approximately one million and were conducted in a range of areas including the US (30 studies), Europe (14), Canada (5) and Australia (2). Patient populations included those with acute myocardial infarction or coronary artery disease (18 studies), hypertension (16), diabetes (8), chronic heart failure (6), atrial fibrillation (5), or mixed cardiovascular disease populations (5). The average age of study participants ranged from 42 to 80 years, and 28% to 63% were female. Medication utilization Forty seven (92%) studies [2,8,9,27-36,38-47,49,51-69, 71-73,75] evaluated cardiovascular medication utilization with 20 (39%) studies specifically evaluating utilization of ASA or other anti-thrombotic agents, 34 (67%) evaluating antihypertensive use, and 11 (22%) evaluating the use of lipid lowering agents (Additional file 1: Table S3). Substantial variation in the use of cardiovascular medications was observed between studies and between rural versus urban sub-populations within each study. Indeed, the absolute difference in the utilization of cardiovascular medications ranged from −46% to +4% in rural versus urban patients for ASA or other anti-thrombotic drugs, −37% to +25% for antihypertensive drugs, and −45% to +8% for lipid lowering agents. Of the 47 studies that evaluated cardiovascular medication use, sufficient data for pooling were available in 34 studies ( cohorts as two cohorts were included in three studies [43,66,72] and three cohorts in one study [71] (Figure 2). In both analyses there was substantial heterogeneity between studies. Although numerous subgroup analyses by setting, drug [43], Rice [66], Wan [71], and Williams [72] provided data for more than one cohort. CI = confidence interval; NA = not applicable; OR = odds ratio. class, disease, age, country, and publication year were undertaken, similar results were observed and these factors only partially explained some of the variation between studies (Table 1; Figures 2 and 3). Moreover, when studies were categorized according to how the outcome was assessed (patient self-report, medical chart review, or administrative data), similar results were observed (Table 1). Additional analyses by study quality, for countries with universal health care systems, and for specific drugs, such as ACEI/ARB, also found similar findings (Additional file 1: Table S4). Among studies reporting data not suitable for meta-analysis, the findings were also consistent in that there was no clear trend towards a reduction or increase in cardiovascular medication utilization between rural and urban patients [2,8,[30][31][32][33]40,49,54,59,63,65,67]. Publication bias was assessed by visually examining funnel plots and no obvious asymmetry was noted (Egger's p value = 0.98). Medication adherence or persistence Six (12%) studies [29,37,46,70,73,74] evaluated cardiovascular medication adherence and two (4%) studies [46,50] evaluated medication persistence. Adherence was measured as the percentage of doses taken [73], the proportion of patients with a medication possession ratio ≥0.8 [46], and was undefined in three studies [37,70,74]. Persistence was measured as the proportion of patients remaining on the same [46,50] or any treatment [46] at the end of the follow up. In five studies, the drugs evaluated were antihypertensive agents, one study evaluated heart failure medications [73], and one assessed ASA or ACEI/ARB adherence [29]. Discussion Our systematic review of the literature found that rural patients were 12% less likely to receive cardiovascular medications than urban patients in unadjusted analyses; however, pooling of data adjusted for patient, practitioner or other factors revealed no differences in the proportions of rural and urban patients receiving therapy. This suggests that differences in these characteristics between rural and urban residents are largely responsible for the discrepancies in medication use observed and is consistent with previous studies showing important differences in the demographics, health behaviors, and overall health of people living in rural and urban areas [1,3,4,7]. In this review, many of the included reports provided little data on the demographics or comorbidities of the rural and urban patient groups, which hindered our ability to assess the similarity of these populations. As a result, it is difficult to draw conclusions from those studies that reported only unadjusted rural-urban comparisons. When medication utilization data were pooled, substantial between-study heterogeneity was observed (I 2 = 97%). Some of this variability could be explained by differences in the setting (hospital, ambulatory care practice or community-based sample), age, and disease state. While most studies adjusted for some clinical characteristics, only some controlled for socioeconomic factors that could also have impacted medication use. Indeed only fourteen studies [27,32,34,42,43,45,47,52,57,58,61,62,66,69] reported adjustment for health insurance and previous studies have shown patients with a chronic condition who lack medication insurance are less likely to take medications or frequently skip doses due to cost [52,66,77]. Similarly, we found no consistent relationship between rural residence and cardiovascular medication adherence or persistence rates based on unadjusted data. The adjusted analysis suggested higher adherence and persistence for rural residents, although there were few studies reporting these outcomes. Rural-urban differences in other health behaviors, such as smoking, exercise, and consumption of fruits and vegetables have been reported [4], and considering the link between adherence and positive health outcomes [78], further study is warranted. Although we conducted an exhaustive search for literature and conducted our systematic review in accordance to the highest reporting standards, our review is not without limitations. Firstly, studies evaluating differences in urban versus rural settings within subgroup analyses may not have been easily identified. Second, as in any systematic review, the findings are limited by the quality of the individual studies. Potential limitations included reporting bias [32,33,40,41,45,49,51,54,71,72], selection bias [9,35,54], lack of generalizability [9,28,36,37,39,50,51,59,71,73,74], limited sample size [28,36,37,39,50,64,73], no adjustment for confounders [8,28,[35][36][37]40,50,54,64,74,75], and poor reporting (STROBE score below the 25th percentile) [8,28,36,37,40,44,[50][51][52]54,59,64,70,71,74]. Third, we accepted a broad range of definitions for rural and urban populations, which may have affected study results. In 25 studies [8,28,34,36,37,40-43,45,46,50,52,53,55-60, 64-66,72,74] no clear definition of rural and urban were provided and various definitions were used in the remaining studies. It is possible some of the heterogeneity observed in our pooled analysis is related to these differences in definitions, although subgroup analysis by studies with defined and undefined rural populations showed similar results. Fourth, between-study heterogeneity was high and not fully explained despite multiple subgroup analyses by setting, drug class, disease, age, country, publication year, and data source (self-report, administrative or medical records). Moreover, clinical indications and classification of disease may also have changed over time increasing the heterogeneity between studies. Last, our review only included adults with established cardiovascular disease as medication utilization for secondary prevention was expected to be high within patients; thus, improving the power to detect differences if differences exist between rural and urban patients. Moreover, as cardiovascular medications are widely prescribed medications in the general adult population, our results would be expected to be highly generalizable. Conclusions In conclusion, we found no consistent relationship between rural versus urban residence and utilization of, or adherence with evidence-based cardiovascular medications among adults with cardiovascular disease or diabetes. There was substantial between-study variation that was only partially explained by the setting, age, and disease. Higher quality evidence is needed to determine if differences in cardiovascular medication utilization and adherence between urban and rural patients truly exist.	2018-04-03T02:42:08.076Z	2014-06-02T00:00:00.000	{ "year": 2014, "sha1": "0da879c5e76359e3aa7b8ae7158da75766da92ba", "oa_license": "CCBY", "oa_url": "https://bmcpublichealth.biomedcentral.com/track/pdf/10.1186/1471-2458-14-544", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "f678cdf45b773dc4cf7142c63f7b46627af1aaa3", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
133665711	pes2o/s2orc	v3-fos-license	Evaluation of optimal reservoir prospectivity using acoustic-impedance model inversion: A case study of an offshore field, western Niger Delta, Nigeria Abstract The evaluation of economic potential of any hydrocarbon field involves the understanding of the reservoir lithofacies and porosity variations. This in turns contributes immensely towards subsequent reservoir management and field development. In this study, integrated 3D seismic data and well log data were employed to assess the quality and prospectivity of the delineated reservoirs (H1–H5) within the OPO field, western Niger Delta using a model-based seismic inversion technique. The model inversion results revealed four distinct sedimentary packages based on the subsurface acoustic impedance properties and shale contents. Low acoustic impedance model values were associated with the delineated hydrocarbon bearing units, denoting their high porosity and good quality. Application of model-based inverted velocity, density and acoustic impedance properties on the generated time slices of reservoirs also revealed a regional fault and prospects within the field. Introduction Determination of hydrocarbon bearing sands is a primary goal of most reservoir characterization projects and efforts are made to increase the confidence levels and reduce to the barest minimum the associated risks in drilling and exploration activities. Seismic inversion technique is principally a sophisticated process of inverting the seismic data into the elastic properties of the earth's subsurface. Elasticity and density are seismic characters of subsurface strata that are principally affected by the subsurface properties of rocks and fluids like lithology, porosity, fractures, textures, permeability, viscosity, fluid type and saturations (Mavko et al., 2009;Bosch et al., 2010). The quantitative characterization of these subtler fluid properties from seismic responses would go a long way to improve hydrocarbon reservoir characterization and reserve estimation within a hydrocarbon field (Russel, 1988;Doyen, 2007). Seismic inversion of the acoustic impedance property primarily involves the conversion of seismic traces into reflection coefficient time series, which are then converted back into acoustic impedance traces (Lavergne and Willim, 1977;Lindseth, 1979). These generated acoustic impedance traces have the capacity to improve the accuracy of geological interpretations (e.g. environments of deposition and stratigraphy) and subsequent correlations with several petrophysical properties derived from the wireline logs (Xinyang et al., 2015). Benefits of seismic acoustic impedance data over conventional seismic data have been discussed by several researchers (Duboz et al., 1998;Connolly, 1999;Latimer et al., 2000;Yilmaz, 2001;Pendrel, 2006). Seismic inversion converts the seismic reflection data into an acoustic impedance section where band-limited seismic reflection data are transformed into quantitative rock properties for effective reservoir identification and description. Where the acquired exploration data involve high quality seismic and better distribution of well control, the interpretations of acoustic impedance model inversion often facilitates an improved estimation of reservoir porosity, acoustic impedance and uncertainty (Pendrel, 2001;Alshuhail et al., 2009;Jalalalhosseini et al., 2015). Seismic acoustic impedance inversion section elucidates subsurface layers thereby enhancing visualization both in terms of layering and vertical resolution unlike reflection coefficient of raw seismic data that reveal only the interface. However, the major limitation of this method is that it suffers major setback when the reservoir thickness falls below a quarter of wavelength (1/4 k). These thin beds and other small targets can be resolved on seismic data in two ways. The first involves increasing the dominant frequency of the stacked data simply by raising the bandwidth of the seismic data. The other involves conducting a phase shift or phase rotation of the seismic data through the use of advanced data processing algorithms and other inversion techniques such as seismic coloured inversion (Oyeyemi et al., 2016;Oyeyemi et al., 2017). The values of acoustic impedance derived from seismic inversion process are suitable to infer zones of high and low porosity within the delineated reservoirs. The relationship between the derived acoustic impedance and porosity is such that when the former is low, the later and the reservoir potential in terms of hydrocarbon saturation would be high (Dolberg et al., 2000;Farajpour et al., 2010;Çemen et al., 2014). Varela et al. (2006) reiterated the importance of acoustic impedance seismic inversion in reduction of uncertainty associated with reservoir production forecast, while Kadkhodaie-IIkhchi et al. (2014) stated categorically that acoustic impedance section from a model-based seismic inversion technique is a robust tool for tight sandstone reservoir characterization. There are several methods for seismic inversion analysis and they are broadly categorized into either deterministic or stochastic process. The deterministic seismic inversion include band-limited, sparse-spike and modelbased techniques. The focus of this study is to use a model-based deterministic seismic inversion technique to evaluate the hydrocarbon potential and prospectivity of the delineated reservoirs within the study area in western Niger Delta. Geological setting The hydrocarbon field of study is situated between Longitudes 5°00 0 E-5°02 0 E and Latitudes 5°50 0 N-5°52 0 N lying within the western parts of the continental margin shallow offshore Niger Delta basin (Fig. 1). The basin is bounded in the South by the Gulf of Guinea and in the North by the cretaceous tectonic elements including the Abakaliki uplifts, Afikpo syncline and the Anambra basin (Fig. 1). The siliciclastic deposits within this basin are of Tertiary age with three lithologic formations termed Akata, Agbada and Benin Formations (Fig. 2). The basal marine shale Akata Formation extends down to the basement and is a pro-delta shale unit with characteristic dark-grey and medium hard with floral fossils within its upper portion. The overlying paralic sequences of Agbada Formation house the oil and gas bearing reservoir units in Niger Delta; this geologic formation is composed of sandstone with interbeds of shale units that are typical of the delta front, distributaries channels and deltaic plain depositional facies (Avbovbo, 1978). Agbada Formation is characterized with increasing shale content from the upper to the lower portion denoting the seaward advance of the Niger Delta basin over some geologic time. The topmost Benin Formation is composed of massive continental plain sands that are highly porous with relatively low minor shale interbeds connoting an alluvial environment of deposition. The predominant structural styles within the Niger Delta are syn-sedimentary structures also referred to as the growth faults, deforming the delta under the Benin continental sandstone facies. These growth faults, generally trending in NE-SW and NW-SE directions (Hosper, 1971), are byproducts of that gravity sliding during the sedimentation of the deltaic deposits and they are polygenic in nature as their complexity increase in a down-dip direction of the delta (Merki, 1972;Corredor et al., 2005). According to Orife and Avbovbo (1982) the stratigraphic traps that are associated with unconformity surfaces within the Niger Delta for oil and gas accumulations are paleochannel fills, crestal accumulations, sand pinch-outs, erosional truncations, canyons fills, incised valleys and lowstand fans. Shale intercalations or parasequence shale units within the Agbada Formation as shown in Fig. 2 have been interpreted to act as reservoir seals within the Niger Delta basin. Seismic inversion Acoustic impedance as a rock property is a product of density and velocity expressed by Eq. (1), both of which can be measured at well locations. Seismic data can also be expressed by Eq. (2) as the convolution of wavelet and reflection coefficient sequence. Where SðtÞ is the synthetic seismic record, WðtÞ is the seismic wavelet, RðtÞ is the reflection coefficient series, and NðtÞ is the random noise. Eq. (3) also shows the sequence of the reflection coefficient on normal incidence base on the assumption that the seismic incidence ray is perpendicular to the rock interface. Where R i , V i and q i are the reflection coefficient, speed and density of the i th layer. In order to generate acoustic-impedance sections from seismic data, one of the major steps is to estimate the inverse wavelet wðtÞ expressed by Eq. (4) which is used to ultimately build the initial model for seismic inversion (Fig. 1); where aðtÞ and dðtÞ are the inverse wavelet and unit impulse function respectively. The reflection coefficient RðtÞ is obtained through a convolution process of seismic trace SðtÞ and the inverse wavelet aðtÞ as formulated in Eq. (5). Reconstruction of the P-impedance section is achieved by using a recursive method of estimating the reflection coefficient series RðtÞ as stated in Eq. (6). Stratigraphy of the Niger Delta and variable density seismic display of the main stratigraphic units (Lawrence et al., 2002). Model-based seismic inversion is a technique that involves the building of a geologically consistent model and then comparing the same model with measured seismic data. The results of the comparison between the observed data and modelled data are then used to iteratively update the model in such a way as to obtain a better match with the observed seismic data. Establishing the mathematical relationship between the initial model data and the seismic data along with the model update through iterations are resolved using two approaches such as application of the general linear inversion (Cook and Schneider, 1983) and the seismic lithologic modelling method based on the work of Gelfand and Larner (1984). The generalized linear inversion technique will compute the geological model that best fits the seismic data using a least squares method. The vector of k model parameters and vector of real data n observations are expressed by Eqs. (7) and (8). The relationship between the model and real data is then expressed in a functional form as shown in Eq. (9). Once this functional relationship is derived between both observed data and the models, any set of model parameters will generate an output. This obvious condition of non-uniqueness is eliminated within the generalized linear inversion merely by analyzing the error between the model output and observed data, the generated model parameters are subsequently perturbing so as to produce an output with inherent lesser error. This process would eventually lead the repeated iterations towards a satisfactory solution. This process can be expressed mathematically as Eq. (10). The M 0 is the initial model, M is true earth model, DM is a change in model parameters, FðMÞ is the seismic data (or observations), FðM 0 Þ is the calculated values from initial model, and @FðM 0 Þ @M is Fig. 3. Survey basemap in the study area showing the seismic survey profile lines (inlines and crosslines) and the available wireline logs. change in the calculated values. The error between the observed data and the computed values is give as Eq. (11) which can be expressed in a matrix form in Eq. (12), where A is the matrix of derivatives with n-rows and k-columns. The solution to this matrix would appear as another matrix given as Eq. (13), and where A À1 is the inverse of the matrix A. Most times observed data than the model parameters (n > k), the matrix A is usually a non-square matrix and may not have a true inverse resulting to what is referred to as an overdetermined case. The above equation can be solve using least square solution expressed in Eq. (14). Dataset and method The geophysical data used for this research consist of poststack 3D seismic of 496 inlines and 780 crosslines with 3000 trace gathers and 751 samples per trace, covering an area of 83.45 km 2 and four digital suites of wireline logs designated as OPO 1-4 (Fig. 3). The depths of the wells range from 2438.4 m for OPO 2 to 2987.04 m for OPO 1. Density, sonic and porosity logs along with the checkshot and well-top data that were available for this research. Reservoir mapping and correlation were carried out using gamma-ray and resistivity logs motifs on the wireline logs. The 3D seismic data was time-migrated with good quality for horizon mapping and seismo-structural interpretation. Model-based seismic inversion technique, despite its inherent problem of nonuniqueness was adopted for this research because it could expedite better resolution compared to the band-limited and sparse-spike inversion techniques. The data preconditioning process was carried out using the procedures outlined by Veeken and Da Silva (2004) and the methodology workflow used for the seismic model-based inversion is presented in Fig. 4. The presented methodology has been attested to by several researchers to have high capacity of generating the subsurface structures from input geophysical data such that the resulting model fits the observation with some measure of reasonable error (Treitel and Lines, 2001). The input data comprising seismic and well-logging data were quality controlled to ensure that they were in suitable format for the seismic inversion. The well data were calibrated with checkshot data with the sole intension of transforming the data from depth in well logs to time domain in seismic, thereby updating the sonic logs. The data from all the well logs, synthetic seismogram and the estimated reflection coefficient series were used for the extraction of seismic wavelet (Fig. 5). High correlation coefficients obtained between synthetic and real seismogram with minimal errors in all the wells location are presented in Fig. 6. The extracted statistical wavelet with zero phase was applied in order to align the reflections on the synthetic with the composite trace. The synthetic seismogram was then stretched and squeezed so as to make events on the synthetic tie with the events on the original seismic data. This process of seismic-to-well tie ultimately produced a new sonic logs, depthtime data logs and well synthetics that properly aligned with that of the seismic data. The synthetic seismogram generated from the field data is presented in Fig. 7. An initial model was later built using the low pass filtered acoustic impedance logs of the wells and the inversion process was then applied to the entire 3D seismic volume. The output of the final inversion results was thus well data constrained for better interpretation. The time slices of inverted seismic window centered about the delineated horizons (reservoirs) were generated and evaluated based on the characteristic inverted elastic properties such as p-wave velocity, density, and acoustic impedance. These is necessary for both qualitative and quantitative analyses of the reservoirs prospectivity using the afore-stated relationships between the elastic properties and petrophysical properties such as volume of shale, porosity and fluid contents. Results and discussion The well logs correlation and analysis as presented in Fig. 7 reveal three lithofacies which are sand, sand-shale and shale facies. These lithofacies formed the sedimentary sequences within which identified hydrocarbon bearing reservoirs (H1-H5) were mapped and correlated across the entire wells (Fig. 7). Sand facies are porous strata that are impregnated with hydrocarbon, their pores are well interconnected and they are therefore permeable. Sand-shale facies are also porous and can accommodate hydrocarbon, but the fluid flow rate through these strata can be very slow as a results of shale intercalations. The pore spaces within the shale facies are not usually interconnected, having very low effective porosity and permeability. Hydrocarbon producibility of a clastic reservoir units such as in the Niger Delta basin can be greatly influenced by shale intercalations. The reservoir sand unit H1 on the well section correlation panel shows evidence of fewer shale intercalation compares to others (Fig. 7). Seismic inversion process is usually adopted to extract from seismic data the subsurface geology. This is done by estimating an interval impedance property from wireline logs data (sonic and density) which links directly to porosity from the seismic data. This is extremely important in that seismic inversion utilizes the power of both depth and spatial resolutions of the well logs and seismic data respectively (Veeken, 2007). The original seismic data and the initial model built are presented in Figs. 8 and 9. This initial model is the prior low frequency model generated from low pass filtered acoustic impedance logs from the wells and extrapolated along all the events in the seismic data. It is the absolute level of acoustic impedance for the seismic data. Four distinct sedimentary packages can be identified on the initial inversion model based on the acoustic impedance contrast. Boundaries between these layers and their internal lithologic contrasts are conspicuous; the first, second, third and fourth package extends from 800 to 1580 ms, 1580 to 2000 ms, 2000 to 2600 ms, and P2600 ms respectively. The five reservoirs H1-H5 are situated within the delineated second package from time 1620 to 2000 (ms) and the amplitude of the acoustic impedance values within this package is relatively low ranging from 19,396 to 22,383 ((ft/s)⁄(g/cc)). Low acoustic impedance values denotes a sand lithologic units with high porosity and hydrocarbon saturation. This line of interpreting porosity as well as fluid contents from acoustic impedance section has contributed immensely to limiting exploration risks and hydrocarbon resource evaluation in challenging reservoirs (Latimer et al., 2000;Huuse and Feary, 2005;Çemen et al., 2014;Kadkhodaie-IIkhchi et al., 2014;Farvour et al., 2015). The base of this sedimentary package can be interpreted as sequence boundary where there is a toplap termination of horizon from beneath at time 2100 (ms). The final inverted seismic section which was obtained using the generalized linear inversion technique (Fig. 10). This technique involves several recursive iterative procedures of recalibrating the logs to seismic data in order to get an optimal wavelet that is used to convolve the initial acoustic impedance model such that the inversion error is drastically reduced. The final seismic inverted section reveals the variations in the sand and shale contents across all the delineated sedimentary packages (Fig. 10) based on the interpreted subsurface acoustic impedance properties. The topmost package coincides with the continental sand of Benin Formation and the high impedance values in some part of this package is not unconnected to the absence of wireline logs data within this area of the seismic section. The underlined second package shows evidence of high net-to-gross sand content with variable shale intercalations increasing from NW to SE portion of the seismic section. This sedimentary package is that of typical Agbada Formation with high potential for hydrocarbon bearing reservoirs that possess high porosity and hydrocarbon saturation. Beneath this package is the basal unit of the Agbada Formation and the upper portion of the Akata Formation. This portion of the Niger Delta has been interpreted by some workers to be where the source rocks are situated (Evamy et al., 1978;Brownfield, 2016). The last sedimentary package is predominantly shale units with extremely high acoustic impedance values depicting very low to zero porosity values. Fig. 11(a-d) shows the inverted model P-wave velocity and density for the 3D seismic time slice corresponding to both H1 and H2 reservoirs with the wells. The essence of these attributes was to delineate any prominent geologic structures and map the hydrocarbon prospects within the field. The P-wave velocity is predominantly low within the range of 7623-9244 (ft/s) for the reservoir sand H1, whereas the north-western to south-western portion of the same reservoir is predominantly low in the inverted model density ( Fig. 9(a-b)). The range of inverted model p-wave velocity values within the portion of reservoir sand H2 where the wells are localized is between 8568 (ft/s) and 9244 (ft/s) which is quite low but is incomparable to that of the reservoir H1. A mappable regional fault is evident on the inverted p-wave velocity time slice maps for both H1 and H2 (Fig. 11a and c). The inverted model density for reservoir H2 is equally higher than that of the reservoir sand H1. These observation is basically tied to the shale contents of both reservoirs. Two prospect (X and Y) are identified as the area where there are high sand fairways within the field using inverted density attributes on the reservoirs (Fig. 11b and d). In terms of the Fig. 11. Inverted P-wave velocity and density attributes over the time slice of the 3D seismic data corresponding to the delineated reservoir (a and b) H1 and (c and d) H2 on seismic section. The arrows show the presence of a regional fault delineated by p-wave velocity attributes. model inverted acoustic impedance for the 3D seismic slices corresponding to the above reservoirs, sand H1 has lower impedance values than sand H2 as presented in Fig. 12(a-b), and this observations further affirm that reservoir H1 is cleaner (with low volume of shale) than reservoir H2. The portion within each reservoir time slice corresponding to the delineated prospects (X and Y) have characteristic low acoustic impedance denoting high porosity and fluid contents. Conclusion The following conclusions are drawn from the research: (i) Inverted acoustic impedance model provides a better improved seismic attributes for accurate subsurface characterization and subsequent reservoir modelling than any the attributes derived from band-limited seismic data. (ii) The generated impedance volume illuminates better the aerial and vertical distribution of the lithology and porosity. The output impedance volume equally yields credible estimations of these petrophysical properties away from well points. (iii) The producing wells and delineated hydrocarbon bearing reservoirs fall within the area where the acoustic impedance values are low. Porosity and fluid saturation values in this area are characteristically high depicting a sand lithological units with intercalations of shale pockets. (iv) The results of the acoustic impedance inversion would provide useful means of mitigating the risks associated with the exploratory efforts within the study area.	2019-01-02T03:45:01.139Z	2017-12-01T00:00:00.000	{ "year": 2017, "sha1": "f65a08907b28fe69e723439ad9ea6abae5909b82", "oa_license": null, "oa_url": "https://www.tandfonline.com/doi/pdf/10.1016/j.nrjag.2017.11.001?needAccess=true", "oa_status": "GOLD", "pdf_src": "TaylorAndFrancis", "pdf_hash": "20e1c88443b718bbe40f4f8ff3445abb72bd5f55", "s2fieldsofstudy": [ "Geology" ], "extfieldsofstudy": [ "Geology" ] }
19163725	pes2o/s2orc	v3-fos-license	Sonosensitive MRI Nanosystems as Cancer Theranostics: A Recent Update In the tireless search for innovative and more efficient cancer therapies, sonosensitive Magnetic Resonance Imaging (MRI) agents play an important role. Basically, these systems consist of nano/microvesicles composed by a biocompatible membrane, responsive to ultrasound-induced thermal or mechanical effects, and an aqueous core, filled up with a MRI detectable probe and a therapeutic agent. They offer the possibility to trigger and monitor in real time drug release in a spatio-temporal domain, with the expectation to predict the therapeutic outcome. In this review, the key items to design sonosensitive MRI agents will be examined and an overview on the different approaches available so far will be given. Due to the extremely wide range of adopted ultrasound settings and formulations conceived, it is hard to compare the numerous preclinical studies reported. However, in general, a significantly better therapeutic outcome was noticed when exploiting ultrasound triggered drug release in comparison to traditional therapies, thus paving the way to the possible clinical translation of optimized sonosensitive MRI agents. INTRODUCTION New insights in cancer biology, along with the advances in early detection and treatment, led to increased life expectancy, and reduced cancer related deaths. However, there is still a strong need for more efficient, precise, and safer therapies. To decrease the systemic toxicity of some chemotherapeutic agents, drug encapsulation into biocompatible nanovesicles, named liposomes, has been envisaged. The FDA approval of Doxil R /Caelyx R in 1995, a doxorubicin liposomal formulation, for the treatment of Kaposi sarcoma, metastatic breast cancer and recurrent ovarian cancer (Barenholz, 2012), boosted the research for optimizing liposomal preparations. Liposomes are highly versatile nanovesicles consisting of a phospholipid bilayer surrounding an aqueous core; these vesicles can vary in size, shape, and lipid composition (Sessa and Weissmann, 1968). Basically, liposomes can carry drugs and/or other molecules, such as imaging or targeting agents, both in their aqueous core and/or in the lipidic membrane.Liposomes are designed to prevent drug extravasation into healthy tissues, prolong blood circulation time, improve drug accumulation, and enhance bioavailability at the target site (Blanco et al., 2015). The accumulation of liposomes into solid tumors is mostly due to peculiarities of the cancerous tissue, namely increased blood supply, enhanced endothelial permeability and reduced lymphatic drainage, resulting into the Enhanced Permeability and Retention (EPR) effect (Maeda et al., 2000). Once accumulated in the tumor, the drug needs to be released from the carrier. For the nanomedicines currently approved in clinics, this crucial step occurs spontaneously, i.e., following the natural degradability of the nanocarrier interacting with tissue components (Rizzitelli et al., 2015). However, to better control this process, liposomes sensitive to endogenous (e.g., pH, redox potential, enzymatic activity) or exogenous (e.g., heat, light, pressure) stimuli can be designed (Guo and Szoka, 2003). Among different possibilities, in the last decade, considerable attention has been devoted to the use of ultrasounds (US) (Pong et al., 2006), as they are already clinically approved tools for imaging and therapy. Moreover, they can be modulated in terms of frequency and intensity according to the specific goal to be achieved. In addition, in view of personalized medicine based approaches, co-encapsulation of an imaging agent inside the vesicle could be envisaged in order to follow the release process and predict the therapeutic outcome. In this respect, Magnetic Resonance Imaging (MRI), due to its outstanding spatial resolution, low invasiveness and absence of limits in tissue penetration could be regarded as the optimal imaging technique. Small paramagnetic molecules can be loaded into liposomes in order to spatio-temporally track drug delivery and/or drug release. In the next paragraphs an overview on the different sonosensitive MRI agents will be presented, with a focus on ultrasound-based trigger mechanisms. Finally, some examples of successful preclinical applications in this field will be reported. US Mechanical Effect The US mechanical effect is based upon a combination of micromassage, cavitation and acoustic streaming. Micromassage refers to US induced cell vibrations, likely affecting tissue fluid interchange and tissue mobility. Cavitation is defined as the phenomena of the formation, growth, and subsequent collapse of microbubbles (Frenkel, 2008). Newly formed microbubbles or administered sonosensitive vesicles can oscillate stably (stable cavitation), inducing a constant fluid flow around the bubble, called microstreaming, that induces stress on cell membranes, and may enhance cell permeability. While, if the bubbles increase more than twofold their size, they violently collapse (inertial cavitation) causing microstreaming, formation of liquid jets and ultrasound shock wave emission, able to disrupt membranes of adjacent cells and create pores in capillary walls (Khokhlova et al., 2015). As cavitation phenomena can induce severe cytotoxic effects, a cavitation level sufficient to release drug form nanovesicles and permeabilize cell membranes but without killing cells should be induced (Pitt et al., 2004). To exploit this wide range of mechanical effects, specific US responsive nano-microsystem have been designed, mainly microbubbles (Hernot and Klibanov, 2008), and liposomes (Schroeder et al., 2007). More details about these systems are provided in section Sonosensitive Systems. US Thermal Effect The thermal effect, recently reviewed by T. Boissenot et al. (2016), is strictly linked to the application of high intensity US. More in details by focusing the US beam in a small area of a target tissue, the power per cross section area becomes extremely high, leading to significant absorption of thermal energy from the beam by the tissue and consequently resulting in local temperature rise. Resulting hyperthermia could be mild (39-42 • C) or high (>43 • C). Mild hyperthermia is generally employed to trigger drug release, as it will be discussed in paragraph Thermosensitive Systems. Whereas, high hyperthermia is mainly exploited to kill or ablate tissues (Diederich and Hynynen, 1999): uterine fibroids, prostate, breast, pancreatic, and liver cancers have been safely and successfully treated with High Intensity Focused Ultrasound (HIFU). In this regard, precise, and constant temperature monitoring of the heated area is of paramount importance; it can be obtained by means of invasive thermocouples or by MRI thermometry. ULTRASOUND RESPONSIVE AND MRI DETECTABLE PROBES So far, a number of reports has been published about the synthesis and optimization of ultrasound responsive and MRI detectable probes. The key components to design such systems are: (i) a lipidic membrane stable in physiological conditions, able to release its content selectively upon US exposure; (ii) a MRI agent, encapsulated within the system, able to report on drug release; (iii) an entrapped drug, that, upon release, carries out the therapeutic effect (Figure 1). Sonosensitive Systems Initially invented as contrast agent for ultrasound imaging and lately reconsidered for drug and gene delivery purposes, microbubbles are micron sized systems made of an external FIGURE 1 \| Schematic illustration of the mechanism of a liposome-based sonosensitive MRI agent. The MRI signal of a low water permeable liposome entrapping a relatively high amount of a MRI agent is almost silent, due to the compartmentalization of the probe. However, when the US stimulation triggers the release of the agent, the MRI contrast activates, thus signaling the release of the drug co-loaded in the nanocarrier. shell encapsulating an inner gas core. These systems display a strong tendency to undergo stable or inertial cavitation upon US exposure. The shell could be made up of lipids, polymers or proteins. The inner core could also be filled with liquid perfluoropropane, then triggered to the gaseous phase when stimulated with acoustic energy. Theranostic microbubbles were prepared by Fan et al. bearing a phospholipidic shell, endowed with a complex of SuperParamagnetic Iron Oxide (SPIO) nanoparticles and doxorubicin, and filled with perfluoropropane (Fan et al., 2016). Similar, nanosized, probes were developed by Cavalli et al. (2015). Palmitic acid was the surfactant of choice to be included in the shell in order to entrap the MRI agent Gd-DOTP. Prednisolone phosphate (PLP) was added as therapeutic agent, while pluronic F68 was used as stabilizing agent. Both the preparations, however, were characterized by limited stability over time (1-3 h). To this regard, liposomes appear promising systems, but, in order to be sonosensitive, special attention has to be dedicated to their membrane composition. TJ Evjen (2011) highlighted the important role played by the phospholipid Packing Parameter (PP) . Non-bilayer forming lipids with large hydrophobic cross section as compared to the polar headgroup (PP >1), like DOPE or DSPE, significantly promote liposome sonosensitivity (Evjen et al., 2013). This effect is associated to a DOPE/DSPE restructuring process: upon US exposure, a transformation from lamellar to reversed hexagonal phase occurs, inducing the formation of pores and/or tubular aggregates, through which the drug can readily leak out (Kang et al., 2014). In order to form liposomes, DOPE (or other phospholipids bearing PP < 1) must be mixed with phospholipids displaying natural tendency to form bilayers (PP∼1), cholesterol, useful to induce mechanical stability by tightening the membrane, and stealth moieties, to prolong blood circulation time. While the amount of cholesterol, saturated phospholipids, and polymers should be limited as they increase membrane stiffness and decrease the tendency to drug leakage, stealth moieties, like DSPE-PEG2000 seems to enhance sonosensitivity by the so called "antenna effect, " favoring the interactions between acoustic waves and the vesicles. Finally, Giustetto et al. demonstrated that also shape, size, and intravesicular composition of liposomes may influence US triggered drug release (Giustetto et al., 2013). Thermosensitive Systems In 1978, Yatvin et al. first described a thermosensitive formulation, consisting of a DPPC/DSPC (3:1) liposome able to selectively release its hydrophilic content when the temperature was raised up of a few degrees above physiological temperature (Yatvin et al., 1978). Since then, many efforts have been made to obtain improved formulations and to accurately monitor temperature variations in the region of interest, where drug release should take place. The fundamental points to be fulfilled to design a good thermosensitive system are:(i) the presence of temperature sensitive phospholipids or polymers in the membrane; (ii) stable encapsulation of drugs at body temperature; (iii) fast and complete drug release upon heat stimulation and (iv) provision of high drug plasma levels during the time span of hyperthermia treatment (Hijnen et al., 2014). The first thermosensitive liposomes (TSL) developed were mainly composed of phosphatidylcholines, bearing a transition temperature (T m ) in the range of mild hyperthermia (40-43 • C). Around T m , the temperature at which a polymer/phospholipid melts from the gel-ordered phase to the liquid-crystalline phase, a significant drug release was achievable due to grain boundaries between the two phases (Landon et al., 2011). In 1994 Unezaki et al. (1994) added 3% PEGylated phospholipids to a DPPC/DSPC 9:1 formulation, resulting in prolonged circulation time, due to PEG stealth effect, and better therapeutic outcome. PEGylated phospholipids were then considered a basic component. A breakthrough occurred in 2000, when Needham and coworkers proposed the inclusion of lysolipids (PP < 1) into the membrane bilayer (DPPC/MPPC/DSPE-PEG2000 90:10:4), obtaining 80% drug release around T m in a few seconds, thanks to the formation of lysolipids-stabilized membrane pores (Needham et al., 2000). This innovation led to the birth of ThermoDox R , a TSL formulation containing lysolipids and doxorubicin, that has reached phase III clinical trials for hepatocellular carcinoma treatment (Bulbake et al., 2017). In recent years a myriad of subtle variations has been made to the above-mentioned formulations and different simulated or experimental models have been developed to predict the complex interplay between liposome properties, tumor perfusion, heating regimen, and therapeutic efficacy (Gasselhuber et al., 2010(Gasselhuber et al., , 2012Lokerse et al., 2016Lokerse et al., , 2017. Interestingly, Banno et al. drew the attention on the importance of TSL formulation to retain liposome stability: they reported that 70% of lysolipids was lost within 1 h post injection of TSL, likely due to interactions with the large lipid membrane pool in vivo, thus inducing non-triggered drug leakage (Banno et al., 2010). The advantage of using HIFU as heating source is the possibility to obtain a fast and localized temperature increase. Moreover, spatial guidance and temperature monitoring are nowadays available thanks to non-invasive MR-HIFU systems. Among different magnetic resonance thermometry techniques available, Proton Resonance Frequency Shift (PRFS) is by far the most employed, with a precision of approximately 1 • C. Using the MR-HIFU platform different focal points can be steered, heating tissues of various volumes, meanwhile sparing vulnerable and crucial structures (Hijnen et al., 2014). It has to be mentioned that MR-thermometry could be perturbed by MRI contrast agents loaded into TSL, due to relaxivity changes induced by the agent accumulation. Conception of correction methods is therefore urgently needed. To overcome this problem, recently Shin et al. proposed the use of perfluorocarbon nanoemulsions as 19 F MR contrast agent in sonosensitive systems, in order not to have interferences with 1 H based PRFS (Shin et al., 2017). However, it seems quite challenging to obtain temporal information about drug release using fluorinated compounds. MRI Reporter Agents Co-encapsulation of MRI agents into sonosensitive nanovesicles, thus obtaining sonosensitive MRI agents, is of paramount importance in order to monitor and assess in real-time the drug release process. The most diffuse approach consists in encapsulating small hydrophilic paramagnetic compounds, based on Gd 3+ or Mn 2+ ions, in the aqueous core of the nanovesicles. Upon the entrapment the MRI contrast is "silenced, " due to reduced exchange rate of water molecules across the liposomal membrane. When the agent is released, the "quenching effect" is removed, allowing the detection of a contrast enhancement. This contrast enhancement can be estimated and used as reporter of the extent of drug delivery (Ponce et al., 2007;de Smet et al., 2011;Tagami et al., 2011;Rizzitelli et al., 2016). The first TSL encapsulating a MR contrast agent was developed by Viglianti et al. who coloaded MnSO 4 and doxorubicin in liposomes and detected a change in relaxivity upon heat-triggered drug release (Viglianti et al., 2004). De Smet et al. instead, entrapped into TSL both the clinically approved contrast agent Gd-HPDO3A and doxorubicin, demonstrating that: (i) the paramagnetic compound did not affect doxorubicin loading and release, (ii) the drug and the imaging agent were released simultaneously upon heating and (iii) Gd 3+ encapsulation within the aqueous core quenched its relaxivity until release occurred (de Smet et al., 2010). However, in both works release was obtained with non-US mediated hyperthermia. Only one year later, the system conceived by de Smet et al. was exploited for MR-HIFU triggered drug release by Negussie et al. co-workers, who demonstrated that upon HIFU stimulation the releases of the drug and the imaging agent were comparable (Negussie et al., 2011), thus paving the way to real-time drug release estimation by MRI using the paramagnetic compound as doxorubicin surrogate. Similar conclusions were drawn by Tagami et al. who co-encapsulated Gd-DTPA and doxorubicin (Tagami et al., 2011), and Rizzitelli et al. who demonstrated the feasibility of stimulating with pulsed Low Intensity Non Focused Ultrasounds (pLINFU) and tracking by MRI drug release from nanovesicles doped with gadoteridol and doxorubicin (Rizzitelli et al., 2014). In 2013, Han et al. developed a sonosensitive system loaded with doxorubicin and endowed of a newly-synthesized Gd-DOTA-DPPE lipid. These vesicles were intended to track drug delivery by MRI, but were unable to report on drug release as Gd 3+ was incorporated in the vesicle membrane, thus preventing the removal of the so-called "quenching effect" (Han et al., 2013). Original approaches included the use of a dysprosium chelate (Fowler et al., 2013) or Chemical Exchange Saturation Transfer (CEST) contrast agents. Langereis and coworkers loaded into TSL both a chemical shift agent (Tm-HPDO3A, for 1 H lipoCEST detection) and a highly fluorinated compound (hexafluoro-phosphate, for 19 F detection). When the two agents were compartmentalized, liposomes could be visualized solely by the CEST effect due to the influence of the paramagnetic shift agent exerted over 19 F-NMR resonance, but once the release was triggered the lipoCEST contrast enhancement vanished, while the 19 F MRI signal appeared, allowing release quantification (Langereis et al., 2009). Delli Castelli et al. entrapped Eu-HPDO3A in ultrasound-sensitive liposomes, resulting in a "quenched" paraCEST effect at 18 ppm, promptly restored following disruption of the liposomal membrane (Delli Castelli et al., 2014). Finally, iron oxide based contrast agents have been employed. For instance, it has been shown that sonosensitive liposomes coated with hydroxyapatite and entrapping nanodots of SPIO in the inorganic shell can release their content upon application of ultrasounds, inducing changes in T * 2 /T 2 MRI contrast (Liu and Huang, 2011). Lorenzato et al. developed temperaturesensitive magnetoliposomes, encapsulating ultrasmall iron oxide particles (USPIO), displaying significant differences in MRI signal enhancement and relaxivities before and after HIFU stimulation (Lorenzato et al., 2013). In the next paragraph the most promising preclinical applications of sonosensitive MRI agents will be reported. PRECLINICAL STUDIES A direct comparison of the performance of preclinical studies involving sonosensitive MRI agents is not an easy task as they vary not only in terms of liposomal formulation but also in the type of ultrasounds employed to trigger the release (Table 1). Basically, we can gather these works into three groups: (i) HIFU exploited to induce hyperthermia to stimulate both the release of the imaging agent and tissue ablation (no drugs encapsulated); (ii) HIFU employed to trigger the release of both the imaging and the therapeutic agent by hyperthermia; (iii) pLINFU used to trigger the release of the drug and the imaging moiety by mechanical effect. A representative example of the first group is the work reported by McDannold et al. dealing with nonstealth TSL loaded with Gd(DTPA-BMA). The liposomes were injected in rabbits bearing VX2 liver tumors for MR monitoring of thermal therapy carried out with HIFU, obtaining a good match to traditional MR thermometry methods only in the liver (McDannold et al., 2004). The aim was to provide a less motionsensitive technique to monitor temperature in real time. Several works report on the design and testing of TSLs loaded with doxorubicin and Gd 3+ complexes, to track drug release induced by US thermal effect (de Smet et al., 2011;Negussie et al., 2011). In the study reported by De Smet et al. TSL coencapsulating doxorubicin and Gd-HPDO3A were administered to rats bearing subcutaneous 9L gliosarcoma tumors. Local hyperthermia (42 • C) was applied for 30 min through HIFU and drug release was monitored with interleaved T 1 mapping of the tumor tissue, finding out a good correlation between released doxorubicin and Gd 3+ (de Smet et al., 2011). In 2016 Lokerse et al. seeking for an optimized Dox-TSL formulation, found that the prediction of liposomal efficiency based merely on in vitro test is challenging (Lokerse et al., 2016). One year later Hijnen et al. compared different MR-HIFU treatment schemes, assessing that a combination protocol of hyperthermia-induced drug delivery with TSL, followed by ablation, resulted in a homogeneous drug distribution and in the highest therapeutic effect (Hijnen et al., 2017). The only preclinical study using pLINFU was published by Rizzitelli et al. In this work, Gd-HPDO3A was loaded in a Doxil R -like preparation and administered to mice bearing subcutaneous breast cancer. The most relevant aspect of this study dealt with the achievement of complete tumor regression following the injection of liposomes and the application of pLINFU for drug release and cell membrane sonoporation. Worth of note is the fact that the used theranostic agent is made of already clinically approved agents (Doxil R , Gd-HPDO3A). The authors outlined that this approach may offer the possibility of predicting the therapeutic outcome in each patient, simply looking at the MRI contrast enhancement 15 min p.i. (Rizzitelli et al., 2016). The latter aspect, also known as "dose painting, " is of crucial importance in view of a more and more personalized medicine. CONCLUSION This minireview underlines the novelty and the potentiality of the topic, as most of the sonosensitive agents were developed in the last few years. Chemistry plays a fundamental role in the selection of the lipids composing the membrane, in order to boost drug release under US stimuli, while stably trapping the content under physio-pathological conditions in the absence of an external stimulus. It is stressed how special attention has to be devoted to the choice of the encapsulated MRI active compound, that has to act as a quantitative reporter of drug release. Regrettably, making critical comparison is rather difficult at present, due to the lack of standardization of transducers and circuits, often highly customized. However, even if a standardization of the US setting protocol is still lacking, mathematical and practical models are under development in order to predict the efficiency of the various formulations, and the obtained in vivo results appear very promising for a future clinical translation. AUTHOR CONTRIBUTIONS FG organized the database and wrote the first draft of the manuscript; ET contributed to the conception and planning of the review, and revised the work critically for important intellectual content. All authors contributed to manuscript revision, read, and approved the submitted version. FUNDING Consorzio Interuniversitario di Ricerca in Chimica dei Metalli nei Sistemi Biologici CIRCMSB is gratefully acknowledged for the support. The present work was granted by Compagnia di San Paolo, grant agreement n • CSTO165439, entitled MRI-guided Therapy for combating Ovarian cancer.	2018-05-07T17:42:24.903Z	2018-05-07T00:00:00.000	{ "year": 2018, "sha1": "ebcf0414e7e4a093beff639f8009044e9177fd6d", "oa_license": "CCBY", "oa_url": "https://www.frontiersin.org/articles/10.3389/fchem.2018.00157/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "ebcf0414e7e4a093beff639f8009044e9177fd6d", "s2fieldsofstudy": [ "Materials Science", "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
209426044	pes2o/s2orc	v3-fos-license	Experience of parents who have suffered a perinatal death in two Spanish hospitals: a qualitative study Background Perinatal grief is a process that affects families in biological, psychological, social and spiritual terms. It is estimated that every year there are 2.7 million perinatal deaths worldwide and 4.43 deaths for every 1000 births in Spain. The aim of this study is to describe and understand the experiences and perceptions of parents who have suffered a perinatal death. Methods A qualitative study based on Gadamer’s hermeneutic phenomenology. The study was conducted in two hospitals in the South of Spain. Thirteen mothers and eight fathers who had suffered a perinatal death in the 5 years prior to the study participated in this study. In-depth interviews were carried out for data collection. Inductive analysis was used to find themes based on the data. Results Eight sub-themes emerged, and they were grouped into three main themes: ‘Perceiving the threat and anticipating the baby’s death: “Something is going wrong in my pregnancy”’; ‘Emotional outpouring: the shock of losing a baby and the pain of giving birth to a stillborn baby’; “We have had a baby”: The need to give an identity to the baby and legitimise grief’. Conclusion The grief suffered after a perinatal death begins with the anticipation of the death, which relates to the mother’s medical history, symptoms and premonitions. The confirmation of the death leads to emotional shock, characterised by pain and suffering. The chance to take part in mourning rituals and give the baby the identity of a deceased baby may help in the grieving and bereavement process. Having empathy for the parents and notifying them of the death straightaway can help ease the pain. Midwives can help in the grieving process by facilitating the farewell rituals, accompanying the family, helping in honouring the memory of the baby, and supporting parents in giving the deceased infant an identity that makes them a family member. This type of death produces social changes on a family level [19]. Relationships with the older children are affected [20][21][22] because the parents may become overprotective [23] of them or even distant [24]. Children experience feelings of guilt, fear and anxiety [25], which puts a strain on the parents' relationship and increases conflicts between them [15,26]. Perinatal grief is framed within a social context in which a perinatal death is not recognised as the death of a baby [27][28][29]. Research shows that after a perinatal death, parents receive inaccurate information too late [30,31] and receive inappropriate comments from healthcare professionals [32]. In Spain, unlike in other countries, there is no standardised care for families suffering perinatal grief, with a great variability of care in daily practice for mothers and fathers who suffer the death of their baby [33]. Some studies have shown that in Spain many parents whose babies died did not have a chance to say goodbye to them and have no memory of them such as a photograph or fingerprints as they were never offered these possibilities [34]. Worden's model defines grief as an adaptation in which "tasks" must be completed by the person who suffers a loss. These "tasks" are: to accept the reality of the loss; work through the pain of the baby's death; adjust to an environment in which the deceased baby is missing; and find an enduring connection with the deceased baby while embarking on a new life [35]. This way of approaching grief gives those suffering a loss an active role in their mourning [36]. Literature on the subject shows both a scarcity of clinical guidelines [37] and the need for more evidence on the phenomenon from the perspective of the mothers and fathers of the deceased baby [38]. Aim The objective of this study is to describe and understand the experiences and perceptions of mothers and fathers who have suffered a perinatal death. Design and setting A qualitative study based on Gadamer's hermeneutic phenomenology was designed. For Gadamer, human experience cannot be understood without language [39]. Understanding participants' stories requires being prepared to be told something through a dialogue, from which meanings emerge. The development of the study followed the phases of a Gadamerian-based research method [40]: 1. To decide whether the research question is pertinent to the methodological assumptions. Perinatal loss is a phenomenon of the lifeworld that can be understood from the perspective of hermeneutic phenomenology. Gadamer's philosophical approach is appropriate for comprehending the experiences of parents who have suffered a perinatal death. 2. To identify the researchers' pre-understanding of the topic (Reflexivity). The pre-understanding of the researchers came from their clinical experience as healthcare professionals who work or have worked in a delivery room as midwives. The study was carried out in Torrevieja Hospital and Vinalopó Hospital, in Alicante, Spain. They are both public hospitals and have an average of 1400 births per year. Population and participants A convenience sample of parents who had suffered a perinatal death was recruited. The histories of perinatal deaths in both hospitals over the last 5 years was consulted. A total of 63 perinatal deaths took place in the hospitals in the 5 years prior to the study. Mothers and fathers who fulfilled the inclusion criteria were called, and the aim of the study was explained to them. They were told that their participation was voluntary and they could choose not to reply or abandon the study at any time. It was explained that the data obtained would be treated confidentially and in accordance with European data protection regulations. The inclusion criteria were: (1) Being a mother or father who has suffered a loss through the perinatal death of their child, from the 22nd week of gestation to the first week of life. (2) The death occurring between 3 months and 5 years prior to the study was decided on as the memory of a very recent loss could be very painful. This would provide a broad sample and ensure that the experience would be remembered with sufficient detail and intensity. (3) The mother or father had to speak Spanish or English. (4) Signing the informed consent form. A total of 13 mothers and 8 fathers were interviewed (see Table 1). Of the 63 cases of perinatal death registered in this period, eleven did not answer the telephone, eight did not speak Spanish or English, nine refused to discuss the subject, and nine claimed they did not have time to be interviewed. Data collection Open, in-depth interviews were used for data collection. The interviews took place between April 2016 and May 2017. The parents were contacted and invited to take part by the main researcher. The interviews were carried out by two researchers who were midwives, and one who was a paediatric nurse. Each interview lasted between 45 and 60 min, and they were audio recorded. The interview started with an open-ended, general question: "What was the experience of losing your baby like?" Subsequently, other questions were asked following the protocol used and based on the natural flow of the conversation. The final question in each interview was: "Do you have anything else you would like to add?" The interviewer also took note of non-verbal signs throughout the interview. During the interviews, participants were reminded of ethical issues and were told that recounting their experience could be helpful in improving care for other parents in their same situation, times were respected, and the emotional and psychological needs of the mothers and fathers were met, for example, by pausing or letting them express their emotions. When the researchers perceived that no new issues were emerging and the same topics were being repeated, it was considered that the saturation of the data had been reached, and the data collection was concluded. Data analysis The following steps were followed for the data analysis [40]: 1. To achieve understanding of the topic through dialogue with the participants. During the interviews, a spontaneous clarification regarding what the participants discussed was achieved by using follow-up questions (e.g., "Could you tell me what you mean when you say that the gynaecologist had a cold manner?") 2. To conduct a conversation between the researchers and the participants through the text. Each transcription was analysed line by line in order to identify meaningful and important phrases and select them as quotes. Each quote was assigned a code that captured its meaning, grouped into units of meaning, sub-themes and themes. Data coding was performed individually by three researchers. They then compared their interpretations so each unit of meaning, theme and subtheme were agreed upon by consensus. Rigour In order to ensure the rigour of the study, the participants were given the opportunity to confirm the transcriptions, units of meaning, sub-themes and themes by reading their answers. Additionally, all the participants' experiences were represented. The study's credibility was complemented by the triangulation of the researchers. Results Eight sub-themes emerged, and they were grouped into three themes that help us to understand the experience of parents who have suffered a perinatal loss ( Table 2). Mothers in the study often had a hunch that something was going wrong with their pregnancy. This feeling might have been related to the difficult process of assisted reproduction, a medical history of miscarriages, a high-risk pregnancy, or noticing a decrease in foetal movements. When a perinatal death is sensed in this way, mothers start assimilating the death right away. "… I was worried, I got the feeling that I was exaggerating it too much (…), but then I started to assimilate the fact that it would end badly" (P-10). "It could end badly". Medical history as a threat and a source of uncertainty Some parents have a long history of infertility treatments or miscarriages. According to the participants, pregnancies achieved through rigorous clinical procedures can leave an emotional footprint on couples, increasing their fears. Having such a background seemed to help parents to quickly assume the possibility of a loss and usually make them more aware of the fragility of the pregnancy. "I had been trying to have children for six years straight. My husband and I would get up at 2:00 a.m. for the treatments, (…) it was very stressful and I felt physical and psychological fatigue. After that, you're always worried that something's not right" (P-5) Some of the pregnancies were considered high-risk because of pre-existing conditions in the mothers such as hypertension or diabetes. These conditions made parents experience the pregnancy in a distressing way. The parents reported frequent visits to emergency services upon noticing any minor issue and having requested follow-up ultrasound scans in order to ease their feelings of uncertainty. "The doctors diagnosed me with a high-risk pregnancy (…). I had to get an ultrasound every month, but on top of that, I often went to the hospital, for aches, pains, bleeding (…). I was afraid something would happen" (P-12). Bureaucracy and administrative language as obstacles in the mourning process. Administrative slowness, misinformation about administrative processes, inappropriate language, referring to the baby as a foetus, denying registry, denial of the existence, identification as a deceased baby. Anticipating the death. From suspicion to confirmation The warning signs of a perinatal death consist mainly of changes in the foetus' movement patterns (from a decrease to a complete absence of movement). This change in movement patterns caused great uncertainty for the mothers that took part in the study. "(…) it had been a few days since I had felt any movements (…). Since I already had an appointment for foetal monitoring, I didn't want to go to the emergency room in case I was overreacting, I didn't want to bother them although I was really worried" (P-14). Some mothers recalled pain and periodic or isolated contractions that they associated with cramping. Sometimes, these contractions led to active labour. In such cases, the physical pain of labour was compounded by the emotional pain and the feeling of helplessness when losing a baby and not being able to do anything to stop it. "The cramps that I felt (…) weren't periodic or regular. I had some cramping, later a little more, then they told me that I had been dilating and my cervix had effaced (…) At that point, there was choice but to go ahead, and that made it even harder, if that's even possible" (P-13). Theme 2. Emotional outpouring: the shock of a baby's death and the pain of giving birth to a stillborn baby The confirmation of a baby's death is the start of a long, hard journey that makes up part of the parents' grief process. The emotional outpouring is produced once the death has been confirmed. The expectations the parents had for their baby are no longer present, and the pain of the loss increases due to uncertainty or loneliness. There were cases in which the babies were born alive and lived for a few hours or days, which resulted in even stronger attachments as well as more intense shock over the subsequent death of the baby. "You feel terrible, devastated, you go in there thinking you're going to be a father, and suddenly your baby is dead" (P-11). Emotional shock upon notification of the baby's death Physicians often inform the parents whose babies passed away in critical care units. Parents recalled a strong emotional shock because, despite the warnings of a high-risk situation, they still hoped that they would be able to take their healthy baby home. " The doctors told me that the next few hours were really important in his development (…) they called us on the phone, and of course, my heart started racing. They told us to come quickly because the baby was worse (…), your world falls apart, and you lose all hope" (P-10). Parents recalled that they felt disbelief, anguish, anger, emptiness, insurmountable pain, and outrage about the loss of their baby. This represents the first step in working through their grief. If the baby's death occurs before birth, the notification of the death can be an extremely delicate situation. The participating parents recounted how they felt when doctors confirmed there was no foetal heartbeat. "I think the gynaecologist should have waited until my husband was there and have given us the news in a different way, but she started to check me with the sonogram, she looked at me and said: 'No, I'm sorry, she doesn't have any vital signs, she's dead.' Just like that " (P-2). According to the parents in the study, non-verbal language can be vital when receiving this devastating news: silence and the facial expressions of the gynaecologists and midwives can tell much. The delay in providing information and the scarcity of explanations also increased parents' suffering. "The gynaecologist made a strange face, and I said, "Something's wrong, isn't it? (…) Please tell me everything's ok." She didn't say anything, but her face said it all." (P-6). In some cases, doctors consulted with a colleague before speaking to parents, which increases the waiting time and causes anxiety, tension, and uncertainty. "I could tell they were not saying anything to me during the ultrasound, then they called the other doctor in, and I got really nervous. And when I saw their faces, I asked, "What is it, is he dead?" because they weren't telling me anything" (P-2). Giving birth to a stillborn baby: a doubly painful labour process for parents When the death occurred before birth, the pain of the loss was increased by the physical labour pain of giving birth to a stillborn baby. In all cases, the mothers and fathers were advised to try a vaginal birth. One father asked for the caesarean section to avoid an even more difficult situation. "I didn't want to see my son born dead, … it broke my heart (…) My wife and I thought that it would just draw out the situation and create unnecessary suffering" (P-18). The feeling of losing the baby, the pain of childbirth, and the fact that midwives may not be immediately available lead to feelings of anger and neglect in some parents. "(...) The pain was unbearable. I had called several times, and no one came until I finally went out into the hallway and yelled to get the midwife. The nurse's aide told me that he was busy, and I responded angrily: "I don't care where he is, but where he should be is in this room with my wife!". (P-21). During such an unpleasant and emotional situation, receiving individualised care from midwives and physicians could become the most important source of comfort for parents. "I was impressed by the delivery room midwife, … the warmth with which she treated my wife. She let me stay with her, she held her hand and spoke to her gently … at such a hard time, that sort of personal treatment was comforting, and even today we still remember it as the most positive thing about that sad experience". (P-21). Loneliness and lack of information as aggravating factors in the pain of the loss The participating parents described various elements that made the grief process even harder. Some mothers were alone at the time they were notified about the death as their partners were not allowed to enter the ultrasound room. "Getting the news without having my husband there, you feel helpless and alone, (…) think about if someone told you something like that without anyone there, even to put their hand on your shoulder to console you, it's shocking, right? It's just wrong". (P-12). Even when surrounded by many people, mothers sometimes felt just as lonely as they did when they were physically alone. As one mother pointed out, making parents who have lost a baby share the same hospital room with other parents that have given birth to a healthy baby increases suffering. "(…) I wasn't in the mood to be there, in the same room where you can hear newborn babies in their cribs, and bottle carts go by, it was frustrating." (P-3). Another element that added to the suffering was not knowing the cause of death or not having a clear explanation about the causes. "They should tell you things as they really are [with emphasis]. (…) But no, instead they tell me that my baby was born tired (…) [Pause]. This is the story the paediatricians told me, that my child was born tired. How can you say that, in that way?" (P-1). The results of the autopsy usually take a long time to come out. This prolonged uncertainty and prevented parents from achieving closure and moving on to the process of readapting to the world without their deceased baby. "On top of that, they make you wait months to get the autopsy back, and waiting such a long time with that same anguish and uncertainty doesn't let you really live or move on." (P-6). On most occasions, the autopsy is inconclusive and does not clarify the cause of death. This deepens parents' suffering and prevents them from getting over feelings of guilt. Parents search for answers to questions such as: What did I do wrong? Why didn't I go to the doctor sooner? The parents demand more information to understand the reasons behind what happened in the hope of finding relief for their pain and getting rid of their feelings of guilt. "He told me, "Everything was fine." And I said, "What do you mean, everything was fine?" (…) That's the explanation I was given (…) Can you believe that? All I wanted to know was the reason why and that I wasn't to blame." (P-14). "I had been waiting for that appointment with the gynaecologist for such a long time..., to continue without answers. It would have helped me to understand why that had happened to me" (P-15). Theme 3. "We have had a baby": the need to give the baby an identity and legitimise grief This theme reflects the parents' need to give their deceased baby an identity as a member of the family and recognise their grief process. Parents valued having the opportunity to say farewell, keep memories of their baby and participate in mourning rituals based on their beliefs. Saying goodbye to the deceased baby, having the baby's footprint, keeping the memory of the baby alive The participating parents valued the gynaecologists' and midwives' efforts to help them to say goodbye to the baby: allowing physical contact, keeping the umbilical cord clamp and the baby's clothes or foot/handprints. In line with our theoretical framework, this eases the task of establishing an enduring connection with the deceased baby. "We appreciate the fact that we were allowed to be with him, see him, touch him (…) it was very hard, but we had to say our last goodbyes because even though he wasn't born alive, he was still our son, and he would be forever." (P-11). "There are people that think that me having pictures of my son on my phone is gruesome. I don't sit there looking at the photo all day, but if I need to, I know I can look at it, and that helps" (P-5). Some parents got distressed because some midwives did not offer them the possibility of keeping some mementos of their babies and recommended not seeing the baby. Not having time to say goodbye leads to suffering. "My daughter was beautiful and (…) they wrapped her up and took her away immediately, they didn't let me see her or hold her at all (…) and I told them, "Wait!" and they didn't wait. They told me it wasn't good for me to spend a long time with her." (P-17). Mourning rituals. The importance of respecting individual beliefs Regarding post-mortem rituals, participating mothers and fathers agreed on the need for each set of parents to be able to act according to their beliefs. The possibility of baptising babies was something that was highly valued among Christian families as it made the process more bearable. However, when the gynaecologists, midwives and priests refused this possibility, the parents suffered because of the lack of legitimate recognition of their babies. "Couldn't they have offered us the possibility of baptising her?...What fault did the baby have that she had left this world so soon, so little? The priest told us that we couldn't have a mass for her, it's not fair that he wouldn't do anything for her, because she was and will continue to be my child." (P-4). Such rituals, as well as choosing a final resting place can help parents to make a long-lasting connection with their babies. It is important that all parents decide how to proceed with the corpse according to their beliefs and needs. "I picked up the ashes, and that was it, and he's with me. In the summer, I go to a different house, and I take my puppies and my son's ashes (...) it's just something that I need (…) to know that he's there." (P-19). "Every week or fortnight, I go up to the cemetery, nobody can take away those five minutes I have with her. (…) Being able to go up there and be with my daughter puts my mind at ease." (P-8). Others parents did not worry about the religious rituals or the final resting place of the baby. Especially in the case of gestational stillborn children, where the hospital took care of the foetus. They felt it re-created the pain, and they avoided the topic completely. "That (stillbirth) was a painful experience, and I don't need to have the remains of my baby to remember her. I remember the experience, but the baby no, because I didn't get to meet her, they didn't give us her body. My wife thinks about that more than I do, but neither of us wishes we had had a ceremony or had kept her remains" (P-21). Bureaucracy and administrative language as obstacles in the mourning process The majority of parents recalled the administrative proceedings after the death to be a slow process. They express that greater diligence and better information on the steps to follow would have made this situation easier. "They should tell you what to do when your baby died, that you have to go to the funeral parlour, what papers they're going to ask you for. They should direct you and guide you … " (P-6). The parents call for more appropriate language when referring to their deceased baby during the administrative processes. Referring to the baby as a "foetus" is considered derogatory and generates pain and suffering. "There were a few things that we didn't like … for example, (in the report) it said 'foetus'. That was pretty painful. No, for us, it's not a foetus (…) it's our baby" (P-20). Furthermore, if the baby was not born alive, parents could not register him/her in the civil registry as a member of the family unit. This takes away the baby's identity and reduces the legitimacy of the grief. "Since he wasn't alive for one minute, neither his birth nor death could be registered, and I just broke down. When I went to the registry, I started crying (…) it's as if my son didn't exist" (P-17). Discussion The objective of this study was to describe and understand the experiences and feelings of mothers and fathers who have suffered a perinatal death. The study findings suggest that some parents perceive "there is something wrong with the pregnancy" and anticipate the foetus loss. This could be influenced by either previous experiences or physical signs that make them suspect. The baby's death triggers an emotional outbreak characterized by shock, pain and sharpened by the feeling of loneliness. Parents need to legitimize grief, giving identity to the dead baby, saying goodbye to him and participating in rituals according to their beliefs. Parents start thinking about a possible perinatal death when they notice a decrease or absence of foetal movement [41][42][43]. However, some women simply had a feeling or a hunch, which may be explained by the strong physical, emotional and spiritual connection between a mother and her baby [44][45][46]. The study conducted by Erlandsson et al. [47] points out the importance of educating mothers on typical intrauterine movement patterns and the gut feelings that a mother gets when she knows something is wrong [31]. These suspicions have been interpreted as a predictor for a death, which precedes the acceptance of such [34]. The majority of intrauterine deaths recounted in our study were diagnosed using ultrasound, which the participating mothers and fathers remember as an agonising and sad process [42]. Parents feel that the process of being informed takes too long, and they tend to assume something is wrong as soon as they see the non-verbal language of healthcare professionals conducting the ultrasound [48,49]. Research suggests that notifying the death in a clear manner is best, taking care to avoid patronising parents and maintain respect for the individual's preferences [50][51][52]. Consistent with other studies, the response of mothers and fathers to a perinatal death is characterised by feelings of anger, shock, disbelief, denial, despair and hopelessness [53,54]. These feelings are stronger when women were not accompanied by their partners at the time of receiving the news of the death. In these cases, feelings of fear, anguish, and dissatisfaction with the care arise [55]. The results point to the need to restore the role of the husband, whose pain during a perinatal death is often relegated to a secondary role by society and healthcare professionals [56]. Our participants found any information regarding the cause of death to be useful, helping them to free themselves from feelings of guilt. However, they often do not find out what the actual cause was [57,58]. In these cases, literature on the topic underscores the importance of healthcare professionals' support [59]. Certain practices that ease pain amongst the mothers in our study coincide with those described in other studies; these practises are: seeing the baby; making time to say goodbye, and keeping mementos such as photographs, footprints or clothing [60,61]. Participants agree that being encouraged to perform mourning rituals and keeping memories of their baby is a positive idea [9,62]. In the case of parents not wishing to do so, healthcare professionals should respect their decision [61]. Furthermore, healthcare professionals should encourage parents' intimacy and avoid placing them in the maternity area as if they were parents of healthy babies [12,63,64]. The absence of care protocols can lead to the refusal or opposition to farewell rituals [17]. Consistent with other studies [65][66][67], the need to give the deceased baby an identity as a member of the family is clear. Referring to the baby as a foetus in the medical reports and denying parents the possibility of registering the baby in civil registries contributes to parents' frustration and "disenfranchised grief" [28]. One of the study's limitations is in its sample selection. The participants recruited for this study had diverse life experiences; some of them had had experiences of miscarriages or infertility whilst others had not. This could have influenced our results, particularly in theme 1. Concerning the data collection, despite having performed individual interviews and reaching data saturation, using focus groups would have contributed to the interaction of the participating mothers and fathers and could have enriched the results. Another limitation is the lower number of men in the sample since they refused more to talk about the subject or were busy with work. A third limitation was the inability to interview non-Spanish and English-speaking mothers and fathers as there were quite a few parents with other languages such as Arabic, Russian or Romanian who could not be interviewed due to the language barrier. Conclusions The findings of this study show that parents who face a perinatal loss tend to anticipate their baby's death as the often perceive that "something was wrong with their pregnancy". This study also shows that the shock of losing a baby and the pain of giving birth to a stillborn baby triggers an emotional outpouring in parents, who feel that they "have had a baby" and need to give him an identity to legitimate the grief. The participating parents have had suspicions, based on their clinical background and anticipated the death of their children before the signs of alarm and foreboding. The participating fathers and mothers report an emotional shock at the death of their baby, which, in the case of antepartum death, is increased by the physical and emotional pain of giving birth to a dead child. Loneliness, lack of empathy and information deepened the feeling of pain from the death of the baby. Parents express the need to give identity to their dead baby and legitimacy to their grieving process. Saying goodbye to the baby and carrying out rituals according to the beliefs of each person favours the grieving process, but the bureaucracy and the clinical-administrative language (of the reports) worsen the feeling of pain. For future research, taking into account the social and family implications that affect perinatal grief, it would be of interest to conduct interviews with family and friends of those who have suffered a perinatal death. This could contribute to understanding the character of this "disenfranchised grief" [27], which participants in this study describe. Acknowledgements We are grateful to all participants for their contribution to the study and sharing their experiences with us. Thanks to the Research Group Health Sciences CTS-451, from the University of Almería, for their support. Authors' contributions MC-A collected and analysed data and wrote the article. CF-S designed the study, analysed data and wrote the article. FRJ-L analysed data and reviewed the study. JG-M designed the study, analysed data and critically reviewed the manuscript. IMF-M collected data and critically reviewed the manuscript. LM-A collected data and critically reviewed the manuscript. JMH-P participated in the data analysis, drafting and critical review of the manuscript. All authors read and approved the final version of the manuscript. Funding Not applicable. Availability of data and materials The datasets generated and analysed during the current study are not publicly available due technical issues. Data is based on recorded interviews and their transcriptions are only available for ATLAS.ti software, but are available from the corresponding author on reasonable request. Ethics approval and consent to participate The study was approved by the Ethics and Research Committee of the Nursing, Physiotherapy and Medicine Department of the University of Almería (protocol number 1401-3/16). Following the Declaration of Helsinki, participants received written information about the research process and objectives. Before commencing the study, written informed consent and permission to record the interview were obtained. Confidentiality and anonymity of all participants were ensured throughout the study. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests.	2019-12-21T14:04:21.096Z	2019-12-19T00:00:00.000	{ "year": 2019, "sha1": "cb59c31dccdd529e919260925f0cdae5b5cf647b", "oa_license": "CCBY", "oa_url": "https://bmcpregnancychildbirth.biomedcentral.com/track/pdf/10.1186/s12884-019-2666-z", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "f683e01ab08b6ca658e035a0815ba91ba537b13b", "s2fieldsofstudy": [ "Medicine", "Sociology" ], "extfieldsofstudy": [ "Medicine" ] }
233175976	pes2o/s2orc	v3-fos-license	Rapidly Improving High Light and High Temperature Tolerances of Cyanobacterial Cell Factories Through the Convenient Introduction of an AtpA-C252F Mutation Photosynthetic biomanufacturing is a promising route for green production of biofuels and biochemicals utilizing carbon dioxide and solar energy. Cyanobacteria are important microbial platforms for constructing photosynthetic cell factories. Toward scaled outdoor cultivations in the future, high light and high temperature tolerances of cyanobacterial chassis strains and cell factories would be determinant properties to be optimized. We proposed a convenient strategy for rapidly improving high light and high temperature tolerances of an important cyanobacterial chassis Synechococcus elongatus PCC 7942 and the derived cell factories. Through introduction and isolation of an AtpA-C252F mutation, PCC 7942 mutants with improved high light and high temperature tolerances could be obtained in only 4 days with an antibiotics-free mode. Adopting this strategy, cellular robustness and sucrose synthesizing capacities of a PCC 7942 cell factory were successfully improved. INTRODUCTION With resource shortages and environmental pollution issues becoming increasingly prominent, photosynthetic biomanufacturing provides important options for the green and sustainable production of biofuels and biochemicals (Melis, 2009;Lu, 2010). Cyanobacteria are potential photosynthetic platforms and have been successfully engineered for production of multiple natural and non-natural products (Oliver and Atsumi, 2014). To put cyanobacterial photosynthetic biomanufacturing technology into practice, high light and high temperature tolerances would be important properties of the cyanobacterial cell factories, aiming to facilitate stable growth and production under stressful conditions during scaled outdoor cultivations (Luan and Lu, 2018). To improve high light and high temperature tolerances of cyanobacterial cells, complex genetic engineering strategies have been explored and adopted to modify or to update the native stress-response protective systems or the photosystems. For example, overexpression of specific heat shock proteins improved high temperature tolerance of cyanobacterial cells (Chaurasia and Apte, 2009;Su et al., 2017), and reduction of light harvesting antenna led to improved tolerances to high light stress (Kirst et al., 2014). However, implementations of these strategies rely on antibiotic resistance selection based on genetic manipulations and require long-term cultivations and passages for isolation of the homozygous mutant, and the entire cycle would usually take several weeks. Previously, we have demonstrated that specific point mutations in F O F 1 ATP synthase α subunit (AtpA), converting the 252 nd amino acid from cysteine to any of the four conjugated amino acids (tyrosine/phenylalanine/tryptophan/histidine), could endow PCC 7942 cells with high temperature and high light resistances (Lou et al., 2018). Inspired by this, we proposed a convenient strategy for rapidly improving high light and high temperature tolerances of PCC 7942 derived cell factories by targeted mutagenesis of AtpA-C252. Compared with previously reported approaches, this new strategy has two potential advantages. First, the direct stress selection (with high light and high temperature) process would eliminate the dependence on introducing and screening antibiotics markers in the recombinant strains. Second, the growth advantages of the desired transformants would reduce the time required for cultivation and segregation. For proof-of-concept, we successfully isolated the PCC 7942 mutant strains with improved high light and high temperature tolerances that could be obtained in only 4 days in an antibiotics-free mode. In addition, cellular robustness and sucrose synthesizing capacities of a previously constructed PCC 7942 cell factory were significantly optimized, increasing the sucrose productivities by nearly onefold. Overview To facilitate rapid mutagenesis of AtpA-C252 and effective isolation of the mutants, we designed a three-step procedure, consisting of transformation, initial screening, and rescreening. As shown in Figure 1, plasmids carrying tailored AtpA fragments would be transformed into PCC 7942 to induce homologous recombination. During the initial screening step, mutant cells obtaining the AtpA-C252 mutagenesis would survive under the selective conditions and form colonies on the agar plates. To exclude false positive results and to confirm the tolerant phenotypes of the mutants, the colonies obtained from the initial screening step would be streaked and cultivated on fresh agar plates under the same selective conditions. The final colonies would be collected for further assays and evaluations. Rapid Isolation of PCC 7942 Mutants Carrying the AtpA-C252F Mutation Among the four AtpA-C252 mutations endowing PCC 7942 with improved tolerances to high light and high temperature (Lou et al., 2018), we selected AtpA-C252F for proof-of-concept of the proposed strategy in this work. A plasmid carrying an AtpA-C252F gene fragment (termed as pSS18 in Figure 1) was constructed and transformed into PCC 7942 cells, and another plasmid carrying wild-type (WT) AtpA was used as a control. PCC 7942 tolerant transformants appeared on BG11 agar plates after only 48 h of cultivation in the selective conditions with high light and high temperature. More than 1,000 transformants were obtained on the screening plate of PCC 7942 (transformed with pSS18, AtpA-C252F), whereas no transformants appeared on PCC 7942 transformed with pSS3 (AtpA WT). Phenotypes of the transformants were further checked by re-screening under the same conditions, and 23 randomly selected transformants (from the initial screening step) all survived in the rescreening step. We further collected the 23 transformants (after the two-round selection) for AtpA sequence analysis, and the Sanger sequencing results revealed all of the strains carrying the AtpA-C252F mutation, indicating that the tolerances to high light and high temperature were endowed by the targeted mutagenesis. In addition, the sequencing results showed that all the mutants carrying the AtpA-C252F mutation existed as homozygous (as shown in Supplementary Figure 1), indicating that the significant growth advantages caused by the AtpA-C252F facilitated effective segregation of the mutated chromosomes. High light and high temperature tolerances of the isolated mutants were further evaluated by liquid cultivation in column photobioreactors. As shown in Figure 2A, the growth of the PCC 7942 mutants carrying the AtpA-C252F mutation was similar to that of the control under normal conditions (30 • C and 50 µmol photons/m 2 /s). While in a stressful environment (44 • C and 400 µmol photons/m 2 /s), the mutant strain exhibited significantly improved adaptabilities compared with the WT control and maintained rapid growth ( Figure 2B). WT cells of PCC 7942 could not survive or grow facing the environmental stress, whereas OD 730 of the AtpA-C252F mutant cells reached up to 7 with a significantly improved growth rate compared with when cultivated in normal conditions (30 • C and 50 µmol photons/m 2 /s). The results indicated that the mutant strain of PCC 7942 obtained through the rapid isolation process displayed significantly improved capacities to tolerate and utilize strong illuminations even when cultivated at high temperatures. AtpA-C252F Mutagenesis Rapidly Improves High Light and High Temperature Tolerances and Sucrose Production Rates of a PCC 7942 Cell Factory We further adopted this strategy to engineer a sucrose synthesizing cyanobacterial cell factory. Previously, we have introduced an Escherichia coli sourced sucrose permease CscB into PCC 7942 to facilitate the secretory production of sucrose under salts stress and overexpressed the native sucrose-phosphate synthase Sps to enhance sucrose synthesis (Duan et al., 2016;Qiao et al., 2018). The final strain FL130 was transformed with the pSS18 plasmid (AtpA-C252F) to confirm the feasibility and effectiveness of our strategy when adopted on cell factories. Following the described procedures, we successfully isolated high light and high temperature tolerant transformants of FL130 ( Figure 3A). The final isolated and verified FL130 mutant (AtpA-C252F) was termed as SZ41, which was able to grow under the conditions of 45 • C with 300 µmol photons/m 2 /s illumination ( Figure 3B). When cultivated under normal conditions of 30 • C and 100 µmol photons/m 2 /s, growths FL130 and SZ41 show minor differences ( Figure 3C, FL130-NTL and SZ41-NTL), and the sucrose production reached up to 1,835 and 2,025 mg/L in 8 days, respectively ( Figure 3D, FL130-NTL and SZ41-NTL). The slight increase of total carbon fixation (the sum of biomass and sucrose) in SZ41 might result from an increased oxygen evolution rate (Figure 4A), indicating that the introduction of AtpA-C252F brought in benefits on the overall efficiency or stability of cellular photosynthesis in PCC 7942 under these conditions (30 • C, 100 µmol photons/m 2 /s, 150 mM NaCl). When stressful high light and high temperature conditions (40 • C and 400 µmol photons/m 2 /s) were adopted, SZ41 exhibited significantly improved growth advantages over FL130. After 4 days of cultivation supplemented with 150 mM NaCl, the cell density of the SZ41 culture broth reached about 3, whereas that of the FL130 was lower than 2.0 ( Figure 3C). The bleaching phenotypes might be caused by the synergy effects of high light and high temperature stress and salts stress on the Synechococcus cells, and during this process, 1.8 and 0.98 g/L of extracellular sucrose were synthesized by Frontiers in Microbiology \| www.frontiersin.org SZ41 and FL130, respectively, under the high light and high temperature conditions ( Figure 3D). The different performances on cell growths and sucrose production between the two strains under the stressful conditions are in accordance with the photosynthesis activities differences revealed from the oxygen evolution rates. With enhanced illumination strengths (and increased temperatures), photosynthesis activities of FL130 and SZ41 were both significantly elevated compared with these under normal conditions, and the oxygen evolution rate of SZ41 would be 40% higher than that of FL130 ( Figure 4A). It is noteworthy that the specific sucrose production on per cell levels was on similar levels between the two strains ( Figure 4B) whether in normal or stressful conditions, indicating that the improved accumulation of sucrose in the SZ41 strain resulted from the optimized cell growth and enhanced carbon fixation caused by the AtpA-C252F mutation, rather than a distribution of a larger portion of biomass to sucrose synthesis. After 4 days of high light and high temperature cultivation, bleaching phenotypes would be observed for both the SZ41 and FL130 cells, and the sucrose production would be terminated. However, it is noteworthy that the AtpA-C252F mutation carrying the SZ41 strain under the stressful conditions could synthesize similar concentrations of sucrose utilizing half of the cultivation term as that obtained under normal conditions (1.8 g/L in 4 days versus 2.0 g/L in 8 days). In addition to sucrose, the SZ41 strain also accumulated higher intracellular glycogen contents than the FL130 under high light and high temperature conditions, confirming the improved capacities for carbohydrates synthesis caused by the AtpA-C252F mutations ( Figure 4C). Both sucrose and glycogen biosyntheses depend on the precursor glucose-1-phosphate, and it has been reported that the metabolic flux through glucose-1-phosphate significantly determines the flexibility of the intracellular carbon distribution of cyanobacteria via a dynamic balance between different metabolic branches . Furthermore, it has been reported that S. elongatus UTEX 2973, the fast-growing strain carrying AtpA-C252Y single nucleotide polymorphism (SNP) compared with PCC 7942, would overaccumulate glycogen to buffer the enhanced carbon flux from the CBB cycle (Song et al., 2016;Tan et al., 2018). Glycogen storage serves as the main carbon sink mechanism of cyanobacterial cells and could play the role of carbon pool for synthesis of desired metabolites; thus, the enhanced glycogen synthesis activities and glycogen contents could be an additional advantage of the strategy developed in this work. We also evaluated the effects of this strategy on engineering an ethanologenic cell factory derived from PCC 7942. A previously optimized ethanologenic pathway consisting of the pyruvate decarboxylase from Zymomonas mobilis and a type II alcohol dehydrogenase from Synechocystis sp. PCC 6803 was introduced into PCC 7942 to generate the ZN44 strain, producing about 0.22 g/L ethanol in 2 days (Supplementary Figure 2). Introduction of the AtpA-C252F mutation into ZN44 significantly improved the growth of the ethanologenic cell factory (ZN45) under high light conditions (OD 730 of 7.2 in ZN45 compared with OD 730 of 5.1 in ZN44 after 2 days of cultivation). However, the ethanol production was just slightly improved from 0.22 to 0.25 g/L, indicating that as for the PCC 7942 derived ethanologenic cell factory, the activities of the ethanol synthesis pathways rather than the total were holding control over the ethanol production capacities, and the increase of carbon fixation could not be effectively rewired into ethanol synthesis, which could be solved through further metabolic engineering. Compared with the recently reported fast-growing and robust cyanobacteria S. elongatus UTEX 2973, PCC 7942 still possesses the advantages of clear genetic background and more convenient genetic manipulation, and dozens of photosynthetic cell factories have been engineered based on this typical model strain. The strategy developed in this work provided a convenient approach to update the related PCC 7942 derived cell factories, which would avoid repeated implementations of the complex metabolic engineering with a new chassis. In addition, the success in this work provided a novel targeted mutagenesis and selective isolation manipulation paradigm for cyanobacterial phenotypes improvements. In recent years, many SNPs have been identified to be related with cellular growth or survival benefits, e.g., acid tolerances (Uchiyama et al., 2015), alcohol tolerances (Hirokawa et al., 2018), high light and high temperature tolerances (Lou et al., 2018), and growth rates (Ungerer et al., 2018). Regarding such mutations endowing cells with specific growth advantages, rapid and convenient pSN44-slr1192-r CTAATTTTTACTATGGCTGA Amplifying the Sp r -P rbcL -pdc ZM -slr1192 DNA fragment pSN44-sp-f GCATGCCCGTTCCATACAGA Amplifying the Sp r -P rbcl -pdc ZM -slr1192 DNA fragment transplantation to photosynthetic cell factories could be expected through similar procedures developed in this manuscript, aiming to optimize the growth or tolerance properties. In addition, to some gene deficiency-related growth advantages (meaning benefits caused by loss of specific gene functions) (Joseph et al., 2014;Kirst et al., 2014), this strategy could be expected to work and facilitate the rapid isolation of the gene-deficient mutants. However, it is still noteworthy that the application of this strategy to other targets might be limited by the significant degrees of the growth advantages endowed by specific mutations, which might not be as effective as that from AtpA-C252 (stress tolerances coupled with growth advantages). To this end, systematic optimization of selective pressure strengths and methods would be required in order to achieve the ideal effects of strain improvements. CONCLUSION High light and high temperature tolerances are important properties for cyanobacterial cell factories, aiming to put the photosynthetic biomanufacturing technology into practice. We developed a rapid and convenient strategy for improving cellular tolerances to high light and high temperature in an important cyanobacterial chassis PCC 7942 derived cell factory by antibiotics-free introduction of an AtpA-C252F mutation. Adopting this strategy, cellular robustness and sucrose synthesizing capacities of the PCC 7942 derived cell factory were significantly improved. Chemicals and Reagents All reagents were purchased from Sigma-Aldrich (United States). Taq and pfu DNA polymerases for PCR were purchased from Transgene Biotech (Beijing, China). T4 DNA ligase and restriction enzymes were purchased from Fermentas (Canada) or New England Biolabs (Japan). The kits for molecular cloning were from Omega Bio-tek (Norcross, GA, United States). Oligonucleotides were synthesized, and DNA sequencing was performed by Tsingke (Qingdao, China). Construction of Plasmids and Strains Escherichia coli DH5α (TaKaRa, Dalian, China) was used as a host for cloning. The strains constructed and used in the present study are listed in Table 1. All the constructed plasmids employed pUC19 as backbone. To construct pSS3 and pSS18, AtpA-encoding gene containing 252C/252F was cloned from the chromosome of PCC 7942 and PCC 7942-C252F via PCR using the primers AtpA-UP500-F-NdeI and AtpA-Down500-R-EcoRI listed in Table 2. PCR products were purified using the E.Z.N.A Cycle Pure Kit (Omega Bio-Tek, Norcross, GA, United States), and then the AtpA/AtpA-C252F fragments were digested with NdeI and EcoRI and inserted into pUC19 for constructing the target plasmids pSS3 and pSS18. To construct the ZN44 strain, a cassette containing the spectinomycin resistance gene, the PrbcL promoter sequence, the Pdc ZM gene, and the slr1192 gene was amplified from a previously constructed pZG25 plasmid (Gao et al., 2012) and ligated with the backbone of another previously constructed plasmid pFL20n (Duan et al., 2016) containing the upstream and downstream sequences of neutral site I (NS1). The final obtained plasmid was termed as pZN44. All the resulting plasmids were then verified by Sanger gene sequencing and are listed in Table 1. Transformation of the PCC 7942 Strains Transformation of PCC 7942 derived cell factory FL130 was performed according to a previously reported method with optimization and quantitation (Lou et al., 2018). Briefly, 2 ml WT or FL130 broth with an OD 730 of 1.0 was centrifuged at 6,000 × g for 5 min and then re-suspended in 250 µl of fresh BG11 medium. Next, 200 ng plasmid was added and incubated in the dark at 150 rpm and 30 • C for 6-24 h. Then, 5 µl of the mixtures (about 4 × 10 6 cells) was plated on solid BG11 plates with 1.5% agar after transformation and incubated at 45 • C and 300 µmol photons/m 2 /s for 2 days. The obtained colonies would be picked, streaked on fresh BG11 plates, and incubated for another 24 h under the same conditions. Genotypes of the final survived transformants would be checked, and the atpA gene would be amplified for Sanger sequencing. Cultivations of Cyanobacterial Strains For phenotypes evaluations, PCC 7942 WT and the derived strains would be cultivated in BG11 medium with column photobioreactors (3 cm diameter), and the temperature and illumination strengths would be set as required. The light would be provided by incandescent lamps (NVC, NDL433SI-28W). Then, 3% (volume to volume) CO 2 air would be bubbled for carbon source supplementations. For liquid cultivation of the WT PCC 7942 and the derived mutant carrying the AtpA-C252F mutation, the normal conditions were set as 30 • C and 50 µmol photons/m 2 /s, and the respective stressful conditions were set as 44 • C and 400 µmol photons/m 2 /s. For sucrose production, 150 mM NaCl and 0.1 mM isopropyl-D-1-thiogalactopyranoside (IPTG) would be added to the culture medium. During this process, normal conditions were set as 30 • C and 100 µmol photons/m 2 /s, whereas stressful conditions were set as 40 • C and 400 µmol photons/m 2 /s. The growth of each cyanobacterial strain was monitored by measuring the optical density at 730 nm (OD 730 ). Dry cell weights (DCWs) were determined following the methods previously described . As for the cultivation of ethanologenic strains, the conditions were also set as 30 • C and 400 µmol photons/m 2 /s, and the ethanol production would be evaluated as previously described . Oxygen Evolution Rates Determination Oxygen concentration was measured using fiber-based and contactless oxygen microsensors (PyroScience, Aachen, Germany). The sensor was calibrated with air-saturated water and de-oxygenated water as 100% and 0% O 2 levels. The culture broths of FL130 or SZ41 were transferred to the respiration vials and placed in the corresponding culture conditions. The oxygen concentration in the respiration vials was continuously monitored by the oxygen microsensor connected to the oxygen logger software. The final photosynthetic oxygen evolution rate was calculated according to the Chla content, which was determined spectrophotometrically at OD 665 and OD 720 in methanol extracts and calculated with the formula: Chla (mg/L) = 12.9447 × (A665 − A720). Determination of Sucrose and Glycogen Contents in PCC 7942 To determine the amount of extracellular sucrose in PCC 7942 culture broth, 1 ml of cyanobacterial culture was centrifuged, and sucrose in the supernatant was measured using the sucrose/D-glucose assay kit (Megazyme). The glycogen contents in PCC 7942 cells were measured as previously introduced with modifications ( Chi et al., 2019). After the glycogen precipitants were hydrolyzed to glucose by treatment with amyloglucosidase at 60 • C for 2 h, glucose in 100 mM sodium acetate (pH 4.5) solution was measured using the sucrose/D-glucose assay kit (Megazyme). DATA AVAILABILITY STATEMENT The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. AUTHOR CONTRIBUTIONS SZha, SZhe, XZ, and YD performed the research project. GL and XL supervised the research project and guided the design of experiments. SZha, GL, and XL drafted and revised the manuscript. All authors contributed to the article and approved the submitted version.	2021-04-08T13:13:04.895Z	2021-04-08T00:00:00.000	{ "year": 2021, "sha1": "c32a2357bd2ca2057223f003e63f559e3a81cc26", "oa_license": "CCBY", "oa_url": "https://www.frontiersin.org/articles/10.3389/fmicb.2021.647164/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "c32a2357bd2ca2057223f003e63f559e3a81cc26", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
253841451	pes2o/s2orc	v3-fos-license	Twenty-four-hour time-use composition and cognitive function in older adults: cross-sectional findings of the ACTIVate study Introduction Physical activity, sedentary behaviour and sleep are associated with cognitive function in older adults. However, these behaviours are not independent, but instead make up exclusive and exhaustive components of the 24-h day. Few studies have investigated associations between 24-h time-use composition and cognitive function in older adults. Of these, none have considered how the quality of sleep, or the context of physical activity and sedentary behaviour may impact these relationships. This study aims to understand how 24-h time-use composition is associated with cognitive function across a range of domains in healthy older adults, and whether the level of recreational physical activity, amount of television (TV) watching, or the quality of sleep impact these potential associations. Methods 384 healthy older adults (age 65.5 ± 3.0 years, 68% female, 63% non-smokers, mean education = 16.5 ± 3.2 years) participated in this study across two Australian sites (Adelaide, n = 207; Newcastle, n = 177). Twenty-four-hour time-use composition was captured using triaxial accelerometry, measured continuously across 7 days. Total time spent watching TV per day was used to capture the context of sedentary behaviours, whilst total time spent in recreational physical activity was used to capture the context of physical activity (i.e., recreational accumulation of physical activity vs. other contexts). Sleep quality was measured using a single item extracted from the Pittsburgh Sleep Quality Index. Cognitive function was measured using a global cognition index (Addenbrooke’s Cognitive Examination III) and four cognitive domain composite scores (derived from five tests of the Cambridge Neuropsychological Test Automated Battery: Paired Associates Learning; One Touch Stockings of Cambridge; Multitasking; Reaction Time; Verbal Recognition Memory). Pairwise correlations were used to describe independent relationships between time use variables and cognitive outcomes. Then, compositional data analysis regression methods were used to quantify associations between cognition and 24-h time-use composition. Results After adjusting for covariates and false discovery rate there were no significant associations between time-use composition and global cognition, long-term memory, short-term memory, executive function, or processing speed outcomes, and no significant interactions between TV watching time, recreational physical activity engagement or sleep quality and time-use composition for any cognitive outcomes. Discussion The findings highlight the importance of considering all activities across the 24-h day against cognitive function in older adults. Future studies should consider investigating these relationships longitudinally to uncover temporal effects. Introduction: Physical activity, sedentary behaviour and sleep are associated with cognitive function in older adults. However, these behaviours are not independent, but instead make up exclusive and exhaustive components of the 24-h day. Few studies have investigated associations between 24-h time-use composition and cognitive function in older adults. Of these, none have considered how the quality of sleep, or the context of physical activity and sedentary behaviour may impact these relationships. This study aims to understand how 24-h time-use composition is associated with cognitive function across a range of domains in healthy older adults, and whether the level of recreational physical activity, amount of television (TV) watching, or the quality of sleep impact these potential associations. Methods: 384 healthy older adults (age 65.5 ± 3.0 years, 68% female, 63% non-smokers, mean education = 16.5 ± 3.2 years) participated in this study across two Australian sites (Adelaide, n = 207; Newcastle, n = 177). Twentyfour-hour time-use composition was captured using triaxial accelerometry, measured continuously across 7 days. Total time spent watching TV per day was used to capture the context of sedentary behaviours, whilst total time spent in recreational physical activity was used to capture the context of physical activity (i.e., recreational accumulation of physical activity vs. other contexts). Sleep quality was measured using a single item extracted from the Pittsburgh Sleep Quality Index. Cognitive function was measured using a global cognition index (Addenbrooke's Cognitive Examination III) and four cognitive domain composite scores (derived from Introduction The positive relationship between physical activity and cognitive function in older adulthood is well documented. Several cross-sectional studies have demonstrated that older adults who engage in higher levels of physical activity have better global cognitive function (Falck et al., 2017b), executive function (Daly et al., 2015), and memory (Xu et al., 2011). Further, some longitudinal (Daly et al., 2015;Rojer et al., 2021) and intervention studies (Lautenschlager et al., 2008;Northey et al., 2018) suggest that physical activity engagement is associated with lower odds of cognitive decline and preserved cognitive functioning in older adults. Conversely, other studies have found no associations between physical activity and cognitive function in older adults after adjusting for sociodemographic and health covariates such as age, sex, education, smoking, body mass index and depression, all of which are risk factors for cognitive decline and dementia (Daimiel et al., 2020;Livingston et al., 2020). Several studies have reported intensity-specific differences in the association between physical activity and cognitive function. Moderate-to-vigorous physical activity (MVPA) is strongly and positively associated with cognitive function (Kerr et al., 2013;Rojer et al., 2021), and associations between engagement in light-intensity physical activity (LPA) and cognitive function have also been reported (Johnson et al., 2016). Additionally, there is evidence to suggest that certain types of physical activity may be more beneficial for cognitive function than others because they differ in level of cognitive engagement. Physical activity may be accumulated in a variety of contexts, including occupational, household, transportation, and recreational modalities (Gill et al., 2015). Activities which constitute recreational physical activity (also known as leisuretime physical activity), such as dancing, gym classes or playing sports, are considered to be more beneficial for cognitive function than other physical activity modalities such as active transport (i.e., walking to the bus stop) or household chores (Phansikar et al., 2019), whilst studies on the impact of occupational physical activity on cognitive function in older adults have presented mixed findings (Rovio et al., 2007;Adam et al., 2013). The benefits associated with recreational physical activity may be partly because many recreational activities require more neuromuscular complexity, higher levels of cognitive engagement, and more social interaction (i.e., through engaging with others during group activities) (Phansikar et al., 2019). However, to date, few studies have assessed differences between recreational physical activity and non-recreational physical activity against cognitive function in older adults, and this warrants further exploration. A limitation of many studies in this field is that they have not considered the interaction between physical activity and other time-use behaviours, sedentary behaviour and sleep, which make up the 24-h day . There is mixed evidence regarding the association between sedentary behaviour and cognitive functioning in older adults. Engaging in excessive daily sedentary behaviour has been negatively associated with cognitive function in older adults in a previous review and longitudinal study (Falck et al., 2017a;Ku et al., 2017). Conversely, both Maasakkers et al. (2020) and Chen et al. (2022) did not find associations between total sedentary time and cognitive function in older adults, although Chen et al. (2022) did report a negative association between prolonged bouts of sedentary behaviour and cognitive outcomes. These discrepant findings may reflect the importance of considering how sedentary behaviour is broken up, or types of sedentary behaviour, rather than operationalizing sedentary behaviour as total undifferentiated sedentary time for cognitive function outcomes (Maasakkers et al., 2020). The context of sedentary behaviour may differentially affect cognitive functioning, such that sedentary behaviours that are more cognitively engaging (such as computer use or workrelated activities) may have different associations with cognitive function outcomes compared to cognitively passive sedentary behaviours [such as television (TV) watching]. For example, several studies have reported positive associations between computer use and cognitive outcomes, and negative associations between TV watching and cognitive outcomes (Kesse-Guyot et al., 2012;Bakrania et al., 2018). Olanrewaju et al. (2020) reported no cross-sectional associations between total sitting time and cognitive function but found significant negative associations between TV watching and verbal memory and fluency. Taken together, when investigating cognitive function outcomes, it may be useful to consider the context of sedentary behaviours (i.e., how mentally engaging they are), in addition to total sedentary time. Sleep is also associated with cognitive function in older adults. Several studies have suggested that too much or too little sleep (i.e., >9 h or <6 h) is negatively associated with cognitive function, often referred to as an "inverted U-shaped" relationship (Gildner et al., 2014;Wild et al., 2018). Other studies found no evidence for these relationships (McSorley et al., 2019) or reported associations with long but not short sleep duration (Faubel et al., 2009;Low et al., 2019;Kondo et al., 2021). Beyond sleep duration, poor sleep quality and efficiency have been linked to poor cognitive function in older adults (Nebes et al., 2009;Miyata et al., 2013). Therefore, in the context of cognitive function in ageing, it is likely important to consider the not only sleep duration, but the whole sleep experience including sleep quality. It is plausible that both the amount of time and the characteristics of that time use are important factors that may affect cognitive function. Furthermore, physical activity, sleep and sedentary behaviour interact to make up the 24-h day, such that increasing time in one behaviour must result in an equal and opposite change in time spent in one or both of the other timeuse behaviours (Dumuid et al., 2020). Intervention studies in which participants only modify one daily activity (e.g., increase physical activity) are not able to disentangle whether any cognitive or other benefits are due to the change in this activity (e.g., the increase in physical activity) or to the compensatory change in time spent in one of the other behaviours (i.e., decreased sedentary behaviour or sleep) (Dumuid et al., 2017;Rosenberger et al., 2019). Recent developments in statistical techniques based on compositional data analysis (CoDA) (Aitchison, 1982) allow the daily composition of all three timeuse behaviours, or time-use composition, to be studied against health outcomes in a single analytical model. CoDA requires time-use variables (e.g., physical activity, sedentary behaviour, and sleep) to be expressed as isometric log ratios that are able to be used in traditional statistical models, such as multiple linear regression analyses. A CoDA approach overcomes the issue of the perfect multicollinearity of 24-h time use data that violates the assumptions of many statistical models. This has previously been (inadequately) overcome by omitting one or more time-use behaviours from statistical models (Dumuid et al., 2017). To our knowledge, only three previous studies on time use and cognitive function in older adults have considered all three time-use behaviours simultaneously in the same statistical model. Briefly, using isotemporal substitution methods, Fanning et al. (2017) reported that replacing time in sedentary behaviour with MVPA or sleep was positively associated with cognitive function, whilst replacing time in sedentary behaviour with LPA had no association with cognitive function. Wei et al. (2021) reported that the associations between time-use composition and cognitive function varied between older adults who achieved <7 h of sleep per night compared to those who achieved >7 h per night. However, these studies relied on a combination of device-based and self-report measures to capture 24-h time use, which may have resulted in the overor under-estimation of time spent in respective behaviours. Only one study has captured 24-h time use using devicebased measures and investigated these relationships using a CoDA approach. Dumuid et al. (2022) reported that 24-h time-use composition was associated with global cognition and executive function in a sample of middle-to-older adults. Predictive modelling indicated that more time in MVPA and less time in LPA was beneficial for cognitive function, and that this relationship was more pronounced in people at higher genetic risk of Alzheimer's disease (carrying an APOE ε4 allele) . Together, these studies highlight the importance of simultaneously considering all behaviours across the 24-h day against cognitive function. However, none have considered the context of time-use behaviours in addition to their duration (e.g., sleep quality and context of physical activity and sedentary behaviour). This study investigates the associations between 24-h timeuse composition of physical activity, sedentary behaviour, and sleep on cognitive function in a large sample of healthy older Australians. It uses data collected during the baseline phase of the ACTIVate study, a 3-year multisite study aiming to optimise daily activity patterns and diet for dementia prevention [see Smith et al. (2022) for study protocol]. In addition to examining the relative duration of these behaviours across the 24-h cycle, we investigate whether the context of physical activity, sedentary behaviour and quality of sleep modify the associations between time-use composition and cognitive outcomes. Ethics The ACTIVate study was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12619001659190) on November 27, 2019. Ethics approval was obtained from the University of South Australia and University of Newcastle Human Research Ethics Committee (202639). All procedures were conducted in accordance with the Declaration of Helsinki. Participant recruitment and screening Eligibility criteria for the ACTIVate study are described in more detail elsewhere (Smith et al., 2022). Briefly, participants met the inclusion criteria if they were aged 60-70 years, fluent in English, had no current clinical diagnosis of dementia, major neurological or psychiatric diagnoses, known intellectual disability, or major physical disability, and presented no contraindications to transcranial magnetic stimulation screening (Rossi et al., 2009). Potential participants were required to undertake a phone screening interview, during which the inclusion criteria above were assessed (via self-report) in additional to completing a Montreal Cognitive Assessment (Blind) (MoCA-B) to screen for dementia (using a cut-off score of <13/22). Power calculations were determined for the larger ACTIVate study based on cross-sectional pilot data, in the absence of longitudinal data on diet and activity compositions in relation to cognitive outcomes. Aiming for 80% power, accounting for the longitudinal design and allowing for attrition and response rate at recruitment, the final sample size of 448 participants was determined [see Smith et al. (2022) for further details]. Device-measured activity patterns Time spent in physical activity, sedentary behaviour and sleep was measured using accelerometry. Participants were asked to wear a triaxial accelerometer on their non-dominant wrist (Axivity AX3) 24 h per day for seven consecutive days, with data recorded at a sampling frequency of 100 Hz. Raw acceleration data were downloaded using the Open Movement GUI software (OmGUI; Newcastle, UK) and further processed using a custom MATLAB graphic user interface developed at the University of South Australia (COBRA; MATLAB R2018B). Time spent in sleep was verified by manual cross-checking of sleep logs completed by participants during the 7 days of recording and visual inspection of the accelerometry trace across the 7 days. Non-wear time was identified as ≥60 min of ≤25 gmin, and was also manually verified against participant diaries (removal of watch recorded as well as sleep information) and by visual inspection of the accelerometry trace. Waking day behaviours were classified as time in MVPA (>93 g-min), LPA (>48 g-min) or sedentary behaviour (<48 g-min) using previously published cut points adjusted for sampling frequency (Hildebrand et al., 2014). Accelerometry data were classified as a "valid wear day" if the accelerometer was worn for at least 10 waking hours. Only participants with three or more valid weekdays and one valid weekend day were included in the analyses. Total time spent in each activity was averaged across the recording period, providing average time (minutes) spent in MVPA, LPA, sedentary behaviour, and sleep per day. Self-report activity measures Television watching The amount of TV watching per day (in minutes) was used as a measure for cognitive engagement during sedentary behaviour. These data were obtained through the Multimedia Activity Recall for Children and Adults (MARCA), a computerised use-of-time recall tool containing over 500 potential daily activities (Ridley et al., 2006). During the MARCA assessment, participants were asked to recall every activity they had engaged in over the 2 previous days (Gomersall et al., 2011). Due to study constraints, not all participants recalled one weekend day and one weekday (two weekdays most commonly recalled). All MARCA phone calls were scheduled during the 7-days that participants were wearing the accelerometer. Total time spent watching TV was averaged across the 2 days of recall and used to categorise participants into low, medium, and high tertiles. Recreational physical activity To investigate whether the context of physical activity influenced associations between time-use composition and cognitive function, we captured total time spent in recreational physical activity using the MARCA assessments (Ridley et al., 2006). Recreational physical activity encompasses play-based activities (e.g., totem tennis, darts, juggling) and sport (e.g., dancing, gym-based exercise, team sports, partner sports and individual sports). Total time spent in recreational physical activity was averaged across the 2 days of recall, and participants were further categorised in to "no recreational physical activity, " "0-30 min of recreational physical activity, " and "30+ minutes of recreational physical activity" categories. These categories were chosen during analysis because (1) there were a considerable number of participants with 0 min of recreational physical activity, and (2) separating those who did engage in recreational physical activity in to <30 min or >30 min allows for sufficient comparison (rather than comparing "none" against "some" recreational physical activity). An exhaustive list of activities that were categorised as recreational physical activity can be found in Supplementary material 1. Sleep quality The Pittsburgh Sleep Quality Index (PSQI) is a subjective measure of sleep quality over the past month (Buysse et al., 1989). The PSQI asks 19 questions which produce seven component scores assessing sleep quality, latency, duration, efficiency, disturbances, use of medications and daytime dysfunction. For the purpose of this study, the sleep quality component was used as a measure of sleep quality (question 6: "During the past month, how would you rate your sleep quality overall?"). Participants rated their sleep using the following scoring system: 0 = "very good"; 1 = "fairly good"; 2 = "fairly bad"; 3 = "very bad." Participants were therefore categorised as having "bad" sleep quality (scoring 2 or 3) or "good" sleep quality (scoring 0 or 1) for final analyses. Cognitive function measures Addenbrooke's Cognitive Examination III The Addenbrooke's Cognitive Examination III (ACE-III) was used as a measure of global cognition. The ACE-III is a brief paper-and-pencil-style cognitive screening tool that assesses cognitive function across five domains as a total score out of 100 that includes scoring of memory, attention/orientation, language, fluency and visuospatial ability. The ACE-III demonstrates high internal reliability (Cronbach's α = 0.88), and high specificity (0.96) and sensitivity (1.00) to detecting dementia using the cut-off score of 88/100 (Hsieh et al., 2013). Cambridge Automated Neuropsychological Test Automated Battery Domain-specific cognitive function was assessed using the Cambridge Automated Neuropsychological Test Automated Battery (CANTAB). CANTAB is a computerised neuropsychological test battery containing a range of cognitive tests across several cognitive domains. CANTAB tests have demonstrated discriminant validity between clinical populations and healthy controls (Saunders and Summers, 2010), and moderate correlations with traditional neuropsychological tests in younger populations (e.g., Trail Making Test with Paired Associates Learning; Animal Fluency and Green Story Delayed Recall with Verbal Recognition Memory) (Smith et al., 2013) and older populations (e.g., Rey complex figure test compared to Paired Associates Learning test) (Lenehan et al., 2015). Test-retest reliability of CANTAB tests ranges from weak to strong (0.56 to.89) (Gonçalves et al., 2016). Participants completed the following tests: Paired Associates Learning (PAL), Reaction Time (RTI), Multitasking Test (MTT), Verbal Recognition Memory (immediate and delayed) (VRM), and One Touch Stockings of Cambridge (OTS). The total time to complete all five CANTAB tasks including a 2-min familiarisation task (Motor Screening Task) was approximately 40 min. For tests where lower scores indicate better performance (e.g., reaction time), scores were reversed. Raw CANTAB test scores were then converted to z-scores. Individual outcome measure z-scores were collated into cognitive composites using the Cattell-Horn-Carroll-Miyake (CHCM) cognitive domain taxonomy as a guiding framework (Webb et al., 2018). Cognitive composites were classified according to the broad domains in the CHCM taxonomy: long-term storage and retrieval (herein referred to as "long-term memory"; short-term and working memory (herein referred to as "short-term memory"); executive function; and processing speed. For tests that were not classified within the CHCM taxonomy (i.e., OTS), tests were allocated to a cognitive domain by consensus among the authorship team. Table 1 displays the outcome measures used to create each cognitive composite. For each outcome measure to be included in the cognitive composite, a selection of inclusion criteria was applied to respective CANTAB tests. Participants who achieved <50% accuracy on tests used in the criteria listed in Supplementary material 2 were excluded from analyses. Cognitive composite scores were not generated for a participant unless all data were available (i.e., four test scores in a four-test composite), and participants met all inclusion criteria within that composite. Covariates Demographic factors that have been associated with increased dementia risk (Livingston et al., 2020) were entered as covariates, including age (years), sex (male, female), education (total years), and smoking status (current smoker, previous smoker, never smoker). Additionally, models were adjusted for site (Adelaide, Newcastle). Statistical analysis Associations between time-use composition and cognitive function All inferential statistics were conducted in R version 4.2 (R Core Team, 2021), and the code used to analyze data is available at https://github.com/MaddisonMellow/Timeuse-cognition-paper. Pearson correlation coefficients between time use variables (MVPA, LPA, sedentary behaviour and sleep) and cognitive outcomes (global cognition, short-term memory, long-term memory, executive function and processing In order to include time-use composition as a variable in linear regression models, the composition of behaviours were expressed as isometric log ratio coordinates (see Dumuid et al. (2017) for extensive overview of this approach). To achieve this for a four-part composition, three isometric log ratios were created: the first represented the log-ratio of one behaviour (e.g., sleep) to the remaining three behaviours (e.g., sedentary behaviour, LPA, MVPA); for the second isometric log-ratio coordinate, sleep was excluded and the log-ratio of the next time-use behaviour in the set (sedentary behaviour) to the remaining two behaviours (LPA and MVPA) was calculated; finally, the third isometric log-ratio coordinate only contained information on the remaining two behaviours (LPA:MVPA). All three isometric log-ratio coordinates were entered into the regression model to represent the entire 24-h time-use composition. Further, all three coordinates were included in ANOVA type II F-tests (described below) to test the null hypothesis that all coefficients of isometric log-ratios were equal to zero. With the aim of producing parsimonious models supported by the data for each outcome variable, backward selection of the multiple linear regression models was employed. The initial (potential) model for each outcome included an intercept, main effects (demographics, time-use composition, TV watching, recreational physical activity level, sleep quality) and the two-way interaction effects of primary interest in timeuse composition with each TV watching, recreational physical activity, and sleep quality. The backward selection process proceeded by simplifying the model by sequential F-tests, stopping model reduction when significance at the level of 0.05 was reached for all remaining terms in the following groups, considering the following order: (i) interaction terms (in order of largest p-value), then (ii) the time-use composition, and (iii) the remaining covariates collectively. Type II F-tests were used to determine variable significance, which assesses variable effects after adjusting for other variables while adhering to the principle of marginality (Langsrud, 2003;Fox and Weisberg, 2019). P-values within each of the final models were adjusted for multiple comparisons using the Benjamini-Hochberg false discovery rate adjustments (Benjamini and Hochberg, 1995). Both adjusted and unadjusted p-values are presented in the results. Interaction terms and reallocations of time Where models containing interaction terms remained statistically significant after false discovery rate adjustment, we planned to plot model-generated one-for-remaining predictive response curves to show how the cognitive outcome measures were associated with meaningful reallocations of time (i.e., in 15-min increments), across different levels of sleep quality, TV watching or recreational PA [see Dumuid et al. (2017) for example]. Participant demographics Four hundred and twenty-six participants were initially included in the dataset. Of these, 21 were removed as they Values are presented as either mean ± SD for numeric variables or count (percentage) for categorical variables. Recreational physical activity (PA) and TV watching data are presented as the mean ± SD minutes per day spent in respective activities. MVPA, moderate-vigorous physical activity; LPA, light physical activity; SB, sedentary behaviour; PME, perceived mental effort; ACE-III, Addenbrooke's Cognitive Examination III. did not have valid accelerometry data, and 21 were removed due to missing covariate data. Thus, the overall final sample included 384 older adults (65.5 ± 3.0 years old, 121 males). Means and standard deviations of demographic and other key variables are presented in Table 2. The final sample were 68% female, 63% non-smokers, with mean 16.5 ± 3.2 years of total education. Over half of participants (n = 207) reported engaging in no recreational physical activity, whilst 26.7% engaged in over 30 min per day. A number of participants were removed from each analysis due to missing cognitive data: total samples for each cognitive outcome included: n = 384 for global cognition; n = 292 for short-term memory; n = 353 for long-term memory; n = 369 for executive functions; n = 358 for processing speed. Associations between time-use composition and cognition Pairwise correlations Pearson correlation coefficients revealed that time spent in sleep was negatively correlated with long term memory (r = -0.11, p = 0.03), time spent in sedentary behaviour was negatively correlated with processing speed (r = -0.13, p = 0.01), and time spent in MVPA was positively correlated with processing speed (r = 0.17, p < 0.01) ( Table 3). Linear regression models Prior to adjustment for false discovery rate, time-use composition was significantly associated with processing speed, such that more time spent in sleep (β = 0.28) or MVPA (β = 0.20) relative to time spent in the remaining behaviours was associated with faster processing speed, whilst more time spent in LPA (β = -0.16) or sedentary behaviour (β = -0.33) was associated with slower processing speed ( Table 4). FIGURE 1 Distribution of the sample time-use compositions. Each grey dot represents a single participant's time-use composition. The black dot represents the average time-use composition of the sample, calculated as the geometric means of each activity, collectively adjusted to sum to 1,440 min. This differs to the arithmetic means, which are presented in Table 2; instead of summing the values and dividing by the number of values "n," the geometric mean multiplies all the values and then takes the nth root. To convert the geometric mean to the compositional mean, the "closure" operator (a function part of Compositions package in R) is applied to the geometric mean, so that the centrality measure reflects the 1,440 min (24 h) in the day. The black dot represents this average composition ("compositional mean") and shows that the average participant spends approximately 18.5% of their day in physical activity, 46.6% of their day in sedentary behaviour, and 35% of their day in sleep. Black ellipses represent 75, 95, and 99% confidence intervals, respectively. Additionally, several covariates were significantly associated with cognitive outcomes: older age was associated with poorer executive function (β = -0.05) and slower processing speed (β = -0.04); site was negatively associated with global cognition (β = -1.43) and positively associated with longterm memory (β = 0.25) and short-term memory (β = 0.17) (i.e., participants from Newcastle had lower global cognition scores and higher long-term and short-term memory scores than Adelaide); higher education (years) was associated with better global cognition (β = 0.22); sex (being female) was negatively associated with executive function (β = -0.23) and engaging in no recreational physical activity (relative to >30 min) was associated with poorer long-term memory (β = -0.38). None of the interaction terms were statistically significant. After false discovery rate adjustment, none of the associations between 24-h time-use composition and cognitive function outcomes were statistically significant. Associations between age and executive functions, recreational physical activity and long-term memory, sex and executive functions, as well as education, site and global cognition, remained significant. Unadjusted and adjusted p-values for all variables across each cognitive outcome are displayed in Table 4. Linear regression outputs for each cognitive outcome can be found in Supplementary material 3. 24-h time use composition and cognitive function in older adults Although there is some evidence that physical activity, sleep, and sedentary behaviour are independently associated with cognitive function in older adults, it remains unclear how the balance of these three behaviours in the 24-h day, or time-use composition, relates to cognitive function outcomes. The current study investigated the cross-sectional associations between 24-h time-use composition and cognitive function in a sample of healthy older adults. We initially explored pairwise correlations to understand the independent and unadjusted associations between time use variables and cognitive outcomes, in which we found that sleep was negatively correlated with long-term memory, sedentary behaviour was negatively correlated with processing speed, and MVPA was positively correlated with processing speed. Subsequently, after adjusting for demographic and health factors that are associated with ACE-III Long-term memory Short-term memory Executive function Processing speed Sleep ( (3,359) 0.04 0.16 F ( n , d ) , F statistic, and numerator and denominator degrees of freedom; adj.p, p-value adjusted for false discovery rate. Bold denotes statistical significance (p ≤ 0.05). *Denotes p-values that remained significant after false discovery rate adjustment. "-" Denotes variables that were not included in final models for respective cognitive outcomes. Interaction terms (for sleep quality, recreational PA or TV watching) were not included in final models for any cognitive outcomes and therefore are not listed in this table. Frontiers in Human Neuroscience 09 frontiersin.org risk of dementia (age, sex, education, smoking status), linear regression models demonstrated 24-h time-use composition was significantly associated with processing speed, but there were no associations with global cognition, long-term memory, shortterm memory or executive function. However, the association between 24-h time use and processing speed was nonsignificant after adjustment for false discovery rate. Together, these findings demonstrate the importance of considering all activities across the 24-h day against cognitive function, as the relationships between time-use behaviours and several cognitive outcomes (assessed via correlations) were likely attenuated after accounting for other time-use behaviours. Few studies have examined 24-h day time use against cognitive function in older adults, and only one of these has used CoDA . Our findings of no associations contradict the few previous studies that reported associations between time-use composition and cognitive function in healthy older adults (Fanning et al., 2017;Wei et al., 2021). There are several important differences which may explain the contradictory findings in the current study, such as the type of analyses conducted (testing associations of overall compositions rather than time-reallocations with cognitive function), and the characteristics of the recruited samples. Both Fanning et al. (2017) and Wei et al. (2021) recruited a low-active sample, whereby participants engaged in approximately 46 min or 36 min of MVPA per day, respectively. In our sample, participants were achieving an average of 89 min/day in MVPA. As such, in a low-active sample, it is likely that small differences in MVPA (or reallocations of time to MVPA) have more potent effects on cognitive function as the level of MVPA is low across participants. For example, in participants who engage in lower levels of MVPA per day (i.e., 30 min/day), reallocating an additional 15 min of MVPA per day equates to a 50% increase in total MVPA, whereas for those with a baseline level of 90 min of MVPA per day, reallocating an additional 15 min is equivalent to a ∼16% increase only. Additionally, Dumuid et al. (2022) recruited a sample from a wider age range (50-80 years), across a range of cardiovascular risk profiles ("low" and "elevated" cardiovascular disease risk), with a subsequently lower level of cognitive function (mean ACE-III score = 91). Taken together it is likely that the recruitment of a highly active, high-performing sample in the current study contributed to the largely null findings. Another relevant consideration may be the difference in measures of time use. Fanning et al. (2017) measured waking time-use behaviours (i.e., sedentary behaviour, LPA and MVPA) using accelerometry, but measured sleep duration using self-report (PSQI). This may have resulted in underor over-estimation of sleep duration. Additionally, Wei et al. (2021) measured time use using the Global Physical Activity Questionnaire, and only included MVPA, walking/bicycling, sedentary behaviour and sleep (duration value obtained during interviews) in their regression models. It is possible both studies did not capture 24-h time use in its entirety. Finally, two of the previous studies used different measures of cognitive function compared to the current study, namely task-switching and working memory paradigms (Fanning et al., 2017), and tests of memory (CERAD word learning test), language (animal fluency), executive function/processing speed (digit symbol substitution test) and global cognition (composite z-score of all tests) (Wei et al., 2021). It is acknowledged that using scores from single cognitive tests may exhibit a higher level of variability in cognitive function, and conversely, creating composite scores containing multiple tests may mask important differences in individual component scores. Although the composites created for this study were guided by a cognitive domain framework, the number of scores contributing to each composite varied (one for global cognition; one for long-term memory; 4 for short-term memory, 5 for executive function; 4 for processing speed). The true variability of cognitive test scores may have been weakened by creating large composites. Sedentary behaviour context, time use and cognitive function A secondary aim of this study was to investigate whether time spent in sedentary behaviours that require low mental engagement, in this instance TV watching, influenced the associations between time-use composition and cognitive function. We found no significant main effect of TV watching on cognitive function, or interaction effect of TV watching on associations between time-use composition and cognitive function. Overall, our non-significant findings do not align with previous studies that have reported positive associations between mentally stimulating sedentary behaviours (i.e., reading and computer use) and cognitive function, and negative associations between passive sedentary behaviours (e.g., TV watching) and cognitive function (Bakrania et al., 2018;Olanrewaju et al., 2020;Nemoto et al., 2022). Previous studies measuring sedentary behaviour types and cognitive function in older adult populations have predominantly used self-report measures (Kesse-Guyot et al., 2012;Hamer and Stamatakis, 2014;Falck et al., 2017a). Although self-report measures are able to capture the context of sedentary behaviours, accumulation of activities in older adults is often intermittent or unstructured, which can contribute to over-reporting of sedentary behaviours (Falck et al., 2017a). Conversely, objective measures such as accelerometry do not provide information on which types of sedentary behaviour people are engaging in, which in the context of cognitive function, may be an important consideration. These limitations have led to the recommendation that future studies use a combination of both objective and subjective measures of sedentary behaviour (Falck et al., 2017a). To our knowledge, this is the first study to combine subjective and device-based measures of sedentary behaviour in the same analytical model against cognitive function in older adults. This is also the first study to do so using a CoDA approach, whereby all time-use behaviours are controlled for. It is likely that our methodological approach contributed to the contradictory findings of this study, as most previous studies have not controlled for other time-use behaviours. Sleep quality, time use and cognitive function We did not detect a significant interaction between 24-h time-use composition and sleep quality in association with any of the cognitive measures. Although this interaction has never been directly tested previously, findings from some studies have suggested that perceived sleep quality is important to consider when exploring the relationship between physical activity, sleep duration and cognitive outcomes (Nakakubo et al., 2017;Yuan et al., 2020;Sewell et al., 2023). The lack of evidence supporting this relationship in the current study may be due to several factors. First, the measure used to capture sleep quality in this study relied on a single item measure, chosen to avoid collinearity between total sleep duration measures from the PSQI and accelerometry measures. Second, most participants in this sample rated their sleep quality as "fairly good" (54%) or "very good" (27%), whilst only 17% of participants reported "fairly bad" and 2% reported "very bad" sleep quality. Therefore, over three quarters of participants were classified as having "good" sleep quality in final analyses. This suggests that either, most of the sample had good sleep quality, or that our singleitem measure did not capture the true variability in sleep quality. Future studies may benefit from capturing sleep quality or disturbances using objective measures such as accelerometry or polysomnography, to better understand the impact of sleep quality on the associations between time-use composition and cognitive outcomes. Additionally, future studies should consider exploring additional measures such as sleep stage duration, sleep efficiency, and spindle activity, which were identified as predictors of cognitive function in older adults in a recent study (Djonlagic et al., 2021). Physical activity context, time use and cognitive function Although there is some evidence to suggest that recreational physical activity is more beneficial for cognitive function compared to active transport or occupational physical activity (Phansikar et al., 2019), few studies have investigated this in older adults. We found a significant main effect of spending no time in recreational physical activity (relative to >30 min per day) with worse long-term memory performance. However, we did not find an interaction between time-use composition and the amount of recreational physical activity with any cognitive outcomes. This may be due in part to the low proportion of participants who reported engaging in recreational physical activity: over half (54%) of participants reported engaging in no recreational physical activity, and only one quarter (27%) engaged in more than 30 min per day. Given that the average time per day spent in MVPA and LPA (as measured by accelerometry) was ∼1.5 and ∼3 h respectively, this may suggest that our participants accumulated their physical activity via other modalities, including active transport (walking, cycling), occupational, or household (i.e., gardening, sweeping). This aligns with evidence that older adults >65 years of age generally prefer walking as their mode of physical activity, whilst older adults aged >74 years more often prefer exercising in social contexts, such as recreational group fitness classes (Amireault et al., 2019). Thus, future studies in this age group should consider physical activity contexts beyond recreational PA. Strengths and limitations To our knowledge, this study was the first to investigate the associations between 24-h time use and cognitive function in older adults using device-based measures (accelerometry) to capture time use, and to additionally consider the role of the context and quality of time-use behaviours in these relationships. A wide range of reliable cognitive tests were used to examine whether 24-h time use is associated with cognitive function across a range of domains. We took a conservative approach to interpreting the findings by adjusting for false discovery rate which is not commonly done in exploratory studies (Jafari and Ansari-Pour, 2019). However, there are several limitations to consider. First, despite efforts to recruit participants who represented a variety of activity patterns and dietary patterns (for the ACTIVate study), the final sample were highly active, highly educated, and subsequently achieved high scores on the cognitive assessments (i.e., average ACE-III score = 95). Second, the cross-sectional nature of this study does not allow causal inferences to be made about the relationships between variables. Third, the CANTAB tests took approximately 40 min in total and were conducted in the same order for all participants, so it may be possible that those conducted toward the end of the battery were impacted by fatigue. However, given the consistency in order of test completion, these effects would be systematic across the cohort. Finally, due to study constraints there were inconsistencies in the days being recalled during the MARCA, which resulted in some participants including a weekend day in their 2-day recall, and others only recalling weekdays. This may have contributed to variability in recall of recreational physical activity and TV watching. Our study identified independent correlations between time use variables and cognitive outcomes, however, linear regression models found no significant associations between 24-h time-use composition and cognitive function outcomes, and no significant interactions between TV watching time, recreational physical activity engagement or sleep quality and time-use composition for any cognitive outcomes. Together these findings highlight the importance of considering all activities across the 24-h day against cognitive function in older adults. Future studies should consider investigating these relationships longitudinally to uncover temporal effects. Data availability statement The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Ethics statement The studies involving human participants were reviewed and approved by University of South Australia and University of Newcastle Human Research Ethics Committees. The patients/participants provided their written informed consent to participate in this study. Author contributions MM contributed to the study design, data analysis, interpretation of results, writing, and drafting of the manuscript. DD and TS contributed to the data analysis, interpretation of results, and drafting of the manuscript. AW contributed to the study design, data collection, interpretation of results, and drafting of the manuscript. TO, FK, and HK contributed to the study design, interpretation of results, and drafting of the manuscript. MH contributed to the study design, data collection, and drafting of the manuscript. JD and MG contributed to the interpretation of results and drafting of the manuscript. AS contributed to study design, data collection, interpretation of results, and drafting of the manuscript. All authors contributed to the article and approved the submitted version. Funding MM was supported by a Dementia Australia Research Foundation Ph.D. scholarship. The ACTIVate study is funded by an NHMRC Boosting Dementia Research Priority Round 5 grant (GNT1171313, $1.23m). DD was supported by an NHMRC Early Career Fellowship (GNT1162166). MG was supported by an Australian Research Council (ARC) fellowship (DE200100575). TS was supported by a Hospital Research Foundation grant (C-PJ-008-Transl-2020) awarded to AS and DD.	2022-11-24T14:43:33.655Z	2022-11-24T00:00:00.000	{ "year": 2022, "sha1": "c0b29af0d1000616b8ea6eb970890bafbddb6bdc", "oa_license": "CCBY", "oa_url": "https://www.frontiersin.org/articles/10.3389/fnhum.2022.1051793/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "072daf5ba5c1a5a363326a63dd904fe8f1f2c910", "s2fieldsofstudy": [ "Psychology" ], "extfieldsofstudy": [ "Medicine" ] }
12105752	pes2o/s2orc	v3-fos-license	Application of Blood-Brain Barrier Permeability Imaging in Global Cerebral Edema BACKGROUND AND PURPOSE: Blood-brain barrier permeability is not routinely evaluated in the clinical setting. Global cerebral edema occurs after SAH and is associated with BBB disruption. Detection of global cerebral edema using current imaging techniques is challenging. Our purpose was to apply blood-brain barrier permeability imaging in patients with global cerebral edema by using extended CT perfusion. MATERIALS AND METHODS: Patients with SAH underwent CTP in the early phase after aneurysmal rupture (days 0–3) and were classified as having global cerebral edema or nonglobal cerebral edema using established noncontrast CT criteria. CTP data were postprocessed into blood-brain barrier permeability quantitative maps of PS (permeability surface-area product), Ktrans (volume transfer constant from blood plasma to extravascular extracellular space), Kep (washout rate constant of the contrast agent from extravascular extracellular space to intravascular space), VE (extravascular extracellular space volume per unit of tissue volume), VP (plasmatic volume per unit of tissue volume), and F (plasma flow) by using Olea Sphere software. Mean values were compared using t tests. RESULTS: Twenty-two patients were included in the analysis. Kep (1.32 versus 1.52, P < .0001), Ktrans (0.15 versus 0.19, P < .0001), VP (0.51 versus 0.57, P = .0007), and F (1176 versus 1329, P = .0001) were decreased in global cerebral edema compared with nonglobal cerebral edema while VE (0.81 versus 0.39, P < .0001) was increased. CONCLUSIONS: Extended CTP was used to evaluate blood-brain barrier permeability in patients with SAH with and without global cerebral edema. Kep is an important indicator of altered blood-brain barrier permeability in patients with decreased blood flow, as Kep is flow-independent. Further study of blood-brain barrier permeability is needed to improve diagnosis and monitoring of global cerebral edema. T he blood-brain barrier plays an important role in many CNS diseases relating to microvascular dysfunction and neuroinflammation, which are known to increase blood-brain barrier permeability (BBBP) in animal models. 1 Obtaining quantitative BBBP information in the clinical setting would constitute a novel diagnostic and therapeutic strategy with widespread application. Global cerebral edema (GCE) represents a major complication in patients with SAH, is an independent predictor of SAH morbidity and mortality, 2,3 and is thought to be related to disruption of the BBB. Moreover, delayed cerebral ischemia represents an independent major complication and poor prognostic factor in patients with SAH. 4 Initial studies have shown that earlier diagnosis of delayed cerebral ischemia before manifestation of clinical symptoms, altered CBF, and NCCT findings is possible with BBBP imaging by using CTP. 5 BBBP evaluation with CTP is a promising technique, especially in patients with SAH, given that these patients are critically ill, require intensive monitoring, and often have contraindications to evaluation with MR imaging. 5 Currently, imaging diagnosis of GCE relies on NCCT demonstrating diffuse effacement of the hemispheric sulci and diffuse obscuration of the hemispheric gray-white matter junction at the level of the centrum semiovale. 3 These findings are challenging to identify in the setting of diffuse SAH and increased intracranial pressure effacing the sulci and cisterns. Moreover, NCCT fails to provide a direct assessment of the hemodynamic and microvascular dysfunction thought to be central to the underlying pathophysiology of GCE. Quantitative imaging of microvascular parameters, such as measures of BBBP, may provide for more accurate detection and monitoring of GCE compared with NCCT and could potentially improve the clinical management and outcomes of patients after SAH. Extended CTP allows measurement of BBBP parameters not currently used in clinical practice. K ep represents the washout rate constant of the contrast agent from the extravascular extracellular space (EES) to the intravascular space (IVS). K ep therefore reflects the flow across the blood vessel wall from the EES to the IVS, and is independent of blood plasma flow. Based on transport equations proposed by Gowland et al 6 : where T is the time constant for transfer across the BBB. Therefore, K ep is thought to be related inversely to BBBP. K trans represents the volume transfer constant from the blood plasma to the EES. PS is the permeability surface-area product and represents the flow across the blood vessel wall from the IVS to the EES. VE represents the volume in the EES per unit of tissue volume, and VP represents the plasmatic volume per unit of tissue volume. F represents the plasmatic flow. In summary, K ep , PS, and K trans can be viewed as the primary BBBP parameters because they reflect flow across the BBB. 6,7 On the other hand, VE, VP, and F are descriptive of hemodynamic conditions on either side of the BBB (the IVS or the EES, respectively). Fig 1 summarizes the relationships of BBBP parameters under normal conditions and in the setting of GCE. CTP has numerous practical advantages that make it highly suitable for critically ill patients: it is widely available, can be performed with most scanners, has few contraindications, and has rapid acquisition and processing times. Assessment of these parameters offers a promising technique to evaluate BBBP in the clinical setting. However, it has not been widely adopted in clinical practice, with few published investigations studying measures of BBBP. 5,8,9 The purpose of our study was to apply BBBP imaging by using extended CTP in the setting of GCE and to quantitatively assess its parameters. We hypothesize that application of extended CTP is able to detect a significant difference in the BBBP parameters in patients with GCE compared with patients without GCE in the early phase after SAH. Population This study was an institutional review board-approved retrospective analysis performed on consecutive patients with SAH enrolled in a prospective diagnostic accuracy trial at our institution. SAH was determined by a combination of NCCT, CTA, DSA, and/or CSF analysis. In this prospective study design, CTP was performed on days 0 -3 after aneurysmal rupture to assess baseline cerebral perfusion and to compare with later examinations for interval change. All patients underwent either endovascular coiling or surgical clipping before CTP. Demographic and clinical data were collected ( Table 1). All patients with SAH with CTP performed on days 0 -3 were included in our study. Exclusion criteria were CTP examinations with severe motion degradation limiting postprocessing methods and unavailable CTP data from archives for postprocessing. Patients were classified into 3 outcome groups: GCE (n ϭ 11), non-GCE (n ϭ 11), and indeterminate (n ϭ 2) based on admission NCCT. GCE was defined using the following established criteria 3 : complete or near-complete effacement of the hemispheric sulci and basal cisterns and bilateral and extensive disruption of the hemispheric gray-white matter junction (including basal ganglia) at the level of the centrum semiovale. 3 Absence of both features resulted in classification as non-GCE. If only 1 feature was present, patients were classified as indeterminate. Classification was performed independently by 2 board-certified neuroradiologists (15 years and 3 years of experience) blinded to all other clinical and imaging data. Consensus reading was provided for adjudication. Interobserver agreement was analyzed using the coefficient. Image Acquisition, Postprocessing, and Quantitative Analysis CTP was performed in the early phase (days 0 -3) after SAH, along with NCCT and CTA, as part of our study trial protocol. NCCT was performed from the foramen magnum to the vertex by using the following parameters: 120 kVp, 250 mAs, 1.0 rotation time, and 5.0-mm collimation. Extended CTP scanning was performed on a 64-section scanner (LightSpeed Discovery HD-750; GE Healthcare, Milwaukee, Wisconsin) with an axial shuttle mode protocol for simultaneous acquisition of CTA and CTP data using the following parameters: 80 kVp, 400 mAs, 0.4 rotation time, 5.0-mm collimation with 17 cine cycles, and 2.8-second interscan delay for the first pass. Second pass technique included 10 cine cycles with a 10-second interscan delay. A total of 90 mL of nonionic contrast was intravenously administered at 4.0 mL/s followed by a 30 mL saline-bolus chaser. Average CT dose index volume was 300 mGy and average dose-length product was 2500 mGy-cm for the entire examination, including NCCT, CTA, and extended CTP. For this study, CTP data were postprocessed off-line into maps of the BBBP parameters (K ep , K trans , PS, F, E [extraction fraction of contrast agent], VE, and VP), CBF, CBV, and MTT by using commercially available software (Olea Sphere V.2.0; Olea Medical, La Ciotat, France), based on previously published recommendations and guidelines. 8,10 Quantitative analysis of CTP data was performed using standardized sampling of the cerebral cortex according to established methods 4,11 with round, continuous ROIs placed along the cerebral cortex. Up to 24 ROIs were included in each CTP section, distributed in the following territories: approximately 4 in the anterior cerebral artery, 12 in the middle cerebral artery, and 4 in the posterior cerebral artery territories. ROIs overlying large cortical vessels with CBF greater than 100 mL/100 g/min were excluded from the statistical analysis. 12 Mean values and 95% CIs were calculated using the remaining ROIs in all CTP sections for each patient. Unpaired t tests were performed to compare K ep , K trans , PS, F, extraction fraction of contrast agent, VE and VP, as well as CBF, CBV, and MTT in the GCE and non-GCE groups. Permeability Modeling Permeability parameters were calculated using the Delayed Lawrence and Lee Model. 13 According to the Delayed Lawrence and Lee Model, K ep represents the washout rate constant of the contrast agent from the EES to the IVS. The formula for K ep is: K ep appears to have an inverse relationship with BBBP, and therefore decreases with increased permeability. Importantly, K ep is independent of flow. 6 K trans equals the blood plasma flow per unit volume of tissue, derived through the formula: where E represents the extraction fraction of contrast agent that leaves the IVS in the first pass of the bolus through the vascular bed. E has no units assigned and has a value between 0 and 1. F represents the plasma flow per unit tissue volume and its units are mL ϫ mL Ϫ1 ϫ min Ϫ1 . K trans is dependent on the blood flow. PS represents the flow across the blood vessel wall from the IVS to the EES. The formula for PS is: where Hct is the hematocrit. According to recommendations contained in the Olea Manual, a default hematocrit level is set to 0.45 for the calculation of all derived perfusion and permeability parameters for which a hematocrit level is required. PS increases with increased permeability; however, PS is also flow-dependent. 6 In summary, K ep , PS, and K trans have previously been used to assess BBBP in patients 9,14 and animal models. 15,16 However, according to the Lawrence and Lee Model derivations above, decreasing K ep most accurately reflects increasing disruption and permeability of BBB, especially in the setting of impaired CBF, such as occurs in GCE. 12 Study Population Twenty-two patients were included in the study after excluding indeterminate patients (n ϭ 2) from the statistical analysis. In the 22 patients, 50% (11/22) were classified as having GCE and 50% (11/22) as non-GCE. There was statistically significant high interobserver reproducibility, with interobserver agreement coefficient of 0.86 (P Ͻ .0001) for sulcal effacement and 0.73 (P Ͻ .0001) for gray-white matter differentiation between the 2 blinded neuroradiologists. Clinical characteristics of the study population are summarized in Table 1. Global BBBP Analysis Quantitative BBBP results along with CTP results are summarized in Table 2. K ep , K trans , VP, and F were significantly decreased, while VE was significantly increased, in the GCE group compared with the non-GCE group. Differences in PS and extraction fraction of contrast agent did not reach statistical significance. However, there was a trend for increased PS in GCE. CBF was significantly decreased in the GCE group compared with the non-GCE group, while differences in CBV and MTT did not reach statistical significance. Fig 2 demonstrates quantitative BBBP maps in a representative patient from the GCE group and a representative patient from the non-GCE group, respectively. DISCUSSION Assessment of BBBP with CTP is a novel technique that has not been adopted in clinical practice despite its numerous practical advantages, including widespread availability and relatively low cost. Quantitative BBBP evaluation in the clinical setting promises new insights into the underlying pathophysiology of many neurologic diseases. The relationship between BBBP and microvascular dysfunction and neuroinflammation plays an important role in numerous conditions, such as GCE. GCE remains a major complication after SAH. Presently, assessment of GCE relies primarily on NCCT findings that are often difficult to identify in the setting of diffuse SAH and increased intracranial pressure. Given the influence of GCE on SAH morbidity and mortality, earlier detection of GCE may improve outcomes in SAH. 3 At present, there are no clinically established methods to assess the pathophysiologic mechanisms underlying GCE, such as microvascular dysfunction, inflammatory effects, and neurotoxic effects of blood. 3 Given that these factors also play a role in BBB dysfunction, we have evaluated the utility of BBBP parameters in the assessment of GCE after SAH. However, these findings may be applicable to a wide spectrum of conditions where cerebral edema is known to occur, such as acute stroke, traumatic and anoxic brain injury, and CNS infections. Our findings suggest that K ep decreases with increased BBBP 14,15 and may represent the most useful parameter in GCE because it is independent of low blood flow conditions. This inverse relationship between K ep and BBBP could be attributed to the underlying pathophysiologic mechanisms, including time kinetics. Kinetic analysis of BBBP changes during focused sonography-induced BBB opening demonstrated initial decrease in K ep with subsequent recovery over time. 16 Therefore, time kinetics are a crucial component in the assessment of the different BBBP parameters. In our study, we measured BBBP at a very early (initial) stage in patients with SAH who develop GCE demonstrating decreased K ep . However, future work will show if K ep recovers on follow-up examinations in these patients. The exact timing of BBBP imaging during the course of the disease is an important factor in understanding K ep in these patients, as it relates to the timing of the BBB breakdown. In contrast, K trans is dependent on blood flow and also decreased with increased BBBP in our study, representing a less reliable parameter for assessment of BBBP in GCE. Furthermore, our findings demonstrate significant differences in VE and VP in patients with GCE. The mathematical derivation of VE and VP is based on the fractions of volume of the EES and IVS, respectively. Fig 1 summarizes the relationships of these BBBP parameters under normal conditions and in the setting of GCE. In the setting of interstitial edema such as GCE, fluid from the IVS (represented by VP) transgresses the BBB into the EES (represented by VE). In GCE, there is an increase in EES as reflected by increased VE and decreased K ep , indicating increased BBBP. Furthermore, our findings also demonstrated significantly decreased K trans and F in patients with GCE. In the literature, CBF impairment recently has been reported in patients with GCE. 12 Our data demonstrated decreased CBF in patients with GCE compared with patients without GCE and is thus concordant with previously published NCCT images demonstrate loss of gray-white matter differentiation and presence of sulcal effacement in the patient with GCE. Corresponding BBBP maps reveal decreased K ep , slightly decreased K trans , and increased VE in the same patient compared with the patient without GCE. studies. 12 Because the derivation of K trans is dependent on F, it is a less reliable parameter in flow-restricted conditions, such as seen in GCE. 17 Given that K ep is independent of F, it may represent a more useful parameter to assess BBBP in the context of GCE, particularly when flow impairment occurs. 12 These findings further contribute to the development of an imaging biomarker of BBBP in flow-restricted conditions such as cerebral edema. In this initial exploratory study, our sample was limited to patients with SAH who had an extended CTP technique performed. Future studies evaluating BBBP parameters in other patient populations such as stroke, hypoxic-ischemic injury, traumatic brain injury, neoplasm, and inflammation will be important to improving our understanding of BBBP in different clinical settings as well as to evaluate BBBP parameters as potential outcome measures in specific disease states. Although there was a trend for increased PS in patients with GCE compared with patients without GCE, statistical significance was not achieved. However, these findings support further prospective work to elucidate the pathophysiologic processes leading to GCE after SAH. CONCLUSIONS Further study of BBBP is needed to improve diagnosis and monitoring of GCE. Validating a potential quantitative imaging biomarker may improve detection and management of GCE and may provide a quantitative end point for future clinical trials aimed at improving SAH outcomes.	2017-10-02T00:40:51.902Z	2016-09-01T00:00:00.000	{ "year": 2016, "sha1": "5c26aaeaa04115e27a356a31fa66673fcf518e49", "oa_license": "CCBY", "oa_url": "http://www.ajnr.org/content/ajnr/37/9/1599.full.pdf", "oa_status": "HYBRID", "pdf_src": "Anansi", "pdf_hash": "5c26aaeaa04115e27a356a31fa66673fcf518e49", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
11478847	pes2o/s2orc	v3-fos-license	Identification of SNAREs involved in synaptotagmin VII-regulated lysosomal exocytosis. Ca2+-regulated exocytosis of lysosomes has been recognized recently as a ubiquitous process, important for the repair of plasma membrane wounds. Lysosomal exocytosis is regulated by synaptotagmin VII, a member of the synaptotagmin family of Ca2+-binding proteins localized on lysosomes. Here we show that Ca2+-dependent interaction of the synaptotagmin VII C(2)A domain with SNAP-23 is facilitated by syntaxin 4. Specific interactions also occurred in cell lysates between the plasma membrane t-SNAREs SNAP-23 and syntaxin 4 and the lysosomal v-SNARE TI-VAMP/VAMP7. Following cytosolic Ca2+ elevation, SDS-resistant complexes containing SNAP-23, syntaxin 4, and TI-VAMP/VAMP7 were detected on membrane fractions. Lysosomal exocytosis was inhibited by the SNARE domains of syntaxin 4 and TI-VAMP/VAMP7 and by cleavage of SNAP-23 with botulinum neurotoxin E, thereby functionally implicating these SNAREs in Ca2+-regulated exocytosis of conventional lysosomes. Conventional lysosomes have been identified recently as the major intracellular compartment that undergoes Ca 2ϩ -triggered exocytosis in non-specialized secretory cells (1)(2)(3). This pathway is utilized by many cell types to reseal plasma membrane wounds (4,5) and is also subverted by the intracellular parasite Trypanosoma cruzi in order to invade mammalian cells (6,7). Previous studies showed that synaptotagmin (Syt) 1 VII, a ubiquitously expressed member of a family of putative Ca 2ϩ sensors for membrane fusion (8,9), is involved in the regulation of lysosomal exocytosis and of membrane repair (2,4,5). Syt VII-deficient mice show defects in cell resealing and develop a form of autoimmune myositis, with inflammation and increased collagen accumulation in the skin and skeletal muscle (5). Synaptotagmins contain two C 2 domains in their cytosolic region, which can bind phospholipids in response to Ca 2ϩ , as well as components of soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complexes (10). SNARE proteins are thought to be key mediators of all intracellular membrane fusion events (11). The most extensively characterized SNARE complex is the one involved in the exocytosis of synaptic vesicles in neurons. This complex consists of synaptobrevin/VAMP2 (vesicle-associated membrane protein 2) on the vesicle membrane, and syntaxin 1 and SNAP-25 (synaptosome-associated protein of 25 kDa) on the plasma membrane (12). Syntaxin 1 and VAMP2 each contribute one ␣-helical domain and SNAP-25 two domains to a parallel fourhelix coiled-coil bundle, which has been proposed to provide the force necessary to bring the opposing bilayers together and cause membrane fusion (13,14). There are multiple ubiquitously expressed isoforms of the three SNAREs involved in synaptic vesicle exocytosis, a finding that has led to the hypothesis that SNAREs might be determinants of the specificity of different fusion events within a cell (15). Syt I specifically interacts with the target SNAREs (t-SNAREs) syntaxin 1 and SNAP-25 (16), and these interactions are thought to be key for its proposed function as a Ca 2ϩ sensor for rapid synaptic vesicle exocytosis (10,17). In this study, we sought to identify SNARE proteins involved in the Ca 2ϩ -triggered fusion of lysosomes with the plasma membrane. In analogy to previous findings on Syt I-regulated synaptic vesicle exocytosis, we hypothesized that the SNARE complex mediating lysosomal exocytosis interacts with Syt VII and consists of a VAMP isoform on the lysosome and syntaxin and/or SNAP-25 isoforms on the plasma membrane. Consistent with this view, our studies identified interactions between Syt VII and a Ca 2ϩ -triggered SNARE complex containing the lysosomal vesicle SNARE (v-SNARE) TI (toxin-insensitive)-VAMP/ VAMP7 (18,19) and the t-SNAREs SNAP-23 (20) and syntaxin 4 (21). Lysosomal exocytosis was specifically inhibited by interfering with the capacity of these SNARE proteins to assemble into a complex, thus confirming their functional role in the process. Immunofluorescence-For staining with antibodies against TI-VAMP/VAMP7 and Lamp-1, NRK cells were grown on coverslips and fixed in 3.7% formaldehyde for 20 min at room temperature, followed by 50 mM NH 4 Cl for 15 min. After permeabilization with 0.01% saponin for 15 min, the cells were incubated with antibodies against Lamp-1 (1:100) and TI-VAMP/VAMP7 (1:100). For staining with antibodies against VAMP7 and synaptotagmin VII, NRK cells were grown on coverslips and fixed in 3.7% paraformaldehyde for 10 min at room temperature. After one wash with 0.1 M glycine, cells were permeabilized with 0.3% Triton X-100, blocked with 1% bovine serum albumin and 0.1% Triton X-100, and incubated with antibodies against TI-VAMP/VAMP7 (1:100) and Syt VII (1:200). Infection of Chinese hamster ovary cells with T. cruzi trypomastigotes was performed for 1 h as described previously (7). Plasma Membrane Isolation-NRK cells were washed twice with PBS and incubated with 15 mM Sulfo-NHS-Biotin (Pierce) in Biotinylation Buffer (10 mM sodium borate, pH 8.8, 150 mM NaCl) for 15 min at room temperature. The reaction was stopped by adding 10 mM NH 4 Cl. The biotinylated cells were then scraped into Cytosolic Lysis Buffer (10 mM HEPES, pH 7.4, 10 mM NaCl, 1 mM KCl, 5 mM NaHCO 3 , 1 mM CaCl 2 , 0.5 mM MgCl 2 , 5 mM EDTA, and protease inhibitor mixture), incubated on ice for 5 min, and disrupted with 25 strokes of a ball-bearing homogenizer. The resulting lysate was spun at 200 ϫ g to pellet nuclei and intact cells. The supernatant was incubated with Ultralink Immobilized Neutravidin beads (Pierce) for 1 h at room temperature. The beads were then washed twice in Cytosolic Lysis Buffer, resuspended in 1ϫ Sample Buffer, and analyzed by SDS-PAGE and immunoblotting. For the detection of SDS-resistant complexes, confluent monolayers of NRK cells were washed in PBS and treated with 1 mM N-ethylmaleimide (NEM) in PBS for 15 min on ice, followed by 2 mM dithiothreitol in PBS for 15 min on ice to inactivate NEM. Cells were scraped in either Cytosolic Lysis Buffer containing 1 mM CaCl 2 or 5 mM EGTA. Total membranes were prepared as described in the plasma membrane purification except that instead of binding to the Neutravidin beads, the homogenized cells were spun at 14,000 ϫ g to pellet all membranes, which were then mixed with 1ϫ SDS Sample Buffer and run on SDS-PAGE. Binding Assays-NRK cells were lysed in Buffer A (10 mM HEPES, NaOH, pH 7.4, 0.1 M NaCl, 1% Nonidet P-40, 2.5 mM MgCl 2 , 1 mM CaCl 2 ), spun in a microcentrifuge at 14,000 rpm for 5 min, and precleared by incubating with glutathione-Sepharose 4B for 2 h at 4°C. The supernatant was incubated overnight with glutathione-Sepharose 4B bound to either GST only or GST fused to the C 2 A domain of Syt VII. The beads were washed three times with Buffer A and eluted with 0.5 M NaCl in Buffer A, followed by an elution in the same buffer containing 5 mM EGTA instead of 1 mM CaCl 2 . The eluted fractions were precipitated with trichloroacetic acid, resuspended in 1 ϫ Sample Buffer, and analyzed by SDS-PAGE. For binding assays to His-tagged C 2 A constructs, the His-tagged C 2 A domain of Syt I or Syt VII was bound to Affi-Gel 10 beads in Buffer A for 4 h at 4°C. The beads were then treated with 0.25 M Tris, pH 8, for 1 h to block unreacted sites. NRK cells were treated with 1 mM NEM for 15 min on ice followed by treatment with 2 mM dithiothreitol for 15 min on ice. The cells were then lysed in Buffer A and incubated with the C 2 A-bound beads for 2 h. The beads were then washed three times in Buffer A, suspended in 1ϫ Sample Buffer, and analyzed by SDS-PAGE followed by immunoblotting. Immunoprecipitation experiments were performed using the Seize X IP kit (Pierce). NRK cell lysates prepared as described above were diluted in an equal volume of Binding/Wash Buffer and added to antibody-coupled beads. After a 1-h rotation at room temperature, the beads were washed three times, and bound proteins were eluted and combined with 3ϫ Sample Buffer and analyzed by SDS-PAGE followed by immunoblotting. Toxin Cleavage-Recombinant His-tagged BoNT/E was activated by treatment with 20 mM dithiothreitol for 30 min at 37°C and incubated with NRK lysates prepared as above for at least 1 h at 37°C. Reactions were stopped by adding 3ϫ SDS Sample Buffer and boiling for 5 min. Streptolysin-O Permeabilization and ␤-Hexosaminidase Secretion-NRK cells were plated in 60-mm dishes at a concentration of 5 ϫ 10 5 per 4 ml of Dulbecco's modified Eagle's medium ϩ 10% FBS and allowed to grow overnight. Cells were washed in ice-cold ␤-hexosaminidase Assay Buffer A (20 mM HEPES, 110 mM NaCl, 5.4 mM KCl, 0.9 mM Na 2 HPO 4 , 10 mM MgCl 2 , 2 mM CaCl 2 , and 11 mM glucose) and treated with streptolysin-O (SLO), obtained from S. Bhakdi, Mainz, Germany) at a concentration of 1 g/ml for 7 min on ice. The cells were then washed twice with ice-cold Low Ionic Strength Medium (LISM, 5 mM NaCl, 4.8 mM KCl, 1.2 mM MgSO 4 , 1.2 mM KH 2 PO 4 , 5 mM NaHCO 3 , 0.5 mM MgCl 2 , 20 mM HEPES/NaOH, pH 7.4, 10 mM glucose, 220 mM sucrose, 0.5% bovine serum albumin), incubated in LISM at 37°C for 5 min in the presence of varying concentrations of recombinant protein, and finally incubated in 5 mM free Mg 2ϩ and 2 mM ATP-containing Buffer B (20 mM HEPES, 40 mM KCl, 100 mM potassium glutamate, and 5 mM EGTA) with or without 1 M free Ca 2ϩ for 10 min at 37°C. The desired concentrations of free Mg 2ϩ and Ca 2ϩ were obtained with a Ca 2ϩ or Mg 2ϩ /EGTA buffering system calculated using the software by Foehr and Warchol. Supernatants were collected, and the cells were lysed in Nonidet P-40 to determine the total amount of ␤-hexosaminidase remaining in the cells. For detection of ␤-hexosaminidase activity, 350 l was incubated for 15 min at 37°C with 50 l of 6 mM 4-methylumbelliferyl-N-acetyl-␤-D-glucosaminide in sodium citrate phosphate buffer, pH 4.5. The reaction was stopped by the addition of 100 l of 2 M Na 2 CO 3 , 1.1 M glycine, and the fluorescence was measured at excitation 365 nm, emission 450 nm. For toxin treatment, cells were incubated in LISM at 37°C for 10 min in the presence of varying concentrations of activated toxin. Electroporation, Surface Staining for Lamp-1, and Flow Cytometry-Confluent monolayers of NRK cells were trypsinized and resuspended in Hanks' balanced salt solution containing 3% FBS and 10 mM HEPES, NaOH, pH 7, at a concentration of 1 ϫ 10 6 /ml. 400 l of this suspension was added to a Gene Pulser Cuvette (Bio-Rad), and cells were electroporated at 300 V and 450 microfarads, followed by a 5-min incubation on ice. After 1 min at 37°C, the cells were returned to ice and stained for surface-exposed Lamp-1. Briefly, cells were resuspended in 50 l of LY1C6 hybridoma supernatant and incubated at 4°C for 30 min. The cells were then washed, fixed with 2% paraformaldehyde, quenched with 10 mM NH 4 Cl, and stained with Alexa Fluor488 goat-anti-mouse antibodies (Molecular Probes) for 30 min, resuspended in 0.5 ml PBS, and analyzed by flow cytometry. A fluorescence-activated cell sorter (FACS), FACSCalibur (BD Biosciences), was used to excite the cells at 488 nm, and the emission was collected through a 530/30-nm bandpass filter. A minimum of 10,000 cells was analyzed in each sample. Data analysis was performed using CellQuest (BD Biosciences). Microinjection and Surface Staining for Lamp-1-Microinjection and staining for surface Lamp-1 was carried out as described previously (4). Briefly, 4.5 ϫ 10 5 /4 ml NRK cells were plated on coverslips and allowed to grow overnight. The coverslips were then transferred to Hanks' balanced salt solution containing 10% FBS and placed on a heated microscope stage at 37°C. Previously activated and/or heat-killed BoNT/E was microinjected into the cells at a concentration of 400 nM in microinjection buffer (150 mM potassium gluconate, 2 mM MgCl 2 , 10 mM HEPES) containing Texas Red dextran (4). The cells were returned to Dulbecco's modified Eagle's medium containing 10% FBS and allowed to recover at 37°C for 2 h, followed by treatment with 10 M ionomycin for 3 min at 37°C, and staining on ice for surface-exposed Lamp-1. Viability Assay-Confluent monolayers of NRK cells were scraped in the presence of buffer alone, 1 M BoNT/E, or heat-inactivated BoNT/E, plated out in serial dilutions of 100 l in a 96-well plate, and allowed to recover at 37°C. After 4 h, 10 l of 12 mM 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (Sigma) was added to each well, and the plate was returned to 37°C. After 4 h, 100 l of a 10% SDS, 0.01 M HCl solution was added to each well, and the plate was incubated overnight at 37°C, and the fluorescence was determined at 570 nm. RESULTS The v-SNARE TI-VAMP/VAMP7 Co-localizes with Syt VII on Lysosomes-The v-SNARE TI-VAMP/VAMP7 (18) was recently shown to be specifically targeted to late endosomes and lysosomes of HeLa cells by its amino-terminal Longin domain (19). This finding is consistent with prior observations in NRK cells, in which Myc-tagged TI-VAMP/VAMP7 was targeted to compartments containing the lysosomal glycoprotein Lamp-1 (25). Immunofluorescence with a monoclonal antibody against TI-VAMP/VAMP7 (22) confirmed that endogenous TI-VAMP/ VAMP7 is present on lysosomes in NRK cells. Extensive overlap of TI-VAMP/VAMP7 with Lamp-1 (Fig. 1A) and Syt VII (Fig. 1B) staining was observed. Previous studies showed that in NRK cells Syt VII is localized on Lamp-1-positive, dense lysosomes, which also contain the processed form of cathepsin L (2). Therefore, these data identified TI-VAMP/VAMP7 as a candidate for involvement in a complex with plasma membrane t-SNAREs to drive lysosomal exocytosis, because it is the only v-SNARE detected so far on the membrane of mature lysosomes in mammalian cells. Consistent with this view, endogenous TI-VAMP/VAMP7 was detected on recently formed intracellular vacuoles containing T. cruzi, the protozoan parasite that utilizes Ca 2ϩ -regulated lysosomal exocytosis for invasion ( Fig. 1C) (6). TI-VAMP/VAMP7 Interacts with the Plasma Membrane t-SNAREs Syntaxin 4 and SNAP-23-Previous studies indicated that syntaxins 2-4 and SNAP-23 are ubiquitously expressed plasma membrane t-SNAREs (21). SNAP-23, a ubiquitously expressed isoform of SNAP-25, has been detected on the plasma membrane of many cell types (26). It was initially cloned as a syntaxin 4-interacting protein, although it was also shown to be capable of binding multiple syntaxin isoforms in vitro, as well as VAMP1 and VAMP2 (20). To determine whether SNAP-23 and the known plasma membrane syntaxins were expressed in NRK cells, we prepared a purified plasma membrane fraction. Monolayers of NRK cells were surfacebiotinylated, and disrupted cells were incubated with avidincoated beads to specifically pull down the plasma membrane. Western blot with isoform-specific antibodies detected SNAP-23, syntaxin 3, and syntaxin 4 on the plasma membrane fraction of NRK cells, whereas TI-VAMP/VAMP7 and syntaxin 2 were absent ( Fig. 2A). Thus, syntaxin 3 and 4 and SNAP-23 represented potential t-SNARE partners for TI-VAMP/VAMP7 in lysosomal exocytosis. To determine which t-SNAREs were the best candidates for this role, co-immunoprecipitation experiments were performed in NRK cell lysates. As shown in Fig. 2B, a monoclonal antibody specific for TI-VAMP/VAMP7 co-immunoprecipitated syntaxin 4, but not syntaxin 2 or 3 or SNAP-23 from total NRK cell lysates. These findings are consistent with recent experiments in HeLa cells, which detected SNARE complexes containing green fluorescent protein-tagged TI-VAMP/VAMP7 and syntaxin 4 (19). In pulldown experiments, we found that endogenous syntaxin 2 and syntaxin 4 bound to GST-SNAP-23 but not to GST alone (Fig. 2C). This interaction seemed to be specific for these isoforms, as syntaxin 3 and syntaxin 6 (a Golgi syntaxin (27)) did not bind to GST-SNAP-23. Co-immunoprecipitation experiments, however, detected a preferential interaction of endogenous SNAP-23 with syntaxin 4 and not syntaxins 2 and 3 (Fig. 2D). Syt VII Interacts with the t-SNAREs Syntaxin 4 and SNAP-23-SNAP-23 has been shown to be involved in several Ca 2ϩregulated exocytic events, including neutrophil degranulation (28), secretion of dense core granules in platelets (29), mast cell degranulation (30), and zymogen granule exocytosis in pancreatic acinar cells (31). In neuroendocrine cells, SNAP-23 was recently shown to interact with Syt VII C 2 A-B in a Ca 2ϩ -dependent manner (32). The C 2 A domain of Syt VII has the capacity to bind SNAP-25-syntaxin 1a heterodimers with high affinity (33), a property not exhibited by the Syt I C 2 A domain (33,34). Consistent with these results, we found that the Syt VII C 2 A domain (previously shown to inhibit lysosomal exocytosis (2)) and not the Syt I C 2 A domain interacts with endogenous SNAP-23 in NRK cells (results not shown) and with GST-SNAP-23 in vitro (Fig. 3, A and B). In the presence of 1 mM Ca 2ϩ , GST-SNAP-23 specifically associated with His-tagged Syt VII C 2 A, but not Syt I C 2 A (Fig. 3A). Most interesting, the interaction of GST-SNAP-23 with His-tagged Syt VII C 2 A was enhanced in the presence of the H3 domain (the region involved in the formation of SNARE coiled-coil bundles) of syntaxin 4 but not of syntaxins 2, 3, or 6 ( Fig. 3B). We next examined the material pulled down from NEMtreated lysates (to inactivate N-ethylmaleimide-sensitive factor and stabilize SNARE complexes (18)) by beads coated with His-tagged C 2 A domains of Syt VII or of Syt I, for the presence of plasma membrane syntaxins which could be part of a SNARE complex with SNAP-23. It was reported previously (8) that the Syt VII C 2 A domain interacts in a Ca 2ϩ -dependent manner with multiple syntaxin isoforms in vitro. Our results confirmed this observation, but we also found that Syt VII C 2 A bound syntaxin 4 more effectively than Syt I C 2 A, whereas syntaxin 2 and 3 seemed to bind both synaptotagmin isoforms equally well (Fig. 3C). Our data thus suggests that syntaxin 4 and SNAP-23 are good candidates for t-SNAREs involved in a SNARE complex mediating lysosomal exocytosis, as both are located on the plasma membrane of NRK cells and specifically interact with Syt VII under conditions that preserve SNARE complexes. Ca 2ϩ Triggers Formation of an SDS-resistant SNARE Complex Containing TI-VAMP/VAMP7, Syntaxin 4, and SNAP-23-One characteristic of SNAREs is their ability to form high molecular weight SDS-resistant complexes that can only be dissociated upon boiling (35). In order to determine whether specific SDS-resistant SNARE complexes were formed upon lysosome exocytosis, we wounded NRK cells by scraping from the dish. This procedure causes plasma membrane wounding and Ca 2ϩ influx in the majority of the cell population, triggering lysosomal exocytosis and rapid resealing (4). NEM-treated NRK cells scraped in the presence of Ca 2ϩ were solubilized directly in SDS sample buffer, a condition that does not allow de novo formation of SNARE complexes in solution (35). Half of each sample was boiled, and the other half was kept at room temperature. Upon SDS-PAGE followed by Western blot, we detected high molecular weight, heat-sensitive complexes recognized by antibodies to syntaxin 4, SNAP-23, and TI-VAMP/ VAMP7 (Fig. 4A, arrowheads). These bands, which migrated with an apparent molecular mass greater than 220 kDa, were only present when cells were scraped in the presence of 1 mM Ca 2ϩ . SDS-resistant SNARE complexes containing endogenous SNAP-25, syntaxin, and synaptobrevin/VAMP2 have been reported to form a ladder of bands varying in size from 60 to 300 kDa in cell lysates (36). Under our conditions, which do not involve massive exocytosis as observed in specialized secretory cells (only about 10% of the total lysosomal population fuses with the plasma membrane upon Ca 2ϩ influx (1)), SDS-resistant complexes of greater than 220 kDa in size were the most readily detectable. Thus, the presence of TI-VAMP/VAMP7, SNAP-23, and syntaxin 4 in a Ca 2ϩ -dependent, SDS-resistant complex is consistent with the participation of these SNAREs in a complex responsible for lysosomal exocytosis. We predict that Syt VII regulates the formation of SNARE complexes promoting lysosomal exocytosis. Therefore, we also examined whether the SNARE proteins found to interact with Syt VII also bound TI-VAMP/VAMP7. As described above, Histagged Syt I or Syt VII C 2 A domains were used in pulldown experiments from NEM-treated NRK lysates. Western blotting detected TI-VAMP/VAMP7 bound to the C 2 A domain of Syt VII and not to the equivalent domain of the brain isoform Syt I (Fig. 4B). In a similar experiment, a GST fusion of the Syt VII C 2 A domain was used in pulldown experiments from NEMtreated NRK lysates. Proteins were eluted from the beads using high salt, followed by an elution with EGTA. Western blotting confirmed that TI-VAMP/VAMP7 bound to GST-Syt VII C 2 A and not to GST alone, and also showed that more TI-VAMP/VAMP7 was recovered in the EGTA elution, indicating a Ca 2ϩ -dependent interaction with Syt VII (Fig. 4C). These data suggest that TI-VAMP/VAMP7 interacts in a Ca 2ϩ -dependent manner with Syt VII, possibly as part of a SNARE complex. FIG. 3. The C 2 A domain of Syt VII interacts with SNAP-23 and syntaxin 4 in NRK cell lysates. A, GST-SNAP-23 or GST alone pulldowns of His-tagged C 2 A domains from Syt VII or Syt I in the presence or absence of Ca 2ϩ , run on SDS-PAGE and stained with Coomassie Blue to visualize bands. B, GST-SNAP-23 pulldown of Histagged Syt VII C 2 A domain in the presence of increasing amounts of His-tagged syntaxins 2-4, 6, or 7 H3 domains. Bound C 2 A was determined through densitometry of Coomassie Blue-stained bands. C, bound and unbound fractions of His-tagged synaptotagmin I or VII C 2 A domain pulldowns and Western blot (WB) with antibodies against syntaxins 2, 3 and 4. FIG. 4. Ca 2؉ influx into wounded NRK cells triggers formation of an SDS-resistant complex containing syntaxin 4, SNAP-23, and TI-VAMP/VAMP7, which interacts with Syt VII. A, boiled and unboiled membrane fractions of NEM-treated NRK cells, Western blotted with antibodies against SNAP-23, syntaxin 4, and TI-VAMP/VAMP7GST. Arrowheads point to the Ca 2ϩ -dependent complexes that migrate at Ͼ220 kDa. B, bound and unbound fractions of His-tagged Syt I or VII C 2 A domain pulldowns from NRK extracts, Western-blotted with anti-TI-VAMP/VAMP7 antibodies. C, GST alone or GST-Syt VII C 2 A domain pulldown from NRK extracts, eluted with 0.5 M NaCl or 5 mM EGTA, and Western-blotted with anti-TI-VAMP/VAMP7 antibodies. Functional Role of TI-VAMP/VAMP7, Syntaxin 4, and SNAP-23 in Lysosomal Exocytosis-A major piece of evidence supporting a role for SNAREs in controlling specific membrane fusion events is the inhibitory effect of recombinant ␣-helical coil domains, the regions involved in the formation of SNARE complexes. In cracked PC12 cells, the SNARE domains of VAMP2 and syntaxin 1a are more effective in inhibiting norepinephrine secretion, when compared with other isoforms (15). We showed previously that the recombinant Syt VII C 2 A domain specifically inhibits Ca 2ϩ -dependent lysosomal exocytosis using a ␤-hexosaminidase secretion assay in SLO-permeabilized NRK cells (2). We examined the effect of adding Histagged SNARE domains of TI-VAMP/VAMP7 or syntaxin 4, or of VAMP4 (37) or syntaxin 6 (27) as controls, to this assay. A dose-dependent inhibition of ␤-hexosaminidase release was observed in the presence of the H3 domain of syntaxin 4, reaching an inhibition of about 37% at 15 M (Fig. 5B). The same domain from the Golgi isoform, syntaxin 6, did not inhibit exocytosis at any concentrations tested. Similar dose-dependent inhibition was seen with the SNARE domain of TI-VAMP/VAMP7, reach-ing an inhibition of about 35% at 12.5 M (Fig. 5A). The SNARE domain of VAMP4, a Golgi VAMP isoform, had no effect. In previous studies (15) using cracked PC12 cells, similarly high concentrations of soluble SNARE domains were required to achieve inhibition of exocytosis. Previous work from our laboratory showed that lysosomal exocytosis has an important role in the resealing of plasma membrane wounds (4). We thus proceeded to verify whether specific recombinant SNARE domains were also inhibitory in a more physiological assay of plasma membrane wounding and repair. Cells were wounded by electroporation, and the rapid exocytosis-mediated repair was measured by detecting surface exposure of a luminal epitope of Lamp-1. Inhibition in Lamp-1 surface staining, quantified by flow cytometry, was detected when NRK cells were electroporated in the presence of the His-tagged syntaxin 4 H3 domain. In contrast, no difference was seen when His-tagged syntaxin 6 H3 domain was added (Fig. 5C). Similarly, we saw an inhibition of Lamp-1 surface exposure when cells were wounded in the presence of the TI-VAMP/VAMP7 SNARE domain, but no inhibition was observed with the equivalent domain from VAMP4 (Fig. 5D). The results of these two independent assays, which measured ␤-hexosaminidase release in SLO-permeabilized cells or the surface exposure of Lamp-1 after resealing of wounded cells, indicate that syntaxin 4 and VAMP7 are functionally involved in Ca 2ϩtriggered lysosomal exocytosis. SNAREs were first identified as candidates for involvement in membrane fusion due to their specific cleavage by clostridial neurotoxins (38,39). Consequently, these neurotoxins have been widely used as tools to dissect the function of specific SNARE proteins. Although botulinum neurotoxin serotype E (BoNT/E) specifically cleaves SNAP-25 at its COOH terminus, human and mouse SNAP-23 were found to be resistant to cleavage (18,40). However, BoNT/E was subsequently shown to be able to cleave canine and rat SNAP-23, also at a COOHterminal site (24,41). To confirm that effective cleavage of SNAP-23 could be achieved in NRK (rat) cells, we monitored the activity of purified His-tagged recombinant BoNT/E on cell extracts by using Western blot. Our antibody recognizes the extreme COOH terminus of SNAP-23, so cleavage is indicated by the disappearance of the SNAP-23 immunoreactive band. Near complete cleavage was observed after 1 h, at concentrations between 10 and 100 nM (Fig. 6A). We proceeded to examine whether SNAP-23 cleavage by BoNT/E had an effect on lysosomal exocytosis. Cells were permeabilized with SLO, and the toxin was added for 10 min at 37°C, and 1 M Ca 2ϩ was added to trigger exocytosis. Secretion of the lysosomal enzyme ␤-hexosaminidase was inhibited in a dose-dependent manner, with an inhibition of 51% reached at 4.4 M BoNT/E. In contrast, no inhibition was observed when the cells were incubated with the same concentrations of heatinactivated toxin (Fig. 6B). Given that only partial cleavage was probably achieved in the short 10-min incubation period allowed by the SLO permeabilization assay, these results strongly suggested that SNAP-23 is essential for lysosomal exocytosis. For this reason we also microinjected the toxin directly into NRK cells, and we looked for an effect on lysosomal exocytosis after a longer incubation period. BoNT/E was microinjected at a concentration of 400 nM, along with Texas Red dextran, to allow for subsequent identification of injected cells. After a 2-h incubation at 37°C to allow full cleavage, the cells were stimulated to undergo lysosomal exocytosis by the addition of 10 M ionomycin for 3 min (1). Exocytosis was monitored by exposure of the luminal epitope of Lamp-1, which we detected through immunofluorescent surface staining of non-permeabilized cells (1,2). All cells that were microinjected with the heat-inactivated toxin showed Lamp-1 surface staining following ionophore treatment. In contrast, injection of the active toxin caused a marked inhibition in Lamp-1 surface exposure, when compared with neighboring uninjected cells (Fig. 6C). In addition, the inhibitory effect of SNAP-23 cleavage on lysosomal exocytosis was shown not to be due to a generalized decrease in cell viability. NRK cells scraped from the dish in the presence of 1 M toxin fully recovered and showed normal viability after 4 h in an 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide assay (data not shown). These results thus confirm that the t-SNARE SNAP-23 plays an essential role in Ca 2ϩ -triggered exocytosis of lysosomes, probably by participating in the formation of a SNARE complex containing syntaxin 4 and TI-VAMP/VAMP7. DISCUSSION In this study we identified the v-SNARE TI-VAMP/VAMP7 and the t-SNAREs SNAP-23 and syntaxin 4 as components of a SNARE complex involved in the Ca 2ϩ -triggered exocytosis of conventional lysosomes. Several independent lines of evidence support this conclusion. First, TI-VAMP/VAMP7 is localized on lysosomes, and in cell lysates it interacts specifically with the plasma membrane t-SNAREs SNAP-23 and syntaxin 4, as well as with the lysosomal exocytosis regulatory protein Syt VII. Second, interaction of SNAP-23 with the C 2 A domain of Syt VII is Ca 2ϩ -dependent and facilitated by the SNARE domain of syntaxin 4. Third, Ca 2ϩ triggers formation of a membraneassociated, high molecular weight SDS-resistant complex containing TI-VAMP/VAMP7, SNAP-23, and syntaxin 4. Fourth, SNARE domains of TI-VAMP/VAMP7 and syntaxin 4, or cleavage of SNAP-23 by BoNT/E, inhibit Ca 2ϩ -triggered lysosomal exocytosis in NRK cells. The SNARE complex formed by syntaxin 1, SNAP-25, and VAMP2 during synaptic vesicle exocytosis resists SDS dissociation due to the formation of a highly stable coiled-coil bundle, which consists of four ␣-helical domains contributed by the three members (35,36). Work with recombinant proteins showed that many different SNARE proteins can form SDSresistant complexes in vitro (42,43), but their ability to associate promiscuously in vivo is significantly more restricted, probably due to the presence of accessory proteins that promote correct pairing (15,44). Thus, although SDS-resistant complexes can be formed de novo after cells are solubilized in Triton X-100, co-immunoprecipitation experiments have proven to be a reliable approach in the identification of components of functional SNARE complexes. In our study, regardless of the antibody used for immunoprecipitation, this approach consistently identified syntaxin 4, and not syntaxin 2 or 3, as a plasma membrane component of a complex also containing SNAP-23 and TI-VAMP/VAMP7. Syntaxin 4 has been implicated in other regulated exocytic trafficking pathways, including the translocation of GLUT4 to the plasma membrane of rat adipose cells. In that study, it was proposed to bind SNAP-23 and VAMP2 and/or VAMP3, forming a specific SNARE complex (45). In RBL-2H3 mast cells, syntaxin 4 was shown to be functionally involved in the exocytosis of secretory granules and to bind SNAP-23 and different VAMP isoforms although not to bind TI-VAMP/VAMP7 (46). In these cells, however, TI-VAMP/VAMP7 was shown to translocate to the plasma membrane upon stimulation (47). In platelets, SNAP-23 as well as syntaxins 2 and 4 have been implicated in ␣-granule release (48,49). Syntaxin 4 is therefore emerging as a multifunctional t-SNARE on the plasma membrane, which along with SNAP-23 appears to be able to mediate the exocytosis of various compartments, possibly by forming complexes with distinct v-SNAREs. After triggering lysosomal exocytosis by scrape-wounding NRK cells, we detected formation of a high molecular weight SDS-resistant complex of greater than 220 kDa, recognized by antibodies to VAMP7, SNAP-23, and syntaxin 4 (Fig. 4A). This complex was Ca 2ϩ -dependent and dissociated by boiling, leading us to conclude that it was likely to correspond to a trans-SNARE complex between the lysosome and the plasma membrane containing TI-VAMP/VAMP7, SNAP-23, and syntaxin 4. In these experiments, however, we also detected a Ca 2ϩ -independent complex that was slightly smaller, containing SNAP-23 and syntaxin 4, but not TI-VAMP/VAMP7. This complex may reflect SNARE complexes involved in constitutive traffic of other vesicles to the plasma membrane, possibly involving a distinct VAMP isoform. Alternatively, these Ca 2ϩindependent complexes may consist of SDS-resistant heterodimers of SNAP-23 and syntaxin 4. Multiple SDS-resistant complexes have been detected in chromaffin cells, and the majority of these complexes contained only SNAP-25 and syntaxin 1 as a stable dimer on the pathway to the ternary SNARE complex (50), with only a small subset containing VAMP2 (51). But it is important to note that a heterogeneous band pattern is also routinely observed with SDS-resistant ternary SNARE complexes from whole cell lysates (36), a phenomenon that has also been attributed to different folding intermediates, as op-posed to multimers of complexes (51). Earlier studies showed that the soluble C 2 A domain of Syt VII, or affinity-purified antibodies against this domain, has a strong inhibitory effect on the Ca 2ϩ -triggered exocytosis of lysosomes, as well as plasma membrane resealing and host cell invasion by T. cruzi (2,4,7). In addition to a Ca 2ϩ -dependent Syt VII-SNAP-23 interaction, consistent with what was previously described in neuroendocrine cells (32), we also detected TI-VAMP/VAMP7 in pulldown assays with the Syt VII C 2 A domain but not with the Syt I C 2 A domain (Fig. 4, B and C). Syt I is known to bind to the heterotrimeric complex consisting of syntaxin 1, SNAP-25, and VAMP2 (52,53), and there is no strong evidence indicating a direct interaction between the C 2 A domain of Syt I and VAMP2. Our data suggests that the interaction we detected was also between Syt VII C 2 A and a SNARE complex containing TI-VAMP/VAMP7, in addition to SNAP-23 and syntaxin 4 (Fig. 4A). Previous studies indicated that lysosomal exocytosis plays an important role in the mechanism used by the human parasite T. cruzi to invade host cells and in the repair of plasma membrane wounds (4,7). A role for the lysosomal SNARE TI-VAMP/VAMP7 in the exocytic events mediating repair is consistent with the deficient resealing observed in Madin-Darby canine kidney cells expressing the dominant-negative amino-terminal Longin domain (54) of TI-VAMP/VAMP7 (results not shown). Identification of specific SNARE proteins involved in this process opens the possibility for specific intervention strategies in trypanosome infections. Our findings also represent, to our knowledge, the first description of synaptotagmin-SNARE protein interactions regulating Ca 2ϩ -triggered exocytosis in non-specialized secretory cells. Our findings suggest that additional members of the synaptotagmin family may also function by regulating the formation of SNARE complexes involved in specific Ca 2ϩ -dependent membrane fusion events.	2018-04-03T00:22:50.719Z	2004-05-07T00:00:00.000	{ "year": 2004, "sha1": "3f0d5e22fad6022d1da362cb7eebcbbb7fd16cfc", "oa_license": "CCBY", "oa_url": "http://www.jbc.org/content/279/19/20471.full.pdf", "oa_status": "HYBRID", "pdf_src": "Adhoc", "pdf_hash": "8f36756a22c137df583f5f6dfd14c2ea93131585", "s2fieldsofstudy": [ "Biology", "Chemistry" ], "extfieldsofstudy": [ "Biology", "Medicine" ] }
255593351	pes2o/s2orc	v3-fos-license	Do health service contacts with community health workers influence the intention to use modern contraceptives among non-users in rural communities? Findings from a cross-sectional study in Nigeria Background Studies in many developing countries have shown that community health workers (CHWs) are valuable for boosting contraceptive knowledge and usage. However, in spite of the evidence, studies in Nigeria have rarely examined whether in the absence of skilled health personnel such as doctors and nurses in rural and remote communities, the health service contacts of non-users with CHWs drive the intention to use modern contraceptives. This study, therefore, examines the extent to which health service contacts with CHWs are associated with the intention to use modern contraceptives among non-users in rural communities of Nigeria. Methods This study adopted a descriptive cross-sectional design. Data were extracted from the most recent Nigeria Demographic and Health Survey (NDHS). The study analyzed a weighted sample of 12,140 rural women. The outcome variable was the intention to use modern contraceptives. The main explanatory variable was health service contacts with CHWs. Statistical analyses were performed at three levels with the aid of Stata version 14. Three multivariable regression models were estimated using an adjusted Odds Ratio (aOR) with a 95% confidence interval. Statistical significance was set at p < 0.05. Results Findings showed that more than a quarter (29.0%) of women intends to use modern contraceptives. Less than one-fifth (15.9%) of the women had health service contacts with CHWs. In Model 1, women who had health service contacts with CHWs were more likely to intend to use modern contraceptives (aOR =1.430, 95% CI: 1.212–1.687). Likewise, in Model 2, women who had health service contacts with CHWs had a higher likelihood of intending to use modern contraceptives (aOR = 1.358, 95% CI: 1.153–1.599). In Model 3, the odds of intention to use modern contraceptives were higher among women who had health service contacts with CHWs (aOR =1.454, 95% CI: 1.240–1.706). Conclusion In rural areas of Nigeria, health service contacts with CHWs are significantly associated with the intention to use modern contraceptives. Family planning programmers should leverage the patronage of CHWs for the purpose of family planning demand generation in rural areas. Background The intention to use a modern contraceptive indicates the proportion of all women of reproductive age who are not currently using a modern method but have the intent to become modern contraceptive users in the future. This indicator is crucial to family planning programming particularly the family planning demand generation in Nigeria for two reasons. One, it provides information about the current level of non-use of modern contraceptives in the country. Evidence suggests that the non-use of modern contraceptives is not only high in Nigeria but also higher in rural areas [1], as well as in the northern parts of the country [2] compared to the urban and southern parts of the country. Two, it provides information about the extent to which current non-users plan to utilize modern contraceptives in the future. Such information drives the expansion of existing family planning service delivery to levels that anticipates and accommodate future demand for modern contraceptives. It also justifies the need for continued research focusing on the intention to use modern contraceptives among childbearing women. In Nigeria, modern contraceptive prevalence remained low among young and older women of reproductive age [3,4] and young and older men [5] though with substantial state and regional variations across the country [6,7]. For instance, while modern contraceptive use was 12% among currently married women, the use was 28% among unmarried women [7]. Also, the prevalence rate was lower than 10% in northern Nigeria compared to more than 15% in southern Nigeria [2,3,8]. Furthermore, modern contraceptive use was 26% among currently married women in urban areas compared to 10% among a similar group of women in rural areas [7]. Evidence also shows that a high level of unmet need for contraception exists in the country [8,9]. For instance, a recent study found that the unmet need for contraception exceeded 20% in many states of Nigeria. The 2018 Nigeria Demographic and Health Survey reported the unmet need for contraception to be 35% in Cross River State, 33% in Edo State, and 10% each in Anambra and Adamawa states [7]. These pose a significant threat to achieving the national contraceptive prevalence target as declared in the current national policy on population for sustainable development [10]. The policy seeks to increase the modern contraceptive prevalence rate to 27% by 2020, and successively gain an additional 2% increase yearly until 2030. The policy was built on the 2014 national family planning blueprint, which revised the national contraceptive prevalence goal from 36 to 27% by 2024 [11,12]. The development of the national family planning blueprint was driven with a view to achieving these targets through improve family planning demand generation and service delivery, particularly in the rural areas of the country where the use of modern contraceptives is lower compared to the urban areas. The family planning blueprint included sets of workable activities intended to generate more demand for family planning. This includes the development of the national family planning communication plan, the rejuvenation of the national health promotion forum, and the launching of the new family planning logo codenamed 'the Green Dot' . These activities utilized several health and non-health resources such as health communication experts, media organizations, the national orientation agency, faith-based organizations, and existing health agencies. These activities have improved public awareness of family planning in the country [12]. Nevertheless, some health resources that could further enhance the adoption of modern contraceptives by current non-users are still underutilized in the existing framework. One such health resource is the pool of Community Health Workers (CHWs). This refers to individuals who live and work in rural and remote communities to provide basic health services [13,14]. In Nigeria, there are four cadres of CHWs, namely, community health officers, community health extension workers, junior community health extension workers, and community resource persons [15]. The CHWs received limited health-related training certified by the Community Health Practitioners Registration Board of Nigeria. This enables CHWs to provide recognized specific primary health care both in a health facility and as well as in the community [16,17], to support the services provided by skilled health personnel such as doctors and nurses/midwives, who in many instances are not readily available in rural and remote communities [18][19][20]. The range of duties performed by CHWs includes but is not limited to supporting maternal and child health [21][22][23][24], helping to control communicable and non-communicable diseases [25], treatment of endemic diseases [26], promoting adequate nutrition [27], providing support during outbreaks of pandemics [17,28], home visits, referral of patients, and disease surveillance [29]. More importantly, studies in many developing countries have shown that CHWs are valuable for boosting contraceptive usage in rural communities [30][31][32]. In Nigeria, a recent study [33] based on the PMA2020 survey in six states investigated the positive impact of CHWs on modern contraceptive use among childbearing women in rural parts of the country. The study revealed that women visited by CHWs reported higher usage of modern contraceptives which corroborates an earlier finding in India that health workers' outreach may influence intention to use contraceptives [34]. In recognition of the value of CHWs in boosting contraceptive prevalence in the country, the Nigeria National Council on Health [NCH] (2012) approved that community health extension workers be allowed to provide injectable contraceptives in communities, which is a practice already found to be impactful in other climes [35,36]. In spite of the evidence of CHWs relevance to contraceptive usage in the country, studies have rarely examined whether in the absence of skilled health personnel such as doctors and nurses in rural and remote communities, the health service contacts of non-users with CHWs drive the intention to use contraceptives. This is important because, in many rural and remote communities of Nigeria, skilled health personnel is not readily available at health facilities in addition to other sundry challenges such as long distances to the facility, lack of essential drugs, and poverty [18,20]. These challenges discourage the use of essential healthcare services in rural areas. The challenges are now aggravated by widespread insecurity of lives and properties in the country [37,38] which has led to increasing non-availability of skilled health personnel in rural areas of the country. Some rural dwellers may thus be limited to contacting CHWs for antenatal, delivery, postnatal checks, family planning, and treatment of diverse ailments. Often the health service contacts take place in non-clinical spaces, and sometimes the CHWs visit homes, markets, and farmlands to provide health education and basic services to the people. Hence, if CHWs have sufficient training and education about modern contraceptives, the health service contacts with non-users could be an avenue to talk about family planning, clear misconceptions about family planning, and addresses other family planning concerns of the rural populace. The study, therefore, examines the extent to which health service contacts with CHWs are associated with the intention to use modern contraceptives among non-users in rural communities. Findings will provide further inputs for strengthening the family planning demand generation activities of the family planning blueprint. It will also shed light on how CHWs could contribute to the achievement of the national contraceptive targets sets for 2030 and 2050 in Nigeria. Design and Data This study adopted a descriptive cross-sectional design using a quantitative approach to determine whether the health service contacts of non-users with CHWS are associated with the intention to use contraceptives. The women's datasets analyzed in the study were extracted from the 2018 Nigeria Demographic and Health Survey (NDHS). The 2018 NDHS is the current (seventh) round of the Demographic and Health Survey (DHS) program implemented in Nigeria by the National Population Commission (NPC) with the collaboration of related national agencies, and the technical support of the Inner-City Fund (ICF) obtained through the DHS Program [7]. The DHSs are conducted across developing countries to build national capacity for the production of demographic and health data, and to provide estimates of demographic and health characteristics required for the monitoring of the Sustainable Development Goals in the participating countries [39,40]. The 2018 NDHS thus provides internationally comparable demographic and health information that remains valid until the next round of the DHS in 2023. Comprehensive information about the methodology of the 2018 NDHS has been published and made available in the public domain via https:// dhspr ogram. com/ pubs/ pdf/ FR359/ FR359. pdf. Population and Sample The 2018 NDHS covered 41,821 women of childbearing age. However, some of the women who were not relevant to the current study were excluded. Women excluded include all urban women (16,984), women currently using a method of contraception (8300), infecund women (572), and women who had health service contacts with skilled health personnel such as doctors and nurses/midwives (3121). Thus, the study analyzed a weighted sample of 12,140 women. This represents all women in the 2018 NDHS who were rural dwellers, reported CHWs as the provider of antenatal care, delivery, or postnatal care, and were not currently using a method of contraception. The sample was weighted using the weighting factors available in the dataset. Research Variables The outcome variable in the study was the intention to use modern contraceptives. This was derived from responses to the question: "Do you think you will use a contraceptive method to delay or avoid pregnancy at any time in the future?" The responses were in three categories, namely, intend to use later, unsure about future use, and do not intend to use modern contraceptives later. The study focused on women who were sure of their intention to use modern contraceptives later. We, therefore re-grouped the outcome variable into two categories, namely, 'intend to use' or 'otherwise' . This measure was consistent with how the intention to use contraceptives was operationalized in existing studies [41][42][43][44]. The main explanatory variable in the study was health service contacts. This was measured among childbearing women who received antenatal care, delivery care, or postnatal care solely from any of the cadres of CHWs and not from skilled health personnel. The variable was grouped into two categories of 'health service contact' or 'no health service contact' Three other sets of explanatory variables were included in the analysis for the purpose of making the study findings robust. One, eight individual demographic and social characteristics were examined. These are maternal age group (15-24, 25-34, and 35 years or older), timing of marriage (early -before 18 years or not early -18 years or older), maternal education (none, primary, secondary, and higher), parity (primiparity [one child], multiparity [two to four children], and grand multiparity [five or more children]), working status (employed or unemployed), fertility desire (wants within 2 years, wants after 2 years, wants but unsure of timing, undecided, and wants no more) and, mass media exposure (low, moderate, and high). This was derived from the frequency of reading newspapers, listening to the radio, or watching television. A total of nine points was generated and subsequently divided into three equal parts to indicate low, moderate, and high exposure. Also, religion (Christianity, Islam, and others) was examined. Previous studies [42,43,[45][46][47] have shown that these variables are important correlates of the intention to use contraceptives. Two, six household characteristics were examined. These are household wealth quintile (poorest, poorer, middle, richer, and richest), partners' education (none, primary, secondary, and higher), financial autonomy (autonomous or not autonomous), and healthcare autonomy (autonomous or not autonomous). Healthcare and financial autonomy were derived from responses to who had the final say on women's own health and spending of the respondents' cash earnings. Women who solely had the final say or jointly with partners were grouped as 'autonomous' and 'not autonomous' if otherwise. Safer sex negotiation (able to negotiate or unable to negotiate) was also included. Women who reported they could refuse sex from their partners or could ask partners to use a condom during intercourse were categorized as 'able to negotiate' . The last household factor included was the decision-maker for the non-use of contraceptives (respondent, partner, joint, and others). These variables are included to reveal the power dynamics within the household in patriarchal societies such as Nigeria, which have been established in existing studies as important influencing factors on women's use or non-use of modern contraceptives [48][49][50][51]. Three, some variables were selected for statistical control in the study. These are attitudes to wife-beating (not acceptable if the respondent did not justify wife-beating in any circumstance or acceptable if the respondent justifies wife-beating in specific circumstances), visitation by a family planning worker in the last 12 months (visited or not visited), pregnancy termination experience (never or ever experienced), and geo-political zone of residence. These variables have been linked to the intention to use modern contraceptives in previous studies [41,42]. Pregnancy termination experience was included because evidence suggests that women who have experienced induced abortion, may develop an interest in the use of contraceptives especially when such women have been exposed to postabortion care and counselling [52,53]. Data Analysis Statistical analyses were performed at three levels with the aid of Stata version 14 [54]. At the first level, descriptive statistics were used to present sample characteristics and the intention to use modern contraceptives. At the second level, the explanatory variables were crosstabulated with the outcome variable to assess variations in the intention to use contraceptives due to variations in the explanatory variables. Two bivariate analyses were further performed to select variables for inclusion in the final stage of data analysis. Firstly, a binary logistic regression using the unadjusted Odds Ratio (uOR) with a 95% confidence interval was used to assess associations between the explanatory and outcome variables. The control variables were excluded from the assessment. Only variables that reveal significance at p < 0.025 were selected. Two, a Variance Inflation Factor (VIF) was performed to check the extent of multicollinearity among the variables. Variables were selected into the multivariable model on the basis of either showing significance by the estimates of the uOR or having a VIF score of less than ten. This was done to maintain the statistical principle that a variable with a VIF score of 10 or higher scores signifies the presence of multicollinearity [55]. At the third stage of data analysis, three multivariable regression models were estimated using an adjusted Odds Ratio (aOR) with a 95% confidence interval. Three models were estimated for the purpose of assessing the strength of health service contacts given the introduction of other variables into the successive models. Model 1 controlled for the individual demographic and social characteristics, while Model 2 controlled for the individual characteristics, and household characteristics. Model 3 was the full model which included all the research variables. Statistical significance was set at p < 0.05. of the respondents. More than a quarter (29.0%) of the respondents intends to use modern contraceptives later. Less than one-fifth of them (15.9%) had health service contact with CHWs, while the majority had no health service contact with CHWs. The proportions of women in older age groups were higher compared to the proportion in the younger age group of 15-24 years. The majority (71.6%) of the respondents reported early marriage while more than a quarter (28.4%) of the women reported otherwise. The proportion of the respondents with a desire for an additional child was high though those who desired to have another child within 2 years were slightly more than those who desired another child after 2 years. The majority of the women (63.3%) had no formal education. Also, the majority (65.3%) were employed at the time of the survey. More than half of the respondents (55.5%) had low mass media exposure. Likewise, more than one-third of them (37.7%) had moderate mass media exposure. Muslim women were dominant among the respondents. Nearly half (46.5%) of the respondents were grand multiparous. Univariate Results Most of the women belong to either the poorest or poorer household wealth groups. The majority (62.3%) of respondents' partners were not formally educated. While the majority (73.7%) of the respondents had no healthcare autonomy, more than half (54.7%) of them had no financial autonomy. Slightly more than two-fifths (42.1%) of the respondents were able to negotiate safer sex with partners. Though the proportion (30.1%) of respondents who were decision-makers for non-use of contraceptives was substantial, the proportion (28.8%) who reported joint decisions with a partner was equally substantial compared to the proportion (18.5%) whose partners decide their contraceptive non-use. The majority of the respondents were not visited by family planning workers in the last 12 months preceding the survey. Though more than half of the women rejected wife-beating under any circumstance more than two-fifths of them justified wife-beating. The majority of the respondents had never experienced a pregnancy termination. Women from the northern region particularly the northwest geo-political zone were preponderant in the sample. Table 2 presents the bivariate findings. Health service contacts were significantly associated with the intention to use modern contraceptives with a higher proportion of the intention to use contraceptives among women who had health service contacts with CHWs compared to women who reported otherwise (38.6% vs. 27.2%). The intention to use contraceptives increased from 34.6 to 35.4% as the maternal age group increased from 15 to 24 years to 25-34 years but decline drastically to 19.5% among women in the advanced age group. The uOR at advanced maternal age (uOR = 0.459; 95% CI: 0.406-0.519) confirms a significant negative association between the maternal age group and the intention to use modern contraceptives. Though the intention to use modern contraceptives was lower among women who reported early marriage compared to those who reported otherwise (28.8% vs. 30.0%), the association was however not significant. Parity and the intention to use modern contraceptives were negatively associated. As women's parity status was increasing from primiparity to multiparity and to grand multiparity, the intention to use modern contraceptives was declining steadily. With the exclusion of women who wanted no more children, the intention to use modern contraceptives was lower (25.3%) among women who wanted another child within the next 2 years compared to women who either wants after 2 years (38.6%), or were undecided (27.5%) or unsure (36.5%) of the timing of future fertility. Bivariate Results Maternal education was positively associated with the intention to use modern contraceptives with a consistent increase in the intention to use modern contraceptives as educational attainment improves. However, the level of intention to use contraceptives dropped as the educational level reached higher education. Working status was significant (uOR = 1.203; 95% CI: 1.072-1.349) and positively associated with the intention to use modern contraceptives with higher intention among employed compared to unemployed women (30.3% vs. 26.5%). Likewise, mass media exposure and the intention to use modern contraceptives were positively related to a consistent increase in the level of intention as exposure to mass media improves. Christian women reported a higher intention to use modern contraceptives compared to women in other religions. Except for the richest household wealth group, the intention to use modern contraceptives increase progressively as the household wealth group improves. This reveals a positive association. Also, the intention to use modern contraceptives increase with improvement in partners' education but the level of intention dropped at higher educational attainment. While healthcare autonomy was positively associated with the intention to use a modern contraceptive, financial autonomy was negatively associated with the intention to use modern contraceptives. Safer sex negotiation relates positively to the intention to use modern contraceptives. In contrast, the association between decision-makers for the non-use of contraceptives and the intention to use contraceptives was inconsistent. Table 3 presents further association of health service contacts and other characteristics with the intention to use modern contraceptives. In Model 1 which controlled for the individual demographic and social characteristics, the odds of the intention to use modern contraceptives were higher among women who had health service contacts with CHWs compared to the intention among women who had no health service contacts with CHWs (aOR = 1.430; 95% CI: 1.212-1.687). In the model, the timing of the first marriage was the only individual characteristic with no significant effect on the intention to use modern contraceptives. With the introduction of household characteristics into Model 2, two noticeable changes were observed. One, the odds of the intention to use modern contraceptives reduced from 1.430 in Model 1 to 1.358 in Model 2 though the odds remain higher among women who had health service contacts with CHWs compared to those who had no health service contacts (aOR = 1.358; 95% CI: 1.153-1.599). Two, religion no longer reveals a significant effect on the intention to use modern contraceptives. As also observed in Model 1, the timing of the first marriage did not show a significant effect on the intention to use modern contraceptives. Two of the household characteristics, namely, household wealth and healthcare autonomy had no significant effects on the intention to use modern contraceptives. In the full model (Model 3), health service contacts were strengthened by the inclusion of the control variables. Women who had health service contacts with CHWs were more likely to intend to use modern contraceptives In the full model, all the individual demographic and social characteristics revealed a significant association with the intention to use modern contraceptives. While maternal age group and religion reduce the odds of the intention to use modern contraceptives, other individual characteristics, namely, the timing of first marriage, parity, fertility desire, maternal education, working status, and exposure to mass media increases the odds of the intention to use modern contraceptives. On one hand, two household characteristics, namely, household wealth and healthcare autonomy remained without statistical significance as observed in the previous model. On the other hand, other household characteristics, namely, partners' education, financial autonomy, safer sex negotiation, and decision-maker for non-use of contraceptives reveal significant association with the intention to use modern contraceptives. Geo-political zone of residence was the only control variable that showed significant effects on the intention to use modern contraceptives with higher odds in the North-East and North-West compared to lower odds in the three Southern zones. Discussion This study examined the extent of the association between health service contacts with CHWs and the intention to use modern contraceptives in settings with limited access to skilled health personnel such as doctors and nurses/midwives. This study builds on existing RC (Reference category), p < 0.05, *p < 0.01 [30][31][32]36] and was carried out to provide more information on an additional initiative that may boost modern contraceptive usage in rural and remote communities of Nigeria in light of the grossly insufficient numbers of skilled health personnel in such communities [19,20], which makes the CHWs an important provider of reproductive health services in such communities [13,23,24]. This presents an opportunity for family planning programmers to leverage the patronage of CHWs for the purpose of family planning demand generation by encouraging CHWs to use every health service contact with the rural populace to talk about family planning, clear family planning misconceptions, and provide enlightened information about the health concerns of family planning usage. Such initiative will go a long way to reduce the existing disparity not only between the rural and urban areas [1] but also current state and regional disparities [2,6] in family planning knowledge and use in the country, and also accelerate the achievement of the national contraceptive targets sets for 2030 and 2050 in Nigeria [10]. The study found a 29.0% prevalence of the intention to use modern contraceptives. When compared to findings in existing studies, the prevalence revealed in the current study could be explained in two possible ways. One, the prevalence found is substantially lower than the 52.2, 44.11, and 44.1% respectively reported in three studies conducted outside Nigeria [41,42,44] which analyzed similar DHS datasets, and lower than the prevalence found in a Nigerian study [47] that also analyzed the NDHS dataset. It is possible to explain the differences in the prevalence by the group of women analyzed in the studies. While this study focused on rural women who only had health service contacts with CHWs, the other studies conducted outside Nigeria included all childbearing women, and the study conducted in Nigeria included parous women in both urban and rural areas of the country. The prevalence as found in this study may thus be expected to differ from those found in studies that covered larger groups of women. Two, the prevalence found in the current study also differs from the prevalence reported in an existing hospital-based study [43]. Usually, hospital-based studies covered a smaller proportion of childbearing women which often makes the result inconsistent with findings in population-based studies. Notwithstanding, the prevalence of the intention to use modern contraceptives among non-users as found in this study represents a key opportunity for family planning demand generation in the country. Since CHWs mostly live in the community, they have an understanding of the reproductive norms and culture of the people. This makes them suitable for presenting family planning in a way that is sensitive to the culture, religion, and tradition of the indigenous people, particularly during health service contacts of rural women and may justify higher intention to use modern contraceptives as found among women who had health service contacts with CHWs in the study. The implementation of the family planning blueprint in Nigeria [11,12] will become more effective by expanding the range of family planning services that CHWs may provide. Incidentally, the existing strategy in the country has empowered CHWs to administer injectable contraceptives [11] but more roles could still be assigned particularly in making resources available for more home visits and family planning outreach programs in rural communities. A recent study in Nigeria [33] confirmed that home visits of CHWs have a strong impact on the uptake of modern contraceptives, which indicates that the strategy is very useful for improving knowledge and the use of modern contraceptives. Though the intention to use modern contraceptives does not automatically translate to future use due to a number of health service factors such as inadequate health services providing contraceptive services, service provision bias, poor private sector participation, and ineffective supply chain management [12], once knowledge is improved, and the intention to use is established, actual use may not face difficult opposition. Beyond health service contacts with CHWs, this study also found that individual demographic and social characteristics such as age, parity, maternal education, work status, mass media exposure, fertility desire, and religion are important drivers of the intention to use modern contraceptives. This provides support for findings in several existing studies conducted in Nigeria and elsewhere [41,46,47]. Also, the study revealed that several relational characteristics such as partners' education, women's financial autonomy, safer sex negotiation, and participation in household decisions may have a strong influence on the intention to use modern contraceptives in line with findings in previous studies [42,43]. The import of these findings is that the socio-demographic context of rural women cannot be exonerated from poor use or non-intention to use modern contraceptives. In most rural communities of Nigeria, the social structure is largely patriarchal which subjugates women's economic and reproductive lives to male authority. A social transformation of society through improved education and economic empowerment of rural women is a veritable platform for changing the family planning situation in rural and remote communities of Nigeria. This point has been stressed in previous studies [48][49][50][51] that call for increasing women's empowerment to improve the sexual and reproductive health of women in developing countries. Strengths and Limitations The literature search confirms that the focus of this study has rarely been examined across developing countries where CHWs play important role in the provision of primary health care. The study thus expands the frontiers of knowledge about the relevance of CHWs in relation to improvement in modern contraceptive usage in rural communities. The use of the DHS datasets in the study provides a basis for the international comparability of the study methods and findings. The principal investigator will be willing to share the do file with any interested researcher to enhance its replication in other climes. It is also important to draw attention to the few drawbacks of the study. One, the health service contacts recognized in the analysis were limited to contacts for maternal healthcare services such as antenatal care, delivery, postnatal check for mothers, or postnatal check for a child. Health services contacts for other health issues were excluded. This was done to ensure that the women concerned are those susceptible to pregnancy and child delivery concerns which makes family planning a relevant matter to them. The inclusion of other health contacts may moderate the findings. Also, the data analyzed did not capture the possibility of some women contacting both doctors/ nurses and CHWs, which may further moderate the findings. In addition, it is important to note that the study did not exclude women based on the recency of last delivery. This is important because the intention to use contraceptives as measured in the study focused on the likely future use of contraceptives and not necessarily the intention to use contraceptives after the last delivery. Two, the cross-sectional nature of the analyzed data did not permit the establishment of cause and effect between health service contacts and the intention to use modern contraceptives. It, however, provides a significant association between them which is important for understanding how health service contacts relate to the intention to use modern contraceptives. Three, the application of the theory of planned behavior may have improved the study findings, however, the theory could not be applied due to the unavailability of variables that could appropriately measure most of the theoretical constructs. Finally, the study only considered a binary measure of health service contacts. The number of contacts was not captured in the analyses. Follow up studies may thus include the number of health service contacts as another exposure variable that may affect the intention to use contraceptives. Conclusion This study examined whether health service contacts with CHWs are significantly associated with the intention to use modern contraceptives among non-users in rural communities of Nigeria. The women's datasets of the most recent Nigeria Demographic and Health Survey were analyzed based on a sample of 12,140 rural women. Findings from the analysis revealed that more than a quarter of the rural women intend to use modern contraceptives, and the odds of intention to use modern contraceptives were higher among women who had health service contacts with CHWs. In rural areas of Nigeria, health service contacts with CHWs are significantly associated with the intention to use modern contraceptives. This implies that family planning programmers should leverage the patronage of CHWs for the purpose of family planning demand generation in rural areas of Nigeria.	2023-01-12T06:16:20.781Z	2023-01-10T00:00:00.000	{ "year": 2023, "sha1": "831bba969af290d67be187cfc22960720f60e5a3", "oa_license": "CCBY", "oa_url": "https://bmchealthservres.biomedcentral.com/counter/pdf/10.1186/s12913-023-09032-3", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "5d68f34e14870efb3de20236469b6090d85e0db6", "s2fieldsofstudy": [ "Medicine", "Political Science" ], "extfieldsofstudy": [ "Medicine" ] }
17991860	pes2o/s2orc	v3-fos-license	Measuring Emotion in Parliamentary Debates with Automated Textual Analysis An impressive breadth of interdisciplinary research suggests that emotions have an influence on human behavior. Nonetheless, we still know very little about the emotional states of those actors whose daily decisions have a lasting impact on our societies: politicians in parliament. We address this question by making use of methods of natural language processing and a digitized corpus of text data spanning a century of parliamentary debates in the United Kingdom. We use this approach to examine changes in aggregate levels of emotional polarity in the British parliament, and to test a hypothesis about the emotional response of politicians to economic recessions. Our findings suggest that, contrary to popular belief, the mood of politicians has become more positive during the past decades, and that variations in emotional polarity can be predicted by the state of the national economy. Introduction Our main goal in this paper is to adapt affective computing methods to the study of political discourse. We develop a methodology to produce domain-specific polarity lexicons and implement this approach using the entire corpus of proceedings of the British House of Commons during the past one hundred years. Next, our paper illustrates the potential of this methodology by tackling a specific question about the emotional states of policy-makers. We argue that politicians not only represent the preferences of their constituents over issues debated in parliament, they also react emotionally to national and world events in a manner that is predictable. In essence, politicians mirror the feelings and apprehensions of civilians in the face of adversity. We test this claim by tracking down the dynamics of politicians' emotional responses during economic hard times. Our empirical results contribute to mounting evidence accumulated in social sciences about the linkages between emotion and human behavior [1][2][3][4]. Our decision to focus on emotions in political discourse was propelled by the rise in importance of two influential streams of literature. The first one concerns the psychology of human behavior. Since at least the 1960s, developments in the field of behavioral economics have brought the psychological aspects of decision-making to the forefront. Seminal works on bounded rationality [5], prospect theory [6,7] and regret theory [8] have all attempted to a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 tackle the observed discrepancies between theories grounded in rationality and actual human behavior. Lending additional credence to this field of research, a recent body of work stressed the conclusion that the emotionality of human decision-making has an intrinsic, neural basis [9][10][11][12]. The progress of neuroscience has even led some scholars to coin the term "neuroeconomics" to speak of its applications in the discipline of economics [13]. Implications for the study of political behavior have been surveyed in [3]. A primary concern behind this study is that making sense of the decisions made by elected politicians entails being able to tap into, in one way or another, their emotions. Until now, addressing this question has been hindered by the apparent difficulty of monitoring politicians from afar, let alone measuring their states of mind. The theory and methods that we introduce in this paper are an attempt to fill that gap, by focusing on large amounts of officially recorded political writings and by employing the practices developed in natural language processing. The second stream of literature stems from computer science and concerns the detection of sentiment and emotion in textual data. More specifically, affective computing refers to a wide range of computational tools for the measurement of emotional states and affective responses communicated by humans either through facial and corporal expressions, oral or written speech [14][15][16][17][18]. These methods are sometimes referred to as emotional prosody or textual affect sensing. Recent work in this field has examined a variety of topics, from the use of emotions in musical lyrics [19] to the spread of happiness in social media [20,21] and the study of medical conditions such as depression in online communities [22][23][24]. The digitization of book collections has also opened the door to linguistic research on a massive scale [25], which includes studies examining the expression of emotions in English literature during the past centuries [26][27][28]. To our knowledge, however, such methodologies had not been implemented on historical corpora of parliamentary debates until now. Of particular relevance for this study are the prior attempts to apply affective computing methods for the study of political and socio-economic topics. Bollen et al., for instance, were able to measure meaningful emotional reactions among Twitter users in the face of real-world events such as the 2008 U.S. presidential campaign, shifts in market indicators of the recession and changes in oil prices [29]. Variations in the public mood as a response to economic and political events have also been studied in a growing number of papers (see e.g. [30,31]). Another recent study has looked at the association between economic indicators and the polarity of words expressed in books during the last century [32]. The researchers found a significant relation between the misery index and the emotional tone of books published in the decade that followed: changes in economic conditions appear to be reflected in the authors' use of language. This body of work provides evidence that individuals react emotionally to macroeconomic conditions, justifying further our interest in finding out whether politicians exhibit similar attitudes inside parliamentary institutions. Materials and Methods Our corpus consists of all available volumes of the British House of Commons' Hansard between 1909 and 2013 inclusive. It contains all the debates, oral questions and oral answers to written questions. The format of the Hansard-the official text archives of debates and speeches-was modified in 1909, when new standards were implemented for the verbatim record of the debates [33]. This is why our corpus begins at that date. Those text documents have been stored using a markup language following linked open data standards as part of the international project Digging into Linked Parliamentary Data (Dilipad). The corpus comprises a total of approximately 956.8 million tokens (i.e. words in a broader sense, including digits and other types of strings), with an average of 9.1 million tokens each year and a standard deviation of 2.2 million tokens. Considering the lemmatized version of the corpus-that is, the roots of words, which avoids duplicate counting of plural and conjugated verb forms-and considering only tokens appearing 10 times or more, UK parliamentarians have used a total vocabulary of 108,506 tokens. The length of parliamentary sessions has increased over time, and so has the corpus size per year: the decade 1910-1919 had an average annual size of 6.9 million tokens, compared to 9.8 million during the decade 2000-2009. Many approaches have been developed for capturing emotions in textual data, using either machine learning classifiers trained on human-annotated corpora or lexicons (dictionaries), that is, lists of words associated with emotions. One of the main challenges with those approaches is that they can rarely be used across domains. Sentiment analysis classifiers trained with corpora from a specific domain were shown to have limited exportability to texts using a different register or genre [34][35][36]. A model trained on social media blurbs written in casual English would fare poorly if applied to parliamentary speeches, more sophisticated and restricted by the general decorum of political institutions. On the other hand, polarity lexicons are often more general, but they still have limitations when applied to corpora across domains. Popular examples of such lexicons include the NRC Word-Emotion Association Lexicon [37], in which words were annotated for eight specific emotions and positive/negative polarity using crowdsourcing, SentiWordNet [38], created using recursive algorithms based on the WordNet database, and the polarity lexicon of OpinionFinder [39]. (Other general-purpose lexicons for affective computing include the General Inquirer [40], the Linguistic Inquiry and Word Count (LIWC) dictionaries [41], Hu and Liu's Opinion Lexicon [42], and WordNet extensions such as WordNet-Affect [43] and Q-WordNet [44].) Like classifiers trained with non-political corpora, these lexicons are not tailored to the analysis of political speeches. Parliaments are associated with expressions that convey specific meanings and interpretations that we need to take into account. For instance, the first three lexicons mentioned above give a negative score to a word such as war, and positive ones to education and health. Yet, a word like war will inevitably be used more frequently in times of war, since the topic needs to be discussed in parliament. Assuming that debates become more negative simply because of the increased presence of this word would be misleading. Similarly, nouns like education and health have different meanings in politics as they relate to policy domains. They also identify specific departments and ministerial functions. An increased usage of the word health would provide little information about the tone of the debates taking place in the House of Commons, as this could merely reflect the presence of a bill about that specific issue on the agenda. In short, we would like to avoid attributing an emotional value to words without considering the fact that they may have a descriptive, domain-specific usage. Creating Domain-Specific Lexicons for Affective Computing To overcome these problems, we rely upon a methodology for textual affect sensing that is adaptive to the domain under study. The general approach that we follow here has been introduced in [45] and a related methodology is discussed in [46]. We start by creating the vector space model of our corpus using the GloVe algorithm [47]. This model converts the vocabulary of our corpus into numerical vectors based on the matrix of word-word co-occurrences. We compute word vectors of 300 dimensions for each combination of lemma and part of speech (e.g. nouns, adjectives, and so forth), using a symmetric context window of 15 tokens. For simplicity, we use the expression "lemma" in what follows to speak of a lemma/part-of-speech pair. The second step of our methodology consists of creating an initial list of 200 seed lemmas capturing positive and negative emotions in the English language (100 lemmas for each pole). We selected these lemmas individually to ensure that they do not have multiple, opposite meanings when used as a specific part of speech, and to exclude terms with domain-specific meanings. Our objective is to use this list of seed lemmas as a starting point for the creation of emotion lexicons adaptive to any domain, given the availability of a large corpus. We provide additional details about these steps of our methodology in S1 Appendix. Specifically, using vector distances to detect word similarities, we attribute to all other lemmas in the vocabulary a score indicating how close they are to each of the two groups of seeds. The formula corresponds to: where kv i k is the norm of vector v i associated with lemma i, and where the seed lemmas for positive and negative emotions are indexed by p = {1, . . ., P} and q = {1, . . ., Q}, respectively. The scores s i are scaled into a [−1, 1] interval reflecting their emotional polarity. We retain the 2000 lemmas with the highest and lowest scores, expanding our lexicon to 4200 words. To illustrate the output of this method, we report the first 20 lemmas with the highest and lowest scores in Table 1. By redistributing those scores to the lemmas across the original corpus, we are able to quantify the mood of parliamentary debates over time, which can be aggregated by session, month, quarter, or year. For the purpose of this study, we create a measure of emotional polarity as follows. (In the remainder of this text, we use the expression "emotional polarity" or "polarity" to speak of the aggregate indicator of emotional words in the British parliament. Concept usage varies across disciplines, and the term "sentiment" is also common to speak of positive and negative emotions in computer science.) Let w it denote lemmas occurring in the text during period t. Next, let 1{w i 2 L} be an indicator function equaling 1 if the lemma w i is present in our lexicon, denoted by L, and 0 otherwise. Multiplying this function with the polarity score s i associated with each lemma i, and summing over all lemmas occurring in a time period t, gives an absolute indicator of the emotional polarity of language during that period. We divide by the unweighted count of lemmas during each period, so that accurate comparisons over time are possible. This means computing where n t is the total number of lemmas in the Hansard during period t. We improve this indicator further by accounting for the location of each lemma within negative sentences. We constructed a parameter, denoted θ it , measuring the valence of lemma i in the speeches of period t. This parameter is set to 1 unless w it is located between a word indicative of a negative clause and a punctuation mark, in which case it equals 0 (words indicating negative clauses include not, no, never, neither and nor). Thereby, we avoid attributing positive scores to an expression such as "not satisfied". Letting y t denote the emotional polarity in the House of Commons at time t, our measure amounts to Higher values of y t indicate more positive debates. The measure of polarity, just like the score variable s i , can be negative or positive. However, positive words are used more frequently in the English language and as a result, aggregate measures will tend to remain in the positive range as the corpus length increases. Our interest lies in the temporal change in y t ; thus, the scaling of that measure is irrelevant. Importantly, by using the count of lemmas as the denominator in Eq (3), we account for the fact that parliamentary sessions may differ in length from one period to the next, and avoid biases caused by the overall frequency of positive words. We assessed the substantive validity of our measures and included more detailed information on this stage in S1 Appendix. In particular, we evaluated our approach against a corpus of film reviews commonly used to assess polarity classifiers, in which users' self-reported scores can be considered a basis for ground validity. We found that our measure of emotional polarity alone can predict up to 80% of film ratings accurately, and classifiers incorporating our indicator as a feature compare favorably to other models for sentiment analysis in the literature. We should point out, however, that our approach does not account for the full structure of sentences. Utterances that our method may not capture easily are complex uses of language such as sarcasm, irony and hyperbole. On the other hand, notice that we rely on word vectors designed to capture the underlying meaning of terms in the Hansard vocabulary. Thus, a positive word commonly used to express the opposite of a speaker's true feelings is likely to be found in a negative context across the corpus, and the resulting scores would capture that semantic ambiguity. A rather straightforward extension to the methodology presented here could also model vectors for entire phrases or sentences, instead of lemmas, using tools for latent semantic analysis such as singular value decomposition [48] or document vectors [49]. For the purpose of this study, we believe that the benefits and simplicity of our approach outweigh its limitations. that politics has become more negative in recent years, although existing research on the tone of political discourse has focused mostly on peripheral evidence. For instance, some have documented a negative tone in the news coverage of politics [50] or a decline in public support for governments [51]. Our results show that the British parliament, on the contrary, has become more positive than it used to be. In fact, the trends appear consistent with the major turns of events of the past century. The first two decades of our sample mark an era of negative polarity in the British House of Commons, in line with the social divide of those turbulent years. The 1910s and 1920s were punctuated by major labor disputes, including a nation-wide strike in 1926, the rise of the workers' movement and the perceived threat of socialism with the arrival into power of a new Labour Party in 1924. Adding to that were the First World War and the Irish War of Independence of 1919-1921. The current tone of parliamentary proceedings seems to be far more optimistic compared to the early 20th Century, in line with the relative peace and stability characterizing the more recent decades (see [52]). Emotional Polarity Trends in the British Parliament We examined this trend in a number of ways to assess its robustness. We first compared the time-series obtained using a number of alternative polarity lexicons. The finding of debates becoming more positive over time resurfaces even when considering indicators based on the three general lexicons mentioned above (see S1 Fig). We also investigated changes in usage for the 200 initial seed lemmas considered to construct our domain-specific lexicon. We found that the relative usage of our negative seed words has actually increased slightly over time, based on data from the Google Books Ngram Viewer. The average relative frequency of our 100 positive seeds has decreased by 9 percent from the decade 1909-1918 to the decade 1999-2008, whereas the average for our 100 negative seeds has increased by 2 percent over the same period. This suggests that the pattern we observed in Fig 1 is not simply derived from the choice of specific seeds to create our lexicon. To investigate further, we tested whether the trend in emotional polarity is explained by an increasing gap between the tone of speeches of members from the party forming the government on one hand, and members of the opposition on the other hand (that is, members of the parliament (MPs) not affiliated with the party in power). We naturally expect the government to be more positive than the opposition, whose members have the duty to question the party in power and challenge its decisions [53]. Notice that we call those groups of MPs "government" and "opposition" for short, even though the government-the executive branch-is the Cabinet. Since Westminster countries are characterized by strong party discipline, it is unlikely that MPs from the party in power would criticize the government, which is why we group them together. When making the distinction between these parliamentary functions, we still find increasing trends in our polarity indicator (Fig 2). Although the party in power remained more positive than the opposition-which is unsurprising-the upward trend of the last few decades appears to generalize to the House of Commons as a whole. Our yearly and quarterly indicators also exhibit the characteristics of long-memory processes, as suggested by the slow decay of their autocorrelation functions (ACF) (S2 Fig). This long-memory feature means that exogenous shocks affecting the mood of politicians can have a lasting effect on the nature of debates over time. The power spectra of the series are reported in S3 Fig. Using linear regressions on a log scale, we estimated parameters from power spectral densities of the form 1/f α , where f is the frequency and α a parameter which should approximate 1 in the presence of pink noise. The fitted value of α is close to 1 when using the quarterly series (1.03, with a 95% confidence interval of [0.86, 1.20]). This parallels a number of other societal processes that have previously been found to be heavy-tailed [54]. Since our measure is the result of a large number of micro-interactions between members of parliament with shared histories, the finding that mood spreads over time following a 1/f process is not surprising. We note, however, that the parameter α in the yearly series is closer to 2 (1.71, with a 95% confidence interval of [1.38, 2.04]), a characteristic of Brownian noise. This has implications for the choice of estimators in the rest of our analysis, as Brownian noise is associated with unit root processes [55]. Accordingly, we tested each time-series for stationarity. Tables A-D in S1 Appendix report the full results. The main measures of emotional polarity used in our empirical analysis below appear to follow a unit root process, based on several specifications of both the Dickey-Fuller and KPSS tests. Empirical Results As explained at the outset, we are interested in explaining changes in emotional polarity over time. In other words, we are considering an equation of motion in discrete time of the type where Δy t is the first-difference operator of our polarity measure, and x t is a measure of national events affecting the mood of politicians. Specifically, we expect that the mood in parliament will respond to economic business cycles, that is, the core periodic transitions between economic recessions and expansions. Our theoretical argument is that those business cycles are a fundamental force affecting most spheres of activity in a polity; hence, recessions are likely to trigger a large amount of stress on workers and businesses that should realistically have repercussions in the House of Commons. The mandate of elected politicians is to represent their constituents. We usually think of this representation in terms of positions to be expressed on issues. In our view however, the emotions of civilians are just as likely to be echoed inside political institutions. This can take the form of harsher questions directed at the government in power and heightened discord during debates over bills and motions. In short, we expect the mood to become negative during recessions, as opposed to periods of economic expansion. Below, we test this expectation using both a binary indicator of recessions and continuous indicators of the national economy. We begin with the computation of autoregressive models including two exogenous components. The first is a simple binary indicator r t equaling 1 for years encompassing a recession, where a recession is defined as two or more consecutive quarters of negative growth in the gross domestic product (GDP), and 0 otherwise. For the time-period where quarterly data are missing, we coded a year as 1 if GDP growth was negative during that year. The full list of recessions appears in S1 Appendix, Table E. In addition to economic cycles, we control for the occurrence of political cycles caused by the periodicity of democratic institutions. We include a second binary indicator e t equaling 1 if a general election has been held during a given year, and 0 otherwise. We considered various other empirical indicators as control variables, in particular binary measures accounting for wars and the party in power, but those turn out to be poor predictors of emotional polarity and have little impact on our results (see S1 Appendix). For the sake of parsimony, we focus on the simpler models here. Our empirical specification can be expressed in terms of the autoregressive-moving average framework including explanatory factors treated as exogenous (ARMAX). The model corresponds to: where l is the lag length and ε t an error term. The first-difference transformation produces a stationary y t series, and our estimators satisfy the usual stability conditions. We report maximum likelihood estimates computed with one and two lags in Table 2. The estimated autoregressive parameters are negative, indicating that random shocks to the rate of change in emotional polarity (y t ) eventually vanish following oscillatory decays. The estimated coefficient for the Recession variable is negative, which is consistent with our expectation that the mood in parliament responds negatively to economic downturns. The value of the coefficient -0.198 in the first model indicates that a recession is associated with a 0.2 point decrease in the annual change in polarity. The estimate is statistically significant at the 95% confidence level in specifications with 1 and 2 autoregressive lags. Conversely, election years appear to increase the positivity of the mood by a similar order of magnitude (0.19). We also computed mean difference tests (t-tests) by considering bivariate relationships one at a time. Table 3 reports the average differences in polarity for recession years, election years, and years at war. Once again, the difference associated with recessions is negative and statistically significant. Continuous Indicators of Economic Conditions For the next step, we investigate the robustness of the relationship between national economy and political mood by considering continuous indicators. We selected four pertinent annual time series available for the entire time period: a measure of labor disputes (the natural logarithm of the number of days lost due to strikes per year), the rate of unemployment, the misery index (the sum of unemployment and inflation), and the rate of growth of the gross domestic product (GDP). The latter may not represent an ideal measure from a scientific standpoint since it merely reflects variations in the size of an economy, rather than its wealth-measures of economic growth in the field of economics are usually per capita indicators, to account for changes in the size of the population. In practice, however, policy-makers routinely rely upon the growth in the GDP to assess whether a country faces a recession or to establish targets. As a result, we expect this particular indicator to be relevant for the British House of Commons. Obviously, those indicators are strongly related to each other, but they capture different dimensions of economic cycles. Unemployment and the misery index are, arguably, more likely to affect constituents in their daily life and as a result, to trigger a response from elected officials. Labor disputes can be viewed as an even more concrete measure of public discontent regarding economic conditions. The idea that labor disputes can be a relevant indicator to model the mood of politicians should make sense for most observers of politics. Strikes and related labor conflicts are disruptive social activities that are fundamentally political, as well as emotionally laden for the actors involved. Although we found signs of a statistical relation between each of these four measures and our indicator of emotional polarity, the labor disputes and misery index series exhibit the clearest association. Both series are superimposed on the emotional polarity indicator in Fig 3. As can be seen, the intensity of labor disputes and the misery index appear counter-cyclical to the polarity of debates in the parliament. In particular, the major recession of 1973-1975 matches a peak in the negativity of debates, a surge in labor conflicts and an unprecedented spike in the level of the misery index caused by soaring inflation. The five key indicators mentioned so far are also plotted using a heat map in S4 big picture of the last century in Britain. The early 20th Century was characterized by intense labor disputes and skyrocketing rates of unemployment, culminating with the Great Depression of the 1930s. Things changed drastically after the second World War, when most of these measures stabilized. Meanwhile, the mood of parliamentary debates became more positive. A second period of turbulence arose in the 1970s and 1980s, decades encompassing two important recessions and new episodes of intense work conflicts. Accordingly, the mood became increasingly negative, before experiencing an upward trend in the recent years. On the other hand, the period encompassing the Second World War seems to shift away from that pattern. The end of the war coincides with relatively positive speeches in parliament, while the country was in fact struggling economically after the prolonged war effort. We explain this discrepancy, however, by the political significance of the victory for Britain. The war period also witnessed a series of coalition governments in the British parliament, which limited opportunities for partisan clashes on the usual issues of contention. Comparing figures visually does not give a definitive idea of the statistical association between these variables. For this reason we perform pairwise Granger causality tests between emotional polarity and these four historical measures. As can be seen in Table 4, one relationship in particular is revealed to be Granger-causal: the one going from labor disputes to emotional polarity. Since both variables are unit root processes, we tested for the existence of a cointegrating vector using the Johansen rank test. We find evidence of cointegration (see Table F in S1 Appendix), which suggests that the temporal trajectories followed by the mood of parliamentarians and the intensity of labor disputes are tied together. We estimated this relation using a vector error correction model (VECM). The fitted cointegrated vector corresponds to where x t is the indicator of labor disputes. Eq (6) suggests that, in the long run, both series tend to stay negatively related, more than a standard deviation apart. Since the variables are moving as a system, this should not be interpreted as a causal effect. One way to assess the dynamics between the number of strikes and the polarity of speeches in the House of Commons is using an orthogonalized impulse response function computed from the VECM estimates (Fig 4). The emotional response to a standard deviation shock in the labor disputes series is negative, converging to a value of approximately -0.18. The series being non-stationary, any random shock is persistent, and so is the estimated effect. Or, put another way, the estimated effect of a labor conflict persists until the initial shock is offset by another shock in the opposite direction. We examined the robustness of this finding using alternative estimators such as a dynamic ordinary least squares (DOLS). Table G in S1 Appendix reports the full set of estimates. This estimator is also suitable for modeling cointegrated relations by assuming the direction of causality, and may be easier to interpret. Using emotional polarity as the endogenous variable, the estimated long-run effect of labor disputes is approximately -0.9, and statistically significant at the 99.9% confidence level using heteroskedasticity and autocorrelation consistent standard errors. In other words, a one standard deviation increase in the levels of labor disputes is associated with a nearly proportional decrease, around 0.9 standard deviation, in the normalized indicator of the polarity of speeches in the British parliament. Together, these results suggest a strong statistical relationship between the two series. Finally, making the distinction between party in power and opposition helps to identify the causal mechanism at work. The bottom part of Table 4 tests the causality of labor disputes for each group of parliamentarians. As can be seen, occurrences of labor conflicts Granger-cause the emotional polarity of opposition parties in the House of Commons, but not that of the party in power. To assess this result, we estimated the emotional response of each group of MPs to a shock in labor disputes, again with VECMs. Using either the yearly or quarterly dataset, the size of the effect of labor disputes appears greater when considering opposition parties (see S5 Fig), which suggests that the emotional response of politicians transits via the opposition first and foremost. This finding lends additional support to our claim that political discourse reflects the mood of the electorate in meaningful ways. The opposition in British democratic institutions was designed to keep the executive accountable by making inquiries in the House of Commons [56]. Economic downturns or labor conflicts are by nature undesirable situations, and it makes sense to think that opposition parties will be the first ones to decry them. Parties in power face different incentives. Being ultimately accountable for the state of affairs in the country, government MPs are better off avoiding outright recognition of social ills whenever they can, which could otherwise undermine their popularity. Our findings help to interpret the inverse relationship observed in Fig 3: labor conflicts make the discourse of opposition parties more negative, which in turns affects the tone of speeches in the parliament as a whole. Conclusion The method discussed in this paper to measure emotion in political discourse has several benefits. It is relatively simple to use, it can be applied to different domains as long as a sufficiently large corpus exists, and it allows scoring the emotional valence of lemmas on a continuous scale. In fact, it is possible to use this approach to generate lexicons measuring different features of language, not just emotional polarity, as long as researchers can provide seed words to serve as a basis for computing word similarities from the vector space models. Applying this method to the corpus of the British Hansard from 1909 to 2013 leads to insightful findings. In particular, we found that politicians' usage of emotional words is significantly related to economic cycles, a result supported by models using both a binary indicator of recession and a more politicized indicator of labor conflicts. The observation that debates have become more positive in tone during the last decades raises an interesting question for social science scholars. Does this trend reflect a more general transformation of attitudes in society, or is it limited to politics? Although expanding the analysis to multiple corpora is a task that fell beyond the scope of the present research, existing work can provide some insight. For instance, one recent article analyzed long-term trends in the Google Books database using affective computing methods, in particular with an indicator based on the difference between the frequency of words associated with joy and sadness [28]. The findings suggest that books published in the late 20th Century had a more positive tone than those released in the 1970s and the 1980s, which parallels the trajectory depicted in Fig 1. Another study mentioned in our introduction found an association between economic misery and the expression of negative emotions in books [32]. The authors argued that economic considerations have become part of the shared experiences that influenced the general culture during the past century. Connecting our findings with these other studies suggests that improvements in living conditions may cause positive mood swings that transcend the political realm. On the other hand, our findings somewhat diverge from a body of literature interested in the economics of happiness. Whether higher incomes make individuals happier is a question that attracted intense scrutiny in the social sciences [57]. To be sure, there is empirical support to the conclusion that better living conditions are associated with higher self-reported levels of happiness in cross-sectional surveys [58][59][60][61]. However, scholars have often failed to observe rising aggregate levels of happiness in the long run, even as countries get richer [57,62]. This paradox contrasts with our findings, which exhibit long-range dependence. The divergence may be related to measurement. Answers to close-ended survey questions at fixed points in time give only a limited range of options to measure the dynamics of emotional states, as opposed to a full text corpus recording the language of politicians on an almost daily basis. Alternatively, there may be meaningful differences in the way various publics react to changes in economic conditions, differences that we do not fully understand yet. Future studies could take advantage of the type of methods presented here as an alternative way to tap into people's emotions, and advance our knowledge on that fundamental question. The findings highlighted in this paper also point to other implications for future research. For instance, if politicians react emotionally to economic downturns, it matters to reassess whether these emotions have in turn an impact on crucial decisions made during those periods. The indicators that we proposed in this paper could be used to pursue fine-grained analyses of this type. Moreover, we found evidence that emotional polarity follows a long-memory process, which is consistent with earlier findings about many social phenomena. To examine the persistence of moods, the Hansard corpus could be used to test for the presence of emotional contagion in the parliamentary network, as was done previously using social media data [63]. We coped with this property of the indicators by considering empirical methods that can accommodate integrated processes, but additional research could provide more insights on this particular question. Overall, given the importance of legislation and the ripple of impacts it begets on societies, we believe that improving our comprehension of the factors that alter the mood of policy-makers is an important research objective for the social sciences.	2016-12-30T08:36:52.502Z	2016-12-22T00:00:00.000	{ "year": 2016, "sha1": "64a739bfab7801e24239d5fab70b5bf05f240f87", "oa_license": "CCBY", "oa_url": "https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0168843&type=printable", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "bbc43a6288d1e148b2cd5821aeb1eac77f2ae41c", "s2fieldsofstudy": [ "Sociology" ], "extfieldsofstudy": [ "Medicine", "Political Science" ] }
268398664	pes2o/s2orc	v3-fos-license	Interstitial lung disease with prolonged fever that occurred during long-term administration of olaparib in a 74-year-old ovarian cancer patient: Radiological features and considerations for preventing delayed diagnosis A 74-year-old woman, who had been receiving olaparib for the treatment of ovarian cancer for more than a year, visited the emergency department complaining of a fever that had lasted for 1 month. She had been taking antipyretics and antibiotics for her fever, but without any effect. Although she had no symptoms other than fever, she had stopped taking olaparib for 1 week before her visit because she had developed anemia caused by myelosuppression from olaparib. After discontinuing olaparib, her maximum body temperature decreased. On admission, chest X-ray revealed no abnormalities, but chest CT showed diffuse ground-glass opacities. Chest CT taken 5 days later showed partial improvement; therefore, we diagnosed her with interstitial lung disease (ILD) associated with olaparib. After short-term steroid treatment, the ground-glass opacities disappeared, and the patient became afebrile. The CT scan taken for tumor evaluation 2 days before the onset of fever showed a few centrilobular nodular opacities and small patchy ground-glass opacities. These findings could indicate early lesions of ILD, but they seemed inconspicuous and nonspecific, and it might have been difficult to diagnose ILD then. To date, few cases of ILD associated with olaparib have been reported. However, based on previous reports, fever is often seen, and CT findings mainly comprise diffuse ground-glass opacities, and in some cases, centrilobular nodular shadows. Thus, in conjunction with the findings of the present case, these characteristics may be representative of olaparib-induced ILD. Introduction Anticancer agents often cause drug-induced lung injuries, the most common of which is interstitial lung disease (ILD) [1] .There are no specific imaging findings for drug-induced ILD, and it resembles other ILDs, such as idiopathic interstitial pneumonia.Therefore, computed tomography (CT) findings of drug-induced ILD are classified as CT patterns of similar ILDs, for example, organizing pneumonia (OP) pattern, hypersensitivity pneumonia (HP) pattern, nonspecific interstitial pneumonia (NSIP) pattern, etc. [2] .Furthermore, the symptoms of drug-induced ILD are usually nonspecific, such as coughing, dyspnea, and fever, which may occur singly or in combination; these symptoms are also common in other lung diseases.Thus, if drug-induced ILD is not considered, a diagnosis may not be made or may be delayed, even if symptoms and abnormal findings are recorded through an interview and chest imaging is performed. We present a case of a woman with ovarian cancer who had been receiving the poly (adenosine diphosphate-ribose) polymerase (PARP) inhibitor olaparib, an oral molecular targeted agent, and experienced drug-induced ILD.This case was diagnosed 1 month after the onset of symptoms and is, therefore, a cautionary case in the diagnosis of drug-induced ILD.The reported incidence of ILD was approximately 1.5% in phase III clinical trials of olaparib maintenance therapy for ovarian cancer [ 3 ,4 ].However, few cases have been reported, and the characteristics of ILD are not well understood.Therefore, we investigated previous literature and discussed the features of olaparib-induced ILD. Case presentation A 74-year-old woman with recurrent ovarian cancer visited the emergency department of our hospital because of a prolonged fever.She had developed a fever of up to 38 • C, which lasted for a month.She had no symptoms other than fever.Initially, she was treated with antipyretics and monitored; however, her condition did not improve.Subsequently, it was thought to be an infection; therefore, levofloxacin was administered, but it had no effect.The patient had been receiving olaparib for over 13 months.Continuous use of olaparib caused myelosuppression, and the anemia progressed gradually.A week before her visit, her hemoglobin level dropped to 6.4 g/dL, for which, she received a blood transfusion and olaparib was discontinued.She continued to have a fever after discontinuing olaparib; however, the peak body temperature decreased. On physical examination, her general condition was good and chest auscultation revealed no abnormalities in the heart or respiratory sounds.Body temperature was 37.3 • C, and oxygen saturation was 95%.Blood examination showed a white blood count of 4130/ μL, lymphocyte count of 540/ μL, hemoglobin level of 8.0 g/dL, platelet count of 83000/ μL, and Creactive protein level of 6.13 mg/dL.Urine examination results were normal.The polymerase chain reaction test did not detect severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).There were no abnormal findings on the chest radiograph ( Fig. 1 ), and a systemic CT scan was performed to investigate the underlying cause of the fever.Chest CT revealed diffuse, faint ground-glass opacities (GGO) in both lung fields ( Fig. 2 ).Because of the myelosuppression, we considered the possibility of an opportunistic infection, such as cytomegalovirus (CMV) pneumonia or Pneumocystis pneumonia, but she tested negative for serum CMV antigen and βd glucan.The serum markers for ILD, including Krebs von den Lungen-6 (KL-6) and surfactant protein-D (SP-D), were elevated to 588 U/mL and 169.0 ng/mL, respectively (normal value: < 500 U/mL and < 110 ng/mL, respectively).Therefore, the cause of ILD was considered other than infectious disease.Based on the CT findings, HP was considered the most likely ILD.Summer-type HP, caused by the inhalation of Trichosporon asahii suspended in the air in houses, is the most common type of HP in Japan; however, she tested negative for the serum antibody for T. asahii .Finally, other etiologies of HP were considered, and olaparib was suspected to cause ILD.After admission, she had a low-grade fever; however, the maximum body temperature had decreased, and CT scans taken 5 days later showed that some of the shadows had disappeared ( Fig. 3 ).Thus, the patient was diagnosed with ILD caused by olaparib.ILD was treated with 30 mg/day of prednisolone, and followup CT showed that the GGO had completely disappeared 7 days later ( Fig. 4 ).Corticosteroid therapy was administered for a short period. The patient underwent periodic CT scans to evaluate the status of ovarian cancer.She underwent a CT scan 2 days before the onset of the fever, and we reviewed the CT scan again.Small patchy GGO and poorly defined nodules were detected in both lungs ( Fig. 5 ).These findings did not exist before, and we speculated that they depicted an early lesion of ILD caused by olaparib.The patient was discharged from the hospital and returned home, and the fever did not recur. Discussion In the present case, the cause of persistent fever could not be identified early.It is sometimes difficult to identify the ori- gin of fever when the patient presents with only fever and no other symptoms.Generally, drug-induced ILD is diagnosed comprehensively after differential diagnosis of similar diseases.Because this patient had a fever, it was important to differentiate the condition from an infectious disease.Respiratory infection was ruled out because the patient was unresponsive to antibiotics.Blood tests were negative for antigens derived from CMV and PCP, which caused diffuse ground-glass opacities, as in this case.HP caused by environmental factors, such as summer-type HP, was ruled out as there was no recurrence after discharge.After discontinuing only olaparib among the multiple oral medications, the fever tended to decline, and the CT findings also partially improved, making it possible to diagnose olaparib-induced ILD. There are several reasons for the delayed diagnosis of the present case.First, a long time (more than a year) had passed since the administration of the causative drug till the onset of symptoms.Antineoplastic drugs often cause drug-induced ILD, which usually occurs within a few days to 6 months of starting administration [5] .Therefore, if drug administration is continued safely for an extended period, the risk of developing drug-induced ILD may be underestimated.In a study on PARP inhibitor-related ILD using the FDA Adverse Events Reporting System database, 170 cases of ILD were collected, of which 152 were associated with olaparib [6] , with a median onset of ILD of 99 days (interquartile range: 77-241, range: 0-477).Although it is rare for patients to develop ILD even after taking olaparib for more than one year, this case shows that it is necessary to be aware of the development of ILD during treatment.Second, in this case, there were no symptoms other than fever.If the patient had presented respiratory symptoms, the time to diagnosis could have been reduced.However, we should be aware that fever may be the only symptom of ILD.Third, a few poorly defined small nodules and small patchy GGO that appeared to be an initial image of ILD on chest CT taken two days before the onset of fever were inconspicuous and appeared as nonspecific shadows that could not easily be thought of as ILD.If the incidental findings on chest CT at the onset of ILD are insignificant, as in this case, they may be overlooked, and it may be difficult to make an early diagnosis.Careful monitoring is required even if the incidental CT finding is a slightly nonspecific shadow. Few cases of olaparib-induced ILD have been published, and detailed information, especially regarding the symptoms and CT findings, is not well understood.Table 1 summarizes the information based on previously published case reports Fig. 4 -(A, B) Follow-up chest CT after corticosteroid treatment for one week.The same region as Fig. 2A and B are presented.A complete disappearance of ground-glass opacities can be seen.[7][8][9][10][11] and the present case.Fever was observed in eight of nine cases and is considered a common symptom of olaparibinduced ILD.Regarding chest CT findings, diffuse GGO was the most common shadow.However, notably, 4 of the 9 cases had centrilobular nodular shadows.When the CT findings of these Fig. 5 -(A, B) Chest CT taken two days before the onset of fever.A few small patchy ground-glass opacities (arrowheads) and faint, poorly defined nodular shadows (arrows) are seen. cases are classified into patterns, the most common radiologic pattern is the HP pattern defined as "small, poorly defined centrilobular nodules with or without widespread areas of ground-glass opacity [2] ."Fever and HP pattern on chest CT may be characteristic of olaparib-induced ILD.Fur- Conclusion When patients present with fever while receiving olaparib, differential diagnoses should include drug-induced ILD and infectious diseases.Drug-induced ILD cannot be ruled out even if the patient presents only with a fever.Olaparib-induced ILD mainly manifests as diffuse GGO or HP patterns and is sometimes accompanied by centrilobular nodules.Faint GGO cannot be clearly recognized on chest radiographs; therefore, chest CT is useful for diagnosis.Early recognition and diagnosis of drug-induced ILD improves patient outcomes, and thus, chest CT should be performed immediately when druginduced ILD is suspected. Patient consent Informed consent for the publication of this report was obtained from the patient. Fig. 1 - Fig. 1 -Chest X-ray at the diagnosis of ILD.It is difficult to determine the presence of ground-glass opacities. Fig. 3 - Fig. 3 -Comparison between chest computed tomography at the presentation to the emergency department (A, C) and 5 days later (12 days after discontinuing olaparib) (B, D).The patchy ground-glass opacities (dotted circles) and poorly defined nodular opacities (arrows) disappeared.Poorly defined nodular opacities represent small, rounded pulmonary opacities with ill-defined margins. Table 1 -Clinical and radiological presentation of previously reported cases of olaparib-induced ILD. case accumulation and research are required to clarify the clinical and radiological features of olaparib-induced ILD.	2024-03-15T16:27:28.935Z	2024-03-08T00:00:00.000	{ "year": 2024, "sha1": "35c51c83f05c203243e3c0b601e79aad0ce78e82", "oa_license": "CCBYNCND", "oa_url": "https://doi.org/10.1016/j.radcr.2024.02.064", "oa_status": "HYBRID", "pdf_src": "PubMedCentral", "pdf_hash": "0fba41e8eb98a09b66ed9277b2c2e8498f2701a1", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [] }
43937463	pes2o/s2orc	v3-fos-license	Green Nanotechnology from Plant Extracts : Synthesis and Characterization of Gold Nanoparticles The advantage of using plants in nanoparticles synthesis is that they are easily available, safe to handle and possess a broad variability of metabolites such as antioxidants, nucleotides and vitamins. The aim of this study was to investigate the effects of Green and Zimbro tea and also Green coconut water as a reducing and stabilizer agent in gold nanoparticle synthesis. The gold nanoparticles were characterized by UV-Vis absorption spectroscopy, X-ray diffraction (XRD), Dynamic light scattering (DLS) and Transmission electron microscopy (TEM) analysis. Their physical stability was determined using a UV-Vis spectrophotometer over several days during storage at room temperature. We observed that green chemical process to obtain gold nanoparticles did not require any external chemicals reagent for stabilization of nanoparticulate. Absorption measurements indicated that the plasmon resonance wavelength appears around 530 nm. X-ray diffractograms of gold nanoparticles evidenced the presence of Au-rich (fcc) phases. TEM analysis showed a homogeneous dispersion of nanoparticles and some agglomerates. Differences in size and shape of the nanoparticles were observed. Zeta potential of AuNPs synthetized in presence of Green tea was −33 mV indicating stability of the synthesized nanoparticles. Introduction Nanotechnology is one of the fastest growing areas of Science and technology.The synthesis of metal nanoparticles is an active area of research in the field of nanotechnology with an exponential progress in biomedical applications including imaging, diagnostics, drug delivery and therapeutics using metal nanoparticles [1] [2]. Several chemical and physical methods have been used for synthesis of nanoparticles.However, a number of methods present disadvantages including use of toxic solvents, high-energy consumption, hazardous products etc.Therefore, there is an essential need to develop environment friendly methods for synthesis of metal nanoparticles.The development of eco-friendly technologies in material synthesis is of considerable importance to expand their biological applications.Nowadays, varieties of green nanoparticles with well-defined chemical composition, size, and morphology have been synthesized by different methods and their applications in many innovative technological areas have been explored [3]- [6]. Due to their unique optical, electrical and catalytic properties, gold nanoparticles (AuNPs) have been used in many areas of life, mainly in medicine, electronics and technologies of manufacturing modern materials [4]- [10].AuNPs have been highly used as agents in biomedical detection and contrasting image due to its properties that match: biocompatibility, bioconjugation and optical properties.Their small size, high surface area and stability at high temperatures make them perfect tool in medical diagnostics, photodynamic therapy as well as in the active transport of drugs, especially for cancer treatment [11], minimizing the damage cell during optical staining or biomarker tagging even as during the delivery of biomolecules/drugs to the target cell/tissue/organ [12].Gold nanoparticles have been employed in radiotherapy, in two modes: to increase local dose deposition in tissue during radiotherapy or as a local emitter of gamma and beta rays.The radioactive properties of gold include: 198 Au (β max = 0.96 MeV; t 1/2 = 2.7 days) and 199 Au (β max = 0.46 MeV; t 1/2 = 3.14 days), making it a strong candidate to therapeutic radio applications [13]- [15].Furthermore, both isotopes have gamma emission that can be used in pharmacokinetic and dosimetric studies. AuNPs are generally unstable because of their high surface and a suitable stabilizer should be added to prevent aggregation [11]. The renewable nature of plant extracts, eco-friendly aqueous medium and mild reaction conditions make the method advantageous over other hazardous methods.In the last years, different kind plants extract and their products have received attention due to its low cost, energy-efficient and nontoxic behavior in approach for synthesis of metal nanoparticles [4].Thus, the focus has turned toward Green nanotechnology.The green synthesis of metal nanoparticles should involve three main steps based on green chemistry perspectives, namely 1) the selection of a biocompatible and nontoxic solvent medium, 2) the selection of environmentally favorable reducing agents, and 3) the selection of nontoxic substances for stabilization of the nanoparticles.Many works have been done using plant extracts to promote the reduction of metal nanoparticles due to the presence of some phytochemicals.They have been extensively used in the treatment of various diseases including human cancer.That substances serve as a dual role: effective reducing agents to reduce gold and also as stabilizers to provide a coating on the gold nanoparticles in a single step [4] [12]- [18].The main phytochemicals responsible have been identified as terpenoids, flavones, ketones, aldehydes, amides and carboxylic acids through Fourier Transform Infra-Red (FT-IR) spectroscopy studies.The main water-soluble phytochemicals are flavones, organic acids and quinones, which are responsible for immediate reduction [4] [7]. The growing evidence towards the health benefits of Camellia sinensis (Green tea) has resulted in extensive studies to understanding the curing power of tea [11].A well-accepted scientific consensus emanating from several scientific investigations is that tea contains high levels of antioxidant polyphenols, including flavonoids, and catechins, all of which scavenge the dangerous free radicals in the body and thus, prevent the progress of various diseases [11] T-AuNPs have been found to be nontoxic as assessed through MTT assays.Natural chemicals, other than gold salts, were used in this truly biogenic, green nanotechnological process thus paving the way for excellent opportunities for their application in molecular imaging and therapy [11]. Nestor et al. employing Green tea extract as reducing and stabilizing agent for production of gold and silver nanostructures in aqueous solution at ambient conditions.Colloidal systems of silver and gold nanoparticles exhibit highly efficient single photon-induced luminescence.This optical response can be manipulated by changing concentrations of metal ions and the quantity of reducing agent, which plays a crucial role in formation, growth and luminescence response of these noble-metal nanostructures [24]. On the other hand, Juniperus communis (Zimbro tea) has been used in various diseases including antidiabetic effect, treatment of kidney infections, as diuretic, stomachic, treatment of rheumatism, cardiac and skin diseases.There are some studies about the constituents of the fruit of J. communis where polyphenols, polyphenol esters, and monoterpenes hydrocarbons [25] [26] were isolated. As an alternative, the coconut shell extract have been reported as antibacterial, and antiviral [27] and anti-inflammatory [28], among others properties.Moreover, a very mild and environment-friendly method for the synthesis of AuNPs from Green coconut shell (mesocarp) extract without any additional capping or stabilizing agents was too reported [29].Paul et al. studied the shell extract of green coconut (Cocos nucifera Linn) for the synthesis of gold nanoparticles at room temperature under very mild condition without any extra stabilizing or capping agents.The size of the synthesized gold nanoparticles could be controlled by varying the concentration of the shell extract [29].The coconut water (C.nucifera) as the reducing agent was employed for Elumai et al. on the synthesis of silver nanoparticles (Ag-NPs).The synthesized Ag-NPs were predominately polydispersed.Fourier Transform Infra-Red (FT-IR) spectroscopy analysis showed that the synthesized nanoparticles were capped with bimolecular compounds, which are responsible for the reduction of silver ions [30]. In our laboratory, there is a high interest in the study of gold nanoparticles synthesis for many application including cancer treating and/or cosmetology.In this context, the aim of this study was investigate the effect of different plants as reducing and stabilizer agent in gold nanoparticle formulation, namely: Green tea, Zimbro tea and Green coconut water.In the other words, we are interested in an environment-friendly method for the synthesis of AuNPs from plants extract without any additional reducing or stabilizing chemical agents.The stabilized colloidal AuNPs was characterized by UV-Vis absorption spectroscopy, energy dispersive X-ray spectroscopy (XRD), Dynamic light scattering (DLS), and Transmission electron microscopy (TEM) analysis.Their physical stability was determined using a UV-Vis spectrophotometer over several days during storage at room temperature. Solution Preparation All chemicals and plant extracts precursors used in the synthesis of gold nanoparticles (AuNPs) were purchased from suppliers: HAuCl 4 •3H 2 O (Fluka), Camellia sinensis, J. communis and Green coconut water were obtained from a local market.For Camellia (Green tea) extract, 1 g of leafs was placed in 50 mL of boiling water and magnetic stirrer for 30 minutes.For J. communis (Zimbro tea) extract, 1 g of fruit was macerated, placed in 50 mL of boiling water and magnetic stirrer for 60 minutes. To 10 mL vials was added 0.1 mL of 0.1 mol•L −1 HAuCl 4 solution (in DI water) 9.9 mL of Green tea, Zimbro tea solution or Green coconut water (in triplicate).The reaction was stirred continuously at 25˚C.The color of the mixture started in pale yellow and become purple-red after 15, 90 and 30 minutes for Green tea, Zimbro tea solution and Green coconut water respectively, indicating the formation of gold nanoparticles.The reaction mixture was stirred for an additional 20 minutes. The gold nanoparticles formed were characterized by UV-Vis absorption spectroscopy, Energy dispersive X-ray spectroscopy (XRD), Dynamic light scattering (DLS) and Transmission electron microscopy (TEM) analysis.Their physical stability was determined using a UV-Vis spectrophotometer over several days during storage at room temperature. UV-Vis Spectroscopy The solutions were characterized by absorption spectroscopy in the UV-Vis Spectrophotometer SpectraMax I3, Soft Max Pro ® 6.4 Microplate Analysis Software to confirm the peak surface plasmon resonance of the gold nanoparticles at λ = 535 nm, approximately.The spectra were recorded at first day and after 1, 2 weeks and finally after 1 month.All samples were diluted with water in each measure to guarantee that the maximum absorbance below 1.5. X-Ray Diffraction (XRD) Analysis X-ray diffraction (XRD) analyses were carried out with a Miniflex II model Rigaku diffractometer using a Cu Kα source (λ = 1.54056Ǻ).The diffractograms were recorded at 2θ in the range 20˚ -90˚ with step size of 0.05˚ and scan time of 2 s per step. Dynamic Light Scattering (DLS) Analysis Dynamic light scattering (DLS) was carried out using Zeta Plus-Zeta Potential Analyzer (Brookhaven Instruments Corporation, Holtsville, NY), which was equipped with a 677 nm laser and dynamic light-scattering (PCS) at 90˚ for particle sizing.The particle size (multimodal size distribution) was determined by measuring the angles at which an incident light beam is scattered as a function of Brownian motion of the colloidal gold particles.1 mL of each sample was filtered in CHROMAFIL ® Xtra PVDF-20/25, until measurement. Transmission Electron Microscopy (TEM) Analysis Transmission electron microscopy (TEM) was carried out using a JEOL JEM-2100 electron microscope, operated at 200 kV.The particle distribution histogram was determined by measuring the particles by Lince program. Results and Discussion UV-vis spectroscopy is one of the most important techniques to determine and evaluate the formation and stability of metal nanoparticles in aqueous solution.Gold nanoparticles produced by Green synthesis did not require any external chemicals agents for the reduction and stabilization of the nanoparticle.Phytochemicals substances presents in tea or coconut water are presumably responsible for the creation of coating on gold nanoparticles and thus, rendering the nanoparticles stable against agglomeration [4].Figure 1 shows the UV-Vis spectra of gold nanoparticle formation at constant concentration of HAuCl 4 (1 mmol•L −1 ) for Green, Zimbro tea and Green coconut water solutions.The absorbance was observed around λ max = 530 -540 nm for all solutions, indicating the formation of AuNPs [4] due to the excitation of the surface plasmon vibrations in the AuNPs, but is possible to find the plasmon band of AuNPs in ranges from 510 to 560 nm [31] [32].It is evident from Figure 1 that in Green and Zimbro tea AuNPs solution was not observed any displacement in λ max (540 or 530 nm respectively), meanwhile had a displacement in around 20 nm over time for Green coconut water AuNPs in λ max , suggesting that Green and Zimbro tea AuNPs were more stable than Green coconut water AuNPs.These results showed that Green coconut water is not efficient enough to stabilizing the gold nanoparticles for longs periods due to the AuNPs agglomeration were observed after 2 weeks.In the other samples, the agglomeration of AuNPs was not observed after almost 1 month. Particle size and size distribution are the most important characteristics of nanoparticle systems.They determine the in vivo distribution, biological fate, toxicity and the targeting ability of nanoparticle systems.In addition, they can also influence the drug loading, drug release and stability of nanoparticles.Many studies have demonstrated that nanoparticles of sub-micron size have a number of advantages over microparticles as a drug delivery system.Generally, nanoparticles have relatively higher intracellular uptake compared to microparticles and available to a wider range of biological targets due to their small size and relative mobility [34]. In Figure 3 the dynamic light scattering (DLS) histogram of multimodal size distribution and zeta sizer corresponding to the AuNPs in Green, Zimbro tea and Green coconut water solutions.The DLS method was employed to determine the size of AuNPs coated with all phytochemicals present on tea.Different particle size distribution curves were observed using three different green stabilizers.The Size distribution and zeta sizer are shown in the Table 1.Difference in size of AuNPs suggesting that AuNPs are coated with phytochemicals with low and high molecular weight [1]. The zeta potential of a nanoparticle is commonly used to characterize the surface charge property of nanoparticles.It reflects the electrical potential of particles and is influenced by the composition of the particle and the medium in which it is dispersed.Nanoparticles with a zeta potential above (±) 30 mV have been shown to be stable in suspension, as the surface charge prevents aggregation of the particles [1] [35].In our results only the zeta potential of AuNPs synthesized with green tea was above 30 mV (−33 mV) indicating the stability of the AuNPs. TEM micrographs of AuNPs, shows the shape, particle size and their histograms distribution (Figure 4).In Figure 4(a), small and larger nanoparticles were formed, showing agglomerates of hexagonal shape and particle size between 40 -70 nm.In Figure 4(b), the agglomeration of AuNPs was lower than the last one and the particle size was between 20 -30 nm showing circular and hexagonal shapes.Circular, triangular, hexagonal and rodshaped nanoparticles were observed in Figure 4(c) showing particle sizes were between 30 -70 nm.The results suggest that smaller particle size synthesize with Zimbro tea is due to it promote the presence of more nucleation site for AuCl 4 -complexation.It is also possible that Zimbro tea components have effectively protection for synthesized nanoparticles thus preventing their aggregation.Whereas for Green tea probably less number of nucleation sites would be present more reduction has taken place at nuclei that leading to formation of bigger particle.This observation is in agreement to Mata et al. who prepared gold nanoparticles using aqueous Plumeria alba flower extract (PAFE).The use of 1% and 5% concentrations of PAFE resulted in two different sizes of P. alba gold nanoparticles, PAGNPs1 and PAGNPs2.The average particle diameters of PAGNPs1 and PAG-NPs2 are found to be 36.05and 20.65 nm, respectively.A possible explanation for the formation of smaller particles in PAGNPs2 could be the presence of the increased number of nucleation sites for AuCl 4 -complexation with the increasing concentration of the PAFE.Whereas in PAGNPs1, less nucleation sites were present this may lead to more reduction at one nucleation and formation of bigger particles [5].Various nanoparticles shapes were too observed by Philip using the leaf extract of Hibiscus rosa sinensis.The size and shape of Au nanoparticles were modulated varying the ratio of metal salt and extract in the reaction medium [36].Paul et al. [29] found AuNPs of spherical, triangular, tetragonal, pentagonal and hexagonal shapes using shell coconut extract.The average size of the AuNPs were 20 -9.5 nm the extent to which increasing the coconut extract concentration.At higher concentration of the shell extract, the polyphenolic compounds, quinones and other chelating phytochemicals present in the shell extract can effectively stabilize the smaller sized AuNPs. The difference in size of particles observed by DLS when compared to TEM is the fact that the measured size also includes the bio-organic compounds enveloping the core of the AuNPs [11]. Gold nanoparticles generated through that process presents agglomerate suggesting that the combination of thearubigins, theaflavins, catechins and various phytochemicals present in green reducing and stabilizer agent studied in this paper serve as excellent stabilizers on nanoparticles and thus, optionally provide shielding from agglomeration [4]. Conclusion In this contribution, gold nanoparticles of different size and shape were synthesized using Green and Zimbro tea and Green coconut water as a reducing and stabilizer agent without adding of different physical and chemical steps.From the point of view of nanotechnology, this was a significant advancement to synthesize gold nanoparticles by economically procedure.The reported syntheses were not only simple and cost-effective but also capable of providing monodisperse, functional gold nanoparticles.It was possible because green reducing and stabilizer agents perform an effectively protection and prevent their aggregation.The AuNPs were confirmed by UV-vis due to the presence of a peak around 535 nm concerning the surface plasmon resonance characteristic of AuNPs.These nanoparticles were stable in water for at least one month, for Green and Zimbro tea, which can be attributed to surface binding of various phytochemicals substances present in tea.Overall, the results suggested that Green tea provided the formation of more stable nanoparticles.These results showed a wide range of particle sizes and shapes.Finally the synthesis of gold nanoparticles was carried out at room temperature, atmospheric pressure and were made in water (universal solvent), indicating a green process that presents a reliable and economic method. [19]-[23].The tea-generated gold nanoparticles (T-AuNPs), have demonstrated remarkable in vitro stability in various buffers including saline, histidine and cysteine solutions.Nune et al. studied T-AuNPs with several phytochemical coatings.The results have shown significant affinity toward prostate (PC-3) and breast (MCF-7) cancer cells.Moreover, results on the cellular internalization of T-AuNPs through endocytosis into the PC-3 and MCF-7 cells are presented.The generation of T-AuNPs follows all principles of green chemistry and Figure 1 . Figure 1.UV-Vis spectra of different gold nanoparticles solutions reduced and stabilized with (a) Green tea, (b) Zimbro tea and (c) Green coconut water at 25˚C. Figure 2 . Figure 2. X-ray diffractograms of Gold nanoparticles reduced and stabilized with Green tea. Figure 3 . Figure 3. Histogram of multimodal size distribution corresponding different gold nanoparticles solutions reduced and stabilized with (a) Green tea, (b) Zimbro tea and (c) Green coconut water. Figure 4 . Figure 4. TEM micrographs and particle size distribution histograms of AuNPs reduced and stabilized with (a) Green tea, (b) Zimbro tea and (c) Green coconut water. Table 1 . Size distribution and zeta potential obtained by DLS of different gold nanoparticles solutions reduced and stabilized with Green tea, Zimbro tea and Green coconut water.	2018-05-25T01:55:16.817Z	2016-07-26T00:00:00.000	{ "year": 2016, "sha1": "cc132a00cd44a6f2f19486fba2b4d15c0917a4a4", "oa_license": "CCBY", "oa_url": "http://www.scirp.org/journal/PaperDownload.aspx?paperID=69898", "oa_status": "GOLD", "pdf_src": "ScienceParseMerged", "pdf_hash": "cc132a00cd44a6f2f19486fba2b4d15c0917a4a4", "s2fieldsofstudy": [ "Materials Science", "Chemistry", "Environmental Science" ], "extfieldsofstudy": [ "Materials Science" ] }
15998438	pes2o/s2orc	v3-fos-license	Expression of transforming growth factor-beta-1 and p27Kip1 in pancreatic adenocarcinomas: relation with cell-cycle-associated proteins and clinicopathologic characteristics Background The purpose of our study was to investigate the immunohistochemical expression of TGF-β1 and p27 in pancreatic adenocarcinomas and to compare the findings with the clinicopathological features and survival. We also aimed to evaluate the expression of TGF-β1 and p27 in the context of other cell cycle and proliferation markers such as cyclin D1 and Ki-67. Methods We examined TGF-β1 and p27 expression immunohistochemically in 63 cases of invasive ductal adenocarcinoma of the pancreas. Standard streptavidin-biotin immunperoxidase method was used for immunostaining and the stained slides were examined microscopically using semiquantitative criteria. Results TGF-β1 stained the cytoplasms of the tumor cells in 43 cases [68.3%]. There was a statistically significant difference among TGF-β1 staining scores in terms of clinicopathologic factors such as blood vessel invasion, stage and distant metastasis [p < 0.05]. Of the 63 tumors evaluated 23 [36.5%] were positive for p27 within the nucleus. An inverse correlation was found between p27 immunoreactivity and grade [p < 0.05]. But no significant correlation was found between p27 and other parameters. Among the patients with survival data 27 patients had RO resections and these cases were considered in survival analysis. In the univariate analysis, neither TGF-β1 nor p27 expression was related with patient survival. Conclusion Our findings suggest that in pancreatic carcinoma, TGF-β1 expression is related to tumor growth and metastasis. But it is not associated with cell cycle proteins. p27 expression is reduced in pancreatic adenocarcinomas and decreased protein levels of p27 may play a role in the differentiation of pancreatic cancer. diagnosis. Even when the tumor is localized, the mean survival time after radical resection varies from 10 to 20 months [1]. The mechanisms of the aggressive growth and metastasis are not yet extensively understood. Several studies indicated that proliferative activity of tumor cells, as well as tumor angiogenesis, inactivation of tumor supressor genes, overexpression of growth factors may play role in pancreatic carcinogenesis and may help to predict patient outcome [2][3][4][5][6][7][8]. Recent studies denoted that alterations in growth factors and growth factor receptors seem to influence the biologic behaviour of pancreatic cancer cells [2]. Growth factors are involved in carcinogenesis, where they influence a variety of functions including cell proliferation, invasion, metastasis, angiogenesis, local immune system functions, and extracellular matrix formation [2]. Growth factors do not only stimulate cell proliferation, but they may also act as growth inhibitors, depending on the cell type and the stimulatory pathway that is involved. Transforming growth factor-β [TGF-β] is such an example, being a growth stimulator in fibroblastic cells with TGF-β receptors, but a negative regulator in epithelial cells. TGF-β belongs to a family of homologous polypeptides that includes three major isoforms [TGF-β1, TGF-β2, TGF-β3]. It has been reported that TGF-β influence different cell functions, including growth, proliferation and differentiation. It can influence cancer growth in various ways, such as by stimulating angiogenesis, suppressing cancerdirected immune functions, increasing the expression of adhesion molecules and extracellular matrix components [9]. Human pancreatic cancer cells may exhibit loss of responsiveness to TGF-β-mediated growth inhibition as a consequence of altered TGF-β expression as well as a result of postreceptor alterations [10]. It has also been demonstrated that TGF-β induced cell cycle arrest can be partially attributed to the regulatory effects of TGF-β on both the expression and activity of cyclin-dependent kinase inhibitors [CDKI] such as p21 and p27. The binding of these inhibitors to spesific cyclin-dependent kinase [CDK] complexes blocks their activity and causes cell cycle arrest [11,12]. Alterations in cell cycle regulatory mechanisms play an important role in the tumor development. Cell cycle progression is regulated by a series of cyclins, CDKs and CDKIs. p27, a member of the Cip/Kip family is a low molecular weight CDKI, which is able to arrest cell cycle progression by complexing CDKs and their activity [13]. Low p27 expression has been reported to be a poor prognostic factor in a variety of human cancers including prostate, lung, squamous cell carcinomas [13][14][15][16][17][18]. In this study, we investigated the immunohistochemical expressions of TGF-β1 and p27 in pancreatic adenocarcinomas and the results were correlated with the clinicopathologic characteristics of the cases and the patients' survival to find out if these factors could be used as an additional predictor of the disease extent and patient outcome. Additionally, we evaluated the expression of TGF-β1 and p27 in the context of other cell cycle and proliferation markers; cyclin D1 and Ki-67. Methods Pancreatic cancer tissues were obtained from 63 patients [36 female, 27 male] undergoing pancreatic surgery for primary pancreatic adenocarcinoma in Dokuz Eylül University Hospital between the years 1996 and 2003. The mean age of the patients was 62 years, with a range of 42-82. The patients were not subjected to any sort of chemotherapy or radiation therapy prior to resection. Surgical procedures consisted of either distal pancreatectomy or Whipple procedure. Some of the patients underwent palliative surgical treatment. The histological diagnosis of each specimen was provided by standard light microscopic evaluation of the routinely processed and parafin embedded tissues. The H&E stained slides of each case were taken from the pathology archieves and reviewed by two pathologists without knowledge of the patients' outcome. The tumors were typed as proposed by WHO and graded as well, moderate or poorly differentiated. International Union Against Cancer [UICC] TNM classification system [sixth edition] was used for staging. Staging was based on the information at the time of diagnosis. Lymphatic vessel, blood vessel and perineural invasions as well as lymph node, resection margin and adjacent organ involvements were reevaluated in each tumor. Analysis of immunohistochemical data TGF-β1 scoring The stained slides were examined microscopically using the following parameters and semiquantitative criteria. Cytoplasmic immunostaining in tumor cells were accepted as positive. For evaluation of the immunohistochemical results, the intensity of the immunoreaction and the number of the immunoreactive cancer cells were scored on the tissue slides. The intensity of the immunoreaction was classified into 4 levels: 0: no staining, 1: weak staining, 2: moderate staining, 3: intense staining. In addition, the percentage of the immunoreactive cancer cells was recorded; 0, negative; 1, less than 33% positive staining; 2, 33% to 66% positive staining, 3, more than 66% positive staining. A combined score of 0 to 6 was assigned. Tumors given a score 0 to 1 were classified as negative; those given a score of 2 were classified as weakly positive; a score of 3 to 4 was considered moderately positive; and a score of 5 to 6 was considered strongly positive [19]. p27 scoring Approximately 500 cells from the selected area were counted, and the number and percentage of cells with nuclear p27 staining were noted. Cells that had either weak or strong nuclear staining were considered to be positive for p27 protein expression. Cells without nuclear staining were considered negative. The percentage of cells expressing p27 was recorded as the ratio of positive cellsto-the total number of cells counted. p27 expression was grouped as high or low. Patients with low p27 expression had less than 30% of the nuclei in the specimen staining positive, while those with high levels of p27 expression had more than 30% of the nuclei staining for the bound antibody [15]. Ki-67 scoring Fifty-two cases were stained with Ki-67. Nuclear immunostaining in tumor cells were accepted as positive. At least 10 fields containing approximately 500 tumor cells were counted each per case at 400 × magnification. The number of Ki-67-staining reactive cells in each field was determined as a percentage of the total number of tumor cells counted [20]. Statistical analysis Data were analyzed by a computer software SPSS for Windows 10.0. A p value <0.05 was considered statistically significant. Immunohistochemical scores were compared among prognostic parameter subgroups by the chi-square and Fishers exact tests. Correlations between non-parametric and parametric values were tested with Kendall's tau-b and Spearman correlation tests respectively. The effect of prognostic and immunohistochemical scores on the patient survival was tested in potentially curatively operated patients who had R0 resections, with Cox regression analysis. Clinicopathologic features Clinicopathologic features of the patients is shown in Table 1. Thirty-six were women, 27 were men. The median age of the patients was 62 years, with a range of 42-82. The size of the resected tumors ranged between 1 and 13 cm [median 4.5 cm]. Sixteen tumors were well differentiated, 33 were moderately differentiated, and 14 were poorly differentiated. Four, 23, 3 and 33 patients had stage I, II, III and IV disease. Thirty-nine [61.9%] patients had one and more than one regional lymph node metastasis and 33 Table 2]. Although the number was so small high grade p27 staining tended to be in low stage tumors [stage grouping was performed as: I and II as low, III and IV as high], but the difference was not statistically significant. On the other hand, there was no association between p27 expression and other clinicopathologic parameters. In statistical analysis, there was also no difference between p27 and cyclin D1 expression and Ki-67. Forty-nine [77.8%] tumors showed positive nuclear staining with cyclin D1 while 3 [4.8%] cases showed no staining. There was a statistically significant difference among cyclin D1 nuclear staining scores in terms of tumor size [p = 0.03]. The other prognostic parameters did not relate significantly with nuclear or cytoplasmic staining with cyclin D1. In the whole group, the mean proliferative index of the tumors were 41.5%, ranging from 2% to 97%. In statistical analysis, no significant relation was found between mean proliferative index of the tumors and prognostic parameters. TGF-β1, p27 and clinical outcome The median survival time was 13.4 months [range, 0-59 months] for low-expressors and the median survival time was 13.7 months [range, 1-38 months] for highexpressors for TGF-β1. The median survival time was 13.6 months [range, 0-59 months] for low-expressors and the median survival time was 13.4 months [range, 6-26 months] for high-expressors for p27. In the univariate analysis, neither TGF-β1 nor p27 expression was related with patient survival. Discussion Several tumor markers are studied to explain the mechanism of pancreatic carcinogenesis. Dysregulation of the normal cell cycle machinery is integrated to the neoplastic process and there is evidence implicating loss of cell cycle control in the development and progression of most human cancers, including pancreatic cancer. Regulation of cell cycle depends on the complex association between cyclin, CDK and CDKIs. According to their effects on cell cycle progression, these regulators qualify as proto-oncogenes [cyclin D1, D2, E] or tumor supressors [p21, p27]. Recently, studies have been performed in demonstration of p21, p27, cyclin D in tumors [17,[22][23][24][25]. p27 is a CDKI that negatively regulate cyclin and its low expression results in proliferation of DNA damaged cells. Recent studies indicate that reduced expression of p27 is an independent predictor of poor outcome in cancers such as oral tongue squamous cell carcinoma, prostate cancer, gastrointestinal tract cancers, and even laryngeal precancerous lesions [14][15][16][17][18]26,27]. Lu et al demonstrated that loss of p27 expression independently predicts a poor prognosis for patients with resectable pancreatic adenocarcinomas [24]. Also anomalous overexpression of p27 in human pancreatic endocrine tumors were reported [25]. Inverse correlation between p27 expression and tumor grade, lymph node metastasis and stage of tumor was reported [14]. In the present study, we found positive staining for p27 in only minority of pancreatic adenocarcinoma cases and the staining was weak indicating that p27 is down-regulated during neoplastic process in pancreas. Additionally, although the difference was not statistically significant, p27 was dominantly expressed in low stage tumors. These results confirmed the previous reports. In a few reports, p27 and p21 staining did not correlate significantly with any clinicopathologic parameters [28,29], but there was an association between p27 and cyclin D1 expression indicating that the balance between the two opposing regulators was important for the end result of cell cycle progression [22]. Del Pizzo and Armengol found a positive correlation between p27 and cyclin E [30,31]. In the present study we could not find any relation between p27 and cyclin D1 expression. It is well established that cell proliferation kinetics are important in predicting the prognosis of various tumors. Recenty, the highly proliferative activity as measured by Ki-67 antigen has been shown to correlate with a reduced p27 expression in various tumors suggesting that decreased expression of p27 may play important role in the increased cell turnover [32]. But the lack of correlation between cell cycle regulators and proliferation index has been also reported in some studies [22,27,31]. In the present study, we did not find any significant association between p27, cyclin D1 and proliferative activity. Cell cycle regulators might be more closely related with differentiation than proliferative activity, so induction of them may perhaps be necessary, but probably not sufficient for inhibition of cell cycle. Growth factors, such as TGF-β and growth factor receptors have been shown to be overexpressed in a variety of cancers [10,33]. TGF-β can act as both a tumor suppresor and as a stimulator of tumor progression, invasion and metastasis. TGF-β inhibits normal cell proliferation and this effect is induced by binding specific cell receptors, type I [TβR-I ALK5 ] and type II [TβR-II]. Both receptors were present in most cancer tissues and presence of them was associated with advanced tumor stage [10]. In pancreatic cancers, all three TGF-β isoforms are expressed at high levels and the presence of TGF-β isoforms is associated with shorter postoperative survival [34]. TGF-β isoforms were also markedly increased in hepatocellular carcinoma and prostate carcinoma [35][36][37]. Overproduction of TGF-β1 and loss of TGF-β-RII expression were found to be associated with poor clinical outcome [10]. Nevertheless several reports have indicated that TGF-β1 expression showed a significant correlation with low grade tumors and with longer survival time [19,29]. In the present study there was a statistically significant difference between TGF-β1 staining scores in terms of clinicopathologic factors such as blood vessel invasion, distant metastasis and stage. Our analysis revealed that overproduction of TGF-β1 correlates with tumor progression and metastasis. In the present study we could not find a correlation between TGF-β1 and p27 expression. TGF-β1 might influence cancer growth in other ways rather than cell cycle proteins. Hashimoto demonstrated a significant correlation between TGF-β1 and p21 expression and stated that the evaluation of TGF-β1 and p21 expression might be an effective tool in the prediction of the prognosis of patients with pancreatic cancer [29]. Grau and Ito also reported that TGF-β1 induces p21 expression in pancreatic cancer cells [11,38]. p21 may play more important role in the biology of pancreatic cancer than p27. We found no correlation between TGF-β1 and other proliferation markers studied. TGF-β1 induction of other CDKIs, such as p15 may partially participate in the inhibition of mitosis. There is conflicting data in the literature about the correlation of TGF-β1 expression with survival in pancreatic cancer. In most series, the presence of TGF-β1 is associated with shorter postoperative survival [34], although in some other studies TGF-β1 expression showed a significant correlation with longer survival time [19,29]. On the other hand, there are reports indicating that TGF-β expression has no association with survival [39]. In the present study, to have the same treatment protocol only the patients which had RO resections were considered in survival analysis and TGF-β1 and p27 expression did not show a significant influence on the survival of patients. The small number of the patients may be the reason not to have definitive conclusions. TGF-β's role in controlling cellular growth and differentiation is very complex. In this study, findings may suggest that TGF-β1 may be a factor in the extention of the disease in pancreatic adenocarcinomas. But, progress may depend on the participation of other regulatory signals. Additional parameters related to prognosis and standard prognostic parameters are needed in the prediction of the disease outcome. Further studies that involve screening a larger number of patients with invasive disease will be required to establish the potential clinical usefulness of cell cyle regulators as prognostic markers in pancreatic cancer. Conclusion In the present study we examined the expression paterns of TGF-β1 and p27, which involved in cell cycle regulation, in pancreatic adenocarcinoma. Our findings suggest that in pancreatic carcinoma, TGF-β1 overexpression is more pronounced in metastatic and progressive ones. p27 expression is reduced in pancreatic adenocarcinomas and it may play a role in the differentiation of pancreatic cancer. Therefore, evaluation of these proteins might be useful in determining the agressive capacity of these tumors at an earlier stage.	2017-06-22T03:39:19.491Z	2005-08-08T00:00:00.000	{ "year": 2005, "sha1": "02a1d04d0745e2e8fd34948329024332c80b4f47", "oa_license": "CCBY", "oa_url": "https://bmccancer.biomedcentral.com/track/pdf/10.1186/1471-2407-5-98", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "02a1d04d0745e2e8fd34948329024332c80b4f47", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
119314137	pes2o/s2orc	v3-fos-license	Combinatorial extension of stable branching rules for classical groups We give new combinatorial formulas for decomposition of the tensor product of integrable highest weight modules over the classical Lie algebras of type $B, C, D$, and the branching decomposition of an integrable highest weight module with respect to a maximal Levi subalgebra of type $A$. This formula is based on a combinatorial model of classical crystals called spinor model. We show that our formulas extend in a bijective way various stable branching rules for classical groups to arbitrary highest weights, including the Littlewood restriction rules. 1. Introduction 1.1. Stable branching rules. Let G be a classical group over the complex numbers with a closed subgroup H. For a finite-dimensional irreducible G-module V , there are various branching rules for the multiplicity of an irreducible H-module in V , which are given as a sum of product of Littlewood-Richardson coefficients. These formulas, which originated in the Littlewood restriction rules [32,33], are often referred to as stable branching rules since they hold in a certain range of highest weights of irreducible modules and depend only on the partitions parameterizing highest weights when the ranks of classical groups are sufficiently large. We refer the reader to [12] for a systematic approach to these formulas and detailed exposition on previous works. The goal of this paper is to give combinatorial extension of various stable branching rules to arbitrary highest weights, that is, to give a combinatorial formula for the branching multiplicity, which holds for arbitrary highest weights and also extends a given stable branching rule in a bijective way. We recall that there are formulas extending stable branching rules to arbitrary highest weights (for example, [7,21,22,23,24,36,41]) where the multiplicities outside a stable range are in general given as an alternating sum of products of Littlewood-Richardson coefficients or determined by certain modification and cancellation of irreducible factors in the decomposition inside a stable range. But compared to these results which are obtained in an algebraic way, relatively not much is known about combinatorial or subtraction-free extension. On the other hand, the theory of crystals [18,19] has made it possible to provide formulas for the branching multiplicities, which are available for arbitrary highest weights and described explicitly in terms of various combinatorial objects (for example, [2,17,20,31]). However, most of the known results do not seem to imply the stable branching rules immediately. So one may ask which combinatorial model for classical crystals explains and extends the stable branching formulas more naturally. Combinatorial extensions. Let us explain our results in details. Our approach is based on the theory of crystal base, and the theory of reductive dual pairs [11]. Let g ∞ be the affine Kac-Moody Lie algebra of type B ∞ , C ∞ , and D ∞ [16], and let l ∞ be its maximal Levi subalgebra of type A. We first give a new formula for decomposition of the tensor product of integrable highest weight g ∞ -modules, and the branching decomposition of an integrable highest weight g ∞ -module with respect to l ∞ . Our formula is given in terms of a combinatorial model for the g ∞ -crystal associated to an integrable highest weight g ∞ -module, which was recently introduced by the author [26,27]. A main advantage of this model, which we call spinor model, is that it is compatible with the combinatorics of usual Young and Littlewood-Richardson tableaux, and hence it turns out to fit into the stability phenomenon of decomposition numbers very nicely. The notion of spinor model is motivated by the duality on the fermionic Fock space F n 2 , where (g ∞ , G n ) forms a reductive dual pair on F n 2 for some complex classical reductive algebraic group G n [11,40]. Let P(G n ) be the set of partitions parameterizing the highest weights for the finite-dimensional irreducible G n -modules appearing in F n 2 . By dual correspondence, we may parameterize the dominant integral weights for g ∞ by P(G n ) for all n. Let T g (λ, n) denote the spinor model of the crystal associated to an integrable highest weight g ∞ -module with highest weight corresponding to λ ∈ P(G n ). Here g = b, c, d is understood as a symbol representing the type of g ∞ . Now, let λ ∈ P(G m+n ), µ ∈ P(G m ), and ν ∈ P(G n ) be given. The multiplicity of T g (λ, m+n) in the tensor product T g (µ, m)⊗T g (ν, n) is given by the cardinality of a subset LR λ µν (g) of T g (ν, n) satisfying certain combinatorial conditions determined by the general theory of crystal base (Proposition 4.4). Hence we may regard elements in LR λ µν (g) as a generalization of Littlewood-Richardson tableaux. Then we show as our first main result that LR λ µν (g) has a nice decomposition into a set of pairs of Littlewood-Richardson tableaux in a stable range. More precisely, we construct an explicit bijection (Theorem 4.5) Here P is the set of partitions, P g is a subset of P given in (4.1), LR α βγ is the set of Littlewood-Richardson tableaux of shape α/β and content γ for α, β, γ ∈ P, λ ′ is the conjugate of λ, and ℓ(λ) is the length of λ. Next, we consider the branching decomposition with respect to l ∞ . Let µ ∈ P(G n ) be given. For λ ∈ P, we denote by LR µ λ (g) the subset of T g (µ, n), which consists of the highest weight elements of the l ∞ -subcrystals in T g (µ, n) isomorphic to the crystal of Young tableaux of shape λ ′ . The cardinality of LR µ λ (g) is the associated branching multiplicity. We show as our second main result that LR µ λ (g) decomposes into a set of Littlewood-Richardson tableaux in a stable range by constructing an explicit bijection Finally, by the reciprocity laws associated to suitable see-saw pairs on F n 2 , the bijections in (1.1) and (1.2) recover well-known stable branching rule for the pairs (I) G m+n ⊃ G m × G n , (II) GL n ⊃ G n (the Littlewood restriction rule), with G ℓ = Sp ℓ , O ℓ (Theorems 5.1 and 5.3), and therefore the sets LR λ µν (g) and LR µ λ (g) extend these stable branching rules to arbitrary highest weights in a bijective way. On the other hand, considering a subset of tableaux in T g (λ) with entries from {1, . . . , k}, we can also describe the tensor product decomposition and branching decomposition for crystals of type B k , C k , and D k in terms of LR λ µν (g) and LR µ λ (g). That is, we have branching rules for the pairs with G ℓ = Sp ℓ , Spin ℓ , Osp 1\|ℓ and k = [ℓ/2], together with the stable limits (1.1) and (1.2), respectively (see Remark 4.10). In this case, the formula (1.1) for (III) looks different from the known result, and the formulas in (III) and (IV) for Spin ℓ -modules with half-integer weights seem to be new. An extension of stable branching formula similar to (1.1) and (1.2) has been obtained in [25] by using a different model of g ∞ -crystals for g = b, c but not for g = d, which is a technically more difficult case. For other combinatorial extensions, we refer the reader to [39] for (II) with G n = Sp n , and [1, Proposition 2.6] for (IV) with G ℓ = Sp ℓ , SO ℓ . Recently, there have been several works studying stability phenomenon from completely different viewpoints from ours. In [6,37,38] (see also references therein), the branching multiplicities in a stable range are studied in the context of a nonsemisimple tensor category of modules over classical Lie algebras of infinite rank. Also in [41], the extension of stable branching multiplicities in (II) has been studied using q-versions of Brauers centralizer algebras. We expect that our combinatorial approach would give a new insight into the structure of representations of classical Lie algebras and groups which are related to the stability phenomenon. 1.3. Application to holomorphic discrete series. We give an interesting application of spinor model to another stability phenomenon. In [12] a systematic approach of stable branching rule is given by using Howe's theory of reductive dual pairs [11], which also partly motivated the work in the present paper. Instead of integrable highest weight g ∞ -modules, a family of infinitedimensional unitarizable representations of classical Lie algebras of finite rank, which appears in the study of dual pair acting on bosonic spaces, is considered in [12], and then the irreducibility of their associated generalized Verma modules in a certain stable range plays a crucial role in giving a unified proof of 10 families of stable branching formulas classified there. We show that the spinor model T g (λ, n) admits a variation to give a combinatorial character formula for these unitarizable representations (Theorem 6.1) thanks to super duality [5]. Hence we obtain the tensor product multiplicity and the branching multiplicity with respect to a maximal Levi subalgebra type A for these irreducible modules. In particular, the branching multiplicity formula implies the irreducibility of associated generalized Verma modules in a stable range called holomorphic discrete series (Corollary 6.6). 1.4. Organization. The paper is organized as follows. In Section 2, we briefly recall necessary background. In Section 3, we review the spinor model of crystals of types BCD. In Section 4, we establish the bijections (1.1) and (1.2). In Section 5, we recover well-known stable branching rules by using the dualities on Fock spaces. In Section 6, we discuss a combinatorial character formula for infinite-dimensional unitarizable representations. Preliminaries 2.1. Lie algebras of type B, C, and D. We assume that our base field is C, and Z + (resp. N) denotes the set of non-negative (resp. positive) integers. Let b ∞ , c ∞ , and d ∞ be the Kac-Moody Lie algebras of type B ∞ , C ∞ , and D ∞ , respectively (see [16]). Following the conventions of g ∞ (g = b, c, d) in [3, Section 2.2], we let · I = Z + : the index set for simple roots, the set of simple coroots, · Λ g i : the fundamental weight with respect to α i , · P = ZΛ g 0 ⊕ i≥1 Zǫ i : the weight lattice, where the Dynkin diagram of the Cartan matrix ( α j , α ∨ i ), Π, and Λ g i are given by We also let l ∞ be the subalgebra of g ∞ associated to { α i \| i ∈ I \ {0} }, which is of type A +∞ (cf. [16]). We assume that its weight lattice is i≥1 Zǫ i ⊂ P . We will also consider the following Kac-Moody Lie superalgebra b • ∞ of infinite rank, whose Dynkin diagram, Π, and the fundamental weights Λ b • i are given by where ② denotes a non-isotropic odd simple root (cf. [15]). We use the same notations for the associated data as in the case for g ∞ (g = b, c, d). For g = b, b • , c, d and k ∈ N, let g k be the subalgebra of g ∞ whose Dynkin diagram corresponds to the simple roots α 0 , . . . , α k−1 . Let l k = l ∞ ∩ g k be the corresponding subalgebra of type A k−1 . Let V and W be modules over a Lie algebra g, where W is irreducible. We define [V : W ] = dim Hom g (W, V ), the multiplicity of W in V . 2.2. Dual pairs. Throughout the paper, G n denotes one of the following complex reductive algebraic groups: Sp n , O n , Spin n and Pin n for n ≥ 2, where n is even for Sp n and Pin n , and n is odd for Spin n . Following [3, Section 2.2], let V λ Gn denote the finite-dimensional irreducible representation of G n corresponding to λ ∈ P(G n ) (see also [8,11]), where Here is the conjugate partition of λ = (λ 1 , λ 2 , . . .). From now on, we mean by (g ∞ , G n ) one of the pairs unless otherwise specified. For λ ∈ P(G n ), we define a dominant integral weight Λ g (λ) for g ∞ by where ǫ = 2 if g = c, and ǫ = 1 otherwise. Note that n P(G n ) parameterizes the set of all dominant integral weights for g ∞ . We denote by L(g ∞ , Λ g (λ)) the irreducible highest weight g ∞ -module with highest weight Λ g (λ). For a positive integer ℓ ≥ 1, let be ℓ pairs of free fermions with non-trivial commutation relations [ψ +,i r , ψ −,j s ] = δ ij δ r+s,0 . Let F ℓ denote the corresponding Fock space generated by the vacuum vector \|0 , which is annihilated by ψ +,i r , ψ −,i s for r, s > 0. We introduce a neutral fermionic field φ(z) = r∈ 1 2 +Z φ r z −r− 1 2 with non-trivial commutation relations [φ r , φ s ] = δ r+s,0 . Denote by F 1 2 the Fock space of φ(z) generated by a vacuum vector that is annihilated by φ r for r > 0. We denote by F ℓ+ 1 2 the tensor product of F ℓ and F 1 2 . Then we have the following (g ∞ , G n )-duality on F n 2 . Proposition 2.1. [40] There exists an action of g ∞ × G n on F n 2 . Furthermore, under this joint action, we have 2.3. Crystals. Let us give a brief review on crystals (see [9,19] for more details). Let g be a Kac-Moody algebra associated to a symmetrizable generalized Cartan matrix A = (a ij ) i,j∈I indexed by a set I. Let P ∨ be the dual weight lattice, P = Hom Z (P ∨ , Z) the weight lattice, Π ∨ = { α ∨ i \| i ∈ I } the set of simple coroots, and Π = { α i \| i ∈ I } the set of simple roots of g such that α j , α ∨ i = a ij for i, j ∈ I. Let U q (g) be the quantized enveloping algebra of g. A g-crystal (or crystal for short) is a set B together with the maps wt : B → P , ε i , ϕ i : B → Z∪{−∞} and e i , f i : B → B ∪{0} (i ∈ I) satisfying certain axioms. For a dominant integral weight Λ for g, we denote by B(g, Λ) the g-crystal associated to an irreducible highest weight U q (g)-module with highest weight Λ. Let B 1 and B 2 be crystals. A tensor product Crystals of Young tableaux and Littlewood-Richardson coefficients. Let P be the set of partitions. For λ = (λ i ) i≥1 ∈ P, we denote by ℓ(λ) the length of λ. We identify λ with a Young diagram as usual. For a skew Young diagram λ/µ, let SST (λ/µ) be the set of semistandard tableaux of shape λ/µ with entries in N. For T ∈ SST (λ/µ), let w(T ) be the column word of T , that is, the word given by reading the entries of T column by column from right to left and from top to bottom in each column. For T ∈ SST (λ) and a ∈ N, we denote by a → T the tableau obtained by the column insertion of a into T (cf. [13]). For a semistandard tableau S, we define (S → T ) = (w(S) → T ). Considering N as a l ∞ -crystal (of type A +∞ ) associated to the natural representation of l ∞ , we may regard SST (λ/µ) as an l ∞ -crystal [20]. In particular, Let LR λ µν be the set of Littlewood-Richardson tableaux corresponding to λ, µ, ν ∈ P, which is the set of tableaux T ∈ SST (λ/µ) with weight ν such that w(T ) is a lattice word (cf. [13]). Let c λ µν = \|LR λ µν \| be the Littlewood-Richardson coefficient corresponding to λ, µ, ν ∈ P. Recall that c λ µν is the number of connected components in SST (µ) ⊗ SST (ν), which is isomorphic to SST (λ) as an l ∞ -crystal. By tensor product rule of crystal, we may regard LR λ µν as the set of T ∈ SST (ν) such that wt(H µ ) + wt(T ) = wt(H λ ), and Spinor model for crystals of types B, C, D In this section, we briefly review the combinatorial model for B(g ∞ , Λ g (λ)) (λ ∈ P(G n )), which was introduced by the author in [26,27] (see also [29,Section 2] for more details on the existence of the crystal B(g ∞ , Λ g (λ)) for U q (g ∞ )). 3.1. Crystals of fundamental representations of g ∞ . In this subsection, we describe a combinatorial model for B(g ∞ , Λ g a ) (a ∈ Z + ). It is given by translating the q-deformed Fock space models for classical Lie algebras due to Hayashi [10] in terms of semistandard tableaux of skew shapes with at most two columns. For a single-columned tableau U , let ht(U ) denote the height of U , and U (i) (resp. U [i]) the ith entry of U from the bottom (resp. the top) for i ≥ 1. For a, b, c ∈ Z + , let λ(a, b, c) = (2 b+c , 1 a )/(1 b ), a skew Young diagram with heights a + c and b + c from the left, where a and b denote the heights of lower and upper single columns in the diagram, respectively. For example, Let T be a tableau of shape λ(a, b, c), each of whose columns is a semistandard tableau. We denote by T L and T R the left and right columns of T , respectively. We also denote by T tail the subtableau of T corresponding to the tail of λ(a, b, c), a lower single column of height a, and denote by T body the subtableau of T above T tail . Suppose that T is semistandard, that is, T ∈ SST (λ(a, b, c)). One may slide down T R by k positions for 0 ≤ k ≤ min{a, b} to have a (not necessarily semistandard) tableau T ′ of shape λ(a − k, b − k, c + k). We define the residue of T to be the maximal k such that T ′ is semistandard, and denote it by r T . For example, r S = 1 and r T = 2, when For a ∈ Z + , let We also put , Here we define the residue r T of T ∈ T sp to be the residue of ht(T ) modulo 2. (Note that if T ∈ SST (λ(0, b, 0)) ⊂ T g (0), then the shape of T is a single column but r T = 0.) Now, let B be one of T g (a) (a ∈ Z + ), T sp , and T d (0), and let T ∈ B be given. First, we define e i T and f i T for i ∈ I \ {0} regarding B as a subset of an l ∞ -crystal λ∈P SST (λ). Next, we define e 0 T and f 0 T as follows: (1) Suppose that g = c and B = T c (a). We define e 0 T to be the tableau obtained from T by removing a domino 1 1 if T has 1 1 on its top, and 0, otherwise. We define f 0 T in a similar way by adding 1 1 on top of T . (2) Suppose that g = b. When B = T sp , we define e 0 T to be the tableau obtained from T by removing 1 if T has 1 on its top, and 0, otherwise. We define f 0 T in a similar way by adding 1 on top of T . When B = T b (a), we definex 0 T (x = e, f ) to bex 0 (T R ⊗ T L ) regarding T sp as a regular sl 2crystal with respect to e 0 and f 0 , and applying the tensor product rule of crystals to B ⊂ (T sp ) ⊗2 . (3) Suppose that g = d. When B = T sp , we define e 0 T to be the tableau obtained from T by removing a domino 1 2 if T has 1 2 on its top, and 0 otherwise. We define f 0 T in a similar way by adding (2). We put where m i is the number of occurrences of i in T (cf. (2.2)), and Then B is a g ∞ -crystal with respect to wt, (1) For g = c, Note that the highest weight element in B is given by H (1 a ) , ∅, H (1) , and H (2) , 3.2. Admissibility. To describe a g ∞ -crystal B(g ∞ , Λ) associated to arbitrary integral dominant weight Λ = Λ g a 1 + · · · + Λ g ar , we embed B(g ∞ , Λ) into the tensor product B(g ∞ , Λ g a 1 ) ⊗ · · · ⊗ B(g ∞ , Λ g ar ), and then describe a connected component of the highest weight element with weight Λ. For this, we need to characterize an explicit condition for to be in the same connected component including the highest weight element of weight Λ g a i + Λ g a j . This condition is called admissibility, which is an analogue of semistandardness between two adjacent columns in type A. For a ∈ Z + , let T ∈ T g (a) be given. Associated to T , we introduce the following two pairs of single-columned tableaux ( L T, R T ) and (T L * , T R * ), which will play a crucial role in describing admissibility. First, we define ( L T, R T ) to be the pair determined by the following algorithm: (1) Let y i = T R (i) for 1 ≤ i ≤ ht(T R ). First, slide down the box y 1 in T R as far as the entry of T L in the same row is no greater than y 1 . If no entry of T L is greater than y, we place y 1 next to the bottom of T L . (2) Next, slide down y 2 until it is above y 1 and the entry of T L in the same row is no greater than y 2 . Repeat the same process with the other boxes y 3 , y 4 , . . . until there is no moving down. Next, when r T = 1, we define (T L * , T R * ) to be the pair determined by the following algorithm: (1) Let x i = T L [i] for 1 ≤ i ≤ ht(T L ). First, slide upward x 1 until the entry of T R in the same row is no smaller than x 1 . If no entry of T L is smaller than x 1 , we place x 1 next to the top of T R . (2) Next, slide upward x 2 until it is below x 1 and the entry of T R in the same row is no smaller than x 2 . Repeat the same process with the other boxes x 3 , x 4 , . . . until there is no moving up. (3) Choose the lowest box y in T R whose left position is empty, and then slide it to the left (there exists at least one such y since r T = 1). (4) Define T L * to be the tableau determined by the boxes x i 's in T L together with y , and define T R * to be the tableau given by the remaining boxes on the right. T L * T R * Note that the pairs (T L , T L * ) and (T R , T R * ) are arranged to share the same bottom lines, respectively. We refer the reader to [26,Section 6] and [27,Section 3] for more details on the well-definedness of ( L T, R T ) and (T L * , T R * ). Definition 3.2. Let a, a ′ ∈ Z + be given with a ≥ a ′ . We say that a pair (T, S) is admissible, and write T ≺ S if it is one of the following cases: for i ≥ 1 (Here ε = 1 if S ∈ T sp− and 0 otherwise, and we assume that which implies that T ≺ S. The following lemma will be used in the next section. Since r T = 0 or r S = 0, we have T L (i + a − a ′ ) ≤ S L (i) for i ≥ 1 by Definition 3.2(1). (2) Suppose first that g = b or c. In this case, we always have r T = r S = 0. We have L S(1) ≤ S R (1) by definition of L S, and hence T R (1) ≤ L S(1) ≤ S R (1) by Definition 3.2(1). Now, we suppose that g = d. If r T = 0 or r S = 0, then we also have T R (1) ≤ S R (1) by the same argument as in g = b or c. So we may assume that r T = r S = 1. Let x = T R (1). Suppose that x moves to the right when we construct T L * . This implies that T L (a) < x < T L (a − 1), and hence R T (a) = x. On the other hand, we have S L * (a ′ ) = S L (a ′ ) ≤ S R (1) or S L * (a ′ ) = S R (1) since r S = 1 depending on whether y with y = S R (1) moves to the left or not when we construct S L * . By Definition 3.2(1), we have x = R T (a) ≤ S L * (a ′ ) ≤ S R (1). Next, suppose that x does not move to the right when we construct T L * . This implies that T R * (1) = x. By Definition 3.2(1), we have x ≤ L S(1) ≤ S R (1). This completes the proof. 3.3. Highest weight crystals. Let (g ∞ , G n ) be one of the dual pairs in (2.1) and λ ∈ P(G n ). Suppose that (g ∞ , G n ) = (d ∞ , O n ). We put Suppose that (g ∞ , G n ) = (d ∞ , O n ). Let q ± and r ± be non-negative integers such that Note that 2L + r ± = n. Then we put Definition 3.5. For λ ∈ P(G n ), we define We regard T g (λ, n) as a g ∞ -crystal by identifying T = (T 1 , T 2 , . . .) ∈ T g (λ, n) with · · · ⊗ T 2 ⊗ T 1 , and regard T g (λ, n) as its subcrystal. Let H λ be the element (T 1 , T 2 , . . .) in T g (λ, n) such that each T k is a highest weight element for all k ≥ 1, or e i T k = 0 for all i ∈ I. Then we have the following by [26,Theorem 7.4] for g = b, c and [27, Theorem 4.4] for g = d. Theorem 3.6. For λ ∈ P(G n ), we have with highest weight element H λ of weight Λ g (λ). Remark 3.7. Following [14], one can see that T b (λ, n) for λ ∈ P(G n ) (G = Pin or Spin) coincides with the crystal graph associated to an integrable highest weight module over b • ∞ with highest weight Λ b (λ). So the tensor product and branching multiplicities for b ∞ are equal to those for b Remark 3.8. For k ∈ N and λ ∈ P(G n ) with λ 1 ≤ k, let T g k (λ, n) be the subset of T g (λ, n) with entries from { 1, . . . , k }. Then T g k (λ, n) is a g k -crystal associated to a finite-dimensional irreducible g k -module with highest weight Λ g (λ) for g = b, b • , c, d with respect to e i and f i for i ∈ {0, . . . , k − 1}. Stability in tensor product and branching decomposition 4.1. Separation Lemma. We put Consider a g ∞ -crystal T g (λ, n) for λ ∈ P(G n ). Let T = (T 1 , . . . , T r ) ∈ T g (λ, n) given. We put w(T) = w(T r ) . . . w(T 1 ), and define L(T) to be the maximal length of a weakly decreasing subword of w(T). We put For g = d and T i ∈ T d (0) or T sp− , we assume that T tail i is the subtableau of T i consisting of the boxes at the bottom, and T body i is the subtableau consisting of the other part of T i . We assume that T body i and T tail i are separated by a common horizontal line. Then we may regard T tail as a tableau of shape λ ′ , where T tail i is the ith column from the left for 1 ≤ i ≤ r, and regard T body as a tableau of shape µ π for some µ ∈ P, when r T 1 = 0 for 1 ≤ i ≤ r. Here µ π denotes a skew Young diagram obtained by 180 • -rotation of µ. Note that T body is not semistandard. The following lemma will play a crucial role in this paper. Lemma 4.3. If L(T) ≤ n 2 , then we have (1) T tail ∈ SST (λ ′ ), and T body ∈ SST (µ π ) for some µ ∈ P g , (2) T ≡ A T body ⊗ T tail , where ≡ A denotes the equivalence as elements of l ∞crystals. Proof. Case 1. Suppose that g = c. Let T = (T 1 , . . . , T n/2 ) ∈ T c (λ, n) with L(T) ≤ n/2. We may regard T tail as a tableau of shape λ ′ and T body as a tableau of shape µ π for some µ ∈ P c . We first have by Lemma 3.4(1) that T tail is semistandard and hence T tail ∈ SST (λ ′ ). Next, we claim that T body is semistandard. Since T body i is a semistandard tableau of shape λ(0, 0, c i ) for some c i ∈ Z + with (T which yields a weakly decreasing subword of w(T) of length n/2 + 1. This is a contradiction, which proves our claim. Therefore, T body ∈ SST (µ π ). Finally, we claim that T ≡ A T body ⊗ T tail . By the semistandardness of T body and Lemma 3.4(2), we have for 2 ≤ i ≤ n/2 Since ≡ A coincides with the usual Knuth equivalence, one can check by (4.3) and considering the column insertion of tableaux that Applying (4.4) successively, we have The proof is almost identical to Case 1. We leave the details to the readers. Case 3. Suppose that g = d. Let T = (T 1 , . . . , T r ) ∈ T d (λ, n) with L(T) ≤ n/2. By Lemma 4.1, we have ℓ(λ) ≤ n/2, and hence n − 2λ ′ 1 ≥ 0. We first claim that r T i = 0 for all 1 ≤ i ≤ r. Suppose that r T i = 1 for some 1 ≤ i ≤ λ ′ 1 , and let j be the smallest one such that r T j = 1. By Lemma 3.4, we have a sequence Note that T R j+1 (1), . . . , T R r (1) are non-empty since T i ≺ T i+1 for j ≤ i ≤ r − 1. Reading from right to left, we get a subword of w(T) with length r + 1. Since 2λ ′ 1 + 2q + + r + = n and r = λ ′ 1 + q + + r + , we have r + 1 > n/2. This is a contradiction, which proves our claim. Now by Lemma 3.4(1), we have T tail ∈ SST (λ ′ ). By the same arguments as in Case 1 we conclude that T body ∈ SST (µ π ) and T ≡ A T body ⊗ T tail . 4.2. Stable tensor product and branching rules. Let µ ∈ P(G m ) and ν ∈ P(G n ) be given. For λ ∈ P(G m+n ), let (4.5) Then we have the following formula of tensor product multiplicity for integrable highest weight g ∞ -modules. Proposition 4.4. For µ ∈ P(G m ), ν ∈ P(G n ), and λ ∈ P(G m+n ), we have Proof. By tensor product rule of crystals, we have LR λ ). Now we have the following stable tensor product rule, which is the first main result in this paper. Since H µ ⊗ T body ≡ A H µ ′ ⊗ T body ≡ A H γ for some γ ∈ P and there exists a unique tableau U ∈ SST (δ) such that U ≡ A T body , we may regard T body as an element in LR γ µ ′ δ (see Section 2.4). Then we have Conversely, suppose that a pair (U, V ) ∈ LR γ µ ′ δ × LR λ ′ γν ′ is given for γ ∈ P and δ ∈ P g . Here we assume that U ∈ SST (δ π ) such that H µ ′ ⊗ U ≡ A H γ , and V ∈ SST (ν ′ ) such that H γ ⊗ U ≡ A H λ ′ . We also assume that g = c since the arguments for g = b, d are very similar. Let µ ∈ P(G n ) be given. For λ ∈ P, let LR µ Proposition 4.7. For µ ∈ P(G n ) and λ ∈ P, we have Proof. By definition, the connected component of T ∈ LR µ λ (g) in T g (µ, n) with respect to e i and f i for i ∈ I \ {0} is isomorphic to SST (λ ′ ) as an l ∞ -crystal. Hence c µ λ (g) is the multiplicities of SST (λ ′ ) or B(l ∞ , λ ′ ) in B(g ∞ , Λ g (λ)). Then we have following stable branching rule, which is the second main result in this paper. Theorem 4.8. Let µ ∈ P(G n ) and λ ∈ P be given. If ℓ(λ) ≤ n 2 , then the map T → T tail gives a bijection Proof. Let T ∈ LR µ λ (g) be given. By definition, Moreover, since T ≡ A T body ⊗ T tail ≡ A H λ ′ , we have T body ≡ A H δ and we may regard T tail as an element in LR λ ′ δµ ′ . Hence T → T tail is an injective map from LR µ λ (g) to δ∈Pg LR λ ′ δµ ′ . Conversely, suppose that V ∈ LR λ ′ δµ ′ is given for δ ∈ P g . Let us assume that g = c since the arguments for g = b, d are similar. Let U be the unique tableau in SST (δ π ) such that U ≡ A H δ . We assume that be the ith column of U from the right, and let V (i) be the ith column of V from the left for 1 ≤ i ≤ n. Then U (i) is empty for 1 ≤ i ≤ n/2, and V (i) is empty for n/2 < i ≤ n since ℓ(δ ′ ), ℓ(µ) ≤ ℓ(λ) ≤ n/2. By the same argument as in Theorem 4.5, we can show that V (i) [1] > U (2i) (1) for all 1 ≤ i ≤ n/2, and hence there exists a unique T ∈ T g (µ, n) such that T body = U and . Therefore T → T tail gives a bijection. Corollary 4.9. Let µ ∈ P(G n ) and λ ∈ P. If ℓ(λ) ≤ n 2 , then Remark 4.10. For k ∈ N and λ ∈ P(G n ) with λ 1 ≤ k, let T g k (λ, n) be as in Remark 3.8. Since T g k (λ, n) for λ ∈ n P(G n ) with λ 1 ≤ k give crystals for all finite-dimensional irreducible g k -modules, we have combinatorial extensions of the stable branching rules for the pairs G ℓ × G ℓ ⊃ G ℓ and G ℓ ⊃ GL [ℓ/2] , where G ℓ = Sp ℓ , Spin ℓ , Osp 1\|ℓ with k = [ℓ/2] by Theorems 4.5 and 4.8 (see also Remark 3.7). 5. Branching rules for classical groups 5.1. Branching rule for G m+n ⊃ G m × G n . Suppose that G = Sp or O, that is, g = c or d. Note that for λ ∈ P(G m+n ), V λ G m+n can be viewed as a G m × G n -module since there is a natural embedding of G m × G n into G m+n . For µ ∈ P(G m ) and Theorem 5.1. For λ ∈ P(G m+n ), µ ∈ P(G m ) and ν ∈ P(G n ), we have where c λ µν (g) is given in (4.5). Furthermore, if ℓ(λ) ≤ 1 2 min{m, n}, then Proof. By Propositions 2.1 and 4.4, we have The second equality follows from Corollary 4.6. Remark 5.2. By the same argument as in Theorem 5.1, we also have a branching rule for the closed subgroup Spin m × Spin n ⊂ Pin m+n for m, n odd in the case when g = b. 5.2. Branching rule for GL n ⊃ G n . Suppose that G = Sp or O. Note that G n is a subgroup of GL n . For λ ∈ P with ℓ(λ) ≤ n, let V λ GLn be the finite-dimensional irreducible GL n -module corresponding to λ, and for µ ∈ P(G n ), let V λ GLn : V µ Gn denote the multiplicity of V µ Gn in V λ GLn as a G n -module. Then we have the following, which extends the Littlewood restriction rule [32,33]. Theorem 5.3. For λ ∈ P with ℓ(λ) ≤ n and µ ∈ P(G n ), we have Proof. We first consider an action of gl n = 1≤i,j≤n Ce ij as operators on F n 2 by left multiplication as follows: For n = 2ℓ and G n = O 2ℓ , put For n = 2ℓ and G n = Sp 2ℓ , put For n = 2ℓ + 1 and G n = O 2ℓ+1 , put e ij to be the same as in (5.1) except we assume that ψ ±,n r = φ ± r . Then one can check that (5.1) and (5.2) define an action of GL n on F n 2 such that its restriction to G n coincides with the action of G n in Proposition 2.1. Recall that there exists an action of l ∞ × GL n on F n 2 such that On the other hand, by Proposition 2.1 we have Combining with (5.3), we have The second equality follows from Corollary 4.9. Example 5.4. Let λ = (2, 2) ∈ P. The decomposition number [V λ GLn : V µ On ] for each µ ∈ P(O n ) is equal to the number of T ∈ T d (µ) such that T ≡ A H λ . Using this fact, it is straightforward to check that For n = 4, we have , and the tableaux T = (T 1 , T 2 ) ∈ T d (µ) such that T ≡ A H λ are as follows: where the dashed line denotes the line separating the body and the tail of T i 's for i = 1, 2. Hence we have V (2,2) It is not difficult to see that the same decomposition of V (2,2) GLn as above holds for n ≥ 4, which is equal to the stable formula in Theorem 5.3. Let us consider the decomposition of V λ GLn (n = 2, 3), where λ is outside the stable range. For n = 3, we have is the unique tableau in T d (0, 0) such that T ≡ A H λ , and hence V (2,2) For n = 4, we have , and the tableaux T = (T 1 , T 2 ) ∈ T d (µ) such that T ≡ A H λ are as follows: Hence we have V (3,1) . of (5.4) is introduced in [28], and its stable limit is closely related with the energy function on a tensor product of finite affine crystals (see for example [30]). It would be interesting to find a combinatorial interpretation of the q-analogue of (5.4) or (5.5) in terms of spinor model. 6. Character of holomorphic discrete series 6.1. Superization of T g (λ, n). Let A be a linearly ordered countable set with a Z 2 -grading A = A 0 ⊔ A 1 . For a skew Young diagram λ/µ, we define the notion of A-semistandard tableaux of shape λ/µ as usual, that is, (1) the entries in each row (resp. column) are weakly increasing from left to right (resp. from top to bottom), (2) the entries in A 0 (resp. A 1 ) are strictly increasing in each column (resp. row). We denote by SST A (λ/µ) the set of A-semistandard tableaux of shape λ/µ. Let x A = { x a \| a ∈ A } be the set of formal commuting variables indexed by A. For λ ∈ P, let s λ (x A ) = T x T A be the super Schur function corresponding to λ, where the sum is over T ∈ SST A (λ) and x T A = a x ma a with m a the number of occurrences of a in T . For λ ∈ P(G n ), one can define T g A (λ, n) by replacing (N-)semistandard tableaux with A-semistandard tableaux in Definition 3.5. (See [26,Section 6] and [27,Section 3] for more details, where T g A (λ, n) is denoted by T g A (λ, n/ǫ).) When A is (a subset of) N, we assume that A 0 = A and it is equipped with the usual linear ordering on N. In particular, we have T g N (λ, n) = T g (λ, n). We define the character of T g A (λ, n) by where z is another formal variable and the sum is over T = (T 1 , T 2 , . . .) ∈ T g A (λ, n). It is shown that under certain choices of A (more precisely, for A with 0 < \|A 0 \| < ∞), { chT g A (λ, n) \| λ ∈ n P(G n ) } gives the characters of a family of irreducible representation of ortho-symplectic Lie superalgebra in a semisimple tensor category [26,Theorem 7.6] and [27,Theorem 4.6]. of stable branching rules for classical groups in a unified way [12]. Nevertheless the proof here is given in a completely different and purely combinatorial way by decomposing T g k (λ, n) into l k -crystals. Moreover the result for g = b • obtained here seems to be new. We put [k] = { 1, . . . , k } ⊂ N as a Z 2 -graded linearly ordered set, which can be viewed as the crystal of the natural representation of l k . As in Section 2.4, one can regard SST [k] (λ/µ) as an l k -crystal for a skew Young diagram λ/µ. For T ∈ SST A (λ/µ), we define T ′ to be the tableau of shape λ ′ /µ ′ obtained by flipping T with respect to the main diagonal and replacing i (resp. i ′ ) with i ′ (resp. i) when A is [k] (resp. [k] ′ ). Proof. Note that T g k (λ, n) can be defined in the same way as in T g (λ, n) except that we replace all ≤ with < in Definition 3.2 since the elements in [k] ′ are of odd degree. Then Lemma 3.4 also holds for (T, S) ∈ T g k (a) × T g k (a ′ ) (a ≥ a ′ ) such that T ≺ S, where ≤ is also replaced with <. This implies that there exists a strictly decreasing subword T tail ℓ(λ) (1) · · · T tail 1 (1), and hence L(T) ≥ ℓ(λ). On the other hand, we have k ≥ L(T) since the entries of T are from [k] ′ . Let us define T tail and T body be defined in the same way as in (4.2). Lemma 6.4. If L(T) ≤ n 2 , then we have (1) T tail ∈ SST [k] ′ (λ ′ ), and T body ∈ SST [k] ′ (µ π ) for some µ ∈ P g ∨ , Proof. The proof is almost parallel to that of Lemma 4.3. Now, we have the following characterization of T g k (λ, n) as an l k -crystal in a stable range, which is the main result in this section. Theorem 6.5. Let λ ∈ P(G n ) k be given. If n ≥ 2k, then we have an isomorphism of l k -crystals which maps T to (T tail ) ′ ⊗ (T body ) ′ . Proof. Let B denote the right-hand side of the above isomorphism. Since 2k ≤ n, we have L(T) ≤ k ≤ n 2 for all T ∈ T g k (λ, n). By Lemma 6.4, the map T → (T tail ) ′ ⊗ (T body ) ′ is a well-defined morphism of l k -crystals from T g k (λ, n) to B. The bijectiveness of the map can be proved by similar arguments as in Theorem 4.5.	2017-12-31T12:05:59.000Z	2015-12-07T00:00:00.000	{ "year": 2015, "sha1": "7c3fe511976f52475d797fc192c98482f2b1782e", "oa_license": "publisher-specific, author manuscript", "oa_url": "https://doi.org/10.1090/tran/7104", "oa_status": "HYBRID", "pdf_src": "Arxiv", "pdf_hash": "7c3fe511976f52475d797fc192c98482f2b1782e", "s2fieldsofstudy": [ "Mathematics" ], "extfieldsofstudy": [ "Mathematics" ] }
252060766	pes2o/s2orc	v3-fos-license	The Forms, Channels and Conditions of Regional Agricultural Carbon Emission Reduction Interaction: A Provincial Perspective in China Agricultural emission reduction is a key objective associated with sustainable agricultural development and a meaningful way to slow down global warming. Based on the comprehensive estimation of agricultural carbon emissions, this study applied the traditional spatial Durbin model (SDM) to analyze the type of regional emission reduction interaction and explore whether it is a direct or an indirect interaction caused by technology spillovers. Moreover, geographic, economic, and technical weights were used to discuss the channels of emission reduction interactions. The partitioned spatial Durbin model was applied to explore the realization conditions of regional emission reduction interactions. We found that: (1) comprehensive emission reduction interactions were identified in various regions of China, including direct and indirect interactions, in which geographic and technical channels were the major pathways for direct and indirect emission reduction interactions, respectively; (2) regions with similar economic development levels are more likely to have direct interactions, whereas regions with low technical levels are more willing to follow the high-tech regions, and the benchmarking effect is noticeable; (3) emission reduction results promoted by economic cooperation may be offset by vicious economic competition between regions, and more emission reduction intervention measures should be given to regions with high economic development levels; (4) to achieve better technological cooperation, regions must have similar technology absorption capabilities and should provide full play to the driving force of technical benchmarks. Introduction The increase in carbon emissions and consequent global warming threaten human survival [1]. Carbon concentration in the atmosphere is increasing at an unprecedented rate [2], leading to severe and irreversible consequences in the climate system [3,4]. Therefore, carbon reduction has gathered the focus of global attention. Agriculture is the primary source of global carbon emissions [5,6]. Studies have revealed that the worldwide food systems are responsible for more than one-third of the global carbon emissions, with approximately two-thirds of the food system emissions originating from the agricultural sector and methane from livestock production and rice cultivation accounting for approximately 35% of food system carbon emissions [7]. Furthermore, agricultural carbon emissions increase by approximately 1% annually [8,9]. Therefore, conducting sustainable production is not only a huge challenge in the agricultural sector, but also an inevitable choice for agricultural development [10]. As a leading agricultural and a major emission country [5,11], China has always considered regional cooperation a crucial way to efficiently decrease emissions. Since the joint prevention and control mechanism of air pollution was proposed in 2010, various 2 of 22 regions have started cooperating in decreasing pollution. In 2015, it was thus proposed to rely on regional integration to attain green and low-carbon development among regions, as well as further deepen regional coordination and cooperation to decrease emissions. In 2020, the dual-carbon goal of "carbon neutrality and carbon peaking" also highlighted the establishment of a regionally coordinated carbon reduction framework to create a synergy of regional emission reductions. Driven by the national emission reduction policy, China accomplished its carbon intensity reduction target ahead of schedule and exceeded in 2018, with rapid momentum of low-carbon development. At the same time, carbon emission reduction technologies are also developing in an orderly manner. On the one hand, advances in new energy technologies and material technologies have greatly increased the proportion of clean energy represented by nuclear energy. On the other hand, negative emission technology has developed rapidly, and there are about 40 CCUS (Carbon Capture, Utilization, and Storage) demonstration projects with a capture capacity of 3 million tons per year. This study aims to demonstrate China's experience in regionally coordinated emission reductions and provide a reference for other countries to reduce carbon emissions. Many studies in various countries and regions have supported a significant spatial correlation of carbon emissions [12][13][14], indicating that reliance on the unilateral actions of individual regions is impossible and, thus, cooperative regionally coordinated emission reduction actions are vital [15]. Moreover, many studies have attributed regional correlations between carbon emissions to economic, technological, and policy associations [16,17], as well as similarities in the energy consumption behavior of micro-subjects and their imitation of environmental behavior [18,19]. Alternatively, regional correlations are attributable to the variation in cross-regional output resulting from the changes in the final demand [20,21]. Some studies investigated the spatial correlation of agricultural carbon emissions to determine whether total agricultural emissions, emission intensity, emission efficiency, or net emissions have spatial spillover effects. The status of the agricultural economy, production structure, technology innovation, labor force, and urbanization affect the spatial correlation of agricultural carbon emissions [22,23], suggesting that "technology spillover" can benefit more regions [24]. The spatial correlation of carbon emissions renders the interaction of regional emission reduction a crucial way to enhance the efficiency of emission reduction. Some studies examined the interaction of carbon emission reduction between countries and regions of the aspects of policy coordination, technology coordination, and enterprise coordination. Regarding policy coordination, establishing emission reduction targets beyond national and domestic regions is the basis for implementing emission reduction cooperation [25], and coordinated policies are more conducive to promoting carbon emission reduction and the development of renewable energy at lower costs than single policies [26,27]. Zhou [28] proposed establishing cross-regional environmental protection policy, thereby breaking the administrative boundary of pollution control and promoting regional coordinated emission reduction. Luqman [29] discussed about improving the implementation effect of the CDM (Clean Development Mechanism) from the standpoint of the dynamic cooperative game, reporting that the introduction of the Shapley value cost allocation scheme could improve international cooperation in carbon emission reduction. Tapia [30] claimed that the carbon trading policy plays a limited role in promoting emission reduction cooperation, and the restriction threshold for carbon trading in developing countries can increase the effect of the policy in promoting regional coordinated emission reduction [31]. Regarding technology coordination, most studies agreed that the development of critical technologies to manage global warming is important to effectively mitigate climatic hazards [32,33]. The spatial network connection of low-carbon innovative technologies provides an opportunity to build a cross-regional synergy mechanism and green innovation development [34], and cross-industry technology research and development (R&D) can effectively enhance the efficiency of carbon emission reduction [21]. Nevertheless, regional coordination of low-carbon technology innovation needs the cooperation of policies and industries [35], and technological collaboration can only improve global collaborative emissions reductions under mean or pessimistic assumptions about the development of key low-carbon technologies and when damage is severe [36]. Regarding enterprise coordination, Wang [37] and Wang [38] claimed that cooperative carbon emission reduction strategy has more advantages than the independent carbon emission reduction approach. This is because enterprises can rationally allocate emission reduction investments, further rationalizing the emission reduction structure of the supply chain. Hau [39] highlighted that external technology R&D cooperation can effectively break through the limitations of internal resources and capabilities of SMEs, thereby exerting a positive impact on carbon emission reduction and energy saving. Mao [40] reported that optimal cooperation in emission reduction can be attained by entering into a revenuesharing agreement between manufacturers and service providers. Generally, coordinated emission reduction between countries and domestic regions has been recognized as a crucial way to decrease emissions. In addition, studies have discussed specific methods of cooperative emission reduction from the aspects of policy coordination, technology coordination, and enterprise coordination. However, two shortcomings persist. First, although studies have examined the spatial correlation of carbon emissions, they have not deeply examined the reasons for the regional coordination of carbon emission reduction from a theoretical standpoint and, thus, cannot summarize the possible strategies in regional coordinated emission reduction. Second, previous studies only proposed the framework of coordinated emission reduction between countries or domestic regions, or investigated the cooperation mechanism from the standpoint of micro-enterprises, but did not deeply analyze the strategic choice of regional coordinated emission reduction-is it the direct interaction of emission reduction behavior or the indirect spillovers of emission reduction technologies? Besides this, these studies failed to answer the best channels and possible conditions for inter-regional emission reduction coordination; thus, these cannot provide practical suggestions for regional emission reduction coordination. Using 2008-2018 panel data from China, this study not only discussed the interactive strategies of regional direct emission reduction (imitation or opposition) and indirect emission reduction (technology radiation and technology learning), but also examined the conditions for coordinated emission reduction from the dimensions of economy, industry, human capital, and technological R&D capabilities. The findings can provide references for effectively promoting regional cooperation in reducing emissions, attaining carbon peaking and carbon neutrality, and eventually decelerating global warming. Theoretical Analysis Regionally coordinated emission reduction can be performed in two ways: direct and indirect emission reduction interactions ( Figure 1). On the one hand, there is a mutual alignment of emission reduction behaviors between regions, which leads to "emission reduction imitation" or "emission reduction opposition," that is, direct emission reduction interaction. Due to the existence of the group effect, the carbon emission decision of a region is not only affected by its own agricultural economic development, policy environment, emission reduction potential, and other factors, but also affected by the emission decision of neighboring regions. As a result, the carbon emission behavior of each region presents a certain law. The geographically neighboring regions face similar economic development policies and environmental regulatory measures. No one wants to be a "poor student" in environmental assessment, and all want to be a "good student" in economic assessment [41]. Therefore, when the emission level of a certain region decreases, the other regions also have stricter emission levels, presenting the emission reduction imitation. When a certain region relaxes its emission level and focuses on economic development, the rest of the region also relax its own emissions. There is often fierce economic competition between regions with similar economic development levels, especially regions with higher economic development levels [42]. In order to compete for or maintain their economic status, they always adjust their strategies according to the actions of their opponents. When opponents slow down their economic development and reduce emissions, the region takes this opportunity to vigorously develop its economy, thus presenting the emission reduction opposition. China takes technological innovation as its development goal, and technological development is also an important direction of emission reduction. Therefore, in order to be consistent with national goals, regions with high R&D capabilities are used as benchmarks, and other regions focus on emission reduction imitation. tween regions. Due to the convenient transportation, the flow cost of production factors such as human capital is lower in the neighboring regions of geographical distance, resulting in knowledge spillover. In regions with close economic relations, technology is diffused through industrial cooperation [43,44] , and in regions with high economic and technological levels, its strong radiation force promotes the diffusion of resources, technology, experience, etc. to other regions, resulting in a "trickle-down effect" [45][46][47], while underdeveloped regions take the initiative to learn advanced technologies to improve carbon productivity. The existence of technological gaps between regions affects the diffusion and absorption of technology, and the effect of technology diffusion and absorption between regions with smaller technological R&D capabilities is better [48]. Agricultural Carbon Emission Agricultural carbon emission sources include five categories: (1) CO2 produced by energy consumption; (2) CO2 produced by farmland utilization; (3) CH4 produced by growing rice and N2O produced from other crops; (4) CH4 and N2O produced by ruminant feeding; and (5) CO2, CH4, and N2O produced by straw burning. The measurement framework of agricultural carbon emission is shown in Figure 2. Indirect emission reduction interaction is a relatively smart and continuous emission reduction interaction. Regions analyze the reasons for emission reduction in other regions and then adjust their behavior to promote emission reduction. Technology radiation and technology learning are the main indirect emission reduction interactive strategies between regions. Due to the convenient transportation, the flow cost of production factors such as human capital is lower in the neighboring regions of geographical distance, resulting in knowledge spillover. In regions with close economic relations, technology is diffused through industrial cooperation [43,44] , and in regions with high economic and technological levels, its strong radiation force promotes the diffusion of resources, technology, experience, etc. to other regions, resulting in a "trickle-down effect" [45][46][47], while underdeveloped regions take the initiative to learn advanced technologies to improve carbon productivity. The existence of technological gaps between regions affects the diffusion and absorption of technology, and the effect of technology diffusion and absorption between regions with smaller technological R&D capabilities is better [48]. Agricultural Carbon Emission Agricultural carbon emission sources include five categories: (1) CO 2 produced by energy consumption; (2) CO 2 produced by farmland utilization; (3) CH 4 produced by growing rice and N 2 O produced from other crops; (4) CH 4 and N 2 O produced by ruminant feeding; and (5) CO 2 , CH 4 , and N 2 O produced by straw burning. The measurement framework of agricultural carbon emission is shown in Figure 2 The emissions of each category can be calculated as follows: where is the total emissions of a specific category; is the emissions of source j belonging to this category; and and represent the activity data and emission factor of source j, respectively. The emission factors can be found in Liu [49], Min [50], Tian [51], and Yao [52]. The GHG effects caused by 1 t of N2O and 1 t of CH4 are equivalent to those caused by 298 t of CO2 (81.2727 t C) and 25 t of CO2 (6.8182 t C), respectively [53] , upon conversion into carbon equivalents. Activity data is shown in Table 1. Spatial Correlation Test Moran's I was used to verify the spatial correlation of agricultural carbon emissions. The global Moran's I can be calculated as follows: The emissions of each category can be calculated as follows: where E i is the total emissions of a specific category; E j is the emissions of source j belonging to this category; and e j and f j represent the activity data and emission factor of source j, respectively. The emission factors can be found in Liu [49], Min [50], Tian [51], and Yao [52]. The GHG effects caused by 1 t of N 2 O and 1 t of CH 4 are equivalent to those caused by 298 t of CO 2 (81.2727 t C) and 25 t of CO 2 (6.8182 t C), respectively [53] , upon conversion into carbon equivalents. Activity data is shown in Table 1. Spatial Correlation Test Moran's I was used to verify the spatial correlation of agricultural carbon emissions. The global Moran's I can be calculated as follows: 6 of 22 The local Moran's I can be calculated as follows: where I and I i are global and local Moran's I, respectively; y i and y j are the total agricultural carbon emissions of provinces i and j, respectively; y is the average carbon emission; ω ij is the element of row i and column j of the spatial weight matrix; n is the number of provinces; and s 2 is the variance of the total agricultural carbon emissions. According to the local agglomeration characteristics of variables, the regions were divided into four categories: high-high carbon emission agglomeration (H-H), low-low carbon emission agglomeration (L-L), high-low carbon emission agglomeration (H-L), and low-high carbon emission agglomeration (L-H). Forms and Channels of Regionally Coordinated Emission Reduction-Classical SDM To investigate the form of regionally coordinated emission reduction, the classical SDM can be used: where AEI nt represents the intensity of the agricultural carbon emissions of the 30 provinces for ten years and PI nt represents agricultural technology innovations. As agricultural patents directly affect agriculture [54] and patent data have strong time continuity [55], the strength of agricultural patent authorization was used to measure agricultural technological innovation. x nt represents control variables, w represents spatial weight matrix, ε nt represents the random error term, µ n represents individual-fixed effects, and υ t represents time-fixed effects. ρ represents the response coefficient of the emission intensity of a province to the emission intensity of other provinces and θ pi represents the response coefficient of emission intensity of a province to the technology innovations of other provinces. If ρ is significant, it implies that there is a direct interaction between regional emission reductions. If θ pi is significant, it indicates that there is an indirect interaction in regional emission reductions. To study the realization channels of the interaction of agricultural emission reduction between regions, geographical (w ijd ), economic (w ije ), and technical difference (w ijt ) weights were selected as follows: where d ij is the spherical distance between the provinces i and j; GDP i and GDP j are the agricultural added value of the provinces i and j, respectively; Tec i and Tec j are the total amount of agricultural patents granted by the provinces i and j, respectively. Previous studies considered that the levels of agricultural economic development [56], the urbanization process [57], government environmental supervision [58], and urban-rural income gap [59] affect agricultural carbon emissions. Therefore, the variables used in the traditional SDM are defined in Table 2. Condition of Regional Direct Emission Reduction Interaction-Partitioned SDM for Agricultural Carbon Emission Intensity Facing agricultural carbon emission reduction behaviors in other regions, the responses of different regions vary [60]. Different levels of economic development in various regions prompt different agricultural economic development policies and environmental regulatory measures by the region. Therefore, local governments adopt imitation or oppositive strategies in agricultural emission reduction. Herein, the partitioned SDM for agricultural carbon emission intensity (AEI) is introduced to analyze realization conditions of the direct emission reduction interaction between regions, whether it is more likely to occur in regions with similar levels of agricultural economic development or if it can also occur in regions with significant differences in agricultural economic development levels, and also to clarify whether it is an imitation or an oppositive strategy. The weight matrix was divided according to agricultural value added per capita (H: regions of higher than the national average; L: regions of lower than the national average). The model can be provided as follows: where ρ HH ,ρ HL ,ρ LH , and ρ LL , represent the interaction of emission reduction strategies among the H-H (high-high agricultural value added per capita agglomeration), H-L (high-low agricultural value added per capita agglomeration), L-H (low-high agricultural value added per capita), and L-L (low-low agricultural value added per capita) regions, respectively. The partitioned weight matrix can be expressed as follows: 3.5. Condition of Regional Indirect Emission Reduction Interaction-Partitioned SDM for Agricultural Patent Intensity (PI) The partitioned SDM for agricultural PI was introduced to analyze the conditions for indirect emission reduction between regions. The weight matrix was divided according to the aggregation level of the agricultural industry, human resource level, and R&D level in each region (H: Regions where the aggregation level of the agricultural industry or the human resources level or R&D level is higher than the national average, L: Regions where the aggregation level of the agricultural industry or the human resources level or R&D level is lower than the national average). The model can be expressed as follows: where θ HH and θ LL represent the impact of technology spillovers on emissions reductions between regions with similar levels of the agricultural industry aggregation, human resources, R&D; and θ LH and θ HL represent the impact of technology spillovers on emissions reductions between regions with gaps in the level of the agricultural industry aggregation, human resources level, and R&D level. A negative coefficient implies that technology spillovers can benefit more regions, bringing "positive effects". In contrast, a positive coefficient implies that technology spillovers cannot benefit other regions and cause "negative effects". Model Selection Since this paper focuses on the interaction of regional emission reduction actions, a spatial econometric model is used. According to the general to special modeling ideas, starting from the SDM, the LR and LM tests are used to judge whether it can be simplified into the spatial lag model (SLM) and the spatial error model (SEM) [61]. From Table 3, the LR-lag and LR-error exhibit significance at 1% and 10% levels, respectively, which implies that the SDM performed better than SLM or SEM. The p-values of the LM-lag (robust) and LM-error (robust) tests are 0, indicating that the SDM could not be reduced to the SLM or the SEM. Therefore, the SDM is considered appropriate. Data Sources In this research, the primary data used, which spanned the 2008-2018 period, corresponded to 30 provinces of China. Hong Kong, Macao, Taiwan, and Tibet were excluded because of missing data. The activity data required for agricultural carbon emission estimation were obtained from the China Energy Statistics Yearbook and the China Rural Statistical Yearbook. The variables for the establishment of the SDM model were obtained from the China Rural Statistical Yearbook, the China Environmental Pollution Statistics Yearbook, and China Patent Database. In 2008, China promulgated the "2008 China Energy Conservation and Emission Reduction" report. The issue of energy conservation and emission reduction was widely publicized, and national action began. Therefore, the data began in 2008. After more than ten years of hard work, China accomplished its carbon intensity reduction target ahead of schedule and exceeded in 2018, so the study period from 2008 to 2018 is more representative for analyzing the effect of regional coordinated emission reduction. Analysis of Agricultural Carbon Emissions and Agricultural Technology Innovations The agricultural carbon emission intensity exhibited a fluctuating downward trend, From the development stage, carbon emission intensity and patent intensity could be divided into two stages. First, during 2008-2011, the carbon emission intensity declined rapidly, and patent intensity increased slowly, with mean growth rates of −9.65% and 15.44%, respectively. During this period, under the dual influence of China's "Eleventh Five-Year Plan" agricultural energy conservation and emission reduction targets and the commitment to reducing carbon intensity, agriculture actively enhanced the use efficiency of energy, chemical fertilizers, pesticides, and other input factors. The average annual growth rate of total carbon emissions is just 2.1%. Moreover, the proposal of the modern agricultural development strategy brought agriculture into a period of rapid development, the average annual growth rate of added value is as high as 12.8%, and the agriculture carbon intensity has declined rapidly. In the second stage (2012-2018), carbon emission intensity declined slowly, and the patent intensity increased rapidly, with mean growth rates of −3.38% and 22.45%, respectively. During this period, China's economic growth slowed down. Augmenting the quality of economic development and the level of agricultural science and technology became the focus of development. The growth rate of agriculture declined, and the average annual growth rate declined to 5.4%, while the total amount of agricultural carbon emissions continued to increase slowly. The contribution of scientific and technological progress in the agricultural economy has increased rapidly. The contribution rate of scientific and technological progress in the agricultural economy in 2018 increased by 6 percentage points compared with 2012. Hence, agricultural carbon intensity decreased slowly, and the agricultural patent intensity increased rapidly. Figure 4a shows that the emission intensities in northeast, northwest, and middle From the development stage, carbon emission intensity and patent intensity could be divided into two stages. First, during 2008-2011, the carbon emission intensity declined rapidly, and patent intensity increased slowly, with mean growth rates of −9.65% and 15.44%, respectively. During this period, under the dual influence of China's "Eleventh Five-Year Plan" agricultural energy conservation and emission reduction targets and the commitment to reducing carbon intensity, agriculture actively enhanced the use efficiency of energy, chemical fertilizers, pesticides, and other input factors. The average annual growth rate of total carbon emissions is just 2.1%. Moreover, the proposal of the modern agricultural development strategy brought agriculture into a period of rapid development, the average annual growth rate of added value is as high as 12.8%, and the agriculture carbon intensity has declined rapidly. In the second stage (2012-2018), carbon emission intensity declined slowly, and the patent intensity increased rapidly, with mean growth rates of −3.38% and 22.45%, respectively. During this period, China's economic growth slowed down. Augmenting the quality of economic development and the level of agricultural science and technology became the focus of development. The growth rate of agriculture declined, and the average annual growth rate declined to 5.4%, while the total amount of agricultural carbon emissions continued to increase slowly. The contribution of scientific and technological progress in the agricultural economy has increased rapidly. The contribution rate of scientific and technological progress in the agricultural economy in 2018 increased by 6 percentage points compared with 2012. Hence, agricultural carbon intensity decreased slowly, and the agricultural patent intensity increased rapidly. Figure 4a shows that the emission intensities in northeast, northwest, and middle reaches of the Yellow River were relatively high. In contrast, those of the eastern and southern coasts were relatively low, exhibiting a characteristic decreasing trend from north to south and west to east. Nonetheless, the regional differences in emission intensity gradually narrowed. Figure 4b shows the regional differences in agricultural technology innovation levels were expanding. The technological innovation level of Beijing and Tianjin on the northern coast and Shanghai on the eastern coast was considerably higher than that of other provinces. The technological innovation exhibited a decreasing trend from south to north and from east to west. The level of agricultural technology innovation gradually improved in all provinces. Table 4 shows that the global Moran's I index of agricultural carbon emission intensity and technology innovation were positive within 2008-2018 and significant at a 99% confidence level. These results indicate that the spatial distribution of emission intensity and technological innovation was not random but crossed regional restrictions and exhibited significant spatial agglomeration characteristics. The global Moran's I index of agricultural carbon emission intensity exhibited a decreasing and then increasing trend, with a mean value of 0.259. Moreover, the global Moran's I index of technological innovation was relatively stable, fluctuating from 0.260 to 0.319. Note: *, and indicate significance at the 1% and 5% levels, respectively. Spatial Correlation Test As shown in Figure 5a,b, agricultural carbon emission intensity formed two clustering categories in the regional spatial distribution. The number of provinces with hotspot clusters (H-H) and coldspot clusters (L-L) increased, with the hotspot clusters of AEI being located in the northwest and middle reaches of the Yellow River. Some coastal provinces, such as Fujian, Guangdong, Guangxi, and Hainan, exhibited coldspot clustering. Compared with clusters formed by agricultural carbon emission intensity, the scale of PI clusters was substantially smaller (Figure 5c,d). Comparing data from 2008 and 2018, the number of provinces with hotspot clusters decreased. In 2008, two hotspot clusters were formed, one in Beijing and Tianjin, and another in Shanghai and Zhejiang. In 2018, only one hotspot cluster remained. Table 4 shows that the global Moran's I index of agricultural carbon emission intensity and technology innovation were positive within 2008-2018 and significant at a 99% confidence level. These results indicate that the spatial distribution of emission intensity and technological innovation was not random but crossed regional restrictions and exhibited significant spatial agglomeration characteristics. The global Moran's I index of agricultural carbon emission intensity exhibited a decreasing and then increasing trend, with a mean value of 0.259. Moreover, the global Moran's I index of technological innovation was relatively stable, fluctuating from 0.260 to 0.319. Note: *, and indicate significance at the 1% and 5% levels, respectively. Spatial Correlation Test As shown in Figure 5a,b, agricultural carbon emission intensity formed two clustering categories in the regional spatial distribution. The number of provinces with hotspot clusters (H-H) and coldspot clusters (L-L) increased, with the hotspot clusters of AEI being located in the northwest and middle reaches of the Yellow River. Some coastal provinces, such as Fujian, Guangdong, Guangxi, and Hainan, exhibited coldspot clustering. Compared with clusters formed by agricultural carbon emission intensity, the scale of PI clusters was substantially smaller (Figure 5c Choice of Regional Agricultural Coordinated Emission Reduction Strategies To test which coordinated emission reduction strategy was active in each province, the results of SDM, SAR, SEM, and ordinary panel model (OPM) were evaluated. Table 5 shows that the SDM, SAR, and SEM coefficients were more significant than the OPM. The spatial econometric model extracted the influence of the independent variable on the dependent variable and the spatial effect of the independent and dependent variables; therefore, the spatial model was better than the OPM. Moreover, the SDM comprehensively analyzed the interaction of emission reduction strategies between regions, exhibiting significant direct and indirect emission reduction interactions between regions. Therefore, the SDM is more suitable. Choice of Regional Agricultural Coordinated Emission Reduction Strategies To test which coordinated emission reduction strategy was active in each province, the results of SDM, SAR, SEM, and ordinary panel model (OPM) were evaluated. Table 5 shows that the SDM, SAR, and SEM coefficients were more significant than the OPM. The spatial econometric model extracted the influence of the independent variable on the dependent variable and the spatial effect of the independent and dependent variables; therefore, the spatial model was better than the OPM. Moreover, the SDM comprehensively analyzed the interaction of emission reduction strategies between regions, exhibiting significant direct and indirect emission reduction interactions between regions. Therefore, the SDM is more suitable. Note: *, , and * indicate significance at the 1%, 5%, and 10% levels, respectively. From the estimated results of the SDM, the spatial autocorrelation parameter "ρ" was significantly positive, indicating a direct emission reduction imitation between regions. A 1% intensity reduction (increase) in the agricultural emissions of a region led to a 0.514% intensity reduction (increase) in the emissions of the surrounding regions. Due to the institutional arrangement combining political centralization and regional economic decentralization in China [62,63], regions face similar agricultural economic development policies and environmental regulatory measures and exhibit "strategy convergence". When the emission of a region increases and the economy develops rapidly, government officials in other regions face tremendous economic assessment pressure. Subsequently, local officials prefer economic growth to increase their chances of promotion. Conversely, when the emissions in a region decrease, government officials in other regions face greater environmental assessment pressure, which leads to the prioritization of environmental protection over economic growth [64]. Agricultural technology innovation exhibited spatial effects at p < 0.05 (θ PI ). If the technological innovation level increased by 1%, the agricultural emission intensity of the surrounding regions were reduced by 0.096%. This shows an indirect strategic interaction between regions and that technology spillovers can benefit more regions. Analysis of the Interaction Channels of Regionally Coordinated Emission Reduction Herein, geographic, economic, and technical weights were used to examine the coordinated emission reduction channels. As shown in Table 6, the coefficients of the spatial lag term (ρ) were all significantly positive. The coefficient was the largest under geographic weight, indicating that the agricultural carbon emission reduction behavior in one region has direct strategic interaction with other regions through three channels, namely, geography, economy, and technology, exhibiting a mimic behavior of emission reduction, with geographic distance being the main channel for strategy imitation. The natural conditions and resource endowments of geographically adjacent regions had high similarities, and the "linkage effect" of carbon emission reduction was noticeable [65][66][67]. Regions with similar levels of economic development face relatively similar economic development policies and environmental regulatory measures given by the state. To become the "top students" in the development of the agricultural economy, local governments observe each other, causing economic competition and emission reduction competition to coexist [42], so that emission reduction behaviors converge. Regions with relatively small technological gaps have two-way exchanges, one-way support, and purchase services centered on technology, rendering the relationship between agricultural carbon emissions closer [41]. The spatial lag coefficients of agricultural technological innovation (θ PI ) is significantly negative under the geographical and technical weights, indicating that the agricultural carbon emission reduction behavior has indirect strategic interactions with other regions through geographical and technological channels. Moreover, industrial and technology clusters in neighboring regions strengthen the geographic spillover of knowledge. The spillover costs control the scope of spillover [68], rendering technology spillover effects more likely to occur between neighboring regions. In addition, because technology spillovers are closely related to regional technology absorption capabilities [69], regions with small technological gaps have similar technology R&D capabilities and technology absorption capabilities. It is easier to achieve emission reduction interactions through technology learning. The spatial lag coefficients of agricultural technological innovation under technological weight was larger, indicating that the technological channel is the primary channel for indirect strategic interaction. Note: *, , and * indicate significance at the 1%, 5%, and 10% levels, respectively. Conditions for Direct Emission Reduction Strategies Interaction As shown in Figure 6, under the geographical weight matrix, ρ HH , ρ HL , ρ LH and ρ LL were all positive at the 1% significance level, indicating that, no matter how the agricultural economy develops, all regions imitated the emission reduction behavior of their surrounding regions. Because regions with low agricultural economic levels face both economic and environmental pressures to prevent becoming laggards, the mimicking behavior of emission reduction among these regions was more prominent. effect" should be fully utilized in the emission reduction interaction to drive more regional emission reductions by setting benchmark regions. Figure 6. Conditions for direct emission reduction strategy interaction. Note: These coefficients are the estimated result of the partition SDM's spatial lag term (ρ). *, and indicate significance at the 1% and 5% levels, respectively. H: regions where per capita agricultural added value is higher than the national average, L: regions where per capita agricultural added value is lower than the national average. Conditions for Indirect Emission Reduction Strategic Interaction As shown in Figure 7, from the perspective of industrial agglomeration conditions, under the geographic and technological channels, the indirect strategy interaction occurred both in regions with similar agricultural industrial agglomeration and in regions with differences in agricultural industrial agglomeration. The effect of industrial cooper- Figure 6. Conditions for direct emission reduction strategy interaction. Note: These coefficients are the estimated result of the partition SDM's spatial lag term (ρ). *, and indicate significance at the 1% and 5% levels, respectively. H: regions where per capita agricultural added value is higher than the national average, L: regions where per capita agricultural added value is lower than the national average. Under the economic weight matrix, only ρ HH and ρ LL were significant, indicating that from the perspective of economic channel, regions with similar levels of agricultural economic development were more likely to have direct emission reduction strategic interactions. ρ HH was significantly negative, indicating that the high economic level regions adopted opposed strategies. This is because in the competition of political performance, when one region focuses on environmental protection, other regions promote economic development, thereby relaxing environmental control [70]. Under the technology difference weight matrix, ρ HH , ρ HL , and ρ LH were all significantly positive, indicating that the imitation strategy was still the mainstay among regions under the technology channel. Furthermore, ρ LH was the largest, indicating that the regions with lower economic development levels had the highest degree of mimicking emission reduction to regions with higher economic development levels. The "benchmarking effect" should be fully utilized in the emission reduction interaction to drive more regional emission reductions by setting benchmark regions. Conditions for Indirect Emission Reduction Strategic Interaction As shown in Figure 7, from the perspective of industrial agglomeration conditions, under the geographic and technological channels, the indirect strategy interaction occurred both in regions with similar agricultural industrial agglomeration and in regions with differences in agricultural industrial agglomeration. The effect of industrial cooperation drives technology sharing, and thus promotes more regions to achieve emission reduction. Under the geographic channel, the spillover effect of agricultural technology innovation in regions with similar industrial agglomeration levels was more prominent (θ HH > θ LL > θ LH > θ HL ). Among them, technology spillovers between regions with high industrial agglomeration levels had the strongest inhibitory effect on agricultural emissions due to the high agricultural industry clusters being primarily concentrated in the central and western regions of China, where agricultural production is relatively large. To achieve green development of agriculture, these regions are more proactive in reducing emissions through technological learning. Under the technology channel, the technology spillover effect of low industry agglomeration regions on high industry agglomeration regions was more robust. The effect of suppressing emissions was the greatest (θ LH was the largest). The main reason is that, under the background of industrial integration, the central and western regions with a high degree of agricultural industry agglomeration have increased cooperation with the eastern regions where the agricultural industry clusters are lower by building cross-regional agricultural industry chains. Thus, through cooperation, they can incorporate advanced technology into agricultural production and improve their level of sustainable agricultural development. From the perspective of human capital conditions, under the three channels, indirect emission reduction strategy interaction was affected by the differences in human capital levels between regions. However, indirect strategic interaction enabled more regions to achieve emission reductions (θ are all negative). Under the geographic channel, both θ HH and θ LL were negative at 1% significance, whereas θ HL and θ LH were not significant, indicating that the difference in the level of human capital in geographically adjacent regions affected the regional sharing of technical emission reduction results driven by the "knowledge spillover effect" with human resources as the carrier. Under the economic and technology channels, θ LH was significant, indicating that the low human capital accumulation region produced knowledge spillovers to the high human capital accumulation region with close economic and technological relations. This is because people always seek better development opportunities and conditions. Under the "Matthew effect," the tendency of people to move to better places is evident. Moreover, technical cooperation was also performed between regions with high human capital, resulting in a strong alliance (θ HH = −0.134). From the perspective of R&D conditions, under the geographic channel, indirect strategic interactions occurred between regions with similar R&D capabilities and regions with large gaps in R&D capabilities (the four coefficients were all significant). The coefficients are all negative, indicating that technology played an active role in reducing emissions. θ HL was greater than θ LH , indicating that technology spillovers from regions with high R&D capabilities to regions with low R&D capabilities had a strong inhibitory effect on agricultural emissions. This is because Beijing, Tianjin, and the eastern coastal regions have provided technical assistance to many central and western regions to jointly increase agricultural productivity, reduce agricultural pollution levels, and create a phenomenon of mutual assistance between the strong and the weak. As the regions with low technological R&D capabilities are the central and western regions, where the agricultural production scale is relatively large, agricultural emissions were also higher. The emission reduction effects of mutual technology spillovers were more pronounced (θ LL was greater than θ HH ). Under the technology channel, only θ HH was significantly negative, indicating that the regional technology absorption capabilities of high R&D capabilities were also relatively similar from the perspective of technological cooperation. Mutual technology spillovers were more likely to occur, thereby presenting a strong cooperation situation. cumulation region produced knowledge spillovers to the high human capital accumulation region with close economic and technological relations. This is because people always seek better development opportunities and conditions. Under the "Matthew effect," the tendency of people to move to better places is evident. Moreover, technical cooperation was also performed between regions with high human capital, resulting in a strong alliance ( = −0.134). From the perspective of R&D conditions, under the geographic channel, indirect strategic interactions occurred between regions with similar R&D capabilities and regions with large gaps in R&D capabilities (the four coefficients were all significant). The coefficients are all negative, indicating that technology played an active role in reducing emissions. was greater than , indicating that technology spillovers from regions with high R&D capabilities to regions with low R&D capabilities had a strong inhibitory effect on agricultural emissions. This is because Beijing, Tianjin, and the eastern coastal regions have provided technical assistance to many central and western regions to jointly increase agricultural productivity, reduce agricultural pollution levels, and create a phenomenon of mutual assistance between the strong and the weak. As the regions with low technological R&D capabilities are the central and western regions, where the agricultural production scale is relatively large, agricultural emissions were also higher. The emission reduction effects of mutual technology spillovers were more pronounced ( was greater than ). Under the technology channel, only was significantly negative, indicating that the regional technology absorption capabilities of high R&D capabilities were also relatively similar from the perspective of technological cooperation. Mutual technology spillovers were more likely to occur, thereby presenting a strong cooperation situation. Conditions for indirect emission reduction interaction. Note: These coefficients are the estimated results of the spatial lag term (θ) of agricultural technological innovation in the partition SDM. *, , and * indicate significance at the 1%, 5%, and 10% levels, respectively. Hi: regions where the aggregation level of the agricultural industry is higher than the national average, Li: regions where the aggregation level of the agricultural industry is lower than the national average. H h : regions where the human resources level is higher than the national average, L h : regions where the human resources level is lower than the national average. Ht: regions where R&D is higher than the national average, Lt: regions where R&D is lower than the national average. Discussion (1) Unlike previous studies focusing on the reasons for the spatial correlation of carbon [12,20,21,23], this study analyzed and summarized the regional emission reduction interaction strategies and found two ways for the interaction of emission reduction between regions in China: (i) direct interaction of emission reduction, in-cluding imitation strategy and opposing strategy, and (ii) technical interaction. From the standpoint of direct interaction, owing to China's relatively strict environmental assessment mechanism, to avoid administrative penalties, regions imitate each other's carbon emission reduction behavior, but for regions with a high level of agricultural economic development, the more similar the level of economic development, the more likely it is to adopt the opposite emission reduction strategy, which differs from positive spatial correlation of carbon emissions found by some scholars [71][72][73][74]. This is because regions with a higher level of agricultural economic development have relatively fierce economic or environmental competition to compete for political performance, either choose the development idea of "economy first, environment second," or choose the development idea of "environment first, economy second," to take the lead in economic assessment or environmental assessment. From the viewpoint of technological interaction, scholars unanimously agreed on the existence of a technological interaction between regions [75][76][77]. Nevertheless, few studies examined the realization conditions of technological interaction. We discussed three conditions of industry, human capital, and R&D capabilities, and deduced three modes of technological interaction. First, "industrial agglomeration leads to technological interaction". Cross-regional industrial agglomeration brings technology-sharing between regions. Geographically adjacent regions are dominated by industrial specialized agglomeration, and regions with similar technological development levels are dominated by industrial synergy agglomeration. Second, "knowledge spillovers lead to technological interaction," which primarily occurs between regions with similar economic or technological levels, and is characterized by the cross-regional flow of human capital, but human capital does not flow from high-level regions to low-level regions. Third, the "technological R&D capability leads to technological interaction," which is manifested as "the strong and the strong cooperating" between regions with high-tech R&D capabilities. The large gap in technological R&D capabilities affects the technology spillover between regions, and the technology threshold effect is apparent. (2) Many studies have discussed the ways of enterprise cooperation and its impact on carbon emission reduction [78][79][80] but the improvement of enterprise cooperation awareness is inseparable from the government's guidance [81]. Apart from this, when the region implements the coordinated joint carbon reduction model, the carbon emission reduction efforts of enterprises can also reach the peak [82], showing that the interaction of carbon emission reduction between regions can send signals to enterprises, and then promote the interaction and cooperation between regional economy, industry, and enterprises. In this study, we focused on exploring what emission reduction interaction strategies have been adopted by various regions in China under the background of regional coordinated emission reduction policies, and used geographic weight, economic weight, and technical weight to comprehensively consider whether regional emission reduction interaction is "vicious interaction" or "benign interaction". Our findings can lay the foundation for promoting the benign interaction between enterprises in the region. For regions that implement the imitation strategy, it is crucial to guide the development of low-carbon technologies of enterprises, drive the low-carbonization of the industry, and establish a "benchmark region for emission reduction". For regions that implement opposing strategies, it is essential to regulate the competition of enterprises, guide the benign interaction between regions, and evade the increase in carbon emissions due to vicious competition. For regions where industrial agglomeration leads to technological interaction, it is essential to promote cooperation between cross-regional enterprises, further promoting technology-sharing and transfer through economy of scale and industrial chain extension. For regions where knowledge spillovers lead to technological interaction, it is essential to guide the wider flow of human capital and promote the sharing of regional emission reduction experience. For regions where technical level leads to technical interaction, it is essential to improve the overall technical R&D ability of the region by augment the technical R&D capabilities of enterprises, thereby decreasing the problems of technical barriers to regional technical interaction. (3) In the field of cooperative emission reduction, unlike most scholars who focused on the interaction of emission reduction between countries, we focused on the interaction of emission reduction between regions. Li [83] pointed out that Belt and Road countries can achieve economic and environmental win-win through international trade, while infrastructure investment and energy cooperation can improve energy efficiency and reduce carbon emissions by promoting advanced technologies and funds transfer [84]. Mina [85] and Shin [86] analyzed the international cooperation of REDD+ projects and found that partnerships are less likely to be created between different organization categories (across-type bridging), but tend more towards cooperation with the same types (within-type bridging). Li [87] emphasized reducing emissions through energy-related aid from high-income countries to low-income countries. Scholars all believed that cooperation is beneficial to emission reduction. Compared with regional cooperation, international cooperation obviously faces more difficulties. Therefore, regional cooperation is more important for a country to achieve emission reduction goals. By studying the emission reduction interaction between regions in China, we found that in order to stimulate emission reduction potential, it is necessary to form emission reduction benchmark regions, to drive adjacent regions to reduce emissions through the "imitation effect," and to promote technology spillovers and technology learning. Spillover should take full advantage of industrial agglomeration and human capital flow, and technology learning should reduce technical barriers. These conclusions provide more comprehensive and feasible recommendations for inter-regional emission reduction synergies in other countries. (4) This study discussed the coordinated strategies for low-carbon emission reduction of Chinese local governments. Currently, China's agriculture is characterized by large-scale, industrialized, and small-scale farmers. Thus, it is not only crucial to examine the implementation path of low-carbon development from a macro-perspective but also perform comprehensive analysis from the farmers' perspective. The better realization of regional agricultural coordinated emission reduction also warrants the cooperation of farmers. To investigate the low-carbon coordination between farmers from a micro-perspective will be the direction of future research. In addition, predicting agricultural carbon emissions under coordinated regional emission reduction, judging whether China's carbon peaking goal can be achieved, and then guiding regions to adjust emission reduction interaction strategies, are also issues worthy of study. Conclusions This study analyzed the forms, channels, and conditions of China's regional emission reduction interactions to extend China's emission reduction experience to other countries. The conclusions are as follows: Overall, relatively comprehensive emission reduction interactions, including direct and indirect interactions caused by technology spillover, were identified in various regions of China, for which the geographic channel was the main pathway for direct emission reduction interactions and the technical channel was the main channel for indirect emission reduction interactions. The differences in economic development levels did not significantly hinder direct emission reduction interactions between regions. The differences in industrial agglomeration levels were not related to indirect emission reduction interactions between regions. In contrast, differences in human capital levels and technological R&D capabilities impacted indirect emission reduction interactions. Finally, the following suggestions are made: (i) Improve the top-level design of emission reduction policies, establish a regional coordinated emission reduction mechanism, and augment emission reduction cooperation. Relying on the coordinated development strategy, enhance the balance of agricultural economic development among regions and prevent the adverse impact of vicious economic competition on carbon emission reduction. In addition, advocate the "rich neighbor" strategy, break down barriers to regional cooperation in emission reduction, and share experience in energy conservation and emission reduction through technical cooperation or financial cooperation. (ii) Establish benchmark regions and take full advantage of the industrial integration strategy to promote technology absorption to its radiation effect on carbon emission reduction. In addition, establish an "economic benchmark" and promote the horizontal integration or vertical integration of industries between benchmark regions and other regions, and then share emission reduction experience and technologies. Besides these, establish "emission reduction benchmark" and use the government's environmental assessment system to guide regions to learn from emission reduction benchmark, thereby stimulating the emission reduction potential of more regions. (iii) Create an excellent technology R&D environment to promote regional technology spillover and absorption. Upgrade the intellectual property protection system, encourage enterprises, universities, and other scientific research entities to carry out technology R&D through preferential policies, such as tax relief, financial subsidies, and financial discounts, and integrate talents, capital, information, and other resources to hasten the promotion and application of technology. Furthermore, regions with low-technology R&D capabilities should increase investment in technology-intensive industries, and make full use of the industrial integration strategy to promote technology absorption, thereby driving emission reductions. (iv) Guide the flow of agricultural technical talents and exerts the "knowledge spillover" effect. Increase government guidance, improve the rate of return of production factors in regions with low human capital through preferential policies, such as taxation, to attract technical talents to flow to regions with low human capital through the "Retain talent through preferential policies" method. Finally, establish a long-term mechanism for talent flow, build a career platform, illustrate the development potential of the region, and attract technical talents to flow to regions with low human capital through the "Retain talent through career development" method.	2022-09-04T15:18:26.299Z	2022-09-01T00:00:00.000	{ "year": 2022, "sha1": "b4ea3248f48163ebd0d7888525a2cea8efed38c9", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/1660-4601/19/17/10905/pdf?version=1662027377", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "eafe4918e50fc2bd2741087fca7d23b3705d40fb", "s2fieldsofstudy": [ "Environmental Science", "Agricultural and Food Sciences", "Economics" ], "extfieldsofstudy": [ "Medicine" ] }
52018426	pes2o/s2orc	v3-fos-license	Sol-Gel Synthesis of Silicon-Doped Lithium Manganese Oxide with Enhanced Reversible Capacity and Cycling Stability A series of silicon-doped lithium manganese oxides were obtained via a sol-gel process. XRD characterization results indicate that the silicon-doped samples retain the spinel structure of LiMn2O4. Electrochemical tests show that introducing silicon ions into the spinel structure can have a great effect on reversible capacity and cycling stability. When cycled at 0.5 C, the optimal Si-doped LiMn2O4 can exhibit a pretty high initial capacity of 140.8 mAh g−1 with excellent retention of 91.1% after 100 cycles, which is higher than that of the LiMn2O4, LiMn1.975Si0.025O4, and LiMn1.925Si0.075O4 samples. Moreover, the optimal Si-doped LiMn2O4 can exhibit 88.3 mAh g−1 with satisfactory cycling performance at 10 C. These satisfactory results are mainly contributed by the more regular and increased MnO6 octahedra and even size distribution in the silicon-doped samples obtained by sol-gel technology. Introduction As green energy, the application of lithium-ion batteries has been extended to various fields in our life [1][2][3]. At present, an increasing number of countries are publishing timetables and road maps for forbidding the sale of traditional fuel vehicles. Against this backdrop, the research and development of lithium-ion batteries is receiving more and more attention at home and abroad. It is generally known that the cathode materials greatly influence the electrochemical performance of lithium-ion batteries. Among appropriate cathode materials, LiMn 2 O 4 possesses the distinct advantages of low price, mature production technology and non-pollution characteristic, and is conducive to sustainable development and large-scale application [4,5]. However, it is a great pity that the poor cycling life cannot satisfy the needs of commercial application of LiMn 2 O 4 . This unwelcome fact is related to Jahn-Teller distortion, manganese dissolution and non-uniform particle-size distribution [6][7][8]. Until now, a large number of optimization strategies have been developed to enhance the electrochemical performance of LiMn 2 O 4 [7,[9][10][11][12][13][14]. According to the reported works [7,15], the surface coating treatment can improve the cycling performance to some degree by inhibiting the dissolution of manganese in the electrolyte. Unfortunately, this strategy cannot fundamentally reduce the negative impacts of the Jahn-Teller distortion effect, and also decreases the discharge capacity [12]. These facts indicate that surface modification is not a top-priority optimization method to enhance the comprehensive performance of LiMn 2 O 4 . Therefore, lots of researchers choose to use the doping strategy to avoid the shortcoming of the surface coating treatment [11,16,17]. Yu et al. [17] prepared Li 1+x Mn 2−x O 4 samples by a solid-state sintering method. The obtained Li 1.06 Mn 1.94 O 4 sample presents better cycling performance because the introduction of lithium ions can weaken the ordering of lithium ions and enhance the structure stability. Xu et al. [18] reported the synthesis of LiZn x Mn 2−x O 4 by a solution combustion method. The research results showed that Zn-doping can enhance the cycling performance by reducing the negative impacts of the Jahn-Teller distortion effect. Furthermore, the LiAl x Mn 2−x O 4 samples synthesized by solution combustion technique present better cycling life, which benefits from the effective inhibition of the Jahn-Teller distortion by Al-doping [19]. These analyses indicate that introducing other cations can actually enhance the cycling life of LiMn 2 O 4 . It should be noted, however, that introducing some monovalent cations, bivalent cations or trivalent cations can produce certain negative effects on reversible capacity because of the decrease of Mn 3+ ions, which has previously been confirmed [11,[20][21][22]. Based on all of the above studies, the introduction of some tetravalent cations has been proposed to effectively enhance the electrochemical performance of LiMn 2 O 4 because this modification strategy can avoid the decrease of Mn 3+ ions [23,24]. Herein, we have successfully obtained a series of silicon-doped lithium manganese oxides (LiMn 2−x Si x O 4 , x ≤ 0.10) by sol-gel technology. The effect of silicon doping content on the structures, morphologies and electrochemical properties of the LiMn 2−x Si x O 4 samples obtained by sol-gel technology is discussed. The results indicate that the optimal silicon-doped sample prepared by sol-gel technology shows pretty high reversible capacity and outstanding cycling life. Materials and Methods The silicon-doped lithium manganese oxides (LiMn 2−x Si x O 4 , x ≤ 0.10) were obtained via a sol-gel process with tetraethoxysilane (TEOS, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) as the dopant. Firstly, stoichiometric lithium hydroxide (0.8812 g) and citric acid (4.4129 g) were dissolved in deionized water (20 mL). Under vigorous stirring, the manganese acetate solution (1.5 M) and mixed solution of TEOS (0.2083 g) and ethanol solution (3.0 mL) were added dropwise into the above-mentioned solution at 50 • C. Then, a certain amount of NH 3 ·H 2 O (Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) was added dropwise into the mixed solution to adjust the pH value to 7-8, and the temperature was adjusted to 70 • C. After continuous stirring for a few hours, a reddish-brown sol was formed, which was then dried at 110 • C. The obtained dried gel was sintered at 450 • C for 4 h and then further sintered at 750 • C for 18 h at a heating and cooling speed of 5 • C·min −1 . To investigate the influence of the Si-doping, an undoped LiMn 2 O 4 spinel was prepared under the same conditions. Results and Discussion To investigate the influence of Si-doping content on the crystalline phase of LiMn 2 O 4 , XRD was performed on the obtained samples. As shown in Figure 1, the characteristic diffraction peaks of all the Li Mn2−x Si x O 4 samples obtained by sol-gel technology agree with that of LiMn 2 O 4 (JCPDS No. 35-0782), suggesting the silicon-doped lithium manganese oxides obtained by sol-gel technology possess the cubic spinel structure, with lithium and manganese ions located at tetrahedral sites (8a) and octahedral sites (16d), respectively [25]. In addition, the (220) characteristic diffraction peak will be observed if the tetrahedral sites are occupied by other cations [26]. However, note that the (220) peak does not appear in the XRD patterns of the silicon-doped LiMn 2 O 4 samples, indicating the substitution of silicon ions for manganese ions. As the silicon doping content increases, the lattice parameter of these samples gradually increases. According to the reported results [24], the silicon-doped spinel presents longer Mn−O bond length and larger MnO 6 octahedra. Moreover, the O−Mn−O angle in the Si-doped spinel presents values closer to 90. These results suggest that introducing some silicon ions leads to the more regular and increased MnO 6 octahedra, which could explain the increase of lattice parameter and cell volume. In addition, Si-doping showed a great influence on the (400) FWHM value and the (311)/(400) intensity ratio. Among all the silicon-doped spinels, the LiMn 1.95 Si 0.05 O 4 sample shows a smaller (400) FWHM value and (311)/(400) intensity ratio than the other samples, which is consistent with the results of previous research [27], suggesting higher crystallinity and longer cycling life. The SEM images of the LiMn 2−x Si x O 4 samples are shown in Figure 2. The undoped LiMn 2 O 4 particles shown in Figure 2a present an uneven size distribution. The corresponding range of particle size is from 0.1 to 1.6 µm. For the Si-doped LiMn 2 O 4 , the introduction of some silicon ions can optimize the mean diameter and size distribution. When the silicon doping content increases, the mean diameter of the LiMn 2−x Si x O 4 (0.025 ≤ x ≤ 0.10) has a decreasing tendency, which may be interpreted as the nucleation rate of silicon-doped samples exceeding the growth of particles with the silicon doping [5,28]. In particular, the LiMn 1.95 Si 0.05 O 4 particles shown in Figure 2c present the most uniform size distribution, which is conducive to the enhancement of cycling life [28,29]. The above-mentioned results suggest that introducing some silicon ions can effectively improve the crystallinity and optimize the size distribution. Figure 3a,b shows the TEM and HRTEM images of the representative LiMn 1.95 Si 0.05 O 4 sample. It can be observed that the growth of sample particles matches the (111) direction, and the lattice fringes of 0.478 nm correspond to the spinel lattice structure [30]. Figure 4. According to these results, we can obtain information regarding the chemical and electronic state. As shown in Figure 4a,c-d, the oxidation states of Li1s, Mn2p and O1s can be inferred from the binding energy peaks, which are consistent with the existing results [31]. it is important to note that the binding energies of Mn2p 3/2 correspond to the Mn 3+ ions (641.7 eV) and Mn 4+ ions (643.1 eV), respectively [32]. However, the Mn2p 3/2 binding energy shown in Figure 4c is at 642.6 eV, suggesting the mixture situation of Mn 3+ and Mn 4+ in the silicon-doped sample obtained by sol-gel technology. Figure 4b presents the XPS spectra of Si2p. We can deduce that the corresponding oxidation state is at 102.1 eV, which is in good agreement with the reported results [24]. .00 V, suggesting that introducing silicon ions did not change the electrochemical redox reaction mechanism, and that all these Si-doped LiMn 2 O 4 sample processes comprise two extraction/insertion processes of lithium ions [33]. Figure 5b presents the cycling life of the LiMn 2−x Si x O 4 (x = 0, 0.025, 0.05, 0.075 and 0.10) samples. The reversible capacity and cycling life of the LiMn 2−x Si x O 4 (x = 0, 0.025, 0.05) samples were remarkably enhanced as the silicon doping content increased, due to the more regular and increased MnO 6 octahedra, which is conducive to the lithium ion diffusion in the electrochemical redox process [24]. However, it should be noted that the introduction of more silicon ions has great negative impact on the reversible capacities of the LiMn 2−x Si x O 4 (x = 0.075, 0.10) samples despite the improvement of cycling life (Figure 5c). These unsatisfying results are principally because introducing more silicon ions can cause a reduction in the tetravalent manganese ions, which is unfavorable to Mn(III)−Mn(IV) interconversion. When the silicon doping content is 0.075 and 0.10, the adverse effect exceeds the positive influence from the more regular and increased MnO 6 octahedra. Therefore, the electrochemical performance the LiMn 2−x Si x O 4 (x = 0.075, 0.10) samples will deteriorate to some extent. In particular, the undoped spinel only delivered 132.7 mAh g −1 with low retention of 62.5% after 100 cycles. These analyses indicate that the introduction of silicon ions dramatically enhances the electrochemical performance of LiMn 2 O 4 . Figure 6a shows the rate performance of the LiMn 2−x Si x O 4 (x = 0, 0.025, 0.05 and 0.075) samples. For all these samples, the increased rate has a great negative impact on the reversible capacity because the high rate seriously interferes with the diffusion process of lithium ions [26]. Among these samples, the The above discussion indicates that the optimal Si-doping amount can produce the best improvement effect on the electrochemical performance on the premise that all the silicon-doped samples involved a small amount of Si 4+ ions. To further explore the rate performance at high rates, the LiMn 2−x Si x O 4 (x = 0, 0.025, 0.05 and 0.075) samples were tested at 10 C, and the corresponding test results are presented in Figure 6b. samples, a lower reversible capacity with worse cycling stability was presented. The above results further confirmed that the best improvement effect was obtained by introducing an optimal amount of Si 4+ ions. Figure 6c,d present the representative discharge curves of the undoped LiMn 2 O 4 and the LiMn 1.95 Si 0.05 O 4 samples at varying rates. As shown here, there are two obvious voltage platforms at 0.2 C and 0.5 C, suggesting the diffusion process of lithium ions [33]. When the rate was further increased, these two potential plateaus gradually show ambiguous boundaries and shift toward lower voltage as the discharge rate increases. This result has a lot to do with the ohmic drop and the polarization effect [9]. Compared with the LiMn 1.95 Si 0.05 O 4 sample, the undoped LiMn 2 O 4 sample showed a lower platform at high rate and a more obvious reduction in capacity. The above analysis indicates that the introduction of some silicon ions can have a positive effect on high rate performance. Figure 7a, the undoped LiMn 2 O 4 possesses two pairs of redox peaks, which correspond to the relevant diffusion process of lithium ions [34]. It is important to note that the redox peak current decreased significantly after 100 cycles, suggesting that the undoped LiMn 2 O 4 sample did not show outstanding cycling performance [35]. Figure 7b presents the results of the LiMn 1.95 Si 0.05 O 4 sample. We can see that there are few significant changes in the peak currents. These results suggest that the introduction of some silicon ions plays an effective role in enhancing lithium ion diffusion. According to the reported results [9,19], the charge transfer resistance (R 2 ) corresponding to the high-frequency semicircle has much to do with cycling life. Therefore, the influence of introducing some silicon ions on the cycling life was studied by a thorough analysis of R 2 values. Table 2 lists the relevant fitting values of R 2 . For the Si-doped spinel, the original R 2 value only reached 61.5 Ω cm 2 and increased to 90.6 Ω cm 2 with a low growth rate of 47.3% after 100 cycles. Compared with the Si-doped spinel, the undoped spinel presents a higher original R 2 value (92.3 Ω·cm 2 ). After 100 cycles, this value could reach up to 302.7 Ω·cm 2 with a very high growth rate of 228.0%. These analyses indicate that introducing some silicon ions can help to decrease the R 2 value, which can promote lithium ion diffusion [18]. Conclusions Silicon-doped lithium manganese oxides were obtained via a sol-gel process. As the optimal Si-doped spinel, the LiSi 0.05 Mn 1.95 O 4 sample possessed a regular surface morphology and an even size distribution. More importantly, it showed much better electrochemical properties than those of the other Si-doped LiMn 2 O 4 samples with a small amount of Si 4+ ions. When cycled at 0.2°C and 0.5°C, the LiMn 1.95 Si 0.05 O 4 sample exhibited 142.5 and 140.8 mAh·g −1 , respectively, which are higher values than those of the LiMn 2 O 4 , LiMn 1.975 Si 0.025 O 4 and LiMn 1.925 Si 0.075 O 4 samples. After 100 cycles, the LiMn 1.95 Si 0.05 O 4 sample could exhibit 128.3 mAh·g −1 with an outstanding retention of 91.1% at 0.5 C. When cycled at 10°C, the initial discharge capacity of the optimal Si-doped LiMn 2 O 4 sample could exhibit 88.3 mAh·g −1 . All of these results suggest that the optimal Si-doping amount can produce the best improvement effect on the electrochemical performance on the premise that all the silicon-doped spinels involved a small amount of Si 4+ ions. Conflicts of Interest: The authors declare no conflict of interest.	2018-08-18T21:15:58.427Z	2018-08-01T00:00:00.000	{ "year": 2018, "sha1": "d426777c75105bf6c2fe45ac9f7b10f7c4a16996", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/1996-1944/11/8/1455/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "d426777c75105bf6c2fe45ac9f7b10f7c4a16996", "s2fieldsofstudy": [ "Materials Science", "Engineering" ], "extfieldsofstudy": [ "Medicine", "Materials Science" ] }
11429017	pes2o/s2orc	v3-fos-license	ASSESSMENT OF SERUM CATECHOLAMINE CONCENTRATIONS IN PATIENTS WITH PHEOCHROMOCYTOMA UNDERGOING VIDEOLAPAROSCOPIC ADRENALECTOMY Introduction: We analyzed the changes in serum catecholamine concentrations, i.e. adrenaline and noradrenaline, in response to surgical stress in patients with pheochromocytoma who undergone videolaparoscopic adrenalectomy. Materials and Methods: Between January 1998 and March 2002, 11 patients underwent 12 videolaparoscopic adrenalectomies. In one case, the adrenalectomy was bilateral. Serum catecholamines were measured at 6 surgical times: T0: control before induction; T1: following the induction, laryngoscopy and intubation sequence; T2: after installing the pneumoperitoneum; T3: during manipulation-exeresis of the pheochromocytoma; T4: following ablation of the pheochromocytoma; T5: in the recovery room following intervention when the patient was extubated and was hemodynamically stable. Results: Mean concentrations of serum noradrenaline were significantly different when the T0 and T2 surgical times were compared (T0: 3161 pg/mL; T2: 40440 pg/mL; p < 0.01), T0 and T3 (T0: 3161 pg/mL; T3: 46021 pg/mL; p < 0.001), T1 and T3 (T1: 5531 pg/mL; T3: 46021 pg/mL; p < 0.01), T2 and T4 (T2: 40440 pg/mL; T4: 10773 pg/mL; p < 0.01) and T3 and T5 (T3: 46021 pg/ mL; T5: 2549 pg/mL; p < 0.001). Mean concentrations of serum adrenaline were significantly different when the T0 and T3 surgical times were compared (T0: 738 pg/mL; T3: 27561 pg/mL; p < 0.01). Conclusion: The pneumoperitoneum significantly increases serum noradrenaline concentrations, manipulation of the adrenal gland significantly increases the serum concentrations of noradrenaline and adrenaline, and the pheochromocytoma ablation significantly decreases serum noradrenaline concentrations. INTRODUCTION Pheochromocytoma is an uncommon and important neoplasia because, despite its rarity, it is associated with catecholamine-induced hypertension, which can be resolved by neoplasia excision.The definitive treatment for pheochromocytoma is surgical ablation of the adrenal gland and/or paragangliomas. Before the 1950s, the peroperative mortality was between 20 and 25% of cases with a preoperative diagnosis of pheochromocytoma and around 50% of cases without a preoperative diagnosis of pheochromocytoma.The lack of proper control for hypertensive crises and cardiac arrhythmias during manipulation and ablation of the pheochromocytoma was responsible for this high mortality.The advances in peroperative control and the preoperative introduction of alpha 1adrenergic blockade have significantly reduced mortality rates (1). The first laparoscopic adrenalectomies were described by Gagner et al. and Higashihara et al. in 1992 (2,3).Studies have shown that videolaparoscopic adrenalectomy offers lower morbidity than open surgery (4,5).The videolaparoscopic technique has become the preferred option for treating adrenal tumors (6), however some doubts remain about the use of videolaparoscopy for management of pheochromocytomas due to cardiovascular risks that are potentially higher when compared to open surgery.Such risks are related to catecholamine release.Factors such as the use of carbon dioxide, the increase in abdominal tension and manipulation of the adrenal gland have been implied in catecholamine release (7). Thus, in order to validate the videolaparoscopic technique in the management of pheochromocytomas, it is important to determine potential changes in serum catecholamine concentrations, as well as the relationship between such changes and higher cardiovascular risk. This study aimed to assess changes in serum catecholamine concentrations, i.e. adrenaline and noradrenaline, in response to surgical stress in patients with pheochromocytoma undergoing videolaparoscopic adrenalectomy. MATERIALS AND METHODS The study was performed between January 1998 and March 2002 in 11 patients undergoing 12 videolaparoscopic adrenalectomies.The study included patients diagnosed with pheochromocytoma, which was confirmed through dosing of serum and urinary catecholamines.Abdominal and pelvic computerized tomography was performed in 8 patients, magnetic resonance imaging in 10 patients and scintigraphy with 131 I-metaiodobenzylguanidine in 10 patients. Blood collections for dosing catecholamines, adrenaline and noradrenaline, by high-pressure liquid chromatography (HPLC) were performed at the following times: T0: control before induction; T1: following the induction, laryngoscopy and intubation sequence; T2: during the creation of pneumoperitoneum; T3: during manipulation-extraction of the pheochromocytoma; T4: following ablation of the pheochromocytoma; T5: in the recovery room following the intervention when the patient was extubated and hemodynamically stable. All patients underwent preoperative cardiovascular assessments, including Doppler echocardiography and 24-hour Holter.Preparation started 15 days before the intervention by associating a alpha 1 blocker (prazosin: alpress LP 5 mg/ day) and a beta 1 blocker (bisoprolol: détensil 10 to 20 mg/day).Three days before the intervention, an intravenous alpha 1 blocker, urapidil (250 mg/ day in continuous perfusion and hourly control of blood pressure), was started as a replacement to oral therapy and maintained until the end of the intervention.Oral pre-anesthetic medication consisted of 5 mg midazolam (short-acting benzodiazepine) associated with 0.5 mg atropine (muscarinic receptor blocker). General anesthesia was standardized as follows: induction with propofol (2-2.5 mg/Kg -1 ) and sufentanil (0.8-1 µg/Kg -1 ) IV for control of blood pressure (BP), orotracheal intubation facilitated by cisatracurium 0.15 mg.Kg -1 ; maintenance with continuous perfusion of sufentanil and cisatracurium and administration of sevoflurane or isoflurane with pure oxygen.Following induction, BP was continuously monitored using an arterial catheter connected to a blood pressure meter (Baxter tm ).Hydration was started with crystalloids 10-15 mL/Kg -1 /h -1 (isotonic saline solution, Ringer solution).Ventilation was adapted in order to maintain the PCO 2 between 35 and 45 mmHg.Urapidil was maintained in a continuous infusion of 10 mg/h -1 until ligation of the adrenal vein.Tension peaks were defined as systolic blood pressure (SBP) over 160 mmHg, and were treated through administering nicardipine (2-4 mg) aiming to maintain SBP between 120-160 mmHg.Episodes of sinus tachycardia, as defined by a heart rate (HR) over 100-120 b/min -1 , were treated by the administration of esmolol 100 mg in order to maintain the HR under 100 b/min -1 .Cardiovascular shock, as defined by SBP lower than 80 mmHg, was treated by administering ephedrine 3-6 mg IV. The adrenalectomies were performed as described by Rocha et al., 2003 (8).Patients were positioned in lateral decubitus opposite the lesion.Four trocars were used -3 10-mm and one 5-mm.The first trocar (10 mm) was introduced by "open" laparoscopy to the lateral margin of the rectus muscle of the abdomen approximately 4 cm cranial to the umbilical scar.The second trocar (10 mm) was placed on the epigastric midline.The third trocar (10 mm) was placed lateral to the first trocar, between the lateral margin of the rectus muscle of the abdomen and the anterior axillary line.The fourth trocar (5 mm) was placed lateral to the third trocar, between the anterior axillary line and the middle axillary line.The pneumoperitoneum was maintained at 12 mmHg.When the left adrenal gland was operated on, the intra-abdominal procedure started with an incision in the parietocolic gutter and dissection of two thirds of the descending colon.The spleen was withdrawn to expose the upper region of the renal cavity.The left renal vein was dissected and the adrenal vein was then identified, dissected and sectioned with metallic clips.Following this, the adrenal gland was dissected on the cleavage plane between the adrenal gland and the kidney.Arteries and occasional small accessory veins were sectioned with clips or after bipolar coagulation around the gland.The adrenalectomy specimen was removed inside an endosac (Endocath 10) through a trocar orifice that was enlarged by 1 cm on each side.The trocar orifices were closed in 2 planes.When the right adrenal gland was operated on, the right liver lobe was withdrawn after sectioning of the triangular ligament.The renal cavity was exposed above the right colic angle.The cava vein was dissected up to the adrenal vein, which was then sectioned with clips.The following surgical times were similar to the left adrenalectomy. The ligation of the adrenal vein was performed early, with minimal previous dissection of the adrenal gland. To statistically analyze the results, Graphpad Prism software was employed, using non-parametric methods: Kruskal-Wallis test and Dunn's multiple comparisons test. The statistical significance value was established at 95 % (p < 0.05). RESULTS No surgical conversion was required.The mean length of intervention was 127 min (75 to 195 min).Blood loss was between 0 and 1000 mL with a mean value of 105 mL. During the creation of the pneumoperitoneum (T2), 6 hypertension peaks were observed, which were associated with sinus tachycardia in 3 cases.Manipulation or exeresis of the adrenal gland (T3) caused 8 hypertension peaks associated with sinus tachycardia in 2 cases. Patient 5 presented atrial and ventricular extrasystoles during these 2 surgical times (T2 and T3) and was treated with intravenous (IV) esmolol. Return to oral diet occurred between the first (D1) and the third (D3), with a mean of 1.9 days.Removal of the drain occurred between D2 and D3.Deambulation was authorized between D1 and D3 with a mean of 2.18 days.Discharge from hospital occurred between D3 and D6 with a mean hospital stay of 3.8 days.There was one damage to the adrenal vein, which was treated during surgery with no COMMENTS Laparoscopy effectively offers advantages for open surgery, that is less severe postoperative pain, early deambulation, reduced hospital stay and prompter return to daily activities (9,5).However, the surgery for management of pheochromocytoma differs from the approach used for other adrenal tumors due to the increased cardiovascular risks during the surgical intervention related to catecholamine release (2,10). This study confirms that creation of pneumoperitoneum and tumor manipulation during videolaparoscopic ablation of pheochromocytoma are accompanied by a significant release of serum catecholamines.This release is probably responsible for hemodynamic disorders such as hypertensive peaks and sinus tachycardia. Insufflation of pneumoperitoneum is associated with an increase in serum catecholamines, either by stimulus to mechanical compression or a change in tumor vascularization (11,12).Additionally, carbon dioxide used to insufflate the pneumoperitoneum can lead to hypercapnia, which would increase sympathetic tonus, thus changing tension levels (13). In the cases evaluated in this study, the creation of the pneumoperitoneum produced an important release of noradrenaline into the blood stream, thus increasing its concentration when compared with preoperative values (p < 0.01).This release was variable and unpredictable between one patient and the other.In this series, pneumoperitoneum increased mean serum noradrenaline concentrations 12.7 times the baseline value and mean serum adrenaline concentrations increased 29.3 times the baseline value.These results are in agreement with the study by Joris et al. (7), who observed an increase in plasma catecholamine concentrations by 7 to 16 times the baseline value following the creation of pneumoperitoneum in patients with pheochromocytoma undergoing videolaparoscopic surgery. Manipulation and ablation of the pheochromocytoma during laparoscopy equally produce an exaggerated release of serum catecholamines (14).In the series described in this study, manipulation and exeresis of pheochromocytoma (T3) evolved with high concentrations of serum noradrenaline and adrenaline.When assessing the mean concentrations of serum noradrenaline (p < 0.001) and adrenaline (p < 0.01), statistical differences were observed between T3 and pre-induction control (T0).In this series, despite early ligation of the adrenal vein, we observed an elevation in catecholamine concentrations during dissection of the adrenal gland. According to some authors, tension variations are less important or equivalent during laparoscopy when compared to laparotomy (11).Joris et al. (15), studied hemodynamic changes relative to pneumoperitoneum with carbon dioxide in 20 healthy patients undergoing elective laparoscopic cholecystectomy.They also assessed the changes in several neurohumoral mediators, which can contribute to hemodynamic changes, such as plasma concentrations of cortisol, catecholamines, vasopressin, renin, endothelin and prostaglandins.Peritoneal insufflation resulted in significant reduction in cardiac output, as well as an increase in blood pressure and systemic and pulmonary vascular resistance.Laparoscopy resulted in a progressive and significant increase in plasma concentrations of cortisol, adrenaline, noradrenaline, renin and vasopressin.Prostaglandins and endothelins showed no significant change.The authors concluded that vasopressin and catecholamines probably measured the increase in systemic vascular resistance during insufflation of pneumoperitoneum with carbon dioxide. Fernandez-Cruz et al. (10), reported that, in the case of pheochromocytomas, videolaparoscopic adrenalectomy is associated with a lesser increase in catecholamine levels in peripheral circulation when compared with laparotomy, and that hypertensive peaks are related to direct manipulation of the adrenal gland.The authors compared 23 videolaparoscopic adrenalectomies (non-functioning adenomas, aldosterone-producing adenoma, Cushing's adenoma and Cushing's disease) using insufflation of pneumoperitoneum with carbon dioxide, with 8 videolaparoscopic adrenalectomies for pheochromocytoma using insufflation of pneumoperitoneum with helium, and with eight 8 adrenalectomies performed by the conventional open approach.They studied the serum changes in catecholamine levels and correlated them with intraoperative cardiovascular disorders in patients with pheochromocytoma.There was no significant difference between videolaparoscopic adrenalectomies due to pheochromocytoma compared to videolaparoscopic adrenalectomies due to other lesions, as regards surgical time, blood loss, hospital stay and return to usual activities.Results for these parameters were unfavorable to the open adrenalectomies group.A major increase in plasma catecholamine levels occurred in patients with pheochromocytoma during tumor manipulation in the videolaparoscopy group (17.4 times for adrenaline and 8.6 times for epinephrine) and in the open surgery group (34.2 times for adrenaline and 13.7 for noradrenaline).Cardiovascular instability was associated with open surgery only. Among the studied cases, in 6 out of 12 adrenalectomies (50%), hypertensive peaks occurred during the creation of pneumoperitoneum (T2) and in 8 out of 12 adrenalectomies (66.6%), hypertensive peaks occurred during manipulation and exeresis of the gland.These hemodynamic changes occurred simultaneously with an increase in catecholamine levels during the creation of pneumoperitoneum and manipulation / exeresis of the gland. Two independent predictive factors for perioperative morbidity are the secreting characteristic of the tumor, and its size (16).An excessive and unpredictable peroperative catecholamine release can lead to a serious clinic condition associated with malignant hypertension, mydriasis, pulmonary edema (17) and even acute heart failure (18). Despite the association between the creation of pneumoperitoneum (T2) and manipulation (T3) of the adrenal gland with an increase in serum rates of noradrenaline and adrenaline, the adrenalectomies were performed with low morbidity.A laparoscopic exeresis of pheochromocytoma can be performed, but it requires proper preoperative preparation and careful perioperative anesthetic surveillance of blood pressure and occasional cardiac arrhythmias.The preoperative preparation is intended to decrease cardiovascular morbidity and includes alpha adrenergic blockade and, if required, beta adrenergic blockade.Hypertensive peaks in pheochromocytoma are related to the stimulation of alpha 1 receptors (19).In the series under study, preoperative control was achieved by previous blockade of alpha 1 receptors by prazosin and urapidil.Beta adrenergic blockade is not systematic and depends on the presence of associated tachycardia (19). Two problems occur during pheochromocytoma surgery, and they require opposite solutions: 1) the catecholamine release during tumor manipulation leads to a risk of paroxysmal hypertension and episodes of sinus tachycardia, which can be effectively treated by associating nicardipine (calcium channel antagonist) and esmolol (selective beta blocker); 2) the significant decrease in catecholamine levels following tumor ablation can, contrarily, cause severe hypotension, which can be worsened by the persistent effects of alpha 1-antagonist drugs that have been introduced during preoperative preparation (19).Thus, the use of alpha 1-antagonists (Urapidil) available for injection with short half-live and short action can be used during the preoperative period, offering a potential solution to both problems (20). In a series with 8 patients, Joris et al. ( 7) used alpha 1-adrenergic blockers for preoperative preparation.During the intervention, they used an infusion of nicardipine (calcium channel blocker) for treating and preventing increases in blood pressure.Six of the 8 patients (75%) showed increased blood pressure higher than 25% of the baseline value during the creation of pneumoperitoneum.The authors reported that hemodynamic changes were easily treated through a continuous infusion of nicardipine associated with a beta blocker.The authors stated that no episode of acute hypotension (blood pressure lower than 60 mmHg) was observed in the 8 patients, probably because they had received vasodilators and/or were normotensive before surgery. In the present series, during 12 adrenalectomies, 5 of the patients (41.6%) had hypotensive episodes and systolic blood pressure lower than 80 mmHg, and were treated by administering bolus ephe-drine and volume reposition with crystalloids.However, no patient in this series presented pressure levels under 60 mmHg.Episodes of hypotension were related to the decrease in serum catecholamine concentrations observed after ablation of the gland.Serum noradrenaline concentrations, following adrenal ablation (T4), significantly decreased (p < 0.01) when compared with serum noradrenaline levels during creation of pneumoperitoneum (T2).When compared at T4 and T2, the decrease in serum adrenaline concentrations was not significant, probably due to the reduced number of study cases and the large variation in serum adrenaline concentration among the patients. This study is in agreement with data from the literature, showing that videolaparoscopic adrenalectomy is feasible for management of pheochromocytomas and presents a low morbidity rate.The creation of pneumoperitoneum and the manipulation of the adrenal gland are related to an increase in serum catecholamines concentrations, and the ablation of the pheochromocytoma is related to a decrease in noradrenaline concentrations.In some cases, the changes in serum catecholamine concentrations correlate with hemodynamic disturbances, which, however, were easily treated. CONCLUSION The barotrauma promoted by installing the pneumoperitoneum with a pressure of 12 mmHg in patients with pheochromocytoma induced a significant increased in serum noradrenaline concentrations.In patients with pheochromocytoma undergoing videolaparoscopic adrenalectomy, surgical stress during manipulation of the adrenal gland promoted significant increases in serum concentrations of noradrenaline and adrenaline.The serum noradrenaline concentrations significantly decreased following ablation of the pheochromocytoma when compared with serum concentrations during creation of pneumoperitoneum and manipulation of adrenal gland. Figure 2 - Figure 2 -Mean serum adrenaline concentration at different surgical times * p < 0.01 comparing T2 to T0. T0: control before induction; T1: following the sequence induction-laryngoscopy-intubation; T2: during creation of pneumoperitoneum; T3: during manipulation-exeresis of pheochromocytoma; T4: following ablation of the pheochromocytoma; T5: in the recovery room following intervention when the patient is extubated and hemodynamically stable.	2017-03-31T05:49:24.534Z	2005-07-01T00:00:00.000	{ "year": 2005, "sha1": "c0b88cebea147355b9369d84b5ce95acc87ed06f", "oa_license": "CCBY", "oa_url": "https://www.scielo.br/j/ibju/a/TDD7JL7zD3RwSdZ8T875RKH/?format=pdf&lang=en", "oa_status": "GOLD", "pdf_src": "ScienceParseMerged", "pdf_hash": "c0b88cebea147355b9369d84b5ce95acc87ed06f", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
532684	pes2o/s2orc	v3-fos-license	Imaging of the Yellow Cameleon 3.6 indicator reveals that elevations in cytosolic Ca2+ follow oscillating increases in growth in root hairs of Arabidopsis. In tip-growing cells, the tip-high Ca(2+) gradient is thought to regulate the activity of components of the growth machinery, including the cytoskeleton, Ca(2+)-dependent regulatory proteins, and the secretory apparatus. In pollen tubes, both the Ca(2+) gradient and cell elongation show oscillatory behavior, reinforcing the link between the two. We report that in growing root hairs of Arabidopsis (Arabidopsis thaliana), an oscillating tip-focused Ca(2+) gradient can be resolved through imaging of a cytosolically expressed Yellow Cameleon 3.6 fluorescence resonance energy transfer-based Ca(2+) sensor. Both elongation of the root hairs and the associated tip-focused Ca(2+) gradient show a similar dynamic character, oscillating with a frequency of 2 to 4 min(-1). Cross-correlation analysis indicates that the Ca(2+) oscillations lag the growth oscillations by 5.3 +/- 0.3 s. However, growth never completely stops, even during the slow cycle of an oscillation, and the concomitant tip Ca(2+) level is always slightly elevated compared with the resting Ca(2+) concentration along the distal shaft, behind the growing tip. Artificially increasing Ca(2+) using the Ca(2+) ionophore A23187 leads to immediate cessation of elongation and thickening of the apical cell wall. In contrast, dissipating the Ca(2+) gradient using either the Ca(2+) channel blocker La(3+) or the Ca(2+) chelator EGTA is accompanied by an increase in the rate of cell expansion and eventual bursting of the root hair tip. These observations are consistent with a model in which the maximal oscillatory increase in cytosolic Ca(2+) is triggered by cell expansion associated with tip growth and plays a role in the subsequent restriction of growth. Introduction Tip growth of cells such as fungal hyphae, algal rhizoids, pollen tubes and root hairs is sustained by targeted secretion of new membrane and wall material to the apical few micrometers of their elongating tips. Turgor is then thought to drive expansion at the cell apex, with the subapical wall resisting these expansive forces. The combination of localized secretion coupled to regulation of wall properties would then lead to the elongated cylindrical morphology of these cells (for review, see Gilroy and Jones, 2000). The tip-focused Ca 2+ gradient, characteristic of tip-growing cells, seems to play an important role in the spatial control of these systems. The cytosolic free Ca 2+ concentration is approximately 100 nM at the base of the polarized cell but rises up to micromolar levels over the apical few micrometers of the expanding tip (reviewed in Bibikova and Gilroy, 2002). This elevated apical Ca 2+ is proposed to provide a spatial determinant for growth by facilitating membrane fusion at the tip and regulating a host of Ca 2+ -dependent proteins required for tip growth. Dissipating the Ca 2+ gradient in pollen tubes, fungal hyphae and root hairs has been shown to disrupt growth (e.g., Clarkson et al., 1988;Miller et al., 1992;Herrmann and Felle, 1995;Wymer et al., 1997), whereas experimentally altering the direction of the gradient leads to redirected growth, with the site of new expansion following where the new Ca 2+ gradient is imposed (e.g. Malho and Trewavas, 1996;Bibikova et al., 1997). Thus, there is strong evidence supporting a regulatory role for the Ca 2+ gradient through imposing spatial control on the site of cell expansion in these tip-growing systems. Indeed, the structure of the apical actin cytoskeleton and its regulatory proteins (villins, gelsolins and actin depolymerizing factors; Smertenko et al., 1998;Tominaga et al., 2000;Allwood et al., 2001;Ketelaar et al., 2003;Fan et al., 2004) are thought to be regulated by the tip high Ca 2+ (Yokota 6 et al., 2005). Similarly, there is evidence that annexins (Blackbourn et al., 1992;Clark et al., 1992;Carroll et al., 1998), phosphoinositide metabolism (Preuss et al., 2006), calmodulin and protein kinases (Moutinho et al., 1998;Yoon et al., 2006) also play roles in sustaining tip growth and are regulated via the cytosolic Ca 2+ gradient. Tip growth in some systems has been shown to oscillate with periods of rapid expansion alternating with slower growth rates. While this has been confirmed for different species of pollen tubes by independent groups (Pierson et al., 1995;Pierson et al, 1996;Messerli and Robinson, 1997;Watahiki et al., 2004;Hwang et al., 2005), it has not been repeated for tip growth in fungal hyphae (Lopez-Franco et al, 1994;Sampson et al. 2003). Oscillating growth in root hairs has only recently been reported (Monshausen et al., 2007) and is under further investigation in this report. One of the most extensively studied tip growing system has been lily pollen tubes. In this system growth oscillations were accompanied by oscillations in the tip focused Ca 2+ gradient (e.g. Holdaway-Clarke et al., 1997;Messerli and Robinson, 1997;Messerli et al., 2000;Watahiki et al., 2004), Interestingly, the periodic increases in the Ca 2+ gradient actually lagged the periodic increases in growth by about 4 s (Messerli et al., 2000). These observations have led to a model for pollen tubes where a mechanically sensitive Ca 2+ channel may be gated via membrane tension during elongation growth. This stretch-activated channel would then support a Ca 2+ increase that follows rather than coincides with maximal cell elongation (Messerli and Robinson, 2007). It has even been suggested that a burst of secretion may precede elongation in Agapanthus pollen (Coelho and Malho, 2006). Lily pollen tubes in culture (reviewed in Messerli et al., 2000) grow 5-7 times faster than 7 equivalent relationship between oscillatory growth and Ca 2+ changes exists for the root hair. Although a stretch-activated Ca 2+ influx channel has been identified in pollen tubes (Dutta and Robinson, 2004), the Ca 2+ channel sustaining the tip focused Ca 2+ gradient in root hairs is thought to be gated by membrane voltage and reactive oxygen species (ROS;Foreman et al., 2003). Thus, whether fluctuations in the root hair apical Ca 2+ gradient occur and are functionally important for regulating cell elongation remain unknown. In this paper we report that root hairs of Arabidopsis exhibit both oscillatory growth and oscillations in the tip-focused Ca 2+ gradient. In addition, we show that the maximum of the Ca 2+ gradient lags the growth maxima by approximately 5 s. Treatments that dissipate the Ca 2+ gradient promote tip expansion and eventual bursting whereas artificially elevating cytosolic Ca 2+ leads to rapid growth arrest. These results indicate that one role for the maximal Ca 2+ levels attained during the oscillatory increase in cytosolic Ca 2+ that accompany root hair growth is actually to limit turgor-driven expansion after each burst of elongation. Oscillations in Cytosolic Ca 2+ Accompany Root Hair Tip Growth To determine the kinetics of the tip focused Ca 2+ gradient in growing root hairs, we used Arabidopsis plants stably transformed with a soluble version of the green fluorescent proteinbased Ca 2+ sensor, yellow cameleon (YC) 3. 6 (Nagai et al., 2004), driven by the CaMV 35S promoter. Expression of this protein had no detectable effect on root hair growth rates, density or general morphology ( Figure 1, and data not shown) indicating it provided an appropriate approach for analysis of root hair growth. We imaged the YC3.6 signal using a Zeiss LSM 510 confocal microscope and found that root hair elongation was sensitive to laser irradiation intensity, with higher levels of irradiation leading to a minor but significant inhibition of elongation rates (reduction from 2.01±0.29 to 1.34±0.26 µm min -1 at 60% laser attenuation n=6, student t-test P<0.001, see Materials and Methods for specifics of the imaging protocol). However, by using lower laser power (90% attenuation) we were able to monitor root hairs for extended periods (>10 min) with frequent sampling (images every 2 s) without significant alteration of root hair growth rate (1.85±0.33 µm min -1 , n=6, student t-test P=0.38) or morphology (Figure 1), allowing us to make measurements of root hair Ca 2+ dynamics during growth. As previously reported (reviewed in Bibikova and Gilroy, 2002), growing root hairs were characterized by a tip-focused Ca 2+ gradient. Invariably, Ca 2+ levels were highest within 1-2 µm of the extreme apex and then rapidly declined with increasing distance from the tip until reaching resting Ca 2+ concentrations approximately 20 µm behind the apex ( Figure 1A-C). While this gradient persisted as long as root hairs continued to grow, our high temporal resolution measurements showed that the magnitude of the gradient oscillated with a frequency of approximately 2-4 peaks per min ( Figure 1B, C). The largest changes in cytoplasmic Ca 2+ occurred at the extreme tip of the root hair whereas Ca 2+ levels in more subapical regions oscillated in phase but with smaller amplitudes ( Figure 1A, B). We were interested in whether these changes in Ca 2+ were associated with alterations in growth rate. Applying high resolution tip tracking software previously used to measure growth of Arabidopsis root hairs (Messerli et al., 1999;Monshausen et al 2007), we were able to confirm that growth rates of YC3.6 expressing root hairs oscillated at the same frequency of 2-4 peaks per min as apical Ca 2+ levels ( Figure 1D) and as in untransformed wild-type root hairs (Monshausen et al 2007). The magnitude of both the oscillations in Ca 2+ and growth rate 9 were variable even within a single root hair ( Figure 1B, supplemental figure 1). While we could not observe a clear relationship between the amplitude of growth peaks vs amplitude of Ca 2+ peaks using linear regression analysis (supplemental Figure 1) our measurements indicated a close temporal relationship between cytoplasmic Ca 2+ and growth where each burst of growth appeared to be followed by a rapid elevation in Ca 2+ ( Figure 1D). Cross correlation analysis comparing the temporal kinetics of Ca 2+ and growth oscillations (Messerli et al., 2000) confirmed that growth peaks most likely preceded Ca 2+ increases by 5.3 ± 0.3 s ( Figure 1E). In agreement with previous observations (e.g. Wymer et al., 1997), non-growing root hairs showed no sustained tip-focused Ca 2+ gradient and no oscillations in Ca 2+ levels could be detected (supplemental Figure 2). Blocking Ca 2+ Entry Leads to Uncontrolled Expansion Our observation that Ca 2+ concentrations increased after growth, parallels results obtained with pollen tubes where it has even been suggested that pulsatile expansion may actually be divorced from the oscillating Ca 2+ gradient (Messerli et al., 2000). To further assess the relationship between Ca 2+ and tip growth, we therefore attempted to manipulate cytoplasmic Ca 2+ levels while simultaneously monitoring cell expansion. Published data indicates that Ca 2+ enters the cytoplasm of tip growing root hairs from the extracellular environment (Herrmann and Felle, 1995;Wymer et al., 1997). To attenuate this influx, we incubated roots with La 3+ , a blocker of Ca 2+ permeable channels. Monitoring Ca 2+ levels during La 3+ treatment showed that 200 µM La 3+ rapidly caused the dissipation of the tip-focused Ca 2+ gradient. Thus, within less than 10 s of treatment, either no difference 10 between apical Ca 2+ levels and those 20 µm from the tip could be observed or a slight decline to below these subapical Ca 2+ levels immediately after La 3+ treatment was observed ( Figure 2A). These observation suggest that the tip focused gradient is largely supported by influx into the cytosol by La 3+ -sensitive channels and collapses very rapidly upon their inhibition whereas, basal Ca 2+ levels are largely maintained under these conditions. Interestingly, however, although inhibited relative to its rate before addition of La 3+ , expansion of the cell apex continued for a few minutes following La 3+ treatment. This expansion accelerated until the root hairs eventually burst at their tips (Table 1, Figure 2B). At higher concentrations of La 3+ (1 mM), almost all growing root hairs ruptured within 10 min of treatment, whereas at lower concentrations of the inhibitor (200 µM), only the vigorously growing root hairs closer to the apex of the root consistently burst; older, more basal root hairs ceased to elongate but continued to swell at the apex for some time (Table 1, data not shown). To ascertain that the growth effects of La 3+ were indeed due to inhibition of Ca 2+ influx rather than unspecific effects, we sought to attenuate Ca 2+ influx by the alternative means of chelating extracellular Ca 2+ . Treatment with 4 mM EGTA also led to rapid bursting of almost all growing root hairs within 10 min of treatment (Table 1). These results indicate that expansion of the root hair tip can be sustained by processes that are not strictly dependent on the high cytoplasmic Ca 2+ concentrations normally found in the root hair apex. It is important here to distinguish between such experimentally-induced expansion/swelling that leads to eventual bursting and the highly controlled cell elongation that is sustained to allow normal growth to occur. The expansion observed in the absence of a clear tip-focused Ca 2+ gradient seemed to occur in an uncontrolled manner leading to cell rupture. Endogenous oscillatory increases in Ca 2+ levels at the growing root hair apex may thus play a role in restricting expansion after each burst of growth to maintain control of the sustained elongation characteristic of tip growth. Increasing the Tip-high Ca 2+ Gradient Arrests Growth To investigate this potential regulatory role of Ca 2+ in restricting growth, we artificially increased cytoplasmic Ca 2+ by treating root hairs with the Ca 2+ ionophore A23187. Figure 3 shows that prior to application of 10 µM A23187, both the tip-focused Ca 2+ gradient and the growth rate oscillated as described in Figure 1. Immediately after treatment, however, Ca 2+ levels rapidly increased ( Figure 3A). The elevated Ca 2+ levels subsequently declined over 30-60s likely due to activation of Ca 2+ homeostatic systems compensating for the increased influx elicited by the ionophore. Interestingly, upon A23187 treatment elongation was arrested within a few seconds of treatment ( Figure 3B). In many root hairs, this high Ca 2+ -induced cessation of cell expansion was accompanied by a thickening of the apical cell wall. This thickening was visible in bright field images as a cap-like structure with altered refractory properties (compare Figure 3C and E). To confirm the nature of this structure, we loaded root hairs with fluorescein diacetate, supplemented the external medium with fluorescein-dextran (10 kDa) and imaged the cell apex using confocal optical sections of less than 1 µm thickness. We found that if image acquisition was performed soon after addition of fluorescein-dextran to the medium, very little of the dye had yet permeated the cell wall. Thus, as both cytosolic fluorescein and extracellular fluorescein were excluded from the wall space, the thickness of the wall became apparent by the absence of fluorescence ( Figure 3F). Although the extent of wall thickening was variable in ionpohore-treated cells, it was most evident in younger root hairs and equivalent thickening was never found in untreated control root hairs ( Figure 3C-F). These observations indicate that while Ca 2+ ionophore treatment rapidly arrested cell expansion of root hairs, secretion of wall material was sustained, leading to a thickening where tip growth-related secretion normally occurs. YFP::RabA4b is thought to mark the apical secretory vesicle machinery in growing Arabidopsis root hairs, dissipating once growth ceases (Preuss et al., 2004(Preuss et al., , 2006. Monitoring the distribution of this YFP marker in root hairs revealed that, as previously reported, RabA4b decorated an apical accumulation of vesicles ( Figure 3H) and this accumulation was maintained despite A23187arrested elongation ( Figure 3J), consistent with the idea that secretion is sustained in these ionophore-treated non-growing cells. DISCUSSION The oscillations in elongation of pollen tubes are thought to reflect either the rapid usage of growth components, that must be re-accumulated to support the next phase of growth, or feedback in either the regulatory or metabolic machinery supporting tip growth (Feijo et al., 2001). Until recently, whether such oscillatory patterns represent an element in the process of root hair tip growth has been unclear. Figure 1 shows that when made with sufficient resolution, oscillations in growth rate could be monitored in root hairs of Arabidopsis with a frequency of 2-4 peaks per min (see also Monshausen et al., 2007). During these oscillations, elongation decelerated and accelerated, with growth rates rising to up to 3 times basal levels. It is important to note that elongation never completely paused during the slowest phase of the oscillation. Thus, the regulatory mechanisms behind the oscillatory growth most likely represent the effects of cellular machinery fine-tuning the rate of expansion. In lily pollen tubes, growth oscillates at approximately 1-3 peaks per min with peak growth rates reaching 13 2-5 times the basal level (Pierson et al., 1996;Messerli and Robinson, 1997). Thus, despite the slower growth rate of the root hair, the kinetics of its growth oscillations closely parallel those of pollen tubes suggesting a possibly conserved oscillatory mechanism. Although genetically encoded Ca 2+ reporters such as the yellow cameleon 2.1 have been used as a non-invasive method to monitor Ca 2+ changes during pollen tube growth (e.g. It is important to note that while the Ca 2+ gradient underwent regular oscillations, there was a statistically significant elevated Ca 2+ level at the tip throughout the oscillatory cycle ( Figure 1). This consistently elevated Ca 2+ concentration is likely to support basal levels of exocytosis during apical growth. Indeed, Figure 3 shows that when Ca 2+ levels were artificially elevated by ionophore treatment, there was an increase in apical wall thickness, consistent with a Ca 2+ -promoted fusion of secretory vesicles leading to exocytosis of wall material. A similar phenomenon has been reported in pollen tubes where Reiss and Herth (1978) observed an apical thickening as growth was arrested. Importantly, our analysis of the organization of the apical secretory membrane system using YFP::RabA4b indicates that even after growth arrest by application of Ca 2+ ionophore and Ca 2+ elevation, the apical secretory machinery remains in place, consistent with the idea that the elevated Ca 2+ level could drive this apparatus to high levels of secretion and so to the wall thickening we observed. However, during normal tip growth of the root hair, oscillatory Ca 2+ increases are superimposed on the stable tip-focused Ca 2+ gradient. The maximum of each Ca 2+ oscillation occurred after an increase in growth, suggesting that while the basal gradient may be supporting sustained growth, the Ca 2+ peaks may play additional role(s) in organizing spatial and temporal aspects of root hair elongation. One possibility is that the increase in Ca 2+ is acting to 'prime' the hair for the subsequent pulse of growth, i.e. acting to prepare the secretory apparatus for the next round of cell expansion. However, the strong crosscorrelation of the Ca 2+ increase to following the period of maximal growth rate, rather than preceding it, suggests a role in processes following the burst of increased growth. Alternatively, elevated Ca 2+ levels could be acting on enzymes that regulate wall structure to rigidify the wall and so help limit turgor-driven expansion. Such dual Ca 2+ -dependent regulation is well characterized for many mammalian regulatory processes where the spatial and temporal dynamics of a change in Ca 2+ can trigger different responses in the same cell. Our observations that (1) blocking Ca 2+ influx into the root hair leads to uncontrolled expansion/bursting, (2) artificially increasing Ca 2+ causes growth arrest, and that (3) the peak of the Ca 2+ oscillation occurs after a burst in growth are all consistent with a role for the maximal component of the oscillatory cytosolic Ca 2+ increase in limiting rather than facilitating expansion. Similar effects have been seen in fungi where the hyphae of Saprolegnia ferax show abnormal growth with enlarged hyphal diameter when they are transferred to nominally zero Ca 2+ media (Jackson and Heath, 1989), suggesting that the restraints on growth may also have changed. Previous analyses of treatments that alter the Ca 2+ gradient in the root hair often reported a cessation of root hair growth (Clarkson et al., 1988;Miller et al., 1992;Herrmann and Felle, 1995;Wymer et al., 1997). For the channel blockers used in these previous experiments, the cessation of growth likely reflects growth conditions, such as much higher Ca 2+ levels in the medium, where the bursting we observe is suppressed. During oscillating growth we have also measured dynamic increases in extracellular ROS and pH that oscillate with a similar frequency as growth but lag growth oscillations by 7-8 s (Monshausen et al., 2007). These extracellular changes are thought to play a role in restricting growth at the tip (pH) and along the shank immediately behind the tip (ROS). Intriguingly, these oscillations in extracellular ROS and pH lag the oscillating increases in the intracellular Ca 2+ gradient we show here. Therefore it is possible that the oscillatory nature of the cytosolic Ca 2+ gradient, and of the extracellular ROS and pH changes may be linked as part of a system to limit growth once an initial burst of elongation has occurred with elevation in Ca 2+ being driven by each growth pulse and itself triggering subsequent ROS and pH response systems to limit further expansion. Such a model fits well with the likely Ca 2+ dependency of the NADPH oxidases, which contain an EF-hand-like Ca 2+ binding domain that appears critical for supporting tip growth and recent data suggesting that ROS and Ca 2+ regulation of growth form a feedback loop to sustain tip growth (Takeda et al., 2008). The spatial and temporal aspects of these three oscillatory parameters in relation to growth are depicted in the model shown in Figure 4. Our observation that La 3+ and EGTA buffering of the medium prevents the formation of the oscillations in tip focused Ca 2+ (Figure 2) supports the idea that influx across the plasma membrane is a key element regulating the dynamics of the gradient, possibly acting as a primer to trigger Ca 2+ release from internal sites, as proposed for pollen (Messerli and Robinson, 1997;Messerli et al., 1999;Messerli and Robinson, 2003). One possible influx mechanism is through Ca 2+ permeable channels directly gated by tension in the plasma membrane, as seen in pollen tubes (Dutta and Robinson, 2004). These pollen tube channels are known to be Gd 3+ -sensitive (Dutta and Robinson, 2004) and the Ca 2+ influx into root hairs is likewise Gd 3+ sensitive (supplemental Figure 3). Alternatively, cytoskeletal elements may play a role in regulating mechanosensitive channel activity, as suggested for pollen tubes (Wang et al., 2004). The relationship between the ROS/hyperpolarization-activated Ca 2+ channel thought to support the gradient in root hairs (e.g. Foreman et al., 2003) and such a mechanical response remains to be defined. It is probable that more than one Ca 2+ -permeable channel exists at the tips of root hairs similar to the tip-growing rhizoids of Fucus that contain two Ca 2+ -permeable channels, one with and one without mechanosensitivity (Taylor et al., 1996). Integration of the activity of these channels may well lie at the heart of the system that must precisely balance the promotion and restriction of turgor-driven expansion to permit root hair elongation without runaway expansion and the associated catastrophic failure of the apical wall. The molecular identity of these channels, how they relate to enhanced exocytosis of wall material and their relationships to the extracellular ROS production and proton transport systems linked to growth restriction (Monshausen et al. 2007) is a major challenge for future research. Seeds of Arabidopsis thaliana Columbia were surface sterilized and germinated on Murashige & Skoog medium (Sigma) supplemented with 1% (w/v) sucrose and 1% (w/v) agar at 21° C under continuous light conditions. Four-day-old seedlings were chosen for experiments. Imaging of Cytosolic Ca 2+ Levels Arabidopsis seedlings expressing the FRET-based Ca 2+ sensor yellow cameleon YC3.6 (Nagai et al., 1999) were transferred to purpose-built cuvettes and mounted as described previously (Monshausen et al. 2007). After several hours of growth in agar containing 0.1 mM KCl, 1 mM NaCl and 0.1 mM CaCl 2 , pH ~6, supplemented with 1% (w/v) sucrose, root hairs were ratio imaged with the Zeiss LSM 510 laser scanning confocal microscope (Carl Zeiss Inc., Thornwood, NY) using a 40x water immersion, 1.2 numerical aperture, C-Apochromat objective. The YC3.6 Ca 2+ sensor was excited with the 458 nm line of the argon laser. The CFP (473-505nm) and FRET-dependent Venus (526-536 nm) emission were collected using a 458 nm primary dichroic mirror and the Meta detector of the microscope. Bright-field images were acquired simultaneously using the transmission detector www.plantphysiol.org on August 17, 2017 -Published by Downloaded from Copyright © 2008 American Society of Plant Biologists. All rights reserved. of the microscope. For time-lapse analysis, images were collected every 2 or 3 s, with each individual image scan lasting 1.57 s. In situ calibration was performed by raising Ca 2+ to saturating levels for YC3.6. This was attempted by treatment with 1 M CaCl 2 , or 50% EtOH or mechanical perturbation just below the threshold for causing cell rupture. The maximum FRET/CFP ratio was attained in response to mechanical perturbation (R max = 2.5). The minimum FRET/CFP ratio (R min = 0.65) was recorded by treating the plants with 1 mM BAPTA-AM (Molecular Probes). Ca 2+ levels were then calculated according to the equation Ca 2+ =K d (R-R min )/(R max -R) 1/n where R represents the FRET/CFP ratio measured during the experiment (Miyawaki et al., 1999), n represents the Hill coefficient which has been determined as 1 for YC3.6, and the K d for Ca 2+ of 250 nM (Nagai et al., 2004). Due to the inherent uncertainties of the precise in vivo K d in such in situ calibrations, the raw FRET/CFP ratio data is included in each figure. We used two different lines of transgenic Arabidopsis Columbia expressing 35Sdriven YC3.6: line 1 was transformed with YC3.6 in the binary vector pGreenII (generous gift of Jeffrey Harper, University of Nevada, Reno). To generate line 2, pGreenII was restrictiondigested with NcoI and EcoRI to obtain YC3.6 with NOS terminator. This fragment was ligated into the Gateway entry vector pENTR11 (Invitrogen) and subsequently recombined into the binary Gateway-compatible destination vector pEarleyGate100 (Earley et al. 2006) according to published protocol (Invitrogen). Because pENTR11 and pEarleyGate100 both contain kanamycin resistance as a selection marker, the backbone of the entry clone was cleaved with PvuII and SspI prior to recombination. Recombined plasmids were transformed into E. coli Mach1 cells (Invitrogen) and clones were selected on LB/kanamycin plates. Recombinant plasmids were then transformed into Agrobacterium tumefaciens via www.plantphysiol.org on August 17, 2017 -Published by Downloaded from Copyright © 2008 American Society of Plant Biologists. All rights reserved. electroporation, followed by transformation of Arabidopsis by floral dip (Clough and Bent, 1998). No differences in the Ca 2+ oscillations were observed between these two independent Arabiopsis lines and neither showed discernable alterations in root hair morphology or growth rate relative to wild type plants. Measurement of Root Hair Growth Bright-field images were collected every 2 s simultaneously with fluorescence images using 458 nm excitation. High resolution growth measurements were made using the computer vision tracking software as previously described (Messerli et al., 1999) providing 1/10 pixel resolution. For high-resolution analysis of root hair tip growth, plants were normally imaged with the root growing through a gel matrix because this approach restricted movement of the main root axis and allowed us to determine minute changes in root hair tip position caused by apical growth. To observe the effect of the Ca 2+ -modulating reagents EGTA, Ca 2+ channel blocker La 3+ and Ca 2+ ionophore A23187 on root hair growth, growth measurements were performed on root hairs immersed in liquid medium. After observing growth of a root hair for several minutes before treatment, the reagent (at 2x concentration) was gently mixed into the medium and growth measurements were continued on the same cell. The use of liquid medium allowed noticeable shifting of the root axis and made measurements of root hair tip growth more difficult. However, this approach afforded the necessary rapid access of the reagents to the root hairs without delays due to diffusion through an agar medium. To study the effect of irradiation intensity on root hair growth, elongation rates were monitored under three different imaging conditions: (i) using the 458 nm line of the argon 20 laser (at 4.7 A tube current, 90% attenuation), a root hair was imaged once at time 0 and again after 5 min; (ii) a root hair was imaged every 2s for 5 min (458 nm at 4.7 A, 90% attenuation); (iii) a root hair was imaged every 2s for 5 min (458 nm at 4.7 A, 60% attenuation). The average growth rate was calculated on the basis of the increase in root hair length during these 5 min. Cross-correlation analysis was performed to determine the temporal relationship between tip-restricted Ca 2+ oscillations and growth oscillations. The correlation coefficient ) was determined as the measurements of the growth oscillations were shifted in time with respect to the Ca 2+ oscillations. SS X and SS Y are the sum of the squares for corresponding Ca 2+ and growth recordings while SP XY is the sum of the products of the two corresponding recordings. A perfect sine wave produces a correlation coefficient of 1 at t = 0 when compared to itself and a value of -1 at t=0 when compared to itself 180 degrees out of phase. As data sets were shifted in time with respect to each other the data points at the tail ends no longer overlapped. These points were removed from the analysis. The temporal resolution of the analysis is the same as that used to acquire the images but there is an offset of one-half of the temporal resolution due to the fact that the growth rate measurements were plotted at the half-way point in between the corresponding images used to determine the growth rate. Monitoring Root Hair Cell Wall Thickening Arabidopsis roots were treated with 5 µM fluorescein diacetate for 5 min. After washing, fluorescein-dextran (10 kDa) was added to the medium until intracellular and 21 extracellular fluorescein fluorescence intensities were approximately equal. Root hair apices were imaged with the Zeiss LSM 510 microscope using the 40x water immersion objective described above. Fluorescein was excited with the 488 nm line of the argon laser. Emission was collected using a 488 nm primary dichroic mirror and a 505 nm long pass filter. Optical sections of less than 1 µm thickness were acquired. Bright-field images were acquired simultaneously using the transmission detector of the microscope. Monitoring YFP::RabA4b localization in Arabidopsis root hairs Root hairs of Arabidopsis expressing YFP-RabA4b (Preuss et al., 2004) were imaged using the Zeiss LSM 510 and the same imaging parameters as described above for fluorescein. Values selected for calculation of averages are depicted as asterisks (peaks) and arrowheads (troughs) in (B). D, Quantitative analysis of root hair growth rates and cytosolic Ca 2+ levels at the root hair apex. Ca 2+ was measured in a ∼30 µm 2 ROI outlined in (A). Representative of n>10 measurements. E, Cross correlation analysis of Ca 2+ oscillations with growth oscillations indicate that the increases in cytosolic Ca 2+ lag increases in growth rate by approximately 5 s. Cross correlation was performed on data from eight separate root hairs. hair. Ca 2+ levels were measured in a ∼30 µm 2 ROI at the root hair apex as outlined in Figure 1A. The rapid increase in Ca 2+ at the end of this recording is due to Ca 2+ eventual bursting of growing root hairs. Arrow denotes Ca 2+ increase due to bursting. Supplementary data Supplementary movie 1. Cytosolic Ca 2+ oscillations during tip growth of an Arabidopsis root hair. Cytosolic Ca 2+ was monitored in plants expressing the soluble Ca 2+ sensor YC3.6 as described in the Materials and Methods. Images were takes every 3 s. Movie duration 3.5 min. Ca 2+ levels have been pseudocolor-coded according to the scale in Figure 1. Supplementary Figure 1. Regression analysis of the relationship between peak Ca 2+ levels during each oscillation and the preceding or following peak in growth rate. Peak Ca 2+ level and preceding and following growth rate were plotted for 8 separate root hairs over 4-5 min of observation each equating to between 12 and 18 growth and Ca 2+ peaks. Regression analysis was performed using Excel. Note that when plotting amplitudes of growth peaks versus subsequent Ca 2+ peaks, the slopes of all trend lines are positive, whereas slopes are more variable for Ca 2+ peaks versus subsequent growth peaks. However, in all cases, the regression coefficients are low, suggesting that either a poor relationship exists between these factors, or that the relationship between them is not a simple linear one. Alternatively, variability in the data likely related to making such measurements in Ca 2+ levels and especially growth rate at the limits of resolution of current technology may also be obscuring the relationship. Supplementary Figure 2. Ca 2+ levels at the apex of a non-growing root hair. Root hairs that had ceased undergoing tip growth in Arabidopsis plants expressing the Ca 2+ sensor YC3.6 targeted to the cytosol were imaged every 3 s. Note the lack of detectable Ca 2+ oscillations at the apex. www.plantphysiol.org on August 17, 2017 -Published by Downloaded from Copyright © 2008 American Society of Plant Biologists. All rights reserved.	2017-07-19T06:49:11.659Z	2008-06-26T00:00:00.000	{ "year": 2008, "sha1": "1adc2899c4e2c22b2e68b2b7187810082d84f740", "oa_license": "CCBY", "oa_url": "http://www.plantphysiol.org/content/plantphysiol/147/4/1690.full.pdf", "oa_status": "HYBRID", "pdf_src": "Adhoc", "pdf_hash": "e7777340e7be84cb723cadf9eccb5a245bb594d9", "s2fieldsofstudy": [ "Biology", "Environmental Science" ], "extfieldsofstudy": [ "Biology", "Medicine" ] }
227208768	pes2o/s2orc	v3-fos-license	A methodology for co-constructing an interdisciplinary model: from model to survey, from survey to model How should computer science and social science collaborate to build a common model? How should they proceed to gather data that is really useful to the modelling? How can they design a survey that is tailored to the target model? This paper aims to answer those crucial questions in the framework of a multidisciplinary research project. This research addresses the issue of co-constructing a model when several disciplines are involved, and is applied to modelling human behaviour immediately after an earthquake. The main contribution of the work is to propose a tool dedicated to multidisciplinary dialogue. It also proposes a reflexive analysis of the enriching intellectual process carried out by the different disciplines involved. Finally, from working with an anthropologist, a complementary view of the multidisciplinary process is given. Introduction A growing number of natural disasters (wildfires, storms, earthquakes, tsunamis, etc.) and other crisis events (terrorist attacks, industrial accidents, etc.) occur every year (EM- DAT, n.d.). Society needs to prepare and appropriately manage these events. Therefore, it is essential to understand how the population and individuals are likely to behave in such unpredictable events. Computer science can provide models of human behaviour and simulators to be used as decision support systems to improve crisis management. However, in order to be realistic, these models need to be informed by data, often gathered from social science researchers. Indeed, previous studies have shown the importance of taking into account human behaviour when planning evacuation and population protection after a natural disaster such as an earthquake (Bernardini et al. 2014;Rojo et al. 2017). Sociology also provides insight about the factors of human behaviour that are relevant in disasters. For instance, Mawson's theory of social attachment states that individuals facing a threat will first seek proximity with their attachment figures (family, friends, familiar places and objects, etc.) rather than seeking protection (Mawson 2005). As a result, factors such as social attachment must be taken into account when modelling human behaviour in disasters (anonymised for review). In previous work, we designed a model of human mobility after an earthquake (anonymised for review). This was a highly multidisciplinary project involving geographers, computer scientists, anthropologists, and field experts. The model was built based on various types of data (both qualitative and quantitative) from various sources, such as official reports, scientific publications, census data and field surveys, etc. Data was collected on: human activities and mobility during earthquakes, especially just after the main shake; social bonds between members of a social group such as between family, friends, and colleagues, according to different European countries; socio-demographic characteristics of the population; population distribution at night and during the day; etc. Collecting, exploiting and integrating such diverse and multidisciplinary data into a model was an extremely challenging task, especially since no methodology existed at the start of the project to support the process. Therefore, we had to create our own methodology as the project unfolded. The goal of this paper is to share this methodology to help future multidisciplinary projects. Multidisciplinary research is notoriously difficult and the problems mentioned above are typical of such work. Although integrating empirical data is essential (Boero 2005, §2.17) no standard methodology exists yet and so each project tends to adopt their own. Most often, tasks are split chronologically and by disciplines, leaving the data collection to the social scientists (in our case, social geographers) at the start of the project, while computer scientists deal with model design afterwards. As a result of this task division, there is very little interaction and feedback between data collection and model design. This is far from optimal and can lead to inconsistencies or gaps in the model. On the contrary, our approach consisted in involving scientists from all disciplines at all stages of the process. We consciously made the decision not to just share information on methods, techniques and procedures used in the various disciplines, but to actively involve all disciplines. This led to more discussions and iterations between the 'data collection' and 'model design' phases, and to an iterative improvement of the model. Concretely, both social scientists and computer scientists participated in the elaboration of a questionnaire for gathering human behaviour data, and both also participated in the design of a conceptual model based on this data. Domain experts then helped with the scenario design and model validation. In this paper, we describe the multidisciplinary methodology that we devised from our practical experience in this project. We believe that it is generic enough to be applied to and benefit other similar multidisciplinary projects. We focus on explaining how scientists from different disciplines worked together in order to gather data and exploit it to design a more "complete" model. Although we discuss validating the methodology, 'model validation', which ensures that the model is an acceptable representation of reality, is out of the scope of this paper. The paper is structured as follows. Section 2 surveys the state of the art on model design and data collection. Section 3 describes our approach and the proposed methodology for multidisciplinary data collection via surveys and iterative model design. Section 4 presents the three results obtained from using our methodology: the questionnaire, the model, and the tool to help multidisciplinary dialogue. Finally, we discuss the methodology through a metaanalysis, and conclude this paper in Section 5. Model design with an interdisciplinary approach Most efforts to include diverse stakeholders in model design have focused on involving users or domain experts. Indeed, the involvement of stakeholders is now well accepted and commonly used in the design of simulation models. In particular, participatory simulation, in which users are involved very early on in the design of the model, has proved to be particularly advantageous in terms of gaining user acceptance of the final tool and in ensuring that a more complete and refined model is achieved (Ramanath and Gilbert 2004). One of the first efforts to include different types of disciplinary knowledge in multi-agent systems was Olivier Barreteau's work (Barreteau 1998) involving farmers concerned with the irrigation of the Senegal River. This work developed into coupling multi-agent systems with a role-playing game (RPG). RPGs have been used as a method to elicit knowledge and to formalise assumptions constituting the model. Starting from a basic conceptual model, the RPG is used to improve the formalisation and act as a communication tool between the model and reality (Barreteau 2003;Bousquet et al. 2002). Companion Modelling (ComMod) is a participatory gaming and simulation approach that uses RPG and simulation models (Bousquet and Trébuil 2005). The ComMod approach is iterative and evolves with the participative process whereby stakeholders are involved in the definition and design of the questions, models, simulations and outputs (Étienne 2013). RPG and companion modelling predominantly include users or stakeholders that are not designers in order to co-construct and refine the model in an iterative way. Our approach differs in two ways. Firstly, we do not focus on using RPG, but use the survey as an instrument to feed the model. Secondly, although users and domain experts helped in coconstructing the model, we focus on how research scientists from different domains can work together in a complementary way to co-construct the model. Other works, such as Cioffi-Revilla 2010, use an approach based on conceptualizing and developing a succession of models with increasing complexity as they approximate the target system. This methodology is particularly appropriate for large, multidisciplinary projects for complex social simulations. However, it considers a target to be modelled through different submodels that may be related. In our case, we consider one model being addressed by the different disciplines. Other works have proposed methodologies to link simulators that stem from different domains (for example agent-based and physics-based) (Kashif et al. 2013). Our work differs in that we are not trying to link simulators, as in a co-simulation approach, but rather to develop one holistic simulator. Data collection In order to develop a model of human behaviour it is important, not only to draw upon social theories from the literature, but also to collect data on human activities and characteristics. In our project we were specifically interested in gathering social data on behaviours adopted just after the earthquake, on social bonds, and on the day and night population distribution. Regarding the latter, spatial and statistical processing of household-travel survey data was extrapolated to give the spatial and temporal distribution of a population over a territory (Roddis and Richardson 1998). Several methods exist to collect social data on behaviours during natural crises. In our project, gathering social data is highly influenced by the objective of the study. Indeed, among the existing methodologies, some may not be adapted to the thematic framework, i.e. natural unpredictable crises. The literature provides a broad idea of the social response to a natural crisis (Provitolo et al. 2011). We also know that the cultural context plays an important role in risk perception and the behaviours adopted during a natural crisis (Palm 1998;Paradise 2005). Nevertheless, we needed to gather data specifically for the application context of our model. The collection of behavioural data is possible with: direct observation, indirect observation (video recordings), and interviews or questionnaires. Direct in situ observation requires a specific protocol covering a social situation and deploying it at a precise moment (Kaplan 2017). The major advantage of direct observation is that the behaviours are not based on declared words (as opposed to surveys by questionnaires or interviews), which makes behaviours more objective. However, the intrinsic unpredictability of earthquakes means that we cannot use this method in this context. Indirect observation consists in analysing video recordings, provided either by video surveillance or by individuals through digital social networks such as YouTube and FaceBook. This method is powerful to qualitatively describe behaviours (Gu et al. 2013;Lambie et al. 2016;Ziu et al. 2016). However, one of its limitations is the statistical representativity of the output since some extreme behaviours may be over-reported/recorded. Another problem comes from the limited view of the scene; some behaviours may be explained by contextual elements that are not recorded by the camera. Among the survey techniques, interviews and questionnaires highlight the distinction between qualitative and quantitative approaches. Those two techniques are based on declarations which may bring some subjectivity to the answers. Surveys via interviews can provide very fine-grained and precise data (Groves et al. 2009). When they are given to small samples, they can uncover in-depth information about a subject. When carrying out interviews, the interviewer seeks to reveal a diversity of opinions or situations. The administration of the survey as well as the processing of answers can be very time-consuming. Among surveys, feedback methods are carried out just after an event. This is done to understand the physical and social causes of the catastrophe and to highlight the dysfunctions in the crisis management in order to improve local prevention methods or national legal framework of risk management. In a scientific context, these methods have been used to collect very precise data on pedestrian mobility behaviours following a flash flood (Ruin et al. 2013) or an earthquake (Rojo et al. 2017), using a spatio-temporal grid or a map (anonymised). Feedback methods, if used to uncover mobility behaviour after an event, may last at least one hour per interviewee. Like interviews, surveys by questionnaires are widely used to gather social data (De Leeuw et al. 2008), especially behavioural data on medium to large samples. The idea is to obtain a statistically representative picture of a situation at a certain time. Questionnaires can be either administered face-to-face or self-administered, i.e. sent by email, mailed or delivered in mailboxes (Jon et al. 2016). In the case of self-administered questionnaires, the survey-carrier should have verified twice that the questions are well understandable so that the answers correspond to the survey's objective. This can be done by testing the questionnaire several times with various people. The evolution of ICT and in particular digital social networks means that questionnaires can be easily distributed and completed online. The online accessibility of the survey facilitates its spread through social media, electronic mail, webpages, newsletters, therefore gathering many answers from a wide audience over a very large area. It also allows surveying a more socially diverse sample by reaching people who would not answer a face-to-face survey. Thus, a large number of responses can be easily gathered (Rhodes et al. 2003;Evans and Mathur 2005). For example, more than 1500 answers to a single questionnaire were obtained after the Ml 5.2 Ubaye earthquake in 2014 (Boisson 2015). Since the survey is not addressed to a specific sample, statistical representativity can be pre-assured by using a filter or with statistical adjustment. The following table summarises the characteristics of the different techniques that can be used to collect social data. Considering modelling needs, online surveys are often the most useful method: they are designed to rapidly gather data over large samples and can be used to collect detailed information if open questions are included. Limitations From the state of the art there are several limitations with current approaches: • The lack of a tangible tool for use in multi-disciplinary work: a multi-disciplinary approach is increasingly common and necessary, yet few, if any, tools exist to support and facilitate efficient cooperation. • The difficulty of interactions between computer science and social science: interactions between computer scientists, who develop the simulator, and social scientists, who collect the domain knowledge, are crucial but often difficult. Good tools are needed to help social scientists visualise and understand what is being developed by the computer scientists, and ensure it matches with their domain expertise. Likewise, computer scientists need to understand the intricacies of data collection. • Limited iterations between data collection and model design: in the absence of such dedicated tools, the iterations will be limited between the scientists of the different fields, often with only a single data collection phase at the start of the project, followed by a single model design phase. If the social scientists are not able to understand the model being designed, they will not be able to comment on it, nor to provide useful feedback to inform its iterative design. • Existing methods/tools for modelling for computer scientists are often not well understood by social scientists; for example, UML and TDF (Evertsz 2015) have been used for agent modelling (Mancheva et al. 2019;Richiardi et al. 2006) but they are hard to master for non computer scientists. • Conversely, existing methods/tools for modelling for social scientists are often not well understood by computer scientists; for example, in the social sciences, models can be "designed/described/represented" by paragraphs, tables or graphs obtained from analysis of surveys or even schemes. • Different vocabulary/tools/methods in different disciplines that have to cooperate. Each discipline has its own vocabulary and some efforts need to be made to facilitate discussions. Experiencing multi-disciplinary cooperation over time helps acquiring new vocabulary from other disciplines or, at least, understanding the way other disciplines define various terms and concepts. Objectives and proposed approach To address these gaps in the literature we provide several contributions to the modelling community: • A methodology, explained in section 3, to bring together approaches from social sciences and computer sciences, for both data collection, via field surveys, and model design. • An illustration of this methodology, detailed in section 4, for a specific case study of earthquakes. In this context, we designed a questionnaire that is both rigorous from a social science point of view, and useful from a computer science point of view for gathering data required for the model and simulation. We also designed and implemented a simulator based on the survey data and other sources. This model takes into account social interactions in the population after an earthquake 1 ). • A meta-analysis of what can be learnt from applying the methodology is explained in section 4. We believe that the methodology is generic and that it can benefit the modelling community and help them overcome some of the difficulties of multidisciplinary work. Method We first describe our proposed generic methodology in sub-section 3.1. We then analyse how we applied it in a specific project, and discuss the challenges we had to overcome (subsection 3.2). This meta-analysis aims to provide insights to other researchers wanting to apply our methodology in their projects. General method -model design The basic steps of the methodology are shown in Figure 1. STEP 1: The process begins by each team -computer scientists and social scientistscollectively defining the objectives of the model. As noted by Nigel Gilbert this is a crucial step in order to ensure that the developed model directly targets the needs of the stakeholders (Gilbert 2004). In the second step, each disciplinary team separately, but in parallel, develops a first draft of the model. In essence, this describes how each team interprets the problem, what critical factors they think should be included, and how these should be conceptualised. STEP 3: Once the two draft models are completed, they are compared in order to identify the common elements of the models, and more importantly the differences: which elements were included in one model but not in the other. This is a highly interactive step with in-depth discussions to explicitly justify the inclusion or exclusion of model elements. The output of this step is a common model, collectively co-constructed by both teams. STEP 4: Each element of the model is then examined from the point of view of the social data needed to support those elements. For example if demographic data, such as the age distribution of the population is needed, discussions follow as to how this data will be obtained. This step obviously uncovers cases where elements should be included in the model, but for which no data is yet available to support their inclusion. Obtaining data for model building has been cited as one of the most challenging aspects of constructing a suitable model (Beck et al. 2014). The proposed method allows the teams to clearly identify missing data. STEP 5: Following the identification of the missing data required to feed the model, the computer science team then develops a questionnaire in order to collect this data. The social science team would appear more suited to this task, but there are several arguments behind the choice of having the computer scientists do it. The first is educational, as it exposes the computer science team to the social science culture of data gathering, and abruptly confronts them with the reality of multi-disciplinary work. The second is more practical: questionnaires developed by social scientists typically cover a (too) wide range of research questions, while computer scientists tend to focus on questions that provide data to feed the model. However, if they follow the KIDS approach (Edmonds and Moss 2004) and design a very descriptive model, it may result in an excessively large number of attributes and behaviours being included in the model. The two teams then interact and the social science team's experience is used to revisit the proposed survey, classify research questions from essential to less interesting, remove unnecessary questions to shorten the questionnaire, and rephrase questions when needed. The output of this step is a first version of the questionnaire. STEP 6: The final step of our methodology consists in verifying that the questions in the resulting survey are necessary and sufficient. This step aims at avoiding inconsistencies between model and data that can result from the traditional approach when questionnaire and model are designed separately, i.e. the survey is carried out at the start of the project while the model is designed at the end. Concretely in this step, the teams check that the questionnaire can actually be used to gather all the data that has been identified when conceptualising the model (sufficient questions) and, as an iterative process, that the model has taken into account all the potential behaviours listed in the questionnaire (necessary questions). This validation step checks that the model and survey are consistent with each other, i.e. checking that all the information necessary to the model can be collected by the survey (no relevant question is missing), and that all information collected via the survey is really useful to the model (no irrelevant question is included). In summary, the output of this step is a common questionnaire whose set of questions is both necessary and sufficient to build the model. Analysis of the multidisciplinary iterative methodology The process of developing a model using this methodology was analysed by an anthropologist by conducting interviews with the researchers and through direct observation. The goal was to extract the relational dynamics during the process of building a common multidisciplinary methodology. Specifically, we aimed to uncover: 1. How researchers went about jointly reshaping their usual tools, namely the survey questionnaire and the multi-agent model? 2. What were the obstacles to mutual understanding of the disciplines involved and their respective methodological requirements? 3. How these obstacles were overcome or resolved by the team? From the data collected, the longer-term objective was to develop a mediation workshop aimed at fostering interdisciplinary dialogue (see 4.3). The goal was not so much about observing the content of the information exchanged, but about observing "science being made" (Latour 1989), with its uncertainties and hesitations in a project where reflexivity was key. Indeed, the construction of the multidisciplinary methodology required researchers to analyse their own science, triggered by questions from colleagues from another discipline. Step by step, each scientist explored the methodological tools, ways of categorising the data, and how the problem was interpreted by the other discipline. Our observation of the lengths and blockages in this process of methodological deconstructionreconstruction allowed us to identify the sensitive points of interdisciplinary work. Indeed, the research team was already used to working together on the design of multi-agent models of an earthquake crisis. As a result of previous multiple collaborations, the teams had reached a common understanding of some of the tools and concepts of each other's domain (e.g. the notions of "risk perception" and "vulnerability" as used in social geography, and how these may be implemented from the point of view of computer science). This meant that the exchanges were relatively fluid. However, this project involved leaving our comfort zone, in order to closely readjust the methodology to the requirements and constraints of each discipline. The moments of hesitation (in this case the development of a common survey) were informative and highlighted important issues: the critical notion of pre-evacuation delay for computer scientists, or the need for exhaustive identification of individual behaviours after a quake for geographers. The individual interviews that were conducted with the members of the team compared actual observations of the process with the researchers' feelings about how they learnt, the obstacles encountered, and the methodological choices that were made. They were also asked about their initial motivations for conducting multidisciplinary projects, the usual difficulties of this type of collaboration, and the expected benefits for them. Like other researchers sharing interdisciplinary experience, the team worked together primarily on the basis of intellectual and relational connivance. Multidisciplinary collaboration requires a strong will to overcome disciplinary differences through the development of a common epistemology: the search for unification through a language, a common object, a shared method requires devoting time and many reformulations (Hervé and Rivière 2015). Thus, many of the project exchanges were initiated by "what do you mean by ...?'' questions, aiming at ensuring mutual understanding. The intrinsic challenges of interdisciplinarity also face another difficulty in constructing and following a project methodology. Beyond science, even though the objective of the work and the steps needed to achieve had been collectively discussed beforehand, they were still potential sources of misunderstanding. In this project, the position of the survey in the methodology and the lack of intermediate steps were questioned a posteriori by the researchers. As a social and technical construction process, scientific activity takes the form of collective actions (Aggeri and Hatchuel 2003). The attention paid to mutual understanding at all stages of a project thus goes beyond the question of interdisciplinarity: it aims at understanding more generally the modalities of group interaction. The work at the heart of a multidisciplinary team finally raises other concerns or even frustrations. These were uncovered during the interviews as: losing one's own identity due to oversimplification, or leaving important concepts aside to facilitate comprehension. The identity of each discipline is at stake here, as well as the limits of interdisciplinary sharing. Discovering other ways of producing knowledge highlights the division between the disciplines. By breaking down the disciplinary boundaries, it was possible to better understand each of the disciplines, their methodological and conceptual particularities, and the possible bridges between disciplines. As Barth (1969) points out, there is no starting homogeneity, but there is always a relation. Identity is an unstable amalgamation of various self-images derived from relationships with others. If there are multiple challenges in the vision and division of scientific work (Freymond et al. 2003), the identity constructions of the sciences and those who practice them are in perpetual movement, as with any other identity process. When faced with other disciplines, we make moves towards them in order to facilitate multidisciplinary work. However we also retain a view of our own discipline and this prompts a reflection concerning its specificities, relevance and legitimacy. There are real distinctions between disciplines, yet the exact divisions are arbitrary, depending on our individual disciplinary perceptions. Because of that, actions of integration, assimilation or acculturation to other disciplines are affected by our view of where those divisions exactly lie. It is the thought of this similarity with intercultural encounters that gave rise to the creation of "The MeetTic of Scientists" workshop described below (see 4.3). Results The multidisciplinary project resulted in a questionnaire, a conceptual model that was implemented as a simulator, and a tool to facilitate multidisciplinary dialogue. A questionnaire to gather behavioural data From the limitations and advantages of the ways to collect social data after an earthquake , we set up an online questionnaire. Despite the opportunity to gather a great amount of social data, we focused exclusively on the data that would be useful for modelling individual behaviours in the case of an earthquake. Questions therefore needed to be formulated in such a way that answers could directly be integrated into the model. The structure of the questionnaire allowed people who had not experienced an earthquake to answer the survey. In order to facilitate statistical processing, most of the questions are closed questions. The questions deal with several aspects: 1) Earthquake experience (if any): the questions aim at relating the past experience of an earthquake, i.e. the spatial and social context of the event, time and intensity of the earthquake, feelings that the event produced, the adopted behaviours (especially those involving mobility) with many details (place where the people went, time to evacuate a building -if relevant, reasons for the evacuation); 2) Preparation strategies and knowledge (people who have never experienced a seismic event can still answer those questions): knowledge of the safe open areas, hierarchy of the social bonds, organisation with kin, neighbours and/or colleagues before or after evacuation, preventative information received from any stakeholder, confidence in the building's resistance to an earthquake; 3) Finally, some socio-demographic questions that characterised the respondent's social profile, for instance age, gender, education, family structure, dependant persons, and mobility capacities. Regarding genericity, the survey may be used in various geographical and cultural contexts, as it was designed using previous experiences of the authors in France, Italy, Argentina and Lebanon. However, some slight adjustments may be necessary regarding the proposed items for closed questions. The survey is accessible in French, English and Arabic. As mentioned previously, the online survey is by definition self-administered. For this reason, the questions should be easy to understand by any respondent, which requires testing the questionnaire several times with various people. Despite this disadvantage, the online accessibility of the survey facilitates its spread through social media, electronic mail, webpages, newsletters, therefore gathering many answers from a wide audience over a very large area. It also allows us to survey a more socially diverse sample by reaching people who would not answer a face-to-face survey. Since the survey is not addressed to a specific sample, some statistical adjustments may be necessary if the sample is not representative of the population. The diffusion of the survey may be achieved by several means. Using the professional and personal networks of the authors may result in an over-representation of the more educated populace. For this reason, posting the URL of the survey on social networks or webpagesespecially on local groups (local authorities, municipalities, associations, etc.) -may help diversing the sample. This survey is accessible online at www.XXX (anonymised for review) and extracts can be found in the annex of this paper. A model to simulate a seismic crisis at the individual scale The questionnaire allowed us to gather a lot of qualitative behaviour data, which we then used to design a conceptual model of human social behaviour immediately after an earthquake. The model was built iteratively, with many discussions between researchers from the different disciplines involved. It is now fully implemented, functional, and published (anonymised for review). The focus of this model is on the role of social attachment in individual behaviour just after the main shake of an earthquake. It shows that individuals will first gather with their relatives or friends before evacuating, which can slow down evacuation but also help in getting more people to safety. A workshop to bootstrap multidisciplinary dialogue Beyond building this model, we want to answer a more general question. How can the observations and reflections that emerged from this project benefit other researchers from different disciplines? How can they reuse this insight to facilitate their exchanges in other circumstances? Our idea was to create a game or a workshop around this theme in order to facilitate exchanges between scientists on the difficulties encountered during an interdisciplinary cooperation. Without pretending to provide a miracle solution to these difficulties, it did at least open a space for discussion, in order to bootstrap dialogue and mutual understanding that would be required to initiate a research project during its structuration phase. Multidisciplinary encounters obey the same rules that underlie other forms of human encounters and interactions. This inspired the main thread of the workshop entitled "The MeetTic of Scientists". A first test session was organised over a half-day period in June 2019, and attended by 29 researchers from 10 different domains. The workshop was one of the activities proposed by a summer school dedicated to multi-agent modelling using a spatial approach. However, it could be applied to any other multidisciplinary community wishing to make an introspective analysis of its practices. In the workshop, multidisciplinarity was considered from the point of view of the human encounter, involving the construction of relations between people from different scientific cultures, with all the challenges and benefits that this implies. Combining several techniques from popular education and aiming to encourage the participation of all, the workshop was designed in three stages: • Step 1: the "MeetTIC affinity". In mono-disciplinary groups, participants are asked to share their values, motivations and expectations for interdisciplinary work, and to draw up a short consensual list. Energising and breaking the ice between the participants, this activity seeks to bring out a panel of strong ideas from each group concerning its relationship to interdisciplinarity. • Step 2: "Thematic Speed Dating: Interdisciplinary Relationships and You". This step aims to generate discussion within multidisciplinary teams around defined themes. Teams stay together but move from theme to theme following a predefined schedule (20 minutes per theme). The results of the discussion on each theme are written on posters. The themes deal with multi-or interdisciplinary first experience, setbacks and disappointments, "love at first sight" experiences, discoveries, requirements ("happiness elixir") of each participant for fruitful multi-or interdisciplinary relationships. The idea here is not to generalise or reach a consensus, but to share individual experience. This is followed by an introspective phase, both individual and collective, where individual experiences of interdisciplinary relationships are shared among each group. • Step 3: collective creation of advertisements ( Fig. 2 provides some examples) for an interdisciplinary meeting, website or newspaper. If the previous steps aimed at better understanding what made sense for everyone, this more creative phase allows to playfully forge a collective dynamic. In mono-disciplinary groups, participants must find strong arguments to invite their counterparts from other disciplines to break down disciplinary barriers and to engage in mutual collaboration. • Step 4: collective sharing. The different advertisements and posters created during the workshop are displayed to share the experience of each group. The workshop is easily replicable and helps to establish a basis for exchange between all participants of a project. It can be of real use in facilitating the start of a new collaboration between researchers from various scientific backgrounds. However, dedicating time to interdisciplinary dialogue is only beneficial if it is renewed multiple times in different forms over the course of the research project. Building a shared vision, and repeatedly adjusting common objectives can only happen over the long term in a reflexive way. The workshop also raised other questions. Interdisciplinarity implies finding a common language to progress together, but what are the real possibilities to share one's ignorance of a subject, a concept, within a group of scientists? To what extent does the search for precision and accuracy, which is undoubtedly one of the strong characteristics of a group of researchers, also result in prohibitively long discussion times for clarifying concepts as required for interdisciplinary dialogue? Discussion and conclusion This original approach produced surprising results. Working in a multi-disciplinary scientific group generated an enhanced attentiveness to each other's work. This reflexive process also provided a step back from our own research practices. As an example, structuring the model was different depending on the discipline: a social scientist arranges the behaviours of the people taking into account their wish of changing their initial activity plan, or depending on whether the people may move or not. Conversely, a computer scientist structures the model with a view to coding/implementation and taking into account different possible architectures to model these wishes. This methodology meant that the time to develop the model was longer than with the traditional approach and it required a lot of energy from both disciplines. However, it has helped the researchers involved to get accustomed to and to appropriate both the methodologies (model design and formalisation and questionnaire) and the results (answers to the questionnaire and model obtained). They also improved, not only their multidisciplinary approach, but the consistency of the implemented model. In conclusion, this paper has made the following contributions: • A novel generic methodology for multidisciplinary model design • A methodology to design a questionnaire dedicated to model conceptualisation • A meta-analysis of the multidisciplinary methodology applied to a specific project (earthquakes) • The design of a workshop to help bootstrapping multidisciplinary dialogue We obtained results for a specific model of earthquakes in the context of a particular project (anonymised for review), but we also showed that our methodology is beyond this unique project, and generalisable. We intend our methodology to benefit other researchers engaged in similar multidisciplinary projects, by saving them time and removing the obstacles involved in this complex but fulfilling process.	2020-11-30T02:01:10.748Z	2020-11-27T00:00:00.000	{ "year": 2020, "sha1": "8962309648a6cbb341adee677a8aeb3ecec9edb2", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "8962309648a6cbb341adee677a8aeb3ecec9edb2", "s2fieldsofstudy": [ "Business" ], "extfieldsofstudy": [ "Computer Science" ] }
250313612	pes2o/s2orc	v3-fos-license	The Dynamics of Viruslike Capsid Assembly and Disassembly Cowpea chlorotic mottle virus (CCMV) is a widely used model for virus replication studies. A major challenge lies in distinguishing between the roles of the interaction between coat proteins and that between the coat proteins and the viral RNA in assembly and disassembly processes. Here, we report on the spontaneous and reversible size conversion of the empty capsids of a CCMV capsid protein functionalized with a hydrophobic elastin-like polypeptide which occurs following a pH jump. We monitor the concentrations of T = 3 and T = 1 capsids as a function of time and show that the time evolution of the conversion from one T number to another is not symmetric: The conversion from T = 1 to T = 3 is a factor of 10 slower than that of T = 3 to T = 1. We explain our experimental findings using a simple model based on classical nucleation theory applied to virus capsids, in which we account for the change in the free protein concentration, as the different types of shells assemble and disassemble by shedding or absorbing single protein subunits. As far as we are aware, this is the first study confirming that both the assembly and disassembly of viruslike shells can be explained through classical nucleation theory, reproducing quantitatively results from time-resolved experiments Optimization of the SEC protocol for studying the self-assembly dynamics during size reduction S10 3 Theoretical methods and discussions S12 3.1 Theory S12 3.2 Numerics S17 3. 3 The fraction of protein dimers in free solution and in capsids S18 3.4 Fitting of the simulation results with respect to different reference points S19 S 2 3.5 1 Experimental materials and methods Materials Ampicillin, chloramphenicol, yeast extract, and peptone were purchased from Sigma-Aldrich/ Merck. Isopropyl-β-D-thiogalactopyranoside (IPTG) was obtained from PanReac AppliChem VWR. Ni-NTA agarose beads were obtained from Qiagen. All buffers were filtered over a 0.2 micron filter prior to use. UV-vis absorbance measurements In order to determine the protein concentrations during experiments, the absorbance at 280 nm was measured with a spectrophotometer ND-1000 and the concentrations were subsequently calculated using the theoretical extinction coefficients. 1 Mass spectrometry Protein mass characterization was performed using a High Resolution LC Size exclusion chromatography (SEC) SEC analysis was performed on a Superose 6 increase 10/300 GL column (GE Healthcare Life Sciences). Analytical measurements were executed on an Agilent 1260 bio-inert HPLC. Samples with a protein concentration of 100 µM were separated on the column at 21 °C with a flow rate of 0.5 mL/min. Running buffer was "pH 7.5, 100 mM NaCl no EDTA buffer" for T = 3 to T = 1 shift and "pH 5.0 buffer" for T = 1 to T = 3 shift (see table S1 for exact compositions). Transmission electron microscopy (TEM) TEM grids (FCF-200-Cu, EMS) were glow-discharged using a Cressington 206 carbon coater and power unit. Protein samples (10 µM, 5 µL) were applied on the glow-discharged grids and incubated for 1 min. The samples were carefully removed using a filter paper. Then, the grid was negatively stained by applying 2% uranyl acetate in water (5 µL). The staining solution was removed after 15 seconds and the grid was allowed to dry for at least 15 minutes. The samples were studied on a FEI Tecnai 20 (type Sphera) (operated at 200 kV, equipped with a LaB6 filament and a FEI BM-Ceta CCD camera). Dynamic light scattering (DLS) measurements DLS measurements were performed on a Malvern Zetasizer Nano ZSP at 21°C, unless stated otherwise. Samples (100 µM, unless stated otherwise) were centrifuged twice prior to analysis. Buffers were filtered prior to use. All measurements were done in triplo, and the average of the triplo measurements was plotted. General protocol for measuring the self-assembly dynamics during size increase (conversion dynamics from T=1 to T=3 particles) For a typical dynamics experiment a 100 µM VW1-VW8 ELP-CCMV coat protein solution (150 µL -1200 µL) in pH 7.5 buffer with 500 mM NaCl was prepared and dialyzed to pH 5.0 buffer at 4 °C (12-14 kDa MWCO). Dialysis buffer (150 mL -200 mL) was changed after 30 minutes and 60 minutes. At different time points during dialysis, 110 µL and/or 5 µL samples were retrieved from the dialysis membrane, spun down for 1 minute at 13400 rpm, and subjected to SEC analysis and/ or TEM analysis respectively. Hereafter, the mixtures were dialyzed at 4 °C either to pH 5.0 buffer (T=3 capsids) or to pH 7.5 buffer, 500 mM NaCl (T=1 capsids) and incubated in the final buffer at 4 °C for up to one week. S 7 At intermediate time points, samples were taken, heated to 21 °C with 1 °C/min and subjected to SEC analysis with pH 5.0 buffer or pH 7.5 buffer with500 mM NaCl as eluent. At the final time point, fractions were collected during a preparative SEC run and the combined capsid fractions were analyzed with SDS-PAGE and TEM. The amount of mEGFP incorporated into the capsids was determined by SDS-PAGE analysis. Gels that were visualized via Coomassie Brilliant Blue staining (Biorad) were analyzed with ImageJ gel analysis software to calculate the loading of capsids with mEGPF. Hereto, the following formula was used: where gel is the intensity of the protein band on the SDS-PAGE gel as determined by ImageJ analysis; mw is the molecular weight of the protein. Optimization of conditions to allow for optimal dynamics In order to study VW1-VW8 ELP-CCMV capsid size shifts, we first investigated the optimal conditions that would allow for the dynamic behavior of the capsids. In preliminary results we observed a size shift from T = 1 capsids to T = 3 capsids during overnight dialysis from pH 7.5 to pH 5.0 at 4 °C ( Figure S1). We, therefore, evaluated with SEC whether the reverse size shift would also take place. We observed that only a partial shift from T = 3 to T = 1 capsids took place during the overnight dialysis to pH 7.5 buffer with 100 mM NaCl ( Figure S1A,B first two chromatograms in each panel). We then proceeded by investigating whether a second overnight dialysis to pH 5.0 would induce a re-shift back to T = 3 capsids as well. Interestingly, this was the case only when the dialysis was performed at 4 °C ( Figure S1B), while at 21 °C no size shift was observed. This indicates that at 21 °C the capsids are much less dynamic than at 4 °C. Although this seems contra-intuitive at first, this observation can be explained by the interactions between the hydrophobic ELP-domains which are much stronger at 21 °C than at 4 °C. So it is highly likely that these interactions in the capsid interior hamper rearrangements of the CP domains in the capsid shell, which are necessary for a size shift. Another factor that was thought to influence the capsid dynamics as a result of ELP interactions is the ionic strength of the buffers used. Previously, we used a pH 7.5 buffer with 100 mM NaCl to completely disassemble other ELP-CCMV variants. Therefore, we hypothesized that this ionic strength would also allow for dynamics within capsids of our more hydrophobic VW1-VW8 ELP-S 8 CCMV variant. As, ideally, we wanted to study the capsid size shifts while varying as few factors as possible, preferably only the pH, we attempted to store T = 3 capsids in pH 5.0 buffer with 100 mM NaCl instead of 500 mM NaCl. However, this, unfortunately, led to the aggregation of the protein already within 16 hours (data not shown). We, therefore, evaluated whether VW1-VW8 ELP-CCMV exhibited dynamic behavior when dialyzed from pH 5.0 buffer with 500 mM NaCl to pH 7.5 buffer with 500 mM NaCl, thus only changing the pH. Unfortunately, only a very small part of the capsids appeared to be shifted in size after 24 hours ( Figure S1C) as compared to dialysis to pH 7.5 buffer with 100 mM NaCl ( Figure S1D), which can again be explained by hydrophobic interactions between ELP domains hampering capsid dynamics at 500 mM NaCl. We, therefore, decided to use a shift from pH 5.0 with 500 mM NaCl to pH 7.5 with 100 mM NaCl and vice versa to study the dynamics during VW1-VW8 ELP-CCMV capsid size decrease and increase respectively as is discussed further in the main text. Optimization of dialysis conditions As described in the previous section, it is necessary to both change the pH and NaCl concentration in order to study size shifts of VW1-VW8 ELP-CCMV capsids. If only the pH would have to be adjusted, this could have been done by adding either HCl or NaOH to the buffer. However, to also adjust the NaCl concentration a buffer exchange step is necessary. Although spin-filtration would be the quickest option and would allow for evaluation of the capsid size upon change of the conditions very quickly, it would also introduce changes in the protein concentration, which could affect the capsid assembly state. As this could complicate our evaluation of capsid size changes as a function of pH, we decided to perform dialysis in order to change the pH and NaCl concentration, despite being a slower process than spin-filtration. During initial experiments, it was observed that there is a large dependency of capsid dynamics on dialysis time while incubation periods at 4 °C were kept constant ( Figure S2A). When a T = 3 capsid solution was dialyzed to pH 7.5 buffer with 100 mM NaCl, a large shift to T = 1 capsids was only observed during SEC analysis when the total dialysis time was more than 30 minutes. This indicated that either the pH or the NaCl concentration changed slower than anticipated during 30 and 60 minutes and the dialysis buffer was either stirred at 150 rpm (blue circles) or not stirred (yellow squares). All measurements were performed in triplicate and data is presented as mean ± standard deviation. S 10 dialysis. As the pH switch during dialysis takes place within minutes, it was suspected that the other variable during dialysis, the NaCl concentration, changed more slowly. The amount of NaCl that is dissolved in an aqueous solution affects the conductivity of that solution, thus conductivity measurements were performed to follow the change of the NaCl concentration during dialysis. Hereto, a mock dialysis with the same ratio between the volume inside the dialysis bag and the solvent volume was performed and the conductivity was monitored over a time course of 4 hours ( Figure S2B). A dialysis time of around 2 hours was necessary to fully convert the NaCl concentration from 500 mM to 100 mM NaCl in the dialysis bag, which explains why dialysis time is such an important determinant of capsid dynamics. Optimization of the SEC protocol for studying the self-assembly dynamics during size reduction As during initial dynamics experiments large quantities of dimers were observed in the SEC chromatograms when pH 7.5 buffer with 100 mM NaCl was the eluent, while these were never observed for VW1-VW8 ELP-CCMV before, the origin of these dimers was evaluated. Hereto, a 100 µM VW1-VW8 ELP-CCMV coat protein solution in pH 5.0 buffer was dialyzed (MWCO 12-14 kDa) to pH 7.5 buffer with 100 mM NaCl at 4 °C overnight and subsequently spiked with known amounts of native ELP-CCMV dimers in the same buffer. DLS and native PAGE were employed to analyze the capsid-dimer mixtures. From the DLS results in Figure S3A and B, it becomes clear that DLS is not sensitive enough to detect dimers in capsid-dimer mixtures, which could be explained by the high scattering of the capsids overpowering any light scattering caused by the much smaller dimers. Therefore, although no dimers are detected with DLS of the dialyzed capsids, this does not confirm that indeed no dimers are present in this capsid solution. We therefore focused on native PAGE analysis. From the results in Figure S3C, it can be appreciated that capsids and dimers can be easily distinguished from each other. Furthermore, based on the band intensities on the gel it can be stated that in the VW1-VW8 ELP-CCMV capsid solution less than 5 % dimers are present. This indicates that the large fractions of dimers that are observed in the SEC chromatogram are most likely the result of some disassembly of VW1-VW8 ELP-CCMV capsids taking place during SEC analysis, which might be caused by the extreme dilution (240 times) during the chromatographic procedure. S 11 As for data analysis purposes, the SEC chromatograms should provide the best representation of the assembly state of VW1-VW8 ELP-CCMV during dynamics, a protocol was developed for inhibiting capsid dynamics prior to SEC analysis. Hereto, a 100 µM VW1-VW8 ELP-CCMV solution in pH 5.0 buffer was dialyzed (MWCO 12-14 kDa) to pH 7.5 buffer with 100 mM NaCl at 4 °C overnight and subsequently incubated with various amounts of NiCl2 for 50 minutes (the duration of one SEC run) at room temperature. 6 From the results in Figure S4A and B, it can be observed that the addition of at least 0.2 equivalents Ni 2+ (relative to the amount of VW1-VW8 ELP-CCMV coat protein concentration) successfully reduces the number of dimers that are observed in the SEC chromatograms, without affecting the T = 3 : T = 1 ratios. With increasing amounts of Ni 2+ also some higher-order structures became visible on the SEC chromatograms (around 7 mL), indicating that high amounts of Ni 2+ alter the protein fractions in the VW1-VW8 ELP-CCMV solution. Therefore, the addition of 0.2 equivalents of Ni 2+ was found to be most suitable to reduce the number of dimers introduced due to dilution on the SEC column while not affecting the protein fractions. To confirm that the addition of this amount of Ni 2+ effectively stops any dynamics, a sample that was dialyzed for 30 minutes was subjected to SEC analysis after 50 minutes or 7 hours of incubation with Ni 2+ (Figure S4C,D). The resulting SEC traces and protein fractions were very similar, indicating that the addition of 0.2 equivalents of Ni 2+ effectively stops the dynamics and stabilizes the samples for prolonged storage prior to SEC analysis. Theory Based on previous experimental, theoretical and computer simulation results, we put forward that nucleation is the underlying mechanism for capsid assembly and disassembly. [7][8][9] To this end, we combine equilibrium theory, borrowed from the physics of supramolecular polymers, and classical nucleation theory. 10 This allows us to calculate the time evolution of the assembly and disassembly of mixtures of capsids, the predictions of which we compare with our experimental results. Note that the assembly kinetics of free subunits into competing capsids with different numbers has been discussed before 11,12 -a kinetic theory of T number conversion has not yet been attempted. S 13 To obtain the thermodynamic parameters required for our kinetic theory, we write the free energy of an aqueous solution in which only free ELP-CCMV subunits and the fully formed capsids are allowed to be present. In the equilibrium theory we ignore the intermediate states, as previous experimental and simulation work, as well as the findings presented in this paper clearly show that they are barely detectable (if at all) and represent short-lived states. 13 i.e., can never exceed * 1 = 1 or * 3 = 3 . Thus, for each type of capsid, there is a critical free protein concentration * below which the concentration of capsids is almost zero as the number of subunits in the capsids, , is large compared to unity. For CCMV, the basic protein subunits are dimers, so 1 = 30 for the = 1 and 3 = 90 for the = 3 capsid. S 14 Using the equilibrium theory described above, we can now set up the kinetic theory of capsid assembly and disassembly within the framework of CNT. 8 The Gibbs free energy of the formation of an incomplete spherical capsid of the species containing = 1, . . . , molecules with a circular rim can be written as / is a dimensionless magnitude of the rim energy, with the radius of the shell and the free energy cost per unit length of the rim. 10 can be estimated as = − where ∈ [0,1] is a geometric factor indicating the average fraction of bonds that a subunit on the rim is missing, which depends on the local coordination number and roughness of the rim. is the effective diameter of a protein unit that is approximated as a disk. Assuming that the surface of a fully formed capsid is covered entirely by capsid proteins, the effective diameter can be written as = ( 2 / 2 ) = * = √ (1 + 2 ) 3/4 is the so-called Zeldovich factor that describes the sharpness of the free energy barrier and that may be interpreted as a measure of the lifetime of the critical nucleus of size * . 13 The attempt or attachment frequency * of the monomers attaching to the critical nucleus depends on the mode of attachment, and may be , e.g., a function of the diffusivity and concentration of the free monomers, the size of critical nucleus, and on some internal molecular time scale associated with the docking process that may depend on conformational switching. 20 For simplicity, we assume that it does not depend on the size of the clusters nor on the concentration. To model the disassembly process, we presume that the initial state constitutes a fully formed where , * represents the free energy barrier for the disassembly of a shell to form monomers. Notice that the dissociation rate depends on , the capsid concentrations of species = 1,3. We shall presume that the attachment frequency associated with the association process is the same as that of the dissociation process, as it describes the same process and we presume it to be independent of the size of the critical nucleus. 21 Because capsids with different numbers have different radius of curvature, we do not allow for a direct transition from one number to another one. In our allowed reaction path pathway, growth or disassembly can only proceed by the shedding or docking of individual protein subunits, which for CCMV constitute coat protein dimers. This is not a far-fetched reaction path, as our experiments show no indication of partially disassembled = 3 particles spontaneously morphing into = 1 particles, or vice versa, = 1 particles opening up to absorb subunits and growing into a = 3 particles. Hence, we presume that, first, one type of capsid disassembles into dimers, second, free dimers reassemble into different capsid sizes following their corresponding assembly nucleation rates. Presuming that kinetic processes are sufficiently slow to allow us to use the expressions for steady-state nucleation rates for association and dissociation, i.e., presume a quasi steady state It is important to realize that the time that it takes to change the and the salt concentration of the buffer solution might not be exactly the same in each experiment. In addition, the lag time for assembly and disassembly of capsids with different sizes are arguably different. Therefore it is difficult to pinpoint the actual "time zero" for each individual experiment. In order to deal with this uncertainty, we start collecting data 30 minutes after the experiment commences. We also assume that the lag times are negligible on the time scale of the experiment, thus we ignore the S 17 first phase of assembly in CNT in which capsids have not started to assemble or disassemble. 10 Numerics The kinetics equations predicted by CNT (Eqs. S7 and S8) are solved by using finite difference methods. Assembly and disassembly nucleation rates at the beginning of the simulation are determined by the initial conditions. The concentrations of capsids and free dimers are calculated at each time step, using the values and nucleation rates at the previous time step. Hence, our time-stepping equations read: Eq. (S12) where = \| , − , \| is the absolute assembly or disassembly rate of capsid size . The simulation continues until full depletion of the unfavorable capsid size. From equilibrium theory and experimental observations, we have to assume that there are some free dimers remaining in the solution before the quench, that is, before the induced shift in and in salt concentration that on the time scale of the experiment is (virtually) instantaneous. Therefore, we invoke a non-zero value as our initial free dimer concentration. Quickly after starting S 18 the simulation, the dimer concentration converges to a fixed concentration relatively close to what must be the smaller critical concentration. Having initially more dimers in the system leads to the fast formation of capsids. On the other hand, a low dimer concentration at the start of the simulation increases the initial disassembly rate. In order to avoid both of these conditions, we choose the initial dimer concentration near the concentration it converges to. It also helps us to avoid any divergence in the simulation as the dissociation rates increase significantly at low dimer concentrations. Based on the dimer concentration at the end of the experiment, which is relatively close to the critical concentration of the more stable species, we approximate the total dimer concentration is around 10 times larger than the critical concentration. Therefore, the overall protein in the unfavorable capsid we set at 10 * , where * = ( * 1 , * 3 ). (See table S4 and S5 for parameter values.) Due to the universality of the phenomena, the model is capable of reproducing the experimental results by using different binding energies. We decide to fix the binding energy of T=1 in our framework and generate experimental results only by changing the binding energy of T=3. This allows us to have a better comparison between the two types of experiment. The fraction of protein dimers in free solution and in capsids In the main text we show results of the fraction of proteins in the two types of capsid, , as we Figure S5), implying that the increase in the protein fraction in capsids is not due to the assembly of the free dimers initially present in the solution. This agrees with what is seen in the experiments, see Figure S7. This supports our suggestion that one capsid size disassembles into free dimers and that these proteins re-assemble into the other capsid size, and that the fraction is the relevant quantity describing the assembly and disassembly kinetics for the problem in hand. Fitting of the simulation results with respect to different reference points We have calibrated the simulation results using a reference point in the data series, as mentioned in the caption for figure 2C and 3C. Here we re-calibrated the same simulation data set with respect to a number of reference points to verify the robustness of our fitting procedure. We find that the curve fits depend only relatively weakly on the choice of reference point ( Figure S6). Unfortunately, our numerical implementation of Classical Nucleation Theory does not allow us to find the fundamental time scale, that is, the time scale associated with the attempt frequency. In spite of this, we are able to show that the disassembly and assembly of the two different capsid sizes can be explained by CNT. Table of parameters All parameters related to the simulation of disassembly of T=3 and assembly of T=1, and vice versa, discussed in the main text and used in our comparison with the experiments are tabulated below.	2022-07-07T06:18:09.994Z	2022-07-06T00:00:00.000	{ "year": 2022, "sha1": "4e9db5a19f83e2ac599ff618c79150d686a94f8e", "oa_license": "CCBY", "oa_url": null, "oa_status": null, "pdf_src": "PubMedCentral", "pdf_hash": "4401917b1b78fab00d99c748e802753b34f217a9", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
54706547	pes2o/s2orc	v3-fos-license	Joint impact of ERP systems and non financial performance indicators on corporate financial performance : Evidence from French listed companies Article history: Received September 18, 2014 Accepted 9 December 2014 Available online December 9 2014 This paper aims to examine the joint impact of Enterprise Resource Planning systems (ERP systems) and Non Financial Performance Indicators (NFPI) on corporate financial performance. Our study is based on a comparative analysis between firms that adopt ERP only, firms that use NFPI only and firms that combining both strategies (ERP and NFPI) during the period from 2001 to 2006.The implementation process remains highly uncertain. In fact, the use of Non Financial performance indicators is an important determinant of corporate financial performance. At the operational level, combining ERP systems with NFPI reflects a long-term business strategy to improve business process. In summary, the ERP and NFPI literatures demonstrate the vital importance of aligning business process, information technologies and key performance indicators with the strategic objectives of the firm. Results support the hypothesis in which firms that combining ERP and NFPI have significantly higher ROA than either ERP-only or NFPI-only firms. Growing Science Ltd. All rights reserved. 4 © 201 ERP systems Non Financial performance Indicators Integrated information systems TQM BSC Introduction This paper investigates whether the corporate financial performance increases following the jointly adoption of Enterprise Resource Planning (ERP) systems and Non Financial Performance Indicators (NFPI).In the last two decades, business environment appreciates the importance of information for the decision-making and the improvement of corporate performance (Drucker, 1995).The advent of information and communication technologies (ICT) enhances the business process by providing organizations with easy and fast access to relevant and real-time information needed to implement the appropriate strategy.Therefore, managers should revise their approaches to evaluate performance.Furthermore, excessive competitiveness and uncertain environment affect performance measurement systems.Today's, the performance valuation is based not only on financial numbers, but also on nonfinancial data (Eccles, 1991).Prior studies show that non financial indicators are an important determinant of corporate financial performance (Kaplan andNorton, 1992, 1996;Ittner andLarcker, 1995, 1998;Banker et al., 2000;Said et al., 2003;Wier et al. 2005;Wiersma, 2008).Hunton et al. (2003) argue that managers concerned to have a management oriented towards the future must implement an integrated information system that will used to trace the past.There are many companies those already vacant traditional information systems to implement integrated information systems, such as ERP systems.Kumar et al. (2002) define the ERP as a "computer application that aims to ensure the integration of the management of the business processes by automating the flow of materials, information and financial resources of the company within any functions in the organization by using a unique database".Therefore, ERP systems are management tools that provide the integration within the organization.Laframboise and Reyes (2005) indicate that ERP as well as an opportunity to reengineering through its ability to track and monitor transactions.Woo (2007) suggests that organizations invest in ERP systems to gain a variety of operational, managerial and strategic benefits.The impact of ERP adoption on financial performance is the main objective of several studies management accounting and information systems (Poston and Grabski, 2001;Hayes et al., 2001;Hunton et al., 2003;Hyvönen, 2007;Wang & Zhang, 2006;Wier et al., 2007). ERP system is a client-server architecture that ensures the decision-making.Grunland and Malmi (2002) argue that ERP is an agent of change within the organization.The implementation of such systems is a long-term strategic investment.Indeed, the use of non-financial performance indicators (NFPI) motivates managers to consider the importance of the behavioural aspect through the organization in the process of highlighting the performance measures (Kaplan & Norton, 1992).In the area of NFPI, prior literature focuses on Balanced Scorecard (BSC), Total Quality Management (TQM), Customer Satisfaction (CS), strategic human resource management and technological innovation.Furthermore, the joint adoption of ERP systems and NFPI is a strategic decision that aims to improve the company's financial performance (Wier et al., 2007). This study contributes to the extant literature by combining two long-term strategic investments and investigates whether the jointly effect of ERP systems and NFPI increases the corporate financial performance.Our empirical analysis is based on comparative study between ERP-firms, NFPI-firms and firms that jointly adopt ERP and NFPI.The sample consists of 102 French listed companies. The remainder of the paper is organized as follows.In the next section we describe the literature review and hypotheses development.Section 3 interested about research design.Section 4 presents results and discussion.Section 5 offers conclusions and future research. ERP systems and financial performance ERP systems are a major technology investment (Hendricks et al., 2007).The implementation of ERP systems ensures the integration between the business processes and information technologies in a set of synchronized procedures and applications that allow to gain a competitive advantage (Laframboise & Reyes, 2005;Wier et al., 2007).Furthermore, the ERP is expected to improve the company's financial performance (Davenport, 1998;Akkermans et al., 2003).Brignall and Ballantine (2004) argue that the ERP adoption is a strategic choice that considers the organization's life cycle complexity and the characteristics of the internal and external business environment.Jensen and Meckling (1976) suggest that the disparity of interests and the information asymmetry between the principal and the agent enhance the conflict of interest into the organization.To reduce agency costs that affect the company's performance, managers should be able to establish a strategic management system.Wang and Zhang (2007) showed that the ERP system reduce information asymmetry. The use of ERPS may result in production and quality improvements in key business areas, such as product reliability, customer service, and knowledge management.Hayes et al. (2001) indicate that the announcement of the implementation of ERP systems has a positive effect on the investors' reaction in the financial market.Therefore, the ERP adoption increases the firm's value and improves its performance.In the same field, Hunton et al. (2003) argue that the company's decision to adopt an ERP system has a significant effect on the strategic behavior of investors.They add that the use of ERP may improve product reliability, customer satisfaction and total quality management.Poston and Grabski (2001) investigate the impact of ERP on the financial performance through a sample of firms that adopt the ERP systems.They found no effect on financial performance.Hunton et al. (2003) conduct a comparative study between 63 ERP adopters and 63 non adopters firms.The authors use the Return On Asset (ROA), Return On Investment (ROI), Return On Sale (ROS) and Asset Turnover (ATO) as a proxy of financial performance.They show that non-ERP firms experience decreases the performance; nonetheless, ERP adoption improves the financial performance.Overall results find that the ERP adoption has a positive and significant effect on the company's financial performance.Similarly, Wier et al. (2007) suggest that the implementation of ERP systems increases the firm's financial performance.They argue that the ERP systems are at the origin of competitive advantage.Therefore, the first hypothesis H1 is presented as follows: Hypothesis H1: The ERP adoption has a positive and significant effect on the company's financial performance. Non-financial performance indicators (NFPI) and financial performance Competitive advantage originates and grows from relationships, competences and knowledge (Kaplan & Norton, 1992).Eccles (1991) argues that non-financial performance indicator is the keystone of strategic management systems.Several studies investigate the rule of using NFPI in the companies' business process (Hughes, 2000;Banker et al., 2000;Ittner & Larcker, 1995;Ittner & Larcker, 1998;Said et al., 2003;Hassabelnaby et al., 2005;Kaplan & Norton, 1992, 1995;Behin & Riley, 1999;Wier et al., 2007;Kallunki et al., 2011).For many companies, management control systems are built based on financial measures.Thus, the emphasis most companies set on financial measures leaves a gap between the development of a strategy and its implementation.Consequently, managers have to relate the financial indicators and non-financial performance indicators to linking long-term strategic objectives with short-term actions (Kaplan & Norton, 1996).Wier et al. (2007) indicate that the NFPI contribute to align management process, strategy and performance.The use of NFPI reflects a longterm business strategy (Ittner et al., 1997).Prior literature investigates whether the NFPI increase the companies' financial performance (Chenhall, 1997;Perera et al., 1997;Ittner & Larcker, 1998;Bihen & Riley, 1999;Said et al., 2003Anderson et al., 2007).Amir and lev (1996) demonstrate that non-financial performance measures are more informative to investors than financial accounting measures in the wireless communication industry.Similarly, Riley et al. (2003) find that non-financial performance metrics is more value relevant than traditional accounting and financial metrics in the airline industry.Chenhall (1997) examines the impact of TQM on performance and show that the implementation of TQM has a positive and significant impact on the organizational performance.Anderson et al. (1997); Perera et al. (1997); Behin and Riley (1999) emphasize the importance of customer satisfaction in the improvement of the company's performance.Ittner and Larcker (1998) suggest, via a study in which measure the influence of customer satisfaction on the company's market value, that the use of non-financial measures improves the firm's stock returns.Said et al. (2003) and Wier et al. (2007) associate the use of NFPI and the company's performance and found that the NFPI is positively and significantly related to financial performance. Prior literature demonstrates that the use of non financial information is at the origin of sustainable competitive advantage, improves the long term financial performance and linking long-term strategic objectives with short-term actions.Based on prior studies, our second hypothesis H2 is presented as follows: Hypothesis H2: The use of non-financial performance indicators has a positive and significant effect on the company's financial performance. Combining ERP and NFPI Literature in managerial accounting and information systems stress the great importance to combining strategies.The combination ensures the strategic alignment of business process and information technology with the firm's long-term strategic objectives (Kaplan and Norton, 1996).The combination of ERP systems with the non financial performance indicators (NFPI) is a response to the needs of managers to have a high degree of monitoring into the operational, managerial and strategic level.Indeed, by combining the ERP and NFPI, managers aim to mitigate the information asymmetry and to enhance the corporate financial performance.The effect of the joint adoption of ERP systems and NFPI on corporate performance has been validated empirically by Laframboise and Reyes (2005) and Wier et al. (2007).Findings found that the combined adoption of ERP systems and NFPI is positively and significantly related to financial performance.This result is more relevant than the ERP adoption only or the use of NFPI only.Thus, our third hypothesis H3 is presented as follows: Hypothesis H3: The joint adoption of ERP systems and NFPI has a positive and significant effect on the company's financial performance. Sample and data Table 1 reports the sample selection.Data was collected from the web-site www.amf-france.orgby examining the annual reports of firms.Our sample consists of French listed companies for the period from 2001 to 2006.We divided the sample into three sub-samples.The sub-sample 1 contains 31 French listed companies which adopt ERP only.The sub-sample 2 contains 33 French listed companies that use non-financial performance indicators (NFPI).The sub-sample 3 consists of 38 companies that jointly adopt ERP and NFPI.ROA: Return On Assets of firm i at time t; ERP: Implementation of ERP by firm i at time t; NFPI: Use of non financial performance indicators of firm i at time t; INT: Interaction variable (ERP × NFPI) of firm i at time t; LEV: Measured using debt-equity ratio of firm i at time t; SIZE: Size measured as the log of total assets of firm i at time t; GROWTH: Growth of firm i at time t; IROA: The industry average of change in ROA of firm i at time t; ACTVOL: The standard deviation of annual return on assets over the previous 5 years of firm i at time t; ε: Error term. Econometric modeling We estimate the following models to test, respectively, the hypotheses H1, H2 and H3: (3) Dependent variable: Financial performance The company's financial performance measure has always been very sensitive in empirical research.Indeed, several measures of performance were developed in the literature.In our study, we retain the Return On Assets (ROA) as a proxy for the variable "financial performance": s TotalAsset Netincome ROA  Independent variables a. ERP systems adopt The ERP system is a dummy variable that takes the value of "1" if the company adopt it and "0" otherwise. b. Use of non-financial performance indicators (NFPI) The NFPI is a dummy variable that takes the value of "1" if the company use it and "0" otherwise. c. Interaction variable (ERP x NFPI) To examine the joint impact of ERP adoption and use of non-financial performance indicators on financial performance we use an interaction variable (ERP x NFPI).This variable takes the value of "1" if the company adopts the two variables at same time and "0" otherwise. Control variables Control variables are leverage, size and growth.Leverage is measured by the Debt to Equity ratio.The size of the company is measured by the natural logarithm of total assets.The Growth is measured by divided the market value of equity plus book value of debt by book value of assets at the beginning of the year. Multivariate analysis In the multivariate analysis, we run two econometric tests namely, the homogeneity test and the Hausman test for the three equations Eq. (1), Eq. ( 2) and Eq.(3). Homogeneity test Table 2 presents the results of the homogeneity test.The existence of specific effects leads to choose the estimation methods.If it's a fixed effect, we retained the within method, if it's a random effect we retain the GLS method.In counter-point, if there is no specific effect, we use the OLS method.The run of this test to the three econometric models shows the existence of specific effects. Hausman test Results of Hausman test are shown in Table 3.To specify the types of retained effects (are fixed or random effects) we computing the Hausman test.This test applied to the three models suggests that the retained effects for all equations are random effects.Then, we use the GLS regression to the Eq. ( 1), Eq. ( 2) and Eq. ( 3). ERP and financial performance To test our first hypothesis H1, we use GLS multivariate regression model.Table 4 reports the results of an analysis regressing ROA on ERP.Our multiple regressions want to investigate the effects of ERP adoption on financial performance.Following the estimation of the equation Eq. (A.1), we find that the adjusted R-square is in the order of 0.1718.This result indicates that the explanatory variables contribute to the explanation of the ROA (proxy of financial performance) at the proportion of 17.18% and the model is globally significant (F-statistic = 3.474; p-value = 0.000).Then, we observe that the coefficient of ERP is positive and statistically significant at the 5% level (β1 = 0.039; t-test = 2.30).This matter stipulates that a one-unit increase in the ERP investment means an improvement of the financial performance by 0.039 units.Among the control variables, the coefficient of LEV is negative and not significant.The SIZE has a positive and significant coefficient at the 5% level (β3 = 0.025; ttest = 2.579).Indeed, the coefficient of the variable GROWTH is negative and statistically not significant. Overall results stress the vital importance of the implementation of ERP systems in the organization.ERP adoption supports business process improvements and enhances decision-making and financial performance of French listed companies.Thus, presented by its characteristics such as integration, standardization, and centralization, the implementation of ERP systems is appreciated by managers since it improves the management control systems and mitigates the information asymmetry.Among the variable SIZE, findings demonstrate that the company's size is an important determinant of the strategic choice to invest in ERP projects a fact that the ERP implementation cost is too expensive.The absence of the relationship between leverage (LEV) and financial performance may be explained by the fact that innovative companies consider that the adoption of an integrated information systems is a vital investment that need funding with debt.Results confirm our hypothesis H1.Our findings are consistent with prior studies of Hunton et al. (2003) and Wier et al. (2007). NFPI and financial performance To test our second hypothesis H2, we use GLS multivariate regression model.Table 5 presents the findings of an analysis regressing ROA on NFPI.Our multiple regressions to investigate the effects of the use of NFPI on financial performance.Following the estimation of the equation Eq. (A.2), we find that the adjusted R-square is in the order of 0.5513.This result indicates that the explanatory variables contribute to the explanation of the ROA at the proportion of 55.13% and the model is globally significant (F-statistic = 7.835; p-value = 0.000).Then, we observe that the NFPI has a positive and statistically significant coefficient at the 5% level (β1 = 0.010; t-test = 2.10).This matter stipulates that a one-unit increase in the use of NFPI reflects an improvement of the financial performance by 0.010 units.Among the control variables, the coefficient of LEV is negative and not significant.The SIZE has a positive and significant coefficient at the 1% level (β3 = 0.021; t-test = 2.84).Then, the coefficient of the variable GROWTH is positive and statistically significant at 5% level (β4 = 0.025; ttest = 2.33). Overall results indicate the vital importance of the use of NFPI in the organization.In fact, the nonfinancial indicators reflect a long-term business strategy.By the implementation of NFPI, managers have to relate the financial indicators and non-financial indicators with the aim to ensure the connection between long-term strategic objectives with short-term actions.Among the variable SIZE, findings demonstrate that the company's size is an important determinant of the strategic choice to use NFPI in management process.Considering the positive and significant relationship between GROWTH and financial performance, it is worth noting that the use of NFPI enhances the company's growth rate which is reflected on the company's financial performance.Findings confirm our hypothesis H2.Our results are consistent with prior studies of Laframboise and Reyes (2005) and Wier et al. (2007). Interaction variable (ERP x NFPI) and financial performance To test our hypothesis H3, we use GLS multivariate regression model.Table 6 summarizes the results of an analysis regressing ROA on the interaction between ERP and NFPI.Our multiple regressions want to investigate the effects of the joint adoption of ERP systems and the use of NFPI on financial performance.Following the estimation of the equation Eq. ( 3), we find that the adjusted R-square is in the order of 0.3606.This result indicates that the explanatory variables contribute to the explanation of the ROA at the proportion of 36.06% and the model is globally significant (F-statistic = 9.643; p-value = 0.000).Then, we observe that the ERP has a positive and statistically significant coefficient at the 5% level (β1 = 1.30; t-test = 2.06).The coefficient of is significantly higher in the Eq. ( 3) than the coefficient of this variable in the Eq. ( 1) and the difference is statistically significant at the 1% level.Then, following the estimation of Eq. ( 3), the coefficient of NFPI is positive and statistically significant at the 1% threshold (β2 = 1.008; t-test = 3.05).The coefficient of the variable NFPI in the Eq. ( 3) is larger than this coefficient in the Eq. ( 2) and the difference is statistically significant at the 5% threshold.The coefficient of the interaction between ERP and NFPI is positively and significantly related to the ROA (β3 = 1.96; t-test = 3.63).Among the control variables, the coefficient of LEV is negative and not significant.The SIZE has a positive and significant coefficient at the 5% level.Then, the coefficient of the variable GROWTH is positive and statistically not significant.Overall results support our hypothesis H3 stipulating that the joint adoption of ERP systems and NFPI has a positive and significant effect on the company's financial performance.Findings are consistent with the studies of Hunton et al. (2003); Laframboise and Reyes (2005); Wier et al. (2007). Conclusions and future research This paper examine whether the joint adoption of ERP systems and non financial performance indicators increases the company's financial performance.Our study is based on a comparative analysis between firms that adopt ERP only, firms that use NFPI only and firms that combining ERP and NFPI during the period from 2001 to 2006.The sample consists of 106 French listed companies.Results show that the ERP adoption only improves the financial performance of French listed companies.In addition, the use of non-financial performance indicators alone increases the firms' financial performance.Whereas, findings suggest that the combination of both strategies is more relevant in the explanation of ROA, as compared to either strategy alone. Combining ERP systems and non-financial performance indicators provides the advantage to linking the long-term strategic objectives with short-term actions (Kaplan and Norton, 1996).Future research should examine the value relevance of combining strategies in the business process.Indeed, it is vital important to associate financial numbers with non-financial data to improve the investors' makingdecision and to understand the complexity of business environment. Wier et al. (2007)ior study ofWier et al. (2007), we develop our econometric models.Authors use the return on assets (ROA) as a proxy of company's financial performance.The Wier et al. model is presented as follows: Table 2 Homogeneity test	2018-12-11T07:06:36.174Z	2015-01-01T00:00:00.000	{ "year": 2015, "sha1": "87e735f499826637d827cc6ec879c7175f62b8be", "oa_license": "CCBY", "oa_url": "https://doi.org/10.5267/j.msl.2014.12.004", "oa_status": "GOLD", "pdf_src": "ScienceParseMerged", "pdf_hash": "87e735f499826637d827cc6ec879c7175f62b8be", "s2fieldsofstudy": [ "Business", "Economics", "Computer Science" ], "extfieldsofstudy": [ "Psychology" ] }
58012282	pes2o/s2orc	v3-fos-license	Biological effects different diameters of Tussah silk fibroin nanofibers on olfactory ensheathing cells Transplantation of olfactory ensheathing cells (OECs) has potential for treating spinal cord and brain injury. However, they are void of an extracellular matrix to support cell growth and migration. Engineering of tissue to mimic the extracellular matrix is a potential solution for neural repair. Tussah silk fibroin (TSF) has good biocompatibility and an Arg-Gly-Asp tripeptide sequence. A small number of studies have assessed the effect of the diameter of TSF nanofibers on cell adhesion, growth and migration. In the present study, TSF nanofibers with a diameter of 400 and 1,200 nm were prepared using electrospinning technology; these were then used as scaffolds for OECs. The structure and morphology of the TSF nanofibers were characterized by scanning electron microscopy (SEM) and Fourier-transform infrared spectroscopy. An inverted-phase contrast microscope and SEM were used to observe the morphology of OECs on the TSF nanofibers. The effect on the adhesion of the cells was observed following crystal violet staining. The phenotype of the cells and the maximum axon length on the scaffolds were evaluated by immunostaining for nerve growth factor receptor p75. Cell proliferation and viability were assessed by an MTT assay and a Live/Dead reagent kit. The migration efficiency of OECs was observed using live-cell microscopy. The results indicated that a 400-nm TSF fiber scaffold was more conducive to OEC adhesion, growth and migration compared with a 1,200-nm TSF scaffold. The phenotype of the OECs was normal, and no difference in OEC phenotype was observe when comparing those on TSF nanofibers to those on PLL. The present study may provide guidance regarding the preparation of tissue-engineered materials for neural repair. Introduction Olfactory ensheathing cells (OECs) have an important role as seed cells in central nervous system (CNS) repair (1)(2)(3). OECs secrete a variety of neurotrophic factors, including nerve growth factors, as well as extracellular matrix molecules, to improve the microenvironment after nerve injury, reduce glial scarring and prevent neuronal apoptosis (4). Previous studies using animal models have also reported that OEC transplantation significantly promoted nerve fiber regeneration and partial functional recovery (5). However, at present, the efficacy of cell transplants in repairing the CNS is not ideal as inflammatory substances and glial scar formation has been demonstrated to inhibit the secretion of extracellular matrix components and cell scaffolds (6). In recent years, promising tissue-engineered biomaterial scaffolds have been demonstrated to have the capacity to improve the repair of CNS injury. The ideal biomaterials should have biodegradability, biocompatibility, excellent mechanical properties and flexibility (7,8). Natural biodegradable materials include chitosan, polypeptide hydrogel, poly L-lactic acid (PLL), chitosan/polyethylene stents and non-cellular scaffolds, which have been considered for the repair of nerve damage (9)(10)(11). Our group has been working on natural Tussah silk fibroin (TSF) biomedical materials for numerous years (12). TSF is a natural protein with a polymer structure. Its amino acid composition is characterized by a large number of arginine-glycine-aspartic acid (RGD) tripeptide sequences. This RGD tripeptide sequence structure is known to facilitate cell adhesion. TSF also has good biocompatibility, with no toxic effects on cells and organisms, and a low likelihood of inflammatory reaction or immune rejection (13)(14)(15)(16). In recent years, the use of this electrospinning technology has led to great progress in the preparation of tissue-engineered materials. The diameter, structure distribution, molecular conformation and crystallinity of TSF may be controlled using electrospinning technology (17). When the fiber diameter is Biological effects different diameters of Tussah silk fibroin nanofibers on olfactory ensheathing cells decreased from microns to nanometers, the material properties are improved accordingly. A previous study by our group indicated that TSF scaffolds prepared by electrostatic spinning have good biocompatibility with OECs and may support their growth and migration (12). However, the study also suggested that the diameter of nanoscale fibers affects the biological effects of OECs. Previous studies have demonstrated that the diameter of electrospun nanofibers has a significant effect on cell behavior (18)(19)(20). In one study, fibers with a diameter of 400 and 1,200 nm were prepared for evaluating their effect on cell behavior (21), and the results suggested that the diameter of the scaffold had a marked impact on neural cell behavior, with a significant increase in the cell-spreading area observed on 400-nm silk fibroin (SF). A significantly enhanced migration efficiency of astrocytes grown on SF scaffolds was demonstrated, which highlighted the effects of SF nanofibers to enhance cell migration. As it was indicated that the diameter of SF nanofibers may be an important factor in the construction of biomimetic microenvironments suitable for cell growth, the present study aimed to assess the ideal diameter of TSF as a tissue engineering scaffold material. Materials and methods Materials and sample preparation. TSF nanofibers were prepared as previously reported (22). In brief, TSF fibers were boiled in 0.5% (w/w) Na 2 CO 3 aqueous solution for 30 min twice for degumming, followed by thorough rinsing with sufficient deionized water to remove the glue-like sericin. The degummed TSF fibers were dissolved in 16 M lithium thiocyanate solution and the mixture was heated with hotplate stirrers at 50˚C for 1 h, followed by dialysis with cellulose tubular membranes in distilled water for 3 days (molecular weight cut-off, 8,000-14,000 kDa). A TSF film was prepared by spreading the TSF solution onto a polyethylene plastic board and leaving it to dry at room temperature. The electrospinning solution was prepared by dissolving the TSF film in hexafluoroisopropanol with hotplate stirrers for 1 week at 25˚C. The spin solution was transferred to a 10-ml syringe with an 18 G needle. The flow rate was 0.5 ml/h, the voltage was gradually increased to 16 kV and the collection distance was 12 cm. TSF nanofibers with diameters of 400 and 1,200 nm were generated using spin solutions with concentrations of 8 and 16wt%, respectively. The as-spun TSF nanofibers were immersed in 75% (v/v) ethanol/water for 30 min to induce a structural change, and then dried at room temperature for 24 h. Finally, the TSF nanofibers were X-rayed prior to use. X-ray diffraction experiments were measured on X'Pert Pro MPD (PANalytical B.V., Almelo, Netherlands) in transmittance mode to investigate the crystalline structure of samples at a wavelength of 0.154 nm. Lastly, the intensity of the incident beam, the sample absorption and the background were corrected for changes (23). Material characterization. The samples were cut into squares of 1 mm 2 and fixed on copper chips. After spraying with gold, they were observed using a scanning electron microscope (SEM) (23,24). In order to calculate the diameter of the nanofibers, SEM images of >100 independent fibers were analyzed by ImageJ software (version 1.47; National Institutes of Health, Bethesda, MD, USA). Fourier Transform Infrared Spectroscopy (FTIR) is a powerful and commonly used tool for secondary structure analysis; the conformational changes and secondary structure of TSF nanofibers after ethanol treatment were determined by FTIR (24,25). In vitro primary cultured OECs on TSF scaffolds. All animal experiments of the present study were approved by the Ethics Committee of the Second Affiliated Hospital of Soochow University (Suzhou, China) and performed in accordance with the Guidelines for the Welfare of Animals of Soochow University (Suzhou, China) (12) as well as the National Institute of Health's Guide for the Care and Use of Laboratory Animals (26). Primary cultures of OECs were prepared from 30 male Sprague Dawley rats (weight, 100-150 g; age, 4-5 weeks; Experimental Animal Center of Soochow University) as reported previously (12). The culture medium for OECs consisted of Dulbecco's modified Eagle's medium (DMEM)/F-12 (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.), 1% glutamine (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) and 2% penicillin-streptomycin (Harbin Pharmaceutical Group Co., Ltd., Harbin, China). The culturing condition was as follows: 37˚C in a humidified atmosphere containing 5% CO 2 . After 10 days in culture, OECs were used for a biocompatibility evaluation with TSF nanofibers. The OECs were detached with 0.1% trypsin (Sigma-Aldrich; Merck KGaA) at 37˚C for 10 min, centrifuged at 157 x g for 5 min and then resuspended. Subsequently, the OECs (1.0x10 5 ) were seeded on coverslips coated with 400-nm or 1,200-nm TSF fibers or PLL (Sigma-Aldrich; Merck KGaA) that had been pre-wetted with culture medium in 35-mm Petri dishes (Corning Inc., Corning, NY, USA), the dimensions of TSF fibers were 10x10x1 mm, the thickness was 1 mm; PLL served as a positive control. The nanofibers were incubated in 1 ml complete culture medium at 37˚C for 30 min prior to seeding. Cell suspension (1 ml) was placed on the fibers in each dish, and 1 ml culture medium was added after 4 h. The culture medium was changed every 3 days. After culture for 4 days, the samples were observed with an inverted phase-contrast microscope and SEM, using the same method of SEM as described above. Immunofluorescence staining and quantitative analysis for OECs on TSF scaffolds. OECs were seeded on the TSF nanofibers or PLL-coated coverslips as described above, and their identity was confirmed by immunostaining. Cells on the fibers and PLL were fixed for 30 min at room temperature with 4% paraformaldehyde (Zhongde Chemical Products Trade Co., Ltd., Tianjin, China) and then blocked for 30 min in PBS containing 3% bovine serum albumin (BSA), 0.2% Triton X-100 and 0.02% NaN 3 (v/v) (all from Sigma-Aldrich; Merck KGaA). The cells were incubated with rabbit polyclonal anti-nerve growth factor receptor (NGFR) p75 (cat. no. BA0514-2; Wuhan Boster Biological Technology, Ltd., Wuhan, China), diluted to 1:200 in PBS, at 4˚C overnight. Following three washes with PBS, cells were incubated with fluorescein isothiocyanate-conjugated goat anti-rabbit secondary antibodies (cat. no. BA1105; Wuhan Boster Biological Technology, Ltd.) diluted 1:70 in PBS for 1 h at room temperature. Following three washes with PBS, cell nuclei were stained with Hoechst 33258 diluted 1:100 in PBS at room temperature for 15 min. Cover slips were washed with PBS, mounted with 50% glycerin in PBS and imaged with an AF6000 fluorescence microscope (Leica Microsystems GmbH, Wetzlar, Germany). To evaluate the growth of OECs on TSF nanofibers, the spreading area and the longest cellular process of OECs on TSF nanofibers were quantified at 1, 4 and 7 days. In brief, at least 10 individual cells from three randomly selected fields were measured with ImageJ 1.47 software on the basis of NGFR p75 immunostaining. At least five coverslips were included for each experimental group. Experiments were repeated three times. Crystal violet staining for cell adhesion assays. The attachment test for OECs on TSF nanofibers was performed as previously described (27). In brief, 500 µl OEC suspension (1x10 5 cells/ml) was seeded into TSF nanofibers and a 24-well PLL-coated plate. Plates were coated with PLL by adding 1 ml 0.1% mg/ml PLL solution to each well at room temperature overnight. The PLL solution was aspirated and the plates were left to dry on a clean bench. OECs were incubated for 1 or 3 h in an atmosphere of 5% CO 2 at 37˚C. The loose and unadhered cells were removed. The wells were gently washed twice with PBS. The adherent cells were fixed with 15% formalin in PBS for 15 min. Once the fixation was completed, the wells were washed twice with PBS and stained with 0.05 g/ml crystal violet at 37˚C for 15 min. In each well, the number of cells was counted in five microscopic fields (magnification, x200) and images were captured under an Olympus microscope (Olympus Corp., Tokyo, Japan). MTT cell proliferation assay. The proliferation of OECs on TSF and PLL was examined at 1, 4, 7 and 10 days. At each time-point, 20 µl MTT (Sigma-Aldrich; Merck KGaA) dissolved in PBS at 5 mg/ml was added to each well, followed by incubation at 37˚C for 4 h. Then, the medium was discarded and 150 µl dimethyl sulfoxide (Sigma-Aldrich; Merck KGaA) was added to each well to dissolve the dark blue crystals with agitation for 10 min. The absorbance was measured at 490 nm with a microplate reader (iMark™ Microplate Absorbance Reader-168-1130; Bio-Rad Laboratories, Inc., Hercules, CA, USA). Live/Dead Kit cell viability analysis. The Live/Dead viability/cytotoxicity kit (cat. no. L3224; Invitrogen; Thermo Fisher Scientific, Inc.), including 0.5 mM calcein AM and 0.5 mM ethidium homodimer-1 (Ethd-1) dissolved in 1 ml DMEM/F-12, was used for quantitative cell viability analysis after 6 days of culture on PLL or TSF scaffolds in a 6-well plate. In live cells, calcein displays bright green fluorescence, while in dead cells, Ethd-1 binding to nucleic acids results in red fluorescence. Live/dead reagent was added to cells, followed by incubation at 37˚C for 30 min and subsequent washing with PBS twice and observation under a fluorescence microscope. In each well, ~3,000 cells were counted from randomly selected fields. Live-cell imaging for migration assay. The migration of OECs on TSF nanofibers was assayed by timelapse video microscopy. An environmental chamber was used to maintain optimal growth conditions for OECs at 37˚C. In brief, coverslips with cells were loaded and observed every 5 min using a 10X objective under a Leica DMI 6000 B microscope for a period of 400 min. The forward migration index (FMI) was calculated as the ratio of forward progress (the net distance a cell progressed) to the total path length (total distance a cell traveled through the field). Statistical analysis. Values are expressed as the mean ± standard error of the mean. Statistical analysis was performed using analysis of variance followed by a Dunn-Bonferroni post-hoc test for multiple group comparisons using GraphPad Prism 6.0 software (GraphPad Software, Inc., La Jolla, CA, USA). P<0.05 was considered to indicate a statistically significant difference. Results Morphology and structure of TSF. In order to assess the role of the nanofiber diameter in regulating cell behavior, TSF nanofibers of different diameter were prepared. As presented in Fig. 1A and B, the diameters of the two types of TSF nanofiber were 400±44 and 1,200±95 nm, respectively; the nanofibers exhibited a smooth surface with micro-sized pores to allow for cell proliferation and penetration. Although native silk fibroin is stable in water, the regenerated silk material is usually post-treated with a solvent to achieve stability. In the present study, 75% ethanol was employed to treat as-spun TSF nanofibers and induce this structural change. As presented in Fig. 1C, the FTIR spectra of TSF nanofibers exhibited absorption bands at 1,625 cm -1 (amide I), 1,520 cm -1 (amide II) and 1,265 cm -1 (amide III), attributed to a β-sheet structure. The crystalline β-sheet structure is the basis for the water stability and strength of the TSF nanofibers. Thus, water-stable TSF nanofibers with different diameters were successfully prepared. Morphology and structure of OECs on TSF. An inverted-phase contrast microscope was used to observe the OECs cultured on PLL and TSF (400 and 1,200 nm) after 4 days ( Fig. 2A-C). Most OECs presented with a bipolar or tripolar morphology, with excellent refraction. The cells on PLL were connected through cell processes. The OECs on TSF followed the fiber direction, and adjacent cells formed connections through neurites. As presented in Fig. 2D and E, SEM revealed that OECs grew on the surface and pores of the TSF microfibers after 4 days of culture. The surfaces of the cells produced different cellular processes, crossing the microporous surface to form intercellular connections. In addition, extracellular matrix production was observed in gaps in the material. Most of the OECs on the 400-nm TSF displayed flat cell bodies and elongated protuberances, with the cells tightly linked. The cellular connections of OECs on 1,200-nm TSF were fewer than those on 400-nm TSF. Immunostaining and growth parameters of OECs on TSF. NGFR p75 staining indicated that OECs on TSF retained their phenotype. Hoechst 33258 staining revealed that the nuclei of the OECs were round or oval; no chromatin condensation or apoptotic body formation was observed and the staining was uniform, indicating that there were no obvious apoptotic phenomena of OECs grown on TSF nanofibers of either diameter. As presented in Fig. 3, the difference in the area of OEC spreading on the two types of TSF nanofiber after 1 day was relatively small (0.9-1.5x10 3 µm 2 ). After 4 days, the area of cell spreading on 400 nm TSF fibers reached 3.8x10 3 µm 2 , which was significantly higher than that on 1,200 nm TSF (3x10 3 µm 2 ; P<0.05). Significant differences in the spreading area between the OECs grown on PLL or 400 nm TSF fibers and the 1,200 nm TSF fibers were obtained at 7 days (P<0.05). Quantitative analysis also indicated that the maximum process length of OECs on 400 nm TSF fibers was significantly longer than that of OECs on 1,200 nm fibers at 4 and 7 days (P<0.05). Cell adhesion of OECs on TSF. The cell adhesion on 400 nm TSF was greater than that on PLL or 1,200 nm TSF (P<0.05). The cell adhesion on PLL was the lowest (Fig. 4). This may be due to TSF containing more RGD sequences. Proliferation and viability of OECs on TSF. As presented in Fig. 5A-C, the morphology of OECs on TSF was normal. The majority of the cells were distributed along the fiber and evenly distributed. According to an MTT assay (Fig. 5D), the absorbance value in the 1,200 nm TSF group was significantly less than that in the 400 nm TSF and PLL groups at 4 days (both P<0.05), and the absorbance value of the 1,200 nm TSF group was significantly less than that of the PLL group at 7 days (P<0.05). However, there were no differences between the groups at 1 and 10 days. These results indicate that, compared with PLL, TSF has good biocompatibility and did not induce any marked cytotoxicity on OECs. The cell viability and death rates were not significantly different in the TSF groups (Fig. 5E). These results indicated that TSF nanofibers successfully supported the survival, growth and proliferation of OECs. Cell migration of OECs on TSF. Cell migration was analyzed by single-cell trajectories. The starting point of each cell migration was standardized to the intersection point of the X-and Y-axes. OECs on PLL migrated in random directions, whereas those on 400-and 1,200-nm TSF moved along the fiber. In order to measure cell motility, the FMI was calculated and analyzed (Fig. 6). Quantitative analysis suggested that the FMI of OECs in the 400-nm TSF group was significantly higher than that in the PLL group (P<0.05). Discussion It has been verified that TSF has good biocompatibility, permeability and biodegradability, with a low inflammatory response and other favorable qualities (28,29). TSF supports cell growth, proliferation and differentiation; in vivo studies have also reported that it facilitates tissue repair (30,31). A further advantage of TSF is that it promotes cell adhesion and migration through the RGD tripeptide sequence, a biometric signal (32). It has been reported that TSF was beneficial for the adhesion and growth of fibroblasts and bone marrow mesenchymal stem cells in a mouse model (33,34). The present study indicates that the diameter of nanofibers has an important role in the arrangement and specific biological behavior of cells. OECs were observed to be well-aligned on 400-nm TSF, while cells were randomly arranged on 1,200-nm TSF. This result suggests that a smaller diameter (400 nm) of TSF nanofibers may promote cell alignment when compared with a larger diameter (1,200 nm). This result is similar that of a previous study by our group on OECs cultured on fabricated SF scaffolds of different diameters (35). The RGD tri-peptide sequence structure of TSF has been demonstrated to facilitate cell adhesion (13). TSF has also been shown to possess good biocompatibility with no toxic effects on cells and organisms, and a low likelihood of inflammatory reaction or immune rejection (15). Through the current study, it was speculated that a smaller diameter of TSF nanofibers is more similar to the microenvironment created by the natural extracellular matrix, Figure 3. Effect of the TSF scaffold diameter on the growth and spread of OECs. Immunocytochemistry staining of OECs with anti-NGFR p75 antibodies (green) and Hoechst 33258 (blue). OECs were cultured on (A) PLL and (B and C) TSF nanofibers with a diameter of (B) 400 nm and (C) 1,200 nm, and were observed at 1, 4 and 7 days (scale bar, 25 µm) using a fluorescence microscope. (D) The spreading area and (E) the longest process of OECS was quantified at 1, 4 and 7 days. The area of OECs on the 400-nm TSF fibers reached 3.8x10 3 µm 2 , which was significantly higher than on 1,200-nm TSF, which reached 3.0x10 3 µm 2 . There were also significant differences in the spread area of cells between PLL fibers and 1,200 nm TSF fibers at 7 days. In addition, the longest process of OECs on 400-nm TSF was significantly longer than that of OECs on 1,200-nm TSF on days 4 and 7. * P<0.05. TSF, Tussah silk fibroin; OECs, olfactory ensheathing cells; PLL, poly-L-lysine. which is suitable for cell survival and proliferation. However, the specific biochemical effects and directional guidance of TSF nanofibers, as well as the underlying mechanisms, still require further study. Based on the above results, it may be speculated that the precise control of the direction of nanofibers may control the arrangement and directional migration of cells to promote the regeneration of axons and the repair of central nerve injury. This may be facilitated by the application of physical or chemical stimuli, including electrical stimulation or growth factor release (36). In conclusion, through electrospinning, a three-dimensional TSF scaffold material with a controllable diameter, smooth surface and uniform pore spacing was prepared. These TSF nanofibers do not affect the phenotype of OECs, and support their adhesion, migration, growth and proliferation. The performance of the 400-nm TSF fibers, including OEC adhesion, proliferation and migration, was improved relative to that of the 1,200-nm TSF fibers. This indicates that TSF nanofibers with a diameter of 400-nm may be a superior scaffold material for repairing nerve injury. However, the mechanisms of this effect require further study. The FMI in the 400-nm TSF group was significantly higher than that in the PLL group. Scale bar, 100 µm. * P<0.05. TSF, Tussah silk fibroin; OECs, olfactory ensheathing cells; PLL, poly-L-lysine; FMI, forward migration index.	2019-01-22T22:34:04.694Z	2018-11-06T00:00:00.000	{ "year": 2018, "sha1": "c694faf75823392ca5b6bb7ebb20c50377facdd8", "oa_license": "CCBYNCND", "oa_url": "https://www.spandidos-publications.com/10.3892/etm.2018.6933/download", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "c694faf75823392ca5b6bb7ebb20c50377facdd8", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [ "Chemistry", "Medicine" ] }
114669856	pes2o/s2orc	v3-fos-license	The School We Want: Evolution or Change? A Consideration about Italian School This article is the result of an experimental research on the syllabuses that have been actually taught in Pisan high schools in the school year 2011-12. Highlighting four aspects heterogeneity, comparative studies, importance of Dante, importance of twentieth century literature I will attempt an analysis of the direction taken by Italian schools after the last educational reform and the remaking of school curricula. Premises of the Analysis The didactics of literature represents an increasingly investigated and current field, whose analysis has become more and more urgent in Italy after the institution of the so-called TFA 2 . In September 2013, during the ADI 3 conference in Rome, I gave a talk based on the analysis of texts examined in Pisa's high schools during the school years 2011-2012 4 , which highlightened a strong heterogeneity in the schools' syllabuses for Italian literature, sometimes even among courses offered by the same school. This study aims to be a continuation of those initial graphs, a more detailed analysis, and above all an attempt to contribute to these years' thriving discussion. I will start by summing up briefly my previous work, from which the current study moves on. I considered the texts read in five Pisan high schools (each specializing in either in humanities, foreign languages and socio-pedagogy, arts, and science with either an artistic or technical curriculum), for a total of 25 third years, 27 fourth years and 26 fifth years 5 . After a somewhat difficult recovery of the sources, that is the programs signed by both teachers and student representatives and handed in to the administrative office at the end of the school year, I proceeded to an analytic scan of the texts according to the following model. -Mathematical competence and basic competence in science and technology -Digital competences -Learning to learn -Social and civic competences -Sense of initiative and entrepreneurship -Awareness and Cultural expressions Eventually, the competences we were used Eventually, the good old study of fundamental notions of literature (facts, dates, names, titles etc.), at times sadly mistaken for superficial knowledge by teachers and students alike, have given way to goals to achieve. The risk is that these macro-areas prove to be empty boxes to be filled with whatever we have and whatever we want rather than what is actually needed. The argument implies important consequences. First of all, from the teacher's training's point of view which, with the reopening of TFA and the perspective of the qualification of teachers through high school leaving diploma alone, is more actual than ever. One need only think that TFA currently plans to dedicate more hours to pedagogic teaching than to disciplinary teaching. As a result, the kind of teachers we are preparing will be more likely to take care of the students' personality rather than of their disciplinary competence. A more than questionable choice 13 , especially considering that the first can easily be protected through the latter. The attention towards disciplinary competence should increase together with the teacher's autonomy, and therefore it should be -in the author's opinion-the main focus of the teacher's training. It goes without saying that the more autonomy teachers have, and therefore the more decisional power they have, the more accurate their disciplinary training should be. In order for a teacher to autonomously determine which authors their students will study, according to a personal criterion that includes some authors and excludes others, they should first of all know their subject in depth in order to be responsible enough to respond for their choice. This should be an obvious fact, which would not only preserve local literature, but also classes: as all teachers know, every class is different and has different needs. Nevertheless, it is necessary to keep in mind the other side of the coin. If, in order to (preserve) pay homage/privilege the local literature of Sardinia, for instance, Grazia Deledda were chosen as one of the compulsory authors to study, everyone would approve. However, if she were chosen over Pirandello, this would be a questionable decision regardless of context. And while Grazia Deledda is, after all, a Nobel prize winner, the priority could be given (and often is given) to less influential authors. There is in fact a general meaning, a value that us, as teaching and hermeneutic community, believe should be safeguarded. There are some authors we believe are fundamental 14 . Hence the intention of ministerial 13 On the actual preparation of TFA teachers, see this interesting article: C. Carminati (2015, February), "Leopardi non era pessimista. Quello che sanno e non sanno i futuri insegnanti", Internazionale. Retrieved from http://www.internazionale.it/opinione/clizia-carminati/2015/02/02/leopardi-non-era-pessimista-quello-che-sanno-enon-sanno-i-futuri-insegnanti 14 As Romano Luperini would say, although refering principally to the choices of twentieth century literature: "Se autonomia alle scuole significa far largo a qualche scrittore locale del Novecento (come suggerisce qualche indirizzo ministeriale) quando non si conoscono quelli dotati di valore nazionale e internazionale, ciò può solo accrescere il caos, la mancanza di vincoli comunitari, lo spappolamento delle società. L'autonomia ha senso solo se è sorretta da un punto di riferimento centrale" ["If schools' autonomy means to make room for local twentieth century authors (as some ministerial guidelines suggest) when students do not know the nationally and internationally influencial one, this could be a source of extra confusion, of loss of communal links, of the melting of society. Autonomy makes sense only if it is supported by a central reference system"], R. , "La questione del canone, la scuola e lo studio del Novecento", Un canone per il terzo millennio, cit., p. 168. Again, Maria Teresa Sarpi, referring to the uncertainty of twentieth century canon as contemplated by older guidelines: "Non è da sottovalutare il pericolo degli spontaneismi troppo disinvolti nelle scelte dei docenti e soprattutto dello smarrimento di valori condivisi e della mancanza di senso dell'attività didattica in campo letterario per pratiche disorganiche, frammentarie, prive di valenza conoscitiva e formativa per la persona e per il cittadino, che hanno radici in questa comunità e in questa realtà storica" ["It is not to underestimate the risk of thoughtless spontaneous choices by the teachers, and especially the risk of losing communal values and the meaning of the didactics of literature in favor of disorganized, fragmentary politics, which lack educational and cognitive value for the person and citizen, who has root in this guidelines and the debate that followed due to the exclusion of authors, who, until then, had been canonical. The issue is that, being these indications absolutely generic, the authors' value is not preserved at all. It is an open discussion. So those who wrote the announcement for the teachers' examination in 2016 seem to disagree with the choices made by their colleagues in 2012, to the point where they include Vittorini and Carducci 15 among the authors' future teachers should study, names not present in the guidelines for students. This could be justified if the teacher was well-prepared on a much wider range of authors, in order to handle the autonomy mentioned earlier. What cannot be justified is the opposite instance: Rebora and Meneghello were suggested by National Indications in 2012, but they were not included in the syllabus for teachers' examinations in 2016. Future teachers therefore must include in the program (as the Ministry states explicitly) two authors they were not required to know when they became teachers. A bizarre situation. The graph shows that 100% of the third year classes read Dante, versus 59,25% of the fourth years and 26,92% of the fifth years. About Dante Year after year, students who read Divine Comedy diminish by a half. It is an important number if we consider that only the fifth year's program is required for passing the high school exam. It means three out of four students will not be examined on Divine Comedy. Needless to say, we are far from the 25 cantos National Indications suggest are covered throughout the three years 16 community and in this historical setting"], M. T. Sarpi (2001), "La letteratura del Novecento a scuola", Un canone per il terzo millennio, cit., pp. 192-193. 15 Exclusions that were decided in a not always peaceful climate. See for example the case of Carducci in L. Curti (2016), "Carducci nelle antologie e nel canone. Breve indagine su una cancellazione" in Antologie d'autore. La tradizione dei florilegi nella letteratura italiana, Atti del Convegno di Roma, 27-29 ottobre 2014, a cura di E. Malato e A. Mazzucchi. Roma, IT: Salerno Editrice, pp. 383-406. 16 From National Guidelines 2012: "E dentro questo quadro, di descrizione e di analisi dei processi culturali -cui concorrerà lo studio della storia, della filosofia, della storia dell'arte, delle discipline scientifiche -che troveranno necessaria collocazione, oltre a Dante (la cui Commedia sarà letta nel corso degli ultimi tre anni, nella misura di almeno 25 canti complessivi), la vicenda plurisecolare della lirica (da Petrarca a Foscolo), la grande stagione della poesia narrativa cavalleresca (Ariosto, Tasso), le varie manifestazioni della prosa, dalla novella al romanzo (da Boccaccio a Manzoni), dal trattato politico a quello scientifico (Machiavelli, Galileo), l'affermarsi della tradizione teatrale (Goldoni, Alfieri)" ["it is in this frame of analysis and description of cultural processes-to which the study of history, philosophy, art history scientific disciplines will contribute-that teachers will make room for the following fields: Dante (at least 25 cantos in three years from Divine Comedy), the plurisecular tradition of lyrical poetry (from Leopardi to Foscolo), the great time of narrative epic poetry (Ariosto, Tasso), different manifestations (the data can be easily deduced: the classes that only read Divine Comedy in the third year cannot have read 25 cantos unless that was the only thing they studied at all). To paraphrase Dante is not easy of course, and if Divine Comedy were more intelligible we would not have had eight centuries of glossaries specialised in it. The language used is difficult, especially in Paradise, certainly the most studied by that 26,92%. This is why it should be featured in the exam, although it is certainly not easy to make room for great Dante, considering how many authors the fifth year students are required to read. The discourse brings to a wider consideration on language. In an Italy where most children and teenagers do not read outside of school 17 , books in middle school and, incredibly, in high school, often report only paragraphed texts 18 . Not Dante or poets from the III century exclusively, but also relatively recent ones such as Manzoni are paraphrased. Unacceptable proposals, not only because it then becomes very hard, if not impossible, to instil the concept of "language matters" and diachronic variety, on which the National guidelines insist, in the student's mind; also because the method does not work: once more the problem is being avoided rather than solved with the proper strategy. It is a matter of principle if not only of method. It could be objected that historical texts are read in contemporary language in many other countries. My answer is: why should we always import the worst of other school systems? Teaching students to read in an historical language helps them to develop competences in problem solving; it helps them not to avoid an issue; it helps them to open their minds. It has become undoubtedly difficult to make students read an historical text. However, language is only partially the cause: students do not even read contemporary classics. A solution could be to introduce intermediate texts as part of a gradual process towards classics in the original language. It is a much harder path for both students and teachers, compared to choosing a text by Boccaccio translated into contemporary Italian, of course. It is also much fairer. May I also be allowed to make another, this time strictly personal, reflection. When I was in middle school -less than fifteen years ago -we had to read the integral version of I Promessi Sposi in Manzoni's original language. Not only we were required to know five or six cantos from Inferno by heart, we also had to do a grammatical, logic and sentence analysis of some of Leopardi's Canti. Has the world changed this much, during the last fifteen years? Comparative Analysis As we have seen, knowledge of foreign languages is one of the competences suggested by Europe in 2006. Our students will be required to reach a B1 level in a chosen language by the end of their studies. Moreover, as a result of the reorganization of high schools in 2010, CLIL (Content and Language Integrated Learning) was introduced. This is just to add more technical data, since globalization and the formation of a supranational entity such as the European Union would be enough to imagine, as years go by, an increase in comparative studies. It is not so. We chose as a point of reference one of the most used manuals of Italian literature, Baldi-Giusso 19 , specifically the volume that is chronologically closer to us, in its 2012 and 1994 editions. We find two chapters dedicated to foreign literature in the most recent of prose, from short stories to novels (Boccaccio, Manzoni), political and scientific treaty (Machiavelli, Galileo), theatre (Goldoni, Alfieri)"]. 17 See: R. Giovannini (2016, May 09), "Metà dei ragazzi italiani non ha letto neanche un libro al di fuori della scuola", La Stampa. Retrieved from http://www.lastampa.it/2016/05/09/societa/met-dei-ragazzi-italiani-non-haletto-neanche-un-libro-al-di-fuori-della-scuola-neAVu9Mwlq1XXd4abFH07K/pagina.html 18 Also Clizia Carminati signals the problem: "Deve, insomma, [il docente universitario] obbligare gli studenti a confrontarsi con il testo, il che significa confrontarsi con un intero la cui struttura spesso è portatrice di significato, e confrontarsi con usi linguistici che possono risultare oggi poco familiari. Ma senza far perdere la coscienza che di lingua italiana si tratta: in Italia non dovrebbe esistere (purtroppo, invece, esiste) un'edizione del Principe di Machiavelli con 'versione in italiano moderno' a fronte" ["[The professor] should force students to face the text, that is facing with a text body whose structure often carries meaning, and face linguistic customs that may be unfamiliar to contemporary speakers. All this without losing awareness of the fact that this is still Italian. In Italy we should not have (but unfortunately we have) editions of Machiavelli's Principe with a contemporary Italian translation"], see: C. Carminati (2015, February). 19 The percentage is not very high, especially in the third year class. We tried to make a further differentiation between Italian and foreign thematic path. A thematic path is in fact often part of a more global vision of literature, sometimes requiring comparative studies. This allows to draw attention to one figure: the foreign authors included in the thematic path of the third year class are about double the number of those included in the fourth year class, regardless of the general minor importance that is given to thematic development. The tendency, as we will see in the next graph, is confirmed in the fifth year classes as well, where the percentage of foreign authors in our (Novecento) is halved, making it an almost exclusive Italian Novecento. The discourse changes in the last year, where we no longer talk of a chronological or thematic paths, but of a percentage of XX-century authors studied. It will be helpful for this purpose to remember what the National Guidelines say: Considering the XX-century resonance in his work, and of the complexity of his position in European XIX-century literature, Leopardi shall be studied at the start of the last year. Always resorting to a real multidisciplinary program, the historical outline, that will cover from the Unification of Italy to the present day, establishes that the student is able to comprehend the relation of the literary system (genres, themes, styles, relationship with audience, new expressive methods), on one side with the course of events that bit by bit modified the social and political layout in Italy, on the other side with the phenomena that characterize modernity and post-modernity, both in and outside Europe. At the core of the syllabus should be authors and texts whose influence on the innovation of forms and genres between XIX and XX century was deeper, and that marked the paths along which poetry and prose would redefine their canons during the XX century. From this perspective, the happenings of poetry, which cannot be reduced to national issues, cannot but move from the work of Baudelaire and from the Italian reception of European symbolism which in Baudelaire has its roots. The influence of Pascoli and D'Annunzio on twentieth century literature makes them fundamental to study; as much as, for what concerns prose, the representation of the "real" in Verga and the breakdown of the novel in Pirandello and Svevo constitute crucial moments in the "tradition of Novecento". In the XX century and until now, the study of poetry, that will decisively start with Ungaretti, Saba and Montale, will contemplate an adequate knowledge of texts chosen among contemporary and subsequent lyric (for instance Rebora, Campana, Luzi, Sereni, Caproni, Zanzotto…). The study of prose, from neorealism to the contemporary period, will comprehend readings from significative authors such as Gadda, Fenoglio, Calvino, P. Levi, and can be integrated with other authors (for instance Pavese, Pasolini, Morante, Meneghello). 26 26 National Guidelines, 2012, "In ragione delle risonanze novecentesche della sua opera e, insieme, della complessità della sua posizione nella letteratura europea del XIX secolo, Leopardi sarà studiato all'inizio dell'ultimo anno. Sempre facendo ricorso ad una reale programmazione multidisciplinare, il disegno storico, che andrà dall'Unità d'Italia ad oggi, prevede che lo studente sia in grado di comprendere la relazione del sistema letterario (generi, temi, stili, rapporto con il pubblico, nuovi mezzi espressivi) da un lato con il corso degli eventi che hanno modificato via via l'assetto sociale e politico italiano e dall'altro lato con i fenomeni che contrassegnano più generalmente la modernità e la postmodernità, osservate in un panorama sufficientemente ampio, europeo ed extraeuropeo. Al centro del percorso saranno gli autori e i testi che più hanno marcato l'innovazione profonda delle forme e dei generi, prodottasi nel passaggio cruciale fra Ottocento e Novecento, segnando le strade lungo le quali la poesia e la prosa ridefiniranno i propri statuti nel corso del XX secolo. Da questo profilo, le vicende della lirica, meno che mai riconducibili ai confini nazionali, non potranno che muovere da Baudelaire e dalla ricezione italiana della stagione simbolista europea che da quello s'inaugura. L'incidenza lungo tutto il Novecento delle voci di Pascoli e d'Annunzio ne rende imprescindibile lo studio; così come, sul versante della narrativa, la rappresentazione del "vero" in Verga e la scomposizione delle forme del romanzo in Pirandello e Svevo costituiscono altrettanti momenti non eludibili del costituirsi della "tradizione del Novecento". Dentro il secolo XX e fino alle soglie dell'attuale, il percorso della poesia, che esordirà con le esperienze decisive di Ungaretti, Saba e Montale, contemplerà un'adeguata conoscenza di testi scelti tra quelli di autori della lirica coeva e successiva (per esempio Rebora, Campana, Luzi, Sereni, Caproni, Zanzotto, …). Il percorso della narrativa, dalla stagione neorealistica ad oggi, comprenderà letture da autori significativi come Gadda, Fenoglio, Calvino, P. Levi e potrà essere integrato da altri autori (per esempio Pavese, Pasolini, Morante, Meneghello).". shing the a dense, that they not read. erring to a hi è.» Così ima prova a poesia di n provano scolastico, nonostante vecento, il don't even ated when ool-leaving ent should nalyzed in amount of e canon of That they do not read books. That they do not buy them, spending their time watching TV instead of scrolling through the pages of a novel. Although we cannot generalize, this is a credible scenario 30 . Let us wonder, then, if reading texts closer to them, in both language and content, could help some potential readers to approach literature. Potential readers who would remain so, maybe even after school, maybe after university. And this should not substitute other readings, less close to students and maybe linguistically complex (such as highly entertaining but hard to understand Orlando Furioso), but only be a new path for those who already fell in love with literature during their school years, and a last hope for those who did not. Who says that someone who reads Una Questione Privata might not autonomously decide to pick up Ariosto? As Luperini states: "The study of (Novecento) is fundamental in order to better understand the great authors of the past, to ask them the most current, culturally stimulating questions. The problem is not quantitative but qualitative. To study (Novecento) means to acquire the sensitivity and culture of contemporaneity and be able to look at the past through the point of view and intellectual questions of the present. This does not mean just adding more authors, but studying those of the past from a different perspective. To study (Novecento) means to review the whole literary heritage from the perspective, the horizon of values, the urgency of the present, not to flatten it on contemporaneity or to just emphasize analogies, but also to be able to appreciate and value the differences. 31	2019-04-15T13:11:51.352Z	2017-05-18T00:00:00.000	{ "year": 2017, "sha1": "2d1f87562f9c50755b9114e072b05ba316f6a768", "oa_license": null, "oa_url": "https://doi.org/10.22606/jaer.2017.22002", "oa_status": "GOLD", "pdf_src": "MergedPDFExtraction", "pdf_hash": "1cff801484e179737421241e00f87b41e1b0addc", "s2fieldsofstudy": [ "Education", "Agricultural and Food Sciences" ], "extfieldsofstudy": [ "Sociology" ] }
151368443	pes2o/s2orc	v3-fos-license	Factors for Suicidal Ideation in Middle School Students by Gender Purpose: This study was conducted to identify the factors of suicidal ideation in middle school students by gender, and to announce the necessity of gaining upon a socio-cultural comprehension-based approach. Methods: Three middle schools were randomly selected from a pool of 21 middle schools in S City, and were stratified by school and grade. A total of 878 students were included in the analysis. Multiple logistic regression analysis was conducted to identify the factors that affect suicidal ideation by gender. Results: Suicidal ideation and depression scores of the female students were over twofold higher than those of the male students. The factor that was found to affect suicidal ideation in the male students was depression while for the female students the factors were visiting suicide-related websites, having friends who attempted to commit suicide, depression, and poor communication with one ’ s parents. Conclusion: The findings in this study support gender differences in suicidal ideation, and suggest that building socio-cultural environments are needed to abate their negative emotions and to help youth find out their reasons to live. INTRODUCTION Suicide has been reported to be the fourth highest cause of death in South Korea since 2005. In terms of age, from the teens to the thirties, suicide has been ranked as the first cause of death in both males and females. It has been the highest (29.1/100,000 persons) and was over twofold higher than the OECD average (12.1/100,000 persons) in a single year [1]. With almost one million people dying from suicide around the world each year, and with a disproportionate impact of such on the world's youth, the global society owes it to the future generations to act now with regard to such matter [2]. Suicidal ideation is distinguished from suicidal attempt and suicide but is linked to them. Although not all cases of suicidal ideation would necessarily lead to suicidal attempt or suicide, suicidal ideation has been geared to-wards suicide and could facilitate suicidal risk and the plausibility of committing suicide [1,3,4]. The ripple effect -that is, that suicide may impact on the lives and mental health of many families and communities, beyond the individual who commits it -dictates the need for collective action to acknowledge and address the problem as well as for commitment to effective interventions supported by political will and resources [2]. Adolescence is a transitional stage of rapid developmental change in the physical, psychological, intellectual, and social realms, which exposes one to increasing selfconsciousness and doubts about one's identity. These rapid changes can induce stressful conditions and may lead to emotional instability and impulsiveness [5][6][7]. According to the 2013 Korean Youth Health Behavior Survey (9th), the level of stress cognition and depression of adolescents is higher than that of adults and is higher in girls than in boys and in high school students than in middle school students [8]. Suicidal ideation was also found to be higher in high school students than in middle school students, but the suicidal attempt rate was found to be higher in middle school students than in high school students [8]. Adolescents who have suicidal ideation may commit suicide in adulthood tenfold more than those who do not have suicidal ideation [7]. Also, suicidal ideation has a different elevation trend by gender, specially, suicidal ideation rate in girls has highly elevated from twelve to fourteen years old as that of boys [9]. Suicide is a concept resulting from suicidal ideation and suicidal attempt [4,10], and suicidal ideation among the youth needs to be given particular attention and public health concern [9]. Besides, unlike in many other countries, the educational and social structures in South Korea have a tendency to foster crooked college admission practices and emphasize earning a degree rather than learning and ability, which has resulted not only in biased education but also in an imbalanced educational culture, particularly examinationoriented education and early childhood education. Hence, adolescents have come to have stressful conditions that could worsen their anxiety and depression [4]. Specially, family environmental stress and academic stress that affect to suicidal ideation in adolescents have been reported in Korean researches [10,17]. Legislation for suicide prevention, including fostering a life-respecting culture, has been enacted in 2011. Concurrently, Korea Suicide Prevention Center has established and implemented various relevant programs. In school, the student health promotion program has come to emphasize mental health, and this trend has expanded generally, but the scope of the program is still limited, and it is focused on preventive and general education, excepting specific strategies. To identify adolescents' mental health status and to connect which can lay down their suicidal ideation, an approach based on socio-cultural comprehension is needed. Although suicidal ideation rate was higher in high school students than in middle school students, this study focused on middle school students considering their conditional differences; elevation of academic stress in general, emotional instability, depression, and decreasing of self-esteem. This study was conducted to identify the factors for sui-cidal ideation in middle school students by gender, and to announce the necessity of gaining upon socio-cultural comprehension based approach. Study Design This study used a cross-sectional survey research design, conducting secondary analysis of screening survey data from Siheung Mental Health Center in Gyeonggi Province, South Korea. The screening survey was conducted to examine the youth's mental health status and to gather basic data for use in coming up with a community-based mental health program. The study received ethical approval from the Institutional Review Board of J Oriental Hospital, which is affiliated with S University (SMJOH, 2010-4). Setting and Sample This study was conducted in three middle schools randomly selected from a pool of 21 middle schools in S City. The sample was stratified by school and grade (3 classes per grade, 9 classes in one school). A total of 18 classes were selected, and a total of 906 students from the three schools agreed to participate in the study. During the data screening, the 28 participants who only partially completed the survey questionnaire were dropped out of the study. As such, a total of 878 participants were included in the analysis. Based on the power calculation using the G*Power program, effect size f 2 =0.15, ⍺=.05, power (1β)=.95, and Odds 1.5, and considering depression the main variable related to suicidal ideation, as in the previous research [19], 417 participants were needed for the multiple logistic regression model. The actual number of participants in this study was sufficient. Procedure Data collection was conducted from November 1 to December 10, 2010. Three middle schools were randomly picked from a pool of 21 middle schools in S City. Three classes were randomly assigned in each grade, and nine classes per school. A document reflecting the research purpose and a request for cooperation was sent to the principal of each school, but direct contact was established with the target participants to obtain their informed consent to participate in the study. After the explanation of the research purpose to them, the students signed the informed-consent form. The ethical-consideration infor-mation that the students' participation was not mandatory and could be withdrawn at any time during the study was given to the students in advance. 1) Suicidal ideation Suicidal ideation was measured using the Korean Suicidal Ideation Questionnaire (KSIQ), which is a modified version [20] of the Reynolds Suicidal Ideation Questionnaire (RSIQ) developed by Reynolds [21]. The 30 items of the RSIQ elicit data about a person's thoughts about and views of suicidal ideation. Each item is scored based on a 7-point Likert scale ranging from 0 to 6. The means across all the items are summed up to obtain the total score (range: 0~180); a higher score indicates a higher level of suicidal ideation. The reliability was .96 for the males and .98 for the females in this study. The cut-off score for suicidal ideation was 20, reported as having the optimal clinical utility based on the results of a clinical study involving adolescents aged 13~18 [22]. In this study, the cut-off score for suicidal ideation was chosen by this clinical utility report. 2) Academic stress Academic stress was measured using the Visual Analog Scale (VAS), a 10cm horizon bar ranging from 0 (left end) to 10 (right end). Each subject marked his/her own academic stress level on the instrument, and a higher mark represents a higher level of academic stress. Most of the other existing academic stress scales are included in general stress inventories or consist of many items, which might elevate the dropout rate or lessen the reliability of the results obtained. VAS is based on the intensity of the respondent's conformity to each of the items, and can evaluate the evaluation object easily, but its validity is still being debated [23]. Hence, by targeting 119 middle school students, partial correlation analysis was conducted between VAS and Adolescent Academic Stress Scale, which is used for the Korean Adolescent Panel Survey [17], to establish concurrent validity (r=.736, p<.001). 3) Depression Depression was measured using Beck Depression Inventory II (BDI II) [24], which consists of 21 items. Each item is scored based on a 4-point Likert scale ranging from 0 (not at all) to 3 (severely). The means across all the items are summed up to obtain the total score (range: 0~63); a higher score indicates a higher level of depression. The reliability was .90 in this study. 4) Self-esteem Self-esteem was measured using VAS, which was described earlier. Each subject marked his/her own self-esteem level on the instrument; a higher mark indicates a higher level of self-esteem. Although the validity of VAS is still being debated [23], the instrument is based on the intensity of the respondent's conformity to each of the items, and can evaluate the evaluation object easily. To address VAS's weakness in terms of validity, targeting 119 middle school students, partial correlation analysis was conducted between VAS and Adolescent Self-Esteem Scale, which is used for the Korean Adolescent Panel Survey [17], to establish concurrent validity (r=.779, p <.001). 5) Communication with one's parents To measure the respondent's level of communication with his/her parents, a subscale (8 items related to communication with one's parents) of Inventory of Parent and Peer Attachment (IPPA) developed by Armsden and Greenberg (1987) [25] was used. Each item is scored based on a 4-point Likert scale ranging from 1 (never) to 4 (almost every day). The means across all the items are summed up to obtain the total score (range: 8~32); a higher score indicates frequent communication with one's parents. The reliability was .78 in this study. Data Analysis Descriptive statistics were used to evaluate the characteristics of the subjects by gender, and the x 2 test was used to compare the differences in suicidal ideation according to the general characteristics and suicide-related variables of the subjects by gender. The independent t-test was used to compare the subjects' academic stress, depression, selfesteem, and communication with their parents according to suicidal ideation by gender. To identify the factors that affect suicidal ideation by gender, multiple logistic regression analysis was conducted. For data analysis, SPSS/WIN 16.0 was used. Characteristics of the Subjects and Suicidal ideation relating variables Except for 10.5% of the subjects, the subjects had midlow school grades (boys: 90.0%; girls: 88.9%). Most of the subjects claimed that their family income was under 2 million Korean won per month. With regard to the subjects' health behavior 39.3% of all the subjects were ad- dicted to the Internet, and in particular. 8.4% had friends who had attempted to commit suicide while 2.1% had family members who had attempted to do the same. And there were gender differences in suicidal ideation, academic stress, and depression. The subject female students had significantly higher scores in suicidal ideation, academic stress, and depression compared to the subject male students (p<.001). Especially, the suicidal ideation in the subject female students was over twofold higher than that in the subject male students (boys: 13.3%; girls: 31.1%)( Table 1). Suicidal Ideation according to the General Characteristics and Suicidal Ideation Relating Variables by Gender Among the subject male students, there were significant group differences in suicidal ideation according to whether or not they visited suicide-related websites (p< .001) and whether or not they had friends who had attempted to commit suicide (p =.008). Among the subject female students, there were also significant group differences in suicidal ideation according to whether or not they were smokers, whether or not they drank alcohol, and whether or not they were addicted to the Internet (p< .001). That is, among both the subject male and female students, those who visited suicide-related websites and who had friends who had attempted to commit suicide were found to have significantly higher suicidal-ideation scores than those who did not visit suicide-related websites and did not have friends who had attempted to commit suicide ( Table 2). Differences in Academic Stress, Depression, Self-esteem, and Communication with Parents according to Suicidal Ideation by Gender Based on the cut-off score for suicidal ideation (20), in both the subject male and female students, the academic stress and depression were significantly higher in the suicidal-ideation group than in the non-suicidal-ideation group while the self-esteem and communication with parents were significantly lower in the suicidal-ideation group than in the non-suicidal-ideation group, whose suicidal-ideation score was under 20 (p<.001)( Table 3). Factors affecting Suicidal Ideation by Gender To identify the factors that affect suicidal ideation in middle school students, multiple logistic analysis was conducted. For independent variables, significant variables both male and female students which were analyzed in descriptive analysis were included. Among suicidal ideation relating variables, visiting suicide related websites and having friends who have attempted suicide For the subject male students, the factor that was found to affect suicidal ideation was depression; a 1 point increase in the depression score led to 1.2-fold higher odds of suicidal ideation (95%CI: 1.13~1.26). For the subject female students, those who visited suicide-related websites had 28-fold higher suicidal-ideation scores than those who did not (95%CI: 2.74~285.74), and those who had friends who had attempted to commit suicide got 3.40-fold higher suicidal-ideation scores than those who did not (95% CI: 1.57~7.38). Also, a 1 point increase in the depression score led to 1.165-fold higher odds of suicidal ideation (95% CI: 1.12~1.22), and a 1 point increase in the communication with parents score led to 0.91-fold lower odds of suicidal ideation (95% CI: 0.86~0.97)( Table 4). DISCUSSION Social interests for suicide have been increased as the highest suicide mortality rate, and many studies on suicidal ideation have been conducted in Korea. This study was conducted not only to identify the factors for suicidal ideation in middle school students by gender, but also to announce the necessity of gaining upon socio-cultural comprehension based approach which offset suicidal ideation in middle school students. There were differences in the suicidal-ideation scores by gender according to the general characteristics of the subjects. Jeon and colleagues [18] reported that conflicts and stress with one's peer group and conflicts with one's parents were the common predictable variables related to suicidal ideation. Conflicts with one's parents were ranked as the factor with the greatest effect in both the subject male and female students, followed by bullying and familial economic status in the subject male students and by conflicts with one's peer group and familial economic status in the subject female students. Other studies reported that familial economic status could adversely affect adolescent mental health, smoking, and alcohol consumption [11]. Also, Internet suicide site accessibility was added to the factors affecting suicidal ideation, as well as conflicts with familial members or friends [2]. Based on the cut-off score for suicidal ideation (20), the academic stress and depression were higher and the selfesteem and communication with one's parents were lower in the suicidal-ideation group than in the non-suicidalideation group. In previous studies, self-esteem [15,16], familial alliance [12,16,18], and social support [2,3,5,15] were found to lessen the danger of suicidal ideation. In this study, self-esteem and communication with one's parents presented a similar effect of abating suicidal ideation. Depression was cited as a critical factor in suicidal ideation psychologically [4,[10][11][12][13], and life events preceded suicidal ideation [2,4,14,26]. As discussed earlier, the suicidal ideation in the subject female students in this study was twofold higher than in the subject male students, and the smoking and alcohol consumption rates were higher in the suicidal-ideation group than in the non-suicidalideation group among the subject female students. It is presumed that smoking and alcohol consumption need to be more carefully observed or monitored in female students than in male students. Suicide impacts the most vulnerable of the world's populations, and places a larger burden on the low-and middle-income countries [2], and this economic vulnerability could be linked to other poor conditions, not only to the factors impacting suicidal ideation among adolescent [5,11,18]. In this study, the characteristics and environment of the subjects regarded that they were socioeconomically disadvantaged: 82.9% of the students lived in a family that their family income were under 4,000,000 won per month (family median income was 4,220,000 won per month in Korea), especially 24.9% were under 2,000,000 won; and Internet-addicted, 39.3%; and low school grades, 45.0%. These socio-economically disadvantaged conditions have resulted in adolescents' depression and suicidal ideation [12]. Also, in suicidal ideation relating variables, gender differences were found in terms of visiting suicide-related websites, having friends who have attempted to commit suicide, academic stress, and depression; more female students visited suicide-related websites and had friends who have attempted to commit suicide, and the academic stress and depression in the female students were significantly higher than those in the male students. It is not irrelevant that in previous studies, the depression level of the female students was higher than that of the male students [4,12,14], and the female students could be more sensitive in terms of stress [12,26]. Phillips et al., [15] reported that negative emotions strongly associated with current and future self-harm in both male and female adolescents, while self-esteem strongly moderated current and future self-harm. It is suggested that negative emotions and response to neutral life events should be attenuated and should hold gender differentiated approach. Shim and Kim [19] revealed that the suicidal ideation relating factors differ by phase; in early youth, stress and depression are the most powerful factors, and in mid-and late youth, the social support and impact of one's relationship with a significant key person determined the level of one's suicidal ideation. On the contrary, Kim [27] found that self-efficacy, which has a protective effect on earlyyouth suicidal ideation, decreases in mid-and late youth while the impact of depression is maintained. Another study presented participation in sports teams in and outside of school protected against suicidal ideation [13]. Considering gender and youth phase, suicidal ideation rate in boys showed indifferences, while the rate in girls showed peak elevation in age twelve to fourteen years old and relieved after age fifteen years old [9]. Although the differences depending on the youth phase and gender have not been verified, it may be appropriate to approach the clients considering the youth phase they are in for the suicidal ideation prevention program's effectiveness. The multiple logistic regression analysis revealed that in the male students, depression was the sole factor that influenced suicidal ideation, while in the female students, visiting suicide-related websites, having friends who have attempted to commit suicide, depression, and communication with one's parents influenced suicidal ideation. Especially, among the female students, those who visited suicide-related websites showed a 28-fold higher suicidalideation risk than those who did not, which is more than quadruple that among the male students. Although Shim and Kim [19] showed that the suicidal ideation relating factors are indifferent by gender, Jeon and colleagues [18] reported that academic performance, family economic state, conflicts with one's parents, bullying, and conflicts with one's friend related with suicidal ideation, and these are prominent in girls than those of boys. And family environmental stress and academic stress that affect to suicidal ideation in adolescents have been reported [10,17]. In this study, academic stress had related in suicidal ideation but didn't affect to suicidal ideation in logistic analysis both male and female students. Further study is needed to identify various stress inducing variables to confirm suicidal ideation affecting factors. In Smahel and colleagues' [28] research, digital media including internet use and technology impacted physical and mental health problems to children. It is presumed that female students who have a negative mental state and strong emotional stress due to their personal conditions, including their familial environment, have a great risk to suicidal ideation. Further research should investigate whether internet addiction and visiting suicide-related website are correlated. In the Guidelines for the Prevention of Suicide, the United Nations cites the following key components of national suicide prevention strategies: clear objectives, relevant risk and protective factors, and effective interventions. Further, the intervention targets should be divided into the general population, the vulnerable subpopulations at risk, and individuals [29]. Also, gatekeeper training, crisis care, scaling up human resources, and access to and management of persons at risk should be considered, as well as restricting access to means of inflicting self-harm and development of policies aimed at reducing the risk [2]. In South Korea, rapid growth and development may induce a work-oriented society and an inordinately education-oriented culture, particularly examination-oriented education and early childhood education. These tight and rigid atmospheres can worsen adolescents' stress; especially, they can aggravate adolescents' depression and provoke negative thinking [4,11]. Also, negative emotions and low self-esteem strongly associated with current and future self-harm in adolescents [15]. Not a few adolescents have faced the danger of suicide; in fact, the leading cause of death among teens at present is suicide in Korea, and the level of stress cognition and depression of adolescents is higher than that of adults [8]. To combat suicide and to prevent suicidal ideation in middle school students, we have to try to take socio-cultural comprehension based approach, and to build an environment which can abate negative emotion and draw out reason for living including making meaningful relations. Limitations The limitations of this study should be presented before planning the later study. First, this study was based on the respondents' self-report in response to each item. And academic stress and self-esteem were assessed through VAS instead of based on a scale with documented reliability, although concurrent validity was applied to address VAS's weakness in validity. If researchers considered another approach in depth, such as focus group interview, we could find more diverse and copious factors which affect to negative conditions of adolescents and intimate conditions that draw in their suicidal ideation. Further research should be conducted to bring more reliable result and evidence to this concern. CONCLUSION Most research on suicidal ideation of adolescents in Korea has focused on analyzing risk factors, and the factors contributing to suicidal ideation by gender have reported inconsistently in each study. This study was conducted to identify the factors for suicidal ideation in middle school students by gender, and to suggest the needs for socio-cultural comprehension based approach which put down their suicidal ideation. Suicidal ideation scores of the female students were over twofold higher than those of the male students. The factor that was found to affect suicidal ideation in the male students was depression while for the female students the factors were visiting suicide-related websites, having friends who attempted to commit suicide, depression, and poor communication with one's parents. Definitely, we should take responsibility in combating suicide by building a socio-cultural environment as well as preventing and managing the suicide risk of adolescents: diminution of negative emotion, making school culture which generate encouraging and consolation, and drawing out the reason for living. It would be possible when family, school, community including policymakers strives to develop diverse socio-cultural environment.	2019-05-10T13:08:55.751Z	2016-12-31T00:00:00.000	{ "year": 2016, "sha1": "ef2681cb91f56941e95f85324168827817f81cca", "oa_license": "CCBYNC", "oa_url": "http://www.jkssh.or.kr/journal/download_pdf.php?doi=10.15434/kssh.2016.29.3.267", "oa_status": "GOLD", "pdf_src": "Adhoc", "pdf_hash": "f9f15b65497ae8b8676066c07e1caba88dda1117", "s2fieldsofstudy": [ "Psychology", "Education" ], "extfieldsofstudy": [ "Psychology" ] }
9363741	pes2o/s2orc	v3-fos-license	Periprosthetic Femur Fracture Occuring after Contralateral Neglected Femoral Neck Fracture Introduction: Periprosthetic fractures of the femur are uncommon, but at times may lead to complications especially in elderly patients. As treatment of these fractures is difficult, prevention by identifying the risk factors is the best way to overcome these complex problems. Case Report: A periprosthetic right femur fracture associated with a neglected left femoral neck fracture in the contralateral femur in a 78-year-old elder woman patient is reported in the present article. We discuss the prevention of periprosthetic fractures after hip arthroplasty and address the risk factors associated with this complication. Conclusion: The present case emphasizes the importance of investigating and treating the cause of sudden onset of restriction on full weight-bearing in the contralateral limb, to prevent periprosthetic femoral fracture after hip arthroplasty in elderly patients. Introduction Periprosthetic fracture generally seen in frail elderly patients with osteoporosis is one of the most devastating complications of hip arthroplasty [1,2,3]. In literature, risk factors associated with periprosthetic fractures are still under debate and investigation [1,3,4,5,6]. Female gender [1,5], implant design [6], higher number of comorbidities [1,5], old age [3,5] and cementless implants [1,3] were suggested to be the risk factors for periprosthetic fractures. However, the results of the previous studies are contradictory and a consensus about the risk factors has not been reached yet. Whatever the aetiology, the present case addresses the importance of sudden onset of restriction on full weight-bearing in the contralateral limb after hip arthroplasty in elderly patients. We report a case of a periprosthetic right femur fracture associated with a neglected contralateral left femoral neck fracture in an elderly patient. Case Report A 78-year-old woman suffering from generalized right leg pain applied to the outpatient department. From her medical history; she underwent total hip arthroplasty six years ago after being diagnosed as having primary osteoarthritis of the right hip. After surgery, she had not reported any problems about her right hip arthroplasty. Suddenly; she had restriction on full weight-bearing in her contralateral left side for seven months. She was prescribed painkillers by her general practitioner and started using a walking stick. Her relatives said that there was no specific trauma that could explain sudden onset of restriction on full weight-bearing in her contralateral left side. They criticized the general practitioner for not investigating her left side pain meticulously. We had no Xray records of the patient during her application to general practitioner seven months ago. While she was walking for seven months after restriction of full weight-bearing of left hip, she had a fall after stumbling on the carpet, and was admitted to the outpatient department of orthopaedics for evaluation of severe right hip pain. The anteroposterior radiograph of the pelvis revealed Vancouver type-C periprosthetic fracture of the right femur and a neglected femoral neck fracture in her left hip (Fig. 1). According to the information gathered from her and her relatives, it was a simple fall that could not explain such a catastrophic periprosthetic fracture in the absence of any predisposing factor such as contralateral full weight-bearing for months. The patient was 78 kg, 165 cm tall with a calculated body mass of 28.65 kg/cm2. She had undergone total hip arthroplasty with expansion cementless acetabular cup and cemented femoral prosthesis previously. After surgery, she had been followed up regularly for three years once a year and there was no complication such as significant leg length discrepancy or rotational abnormalities. Later she was lost to follow-up as she had no complication during those three years. She applied to outpatient department with periprosthetic fracture of the right femur and a neglected femoral neck fracture in her left hip. Preoperative routine procedure was done by planning revision arthroplasty of the right hip. The only comorbidity was wellcontrolled blood pressure. C-reactive protein level and erythrocyte sedimentation rate were in normal range. She underwent revision arthroplasty of the right hip by using the previous anterolateral incision [modified Watson-Jones]. Both acetabular and femoral components were removed. Cementless acetebular cup with constrained polyethylene liner and 40 mm proximal body, 12/160 mm distal stem of fully porous coated long femoral prosthesis was implanted during revision surgery (Fig. 2). Five dall-miles cables were used to stabilize the femoral component. There were no post-operative complications and the patient was ambulated with partial weight-bearing on the left side and as much as she tolerated on the right side with the aid of a walker on the third day postoperatively. Hip musculature strengthening physical therapy was applied during her follow-up. Although total hip arthroplasty was offered for neglected contralateral femoral neck fracture, patient and her relatives were not keen for a second operation due to advanced age of the patient and also because she had relatively less complaints with her left hip. Weight-bearing had been increased incrementally during the follow-up of six and twelve weeks. The patient was ambulated with full weight-bearing six months after surgery. After seven months of neglected contralateral femoral neck fracture, patient was encouraged for weight-bearing on that side as much as her pain allowed her to bear. During regular follow-ups, she continued to have restricted weight-bearing in the contralateral side and was consistently reminded of the necessity for a total hip arthroplasty for that side at eachfollow-up visit. Discussion Prevention of late periprosthetic femoral fracture is suggested to be best accomplished through routine follow-up and intervention of risk factors [7,8]. Identification of the underlying reasons can show a reasonable way for interventions to reduce the risk of such fractures. However, predisposing factors for periprosthetic femoral fractures are still a matter of debate due to the contradictory results of the previous studies [1,3,4,5,9]. Singh et al. [1] found that female gender was associated with a higher risk of postoperative periprosthetic fractures. Cook et al. [3] suggested that there was no association between the occurrence of periprosthetic fractures and gender, and two other studies supported the results of this study [4,9]. Conflicting results were reported concerning the age. Sarvilinna et al. [9] concluded that age was not associated with an increased risk of periprosthetic fractures. While Cook et al. [3] emphasized an increase in periprosthetic fracture rates with increasing age; Sarvilinna et al. [9] suggested a decrease with age. Franklin et al. [7] evaluated the risk factors of periprosthetic fractures in their review article and www.jocr.co.in stressed the debated results in the literature. As mortality and morbidity rates after periprosthetic femoral fractures are seriously high, it is important to identify more accurate and unarguable risk factors for periprosthetic fractures. To our knowledge, there has been no case report of periprosthetic femoral fracture after neglected contralateral femoral neck fracture in English literature. In the present case, as an unarguable risk factor for periprosthetic femoral fractures, excessive loading of osteoporotic bone was not taken notice of, by neglecting the contralateral femoral neck fracture. Yun et al. [10] suggested that periprosthetic femoral fractures were associated with contralateral hip diseases such as degenerative arthritis in the presence of lysis around the femoral component. Based on their results from five patients with previous total hip arthroplasty, they recommended timely elective hip surgery for the contralateral hip in case of restricted use of the associated limb secondary to degenerative diseases [10]. Besides all the debated risk factors mentioned above and degenerative diseases of the contralateral hip, sudden onset of loss of weight-bearing of the contralateral limb is an absolute risk factor for periprosthetic femur fractures via increasing stress over the osteoporotic bone. The present case supports the suggestion of Yun et al. [10] and addresses the importance of pathologies in the contralateral hip if there is a history of total hip arthroplasty especially in elderly patients. As a result, in case of non-weight bearing of the contralateral limb, we recommend activity limitation in patients until the aetiology is found and treated. Conclusion Excessive loading of osteoporotic bone, restriction of full weight of contralateral hip due to neglected fracture or osteoarthritis of hip, lysis around the femoral component may cause periprosthetic femoral fractures. We strictly advise meticulous investigation of the cause of a sudden onset restriction of full weight-bearing in the contralateral limb to prevent periprosthetic femoral fractures after hip arthroplasty in elderly patients. In the elderly patients, sudden onset restriction of full weight-bearing in the contralateral limb should be meticulously investigated to prevent periprosthetic femoral fracture if the patient has/had previous hip arthroplasty history.	2017-11-03T13:23:13.283Z	2016-01-01T00:00:00.000	{ "year": 2016, "sha1": "3240f6af6e1c9011a1d0c208e14ace2c0874b3ed", "oa_license": "CCBYNCSA", "oa_url": null, "oa_status": null, "pdf_src": "PubMedCentral", "pdf_hash": "3240f6af6e1c9011a1d0c208e14ace2c0874b3ed", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
8486003	pes2o/s2orc	v3-fos-license	phi-LSTM: A Phrase-based Hierarchical LSTM Model for Image Captioning A picture is worth a thousand words. Not until recently, however, we noticed some success stories in understanding of visual scenes: a model that is able to detect/name objects, describe their attributes, and recognize their relationships/interactions. In this paper, we propose a phrase-based hierarchical Long Short-Term Memory (phi-LSTM) model to generate image description. The proposed model encodes sentence as a sequence of combination of phrases and words, instead of a sequence of words alone as in those conventional solutions. The two levels of this model are dedicated to i) learn to generate image relevant noun phrases, and ii) produce appropriate image description from the phrases and other words in the corpus. Adopting a convolutional neural network to learn image features and the LSTM to learn the word sequence in a sentence, the proposed model has shown better or competitive results in comparison to the state-of-the-art models on Flickr8k and Flickr30k datasets. Introduction Automatic caption/description generation from images is a challenging problem that requires a combination of visual information and linguistic as illustrated in Fig. 1. In other words, it requires not only complete image understanding, but also sophisticated natural language generation [1][2][3][4]. This is what makes it such an interesting task that has been embraced by both the computer vision and natural language processing communities. One of the most common models applied for automatic caption generation is a neural network model that composes of two sub-networks [5][6][7][8][9][10], where a convolutional neural network (CNN) [11] is used to obtain feature representation of an image; while a recurrent neural network (RNN) 1 is applied to encode and generate its caption description. In particular, Long Short-Term Memory (LSTM) model [12] has emerged as the most popular architecture among RNN, as it has the ability to capture long-term dependency and preserve sequence. Although sequential model is appropriate for processing sentential data, it does not capture any other syntactic structure of language at all. Nevertheless, it is undeniable that sentence structure is one of the prominent characteristics of language, and Victor Yngve -an influential contributor in linguistic theory stated in 1960 that"language structure involving, in some form or other, a phrasestructure hierarchy, or immediate constituent organization" [13]. Moreover, Tai et al. [14] proved that a tree-structured LSTM model that incorporates syntactic interpretation of sentence structure, can learn the semantic relatedness between sentences better than a pure sequential LSTM alone. This gives rise to question of whether is it a good idea to disregard other syntax of language in the task of generating image description. In this paper, we would like to investigate the capability of a phrase-based language model in generating image caption as compared to the sequential language model such as [6]. To this end, we design a novel phrase-based hierarchical LSTM model, namely phi-LSTM to encode image description in three stages -chunking of training caption, image-relevant phrases composition as a vector representation and finally, sentence encoding with image, words and phrases. As opposed to those conventional RNN language models which process sentence as a sequence of words, our proposed method takes noun phrase as a unit in the sentence, and thus processes the sentential data as a sequence of combination of both words and phrases together. Fig. 2 illustrates the difference between the conventional RNN language model and our proposal with an example. Both phrases and sentences in our proposed model are learned with two different sets of LSTM parameters, each models the probability distribution of word conditions on previous context and image. Such design is motivated by the observation that some words are more prone to appear in phrase, while other words are more likely to be used to link phrases. In order to train the proposed model, a new perplexity based cost function is defined. Experimental results using two publicly available datasets (Flickr8k [15] and Flickr30k [16]), and a comparison to the state-of-the-art results [5-7, 9, 33] have shown the efficacy of our proposed method. Related Works The image description generation task is generally inspired by two lines of research, which are (i) the learning of cross-modality transition or representation between image and language, and (ii) the description generation approaches. Multimodal Representation and Transition To model the relationship between image and language, some works associate both modalities by embedding their representations into a common space [17][18][19][20]. First, they obtain the image features using a visual model like CNN [18,19], as well as the representation of sentence with a language model such as recursive neural network [19]. Then, both of them are embedded into a common multimodal space and the whole model is learned with ranking objective for image and sentence retrieval task. This framework was also tested at object level by Karpathy et al. [20] and proved to yield better results for the image and sentence bi-directional retrieval task. Besides that, there are works that learn the probability density over multimodal inputs using various statistical approaches. These include Deep Boltzmann Machines [21], topic models [22], log-bilinear neural language model [8,23] and recurrent neural networks [5][6][7] etc. Such approaches fuse different input modalities together to obtain a unified representation of the inputs. It is notable to mention that there are also some works which do not explicitly learn the multimodal representation between image and language, but transit between modalities with retrieval approach. For example, Kuznetsova et al. [24] retrieve images similar to the query image from their database, and extract useful language segments (such as phrases) from the descriptions of the retrieved images. Description Generation On the other hand, caption generation approaches can generally be grouped into three categories as below: Template-based. These approaches generate sentence from a fixed template [25][26][27][28][29]. For example, Farhadi et al. [25] infer a single triplet of object, action and scene from an image and convert it into a sentence with fixed template. Kulkarni et al. [26] use complex graph of detections to infer elements in sentence with conditional random field (CRF), but the generation of sentences is still based on the template. Mitchell et al. [28] and Gupta et al. [29] use a more powerful language parsing model to produce image description. In overall, all these approaches generate description which is syntactically correct, but rigid and not flexible. Composition Method. These approaches extract components related to the images and stitch them up to form a sentence [24,30,31]. Description generated in such manner is broader and more expressive compared to the template-based approach, but is more computationally expensive at test time due to its nonparametric nature. Neural Network. These approaches produce description by modeling the conditional probability of a word given multimodal inputs. For instance, Kiros et al. [8,23] developed multimodal log-bilinear neural language model for sentence generation based on context and image feature. However, it has a fixed window context. The other popular model is recurrent neural network [5-7, 9, 32], due to its ability to process arbitrary length of sequential inputs such as sequence of words. This model is usually connected with a deep CNN that generates image features. The variants on how this sub-network is connected to the RNN have been investigated by different researchers. For instance, the multimodal recurrent neural network proposed by Mao et al. [5] introduces a multimodal layer at each time step of the RNN, before the softmax prediction of words. Vinyals et al. [6] treat the sentence generation task as a machine translation problem from image to English, and thus image feature is employed in the first step of the sequence trained with their LSTM RNN model. Relation to Our Work Automatic image caption generated via template-based [25][26][27][28][29] and composition methods [24,30,31] are typically two-stage approaches, where relevant elements such as objects (noun phrases) and relations (verb and prepositional phrases) are generated first before a full descriptive sentence is formed with the phrases. With the capability of LSTM model in processing long sequence of words, neural network based method that uses a two-stage approach deem unnecessary. However, we are still interested to find out how sequential model with phrase as a unit of sequence performs. The closest work related to ours is the one proposed by Lebret et al. [33]. They obtain phrase representation with simple word vector addition and learn its relevancy with image by training with negative samples. Sentence is then generated as a sequence of phrases, predicted using a statistical framework conditioned on previous phrases and its chunking tags. While their aim was to design a phrase-based model that is simpler than RNN, we intend to compare RNN phrase-based model with its sequential counterpart. Hence, our proposed model generates phrases and recomposes them into sentence with two sub-networks of LSTM, which are linked to form a hierarchical structure as shown in Fig. 2 Our Proposed phi-LSTM Model This section details how the proposed method encodes image description in three stages -i) chunking of image description, ii) encode words and phrases into distributed representations, and finally iii) encodes sentence with the phi-LSTM model. Phrase Chunking A quick overview on the structure of image descriptions reveals that, key elements which made up the majority of captions are usually noun phrases that describe the content of the image, which can be either objects or scene. These elements are linked with verb and prepositional phrases. Thus, noun phrase essentially covers over half of the corpus in a language model trained to generate image description. And so, in this paper, our idea is to partition the learning of noun phrase and sentence structure so that they can be processed more evenly, compared to extracting all phrases without considering their part of speech tag. To identify noun phrases from a training sentence, we adopt the dependency parse with refinement using Stanford CoreNLP tool [34], which provides good semantic representation over a sentence by providing structural relationships between words. Though it does not chunk sentence directly as in constituency parse and other chunking tools, the pattern of noun phrase extracted is more flexible as we can select desirable structural relations. The relations we selected are: determiner relation (det), numeric modifier (nummod ), adjectival modifier (amod ), adverbial modifier (advmod ), but is selected only when the meaning of adjective term is modified, e.g. "dimly lit room", compound (compound ), nominal modifier for possessive alteration (nmod:of & nmod:poss). Note that the dependency parse only extracts triplet made up of a governor word and a dependent word linked with a relation. So, in order to form phrase chunk with the dependency parse, we made some refinements as illustrated in Fig. 3. The triplets of selected relations in a sentence are first located, and those consecutive words (as highlighted in the figure, e.g. "the", "man") are grouped as a single phrase, while the standalone word (e.g. "in") will remain as a unit in the sentence. Compositional Vector Representation of Phrase This section describes how compositional vector representation of a phrase is computed, given an image. Image Representation. A 16-layer VggNet [35] pre-trained on ImageNet [36] classification task is applied to learn image feature in this work. Word Embedding. Given a dictionary W with a total of V vocabulary, where word w ∈ W denotes word in the dictionary, a word embedding matrix W e ∈ R K×V is defined to encode each word into a K -dimensional vector representation, x. Hence, an image description with words w 1 · · · w M will correspond to vectors x 1 · · · x M accordingly. Composition of Phrase Vector Representation. For each phrase extracted from the sentence, a LSTM-based RNN model similar to [6] is used to encode its sequence as shown in Fig. 4. Similar to [6], we treat the sequential modeling from image to phrasal description as a machine translation task, where the embedded image vector is inputted to the RNN on the first time step, followed by a start token x sp ∈ R K indicating the translation process. It is trained to predict the next word at each time step by outputting p tp+1 ∈ R K×V , which is modeled as the probability distribution over all words in the corpus. The last word of the phrase will predict an end token. So, given a phrase P which is made up by L words, the input x tp at each time step are: For a LSTM unit at time step t p , let i tp , f tp , o tp , c tp and h tp denote the input gate, forget gate, output gate, memory cell and hidden state at the time step respectively. Thus, the LSTM transition equations are: Here, σ denotes a logistic sigmoid function while denotes elementwise mul- Intuitively, each gating unit controls the extent of information updated, forgotten and forward-propagated while the memory cell holds the unit internal memory regarding the information processed up to current time step. The hidden state is therefore a gated, partial view of the memory cell of the unit. At each time step, the probability distribution of words outputted is equivalent to the conditional probability of word given the previous words and image, P (w t \|w 1:t−1 , I). On the other hand, the hidden state at the last time step L is used as the compositional vector representation of the phrase, z ∈ R K , where z = h L . Encoding of Image Description Once the compositional vector of phrases are obtained, they are linked with the remaining words in the sentence using another LSTM-based RNN model as shown in Fig. 5. Another start token x ss ∈ R K and image representation with W is ∈ R K×D and bias b is ∈ R K as embedding parameters. Hence, the input units of the LSTM in this level will be the image representation v s , start token x ss , followed by either compositional vector of phrase z or word vector x in accordance to the sequence of its description. For simplicity purpose, the arranged input sequence will be referred as y. Therefore, given the example in Fig. 4-5, the LSTM input sequence of the sentence will be {v s , x ss , y 1 ...y N } where N = 8, and it is equivalent to sequence {v s , x ss , z 1 , x 3 , z 2 , x 7 , x 8 , x 9 , x 10 , z 3 }, as in Fig. 5. Note that a phrase token is added to the vocabulary, so that the model can predict it as an output when the next input is a noun phrase. The encoding of the sentence is similar to the phrase vector composition. Eq. 3-9 are applied here using y ts as input instead of x tp , where t p and t s represent time step in phrase and sentence respectively. A new set of model parameters with same dimensional size is used in this hierarchical level. Training the phi-LSTM Model The proposed phi-LSTM model is trained with log-likelihood objective function computed from the perplexity 2 of sentence conditioned on its corresponding image in the training set. Given an image I and its description S, let R be the number of phrases of the sentence, P i correspond to the number of LSTM blocks processed to get the compositional vector of phrase i, Q is the length of composite sequence of sentence S, while p tp and p ts are the probability output of LSTM block at time step t p − 1 and t s − 1 for phrase and sentence level respectively. The perplexity of sentence S given its image I is where Hence, with M number of training samples, the cost function of our model is: where It is the average log-likelihood of word given their previous context and the image described, summed with a regularization term, λ θ · θ 2 2 , average over the number of training samples. Here, θ is the parameters of the model. This objective however, does not discern on the appropriateness of different inputs at each time step. So, given multiple possible inputs, it is unable to distinguish which phrase is the most probable input at that particular time step during the decoding stage. That is, when a phrase token is inferred as the next input, all possible phrases will be inputted in the next time step. The candidate sequences are then ranked according to their perplexity up to this time step, where only those with high probability are kept. Unfortunately, this is problematic because subject in an image usually has much lower perplexity as compared to object and scene. Thus, such algorithm will end up generating description made up of only variants of subject noun phrases. To overcome this limitation, we introduce a phrase selection objective during the training stage. At all time steps when an input is a phrase, H number of randomly selected phrases that are different from the ground truth input is feed into the phi-LSTM model as shown in Fig. 6. The model will then produce two outputs, which are the next word prediction solely based on the actual input, and a classifier output that distinguishes the actual one from the rest. Though the number of inputs at these time steps increases, the memory cell and hidden state that is carried to the next time step keep only information of the actual input. The cost function for phrase selection objective of a sentence is where P is the set of all time steps where the input is phrase, h tsk is the hidden state output at time step t s from input k, and y tsk is its label which is +1 for the actual input and -1 for the false inputs. W ps ∈ R K×1 is trainable parameters for the classifier while κ tsk scales and normalizes the objective based on the number of actual and false inputs at each time step. The overall objective function is then This cost function is minimized and backpropagated with RMSprop optimizer [37] and trained in a minibatch of 100 image-sentence pair per iteration. We crossvalidate the learning rate and weight decay depending on dataset, and dropout regularization [38] is employed over the LSTM parameters during training to avoid overfitting. Image Caption Generation Generation of textual description using the phi-LSTM model given an image is similar to other statistical language models, except that the image relevant phrases are generated first in the lower hierarchical level of the proposed model. Here, embedded image feature of the given image followed by the start token of phrase are inputted into the model, acting as the initial context required for phrase generation. Then, the probability distribution of the next word over the vocabulary is obtained at each time step given the previous contexts, and the word with the maximum probability is picked and fed into the model again to predict the subsequent word. This process is repeated until the end token for phrase is inferred. As we usually need multiple phrases to generate a sentence, beam search scheme is applied and the top K phrases generated are kept as the candidates to form the sentence. To generate a description from the phrases, the upper hierarchical level of the phi-LSTM model is applied in a similar fashion. When a phrase token is inferred, K phrases generated earlier are used as the inputs for the next time step. Keeping only those phrases which generate positive result with the phrase selection objective, inference on the next word given the previous context and the selected phrases is performed again. This process iterates until the end token is inferred by the model. Some constraints are added here, which are i) each predicted phrase may only appears once in a sentence, ii) maximum number of unit (word or phrase) that made up a sentence is limited to 20, iii) maximum number of words forming a phrase is limited to 10, and iv) generated phrases with perplexity higher than threshold T are discarded. Datasets The proposed phi-LSTM model is tested on two benchmark datasets -Flickr8k [15] and Flickr30k [16], and compared to the state-of-the-art methods [5][6][7]9,33]. These datasets consist of 8000 and 31000 images respectively, each annotated with five ground truth descriptions from crowd sourcing. For both datasets, 1000 images are selected for validation and another 1000 images are selected for testing; while the rest are used for training. All sentences are converted to lower case, with frequently occurring punctuations removed and word that occurs less than 5 times (Flickr8k) or 8 times (Flickr30k) in the training data discarded. The punctuations are removed so that the image descriptions are consistent with the data shared by Karpathy and L. Fei-Fei [7]. Results Evaluated with Automatic Metric Sentence generated using the phi-LSTM model is evaluated with automatic metric known as the bilingual evaluation understudy (BLEU) [39]. It computes the n-gram co-occurrence statistic between the generated description and multiple reference sentences by measuring the n-gram precision quality. It is the most commonly used metric in this literature. Table 1 shows the performance of our proposed model in comparison to the current state-of-the-art methods. NIC [6] which is used as our baseline is a reimplementation, and thus its BLEU score reported here is slightly different from the original work. Our proposed model performs better or comparable to the state-of-the-art methods on both Flickr8k and Flickr30k datasets. In particular, we outperform our baseline on both datasets, as well as PbIC [33] -a work that is very similar to us on Flickr30k dataset by at least 5-10%. As mentioned in Section 5, we generate K phrases from each image and discard those with perplexity higher than a threshold value T, when generating the image caption. In order to understand how these two parameters affect our generated sentence, we use different K and T to generate the image caption with our proposed model trained on the Flickr30k dataset. Changes of the BLEU score against T and K are plotted in Fig. 7. It is shown that K does not have a significant effect on the BLEU score, when T is set to below 5.5. On the other hand, unigram and bi-gram BLEU scores improve with lower perplexity threshold, in contrast to tri-gram and 4-gram BLEU scores that reach an optimum value when T =5.2. This is because the initial (few) generated phrases with the lowest perplexity are usually different variations of phrase describing the same entity, such as 'a man' and 'a person'. Sentence made with only such phrases has higher chance to match with the reference descriptions, but it would hardly get a match on tri-gram and 4-gram. In order to avoid generating caption made from only repetition of similar phrases, we select T and K which yield the highest 4-gram BLEU score, which are T =6.5 and K =6 on Flickr8k dataset, and T =5.2 and K =5 on Flickr30k dataset. A few examples are shown in Fig. 8. Comparison of phi-LSTM with Its Sequence Model Counterpart To compare the differences between a phrase-based hierarchical model and a pure sequence model in generating image caption, the phi-LSTM model and NIC [6] are both implemented using the same training strategy and parameter tuning. We are interested to know how well the corpus is trained by both models. Using Fig. 7: Effect of perplexity threshold T and maximum number of phrases used for generating sentence, K on BLEU score. the Flickr8k dataset, we computed the corpus information of i) the training data, ii) the reference sentences in the test data and iii) the generated captions as tabulated in Table 2. We remove words that occur less than 5 times in the training data, and it results in 4833 words being removed. However, this reduction in term of word count is only 2.48%. Furthermore, even though the model is evaluated in comparison to all reference sentences in the test data, there are actually 1228 words within the references that are not in our training corpus. Thus, it is impossible for the model to predict those words, and this is a limitation on scoring with references in all language models. For a better comparison with the 1000 generated captions, we also compute another reference corpus based on the first sentence of each test image. From Table 2, it can be seen that even though there are at least 1187 possible words to be inferred with images in the test set, the generated descriptions are made up from only 128 and 154 words in NIC [6] and phi-LSTM model, respectively. These numbers show that the actual number of words learned by these two models are barely 10%, suggesting more research is necessary to improve the learning efficiency in this field. Nevertheless, it shows that introducing the phrase-based structure in sequential model still improves the diversity of caption generated. To get further insight on how the word occurrence in the training corpus affects the word prediction when generating caption, we record the top five, most trained words that are missing from the corpus of generated captions, and the top five, least trained words that are predicted by both models when generating description, as shown in Table 3. We consider only those words that appear in the reference sentences to ensure that these words are related to the images in the test data. It appears that the phrase-based model is able to infer more words which are less trained, compared to the sequence model. Among the top five words that are not predicted, even though they have high occurrence in the training corpus, it can be seen that those words are either not very observable in the images, or are more probable to be described with other alternative. For example, the is a more probable alternative of another. A few examples of the image description generated with our proposed model and NIC model [6] are shown in Fig. 9. It can be seen that both models are comparable qualitatively. An interesting example is shown in the first image where our model mis-recognizes the statue as a person, but is able to infer the total number of "persons" within the image. The incorrect recognition stems from insufficient training data on the word statue in the Flickr8k dataset, as it only occurs for 48 times, which is about 0.015% in the training corpus. Additional Results In order to further demonstrate the capability of our proposed model -the phi-LSTM, additional results from the test set of both Flickr8k and Flickr30k datasets are shown in Fig. 10 and Fig. 11, respectively. The results are selected such that images with very similar content are not repeatedly reported. 10 shows the outputs from the Flickr8k dataset. In the first row, it can be seen that our proposed model is able to distinguish different actions performed by the same subject (i.e. dog), from "playing in the field" to "racing" to "jumping to catch a toy". In the second row, we demonstrate the capability of the proposed phi-LSTM model in identifying three different sports with very similar appearance in action. In particular, our model managed to detect and recognize a bicycle in the third image, even though the size of the bicycle is very small. Beside that, we also show that our proposed model is able to determine the number of subject(s) to certain extent. For example, it can identify "two dogs" and "a group of women". Fig. 11 presents the outputs from the Flickr30k. Images in first row show three running actions performed by a dog and a horse in different scenes, in which the captions generated by our proposed model have correctly described them. Then, all images in the second row and the first image in the third row once again demonstrate the capability of the phi-LSTM in identifying subjects, number of subjects and scene correctly. The last two images in the third row show that our proposed model is capable of recognizing a bike, regardless the object is displayed in a partial view or a complete view. Lastly, all images in the final row display different subjects in the water, and our proposed method is able to describe each of the subjects correctly (i.e. girl, man and surfer). Also, note that these results show that the captions generated from our proposed model are in free form, instead of fixed template like subject-verb-object or subject-action-scene. Some descriptions may describe the scene while others may not, and verb is also an optional in the description generated. The only recurring element is the subject, which is essential in the task of image description. Fig. 12 shows some examples of our proposed method that have some errors in the generated captions, such as the number of subjects, actions, negligence of simultaneous action performed by subject and more specific object etc. However, it is still able to generate description that is somewhat related to the image. From our investigation, in any case, there are hardly any generated captions that infer a totally unrelated subject in the test set. Conclusion In this paper, we present the phi-LSTM model, which is a neural network model trained to generate reasonable description on image. The model consists of a CNN sub-network connected to a two-hierarchical level RNN, in which the lower level encodes noun phrases relevant to the image; while the upper level learns the sequence of words describing the image, with phrases encoded in the lower level as a unit. A phrase selection objective is coupled when encoding the sentence. It is designed to aid the generation of caption from relevant phrases. This design preserves syntax of sentence better, by treating it as a sequence of phrases and words instead of a sequence of words alone. Such adaptation also splits the content to be learned by the model into two, which are stored in two sets of parameters. Thus, it can generate sentence which is more accurate and with more diverse corpus, as compared to a pure sequence model.	2016-08-26T08:34:55.694Z	2016-08-20T00:00:00.000	{ "year": 2016, "sha1": "e07183a8d585a9c5860ea712095fbff6ab893103", "oa_license": null, "oa_url": "http://arxiv.org/pdf/1608.05813", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "cb4fd4651f0ddd931173ff00cbd23d8e2edaa4e8", "s2fieldsofstudy": [ "Computer Science" ], "extfieldsofstudy": [ "Computer Science" ] }
119257429	pes2o/s2orc	v3-fos-license	Pressure dependence of two-level systems in disordered atomic chain The dependence of two-level systems in disordered atomic chain on pressure, both positive and negative was studied numerically. The disorder was produced through the use of interatomic pair potentials having more than one energy minimum. It was found that there exists a correlation between the energy separation of the minima of two-level systems Delta and the variation of this separation with pressure. The correlation may have either positive or negative sign, implying that the asymmetry of two-level systems may in average increase or decrease with pressure depending on the interplay of different interactions between atoms in disordered state. The values of Delta depend on the sign of pressure. I. INTRODUCTION As it was ascertained in last years a number of anomalous properties of glasses at low temperatures can be explained supposing that there exist anomalous low-energy excitations: two-level (tunneling) systems (TLSs) [1,2] and quasi-localized modes [3][4][5]. A characteristic property of these excitations is their remarkable sensitivity to different impacts including external pressure P . The pressure dependence of TLSs in glasses, on the basis of rather general arguments, was considered by Phillips [6]. He concluded that the anomalous (negative) value of the Grüneisen constant for vitreous silica may be explained supposing that the TLSs, when pressure is applied, have a slight tendency to become less asymmetric. Such a tendency means that the energy difference between the two potential energy minima of a TLS ∆ and the derivative of ∆ with respect to P are to some extent correlated, so that the mean value ∆ ∂∆/∂P differs from zero (and is negative). In [7], it was shown that the temperature cycling effect on the width of spectral holes, burned in a dye-doped polymer glass at high pressure and low temperature can be understood if one assumes that under external pressure the higher minimum in the potential energy of TLSs gains more energy than the lower minimum. This means that in polymer glasses there also exists a correlation between ∆ and ∂∆/∂P ; however, it is of the opposite (i.e. positive) sign, so that TLSs in these glassy materials become more asymmetric with increasing pressure. Although the existence of correlation between ∆ and ∂∆/∂P in amorphous solids is expected, it is of interest to verify this property of disordered state by direct molecular dynamics simulation. For checking this correlation numerically, one must bear in mind that the concentration of TLSs in amorphous state is usually rather low and therefore quite large clusters of atoms should be involved in computations, making them fairly intricate and * Electronic address: shell@fi.tartu.ee time consuming. A relatively simple opportunity here is to compute one-dimensional (1D) disordered system consisting of a reasonably moderate number of atoms. In this study, we have carried out the computations of a disordered atomic chain. To produce a disordered state, special pair potentials having more than one energy minimum were employed. In this model, already in a single-component case, one gets the states of the chain with interatomic bonds of different length. In our computations we employed the pair potentials of two types: the potential Z1 in Ref. [8] with Friedel oscillations and the Schilling piecewise parabolic double-well potential [9], which has a discontinuity of its first derivative at the barrier separating the minima. II. DISORDERED CHAIN WITH SMOOTH MULTI-WELL PAIR-POTENTIAL Let us first discuss a monatomic chain with the pair potential Z1 [8] V (r) = a e αr cos (2kr)/r 3 where r is the distance between atoms in proper units. Values of the parameters a = 1.58, α = −0.22, b = 4.2 · 10 8 , σ = 0.331, n = 18 are the same as those given in [8] for the potential Z1. The quantity V 0 is a constant (we take V 0 = 0). This potential has the sequence of minima and maxima describing the Friedel oscillations. This-type potentials are typical for metals where the oscillations describe the effect of screening of the electric field by free electrons. The parameter k determining the period of oscillations is given by the Fermi wave vector. This parameter will be varied below to show that one can observe distinct kinds of response of the chain to the applied pressure. In our computations we examined the chain consisting of 400 atoms. We took into account the interactions between 24 nearest atoms (inclusion of more interactions did not noticeably change the results). The disordered state was generated by randomly selecting 20 percent of the nearest neighbor pairs of atoms and placing them at the distance, corresponding to the second minimum (situated at r ≈ 1.8) of the potential given by Eq. (1). All the other nearest neighbor pairs were initially placed at the distance (r ≈ 1.1), corresponding to the first (main) minimum. The initial configuration of atoms was allowed to relax to the local potential minimum of the chain. In order to find the relaxed configuration, the equations of motion of atoms were integrated using the forth-order Runge-Kutta algorithm. The motion towards the relaxed configuration was accomplished step by step, zeroing of the velocities of all atoms at every five to ten time steps. This procedure was repeated until only negligibly small changes (less than 10 −8 in our units) in the positions of all atoms were observed, indicating that the final configuration of the chain was reached. TLSs in the present model are located at the boundaries of "dense" and "rarefied" islands (groups of atoms with small (r ≈ 1.1) and large (r ≈ 1.8) distance, respectively). The tunneling transition in a TLS corresponds to the motion of the atom(s) located at the boundary of short and long bonds, resulting in the permutation of two bonds, i.e. in the displacement of a long bond to another position; see, e.g. the changes in the fragment of the atomic chain in the picture below -displacements of atoms (and bonds) cause the transitions between the configurations presented in the different lines of the picture. Our first computation was done for the chain with free ends. In the first run, the positions of atoms of the relaxed initial configuration were determined and the potential energy of the chain and its length L 0 were found. After the first run, one of the long bonds was displaced to one of the nearest positions (this displacement corresponds to the single-particle tunneling transition). Then the new coordinates of atoms in the relaxed configuration and the new length were computed and the new potential energy of the chain found. The difference of this energy value from the value corresponding to the initial relaxed configuration gave us the magnitude of ∆. The described runs were repeated with displacements of all other long bonds to the nearest positions, and the values of ∆ of all other single-particle TLSs were found. Our computations revealed that the largest \|∆\| belongs to the TLSs which correspond to the configurations presented in the first and second lines of the picture. If we mark the short bond by 1 and the long bond by 2, then the back and forth tunneling transitions in these TLSs can be denoted as (1 1 2 2 1) ↔ (1 2 1 2 1). The next largest \|∆\| belongs to the TLSs which correspond to the configurations presented in the second and third lines of the picture (the tunneling transition in these TLSs can be denoted as (1 2 1 2 1) ↔ (2 1 1 2 1)). The configurations with more than two adjacent long bonds, e.g., (1 2 2 2 1), (1 2 2 2 2 1), etc., lead also to the TLSs with a comparable \|∆\|. The concentration of these structures in our case is the smaller the larger is the "rarefied" island. In the model under consideration, the structures with extra long bonds corresponding to the third, the fourth, etc. minimum of the potential given by Eq. (1) can also exist. However, these structures were ignored relying on the physical argument of their weak stability. The energy difference ∆ of a TLS depends also on the degree of disorder. This dependence is caused by the interaction of the described above central atoms of the TLS with the surrounding atoms. This conclusion is supported by consideration of the chains with lower and higher concentration of long bonds. Namely, it was found that in the first case the dispersion of the values of ∆ of the TLSs with the same configuration of the central atoms is diminished and in the second case enlarged. We have also studied the cooperative TLSs that correspond to a simultaneous displacement of pairs and larger groups of atoms (e.g. tunneling transitions (1 1 2 2 1 1) ↔ (2 1 1 2 1 1), (1 2 1 2 1 2) ↔ (2 1 2 1 2 1), etc). It was found that the main conclusions which were drawn above for single-particle TLSs, hold also for cooperative TLSs. However, as more particles have to move, the effective mass of the tunneling group of atoms is increased, which implies rather long relaxation times. Therefore, the contribution of such TLSs to low-temperature characteristics can be neglected on a time scale much shorter than the corresponding relaxation time [10]. B. Isobaric and isochoric tunneling transitions The difference of the lengths of different configurations of the chain with free ends is very small as compared to L 0 . The larger the L 0 is the smaller is the pressure and the work required for bringing the lengths of these configurations to the same length L 0 . Thus, for sufficiently large L 0 , the energies of the configurations of the chain with free ends are practically equal to the energies of the corresponding configurations of the chain with fixed length L 0 . The same is true for the corresponding values of ∆. An analogous situation exists for isobaric and isochoric tunneling transitions for nonzero pressure. An external pressure applied to the ends of the initial configuration of the chain causes the change of the length of the chain from L 0 to L. In the isobaric case (the pressure P is fixed) after a tunneling transition the latter length in its turn changes from L to L + dL. The difference ∆ of the energy of corresponding configurations includes the term P dL. The work required to bring these configurations to the same length L, i.e. to the isochoric case, is also P dL if we assume the change of P required for bringing the lengths of these configurations to the same length L to be negligibly small. The larger the chain is the better this assumption is fulfilled, and the smaller is the difference between the isobaric and isochoric ∆ values. In our case of the chain of 400 atoms the difference is less than 0.1 %. Here, following [6] we consider the isochoric tunneling transitions. We take pressure P in reduced dimensionless units according to the relation P ≡ (L 0 − L)/L 0 . The case P = 0 corresponds to the free ends of the initial (i.e. prior to the tunneling transition) configuration of the chain. C. Effect of compression In the next step the effect of positive pressure (in our case a uniaxial compression of the chain) was examined. Here the positions of the end atoms were fixed so that the length of the chain L would be the same (in the P = 0 case) or shorter (in P > 0 case) than the length L 0 of the initial relaxed configuration of the chain with free ends. The initial positions of all other atoms were chosen so that the distances between atoms would be uniformly reduced. By using this initial condition, the relaxed positions of all by end atoms were computed and the potential energy of the chain found. Then in a similar manner as in the runs without pressure, one of the long bonds was displaced to another (nearest) position, the new relaxed configuration was computed and the new potential energy found. The difference between this energy value and the value corresponding to the initial relaxed configuration gave us the energy difference of the two minima of a TLS at a pressure P , denoted by ∆(P ). The described runs were also repeated with displacements of all other long bonds, and the values of ∆(P ) of all other TLSs were found. Figures 1 and 2 show the obtained dependence of ∆(P ) − ∆(0) on ∆(0) for different TLSs at a small pressure P = 0.001 in the case of the potential Z1 given by Eq. (1) with the parameters k = 4.12 and k = 4.5, respectively. From these figures it is clearly seen that ∆(0) and ∆(P ) − ∆(0) are correlated. The sign of the correlation may be different: in the case k = 4.12 (Fig. 1) the correlation is positive (TLSs become more asymmetric under pressure), in the case k = 4.5 (Fig. 2) it is negative. The correlation is almost complete. However, there exist intermediate cases (with overall weak dependence of ∆ on P ) where the correlation is insignificant (see, e.g., Fig. 3). Note further that in the case k = 4.12 the minimum of a TLS with a higher energy corresponds to a larger length of the chain with free ends. However in the case k = 4.5 the relation between energy and length is just the opposite: the minimum with larger energy corresponds to a smaller length. In both cases the larger is the length, the bigger is the change (increasing) of the energy with pressure. This can be understood as follows -to compress the chain extra energy is needed, which is the larger the larger is compression. Unlike the case of low P values, in the case of high P values the correlation between ∆ and ∂∆/∂P was found to be always positive. In the region of high enough pressures the TLSs are more asymmetric than at zero pressure. It is clearly seen in Fig. 4, where the dependence of the energy differences ∆ of two TLSs of the types (1 1 2 2 1) ↔ (1 2 1 2 1) and (1 2 1 2 1 dimensionless pressure P is given for the chain with the same interaction parameters as in Figs. 1 and 2. The upper values of P in Fig. 4 correspond to the critical pressure at which the abrupt changes in a chain begin to occur. Such changes take place when some long bonds transform into the short bonds, which results in disappearance of respective TLSs. Obviously, this happens in such a manner that TLSs gradually increase their asymmetry. With further increase of pressure more and more TLSs disappear. At sufficiently high pressure P ∼ 0.1 there remain no TLSs at all and the chain becomes ordered. The presented results are in agreement with the experiments [11][12][13], where a reduction in the number of TLSs and soft localized modes with pressure in glasses was observed. The reduction in the number of states of a disordered structure with the growth of its density was earlier found theoretically in [14] by an analytical study of a chain of particles with Schilling piecewise parabolic double-well potential [9]. An analogous result was obtained in [15] by the molecular dynamics simulation of 32 and 108 particles with the finite-range Lennard-Jones interaction between them. D. Effect of stretching It is reasonable to expect that the response of a glass to the volume contraction and dilatation should be different [16,17]. If so then the value of ∆(P ) should depend on the sign of P . In order to control this statement we performed computations of TLSs also for negative P (in our case for uniaxial stretching of the chain). We have found that, indeed, the value of the energy ∆ of a TLS depends on the sign of pressure, which can clearly be seen in Fig. 4. Only in the case of a large \|P \| the change of ∆ with \|P \| for positive and negative P is similar: in both cases ∆ increases with \|P \|. It was also found that at negative pressure P ≈ −0.02 the chain breaks down. III. DISORDERED CHAIN WITH PIECEWISE PARABOLIC DOUBLE-WELL POTENTIAL Lastly we have performed computations of the pressure dependence of 1D disordered chain with the Schillingtype piecewise parabolic double-well potential for nearest neighbors [9,14] V 1 (r) = V 0 +a 1 (r−r 1 ) 2 Θ(r 0 −r)+[a 2 (r−r 2 ) 2 +b]Θ(r−r 0 ), (2) where Θ(x) is the Heaviside step function. In [9,14] only the case a 1 = a 2 was considered. Here the values of parameters V 0 = −0.743, a 1 = 40, a 2 = 5.4, b = 0.6, r 0 = 1.357, r 1 = 1.13, r 2 = 1.88 were chosen such that the minima of the potential given by Eq. (2) would coincide with the first two minima of the used above potential Z1. In the model [9,14], besides the interaction between nearest neighbors, also the interaction with next-nearest neighbors is taken into account. This interaction is represented by a parabolic potential We took c = 0.05 or c = −0.05 and r 3 = 4.0. The results of computations of TLSs for small dimensionless pressure P = 0.001 are shown in Fig. 5. It is clearly seen that the quantities ∆(0) and ∆(P ) − ∆(0) are correlated, and the correlation is complete and negative. The variance of the TLSs parameters is very small -much smaller than in the case of the potential Z1. This is a consequence of a strong localization of the Schilling interactions. We have also performed computations for other values of the parameters of the potentials V 1 (r) and V 2 (r). It appeared that the sign of the difference a 2 −a 1 determines the sign of the correlation between ∆ and ∂∆/∂P : if a 1 > a 2 then the sign of the correlation is negative, thereby if a 1 < a 2 then the sign is positive. In the case a 1 = a 2 the values of ∆ do not depend on P . In the large P limit all the TLSs disappear. IV. CONCLUSION Molecular dynamics simulations of 1D disordered state unambiguously testify that for two-level systems there exists a correlation between the magnitudes of ∆ and their alterations under external pressure. The correlation may be either positive or negative sign. This signifies that the asymmetry of two-level systems may in average increase or decrease with pressure, depending on the interplay of interactions between particles in the disordered state. The response of TLSs to external pressure depends on the sign of pressure.	2011-04-18T11:54:54.000Z	2009-11-23T00:00:00.000	{ "year": 2011, "sha1": "b4746f9c6d689e855b4a91852a9c320cd1d6a9cb", "oa_license": null, "oa_url": "http://arxiv.org/pdf/1104.3442", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "b4746f9c6d689e855b4a91852a9c320cd1d6a9cb", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics" ] }
258375184	pes2o/s2orc	v3-fos-license	Applying systems thinking to unravel the mechanisms underlying orthostatic hypotension related fall risk Orthostatic hypotension (OH) is an established and common cardiovascular risk factor for falls. An in-depth understanding of the various interacting pathophysiological pathways contributing to OH-related falls is essential to guide improvements in diagnostic and treatment opportunities. We applied systems thinking to multidisciplinary map out causal mechanisms and risk factors. For this, we used group model building (GMB) to develop a causal loop diagram (CLD). The GMB was based on the input of experts from multiple domains related to OH and falls and all proposed mechanisms were supported by scientific literature. Our CLD is a conceptual representation of factors involved in OH-related falls, and their interrelatedness. Network analysis and feedback loops were applied to analyze and interpret the CLD, and quantitatively summarize the function and relative importance of the variables. Our CLD contains 50 variables distributed over three intrinsic domains (cerebral, cardiovascular, and musculoskeletal), and an extrinsic domain (e.g., medications). Between the variables, 181 connections and 65 feedback loops were identified. Decreased cerebral blood flow, low blood pressure, impaired baroreflex activity, and physical inactivity were identified as key factors involved in OH-related falls, based on their high centralities. Our CLD reflects the multifactorial pathophysiology of OH-related falls. It enables us to identify key elements, suggesting their potential for new diagnostic and treatment approaches in fall prevention. The interactive online CLD renders it suitable for both research and educational purposes and this CLD is the first step in the development of a computational model for simulating the effects of risk factors on falls. Supplementary Information The online version contains supplementary material available at 10.1007/s11357-023-00802-9. Introduction Falls among older adults are a major and increasing public health problem.Annually, one-third of individuals over the age of 65 falls at least once, and 20% of these falls lead to severe injuries [1][2][3].Falls are often multifactorial.In unexplained and recurrent falls, cardiovascular diseases are relatively frequently present, but not always recognized [4,5]. Orthostatic hypotension (OH) is an established and one of the most common cardiovascular risk factors for falls [6,7].Thus, as recommended by the recently published World Guidelines on Falls Prevention, assessment and treatment of OH is a standard component of the multifactorial fall prevention approach [4].A recent systematic review and meta-analysis [8] showed an almost doubled risk of falls in older adults for (OR 1.73; 95% CI 1.50-1.99).The incidence of OH increases with age and has been shown to contribute to up to one-third of the fall incidents in older individuals [6,7,9].Analogous to falls, OH has a multifactorial etiology [10,11], making OHrelated falls a particularly complex health concern.The best-known pathway from OH to falls involves the direct effect of inadequate brain perfusion upon standing [12].However, various other contributing mechanisms are involved and needs to be considered in older adults.For example, cerebral white matter lesions (resulting from recurrent episodes of cerebral hypoperfusion due to OH) may lead to motor dysfunction or cognitive impairment, which in turn contribute to falls indirectly [13].Also, OH is linked to poorer physical functioning, which is a fall risk factor [14].However, most studies in the field are observational in nature, and generally focused on single pathophysiological routes.A comprehensive understanding of the interactions between causes is lacking and these studies do not capture the complex interactions of other intrinsic and extrinsic risk factors superimposed on the normal aging process of the individual.An in-depth understanding of the various interacting pathophysiological pathways contributing to OH-related falls is essential to identify critical fall preventive factors and such an overarching understanding requires interdisciplinary collaboration between disciplines that now mainly focus on single pathophysiological routes. Applying systems thinking offers a methodology to understand the behavior of complex systems [15][16][17][18] and thus be helpful in unraveling the interactions and pathways between OH and falls.An important concept in systems thinking is the causal loop Vol.: (0123456789) diagram (CLD) [18]: a conceptual model of relevant mechanisms and interactions developed by experts in the field, which can highlight the dynamic nature of an issue and help explore the multiple, interacting feedback mechanisms within a system of interest [19].This can lead to an improved understanding of a complex system, such as OH-related falls.Thus, the aim of this study was to develop an expert consensus CLD on OH-related falls, by combining and weighing existing and evolving evidence of the processes involved and to identify those factors with the highest fall preventive potential for use in clinical practice. Group model building Our CLD was developed by following the structure of Group Model Building (GMB) approach, a participatory system thinking approach in which experts engage in the process of developing conceptual or computational models [20][21][22].In GMB, the perceptions and knowledge of experts are elicited and captured in a shared model, which is the result of consensus in the group, and the resulting model is a summary of explicit, tested, and integrated knowledge of the group [17,21,22].For the formation of our (national and international) expert group (Supplementary Table 1), we identified the clinical areas of knowledge relevant to OH-related falls, namely internal medicine, geriatrics, physiology, clinical pharmacology, rehabilitation medicine, cardiology, and neurology as well as methodological experts from the field of computational science, knowledgeable on the clinical subject.We aimed to include a variety of expertise, but also ensure some overlap.The experts were derived from established groups, namely the Dutch Syncope Society and the European Geriatric Medicine Society (EuGMS) Special Interest Group on Falls and Fractures.Methodological experts (computational science) were invited through the network of consortium members.This resulted in a total of 16 national and international experts.The study was conducted according to the principles in the Declaration of Helsinki.The study was exempted from ethical approval as there was no infringement of the physical and/or psychological integrity of the participants.Before participating in the project, the experts received written information about the background, aims, and process of the project.Consent to participate was given before entering the project. First, we held three general GMB sessions resulting in the backbone structure of the CLD.Based on the first findings, the variables were categorized into three intrinsic domains (cerebral, cardiovascular, and musculoskeletal), and one extrinsic domain that included variables devoid of incoming connections (including relevant medications, temperature, and alcohol variables).Subsequently, nine domainspecific subgroup sessions were organized based on the main research topics of the experts.The meetings were organized over the course of one year (2021), with several weeks in between.Length of the meetings varied between 1.5 and 3 h. During the GMB meetings, a facilitator led the GMB discussions, and a computational modeler sketched the connections in the CLD (Vensim, Ventana Systems Inc. 2022) in real time.A graphical representation of the GMB approach for developing a CLD is shown in Fig. 1.First, each expert was encouraged to propose three variables the expert deemed most relevant, and to suggest how these variables would be linked to other variables in the CLD.After that, a plenary discussion followed until a consensus was reached on the new variables and connections.All connections were confirmed by scientific evidence from the literature, where after each meeting experts were asked to provide relevant literature for their included variables and connections and a comprehensive literature search was conducted to find relevant literature for the included variables and connections.If after reviewing the published literature and consultations with the experts there remained uncertainties about the validity of connections, the connections were considered hypothetical. Causal loop diagram A CLD is a graphical representation of different pathways describing the factors (called nodes) and their interrelatedness (called connections, both known and hypothesized) relevant to complex problems under study [18,23].Positively connected variables in the CLD (displayed as "+") follow the same direction: when a causal variable increases, the variable it is linked to also increases; if the variable decreases, its linked variable also decreases.Negatively connected variables Vol:.(1234567890) ("−") have an opposite direction; when a causal variable increases, the variable it is linked to decreases, and vice versa [18,24].For some connections, the polarity can change depending on certain conditions.For example, the negative polarity from blood pressure to stroke volume can change when blood pressure gets too high (Frank-Starling mechanism) [25].In general, when heart rate increases, stroke volume increases as well.However, when heart rate increases too high, stroke volume decreases [26].Connections like these have double polarity in the CLD ("+/−"). Feedback loops are important features within a CLD [18].Reinforcing feedback loops (displayed as "R") accelerate/strengthen change and potentially disturb the system, whereas balancing feedback loops ("B" sign) counteract change and promote stabilization of the system [18,23,24].Hypothesized loops ("H" sign) can be either reinforcing or balancing but contain connections for which evidence from the literature is limited or evolving.Loops up to length 5 were identified, using Vensim [24], and the potential clinical relevance was described in a narrative in the results section.Also, an interactive (visualization) of the CLD was created in Kumu (2022; https:// kumuio/ 2022) [27] to allow visual inspection and analysis of the CLD, and to easily navigate the CLD and the underlying scientific evidence for variables and links between variables. Network analysis The resulting CLD is represented as a graph or network of relationships among a set of variables and thus can be interpreted to form a network structure.Network analysis provides a suite of quantitative techniques that can summarize the structure of a network and quantify the importance of its elements [27,28].Network analysis on the CLD may facilitate the identification of the key drivers in CLDs by quantifying their structural importance in the system [27].Although the identification of feedback loops can be seen as a form of analyzing the network [28], betweenness centrality (BC) and closeness centrality (CC) are frequently used measures for network analysis [29].BC measures the extent to which each variable (node) lies on the shortest paths between other variables (nodes) in the network [27,30].High BC variables might therefore have a mediating function, making them potential targets for interventions [18,[29][30][31].As a complementary, CC measures how close a variable (node) is to the other variables (nodes) in a network [30].It may relate to the speed or efficiency with which one variable connects to other variables [29,31].Variables with high CC have the shortest distance to many other variables, suggesting that they may rapidly exert their effects in the network.Variables with both high BC and CC may play a central role in the CLD and could be relevant to inform potential intervention approaches [18,29].Analysis and interactive visualization of the network analysis was conducted using the Kumu [], which applies well-established algorithms for computing network statistics, based on the algorithm for calculating shortest paths provided by Freeman and Brandes [30][31][32]. Results Our CLD contains 50 variables and 181 connections between them (Supplementary Fig. 1).An interactive version of the CLD can be found online (https:// falls cld.kumu.io/ under stand ing-the-multi causa lity-betwe en-ortho static-hypot ension-and-falls-19a96 c2e-a7a1-47b0-a3bf-53d6e a837d de).Three main intrinsic domains (cerebral, cardiovascular, and musculoskeletal) of causal loop diagram for orthostatic hypotension-related falls are shown in Fig. 2. Variable definitions and supporting literature evidence for the connections can be found in Supplementary Tables 2 and 3. We identified 65 feedback loops (37 reinforcing, 21 balancing, and 7 hypothesized) in the CLD.An overview of all the identified feedback loops can be found in Supplementary Table 4.An overview of betweenness and closeness centrality for the variables in the CLD is shown in Fig. 3.In this section, we first describe the main physiological balance system and then report the results from three intrinsic domains (cerebral, cardiovascular, and musculoskeletal), and one extrinsic domain. The main physiological balancing systems that ensure the maintenance of cerebral perfusion upon standing are shown in B1, B3, and B11 (Fig. 4a, b and Supplementary Table 4).A change from supine to standing position leads to pooling of blood (up to 1L) in the lower extremities and splanchnic vasculature, decreasing venous return and stroke volume and subsequently causing a drop in blood pressure.In order to restore blood pressure and maintain adequate cerebral blood flow, different balancing physiological/homeostatic effects act in concert [33]: when blood pressure drops due to orthostatic challenge/ gravitational effects, baroreceptors, as part of the short-term blood pressure regulation, are unloaded to restore blood pressure (B9, Fig. 4b and Supplementary Table 4).Also, with (orthostatic) blood pressure drops, cerebral blood flow decreases, and cerebral oxygenation decreases, triggering diminished baroreflex firing, resulting in less vasodilation, therefore reducing the occurrence of OH (B9, B10, B11, Fig. 4b and Supplementary Table 4). Cerebral domain In the cerebral domain, cerebral blood flow has high BC (Fig. 3), connecting to various variables within Fig. 2 Three main intrinsic domains (cerebral, cardiovascular, and musculoskeletal) of causal loop diagram for orthostatic hypotension-related falls.The variables of the diagram were categorized into three intrinsic domains: cerebral (in red), cardiovascular (in green), and musculoskeletal (in orange) based on the organ system, and the two key variables (in blue), with the (causal) connections between these variables.A positive connection (+) represents an effect in the same direction, e.g., an increase/decrease in "X" causes similar change in "Y," whereas a negative connection (−) represents an effect in the opposite direction, e.g., an increase/decrease in "X" causes opposite change in "Y."A hypothesized connection is shown as a dotted line.Reinforcing feedback loops are indicated with "R," balancing feedback loops with "B" and feedback loops that contained hypothesized (dotted) connection with "H."The size of the variables is scaled by their betweenness centrality.An online interactive version is available at https:// falls cld.kumu.io/ under stand ing-the-multi causa lity-betwe en-ortho static-hypot ensionand-falls-19a96 c2e-a7a1-47b0-a3bf-53d6e a837d de Vol:.(1234567890) and across domains (Fig. 4 and Supplementary Table 3 and 4).Given its high centrality, decreased cerebral blood flow could be a central factor from the cerebral domain in OH-related falls and may play a vital role in (pre)syncope.Cerebral blood flow directly influences cerebral oxygenation (positive polarity) with a (negative) connection to falls.When compensatory mechanisms are adequate, however, a decrease in cerebral oxygenation does not lead to a fall (e.g., loop B1, B10-B11, Supplementary Table 4). Similarly, cerebral autoregulation is the mechanism that aims to stabilize or restore cerebral blood flow when there are changes in blood pressure.As can be seen in the CLD, baroreflex activity, blood pressure, arterial stiffness, and carbon dioxide (CO 2 ) influence cerebral autoregulation.A decrease in cerebral blood flow decreases cerebral oxygenation, which decreases baroreflex activity, and consequently (hypothetically) activates cerebral autoregulation (hypothesized reinforcing loop H6, Fig. 4a, Supplementary Table 4).In turn, this increases cerebral vascular resistance, which further decreases cerebral blood flow (static autoregulation).It is controversial whether older age by itself is related to decreased function of cerebral autoregulation.Although mixed data have been published, there is some evidence suggesting that with aging, cerebral autoregulation is negatively affected, but that these changes (if any) are only small [34,35].With older age, baroreflexes become less sensitive and there is slowing of blood pressure recovery [34,36], rendering older adults to be more vulnerable for blood pressure drops than their younger counterparts, especially when they are dehydrated and/or use vasodilating medications (e.g., nitrates and alpha blocking agents).As a result, it puts older adults at greater risk of changes in cerebral blood flow and consequently falls.In addition, carbon dioxide (CO 2 ) is well-known for exerting potent cerebral blood flow responses [40] (B4 and B5, Supplementary Table 4), but its role is highly complex, and related to the speed in which blood pressure drops evolve. Several hypothesized connections (involved in feedback loops) in the cerebral domain were proposed by experts.For instance, microvascular cerebral lesions may also contribute to OH-related falls (hypothetical H1-H5, Supplementary Table 4).In H2 (Fig. 4a), there is a direct reinforcing feedback loop from cerebral blood flow to microvascular lesions, which presents the short-term/immediate negative effect of reduced cerebral blood flow (due to a drop in blood pressure) [37,38].Hypothetical loop H4 presents the long-term effect where repeatedly decreased cerebral oxygenation causes microvascular cerebral lesions, which negatively influences neurocognitive functioning [39,40].A recent prospective study, however, did not confirm that OH resulted in white matter lesions [41].Cerebral lesions and cognitive disorders are also associated with fall risk [42]. Cardiovascular domain In the cardiovascular domain, hypertension is a variable with high BC and CC (Fig. 3).As such, it may play an essential role in OH-related falls.Noticeably, baroreflex activity is involved in multiple feedback loops (Supplementary Table 4), both reinforcing and balancing of the short-term blood pressure regulation.The central role of baroreflex activity and its engagement in multiple feedback loops makes it vulnerable to disruption of the system.For instance, in R9 (Fig. 4b; Supplementary Table 4), the orthostatic blood pressure drop stimulates release of vasopressin, which in turn increases baroreflex activity and results in blood pressure restoration [43][44][45].In B11 (Fig. 4b; Supplementary Table 4), a blood pressure drop decreases cerebral blood flow and cerebral oxygenation.Baroreflex activity decreases and vasoconstriction in turn restores blood pressure.Inadequate baroreflex effects can have a significant contribution to OH-related falls.Multiple variables in our CLD (e.g., arterial stiffness, autonomic nervous system failure, and medications) can negatively influence baroreflex functioning.For example, in autonomic failure, compensatory increase in heart rate and vasoconstriction are diminished/inadequate, potentially leading to OH.As seen in the CLD, cardiovascular diseases/conditions can have disruptive (reinforcing) effects when concomitantly present (R18-R26, Supplementary Table 4). Musculoskeletal domain In the musculoskeletal domain, physical activity shows high centrality with the highest BC and high CC (Fig. 3).As a "mediator that connects the different organ systems," physical inactivity influences both OH and falls through muscle strength, gait/balance performance, and cardiovascular variables (e.g., heart rate and cardiac output).Also note that physical activity is part of all the feedback loops in the musculoskeletal domain.All identified feedback loops in this domain (Supplementary Table 4) are reinforcing and highly interrelated, implicating that changes in variables have the potential to disrupt or strengthen the system, not only in the musculoskeletal domain, but also in other domains.Physical activity and muscle strength play a central role in the musculoskeletal domain of our CLD and have close connections to factors from the other domains.For example, in R31 (Fig. 4c; Supplementary Table 4), decreased muscle strength (e.g., in sarcopenia) negatively affects OH by increased/prolonged venous pooling, resulting in decreased stroke volume and cardiac output; in turn, this reinforces the development of OH.Similarly, R29 is nested in R31, indicating muscle strength also has a reinforcing negative influence on OH (Fig. 4c; Supplementary Table 4).Furthermore, R30 (Supplementary Table 4) shows how falls can negatively influence fall risk through changes in muscle strength: falls can lead to fear of falling [46,47], and consequently, more physical inactivity with further loss of muscle mass [48]. Extrinsic domain The extrinsic domain contains different medications, temperature, and alcohol intake.Most of the extrinsic variables have relatively high CC compared to the intrinsic variables, suggesting that the system can be influenced by external factors.Specifically, medications (e.g., antipsychotics, antidepressants, and betablocking agents) have high CC (Fig. 3), indicating Vol.: (0123456789) that they may be contributors to OH-related falls.Moreover, medications, alcohol, and temperature influence various variables in our CLD across all three intrinsic domains.For instance, high temperature causes vasodilation and vasopressin release, both involved in cerebral blood flow feedback loops (B3, B8, and R6, Supplementary Table 4).In addition, medications with vasodilating properties and alcohol can promote venous pooling or vasodilation contributing to OH (R31, Fig. 4c).Considering these variables are "extrinsic" to the system, they may be modifiable risk factors and thus can be viewed as promising starting points for interventions (e.g., performing a medication review, and switching culprit medications to safer alternatives). Discussion We developed a comprehensive conceptual model of 50 variables involved in OH-related falls and identified which of these are most relevant.In the cerebral domain, we identified cerebral blood flow as a key factor in OH-related falls based on its high centrality.Cerebral hypoperfusion and reduced cerebral oxygenation contribute to OH and (pre)syncopal symptoms including falls [49].There are several noninvasive diagnostic techniques to measure these variables.For example, cerebral blood flow can be assessed with transcranial Doppler, and cerebral oxygenation can be measured through near-infrared spectroscopy (NIRS).With NIRS, information related to regional cerebral blood flow and oxygenation can be captured real time [50].For example, in patients with unexplained syncope, NIRS measurements showed a significant decrease in frontal cerebral tissue oxygenation saturation during the head-up tilt test, and loss of consciousness when cerebral tissue oxygenation fell below 60% [51].Whether NIRS measurements are of benefit in falls prevention needs to be confirmed. In the cardiovascular domain, we identified blood pressure as an essential "mediator" in OH-related falls based on its high BC.On the one hand, hypotension can cause or aggravate OH, but the same is true for uncontrolled hypertension [52].Uncontrolled hypertension also increases the risk of cardiovascular complications (e.g., myocardial infarction and heart failure with reduced ejection fraction) [53].Therefore, hypertension should be adequately treated in older patients, even in those at risk of falling.This was illustrated by the SPRINT trial [54], in which intensive blood pressure lowering was shown to be effective in preventing major cardiovascular events also demonstrated in the older (>75 years) participants with hypertension.In their study group, targeting systolic blood pressure to <120 mmHg appeared safe, without increasing the risk of (injurious) falls or syncope [55].This was also demonstrated in the STEP study [56].However, these studies excluded the frailer older adults [56].In TILDA (The Irish Longitudinal Study on Aging) cohort (≥75 years of age), the authors demonstrated a 5-fold higher rate of injurious falls/syncope in participants who did not meet the SPRINT inclusion criteria compared to the relatively healthy SPRINT participants [56].Data from observational studies consistently show that too strict hypertension treatment may lead to OH and OHrelated falls [6]. In concurrence with this, a recent systematic review demonstrated that withdrawal of antihypertensive medications in older people is safe [57].In line with this, a non-randomized trial demonstrated that in frail older adults, OH can be improved by deprescribing antihypertensives, resulting in an reduction in OH-related falls risk [58].It is therefore important to make personalized decisions in patients with hypertension and take patient characteristics such as frailty into account [59] Besides blood pressure, we also indicated the importance of impaired baroreflex activity as a key element in OH-related falls literature [43][44][45]60].Baroreflex activity is linked to stroke volume and heart rate.Because common age-related diseases/ conditions (e.g., heart failure, autonomic failure [44]) and commonly used medications in older adults (e.g., beta-blockers) have effects on stroke volume and/or heart rate, baroreflex activity may be an important target for fall preventive interventions in older adults. We identified that in the musculoskeletal domain, all feedback loops involved physical activity, and that muscle strength and gait/balance performance are important contributing factors in OH-related falls proposed by experts.This is not surprising, as physical counter maneuvers (e.g., leg-crossing and squatting) are well-established cornerstones of OH management, and exercise interventions (especially those targeting gait, balance, and muscle strength) have been proven their efficacy in falls prevention [61,62].Literature suggests that increased muscle tensing likely reduces OH [61][62][63], but further research is necessary to confirm this. Strengths and limitations To our knowledge, we are the first to develop a comprehensive CLD within fall research.This approach is relatively novel in medical science and particularly suitable for answering complex clinical research questions [18,21] as needed to study falls prevention in older adults.We characterized the complex pathophysiological pathways (and their interrelatedness) involved in OH-related falls.We combined evidence from the literature with expert knowledge from the fields of internal medicine, geriatrics, physiology, clinical pharmacology, cardiology, neurology, rehabilitation medicine, and computational science as input for our CLD.We applied network analysis to summarize and quantify the structure of our CLD, and thus generated original data.In the field of falls prevention, we are the first to develop a CLD and perform network analysis to quantify the strength of the connections and rank the importance of the variables.Our findings add to the knowledge base that until now has been confirmed to single pathways [8,14].To the best of our knowledge, in geriatric medicine, CLDs have only been developed in the field of cognition [18,64].In their CLD paper on Alzheimers's disease, Uleman et al. also conducted network analysis and feedback loop analysis to rank the importance of variables and further analyze their CLD [18]. Our study has several limitations.First, our aim was to capture and understand the most important mechanisms between OH and falls, based on the expertise in our project group.As a result, this CLD may not be a complete conceptual model of the topic.Second, the most important variables in this conceptual model are mainly determined based on static structural features (network analysis), which may not relate to the variables' causal and dynamic importance in physiology.In addition, the CLD methodology, without translating into a stock and flow model, is unable to account for time-dependent variations [15], whereas these variations are relevant to the pathways in our model.For example, cerebral autoregulation depends on the speed of blood pressure changes (slow blood pressure changes are more effectively buffered than fast changes, whereas extremely fast changes cannot be buffered at all) [32]. Implications for clinical practice and future directives Our CLD provides a comprehensive overview of the complex and multifactorial pathways involved in OH-related falls.The CLD is freely accessible online (https:// falls cld.kumu.io/ under stand ing-the-multi causa lity-betwe en-ortho static-hypot ension-and-falls-19a96 c2e-a7a1-47b0-a3bf-53d6e a837d de) and interactive in nature: by hovering, pathways of interest can be selected, while providing literature evidence for them.Therefore, our CLD can be used by healthcare professionals to reason about patients' etiology and can also be used to explain to patients how their individual risk factors contribute to their OH-related falls risk, and how this risk can be mitigated by interventions aimed at reducing OH [19,21]. Our CLD can also inform researchers in the field of falls prevention on the most promising targets for interventions aimed at reducing fall risk in older people.This CLD can be considered a contemporary starting point for the research community.The field is evolving rapidly, however, and evidence on OHrelated falls continues to increase [8,65].We encourage colleagues to enhance and update our CLD based on the latest evidence, for example on our hypothesized connections/pathways, especially in the cerebral domain (e.g., the role of measuring cerebral blood flow in falls prevention) [34,51]. We used this CLD to identify the centrality of variables involved in OH-related falls and analyzed these variables mainly qualitatively with feedback loops [16,18,27,66].Next, we will develop a quantitative systems dynamics model of this CLD [17].This model will enable us to capture the dynamics of the processes related to OH and falls and estimate the effect size of fall preventive intervention strategies [67][68][69]. Conclusion This CLD shows the relevance and feasibility of applying systems thinking and GMB to capture the complex and multifactorial pathophysiology of OHrelated falls by combining state-of-the-art knowledge from the fields of geriatric medicine, neurology, Vol.: (0123456789) cardiology, physiology, rehabilitation medicine, and clinical pharmacology.Our CLD increased our understanding of this highly complex and major health care problem.We identified cerebral blood flow, blood pressure, baroreflex activity, and physical activity as the key elements in OH-related falls, suggesting their potential for new diagnostic and treatment approaches in fall prevention.Our CLD combines and adds to existing knowledge on single pathways and feedback mechanisms that do not fully account for the interacting behavior of a multicausal network.This CLD will be used for the development of a system dynamics model, which means it can become an important tool for simulating and predicting falls. Fig. 1 Fig. 1 Schematic representation of the process for group model building to develop a causal loop diagram Fig. 3 Fig. 4 a Fig. 3 The betweenness and closeness centrality for the variables in the causal loop diagram.The variables are ranked descending based on their betweenness centrality results	2023-04-29T06:18:12.680Z	2023-04-28T00:00:00.000	{ "year": 2023, "sha1": "cb6b5f395fa45ff73ae71e74d5b1ab4de3d702d4", "oa_license": "CCBY", "oa_url": "https://link.springer.com/content/pdf/10.1007/s11357-023-00802-9.pdf", "oa_status": "HYBRID", "pdf_src": "Springer", "pdf_hash": "4b68ca1f153d986f44252efe6f7418cbba4977f0", "s2fieldsofstudy": [ "Psychology" ], "extfieldsofstudy": [ "Medicine" ] }
237638909	pes2o/s2orc	v3-fos-license	Dense core vesicle markers in CSF and cortical tissues of patients with Alzheimer’s disease Background New fluid biomarkers for Alzheimer's disease (AD) that reveal synaptic and neural network dysfunctions are needed for clinical practice and therapeutic trial design. Dense core vesicle (DCV) cargos are promising cerebrospinal fluid (CSF) indicators of synaptic failure in AD patients. However, their value as biomarkers has not yet been determined. Methods Immunoassays were performed to analyze the secretory proteins prohormone convertases PC1/3 and PC2, carboxypeptidase E (CPE), secretogranins SgIII and SgII, and Cystatin C in the cerebral cortex (n = 45, provided by Bellvitge University Hospital) and CSF samples (n = 66, provided by The Sant Pau Initiative on Neurodegeneration cohort) from AD patients (n = 56) and age-matched controls (n = 55). Results In AD tissues, most DCV proteins were aberrantly accumulated in dystrophic neurites and activated astrocytes, whereas PC1/3, PC2 and CPE were also specifically accumulated in hippocampal granulovacuolar degeneration bodies. AD individuals displayed an overall decline of secretory proteins in the CSF. Interestingly, in AD patients, the CSF levels of prohormone convertases strongly correlated inversely with those of neurodegeneration markers and directly with cognitive impairment status. Conclusions These results demonstrate marked alterations of neuronal-specific prohormone convertases in CSF and cortical tissues of AD patients. The neuronal DCV cargos are biomarker candidates for synaptic dysfunction and neurodegeneration in AD. Supplementary Information The online version contains supplementary material available at 10.1186/s40035-021-00263-0. dystrophic neurites. In addition to the loss of neural circuitry and neuronal cell bodies, cerebral amyloid angiopathy, granulovacuolar degeneration (GVD), and Hirano bodies with glial activation have also been suggested as part of AD pathology [8]. Reliable cerebrospinal fluid (CSF) signatures have been developed for plaque and tangle pathologies and for associated neurodegenerative processes (reduced Aβ and increased phosphorylated tau (P-tau) and total tau (T-tau) levels, respectively). These core CSF biomarkers are currently routinely used in clinical practice for diagnosis of AD in patients at the mild cognitive impairment or dementia stage of the disease [9]. Additionally, several reports have proposed certain microglial and astrocytic proteins, such as TREM2 and YKL-40, as biomarkers of glial activation in AD [10][11][12][13]. Importantly, as decreased synapses are a major quantitative correlate of loss of memory and cognition in AD brains [7], much attention is being focused on biomarkers to detect the degree of synaptic dysfunction and degeneration in early stages of AD. Novel CSF biomarker candidates for synaptic pathology include axonal and pre-and postsynaptic proteins, such as SNAP-25, Syntaxin 1B, neurogranin and neurofilament light chain [14][15][16][17]. Other potential CSF biomarkers for synaptic alterations in AD consist of released cargos of the so-called dense core vesicles (DCVs). As synaptic vesicles (SVs), DCVs belong to the regulated secretory pathway in neurons and also, presumably, in astrocytes [18,19]. In response to membrane depolarization, DCVs release a variety of neuropeptides and growth factors (e.g. brain-derived neurotrophic factor [BDNF]) as well as distinctive molecular components, such as chromogranins (Cg), secretogranins (Sg), and processing enzymes [18]. Although DCVs do not accumulate within synaptic compartments as do SVs, the vast majority of fusion events occur at synaptic boutons and axons [20,21]. Interestingly, a major and ubiquitous constituent of the DCV matrix, CgA, was one of the first biochemical biomarker candidates for AD synaptic degeneration [22,23]. Recent advances in proteomic technology have identified secreted DCV proteins in CSF screens for AD patients, such as the prototype granins CgA and CgB and the non-classical granin SgVII (usually called VGF, the nerve growth factor inducible protein VGF) [24,25]. These observations suggest that DCV proteins may be promising biomarkers of synaptic loss in AD. However, changes in DCV cargos in AD CSF observed in different proteomic and immunological analyses offer variable results, or have not been completely validated [26,27]. Here, we investigated changes in DCV proteins in the CSF and brain samples of AD patients. Specifically, we examined the neuronal prohormone convertases PC1/3 and PC2, and the neuronal and astroglial carboxypeptidase E (CPE), SgIII and SgII [28][29][30][31][32][33]. Additionally, we examined the secretory protein Cystatin C (CysC), a neuronal and astrocytic protease inhibitor involved in AD, located mainly in lysosomes and possibly in DCVs [34,35]. Brain tissues Post-mortem human AD (n = 23) and non-AD (n = 22) brain samples (aged were obtained from the Institute of Neuropathology Brain Bank IDIBELL Hospital Universitari de Bellvitge (Hospitalet de Llobregat, Spain) under an agreement with the local ethics committee; demographic data are presented in Table 1. Individuals were selected based on the post-mortem diagnosis of AD according to the ABC score (A, Amyloid phases Thal; B, Braak stages of neurofibrillary tangle pathology; C, CERAD stages) following the National Institute on Aging-Alzheimer's Association (NIA-AA) clinical research criteria [36]. The AD-diagnosed subjects corresponded to A3-A4, V-VI (B3) and C3, whereas the non-AD subjects were A0, 0-I/II (B0-BI) and C0. Lumbar CSF samples Both AD patients and control subjects (healthy volunteers) were recruited from the SPIN cohort in the Memory Unit at the Hospital de la Santa Creu i Sant Pau [37]. This study was approved by the local ethics committee and was carried out in accordance with the Declaration of Helsinki. All patients (or their nearest relatives) and controls gave informed consent to participate in the study. Extensive clinical, neuropsychological, MRI and molecular examinations were performed in all subjects. CSF samples (n = 66) were collected by lumbar puncture between 9 am and noon. Centrifugation at 4 °C for 10 min at 2000 × g and storage of 500 μl aliquots in polypropylene tubes at − 80 °C were accomplished within 1 h after collection. All AD patients fulfilled clinical criteria for probable AD according to the revised NIA-AA criteria [38] and had a CSF biomarker profile consisting of decreased Aβ 1-42 plus high T-tau and P-tau levels (ELISA tests from Innogenetics, Ghent, Belgium), indicating high likelihood of being due to AD. The cut-off values we used to define our AD cohort in this study were 550 pg/ml for Aβ 1-42 , 350 pg/ml for T-tau, and 61 pg/ml for P-tau [39]. The control group was defined according to the following criteria: objective cognitive performance within the normal range (performance within 1.5 standard deviation) on all tests from a specific test battery, clinical dementia rating scale score of 0, no significant psychiatric symptoms or previous neurological disease, and a non-pathological CSF biomarker profile. The average Mini-Mental State Examination (MMSE) score was 21.6 ± 4.4 for AD patients, whereas control subjects had a score of 28 or higher. AD and control groups were well matched for age at CSF collection. Demographic information is presented in Table 2. Immunohistochemistry Human brain samples were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4, by immersion. After being cryoprotected in a 30% sucrose solution, the samples were frozen and sectioned with a cryostat. For peroxidase immunohistochemistry, histological sections were blocked for 30 min in phosphate buffered saline (PBS) supplemented with 10% methanol and 3% H 2 O 2 and then washed in PBS. Pretreatment with formic acid was used to enhance labeling of plaques. To avoid nonspecific binding, brain sections were blocked in PBS containing 10% serum, 0.2% glycine, 0.1% Triton X-100, and 0.2% gelatin for 1 h at room temperature. Incubations with the primary antibodies were carried out overnight at 4 °C in PBS containing 1% fetal bovine serum, 0.1% Triton X-100 and 0.2% gelatin. Bound antibodies were detected using the avidin-biotin-peroxidase system (Vectastain ABC kit, Vector Laboratories Inc., Burlingame, CA). PBS containing 0.05% diaminobenzidine (DAB) and 0.01% H 2 O 2 was used to visualize the peroxidase complex. Sections were mounted, dehydrated, and coverslipped in Eukitt ® (Sigma-Aldrich, Diesenhofen, Germany). Semiquantitative analysis of CysC immunoreactivity on DAB-stained slices was carried out on the outer and inner layers of the AD and control parietal cortices (n = 3, each). Images at 20 × magnification (2 per section, 2 sections per sample) were randomly taken using identical acquisition microscope parameters. The signal intensity of 0.1 × 0.1 mm 2 fields was measured using the ImageJ ® software (NIH, Bethesda, MD). Levels of CysC were initially quantified as arbitrary density units and subsequently AD values were expressed as a percent change from control measurement. Double-labeling fluorescent immunohistochemistry was carried out by incubation with different fluorochrome-conjugated secondary antibodies (Alexa Fluor 488 and Alexa Fluor 568; Molecular Probes, Eugene, OR), and cell nuclei were stained with Hoechst (Sigma-Aldrich, Diesenhofen, Germany). Endogenous autofluorescence was avoided using Sudan Black B (Sigma-Aldrich, Diesenhofen, Germany). Sections were mounted in Mowiol (Merk Chemicals Ltd., Nottingham, UK) and visualized with a Leica TCS SPE scanning confocal microscope. Colocalization analyses of DCV proteins with GVD markers were performed by ImageJ software, using the threshold-corrected Mander's correlation coeficient, which ranges between 1 (highcolocalization) and 0 (low-colocalization). The specificity of the immunolabeling was checked omitting primary antibodies, using nonspecific IgG instead of them and using a previous incubation of the primary antibodies with an excess of antigen (Proteintech Group Inc., Deansgate, Manchester, UK). No immunostaining was observed in these conditions. Western blotting and radioimmunoassay Equal volumes of lumbar CSF samples were collected in vials and brain tissues were homogenized in ice-cold lysis buffer containing 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM MgCl 2 ,1 mM EGTA, 1% Triton X-100, and a protease inhibitor cocktail (Roche Diagnostics GmbH, Mannheim, Germany). The CSF samples and postnuclear brain lysates were mixed with 3 × sample buffer (188 mM Tris-HCl, 30% glycerol, 3% SDS, 0.01% Bromophenol Blue, and 15% β-mercaptoethanol), electrophoresed in 10%-12% SDS-acrylamide gel (Bio-Rad Laboratories, Hercules, CA), and then transferred to polyvinylidene difluoride immobilization membranes (Whatman ® Schleicher & Schuell, Keene, NH). Tissue results were normalized for total protein content (data obtained from Ponceau staining scans). Membranes were blocked in 5% nonfat milk powder solution in Tris-buffered saline and Tween-20 (TBST; 140 mM NaCl, 10 mM Tris/HCl, pH 7.4 and 0.1% Tween-20) for 1 h at room temperature and then incubated with primary antibodies in blocking solution overnight at 4 °C. After numerous washes in blocking buffer, the membranes were incubated for 1 h with horseradish peroxidase-conjugated secondary antibodies (Bio-Rad Laboratories, Hercules, CA). Enhanced chemiluminescence reagents ECL ™ (GE Healthcare, Buckinghamshire, UK) and X-ray films (AGFA) were used to visualize bound antibodies. Blot images were taken with a scanner and densitometric values were obtained using a Java-based image processing software (ImageJ ® software). The amount of total proteins in brain homogenates (10-15 µg) and CSF volume (2-5 µl) and exposure times of the films were empirically determined to obtain the adequate linear range for quantitative analysis. For protein forms with different electrophoretic mobilities, a single average exposure of the film is shown in the figures, whereas quantitative analysis of each form was performed at different and suitable amounts of total protein/ volume and exposure time of the films. The SgII/secretoneurin radioimmunoassay was performed as described previously [41]. In brief, samples were denatured to avoid protein degradation and antiserum was added in the radioimmunoassay buffer and incubated for 24 h at 4 ºC. Then iodinated secretoneurin (200 cpm/μL) was added and samples were incubated for an additional 24 h at 4 ºC. The non-bound tracer was separated by dextran-coated charcoal absorption and pulled down by centrifugation. The iodinated tracer remaining in the supernatant fraction was quantified, with a detection limit of 1-2 fmol. Statistical analysis Quantitative data were statistically analyzed using GraphPad Prism 6.01 ® software (GraphPad Software, San Diego, CA) with t-test and Mann-Whitney test. All data are presented as the means ± SEM. Significance was set at P < 0.05. Correlations of the measured values were examined using the Bonferroni-corrected Spearman correlation coefficient. A significance cutoff of P ≤ 0.0019 based on Bonferroni correction of multiple comparisons and P ≤ 0.05 in non-corrected comparisons were applied. DCV markers are neuropathologically accumulated in dystrophic neurites and GVD bodies in the AD cerebral cortex To examine DCV changes in the cerebral cortex of AD patients, we performed Western blotting and immunostaining analyses in homogenates and histological sections from AD patients and age-matched controls. PC1/3, PC2, CPE, SgIII and SgII were abundantly found by immunoblotting in cortical samples (Additional file 1: Fig. S1). In addition to the unprocessed and mature forms, higher and lower electrophoretic mobility bands were detected with polyclonal antibodies against DCV proteins, likely corresponding to the described aggregated and cleaved species [30][31][32]. A robust detection for all of these proteins was noted in CSF samples, essentially their precursor and mature forms. Volumes as small as 1 µL were enough to detect some of these proteins in the CSF (i.e. SgIII). As previously reported, only the monomeric form of the secretory protease inhibitor CysC was detected in the CSF (~ 14 kDa), whereas cortical tissues also displayed oligomeric/aggregated species [48,49]. When performing analyses of similar CSF volumes, levels of membrane and cytosolic proteins such as synaptophysin (Syp) and β-actin (Additional file 1: Fig. S1), as well as the potential AD biomarker neurofilament light chain (data not shown), were under the detection limit. Thus, besides the abundance in brain tissues, DCV proteins are broadly detected in the CSF, feasibly corresponding with released species. Next, we examined the levels of distinct molecular forms of secretory proteins in AD brains by immunoblotting (n = 7 per tissue and condition) ( Fig. 1 and Additional file 1: Fig. S2). Comparing DCV protein amounts (normalized to membrane-transferred total protein) in hippocampus and parietal cortex of AD cases with controls, we only found changes for the 72-kDa precursor form of PC2 (proPC2). The proPC2 species was increased 1.7-fold in parietal cortex (P = 0.038) and 2.4-fold in hippocampus (P = 0.0006). Interestingly, the ~ 28 kDa form of CysC was significantly decreased in the parietal cortex of AD patients (35%, P = 0.038). This electrophoretic band probably corresponds to dimeric or aggregated forms of CysC, which have been associated with amyloid damage [48,49]. Finally, a prominent reduction was evident in the content of the SV-specific integral protein Syp both in parietal cortex (30%, P = 0.042) and in hippocampus (50%, P = 0.017) of AD patients. These results show that the global levels of different DCV markers are mostly preserved in AD cortices, whereas ubiquitous SV proteins are dramatically reduced. To examine the alterations of DCV proteins in situ, we performed peroxidase immunohistochemical and confocal immunofluorescence analyses in brain sections from control (n = 8) and AD cases (n = 9). In controls, PC1/3, PC2, CPE, SgIII, SgII and CysC were widely distributed throughout the cerebral cortex, consistent with previous studies [45,[50][51][52] (Fig. 2). All of these proteins were abundant in neuronal populations of the neocortex and hippocampus (Fig. 2a), although a differential and robust immunolabeling for CPE was detected filling dendritic shafts (Fig. 2a, c). Immunostaining for SgIII, CPE, and CysC was also detected in GFAP + astrocytes (Fig. 2b), whereas prohormone convertases were absent in glial cells. In neurons, subcellular structures labeled for secretory proteins included somatic and dendritic granules, varicose fibers, and terminal-like buttons, as we previously reported [45]. Granular compartments immunostained for typical DCV markers were negative for the lysosomal marker LAMP1 (not shown), whereas CysC immunofluorescence was largely associated with LAMP1 + structures (Fig. 2c). Immunostaining alterations for secretory proteins in AD cortices were mainly associated with senile plaques (Fig. 3a). All typical DCV components were aberrantly accumulated, to different degrees, in AT8 + dystrophic neurites surrounding amyloid plaques, in the parietal cortex and hippocampus (in each examined AD cases) (representative examples are shown in Fig. 3b, c). In some samples, intense immunolabeling for SgIII and CysC was detected in reactive astrocytes close to the amyloid deposits (data not shown), as described elsewhere [45,51]. No major changes were detected for secretory proteins in non-plaque areas of the AD brains, including tangle-bearing neuronal somata identified by the AT8 antigen (excepting cells showing GVD, see below). Only a slight, but consistent, increase in CysC was found in pyramidal neurons located in outer and inner layers of the AD parietal cortex (Additional file 1: Fig. S3). Unexpectedly, we found GVD-shaped structures immunolabelled for some DCV proteins in CA1 pyramidal somata (and to a lesser extent in CA2 and subiculum) in each of the four examined AD hippocampi (Fig. 4a). These granules had strong immunostaining for PC2 and moderate immunostaining for PC1/3, whereas only one AD case displayed scarce CPE + large granules (Fig. 4a). An average of 241 ± 4 CA1 pyramidal neurons containing aberrant PC2 + granules per mm 2 was observed in AD (n = 4), whereas age-matched controls were virtually devoid of these (a few cells, and in only 1 out of 5 cases). Double labeling of PC1/3, PC2 or CPE and the established GVD markers CK1δ and CHMP2B substantiated the neuropathological identity of these large granules (Fig. 4b). Mander's correlation coefficient for CK1δ overlapping with PC2 was 0.55 ± 0.13 and for PC2 overlapping with CK1δ was 0.62 ± 0.20 (n = 12), whereas Mander's coefficient for CHMP2B overlapping with PC1/3 was 0.43 ± 0.25 and for PC1/3 overlapping with CHMP2B was 0.54 ± 0.28 (n = 10). These values indicate good colocalization between DCV proteins and GVD bodies in the AD hippocampus. Additionally, an atypical colocalization of PC2 and PC1/3 with the lysosomal marker LAMP1 was also detected in GVD bodies (Fig. 4b). Interestingly, the PC2-and PC1/3immunolabelled GVD inclusions were associated with neurofibrillary tangle pathology in AD samples (Fig. 4c, d). Most pyramidal neurons displaying GVD bodies positive for PC2 (81%) and PC1/3 (77%) also contained hyperphosphorylated tau (AT8 + ). Although the GVD inclusions positive for DCV proteins were detected in neurons containing mature neurofibrillary tangles, the DCV-labeled aberrant granules were more often present in early-stages of neurofibrillary tangle formation (Fig. 4c, d). Taken together, these results show that an aberrant accumulation of DCV components is associated with two neuropathological hallmarks of the AD cerebral cortex, plaque-related dystrophic neurites and GVD bodies. Decreased levels of secretory DCV cargos in CSF from AD patients are associated with neurodegeneration markers Next, we investigated the levels of DCV proteins in 66 CSF samples from phenotypically well-characterized AD patients (n = 33) and cognitively normal controls (n = 33) with immunoblotting and radioimmunoassay methods. No differences in the CSF total protein content were found between AD and controls, either by Bradford assay or Coomassie staining of SDS-PAGE gels (Fig. 5a). However, levels of most molecular forms of DCV proteins were decreased in the CSF from AD patients compared with those from cognitively normal controls (Figs. 5b Fig. 5b. The most prominent reduction in AD was found for unprocessed PC1/3 (− 42% ± 5.7%, P = 0.0003). Although levels of the mature form were apparently lower in AD, they did not reach statistical significance (P = 0.075) (Fig. 6). When we quantified the levels of mature and unprocessed forms of PC1/3 together, a significant decrease was detected in the AD group (− 24%, P = 0.03). Consistently, we also detected an apparent PC1/3 cleaved form displaying a lower electrophoretic mobility identified by the polyclonal antibodies used in this study. When analyzing this molecular species together with the other two bands, we found decreased levels of total PC1/3 in the CSF from AD patients compared to controls (P = 0.05). In addition, each of the two bands detected by PC2 antibodies that represent precursor and mature forms, was statistically decreased in AD samples (− 16%, P = 0.024 and − 12%, P = 0.011, respectively; Fig. 6), as well as in a pooled quantification (− 18%, P = 0.0009). The DCV-contained exopeptidase CPE was reduced by ~ 20% in the CSF of the patient cohort (P = 0.016). Regarding members of the granin family, the most abundant form of SgIII in the CSF, proSgIII, was decreased in AD (P = 0.047), whereas no significant differences were detected for the processed protein. An analysis of the two forms showed no differences between AD and non-AD individuals (P = 0.612). We also determined the CSF levels of SgII/secretoneurin with an radioimmunoassay assay. No statistical differences were found between AD and control groups. Finally, a reduction of about 33% was observed for the secretory protein CysC in the CSF of AD patients (P = 0.001). To investigate the possible relationships of secretory protein forms with age, cognitive status (MMSE), and the core CSF biomarkers of AD (Aβ 1-42 , P-tau and T-tau), Spearman correlation analysis was carried out in the AD cohort (33 individuals) (Fig. 7). No molecular forms of secretory proteins were significantly correlated with age or Aβ 1-42 (Fig. 7). In contrast, we detected strong correlations, positive between the mature PC2 form and MMSE score (P = 0.0001, r = 0.6449) and inverse between PC2 and T-tau (P = 0.0004, r = − 0.5837), applying Bonferroni correction. Non-corrected analysis further evidenced an inverse correlation of P-tau with PC2 and PC1/3 and T-tau with PC1/3, and a direct correlation of MMSE with PC1/3. Additionally, corrected analysis evidenced positive correlations between levels of P-tau and CPE (P = 0.0013, r = 0.5377) and the precursor form of SgIII (P = 0.0019, r = 0.5216), whereas non-corrected analysis showed a direct correlation of levels of CPE, CysC and different forms of SgIII with P-tau and T-tau (Fig. 7). Discussion The present findings show that the levels of neuronal prohormone convertases and other DCV proteins decrease in the CSF of AD patients, and that this decrease correlates both with the cognitive impairment status and with established CSF neurodegenerative markers. In AD brains, these proteins accumulate in plaque-surrounding dystrophic neurites and reactive astroglia, whereas PC1/3, PC2 and CPE are also remarkably accumulated in hippocampal GVD bodies. In contrast to the global loss of major SV components (e.g. Syp and Rab3A) [53,54], we show that the total levels of DCV proteins remain mostly invariable in the cerebral cortex of AD patients. Only proPC2 shows a reliable increase in AD samples, mainly in the hippocampus. These observations support our previous data by Plá et al. [45] on SgIII and CPE, but differ from others [25,52,55]. Yakovleva et al. [55] have reported an increase in the PC2 form in AD parietal cortex, whereas no changes of PC1/3 and PC2 or decrease of PC1/3 has been shown in the injured frontal cortex [25,52]. Because SV markers are virtually restricted to presynaptic terminals, their global alterations likely reflect variations in synaptic contacts. However, the broad distribution of DCV proteins makes it impossible to associate their total levels with synaptic changes. In fact, in situ immunodetection (Figs. 2, 3, 4) reveals dramatic alterations of these markers at cellular and subcellular levels that are hidden in global analyses ( Fig. 1 and Additional file 1: Fig. S2). In the AD cerebral cortex, we found no apparent changes in DCV proteins in normal-shaped axons and terminal-like buttons. However, we detected accumulations of DCV cargos in somata and projections of neurons and astrocytes associated with the degenerative and inflammatory progression of the disease. Remarkably, we detected deposits of DCV components (mainly PC2 and PC1/3) in GVD bodies in pyramidal neurons in AD brain. These inclusions are believed to be aberrant lysosomes, exhibiting a central dense core, located in neurons of hippocampus and other brain areas in patients with different primary tauopathies, especially in AD [56][57][58]. Our results showing LAMP1-surrounded PC2 + inclusions in cells that contain immature neurofibrillary tangles are consistent with the idea that the emergence of GVD granules precedes the appearance of fully-formed tangles in degenerating neurons [56]. To our knowledge, the present study is the first to provide evidence of a DCV cargo in these neuropathological granules. A plethora of proteins has been identified in GVD bodies [56,[59][60][61], including cytoskeletal components and proteins of the unfolded protein response, autophagic and transduction pathways, cell stress and apoptosis, the necrosome complex, and the RNA-binding proteins, but not DCV components. Therefore, our results involve the regulated secretory pathway in AD-related GVD. This is supported by reports showing the presence of the SV protein TMEM230, the Golgi apparatus Golgin A4, and components of the endoplasmic reticulum unfolded protein response in GVD bodies [56,[61][62][63]. Because in addition to the hippocampus, the parietal cortex can also display GVD [58], our results of elevated levels of the precursor form of PC2 in AD homogenates point to non-processed PC2 as the main form in GVD bodies, which may be related to the high tendency of proPC2 to form large aggregates [64]. Table summarizing Bonferroni-corrected Spearman correlations of secretory protein forms with age, MMSE score, Aβ1-42, P-tau, and T-tau. Significant correlations (P < 0.0019) are highlighted in green. Non-corrected significant correlations (P < 0.05) are highlighted in yellow. Bottom Significant correlations are represented in scatter plots Another neuronal degenerative structure displaying prominent alteration of DCV cargos is the dystrophic neurite. Since the report of striking amounts of CgA in AD dystrophic neurites, almost every DCV marker has been detected as accumulating in these degenerative structures, including transmitter cargos (e.g. BDNF) and intrinsic molecular machinery (i.e. granin family members) but not SV components [22,45,50,54,[65][66][67]. In this study, we show the previously unrevealed accumulation of PC1/3 and PC2 in amyloid-surrounding dystrophic neurites. Thus, dystrophic neurites are a significant location for DCV retention. The recent work of Sadleir et al. [68] showing a dramatic microtubule and trafficking disruption in dystrophic neurites associated with amyloid plaques strongly substantiates the aberrant accumulation of DCVs in these swollen structures and, therefore, suggests an important impairment in their vesicular cargo release. Lastly, in addition to the neuronal changes, the increased immunostaining of SgIII and CysC in peri-plaque astrocytes is consistent with previous studies and likely corresponds with transcriptional changes in activated cells and/or an Aβ-induced deficiency in glial secretion [29,33,45,51,69]. In the CSF of AD patients, we detected that five out of the six analyzed proteins are significantly lowered compared with controls. Decreased levels of PC1/3, PC2 and SgIII have very recently been found in large-scale proteomic CSF screens from AD patients, but this has not been immunologically verified yet [24,[70][71][72][73][74], whereas variable results for CysC, CPE and SgII have been obtained [24,71,[75][76][77][78]. In general, decreased levels of different DCV cargos (i.e. CgA, VGF and 7B2) have been found in AD CSF, with only a few reports showing opposing results [26,78]. Several explanations might be offered for these discrepancies. First, the work of Duits et al. [78], illustrating high levels of CgA, SgII and VGF in CSF of patients with mild cognitive impairment progressing to AD, suggests that early events in the AD pathophysiological cascade can display opposite changes in DCV cargos in later stages. This agrees with earlier studies showing that the synaptic decline in early stages of AD is compensated by an increase in synaptic size of the remaining contacts [5]. Therefore, failure in peptidergic neurotransmission, presumably occurring in dystrophic neurites, could trigger compensatory production of DCVs in unaffected terminals. In addition, taking into account that cargos such as CPE, SgIII and CysC are robustly produced by neurons and non-neuronal cells, CSF levels could depend on the balance between neuronal impairment and glial activation in different AD stages. Since the CSF likely reflects the composition of the extracellular milieu in the central nervous system, decreased levels of secretory proteins in the CSF of patients could contribute to the pathophysiology of the disease. Based on experimental evidence, low levels of neuroactive peptides, such as VGF and BDNF, may contribute to memory and cognitive impairments, whereas decreased levels of CysC and the prohormone convertase-binding proteins 7B2 and proSAAS could promote amyloidogenesis [34,66,67,79,80]. Several mechanisms may be put forward to underlie the occurrence of low concentrations of secretory proteins in AD CSF. Although alterations in the blood-brain barrier and in extracellular peptide degradation cannot be ruled out, both transcriptional changes and release impairments may underlie the decline of secretory components in AD CSF [25,33,68,[81][82][83]. Additionally, the decreased levels of DCV proteins in AD CSF may reflect the synaptic and neuronal loss occurring in the disease. The present results of robust inverse correlation of the neuronal prohormone convertases with tauopathy markers and cognitive impairment strongly suggest that the reduced levels of these DCV cargos correspond with the spreading neural damage. Conversely, the levels of proteins expressed by both neurons and astrocytes (e.g. SgIII) directly correlate with neurodegeneration markers in AD. Within the AD cohort, besides impairment of the neuronally produced forms, it is conceivable that production of secretory proteins by activated astrocytes increases as neuronal damage worsens [29,45,51,69]. Furthermore, the connection of secretory proteins with neurodegeneration, but not with amyloid plaque pathology, is consistent with recent proteomic studies in non-AD neurodegenerative diseases, such as frontotemporal dementia, where a decline of CgA, VGF and CysC occurs in the CSF of patients [84][85][86]. The current diagnosis methods for AD include cognitive tests, neuroimaging techniques and CSF assays. However, there is still no clinical strategy available for the precise and early detection of AD. Some limitations involve disease heterogeneity, co-morbidities and overlapping between different neurodegenerative disorders [87]. Among new potential biomarkers, those that monitor synaptic alterations are needed for prognosis of disease progression and evaluation of effects of drugs on AD [88]. Because the altered DCV proteins in CSF have been implicated in synaptopathies beyond neurodegenerative diseases, such as delirium and schizophrenia [89,90], DCV cargos could be used as complementary biomarkers to follow synaptic dysfunction and loss and neurodegeneration in AD.	2021-09-27T13:30:48.934Z	2021-09-26T00:00:00.000	{ "year": 2021, "sha1": "cdf6defc0a9af1ee72b1683c41cbdc1d5afc9ec7", "oa_license": "CCBY", "oa_url": "https://translationalneurodegeneration.biomedcentral.com/track/pdf/10.1186/s40035-021-00263-0", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "31b6daa02a9338f1a35905fe3e0fa764abc920c4", "s2fieldsofstudy": [ "Biology", "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
259700027	pes2o/s2orc	v3-fos-license	Effects of insecticides and repellents on the spread of ‘Candidatus Phytoplasma solani’ under laboratory and field conditions Recent outbreaks of ‘Candidatus Phytoplasma solani’ resulted in severe losses in potatoes, vegetable crops and grapevines in certain regions of Austria and constituted a major challenge for seed potato production. Therefore, the effects of various insecticides and insect deterrents on pathogen spread were studied both in laboratory and field experiments from 2018 to 2021. In laboratory transmission experiments, field captured Hyalesthes obsoletus were caged on differently treated Catharanthus roseus for five days. The insecticides lambda-cyhalothrin, deltamethrin, esfenvalerate, acetamiprid and chlorpyriphos showed the most rapid impact on insect survival and fully prevented phytoplasma transmission. The particle film forming products kaolin and diatomaceous earth had some effect. A transfer of the promising laboratory results to potato fields, however, was achieved to a limited extent only. Treatments with pyrethroids and acetamiprid every 8–10 days over the flight period of H. obsoletus roughly halved the number of symptomatic plants and tubers in case of moderately susceptible varieties and moderate infection pressure. In the event of susceptible varieties and high disease pressure, treatment effects were hardy discernible. In practical terms, the experiments indicate that insecticide applications alone are not sufficient to mitigate the disease. Spraying of diatomaceous earth and mineral oil did not affect disease incidence in the field. Supplementary Information The online version contains supplementary material available at 10.1007/s41348-023-00768-y. Introduction 'Candidatus Phytoplasma solani' (Quaglino et al. 2013), also known as Stolbur phytoplasma, is a pathogen affecting a wide range of cultivated and wild plant species. Diseases such as Bois noir (BN) of grapevines, stolbur of potatoes and many vegetable species, maize redness and lavender decline are associated with this agent. The phytoplasma is prevalent all over Europe and the Mediterranean area (COST FA 0807 2014;CABI 2022). The spread of the pathogen is linked to certain plant species, which serve both as phytoplasma reservoir and as developmental hosts for the transmitting insect species. Development of vector instars on infected host plants results in infectious vector adults, which transmit the pathogen to various crops. Studies in different regions of Europe have identified several distinct epidemiological cycles, each of them associated with specific plant hosts and insect vectors. So far, the Cixiid Hyalesthes obsoletus is considered as the principal vector species. In central Europe, from the middle of the twentieth century onwards, field bindweed (Convolvulus arvensis) acted as dual host for vector reproduction and for the pathogen (Brčák 1979;Maixner et al. 1995), whereas in Italy Urtica dioica was the relevant reservoir plant species (Lessio et al. 2007). During the last 15-20 years, however, U. dioica associated H. obsoletus populations were oftentimes also detected in central Europe (Jović et al. 2019). In France, H. obsoletus exploits lavender as developmental host and transmits specific 'Ca. P. solani' strains inducing lavender decline (Sémétey et al. 2018). In addition, H. obsoletus populations associated with Vitex agnus-castus and Crepis foetida were reported in South-East Europe (Kosovac et al. 2016(Kosovac et al. , 2018(Kosovac et al. , 2019. Several recent molecular studies indicate that distinct, genetically discernible H. obsoletus populations 1 3 are specifically associated with the different host plant species (Johannesen et al. 2008;Kosovac et al. 2018;Jović et al. 2019 for review). In addition to H. obsoletus, several other Cixiidae planthoppers, as well as leafhoppers from the families Cicadellidae (Subfamilies Deltocephalinae, Agalliinae, Aphrodinae), Aphrophoridae and Delphacidae, are known or are likely to transmit the pathogen. These species might play some role in pathogen transmission or might be of local importance (Jović et al. 2019 for review;Quaglino et al. 2019;Riedle-Bauer et al. 2006Battle et al. 2008). Langer and Maixner (2004) observed that the 'Ca. P. solani' types specifically associated either with nettles or with bindweeds can be discriminated by analysis of the phytoplasma elongation factor Tu (tuf) gene. In the following years, a number of further 'Ca. P. solani' genes, such as secY, stamp and vmp1, have been characterised and found suitable for epidemiological studies. Nowadays, a fine tuned allocation of the phytoplasma types involved in a given disease outbreak to the respective disease cycle is possible with these marker genes (Cimerman et al. 2009;Fabre et al. 2011a, b;Johannesen et al. 2012;Aryan et al. 2014). Past and recent scientific reports suggest a wide distribution of 'Ca P. solani' in Solanaceous crops in central and eastern Europe. E.g. stolbur outbreaks from 1921Europe. E.g. stolbur outbreaks from to 1924Europe. E.g. stolbur outbreaks from and from 1932Europe. E.g. stolbur outbreaks from to 1934 in Slovakia seem likely and until the 1950 stolbur regularly occurred in former Czechoslovakia (Valenta 1953;Navrátil et al. 2009). At the same time, in Austria, Wenzl (1955) observed the outbreak of a "hairy sprout degeneration" of potatoes with incidences of up to 33% symptomatic potato plants in the field and up to 17.7% gummy tubers. More recently, namely around 2003, a severe stolbur outbreak affected celery production in North Eastern Italy (Trieste region, Carraro et al. 2008) and caused losses of up to 100%. In south Moravia, in 2006-2008 the pathogen severely damaged tomato production (Navrátil et al. 2009). A survey on widespread phytoplasma infections of potato crops in Romania and southern Russia in 2011 identified the stolbur phytoplasma as the major cause (Ember et al. 2011). From around 2016 onwards, the pathogen accounted for extreme losses in potatoes, vegetable crops and grapevines in various regions in East Austria. In potato, field symptoms included reddening and upward rolling of the top leaves and aerial tubers as well as wilting and withering of affected plants. Tubers were deformed, showed a soft and gummy consistency and developed hairy sprouts. Tomatoes produced discoloured, deformed leaves and greened flowers. Infected celery and Chinese cabbage plants turned yellow or red and declined (Brader and Riedle-Bauer, unpublished). Rates of symptomatic potato plants in the field reached up to 80%, and during stocking up to 40% of the harvested tubers proved unsaleable (Kamptner, personal communication). Phytoplasma infections caused up to 100% losses in celery fields as well as in tomato seed production. Molecular analyses of the tuf and stamp gene linked the involved phytoplasma types to bindweed related disease cycles. Extensive field observations revealed widespread bindweed associated populations of H. obsoletus in the affected regions (Brader and Riedle-Bauer unpublished). The presence of relevant insect populations on bindweed contrasted sharply to previous observations in Austria. In extensive vector monitorings related to BN outbreaks in grapevine from around 2003 onwards, bindweed associated H. obsoletus populations had been detected to a very limited extent only (Riedle-Bauer et al. 2006Aryan et al. 2014;Johannesen and Riedle-Bauer 2014). In general, management of phytoplasma diseases is challenging and in most cases conventional control strategies are not directly directed against the pathogen. Depending on the given situation, they focus on eradication of the infected reservoir plants and insecticide treatments against the vectors. In consequence, phytoplasma management could possibly have relevant economic and environmental implications . In case of bindweeds, representing one of the principal sources for 'Ca P. solani', an eradication or extensive suppression is hardly feasible. This weed species can invade and persist in highly diverse habitats, so it is widespread not only in crop fields but also in their surroundings. Away from cultivated areas, in most cases, no weed control takes place. Within the fields and along field margins, vegetative reproduction through adventitious buds on roots, root parts and long-lived seeds complicate successful bindweed control. Any kind of mechanical disturbance, e.g. in the course of mechanical weeding or mowing works can aggravate the problem due to spreading of vegetative propagules (Davis et al. 2018). The effect of conventional insecticides on phytoplasma spread is in general not well studied and a subject of controversial debate. Principally, in case of a vector transmittable disease, where infected vectors migrate into the crop sites, two aspects must be considered. One goal of the treatments is to keep vector populations inside the fields as low as possible and consequently reduce pathogen spread. Restriction of vector populations inside a crop, however, does not protect the first crop plant(s) reached by a vector insect migrating into the field. In this respect, insecticides actually preventing pathogen inoculation by disrupting the insect feeding behaviour in a shorter time than the minimum inoculation access period (IAP) are required. In earlier examinations, insecticide treatments reduced the spread of the pathogen within the crop but even frequent applications did not prevent the appearance of the disease (Weintraub 2007). Studies on psyllids as vectors of European Stone fruit yellows phytoplasma ('Candidatus Phytoplasma prunorum') suggested that some insecticide agents, particularly pyrethroids, act quickly and could therefore have a direct effect on phytoplasma transmission (Paleskić et al. 2017). This presumption is strengthened by the fact that insecticide treatments have been used for years to control the spread of apple proliferation ('Candidatus Phytoplasma mali') in Northern Italy (Baldessari et al. 2010;Österreicher and Unterthurner 2015). Šafářová et al. (2016) reported lower presence of psyllid species known as vectors of 'Ca. P. mali' and lower numbers of infected apple trees in orchards under integrated management as compared to organic management. The recent Stolbur outbreak in Austria raised the question, to which extent pathogen spread could be restricted by insecticide or insect repellent treatments. This question was of interest on one hand for producers of consumer potatoes and vegetables but, particularly, for seed potato growers. Therefore, we studied the effects of various insecticides and insect deterrents and analysed their suitability for a future sustainable Stolbur management. Laboratory experiments focused on the effects of the test agents on mortality of H. obsoletus and on phytoplasma transmission. The most promising active compounds were included in a series of field experiments implemented over three years in potato. Treatment effects on vector density, disease symptoms on plants and tubers as well as on phytoplasma infections of tubers were investigated. Evaluation of plant protection products The effects of a range of insecticides and insect deterrents on insect mortality and pathogen transmission were evaluated in pot cage experiments including Catharanthus roseus as test plants and field collected H. obsoletus. The experiments took place in June 2018-2020. C. roseus (cv. 'Sorbas Reinweiß', Austrosaat, Vienna, Austria), were grown from cuttings under laboratory conditions (22-24 °C, L:D 16:8) until six to eight leaf stage. Table 1 illustrates the tested agents, the manufacturers and the used concentrations. The concentrations of the test compounds were derived from the product registrations, namely the application rates, and the spray volume per hectare. The application of test compounds and the efficacy tests took place under outdoor conditions but protected from rain. Test plants received spraying until run off by aid of a 1.25 l hand sprayer (Birchmeier, Stetten, CH). In order to ensure even wetting, spraying started at the uppermost leaves working towards the base. The spraying residues were allowed to air dry for two to four hours. In case of acetamiprid, additional tests with fourday-old spraying residues in case of lambda cyhalotrin and flupyradifuron tests with five-day-old spraying residues were also included in the study ("aged residues", PCR analysis DNA extraction from C. roseus and H. obsoletus was carried out by a CTAB procedure as described previously (Maixner et al. 1995). Each plant or insect was analysed individually. In case of leaf samples, each sample consisted of midrib and petiole tissue of three leaves. DNA extraction from potato tubers was carried out by a Nucleon Phytopure genomic DNA extraction kits according to the protocol of the manufacturer (GE Healthcare, UK). For each tuber, four 10 × 5 × 5 mm sized pieces including skin of washed potatoes were cut and ground in liquid nitrogen with pestle and mortar. 0.1 g of the obtained powder was used for extraction in 1.5 mL tubes with half amounts as in the protocol and with an additional RNAse step (1 µl of 5 mg/mL DNAse and Proteinase free RNAse from Thermo Fisher Scientific (UK) for 30 min at 37 °C) after cell lysis. Plant samples were analysed for 'Ca. P. solani' by a nested PCR procedure including the primers P1/P7 (Deng and Hiruki 1991;Schneider et al. 1995) and STOLF/STOLR (Maixner et al. 1995) or using the nested Stamp protocol using (Stamp-F, Stamp-R0, Stamp-F1 and Stamp-R1; Fabre et al. 2011b). Phytoplasma presence in H. obsoletus was determined by real time PCR following the method described by Christensen et al. (2004). Treatment effects in the field In principle, field experiments were carried out in Rottersdorf, Maissau, Unterhautzenthal and Oberfellabrunn in Lower Austria in commercial potato farms in three consecutive years (2019-2021). However, in Unterhautzenthal 2019 and 2020, there were no signs of the disease and no H. obsoletus, in Oberfellabrunn in 2019, severe drought stress impeded the evaluation of the experiment. In consequence, the current paper focuses on our experiments in Maissau 2019-2021 and in Rottersdorf 2019 and 2020. The linear distance of the two test sites is about 34 km. The distance at on site but between the different years are less than 1 km. Details on the experimental fields and the spraying conditions are outlined in Table 2. Due to the suspected rapid movement of H. obsoletus in the fields, we considered large treatment plots as necessary. Moreover, the farms were equipped with standard spraying devices only. Therefore, we opted for experiments in a strip design including two repetitions per treatment, location and year, except for Maissau 2021 where due to the shape of the field 2 × 2 blocks were laid out (Online resources 2-6). The length of the fields varied from 275 to 330 m, the plot width was adjusted to the spray width of the respective field sprayer (27 m in Maissau, 28 m in Rottersdorf,). The included insecticides were selected according to the results of the laboratory experiments and a given registration of the respective compound for use in potato. In addition, treatments with diatomaceous earth and a paraffin oil were included in the experiments. The start of the treatments was determined according to the H. obsoletus monitoring carried out by aid of vacuum sampling at roadsides and field margins at several locations in Southern Weinviertel (Naglern1, Naglern 2, Weinsteig, Lower Austria, outlined in Online resource 1). The first application was scheduled one week after the first specimen were detected. Experiments 2019 In Maissau (Online resource 2) treatments started on June 13, the included potato variety was 'Eurostarch'. The farmer decided to apply six treatments against H. obsoletus, the last application was on July 16 (Table 3). For determination of field symptoms in each plot the first 50 plants in rows six and seven on the left hand side of the driving lane were visually classified into the categories healthy (0) and symptomatic (1). Visual inspections were carried out on July, 17, July 25 and August 19. At harvest, in each case circa100 tubers per plot were collected both from the field edge and from the centre of the field. On October 15, 100 tubers per experimental plot both from the field edge and the centre of the field were analyzed visually and by squeezing with fingers and classified into the categories healthy (0) and symptomatic (1). Every tenth tuber was collected for PCR analysis. In Rottersdorf, the test field (Online resource 3) was planted with the variety 'Tosca'. The first application out of four was on June 13 and the last one on July 5 (Table 4). Visual inspections on July 16, July 31 and August 21 were carried out as described for Maissau, except that only 40-42 plants per row (rows six and seven on the left-hand side of the driving lane) were inspected. On October 1, the harvest of each treatment (both repetitions together) was transferred into four distinct bulk bins. Tuber analysis (100 tubers per bin) and sampling for PCR analysis were carried out on October 18 as described above. At both locations, from the beginning of the treatments onwards, the presence of H. obsoletus in the field was monitored by yellow sticky traps (25 × 40 cm, Horiver, Biohelp, Vienna, A) mounted in the rows transversely to the direction of the rows. The traps were installed at a distance of 5 m and 20 m from the field edge. Per plot and distance from the field edge, two traps were installed and changed in five to seven day intervals (in total 24 traps per time interval; Online resources 2 and 3). Experiments 2020 In Maissau the experimental field (Online resource 4) was planted with the variety 'Nafida'. Treatments started on June 23, until July 21, five treatments were applied (Table 3). Visual inspections on July 31 and August 14 revealed considerably lower infection rates than in 2019, therefore the number of inspected plants in each plot was increased to 240. At harvest on October 1, 200 tubers per plot were randomly collected for analysis as described for 2019. A separate harvest of tubers from the field margin and the middle of the field was not maintained as no difference had been recorded in 2019. In Rottersdorf (Online resource 5), the experimental field was planted with the variety 'Belmonda'. The treatments started on June 19, until July 15 in total five applications were carried out (Table 4). Visual inspections in the field took place as described on July 30 and on August 14. Tubers were harvested on October 6 and investigated on Oktober 21 as described for 2019. Every tenth tuber was collected for PCR-analysis. Numbers of yellow sticky traps per sampling interval were reduced to five in Maissau and four in Rottersdorf and mounted as outlined in Online resources 4 and 5. Experiments 2021 The experiments took place in Maissau only and included the variety 'Eurostarch'. The first treatment was applied on June 21. In total four treatments directed against stolbur vectors were carried out, the last one on July 29 (Table 3). Per plot, one yellow sticky trap was mounted in a distance of 20 m from the field edge and changed weekly (Online resource 6). Visual analysis of the plants in the field was carried out as described on July 31 and August 19. The experimental plots were harvested on October 11. Due to the low incidence of field symptoms, the number of analysed tubers was increased to 400 per experimental plot. The analysis of the tubers took place as described for the previous years on December 29. At the same time, every tenth tuber was collected for subsequent PCR analysis. Statistical analyses All statistical analyses were performed individually for each test year and location by aid of the statistics program SPSS 26.0 (SPSS, IBM, Vienna, A). The effect of the treatments on insect survival in the laboratory was studied by aid of generalized linear models including insect survival (dead = 1, alive = 0) as the outcome, the model type binary logit, and the explanatory variable tested compound. We considered the factor hours post exposure as crucial for the subsequent choice of compounds for the field experiments. Therefore, individual models were run for all selected exposure durations instead of including duration of exposure as a factor in a joint model (Table 5). To identify treatment effects on the presence of disease symptoms on potato plants in the field and on potato tubers, we calculated generalized linear models for the response variable presence of disease symptoms (symptomatic = 1, healthy = 0). The model type binary logit, and the categorical explanatory variables (1) treatment, (2) date of scoring (in case of plants in the field only, not for tubers), (3) position within the layout of the experiment (position 1, position 2, except for analysis of tubers in Rottersdorf, online resources 2-6) and (4) area of the field from which the tubers had been harvested (from middle or edge of the field, for Maissau 2019 only) were included (Tables 6 and 7). In 2019, insect counts on yellow sticky traps were analysed by aid of generalized Poisson models comprising the response variable number of H. obsoletus per trap and the categorical explanatory variables (1) treatment, (2) sampling (3) distance from the edge of the field (5 m versus 20 m). We tested for the main effects as offered by the programme. Where appropriate, least significant difference (LSD) tests were calculated to analyse differences among the effects of the studied explanatory variables. For the LSD tests a 0.05 level of significance was applied. Evaluation of plant protection agents under laboratory conditions Already one hour after insect release a clear and statistically significant treatment effect on insect survival was recorded (Wald χ 2 = 360.65, df = 21; p = 0.00). All tested compounds significantly reduced insect survival as compared to the control, except cyantraniliprole 0.04%, cyantraniliprole 0.25%, spirotetramate and diatomaceous earth. Lambda-cyhalothrin and deltamethrin were significantly more effective than all other tested compounds. Flupyradifuron, acetamiprid 0.08% and acetamiprid 0.033% also proved highly effective. After three hours of exposure, in addition to lambda-cyhalothrin and deltamethrin an insect mortality at or near to 100% was observed for the compounds acetamiprid (for both concentrations), sulfoxaflor and flupyradifuron. Five-day-old residues of lambda-cyhalothrin and four-day-old residues of acetamiprid in both concentrations still significantly enhanced mortality as compared to the untreated control (Fig. 1, Table 5). All C. roseus plants included in the laboratory experiments were analysed for phytoplasma infection by PCR with primers P1/P7 and StolR/StolF after the end of the experiment. As illustrated in Fig. 2, all test plants treated with lambda-cyhalothrin, deltamethrin, acetamiprid, chlorpyriphos and esfenvalerate remained healthy, whereas 53% of the control plants became infected. The highest infection rates were recorded for quassia (57%), spinosad (71%) and spirotetramate (86%). Out of the test plants treated with particle films (kaolin, diatomaceous earth) 37% were found infected. All PCR positive plants developed classical Maissau 2019 On July 17, 10.0% and 14.0% of the plants in the insecticide treated plots, 12.0% and 13.0% of the plants in the mineral oil treated plots and 14.0% and 18.0% of the plants in the control plots were symptomatic. On July 25, the infection rates rose constantly to reach 15.0% to 20.0% of symptomatic plants in the insecticide treated plots, 21.0% to 34.0% in the mineral oil treated plots and 36.0% to 42.0% in the control plots. On August 19, the control plots showed infection rates between 63 and 65%, in the insecticide plots 32% and 37%, in the mineral oil plots 73% and 50% of the plants were visually diseased (Fig. 3). Generalized linear models indicated a significant effect of the factors treatment and scoring date (treatment: Wald χ 2 = 51.98.08. df = 2, p = 0.000; scoring date: Wald χ 2 = 211.85, df = 1, p = 0.000) but no effect of the factor position within the experimental plot. Pairwise comparisons of the treatments revealed a significantly lower rate of symptomatic plants in the insecticide treatments, whereas mineral oil had no effect as compared to the untreated control. Pairwise comparisons of scoring dates showed a significant increase of diseased plants in the course of the experiment ( Table 6). Rates of symptomatic tubers in the insecticide treatments (20.0% and 18.6% at the field edge, 20.6% and 14.9% in the middle of the field, respectively) were somewhat lower than in the control plots (28.4% and 26.0% at the field edge, 22.5% and 23.5% in the middle of the field) and the mineral oil plots (31.1% and 23.1% at the field edge, 27.5% in the middle of the field). Statistical analysis indicated a significant influence of the factor treatment (Wald χ 2 = 7.81, df = 2, p = 0.02) but no effect of the factors position within the experimental field and area in the field (edge, middle), from which the tubers had been harvested (Table 6, Online resource 7). PCR analysis of the tubers by nested Stamp PCR revealed an average infection rate of 30% for the control treatment, 22.5% for the insecticide treatment and 50% for the diatomaceous earth treatment. Analysis of planthoppers in the field by yellow sticky traps showed a relevant insect density from the sampling period June24-June 29 onwards. Statistical analysis indicated a significant effect of the sampling date (Wald χ 2 = 55.00, df = 4, p = 0.000) on insect numbers on the traps but no effect of the treatment and the distance to the field edge at which the yellow sticky traps had been mounted (Table 6, Fig. 4). Rottersdorf 2019 At the visual inspection on July 16, rates of symptomatic plants in the insecticide plots were 9.4% and 10.6%, respectively, in the diatomaceous earth plots 17.6% and 16.5%, in the control plots 9.6% and 32.5%. On July 31, infection rates in the insecticide plots had risen to 28.2% and 18.6%, in the diatomaceous earth plots to 50.0% and 25.6%, and in the control plots to 51.2% and 52.3%. At the last visual inspection (August 21), infection rates of all plots exceeded 75%, except the plot Ins1 with a rate of 54.7% of symptomatic plants. The generalized linear model indicated a significant effect of the factors treatment, scoring date and position within the experimental design on rates of symptomatic plants (Treatment: Wald χ 2 = 48.56. df = 2, p = 0.000; scoring date: Wald χ 2 = 369.53, df = 2, p = 0.000; position: Wald χ 2 = 7.75 df = 1, p = 0.000; Fig. 5, Table 7). The treatments had no effect on the health status of the tubers (Online resource 8, Table 7). Rates of symptomatic tubers of 17.0 and 26.0% were recorded for the control plots, 14.0 and 24.0% for the diatomaceous earth plots, 13.0 and 21.0% for the insecticide plots. PCR analysis confirmed the low treatment effect on the health status of the tubers. Rates of positive tubers with nested Stamp PCR were 57.7%, 82.5% and 65.0% for the insecticide treatment, the diatomaceous earth treatment and the experimental control, respectively. Maissau 2020 Symptoms ascribable to 'Ca. P. solani' increased from single plants on July 31 to a disease incidence of 0.8% and 4.2% in the insecticide plots and 8.3% and 4.6% in the control plots two weeks later. The generalized linear model proved significant impacts of the factors treatment and scoring date on rates of symptomatic plants in the field. (Treatment: Wald χ 2 = 9.74, df = 1, p = 0.002; scoring date: Wald χ 2 = 18.61, df = 1, p = 0.000), but no effect Table 6). Visual assessment of tubers after 4 month of storage showed no significant difference between treatments, rates of 3.0% and 4.0% of soft tubers in insecticide versus 3% and 4% in control treatments were recorded (Online resource 10). PCR using nested Stamp primers and yielded 15% positive samples in the insecticide and 17.5% positives in the control samples. Sporadically, H. obsoletus individuals were identified on the yellow sticky traps, the first catches originated from the sampling interval June 24 to July 1 (Online resource 11). Rottersdorf 2020 The field observations on July 30 revealed few symptomatic plants only. On August 14, rates of symptomatic plants in the insecticide treatment of 8.3% and 25.8%, in the control plots of 14.2% and 40.0% were determined (Online resource 12). Statistical analysis identified the treatment, the scoring date and the position of the plot within the experimental setup as factors significantly affecting the health status of the plants (Treatment: Wald χ 2 = 5.73 df = 1, p = 0.018; scoring date: Wald χ 2 = 54.69, df = 1, p = 0.000; position: Wald χ 2 = 26.76. df = 1, p = 0.000; Table 7). Rates of symptomatic tubers were lower in the insecticide treatment, ranging from 3 to 5% in the insecticide treated plots and from 4 to 10% in the control plots (Fig. 7). The generalized linear model indicated a significant influence of the treatment on the rate of symptomatic tubers (Wald χ 2 = 6.49, df = 1, p = 0.011; Table 7). PCR using a nested approach targeting the stamp gene confirmed these trends with 17.5% Stamp positive results in the control and 10% in the insecticide. Few H. obsoletus were captured on the yellow sticky traps, at maximum two individuals per trap were recorded from the sampling interval July 1-July 9 onwards (Online resource 13). Maissau 2021 No visually discernible symptoms of stolbur were recorded in any of the examined plants during the visual inspections on July 30 and on August 19. In the whole field only five symptomatic plants were observed, all outside the examination area. Inspection of the tubers led to rates of symptomatic tubers of 2.0 and 4.0% for the insecticide treatment and 3.5 and 7.0% for the control, respectively (Online resource 14). The generalized linear model indicated a significant effect the factors treatment and position within the experimental design on tuber symptoms (Treatment Wald χ 2 = 5.02. df = 1, p = 0.025; position: Wald χ 2 = 7.39. df = 1, p = 0.007; Table 6). With nested Stamp PCR only a single positive tuber could be detected in control 2. Not a single H. obsoletus individual was present on any of the yellow sticky traps. Discussion To the best of our knowledge, this is the first report on the efficacy of insecticides and repellents in preventing the spread of 'Ca. P. solani' including both laboratory and field experiments. The first objective of phytoplasma management by insecticides is to keep the vector population inside a field as low as possible and in that way to reduce pathogen transmission. In our laboratory study, all of the tested insecticides except spirotetramate significantly reduced insect survival as compared to the control. They might therefore more or less efficiently reduce insect survival and in consequence phytoplasma spread. Out of the tested compounds, especially the pyrethroids and the neonicotinoid acetamiprid caused high vector mortality, even when the spraying residues were already 4-5 days old. A mere reduction of the insect populations inside the crop, however, is likely not enough for a sufficient pathogen restriction because the insect vectors constantly move into the potato fields from outside. In consequence, compounds protecting plants from becoming infected by disrupting the insect feeding behaviour in a shorter time than the minimum IAP are desirable. The IAP for H. obsoletus transmitting 'Ca. P. solani' is not well studied, in case of transmission to grapevine a minimum IAP of 3-6 h has been estimated (Bressan et al. 2007). A higher susceptibility of potato or vegetable species to the pathogen and, in consequence, a shorter IAP cannot be ruled out. It seems likely that the quicker the insecticide prevents vector feeding the greater the chance to reduce pathogen transmission. In our experiments, the compounds lambda cyhalothrin and deltamethrin led to an insect knock-down followed by mortality of Fig. 7 Rottersdorf 2020: Assessment of tubers for symptoms of 'Ca. P. solani'. Applied compounds: Ins: Lambda-Cyhalothrin + Acetamiprid, Cypermethrin, Esfenvalerate 100% or close to 100% already after one hour of exposure. Within three hours, also acetamiprid in both concentrations, flupyradifuron and sulfoxaflor largely killed the planthoppers. Aged residues of acetamiprid, lamda-cyhalothrin and flupyradifuron still acted quickly. Data on insect mortality largely coincided with the phytoplasma transmission in these experiments. In tests including the above mentioned compounds and esfenvalerate all or almost all C.roseus remained healthy. In contrast, plants treated with less effective and/or slower acting compounds such as cyantraniliprole and spinosad became infected roughly to the same extent as the experimental control. Interestingly, the highest rate of infected test plants was recorded for spirotetramate, although, likely to the low number of tests, no statistically significant difference to the water control was observed (data on statistical analysis not shown). High rates of diseased plants also occurred in experiments with the putative feeding deterrent quassia. These observations raise the question if treatments inducing feeding deterrence via changes in taste and smell but without affecting insect survival lead to an adverse effect. Such treatments could prompt the insects to an enhanced search behaviour and finally increase phytoplasma transmission. Comparable effects of insecticides on phytoplasma transmission were reported for laboratory studies on the leafhopper Macrosteles quadripunctulatus vectoring chrysanthemum yellows phytoplasma (CYP). In transmission experiments including imidacloprid and organophosphates potted Chrysanthemum carinatum was significantly protected from infection (Saracco et al. 2008). All in all our laboratory study, together with literature reports, proves that insecticide applications might influence phytoplasma transmission, provided that the active agents (and their spray deposits) are (still) effective and fast acting. Due to the difficulties in collecting a sufficient number of insects in the field, the laboratory study comprised one test procedure only. Admittedly, this approach had deficiencies in considering the different modes of action of the test agents. The used scheme, namely the insect release on the treated test plants two to four hours after application of the test products, was likely most suitable for identifying the effects of compounds with a rapid contact action. This was e.g. the case for the pyrethroids, which, upon contact, result in a rapid knock down and death of the target insects (Lybrand et al. 2020). The pyrethroid lambda-cyhalotrin was applied as encapsulated formulation Karate Zeon with enough stability to show knock down and high mortality still after 5 days. Acetamiprid, flupyradifurone and sulfloxaflor are compounds with contact effect and systemic translocation in the plant (Yamada et al. 1999;Nauen et al. 2015;Bacci et al. 2018). Accordingly, all these agents were effective in the setup of the current study. The procedure also revealed an effect of flonicamid on H. obsoletus survival and pathogen transmission. This compound is particularly active against hemipterous pests and inhibits the insects' feeding behaviour within a short time (Morita et al. 2007). Likely, the test scheme was also adequate for evaluating the effects of the included organic insecticides. At least some direct contact effect through the body surface has been reported for spinosad as well as for Quassia amara extract (Grdiša and Gršić 2013;Bacci et al. 2016), albeit these agents were not very effective against H. obsoletus in the actual study. Azadirachtin is considered as pronounced antifeedant provoking behavioural avoidance in many insect species (Kilani-Morakchi et al. 2021), a fact that might have resulted in the relatively low rate of infected C. roseus in the current experiments. In contrast, out test set up was less suitable to identify effects of active compounds with predominantly systemic activity. This was particularly the case for spirotetramat, which becomes effective after absorption and translocation in the plant, a process requiring more than several hours to days. Moreover, it particularly affects juvenile stages and adult fecundity (Brück et al. 2009). However, in the current study, the insects were released hours after application of the test compounds, the complete observation period lasted five days only, and due to the subterraneous lifestyle of the H. obsoletus instars, only adults were included. Likely, these factors contributed to the low effect of this agent in the current tests. The compound chlorantraniliprole mainly acts by ingestion and is to some extent transported in the plant (Lahm et al. 2009). In the current study, the effect on H. obsoletus and pathogen transmission was relatively low, likely, also in this case, due to a lack of time for systemic movement of the compound within the test plants. Some reducing effect on the rate of infected C. roseus as compared to the control was also observed for the particle film forming compounds kaolin and diatomaceous earth. For a long time it has been known that kaolin particle film coating results in a hostile environment for insects and in consequence reduce insect movement, feeding and egg-laying (Glenn and Puterka 2005). Diatomaceous earth is based on fossilised unicellular algae and contains amorphous silicon dioxide. The porous particles have an abrasive effect and a high capacity for absorption of lipids. Contact damages the insects 'wax coating' and leads to desiccation. In addition, the compound can interfere with the insects' settling, probing and feeding behaviour (Korunic 1998). Simultaneously with and subsequently to the laboratory experiments, we made efforts to transfer the achieved laboratory results to field conditions. In 2019, the insecticide applications comprised the compounds lambda -cyhalotrin and esfenvalerate. We included pyrethroids only, despite the fact that in practice active agents from different IRAC (Insecticide Resistance Action Committee) groups ought to be selected. However, at the start of the field experiments, the laboratory results outlined above were only partly available and we regarded pyrethroids due to their knock down effect as particularly promising. Other compounds, such as chlorpyriphos, were ruled out because of their expiring registration. Mineral oil and diatomaceous earth, were included in the study because of their lower environmental impact. Both in Maissau and in Rottersdorf in 2019, at the beginning of the experiments during the visual assessments in July, the insecticide effect seemed promising. The treated plots looked healthier and the insecticide treatment reduced the rates of infected plants roughly by half. Until the middle of August, however, in Rottersdorf the vast majority of plants in all plots had collapsed. In Maissau, at the time, the rate of infected plants was almost halved in comparison to the experimental control, but the rate of diseased plants still reached 35%. In Rottersdorf, the high share of symptomatic plants in the insecticide treatment was also reflected in the status of the tubers. No or only marginal differences to the experimental control were evident, neither in respect to gummy tubers nor in respect to PCR results. In Maissau, statistical analysis proved a significant insecticide effect on the health status of the tubers, but from a practical point of view, the rate of gummy tubers (on average 17%) was too high also in this location. Rates of PCR-positive tubers were even slightly higher (22.5%). The particle films formed by diatomaceous earth had no or almost no effect from the first visual scoring onwards. Similar results were obtained for the mineral oil treatments. In 2020, based on the experiments in 2019 and on the registration situation, treatments included the insecticide compounds lambda-cyhalothrin, acetamiprid and cypermethrin. In this year, however, all over Lower Austria disease incidence and severity sharply declined in comparison to the years before (data not shown). Also in our two experiments, insect populations were smaller and the flight period started later than in 2019. In Maissau, disease incidence in the field and rates of symptomatic tubers were low and marginally influenced by the treatments. In Rottersdorf, despite or maybe because of the lower infection pressure a treatment effect was clearly visible. Both, the incidence of field symptoms and the share of gummy tubers were roughly halved. In Maissau in 2021, the infection pressure was even lower than in 2020. Not a single vector individual stuck on the yellow traps and no symptomatic plants occurred in the rows previously defined for visual scoring. Sporadically, single infected plants were observed in other parts of the field only. Nevertheless, a certain treatment effect on the rates of gummy tubers was visible and confirmed by PCR analysis. Over all experiments, a good agreement between captures of H. obsoletus and disease incidences indicated that this insect species was the most relevant pathogen vector. In 2019, constant insect captures on the yellow sticky traps resulted in high disease rates, whereas in 2021, an obviously low insect population (without captures on yellow sticky traps) was related to low numbers of infected plants. In single cases, other potential vectors, namely Reptalus panzeri, R. quinquecostatus and Pentastiridius leporinus were also present on the yellow sticky traps (data not shown), however, due to their rare occurrence, their impact on disease spread in the test fields was likely negligible. Overall, the results of the current study demonstrate that the insecticide treatments influenced phytoplasma transmission also in field conditions. Maintaining plant health in the course of the vegetation period, however, proved challenging. Putatively, hot temperatures and strong solar radiation in July reduced the insecticide efficacy in a shorter time than the interval of eight days between the treatments. In consequence, the aged spray residues provided no full protection against the pathogen transmission. During the long insect flight period with a constant infection risk the share of diseased plants grew steadily. Moreover, pyrethroids, the predominantly used insecticides in the current study have a contact effect only. In case of spraying densely growing potatoes, a complete coverage with the spray liquid is difficult to achieve, particularly on the stems and the midribs on the underside of the leaf. The vector insects, however, prefer these parts of the plant for feeding. In consequence, an incomplete coverage of the plants with the insecticide might have contributed to the infection rates. Divergent results of the experiments in Maissau and Rottersdorf in 2019 are likely attributable to the two different potato varieties. In Rottersdorf, the planted variety was 'Tosca', which proved as extremely susceptible in the course of this stolbur outbreak. Eurostarch, the variety in Maissau, turned out as moderately susceptible (Brader, unpublished). Comparing the starting times of the treatments with the insect captures in the field in 2019, it emerged, that in all experiments the insecticide treatment had started around 7-10 days earlier than relevant insect populations occurred in the field. The reason for the time lag between the start of the flight period at the roadsides analysed for determination of insect development and insect presence in the potato fields remained unclear. Likely, in that year, the first insecticide applications could have been omitted without a relevant reduction in efficacy. Treatments with mineral oil and diatomaceous earth were largely ineffective. Already in the cage experiments, despite a full coverage of the plants, the effect of the particle films was relatively low. Likely, in the field the effect of diatomaceous earth was additionally impaired by the fact that a dense coverage of the entire plants with the spray suspension was not achieved using the spraying equipment available at the two commercial farms. Although mineral oils are known to obstruct the insects' tracheal structures and provoke structural changes in the cuticle and residual oil films discourage insects from feeding (Buteler and Stadler 2011), these effects were not sufficient to prevent the transmission of 'Ca. P. solani' in the field. Admittedly, with respect to statistical processing of the data, in all our experiments higher numbers of repetitions would have been desirable. However, as already outlined, the trials were carried out in commercial farms, where higher numbers of plots were not feasible due to temporal and technical restrictions. From a practical point of view, our experiments showed that treatments with the insecticides alone are certainly not sufficient to manage potato stolbur in years with high disease pressure and in case of susceptible varieties. A little more encouraging results might be expected in situations with lower disease pressure and less susceptible varieties. In the last two or three years in Austria, we witnessed a rapid decline of the bindweed associated H. obsoletus populations and correspondingly, an endemic disease phase. Therefore, we were unable to clarify the question whether application of insecticides could and should be one component of a future management programme together with other measures, particularly the planting of less susceptible varieties. The results of this study and further field observations during the epidemic disease phase in Austria indicate a great variability of potato varieties in respect to their stolbur tolerance (data not shown). In any case, the rather dense spraying intervals and the considerable amount of plant protection agents used in the current study likely entailed a broad range of side effects on the environment, while a substantial or even complete disease suppression was not achieved. Further studies must clarify if insecticide treatment, for example a reduced number of applications right before or at the flying peak of H. obsoletus, could be a justifiable measure leading to a proportionate disease reduction.	2023-07-12T05:27:25.146Z	2023-07-04T00:00:00.000	{ "year": 2023, "sha1": "827a0a37bee0def57aed9aa20deb335130d3cbd5", "oa_license": "CCBY", "oa_url": "https://link.springer.com/content/pdf/10.1007/s41348-023-00768-y.pdf", "oa_status": "HYBRID", "pdf_src": "Springer", "pdf_hash": "d4325f1b3a6728f2bd1dd7df2b07c3eb2e528148", "s2fieldsofstudy": [ "Environmental Science", "Agricultural and Food Sciences" ], "extfieldsofstudy": [ "Medicine" ] }
257928422	pes2o/s2orc	v3-fos-license	Interrelationship between muscle fitness in childhood and bone mineral density in adulthood: mediation analysis of muscle fitness in adulthood Background This study was aimed to examine the relationship between muscular fitness indicators in childhood and areal bone mineral density (aBMD) in adulthood and to verify whether the relationship is mediated by performance on muscular fitness indicators in adulthood. Methods A sample of 138 healthy adults (69 males; 22.3 years) were followed after a previous assessment at the age of 7–10 years. Stature, body mass and muscular fitness indicators (handgrip strength, standing long jump and sit-ups tests) were assessed in childhood and adulthood. Additionally, total body, upper limbs, lower limbs, right femoral neck and lumbar spine aBMD was assessed in adulthood using dual X-ray absorptiometry. Analysis included descriptive statistics; t-test or Mann-Whitney U-test for comparison between males and females, multiple linear regression for the prediction aBMD from muscular fitness indicators in childhood, mediation analysis of the respective muscular fitness indicators in adulthood and the relationship between muscular fitness indicators in childhood and aBMD. Results Males were stronger compared to females regarding muscular fitness indicators in childhood and adulthood, and presented higher mean values for aBMD in adulthood, except for lumbar spine (p < 0.05). Regression analysis revealed that some muscular fitness indicators in childhood showed significant positive relationship with bone health indicators in adulthood, such as: handgrip strength and total body aBMD (β = 0.005; R2 = 0.35; p = 0.040) and upper limbs aBMD (β = 0.005; R2 = 0.55; p = 0.019); and sit-ups test was a significant predictors of lumbar spine BMD (β = 0.003; R2 = 0.06; p = 0.039). Mediation analysis pointed out the following: adulthood handgrip strength mediated relationships between childhood handgrip strength and total aBMD (indirect effect (IE) = 0.0025; 95%CI = 0.0005–0.0048), and upper limbs aBMD (IE = 0.0040; 95%CI = 0.0017–0.0069). Conclusions Muscular fitness indicators in childhood showed significant relationship with bone health indicators in adulthood and the sit-ups test in childhood had direct effect on lumbar spine aBMD in adulthood. Adulthood handgrip strength mediated the relationship between childhood handgrip strength and total body and upper limb aBMD, pointing out that muscular fitness in childhood may be a aBMD determinant in adulthood, especially when higher muscle fitness performance is maintained in adulthood. Introduction Peak bone mass (PBM) is reached in early adulthood and is interpreted as the achievement of the highest expression of bone tissue [1]. It is considered a relevant indicator in prediction of osteoporosis and fractures in the course of aging [2,3]. The literature suggests that an increment of 10% in PBM tends to delay the onset of osteoporosis by 13 years [4]. In addition, epidemiological evidences [5] concluded that this increase of 10% in PBM during the first decades of life is associated to a decrement of 50% in the risk of fracture among elderly women. By inference, optimization of the bone mass gain during growth seems central to mitigate the consequences of the physiological loss with aging. Several factors contribute to PBM such as genetics, bone status during childhood, endocrine regulators, interaction of bone tissue with other tissues, lifestyle factors, chronic diseases during childhood, and others [6][7][8]. Meantime, the trajectory of bone mass gains is similar to that of linear growth and believed to be sensitive to modifiable or non-modifiable determinants during this period [9]. There is evidence of the effect of physical activity [7] and improved muscular fitness on bone mass [10]. Bones are exposed to muscular action, stronger muscles exert more tension on bones and consequent mechanical adaptations. The mechanical adaptations of bones according to what type of exercise they are submitted to are explained by mechanostatic [11], mechanosensation and transduction theories [12]. Positive relationship between muscle fitness and bone health indicators between children and adolescents aged 8-18 years, has been previously demonstrated in crosssectional studies [13,14], but evidence that muscular fitness in childhood and adolescence is a determinant of bone health in adulthood is still limited [15]. Recently, García-Hermoso et al. [16] performed a meta-analysis to conclude for a moderate effect of muscle fitness during childhood and adolescence on follow-up studies assessing areal bone mineral density (aBMD). Although evidence about the impact of mechanical stress through muscle action on bone modulation [17], it is not clear whether the advantages obtained are maintained when the physical stimulus ceases [18,19]. A few studies observed the maintenance of benefits from exposure to mechanical loads even years after physical activity has ceased [20,21]. In contrast, other studies showed that gains were not maintained after activity cessation or reduction [22,23]. Adult males and females who had been physically active during adolescence demonstrated bone mineral content 8 to 10% higher compared to inactive or moderately active peers during adolescence, and also had higher physical activity scores in adulthood [24]. Meantime, the association between muscle fitness indicators (MFI) in circumpubertal years and bone strength variables in adulthood could be attenuated after the model was adjusted for the performance of muscle fitness indicators in adulthood [25]. Thus, it is of interest to identify the direct, indirect or mediated effects by performance in MFI in adulthood on the relationship between muscle fitness in childhood and aBMD in adulthood. The current study is aimed to examine the relation-ship between MFI in childhood and different aBMD regions of interest in adulthood and to verify whether this relationship is mediated by performance of MFI in adulthood. Procedures and ethics committee This prospective study was initially designed to examine growth, maturation and physical fitness in schoolchildren aged 7 to 10 years recruited from a private school in Londrina (Paraná, Brazil), with a mixed longitudinal design with four birth cohorts (1992,1993,1994,1995) . Criteria of sampling were an α of 95%, a statistical power of 80%, and an error of 5% as previously described [26]. Parents or legal guardians of 1052 children signed an informed consent, and participants were informed that their participation was voluntary. Data were collected during Physical Education requiring three visits of the research team to elementary school within two weeks in all follow-up years. Participants of the preceding sample were contacted 15 years later to be reassessed as part of the "Physical fitness and practice of sports in childhood and adolescence and behavioral risk factors in adulthood" (ethics committee of Londrina State University approved the research: process N 1.340.735, November 27, 2015). The baseline and follow-up have been previously described [27]. Sample Inclusion criteria for the current study were: (i) not being injured or physically limited (as, for example, asthma); (ii) have at least one baseline measure for fitness tests to determined MFI were assessed; (iii) completed the same muscular fitness indcators battery as adults in addition to dual energy x-ray absorptiometry. Exclusion criteria were: participants under frequent use of medication to treat any disease that could interfere with the study variables ( Fig. 1). A total of 138 healthy young adults aged 18-25 years were eligible to participate in the study (50% males). Examination of dropouts compared to children who were assessed at baseline and follow-up is summarized in Table 1. Differences between children who participated as adults and those who were evaluated only at baseline did not indicate significant differences in variables, except for the sit-up test in males (p = 0.02). Anthropometry Anthropometry was assessed according to procedures described by Gordon et al. [28]. Body mass was measured on a digital platform scale with precision of 0.05 kg. Harpenden portable stadiometer with 0.1 cm precision was used to measure height. Subsequently, body mass index (BMI) was calculated and expressed in kg per m 2 . Childhood BMI z-score was calculated using reference values from the CDC Growth Charts [29] and was used as an adjustment variable in the statistical analyses. Muscular fitness indicators (MFI) Muscle fitness is understood as the capacity to perform work against resistance and involves maximum isometric or dynamic strength, isokinetics, muscular endurance and power [10]. The hand grip strength test (HS), strength measure, was completed according to procedures described by Soares and Sessa [30], using Jamar Hydraulic Dynamometer (Sammons and Preston Scientific Industries Inc.) with precision of 1 kilogram force (kgf ). Three measurements were performed and the best score was retained for analysis. Standing long jump test (SLJ), explosive strength indicator, corresponded to the maximal horizontal jumping performance performed from a starting position with participant with parallel feet. Three attempts were performed with the best scores in cm retained for analysis [30]. Sit-ups test, endurance strength indicator, required a mat and a stopwatch. With participants in dorsal decubitus, hips and knees flexed, feet soles facing the ground, arms crossing the thorax, hands supported on shoulders, the evaluator was holding the feet of participants who were instructed to perform the maximum number of trunk elevation including a contact of the forearms with the thighs and return to the initial position, the test was performed only once for a period of 60 seconds and the total number of repetitions was used in the analyses [31]. For the analysis, the total number of repetitions performed on a single trial was recorded. In addition to the performance in each indicator of muscular fitness during the childhood period, an index called childhood muscular fitness z-Score (CMF z-Score) was calculated by adding the standardized z-score value (individual value -mean / standard deviation) of the performance in each indicator of muscular fitness (HS + SLJ + Sit-ups). Regarding the quality control of observed data, the muscular fitness indicators of 20 adults randomly selected, after an interval of 7 days, were analyzed. Intraclass correlation coefficients for intra-observer reliability were: HS (ICC = 0.98), SLJ (ICC = 0.98) and sit-ups (ICC = 0.90). Dual energy X-ray absorptiometry (DXA) Participants were positioned on the table in supine position with the body aligned along with the central axis. A single certified technician completed the scans using DXA (Lunar DPX-MD+, GE Lunar Corporation, 726 Heartland Trail, Madison, WI 53717 − 1915 USA). Data were obtained using the software recommended by the manufacturer. Scans allowed calculations for aBMD of total body, lumbar spine (L1-L4), upper limbs, lower limbs, right femoral neck. The equipment was previously calibrated according to manufacturer. Full body scan was performed with participants in supine position and aligned, holding still for approximately 15 to 20 minutes. For the lumbar region, individuals were also positioned in dorsal decubitus, with legs placed on a block forming a 90-degree angle in relation to the table, with the intention of straightening the lumbar spine. For the proximal femur examination, keeping the patient positioned in dorsal decubitus, a triangular support was used to immobilize the lower limbs after internal rotation and adequate positioning of the femur, in order to capture the femoral neck region of interest. Data analysis Descriptive statistics of the sample were summarized in Table 2, separately for baseline and follow-up and sex. Comparisons between males and females were determined using independent t-test at baseline and followup. Linear regression using the enter method was used to analyze the relationship between MFI measured in childhood (baseline) and aBMD of different regions of interest measured in adulthood (follow-up). Regression analyses were adjusted for sex, chronological age, and childhood BMI z-score. Mediation analysis was performed on the MFI in childhood that showed statistically significant relationship with bone health indicators (BHI), considering the respective MFI in adulthood as the mediator variable. Mediation analysis followed the principles of Baron and Kenny [32] using the PROCESS 3.0 macro by Andrew F. Hayes, where "a" reflected the relationship between independent variable and the proposed mediator variable, "b" was the effect of the mediator variable on the dependent variable, partializing the effect of the independent variable, "c'" represented the direct impact of the independent variable on the dependent variable, and, "c'" represented the total effect of the independent variable on the dependent variable. The indirect effect is the product of "a" and "b" and quantifies the effect of the independent variable on the dependent variable through the mediator variable. Mediation assumptions were confirmed, the confidence interval of the indirect effect was estimated by the bootstrapping technique (5000, resampling), and unstandardized parameters were used to describe Betas. The mediation proportion was estimated by calculating 1 -(direct effect/total effect). Data were analyzed using SPSS version 25.0. The significance level adopted was 5%. Table 2 summarize the mean and standard deviation of males and females separately at the baseline and followup. At baseline, although boys and girls did not differ in terms of body size given by stature, body mass and BMI, significant differences were noted for the MFI (HS, p < 0.01; SLJ, p < 0.01; sit-ups, p < 0.01). Meantime, among the adult sample, males were heavier (p < 0.001), taller (p < 0.001), stronger (HS, p < 0.001; SLJ, p < 0.001; sit-ups, p < 0.001). Regarding aBMD that was uniquely assessed among adults, sex differences were significant for total body (p < 0.001), upper limbs (p < 0.001), lower limbs (p < 0.001), right femoral neck (p < 0.001). Results The association between MFI in childhood and aBMD of the different regions of interest in adulthood and significant values were obtained (Table 3). In the multiple linear regression procedure, adjustments for sex, age, and childhood BMI z-score were considered. Muscular fitness in childhood HS was significant predictors of total body aBMD (p = 0.040), trunk aBMD (p = 0.017) and upper limbs aBMD (p = 0.019). The sit-ups test was significant predictors of lumbar spine aBMD (p = 0.039) and trunk aBMD (p = 0.036), and the CMF z-Score was significant predictors of trunk aBMD (p = 0.040). The other relationships were not statistically significant. Subsequently to this identification, the respective MFI measured at follow-up, i.e., in early adulthood, was evaluated as mediators in the relationship between MFI in childhood and BHI in adulthood. The mediation analyses can be seen in Fig. 2 only in the models that attended the statistical assumptions. Considering the coefficients of regression equations in the mediation analysis identified as a, b, c and c' , as well as the significance level from the confidence interval of the indirect effect (a and b), it was observed that the mediator effect (indirect effect) between childhood HS and total body aBMD (β = 0.0025; 95% CI = 0.0005-0.0048), upper limbs aBMD (β = 0.0040; 95% CI = 0.0017-0.0069) was significant. Adulthood HS mediated approximately 52% of the relationship between childhood HS and total aBMD and 83% of the relationship between childhood HS and aBMD of the upper limbs (Fig. 2). No mediator effect (indirect effect) was observed between childhood sit-ups with lumbar spine aBMD. Discussion The aim of the present study was to verify the relationship between MFI in childhood and aBMD of the different regions of interest in adulthood and to verify whether this relationship is mediated by performance in MFI in adulthood. The main finding of this study is that adulthood HS plays a mediating role in the relationship between childhood HS and total and upper limbs aBMD, pointing out that muscle fitness in childhood may be important aBMD determinants in adulthood, especially when muscle fitness performance is maintained in the transition to adulthood. Direct effect was observed between childhood sit-ups with lumbar spine aBMD. The magnitude of the relationship between MFI in childhood with aBMD in adulthood found in the present study is low to moderate, according to results observed in literature [33,34]. When considering sex, chronological age, childhood BMI z-score, and MFI in determining BHI, models explained from 6% (sit-ups test and lumbar spine aBMD) to 55% (HS and upper limbs aBMD). Although the magnitude of the relationship and the explanation of some models are discrete, it is important to point out some considerations. There is a wide variety of determinants that influence PBM [7], even if low, making these results worthy of attention. Lower correlation coefficients are generally found in longitudinal studies as the time interval between measurements increases, and in this study, this interval is on average 13 years, so discrete coefficients are assumed [33,35]. Finally, small changes impact bone strength and postpone weakening that comes with osteoporosis [4,7]. Few studies have attempted to observe these relationships longitudinally, mainly in childhood. Foley et al. [25] evaluated the relationship between physical fitness in childhood with adulthood BHI by calcaneal quantitative ultrasonography measurements, and found positive and significant association between standing long jump and BHI in females; however, when performance in the standing long jump in adulthood was controlled, this association was not significant. A few other longitudinal studies evaluated the relationship between BHI obtained by DXA in adulthood and MFI in adolescence [33,34,36,37]. These studies are heterogeneous regarding the regions of interest measured by DXA and methods used such as MFI, but in general, they point to significant, positive, and low to moderate magnitude relationships. Additionally, as conducted by Kemper et al. [34], this study controlled regression analyses with sex and also obtained significant relationships between childhood MFI and adulthood BHI. There is a need to control the variable sex in analyses involving muscle fitness and aBMD, due to hormonal aspects, body size [38] and preference for specific physical activities that are inherent to sex and may not reach thresholds that stimulate osteogenesis. With regard to SLJ, muscle fitness indicator, showed no statistically significant relationship with aBMD for any of the regions of interest, although some studies have reported positive association [27,41], others have reported negative association [40] and some found no association [41]. The aspect that may explain this result in motor tests involving body displacement to estimate muscle strength, is the body weight, which can play a key role in performance during the execution of tests, and those with weight values below average may perform better [40]. In addition, SLJ requires greater motor efficiency, so force generation with lower limb speed may be affected by poor motor coordination [40,10]. On the other hand, relationships found between HS and upper limbs aBMD and sit-ups with lumbar spine aBMD can be understood through the mechanostatic theory, via interaction of bones with muscles [11]. The role of HS in childhood and adults with aBMD is worth highlighting. This relationship between HS and total body aBMD may be explained, in part, by the ability of the handgrip test to represent the individual's overall strength level [42]. Furthermore, mechanical and biological stimuli can trigger generalized systemic endocrine effects, such as myokines in bone metabolism, which could explain the interaction of bone and muscle tissue even at anatomically distant sites [43]. With regard to mediation analysis, mediation of adulthood HS in relationships between childhood HS and total body and upper limbs aBMD and direct effect of childhood sit-ups on lumbar spine aBMD were found. This result allows concluding that muscle fitness performance in childhood provides higher adulthood aBMD, especially when higher muscle fitness performance is maintained in adulthood. Thus, the stimuli need to be constant in order to maintain the osteogenic effects. Furthermore, these relationships were longitudinally examined, while other studies have investigated mediator variables in similar relationships with cross-sectional design, such as the work by Torres-Costoso et al. [40] and Vicente-Rodríguez et al. [44]. A recent study found a mediator effect of muscle mass on the relationship between physical activity in child-hood and adolescence and bone parameters at the age of 17 years, highlighting the important effect of muscle fitness on BHI [45]. Studies have sought to analyze the mediator effect of muscular fitness on the association between other factors such as eating disorders [46], sports involvement [47], vitamin D [48], physical activity [49] and BHI. Finally, direct relationship of sit-ups in childhood with aBMD of the lumbar spine in adulthood was observed, sit-ups test performance did not play a mediating role in the relationship of equivalents in childhood. The direct effect of childhood MFI on adulthood aBMD highlight the importance of improving muscle fitness from childhood. The strength of this study is its longitudinal design, capable of partially inferring the causal relationship among variables analyzed. Follow-up studies can have a dropout effect, and in this analysis, they were negligible and not significant. Another potential was the assessment of bone variables and application of motor tests that analyze different body regions, allowing close and distant associations of anatomical specificities. Limitations include sample size, which made stratification by sex unfeasible, but allowed controlling sex in the analyses; the application of motor tests in the identification of muscle fitness, which suffers interference from other variables that could not be controlled; obtaining aBMD only at the adult moment and lack of control of other confounding variables, such as calcium, vitamin D intake, and practice of physical activity between childhood and adulthood. Future studies should invest in assessing and training muscular fitness to promote bone health at all stages of the life cycle, seeking to adequate the dose response needed to obtain advantages and to seek to maintain health benefits already obtained. Conclusion It was possible to conclude that some MFI in childhood showed significative relationship with BHI in adulthood, such as between HS and total body and upper limbs aBMD, sit-ups test and lumbar spine aBMD. In addition, the sit-ups test in childhood had a direct effect on lumbar spine aBMD in adulthood, highlight the importance of improving muscle fitness from childhood. Adulthood HS mediated the relationship between childhood HS and total body and upper limbers aBMD, pointing out that muscle fitness in childhood may be an aBMD determinant in adulthood, especially when higher muscle fitness performance in adulthood is maintained.	2023-04-05T13:54:19.925Z	2023-04-04T00:00:00.000	{ "year": 2023, "sha1": "180560a895c620e22303f4d730504d6a2fd31e13", "oa_license": "CCBY", "oa_url": "https://bmcpublichealth.biomedcentral.com/counter/pdf/10.1186/s12889-023-15545-7", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "180560a895c620e22303f4d730504d6a2fd31e13", "s2fieldsofstudy": [ "Psychology", "Biology", "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
253941857	pes2o/s2orc	v3-fos-license	A Possibility of Vasospastic Angina after mRNA COVID-19 Vaccination We report a case of vasospastic angina (VSA) following COVID-19 mRNA vaccination. Despite the widespread occurrence of myocarditis, there have been few reports of post-vaccinal VSA. A 41-year-old male patient was referred for chest pain at rest following mRNA vaccination; he had never experienced chest pain prior to vaccination. He was diagnosed by an acetylcholine (Ach) provocation test that showed multivessel vasospasm. After the initiation of treatment with a calcium channel blocker and nitrate, no further exacerbation of chest pain was observed. To our knowledge, this constitutes the first reported case of VSA proven by Ach provocation test after COVID-19 vaccination. The vaccination may increase coronary artery spasticity. VSA should be ruled out in post-vaccine new onset resting chest pain. Case Presentation A 41-year-old Japanese male smoker with no medical history received the first dose of the mRNA-BNT162b2 COVID-19 vaccine (Pfizer-BioNTech, Cambridge, MA, USA). One week after vaccination, chest pain and palpitations appeared after drinking alcohol; he had no previous documented chest pain prior to his COVID-19 vaccination. His chest pain repeated exacerbations and remissions at rest and he was admitted to our hospital on the 11th day after vaccination. On admission, his body temperature was 37.2 • C, blood pressure was 126/80 mmHg, heart rate was 98/min, and oxygen saturation was 98% on room air. When he arrived at our hospital, his chest pain was resolved, and his electrocardiogram (ECG) and transthoracic echocardiogram were normal. His blood test results were within normal limits, including creatine kinase (124 IU/L), high-sensitivity cardiac troponin T (0.003 ng/mL), potassium (3.7 mEq/L), creatinine (0.83 mg/dL), and C-reactive protein (0.04 mg/dL) on admission. After he developed chest pain at rest and after alcohol consumption, we suspected VSA. Emergency coronary angiography revealed no coronary artery stenosis. However, the acetylcholine (Ach) provocation test showed multivessel vasospasm with chest pain (left coronary artery (LCA), segments #7 99% stenosis and #12 total occlusion at 100 µg dose into Vaccines 2022, 10, 1998 2 of 6 LCA) with ST elevation in ECG leads I, aVL and large T wave in leads V1-V4 ( Figure 1A-E). He was diagnosed with VSA because his findings fulfilled all criteria (i) reproduction of the usual chest pain, (ii) ischemic ECG changes, and (iii) 90% vasoconstriction on angiography. Vaccines 2022, 10, x 2 of 6 suspected VSA. Emergency coronary angiography revealed no coronary artery stenosis. However, the acetylcholine (Ach) provocation test showed multivessel vasospasm with chest pain (left coronary artery (LCA), segments #7 99% stenosis and #12 total occlusion at 100 µ g dose into LCA) with ST elevation in ECG leads I, aVL and large T wave in leads V1-V4 ( Figure 1A-E). He was diagnosed with VSA because his findings fulfilled all criteria (i) reproduction of the usual chest pain, (ii) ischemic ECG changes, and (iii) 90% vasoconstriction on angiography. Therefore, we initiated a calcium channel blocker (CCB) with nitrate and he was discharged the next day. Antibody titers of various viruses, including SARS-CoV-2, were negative. After the initiation of treatment with a CCB and nitrate, no further exacerbation of chest pain was observed. Therefore, we initiated a calcium channel blocker (CCB) with nitrate and he was discharged the next day. Antibody titers of various viruses, including SARS-CoV-2, were negative. After the initiation of treatment with a CCB and nitrate, no further exacerbation of chest pain was observed. Discussion To our knowledge, this constitutes the first reported case of VSA proven by Ach provocation test after COVID-19 vaccination. He was diagnosed with VSA because his Ach provocation test findings fulfilled all criteria: (i) reproduction of the usual chest pain, (ii) ischemic ECG changes, and (iii) 90% vasoconstriction on angiography [14]. Additionally, after the initiation of treatment with a CCB and nitrate, no further exacerbation of chest pain was observed. Chest discomfort is present in 3% of cases after receiving the mRNA vaccine [15]. Myocarditis, acute coronary syndrome, and VSA have been reported as potential causes of chest pain associated with mRNA vaccination against COVID-19 [1,2]. However, reported cases of VSA following the COVID-19 mRNA vaccine are scarce and the true incidence remains unknown. Acute coronary syndromes (ACS) and angina pectoris in the elderly are reported to often present as atypical chest pain and more cardiac failure [16,17]. Therefore, it is important not to rule out ACS or angina even in the presence of atypical chest pain after vaccination in the elderly. Notably, a flowchart of the differential diagnoses of chest discomfort and palpitation after COVID-19 vaccination has been reported [18]. Regarding the relationship between vaccination and VSA, several mechanisms may induce VSA. The spike protein of the SARS-CoV-2 virus or the mRNA component of the vaccine binds to the angiotensin-converting enzyme 2 (ACE2) receptor, which leads to the downregulation of ACE2 (Figure 2) [3,4]. ACE2 converts the potent vasoconstrictor angiotensin II (AngII) to the vasodilator angiotensin 1-7. Hence, the downregulation of ACE2 activates AngII, which leads to vasoconstriction and also produces interleukin (IL)-6 [3,5]. Various inflammatory stimuli, including AngII and IL-6, upregulate the activity of Rho-kinase. The activation of Rho-kinase plays a central role in coronary artery spasms caused by the hypercontraction of vascular smooth muscle cells (Figure 2) [19]. To our knowledge, this constitutes the first reported case of VSA proven by Ach provocation test after COVID-19 vaccination. He was diagnosed with VSA because his Ach provocation test findings fulfilled all criteria: (i) reproduction of the usual chest pain, (ii) ischemic ECG changes, and (iii) 90% vasoconstriction on angiography [14]. Additionally, after the initiation of treatment with a CCB and nitrate, no further exacerbation of chest pain was observed. Chest discomfort is present in 3% of cases after receiving the mRNA vaccine [15]. Myocarditis, acute coronary syndrome, and VSA have been reported as potential causes of chest pain associated with mRNA vaccination against COVID-19 [1,2]. However, reported cases of VSA following the COVID-19 mRNA vaccine are scarce and the true incidence remains unknown. Acute coronary syndromes (ACS) and angina pectoris in the elderly are reported to often present as atypical chest pain and more cardiac failure [16,17]. Therefore, it is important not to rule out ACS or angina even in the presence of atypical chest pain after vaccination in the elderly. Notably, a flowchart of the differential diagnoses of chest discomfort and palpitation after COVID-19 vaccination has been reported [18]. Regarding the relationship between vaccination and VSA, several mechanisms may induce VSA. The spike protein of the SARS-CoV-2 virus or the mRNA component of the vaccine binds to the angiotensin-converting enzyme 2 (ACE2) receptor, which leads to the downregulation of ACE2 (Figure 2) [3,4]. ACE2 converts the potent vasoconstrictor angiotensin II (AngII) to the vasodilator angiotensin 1-7. Hence, the downregulation of ACE2 activates AngII, which leads to vasoconstriction and also produces interleukin (IL)-6 [3,5]. Various inflammatory stimuli, including AngII and IL-6, upregulate the activity of Rhokinase. The activation of Rho-kinase plays a central role in coronary artery spasms caused by the hypercontraction of vascular smooth muscle cells (Figure 2) [19]. Other plausible mechanisms are the involvement of inflammatory cytokines and allergic reactions. While lipid nanoparticles (LNPs) play an important role as a delivery system for mRNA vaccines, cationic lipids in LNPs themselves reportedly cause inflammatory cytokines such as interleukin (IL)-6 and IL-1β (Figure 2) [7][8][9]. The IL6 genotype has been reported to affect the prevalence of VSA [10]. Four post-mRNA vaccine autopsy cases have recently been reported to express genes that upregulate inflammatory cytokine signaling compared with controls [6]. Inflammatory cytokines inhibit endothelial nitric oxide and increase Rho-kinase activity, causing coronary artery spasms ( Figure 2) [11,19]. Other plausible mechanisms are the involvement of inflammatory cytokines and allergic reactions. While lipid nanoparticles (LNPs) play an important role as a delivery system for mRNA vaccines, cationic lipids in LNPs themselves reportedly cause inflammatory cytokines such as interleukin (IL)-6 and IL-1β (Figure 2) [7][8][9]. The IL6 genotype has been reported to affect the prevalence of VSA [10]. Four post-mRNA vaccine autopsy cases have recently been reported to express genes that upregulate inflammatory cytokine signaling compared with controls [6]. Inflammatory cytokines inhibit endothelial nitric oxide and increase Rho-kinase activity, causing coronary artery spasms ( Figure 2) [11,19]. East Asians are genetically predisposed to VSA since 40% are aldehyde dehydrogenase 2 (ALDH2)2 carriers with deficient ALDH2 activity [20]. In ALDH22 carriers, alcohol intake and tobacco smoking increase toxic aldehydes, which cause endothelial dysfunction ( Figure 2). Although we did not examine the ALDH2 genotype of the patient, we speculated that he was an ALDH22 carrier since he had alcohol flushing syndrome (AFS). AFS is associated with ALDH22 (odds ratio 51.6) [20]. He had never experienced chest pain prior to receiving the COVID-19 mRNA vaccine; therefore, post-vaccinal alcohol consumption may be a factor that triggered VSA. COVID-19 mRNA vaccine components, including polyethylene glycol, LNPs, etc., may cause an allergic reaction, resulting in a coronary vasospasm called Kounis syndrome [2,12]. Mast cells activated by the vaccine antigen induce histamine with or without immunoglobulin E [12], and histamine reportedly induces coronary vasospasm [13]. β1adrenoceptors (ARs) are predominantly found in the coronary epithelium, while β2-ARs are found in the microvascular [21]. It has recently been reported that Kounis syndrome is more likely to occur when β2-ARs, but not β1, are blocked [22]. Therefore, patients taking nonselective β-ARs antagonists such as propranolol should be aware of Kounis syndrome following vaccination. Although this case highly suggested the relationship between COVID-19 mRNA vaccine and VSA, VSA and microvascular dysfunction, not epicardial coronary artery stenosis has also been reported in COVID-19 patients [23,24]. Therefore, the benefit-risk balance should always be considered when deciding whether or not to vaccinate. Conclusions This case demonstrates a possibility of VSA after COVID-19 mRNA vaccination. VSA should be ruled out in post-vaccine new onset resting chest discomfort and palpitation, especially in cases with repeated exacerbations and remissions. Informed Consent Statement: Written informed consent was obtained from the subject for the publication of this case report. Data Availability Statement: Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study. Conflicts of Interest: The authors declare no conflict of interest.	2022-11-26T16:58:58.380Z	2022-11-24T00:00:00.000	{ "year": 2022, "sha1": "f4cfed2e4ad4849c19e7391404c1efa181dc870f", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2076-393X/10/12/1998/pdf?version=1669285111", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "5282b7ad377db5d4b68100e4078871f26fb74935", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
218538886	pes2o/s2orc	v3-fos-license	Contribution to COVID-19 spread modelling: a physical phenomenological dissipative formalism In this study, we propose an evolution law of COVID-19 transmission. An infinite ordered lattice represents population. Epidemic evolution is represented by a wave-like free spread starting from a first case as an epicentre. Free energy of the virus on a given day is defined equal to the natural logarithm of active infected cases number. We postulate a form of free energy built using thermodynamics of irreversible processes in analogy to isotherm wave propagation in solids and non-local elastic damage behaviour of materials. The proposed expression of daily free energy rate leads to dissipation of propagation introducing a parameter quantifying measures taking by governments to restrict transmission. Entropy daily rate representing disorder produced in the initial system is also explicitly defined. In this context, a simple law of evolution of infected cases as function of time is given in an iterative form. The model predicts different effects on peak of infected cases Imax and epidemic period, including effects of population size N, effects of measures taking to restrict spread, effects of population density and effect of a parameter T similar to absolute temperature in thermodynamics. Different effects are presented first. The model is then applied to epidemic spread in Tunisia and compared with data registered since the report of the first confirmed case on March 2, 2020. It is shown that the low epidemic size in Tunisia is essentially due to a low population density and relatively strict restriction measures including lockdown and quarantine. Introduction First cases of pneumonia unknown etiologies have been declared in Wuhan, China, since December 8, 2019. Pneumonia starts with severe acute respiratory infection symptoms and some cases developed acute respiratory distress syndrome with failure complications. On January 7, 2020, Chinese centre for disease control and prevention identified a new coronavirus . COVID-19 is a human coronavirus include in the gender beta coroanvirus group 2b, family coronaviridae. It is the third strain of virus of the coronavirus family (CoV), isolated in humans in the context of an epidemic after SARS-CoV in China (2002) and MERS-CoV in Saudi Arabia (2012). Examination of the COVID-19 genome showed genetic similarity to SARS-CoV about 79.5%. Human to human transmission takes place by either respiratory droplets or close contacts. According to the world health organization, COVID-19 is a virus with unique characteristics that causes respiratory disease and which spreads via oral and nasal droplets (Kolifarhood et al. 2020). On March 2, the first case has been declared in Tunisia. Actually, COVID-19 is causing a disease representing a planetary problem for public health and negative impact on humanity (Boccaletti et al. 2020). The objective of this paper is to propose a simple model to predict COVID-19 transmission using early data of the outbreak. Majority of epidemic transmission models are based on compartmental mathematical models dividing population in different interacting groups and assuming different rates of transmission between them. Solutions are conducted using integration of differential equations and principle of conservation (Kermack and McKendrick 1927). Population is generally assumed as a closed system, the probabilistic formalism of transmission between individuals of different groups leads to saturation and a population size effect on epidemic size and epidemic period. Models that are more sophisticated include also Monte Carlo numerical simulations for stochastic models and more realistic epidemic networks. See for example a review by House et al. (2013) and recent studies by Kim et al. (2020), Liang (2020) and among several others for COVID-19 modelling. Epidemic networks and lattice methods have their origin in social science and computer science (see for example a review by Keeling and Eames 2005). Lattice models are representation of an ordered network in which epidemic transmission is similar to a wave-like spread in regular grid representing connected individuals. Epidemic starts from an epicentre and spreads out in a roughly circular manner. Figure 1 illustrates an example in two dimensions. Lattice models are suitable for example for forest-fire models (Bak et al. 1990) where nodes represent trees that burn leaving empty sites. Keeling and Eames (2005) interpreted this representation as similar to epidemic transmission. This paper proposes, in this context, a phenomenological model of propagation and dissipation in analogy to elastic wave propagation and a size and temperature-dependent elastic damage material model (Ben Hassine et al. 2019;Limam et al. 2014). The advantage of a formalism inspired from thermodynamics is that different effects emerge from principles. Materials and methods Wave propagation is particularly studied in biomechanical applications for example in evaluation of dental implant stability (Vayron et al. 2015). In addition, elastic damage models are considered in geomechanics and biomechanics to model shock waves dissipation. For example, Nelms et al. (2017) applied a finite element elasto-plastic damage model in order to evaluate mechanical shock waves decay in cement microstructure. Fovargue et al. (2018) developed a model for kidney stone fragmentation in shock wave lithotripsy assuming elastic damage behaviour with reduction in the apparent Young modulus, in a context of thermodynamics of generalized standard materials behaviour (Lemaître and Chaboche 1978). Non-local damage mechanics are particularly considered to model shock waves dissipation. For example, Lu et al. (2005) developed a non-local damage approach to model damage wave slow propagation in solids. Non-local damage approaches predict a size effect on macroscopic constitutive behaviour (Pijaudier-Cabot and Bazant 1987). Size effect phenomenon was experimentally observed in biomechanics since the time of Galileo. Furthermore, Limam et al. (2014) andBen Hassine et al. (2019) proposed that, for isotherm processes, a scaling law should be associated with specific entropy S depending on size N and that the specific free energy should be written = U − TS(N) . In fact, internal energy is extensive and specific internal energy U should be independent of N. On the contrary, entropy is non-additive at small scales and become additive as size N tends to infinity (Tsallis 2009). Irreversible isotherm processes assume equilibrium at successive states and allow heat transfer in agreement with Fourier law. For this isotherm case, specific free energy linearly decreases as temperature increases. This was confirmed experimentally. We can cite, for example, structural health monitoring related to concrete and composite structures (Wang et al. 2018;Moll et al. 2019). On the contrary, Carlioz et al. (2019) showed that material damage corresponding to high velocity process and to sudden and abrupt nature of a crack nucleation process is adiabatic rather than isothermal with temperature rising at crack tip during dynamic crack propagation due to heat accumulation at solid boundaries. In this paper, in the case of epidemic transmission context, isotherm open system hypothesis will be considered in analogy to isothermal damage process allowing heat transfer. Physicians who modelled epidemics spread using information entropy concept and epidemic thermodynamics, assumed this hypothesis. For example Koivu-Jolmaa and Annilaa (2018) proposed an isotherm natural process based on thermodynamics of open systems to predict Ebola virus transmission using statistical mechanics and analogy to chemical potentials. Tsallis and Tirnakli (2020) developed analytical model to predict COVID-19 transmission based on non-extensive Tsallis Entropy. The proposed scientific method proposed in this paper is developed in the same context of isotherm open systems considered for example by Koivu-Jolmaa and Annilaa (2018) and Tsallis and Tirnakli (2020), but with a free energy postulated equal to the natural logarithm of active infected cases and written in the form U − TS(N) . The proposed approach can be viewed as part of epidemic physical models. It represents also a complement to compartmental mathematical models and complex network based on sophisticated approaches taking in to account real interaction between different groups and heterogeneous fields of motion and density depicted in real maps (Gomez et al. 2020). We adopt a method based on an empirical macroscopic description of epidemic spread inspired from mechanics and thermodynamics. It is noted that phenomena related to effects of ambient temperature or population size on COVID-19 epidemic transmission were recently reported in the literature. The proposed model is presented and explained in Sect. 3. In Sect. 4, the model is applied to study different effects including measures to restrict spread, effect of population density and size and effect of a parameter T similar to absolute temperature in thermodynamics. Epidemic evolution in Tunisia is also analysed. We consider for comparison, data from national observer for new and emergent diseases (https ://www.onmne .tn) until June 12. We present also an "Appendix". Firstly, this appendix aims to present the considered mechanical model (Ben Hassine et al. 2019). We present the simplified case of unidirectional behaviour, which is macroscopic, and size dependent. This model is therefore non-local in the sense of damage mechanic. Secondly, we present some phenomenological similarities between virus spread and mechanical wave spread. Finally, this appendix includes also a table with both, epidemiology and damage mechanics to list the similarities, parameters, Equations local or non-local and their interpretation. Theory Consider first a simple mathematical model given by Eqs. (1) and (2). I n is the number of infected people on day n . Theses Equations correspond to epidemic theoretical free transmission in a population of size N defined in a finite roughly circular convex domain part of a perfect infinite ordered lattice of connected people and starting from an epicentre I 0 = 1 belonging to the domain. Population density is inversely proportional to the square of distance d as depicted in Fig. 1. Coefficient C defined by Eq. (1) is considered as an intrinsic characteristic of population density and independent of population size. It is clear that it decreases when distance d increases which means that when population density decreases. It represents the number of transmission between every infected person at wave front to other persons. A theoretical free transmission in the lattice corresponds to the linear curve with a slope ln[C] in a semi-logarithmic scale as depicted in Fig. 2a, b for a domain representing, for example, a typical dense city with N = 12 Million and C = 1.62. In that case, population size N will be reached at n s = ln(N) ln(C) , on day 34; meanwhile epidemic will continue to propagate in the lattice outside the population domain, as the considered population is fixed but transmission to the outside was made possible by hypotheses. In reality transmission, process is dissipative and the linear curve of slope ln[C] corresponds only to early stage. It can be deduced, for example, from initial data curves fitting in a semi-logarithmic curve as shown in Fig. 2. This coefficient should be reduced by measures imposed by governments including lockdown and quarantine. It is reduced also by population behaviour including social distancing, personal hygiene, for example by wearing a mask in fear of the spread of the virus. This is similar to the effect described by, for example, Kim et al. (2020) or Liang (2020).This behaviour emerges naturally after first deaths inducing a disorder and long range interaction in the initial lattice of Fig. 1, which means physically that entropy should increase. The idea of the proposed model can be highlighted when we made an empirical analogy of the linear curve as depicted in Fig. 2 with energy as function of time of elastic wave propagation in a rod obeying Hooke's law of elasticity and submitted to harmonic imposed power. Therefore, we defined by analogy free energy of the virus spread by = ln I n which gives a constant daily rate in the case of linear curve as depicted in Fig. 2 and defined by Eqs. (1) and (2). In reality, material behaviour as epidemic spread is dissipative and wave velocity will decrease due to material damage. In analogy to reduction in Young modulus of elasticity in damage mechanics (Kachanov 1958), we should introduce a reduction in C in a semi-logarithmic scale. Therefore, we postulate the evolution model of infected cases I n on day n, given by Eqs. (3-7), where T is a parameter similar to absolute temperature in thermodynamics and where a first case is I 0 = 1. In order to give a physical sense to these Equations, we define first a free energy rate of the virus noted n+1 − n = Δ given by Eq. (8). In a thermodynamically consistent isotherm framework, Helmholtz free energy is defined by the rate Δ = ΔU − TΔS , where ΔU is internal energy rate and ΔS is entropy rate given, respectively, by Eqs. (9) and (10) in the case of virus spread analogy and identified from Eq. (8). (3) (4) D n = 0 for n ≤ 21 and D n = J n−21 J n+1 for n > 21 (5) F n = 0 for n ≤ 21 and F n = J n N for n > 21 (7) C(n) = C for n ≤ 21 and C(n) = CR for n > 21 In order to consider lockdown and quarantine effect, C is decreased from day 22 and noted CR, a coefficient between 1 and C (Eq. 7). Equation (4) is introduced to model recovering or death from day 22 with parameter D n . This is justified by recent studies reporting that observed duration of viral shedding among survivors was between 8 and 37 days ). This means an average of 22 days corresponding to a first death. This means also different behaviours of governments and people for different time intervals n ≤ 21 and n > 21 before and after this event. Equations (4, 6 and 7) define different coefficients C and CR to distinguish between no measures before the event and measures taking after the event. Evolution law of damage defined by these Equations is independent of population size and temperature and can be considered as an intrinsic property of virus spread but dependent upon people density through coefficient C and measures taking through coefficient CR. C and CR are therefore depend on different cities or countries. Equations (3) to (7) were inspired using similarities with non-local damage behaviour as shown in Table 1. As explained in the "Appendix" non-local damage predicts a size effect and local damage does not. In Table 1, the particular case of local damage behaviour corresponding to T = 0 is also presented, where Eq. (11) is obtained from Eq. (3) using Eqs. (4) and (6). We can underline here that Eq. (11) is a conservation Equation eliminating recovered or deceased individuals from infected individuals to obtain only active infected cases with a delay of 22 days. This Equation is similar to Equations defined by compartmental models as for example SIR or SEIR models. Meanwhile, the advantage of the present physical approach is that some intrinsic characteristics of virus spreading are considered in the model, including local interaction between individuals leading to its transmission and non-local interactions leading to its dissipation due to an entropic effect as explained in analogy to the non-local damage model presented in the "Appendix". Its application with CR = 1.15 ignoring entropy dissipative effect which means with T = 0, leads to the trilinear curve presented in Fig. 2a, b where population size will be reached with an epidemic period n s = 89 days for the considered example in Fig. 2. Figure 2a shows also the theoretical case of an ideal lockdown with CR = 1 and T = 0. This leads to an epidemic period of 40 days due to a recovering rate higher than infection rate in that case with an epidemic peak of 11,200, reached on day 21 and independent of N. Evolution law of damage given by Eqs. (4, 6 and 7) is defined exclusively by C and CR and remains (10) ΔS = ln 1 + F n (11) independent of population size. This evolution law of damage induces a decreasing of internal energy rate defined by Eq. (9) which remains also independent of population size and temperature and equal to free energy rate when T = 0. Effects of T and N are rather due to entropy production. Similar hypothesis was considered for a damage evolution law as an intrinsic characteristic of the material independent of specimens size and temperature see Ben Hassine et al. (2019) and (Limam et al. (2014). Equation (5) is introduced to consider population size effect, also from day 21. It is worth mentioning that the considered free energy is choosing with an entropy rate ΔS = ln 1 + F n , null before damage initiation (F n = 0 for (n ≤ 21)) and always positive, which means that entropy increases according to the second law of thermodynamics and contributes to dissipate free energy of the virus. Theoretical free transmission in an ordered lattice case given by Eqs. (1) and (2) can be obtained when considering CR = C, and T = 0, in analogy to absolute zero state in thermodynamics, where entropy effect vanishes. Parameter T should be understood as for example hygiene measures in the system which can be linked also to ultraviolet rays increasing with ambient temperature rising. When increased it contributes to increase entropy effect and consequently to decrease free energy rate and epidemic spread. This is in agreement with recent environment studies shown also through statistical analysis of data that transmission decreases as ambient temperature increases, see for example Prata et al. (2020) and Liu et al. (2020). Figure 2b shows an example applying the proposed model with N = 12 Million, C = 1.62, CR = 1.15 and T = 7. Introducing entropic effect, epidemic size is decreased with a peak of 79,000 infected cases and an epidemic period of 56 days. Parameters were chosen to give an order of epidemic comparable to a dense city like Wuhan (Liang 2020;; . C was identified from the first slope of data using a regression analysis between day 8 and day 21. This choice of linear behaviour was not arbitrary. In fact, we assumed that virus spread similarly to material behaviour is linear at early stages of loading. Similarities between local elasticity and linear spread are depicted in Table 1 ("Appendix"). Figure 3 presents an example of the model applied first with N = 12Million, C = 1.62, CR = 1.15 and T = 7 in a Cartesian scale and then with the same parameters but more restricted measures traduced by a reduction in C on day 21 to CR = 1.1. It can be observed that CR = 1.1 describes stricter measures that induce a decreasing of epidemic size and slightly increase epidemic period. The maximum of positive cases Imax is decreased from 79,000 to 46,000. Figure 3 presents Free energy of the virus n = ln I n = n ln (C) Restriction measures (CR) effect Specific internal power proportional to Young modulus t ∝ E Free energy rate of the virus Δ n = n+1 − n n+1−n = ln also real data of infected active cases in Wuhan, China (Liang 2020;, until epidemic wave end on March 2020. Real data show that I max is equal to 57,900 cases, which means that CR is effectively between 1.1 and 1.15. It is observed also that the epidemic real period is larger than the predicted one. This can be explained by a possibility of a second epicentre with a shift in time and space from the first one. This can be confirmed also by the shift in the data curve shape. It can be noted here that the model assumes a unique epicentre. A superposition can be made to include the effect of a second epicentre. Figure 4 shows effect of population size N on infected cases in a Cartesian scale. C, CR and T are fixed and the size N is changed from 3 to 12 and 24 Million. It is deduced that when population size increases epidemic size and period increase, with respective maximums I max of 25,000, 79,000 and 128,000 reached, respectively, on days 30, 39 and 42with, respectively, epidemic periods of about 41, 56 and 60 days. When N increases the ratio (I max /N) decreases and, respectively, given by 0.83%, 0.66% and 0.53%. Figure 4 presents also real data of infected active cases in Wuhan, China. Comparison shows an agreement with data. Estimated population size in Wuhan is 11 Million. Figure 5 presents data in Cartesian scale in Tunisia until April 25. On this date, our ministry of health reported 38 deaths and 194 recovered cases. The model is depicted and reproduces actual data tendencies, considering C = 1.3 corresponding to initial data fitting in semi-logarithmic scale. It represents quarantine and lockdown effects thereafter with CR = 1.115.The model reproduces data tendencies with I max = 912 and an epidemic period of 92 days, which means an epidemic spread end at the beginning of June if the same measures are maintained. It is noted that a reduction in C from 1.62 to 1.3 induces a reduction in epidemic peak of about 50 times as deduced when comparing Fig. 5 with Fig. 3. Furthermore, it is noted in Fig. 5 that parameter T when changed from 7 to 25 slightly decreases epidemic. Furthermore, Fig. 6 presents updated real data until June 12. A good agreement is observed between data and the model. Discussion An evolution law of COVID-19 based on analogy with wave propagation in elastic solids and a non-local damage model is proposed. The key coefficient C is obtained by a linear fitting of initial slope of data in a semi-logarithmic scale between day 8 and day 21. Results are very sensitive to this parameter, considered as an intrinsic parameter of population density. For populations of a comparable size, but with coefficients, respectively, C = 1.3 and C = 1.62, results show that epidemic size can increase very fast. The second important parameter is CR. It traduces measures like lockdown and quarantine. When controlled, which means decreased, it decreases epidemic size. Conclusion The low predicted epidemic size in Tunisia is essentially due to a low population density (C = 1.3) and strict restriction measures (CR = 1.115). Population density is inversely proportional to the square of distance d depicted in Fig. 1 which explains its important effect on transmission. Information exchange, democratization of access to knowledge and long-range non-local interaction between humans at a planetary level contribute to dissipate pandemic and help humanity to be prepared to a possibility of a second wave. The model was applied to unidirectional isotherm behaviour of concrete under quasi-static compression. It was validated using comparison with experimental tests. It was derived in a context of generalized standard behaviour of material verifying Clausius-Duhem inequality (Limam et al. 2014). Constitutive macroscopic behaviour (Eq. 12) exhibits a size and ambient temperature effects on macroscopic strength and secant modulus of elasticity with decreasing as size and temperature increase. Consider a rod of length L and imposed displacements u(0) and u(L) at its boundaries. This macroscopic behaviour derives from a chosen specific free energy L given by Eq. (13) where D is damage parameter between 0 and 1, E is Young modulus and k is an increasing positive function with horizontal asymptote. A unique evolution law of damage parameter D as function of loading history was considered. This law is independent of the size L and absolute temperature. Evolution law of damage was considered as an intrinsic characteristic of material. It is important to note here that specific free energy was obtained through a second-order Taylor series development ln (1 + D) ≈ D − D 2 2 and can be written as follows: This specific free energy L = U − TS(L) was defined as function of a specific internal energy U and specific entropy S given, respectively, by Eqs. (15) and (16). Specific internal energy U is independent of L and internal energy is proportional to L; meanwhile specific entropy S(L) depends on L and is size dependent. This hypothesis is based on physical principles. For discrete physical systems with weak correlation between microstates, entropy associated with a given macro-state is not necessary additive at small sizes. However, it becomes additive for large sizes. This means also that specific entropy increases as function of system size and tends asymptotically to a constant for large sizes (Tsallis 2009). In Eq. (12), stress is the sum of tow components the first one is derived from the internal energy and given by Eq. (17). The second component is an entropic stress given by Eq. (18) Physicians state that entropic forces are long range (March's principle). This means that long-range non-local interactions should be defined in order to obtain macroscopically Eq. (18). Figure 1 is a simple representation drawn to understand phenomenologically virus spread. The key ideas to explain virus spread is that: 1. Without exchange of information between individuals, virus is spreading by transmission or diffusion due to close contact. This means that transmission is a local phenomenon. This means also that persons represented by nodes in Fig. 1 are not informed about what happens far from them. 2. Long range exchange of information and interaction between persons induce a non-local entropic force opposite to virus spread in analogy to mechanical entropic stress given by Eq. (18). Dissipation of spread is therefore a non-local phenomenon. Table 1 presents, firstly, similarities between elastic wave propagation in a semi-infinite rod submitted to a constant power and free linear spread. Similarities are extended thereafter to shock wave propagation in a semi-infinite rod modelled as a non-local damage medium. Resolution of dynamics with non-local equations remains possible but complicated. The presented Equations are therefore phenomenological and macroscopic.	2020-05-08T13:04:46.059Z	2020-05-08T00:00:00.000	{ "year": 2020, "sha1": "44570a1d1af67c151fe66b25d28950a9fc21a90d", "oa_license": null, "oa_url": "https://link.springer.com/content/pdf/10.1007/s10237-020-01387-4.pdf", "oa_status": "BRONZE", "pdf_src": "PubMedCentral", "pdf_hash": "274457434db6ce2bff394ba848ea1cc8e9173279", "s2fieldsofstudy": [ "Physics", "Environmental Science" ], "extfieldsofstudy": [ "Mathematics", "Medicine" ] }
6787682	pes2o/s2orc	v3-fos-license	Fast Generation of Translation Forest for Large-Scale SMT Discriminative Training Although discriminative training guarantees to improve statistical machine translation by incorporating a large amount of overlapping features, it is hard to scale up to large data due to decoding complexity. We propose a new al-gorithm to generate translation forest of training data in linear time with the help of word alignment. Our algorithm also alleviates the oracle selection problem by ensuring that a forest always contains derivations that exactly yield the reference translation. With millions of features trained on 519K sentences in 0.03 second per sentence, our system achieves signiﬁcant improvement by 0.84 B LEU over the baseline system on the NIST Chinese-English test sets. Introduction Discriminative model (Och and Ney, 2002) can easily incorporate non-independent and overlapping features, and has been dominating the research field of statistical machine translation (SMT) in the last decade. Recent work have shown that SMT benefits a lot from exploiting large amount of features (Liang et al., 2006;Tillmann and Zhang, 2006;Watanabe et al., 2007;Chiang et al., 2009). However, the training of the large number of features was always restricted in fairly small data sets. Some systems limit the number of training examples, while others use short sentences to maintain efficiency. Overfitting problem often comes when training many features on a small data (Watanabe et al., 2007;Chiang et al., 2009). Obviously, using much more data can alleviate such problem. Furthermore, large data also enables us to globally train millions of sparse lexical features which offer accurate clues for SMT. Despite these advantages, to the best of our knowledge, no previous discriminative training paradigms scale up to use a large amount of training data. The main obstacle comes from the complexity of packed forests or n-best lists generation which requires to search through all possible translations of each training example, which is computationally prohibitive in practice for SMT. To make normalization efficient, contrastive estimation (Smith and Eisner, 2005;Poon et al., 2009) introduce neighborhood for unsupervised log-linear model, and has presented positive results in various tasks. Motivated by these work, we use a translation forest (Section 3) which contains both "reference" derivations that potentially yield the reference translation and also neighboring "non-reference" derivations that fail to produce the reference translation. 1 However, the complexity of generating this translation forest is up to O(n 6 ), because we still need biparsing to create the reference derivations. Consequently, we propose a method to fast generate a subset of the forest. The key idea (Section 4) is to initialize a reference derivation tree with maximum score by the help of word alignment, and then traverse the tree to generate the subset forest in linear time. Besides the efficiency improvement, such a forest allows us to train the model without resort- r 1 X ⇒ ⟨X 1 bei X 2 , X 1 was X 2 ⟩ e 2 r 2 X ⇒ ⟨qiangshou bei X 1 , the gunman was X 1 ⟩ e 3 r 3 X ⇒ ⟨jingfang X 1 , X 1 by the police⟩ e 4 r 4 X ⇒ ⟨jingfang X 1 , police X 1 ⟩ e 5 r 5 X ⇒ ⟨qiangshou, the gunman⟩ e 6 r 6 X ⇒ ⟨jibi, shot dead⟩ Figure 1: A translation forest which is the running example throughout this paper. The reference translation is "the gunman was killed by the police". (1) Solid hyperedges denote a "reference" derivation tree t 1 which exactly yields the reference translation. (2) Replacing e 3 in t 1 with e 4 results a competing non-reference derivation t 2 , which fails to swap the order of X 3,4 . (3) Removing e 1 and e 5 in t 1 and adding e 2 leads to another reference derivation t 3 . Generally, this is done by deleting a node X 0,1 . ing to constructing the oracle reference (Liang et al., 2006;Watanabe et al., 2007;Chiang et al., 2009), which is non-trivial for SMT and needs to be determined experimentally. Given such forests, we globally learn a log-linear model using stochastic gradient descend (Section 5). Overall, both the generation of forests and the training algorithm are scalable, enabling us to train millions of features on large-scale data. To show the effect of our framework, we globally train millions of word level context features motivated by word sense disambiguation (Chan et al., 2007) together with the features used in traditional SMT system (Section 6). Training on 519K sentence pairs in 0.03 seconds per sentence, we achieve significantly improvement over the traditional pipeline by 0.84 BLEU. Synchronous Context Free Grammar We work on synchronous context free grammar (SCFG) (Chiang, 2007) based translation. The elementary structures in an SCFG are rewrite rules of the form: X ⇒ ⟨γ, α⟩ where γ and α are strings of terminals and nonterminals. We call γ and α as the source side and the target side of rule respectively. Here a rule means a phrase translation (Koehn et al., 2003) or a translation pair that contains nonterminals. We call a sequence of translation steps as a derivation. In context of SCFG, a derivation is a se-quence of SCFG rules {r i }. Translation forest (Mi et al., 2008;) is a compact representation of all the derivations for a given sentence under an SCFG (see Figure 1). A tree t in the forest corresponds to a derivation. In our paper, tree means the same as derivation. More formally, a forest is a pair ⟨V, E⟩, where V is the set of nodes, E is the set of hyperedge. For a given source sentence f = f n 1 , Each node v ∈ V is in the form X i,j , which denotes the recognition of nonterminal X spanning the substring from the i through j (that is f i+1 ...f j ). Each hyperedge e ∈ E connects a set of antecedent to a single consequent node and corresponds to an SCFG rule r(e). Our Translation Forest We use a translation forest that contains both "reference" derivations that potentially yield the reference translation and also some neighboring "nonreference" derivations that fail to produce the reference translation. Therefore, our forest only represents some of the derivations for a sentence given an SCFG rule table. The motivation of using such a forest is efficiency. However, since this space contains both "good" and "bad" translations, it still provides evidences for discriminative training. First see the example in Figure 1. The derivation tree t 1 represented by solid hyperedges is a reference derivation. We can construct a non-reference derivation by making small change to t 1 . By replacing the e 3 of t 1 with e 4 , we obtain a non-reference deriva-881 tion tree t 2 . Considering the rules in each derivation, the difference between t 1 and t 2 lies in r 3 and r 4 . Although r 3 has a same source side with r 4 , it produces a different translation. While r 3 provides a swapping translation, r 4 generates a monotone translation. Thus, the derivation t 2 fails to move the subject "police" to the behind of verb "shot dead", resulting a wrong translation "the gunman was police shot dead". Given such derivations, we hope that the discriminative model is capable to explain why should use a reordering rule in this context. Generally, our forest contains all the reference derivations RT for a sentence given a rule table, and some neighboring non-reference derivations N T , which can be defined from RT . More formally, we call two hyperedges e 1 and e 2 are competing hyperedges, if their corresponding rules r(e 1 ) = ⟨γ 1 , α 1 ⟩ and r(e 2 ) = ⟨γ 2 , α 2 ⟩ : This means they give different translations for a same source side. We use C(e) to represent the set of competing hyperedges of e. Two derivations t 1 = ⟨V 1 , E 1 ⟩ and t 2 = ⟨V 2 , E 2 ⟩ are competing derivations if there exists e 1 ∈ E 1 and e 2 ∈ E 2 : 2 In other words, derivations t 1 and t 2 only differ in e 1 and e 2 , and these two hyperedges are competing hyperedges. We use C(t) to represent the set of competing derivations of tree t, and C(t,e) to represent the set of competing derivations of t if the competition occurs in hyperedge e in t. Given a rule table, the set of reference derivations RT for a sentence is determined. Then, the set of non-reference derivations N T can be defined from RT : Overall, our forest is the compact representation of RT and N T . for v ∈ t in post order do 4: e ← incoming edge of v 5: append C(t, e) to list; 6: for u ∈ child(v) from left to right do 7: if t n ̸ = t then 9: append t n to list 10: return t,list Fast Generation It is still slow to calculate the entire forest defined in Section 3, therefore we use a greedy decoding for fast generating a subset of the forest. Starting form a reference derivation, we try to slightly change the derivation into a new reference derivation. During this process, we collect the competing derivations of reference derivations. We describe the details of local operators for changing a derivation in section 4.1, and then introduce the creation of initial reference derivation with max score in Section 4.2. For example, given derivation t 1 , we delete the node X 0,1 and the related hyperedge e 1 and e 5 . Fixing the other nodes and edges, we try to add a new edge e 2 to create a new reference translation. In this case, if rule r 2 really exists in our rule table, we get a new reference derivation t 3 . After constructing t 3 , we first collect the new tree and C(t 3 , e 2 ). Then, we will move to t 3 , if the score of t 3 is higher than t 2 . Notably, if r 2 does not exist in the rule table, we fail to create a new reference derivation. In such case, we keep the origin derivation unchanged. Algorithm 1 shows the process of generation. 3 The input is a reference derivation t, and the output is a new derivation and the generated derivations. Figure 1. Although here only shows the nodes, we also need to change relative edges actually. (1) Applying lexicalize operator on the non-terminal node X 0,1 in (a) results a new derivation shown in (b). (2) When visiting bei in (b), the generalize operator changes the derivation into (c). The list used for storing forest is initialized with the input tree (line 2). We visit the nodes in t in postorder (line 3). For each node v, we first append the competing derivations C(t,e) to list, where e is incoming edge of v (lines 4-5). Then, we apply operators on the child nodes of v from left to right (lines 6-13). The operators returns a reference derivation t n (line 7). If it is new (line 8), we collect both the t n (line 9), and also the competing derivations C(t n , e ′ ) of the new derivation on those edges e ′ which only occur in the new derivation (lines 10-11). Finally, if the new derivation has a larger score, we will replace the origin derivation with new one (lines 12-13). Although there is a two-level loop for visiting nodes (line 3 and 6), each node is visited only one time in the inner loops. Thus, the complexity is linear with the number of nodes #node. Considering that the number of source word (also leaf node here) is less than the total number of nodes and is more than ⌈(#node + 1)/2⌉, the time complexity of the process is also linear with the number of source word. Lexicalize and Generalize The function OPERATE in Algorithm 1 uses two operators to change a node: lexicalize and generalize. Figure 2 shows the effects of the two operators. The lexicalize operator works on nonterminal nodes. It moves away a nonterminal node and attaches the children of current node to its parent. In Figure 2(b), the node X 0,1 is deleted, requiring a more lexicalized rule to be applied to the parent node X 0,4 (one more terminal in the source side). We constrain the lexicalize operator to apply on pre-terminal nodes whose children are all terminal nodes. In contrast, the generalize operator works on terminal nodes and inserts a nonterminal node between current node and its parent node. This operator generalizes over the continuous terminal sibling nodes left to the current node (including the current node). Generalizing the node bei in Figure 2(b) results Figure 2(c). A new node X 0,2 is inserted as the parent of node qiangshou and node bei. Notably, there are two steps when apply an operator. Suppose we want to lexicalize the node X 0,1 in t 1 of Figure 1, we first delete the node X 0,1 and related edge e 1 and e 5 , then we try to add the new edge e 2 . Since rule table is fixed, the second step is a process of decoding. Therefore, sometimes we may fail to create a new reference derivation (like r 2 may not exist in the rule table). In such case, we keep the origin derivation unchanged. The changes made by the two operators are local. Considering the change of rules, the lexicalize operator deletes two rules and adds one new rule, while the generalize operator deletes one rule and adds two new rules. Such local changes provide us with a way to incrementally calculate the scores of new derivations. We use this method motivated by Gibbs Sampler ) which has been used for efficiently learning rules. The different lies in that we use the operator for decoding where the rule table is fixing. Initialize a Reference Derivation The generation starts from an initial reference derivation with max score. This requires bi-parsing (Dyer, 2010) over the source sentence f and the reference translation e. In practice, we may face three problems. First is efficiency problem. Exhaustive search over the space under SCFG requires O(\|f \| 3 \|e\| 3 ). 883 To parse quickly, we only visit the tight consistent (Zhang et al., 2008) bi-spans with the help of word alignment a. Only visiting tight consistent spans greatly speeds up bi-parsing. Besides efficiency, adoption of this constraint receives support from the fact that heuristic SCFG rule extraction only extracts tight consistent initial phrases (Chiang, 2007). Second is degenerate problem. If we only use the features as traditional SCFG systems, the biparsing may end with a derivation consists of some giant rules or rules with rare source/target sides, which is called degenerate solution (DeNero et al., 2006). That is because the translation rules with rare source/target sides always receive a very high translation probability. We add a prior score log(#rule) for each rule, where #rule is the number of occurrence of a rule, to reward frequent reusable rules and derivations with more rules. Finally, we may fail to create reference derivations due to the limitation in rule extraction. We create minimum trees for (f , e, a) using shift-reduce (Zhang et al., 2008). Some minimum rules in the trees may be illegal according to the definition of Chiang (2007). We also add these rules to the rule table, so as to make sure every sentence is reachable given the rule table. A source sentence is reachable given a rule table if reference derivations exists. We refer these rules as added rules. However, this may introduce rules with more than two variables and increase the complexity of bi-parsing. To tackle this problem, we initialize the chart with minimum parallel tree from the Zhang et al. (2008) algorithm, ensuring that the bi-parsing has at least one path to create a reference derivation. Then we only need to consider the traditional rules during bi-parsing. Training We use the forest to train a log-linear model with a latent variable as describe in . The probability p(e\|f ) is the sum over all possible derivations: where △(e, f ) is the set of all possible derivations that translate f into e and t is one such derivation. 4 4 Although the derivation is typically represent as d, we denotes it by t since our paper use tree to represent derivation. (∆L(w i , t n ), t n ) ←GENERATE(t n ) 11: This model defines the conditional probability of a derivation t and the corresponding translation e given a source sentence f as: where the partition function is The partition function is approximated by our forest, which is labeled asZ(f ), and the derivations that produce reference translation is approximated by reference derivations inZ(f ). We estimate the parameters in log-linear model using maximum a posteriori (MAP) estimator. It maximizes the likelihood of the bilingual corpus S = {f n , e n } N n=1 , penalized using a gaussian prior (L2 norm) with the probability density function p 0 (λ i ) ∝ exp(−λ 2 i /2σ 2 ). We set σ 2 to 1.0 in our experiments. This results in the following gradient: We use an online learning algorithm to train the parameters. We implement stochastic gradient descent (SGD) recommended by Bottou. 5 The dynamic learning rate we use is N (i+i 0 ) , where N is the number of training example, i is the training iteration, and i 0 is a constant number used to get a initial learning rate, which is determined by calibration. Algorithm 2 shows the entire process. We first create an initial reference derivation for every training examples using bi-parsing (lines 4-5), and then online learn the parameters using SGD (lines 6-12). We use the GENERATE function to calculate the gradient. In practice, instead of storing all the derivations in a list, we traverse the tree twice. The first time is calculating the partition function, and the second time calculates the gradient normalized by partition function. During training, we also change the derivations (line 10). When training is finished after M epochs, the algorithm returns an averaged weight vector (Collins, 2002) to avoid overfitting (line 13). We use a development set to select total epoch m, which is set as M = 5 in our experiments. Experiments Our method is able to train a large number of features on large data. We use a set of word context features motivated by word sense disambiguation (Chan et al., 2007) to test scalability. A word level context feature is a triple (f, e, f +1 ), which counts the number of time that f is aligned to e and f +1 occurs to the right of f . Triple (f, e, f −1 ) is similar except that f −1 locates to the left of f . We retain word alignment information in the extracted rules to exploit such features. To demonstrate the importance of scaling up the size of training data and the effect of our method, we compare three types of training configurations which differ in the size of features and data. MERT. We use MERT (Och, 2003) to training 8 features on a small data. The 8 features is the same as Chiang (2007) including 4 rule scores (direct and reverse translation scores; direct and reverse lexical translation scores); 1 target side language model score; 3 penalties for word counts, extracted rules and glue rule. Actually, traditional pipeline often uses such configuration. Perceptron. We also learn thousands of context word features together with the 8 traditional features on a small data using perceptron. Following (Chiang et al., 2009) coder to generate n-best lists for training. The complexity of CKY decoding limits the training data into a small size. We fix the 8 traditional feature weights as MERT to get a comparable results as MERT. Our Method. Finally, we use our method to train millions of features on large data. The use of large data promises us to use full vocabulary of training data for the context word features, which results millions of fully lexicalized context features. During decoding, when a context feature does not exit, we simply ignore it. The weights of 8 traditional features are fixed the same as MERT also. We fix these weights because the translation feature weights fluctuate intensely during online learning. The main reason may come from the degeneration solution mentioned in Section 4.2, where rare rules with very high translation probability are selected as the reference derivations. Another reason could be the fact that translation features are dense intensify the fluctuation. We leave learning without fixing the 8 feature weights to future work. Data We focus on the Chinese-to-English translation task in this paper. The bilingual corpus we use contains 519, 359 sentence pairs, with an average length of 16.5 in source side and 20.3 in target side, where 186, 810 sentence pairs (36%) are reachable (without added rules in Section 4.2). The monolingual data includes the Xinhua portion of the GIGAWORD corpus, which contains 238M English words. We use the NIST evaluation sets of 2002 (MT02) as our development set, and sets of MT03/MT04/MT05 as test sets. Table 2 shows the statistics of all bilingual corpus. We use GIZA++ (Och and Ney, 2003) Table 2: Effect of our method comparing with MERT and perceptron in terms of BLEU. We also compare our fast generation method with different data (only reachable or full data). #Data is the size of data for training the feature weights. * means significantly (Koehn, 2004) better than MERT (p < 0.01). word alignment in both directions, and grow-diagfinal-and (Koehn et al., 2003) to generate symmetric word alignment. We extract SCFG rules as described in Chiang (2007) and also added rules (Section 4.2). Our algorithm runs on the entire training data, which requires to load all the rules into the memory. To fit within memory, we cut off those composed rules which only happen once in the training data. Here a composed rule is a rule that can be produced by any other extracted rules. A 4-grams language model is trained by the SRILM toolkit (Stolcke, 2002). Case-insensitive NIST BLEU4 (Papineni et al., 2002) is used to measure translation performance. The training data comes from a subset of the LDC data including LDC2002E18, LDC2003E07, LDC2003E14, Hansards portion of LDC2004T07, LDC2004T08 and LDC2005T06. Since the rule table of the entire data is too large to be loaded to the memory (even drop one-count rules), we remove many sentence pairs to create a much smaller data yet having a comparable performance with the entire data. The intuition lies in that if most of the source words of a sentence need to be translated by the added rules, then the word alignment may be highly crossed and the sentence may be useless. We create minimum rules from a sentence pair, and count the number of source words in those minimum rules that are added rules. For example, suppose the result minimum rules of a sentence contain r 3 which is an added rule, then we count 1 time for the sentence. If the number of such source word is more than 10% of the total number, we will drop the sentence pair. We compare the performances of MERT setting on three bilingual data: the entire data that contains 42.3M Chinese and 48.2M English words; 519K data that contains 8.6M Chinese and 10.6M English words; FBIS (LDC2003E14) parts that contains 6.9M Chinese and 9.1M English words. They produce 33.11/32.32/30.47 BLEU tested on MT05 respectively. The performance of 519K data is comparable with that of entire data, and much higher than that of FBIS data. Table 3 shows the performance of the three different training configurations. The training of MERT and perceptron run on MT02. For our method, we compare two different training sets: one is trained on all 519K sentence pairs, the other only uses 186K reachable sentences. Although the perceptron system exploits 2.4K features, it fails to produce stable improvements over MERT. The reason may come from overfitting, since the training data for perceptron contains only 878 sentences. However, when use our method to learn the word context feature on the 519K data, we significantly improve the performance by 0.84 points on the entire test sets (ALL). The improvements range from 0.60 to 1.16 points on MT03-05. Because we use the full vocabulary, the number of features increased into 13.9 millions, which is impractical to be trained on the small development set. These results confirm the necessity of exploiting more features and learning the parameters on large data. Meanwhile, such results also demonstrate that we can benefits from the forest generated by our fast method instead of traditional CKY algorithm. Result Not surprisingly, the improvements are smaller when only use 186K reachable sentences. Sometimes we even fail to gain significant improvement. This verifies our motivation to guarantee all sentence 886 are reachable, so as to use all training data. Speed How about the speed of our framework? Our method learns in 32 mlliseconds/sentence. Figure 3 shows training times (including forest generation and SGD training) versus sentence length. The plot confirms that our training algorithm scales linearly. If we use n-best lists which generated by CKY decoder as MERT, it takes about 3105 milliseconds/sentence for producing 100-best lists. Our method accelerates the speed about 97 times (even though we search twice to calculate the gradient). This shows the efficiency of our method. The procedure of training includes two steps. (1) Bi-parsing to initialize a reference derivation with max score. (2) Training procedure which generates a set of derivations to calculate the gradient and update parameters. Step (1) only runs once. The average time of processing a sentence for each step is about 9.5 milliseconds and 30.2 milliseconds respectively. For simplicity we do not compress the generated derivations into forests, therefore the size of resulting derivations is fairly small, which is about 265.8 for each sentence on average, where 6.1 of them are reference derivations. Furthermore, we use lexicalize operator more often than generalize operator (the ration between them is 1.5 to 1). Lexicalize operator is used more frequently mainly dues to that the reference derivations are initialized with reusable (thus small) rules. Related Work Minimum error rate training (Och, 2003) is perhaps the most popular discriminative training for SMT. However, it fails to scale to large number of features. Researchers have propose many learning algorithms to train many features: perceptron (Shen et al., 2004;Liang et al., 2006), minimum risk (Smith and Eisner, 2006;, MIRA (Watanabe et al., 2007;Chiang et al., 2009), gradient descent . The complexity of n-best lists or packed forests generation hamper these algorithms to scale to a large amount of data. For efficiency, we only use neighboring derivations for training. Such motivation is same as contrastive estimation (Smith and Eisner, 2005;Poon et al., 2009). The difference lies in that the previous work actually care about their latent variables (pos tags, segmentation, dependency trees, etc), while we are only interested in their marginal distribution. Furthermore, we focus on how to fast generate translation forest for training. The local operators lexicalize/generalize are use for greedy decoding. The idea is related to "pegging" algorithm (Brown et al., 1993) and greedy decoding (Germann et al., 2001). Such types of local operators are also used in Gibbs sampler for synchronous grammar induction . apply our forest on other learning algorithms. Finally, we hope to exploit more features such as reordering features and syntactic features so as to further improve the performance.	2014-07-01T00:00:00.000Z	2011-07-27T00:00:00.000	{ "year": 2011, "sha1": "2887e349875dfc30a6e7666424d83abc40c7fd53", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "ACL", "pdf_hash": "4bec897fb3bea4c5e29b692bd0894a6a91919772", "s2fieldsofstudy": [ "Computer Science" ], "extfieldsofstudy": [ "Computer Science" ] }
13356089	pes2o/s2orc	v3-fos-license	Chemical Composition profile of Acacia nilotica Seed Growing Wild in South of Iran Acacia Nilotica is a pioneer species, relatively high in bioactive secondary compound and are important for a variety of functions is economically used as a source of tannins, gums, timber, fuel and fodder. Babul plant is therapeutic used as Anti-cancer, anti tumours, Antiscorbutic, Astringent, antioxidant, Natriuretic, Antispasmodial, Diuretic, Intestinal pains and diarrhea, Nerve stimulant, Cold, Congestion, Coughs, Dysenter, Fever, Hemorrhages, Leucorrhea, Ophthalmia and Sclerosis. The aim of this study was determination of proximate composition, mineral elements (Calcium, Potassium, Iron, Zinc, Sodium, Selenium and Copper) contents in this endemic Iranian seed. A. Nilotica seeds (mature, dry) were collected in August 2014 from Sarkhun village, Bandar Abbas, Hormozgan Province, Iran. Specimen was identified by R. Asadpour and voucher was deposited in the Herbarium of Faculty of Pharmacy, Pharmaceutical Sciences Branch, Islamic Azad University (IAUPS). The samples were analyzed by wet digestion method and analysis of mineral element contents analyzed by Atomic Absorption Spectrophotometer in Research Laboratory in Pharmaceutical Sciences Branch, Islamic Azad University. Obviously in A. Nilotica the contents of the some trace and essential mineral element are high in comparison by other seeds studied in other countries. The value of potassium, iron and Zinc, Copper and Manganese in A. Nilotica were 2.1, 203.1, 108.7, 322.7 and 1.09 (g/100g DW) based on dry weight respectively. The oil, crude protein and crude fiber contents in this edible seed were found to be 4.1, 25.3 and 28.4 % based on the fresh weight respectively. In this study, the nutritive value of A. Nilotica mature seed native to south of Iran was determined and results revealed that it is so rich in mineral elements especially Iron and Potassium. Traditionally the plant used widely for the treatment of various ailments, but scientifically few of them was screened out. Therefore the scientific studies should be conducted to investigate the unexploited potential of Acacia Nilotica (L.).Our results revealed that it could be recommended as a dietary supplement for people who need essential mineral elements. tropical tree with rapid growth and ten year period of yield which can grow extensively in the seashores of southern Iran 3 .A. Nilotica is a multipurpose tree with extensive distribution from Egypt to Mauritania and South Africa in Africa and from East Asia to India, Pakistan and Iran in Asia. A. Nilotica is a pioneer species, relatively high in bioactive secondary compound and is important for a variety of functions is economically used as a source of tannins, gums, timber, fuel and fodder [4][5][6] .It grows in semi-arid, hot and wet regions such as the Persian Gulf, Oman Sea and in Boushehr Province, Hormozgan Province and Sistan and Baluchestan Province (Chahbehar, Iranshahr and Nikshahre) as well as in deep loam soils.A. Nilotica plantation was started from 1984 in Chahbehar and Dashteyari 3 .A. Nilotica is a pantropical and subtropical genus with species abundant throughout Asia, Australia, Africa and America.A.nilotica occurs naturally and is imperative in traditional rural and agro-pastoral systems 7 .A. nilotica is recognized by the following names: Acacia, Acacia Arabica, Babhul -Hindi and Napalese, Babla -Bengali, Babool -Unani, Babool Baum -German, Babhoola -Sanskrit, Babul, Babul Tree, Huanlong Kyain -Burmese, Kikar, Mughilan -Arabian Indogom -Japenese and Ummughiion -Persian 8 .A. Nilotica is an imperative multipurpose plant that has been used broadly for the treatment of various diseases 9 .The plant is therapeutic used as Anti-cancer, anti tumours, Antiscorbutic, Astringent, anti-oxidant, Natriuretic, Antispasmodial, Diuretic, Intestinal pains and diarrhea, Nerve stimulant, Cold, Congestion, Coughs, Dysenter, Fever, Hemorrhages, Leucorrhea, Ophthalmia and Sclerosis 10 .In other studies it has been reported that Acacia species contains secondary metabolites including amines and alkaloids, cyanogenic glycosides, cyclitols, fatty acids and seed oils, fluoroacetate, gums, nonprotein amino acids, terpenes (including essential oils, diterpenes, phytosterol and triterpene genins and saponins), hydrolysable tannins, flavonoids and condensed tannins 11,12 .The plant is richer source of cysteine, methionine, threonine, lysine, tryptophan, Potassium, phosphorus, magnesium, iron and manganese 9 .As evident from literature, there was no phytochemical investigation on A. Nilotica seed growing wild in south of Iran.Hence the current study includes analyses of mineral and essential elements in A. Nilotica seed oil to evaluate its nutritive potential value as a new source of enriched seed for nutritional purposes. Plant material A. nilotica seeds were collected in August 2014 from Sarkhun village, Bandar Abbas, Hormozgan Province, Iran: (27°23'34" N 56°23'59" E, 100m).Specimen was identified by R. Asadpour and voucher was deposited in the Herbarium of Faculty of Pharmacy, Pharmaceutical Sciences Branch, Islamic Azad University (IAUPS) Tehran.The area is mountainous region that located among plains and hills.The region's geographical is located in the north of Bandar Abbas( figure 1).Plain part of the region includes much of the southern, eastern and northern part of the strip consisted of alkaline and saline soils contain large amounts of soluble salts such as chloride, sulfate and carbonate of Ca, Mg, sodium, and potassium 13,14 . Moisture Content All grounded A. nilotica mature dry seeds samples were oven dried at 60 o C for 36 hours until a constant weight were obtained.The moisture contents were expressed as loss in weights of the wet samples [15][16][17][18] . Crude Fiber Five grams of the grounded A. nilotica mature dry seeds samples were digested in 50 ml of 1.25% H 2 SO 4 .The solutions were boiled for 45 minutes and then were filtered and washed with hot distilled water.The filtrates were digested in 50 ml of 1.25% Sodium Hydroxide solutions.For 50 minutes these solutions were heated, filtered and washed with hot deionized water and over dried.The final oven-dried residues were ignited in a furnace at 550 o C. The weights of the left after ignition were measured as the fiber contents and were expressed in term of the weights of the samples before ignition 19 . Crude Protein The protein nitrogen in one gram of the dried samples were converted to ammonium sulphate by digestion with concentrated H 2 SO 4 (Merck 96.5%) and in the presence of CuSO 4 and K 2 SO 4 [20][21] .The solutions were heated and the ammonia evolved were steam distilled into Boric acid 2%.The nitrogen from ammonia were deduced from the titrations of the trapped ammonia with 0.1M HCl with Tashirus indicator (methyl red: methylene blue 2:1) until a purplish pink color were obtained.Crude proteins were calculated by multiplying the valve of the deduced nitrogen by the factor 6.25 mg [22][23][24] . Ash Content One gram of the oven-dried samples in powder from was placed in acid washed crucible by known weight.They were ignited in a muffle furnace for 5 hours at 550 o C.After cooling crucibles they were weighed and the ash contents were expressed in terms of the oven-dried weight of the sample 19 . Zinc, Manganese, Copper and Potassium Determination For Zinc, Manganese, Copper and Selenium concentration in A. Nilotica , powered seed samples were dried in oven for 48 hours at a temperature of 85°C.The samples were then ground and sieved through 0.5 mm sieve.The powdered samples then subjected to the acid digestion using concentrated nitric acid (65% Merck), Sulfuric acid (96.5% Merck) and per chloric acid (70% sigma).Analar grade hydrogen peroxide (about 30%) also was used for the digestion.Application of concentrated HNO 3 along with thirty percent hydrogen peroxide H 2 O 2 (Merck) for mineralization of samples to the complete digestion of samples 25- 27 following Environmental Protection Agency (EPA) Method 3052 was done. Two gram of air-dried of each homogeneously A. Nilotica samples accurately weighed and 30.0 mL of the digestion mixture (3 parts by weight of nitric acid: 1 parts of Sulfuric acid & 3 parts by weight perchloric acid) and heated slowly by an oven and then rise the temperature.The remaining dry inorganic residues were dissolved in 30.0 mL of concentrated nitric acid and the solution used for the determination of trace and essential mineral elements.Blanks and samples were also processed and analyzed simultaneously.All the chemicals used were of analytical grade (AR).Standardized international protocols were followed for the preparation of material and analysis of heavy metals contents 28- 32 .The samples were analyzed by Flame Emission Spectrophotometer Model AA-6200 (Shimadzu, Japan) using an air-acetylene, flame temperature: 2800°C, acetylene pressure: 0.9-1.0bar, air pressure: 4.5-5 bar, reading time: 1-10 sec (max 60 sec), flow time: 3-4 sec (max 10 sec), using at least five standard solutions for each metal and determination of potassium content was followed by FDA Elemental analysis 33 In order to verify of reliability of the measuring apparatus, periodic testing of standard solutions was performed .The accuracy was checked using quality control test for fungi and their substrate samples to show the degree of agreement between the standard values and measured values; the difference was less than 5%. Iron Determination The aliquot was passed through the atomic absorption spectrophotometer to read the iron concentration.Standards were prepared with a standard stock of 10 mg/L using ferrous ammonium sulphate where 3 -60 ml of iron standard solution (10 mg /L) were placed in stepwise volumes in 100 ml volumetric flasks. 2 ml of hydrochloric acid were added and then brought to the volume with distilled water.The concentration of iron in the aliquot was measured using the atomic absorption spectrophotometer in mg/L.The whole procedure was replicated three times 34,35 . Calcium, Sodium and Magnesium Determination 5 ml of the aliquot were placed in a titration flask using a pipette and diluted to 100 ml with distilled water and subsequently 15 ml of buffer solution, ten drops of Eriochrome black T indicator and 2 ml of triethanolamine were added.The mixture was titrated with Ethylene-Diamine-Tetra-Acetate (EDTA) solution from red to clear blue 36 . Selenium Determination Stock standard solutions for selenium were 1000 g /mL solution.All reagents and standards were of analytical grade ( Merck, Germany) .The palladium matrix modifier solution was prepared by the dilution (10 g/ L) Pd(NO 3 ) 2 and iridium AA standard solution, 1000 g/ mL in 20% HCl , 0.1 % V/ V nitric acid prepared by dilution trace pure 65 % nitric acid and 0.1 % Triton X-100 were used.Doubly distilled water was used in all operations.The samples were analyzed by Flame Emission Spectrophotometer Model AA-6200 (Shimadzu, Japan).The analyze performed according by Analytical Method ATSRD 30,35,36 . RESULTS The mean content of trace and essential mineral elements (g/100g DW) in the mature dry seed of Acacia Nilotica samples is shown in table 1.The samples were analyzed by wet digestion method and standardized international protocols were followed for the preparation of material and analysis of mineral contents and analyzed by Atomic Absorption Spectrophotometer in Research Laboratory in Pharmaceutical Sciences Branch, Islamic Azad University. The order depending on the contents of trace metal and essential elements (g/ 100 g) in A. Nilotica samples in Hormozgan-Iran studied regions was: DISCUSSION Obviously in A. nilotica the amount of the some mineral element contents such as copper, iron, zinc and calcium are high in comparison by other seeds studied in other countries. Copper has the role of assisting in the formation of haemoglobin, helping to prevent anemia as well as being involved in several enzymes.Iron is the central metal in the haemoglobin molecule for oxygen transport in the blood and is portion of myoglobin located in muscles.Manganese is one of the co-factors in a number of enzymes as is molybdenum.Selenium has several roles such as regulating the thyroid hormone as well as being part of an enzyme that protects against oxidation 37 , Selenium has also been reported as assisting in deactivating heavy metals.Calcium is responsible for strong bones and teeth and accounts for ninety percent of the calcium in the body whereas the other one percent is circulating in fluids in order to ionize calcium.The metal's function is related to transmitting nerve impulses; contractions of muscles; blood clotting; activation of some enzyme reactions and secretion of hormones Magnesium has many roles including supporting the functioning of the immune system; assists in preventing dental decay by retaining the calcium in tooth enamel; it has an important role in the synthesis of proteins, fat, nucleic acids; glucose metabolism as well as membrane transport system of cells.Magnesium also plays a role in muscle contraction and cell integrity.Potassium and sodium work together in muscle contraction nerve transmission.Sodium is important in muscle contraction and nerve transmission Sodium ions are the main regulators of extra cellular fluid and volume (37) .Zinc is an essential trace element and plays an important role in various cell processes including normal growth, brain development, behavioral response, bone formation and wound healing.Zinc deficient diabetics fail to improve their power of sensitivity and it cause loss of sense of touch and smell 14,38,39 .This is obviously seen in the crude protein content of the samples of Iranian Acacia nilotica examined in this study which reached 25.3 and this value is significantly higher and superior than other Acacia nilotica reported in other countries especially in Africa samples.Rubanza et al., in 2005 stated that the Acacia nilotica seeds had protein (19%) and contain more fibre 40 and Abdalla et al in 2014 in Sudan reported 21.4% crude protein and the crude fiber content of Acacia nilotica fruit in this study is 30.12%,fat (ether extract) is 24.77%,DM is 93.71% and the ash content is 11.76% in A. Nilotica grown in Sudan 41 .Our results show that Acacia nilotica seeds from Hormozgan :south province in Iran have more crude protein and ash and less crud fiber and fat.The protein content for this seed is high and it could be used as dietary supplement for people who need a lot of protein and most importantly for those who require plant protein especially people suffering from hypertension.They can also be incorporated into animal feed to increase the protein content.The Recommended Dietary Allowance (RDA) for protein is 0.8 g/kg body weight for adults, set by the Institute of Medicine, and is based on the consumption of good-quality protein (U.S. Department of Health and Human Services 2006). According to the most recent statistics from the American Cancer Society, more than 1.5 million new cancer cases are diagnosed annually (American Cancer Society 2010). The "ash content" is a measure of the total amount of minerals present within a food, whereas the "mineral content" is a measure of the amount of specific inorganic components present within a food, such as Ca, Na and K. Determination of the ash and mineral content of foods is important for a number of reasons such as nutritional labeling: The concentration and type of minerals present must often be stipulated on the label of a food and quality: The quality of many foods depends on the concentration and type of minerals they contain, including their taste, appearance, texture and stability 19 . CONCLUSION In this study, the nutritive value of A. Nilotica mature seed native to south of Iran was determined and results revealed that it is so rich in crude protein and some trace and essential mineral elements especially Iron, Zinc, Copper and Potassium.Traditionally the plant used widely for the treatment of various ailments, but scientifically few of them was screened out.Therefore the scientific studies should be conducted to investigate the unexploited potential of Acacia Nilotica (L.).Our results revealed that it could be recommended as a dietary supplement for people who need essential mineral elements.In conclusion the present study revealed that the seed oil of A. Nilotica growing in south of Iran could be a new source of high protein and mineral elements and its full potential should be exploited.The use of this seed is of potential economic benefit to the poor native population of the areas where it is cultivated.Hence the seed protein and minerals of A. Nilotica could be a new source of edible vegetable after the future toxicological studies. Table 1 : The Mean content (g/100g DW) composition of the mature dry seeds of A. Nilotica from Hormozgan Province, Iran *SD = Standard Deviation	2018-03-28T17:12:05.510Z	2015-06-20T00:00:00.000	{ "year": 2015, "sha1": "0e1a0e7a0f0e19a155d9d83b42f44da95106e128", "oa_license": "CCBY", "oa_url": "https://doi.org/10.13005/ojc/310251", "oa_status": "HYBRID", "pdf_src": "ScienceParseMerged", "pdf_hash": "0e1a0e7a0f0e19a155d9d83b42f44da95106e128", "s2fieldsofstudy": [ "Chemistry", "Environmental Science" ], "extfieldsofstudy": [ "Chemistry" ] }
55584706	pes2o/s2orc	v3-fos-license	Simulations of Early Baryonic Structure Formation with Stream Velocity: II. The Gas Fraction Understanding the gas content of high redshift halos is crucial for studying the formation of the first generation of galaxies and reionization. Recently, Tseliakhovich&Hirata showed that the relative"stream"velocity between the dark matter and baryons at the time of recombination - formally a second order effect, but an unusually large one - can influence the later structure formation history of the Universe. We quantify the effect of the stream velocity on the so-called"characteristic mass"- the minimum mass of a dark matter halo capable of retaining most of its baryons throughout its formation epoch - using three different high-resolution sets of cosmological simulations (with separate transfer functions for baryons and dark matter) that vary in box size, particle number, and the value of the relative velocity between the dark matter and baryons. In order to understand this effect theoretically, we generalize the linear theory filtering mass to properly account for the difference between the dark matter and baryonic density fluctuation evolution induced by the stream velocity. We show that the new filtering mass provides an accurate estimate for the characteristic mass, while other theoretical ansatzes for the characteristic mass are substantially less precise. INTRODUCTION Gas rich dark matter halos in the early universe serve as a nurturing ground for dwarf galaxies (e.g., Ricotti et al. 2002b,a;Bromm et al. 2002Bromm et al. , 1999Abel et al. 2002;Naoz et al. 2006;Yoshida et al. 2006Yoshida et al. , 2008Greif et al. 2010;Clark et al. 2011;Bromm & Yoshida 2011, and references therein). Their properties are important to quantify, as they are responsible for metal pollution and ionizing radiation at the onset of structure formation (e.g., Shapiro et al. 2004;Ciardi et al. 2006;Hoeft et al. 2006;Gnedin et al. 2008;Okamoto et al. 2008;Trenti & Stiavelli 2009). More than that, even if the smallest of gas rich halos are too small for efficiently cooling via atomic hydrogen lines and may not host actual galaxies, these "mini-halos" may produce a 21-cm signature in future radio observations (Kuhlen et al. (2006); Shapiro et al. (2006); Naoz & Barkana (2008) but see Furlanetto & Oh (2006)) and might block some of the ionizing radiation, causing an overall delay in the initial progress of reionization (e.g., Barkana & Loeb 2002;Iliev et al. 2003aIliev et al. , 2005McQuinn et al. 2007). Thus, the evolution of the gas fraction of dark matter halos at various epochs during the early evolution of the universe is of prime importance. Recently, Tseliakhovich & Hirata (2010) showed that not only the amplitudes of the dark matter and bary-snaoz@cfa.harvard.edu 1 onic density fluctuations were different at early times, but also were their velocities. After recombination, the sound speed of the baryons dropped dramatically, while the dark matter velocity remained high -thus, the relative velocity of baryons with respect to the dark matter became supersonic. Tseliakhovich & Hirata (2010) also showed that this relative velocity between the baryons and the dark matter remained coherent on scales of a few mega-parsec and was of the order of ∼ 30 km sec −1 at the time of recombination. This relative velocity is often called the "stream velocity" in the literature, and throughout this paper we will use this term. The stream velocity effect has previously been overlooked, because the velocity terms are formally of the second order in the perturbation theory and should be neglected in the linear approximation. However, this second order effect is unusually large, resulting in the numerically nonnegligible suppression of power at mass scales that correspond to the first bound objects in the Universe (e.g., Yoshida et al. 2003a). Using the Press-Schechter (Press & Schechter 1974) formalism, Tseliakhovich & Hirata (2010) showed that the number density of halos is reduced by more than 60% for halos with M = 10 6 M ⊙ at z = 40. In a subsequent paper, Tseliakhovich et al. (2010) also included the baryonic temperature fluctuations following Naoz & Barkana (2005). They found that the stream velocity also resulted in much higher "characteristic" mass -the minimum mass for a dark matter halo capable of retaining most of its gas -as compared to the case without the stream velocity (e.g. Naoz & Barkana 2007). As has been shown in subsequent studies, the stream velocity effect has important implications on the first structures (Stacy et al. 2011;Maio et al. 2011;Greif et al. 2011;Naoz et al. 2012;Fialkov et al. 2011;Bovy & Dvorkin 2012) and may also affect the redshifted cosmological 21-cm signal (Dalal et al. 2010;Bittner & Loeb 2011;Yoo et al. 2011;Visbal et al. 2012;. In this paper we explore the effect of stream velocity on the gas fraction in dark matter halos and compare the simulation results to the predictions from the linear theory (e.g. Tseliakhovich et al. 2010). In our first paper (Naoz et al. 2012, hereafter Paper I) we quantified the stream velocity effect on the evolution of the halo mass function with cosmological simulations. We used three different sets of high resolution simulations in order to study the stream velocity effect systematically, thus understanding the overall trends (instead of concentrating on specific halos). We used a set of simulations with different box sizes, particle numbers, and the values for the stream velocity to analyze the suppression of the structure formation as a function of the stream velocity. In Paper I we found that the total number density of halos is suppressed by ∼ 20% at z = 25 in regions of the universe that happen to have v bc = 1σ vbc , where σ vbc is the (scale independent) rms fluctuation of the stream velocity on small scales. In rare patches where v bc = 3.4σ vbc , the relative suppression at the same redshift reaches 50%, remaining at or above the 30% level all the way to z = 11. Perhaps the most interesting phenomenon that we found was the high abundance of "empty halos", i.e., halos that had their gas fractions below half of the cosmic mean baryonic fractionf b . Specifically, we found that for v bc = 1σ vbc all halos below 10 5 M ⊙ are empty at z ≥ 19. As a result, the high abundance of empty halos can significantly delay the formation of gas rich "minihalos" and the first galaxies. In this paper we investigate the effect of the stream velocity on the gas fraction in halos. In particular, we quantify the dependence of the characteristic mass on the magnitude of the stream velocity. For completeness we first describe the parameters and initial conditions of our simulations in §2. We present our results and analysis of the gas fraction in halos and comparison to the linear approximation in Section 3. Finally we offer a brief discussion in §4. Basic Parameters and Settings In this work we use a parallel N -body/hydrodynamics code GADGET-2 (Springel et al. 2001;Springel 2005). Below we describe the general features of our 3 simulation sets, which are also summarized in table 1. 1. The first set, named "N = 256", uses a total of 2 × 256 3 dark matter and gas particles within a cubic box of 200 comoving kpc on a side. To realize statistically significant number of halos in such a small box, we artificially increase gravitational clustering in the simulation by setting σ 8 = 1.4. We choose this box size so that a 10 4 M ⊙ halo is resolved with ∼ 500 particles -the value needed to estimate the halo gas fraction reliably (Naoz et al. 2009). The gravitational softening is set to be 40 comoving pc, well below the virial radius of a 10 4 M ⊙ halo (∼ 680 comoving pc). All the simulations in this set are initialized at z = 199. 2. The second set, named "N = 512", uses a total of 2 × 512 3 dark matter and gas particles within a cubic box with the size of 700 kpc. In this set we also artificially increase σ 8 to 1.4. With these parameters, a halo with 500 dark matter particles has a mass of ∼ 5 × 10 4 M ⊙ . The softening length is set to be 68 comoving pc. All the simulations in this set are initialized at z = 199. 3. The final set of simulations uses 2 × 768 3 dark matter and gas particles (which we name the "N = 768" set) in a 2 Mpc box, and starts at z = 99. For these parameters a halo with 500 dark matter particle has a mass of ∼ 10 5 M ⊙ . The softening length is set to be 0.2 comoving kpc. We use the "correct" value of σ 8 = 0.82 for this simulation set. In each simulation set, we explore a range of the values for the stream velocity (see table 1). 2.2. Initial Conditions As has been shown by Naoz et al. (2006) and Naoz & Barkana (2007), setting up initial conditions for cosmological simulations on small spatial scales is a delicate issue. High accuracy in initial conditions is crucial for accurately predicting the halo mass function in the lowest mass regime (M 10 7 M ⊙ ). Following Naoz et al. (2011), we generate separate transfer functions for dark matter and baryons as described in Naoz & Barkana (2005). Ideally, the stream velocities should be realized in the initial conditions in a self-consistent way with the transfer functions that are calculated up to the second order in the perturbation theory (c. Instead, similar to all previous simulation studies, we used the transfer functions computed in the linear approximation. We account for that choice in §3.3, when we compare our simulations to the perturbations theory, and thus our results remain self-consistent. For all runs, glass-like initial conditions were generated using Zel'dovich approximation. For baryons, we have used a glass file with positions shifted by a random vector, thus removing artificial coupling between nearby dark matter and gas particles (Yoshida et al. 2003b). We note that we have used the same phases for dark matter and baryons in all of our simulations, since we showed in Paper I that the spatial shift between baryons and dark matter is unimportant. Tseliakhovich & Hirata (2010) demonstrated that, while the stream velocity varies in space, its coherence length is quite large, many Mpc. Hence, on scale of our simulation boxes, it can be treated as constant bulk motion of baryons with respect to the dark matter. We include the effect of stream velocity by adding, at the initial redshift, an additional velocity to the x component of the baryons velocity vector. We test a range of values for the stream velocity, which is convenient to quantify in terms of its rms value on small scales, σ vbc . Specifically, we test v bc = 1σ vbc through v bc = 3.4σ vbc for all the simulations sets (see table 1). Halo Definition We locate dark matter halos by running a friends-offriends group finder algorithm with a linking parameter of 0.2 (only for the dark matter component). We use the identified particle groups to find the center of mass of each halo. After the center is located, we calculate density profiles of dark matter and baryons separately, assuming a spherical halo and using 2000 radial bins between r min = 0 kpc and r max = 20 kpc. Using the density profiles, we find the virial radius r vir at which the total overdensity is 200 times the mean background density, and compute the mass and the gas fraction of each halo within that radius. Recently, O'Leary & McQuinn (2012) used an unconventional definition for halos by using the highest baryonic density peaks as the center their halos. This method preferentially results in larger gas fraction for σ vbc = 0, compare to our more conservative method. Recently, More et al. (2011) showed that halos identified by the friends-of-friends algorithm enclose an average overdensity that is substantially larger than 200, and its specific value depends on the halo concentration. In our approach we use the friends-of-friends algorithm only to find the center of mass of a halo, and compute the actual halo mass using the spherical overdensity of 200. We only retain halos that contain at least 500 dark matter particles within their virial radii. The choice allows us to estimate halo masses to about 15% precision (Trenti et al. 2010) and to estimate halo gas fractions reliably to a similar level of accuracy (Naoz et al. 2009). However, for some of our fit calculations we also include halos with the number of particles as low as 100; if we do that, we assign a lower weight in the fit to these halos according to the resolution study done in Naoz et al. (2009), see Appendix A for more details. RESULTS Let us consider the various scales involved in the formation of cosmic structure. On large scales gravity dominates other forces and gas pressure can be neglected. On small scales, on the other hand, the pressure dominates gravity and prevents baryon density fluctuations from growing together with the dark matter fluctuations. The relative force balance at a given time can be characterized by the "filtering scale" (Gnedin & Hui 1998) -a physical scale above which a small gas perturbation can grow due to gravity overcoming the pressure gradient. In the non-evolving background, the filtering scale coincides with the classical Jeans (1928) scale, but in the expanding universe the two scales typically differ by a significant factor. Immediately after recombination Compton scattering of Cosmic Microwave Background (CMB) photons on the residual free electrons after cosmic recombination kept the gas temperature coupled to that of the CMB, the Jeans mass was constant in time and equal to the filtering scale. However, at z ∼ 130, the gas temperature decoupled from the CMB temperature, the Jeans mass began to decrease with time as the gas cooled adiabatically, and the filtering scale lagged behind the Jeans scale. Based on results from an early numerical simulations, Gnedin (2000) suggested that the filtering mass also describes the largest halo mass whose gas content is significantly suppressed compared to the cosmic baryon fraction. The latter mass scale, commonly called the "characteristic mass", is defined as the halo mass for which the enclosed baryon fraction equals half of the cosmic mean. Thus, the characteristic mass distinguishes between gas-rich and gas-poor halos. Many semianalytical models of dwarfs galaxies use the characteristic mass scale in order to estimate the gas fraction in halos (e.g., Bullock et al. 2000;Benson et al. 2002a,b;Somerville 2002;Bovy & Dvorkin 2012). Theoretically this sets an approximate minimum value on the mass that can still form stars. Non-linear Behavior: the Characteristic Mass For halos, Gnedin (2000) defined a characteristic mass M c for which a halo contains half the mean cosmic baryon fraction f b . In his simulation he found the mean gas fraction in halos of a given total mass M , and fitted the simulation results to the following formula: where f b,0 is the gas fraction in the high-mass limit 7 . In this function, a higher α causes a sharper transition between the high-mass (constant f g ) limit and the lowmass limit (assumed to be f g ∝ M 3 ). Gnedin (2000) found a good fit for α = 1, with a characteristic mass that in fact equaled the filtering mass by his definition. The characteristic mass is essentially a non-linear version of the filtering mass, and so it also measures the competition between gravity and pressure. At high masses, where pressure is unimportant, f g → f b,0 , while the low mass tail is determined by the suppression of gas accretion by gas pressure. Naoz et al. (2009) found that 7 Following Naoz et al. (2011) we defined the high-mass bin as the largest 5% halo mass, or the largest 5 halos (if the larges 5% consists of less than 5 halos). These halos are usually gas rich and if when we miss them it causes to an underestimate of the gas fraction at the larger mass bin. Naoz et al. (2011) found this method to be consistent with setting a low linking parameter, and with varying the halo radii between r 100 , r 200 and r 500 . This way we overcome the disadvantages of assuming a spherical halo, which misses gas rich halos which undergoes mergers. , and 15 (from bottom to top panels) for 4 values of the stream velocity v bc = 0, v bc = 1σ vbc , v bc = 1.7σ vbc , and v bc = 3.4σ vbc (blue squares, purple diamonds, brown stars and red triangles, respectively). We also show the best fit as evaluated from equations (1) and (2), with dotted and solid lines respectively. Note for the case of v bc = 3.4σ vbc , at z = 25, about 2/3 of the halos have less than 1% of gas in them, and f b,0 is very low. Thus no reliable fit could be found, and therefore we did not show the resulted red line in that case. (See also Figure 7 for the gas fraction behavior for all halos, i.e., not binned.) the filtering mass from linear theory (calculated in a self consistent way) is consistent with the characteristic mass fitted from the simulations, for two (pre-reionization) scenarios that they tested: a case with no stellar heating and a case of a sudden flash of stellar heating at a given redshift. In a followup paper, Naoz et al. (2011) found the same agreement between the the linear and non-linear theory, and showed that alternative initial conditions models yield a different (higher by about 50 per cent) minimum mass (both the linear and non-linear), since the system retains a memory of the initial conditions. For clarity, we emphasize that the statement (M c = M F ) refers to our definition of M F in equation (9). In Figure 1 we present the gas fraction as a function of halo mass for our "N=512" simulation set 8 . As one can see, the halo gas fraction drops dramatically at lower halo masses for large values of the stream velocity. This trend introduces a qualitatively different behavior in the gas fraction as a function of halo mass, which is not captured by equation (1). The best fit ansatz [Eq. (1)] is shown with dotted lines in Fig. 1. It clearly does not capture the behavior of the gas fraction as a function of halo mass for v bc = 0. Therefore we introduce a new fitting formula for the gas fraction as a function of halo mass, The new fitting formula reduces to Equation (1) for γ = 3β = α. Although we add another free parameter to the fit, finding the best fit model presents some hurdles which we discuss in the Appendix. The best-fit value of M c from Equation (2) is the same as the value from Equation (1) for the v bc = 0 case, as can be expected from the fact that Equation (1) provides a good fit to the simulation results in the v bc = 0 case. It is interesting, however, that the best fit values for M c from Equation (2) are consistent with the best fit values of M c from Equation (1), even for large values of v bc , as can be seen in Figure 8. However, the new fitting formula gives a better overall fit for high v bc values, especially in the high mass limit. , v bc = 1σ vbc , v bc = 1.7σ vbc , and v bc = 3.4σ vbc (blue squares, purple diamonds, brown stars and red triangles respectively) for our N = 762 simulation set. We also show the evolution of the fully self-consistent filtering mass (Equation 7) with solid curves. The error bars are the maximum 1 − σ from the two fit models (see Appendix A). Figures 2-4 show the best-fit values for the characteristic mass and limiting baryon fraction f b,0 at a range of redshifts. At the highest redshifts and large values of σ vbc most of the halos are empty halos (i.e., halos with gas fraction lower then the half of the mean cosmic baryonic fraction, see Paper I, figures 2-4). This of course means that there is no apparent trend similar to Figure 1, and therefore, no convergence could be achieved in the fitting procedure and the parameters of the fit cannot be measured. Furthermore, as has been noted in Paper I, the N = 256 set suffers from poor statistics for M 10 5 M ⊙ , resulting in large error bars for the best-fit values of the parameters. We show the results of this run to caution the interpretations of previous (and perhaps some future) which employ very small boxes. An important point to make is that at low redshifts the baryon fraction at the highest mass bins (for all runs) approaches the same value irrespectively of the magnitude of the stream velocity. This is, of course, expected, as the global trend for the baryon fraction is to approach the cosmic mean (see Appendix A, Figure 9), but still below it at high redshift (even at large scales). This is because the baryons still did not fall into the Dark Matter potential wells (see Naoz & Barkana 2007;Barkana & Loeb 2011, and the Appendix for further discussion). However, in the case of the largest stream velocity we consider, v bc = 3.4σ vbc , the N = 256 and N = 512 simulation sets do not quite reach the cosmic mean values expected. This is most likely due to the low abundance of the most massive halos, since some baryons are in fact bound to the halo and we miss them due to our halo finder algorithm ( §2.3), as can be seen in Figure 4 of Paper I. The N = 768 simulation set, however, does not suffer from that incomplete convergence. We find a simple fit for the gas fraction in the last mass bin, f b,0 , for the low redshift limit as a function of the stream velocity Note that throughout the paper we compare between different simulations using different σ 8 values (i.e., σ 8 = 0.82 for the N = 768 set and σ 8 = 1.4 for the N = 512 and N = 256 sets). As was shown in Paper I the suppression of the halo mass function, due to the stream velocity, compare to the no stream velocity case is independent on σ 8 (see Figures 5 and 6 in Paper I). Since the characteristic mass describes a relative suppression of the gas fraction in small scales compare to large scales, we find that this quantity is independent on σ 8 as well (where the different simulations gave a consistent value of M c ). This is further supported by the agreement to linear theory (see below). Thus, increasing σ 8 only raised the clustering amplitude that enlarges the sample of simulated halos. Linear Theory Predictions: the Filtering Mass In the linear approximation, the filtering mass, first defined by Gnedin & Hui (1998), describes the highest mass scale on which the baryon density fluctuations are suppressed significantly compared to the dark matter fluctuations. Gnedin & Hui (1998) only considered the low redshift case, where the baryonic and dark matter fluctuations have the same amplitude at large scales. Naoz & Barkana (2007) relaxed that assumption and extended the computation of the filtering mass to early times, during which the amplitude of the baryonic fluctuations is below the amplitude of the dark matter fluctuations even on large scales. Both studies, however, only considered a case of zero stream velocity. In order to extend the derivation of the filtering mass to the non-zero stream velocity case, we first introduce the coupled second order differential equations that govern the evolution of the density fluctuations of the dark matter (δ dm ), and the baryons (δ b ) and the baryon temperature (δ T ): where Ω m is the present day matter density as a fraction of the critical density, k is the comoving wavenumber, a is the scale factor, µ is the mean molecular weight, H 0 is the present day value of the Hubble parameter H, andT and δ T are the mean baryon temperature and its dimensionless fluctuation, respectively. These equations are a compact form of equations 5 in Tseliakhovich & Hirata (2010), where we used the fact that v bc ∝ 1/a, and the baryon equation includes the pressure term whose form comes from the equation of state of an ideal gas. The linear evolution of the temperature fluctuations is given by (Barkana & Loeb 2005;Naoz & Barkana 2005) where x e (t) is the free electron fraction as a function of cosmic time t, δ γ is the photon density fluctuation, t γ = 8.55 × 10 −13 yr −1 , and T γ and δ Tγ are the mean photon temperature and its dimensionless fluctuation, respectively. Equation (6) describes the evolution of the gas temperature in the post-recombination era, but before formation of first galaxies, when the only external heating arises from Compton scattering of the remaining free electrons on the CMB photons. The first term in Equation (6) comes from the adiabatic cooling or heating of the gas, while the second term is the result of the Compton interaction. In the top right panel of Figure 5 we show an example of the solution of Equations (4). We plot the ratio for δ b /δ tot as a function of the wavenumber k for the fully self-consistent linear calculation, i.e., starting at the time of recombination and using the exact transfer functions from Tseliakhovich & Hirata (2010). We consider cases with v bc = 0, v bc = σ vbc , and v bc = 2σ vbc at z = 10 and z = 25. For larger v bc values, the drop in δ b /δ tot occurs at larger scales, i.e., the suppression of the baryonic perturbations relative to the total matter fluctuations shifts to larger masses. Naoz et al. (2009) and Naoz et al. (2011), showed that the characteristic mass is in a good agreement with the filtering mass, regardless of the initial conditions, or even if heating is involved, as long as the filtering mass is calculated self consistently. Motivated by these results we set to find a filtering mass that can be calculated self consistently and that will present the excepted agreement with the simulations. Following Naoz & Barkana (2007), we re-define the filtering scale (specifically, the filtering wavenumber k F ) to include the stream velocity effect as where ν = v bc /σ vbc and σ vbc is the (scale-independent) rms of the stream velocity at small scales. The parameter r LSS (a negative quantity) describes the relative difference between δ b and δ tot on large scales (Naoz & Barkana 2007), i.e., where ∆ = δ b − δ tot (see also Barkana & Loeb 2005). The filtering mass is defined from k F simply as: whereρ 0 is the mean matter density today. To find k F in a general case, we write it in the form where u(t) is to be determined. Then, using equation (7), we expand the baryonic fluctuation as a function of wavenumber k, where ∆ LSS ≡ r LSS δ tot [eq. (8)] obeys the following equation to the first order of k, Note that in the case of v bc = 0, the linear term of k has a zero coefficient, and thus the right hand side of this equation is simply zero (see Barkana & Loeb 2005;Naoz & Barkana 2007). Substituting the expansion from equation (11) into equation (6), and using equations (4) and (12), we obtain an equation for u: In the limit of v bc = 0 (i.e., ν = 0) this equation reduces to equation (12) of Naoz & Barkana (2007), and thus results in the same filtering mass found in that study. We can solve Equation (13) to find u(t), where v bc,rec is the stream velocity at the moment of recombination t rec , so that v bc (t) = v bc,rec a(t rec )/a(t), andk = k/k is the unit wavenumber vector. In the bottom panel of Figure 5 we show the evolution of the filtering mass as a function of redshift for v bc = 0, 1 and 2σ vbc . The values of the filtering mass for v bc = 0 as defined by Equation (7) are larger up to an order of magnitude at high redshifts as compared to the definition of Tseliakhovich et al. (2010). We emphasize that this difference is entirely due to the different definition of the filtering scale, not due to any error in Tseliakhovich et al. (2010) calculations. The filtering scale k F can also be obtained simply by fitting equation (7) to the calculated values of δ dm and δ b , using equations (4)-(6). Naoz & Barkana (2007) found a functional form that can be used to produce a good fit for the drop of the wavenumber. Generalizing it to the case of stream velocity we write: and n must be adjusted at each redshift. In the top left panel of Figure 5 we compare this fitting formula to the fully self-consistent linear calculation (long dashed line) for which reproduce the drop of δ b /δ tot as a function of k fairly well. For the example considered in the figure, i.e., z = 15 and v bc = 1σ vbc , we find n = 0.46 and k F = 253.9 Mpc −1 . We also show the resulted fit using the second order in k approximation, i.e. Eq. (7). Comparison Between the Linear Theory Predictions and the Nonlinear Results In order to compare the filtering mass to the characteristic mass we calculated the filtering mass in a self consistent way, as was done in Naoz et al. (2009Naoz et al. ( , 2011. In other words we use the transfer function from Naoz & Barkana (2005) with a boosted velocity for the baryons at z = 99 (z = 199) for the N = 768 (N = 512 and N = 256) set as initial conditions. We then evolve the dark matter and baryon in time according to equations (4)-(6). We note that in all our calculations we included the fact that the boost of the velocity was included in the simulation only in one axis, thus terms which are proportional to v bc · k are reduced by a factor 3 compare to the global average. Our simulation sets N = 256 and N = 512 are initialized at redshift 199 at which Compton heating by the CMB photons significantly affects the evolution of the linear modes and specifically the filtering mass (Naoz & Barkana 2005. However, GADGET-2 does not include CMB Compton heating. Hence, to compare apples and apples, we neglected the Compton heating contribution to the filtering mass when comparing these two simulation sets to the linear approxima-tion. This is the reason that our values of the filtering mass in Figures 3 and 4 are lower than the values for the no stream velocity case in Figure 3 of Naoz & Barkana (2007). We show the linear theory filtering mass as a function of redshift for all the cases we consider in Figures 2-4. The filtering mass is consistent with the characteristic mass within our fit errors for all simulation sets. Therefore, we conclude in agreement with Naoz et al. (2009Naoz et al. ( , 2011, that the evolution of the characteristic mass can be understood using the linear approximation predictions for the filtering mass. Note that we compare between the filtering mass and M c results from simulation runs using different σ 8 . This is possible since the filtering mass definition is independent on σ 8 because it describes the ratio between the two density fluctuations (thus the normalization of the powers simply cancels out). Recently Tseliakhovich et al. (2010) showed that including the effects of the relative velocity between the dark matter and the baryons at recombinations results in a higher filtering mass as compared to the case of Fig. 3), but now compared with previous theoretical models for the characteristic mass. Solid lines show Naoz & Barkana (2007) definition of the filtering mass (Eq. 16) that neglects the stream velocity contribution. We also plot M ef f (dashed lines) and Mesc (dotted lines) as alternative models, see text for details. v bc = 0 (by about order of magnitude for the global average). Naoz & Barkana (2007) defined the filtering scale in the case of v bc = 0 as: Comparison with the Previous Definition of the Filtering Mass and Other Mass Scales In Figure 6 we show thus defined filtering mass for the N = 512 simulation set. As can be seen, the Naoz & Barkana (2007) definition underestimates the characteristic mass in the high v bc limit, since it neglects the difference between the dark matter and baryons density fluctuations. Recently Stacy et al. (2011) andNaiman et al. (2011) suggested that, given a high initial stream velocity, the baryon evolution is dominated by the relative motion of dark matter and gas, and thus the gas sound speed should be replaced with the effective sound speed v eff , where v bc (z) = v bc,0 /(1+z), in the Jeans mass definition. This effective Jeans scale k J,eff can be written as where G is the gravitational constant andρ m is the average density. The effective Jean mass M eff associated with this scale length is simply We show this mass scale in figures 6 with dashed lines. As one can see, this mass scale overestimates the characteristic mass at all redshifts and for all values of the stream velocity that we considered. This is not unexpected, since the Jeans mass always overestimates the scale at which pressure starts to overtake gravity in the expanding background (Gnedin & Hui 1998). The evolution of the characteristic mass as a function of redshift can also be modeled by considering the escape velocity of the gas. Given a mass of a halo M , the escape velocity is simply v esc = 2GM/r, where r is the (comoving) virial radius of the halo. For example, for a halo of 10 5 M ⊙ the escape velocity is about 0.77 km sec −1 , while the stream velocity for v bc = 3.4σ vbc at z = 15 is 1.6 km sec −1 and at z = 25 is 2.6 km sec −1 . Thus, it is not surprising that halos below 10 5 M ⊙ are empty in that redshift range -the stream velocity is simply much larger then the halo escape velocity, so the dark matter halo is unable to accrete any gas. We can, thus, estimate a halo mass M esc below which the stream velocity is larger then the escape velocity, where ∆ c = 200 is the virial overdensity and v bc (z) is the stream velocity at redshift z. We show this limit in Figure 6 with dotted lines. Equation (20) predicts a much stronger evolution of the characteristic mass than is actually observed in our simulations and provides a poor fit to simulations results. CONCLUSIONS We have used three-dimensional hydrodynamical simulations to investigate the effects of stream velocity on the gas fraction in high redshift halos. In a companion paper Naoz et al. (2012), we studied the effect of the stream velocity on the total halo mass function, In this work we focus on the effect of the stream velocity on the gas fraction in halos and on the evolution of the characteristic mass, and compare the simulation results to the linear approximation. In a first improvement over the earlier results, we introduce a new fitting formula (Eq. 2) which offers a much better fit to the gas fraction as a function of halo mass at a given redshift in the limit of large stream velocities, while returning essentially the same values of the characteristic mass M c as the previously used functional form (see Figures 1, 7 and 8, and see Appendix A). Previous studies (Naoz et al. 2009(Naoz et al. , 2011 showed that a quantity defined in the linear approximation, the filtering scale (Gnedin & Hui 1998), provides a good match to the nonlinear characteristic mass measured in numerical simulations. We introduce a new definition for the linear filtering mass that accounts for two effects neglected in Gnedin & Hui (1998): the deviation of the amplitude of baryonic fluctuations from the dark matter fluctuations on large scales (considered first by Naoz & Barkana 2007) and the stream velocity between the dark matter and baryons on small scales, which we include in the definition of the filtering mass for the first time in this paper. The latter effect may result in the filtering mass being up to an order of magnitude larger at high redshifts for high values of the stream velocity, as compared to the case when the stream velocity is neglected. Finally, in comparing our simulations results to the linear calculation (using our new definition of the filtering mass), we find that the filtering mass (i.e. a linear quantity) offers an accurate match to the actual nonlinear characteristic mass measured from the simulations, at all redshifts and for all values of the stream velocity that we simulated. On the contrary, previous theoretical models that used as the characteristic mass scale either the halo mass with the escape velocity equal to the stream velocity or the Jeans mass for the "effective" gas sound speed provide only poor fits to the simulation results. It has been suggested in the literature that gas rich low mass halos may play an important role in cosmic reionization, and that they can produce distinct 21-cm signatures (Kuhlen et al. (2006); Shapiro et al. (2006); Naoz & Barkana (2008) but see Furlanetto & Oh (2006)). For example, minihalos (halos of mass ∼ 10 6 M ⊙ ) can potentially block ionizing radiation and induce an overall delay in the initial progress of reionization (e.g., Shapiro & Giroux 1987;Barkana & Loeb 2002;Iliev et al. 2003b;Shapiro et al. 2004;Iliev et al. 2005;McQuinn et al. 2007). However, our results here suggest that at high redshifts the stream velocity effect results in large variations in the characteristic mass -i.e. the minimum mass of a gas rich halo. Thus, if reionization started sufficiently early (Yoshida et al. 2007), in patches of the universe where the stream velocity is large there were fewer gas rich halos that can absorb ionizing photons. Hence, in these patches the delay of the reionization caused by minihalos would be less than in regions that happen to have a small value of the stream velocity and, hence, a large abundance of minihalos. Therefore, not only the formation of the first generations of galaxies may be affected by the stream velocity effect, but also the whole process of reionization may proceed differently in regions with very different stream velocities. This effect has been considered recently by Visbal et al. (2012) and , but our results indicate that it can even be stronger than previously estimated. ACKNOWLEDGMENTS We thank Avi Loeb, Rennan Barkana, Andrey Kravtsov, Neal Dalal, Will Farr, Matt McQuinn and Dmitriy Tseliakhovich for useful discussions. We thank Dmitriy Tseliakhovich for providing his code. We also thank Yoram Lithwick for the use of his allocation time on the computer cluster Quest. This research was supported in part through the computational resources and staff contributions provided by Information Technology at Northwestern University as part of its shared cluster program, Quest. We show all of the halo of which N h ≥ 100 (grey points) as well as the binned data points for N h ≥ 500. We consider the various values of the stream velocity (from left to right). v bc = 0, v bc = 1σ vbc , v bc = 1.7σ vbc , and v bc = 3.4σ vbc . Note for the case of v bc = 3.4σ vbc , at z = 25, about 2/3 of the halos have less than 1% of gas in them, and f b,0 is very low. Thus no reliable fit could be found, and therefore we did not show the resulted red line in that case. Yoshida, N., Sugiyama, N., & Hernquist, L. 2003b, MNRAS, 344, 481 APPENDIX FIT CALCULATION As can see in Figure 1 (dotted lines), the fitted formula found by Gnedin (2000), i.e., eq. (1), dose not capture the behavior of the gas fraction as a function of mass for the cases of non-negligible stream velocity. In our search for better fitting formula we found that the low mass tail has a significant effect on the fit for large v bc values. It is not surprising since the stream velocity deprives the low mass halos of gas at high redshift, as we showed in Paper I. Therefore, this induce a large dependency on the behavior of the gas fraction in the low mass tail. However, the gas fraction of the low mass halos with less than 500 particles is poorly constrained (as was shown in Naoz et al. 2009). Therefore, we introduce a weight function, that account for the errors in estimating the gas fraction as a function of the number of particle in a halo, N h . We use Naoz et al. (2009) resolution study, their figure 4, and assume a scatter of 20% for halos that have more than 500 particles. For smaller number of particles per halo we adopted a simply linear function of the error of the gas fraction as can be estimated from Naoz et al. (2009), figure 4. Thus the weight function has the following form: Using this weight function we find the fit, using the forma formula [eq. (1)] and the new formula eq. (2). In figure 7 we show an example for the gas fraction as a function of mass, and the two fits as in Figure 1 (solid lines for the new formula and dotted lines for the former one). Here we also show all of the points we considered in evaluating the fit (i.e., N h ≥ 100). We note that the χ 2 for the new formula fit is for most cases higher (closer to one) than the χ 2 of the old formula, both lower than unity. (1), and the color points are the fit for the new model, using eq. (2), calculated with minimum of 100 particles per halo (see text). We consider (from left to right) the N = 256, 512 and N = 768 sets. We consider the various values of the stream velocity v bc = 0, v bc = 1σ vbc , v bc = 1.7σ vbc , and v bc = 3.4σ vbc (blue squares, purple diamonds, brown stars and red triangles respectively). The ultimate goal of the fitting process is to find the characteristic mass M c . An important test is to compare the resulted M c from the two fitting formulae. The different values achieved for the two models are shown in Figure 8. As depicted in this Figure, the M c for the different values of v bc converge over the different ranges of redshift for the N = 512 and N = 768 sets. However, or the N = 256 set the new formula produces a systematically low value for M c even for the v bc = 0 case. As mentioned in Paper I, this run suffers from low statistic, particularly in the large mass tail,. Furthermore, there was no convergence of the gas fraction, these are the main cause for this systematics. Therefore, based on the N = 512 and N = 768 sets, we conclude that although the new formula produces somewhat better fit, the final M c results did not changed by much. This is not surprising since the meaning of the two models is the same. As shown in Figure 7, the new formula results in a plateau at the high mass tail, which indicates a possible degeneracy between the fitting parameters. This of course present a problem in evaluating the values and errors of M c . However, as shown in Figure 8 the values of M c from the two fitting models are consistent. Therefore, in evaluating the errors of the fitting values of M c we choose to be conservative and select the larger values between the 1 − σ errors from the two models. In addition we have used bootstrap method for the new model, in some cases of the N = 512 set to test our evaluation of the errors 9 , and found that they are consistent with choosing the maximum 1 − σ from the two fit models. We show the complete best fit parameters for the 512 set in table 2, they are similar for the other sets. Note that for high redshift and large v bc values the best fit parameters are poorly constrain. Specifically the parameters β and γ from equation (2) are sometimes so poorly constrains (i.e., more then an order of magnitude) that we omit the errors from the table, thus the symbol "−−" in table 2 means error larger then an order of magnitude, for those cases χ 2 was close to zero. Note that in our calculation of the fit we used f b,0 which is the gas fraction in the high mass tail. This value is lower than the mean cosmic baryonic fractionf b , since the baryons are lagging behind the dark matter even at high redshifts (see Naoz & Barkana 2007;Barkana & Loeb 2011). As shown in Naoz et al. (2011, fig . 4), this causes a higher f b,0 for the initial conditions that assumes δ b = δ dm than the f b,0 resulted from the smother baryonic initial conditions, Fig. 9.-Comparison between the new fit model and the old one while using f b,0 andf b . We consider the N = 512 run for z = 15 representative example. We compare between usingf b value, solid (dotted) grey lines for the new (old) model. We also show the fit while using the f b,0 value, solid (dotted) color lines for the new (old) model. We consider the various values of the stream velocity v bc = 0, v bc = 1σ vbc , v bc = 1.7σ vbc , and v bc = 3.4σ vbc (blue squares, purple diamonds, brown stars and red triangles respectively). as explored in their other two initial conditions models 10 . In addition to the physical reason, a numerical reason may arise, since some baryons are in fact bound to the halo and we miss them due to our halo finder algorithm ( §2.3). As can be seen in Figure 9 assuming f b,0 →f b , grey lines, by either of the models results in a worse fit than using the calculated f b,0 .	2012-12-21T21:34:22.000Z	2012-07-23T00:00:00.000	{ "year": 2012, "sha1": "b621f6a591c0ce2869f066fc70ea0b59665d40be", "oa_license": null, "oa_url": "http://arxiv.org/pdf/1207.5515", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "b621f6a591c0ce2869f066fc70ea0b59665d40be", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics" ] }
256306146	pes2o/s2orc	v3-fos-license	Fabrication of silver nanoparticles-deposited fabrics as a potential candidate for the development of reusable facemasks and evaluation of their performance Recently, wearing facemasks in public has been raised due to the coronavirus disease 2019 epidemic worldwide. However, the performance and effectiveness of many existing products have raised significant concerns among people and professionals. Therefore, greater attempts have been focused recently to increase the efficacy of these products scientifically and industrially. In this respect, doping or impregnating facemask fabrics with metallic substances or nanoparticles like silver nanoparticles has been proposed. So, in the present study, we aimed to sonochemically coat silver nanoparticles on the non-woven Spunbond substrates at different sonication times and concentrations to develop antibacterial and antiviral facemask. The coated substrates were characterized using Field Emission Scanning Electron Microscope, Energy Dispersive X-Ray, X-ray diffraction, and Thermogravimetry analysis. The amount of silver released from the coated substrates was measured by atomic absorption spectroscopy. The filtration efficiency, pressure drop, and electrical conductivity of the coated samples were also investigated. The antibacterial activity of fabrics was evaluated against Escherichia coli and Staphylococcus aureus. Cellular viability of samples assessed by MTT and brine shrimp lethality tests. The results revealed that the higher sonication times and precursor concentrations result in a higher and more stable coating, larger particle size, wider particle size distribution, and lower content of released silver. Coated fabrics also revealed enhanced filtration efficiency (against nanosize particles), desired pressure drop, and antibacterial activity without significant cytotoxicity toward HEK 293 cells and Artemia nauplii. As a result, the coated fabrics could find potential applications in the development of facemasks for protection against different pathogenic entities. www.nature.com/scientificreports/ (AAS) measurements, respectively. By knowing the effective antibacterial and antiviral properties of SNPs, we investigated the filtration efficiency (FE), pressure drop, and electrical conductivity of the resultant fabrics for developing facemasks suitable for different medical and general applications. The antibacterial activity of samples was also investigated by Escherichia coli (E. coli) as Gram-negative and Staphylococcus aureus (S. aureus) as Gram-positive bacteria, and cell viability was assessed using MTT and brine shrimp lethality tests using Artemia nauplii (A. salina). Methods. Preparation of coated fabrics. SNPs-coated Spunbond fabrics were coated as described in the literature with some modifications 42 . Briefly, a 500 ml of water/ethanol/ethylene glycol solution (10:7:3) with different concentrations of AgNO 3 was purged under Argon (80 mL/min) for one hour to remove the trace amount of O 2 /Air from the reaction vessel in the presence of 6 plies of fabrics. Then, the solution was irradiated under the flow of Argon at different times with a high-intensity probe sonicator (Titanium horn, 20 kHz, 400 W at 60% efficiency). After 2 min, a 28-30% Ammonia solution was introduced to the reaction medium. The temperature of the reaction was adjusted around 25-30 °C by a cooling bath. At the end of the reaction, the products were isolated and washed three times with distilled water and ethanol to remove the remaining ammonia and finally oven-dried in a vacuum oven. Different reaction conditions in terms of precursor concentrations and reaction times are used for coating the fabrics, as listed in Table 1. Characterization of fabrics. SEM and EDX analysis. SEM technique was used to investigate the morphology of the deposited SNPs on the Spunbond fabrics. A small section of each fabric was stuck on glass slides, placed at the SEM holder, and sputter-coated for 90 s with Gold for better conductivity during imaging. Then observations were performed using FESEM, MIRA3 TESCAN in 20.0 kV and high-vacuum mode. Elemental analysis of the coated samples was conducted using an EDX detector (EDS, MIRA3 TESCAN) attached to the FESEM machine 44 . The particle size and particle size distribution of coated samples were measured by ImageJ 1.52v software (http:// imagej. nih. gov/ ij/). XRD analysis. To evaluate the effect of the coating process and deposited SNPs on the crystallinity of the fabrics, as well as to confirm the presence of SNPs on the fabrics, the uncoated and coated fabrics were analyzed by a wide-angle X-ray diffractometer (EQuinox 3000, INEL, France). The spectra were recorded at a scan rate of 2.4 min -1 in the range of 2θ = 5-80°4 5 . TGA analysis. The total content of SNPs on the fabrics was measured by TGA analysis (TA universal model V1.7F) by heating them from 30 to 800 °C at a heating rate of 10 °C/min under an N 2 atmosphere. The effect of the coating procedure and coated SNPs on the fabric's mechanical properties have also been evaluated 46 . Mechanical properties. Stress-strain curves were obtained according to ASTM D8822 using a universal tensile testing machine (Santam, Iran) at room temperature. Strips with 6 × 4 cm 2 sizes were cut off from each sample, mounted into the grips and stretched with a 10 mm/min strain rate until breakage. Two specimens of each fabric before and after coating were tested 42 . Filtration efficiency and air flowability. The FE of the coated fabrics was measured by a lab-scale filtration devise against airflow and different particle sizes ranging from 0.3 to 3.0 µm and compared with uncoated samples. Briefly, samples were cut into square sheets of 10 × 10 cm 2 and firmly fixed between two chambers of the device. By introducing airflow containing different particle sizes in one chamber, the sample's FE and air www.nature.com/scientificreports/ exchange capability were measured in another chamber by a particle mass counter device 47,48 . All measurements were carried out in triplicates, and the average values were calculated. Conductivity test. The conductivity of the coated fabrics was measured using a four-probe resistance IV meter. For this purpose, the coated and uncoated fabrics were cut into rectangular specimens of 20 mm in length and 10 mm in width and soaked in deionized water. Then the excess water was removed with a filter paper to perform a conductivity test in a wet state 49 . The electrical conductivity of three individual samples from each fabric was measured, and the average values were calculated. In vitro release study. The content of the silver released from coated samples was measured by AAS analysis using novAA 350 (Analytik Jena, Jena, Germany). For this purpose, 2 × 2 cm 2 of coated fabrics were accurately weighed and then placed in 50 ml distilled water as the release medium (pH 6) with a magnet bar. Then, aliquots of 5 mL of solution media were withdrawn at intervals of 24, 48, and 72 h and replaced by the same volume of freshwater. Then, the silver ion concentration in the solution was measured by the AAS method. All measurements were performed in triplicates, and the average values were calculated 50 . Bactericidal assay. The antibacterial activity of coated fabrics was performed against S. aureus (ATCC 25,923) and E. coli (ATCC 25,922), as described elsewhere 42 . Both strains were obtained from the Iranian Research Organization for Science and Technology. A typical protocol was used as follows: First, the bacteria cultures were prepared using nutrient agar modified (QUELAB QB-39-3504) overnight. Then, the cultures were transferred into a nutrient broth (NB) and incubated at 37 °C with aeration. After reaching the logarithmic phase, they were centrifuged and washed to yield a final bacterial concentration of approximately 10 8 CFU ml −1 . Next, 500 μL of the strain cells containing 10 8 CFU ml −1 of each strain was transferred into a vial containing 4.5 ml of a saline solution and coated fabrics (1 cm × 1 cm). Besides test samples, saline without the sample and saline with an uncoated sample was included in the experiment as control groups. After incubation of bacterial suspensions at 37 °C for four hours, 100 μl of each sample at specific time intervals (t = 0, 1, and 3 h) was transferred onto nutrient agar plates after dilution with saline. The plates were incubated overnight at 37 °C, and viable bacteria were counted. The number of initial CFU (N0) and final CFU at the specific times (N) were used to determine the bacterial survival fraction (N/N 0 ). The antibacterial activity of fabrics were performed in triplicates, and the average values were calculated. MTT assay. The cytotoxicity effect of silver-coated fabrics was conducted by MTT assay as described by Vijayakumar 51 . Briefly, 1 × 10 5 mL −1 cells (HEK 293) were seeded in a 96-well plate in their exponential growth phase and were incubated overnight at 37 °C and 5% CO 2 in a humidified atmosphere. Statistical analysis. The statistical analysis was assessed by Student's t-test or One-Way ANOVA. P < 0.05 indicated the statistical significance. The form of mean ± standard deviation (SD) was used to convey the data. The data analysis software was SPSS v26 (IBM SPSS Statistics, Chicago, IL, USA). Results and discussion In this study, we coated Spunbond fabrics with SNPs at different concentrations and times and evaluated their physicochemical and functional properties to fabricate antibacterial and antiviral facemasks (Fig. 1A). Ethylene glycol was used as a reducing agent for producing SNPs and subsequent deposition on the fabrics. In this process, the [Ag(NH 3 ) 2 ] + complex formation takes place by the addition of ammonia solution to the reaction medium. Because of the large constant equilibrium of the formation of this complex, the concentration of silver ions remains in a small amount in equilibrium with the complex. Therefore, the ammonia concentration and Ag + /NH 3 ratio in the reaction medium is an important factor in reducing the silver ions to the desired particle www.nature.com/scientificreports/ size. As a result, in this study, the ratio of Ag + /NH 3 was kept constant, and only the effect of different concentrations of silver and reaction times were evaluated for obtaining the efficient coating of SNPs on the fabrics. The images of all uncoated and coated fabrics are depicted in Fig. 1B. Different concentrations of silver precursor in the reaction medium and different coating times significantly affect coating efficiency, which is obvious from the color of the coated fabrics. Obviously, by increasing the silver concentration and reaction time, the color of fabrics turned darker because of the higher loading of SNPs on the fabrics. The deposited silver content on the fabrics for different precursor concentrations (Samples 0, 1, 2, and 3) and different times (Samples 0, 2, and 6) was determined by TGA analysis and illustrated in Fig. 2. From respective TGA diagrams for each sample, it is evident that the content of deposited silver on the surface of the fabrics depends on the silver nitrate concentration and www.nature.com/scientificreports/ reaction time. The weight percent of deposited silvers varies from 0.1%, 3.2%, 3.4% and 8.5% for 1, 3, 6 and 2 samples, respectively. The morphological characteristics of the samples were investigated by the SEM technique (Fig. 3). The SEM images in Fig. 3a support the results obtained from the TGA analysis. It was shown that the content of deposited silver on the fabrics is in direct correlation with the precursor concentration and reaction time. With increasing the precursor concentration from 2 mM (sample 1) to 50 mM (sample 4) and reaction time from 25 min (sample 4) to 50 min (sample 8), the content of deposited silver on the fabrics was increased. However, coated sample 2 showed unusual results with the highest SNPs coating over all samples. This figure also highlights the differences between coated (Samples 1, 4, and 8) and uncoated fabrics (Sample 0). Figure 3b shows the particle size and particle size distribution of coated samples measured by ImageJ software. The particle size of the deposited silver on the fabrics is on the nanoscale. It is evident that increasing the precursor concentration and sonication time results in the formation of larger particles and higher deposition of NSs (100.14 ± 8, 105.97 ± 14, and 125.12 ± 21 nm for samples 1, 4, and 8, respectively). On the other hand, the particle size distribution was increased with increasing sonication time. These results are in good agreement with previous studies where other precursors and substrates were coated with the ultrasonication method 42,53 . EDX spectra recorded from the uncoated and coated samples are shown in Fig. 4. It is clear that the coated fabrics have the weight percentage of silver as 0.78%, 66.35%, and 47.07% for 1, 4, and 8 samples, respectively. These results again indicate the increases in the silver deposition on the fabrics by increasing the precursor concentration and reaction time. XRD measurements of coated and uncoated fabrics have been shown in Fig. 5a. Uncoated sample 0 and coated sample 1 did not show any characteristic diffraction peaks of silver at all 2-theta scales. Coated sample 1 showed a very low silver content (0.1% from TGA analysis), so the characteristic diffraction peaks of silver are absent in the XRD pattern of this sample. On the other hand, the coated sample 4 showed the characteristic diffraction peaks of silver at 2θ values of 38.1°, 44.45°, 64.55°, and 77.40° corresponding to the (111), (200), (220), and (311) planes of metallic silver, respectively 54,55 . The mechanical properties of coated and uncoated samples were also evaluated to determine the effect of the coating procedure on the mechanical behavior of samples (Fig. 5b). As shown in the figure, silver-coated fabrics (Samples 1, 2, and 6) show a somewhat more brittle behavior than the uncoated fabric (Sample 0). This behavior was higher for coated fabrics with higher concentrations of precursors at the www.nature.com/scientificreports/ www.nature.com/scientificreports/ same reaction time (Sample 2 and 6 vs. sample 1) and longer reaction times at the same precursor concentration (Sample 2 vs. Sample 6). On the other hand, the tensile force for the coated samples 1, 2, and 6 was surprisingly about 27%, 55%, and 46% higher than the uncoated sample, respectively. These results are in direct correlation with precursor concentration and reaction time. However, previous studies have shown both consistent and inconsistent results with the results obtained in this study. In some of these studies, the coated samples showed less tensile force than uncoated samples 42 . However, the other studies corroborated the result of the present study as the coating process has led to an increase in the tensile force. The latter is attributed to the enlargement of the fiber diameter due to the ductile nature of silver metal 56 . Evaluation of the drop pressure of the fabrics resulted in the air flowability of 30 LPM for both coated and uncoated samples, which support normal and easy breath [57][58][59] . On the other hand, the FE of the coated samples confirmed the adequate performance of the resultant fabrics in the filtration of particles of different sizes (Fig. 6a). The uncoated sample showed a 52.75-76.88% FE for particle sizes in the range of 0.3 to 3.0 µm, while samples 2 and 6 resulted in 78.94-89.32% and 81.24-91.60% FEs for particles in this size range, respectively. These results are significantly higher than the values reported for many commercial and handmade facemasks and comparable to N95 masks that are currently used by the public [60][61][62][63][64] . On the other hand, the fabricated facemasks in the present study possess additional advantages by providing antibacterial activity and improved drop pressure compared with N95 masks. Increasing the FE by deposition of SNPs attributed to the increase in the tortuosity of the fabrics by filling the open and connected voids. PP meltblown fabrics were commonly used to fabricate traditional and medical facemasks and gowns. The FE of these products relies on their static electricity originating from the meltdown process. However, the efficacy of electrostatic filtration drops drastically in the moist environment and by prolonged use or re-use. Moreover, smaller particles such as bacteria and virions can easily pass through these fabrics due to their larger pore diameter. In this regard, the study conducted by Dnyanmote et al. indicated that the diameter of the aerosol-carrying virus is bigger than 2.5 µm, and a membrane with a pore diameter of 0.18-0.5 µm could provide a highly effective filtration. Finally, they concluded that the doping or impregnating these scaffolds with metallic NPs like SNPs could enhance the FE of the resultant fabrics against COVID 19 65 . Besides the positive effect on the FE of the fabrics, other studies assessed the antiviral activity of SNPs against coronavirus. For example, Magomedow et al. examined the possible mechanisms by which SNPs could influence coronavirus using computer quantum-chemical modeling. The results revealed that the formation of the "tryptophan-SNPs" complex (E = -5856.83 kcal/mol) is the most energy-efficient interaction among other possible interactions. Moreover, the most stable complex was the "cysteine-SNPs" complex (Delta E = 0.16 a.u.). Therefore, they concluded that SNPs could exert antiviral activity through the interaction with tryptophan and cysteine amino acids of the coronavirus spike proteins 66 . Other studies also reported similar results about SNP's antiviral activity [67][68][69] . These findings clearly indicate the potential application of the coated samples in developing efficient facemasks or gowns for different situations like the COVID-19 pandemic and other medical and industrial applications. The electrical conductivity of the fabrics was examined using a four-probe resistance meter in the wet state, and the results are depicted in Fig. 6b. As we can see, the conductivity of the fabrics increases by increasing the content of SNPs deposited on fabrics. In this respect, samples 4 and 8 showed conductivity of 1.7 × 10 -4 S•mm -1 and 0.98 × 10 -4 S•mm -1 , as a result of higher precursor concentration and sonication time, respectively. It is worth mentioning that the electrical conductivity of sample 8 is nearly 12-times compared to uncoated fabrics. On the other hand, the results clearly indicate that the fabrics possess high conductivity in the wet state, which is highly beneficial for different applications. As noted above, the electrostatic properties of the PP meltblown fabrics play www.nature.com/scientificreports/ an important role in providing effective FE against environmental and biological pollutants. However, these fabrics significantly lose their effectiveness upon contact with water or a moist environment. In this regard, major attempts have been made to fabricate electroceutical fabrics that maintain an electric field in a moist environment or wirelessly generate a low level of electricity in the presence of moisture. For example, Ghatak et al. fabricated an electroceutical polyester fabric printed with alternating circular regions of Ag and Zn dots and evaluated its performance in deactivating porcine respiratory coronavirus AR310 particles. The results showed that the resultant fabrics significantly lower the zeta potential of the virion in one minute of contact and lead to the eradication of its infectivity through destabilization of its electrokinetic properties and subsequent aggregation 70 . The electroceutical fabrics are also expected to find potential applications in other fields, such as wound healing [71][72][73] . In the next step, the coated and uncoated fabrics were subjected to in vitro release study for evaluation of the coating's stability. The 2 × 2 cm 2 specimens were immersed in 50 ml distilled water with a magnet bar, and 5 ml of the resultant silver-containing solutions were subjected to AAS analysis after 24, 48, and 72 h intervals to determine the released content of silver. Referring to Fig. 7, the in vitro accumulative release of SNPs shows a relatively time-dependent behavior over 72 h. However, regarding the cumulative nature of the study and relatively constant increase in silver content at specific time intervals, it can be concluded that the release profile of silver has been constant over time. It is evident that the accumulative release content decreases with increases in the reaction time (Samples 1-4 vs. Samples 5-8) and increases with increasing the precursor concentration (Samples 1 to 4 and Samples 5 to 8). In other words, higher precursor concentrations and reaction times lead to strong adhesion of the NPs to the fabric surface, suggesting a more stable coating. However, more SNPs were deposited on the fabrics in higher precursor concentrations and reaction times, resulting in a higher amount of silver released into the medium. In a typical sonochemical process, in situ generations of NPs occurs in the reaction medium and follows by simultaneous adhering to fabrics by ultrasonically generated fluidic microjets. Anchoring of metal NPs to the fabric surface by physical, chemical, or other interactions in the sonochemical process has been reported in previous studies 74 . However, according to the present study results, it seems that the nature of metallic NP bindings to the surface of the fabrics is of a physical type (Fig. 8). First, a metallic monolayer covers the surface by physically anchoring the SNPs to the fabric surface. Next, metallic NPs deposit layer-bylayer (hypothetical) on the fabric surface by strong metal bonds. Because higher sonication times lead to more silver deposition, stronger binding of NPs could be achieved for these samples, resulting in a lower amount of silver released from fabrics. On the other hand, the strong deposition of SNPs in longer reaction times somehow enhances the interaction of the first layer with the fabric surface. Physical bonding between the fabric surface and the metal or metal oxide NPs rather than chemical bond formation was also reported by Perelshtein et al. 75 . The minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) of SNPs were reported as 3.12 ppm and 6.25 ppm for E. coli, and 0.625 ppm (both MIC and MBC) for S. aureus, respectively, which are in the range of silver amounts released form coated fabrics (1-6 ppm) 76,77 . Therefore, in the next step the antibacterial activity of SNP-coated fabrics was carried out against S. aureus and E. coli as a Gram-positive and Gram-negative bacteria. The results have been summarized in Table 2 and depicted in Fig. 9. Efficient antibacterial activity against both bacterial strains was observed for all coated fabrics. As shown in Table 2, three hours of incubation led to total inhibition of both bacteria, while only in the case of E. coli, 100% bactericidal activity was observed within the first hour. However, a 100% reduction for S. aureus was achieved only after 3 h of incubation. This difference in the results between Gram-negative and Gram-positive bacteria is attributed to www.nature.com/scientificreports/ the SNPs bactericidal mechanism and differences in the cell membrane structure of these two strains of bacteria. According to the generally accepted mechanism, SNPs continuously release silver ions that attach to bacteria's cell wall and cytoplasmic membrane through electrostatic interactions and high affinity to sulfur proteins. Thereby, increasing the membrane permeability and changing the cell functionality eventually leads to cell death [78][79][80] . On the other hand, Gram-positive and Gram-negative bacteria are very different in structure, morphology, and cell wall components. For example, due to the high thickness of the cell wall and dense peptidoglycan layer, the Gram-positive bacterias show more protection against antibacterial agents than Gram-negative bacteria, which are mostly made of tightly packed lipopolysaccharides (LPS). These differences can lead to differences observed in antibacterial testing for these two types of bacteria. However, as shown here, fabrics coated with SNPs are effective for both groups of these bacteria. The cytotoxic effects of the coated fabrics were evaluated by MTT assay by exposing HEK 293 cell lines to different concentrations of the SNPs solutions obtained from in vitro release study at different time intervals (Fig. 10). As shown in Fig. 10, all coated samples showed high safety over test times. For all samples, the viability percentage of HEK 293 cells decreased in a time-dependent manner at 24 h, 48 h, and 72 h, indicating that with increasing the release time, a higher amount of SNPs are released from fabrics. On the other hand, the cell viability in samples 1-4 was higher than in samples 5-8, which is reasonably attributed to a lower amount of silver deposited on the fabrics and subsequently released into the medium. From this point of view, the viability percent of HEK 293 cells also exhibits a concentration-dependent manner. In other words, cell viability has decreased by increasing the SNPs liberation, which, in turn, depends on the initial silver precursor concentration and sonication time. Therefore, we can conclude that higher sonication time results in sturdy and stable nanomaterial coatings. The brine shrimp lethality assay was also conducted to evaluate the cytotoxicity of released SNPs from coated fabrics. Figure 11 shows similar results obtained from the MTT cytotoxicity test. Therefore, the viability percentage of A. Salina decreases in a concentration-dependent and time-dependent manner. In other words, with increasing the test time, the survival rate of A. Salina has declined. In addition, with increasing the concentration of released SNPs during 24-72 h, the survival rate of A. Salina has been decreased. Again, we can see that the viability percentage of the A. Salina population in samples 1-4 is higher than in samples 5-8. As mentioned above, stable SNPs coatings occur in the samples prepared in higher precursor concentrations and sonication time. However, the amount of deposited SNPs is also higher in these samples, which results in a higher amount of www.nature.com/scientificreports/ silver release and reduced viability of A. Salina. Therefore, the lower amount of deposited SNPs in these samples (1)(2)(3)(4) resulted in an unexpected more A. Salina surviving rate. Conclusion The present study demonstrates the sonochemically coating of SNPs on the surface of Spunbond fabrics in different precursor concentrations and reaction times for the development of antibacterial facemask fabrics. The results revealed that this one-step and straightforward method results in remarkably stable coatings by varying the effective parameters that control the extent and quality of coatings on fabrics. The results also showed that the content of deposited SNPs on the fabrics increases with increasing the precursor concentration and reaction time. On the other hand, increasing the sonication time and precursor concentration has increased the particle size and particle size distribution of deposited silver. Further investigation of the coated fabrics in the in vitro release study revealed that higher reaction time results in stable and sturdy coatings and lower content of released silver. Physical bonding between the SNPs and fabrics and metallic bonding between SNPs in a higher amount of deposited silver are speculated for these findings. The particulate-induced FE and enhanced electrical conductivity of the coated textiles provided suitable clearance of particles of different sizes and offer desired air flowability characteristics. The coated fabrics also demonstrated highly efficient antibacterial activity against Gram-positive and Gram-negative bacteria. Investigation of cytotoxicity effects of coated fabrics by MTT assay and brine shrimp lethality test showed high safety for HEK 293 cells and Artemia nauplii. However, the higher concentration of precursor resulted in higher cytotoxicity effect in Artemia nauplii test via high rate of drug release, which should be considered and resolved in future studies. Consequently, it is revealed that a higher concentration of silver precursor and sonication times leads to a strong and uniform coating of SNPs, lower liberation of silver from coated fabrics, and higher antibacterial activity and safety for these fabrics. However, additional studies are required to assess the in vitro and in vivo antiviral activity of the coated samples. In the end, we think that the sonochemically coating technique could be applied for preparing products with enhanced properties and applications, including air filtration, antibacterial applications, food packaging, textile industries, etc. www.nature.com/scientificreports/ www.nature.com/scientificreports/	2023-01-28T14:44:03.305Z	2023-01-28T00:00:00.000	{ "year": 2023, "sha1": "03649147199005bd20b26051ad6eef735c2a3e4b", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Springer", "pdf_hash": "03649147199005bd20b26051ad6eef735c2a3e4b", "s2fieldsofstudy": [ "Materials Science" ], "extfieldsofstudy": [ "Medicine" ] }
236301304	pes2o/s2orc	v3-fos-license	Inﬂuence of Fertilization and Mycorrhizae on the Nutritional Status of Rhododendron ( Rhododendron hybridum ) in a Nursery : Background: This study of the large-ﬂowered azalea cv. ‘Anneke’ investigated the impact of two factors, i.e., methods of fertilization and mycorrhization, on the nutritional status of plants during three years of nursery cultivation. Methods: Single mineral fertilizers, a slow-release fertilizer Hortiform pH (SRF), and fertigation in combination with mycorrhization of plants, were applied. Plant roots were inoculated with fungi from the genera Oidiodendron and Hymenoscyphus sp. The nutritional status of the large-ﬂowered azalea in the ﬁrst three years of cultivation was assessed based on macroelements. Results: The analyses revealed signiﬁcantly higher content of nitrogen, phosphorus and calcium in the leaves of plants inoculated with fungal mycelium. A beneﬁcial effect of plant mycorrhization on plant nutritional status, i.e., higher levels of nitrogen, phosphorus, potassium, calcium, and magnesium, was noted in the second and third years of azalea cultivation. Conclusions: Signiﬁcant amounts of the nutrients were utilized in the middle of the growing season and almost fully utilized after the season. Hence, the necessity to supplement nutrients in each subsequent year of plant vegetation was postulated. Fertigation was shown to require further improvement of the nutrient solution. Introduction Azaleas are permanent elements of gardens and parks. Their flowers provide a wonderful color palette in late May and early June. Noteworthy is the high frost resistance of azaleas, which are, therefore, fully suitable for cultivation in temperate climate conditions. Plants from the family Ericaceae: Rhododendron, Andromeda, Calluna, Erica, Kalmia, and Pieris have low nutritional requirements [1][2][3], as nutrient-poor soils are their natural habitat. The plants take up nutrients from raw humus, which contains poorly decomposed plant parts with a small amount of mineral substances [4]. Although nutrient-poor localities are natural habitats for Ericaceae plants, nursery practice and investigations indicate that supplementation with appropriate amounts of nutrients stimulates these plants to grow intensively [5]. The substrate used in an ornamental plant nursery is garden peat containing trace amounts of nutrients, which should therefore be supplied to plants in sufficient amounts for proper growth and development. In practice, single mineral fertilizers, multicomponent fertilizers and slow-release fertilizers are used for fertilization. In recent years there has been dynamic growth in the cultivation of Ericaceae plants, including the large-flowered azalea [6]. Fertilization of Ericaceae plants should meet the specific soil pH requirements of these plants. The basic fertilization method involves the use of single mineral fertilizers. However, fertilizers with controlled release of nutrients (CRF = Controlled Release Fertilizers) or slow release of nutrients (SRF = Slow Release Fertilizer) together with fertigation, are used more frequently [7]. Although fertigation has been applied primarily in the cultivation of plants under cover to date, this treatment is increasingly being used as supplementation or replacement of mineral fertilization [8]. In their natural habitats, Ericaceae plants establish symbiosis with mycorrhizal fungi. This type of mycorrhiza is referred to as ericoid mycorrhiza (ERM) [9]. Ericoid mycorrhizae are formed between the root hairs of Ericaceae plants and hyphae of Hymenoscyphus ericae and fungi from the genus Oidiodendron, e.g., O. griseum, O. maius, O. cerealia, and O. rhodogenum from the subclass Ascomycotina [10][11][12][13]. The roots of Ericaceae plants do not develop hairs, and the mycorrhizal mycelium is believed to take over their functions. The mycorrhiza stimulates the growth of the host plant via a better supply of such nutrients as nitrogen, phosphorus, and microelements like zinc and copper, and increases the resistance of plants to stress factors, increasing the intensity of photosynthesis [14][15][16][17][18][19]. Tested mycorrhizal vaccines stimulated the growth and number of leaves of the rhododendron plants [20]. Mycorrhizae influence better uptake of compounds unavailable to plants, making them available, especially of phosphorus [21][22][23]. It is assumed that mycorrhizal fungi supply 80% of plants with nitrogen and phosphorus. There are approx. 50,000 species of fungi forming mycorrhizal communities from approx. 250,000 plant species [24]. Significant increases in P, N, Fe, Zn and Cu concentrations have been noted in the mycorrhizal roots, even when plants were grown in a medium with a high concentration of CaCO 3 [25]. ERM fungi have the ability to decompose cellulose, hemicellulose, pectins and lignins. Due to their ability to produce proteases, they can release and utilize proteins as the only nitrogen source [26]. This trait is highly important for sustenance of vegetation in nutrient-poor habitats. Despite the growing popularity of Ericaceae plants, there is no specific information about the impact of fertilization and mycorrhizae on the growth and development of Ericaceae plants cultivated in nurseries. The aim of the study was to show the effect of the fertilization method and plant mycorrhization on the nutritional status of large-flowered azalea cv. 'Anneke' during the first three years of cultivation in containers in a nursery. The presented results are a continuation of the previously published work on the influence of fertilization and mycorrhizae on growth and development of rhododendron [27]. Materials and Methods The investigations of the large-flowered azalea were carried out in controlled conditions at the Lublin-Felin Experimental Station (51 • 13 36.9 N, λ = 22 • 37 56.8 E). The research material was the azalea (Rhododendron hybridum) cv. 'Anneke' representing the group of Knap-Hill-Exbury hybrids. The experiment was carried out in a two-factor design. The plants were grown in a 2.0 L container (1st study year) and in a 4.0 L container (2nd and 3rd study years). Each combination had eight replications. In vitro propagated cuttings were the starting experimental material. The plants were grown in a high-peat substrate with contents of N-min., P, K and Mg below 10 mg·L, an EC of 0.09 and pH 4.2. The effects of the following factors were assessed: Each of these fertilization methods provided the plants with the same amounts of nutrients: 0.52 g/L/year N, 0.10 g/L/year P, 0.38 g/L/year K, 0.05 g/L/year Mg and 0.90 g/L/year S. The other factor studied was the mycorrhization of the plants: mycelium-uninoculated plants as control (M-). In the first study year, half of each combination was treated with the mycorrhizal inoculation before the transfer of the plants into containers. The roots of these plants were soaked in an aqueous inoculum solution containing Oidiodendron fungi and Hymenoscyphus sp. (M+), and the other plants were not inoculated (M-). The investigations were carried out in an open area from April to November. The soil under the plants was covered with black nursery mat. Plant material samples (fully developed leaves from the middle part of a one-year shoot) were collected in the first ten days of August in each study year. The plant material was dried at 50 • C and ground. The following parameters were determined: Total N after combustion in concentrated H 2 SO 4 using the Kjeldahl method with the Kjeltec System 2002 Distilling Unit apparatus After incineration of the plant mass at 200 • C followed by dry combustion at 450 • C and cooling, the ash was supplemented with diluted hydrochloric acid in a ratio of 1:2. In this extract, the following parameters were determined: Phosphorus (P), colorimetrically with the vanadomolybdate method K, Ca, and Mg, with the ASA method (Perkin-Elmer, Analyst 300) [28] Substrate samples intended for chemical analyses were collected with a shortened Egner sampler from the containers in each study year in two terms: (I) in the first ten days of August, and (II) in the second ten days of October. The following parameters were determined in the 0.03 M CH 3 COOH extract at the solution-to-substrate ratio of 1:10: The results were analyzed statistically using analysis of variance for a two-factor experiment. Each year was analyzed separately with Tukey's test at the significance level of α = 0.05. Results and Discussion The assessment of the nutritional status of the large-flowered azalea in the first three years of cultivation was based on the content of total nitrogen, phosphorus, potassium, magnesium and calcium in plant leaves. The analyses detected average nitrogen content in the range of 12.5-23.5 g N-tot.·kg −1 d.m.) ( Table 1). Its lowest content was determined in plants subjected to fertigation, and the highest level was found in the slow-release fertilizer Hortiform pH (SRF) variant ( Figure 1). Moreover, there was a significant effect of the mycelium inoculation on nitrogen content, which was significantly higher in the mycelium-inoculated plants than in the uninoculated ones ( Figure 2). This correlation was confirmed in studies conducted by [29]. The available literature demonstrates a wide range of nitrogen nutrition of azaleas: 1.96-2.24% N [30], and 1.64% N [1]. In a study carried out by [31], a level of 1.88-2.20% N was shown to be the optimal range of nitrogen fertilization of azaleas. Aendekerk [1] reported symptoms of nitrogen deficiency only at a content of 0.67% N. The present study showed that fertilization with Hortiform pH ensured the optimal nitrogen status of the plants, with the content of the element in the range of 19.1-23.5 g N·kg −1 d.m. In turn, plants fertilized through fertigation contained lower levels of nitrogen (12.5-17.5 g N·kg −1 d.m.). Nevertheless, no symptoms of nitrogen deficiency were observed in the plant leaves. The results were analyzed statistically using analysis of variance for a two-factor experiment. Each year was analyzed separately with Tukey's test at the significance level of α = 0.05. Results and Discussion The assessment of the nutritional status of the large-flowered azalea in the first three years of cultivation was based on the content of total nitrogen, phosphorus, potassium, magnesium and calcium in plant leaves. The analyses detected average nitrogen content in the range of 12.5-23.5 g N-tot.·kg −1 d. m.) ( Table 1). Its lowest content was determined in plants subjected to fertigation, and the highest level was found in the slow-release fertilizer Hortiform pH (SRF) variant ( Figure 1). Moreover, there was a significant effect of the mycelium inoculation on nitrogen content, which was significantly higher in the mycelium-inoculated plants than in the uninoculated ones ( Figure 2). This correlation was confirmed in studies conducted by [29]. The available literature demonstrates a wide range of nitrogen nutrition of azaleas: 1.96-2.24% N [30], and 1.64% N [1]. In a study carried out by [31], a level of 1.88-2.20% N was shown to be the optimal range of nitrogen fertilization of azaleas. Aendekerk [1] reported symptoms of nitrogen deficiency only at a content of 0.67% N. The present study showed that fertilization with Hortiform pH ensured the optimal nitrogen status of the plants, with the content of the element in the range of 19.1-23.5 g N ·kg −1 d. m. In turn, plants fertilized through fertigation contained lower levels of nitrogen (12.5-17.5 gN·kg −1 d.m.). Nevertheless, no symptoms of nitrogen deficiency were observed in the plant leaves. The phosphorus content in the azalea leaves was on average 2.0-3.6 g P ·kg −1 d. m. (Table 1). Significantly higher contents of this element were detected in the leaves of plants receiving single fertilizers (control) compared to those fertilized with SRF and through fertigation. A similar significant effect of mycelium inoculation on phosphorus content was demonstrated ( Figure 2) As reported by Aendekerk [1], the optimal phosphorus content is 0.37% P, whereas Kreij et al. [30] found 0.30-0.50% P in azalea leaves. Li et al. [32] found 0.08% P in plants fertilized without N increasing to 0.14% P in plants fertilized with 20 mm N, and the values depended on the type of container. All types of fertilization ensured optimal plant nutrition with phosphorus. Noteworthy is the gradual The phosphorus content in the azalea leaves was on average 2.0-3.6 g P·kg −1 d.m. (Table 1). Significantly higher contents of this element were detected in the leaves of plants receiving single fertilizers (control) compared to those fertilized with SRF and through fertigation. A similar significant effect of mycelium inoculation on phosphorus content was demonstrated ( Figure 2) As reported by Aendekerk [1], the optimal phosphorus content is 0.37% P, whereas Kreij et al. [30] found 0.30-0.50% P in azalea leaves. Li et al. [32] found 0.08% P in plants fertilized without N increasing to 0.14% P in plants fertilized with 20 mm N, and the values depended on the type of container. All types of fertilization ensured optimal plant nutrition with phosphorus. Noteworthy is the gradual increase in the phosphorus content of plants with years of cultivation. In the available literature, phosphorus content below 0.15% P is considered critical. No such content was recorded in these studies, and no such value was recorded in the present study. Ristvey et al. [33] suggested that only very low constant levels of phosphorus are needed to support the growth of young azaleas. In many fertilizers, the N: P ratio significantly exceeds plant requirements, and the optimal ratio of nitrogen to potassium in the substrate should be 20:1. The potassium content in the leaves of the large-flowered azalea cv. 'Anneke' was on average 5.7-8.8 g K·kg −1 d.m. in the mycelium-uninoculated variants and 6.7-7.4 g K·kg −1 d.m. in the inoculated plants. The highest potassium level was determined in the leaves of control plants and the lowest content was detected in plants fertilized through fertigation. Noteworthy are the lower differences in the potassium content in the inoculated plants ( Table 1). The standard potassium content in azalea leaves was estimated at 0.37% by Aendekerk [1] and 0.78-0.98% by Kreij et al. [30]. These results suggest a highly varied optimal range of potassium nutrition in these plants. The plants in the present study did not show symptoms of potassium deficiency, and the content of this element in the range of 5.0-9.0 g K·kg −1 d.m. should be regarded as optimal. The calcium content in the leaves of the large-flowered azalea cv. 'Anneke' ranged from 7.5 to 10.9 g Ca·kg −1 d.m. (Table 1). Its highest level was determined in plants fertilized with the slow-acting fertilizer (Hortiform pH). It was slightly lower in the fertigation variant, and lowest in the single-fertilizer treatment (control) (Figure 1). Moreover, mycelium inoculation contributed to a significant increase in the calcium content in the plants (Figure 2). The standard calcium content in azalea leaves was shown to be 1.60-2.00% by Kreij et al. [30], 0.72-1.39% by Aendekerka [1] and 0.60-1.20% by Michałojć and Koter [31]. These data indicate a very wide optimal range of calcium nutrition in plants. Our observations of the plants carried out during the growing season did not reveal any symptoms of calcium deficiency. The content of magnesium in the leaves of the large-flowered azalea cv. 'Anneke' was in the range of 1.5-2.6 g Mg·kg −1 d.m. There was a slight variation in the magnesium content between the study years ( Table 1). The fertilization methods and mycelium inoculation had no significant effect on the magnesium content in the plants (Figures 1 and 2). The standard magnesium content in azalea leaves were estimated at 0.60% by Aendekerk [1], 0.17-0.33% by Kreij et al. [30] and 0.14-0.25% by Michałojć and Koter [31]. In turn, the critical magnesium contents in Ericaceae plants were 0.07% and 0.09%, as reported by Kreij et al. [30] and Aendekerka [1], respectively. The fertilization applied in the present study ensured optimal magnesium level in the plants. The assessment of the effect of the mycorrhization inoculum with Oidiodendron and Hynemoscyphus sp. fungi on the nutritional status of the large-flowered azalea showed similar levels of the analyzed nutrients in the inoculated and uninoculated plants in the first cultivation year. As shown in earlier studies, Michałojć et al. [27] mycorrhizal frequency on ERM inoculated roots of rhododendron ranged from 36% to 65%, and without inoculation it varied from 4% to 9%. There was a higher frequency of ERM on the roots of plants fertilized with the slow-release fertilizer (SRF) than nourished by fertigation and with monocomponent fertilizers. A beneficial effect reflected in higher nitrogen, phosphorus, potassium, calcium and magnesium content was noted in the second and third years of cultivation of the 'Anneke' plants. Moreover, significantly higher levels of nitrogen, phosphorus and calcium were determined in the leaves of the mycelium-inoculated plants, whereas there was no clear effect on the content of potassium and magnesium (Table 1). A study conducted by Nowak [34] reported no significant effect of mycorrhizae on the content of nitrogen, phosphorus, potassium, and calcium, but a higher level of magnesium was noted. In turn, Smith and Read [14] and Konieczny and Kowalska [18] confirmed the beneficial effect of mycorrhizae on the phosphorus and nitrogen content in plants. The same amounts of nutrients were supplied to the plant substrate in all combinations The analyses of the substrate were conducted to determine nutrient richness in the middle of the growing season (1st ten days of August) and after the growing season (second ten days of October). The present results demonstrated high nitrogen utilization by plants already in the middle of the growing season, and traces of this element after its end (Figure 3). In the analyzed period, the reduced content of mineral nitrogen was determined in the substrate of azaleas with mycorrhizal-inoculated roots (Figure 4). This proves that the inoculated plants took up nitrogen from the substrate more efficiently, as evidenced by the higher nitrogen content in the inoculated plants. Numerous studies indicate a beneficial effect of mycorrhizal fungi on nitrogen uptake from the substrate [36][37][38][39]. The highest content of Nmin. in the substrate was found in the slow-release fertilizer Hortiform pH variant, whereas the lowest value was determined in the fertigation combination. These results indicate that slow-release fertilizers successfully supply nitrogen and ensured an appropriate degree of plant nutrition with this element. The fertilization of azaleas through fertigation requires The present results demonstrated high nitrogen utilization by plants already in the middle of the growing season, and traces of this element after its end (Figure 3). In the analyzed period, the reduced content of mineral nitrogen was determined in the substrate of azaleas with mycorrhizal-inoculated roots (Figure 4). This proves that the inoculated plants took up nitrogen from the substrate more efficiently, as evidenced by the higher nitrogen content in the inoculated plants. Numerous studies indicate a beneficial effect of mycorrhizal fungi on nitrogen uptake from the substrate [36][37][38][39]. The highest content of N-min. in the substrate was found in the slow-release fertilizer Hortiform pH variant, whereas the lowest value was determined in the fertigation combination. These results indicate that slow-release fertilizers successfully supply nitrogen and ensured an appropriate degree of plant nutrition with this element. The fertilization of azaleas through fertigation requires further research. A study conducted by Michałojć et al. [27] demonstrated the lowest growth and development of shoots after application of fertigation, whereas the present study showed the lowest content of nutrients in plant leaves. In the substrate used for cultivation of azaleas, higher content of available phosphorus was noted after the application of single fertilizers (control), which indicates that the plants absorbed it at a slower rate (Figure 3). Additionally, its lower content was determined in the substrate of plants treated with the mycorrhizal inoculum, compared with the uninoculated substrate, which indicates that the plants utilized the element more efficiently (Figure 4). Noteworthy is that trace amounts of this component were found in the first and second analysis terms. This proved phosphorus uptake from the substrate by plants, as confirmed by its higher content in the plant leaves. Mycorrhizal fungi have been reported to enhance phosphorus uptake by plants [37]. A higher potassium level in the substrate was recorded in both periods after the application of single mineral fertilizers, whereas the lowest content was found in the fertigation variant (Figure 3). Moreover, mycorrhization resulted in higher utilization of potassium by the plants; hence, its lower content in the substrate from the cultivation of the inoculated plants (Figure 4). The application of mycorrhizae has a positive effect on the uptake of macroelements, including potassium, by plants [40]. This was fully confirmed in the present study. Calcium is an element regulating soil/substrate pH in addition to its physiological functions in plants. In the first analysis term in the present study, the highest calcium In the substrate used for cultivation of azaleas, higher content of available phosphorus was noted after the application of single fertilizers (control), which indicates that the plants absorbed it at a slower rate (Figure 3). Additionally, its lower content was determined in the substrate of plants treated with the mycorrhizal inoculum, compared with the uninoculated substrate, which indicates that the plants utilized the element more efficiently (Figure 4). Noteworthy is that trace amounts of this component were found in the first and second analysis terms. This proved phosphorus uptake from the substrate by plants, as confirmed by its higher content in the plant leaves. Mycorrhizal fungi have been reported to enhance phosphorus uptake by plants [37]. A higher potassium level in the substrate was recorded in both periods after the application of single mineral fertilizers, whereas the lowest content was found in the fertigation variant ( Figure 3). Moreover, mycorrhization resulted in higher utilization of potassium by the plants; hence, its lower content in the substrate from the cultivation of the inoculated plants ( Figure 4). The application of mycorrhizae has a positive effect on the uptake of macroelements, including potassium, by plants [40]. This was fully confirmed in the present study. Calcium is an element regulating soil/substrate pH in addition to its physiological functions in plants. In the first analysis term in the present study, the highest calcium content in the azalea substrate was shown for the slow-release fertilizer Hortiform pH combinations. Slightly lower levels were determined after the application of single fertilizers, and the lowest values were detected in the fertigation variant. In turn, the calcium content in the substrate in the second analysis term was more homogeneous (Figure 3). As in the case of the nutrients discussed above, a lower level of the element was found in the substrate of the mycorrhiza-inoculated plants compared to the uninoculated combinations ( Figure 4). This proves the beneficial effect of mycorrhizal fungi on calcium uptake from the substrate. Similar findings have been reported in other studies [40]. Regardless of the term of the analyses, the highest magnesium content in the substrate was recorded in the fertigation combinations and the lowest value was noted in control variants ( Figure 3). In turn, in the middle of the growing season (term I), the content of magnesium in the substrate was similar, whereas lower amounts of the element were determined in the substrate of the inoculated plants ( Figure 4). Studies conducted by [37,40] indicated a beneficial effect of mycorrhizal fungi on the uptake of magnesium by plants. The presence of sulfur in the substrate of the analyzed plants derived from the sulfate fertilizers applied. Throughout the study, the content of the element in the middle of the growing season ranged on average from 80 mg to 138 mg S-SO 4 ·L −1 . After the growing season (term II), its content was substantially lower (18-62 mg S-SO 4 ·L −1 ). The highest sulfur level was demonstrated in the Hortiform pH-fertilized combinations, and the lowest amount was noted in the fertigation variant ( Figure 3). Higher efficiency of sulfur utilization was noted in the mycorrhizal-inoculated combinations, as evidenced by its lower content in the substrate ( Figure 4). As specified by Komosa [41], the standard sulfur content in the substrate for ornamental plants with low nutritional requirements was established at 30 mg to 50 mg·S-SO 4 L −1 . Studies conducted by [37,40] also confirmed the beneficial effect of mycorrhiza on the uptake of this nutrient by plants. Azaleas have specific requirements in terms of soil/substrate reaction. Peat with a pH value of 4.6 was used as a substrate in the study. During the vegetation period, the reaction of the substrate fluctuated from 4.09 to 5.79 in the individual study years ( Table 2). The optimal reaction of the substrate is in the range of pH 4.5-5.8 [35]. The fertilization methods applied in the study ensured the optimum pH of the substrate in the azalea cultivation. Cultivation of plants in containers is associated with a risk of excessive salt concentration in the substrate. The present experiments showed that the doses and methods for application of the fertilizers did not elevate salt concentrations in the substrate. The salt concentration in the azalea substrate during the study period was on average 0.47 mS·cm −1 in the middle of the growing season and 0.38 mS·cm −1 at the end ( Table 2). The acceptable salinity (EC) for ornamental plants with low and medium nutritional requirements ranges from 0.80 to 1.10 mS·cm −1 [41] Furthermore, the presence of mycelium in the plant root zone contributes to elimination of excessive salt concentration in the substrate and, consequently, its negative impact on plants [42]. 1. As indicated in the first three years of azalea cultivation in containers, single mineral fertilizers and Hortiform pH (SRF) ensured an appropriate plant nutritional status in terms of the content of nitrogen, phosphorus, potassium, calcium and magnesium, while fertigation required further improvement of the nutrient solution. 2. Plant mycorrhization exerted a beneficial effect on plant nutrition, as higher contents of macro elements in plants were demonstrated in the second and third years of azalea cultivation. 3. The nutrients applied were largely utilized by the plants by the middle of the growing season, and almost complete utilization of the elements was noted at the end of the growing season. Therefore, the nutrients need to be supplemented in each subsequent year of vegetation.	2021-07-26T00:05:55.973Z	2021-06-10T00:00:00.000	{ "year": 2021, "sha1": "2eb6e29251e7e630728819bbfff77a734c59e1b2", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2077-0472/11/6/538/pdf", "oa_status": "GOLD", "pdf_src": "ScienceParsePlus", "pdf_hash": "952fb67282de0307b6b481b844bf705fca690c0b", "s2fieldsofstudy": [ "Agricultural and Food Sciences" ], "extfieldsofstudy": [ "Biology" ] }
18268258	pes2o/s2orc	v3-fos-license	Intraductal Lipid-Rich Carcinoma of the Breast with a Component of Glycogen-Rich Carcinoma We report a rare case of intraductal lipid-rich carcinoma of the breast with a component of glycogen-rich carcinoma. An impalpable tumor that was revealed by mammography and magnetic resonance imaging was excised. Histologic examination showed vacuolated neoplastic cells in the mammary ducts, and electron microscopy confirmed lipid droplets in the cytoplasm. The coexistence of glycogen-rich carcinoma was shown. Lipid-rich carcinoma that is coexistent with glycogen-rich carcinoma is rare, and most lipid-rich carcinomas are invasive. Intraductal lipid-rich carcinoma is difficult to detect without echography or mammography. INTRODUCTION Normal epithelial cells of the mammary gland are able to synthesize lipids, carbohydrates, and proteins. Breast cancer is histologically classified into subtypes according to the substances that the carcinoma cells contain-such as lipid-rich carcinoma and glycogen-rich carcinoma, while the types of carcinoma are mainly ductal carcinoma and lobular carcinoma. Aboumrad et al. [1] reported the first case of mammary carcinoma that contained lipids in the cytoplasm, which was diagnosed as lipid-secreting carcinoma. Lipid-rich carcinoma is a rare mammary cancer, and glycogen-rich carcinoma is also relatively rare. To the best of our knowledge, only 1 other case of invasive mammary carcinoma with composite histologic findings of these 2 carcinomas has been described [2]. We encountered a case of intraductal carcinoma with the coexistence of these 2 rare subtypes of mammary carcinoma, and performed a histological and ultrastructural examination of the lesion. CASE REPORT A Japanese woman in her fifth decade visited our hospital in order to be examined and medically treated for multiple small linear calcifications that were found in her left breast by mammography. No tumor was detected on palpation. The results of mammography, echography, and magnetic resonance imaging (Figure 1), suggested the presence of malignancy. On aspiration cytology, no cells were obtained. Breast-conservation surgery was performed. The tumor was macroscopically indistinct and had a diameter of 30 mm. The intraductal carcinoma was extended, and invasion to the surrounding tissue was not found (Figure 2). Small foci of linear calcification and necrosis were seen in the ducts that were occupied by tumor cells, and these exhibited a comedo pattern. There were no abundant secretory materials. In three-fourths of the lesions, the tumor cells had clear and abundant cytoplasm and compressed crescent-shaped or oval nuclei ( Figure 3). The cytoplasm was vacuolated. Among these cells, some had atypical large nuclei with distinct nucleoli, and some had small round nuclei and modest eosinophilic cytoplasm-so-called "fried-egg cells" (Figure 4A). Each of the components presented different characteristics. The former were negative for periodic acid-Schiff stain (PAS), and the latter were positive. After treatment with diastase, the cells were negative for PAS ( Figure 4B, C). Immunohistochemically, both types of cells were diffusely positive for cytokeratin (AE1/AE3) and focally positive for gross cystic disease fluid protein-15 (GCDFP-15), and were negative for vimentin, S-100 protein, smooth muscle actin and CD10. The carcinoma cells of both lipid-rich carcinoma and glycogen-rich carcinoma are immunohistochemically positive for estrogen (ER) and progesterone receptors, and negative for HER2. An ultrastructural study revealed lipid droplets in the cytoplasm of the vacuolated cells ( Figure 1). However, glycogen particles were not shown. No nodal metastasis was found. DISCUSSION Lipid-rich carcinoma is a rare subtype of breast cancer that is histologically characterized by cells with numerous large and small vacuoles in their cytoplasm. The cells of lipid-rich carcinoma are divided into the following 3 forms: the histiocytoid type, the sebaceous type, and the type with apocrine extrusion of nuclei [3]. Two or 3 of these types often exist together, as in this case, where the histiocytoid and sebaceous types were observed. Fat staining of cryostat sections revealed the presence of a large amount of lipid within the cytoplasm. Lipid accumulation was also shown in an ultrastructural study. Nevertheless, the definition of lipid-rich carcinoma is controversial because it is unclear as to what percentage of vacuolated cells and which type of origin of the lipid vacuoles should confirm the diagnosis [4]. In our case, about 75% of the tumor cells had lipid-rich cytoplasm. Ramos and Taylor [5] have shown by electron microscopy that lipids originate from a secretory product of the neoplastic cells, and they do not consist of a degenerative material. They described the secretory vacuoles as being close to a markedly prominent Golgi apparatus, and the absence of autophagic vacuoles and the presence of prominent rough endoplasmic reticulum supported this speculation. Vera-Sempere and Llombart-Bosch [6] suggested a similar origin of the secretory product. However, Wrba et al. [7] reported no immunohistochemical or ultrastructural findings that could support the lipid secretion of carcinoma cells in their cases. Therefore, in their opinion, the term lipid-rich carcinoma, rather than lipid-secreting carcinoma, should be used, unless there is evidence of active lipid secretion. Neoplasms of the breast that have clear cytoplasm occur in a variety of diseases, such as lipid-rich carcinoma, glycogenrich carcinoma, apocrine carcinoma, secretory carcinoma, clearcell myoepithelioma, and clear-cell adenomyoepithelioma. In addition, soft-tissue neoplasms of the breast and chest wall, for example, clear-cell sarcoma, clear-cell rhabdomyosarcoma, and alveolar soft-part sarcoma, also need to be discriminated [8,9]. Skin tumors with clear cytoplasm must be distinguished too. Metastatic renal cancer of the clear-cell type of carcinoma has been reported, but they are highly rare [10]. In breast carcinomas, PAS stain and immunostaining are useful in the differential diagnosis. For example, glycogen-rich carcinoma and apocrine carcinoma are positive for PAS, and lipid-rich carcinoma is negative. Glycogen granules in glycogen-rich carcinoma become negative for PAS after diastase digestion; on the other hand, granular cytoplasm of apocrine carcinoma keeps positive for PAS after that treatment. Positive immunostain for the S-100 protein has been reported to be useful in the diagnosis of lipid-rich carcinoma [11], although in our case, immunostaining for this protein was negative; however, this is not useful in the differentiation between lipid-rich carcinoma and myoepithelial tumors. Immunostaining for smooth muscle actin, which is reactive in myoepithelial cells, is helpful in such cases. S-100 protein is frequently expressed in secretory carcinoma, too. But, most of secretory carcinomas are negative for ER. Soft-tissue tumors that have clear cytoplasm are ruled out by immunoreactivity for vimentin and negative stains for AE1/ AE3 and GCDFP-15. GCDFP-15 has been reported to be useful in differentiating between primary breast carcinoma and metastatic carcinoma [12]. Metastatic clear cell renal cell carcinoma is positive for vimentin and CD10, and negative for GCDFP-15. In addition, GCDFP-15 is useful in the differential diagnoses among some primary breast carcinomas. Apocrine carcinoma indicates diffuse and intensive positive for GCDFP-15, but secretory carcinoma indicates negative for it. A lesion with a combination of lipid-rich carcinoma and glycogen-rich carcinoma was reported by Kovacs and Krutsay [2]. In the present lesion, some of the carcinoma cells had glycogen in their cytoplasm. These cells had the same histological characteristic features as glycogen-rich carcinoma in that the cells had round nuclei at the center of pale eosinophilic cytoplasm and were positive for PAS and sensitive to diastase digestion. Glycogen, which can be demonstrated by electron microscopy [13], was not revealed in this case. This may have been because the electron microscopy was performed on paraffinembedded specimens. Both lipid-rich carcinoma and glycogen-rich carcinoma are the diagnoses that are determined by the metabolic products in the cytoplasm, and these are different from the diagnoses that are determined by the cellular types. However, diagnoses must be determined on the basis of not only examination of the intracytoplasmic products but also the histologic findings, because a varied amount of lipid and glycogen accumulation in cytoplasm was found in 30% and 85%, respectively, of mammary carcinomas [14,15]. Glycogen has also been demonstrated in mammary squamous cell carcinoma. Histologic findings of clear, foamy, or vacuolated cytoplasm are important for the diagnosis of lipid-rich carcinoma, and the findings of a fried-egg appearance are important for the diagnosis of glycogen-rich carcinoma. Clinically, early detection of mammary lipid-rich carcinoma is presumed to be difficult. Except for a few reports that did not mention whether the carcinoma was invasive or intraductal, all reported cases of lipid-rich carcinoma were invasive. The ratio of intraductal carcinoma to invasive carcinoma in lipidrich carcinoma was lower than that in ductal carcinoma. This is not only because of the low incidence of lipid-rich carcinoma but also because of the difficulty in detecting it. In the present case, the tumor was not palpable. Calcifications that were detected by medical examination led to the early discovery in this case.	2014-10-01T00:00:00.000Z	2012-03-01T00:00:00.000	{ "year": 2012, "sha1": "022c8f310abdd01bbbcef8de2b4a5da262cc4890", "oa_license": "CCBYNC", "oa_url": "https://doi.org/10.4048/jbc.2012.15.1.135", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "022c8f310abdd01bbbcef8de2b4a5da262cc4890", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
10426847	pes2o/s2orc	v3-fos-license	Treatment of overactive bladder in the aging population: focus on darifenacin. Anticholinergics are commonly used in primary and secondary care settings for the treatment of overactive bladder syndrome. The number of anticholinergic drugs available on the market is increasing and various studies, both observational and randomized controlled trials, have evaluated effectiveness of the different preparations available. When anticholinergic therapy is prescribed, there is still uncertainty about which anticholinergic drugs are most effective, at which dose, and by which route of administration. There is also uncertainty about the role of anticholinergic drugs in different patient groups, particularly in the elderly. The rationale for using anticholinergic drugs in the treatment of overactive bladder syndrome is to block the parasympathetic acetylcholine pathway and thus abolish or reduce the intensity of detrusor muscle contraction. There are currently five recognized subtypes of muscarinic receptor; the M1, M2, and M3 subtypes are of interest in bladder activity. Muscarinic receptors are found in other parts of the body, eg, in the gut, salivary glands, tear ducts. Side effects associated with non-selective antimuscarinics can be particularly distressing in the elderly. The development of bladder selective M3 specific antagonists has the advantage of providing increased efficacy with minimal side effects. Darifenacin is one such preparation. The aim of this review is to assess the pharmacology, interactions and the safety and tolerability of darifenacin in the treatment of overactive bladder in the elderly population with particular reference to clinical trial data available. Introduction Overactive bladder (OAB) is a common condition characterized by urinary urgency, with or without urge incontinence usually with frequency and nocturia (Abrams et al 2002). Epidemiological studies from the US have reported the prevalence of OAB to be 16.9% (Stewart et al 2003); it can affect men and women alike, but the prevalence increases with age. OAB can have a detrimental effect on physical functioning and psychological well-being, as well as signifi cantly reducing quality of life (Abrams et al 2000). In patients under 25 the prevalence is 4.8% and in those over the age of 65 years 30.9% (Milsom et al 2001;Stewart et al 2003). In the elderly the commonest manifestation is urinary incontinence (Castleden et al 1981). Antimuscarinic drug therapy, in conjunction with behavioral therapy such as bladder retraining, remains the fi rst line of management in patients with irritative bladder symptoms or OAB. Several antimuscarinic agents are currently available for the treatment of OAB in adults. The antimuscarinics all appear to exert their clinical effect through inhibition of the bladder muscarinic receptors, but they vary in their functional profi le and can be associated with troublesome side effects such as dry mouth, constipation, somnolence, and blurred vision. In a questionnaire follow-up study of women with detrusor overactivity, just 5.5% were cured of their urinary symptoms Treatment of overactive bladder in the aging population: focus on darifenacin and only 18.2% of women continued drug therapy for more than 6 months (Kelleher et al 1997). The development of relatively bladder selective antimuscarinic drugs such as tolterodine has helped to reduce adverse effects, but the development of darifenacin and the recent marketing of solifenacin represent the fi rst highly selective bladder specifi c anti-muscarinic (M 3 specifi c antimuscarinics) agents for the management of OAB. While effi cacy of antimuscarinics has been demonstrated in adult populations (including patients >65 years of age), few studies have been reported specifi cally in a geriatric population, and antimuscarinics are often underutilized in the elderly despite the marked increase in the prevalence of OAB in this age group. One explanation for this apparent under-use of an effective treatment option may be concerns about the frequency of anticholinergic adverse events, such as dry mouth; the likelihood of detrimental central nervous system (CNS) effects, including cognitive impairment and sleep disturbances; and the potential for harmful interactions with existing pharmacotherapy. Muscarinic receptors and effects of aging Molecular cloning studies have revealed the existence of five genes for muscarinic receptors in rats and humans. This corresponds to five subtypes of muscarinic receptors, M 1 -M 5 (Caulfield and Birdsall 1998). The distribution of the different subtypes in the human body varies and so does their functional importance. Table 1 shows the distribution of different muscarinic receptors in the body. Detrusor muscle from various species including man contains M 2 and M 3 receptors. Although the M 2 receptors predominate (approximately M 2 :M 3 :2:1), in humans it is the M 3 receptors that are thought to cause a direct muscle contraction via phosphoinositide hydrolysis (Harriss et al 1995), whereas M 1 receptors have never been demonstrated (Yamaguchi et al 1996). It is thought that the M 3 receptor is responsible for the normal micturition contraction (Caulfi eld and Birdsall 1996). The M 2 receptors may become more important in mediating detrusor contractions in certain disease states such as neurogenic bladder dysfunction (Braverman et al 1998). Animal studies show there is no change in the relative abundance of M 2 and M 3 receptors with advancing age; this is accompanied by only minor if any alterations in receptor responsiveness (Schneider et al 2005). Pharmacology: darifenacin Darifenacin is a highly specifi c M 3 receptor antagonist which is used in the US and Europe, and is yet to be launched in the UK. It has been shown to have a higher degree of selectivity for the M 3 receptors (Tables 2 and 3) than the commonly used antimuscarinics currently available and has activity both in vitro and in vivo (Quinn 2006). It may be expected that drugs selective for the M 3 receptors in the bladder have greater clinical effi cacy in detrusor overactivity. It has also been shown to have a reduced adverse side effect profi le (Anderson 2006). Darifenacin was acquired by Novartis in May 2003 and is expected to be marketed as Enablex in the UK in the near future. Effects in vivo In bladder and salivation models of conscious rats, darifenacin has been compared with oxybutynin (Williamson et al 2006). The micturition interval and volume as well as the peak micturition pressure were assessed before and an hour after injection with the test compound (darifenacin 0.1-0.3 mg/kg; oxybutynin 0.1-3.0 mg/kg). Dose-related decreases were seen in all micturition parameters, which were significantly greater than those seen with oxybutynin (p<0.05), although the effect was equipotent on salivation. Effects on CNS The side effects of antimuscarinics on the CNS have been well documented. Adverse effects such as somnolence and cognitive impairment are a result of M 1 blockade in the brain. The role of M 1 receptors in cognitive function is well accepted, and is the basis of the therapeutic principle in the treatment of Alzheimer's disease (Anderson 2006). The elderly therefore are particularly at risk of cognitive side effects with non-specifi c antimuscarinics hence the reluctance to use them in the elderly population. To assess the effects of darifenacin controlled-release and oxybutynin extended-release (ER) on cognitive function (particularly memory) 150 patients ≥60 years old, were randomized to darifenacin, oxybutynin ER or placebo in a multicenter, double-blind, double-dummy, parallel-group, 3-week study . Doses were administered according to US labels: oxybutynin ER 10 mg once daily (od), increasing to 15 mg od in weeks 1 and 2, then 15 mg od in week 3. The primary end point was accuracy on the Name-Face Association Test (delayed recall) at week 3. While darifenacin had no signifi cant effects on memory versus placebo, oxybutynin ER caused signifi cant memory deterioration (magnitude of effect comparable with brain aging of 10 years). Similar studies in younger populations testing the effects of darifenacin on cognitive function at doses of 7.5 mg or 15 mg, did not produce any detectable effects relative to placebo . The sparing of cognitive function in relation to the use of darifenacin (and trospium) is an attractive result of the M 1 -sparing effect of darifenacin and is therefore potentially very useful in the elderly. Effects on the cardiovascular system and the electrocardiogram A study in the chioralose-urethane anesthetized beagle dog assessed the potency of darifenacin, tolterodine, oxybutynin, and propiverine on pelvic nerve-stimulated bladder contractions, trigeminal nerve stimulated salivation, and heart rate (Gupta et al 2006). Darifenacin did not increase the heart rate at doses required to maximally inhibit bladder contractility (100 μg/kg), whereas tolterodine increased heart rate at doses within the range needed to inhibit contractility. In the study by Nichols and colleagues (2006), darifenacin (7.5 mg and 15 mg) had no signifi cant effect on heart rate or heart rate variability, whereas dicyclomine (dicycloverine) signifi cantly reduced heart rate (−4.79 bpm, p=0.003) and increased variability (12%, p=0.005) compared with placebo. No changes in either the vital signs or electrocardiogram (ECG) were found to be due to darifenacin. Prolongation of QT interval on an ECG is a valuable predictor of a drug's ability to cause potentially fatal ventricular tachyarrhythmia. Serra and colleagues (2005) found there was no signifi cant increase in QT interval with darifenacin treatment compared with placebo. Pharmacokinetics The pharmacokinetics of darifenacin have been determined in human volunteers after oral administration of the drug and studied in animal models after oral and intravenous administration (Beaumont et al 1998). Pooled data have been used to model population pharmacokinetics (Kerbusch et al 2003). Plasma concentrations were calculated using (Anderson 2006) C-labelled darifenacin 5 mg dose, after 5 days of achieving a steady-state plasma concentration. Darifenacin is well absorbed in humans. Unchanged darifenacin is present in very small amounts in the feces, showing that it is almost completely absorbed from the gastro-intestinal tract (Beaumont et al 1998). Compared with immediate release preparations, higher bioavailability has been noted with extended release preparations (Kerbusch et al 2003). This may be due to differing activity of the metabolizing mechanisms in the lower intestinal tract (Paine et al 1997). Food has no effect on absorption (Nichols 2006). Darifenacin is highly protein-bound in the plasma (>96% in all species tested; 98% in man) (Beaumont et al 1998). Volumes of distribution are higher than body water in the animal models after intravenous administration (12 L/kg, 7 L/kg, and 6.8 L/kg in mouse, rat, and dog, respectively). Darifenacin has an effi cient hepatic metabolism (Beaumont et al 1998) and this is demonstrated by plasma clearance values that are high relative to hepatic blood fl ow. This leads to a short half-life in man. Darifenacin has three principal mechanisms of metabolism: monohydroxylation in either the dihydrobenzfuran or diphenylacetamide moieties; opening of the dihydrobenzfuran ring; or N-dealkylation at the pyrrolidine nitrogen. The main metabolite, UK-148993 (Beaumont et al 1998), is formed primarily through hydroxylation of darifenacin by cytochrome P4503A4 (CYP3A4) and cytochrome P4502D6 (CYP2D6) and is 50-fold less potent than the parent drug (Kerbusch et al 2004). There is a great deal of polymorphism in CYP2D6, and individuals with poor CYP2D6 metabolism may have higher plasma levels of the drug, with 20% reduction in clearance of the drug, and 52% increase in bioavailability (Kerbusch et al 2003) when compared with homozygote extensive metabolizers. Japanese subjects have a 56% lower availability of darifenacin than Caucasians (Kerbusch et al 2003). No other racial differences in bioavailability were seen, indicating that the difference is due to increased fi rst pass metabolism in the liver (by CYP3A4 and CYP2D6). Distribution of alleles pertaining to CYP2D6 is known to differ between Japanese and Caucasian races (Xie et al 2001), but when the CYP2D6 activity was controlled for separately, and knowing that no ethnic differences seem to exist in CYP3A4 activity (Xie et al 2001), it appears that other as yet unidentifi ed factors are involved. Women have a 31% lower rate of clearance of darifenacin than men (Kerbusch et al 2003), which is felt to be due to a lower activity of CYP2D6 in women (Tanaka 1999) and which would seem therefore to draw parallels with the CYP2D6 poor-metabolizers previously alluded to (despite a higher activity of CYP3A4 in women [Tanaka 1999]). Biotransformation by the cytochrome P450 (CYP450) system is an important step in the activation or elimination of a large number of drugs, including oxybutynin, tolterodine, darifenacin, and solifenacin, raising the possibility of clinically relevant and potentially serious drug interactions. In elderly patients, such interactions are of particular relevance given the potential for declining activity of certain members of the CYP450 family combined with decreased hepatic blood fl ow, which can reduce fi rst-pass metabolism and thus the bioavailability of drugs metabolized via this route. After administration of 14C-labelled darifenacin, most of the radioactive dose was recovered within 48 h, balanced between urine and feces in humans (Beaumont et al 1998). Approximately 44% of a dose was recovered from the feces and 58% from the urine, in both cases predominantly as metabolites. Drug interactions Data assessing drug and food interactions of darifenacin are relatively scarce. Kerbusch and colleagues (2003) have looked at the effect of ketoconazole (an azole used in the treatment of fungal conditions) and erythromycin (a macrolide antibiotic) on the metabolism of darifenacin. Both of these agents are CYP3A4 inhibitors, although ketoconazole is 50 times more potent than erythromycin. Use of ketoconazole reduced clearance of darifenacin by 68% (Kerbusch et al 2003) and increased the bioavailability of the drug to 100%. Erythromycin had no effect on clearance but increased bioavailability to 97%. Ketoconazole and erythromycin also affected levels of the active darifenacin metabolite, although this was thought to be due to a reduction in metabolism from the parent drug, rather than an increase in clearance of the metabolite. Caution is therefore warranted in the use of darifenacin in patients taking CYP3A4 inhibitors. Potent inhibitors should be avoided in view of their combined effect on clearance and metabolism. Higher doses of darifenacin are generally well tolerated, and so modest increases in drug concentration with less potent inhibitors are not clinically relevant. Effi cacy studies Darifenacin has been the subject of several well conducted and extensive trials with rigorous entry criteria. Analysis of pooled data from three such studies enrolled 1059 patients (85% of whom were female; age range 19-88 years) with at least 6-month history of overactive bladder symptoms (urgency, frequency, and urge incontinence) (Chapple et al 2005). All had a 2-week washout and a 2-week placebo run-in, prior to commencing active participation. Participants were randomized to receive either darifenacin-controlled release 7.5 mg or matched placebo (n=335/271, respectively) or darifenacin controlled release 15 mg or matched placebo (n=330/384, respectively). Outcome data were collected by use of an electronic diary (hand-held computerized diaries have been developed to overcome the lack of patient compliance that has been noted in many studies [Rabin et al 1993]; patients felt that their symptoms were more properly refl ected by the computerized version, felt more motivated to provide data, and found it easier to remember to enter data). Both doses of darifenacin were signifi cantly superior to placebo in alleviating the symptoms associated with OAB (Table 4), however, a marked placebo response was noted. Darifenacin (15 mg controlled release) has also been compared with oxybutynin (5 mg TDS) and matched placebo (Zinner et al 2005). Seventy-six patients (93% female) had detrusor overactivity (urodynamically verifi ed OAB) with ≥4 urinary incontinence (UI) episodes per week and a frequency of >8 micturitions per day; all Darifenacin in the elderly underwent a 2-week run-in period prior to 2 weeks of treatment in each arm, in a randomized sequence. Outcomes were assessed using a paper diary ( Table 5). The results revealed comparable effi cacy with oxybutynin, in terms of signifi cant improvement of the major symptoms of OAB, when compared with placebo. Rapid onset of action of darifenacin is a useful attribute and has been assessed in a well conducted study (Haab et al 2004). In this study, 561 patients (age range 19-88 years; 85% female) were enrolled with OAB symptoms of more than 6-month duration. After a 2-week washout and a 2week placebo run-in, the participants were randomized in a 1:4:2:3 ratio to 3.75 mg darifenacin, 7.5 mg darifenacin, 15 mg darifenacin, or placebo. Significant and rapid onset of action was noted by 2 weeks, although evolving benefit was seen up to 12 weeks in this study. Benefits to patients taking darifenacin have been shown, not only in terms of symptom improvement but also by demonstrable improvements in quality of life, as assessed by the King's Health Questionnaire (Chapple 2006). The low incidence of cognitive side effects has been discussed as a benefit in the elderly. Efficacy in an elderly population has been studied by Foote and colleagues (2005), by analysis of pooled data. This included 317 patients aged 65 years or more, in one of three multicenter, randomized, doubleblind, placebo controlled studies. All had OAB symptoms of more than 6-month duration. All had a 2-week washout (only if needed) and a 2-week drug-free/placebo runin, prior to receiving once daily orally administered controlled-release darifenacin 7.5 mg (n=97) and matching placebo (n=72), or controlled-release darifenacin 15 mg (n=109) and matching placebo (n=108). At 12 weeks, both doses of darifenacin were significantly superior to placebo (p<0.05) in improving all OAB parameters studied, confirming efficacy in an elderly population (Table 6). In a comparison with tolterodine immediate-release (Foote 2006), dry mouth was the commonest side effect (darifenacin 7.5 mg 20.2%; darifenacin 15 mg 35%; tolterodine 26.9%; placebo 8%) in all groups. Constipation (darifenacin 7.5 mg 14.8%; darifenacin 15 mg 21%; tolterodine 10.3%; placebo 5.4%) and dyspepsia (darifenacin 7.5 mg 1.8%; darifenacin 15 mg 7.5%; tolterodine 7.6%; placebo 1.5%) were also reported. Dizziness and somnolence were reported, but the numbers were very small and no signifi cant differences were detected in the different groups. All reported side effects are transient and tolerable. Even though constipation is one of the commonest side effects, the use of laxatives and stool softeners is low and comparable with their use in the placebo groups. Treatment-related adverse events infrequently lead to withdrawal . Studies to date have excluded patients with hepatic disease (Haab et al 2004) in keeping with the nature of the metabolism of darifenacin. However, pharmacokinetic analyses have shown a 4.7-fold increase in drug concentrations with moderate hepatic impairment (Croom and Keating 2004). If being used, dose adjustment of darifenacin would be necessary in this group of patients. No clinically signifi cant change in dose concentrations arose as a result of mild hepatic impairment, or renal impairment of varying degrees (Croom and Keating 2004), and so no dose adjustment is necessary. Patient perspectives Urinary incontinence is an important contributor to the complications and economic cost of OAB for both community-dwelling and institutionalized elderly individuals. Many patients with OAB do not seek treatment because of embarrassment, fear of surgery, or the misperceptions that the problem is untreatable, or is a normal and inevitable consequence of aging. Nonpharmacological therapies improve bladder control by modifying lifestyle and behavior to prevent urine loss. This requires patient and caregiver motivation and can be time consuming. However optimal results may be obtained by combining these strategies with pharmacotherapy or by means of pharmacotherapy alone. Important considerations for elderly patients include tolerability, absence of drug interactions, and the availability of a range of dosages to tailor treatment to individual patients. As demonstrated in this review of literature, darifenacin favorably fulfi lls these criteria. These results are supported by signifi cant improvements in quality of life, which have paralleled the overactive bladder symptom reductions (Haab 2005;Chapple 2006). Both fi xed and fl exible darifenacin dosing regimens produce these benefi cial effects, which extend to the more vulnerable population of older patients. There may also be a place for incorporating patientinitiated dose adjustment into the protocol MacDiarmid 2006a, 2006b). The primary care physician can then effectively manage adverse events associated with OAB without compromising effi cacy. Given some control in the process, patients are willing to tolerate certain side effects in exchange for symptom relief. Conclusion Evidence would suggest that darifenacin is an antimuscarinic drug particularly useful in the elderly population and therefore a useful addition to the current range of medications available on the market. Being an M 3 -specifi c antagonist, it is part of an important advance in the development of drugs for the treatment of OAB. Data from clinical trials have shown that darifenacin is effi cacious and well tolerated in addition to providing improvements in lower urinary tract symptoms with minimal CNS side effects. It also results in signifi cant improvements in quality of life scores. The dosage of drug also represents a compromise between effi cacy and tolerability, and thus, it is important, particularly in the clinical setting, to have a range of doses that may then be tailored to patients' individual needs. Darifenacin has been compared with immediate-release tolterodine, and further studies are warranted to compare the drug with the newer QD preparations of both oxybutynin and tolterodine to examine whether the theoretical advantages of an M 3 -specifi c antagonist translate into clinical practice. Compliance with medical therapy remains a significant problem in the treatment of the overactive bladder. When selecting an antimuscarinic agent for the management of an elderly patient presenting with OAB, in addition to considering evidence of clinical effi cacy and tolerability, issues of safety specifi c to an older population should be borne in mind. In particular, the likelihood of detrimental CNS effects should be considered, including cognitive impairment and sleep disturbances, secondary to anticholinergic load. Oxybutynin and tolterodine have both been associated with cognitive adverse events and effects on sleep architecture and quality. In contrast, darifenacin does not appear to be associated with cognitive adverse events and does not negatively affect sleep architecture or quality though it is just as if not more effi cacious . Darifenacin in the treatment of the OAB represents a pharmacological advance as one of a new type and class of drug. This new addition to the armamentarium available for the management of OAB is welcome to both clinician and patient alike, and the evidence so far would suggest that darifenacin should have a major effect in this diffi cult-tomanage condition.	2014-10-01T00:00:00.000Z	2006-12-01T00:00:00.000	{ "year": 2006, "sha1": "40186e6af20f076dd8d3b2fce577d4e418daf28c", "oa_license": "CCBYNC", "oa_url": "https://doi.org/10.2147/ciia.2006.1.4.309", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "44002615df9344474d8df4347184bbcfea343664", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
55596542	pes2o/s2orc	v3-fos-license	IP-10 AS A PREDICTOR OF TREATMENT RESPONSE IN CHRONIC HEPATITIS C PATIENTS Nabil Omar 1 , Khalid Gamil 1 , Mohamed Abdel-Samiee 1 and Mary Naguib 2 . 1. Hepato-Gastroenterology, National Liver Institute, Menoufia University, Shebin El-Kom, Menoufia, Egypt. 2. Clinical and Chemical Pathology department, National Liver Institute, Menoufia University, Shebin El-Kom, Menoufia, Egypt. ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History ISSN: 2320-5407 Int. J. Adv. Res. 5(3), 752-766 754 A rapid virological response (RVR) was defined as undetectable HCV RNA in serum at week 4 of therapy. Early virological response (EVR) was defined as serum HCV RNA negativity or any >2 log10 decline in HCV RNA levels at week 12 of therapy compared with baseline. Patients with sustained virological response (SVR) were those with undetectable HCV RNA in serum 24 weeks after stopping therapy. Treatment failures included patients who had a < 2 log10 drop in viral load at week 12 as compared to baseline, those whose HCV RNA was still detectable at week 24 (i.e. nonresponders), and those who had undetectable HCV RNA at the end of therapy but detectable HCV RNA at 24 weeks after cessation of therapy (i.e. relapsers) (Ghany et al., 2009). Laboratory Assessment:-Base line investigation; blood samples were taken before starting treatment. They were tested for: HCV RNA using the COBAS® TaqMan® HCV Quantitative Test (Roche Molecular Diagnostics, CA, USA) with lower limit of quantitation of 15 IU/ml; complete blood count (CBC) using Sysmex K-21 (Sysmex Corporation, Japan); blood sugar, liver & kidney function tests using fully automated Beckman Coulter chemistry analyzer Au 480 (Beckman Coulter Inc., CA, USA); alpha fetoprotein (AFP) & TSH hormone using (Advia Centaur CP immunoassay system, Siemens Healthcare Diagnostics Inc, USA) Repeated measurements of liver function tests, complete blood count and HCV RNA were done on weeks (12, 24, and 48) during treatment and at 6 months after stopping therapy. End Points:-SVR response was considered as primary end point. Non-response and/or relapse were considered as the secondary end point. Liver biopsy:-Ultrasound-guided percutaneous liver biopsy was performed to determine the degree of liver fibrosis at baseline. The histopathological assessment of necro-inflammatory grade and fibrosis stage was recorded using Metavir scoring system (Fibrosis was staged on a 0-4 scale. Activity was graded as: A0, none; A1, mild; A2, moderate; and A3, severe. Fibrosis stages ≥ F2 were considered significant) before commencing treatment (Bedossa et al., 1996). Radiological investigations:- All patient had abdominal ultrasonography; liver stiffness measurement (LSM) by transient Elastography using fibroscan apparatus ((Echosens, Paris, France) done initially within a week of liver biopsy and at the end of followup period (Andersen et al., 2009). Statistical analysis:-Data was statistically analyzed using IBM® SPSS® Statistics® version 21 for Windows. Data are expressed as mean ±standard deviation, number with column percentage and the median ± range or 95% confidence interval for non-parametric data. All p-values are 2 tailed, with values <0.05 considered statistically significant, p =0.01 is highly significant and p =0.001 is very highly significant. Comparisons between two groups were performed using the Student's t-test for parametric data ''normally distributed", and Mann-Whitney test for nonparametric data ''not normally distributed". CHI-squared test (χ2) and Fisher exact test for categorical data analysis. Regression analysis is used to find how one set of data relates to another. Univariate and multivariate binary logistic regression were done for detecting the predictors of the event. The area under the receiver operating characteristic (AUROC) curve analysis was used for detection of the cutoff value of the proposed tests. An AUROC value of 0.90-1.0 indicated excellent, 0.80-0.89 good, 0.70-0.79 fair, 0.60-0.69 poor and 0.50-0.59 no useful performance for discrimination of the outcome under assessment. Sensitivity: the proportion of patients with disease who are correctly identified. Specificity: the proportion of patients without disease who are correctly identified. Positive predictive value: the proportion of patients with positive test results who are correctly diagnosed. Negative predictive value: the proportion of patients with negative test results who are correctly diagnosed. Results:- This was a prospective randomized controlled study which was conducted on 70 chronic hepatitis C patients and 10 healthy persons enrolled as a control group. They were recruited from the outpatient clinic of Hepatology department, National Liver Institute, Menoufia University. Comparison between control group and treated group:-There was statistically significant difference between both groups regarding Albumin ( The mean ± SD of transient elastography was (8.9±4.8) kpa in the treated group patients before starting treatment. The base-line HCV RNA level was (1548683.5±1955482.9) IU/ml in treated group (table 1, Figure 1) Comparison between sustained vs. non sustained virus responders:-Sustained virus response (SVR) comprise 57.1% of chronic hepatitis C patients as 40 patients achieved SVR while 30 patients were Non-SVR achievers, who were further subdivided to 20 non responders who did not achieve the end of treatment (EOT) and 10 relapsers (achieved EOT but not SVR). Comparison between patients with non-significant fibrosis and patients with significant fibrosis:-There was statistically significant difference between both groups as regarding age which was significantly higher in patients with significant fibrosis than in those with non-significant fibrosis (47.5 ±7.8 vs.39.7 ±10.6 years old). Prothrombin conc. was significantly lower in patients with significant fibrosis (83.7 ±10.2 vs. 89.3 ±8.6) and INR was significantly higher in patients with significant fibrosis than in those with non-significant fibrosis (1.2 ±0.1 vs. 1.1 ±0.1). On the other hands, there was no statistical significance between two groups concerning IP10, Bilirubin, Albumin, AST, ALT, Creatinine, Hb, WBCs, Platelets, AFP, TSH and base-line HCV RNA level (table 7). However, there was a significant difference between the two groups regarding SVR as 11 out of 15 patients (73.3%) with significant fibrosis were Non-SVR achievers vs. 19 out of 55 patients (34.5 %) with non-significant fibrosis were Non-SVR achievers (P = 0.009) (table8). Discussion:- There are many estimates of the number of people in Egypt that are infected. Many publications suggest that nearly 14 % of the people in Egypt were infected. This is ten times greater than in any other country in the world (Sievert et al., 2011). The genotype 4 is predominant in 91% of these patients (Guerra et al., 2012). The combination therapy with interferon-α and ribavirin, neutralize the virus after 6 months in 40-50% of the infection cases with genotype 1 and in 80% of the infection cases with genotype 2 and 3 (Schuppan et al., 2003). Infections of HCV genotype 4 is, as genotype 1, relatively resistant to the interferon-α/ribavirin combination therapy medical treatment (El Makhzangy et al., 2009). A major paradigm shift happened in HCV treatment, with the advent of highly effective, simplified and short duration (12 weeks) oral DAA-based regimens (Dore et al., 2012). However, the effect of these new therapies, even in developed countries, will be modest without expanded access to treatment (Thomas et al., 2010). The aim of this study was to assess the efficacy of serum IP-10 as a predictor of SVR in chronic hepatitis C virus Egyptian patients receiving Peg-IFN and RBV therapy. This was a prospective study which was conducted on 70 patients with chronic hepatitis C. In addition, 10 healthy persons with matched age and sex were enrolled as a control group. Patients were categorized into two groups according to their sustained virological response (SVR reported that menopause is related to an increased severity of liver fibrosis, and a lower likelihood of response to therapy with peg-IFN and ribavirin. The decreased rate of a complete response to IFNα treatment may correspond to decrease in estrogen levels. Interleukin 1, associated with an inflammatory response, is stimulated by low concentrations of estrogen and progesterone. A low concentration of estrogen allows peripheral blood monocytes to secrete more interleukin 1. The spontaneous production of interleukin 1 by peripheral mononuclear cells has been shown to be significantly higher in patients with CHC than in healthy control individuals, and decreased in those with a complete response after the administration of IFNα. This cytokine production may alter the effectiveness of IFNα treatment in perimenopausal and menopausal women with CHC infection (Di Marco et al., 2013). 762 It is well known that advanced fibrosis or cirrhosis is a poor prognostic factor for response to antiviral therapy compared with no or minimal fibrosis In agreement with these studies; we found that SVR was significantly higher in patients with non-significant fibrosis than in patients with significant fibrosis as (73.3% vs. 34.5 %; P = 0.009). In addition, we found that the mean of Transient Elastography was significantly higher in Non-SVR achievers (10.9 ±6.3) KPa than in the SVR achievers (7.4 ±2.5) Kpa. This was matched with Stasi et al. (2015) who showed that liver stiffness (LS) >12 Kpa prior to the initiation of dual therapy was significantly associated with poor response to therapy (p <0.025). LS >12 Kpa should be considered a strong prognostic indicator of non-response to anti HCV treatment (Martinez et al., 2012). In our work, we found that RVR was a significant predictor of SVR. Thirty seven patients (92.5 %) of fourty SVR achievers were RVR while twenty one patients (70%) of thirty Non-SVR achievers were RVR (P = 0.023). This was coincided with a study conducted by Ferenci et al. (2004), he observed that patients infected with HCV genotypes 2 or 3 achieved RVR in higher proportions than patients infected with genotype 1. However, regardless of the HCV genotype, patients who reached RVR have the highest rates of SVR. In the study by (Fried et al., 2011) RVR was achieved by 16% of patients with genotype 1, 71% of genotype 2 and 60% of genotype 3. Among individuals who reached RVR, the SVR rate was high across all HCV-genotypes and ranged from 88% to 100% (genotypes 1-4) (Mangia et al., 2010). Also in our study, we found that there was statistically significant difference between SVR achievers and Non-SVR achievers as regarding achievement of EVR as 39 (97.5 %) patients who achieved EVR were SVR achievers (P = 0.001). This was matched with a study from China which showed that patients without EVR rarely achieve SVR (Reddy et al., 2005). In our work we found that, AFP level, was higher in Non-SVR achievers than in SVR achievers ( In our study, IP10 level was 260.7 ±165.2 pg/ml in the treated group. High serum IP-10 levels were associated with a failure to achieve SVR. IP-10 levels were significantly lower in patients who achieved SVR pg/ml than in Non-SVR achievers (159.9 ±55.2 vs. 395 ±167.9 pg/ml; p=0.001). While there was no statistically significant difference between patients with significant and non-significant fibrosis regarding IP -10 (P=0.5). This was matched with a study done in Cairo University, Egypt which included 80 treatment naive HCV patients. Pretreatment serum IP-10 levels were significantly lower in patients who achieved SVR than in non-responders and they concluded that low pretreatment serum IP-10 is a favorable predictor of response to antiviral HCV therapy in Egyptian patients Diago et al. (2006) reported the association of serum IP-10 levels with SVR to (PEG-IFN/RBV) therapy in patients with genotype 1 chronic HCV. They showed that the levels of IP-10 were lower in SVR patients. Noteworthy, after successful antiviral therapy, serum IP-10 concentrations decreased to levels lower than baseline, whereas they were unchanged in non-responders, suggesting that HCV itself may be responsible for elevated serum IP-10 763 concentrations found in HCV-infected patients, and thus, pretreatment serum IP-10 is an independent predictive factor of SVR in HCV patients infected by genotype 1. On contrary; El Razikyet al (2015) found that the pretreatment serum IP-10 levels were not significantly different in relation to different grades of necro-inflammatory activity and fibrosis stages .Verifying the predictive value of pretreatment serum IP-10 levels, their study did not find a significant relation to response at week 12, 24, 48, and 72. Concentrations lower than 594.1 pg/mL had a positive predictive value of 86.8% in their study population. Also, Reiberger et al in 2008 did not find a clear association between IP-10 levels before or during treatment and SVR. Also, Yoneda et al in 2011 did not confirm the association between a low baseline IP-10 level and SVR. Other reports confirmed these findings; no difference was seen in baseline IP-10 levels between CHC patients with or without RVR (Falconer et al., 2010) or with or without SVR (Wan et al., 2009). IP-10 has a chemotactic function on different cell types following binding to its receptor .IP-10 recruits T lymphocyte subsets expressing the CXCR3 receptor, including activated T lymphocytes of the T helper type 1 phenotype. Also ,cytotoxic T cells and NK cells express CXCR3 and are targeted by the IP-10 chemotactic effect (Zeremski et al., 2008). IP-10 is upregulated in the liver cells infected by HCV since strong IP-10 staining was found in the cytoplasm of the hepatocytes but not in other liver cells. It has been suggested that HCV itself may be responsible for elevated serum IP-10 levels found in HCV infected patients. HCV proteins such as NS5A and core, alone or in combination with pro-inflammatory cytokines, can induce IP-10 gene expression and secretion in human hepatocyte derived cells, leading to the accumulation of CXCR3 expressing T lymphocytes in the liver. IP-10 expression may also provide important retention signals, resulting in the observed accumulation of T lymphocytes (Apolinario et al., 2005). Thus, the mechanism by which IP-10 induces treatment failure may be explained by its role in the recruitment of effector Th1 lymphocytes into the liver cells of chronic HCV patients and its potential in contributing to the host immune response against the virus as well as to the disease progression. Also, IP-10 modulates the efficacy of IFN-α plus ribavirin therapy by enhancing the expression of the HCV NS5A protein which induces IL-8 (CXC chemokine) secretion. IL-8 was also found to be associated with the inhibition of the antiviral effects of IFN-α (Zeremski et al., 2008). We can assume that HCV and IP-10 have synergistic effects towards each other; HCV induces IP-10 secretion by the hepatocytes, and IP-10 stimulates viral replication and inhibits the antiviral effects of IFN-α (Apolinario et al., 2005). It was found that circulating CXCL10 is indeed processed into the shorter form and turns it into a CXCR3 antagonist. It was demonstrated that the short, antagonist form of CXCL10 predominates in the plasma of chronically infected patients who are destined to fail anti-HCV therapy and thus is responsible for the lack of desired antiviral effects of IP-10 and could play an important role in pathology (Casrouge et al, 2011). Thus, IP-10 is a strong negative predictor of response to Peg-IFN/Ribavirin therapy in HCV-4 mono-infected patients. This highlights the need to consider this factor in the individualization of treatment, and augments the level of predictiveness of IL28B polymorphisms for final treatment outcome (Charles and Dustin, 2011). As IP-10 may play a role in the mechanism of failure of antiviral therapy, interventions neutralizing endogenous IP-10 or blocking the function of its receptor, CXCR3, may provide new strategies to improve the treatment outcome of these difficultto-cure patients (Charles and Dustin, 2011). Our study has limitations as it was conducted on a small number of patients and this population contained small percentage of patients with significant viremia and significant fibrosis. Data may not be generalizable to other study populations. Also, we did not analyze the IL-28 B polymorphism in our cohort which may interact with our results. In conclusion, we found a strong association between low pretreatment IP-10 levels and SVR in HCV infected Egyptian patients. Pretreatment serum IP-10 level should be included as one of the predictor of response to therapy. 764 Attempts to neutralize IP-10 or block IP-10 receptor should be considered as additional strategies to further improve the outcome of the antiviral therapy. Further studies with larger samples are needed to determine the association of IP-10 and HCV treatment response especially with DAA therapy.	2019-04-22T13:05:32.768Z	2017-03-31T00:00:00.000	{ "year": 2017, "sha1": "b219e626992432eb7a12413bd5c6048f1f93c47a", "oa_license": "CCBY", "oa_url": "http://www.journalijar.com/uploads/348_IJAR-16154.pdf", "oa_status": "HYBRID", "pdf_src": "MergedPDFExtraction", "pdf_hash": "9896e5412cb32ef26e78daf47aadaaced5e3f275", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Mathematics" ] }
259225839	pes2o/s2orc	v3-fos-license	Transcriptomic analysis provides insight into the genetic regulation of shade avoidance in Aegilops tauschii Background Weeds are not only economically important but also fascinating models for studying the adaptation of species in human-mediated environments. Aegilops tauschii is the D-genome donor species of common wheat but is also a weed that influences wheat production. How shading stress caused by adjacent wheat plants affects Ae. tauschii growth is a fundamental scientific question but is also important in agriculture, such as for weed control and wheat breeding. Result The present study indicated that shade avoidance is a strategy of Ae. tauschii in response to shading stress. Ae. tauschii plants exhibited growth increases in specific organs, such as stem and leaf elongation, to avoid shading. However, these changes were accompanied by sacrificing the growth of other parts of the plants, such as a reduction in tiller number. The two reverse phenotype responses seem to be formed by systemically regulating the expression of different genes. Fifty-six genes involved in the regulation of cell division and cell expansion were found to be downregulated, and one key upstream negative regulator (RPK2) of cell division was upregulated under shading stress. On the other hand, the upregulated genes under shading stress were mainly enriched in protein serine/threonine kinase activity and carbon metabolism, which are associated with cell enlargement, signal transduction and energy supply. The transcription factor WRKY72 may be important in regulating genes in response to shading stress, which can be used as a prior candidate gene for further study on the genetic regulation of shade avoidance. Conclusions This study sheds new light on the gene expression changes and molecular processes involved in the response and avoidance of Ae. tauschii to shading stress, which may aid more effective development of shading stress avoidance or cultivars in wheat and other crops in the future. Supplementary Information The online version contains supplementary material available at 10.1186/s12870-023-04348-y. Background The emergence of agriculture created habitats not only for intentionally cultivated plants (crops) but also for unwanted species (weeds) that adapted to exploit humanmediated environments. Agricultural weeds are among the great survivors of the plant kingdom, to persist and thrive in the face of human persecution or attempts to kill [1]. In evolution, weeds and crops may interact. For instance, all present-day cultivated varieties of Asian rice (Oryza sativa) were domesticated from the wild species O. rufipogon ~ 10 000 years ago [2,3]. However, de novo weedy rice strains were also generated through domestication during the history of rice cultivation [4][5][6]. Furthermore, gene introgression from cultivated rice into weedy rice has led to the formation of a genetically and morphologically variable group, enhancing the adaptation of weedy strains to diverse anthropogenic environments [7]. Weeds can contribute to the evolution of polyploid crops. A well-known example is common wheat (Triticum aestivum, 2n = 6x = 42, AABBDD). This crop originated from the hybridization between cultivated wheat T. turgidum (AABB) and the weed Aegilops tauschii (DD) ~ 10,000 years ago [8,9]. T. turgidum cultivation is still associated with weedy Ae. tauschii in Middle Eastern agroecosystems, which is thought to be the birthplace of common wheat [10]. The incorporation of the D-genome of Ae. tauschii is crucial for the success of bread wheat as the largest acreage crop in the world since it confers the potential to make diverse food products and a broad adaptability to diverse environments [11]. Natural introgression from other weedy species, such as wild emmer wheat, increases the genetic diversity of common wheat after origination [12]. Artificial introgression of genes from wild species into wheat has been widely exploited as a popular tool in modern breeding [13]. Wheat-Ae. tauschii is a fascinating model to study the growth competition of polyploid crops with progenitor species. Ae. tauschii is an invasive weed affecting wheat production, but controlling it is difficult due to its close evolutionary relationship with wheat. For example, in China, before the 1990s, this species was only sporadically observed in wheat fields in three provinces [14]. However, it escaped control management and quickly expanded to new habitats. By 2007, it occurred in eight provinces with a damage area of ~ 3.3 × 10 5 ha [15]. Malignant weeds can cause up to 50-80% yield loss [16,17]. To better control this weed, it is crucial to understand the adaptation mechanism of its competition with wheat. Competition for sunlight in plants can be particularly fierce since photosynthesis is the only source of energy [18]. Shade avoidance and shade tolerance are two main strategies to adapt to competition for light changes caused by adjacent vegetation [19]. Here, the phenotypes and global gene transcription profiles of Ae. tauschii when grown alone and co-planted with wheat were compared. The picture that emerges is that Ae. tauschii exhibited phenotypic shade-avoidance responses. Correspondingly, gene transcription changed to adapt to competition. This study provides fundamental insight into the regulatory process underlying shade avoidance in Ae. tauschii and facilitates the identification of genes involved in shade-avoidance mechanisms. Plant materials and experimental setup The common wheat variety Shumai969 (bred by Dengcai Liu, Triticeae Research Institute of Sichuan Agricultural University, China) [13] and four Ae. tauschii accessions, two from subspecies tauschii (AS71 and AS77) and two from subspecies strangulata (PI431599 and PI431602), were used in this study. AS77 was collected from the wheat fields of Hennan Province, China [14]. The remaining three were collected from the natural distribution areas of the species, AS71 from Xinjiang, China, PI431599 from Azerbaija, and PI431602 from Turkmenistan. AS71 and AS77 were formally identified by Chi Yen (the Triticeae Research Institute of Sichuan Agricultural University, China) and provided by Sichuan Agricultural University. PI431599 and PI431602 were formally identified by N. I. Vavilov (Institute of Plant Industry, Former Soviet Union) and provided by the USDA National Small Grain Collection. All lines were planted at the Wenjiang Experimental Station (103°51′E, 30°43′N) of Sichuan Agricultural University in the 2017-2018 cropping season. The former crop was rice. To maintain consistent growth conditions, all the materials were planted within an area of ~ 200 m 2 . The experiments exploited two planting patterns, i.e., mono-cropping (MC) of Ae. tauschii and inter-cropping (IC) of wheat-Ae. tauschii. Each MC Ae. tauschii accession and each IC wheat-Ae. tauschii combination were planted in 5 × 5 rows with a row length of 2.0 m and row spacing of 30 cm. For the MC growth condition, five Ae. tauschii seeds were spaced and sown in each row (Fig. S1). For the IC growth condition, five Ae. tauschii seeds were inter-sown with 10 wheat seeds in each row in 2017. The sowing date was 4 November 2017. Phenotypic measurements The measured traits at the seedling stage included plant height, leaf length and tiller number, and measurements were performed on 12 March 2018. The measured traits at the adult stage included plant height, flag leaf length, tiller number, spike length, number of spikelets, internode length, seed setting rate, heading time and flowering time. The plant height, leaf length, spike length, number of spikelets, internode length, and seed setting rate were the average values of the three highest tillers for each plant. All the measured plants were used to compare the difference in Ae. tauschii between MC and IC conditions. Significant differences were determined by Student's t tests. RNA sequencing On 12 March 2018, samples from four accessions under MC and IC were taken for RNA sequencing when the Ae. tauschii plants at vegetative periods were climbing for mono-cropping but erect for intercropping. For each treatment, 2-4 biological replicates were set. Whole plants harvested from the field were immediately snap-frozen in liquid nitrogen and then stored at -80 °C. Total RNA was extracted from the samples excluding roots using an RNAprep Pure Plant kit (TIANGEN, Beijing, China). Sequencing libraries were generated using the NEBNext® Ultra™ RNA Library Prep Kit for Illumina (New England Biolabs, USA). The libraries were sequenced using a HiSeq 2500 platform (Illumina, San Diego, CA, USA) following the standard protocol. RNA concentration and purity were measured using a Nano-Drop 2000 (Thermo Fisher Scientific, Wilmington, DE). RNA integrity was assessed using the RNA Nano 6000 Assay Kit of the Agilent Bioanalyzer 2100 system (Agilent Technologies, CA, USA). The clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v4-cBot-HS (Illumina). After cluster generation, the prepared libraries were sequenced on an Illumina platform, and paired-end reads were generated. Contaminated and low-quality reads were discarded by imposing a Q30 threshold of 90% and a maximum of 0.2% ambiguous base calls. Reads were mapped to the Ae. tauschii reference genome (https://www.ncbi.nlm.nih. gov/assembly/GCF_001957025.1) using HISAT2 with default settings for parameters [20]. FPKM (Fragments Per Kilobase of transcript per Million fragments mapped) was used to quantitatively estimate the value of gene expression [21]. Differential expression analysis Differential expression analysis was performed using the DESeq2 R package [22]. Differentially expressed genes (DEGs) of each accession under competitive stress were determined with a false discovery rate (FDR) threshold < 0.05 and \|log2FC\| ≥ 1 (FC means fold change). The gene with an FPKM value of 0 was given a value close to 0 to calculate the fold change. To discover common DEGs shared by all the analysed accessions under inter-cropping conditions, K-means analysis was performed. The k value was set to 30 according to the gene expression patterns. GO and KEGG enrichment Gene Ontology (GO) enrichment analysis of DEGs was implemented by GOseq, which is an R package based on the Wallenius noncentral hypergeometric distribution [23]. KOBAS software was used to test the statistical enrichment of target genes in KEGG pathways [24]. GO categories and KEGG pathways with P values ≤ 0.05 were defined as significantly enriched. Weighted gene co-expression network analysis (WGCNA) Co-expression networks were built using weighted gene co-expression network analysis (WGCNA) in BMK-Cloud (www.biocloud.net) [25]. The parameters used in the WGCNA were as follows: FPKM ≥ 1; cv (Variation of FPKM) ≥ 0.5; hierarchal clustering tree: dynamic hybrid tree cut algorithm; power: 13; minimum module size: 30; minimum height for merging modules: 0.31455. The candidate co-expression network was visualized by Cytoscape (version 3.4.0, released on May 13, 2016). In the co-expression network, the 10 genes with the highest degree of connectivity were regarded as hub genes. In the net, circular nodes represent genes, and edges represent connections. Quantitative accuracy analysis of RNA-seq Eighteen genes were randomly selected and verified by quantitative real-time PCR (qRT-PCR). The TaKaRa Prime Script TMRT Reagent Kit with gDNA Eraser (Perfect Real Time; TaKaRa, Shiga, Japan) was used to synthesize cDNA according to the manufacturer's instructions. qRT-PCR was performed using SYBR Premix Ex Taq™ II (TaKaRa, Shiga, Japan). The 20 µl mixtures for PCRs consisted of 10 µl of 2× SYBR Green II Mix, 0.4 µl of each forward and reverse primer, 2 µl of cDNA, and 7.2 µl of ddH2O. The PCR program was 94 °C for 5 min, followed by 35 cycles of 94 °C for 30 s, 58 °C for 30 s, and 72 °C for 30 s. Three biological replicates were conducted for each sample. The relative expression levels were calculated using the 2 −ΔΔC t method [26]. Specific primers for qRT-PCR were designed using Primer 5.0 software (Table S1). GAPDH was used as a reference gene [27]. Growth in response to shading stress Four Ae. tauschii accessions belonging to subspecies tauschii (AS71 and AS77) and subspecies strangulata (PI431599 and PI431602) were used in this study. They were planted in an experimental field under mono-cropping (MC) and inter-cropping (IC) patterns (Fig. S1). Under IC, the plants of Ae. tauschii at the seedling stage were lower than those of the surrounding wheat plants, resulting in shading stress. The four analysed Ae. tauschii accessions in IC exhibited similar responses, including a more erect position of plants and elongated leaves but accompanied by reduced root growth and fewer tillers compared to MC (Fig. 1). In IC, the average height of the four materials increased by 14.7 cm, and the average leaf length increased by 4.2 cm. This indicated that the shading stress from wheat strongly changed the growth of Ae. tauschii seedlings regardless of genotype or taxon. We also compared the phenotypic changes of plants between MC and IC at the adult stage. The four Ae. tauschii accessions in IC all exhibited increased plant height, flag leaf length and seed setting rate but fewer tillers and roots (Fig. 2). For instance, the number of tiller changes was very evident (MC vs. IC: 54 vs. 11 for AS71, 82 vs. 15 for AS77, 98 vs. 33 for PI431599, 71 vs. 18 for PI431602). The average height of the four materials increased 26.4 cm, with subspecies tauschii increasing 28.3 cm and subspecies strangulata increasing 24.5 cm. The average length of internodes increased 5.9 cm, 4.6 cm, 5.2 cm, 4.9 cm and 3.8 cm from the first to the fifth internodes. The increase in plant height was actually caused by the increase in internode length. Furthermore, the average increase in the seed setting rate of the four materials was 27%. However, the phenotypic changes in other traits, including heading time, flowering time, spike length and number of spikelets, were dependent on Ae. tauschii accessions (Fig. 2). Among them, the average heading time was approximately 3 days earlier and the average flowering time was approximately 2 days earlier for the four materials under IC. Transcriptome analysis To decipher the gene expression responses under shading stress, we completed RNA-Seq analysis of Ae. tauschii at vegetative periods when plant morphology was distinct between MC and IC (Fig. 1A). RNA-Seq analysis was applied to 23 RNA samples, mostly showing high correlation coefficients between biological replicates (Fig. S2). To further test the accuracy of RNA-seq quantification, we randomly selected 18 expressed genes for qRT-PCR analysis. The linear regression of the expression level using the data from qRT-PCR and RNA-seq was then analysed. The R square was 0.93 (Fig. 3A), inferring that the results of RNA-seq quantitative analysis are trustworthy. On average, the Q30 base percentage was 93.92% or above. The clean reads of each sample were sequenced with the specified reference genome, and the alignment efficiency ranged from 79.75 to 94.64%. Most (83.13%) of the reads were uniquely mapped (Table S2). Gene expression under growth shading stress To find Ae. tauschii genes that responded to shading stress caused by wheat in the seedling stage, we first analysed the DEGs between MC and IC. There were 4294 DEGs (Table S3), and the number was different among the four Ae. tauschii accessions varied from 723 in PI431599 to 2,749 in PI431602 (Fig. 3B). To detect the expression profiles of DEGs, we first carried out a simple cluster analysis of DEGs, found that it was difficult to find the common response genes (Fig. 3C), and then performed K-means analysis. Considering that the four accessions exhibited obvious shade-avoidance responses to light competition, this analysis focused on the common response genes, which were shared by the four Ae. tauschii accessions and either upregulated or downregulated. There were 631 common response genes (Table S4). Of these, 193 had higher expression levels in IC than in MC (Fig. 3D). However, many more genes (438) were downregulated in IC (Fig. 3E). Co-expression network To identify vital genes of Ae. tauschii specific to shading stress, the 4294 DEGs were used for co-expression network analysis using WGCNA (Fig. 6A). This analysis revealed 11 modules, i.e., 11 highly connected gene clusters. The brown module was positively correlated with the shading response (Fig. 6B, Table S6). GO analysis indicated that they were enriched in protein serine/ threonine kinase activity, protein phosphorylation, cinnamic acid biosynthetic process, chlorophyll catabolic process and phenylalanine ammonia-lyase activity (Fig. 6C). The 10 genes with the highest number of nodes were regarded as hub genes (Fig. 6D, Table S7). The gene with the most nodes encoded a protein of unknown function (LOC109774035, 66 nodes). A gene encoding WRKY transcription factor 72 (LOC109776904, 24 nodes) was among the top 10 hub genes as well as in shading response common genes. There were 71 subgenes under WRKY72 node, including 11 protein kinases, which were 2 serine/threonine-protein kinase RIPK, 2 wall-associated receptor kinase 2 (WAK2), 2 cysteine-rich receptor-like protein kinase (CRK10 and CRK6), 2 receptor-like protein kinase FERONIA, 2 L-type lectin-domain containing receptor kinase (LecRK-IX.1 and LecRK-IV.1) and Rust resistance kinase Lr10 (LRK10). Another hub gene, UDP glycosyltransferase 83A1 (LOC109776791, 47 nodes), had 47 subgenes in its network, including 8 protein kinases, all of which were the same as those in the Discussion Crops and weeds unavoidably face competition for resources such as sunlight. How low light stress caused by the shading of adjacent plants affects plant growth is not only a fundamental scientific question but is also of vital importance for agriculture, such as in weed control and crop breeding. The monocot Ae. tauschii is the D-genome donor of bread wheat [8], but it is also a weed found wheat fields. The present study indicated that shade avoidance is a strategy of Ae. tauschii plants to respond to low light stress. Shade-avoidance responses are associated with phenotypic changes that favour plants to obtain more light resources, which are called shade-avoidance syndromes [19,28,29]. Ae. tauschii altered growth in specific organs to avoid the shading of surrounding wheat plants in the field, such as forming a more erect position of plants and increasing stem and leaf length. However, these changes were accompanied by sacrificing the growth of other parts of the plants, such as fewer tillers and roots. The obvious reverse effects of shading on plants, i.e., growth increases in some parts of a plant but decreases in other parts of the plant were also observed in other plant systems, such as the dicot Arabidopsis [29,30]. It is interesting to understand how the dual responses of shaded plants are harmonized. Intuitively, the growth increase and decrease should be controlled by independent strategies by which plants respond to low light stress. A decrease in plant growth is a typical response to abiotic stresses such as severe drought and salt. The trade-off between growth and stress resistance is usually explained by energy/carbon limitations, since plants under stress divert substantial resources away from growth and towards a stress response [31,32]. However, increasing evidence indicates that the trade-off mainly results from the active suppression of growth by stress signalling pathways [33][34][35][36]. Plant growth depends on cell growth, which is the process by which cells accumulate mass by cell division and increase in physical size by cell enlargement, and abiotic stress often impedes plant growth by repressing both cell division and cell expansion [37]. Here, an obvious transcriptomic response to all the analysed Ae. tauschii accessions under shading stress was that fewer genes were upregulated (193) than downregulated (438). Interestingly, ~ 15% (66) of downregulated genes were enriched in the regulation of cell division, including Each row corresponds to a module. The name of modules is indicated on the left. Each column corresponds to a specific sample. The colour of each cell at the row-column intersection indicates the correlation coefficient between the module and sample. A high degree of correlation between a specific module and sample is indicated by red. C The top enriched GO categories in brown module genes. D Co-expression network in the brown module with a KME higher than 0.9 visualized by Cytoscape 12 for cell cycle regulation and 44 for DNA replication, recombination and repair. Of these, 12 were cyclins (2 CYCA3 and CYCD2, CYCD4, CYCD5, CCNF) and DNA replication licensing factors (MCM2-MCM7). Cyclins are key molecular drivers of the cell cycle, and their downregulation inhibits cell division under drought and salt stresses [37]. MCMs play important roles in DNA replication, and their downregulation is associated with abiotic resistance [38][39][40]. Cell division suppression could result in a reduction in the number of specific organs, such as tillers and roots, of Ae. tauschii. In addition to the downregulation of genes directly involved in cell division, some upstream genes involved in the negative regulation of cell division were also identified. For instance, the receptor-like kinase RPK2 is upregulated under shading stress (Fig. 5A). It has been confirmed to be involved in the maintenance of the root apical meristem by controlling cell proliferation and affecting meristem size [41]. The roots of RPK2-overexpressing transgenic lines were diminished compared with those of the wild type [42]. The results suggested that shading stress signals may systematically activate gene systems to inhibit growth. Although cell division and cell enlargement frequently go together, in some cases growth may be due mostly to cell enlargement. For instance, the gibberellin (GA)induced growth of lettuce hypocotyls is primarily due to cell elongation [43]. Cell elongation could be exploited to explain the growth elongation of specific organs, such as stem and leaf elongation, which is the strategy of Ae. tauschii plants to obtain more light resources under shading stress. Consistent with this presumption, 16 out of 19 upregulated Ae. tauschii genes enriched in protein serine/threonine kinase activity under shading conditions belonged to RLK/Pelle family. RLK/Pelle proteins have been implicated in the mechanical properties of the cell wall, which is vital for cell expansion [44]. For instance, wall-associated kinases (WAKs) are a subclass of RLKs and are linked to cell elongation [45,46]. The upregulation of WAK-like kinase 4 gene expression was associated with stress responses, such as to salt [47,48]. In addition, BRL1 (BRASSINOSTEROID RECEPTOR-LIKE1) is the main receptor of the brassinosteroid hormone and is expressed in vascular tissues and regulates shoot vascular development [49,50]. Loss-of-function mutants bri1 in Arabidopsis showed severe dwarfism; in contrast, overexpression of BRL1 manifested as shoot elongation [51]. The upregulation of BRL1 in Ae. tauschii under shading stress suggested that the brassinosteroid (BR) hormone may be involved in the formation of shade-avoidance syndromes. Carbon metabolism is intimately linked to growth and stress responses, and tight control of their fluxes in cellular metabolism and throughout the plant is essential to ensure survival and reproduction under environmental constraints [52]. Abiotic stress affects energy supply. A study of shaded maize leaves found that the levels of amino acids, carbohydrates, lipids, nucleotides and related metabolites increased significantly under carbon starvation [53]. When stressed, to ensure an adequate energy supply, stressed plants respond to carbon metabolism, mainly by transferring a large amount of energy to stress resistance to prevent or repair stress-induced damage and maintain intracellular homeostasis [31,32,54,55]. Five genes involved in energy production and conversion were upregulated in IC. In response to abiotic stress, plants divert substantial resources to resist stress to maintain cellular homeostasis. To ensure an adequate energy supply, stressed plants will generate a response to carbon consumption. In the co-expression network, hub genes are expected to play an important role in shading stress. Of the 10 identified hub genes, the function of the top two genes (LOC109774035 and LOC109774111) in the connectivity ranking was unknown. Therefore, they are interesting candidates for further functional characterization. The third most interactive gene was UDP glycosyltransferase 83A1 (UGT83A1, LOC109776791). UGT83A1 glycosylates most of the lignin precursors and flavonoids, and its overexpressing lines showed strong abiotic stress tolerance [56], so it could potentially play key roles in coping with shading stress in Ae. tauschii. Another highly connected gene in the network was WRKY transcription factor (WRKY72, LOC109776904). WRKYs have multiple roles, including plant development, abiotic stress, hormone signalling, and primary and secondary metabolism [57][58][59][60][61]. WRKYs act as activators of the same ABAinducible promoter and are related to the induction of abscisic acid/stress-related genes [62,63]. During the shade-avoidance response, WRKY26, 45, and 75 restrict root growth and development [64]. Since WRKY was regarded as both a hub gene and a common gene, it could play an important role in balancing growth and shading stress by reprogramming gene expression, such as cell cycle-and serine/threonine kinase-related genes. Conclusion The use of four Ae. tauschii accessions allowed for the detection of a robust set of genes that play a role in the shade-avoidance response. Defending against stress and actively inhibiting growth are two complementary strategies for plants to cope with adverse environments [35]. When the accessions of Ae. tauschii are under shading stress by wheat, the response to low light may mainly adopt two modes. On the one hand, to avoid the harsh environment of shade, plants grow taller and longer by increasing energy supply (carbon metabolism). On the other hand, stress defence is activated, and growth is inhibited. Abiotic stress usually hinders plant growth by inhibiting cell division and cell expansion, such as by reducing tillers to a certain extent (Fig. 7). This is an adaptive strategy to maximize survival. Since WRKY was regarded as both a hub gene and a common gene, it could be regarded as a candidate gene for further study. By introducing the hub gene into common wheat, the weed-like characteristics of Ae. tauschii can be incorporated, enabling the wheat plants to compete more effectively with weeds in harsh environments, leading to weed control. Due to experimental limitations, the aboveground tissues was taken for transcriptome sequencing in this experiment, and although the mechanism of shading stress specifically affecting stem and leaf tissues was not extensively explored, this study sheds new light on the gene expression changes and molecular processes involved in the response and avoidance of shading stress in the overall aboveground parts of Ae. tauschii, which may aid more effective development of shading stress avoidance or cultivars in wheat and other crops in the future.	2023-06-23T13:52:39.606Z	2023-06-23T00:00:00.000	{ "year": 2023, "sha1": "62a1d3f90fcbad57751ae74bbcb71b7d12b2d09f", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Springer", "pdf_hash": "62a1d3f90fcbad57751ae74bbcb71b7d12b2d09f", "s2fieldsofstudy": [ "Biology", "Environmental Science", "Agricultural and Food Sciences" ], "extfieldsofstudy": [ "Medicine" ] }
214039138	pes2o/s2orc	v3-fos-license	Based on ZigBee and GIS technology to explore the real-time dynamic monitoring system of urban atmospheric environment Good ecological Environment and Atmospheric Environment is an important part of human healthy growth and survival. With the acceleration of the urbanization process, bringing people a lot of material and spiritual enjoyment, but also increased water, soil and atmospheric environment and other serious pollution problems. Under the influence of greenhouse effect, global warming further aggravates the harm to people’s physical and mental health. In this paper, based on ZigBee and GIS technology, the content of real-time dynamic monitoring of urban atmospheric environment is deeply analyzed, and the positioning and targeted warning of pollution sources of solid pollution particles and pollution gases in the city are carried out to further prevent the spread of pollution accidents. Abstract. Good ecological Environment and Atmospheric Environment is an important part of human healthy growth and survival. With the acceleration of the urbanization process, bringing people a lot of material and spiritual enjoyment, but also increased water, soil and atmospheric environment and other serious pollution problems. Under the influence of greenhouse effect, global warming further aggravates the harm to people's physical and mental health. In this paper, based on ZigBee and GIS technology, the content of real-time dynamic monitoring of urban atmospheric environment is deeply analyzed, and the positioning and targeted warning of pollution sources of solid pollution particles and pollution gases in the city are carried out to further prevent the spread of pollution accidents. Development background and status quo of environmental monitoring technology As one of the technical contents with far-reaching influence on the development of the times, environmental monitoring technology has been officially appeared in the public field of vision since 1950s. In the four different stages of environmental monitoring technology change, from chemical analysis to pollution source monitoring, from automatic monitoring to remote sensing geographic information measurement and control, environmental monitoring has changed from passive to active, from artificial to intelligent. Nowadays, the DHT11 type temperature and humidity sensor, which is used in ZigBee technology, can realize the above technology very well, see figure 1 for more details. Although great achievements and progress have been made in monitoring technology, in the process of its application in China, such as fig. 2, the comparison of atmospheric monitoring data in China shows that only first-tier cities or municipalities directly under the Central Government have promulgated and implemented the relevant regulations and methods of environmental monitoring. At the same time, the traditional methods of on-line monitoring of environmental data can not match the present situation of the development of our country. Therefore, it is necessary to establish a set of comprehensive environmental monitoring information system with many functions and characteristics, such as automation and networking [1] Existing problems of urban atmospheric environmental monitoring system At present, the technical means of our country only collect and transmit the different data of polluted atmosphere in the human living environment. It is impossible to make effective mining and in-depth analysis of the data. At the same time, the data acquisition and analysis method based on computer control improves the data storage space of environmental monitoring to a certain extent, but the analysis of the data processing results does not produce practical effect and significance. Therefore, according to the problems and disadvantages of the above technology, based on ZigBee and GIS technology, ZigBee technology, as shown in figure 3, can deeply mine and analyze the data on the basis of timely collection and efficient transmission of environmental data information, and provide better technical service support for the environment according to the servers and software supporting the terminal and mobile phone. Provide more efficient and convenient technical support to the public . The router2 The router n The sensing layer is a hardware sensor network system in the process of monitoring the urban atmospheric environment. Figure 5 shows that on the basis of ZigBee technology, wireless transmission is carried out through GPRS and ad hoc network technology. Because the city has certain regional space in the whole spatial layout, in the process of setting up the perceptual layer system, it is necessary to divide the population, area and function of the regional space. Figure 6 shows and transmits and interacts between the router and the coordinator through the sensor subnet in ZigBee, so as to further achieve the purpose of environmental monitoring [2]. Research on the specific application mode of ZigBee and GIS technology The design and construction of the real-time dynamic monitoring system for urban atmospheric environment based on ZigBee and GIS, the specific content is the targeted division and optimization of the application system. In short, based on the system division of the application layer and the sensing layer, in order to further save the economic cost, the related laying method of the static environmental monitoring point is adopted [3]. First of all, as shown in figure 7, according to the street direction of the city, and for the polluted areas and iconic buildings in the city and other related nodes, as the dense area of node collection. Then the tree network structure diagram is formed. In the process of setting up the nodes of each sub-network, under the action of the router and the coordinator, the collected data is transmitted through the connection of the sensor, and finally the functions of data storage, statistics and publication are completed. Fig.5 Sensing system diagram In order to monitor and collect the atmospheric environment more intuitively and concretely, the wireless sensor network nodes are arranged in the sensing layer on the basis of the basic division of the geographical space area and through the different functional areas. The spatial data of each area of the city are monitored and collected in real time, and the analysis is carried out on the basis of centralized statistics. Finally, the display and layout of the thematic content map are formed in the space. As shown in figure 8, the system is designed by ZIGBEE technology, and the results are as follows. System design Therefore, in the process of developing and designing the application layer of the system, in order to match with the perceptual layer, the content of GIS is also collected. In the process of research and development of general application layer system technology, ESRI ArcGIS Sever9.3 and other related technologies are often used to serve and develop the global positioning of geographical system, and Java and other related program technologies are used as the main body of language application and development, so as to realize the real-time and dynamic monitoring of the environmental pollution on which people depend for their survival in urban life, and to provide reliable and real reason for the content of environmental control in the later stage. On data support [4] . Data content division of urban environment dynamic system In the monitoring of urban environment dynamic system, the data content of the system is more complex. From the category attribute, it can be divided into four parts:basic geographical space, dynamic realtime monitoring, environmental governance, comprehensive analysis, etc. [5]. As shown in Fig. 9, the basic data analysis of urban environment dynamics shows that the basic data is based on the application of geospatial data, while controlling the state of node information of different regional network layout in the city, the spatial geographic information content of the whole city is analyzed. Fig. 7 basic data of urban environmental dynamics Figure 10, urban environment dynamic real-time monitoring diagram, we can see that dynamic realtime monitoring is based on the collection of environmental database, the operation status of different network nodes in the environmental database is monitored and managed, and the good condition statistics and analysis are carried out. Environmental governance is to evaluate the network operation and the layout of relevant nodes for the prediction of environmental conditions and the analysis of the management mode;The comprehensive analysis system is based on the analysis of the above environmental secretary system, through the status and situation of network nodes in different urban functional areas to sort out, count and divide, and then get the most scientific and reasonable adjustment mode method [6]. Conclusion The process of urbanization is accelerating, and the pollution and destruction of ecological environment have become an important factor affecting human survival and development. Therefore, only by continuously improving the dynamic monitoring technology of the urban air monitoring system, can the monitoring and prevention of atmospheric environmental pollution be improved while the resource input cost is constantly reduced. Based on the operating principles of ZigBee and GIS technology, this paper realizes the data transmission and interaction of different system layers in the monitoring system through the node layout of the global positioning system, and improves the visualization in the monitoring process. Therefore, in the future development of the industry, its technical principles and data theory has a high promotion value and role.	2019-12-05T09:25:15.549Z	2019-11-01T00:00:00.000	{ "year": 2019, "sha1": "0274babe1d6be5f5581237c932e1153c35412482", "oa_license": null, "oa_url": "https://doi.org/10.1088/1742-6596/1345/6/062057", "oa_status": "GOLD", "pdf_src": "IOP", "pdf_hash": "7164f84bb2cbaec2734fec2b57c430d84c507e8d", "s2fieldsofstudy": [], "extfieldsofstudy": [ "Computer Science" ] }
252033233	pes2o/s2orc	v3-fos-license	Type 1–3 Canal Configuration in the Buccal Root of a Maxillary Second Molar Objectives A major challenge in root canal treatment is the complex and variable root canal system. There are several methods of classification of root canal systems. Herein, we report a case of a maxillary second molar, the root canal system of which could not be classified according to any of the previous methods. Case Report. Cone-beam computed tomography (CBCT), used to visualize the root canal system of a maxillary left second molar with fused roots, revealed a type 1–3 root canal system in the buccal root and a type 1–1 root canal system in the palatal root. A dental operating microscope was used throughout the treatment, and the root canals were successfully obturated. Conclusion To the best of our knowledge, this is the first report on the classification of buccal roots of maxillary second molars. Clinical Significance. A dental operating microscope and CBCT play a vital role in mapping the root canal system to ensure that no canal is missed. Introduction A major challenge in root canal therapy is the existence of complex and variable root canal systems. To fully understand the internal morphological changes in different teeth, several studies have been conducted, and root canal systems have been classified into different categories [1,2]. This provides the dentist with information about the root canal system. When encountering root canal systems beyond the classification, the dental operating microscope (DOM) is an effective device to help dentists identify specific anatomic features of the root canal [3]. The minimum distance that the human eye can distinguish is 0.1 mm, and more details can be visualized using a DOM. Moreover, use of a DOM ensures that the ultrasonic tip accurately removes the calcification that blocks the root canal orifice, reduces unnecessary grinding of healthy dentin, and improves the fracture resistance of the teeth [4]. In addition, cone-beam computed tomography (CBCT) can predict root canal location in teeth with unusual root canal anatomy [5]. Vertucci described three types of root canal systems in maxillary second molars. The variation mainly occurs in the mesiobuccal root, which is divided into type I, type II, and type IV [1]. Regarding the maxillary second molar, some studies have reported three canals in the mesiobuccal root [6], while other studies have reported two canals in the palatal root [7][8][9][10]. Herein, we present a case of a maxillary left second molar with a trifurcated buccal root canal. The DOM and CBCT showed two canals at the orifice level; however, the buccal canal trifurcated at the middle third of the root. To the best of our knowledge, there is no relevant report on the classification of buccal roots of the maxillary second molar. Case Presentation A 44-year-old woman presented to the Department of Operative Dentistry and Endodontics with pain, both spontaneous and to extreme temperatures, on the left side of the face, two days prior to her visit. Clinical examination revealed dental restoration in teeth #15, which had been placed one year prior at a different hospital. Tooth #15 had severe pain on percussion, and a severe lingering painful response to cold water, whereas tooth #14 was only tender to percussion. The periodontal conditions of teeth #14 and #15 were normal. Tooth #15 was diagnosed with irreversible pulpitis. The patient was informed about the conditions of her teeth and was advised to undergo root canal treatment (RCT) on tooth #15. The patient provided informed consent for RCT. The preoperative radiograph ( Figure 1) showed dental restoration approximating the pulp chamber in tooth #15. After administration of local anesthesia, an access cavity was prepared in tooth #15 and paraformaldehyde was placed into the pulp chamber to inactivate the dental pulp. Seven days later, the temporary seal was removed, and the root canals were visualized under a DOM (Zumax, Suzhou, China). A #10 K file (Dentsply Sirona, York, PA) was used to explore the root canals under 5× magnification using the DOM. The resistance to exploration was significantly lesser in the buccal root canal than in the palatal root canal. Next, the magnification was set to 20×, and a C-shaped root canal was observed in the buccal root with some dentin blocking the direct access. An ultrasonic working tip (ET20, Satelec, France) was used to remove part of the dentin; four root canals were found, three of which were in the buccal root and had a common orifice, and one was in the palatal root ( Figure 2). The patient was recommended to undergo CBCT (NewTom, Italy) because of the morphological variation in tooth #15 to avoid missing other canals. CBCT showed that the root canal was trifurcated at the middle third of the fused buccal root, and the palatal root contained a single independent root canal ( Figure 3). An electronic apex locator (Propex II, Dentsply, USA) was used to measure the length of the root canal, and pathfiles #13, #16, and #19 (Dentsply Sirona, York, PA) were used sequentially to dredge the root canals. The canals were prepared using #20 and #25 Protaper-next nickel-titanium rotary instruments (Dentsply Sirona, York, PA). Ethylenediaminetetraacetic acid (EDTA; Glyde, Dentsply, USA) was used for each root canal preparation. Frequent irrigation was performed using 2.5% NaOCl with P5 Newtron ultrasonic agi-tation (Satelec, France). The canals were rinsed, dried, and filled with calcium hydroxide paste. The access opening was sealed with a zinc oxide and eugenol dressing. Seven days later, the canals were emptied and copiously flushed with 2.5% NaOCl. Ultrasonic irrigation of the root canals was performed, followed by drying with paper points. Master gutta-percha cones were selected for all canals (Figure 4a), and the canals were obturated using iRoot SP (Innovative BioCreamix, Vancouver, Canada), using the single-cone technique. Two weeks later, a computer-aided design/manufacturing ceramic crown was fabricated on tooth #15. A final radiograph (Figure 4b) was taken to confirm the completeness and extension of the root filling. Discussion Despite the high success rate achieved in RCT, morphological variations in the root canal are an unresolved mystery. Variations in root canal anatomy are closely associated with failure of RCT because of the difficulty in locating, cleaning, and filling the aberrant canals [1,2,[6][7][8][9][10][11][12]. Owing to the many dissimilarities in selection, the results of previous studies cannot be compared directly. In an in vitro study, 42.25% of 187 extracted maxillary second molars had fused roots. Among teeth with fused roots, those with three-root fusion had a high frequency of merged canals [10]. Another study using CBCT to evaluate the incidence of root fusion and root-canal fusion in 4120 molars in vivo found that a complex root canal system is often present in fused roots. In this study, the incidence of root fusion and root-canal fusion in the maxillary second molar was 25.2% and 8.6%, respectively [13]. Another in vivo study using CBCT showed a 23.9% incidence of fused roots in maxillary second molars, with root-canal fusion within fused roots observed in 10.6% of cases [11]. In our patient, CBCT showed a three-root fusion in tooth #15. Ordinola-Zapata et al. reported that maxillary second molars with fused roots have a high incidence of merged and C-shaped canals [12]. In our case, tooth #15 showed a distinctive variation. Two separate root canals originated from the pulpal floor: one buccal and one palatal canal. Under the DOM, a 3 Case Reports in Dentistry C-shapedconfiguration was visualized till the middle third of the buccal root, which trifurcated in the mesial, buccal, and distal directions and formed three independent root canals with independent apical foramina. In this case, the treatment was difficult because we had to locate and clean the three buccal canals and put three gutta-percha cones through the same orifice at the same time. Fortunately, DOMs are being increasingly used for endodontic treatment, and are especially useful in such cases. CBCT can accurately image the maxillofacial hard tissue and confirm the internal anatomy of the teeth that need RCT [14][15][16]. At present, irrespective of the instruments used, some areas of the root canal cannot be cleaned. Therefore, chemical disinfection of root canals is the key to successful treatment [17]. In our case, EDTA was used to remove the smear layer and NaOCl and ultrasonic irrigation were used to reduce bacterial load and dissolve organic tissues, thereby improving the success rate. Conclusion Finding and treating additional root canals can improve the success rate of RCT. Therefore, clinicians should be aware of the typical root canal configuration. On encountering an aberration, clinicians should ask questions and use the DOM and CBCT in a timely manner to deal with various complex situations. Particularly, during RCT of the maxillary second molar, if a fused root canal is suspected on radiographs, it is recommended that the entire treatment be performed under a DOM. CBCT should be performed to confirm the root shape and number of root canals to avoid missing root canals. We believe that the prognosis of RCT Case Reports in Dentistry can be improved with standardized treatment combined with the use of advanced equipment. Conflicts of Interest The authors declare that they have no conflicts of interest.	2022-09-03T15:15:30.060Z	2022-09-01T00:00:00.000	{ "year": 2022, "sha1": "b412b289faf908b5ea74c4e85efeea1858a5b4f4", "oa_license": "CCBY", "oa_url": "https://downloads.hindawi.com/journals/crid/2022/8991579.pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "70a0a31605c5249c1d3768b0dbfbe4fec9f7408c", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [] }
252748093	pes2o/s2orc	v3-fos-license	Energy Loss After Daily Role Stress and Work Incivility: Caring for Oneself with Emotional Wellness The present study seeks to build on burnout research by examining daily fluctuations of role stress and work incivility, and their impact on employees’ energy loss. Optimism and recovery (i.e., psychological detachment and relaxation), two mechanisms that allow workers’ self-care and self-defence from these toxic conditions when faced by these job stressors, were included. In a daily study, 117 service sector workers completed surveys three times a day, over a period of one working week. Hierarchical linear modeling revealed work incivility as predictor of daily emotional exhaustion. Optimism and recovery play different roles in protecting workers from daily energy loss. Daily optimism increased employees’energy and decreased emotional exhaustion and negative affect at night. It also moderated the relationship between work incivility and positive affect at night. The results on psychological detachment supported the stressor-detachment model (Sonnentag, 2010), in which psychological detachment from work during nonworking time is not only a direct predictor of increased energy, but could similarly buffer the negative impact of role stress and work incivility. Relaxation basically showed main effects in predicting emotional exhaustion (inversely) and positive affect (directly). Our findings suggest two main implications: (1) the necessity for implemention of workplace policies to prevent role stress and work incivility in reducing daily loss of energy. (2) Training workers in self-care programmes focusing on optimism and recovery can provide early steps toward organizational change and employee daily well-being. Introduction Psychosocial risks have important adverse consequences on the health of people, organisations and economies (Schaufeli & Enzmann, 1998). Employees are undergoing turbulent times at work not only because of the high levels of psychosocial stressors such as workload, role stress and workplace mistreatment but also because of the high job insecurity and burnout (Jiang & Probst, 2017;Karatepe et al., 2020). Paradoxically, we know more about the important benefits of occupational health and uch research points out the importance of unceasing organizational efforts to resolve this problem, including more integrated perspectives to understand employees' experiences both in and outside of the work environment, as well as caring practices for employees, their families and organizations (Braganza et al., 2020;Tetrick & Winslow, 2015). From this perspective, it makes sense for the employees to identify proactive self-care behaviours, to effectively solve work problems and remain healthy, and to reduce the negative consequences in their personal lives. Workers also need to be physically and psychologically fit in order to maintain high levels of energy, effort, and engagement over time (Bakker, 2011;Casper & Sonnentag, 2020, Sonnentag, 2015 and to competently resolve daily work problems with a sense of commitment. However, little research has been carried out concerning the study of energy loss due to the combination of two relevant stressors in the field of burnout, such as role stress and work incivility, specifically taking into account personal resources and its mechanisms as workers' recovery experiences, controlling as well the effect of optimism. Surprisingly, despite its benefits, optimism has received little attention in work settings. Therefore, the current study explores the relationships between role stress, work incivility and energy loss-including emotional exhaustion and affect-using a diary design. This study also examines the underlying mechanisms of the association between theses variables, incluiding personal resources (i.e., optimism and recovery). From this perspective, we study the specific effect of the analyzed variables, i.e. whether the relationship of role stress and work incivility differs when the moment of the day is controlled and its predictive effect depends on the criteria variable (i.e. exhaustion and negative and positive affect at night). In addition to this, considering both positive and negative criteria variables allow us to deepen in the knowledge of preventative programs. Thus, taking into consideration optimism and recovery as protective factors provide us insight into how these variables could help to deal with negative situations by re-establishing energy levels. On the contrary, this allow us to study whether the analyzed job stressors may have a profound impact on workers' well-being. The knowledge about how role stress and work incivility affect at a daily level lead to determine how to prevent this within the workers' routine. These facts also boost the development of changes that organizations and workers need to put in practice. Working with the optimism as a trained skill help to protect the workers' well-being, facilitating healthy environments and positive expectations, similar to recovery experience and its role in the increase of energy levels. Hence, both constitute resources that could be used in a personal and organizational level and boost the understanding of how employees can become positive, energetic and engaged, even when facing undesirable work conditions. Theoretical framework and hypotheses Our study is built on the occupational well-being theory of the job demands-resources model (JD-R; Bakker & Demerouti 2007;2017). According to this model, daily job stressors and workplace mistreatment (i.e., role stress and work incivility) explain the daily loss of energy through high exhaustion and negative affect, and low positive affect at night. Within the theories of organizational stress, exhaustion is described as a loss of energy and this loss is well-known derived from emotional process (Cooper, 1998). Exhaustion linked with negative emotions involve a loss of energy that could weaken the person. Such losses of energy are critical because they are associated with burnout (Maslach et al., 1997). Burnout is a syndrome consisting predominantly of chronic exhaustion and negative attitudes toward work and it predicts adverse functioning in the workplace (Bakker et al., 2014). Furthermore, personal resources are expected to buffer the undesirable impact of job demands on strain, and increase the desirable power of job demands on commitment. Research has provided only reduced support for this suggestion, which means that more research is needed to test the potential interacctions of personal resources (Bakker & Demerouti, 2016;Bakker & van Wingerden 2020). From this perspective, our proposal aims to study how this composition of high job demands and low personal resources (JD-R model; Xanthopoulou et al., 2009) constitutes a work environment of high stress, possibly leading to a daily emotional exhaustion and lack of energy. This is a process where the workers' job stress may depend on the job demands and resources of the daily work routine (Bakker, 2014), being the work environment a critical factor to understand the workers' well-being (Cooper, 1998). However, workers do not linearly or simply react to the work environment (Bolger & Zuckerman, 1995). In this sense, the characteristics of the work and how these impact on them could actively intervene, for example, through self-regulation strategies, recovery experiences and their own personality (Prem, 2017). Role stress and work incivility Role stress is one of the most important job stressors predicting job-related exhaustion and negative affect (Garrosa et al., 2011;Jin et al., 2020;Lee & Ashforth, 1996;Schmidt et al., 2014;Shaufeli & Enzmann, 1998). Thus, from the perspective of the JD-R theory, this stressor is considered a job demand that involves continued emotional and cognitive effort (Demerouti et al., 2001), and its associated with physiological and psychological costs (e.g., fatigue, depression), and increases the probability of burnout (Alarcon, 2011;Bakker & Schaufeli, 2000). Emotional exhaustion is often cited as a consequence of role stress and energy loss, thus experiences of stress finally diminish emotional resources (Posig & Kickul, 2003). In addition, role stress in women is related to discrepancy strain (Gillespie & Eisler, 1992), which explains the relationship of role congruity with beliefs about expectations for women (Eagly & Karau, 2002). As such, altogether the evidence suggests the likelihood of a direct relation between these variables. This hypothesis help to understand the role of time concerning the drawbacks of daily role stress in the workers' well-being at bedtime. In this sense, we deepen understanding of how daily symptoms of exhaustion related to the daily role stress suffered could have a long-lasting and severe levels of exhaustion and energy loss through an accumulative process. Hypothesis 1 Day-specific role stress will be positively related to: (H1a) exhaustion at bedtime and (H1b) negative affect (NA) at bedtime, and negatively related to (H1c) positive affect (PA) at bedtime. Workplace mistreatment has received an increasing amount of attention from researchers (Aquino & Thau, 2009;Barling et al., 2009;Garrosa et al., 2015;Li et al., 2020;Yang et al., 2012). Workplace mistreatment comprises different forms, ranging from incivility to physical aggression. In this study, we are specifically interested in workplace incivility, a lowintensity mistreatment with an ambiguous intent to harm the target and violating norms for mutual respect at the workplace (Andersson & Pearson, 1999). Despite its vague intention, it has important consequences (e.g., burnout, dissatisfaction, lack of creativity and cooperation, psychological distress, negative affect) and, according to Cortina's theory, it involves selective incivility toward women and could be considered a "modern" expression of sexism in the workplace (Cortina et al., 2011). From the perspective of JD-R theory, workplace incivility is a job demand, an interpersonal stressor that most likely demands and depletes resources and energy at work (Rhee et al., 2017). Considering the previous arguments, we predict that when employees experience daily work incivility, they are exhausted at bedtime. This hypothesis provides information on how workers would use their resources to face the threatening situations and defend them from work incivility (e.g., disrespect). This daily work incivility is associated with negative emotions, including as well the time effect to know more about this mechanism among these variables. Specifically, work incivility may hurt psychological needs like the need to be both respected and appreciated by others (Baumeister & Leary, 1995) and, consequently, initiate different emotional and affective responses with daily consequences in energy loss (i.e., high exhaustion, high NA, and low PA, everyday). Studies where short-term effects are considered can help us to understand that daily consequences are not small and provide information about how everyday incidents disturb us. Hypothesis 2 Day-specific work incivility will be positively related to (H2a) exhaustion at bedtime and (H2b) NA at bedtime, and negatively related to (H2c) PA at bedtime. In view of these undesirable job stressors, we wished to investigate mechanisms that assist workers' self-care and self-defence from these toxic conditions. Emotional wellness: Optimism and recovering In this study, we focused on two variables that are well known to have a positive influence on emotional wellness: Optimism and recovery. According to JD-R theory and its further developments and applications, these personal resources provide resilience and protection, as they an positively moderate the relationship between the loss of energy, derived from daily job stressors, and negative consequences (Xanthopoulou et al., 2009;Kinnunen et al., 2011). As an example, Van Yperen and Snijders (2000) have observed the moderator role of self-efficacy and optimism in the relationship between job demands and psychological health symptoms. It seems that in situations of work overload and high job demands, personal resources diminish the workers' negative perception of the stressors suffered due to the effort of work (Buruck et al., 2016;Salminen et al., 2014). These studies suggest that workers with high levels of personal resources posit a great domain over themselves, and in turn, this helps them to effectively handle hazardous working conditions such high workload and work incivility, avoiding the negative outcomes. Thus, personal resources such as optimism and recovery not only could contribute to achievi a positive environment, but also could determine the way that workers perceive, react and contribute to the work environment. Optimism research reflects a broad view of self-regulation, planning and managing achievable goals, and positive expectancies when confronted with obstacles (Carver & Scheier, 2014;Neff & Geers, 2013) found that optimists engaged in more constructive problem solving than did pessimists. If a person, for instance, feels insecure and susceptible, problems are more difficult to resolve, and she or he may feel stress and disengagement (Carver & Scheier, 1998;Garrosa et al., 2011;Mappamiring & Putra, 2021). Additionally, optimists have a better profile of emotional responses to adversity and their coping responses are healthier. Interest in daily optimism at the workplace is increasing due to its positive relation with the dynamics of well-being (Casper et al., 2019;Littman-Ovadia & Nir, 2014;Sonnentag, 2015), however, there are few studies about it. Therefore, optimism after work may also act as a buffer, reducing the consequences of daily job stresors, leading to protected levels of energy, and increasing PA, as we propose in the next hypotheses: Hypothesis 3 Day-specific levels of optimism will be negatively related to (H3a) exhaustion at bedtime and (H3b) NA at bedtime, and positively related to (H3c) PA at bedtime. Hypothesis 4 Optimism after working hours will moderate the within-person spillover of (H4a) exhaustion at bedtime and (H4b) NA at bedtime, and (H4c) it will increase PA at bedtime. The relation between job stressors and energy loss will be stronger when optimism is low than when optimims is high. Lack of daily recovery leads to an accumulation of exhaustion and fatigue, and subsequently to chronic loss of energy and illness (Geurts & Sonnentag, 2006). In contrast, effective recovery increases energy, productive behaviours and well-being (Parker et al., 2020;Sonnentag, 2003;Xanthopoulou et al., 2014). Research about recovery from work can help to understand how employees can become energetic, engaged, and healthy, even when confronting poor work environments. In addition, recovery can moderate the daily effect of spillover from work to home (Sonnentag & Fritz, 2015;Trougakos et al., 2008). This mechanism refers to the process of decreasing or eliminating physical and psychological symptoms of strain produced by job demands and job stressors (Craig & Cooper, 1992;Meijman & Mulder, 1998), by distancing oneself from these negative emotions in one's personal life. Therefore, this process is important to keep an inner balance and energy in other personal areas, and also to distancing oneself from a problem to better understand it and choose the best actions to resolve a toxic situation. Sonnentag and Fritz (2015) consider psychological detachment as a core recovery experience. For example, in Spanish workers, experiences of relaxation have produced important benefits in the intensification of positive emotions (Garrosa et al., 2013). This may occur through the prolonged activation of the functional system, neutralizing the effects of negative affect (Parker et al., 2020;Sonnentag & Fritz, 2007;Xanthopoulou et al., 2014). Thus, recovery is understood as a way of energy activation. In this vein, the workers have the needed energy to healthfully handle job demands (Quinn et al., 2012). This fact of workers feeling recovered and rested may lead them to be more proactive and adaptatively face job demands (Lazarus & Folkman, 1991). In other words, workers with high punctuation in recovery are likely to face job demands as their energy levels allow them to make an effort in solving job problems. Also, energy activation could amplify the thoughts repertoire with innovative and creative ideas (Quinn et al., 2012), resulting in a protector of the negative consequences of daily stress derived from high workload and work incivility. From the perspective of JD-R theory, recovery experiences constitute a personal resource by which the energy levels invested in the working day are recovered (Moreno-Jiménez et al., 2012). We propose the following hypotheses: Hypothesis 5 Day-specific levels of psychological detachment after working hours will be negatively related to (H5a) exhaustion at bedtime and (H5b) NA at bedtime, and positively related to (H5c) PA at bedtime. Hypothesis 6 Day-specific levels of relaxation after working hours will be negatively related to (H6a) exhaustion at bedtime and (H6b) NA at bedtime, and positively related to (H6c) PA at bedtime. Hypothesis 7 Recovery from work will moderate the within-person spillover of (H7a) exhaustion at bedtime, (H7b) NA at bedtime, and (H7c) PA at bedtime. The relation between job stressors and energy loss will be stronger when recovery is low than when recovery is high. To sum up, this study will contribute to the occupational health and personality literature in at least four ways. Firstly, by conducting a daily study, we will contribute to the burnout literature by shedding light on how employees through role stress and work incivility may affect personal resources and energy loss, with a day-to-day consequences. Secondly, since we utilize both an inter-individual and an intra-individual perspective, we will postulate insights on how frequency of emotional wellness actions impact the magnitude of change in levels of energy, thus increasing our understanding of how optimism and recovery (i.e., psychogical detachment and relaxation) fluctuate, and if these variables could diminish the daily consequences of role stress and work incivility. Research on flexibility patterns of different types of resources is scarce and critical to further increase our understanding of dynamic processes that occur at work (Casper et al., 2019;Salanova et al., 2010;Prem 2017). The introduction of this intra-individual approach and diary design in the field of occupational health allows us to go deeper into workers' psychological processes (i.e. the sequence of their psychological states) and to identify antecedents and consequences of the central phenomena we are exploring, taking into account their temporal dimension (Navarro et al., 2015). It is important to note that when we approach the study of workers' psychological processes, as is the case with this study, the emphasis has to be on states rather than traits. In this sense, diary studies are one of the most appropriate methodologies to assess those states and their relationships over time, and in the workers' natural contexts (Navarro et al., 2015). Thirdly, because we examine three variables for emotional wellness (i.e., optimism, psychological detachment and relaxation), we will elucidate their different processes in the workplace context. Fourthly, we expand the Job Demands-Resources theory (JD-R; Bakker & Demerouti, 2014;Bakker & Demerouti 2017) to develop a theoretical framework that allows us to explain how employees can improve daily energy at work through optimism and recovery, and to reduce job stressors. Figure 1 represents the proposed research model. Procedure and sample The participants of this study were employees in the service sector from a broad range of professional backgrounds, including trade retail, health, social work, consulting and education. We recruited study participants by contacting them through different associations from this sector in Madrid. After the associations had expressed interest in the study, information packages describing the study procedure were sent to the employees. As an incentive for participation, we offered feedback about the study results. Study participants filled in the general questionnaire and subsequently, they completed daily questionnaires three times a day, for five consecutive working days (Monday-Friday): (1) a morning item of sleep quality (to be completed after waking up in the morning, before going to work), (2) an afternoon questionnaire (to be completed after work), and (3) a night time questionnaire (to be completed before bedtime). Participants received clear instructions on how to complete the questionnaires at these three times, and the researchers emphasized the importance of following this procedure, encouraging the use of reminder alarms (e.g., watch and phone alarms) and maintaining regular contact (Green et al., 2006). Of the 135 survey packages distributed, 121 were returned (86.6% response rate). In 4 of these diaries, the participants did not fill in all the days, so finally, 117 general questionnaires and diaries were used for the current study. The final sample comprised 117 workers (74% female and 26% male). Mean age was 33.7 years (SD = 9.1), and mean professional experience was 18.0 years (SD = 39.6). Mean working time was 40.2 h per week (SD = 8.1), 51.3% reported having contact with people more than 80% of their working day, and 62% had a university degree. Measures We collected the data with the same general and daily pencil-and-paper questionnaires. Specifically, we assessed (1) daily sleep quality in the morning, (2) role stress and work incivility at work in the afternoon, and (3) optimism and recovery from work after working hours with the bedtime survey. Exhaustion, negative and positive affect were also reported at night. Daily measures of all the variables used modifications of items from the corresponding general-scale, which were reworded for daily administration. Moreover, for daily measures, we used the same response categories as for the general measure. This method of developing state-level analogues of general measures has been used successfully in the past (Blanco-Donoso et al., 2015;Nezlek, 2012;Parker et al., 2020). Table 1 shows means, standard deviations, zero-order correlations between study variables, as well as the means of Cronbach's alphas averaged across the days of data collection. Daily role stress at work We measured role stress at work with the end-of workday survey, using scale items from the Role Stress Scale developed by Rizzo et al., (1970). Participants had to report how they had experienced role stress "today, during work" by responding to the items on 4-point Likert scales from 1 (not true at all) to 4 (totally true). In order to limit participant burden, we used 8 items to assess role stress (e.g.,''I have had to work under vague directives or orders.", "I think I have had many tasks in the same time."). Daily incivility at work We used 6-items from the Workplace Incivility Scale (WIS; Cortina et al., 2001) to measure personal experiences of uncivil behavior. Participants responded on a 5-point scale from 1 (not true at all) to 5 (totally true), describing whether they had experienced each behavior by a coworker or supervisor (e.g., "Today you have been in situations in which one of your superiors or coworkers doubted your judgment on a matter over which you had responsibility. Gave you hostile looks, stares, or sneers"). Optimism We assessed optimism with the bedtime survey, using the 10-item Life Orientation Test-Revised (LOT-R; Scheier et al., 1994). This time, we instructed our participants to report how they felt at this moment. Each item is rated on a 7-point Likert scale ranging from 1 (not true at all) to 7 (totally true). A sample item is: "I don't get upset too easily". Recovery from work We measured psychological detachment and relaxation from work after working hours with the bedtime survey, using the 6-item Recovery Experience Questionnaire (Sonnentag & Fritz, 2007) in its Spanish version (Sanz-Vergel et al., 2010), rated on 8-point Likert items from 1 (not true at all) to 8 (totally true). A sample item is: "Tonight, I forgot about work". General and daily exhaustion The Job-Related Exhaustion Scale (Wharton, 1993) provided a measure of emotional exhaustion, the primary dimension of job burnout (Lee & Ashforth, 1996;Wright & Cropanzano, 1998). The 6-item scale refers to the frequency of feeling "emotionally drained" and "burnt out", and ranges from 1 (never felt this way) to 6 (felt this way every day). The scale was modified so that the items refer to nighttime. An example item is: ''At this moment, I feel burnt out''. General and daily affect Positive and negative affect were measured using the 10-item International Positive and Negative Affect Schedule Short-Form (I-PANAS-SF; Thompson 2007) with five items measuring PA and five items measuring NA. Participants rated the extent to which they generally experience each of the 10 emotions on a 5-point scale ranging from 1 (never) to 5 (always). The scale was modified so that PA and NA referred to nighttime. An example item is: ''At this moment, I feel inspired''. For all these expected relationships, we took into account the potential effects of different control variables: Sex, the trait aspect of the corresponding criterion variable and sleep quality. Sleep quality was evaluated in the morning with the item "How do you evaluate your last night's sleep?" derived from the Pittsburgh Sleep Quality Index (Buysse et al., 1989), rated on a 6-point Likert scale ranging from 1 (very bad) to 6 (very good). This measure correlates highly with the full Pittsburgh Sleep Quality Index (Sonnentag & Binnewies, 2013). Since working conditions are currently different depending on sex, it is necessary to control this variable, the general measure of the criterion variable is used to carry out multilevel analysis, and sleep quality is the typical control variable in burnout studies as it affects to daily energy levels. Data analysis Day-level data were nested within persons. We then analyzed the data with a randomcoefficient modeling approach (Bryk & Raudenbush, 1992) using the software package MLwiN (Rasbash et al., 2000). To guarantee that the analyses reflected our concern about within-person variation and used unbiased estimates of the day-level variables (Enders & Tofighi, 2007), we centered day-level predictor and moderator variables at the person mean. We tested the proposed hypotheses in a sequence of multilevel models, with the first ones including the control variables (Model 1), then, the main effects of role stress and incivility at work (Model 2), then, the main effects of the moderator variables of optimism (Model 3) and recovery (Model 4), and then the interaction effects (Model 5). We compared these models with Chi-square tests, using the difference of the − 2 × log likelihoods of the nested models (with degrees of freedom corresponding to the number of new parameters added). Preliminary analyses Prior to testing our hypotheses, we scanned the degree of within-person and between-person variations of our day-level variables. Table 1 shows the percentage of within-person variance (1-ICC = 1 − (between-person variance /[between-person variance + within-person variance]) derived from intercept-only models. All predictor variables showed an ICC above 25% (Hox & Roberts, 2011) with the exception of exhaustion at night (20%). Largely, these findings suggest that a substantial portion of the variance in our variables can be attributed to within-person variation across the five days. We have also considered exhaustion at night in a multilevel approach because, according to Nezlek (2008), even when the ICC approaches 0, multilevel modeling might still be essential, and this circumstance was coherent at the theoretical level. Exhaustion at bedtime. For exhaustion at bedtime as an outcome measure (see Table 2), Model 1, with sleep quality, sex, and general exhaustion as control variables, improved the null model (difference − 2XLog = 98.58, df = 3, p < .001) with sex and general exhaustion as positive predictor variables (t = 2.23, p < .01, and t = 8.36, p < .001, respectively). By adding daily role stress and incivility at work (Model 2), the fit improved (difference − 2XLog = 28.38, df = 2, p < .01), with positive relationships with exhaustion at bedtime, role stress (t = 3.16, p < .01) and work incivility (t = 3.88, p < .01). Model 3, which included optimism, improved the fit, and showed a negative association with exhaustion at bedtime (t = -2.33, p < .01). Model 4 significantly improved the fit (difference − 2XLog = 36.24, df = 2, p < .001), also showing a negative relationship between relaxation and exhaustion at night (t = -4.46, p < .001). In Model 5, the effect of interaction term between role stress and psychological detachment increased the fit (t = -2.09, p < .05). To gain more insight into the nature of this interaction effect, we conducted simple slope tests, using the multilevel tool developed by Preacher et al., (2006). These simple slope tests showed that role stress was more positively related to exhaustion at night on days when people experienced less psychological detachment (γ = 0.457, SE = 0.118, z = 3.84, p < .001), whereas they were unrelated on days when psychological detachment was high (γ = -0.136, SE = 0.180, z = − 0.75, ns) (see Fig. 1). The hypothesis about the interaction between role stress and psychological detachment was also supported. In addition, when this interaction is significant, rol stress lost its significance as predictor. NA at bedtime. When NA at bedtime was included as an outcome variable (see Table 3), Model 1 showed a better fit than the null model, with sleep quality and general NA as predictors (t = -2.39, p < .05 and t = 2.94, p < .01, respectively). Model 2 improved the previous models, but role stress and work incivility had no significant influence on NA at bedtime. Model 3 increased the variance (-2XLog difference = 8.24, df = 1, p < .01) through optimism as a negative predictor of NA at night (t = -2.91, p < .01). Model 4 included recovery experiences, but there were no significant predictors. Finally, Model 5 increased the fit, with psychological detachment as a moderator of the relationship between work incivility and NA at night (t = -2.17, p < .01). Simple slope tests showed that workplace incivility was more positively related to NA at night on days when people experienced less psychological detachment (γ = 0.670, SE = 0.254, z = 2.64, p < .01), whereas they were unrelated on days when psychological detachment was high (γ = -0.362, SE = 0.242, z = -1.49, ns) (see Fig. 2). Overall, the results supported the hypothesis about the interaction, because psychological detachment is a negative predictor of NA at bedtime, with an interaction between incivility and psychological detachment. PA at bedtime. Lastly, with PA at bedtime as an outcome measure (see Table 4), Model 1 fit the data better than the null model, and again, sex was a strong negative predictor of PA at bedtime (t =-3.17, p < .01), and general level of PA was a predictor of PA at bedtime (t = 2.16, p < .05). Role stress, work incivility, optimism included in Models 2 and 3 sequentially, were nonsignificant. Model 4 showed a highly significant improvement over the previous models (difference − 2XLog = 30.26, df = 2, p < .001), with relaxation as a robust predictor of PA at bedtime (t = 4.94, p < .001). Model 5 also increased the fit, and there was a significant effect of the interaction between work incivility and optimism (t = 2.00, p < .05). Simple slope tests (see Fig. 3) indicated that workplace incivility was more positively related to PA at night on days when people experienced more optimism (γ = 0.666, SE = 0.318, z = 2.09, p < .05), whereas they were unrelated on days when optimism was low (γ = -0.573, SE = 0.341, z = -1.67, ns). These results only supported the moderator effect between optimism and work incivility. Discussion Our day-level study showed that predictor variables that are implicated in daily energy have different components in relation to variability, affect symmetry that involve the reactions to positive events and experiences are reflected mainly in positive affective states and reactions to negative events and experiences are reflected mainly in negative affective states (Sonnentag, 2015) and the involved mechanism. These findings show that exhaustion is more stable and does not undergo so many fluctuations at the daily level as NA and PA. Daily exhaustion depends to a great extent on the general levels of chronic exhaustion. Exhausted workers are probably already burnt out (Bakker et al., 2014), disengaged and are lacking energy (Casper & Sonnentag, 2020;Demerouti et al., 2001). Our study provides new insights, showing that daily work incivility presented a significant effect in daily exhaustion, even more than role stress, and could have negative consequences for workers' emotional health. An intensification of physiological dysregulation is likely to accelerate the development of disease (McEwen & Wingfield, 2003), especially when workers are already exhausted in a general way and the effect of work incivility adds up (e.g., disrespect among other elements), producing consequences for daily exhaustion at night time. Exhaustion fluctuates within shorter periods of time in the week and can increase over longer periods of time. In this sense, this accumulative increase of exhaustion may clarify the relation between burnout and depressive symptoms and life dissatisfaction (Hakanen & Schaufeli, 2012). Exhaustion compounds sad affect and, according to Teasdale's (1988) work, these feelings can lead to a self-defeating cycle of escalating negative thoughts and emotions, which, in turn, can increase the risk of an eventual major depressive episode (Beevers & Carver, 2003;Lau et al., 2004). It has been confirmed that burnout predicted antidepressant treatment, with a stronger association in men than women (Madsen et al., 2015). This study extends earlier burnout research and shows that, to prevent exhaustion at the day level, zero tolerance of work incivility and respect for both the self and others in everyday tasks are neededas, once workers experience loss of energy, the recovery is complex. Daily work incivility activates a stress reaction (exhaustion at night-time) that remains persistent, even after working hours have ended and the initial stressor is no longer present (Sonnentag, 2015). Concerning the included variables, only daily relaxation and optimism contribute to the mitigation of this adverse cycle in workers' everyday life, for example, by facilitating positive expectation, effective mood repair, adaptive self-regulation strategies, as well as enhanced psychological functioning (Bakker & van Wingerden, 2020;Carver & Connor-Smith, 2010;Casper et al., 2019;McHale et al., 2015;Scheier & Carver, 1992). Relaxation facilitates recovery because it limits the prolonged activation of the functional system and counteracts the effects of emotional exhaustion (Parker et al., 2020;Sonnentag & Fritz, 2007). These results contribute to the notions that asymmetry is not persistent, and that personal resources can also contribute to offset the negative strain process (Sonnentag, 2015). Thus, daily emotional wellness had direct effects on daily exhaustion at bedtime, specifi- cally with optimism and relaxation. Also, a moderator effect of recovery through detachment in the relationship between role stress and exhaustion at bedtime was found. Unexpected results revealed that job stressors had no direct effect on NA and PA at bedtime. According to Friedman and Kern (2014), NA would be more related to individual dispositions. These results, especially those non-significant, support that emotional exhaustion is a consequence more related to stressful job situations such as work incivility, whereas NA and PA are more related to workers' capacity of self-regulation and their recovery possibilities. Our findings support this idea and extend the power of daily optimism to its negative association with NA at night. An explanation for this could be that optimists work determinedly to counteract these negative moods and achieve mood repair, which contributes to enriched emotional functioning and prevents burnout (Garrosa et al., 2011;Mappamiring & Putra, 2021;McHale et al., 2015). This study extends previous research showing that these mechanisms occur at the daily level and are essential to maintain energy during the working week and not allow NA to affect one's personal life. Furthermore, daily optimism is more likely to develop positive evaluations, such as the belief that their problems will have solution, which in turn reduce the impact of NA. Regarding PA, we observed a major effect of symmetry due to the positive valence of the predictor variables and a valuable relationship with proactive behaviors such as relaxation, which has also been shown to predict an increase in work engagement (Simbula et al., 2011;Sonnentag, 2003). Furthermore, our results show that relaxation is an important resistance factor against energy loss. However, the result of no direct relationship between optimism and PA was unexpected. This may be partially due to the multilevel study methodology employed, in which multicollinearity is diminished. However, optimism resulted in being a positive predictor in the interaction with work incivility, observing its protective effect as an emotional equilibrium. This protective effect maintaining PA could be explained because optimistic workers see negative events (i.e., work incivility) as not being their fault. They also see them as being flukes that have nothing to do with other areas of their lives or future events, linked with the use of better regulation strategies and effective coping skills against job stressors. The moderating effect of psychological detachment as a recovery experience from adverse conditions is confirmed, as it buffers the negative effects of role stress in the face of exhaustion, and of incivility and NA. However, it had no effect on the increase of energy through PA. On the days when workers did not achieve psychological detachment, role stress was more positively related to exhaustion at night. Equivalently, workplace incivility was more strongly associated with NA at night on days when people experienced less psychological detachment. These results support the stressor-detachment model (Sonnentag, 2010), which states that psychological detachment from work during nonworking time is not only a direct predictor of increased well-being and energy but could also buffer the negative impact of job stressors (Sonnentag & Fritz, 2015). Thus, our study extends previous research by revealing that the effect of psychological detachment can moderate the daily effect of role stress and work incivility. This may occur through an attentional process, for example, by focusing thoughts toward other more beneficial life areas which in turn increases self-confidence to cope with job stressors and decreases negative thoughts during nonworking time. This is particularly important in the face of work incivility, such as low-intensity mistreatment, which undeniably has consequences on self-esteem (Andersson & Pearson, 1999) and which may be a precursor of other forms of aggression of greater intensity, such as bullying (Cortina & Magley, 2003). By disconnecting, a worker's selfesteem does not suffer as much, and he or she may be capable to stop being mistreated at the workplace. This study provides more evidence about relaxation as a positive experience at the emotional and energy level. In our study, relaxation had a main effect, contributing to avoiding exhaustion. After general exhaustion, relaxation is the most significant predictor, as well as the most significant at the level of direct effects of PA at night, but it does not have moderator effects. Our study shows that psychological detachment and relaxation are predictors of daily energy, even after controlling for daily optimism, but they play different roles in this process because they intervene with different mechanisms. Relaxation would have a direct effect of reducing activation and increasing PA, perhaps contributing to, or ameliorating physiological dysregulation and thereby playing a key role in exacerbating or counteracting allostatic load/overload (McEwen, 2007). Therefore, processes that reduce this prolonged activation are crucial in order to restore an organism's pre-stress state, whereas psychological detachment would be a buffer. Relaxation probably generates the prior emotional state necessary to distance oneself psychologically, thereby contributing to breaking ruminative thoughts associated with depression (Nolen-Hoeksema et al., 1994). Some people have trouble disconnecting, probably because it requires a prior state of calmness and vitality that can be achieved through relaxation. This process would facilitate distancing and could be considered a precedent in future studies. It would be interesting to explore this relationship to include practical resources to help workers to distance themselves from their work in stress prevention programs. Sex is a significant predictor of exhaustion, and PA. Our results show higher scores on exhaustion and lower PA scores for female workers at the daily level. This is in line with results found in similar studies (Marinaccio et al., 2013). Future studies should investigate whether the cause is women' worse working conditions and/or that more difficulties to practice recovery. In addition to this, gender perspective -based program for women would take a relevant place in order to prevent these work stress-related issues. In this sense, the promotion of optimism as self-efficacy beliefs is related to important cognitions that may protect women from negative outcomes of job stress (Brown & Harris, 2012;Mappamiring & Putra, 2021). Women with a positive viewpoint have less stress, better creative problemsolving skills, and better health outcomes than less optimistic women. Similarly, the uses of time and work-life balance for women still have more barriers (Clercq & Brieger, 2021). In the same vein, poor quality sleep also explains NA, because sleep helps to preserve emotional well-being (Sonnentag & Binnewies, 2013) and could promote physiological balance and resilience (McEwen, 2006). In short, enhancing women's skills to better deal with time to boost recovery and co-responsibility in men are necessary aspects. Finally, daily optimism has been shown to be an important variable as a driver of energy, both at the level of main effect, decreasing exhaustion and NA, as well as moderating the relation between work incivility and PA. It has been observed that workers with high optimism maintain a high level of PA at nighttime, even with high work incivility. Nevertheless, it would be interesting to confirm the long-term effect of this interaction and determine whether it is related to actions such as denouncing these toxic situations at work, or whether the prolonged strain could end up exhausting the worker. Ultimately, whether it is an expression of active resource to achieving positive outcomes in the face of adversity, like coping mechanisms that operate at the time of trauma (Karatsoreos & McEwen, 2011). Concretely, studies have supported optimism's protective mechanisms against stress and its associations with skills to connect and obtain social support in stressful situations (Andersson, 2012). Thus, social support can both help define the extent to which the event is appraised as stressful and mitigate negative outcomes. We understand that feeling socially supported can protect individuals' well-being from the stress associated with incivility. In conclusion, this study contributes to the research on burnout and recovery, revealing the daily and undesirable loss of energy in workers due to work incivility and role stress, perhaps with chronic consequences. Conversely, the combination of optimism, relaxation and psychological detachment can contribute to reducing stressful circumstances and to improving daily energy. Likewise, the moderator resources help to reduce the consequences of negative emotions, daily spillover from work to home. For example, daily psychological detachment helps in the process of decreasing exhaustion and NA related to work incivility and role stress. This role is relevant even after controlling for optimism. Daily optimism could also increase energy through the increase in PA. In sum, these mechanisms are relevant to maintain an adequate energy balance and more healthy lifestyles in our personal life, and to distance oneself from work stress and improve active responses to adverse situations, such as workplace mistreatment and role stress, thereby avoiding energy loss, which predicts depressive symptoms and life dissatisfaction. Limitations and implications for future research The current study has some limitations: Firstly, we assessed all the data with self-reports because they are suitable to assess private experiences (Conway & Lance, 2010), but they involve the problem of common-method variance. However, collecting data with general and daily surveys and using person-centered scores in the analyses ensured the measurement of predictor and criterion variables at different time points each day (morning, afternoon, and night) and included control variables according to previous research. Although the temporality dimensions we focused in this article are limited, they provide us a valuable field to explicitly consider the role of time in work settings and health of workers. Future research should collect an in-depth analysis of time more extensively with the incorporation of the multiple dimensions of temporality (George & Jones, 2000), data from several sources (i.e., colleagues, family, and clients), more objective variables (e.g., biomarkers) and the extension of research findings from a study conducted on a sample to the population at large. Secondly, we used survey packages, following previous daily studies, which have the advantage of being low-cost, easy to administer, and being psychometrically equivalent to data obtained through electronic assessment, but they also have some disadvantages. For this reason, the participants received clear instructions to complete the questionnaire in the specified three moments and, for each moment, they had to indicate the exact time they finished. They were also encouraged to use reminder alarms (e.g., watch and phone alarms) and to maintain regular contact (Green et al., 2006). Although all participants reported a completion time within the corresponding hours, future research could use more creative procedures and a longer assessment period (i.e., two weeks or more) to determine how consistent these processes are over time (Schwarz, 2012). Thirdly, we explored the moderating role of daily optimism and recovery. Future research should also explore mediation within the multilevel context, for example, the sequential relation between relaxation, psychological detachment, and energy. Finally, like most researchers, we assumed that the relationships between the within-person variables are linear. However, many psychological variables do not follow a linear pattern (Burke et al., 2007). It is possible, for example, that the effect of work incivility is nonlinear because different uncivil behaviors are not equivalent. Further research on modeling nonlinear effects is necessary (Deboeck, 2011;Ladstätter et al., 2015). Practical implications Primary interventions may be more operative and effective in the workplace because, when loss of energy occurs, there are negative consequences on daily well-being, which can become chronic. In this case, recapturing initial motivation is more difficult. Therapeutic assistance may be required in this situation, revealing unsuccessful prevention of psychosocial risks. In addition to creating a healthy workplace with a suitable adaptation between demands and resources in the workers and the organization, a clear commitment to ethical values (i.e., zero tolerance of workplace mistreatment and discrimination) and supportive relationships in organizations will help workers to feel healthier (Morrison & Cooper-Thomas, 2013). The implementation of workplace wellness programs including learning skills to develop psychological detachment, personal resources such as optimism, and the use of relaxation techniques could be a self-care practice. Such a wellness program should necessarily include a better adaptation of working time to contribute to an optimal balance of work and life. Organizations can also generate principles for respectful and caring interpersonal interactions, growth in cooperating work and collective goals among employees, and build social support and commitment.	2022-10-07T15:03:45.145Z	2022-10-05T00:00:00.000	{ "year": 2022, "sha1": "2df283d43b847751e5647da4453773404c870762", "oa_license": "CCBY", "oa_url": "https://link.springer.com/content/pdf/10.1007/s10902-022-00570-x.pdf", "oa_status": "HYBRID", "pdf_src": "Springer", "pdf_hash": "de48f85f8cb864f9694f8dbf98e66b7bfa3baf46", "s2fieldsofstudy": [ "Psychology" ], "extfieldsofstudy": [] }

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.