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
259368439	pes2o/s2orc	v3-fos-license	Laparoscopic Repair of Bochdalek Hernia: A Rare Presentation of Abdominal Pain in the Elderly Bochdalek hernias (BHs) are rare, and the presentation, diagnosis, and management of them can be complex. We present a 70-year-old man presenting with left flank pain who underwent a successful laparoscopic repair of BH with mesh placement. Introduction One of the many achievements of anatomist Vincent Bochdalek was to describe a peculiar hernia that arose as an error of diaphragmatic development. It is known as the 'Bochdalek hernia' (BH), this congenital diaphragmatic hernia is the protrusion of tissue through a defect in the lumbocostal triangle of the diaphragm, also known as the foramen of Bochdalek. This foramen is a result of the improper fusion of the septum transversum and the pleuroperitoneal folds during the development of the hernia in gestation [1]. Case Presentation A 70-year-old man presented with increasing pain in the left flank on a background of a known BH. The hernia was discovered five years before and managed non-operatively as the patient lived rurally and had been asymptomatic. Contemporary imaging via computed tomography (CT) demonstrated that the hernia was now larger, and that there was a 15 mm defect in the posterior hemidiaphragm on the left side, with herniation of intra-abdominal fat, most likely from the retroperitoneum ( Figure 1). The herniation measured at least 14 mm craniocaudally and about 66 mm in maximal transverse dimension. Given the patient was having ongoing back and left flank pain, likely due to strangulation of the herniated intraabdominal fat, a laparoscopic diaphragmatic hernia repair of the foramen of Bochdalek was performed. Following the establishment of pneumoperitoneum and safe laparoscopic port placement, the short gastric vessels were divided to enter the lesser sac. The lateral and superior borders of the spleen were mobilised to improve the exposure of the diaphragm. The hernia was identified; the main findings were a small defect approximately 3 cm × 3 cm over the left dome of the diaphragm behind the left kidney, containing fat ( Figure 2), which was reduced. The defect was closed with interrupted prolene sutures. A polyester mesh was placed and fixated with surgical glue (Figure 3). The patient went on to make a full recovery and was eventually discharged from outpatient surgical follow-up. Discussion The incidence of BH in infants is approximately 1 in 4000 [2]. In contrast, the incidence of BH in adults is estimated to be as low as 1 in 7000 [3]. One reason that may explain this difference is that a large foramen of Bochdalek may lead to herniation during prenatal development, which results in presentation shortly after birth due to compression-induced pulmonary hypoplasia [4]. A small foramen, a limiting pleural sac, or an omental plug in the defect may explain why some are present later in adulthood [5]. Asymptomatic BH occurs in approximately 0.17% of the adult population [5]. Otherwise, symptoms are variable. Common complaints are pain or discomfort in the chest and abdomen and exertional dyspnoea. Severe pain with nausea and vomiting suggests incarceration or strangulation. The main precipitants are physical exertion, pregnancy, defecation, or exacerbation of asthma. Abdominal symptoms generally occur first, followed by respiratory symptoms [5]. The diagnosis of BH is challenging due to its rarity and its non-specific presentation in adulthood [3]. Case Reports in Surgery Diagnosis may be obtained from plain radiographs or CT. A chest X-ray may demonstrate gas and organs above the diaphragm, whereas a CT is accurate in identifying diaphragmatic defects and signs of ischaemia [5]. The usual CT findings are fat or soft tissue over the upper posterolateral surface of the diaphragm, or a mass next to the defect [6]. If a BH is suspected then it is recommended that a CT with contrast should be performed due to its sensitivity (50-78%) and specificity (100%) in visualising focal diaphragmatic defects [3] and identifying strangulation [1]. Due to its presentation, delayed and missed diagnosis are common and can result in lethal complications. Once discovered, patients should undergo surgical repair as soon as possible, as mortality and complications are as high as 32% in emergency surgery, whereas the mortality rate for elective surgery is less than 3%. This makes early diagnosis vital in reducing the morbidity and mortality of the disease [7]. The main complications of BH are obstruction, strangulation, perforation, and other coexisting anomalies. It was noted that up to 18% of cases were complicated by lifethreatening strangulation [1]. Given that strangulation rates of inguinal hernias are as low as 0.3%, there is a noteworthy risk of leaving BH untreated [8]. In the literature, both transthoracic and transabdominal approaches to surgical repair have led to similar outcomes [1]. Laparotomy was the most common approach and was performed in 38% of cases. Laparoscopic repair is increasing in popularity and can be performed with short hospital stays and complication rates as low as 7% [9]. Conclusion Given the potential complications, this study aims to educate surgeons about BH, especially since it is an uncommon diagnosis that can present non-specifically. Surgeons should consider a lower threshold for operating on these hernias in the elective setting to prevent the occurrence of avoidable complications. Data Availability All patient relevant data are available within the article. Consent Informed consent was obtained from the patient. Conflicts of Interest The author(s) declare(s) that they have no conflicts of interest. Authors' Contributions CS and JL should be considered co-first authors. SM contributed to the study concept. CS and JL contributed to the acquisition of data. CS, JL, and EW contributed to the writing of the manuscript. EW and SM contributed to the editing of the manuscript.	2023-07-09T05:08:33.599Z	2023-06-30T00:00:00.000	{ "year": 2023, "sha1": "dd786babe7f901582351e0d6b11f4a854704e535", "oa_license": "CCBY", "oa_url": "https://downloads.hindawi.com/journals/cris/2023/5361609.pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "dd786babe7f901582351e0d6b11f4a854704e535", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
264077065	pes2o/s2orc	v3-fos-license	Case Report: Efficacy of albumin dialysis for the reversal of bile cast nephropathy-induced acute kidney injury Background Bile cast nephropathy (BCN) is an underdiagnosed renal complication associated with severe hyperbilirubinemia and is seen in patients with liver failure who have cholestatic complications. BCN-induced acute kidney injury (AKI) can require hemodialysis (HD), and the molecular adsorbent recirculating system (MARS) is a potentially useful therapeutic option. Case summary A 57-year-old male presented with jaundice persisting for 1 month, with laboratory test results indicative of hyperbilirubinemia and AKI. Abdominal imaging and a biopsy confirmed biliary ductal dilation secondary to a pancreatic head mass. The patient had rapidly progressive renal failure and refractory hyperbilirubinemia, despite biliary decompression, and was started on HD. Subsequent therapy with albumin dialysis therapy using MARS was successful in reversing the AKI, the cessation of HD, and the restoration of native renal function. Conclusion In the setting of BCN-induced AKI, timely initiation of MARS can provide a useful therapeutic strategy to reverse renal dysfunction and facilitate intrinsic renal recovery. Introduction Bile cast nephropathy (BCN) is a rare and potentially life-threatening renal complication associated with severe hyperbilirubinemia, particularly in patients with cholestatic conditions or those experiencing liver failure (1)(2)(3).It is characterized by the formation of bile casts within the renal tubules, leading to acute kidney injury (AKI) and, in severe cases, the need for renal replacement therapy.The pathogenesis of BCN is not yet fully understood, but it is believed to involve the precipitation of bile acids, bilirubin, other bile components within the renal tubules, causing tubular obstruction, inflammation, and oxidative stress (2,3).Typically a diagnosis of exclusion, and one confirmed by renal biopsy, BCN is felt to be underdiagnosed.A recent study retrospectively evaluating renal biopsies found that BCN was present in almost 45%-72% of autopsies of patients with cirrhosis (4,5). The management of BCN presents a clinical challenge as current therapies primarily focus on addressing the underlying source of cholestasis and providing supportive care for the associated renal dysfunction.Treatment strategies include optimizing fluid balance, ensuring appropriate electrolyte management, and addressing the precipitating factors of liver dysfunction, such as obstructive jaundice, infection, or drug toxicity (2,3).In some cases, renal replacement therapy such as hemodialysis (HD) may be necessary to manage severe AKI secondary to BCN.In refractory cases, extracorporeal liver support devices utilizing albumin dialysis may help to reverse AKI (6,7). The molecular adsorbent recirculating system (MARS) is an extracorporeal liver support system that has shown promise in the treatment of various liver diseases, including acute liver failure (ALF), acute-on-chronic liver failure as a bridge to transplant, and intractable pruritus (8)(9)(10).MARS functions by removing albuminbound toxins, including bilirubin and bile acids, from the patient's blood through a combination of continuous renal replacement therapy (CRRT) and the addition of adsorption therapy via the MARS circuit (11,12) (Figure 1).With regard to BCN-associated AKI, the strategy of using MARS aims to reduce the burden of cholestasis to mitigate the toxic effects of bilirubin on the renal tubules and potentially promote renal recovery.The application of MARS in the treatment of BCN has been reported in a few case reports with promising results (6,7). In this case report, we describe a patient with obstructive jaundice and BCN requiring HD who was successfully treated using MARS.We will discuss the presentation, diagnosis, and management of the patient's BCN and the potential benefits of MARS therapy for patients with similar clinical profiles.Furthermore, we will explore the current literature on the utility of MARS for treating BCN and other liver-related renal complications, highlighting the potential for this innovative therapy to improve patient outcomes and reduce the need for renal replacement therapy. Case description A 57-year-old male with diffuse large B-cell lymphoma who had been treated with chemotherapy, but was in remission for over 10 years, presented with jaundice persisting for 1 month, fatigue, dark urine, and pruritus.On admission, the patient had scleral icterus, with laboratory test results notable for a total bilirubin level of 22.3 mg/dL (381.3 mmol/L) and a serum creatinine level of 2.0 mg/dL (176.8 mmol/L) with an unclear baseline.An MRI scan showed biliary ductal dilation and signal abnormality of the pancreatic head threatening malignancy.Subsequent endoscopic retrograde cholangiopancreatogram (ERCP) on hospital day 6 demonstrated a high-grade stricture at the distal common bile duct threatening external pancreatic mass compression with a subsequent unsuccessful biliary stent placement for decompression.Brushings were used as diagnostics for pancreatic adenocarcinoma.On hospital day 7, an internal/external biliary drain for biliary decompression was successfully placed via interventional radiology.Despite this, the patient had persistent hyperbilirubinemia that initially improved to 13.2 mg/dL (225.7 mmol/L) before increasing back to 21.9 mg/dL (374.5 mmol/L).The patient also had worsening renal function up to a maximum creatinine level of 6.3 mg/dL (556.9 mmol/L), with oliguria progressing to anuric renal failure requiring the initiation of high-flux HD with conventional dialysate on day 14.The workup to investigate secondary causes of liver and renal dysfunction, including serum and urine protein electrophoresis, viral and autoimmune hepatitis laboratory tests, urinalysis and culture, 24-h urine protein, bone marrow biopsy, and blood cultures, was unrevealing.A renal biopsy demonstrated acute tubular necrosis secondary to BCN (Figure 2).Given these findings, the patient was treated with sessions of MARS therapy.Figure 3 shows the total bilirubin levels decreasing after the MARS therapy.On hospital day 60, the patient was discharged with a total bilirubin level of 5.9 mg/dL (100.9 mmol/L).The patient still required HD at discharge.One month after discharge, the patient's urine output had increased, and his creatinine level had stabilized without HD at 1.5 mg/dL (132.6 mmol/L), and HD was discontinued.With the resolution of his hyperbilirubinemia and renal failure, he was able to commence chemotherapy and radiation without further delay, and he achieved a good clinical response with minimal side effects.At the 9-month follow-up he had maintained an Eastern Cooperative Oncology Group (ECOG) performance status score of 0 to 1 but was found to have small metastatic lesions within the liver and so was restarted on chemotherapy. Discussion Initially introduced in 1993 by two German nephrologists, Mitzner and Stange, MARS is an extracorporeal liver support system with applications in the treatment of various liver diseases, including ALF, acute-on-chronic liver failure as a bridge to transplant, and intractable pruritus, as well as drug toxicities (8)(9)(10)12).MARS gained accepted clinical use in the late 1990s.MARS works owing to the blood flowing countercurrently to an albumin dialysate solution that absorbs albumin-bound toxins.The albumin is then regenerated as it passes through activated charcoal and an anion exchange resin, while water-soluble toxins are cleared using a conventional dialysis solution (13).Other extracorporeal liver support devices exist, such as Prometheus ® and single-pass albumin dialysis (SCAD).Prometheus uses fractioned plasma separation and adsorption in which a conventional dialysis solution is used, although the patient's blood directly passes through an albumin-permeable filter when albumin-bound toxins are cleared (14,15).SCAD uses a 4.4% albumin solution, versus a 20% albumin solution in MARS, as a direct dialysate solution in addition to a conventional dialysate solution to remove both water and albumin-bound toxins (16). In this case report, we present a patient with BCN who was successfully treated using MARS in a critical care setting.Patients with ALF may present similarly with symptoms related to hyperbilirubinemia.MARS and other forms of extracorporeal albumin dialysis may be considered in ALF as a bridge to liver transplant (17).A multicenter randomized controlled trial of MARS in patients with ALF for emergency liver transplant showed no clear survival benefit of MARS therapy (n = 54) at 6 months and 1 year when compared with standard medical therapy (n = 49).However, in the secondary analysis, patients who received three or more sessions of MARS had significantly improved survival rates compared with patients who received fewer than three treatment sessions, suggesting that three or more sessions may be beneficial in patients with ALF (18,19).More recently, a retrospective cohort study that compared patients with ALF who received MARS (n = 104) with propensity score-matched controls undergoing standard medical therapy (n = 416) found MARS to be associated with improved 21-day transplant-free survival rates, biochemical variables, and hemodynamics (9). The early use of MARS in the management of BCN appears to be a promising therapeutic approach, as demonstrated by our patient's clinical improvement and renal recovery.Albumin dialysis is effective in reducing bilirubin levels and likely sequelae of high bilirubin burdens, as well as in patients with hepatic encephalopathy (17,20).By removing albumin-bound toxins, including bile acids and bilirubin, from the patient's blood through a combination of dialysis and adsorption, MARS is believed to remove toxin burdens to help promote renal recovery in cases of BCN-associated AKI. Our patient's hyperbilirubinemia did initially improve slightly on biliary decompression with placement of an internal/external biliary drain on hospital day 7, but it began to increase again on day 17 to day 33 (Figure 3).This is potentially explained by the impaired bile acid excretion in the setting of renal failure from BCN, in addition to hepatic synthetic dysfunction from hyperbilirubinemia.Although bilirubin levels did improve, levels plateaued as intermittent hemodialysis does not eliminate albumin-bound toxins such as serum bilirubin (21).MARS therapy was considered and ultimately started to reverse the severe hyperbilirubinemia.We believe that MARS was critical in mitigating the burden of hyperbilirubinemia exacerbating the severity of BCN, thereby facilitating subsequent renal recovery.The ongoing reduction in hyperbilirubinemia following the MARS therapy was likely from improvements in both renal and hepatic function. Although multiple cases of BCN-associated AKI have been reported in the literature, it is likely underdiagnosed because BCN-associated AKI requires a renal biopsy for confirmation (4,5).No current therapeutic guidelines exist for the treatment of BCN-associated AKI.Initial interventions, such as relieving biliary obstruction via ERCP with stent placement and HD, can be useful especially if implemented early on, although it is unclear how this affects the natural disease course (1).MARS has been used as a treatment strategy in one other case in our literature search for BCN-associated AKIs, although another case of cholestasis leading to renal dysfunction improved with MARS therapy (Table 1) (6,7).This strategy aims to reduce the burden of cholestasis in the hope of minimizing and eliminating the toxic effects of bilirubin on the renal tubules.This is similar to a more common indication for MARS therapy, refractory pruritus, in which patients with severe hyperbilirubinemia experience decreasing bilirubin levels and improvement of their pruritus (8).Given the renal recovery in our patient, patients with renal failure with suspected or proven BCN should be considered early in their disease course for albumin dialysis therapy using MARS. Funding The author(s) declare that no financial support was received for the research, authorship, and/or publication of this article. 1 ( FIGURE 1 (A) Details of the molecular adsorbent recirculating system (MARS) circuit, which involves a dialysate of albumin solution passing through a MARS dialysis membrane and then subsequently through a CRRT membrane, charcoal filter, and anion exchanger.(B) Diagram of the MARS membrane, which illustrates the albumin dialysate, and the passage of both water-soluble low-molecular-weight and larger albumin-bound toxins across the membrane.CCRT, continuous renal replacement therapy; TBG, thyroxine-binding globulin.Reproduced with permission from Baxter. FIGURE 2 FIGURE 2Light microscopy of renal biopsies stained with hematoxylin and eosin and showing bile casts with sclerotic glomeruli.Mild interstitial fibrosis and tubular atrophy are present, involving 10% of the cortical surface (grade 1 IFTA).Intratubular green-tinged casts (blue arrow) are present.Green pigment is also seen within the proximal tubular epithelium cytoplasm (black arrow).IFTA, interstitial fibrosis and tubular atrophy. FIGURE 3 FIGURE 3Timeline of days in the hospital and trends in the levels of creatinine and total bilirubin (mg/dL).	2023-10-14T16:03:48.328Z	2023-10-10T00:00:00.000	{ "year": 2023, "sha1": "c58ab9db646d8a3a34abd7e4049a63657de8d7bc", "oa_license": "CCBY", "oa_url": "https://www.frontiersin.org/articles/10.3389/fneph.2023.1256672/pdf?isPublishedV2=False", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "91734b22dbdeca3b02ad5a9a7f659255439f8163", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [] }
237262272	pes2o/s2orc	v3-fos-license	Application of virtual reality on non-drug behavioral management of short-term dental procedure in children Background Due to the inherent characteristics of immersion, imagination, and interactivity in virtual reality (VR), it might be suitable for non-drug behavior management of children in dental clinics. The purpose of this trial was to measure the role of VR distraction on behavior management in short-term dental procedures in children. Methods A randomized clinical trial design was carried out on 120 children aged between 4 and 8 years to identify the comparative efficacy of VR and tell-show-do (TSD) to improve behavioral management during dental procedures. The primary outcomes were evaluated anxiety, pain, and compliance scores in perioperative children. The levels of operative anxiety and pain were assessed using the Children’s Fear Survey Schedule-Dental Subscale (CFSS-DS) and Wong Baker FACES Pain Rating Scale (WBFS), respectively. The Frankl Behavior Rating Scale (FBRS) was tested before and during dental procedures. The length of the dental procedure was compared between both groups after treatment. Results The average anxiety and behavioral scores of the VR group significantly reduced compared with the control. The decreased anxiety score for the VR group and control group were 8 (7, 11) and 5 (5, 7), p < 0.05. The compliance scores of the control group during treatment were 3 (2, 3), and the same in the VR intervention were 3 (3, 4), p = 0.02. A significant reduction in pain was observed when using VR distraction (p < 0.05). Comparing the length of the dental procedure, the VR group (19.0 2 ± 5.32 min) had a shorter treatment time than the control group (27.80 ± 10.40 min). Conclusion The use of VR significantly reduced the anxiety and pain of children and the length of the dental procedure and improved the compliance of children that underwent short-term dental procedures without an adverse reaction. Trial registration Chinese Clinical Trial Registry, ChiCTR2000029802. Registered on February 14, 2020 Introduction Successful behavioral management might be related to the following two variables: anxiety, which is a psychological state and can be modified and controlled with psychological techniques, and pain, which is an unpleasant sensory and emotional experience [1]. For children's cognition levels during a dental procedure, it is necessary to conduct behavioral management; therefore, the children can cooperate with the pediadontist to complete the treatment. In Southwest China, a substantial proportion of children that require dental care are unable to collaborate well with doctors and nurses due to dental anxiety (DA). The increase in pain, tension, and fear-related behaviors during dental procedures was defined as DA; it can be expressed as rapid heart rate, muscle tension, and even syncope [2]. DA refers to a universal level of stress that is characteristic of an individual and might have a constant level during the life span. The emotion interferes significantly with personal daily life, career development, or relationships. Various studies found that the incidence of DA was 20-43%, which depended on the age of the child. In a recent survey, DA produced the majority of clinical problems in pediatric dental treatment [3]. For the treatment of DA, drugs were generally used in the Department of Anesthesiology of Stomatological Hospital affiliated to Chongqing Medical University (Yubei District, Chongqing, China), such as nitrous oxide (N 2 O) and sevoflurane. The N 2 O sedation can only be applied for patients that score three or four on the Frankl Behavior Rating Scale (FBRS) [4,5]. Because the children are awake and have no other way to distract their attention, many children do not cooperate with pedodontists during a dental procedure. The child's parents might be concerned about possible damage from sevoflurane inhalation anesthesia, because the children and their families need to carry out general anesthesiarelated preoperative preparation: waiting for the appointment, fasting and drinking, risk of apnea, and recovery after anesthesia. Therefore, some children did not receive timely treatment, which results in lifelong poor oral health, and can cause long-term psychological trauma [6]. Therefore, timely and effective management of DA is central to improving the mental and physical health of children with oral diseases. Virtual reality technology (VR) creates a highly realistic virtual, three-dimensional (3D) environment that provides various sense stimulate (e.g., sense of vision, sense of hearing, touch, and sense of smell) for the user to escape the real world [7,8]. By stimulating visual, auditory, and proprioceptive sensations, VR acts as a distraction to interfere with the patient's handling of noxious stimuli [8]. In the last decades, VR has been applied to different healthcare settings. In particular, VR has reported in many clinical trials, such as trauma rehabilitation [9,10], are of burns [11,12], cancer treatment [13], operation training [14], and weight-related disorders [15]. Sato and Sarig-Bahat performed VR on complex regional pain syndromes [16] and chronic neck pain [17]. The analgesic effects of VR distraction reduce negative emotions (i.e., anxiety) and lead to positive emotions [18]. In some studies, VR distraction has been used to relieve pain and anxiety [7,8,11,[19][20][21][22][23]. Similarly, VR has been used as a distraction intervention to relieve pain during the perioperative period in dental surgery [19,23]. The application of VR technology in the behavioral management of children's dental procedures is limited. In addition, recent reports have stated that it reduces pain and anxiety in the dental setting and procedures [24,25], but only for single dental procedures and the suitable duration of the treatment involved needs to be explored. In this study, the role of VR in the nondrug behavioral management of children is measured with short-term and simple dental procedures. Setting and patients This randomized clinical trial recruited 120 preschoolers aged 4-8 years who came to the Stomatological Hospital of Chongqing Medical University for dental treatment. This study followed the Declaration of Helsinki on medical protocol and ethics and the Regional Ethical Review Board of the Stomatological Hospital of Chongqing Medical University approved the trial. Inclusion criteria Consenting children (aged 4-8 years, ASA I-II) with a Children's Fear Survey Schedule-Dental Subscale (CFSS-DS) questionnaire > 19 [26]. The time of the dental procedure (caries treatment, extraction of deciduous teeth, incision of abscess, and root canal therapy) was expected to be < 30 min. Exclusion criteria The exclusion criteria were children or their families that could not agree, and their families were concerned that VR could have an impact on the eyes of the child and for other reasons that could interfere with wearing the VR glasses, such as those that required glasses for myopia. Since VR might cause motion sickness in some users, children were excluded with a history of motor diseases, motor nausea, or vomiting. Children with a history of epileptic or epileptic seizures were excluded, because there are some reports that VR has a theoretical risk of inducing seizures. Unpleasant treatment experience increases anxiety and pain during the following dental sessions, which results in increased pain perception. Therefore, in this study, subjects were excluded if they had a previous serious dental experience [27]. If a child had a serious fear or severe movement during the intervention, the trial was terminated immediately. Figure 1 shows the CONSORT Flow Diagram for the trial. Technical specifications HTC (Hsinchu, Taiwan, China)'s VIVE VR helmet was used, which is commercial, widely used, with a short delay time for video scenes, and is not prone to head vertigo. The VIVE is composed of 32 sensors for 360°m otion tracking, two 2160 × 1200 combined resolution AMOLED screens, and a 90-Hz refresh rate. The helmet was connected to an ASUS Game notebook with an Intel Core i7-8820K processor, 16-GB RAM, and an NVIDIA GeForce GTX 1070 graphics card. The virtual environment allowed the user to navigate naturally, which was created through a 110°field of view for immersion. Procedure Patients were randomly allocated to two conditions using randomization software (STATA software version 15.1). Eligible children's parents or caregivers were informed about the trial by the anesthetists, and informed consent was obtained preoperatively. When they had agreed to participate in the study, personal medical data were collected by researchers and the baseline anxiety was assessed by CFSS-DS (T0 = time after signing the consent). Then, the anesthetist nurse randomly allocated children to the VR intervention or to the control group (children only received tell-show-do (TSD) as usual). Block randomization was performed based on the type of dental procedure: caries treatment, extraction of deciduous teeth, incision of abscess, and root canal therapy. After randomization, the VR intervention took place in a separate room under the guidance of the nurse anesthetist, and children in the TSD group were admitted to the other room. Both groups were treated by experienced pediatric dentists. Figure 2 shows an example of the dental procedure using HTC's VIVE helmet. The assessment per time point was performed. FBRS was scored before intervention (T1 = 5 min before dental procedure) and remeasured at the moment of local anesthetic injection (T2). CFSS-DS and patient satisfaction (PS) scores were performed 5 m after the end of treatment (T3). Because VR intervention could cause adverse events, dizziness, nausea, vomiting, and epilepsy, they were followed-up during treatment. In addition, changes in heart rate and peripheral capillary oxygen saturation between both groups before, during, and after dental treatment were measured. VR intervention Before the dental procedure, the patient was shown the corresponding scenes, specific inducers, and background music that was developed with psychologists, which could attract their attention and relax them. The children lay down on the dental chair and did not shake their heads left and right, which caused the treatment to be interrupted. The nurse anesthetist put the helmet and earphones on the children, who entered a virtual world where they could follow a set route and watch different information expressed in the scenes. The story began in a shallow sea world with a soft environment. The undersea world could only be saved when the undersea creatures shared their most precious things to nourish a rare pearl. First, "I" am in a shell, and a little elf introduces the creatures of the sea and their precious spirits. Second, a sea anemone protected the clown fish when the clown fish reduces the surface material of the anemone, and they shared this precious friendship by helping each other (Fig. 3a). Then, the children were introduced to the ancient precious and tenacious vitality of animals: turtles (Fig. 3b), parrot fish solidarity (Fig. 3c), and the dolphin's helpful spirit (Fig. 3d). Scenes were switched for the children to introduce a beautiful and dangerous jellyfish, which shared its valuable ability to predict storms (Fig. 3e). Finally, the pearl was born from the precious spiritual nourishment from thousands of marine creatures and the undersea world was restored to its former peace and tranquility (Fig. 3f). In addition, the virtual environment was displayed on the ASUS notebook, and therefore, the accompanying families could see what the child was viewing. Primary outcomes The primary outcomes were CFSS-DS, WBFS, and FBRS in the VR group, compared with those in the TSD group. The CFSS-DS was monitored at T0 and T3, while FBRS was evaluated at T2 and T3, and WBFS (score was from 0 to 10; 0 = no pain, 10 = extreme pain) was recorded at T2 and T3 points. Statistical analyses Data were analyzed using SPSS software version 22 (IBM, Chicago, USA). The statistically significant difference was set at p < 0.05. A Chi-squared test was used to assess gender difference, ASA physical status, type of dental procedure, and local anesthetic between both groups. All quantitative variables were presented as mean ± standard deviation, and the analysis of variance or a nonparametric test was performed for comparisons according to the data distribution. Mann-Whitney U tests were used to compare the downward trend for the level of anxiety and FBRS. Results The trial was composed of 25 girls and 35 boys in the VR group and 32 girls and 28 boys in the TSD group, respectively. The mean ages in the VR and TSD groups were 5.59 ± 0.92 and 5.66 ± 0.99 years, p = 0.69. Similarly, there were no significant differences between both groups in the means for gender, ASA physical status, type of dental procedure, and location and type of dental procedure (Pearson's Chi-squared test, p > 0.05; Table 1). The mean anxiety scores decreased significantly after VR distraction. In the VR group, the CFSS-DS was 34.17 ± 5.81 before the intervention, which decreased to 24.77 ± 6.98 after VR distraction. In the TSD group, the anxiety scores were 34.08 ± 8.42 and 27.98 ± 7.41, respectively (n = 60). The anxiety score between VR intervention and TSD were statistically different after a dental procedure ( Table 2). Table 2 shows the downward trend of variations in anxiety scores between both groups and the differences were statistically significant. The Mann-Whitney U tests of FBRS for the VR group in the pre-intervention were 2 (2, 3). Similarly, the T2 of VR distraction treatment were 3 (3, 4), which represented that VR distraction improved patient compliance. In the control group, the FBRS before intervention and the time of the maximum procedure pain were 2 (2, 3) and 3 (2, 3), respectively. This indicated that VR increased compliance (p = 0.02; Table 2). Five children (8.33%) in the VR group had severe intraoperative anxiety and stopped treatment in fear, compared with 11 cases (18.33%) in the TSD group (p < 0.05). The CFSS-DS score of the children after the exclusion of the previous patient found that VR interference relative to the control group could significantly alleviate the anxiety of the children, p < 0.05. The FBRS of both groups at T2 had no statistical difference. During the operation, the pain score of the VR group (1.58 ± 1.08) was lower than that of the control group (2.86 ± 0.96). In addition, the results showed that the VR intervention at the end of treatment, related to the control of operation pain, worked best (p < 0.001) in the patients. Comparing the length of the dental procedure, the VR group (19.02 ± 5.32 min) had a shorter treatment time than the control group (27.80 ± 10.40 min). The results of this trial indicated the decreased treatment time was particularly significant in carie treatment and root canal therapy and was statistically significant (p < 0.001). Table 2 shows that the overall PS of the dental procedure with VR intervention (88.33 ± 7.15) was significantly higher than the TSD group (76.78 ± 8.49). After monitoring the physiological signs, the VR group had decreased heart rates and the control group had the opposite (p < 0.05; Fig. 4). There was no significant difference in the SPO 2 , which was detected before and after the intervention between both groups. Discussion Our group has been engaged in oral-related sedation and analgesia for a long time; therefore, we considered whether new methods could be used to relieve the DA in this group of children [28][29][30]. Therefore, this trial focused on the effectiveness of compliance changes during VR distraction in 4-8-year-old children during a short invasive dental treatment [6]. The results of this trial supported that VR was associated with a larger decrease in behavioral avoidance compared with TDS intervention. The anxiety score of both groups before and after the intervention was monitored, and the VR group decreased more than the control group; however, the compliance score of the children increased. Aminabadi et al. reported that VR eyeglasses successfully decreased pain perception and the anxiety state during dental treatment in 4-6-year-old children [31]. Similarly, Shetty observed that VR distraction could be used as a successful behavior modification method in 5-8-year-old children that underwent short invasive dental treatments [32]. Therefore, this trial and references supported that immersion in a virtual environment could help to control DA, improve compliance, and relieve pain during pediatric dental treatment in children [31,32]. Researchers showed that the effects of VR techniques on pain perception were beyond simple distraction [33]. In addition, by WBFS Wong Baker FACES Pain Rating Scale f PS Patient satisfaction T0, signature of informed consent statement; T1, before intervention; T2, the moment of local anesthetic injection; T3, the end of treatment *p-value < 0.05 statistically significant diverting attention from an unpleasant environment setting to a pleasant and absorbing virtual world, VR diminished the patient's physical pain experience [10,34,35]. By relieving anxiety and reducing pain, children cooperated better with the treatments, which included caries treatment and root canal therapy. In addition, the treatment time for the VR group was significantly shorter than the control group. In a previous clinical treatment, general anesthesia by inhaling sevoflurane was an invasive medical method that has been used on most children, who had difficulty in adapting to TSD behavioral induction. Some children might have received delayed treatment because of family concerns about general anesthesia or their lack of cooperation with behavioral induction, which resulted in poor prognosis and long-term tooth problems. This trial suggested that the customized VR content applied in this study allowed children to be treated as quickly as possible and might reduce the frequency of patients seeking medical treatment and reduce the number of outpatients. In China, the ratio of health workers to people is significantly lower than the global average, especially pediatricians [36]. This could be beneficial for doctors and patients. This trial has several limitations. First, the scales for anxiety and pain in children were relatively single, different clinometric tools, and assessments were not included. Second, other factors that affect children's behavior were not accounted for. Some researchers reported that several factors, such as age, gender, type of dental treatment, parental anxiety, and socioeconomic status were associated with anxiety should be assessed, because these might influence the efficacy of VR [37]. The evidence reported that older children considered the VR technique as a very simple game, and they had a have a lower level of distraction [33]. This study identified five children who were uncomfortable with VR distraction and terminated the trial, considering that some of them were possible terminated for these reasons. But the average age analysis of the five children was not statistically significant. In addition, their anxiety scores were > 35 before and after treatment. Third, the pedodontists and the patients were not blinded to the interventions due to the apparent difference between both groups. During future work, the deficiency could be overcome by watching different VR animations in different groups. Fourth, the VR in this trial only had one animation content, which might be considered a limitation, because the stimuli that trigger DA might differ for each individual. Based on the current data that uses VR, anxiety score, and mobile internet APP to collect family members' awareness and anxiety about dental diseases, to achieve fewer generic intervention scenarios the next step could be hierarchical customization of VR content. In addition, VR should improve and its use increased in the children's department of stomatology. Finally, a set Fig. 4 Changes in heart rate before and during the dental treatment between both groups. VR, virtual reality group; TSD, tell-show-do group; T1, 5 min before dental procedure; T2, the moment of local anesthetic injection. a,b p-value < 0.05 of intelligent biofeedback mechanisms should be formed to achieve closed-loop control of anxiety and pain in children; therefore, children will receive rapid, timely, and comfortable oral treatment. This trial focused on the effectiveness of compliance changes in VR distraction in 4-8-year-old children during short-term invasive dental procedures. The VR distraction used nonintrusive methods; therefore, the children's parents or caregivers have not been concerned that general anesthesia might affect intelligence and learning ability. The adverse effects on vision and hearing have not been reported in this study and literature [31,[38][39][40]. Even with repeated dental treatments, parents and children are much more receptive to other methods.	2021-08-23T13:20:11.385Z	2021-08-23T00:00:00.000	{ "year": 2021, "sha1": "51644e5760e7a405e243b8ed3832ebda1dac008a", "oa_license": "CCBY", "oa_url": "https://trialsjournal.biomedcentral.com/track/pdf/10.1186/s13063-021-05540-x", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "51644e5760e7a405e243b8ed3832ebda1dac008a", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
199072097	pes2o/s2orc	v3-fos-license	Sensory Attributes of Market Milk, Yogurt, and Kefir Supplemented with Various Concentrations of Aronia melanocarpa (black chokeberry) Powder: A Preliminary Study Aronia melanocarpa (black chokeberry) powder containing various bioactive compounds is widely used in the food industry. We examined the sensory attributes (taste, color, flavor, texture, and overall acceptability) of market milk, yogurt, and Kefir containing different concentrations of Aronia melanocarpa powder. In market milk, the scores of all categories except color were the same or lower in market milk containing 0.5%, 1.0%, 1.5%, or 2% Aronia melanocarpa powder than the control group (0%). No statistical difference in color and flavor was observed between Kefir and yogurt. However, there was a statistical difference between the control and treated groups in taste, texture and, overall acceptability (p < 0.05). As the content of Aronia melanocarpa powder increased, color, flavor, taste, and overall acceptability scores generally decreased. Yogurt and Kefir, with the addition of 1% Aronia melanocarpa powder, displayed the highest scores. Especially Aronia melanocarpa (black chokeberry) was cultivated for medicine use in the former Soviet Union from the middle of the 20th century, hence researchers had a focus on Aronia melanocarpa (black chokeberry) (Borowska et al., 2017;Sic Zlabur et al., 2017;Kim et al., 2018). Aronia melanocarpa (black chokeberry) had much amounts of phenolic components than those of the other berries (Chrubasik, et al., 2010;Kim et al., 2018). And the products of Aronia melanocarpa (black chokeberry) were sold the form of powder (Chrubasik, et al., 2010;Kim et al., 2018). Hence, it could be directly applied for adding Aronia melanocarpa (black chokeberry) to dairy products for upgrading nutritional quality and sensory attributes (Nguyen and Hwang, 2016;Jurikova et al., 2017;Kim et al., 2018). Then, the ultimate goal of this study was to manufacture market milk, Yogurt and Kefir the J Milk Sci Biotechnol Vol. 37, No. 2 ｜109 bioactive Yogurt added with Aronia melanocarpa (black chokeberry) powder for improving sensory attributes. In this experiment, among various physicochemical characteristics of dairy products, sensory attributes of market milk, Yogurt and Kefir added with Aronia melanocarpa (black chokeberry) powder were analyzed. Aronia melanocarpa (black chokeberry) powder The power type of Aronia melanocarpa (black chokeberry) was produced in Kiantama OY (Finland), and also was the food additive grade used in this study. The preparation of Kefir added with Aronia melanocarpa (black chokeberry) powder According to the instruction given by , Kefir was manufacture by Kefir grains obtained from Center for KU Food Safety, College of Veterinary Medicine, Konkuk University in Seoul, Korea. Then Kefir samples were added with different concentration of Aronia melanocarpa (black chokeberry) powder (control as 0%, 0.5%, 1.0%, 1.5%, and 2.0%). The final Kefir products were cooled at 4℃ and then stored in a refrigerator. The sensory attributes of market milk, Yogurt, and Kefir added with Aronia melanocarpa (black chokeberry) powder The sensory attributes was analyzed by 15 trained panelists between 20 and 50 years of age who gave written informed consent after explanation of risk-benefits of participation prior to the study. The samples was randomly served in single-use plastic cups (10±1 mL) about 8±1℃. All assessors completed a test assessment form to compare the 5 different sensory attributes -(1) taste, (2) color, (3) flavor, (4) texture and (5) overall acceptability -by a 5-point hedonic scale (1 as extremely poor; 2 as poor; 3 as fair; 4 as good; and 5 as excellent). Statistical analysis Statistical analysis was performed by GraphPad Prism 5.0 (GraphPad Softward Inc., San Diegeo, CA, USA), and statistical significance was considered at p<0.05. In this study, all data were showed as means±standard deviation which were performed by two separate experiments with triplicate assays. According to statistical analysis of the sensory attributes, there was statistical difference in taste, color, flavor, texture and overall acceptability between treated group and control group (p<0.05). Consequently, among the treated group compared with the control group (0%), the scores of taste, flavor, texture and overall acceptability were the same or lower in Aronia melanocarpa (black chokeberry) powder-containing market milk with 0.5%, 1.0%, 1.5%, and 2%. According to statistical analysis of the sensory attributes, there was not any statistical difference in color and flavor, but in taste, texture and overall acceptability between control group and treated group (p<0.05). Consequently, among the treated group compared with the control group (0%), the scores of taste, flavor, texture and overall acceptability were the same or lower in Aronia melanocarpa (black chokeberry) powder-containing Kefir with 0.5%, 1.0%, 1.5%, and 2%. In general, since Aronia melanocarpa (black chokeberry) had a astringent taste in itself, more than 3% Aronia melanocarpa (black chokeberry) juice could not be used for making various dairy products including Yogurt (Nguyen and Hwang, 2016). According to a study by Nguyen and Hwang (2016), they reported that there were not any statistical differences in several sensory attributes between yogurt added with Aronia melanocarpa (black chokeberry) and normal yogurt (control as 0%). the results of this study showed a similar tendency in previous study (Nguyen and Hwang, 2016). Therefore, future studies will need to find out not only how to maintain a good sensory Fig. 4. Comparison of the sensory attributes of Kefir added with Aronia melanocarpa (black chokeberry) powder (control as 0%, 0.5%, 1.0%, 1.5%, and 2.0%). attributes of various dairy products added with Aronia melanocarpa (black chokeberry) powder, but also how to improve human health.	2019-08-02T16:38:22.942Z	2019-06-01T00:00:00.000	{ "year": 2019, "sha1": "ef862efa587d0ee422add8dba517b35c355dddf6", "oa_license": "CCBYNC", "oa_url": "https://www.ejmsb.org/download/download_pdf?pid=jmsb-37-2-108", "oa_status": "GOLD", "pdf_src": "Adhoc", "pdf_hash": "47725dff0f4161279337e1fbd5d8767bc5867b2d", "s2fieldsofstudy": [ "Agricultural And Food Sciences" ], "extfieldsofstudy": [ "Chemistry" ] }
62841093	pes2o/s2orc	v3-fos-license	Enhanced persistent photoconductivity in $\delta$-doped LaAlO$_{3}$/SrTiO$_{3}$ heterostructures We report the effect of $\delta$-doping at LaAlO$_{3}$/SrTiO$_{3}$ interface with LaMnO$_{3}$ monolayers on the photoconducting (PC) state. The PC is realized by exposing the samples to broad band optical radiation of a quartz lamp and 325 and 441 nm lines of a He-Cd laser. Along with the significant modification in electrical transport which drives the pure LaAlO$_{3}$/SrTiO$_{3}$ interface from metal-to-insulator with increasing LaMnO$_{3}$ sub-monolayer thickness, we also observe an enhancement in the photo-response and relaxation time constant. Possible scenario for the PC based on defect-clusters, random potential fluctuations and large lattice relaxation models have been discussed. For pure LaAlO$_{3}$/SrTiO$_{3}$, the photoconductivity appears to originate from inter-band transitions between Ti-derived $3d$ bands which are $e_{g}$ in character and O 2p - Ti $t_{2g}$ hybridized bands. The band structure changes significantly when fractional layers of LaMnO$_{3}$ are introduced. Here the Mn $e_{g}$ bands ($\approx1.5$ eV above the Fermi energy) within the photo-conducting gap lead to a reduction in the photo-excitation energy and a gain in overall photoconductivity. I. INTRODUCTION The origin of the two-dimensional electron gas (2DEG) at the interface of artificially tailored oxide heterostructures 1-5 is attributed to both intrinsic and extrinsic factors. While the former is accounted for the polar catastrophe model, 6,7 extrinsic factors are associated with defects, such as oxygen vacancies [8][9][10] and inter-site cation mixing 11,12 that are introduced in the system during film growth. 13,14 The latter argument is substantiated by the observation of a clear dependence of the interface conductivity on pressure during the film growth. 6,14 Furthermore, the observation of persistent photo-conductivity (PPC) with large relaxation time also points to the role of defect induced states in these oxide heterostructures. [15][16][17] However, there also exists conclusive evidence that defects in the form of O-vacancies in the SrTiO 3 substrate are not the only responsible factor that leads to 2DEG at the interface. Had so been the case, then irrespective of the nature of the films, a 2DEG would have been observed for most of the perovskite oxides deposited on SrTiO 3 , including for LaCrO 3 and LaMnO 3 . Beyond, we note that doping induces electronic phase transition and metallicity in both LaCrO 3 and LaMnO 3 . [18][19][20] On the other hand, transport measurements find that the LaCrO 3 and LaMnO 3 films deposited on TiO 2 terminated SrTiO 3 substrate results in no 2DEG at the interface, irrespective of the film thickness and deposition conditions. 21,22 In general, a 2DEG state is observed at the interface of a non-polar (TiO 2 terminated SrTiO 3 ) and a polar material (such as LaAlO 3 ) motif. Under these circumstances, the in-built electric field at the interface causes bending of conduction and valence band edges of SrTiO 3 , leading to a triangular potential well which is filled by electrons transferred from the LaAlO 3 over-layers. [23][24][25][26][27][28] Although, this model explains many of the experimental observations, it fails to account for the observed charge carrier density that has its dependence on the thickness of the deposited LaAlO 3 film. 4,5 Beyond, the model also has little support from the photo-emission spectroscopy studies. [29][30][31][32] A fundamental understanding of the underlying mechanism for the formation of the 2DEG therefore requires use of perturbative techniques, which will lead to injection of additional charge carriers at the interface and selective and controlled modification of the interface chemistry. While the former is realized by electrostatic gating and photo-excitation, 4,15,17 the latter can be achieved by δ-doping at the interface, a concept commonly used in III-V compound semiconductor quantum wells. 33 The effect of electrostatic gating of LaAlO 3 /SrTiO 3 interface has been studied extensively. The gate field either draws charges towards the interface from the over-layers or pushes charge towards it resulting in a gate controlled metal-insulator phase transition. 4,34 It has also been shown earlier that the electronic transport in oxide heterostructures can be altered significantly on exposure to electromagnetic radiation of optical frequency making these potential candidates in optoelectronic applications. [15][16][17] For the conducting interface of LaTiO 3 /SrTiO 3 and LaAlO 3 /SrTiO 3 , the PPC is seen in the spectral range of 300 − 400 nm. It is also noticed that the magnitude and relaxation dynamics of the photo-conducting state depends on the growth temperature of LaTiO 3 and LaAlO 3 , but is independent of the number of the film overlayers, since the response is only energetically close to the band gap of SrTiO 3 . It has been previously shown that SrTiO 3 (single crystal and thin films) displays wide range of properties such as ferroelectric, 35,36 persistence photoconductivity 37 etc. Therefore, it is quite evident that the SrTiO 3 substrate plays a crucial role in determining the electronic and optical properties in these heterostructures. The PPC seen in oxide interfaces is quite similar to that reported in the III-V semiconductor heterostructures. 38,39 In anal-ogy with the latter systems, the slow decay of the photocurrent in oxides can be associated with the defect induced states in the SrTiO 3 substrate. The longer lifetime of the photo-induced carriers is largely attributed to the less dispersive Ti 3d bands, which make up the conduction band 17 in these systems. We have observed a systematic metal-to-insulator transition in LaAlO 3 /SrTiO 3 system by controlled δdoping at the interface with LaMnO 3 monolayers. 40 Our present experiment rules out the possibility of interfacial intermixing as the cause of metallic conduction in LaAlO 3 /SrTiO 3 system, because such a reaction of La/Sr at the interface would lead to formation of La 1−x Sr x TiO 3 which should be conducting based on the simple valence argument. Our experiment also rules out the reduction of SrTiO 3 as the cause of 2DEG, as reduction would take place even in the case of δ-LaMnO 3 monolayer deposition, resulting in interfacial conductivity irrespective of the Mn concentration at the interface. While the present experiment validates the polarization catastrophe argument for the 2DEG formation, it remains to be seen what would happen to photo-conductivity when transfer of electrons from LaAlO 3 overlayers are inhibited by the δ-LaMnO 3 monolayer. Here we present the results of photo-conductivity measurements on the δ-doped LaAlO 3 /SrTiO 3 heterostructures, and compared them with the photo-response of the undoped LaAlO 3 /SrTiO 3 system. The δ-doping is accomplished by means of a selective deposition of fractional monolayers of LaMnO 3 on the TiO 2 terminated SrTiO 3 surface, followed by layer-by-layer growth of the LaAlO 3 over the LaMnO 3 layer. It has been shown previously that LaMnO 3 films on SrTiO 3 remain nonconducting irrespective of their thickness. 22 Therefore, a controlled deposition of Mn ions at the LaAlO 3 /SrTiO 3 interface is expected to provide a method to control the electronic phases and phase transition. In accordance, we find that the resistivity of the δ-doped LaAlO 3 /SrTiO 3 increases non-trivially as the δ-LaMnO 3 monolayer at the interface is thickened. Beyond, the results of photoconductivity measurements reported here also include the following: (i) A larger photo-response with increasing δ-doped LaMnO 3 interfacial layer thickness and, (ii) persistent photo-conductivity following a stretched exponential behavior with its decay constant increasing in proportion with the LaMnO 3 monolayer thickness. Explanation to these results are sought in terms of a few well established models and also partly from the band structure calculations. II. EXPERIMENTAL DETAILS The samples were grown by Pulsed Laser Deposition at 800 o C and the thickness of LaAlO 3 deposited on TiO 2 terminated (001) SrTiO 3 was kept at 20 unit-cell (uc). The deposition procedure, conditions and growth parameters are described in our earlier reports. 16,17,40 The δ-doping of the interface has been achieved by controlled growth (≃ 0.012 -0.015 nm/s) of a LaMnO 3 submonolayer prior to commencing deposition of LaAlO 3 film. In its bulk form, LaMnO 3 is a strongly correlated anti-ferromagnetic insulator. A schematic of the interfacial δ-doping with LaMnO 3 is shown in the inset of Fig.1. To perform the transport measurements, Ag/Cr electrodes were deposited on the films (drain and source) and back of the substrate (gate) by thermal evaporation using shadow masks. Linear current versus voltage characteristics between drain and source confirm the Ohmic behavior of the electrodes, while the leakage current between the gate and source was less than 10 nA. The photo-excitation experiments were carried out in a closed cycle helium optical cryostat where the samples were exposed to the broad band radiation of a Xenon lamp, whose spectrum contains ≃ 3.5% of UV radiation, and two single wavelength (441 and 325 nm) lines of a He-Cd laser through a quartz window. A. Electrical Resistivity The temperature dependence of the sheet resistance (R ) of the LaAlO 3 /SrTiO 3 and that of the δ-LaMnO 3 doped samples is shown in Fig.1. Consistent with the previous reports, a 20 unit cell thick LaAlO 3 film on TiO 2 terminated SrTiO 3 substrate showed metallic characteristics. 16,17,40 However, on embedding LaMnO 3 sub-monolayer at the interface, we observe a systematic transition to insulating state. We have also noticed that the critical thickness of the LaAlO 3 over-layer required to induce 2DEG in δ-LaMnO 3 doped heterostructures is proportional to the sub-monolayer thickness of the δ-LaMnO 3 . 40 This study also revealed that ≃ 0.6 uc thick LaMnO 3 drives the interface to insulating state when the LaAlO 3 overlayer thickness was of 20 uc. We note that the critical thickness of LaMnO 3 to make the interface insulating is much smaller when LaAlO 3 layer is only 10 uc. The plot of R (T) for δ-doped LaMnO 3 with submonolayer thickness is shown in Fig.1 shows the time evolution of resistance of three samples with δ ≃ 0, 0.2 and 0.5 ML during the period when sample was exposed to the light and then allowed to recover in dark at 20 K. To better understand the recovery process, we have defined a normalized resistance ∆R/R D , where ∆R = R(t)-R D with R(t) being the resistance at time t and R D the resistance in dark. The details of these measurement have been described in an earlier report. 17 Fig. 2(b & c) reveal that the relative change in the resistance on photo-exposure The LaAlO3/SrTiO3 sample shows a metallic behavior as the temperature is lowered to ≃ 30 K followed by an upturn, while the sample with 0.6 ML LaMnO3 at the interface shows insulating behavior. 40 The resistance goes beyond the measurement limit below ≃120 K. increases with δ-LaMnO 3 layer thickness, while the recovery process slows down. It is important to point out here that the most change in the resistance on photoexposure is triggered by the ultra-violet (UV) component of the quartz halogen lamp radiation. This fact is further established when we expose the sample to a monochromatic radiations (λ = 325 and 441 nm) of He-Cd laser. These results are shown in the inset of Fig. 2(b), from which we find that relative change in the resistance of all heterostructures under consideration is significantly suppressed for 441 nm radiation, in comparison to that observed under the 325 nm radiation. In Table I, we list the values of ∆R/R D for all three samples achieved under 325 nm, 441 nm and xenon lamp radiation. The comparative study as highlighted in Table I shows that the overall photo-response increases with increasing δ doping of LaMnO 3 at the interface. Also, our data suggests that the threshold wavelength to induce photo-conductivity in these heterostructures shift towards higher wavelengths with δ doping. This effect has been qualitatively understood by means of band structure calculations, which is discussed in the next section. In Fig. 2(c) we show the ∆R/R D data taken at exp − t τ β with 0≤β ≤ 1, and, τ being the relaxation time constant. A fit (shown by the solid lines in Fig. 2(b)) yields τ varying from 1700 to 3000 seconds as the thickness of the δ-layer is increased to 0.5 uc. Correspondingly, the exponent β goes from 0.2 to 0.8. Further, the recovery dynamics depend strongly on temperature. In order to understand the recovery process and the mechanisms involved, we fit the data to Arrhenius equation given as τ = τ 0 exp − ∆U kB T , where ∆U and k B are the activation energy for detrapping the photo-generated carriers and Boltzmann constant, respectively. The plots of ln (τ ) against 1 T are shown in Fig.3. Clearly, two distinct temperature regions of activation can be seen in the Arrhenius plot. At low temperatures T < 100 K, the calculated value of ∆U is in the range 1 − 2 meV, for all samples. However, at higher temperatures, the ∆U for 0.5 uc thick LaMnO 3 is estimated as ≃ 22 meV, which is almost three times higher in comparison to the activation energy of LaAlO 3 /SrTiO 3 system (≃ 8 meV). A sudden change in the slope about 100 K is observed in Arrhenous plot (Fig. 3). This temperature is very close to the structural phase transition of SrTiO 3 (∼ 105 K). We believe that such a phase transition would be of lesser significance, as the cubic to tetragonal transition of SrTiO 3 is partly due to its latent paraelectric nature, which is intimately associated with the d 0 -ness of the Ti ions in SrTiO 3 . In this case, the interface modifies the occupation of Ti with extra electrons being doped and therefore such a transition is not expected in the SrTiO 3 motif near to the interface, although one may expect the same to happen deep down in the substrate. The electrical conductivity of these structures undergoes strong modulation when a gate field is applied. It has been shown earlier that the electrostatic and photon fields act on two different sets of charge carriers. 17 While the recovery time from the photo-illuminated state stretches over several hours, the system was found to recover to the unperturbed normal state with in microseconds after switching off the gate field. Hence, the role of δ-layer influencing the migration of these two sets of carriers needs to be examined. We have studied the influence of electric field perturbation on the photo-induced relaxation processes of the δ-doped LaMnO 3 interfaces in comparison with that of the pure LaAlO 3 /SrTiO 3 system. In Fig.4 we illustrate our measurement scheme to study the effects of a gate electric field on the recovery process. At point 'A' of the plot in Fig. 4 the sample is exposed to light, which triggers a sharp drop in resistance followed by saturation. At point 'B' the illumination is turned off and the recovery dynamics thereafter is examined by applying a gate field (E g ) of −2kV/cm after a certain time gap, which has been changed from t 1 to t 4 in four sets of exposure-recovery experiments. We find that the resis- tance changes spontaneously on gating. While positive E g enhances the channel conductivity, negative E g tend to reduce it. For better understanding of the effect of gate field alone, we have measured source to drain I-V characteristic of these samples under different gate fields. The slope of these linear curves at the origin gives the zerobias conductance G V (0) which is plotted in the inset of Fig. 4. We observe an order in magnitude decreased in the G V (0), in the δ = 0.5 ML sample. C. Electronic structure The threshold energy required to impart photoconductivity in LaAlO 3 /SrTiO 3 corresponds to that of near ultra-violet wavelength. Numerical calculations based on density functional theory are consistent with this picture. 17 The optical transitions are associated to those states which are in the range ±2 eV above and below the Fermi energy, thereby providing the estimate of photo-conducting threshold wavelength as < 500 nm. However, the photo-response of the δ-doped LaAlO 3 /SrTiO 3 heterostructures show subtle signatures of the photo-response threshold shifting to lower photon energy, which suggests a modification in the band structure caused by the Mn ions in the vicinity of the interface. In Fig.5, we show the band structure of the δ-doped LaAlO 3 /SrTiO 3 . A fully relaxed super-cell of 2×2×9 dimension with 4.5 unit-cells of SrTiO 3 as substrate with TiO 2 termination is modeled in the present study. The δ-doping with 0.5 ML thick LaMnO 3 is modeled by Mn-Al chemical disorder near to the interface. The Brillouin zone integration were carried out on a 11×11×2 k−space grid with the WIEN2K 41 parameters RK max and G max to be as 7 and 24, respectively. The numerical details of the structure optimization, self-consistent total energy and that of the electronic structure are similar to those described in Ref. 17. Both in local density approximation (LDA) and generalized gradient approximation (GGA), Mn ions modeled at the interface of LaAlO 3 /SrTiO 3 system yield a metallic state, with relatively high density of states, with the prominent states being the Mn 3d states. It is well known that such spurious states at the Fermi energy are due consequence of the missing Coulomb correlation term in the Hamiltonian. In the present case, the Coulomb correlation effects (U ef f ) of the Mn 3d electrons are taken into consideration by the LDA+U ef f Hamiltonian, with U ef f = 8 eV. Before we discuss the band structure of the δ-doped In case of δ-doped LaAlO 3 /SrTiO 3 with LaMnO 3 monolayers, the band structure as shown in Fig.5 shows localized Mn 3d z 2 states (e g states) positioned ≃1.5 eV above Fermi energy. The corresponding Mn 3d xy states (t 2g ) in the valence band are relatively more dispersed over a wider energy range. Thus, it is evident that the positioning of the Mn e g states within the photo-conducting gap leads to a decrease in the photo-excitation energy, and therefore is expected to increase the overall photoresponse. Thus, for small wavelength induced excitations one find the transition to occur between the states that lie between ±2eV of the Fermi energy, while the small photo-generation of carriers at the 441 nm is associated with the Ti 3d xy states near the Fermi energy and the Mn e g bands which are relatively 1.5 eV above the Fermi energy. The computed band structure also qualitatively explains the relatively slow relaxation of the photogenerated carriers to the normal state in the δ-doped systems with that of the pure LaAlO 3 /SrTiO 3 . The energy-time relation of the uncertainty principle suggests that the life-time of photo-generated carriers in a given band would be inversely proportional to the energy band width. Along these perspectives, finding that the Mn e g bands which are highly localized and positioned at 1.5 eV above the Fermi energy would serve as localized traps for the photo-induced charge carriers, thereby increasing the relaxation time in δ-doped systems in comparison with the parent system. IV. DISCUSSION For abrupt interfaces of oxide heterostructures, the polar catastrophe model is quite robust to account for the origin of 2DEG. The model asserts on band bending effects via electronic reconstruction due to the in-built electric field at the interface of the polar (LaAlO 3 )−nonpolar (SrTiO 3 ) materials. PPC in these systems can therefore be related to the model proposed by Queisser and Theodorou (QT), where its emergence is associated with the macroscopic potential barriers induced by the band bending effects. 42 Such potential barriers lead to the PPC effect by spatial separation of the photo-generated electrons and holes. With one type of carrier being trapped, the other remains free and causes excess conductivity. 43 The model can be verified by studying the excitationenergy dependence of PPC. For this we used the two excitation energies afforded from He-Cd laser, which gives one of the excitation energy (2.8 eV) less while the other one larger (3.8 eV) than the band gap energy of SrTiO 3 . The results are displayed in Table-I and inset of Fig. 2(b). For pure LaAlO 3 /SrTiO 3 heterostructures, we observed very small change in the resistance and also weak signature of PPC with the photo-excitation below the absorption edge, while for δ-doped samples a large PPC effect could be seen, which increases monotonically with the δ-layer thickness. These results infer that the PPC of undoped sample have little contribution arising from spatial separation of the photo-generated carries, primarily due to the band bending effects. While in the case of δ-doped systems, this effect seems to be less prominent. The change in the resistance of δ-doped systems with lower excitation can be attributed to the finite absorption of incident illumination by in-gap states present in the SrTiO 3 band gap and reduce the possibility of band bending scenario. It has also been shown the 2DEG confinement can be explained by the formation of metal induced gap states in the band gap of SrTiO 3 rather than the band bending 32 and no measurable band bending is observed in La 1−x Al 1+x O 3 /SrTiO 3 (001) heterojunctions. 44 Thus, the model based on the spatial separation of the photoexcited electrons and holes by macroscopic potential barrier due to band bending looks less appropriate to account for the PPC here. However, if the band bending is due to the O-vacancies, then the associated barrier would spatially separate the electrons that constitute the 2DEG. Although, this can lead to some justification of the QT model to the observed PPC effects in the present hetero-structured systems, the observed decay kinetics does not seem to support the model as a predominant process for inducing PPC. The above picture of band bending is further established by the analysis of the PPC decay dynamics as proposed in QT model for artificially tailored heterostructures. 42 We tried to fit the decay to a logarithmic behavior of the form, ∆R/R D = a − b * ln(1 + t/τ ), where a and b are constants, with respect to time. We find that the logarithmic fit becomes poorer and deviates significantly from the measured data as the δ-layer thickness is increased. However, the undoped system still displays a better χ 2 tolerance to the fit and low residual values as compared to the δ-doped samples, which strengthen our argument. Various other models have also been proposed to account for the PPC effects in solid state materials. 42,[45][46][47][48][49] Earliest of all was associated to inhomogeneities in the samples, caused either by bonding configurations, spatial variation in the chemical composition and/or in local charge densities. A much more advanced model was due to Theodorou and Symeonides, 50 who proposed the role of defect clusters in the sample. Such defects may be inevitable and are generally introduced during the growth itself. On preliminary grounds, the defect-cluster model may look appropriate to account for PPC in these heterostructures, as the synthesis technique employed in depositing δ-doped LaAlO 3 films on SrTiO 3 substrates are by means of the laser ablation. This highly energetic deposition process could result in clusterization of the Mn and Al ions forming a disordered LaAl 1−x Mn x O 3 monolayer. In cases where the local inhomogeneities lead to a density of the defects which is larger than the carrier density, a macroscopic potential barrier may be formed. Such defect-clusters may host traps, and upon electron capture would become charged. In turn, this would induce a charge of opposite polarity around the cluster, thus spatially separating the electron-holes pairs, impeding their recombination and resulting in PPC. 50,51 However, the defect-cluster model predicts larger PPC in proportion to the thickness of the conducting layer. The latter, however, is in contrast with our observation in δ-doped LaMnO 3 in the LaAlO 3 /SrTiO 3 interface. We find that as the δ-doped sub-monolayer thickness is increased (larger Al/Mn disorder), the 2DEG conductivity decreases, but with an enhancement in the PPC. Thus, the defect-cluster model also seems less appropriate to account for the larger PPC seen in the δ-doped LaMnO 3 in the samples. A relevant model that appears to describe the PPC in the present case is based on potential fluctuations in the material. 46,[52][53][54][55][56] Here the spatial separation between photo-excited charge carriers by random local-potential fluctuations induced by compositional variations is held responsible for PPC. 46 Within the framework of the polar catastrophe model, the origin of 2DEG is due to electronic reconstruction of the Ti ions at the interface. Thus, the 2DEG wave-function would be thought as a composition of Ti 3+ and Ti 4+ states. Moreover, the Mn ions in the monolayer would also exhibit multiple valence states. Certainly therefore, an energy barrier emanating from such a charge fluctuation would also act as a source of potential fluctuations. Besides, there is also a likelihood of chemical disorder due to random distribution of Al and Mn ions in the delta doped layer. Therefore, both chemical and valence state disorder can induce an uneven potential landscape at the interface of LaAlO 3 /SrTiO 3 with δ-doping of the interface. However, this model predicts PPC to be observed at high temperatures, with a well defined transition temperature, which is in contrast with our observations, where enhanced PPC in seen at lower temperatures. But, we note that the decay kinetics predicted by the random potential fluctuation model is similar to that of a stretched exponential function. The lattice relaxation model 49,57,58 is one of the most widely accepted description for the PPC effects in semiconductor heterostructures. The model stipulates photoexcitation of carriers from defect induced deep-level traps . A barrier is created by lattice relaxation thereby preventing the recapture of mobile carriers. The source of the lattice relaxation could be ionic mismatch and/or vacancies. However, in the present context, mismatch in the lattice constants and/or that in the constituents ionic-radii, does not seems to be relevant. Therefore, the structural relaxation would be mainly due to oxygen vacancies, and hence these defects render a negative-U center. At low temperatures, thermal energy is not sufficient enough to overcome the barrier, and thus the photogenerated carriers remain in these shallow states, resulting in PPC. The PPC buildup and decay kinetics, in general, fit to a stretched exponential curve, which is similar to our observations for the oxide heterostructures. 16,17 Due to the inherent propensity of LaMnO 3 to attract oxygen, its presence in the vicinity of the interface can induce more vacancies in the substrate leading to a nonequilibrium chemical configuration. Thus, while the photo-ionization of oxygen specific defects in the quantum well region appears to be the dominant mechanism for PPC in oxide heterostructures, one cannot rule out the contribution of the spatially separated charges across the interface facilitated by the factors that have been mentioned above. Moreover, as both Mn and Al occupy the substitutional body-centered site of the perovskite structure, the role of Mn ions on the lattice can also be partly inferred from the PPC data. It has been argued in the case of III-V and II-IV semiconductor heterostructures, and also in oxide heterostructures that the 2DEG at the interface is derived out of lattice strain. Within the realms of alloy theory, the solubility of Mn ions in LaAlO 3 matrix, infers to a lowering of lattice strain . In general, reduction of strain decrease the activation energy with Mn at% in the monolayer, which, however, is in contradiction with our observation. Thus, the increase in activation energy with δ-doping at the interface may be attributed to the enhanced electron-electron interaction between the electrons in the trap and the unpaired electrons of the Mn ion. Such an event lead to a narrowing of the optical trap potential, which would narrow with increasing Mn at% at the interface, leading to a higher repulsive barrier for the electrons to return to the oxygen mediated defect centers, after photo-ionization. However, we note that there does not exist to the best of our knowledge, a scheme to estimate the strength of Coulomb scattering from the presence of the deep traps in the quantum well and its effect on transport properties. However, it is reported that in ZnSe/(Zn,Cd,Mn)Se heterostructured systems, the 2DEG exhibits a strong PPC that becomes more pronounced when the Mn 2+ concentration is increased. 59 The enhanced PPC in this system is also accompanied by a decrease in sample mobility, suggesting a connection between the deep traps introduced by the magnetic Mn ions and enhanced Coulomb scattering. V. CONCLUSION In summary, we have investigated the effect of δ-doping on the photo-response of LaAlO 3 /SrTiO 3 interfaces. The photo-response of the pure LaAlO 3 /SrTiO 3 heterostructure is found to be sensitive to near ultra-violet radiation which shifts towards the lower photon energy upon doping the interface with LaMnO 3 . The doped samples show relatively large photo-response and time constant of recovery in comparison to the undoped sample. Based on theoretical calculations, we establish that the slow relaxation arises due to localized Mn e g states (3d z 2 ) which are situated ∼ 1.5 eV above the Fermi level. The positioning of these Mn 3d e g bands within the photo-conducting gap shifts the photo-response threshold to higher photon wavelengths with increasing δ-doping in the heterostructures. We also have made an attempt to understand the decay dynamics of the photo-conducting state . Our experimental findings demonstrate that the defect-cluster and random fluctuation models are less appropriate to describe the large photo-response and high values of activation energies for the recovery seen in δ-doped samples. On other hand, the lattice relaxation model is found to be in better agreement. Moreover, our arguments also suggest that the increase in the activation energy could be attributed to the strong electron-electron interaction of the Mn-ions in the δ-doped monolayer at the interface.	2014-05-08T16:25:30.000Z	2014-03-01T00:00:00.000	{ "year": 2014, "sha1": "8956506a67f678902a775447c56574fa32bc1c35", "oa_license": null, "oa_url": "http://arxiv.org/pdf/1405.2005", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "8956506a67f678902a775447c56574fa32bc1c35", "s2fieldsofstudy": [ "Physics", "Materials Science" ], "extfieldsofstudy": [ "Materials Science", "Physics" ] }
254017617	pes2o/s2orc	v3-fos-license	Density-Induced Hadron-Quark Crossover via the Formation of Cooper Triples We discuss the hadron--quark crossover accompanied by the formation of Cooper triples (three-body counterpart of Cooper pairs) by analogy with the Bose--Einstein condensate to Bardeen--Cooper--Schrieffer crossover in two-component fermionic systems. Such a crossover is different from a phase transition, which often involves symmetry breaking. We calculate the in-medium three-body energy from the three-body $T$-matrix with a phenomenological three-body force characterizing a bound hadronic state in vacuum. With increasing density, the hadronic bound-state pole smoothly undergoes a crossover toward the Cooper triple phase where the in-medium three-body clusters coexist with the quark Fermi sea. The relation to the quarkyonic matter model can also be found in a natural manner. FIG. 1. Schematics of the density-induced two-and three-body crossovers. While the gases of tightly bound states (i.e., dimers and trimers) exhibit smaller cluster size ℓ compared to the mean interparticle distance d ∼ k −1 F in the low-density regime, these bound states start to overlap with each other (d < ∼ ℓ) with increasing density (i.e., decreasing d), and gradually change into Cooper pairs and triples. In ultracold Fermi gases near the Feshbach resonance, such an in-medium two-body crossover is triggered by the change in the attractive interaction, where ℓ increases when the attractive interaction decreases. three-component Fermi gases and the hadron-quark continuity in the three-color QCD has been pointed out theoretically [34,35]. To understand a non-trivial state of matter in the Fermi-degenerate state with three-body correlations, one may borrow the knowledge of the BCS-BEC crossover physics well established in cold-atom communities. In this regard, just as Cooper pairs are formed in the BCS side of the density-induced BEC-BCS crossover [36], one may imagine its three-body counterpart, that is, a Cooper triple [37][38][39][40][41][42], in a dense Fermi-degenerate regime but with non-negligible three-body correlations. Figure 1 shows a comparison between the two-body crossover (i.e., BEC-BCS crossover) and the three-body crossover accompanied by Cooper triples when the number density increases or the attractive interaction decreases. First, let us review the famous BEC-BCS-crossover case; if the density increases, the mean interparticle distance given by d ∼ k −1 F decreases, where k F is the Fermi momentum. In the dilute regime, d is much longer than the cluster size ℓ, which is determined by the interaction (e.g., scattering length) [22]. Such a state is well described by the gas of two-body molecules, and undergoes the BEC at low temperature. At sufficiently high density, d becomes smaller than the cluster size ℓ, indicating the tremendous overlap among clusters, that is, the formation of Cooper pairs. These Cooper pairs are repeatedly formed and dissociated by coexisting with the underlying Fermi sphere, although we do not draw such a dynamical picture explicitly in Figure 1. While Cooper pairs form the BCS condensate, which spontaneously breaks U(1) gauge symmetry, in conventional superconductors, we note that the spontaneous symmetry breaking is not necessary to form Cooper pairs themselves, as the so-called preformed Cooper pairs are found in strongly correlated or disordered superconductors above the critical temperature [43,44]. Let us now move to the three-body case, as shown in the lower side of Figure 1. One would see tightly bound threebody states with the cluster size ℓ if the strong two-or three-body attractive interactions exist in three-color fermionic systems. When the number density increases, these trimer states start to overlap at certain densities, and eventually may form an unconventional state that can be analogous to the Cooper pairing state. In the case of QCD, we recall again that such a three-body state should be kept due to the SU(3) gauge invariance. These overlapped three-body states in medium are similar to Cooper triples, which are the three-body counterpart of Cooper pairs predicted in the context of cold atomic physics [37][38][39][40]42]. The Cooper triple state does not necessarily form a condensate (however, note Ref. [45]), but resulting three-body correlations can modify the equation of state substantially [40,46]. In this paper, motivated by the recent astronomical observations of compact stars as well as the theoretical progress in the in-medium three-body correlations in cold atomic systems, we discuss the possible scenario of the density-induced hadron-quark crossover accompanied by the formation of Cooper triples using a simplified toy model with three-quark attractive interaction responsible for the baryon formation. We show that the in-medium three-body energy obtained from the three-body T -matrix exhibits a unique crossover from baryonic tightly bound states to Cooper triples with increasing quark chemical potential. This paper is organized as follows. In Section II, we introduce the theoretical model and explain how to address the crossover via the three-body T -matrix approach [40]. In Section III, we show the numerical result for the in-medium three-body energy and discuss the density-induced hadron-quark crossover. We summarize this paper in Section IV. Throughout this paper, we use units in which = k B = 1 and the system volume is set to unity. II. MODEL AND METHODS In a way similar to the study of quark-gluon plasmas in Ref. [47], we start from the Hamiltonian based on the Thompson scheme [48], as given by where ψ p,j is the fermionic annihilation (creation) operator with three colors (j = r, g, b), and ε p is the relativistic dispersion ε p = p 2 + m 2 q with the nonzero constituent quark mass m q ≃ 0.34 GeV. For simplicity, we neglect the spin and flavor degrees of freedom, which are not important for our purpose and will be considered in the future work. The second term of Equation (1) is the three-body attractive interaction responsible for the color-singlet baryon formation. While the coupling constant should be momentum-dependent due to the color confinement [49][50][51] as well as relativistic effects (the projection operator for the positive energy states and the coupling with negative-energy states) [47,48], we employ the constant coupling V k,q,k ′ ,q ′ ,P ≃ V 3 for the purpose of qualitatively understanding the hadron-quark crossover accompanied by the Cooper triple formation in the relatively high-density regime. We will adjust the value of V 3 later in such a way as to roughly reproduce the baryon mass in vacuum M B as ≃ 0.91 GeV. Incidentally, we note that non-relativistic or relativistic quark models with three-body forces have been introduced in Refs. [52][53][54]. We obtain the in-medium three-body T -matrix at the zero center-of-mass momentum [40], where Ω + = Ω + iδ is the three-body energy with a small imaginary part iδ. The in-medium three-body propagator Ξ(Ω) reads [55,56] where is the Fermi distribution function. We note that while the antiquarks are ignored in our toy model, such antiparticle contributions involve the antiparticle distribution functionf k = 1 exp , which can be small at large chemical potential and low temperature. We set V 3 = −2.16 × 10 2 (GeV) −5 such that T 3 (Ω) has a pole at Ω ≃ M B in vacuum. The cutoff and the small imaginary energy for the analytic continuation are adapted as Λ = 2 GeV and δ = 2 × 10 −2 GeV. We note that the continuum bottom of three quarks is given by Then, the in-medium three-body binding energy E M B (i.e., energy gain due to the formation of the in-medium threebody cluster) can be defined as which is the pole of the in-medium three-body T -matrix. In the vacuum case, such an energy gain is given by The in-medium three-body energy in the hadron-quark crossover regime at different temperatures. For reference, we plot Ω − E th = 3µ − 3mq, that is, Ω = 0 (at the Fermi level). The lower panel presents the schematic comparison between the single-particle energy ε k = k 2 + m 2 q and the three-body energy per particle (Ω pole − E th )/3 (dashed line) along the hadron-quark crossover. Figure 2 shows the in-medium three-body energy pole Ω pole measured from the bottom of the continuum E th = 3m q − 3µ as a function of µ at different temperatures T . One can see that the the three-body energy located in Ω pole − E th ≃ M B − 3m q ≡ −E B (i.e., E M B ≃ E B ) at µ = 0 moves toward the continuum bottom Ω pole − E th = 0 when µ increases. Moreover, this three-body pole remains at larger µ and locates below the Fermi level given by Ω − E th = 3µ − 3m q , indicating the coexistence of the quark Fermi sea and in-medium three-body clusters. This result manifests a picture of the smooth crossover from the tightly bound baryonic trimer state to the Cooper triple state, being analogous to the density-induced BEC-BCS crossover [21,36,57]. While the existence of the positive pole in the continuum is in contrast to the non-relativistic one-dimensional case [40], the similar behavior of the three-body energy can be found in the variational approach to the non-relativistic Fermi gases in three dimensions [39]. At higher temperatures, Ω pole is shifted toward an upper energy level around the crossover because of the thermal agitation effect on the binding energy [40,58]. However, one can see that the three-body pole is still located below the Fermi level (Ω − E th < −E th , i.e., Ω < 0) at low temperatures, implying the presence of the stable three-body bound state in medium. III. RESULTS AND DISCUSSION To see the physical picture of this crossover from a different point of view, we show the schematic single-particle energy level diagram in the lower panel of Figure 2. At µ = 0, the three-body bound state is located in the energy level −E B /3 measured from the continuum bottom ε k=0 = m q . At finite µ, this three-body bound state is shifted toward an upper energy level because of the reduced binding energy associated with the Pauli blocking effect. However, such a Fermi surface effect does not suppress the three-body pole, but eventually leads to the in-medium bound state in continuum at high density, as in the case of the famous Cooper pairing in superconductors. This state can be regarded as the Cooper triple phase, but we note that there are no clear boundaries between the hadronic bound-trimer phase and Cooper triple phase, a feature similar to the BEC-BCS crossover [39,40]. While we have fixed m q = 0.34 GeV in Figure 2, it may exhibit density dependence through the quark-antiquark pairing [59]. In Figure 3, we show a comparison between the results with massive (m q = 0.34 GeV) and massless (m q = 0) quarks; in the latter case, the chiral symmetry is assumed to be restored due to the Pauli blocking effect on the quark-antiquark pairing. While the result with m q = 0 (dashed curve) is found to be larger than that with m q = 0.34 GeV (solid curve), this difference can be approximately given by 3m q with m q = 0.34 GeV. On the other hand, even the massless case is well below the corresponding Fermi level (3µ in the absence of m q ) at high density, where the Cooper triple phase is expected to occur. For more realistic case, m q changes from the in-vacuum value down to the current quark mass across the evolution of µ, leading to the in-medium three-body pole in between. In the high-density regime, the three-body pole measured from the Fermi level (i.e., 3µ − 3m q − (Ω pole − E th ) ≡ −Ω pole ) can be regarded as the energy gain of the formation of a Cooper triple. The ratio of this energy gain to the Fermi energy may well become smaller as the density increases. However, the non-local properties of the three-body force, if remaining, would be important in such a high-density regime. Indeed, if the three-body force is replaced by the linear potential V qqq (R) ≃ σR [49][50][51], where σ is the string tension and R is the characteristic length scale for the three-body state, the energy gain might be corrected due to a significant spatial overlap of Cooper triples. Combined effects of the density dependence of the dynamical quark mass and of the linear-shaped three-quark interaction would cause the three-body energy pole to increase more rapidly. This will be left for an interesting future work. We remark in passing that at asymptotically high density, the confining three-body force would be suppressed, while two-body interactions due to one-gluon exchange would take over. Even in the density regime of interest here, instanton-induced two-body interactions would take effect. Such two-body interactions would lead to quark Cooper pairs and hence color superconductivity [60], which are ignored in the present analysis. Finally, to see the relevance of our results to neutron star physics, we discuss the connection between the present work and the phenomenological model of quarkyonic matter developed by McLerran and Reddy [14], where the peak structure of the speed of sound has been demonstrated as expected from recent astrophysical observations [12]. In the high-density regime, the baryonic excitation near the quark Fermi surface can be regarded as the Cooper triples. Indeed, in this model, the quark number density may be given by where g i and ∆ are the degeneracy (e.g., spin and isospin degrees of freedom) and the energy shell of baryons (corresponding to the Debye-frequency-like window of Cooper triples [45]), respectively. k c = k F + ∆ ≃ µ is the upper bound of the momentum for the formation of Cooper triples. While k c is called the Fermi momentum of baryons in Ref. [14], k c is not necessarily equal to such a momentum in our model because Cooper triples are no longer point-like fermions in the high-density regime. However, k c plays a similar role in the two different models. The last two terms in the parenthesis of Equation (7), which are regarded as the quark number density in baryons in Ref. [14], correspond to the quark number density in Cooper triples. Let us now set m q to zero for simplicity. Then, the Fermi degenerate state can be filled up to the threshold of Cooper triples given by (noting that 3µ + Ω pole < 3µ, and thus, Ω pole < 0 at high density, as shown in Figure 3). Therefore, as depicted in Figure 4, one may find analogical relations between the present system and the phenomenological model in Ref. [14] as 4. Correspondence between our model and quarkyonic matter proposed in Ref. [14]. While the left figure shows the quark dispersion (solid line) with the threshold of Cooper triples µ − \|Ω pole \|/3, the right figure shows the momentum configuration of quarkyonic matter, where the baryonic excitation (analogous to Cooper triples) can be found in the Fermi shell of baryons ∆ near the quark Fermi surface. In this way, these results imply that the possible peak in the speed of sound can be regarded as the consequence of the Cooper triple formation. Moreover, we also note that a related anomaly of the compressibility can be found near the region where the in-medium three-body binding energy rapidly changes in one-dimensional Fermi gases with the scale anomaly [40]. However, detailed investigations of thermodynamic properties within more realistic models would be needed to confirm such a scenario, which is left for interesting future work. IV. CONCLUSIONS In this paper, we have discussed the density-induced hadron-quark crossover via the formation of Cooper triples by analogy with BEC-BCS crossover realized in cold atomic and in condensed matter systems. Applying the inmedium three-body T -matrix approach to the toy model, where relativistic quarks interact with each other via the phenomenological three-body forces leading to the baryonic three-body bound state in vacuum, we show that the in-medium three-body pole exhibits a smooth crossover from the state of baryonic three-body trimers to the state of Cooper triples with increasing chemical potential, which can be regarded as the three-body counterpart of the densityinduced BEC-BCS crossover. A relation of the present system with the quarkyonic matter model [14] exhibiting the peak of the speed of sound has also been discussed. For future perspective, it is interesting to examine the equation of state and the speed of sound in a more realistic model including, e.g., density dependence of the dynamical quark mass, Pauli blocking effects among bound clusters [61], intercluster repulsion due to quark exchange [62], effects of gluons, and competition with diquark formation and color superconductivity [60]. In the presence of diquarks, which play a role in the baryon structure in vacuum [63,64], and the axial anomaly [65], the Nambu-Gorkov formalism, which helps avoid the breakdown of the T -matrix approach [66,67], would be useful. In the presence of additional degrees of freedom such as spin and flavor, dibaryonic superconducting quark matter might also appear at low temperature [68,69]. For ab initio calculations to elucidate the hadron-quark crossover, the complex Langevin method would be one of the promising methods to address finite-density QCD by avoiding the sign problem, as demonstrated in Ref. [70,71]. The complex Langevin method can also be used for clarifying multi-body correlations in ultracold Fermi gases [72]. ACKNOWLEDGMENTS The authors thank E. Nakano for useful discussions. This research was funded by Grants-in-Aid for Scientific Research provided by JSPS through No. 18H05406, No. 20K14480, No. 22H01158, and No. 22K13981.	2022-11-28T06:42:08.601Z	2022-11-25T00:00:00.000	{ "year": 2022, "sha1": "78a96ffc8ed6ba7eae75ad4c20b464c05162042d", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2073-8994/15/2/333/pdf?version=1674636079", "oa_status": "GOLD", "pdf_src": "Arxiv", "pdf_hash": "4908633a6476b0f4d3c78e953898cd9ec835fe83", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Computer Science", "Physics" ] }
258379667	pes2o/s2orc	v3-fos-license	The fruit morphometric variation and fruit type evolution of the stone oaks (Fagaceae, Lithocarpus) Background The great species diversity of Lithocarpus is associated with interspecific fruit morphological variation, represented by acorn (AC) and enclosed receptacle (ER) fruit types. Species representing both fruit types co-occur in the same forests and share two distribution centers in southern China and southeastern Asia. The predation selection hypothesis suggests that the fruit morphological mechanical trade-off between two fruit types could represent divergent dispersal strategies under varied predation pressures. By integrating phylogenetic construction with fruit morphometric study, we tried to verify the predation selection hypothesis and elucidate the fruit type evolution of Lithocarpus, which is critical in interpreting the distribution and diversification of the genus. Results We identified the functional trade-off between the two fruit types: ER species have bigger seeds which are enclosed mainly by receptacle representing stronger physical defense; whereas the seeds of AC species are smaller and enclosed mainly by thin pericarp representing lower mechanical protection. Despite some reversals from ER back to AC, the ancestral state reconstruction in combination with thermal analysis supports the hypothesis that ER is the derived fruit type from AC-like ancestors independently across all clades. Conclusion Our results support the predation selection hypothesis by verifying the mechanical trade-off between the two fruit types. We propose a divergent selection theory for the two fruit types: the seed size and mechanical defense of AC species become smaller, whereas those of ER species become larger and require more morphological modifications in the receptacle. This signified the importance of the receptacle in differentiating the two fruit types and in the fruit morphological modification through evolutionary time. We found that the ER-type species evolved independently in all clades and across varied climates from tropical to warm temperate regions. As ER fruits are products of convergent evolution, we propose to examine the predation and dispersal variation between two fruit types in the future to verify whether predation selection is the reason behind fruit type evolution of the stone oaks. Supplementary Information The online version contains supplementary material available at 10.1186/s12870-023-04237-4. become independent [3,4]. Meanwhile, with 10-1000 times greater caloric value than most wind-dispersed species [5], nuts became nutritious food sources for animals. Two relationships were formed between nuts and their predators: a mutualistic relationship by providing nutritional reward for the service of seed dispersal by a vertebrate [6]; and antagonism with insects as the most important pests causing pre-dispersal seed predation [5]. The nut characteristics, i.e., morphology, seed chemistry, and physical defense, interact with each other and influence seed predation and dispersal. Meanwhile, the behavior and choices of insects and vertebrates in turn significantly impact the evolution of nut characteristics [5,[7][8][9][10][11][12]. Therefore, the fruit-animal interaction is crucial in the evolution of Fagales nuts. Lithocarpus Blume under Fagaceae serves as an interesting study model. Over 320 noted Lithocarpus species with great fruit morphological variation [13] are widely distributed from far eastern India to southern China, throughout Indochina, southern Japan, and extends through the Malayan Archipelago to Papua New Guinea [14,15]. Their nuts are important food sources for invertebrate predators such as weevils, moths, and wasps [16,17] as well as dispersers such as scatter-hoarding rodents [18][19][20] in the tropical and subtropical forests [17] of these regions, indicating that a long term fruit-animal coevolutionary relationship influences the diversification of Lithocarpus. The nut of Lithocarpus is a seed surrounded by the husk (composed of pericarp and receptacle) [17], which is further subtended by a cup or disk shaped cupule (a modified sterile branch) (Fig. 1a & d) [21,22]. The marked interspecific variation in fruit and cupule structures has been recognized as important diagnostic characteristics of stone oaks [13][14][15]23]. Cannon and Manos [24] recognized two fruit types embodying the interspecific variation of 21 tropical Lithocarpus species, the acorn (AC) and the enclosed receptacle (ER) fruit types ( Fig. 1). Based on visual examination, the two fruit types are morphologically distinct from each other: morphologically similar to the Quercus acorns, AC fruits are smaller-seeded and mainly enclosed by thin pericarp, Fig. 1 The structure of two fruit types and collected morphometric parameters on the longitudinal sectioned fruit. a-c and d to f represent AC and ER fruit respectively. The black scale bar and the white line in each image represent 1 cm and the longitudinal axis respectively. The original structure is displayed on the left, and the diagrammatic representation is on the right of the longitudinal axis. In a & d, the structures of cupule (C), receptacle (R), pericarp (P) and seed (S) were depicted on the left side of the fruit longitudinal section. In b & e, A p , A r , and A s are the area of pericarp, receptacle and seed respectively, which were parameters for estimating the volume of pericarp (V p ), receptacle (V r ) and seed space (V s ). In c & f, L s , L p , and L r are length of seed space, the internal lengths of pericarp and receptacle respectively, which were collected for estimating the coverage to seed by pericarp (S p ) and receptacle (S r ) and seed surface area (S s ) with receptacle present at the base; ER fruits are unique to the genus, larger-seeded and mainly enclosed by extended, thickened, and lignified receptacle, whereas the pericarp is greatly reduced and set on the top [24]. Fruit development study [17] identified that the fruit morphological distinction between the two types mainly resulted from the heterochrony between the receptacle and pericarp in the later fruit developmental process [17]. A potential physical-chemical trade-off between the two fruit types was proposed by examining the seed chemistry of six stone oak species [25]: AC-type species has a high concentration of antifeedants (fibers) in the seeds as chemical defense; whereas the thickened and lignified receptacle of ER-type fruits represents effective mechanical protection for the more nutritious seed. Later, it was identified that these two fruit types could represent the interspecific morphological variation of Lithocarpus on the genus level [26]. And even though AC-type species are more ubiquitous than ER-type species (roughly 3/1 ratio), they share two distribution centers (southern China and Southeast Asia) and often co-occur in the same forests [26,27]. Examining the evolutionary background of these two fruit types is critical in understanding the diversification and distribution of stone oaks. The predation selection hypothesis was proposed to elucidate the fruit type evolution of Lithocarpus. Seed predation and dispersal [28] strongly impact plant regeneration, distribution, and evolution [5]. The predation selection hypothesis suggests that the two fruit types represent different dispersal strategies, and their morphological distinctions are results of varied predation pressures throughout evolutionary history [25]. Smaller-sized AC fruits could attract a wider range of dispersers, but their weaker physical defense could cause higher pre-dispersal predation [26]. In contrast, ER fruits with thicker husk could inhibit insect infestation more effectively, but they can only be dispersed by larger-sized vertebrates due to their larger fruit size. However, two aspects make it challenging to test the predation selection hypothesis. First, the limited information on insect predators and vertebrate dispersers of Lithocarpus fruits [26] impedes testifying this hypothesis. Second, the distinctions in seed size, fruit morphology, and mechanical defense between the two fruit types were descriptive with limited sample sizes [24,25], as measuring the mechanical defense has been challenging. For instance, measuring the thickness of the fruit husk by calipers would include fruit positioning errors, and ignore the density variation between the receptacle and pericarp. Applying digital force gauge (SHSIWI Model SJX-200-500) to crack the husk also generates results with big variation due to fruit positioning and fruit rotating during measurement (unpublished data by Xi Chen). Applying morphometric analysis to estimate the size of varied fruit parts is a new approach to measuring the mechanical defense, which offers reliable volumetric estimates and could avoid the problems caused by direct measurements [26]. So performing the morphometric analysis with adequate sample size is necessary for clarifying the morphometric distinction and the mechanical trade-off between the two fruit types, which is crucial for verifying the predation selection hypothesis. The unresolved phylogeny of Lithocarpus is another impediment in understanding the fruit type evolution of Lithocarpus. Interfertility, frequent interspecific gene flow, large effective population size, and so on, [29] are common difficulties in phylogenetic analysis of Fagaceae. The phylogenetic construction with 21 Bornean species [24] is the pioneer study to examine the fruit type evolution of Lithocarpus, which suggests ER species could have speciated from AC species in two independent linages. But the relatively low species number and the omission of subtropical species were major limitations of this study. Recently, the sequence analysis of chloroplast DNA (atpB-rbcL) and nrITS of 64 species by Yang et al. [30] provides a more complete phylogenetic relationship of Lithocarpus. Nevertheless, their phylogeny has limited support and the fruit type differentiation along evolutionary history remains unexamined. So it is essential to reconstruct a better-resolved phylogeny of Lithocarpus by employing more gene fragments while including more species which cover the entire distribution range of the genus. In this study, we integrated phylogenetic construction and fruit morphometric study with adequate sample size to elucidate the fruit type evolution of Lithocarpus. We first carried out a fruit morphometric study of 168 Lithocarpus species to describe the interspecific and intraspecific variations of various fruit parts, and tested the morphological and mechanical trade-off between the two fruit types: the small seeds of AC-type species are mainly defended by pericarp representing low physical defense; whereas the large seeds of ER-type species are protected by thickened and lignified receptacles representing strong physical defense [25]. By coupling morphometric and phylogenetic analysis, we then investigated the evolution of fruit type and fruit morphometrics on a subset with phylogeny of 72 Lithocarpus species based on 5 chloroplast DNA and nrITS to test the hypothesis that ER is the derived fruit type from AC-like ancestors [24]. Lastly, to estimate the centers of origin for AC-and ER-type species, we also inferred the thermal diversification center from the present-day geographical distribution of Lithocarpus species. Our study is crucial in clarifying the morphological and mechanical trade-off between the two fruit types, and in verifying the predation selection hypothesis, which would bring about a deeper understanding in the fruit-animal interactions and fruit type evolution of stone oaks. Morphometric distinctions between the two fruit types The 168 study species were classified into 138 AC-and 30 ER-type species (Fig. 1, Table 1), around a 4:1 ratio. Compared to that of AC-type species, the fruit husk of ER species is generally lignified [17,24,25]. Here, assuming zero density difference, the fruit husk volume (V h ) would represent total mechanical investment, and the fruit husk and seed surface area ratio (V h / S s ) would represent the relative physical protection. Based on linear mixed-effect modeling, both the seed volume (V s ) and mechanical defense (V h / S s ) of ER species were significantly larger than those of AC species (Table 2). By performing major regression analysis to estimate the constant and allometric parameters and test the fruit type (AC or ER) dependence on them, we also identified a partition in mechanical defense between the two fruit types. The allometric exponents of AC and ER species in pericarp and receptacle coverage to the seed allometry were significantly different from each other (0.71 and 1.39, p = 1.4810 -7 ), and the allometric constants of which were 1.93 and 0.46, respectively (Fig. 2a). Similarly, the allometric exponents of AC and ER species in pericarp and receptacle volume allometry were also significantly different (0.88 and 0.94, p = 8.3510 -3 ), and the allometric constant of which were 1.77 and 0.65, respectively (Fig. 2b). These results suggest that the pericarp and receptacle provided major physical defense for the seeds of AC and ER species, respectively. Combining the results above, the stronger physical defense of ER fruits was mainly contributed by the receptacle tissue. The morphometric variation of AC and ER fruits Representing total mechanical defense, fruit husk volume increased with species' seed size (Fig. 2c). Receptacle volume increased accordingly with pericarp volume for most of the 168 species (Fig. 3c & d). However, two exceptions indicated negative allometric slope: L. ferrugineus is a small-sized AC species with extremely thin pericarp (Fig. S1 d-f ), and L. javensis is an ER species with great interspecific variation (Fig. S1 a-c). The morphometric variation between fruit types was greater than interspecific variation within each fruit type ( Fig. 2 and 3). Evolution of fruit types and fruit morphology Topologies of cpDNA + nrITS (72 species), cpDNA (58 species), and nrITS (66 species) phylogenetic trees were consistent with one another (Fig. 4 & S2, Table S1 & S2). Topologies of 72 species' phylogeny reconstructed by ML (Fig. 4) and Bayesian analyses (Fig. S3) were also largely congruent with each other. We applied the fully bifurcating ML tree to perform ancestral state reconstructions. The reconstruction showed a major transition from AC to ER fruit in the deeper branches, and after which there were several reversals to the ancestral state (AC). Besides the ER-type species that retained their fruit type, there were also ER fruits derived from acorn-like fruits in the clades independently (Figs. 4 & 5). This was further supported by the repetitive occurrence of the ER fruit type in all clades and subclades of the phylogenetic tree reported by Yang et al. (Fig. S4) [30]. Fruit types showed low phylogenetic signal (K = 0.24), suggesting that ER-type species are phylogenetically distantly related to each other. The low phylogenetic signal of fruit types was further supported by the low phylogenetic signals (K < 1) of all six morphometric dimensions (S p , S r , S s , V p , V r and V s ) (Table S3). Ancestral state reconstruction of the morphometric dimensions showed that around half of the AC-type (28 out of 52) species reduced in seed size throughout the evolutionary time course, while half remained the same (Fig. 6). The increase in seed size was only observed for 5 ER-type species, while the seed size for the rest 15 species remained the same. The fruit husk volume representing physical defense changed accordingly with the seed size for both fruit types (Fig. 6). Congruent with the finding that the receptacle is the main tissue differentiating the morphology of the two fruit types (Tables 1 & 2), the reduction and increase in the receptacle coverage to the seed was observed in both AC-and ER-type species (Fig. 7). Besides L. jacobsii, the only exception exhibiting an increase in pericarp coverage, the pericarp coverage for the rest AC-type species remained the same. More ER-type species exhibited decrease rather than increase in pericarp coverage to the seed (8 compared to 4 species). The evolutionary changes in pericarp and receptacle volumes for both fruit types exhibited a similar pattern to that of the seed coverage by pericarp and receptacle (Fig. S5). There was an increase in both receptacle and pericarp coverage to the seed (Fig. 7) for four ER species, i.e., L. pachylepis, L. lampadarius, L. revolutus and L. pulcher (Fig. S6). These four species shared an intermediate fruit morphology between AC and ER fruits: their receptacle was thickened and extended while the pericarp was not obviously reduced. Unlike two fruit types with clear morphological distinctions, this intermediate morphology suggested that these four species were heavily protected by both tissues. The thermal distribution of two fruit types Based on the distribution median of present-day mean annual temperature (MAT) (Fig. 4, Table 3 & S4), the thermal distribution centers of the 72 examined species were mostly concentrated in the tropics (45 species), followed by warm temperate forests (21 species), and subtropical forests (6 species). The tropical-centered species were mainly clustered into two major lineages. The thermal distribution centers for the other two major lineages were warm temperate forests mainly, with a few species occasionally appearing in the tropics and in the subtropics (Fig. 4). For ER-type species, the thermal distribution centers for 14 and 6 species were tropical and warm temperate regions, respectively. The thermal distribution of AC species was mostly in the tropics (31 species), with 15 species in warm temperate and 6 species in subtropical regions. Except for 13 species, the thermal distribution centers based on the median annual temperature were mostly congruent with those based on MAT (Table 3 & S4). Among the exceptions, the thermal distribution centers of 11 species based on median and mean annual temperatures were tropical and subtropical, respectively, indicating a negative skewness in distribution towards the colder region; and the thermal distribution centers of the remaining two species (L. calophyllus and L. corneus) based on median and mean annual temperatures were warm temperate and subtropical regions, respectively, indicating a positive skewness in distribution towards warmer climate. Discussion The marked interspecific fruit morphological variation and the two major fruit types of stone oaks represent the species diversity of the genus [17,27]. Examining fruit morphological variation and phylogenetic relationship is important in understanding the diversification and fruit Fig. 2 The fruit morphological variation among the 168 species and between the two fruit types. a the levels of seed coverage by pericarp and receptacle. b the volumes of pericarp and receptacle. c the volumes of fruit wall and seed. AC and ER stand for acorn and enclosed fruit type respectively. Each circle represents one species Fig. 3 The intraspecific variations of fruit morphological dimensions within the two fruit types. a the seed coverage by pericarp and receptacle of the AC species. b the seed coverage by pericarp and receptacle of the ER species. c the volume of pericarp and receptacle of the AC species. d the volume of pericarp and receptacle of the ER species. Each circle represents a single fruit, and colors correspond to the 51 AC and 11 ER species with no less than eight fruits respectively evolution of Lithocarpus. We studied the morphometrics of 2,613 fruit specimens of 168 species, which represent over half of the species of Lithocarpus and encompass a wide geographic distribution. Despite the frequent cooccurrence of the two fruit types [26,27], AC-type species are more diverse than ER-type species (4:1 ratio, Table 1), which is consistent with previous study [27]. As the morphological variation between the fruit types was larger than within each fruit type or within each species (Figs. 2 & 3), the two fruit types are key to representing fruit morphological variation of the genus. Instead of applying direct measurements involving various errors, the fruit morphometric analysis successfully quantified the volume of varied fruit parts and identified a major fruit morphological and mechanical trade-off between the two fruit types: smaller-seeded AC fruits were mainly enclosed within thin pericarp, representing weaker mechanical protection; larger-seeded ER fruits, on the other hand, were covered by thickened and lignified receptacle tissue, indicating stronger mechanical defense ( Table 2 and Fig. 2). The mechanical trade-off between the two fruit types serves as strong evidence for the predation selection hypothesis. To further testify this hypothesis, examining the seed predation and dispersal variation between AC-and ER-type species is essential. One thing to point out is that the actual mechanical defense difference between the two fruit types should be even higher than our estimate, as the observed lignification in receptacle tissue in ER-types species was not assessed by the morphometric method. We identified that receptacle variation was the major contributor to the fruit-type morphometric distinctions, and this was more obvious for smaller fruits (Fig. 2a & b). Larger-fruited species sometimes exhibit an intermediate morphology in which the seeds were covered by both pericarp and receptacle (Fig. S4), as they require high mechanical investment to effectively defend their seeds. The morphometric and molecular phylogeny study of 21 Bornean Lithocarpus species [24] suggested that the ER fruit is the derived fruit type that has occurred on at least two separate lineages for tropical stone oaks. Despite a few reversals from ER to AC, we found that ER fruits derived independently multiple times, across all clades on the phylogeny of 72 species, and across climatic ranges from tropical to warm-temperate regions (Figs. 4 & 5). Resembling the common acorn of Quercus in general, AC fruit represents the ancestral form of Lithocarpus, which has experienced less evolutionary modification in pericarp development (Fig. 7 & S5). This result supports the hypothesis that ER is the derived fruit type from the AC-like ancestors [24]. Even though we employed four additional cpDNA fragments compared to the previous study [30], some resulting topologies remained with limited support (Fig. 4). However, the repeated occurrence of ER fruits on our phylogenetic tree and previous phylogenetic trees [24,30] suggests the convergent evolution of ER fruits on different lineages across different regions (Fig. 4 & S4). In addition, the low phylogenetic signal of the two fruit types suggests that ER-type species are distantly related to each other, which further supports the convergent evolution of ER-type species. As ER-type species are signified by their thickened and lignified fruit husk, Cannon and Manos [24] suggested that the ER fruit was a result of strong directional selection toward mechanical protection against seed predators. Based on the two-fruit-type morphometric distinction ( , we propose to modify the directional selection theory to divergent selection of the two fruit types. In their evolutionary history, AC-type species tended to reduce their seed size, mechanical defense and receptacle coverage, with little change in pericarp; whereas ER-type species tended to increase in seed size, mechanical defense, with major modification in receptacle. Compared to the little variation for AC-type species, the evolutionary change of pericarp varied greatly among ER-type species (Fig. 7 & S5). This indicates that more modifications in receptacle development are required to achieve the fruit husk change (especially for ER-type species), which signifies the important role of pericarp in the evolution of fruit morphology. The biogeographical study of 91 Chinese stone oaks signifies the impact of mean annual temperature on the fruit morphometric variation [27], so inferring the thermal diversification center based on current geographical distribution is critical in understanding the diversification and fruit-type differentiation of Lithocarpus. The thermal distribution pattern of 72 species (Fig. 4) suggests Ancestral state reconstruction of fruit type based on the ML phylogeny of 72 Lithocarpus species with cpDNA + nrITS. The fruit type of each species is coded by a colored dot as explained in the legend. Pie charts at nodes show the relative proportion of alternative ancestral state estimated in RASP. AC and ER represent acorn and enclosed receptacle fruit type respectively. AC/ER represents it could be either AC or ER status. The asteroid mark represents hidden probabilities that the phylogenetic radiation of species-rich Lithocarpus involved two geographic speciation centers: tropical versus warm temperate regions. The AC-type species are more diverse with greater distribution ranges that extend from warm-temperate to tropical regions, which could be related to their longer evolutionary history; whereas the later appeared ER-type species have smaller distribution ranges as they are absent in subtropical climates. The thermal distribution centers based on mean and median annual temperatures were consistent with each other (Table 3 & S4), which further supports the tropical and warm temperate speciation centers. Combined with previous phylogenetic results, we conclude that the ER fruit type has evolved independently multiple times across various lineages in both tropical and subtropical-warm temperate regions. As we found ER-type Lithocarpus species to be polyphyletic, environment could be a plausible driver for the fruit type evolution and differentiation in the genus. Evolution of fruits mirrors the co-evolutionary history between fruits and animals. During seed predation and dispersal, fruits associated with insects represent antagonism, whereas those associated with vertebrates represent mutualism [28]. Pre-dispersal seed predation by insects causes seed death or damage, which negatively impacts seed dispersal, germination, and successful seedling establishment [28,31,32]. On the other hand, vertebrates such as rodents are effective seed dispersers as their forgetfulness could reduce seed consumption and enhance seed germination as well as seedling establishment [5,17,20,33]. The various fruit characteristics, i.e., morphology, seed chemistry, and physical defense, act as an entity interacting with their predators and dispersers, which shapes the co-evolution between fruits and animals [7][8][9][10][11][12]. The predation selection hypothesis proposes that the morphological and mechanical tradeoff between the two fruit types could be results under different predation pressure: smaller-sized AC fruits attract a wider range of dispersers, but their weaker physical defense could cause higher pre-dispersal predation, whereas ER fruits with thicker husk could inhibit insect infestation more effectively, but their larger fruit could only be dispersed by larger-body sized vertebrates [26]. The mechanical trade-off between AC and ER fruit types that we identified partially supports the predation selection hypothesis, but the variation in predation and dispersal between the two fruit types is still the missing link. The common insect herbivores of both fruit types (Kira, 1991). One aster mark () indicated that thermal region identified based on median MAT is hotter than that of mean MAT; two aster mark (*) indicated that thermal region identified based on mean MAT is hotter than that of median MAT of Lithocarpus are weevils, gall wasps, bark beetles, and crane flies [16,17], but whether there is a difference in pre-dispersal predation between the two fruit types is yet to be examined. In terms of the seed dispersers, while rats and squirrels are known to disperse AC fruits [18][19][20], the species that disperse ER fruits are currently not well documented. Thus, to further test the predation selection hypothesis, a detailed study of the insect and vertebrate predation and seed dispersal of AC and ER fruits is crucial in clarifying the predation selection hypothesis in the future. Conclusion Overall, our study provides important information for understanding the fruit morphometric and fruit type evolution of stone oaks. We examined the morphological and mechanical trade-off between the two fruit types: the larger seeds of ER fruits are mainly enclosed by receptacle tissue representing stronger physical defense, whereas the smaller seeds of AC fruits are enclosed by thin pericarp representing lower physical protection. The mechanical trade-off between the two fruit types serves as evidence for the predation selection hypothesis. The phylogenetic analysis supports the hypothesis that ER is the derived fruit type from AC-like ancestors independently across all clades. Instead of directional selection of ER species [24], we propose a divergent selection theory for two fruit types: the seed size and mechanical defense of AC fruits were reduced, whereas those of ER fruits increased and required more morphological modifications in receptacle through evolutionary time. This result signifies the important role of the receptacle in the fruit type evolution. Lastly, we found that the phylogenetic radiation of Lithocarpus involved two geographical speciation centers: tropical and subtropical-warm temperate regions. As ER fruits are results of convergent evolution, we propose to examine whether there are different types of predation pressure acting as drivers for the evolution of the two fruit types across climatic regions in the future. Fruit morphometrics data collection Sampling design and fruit image preparation In total, we examined the morphometrics of 2,613 mature fruits of 168 species, including 98 species and 595 fruit samples which were applied in the previous fruit morphometric study [26]. The 168 species represent about half of the recorded Lithocarpus species and encompass a wide range of morphological variation and geographic distribution ( Table 1). The majority of the fruit specimens were collections from six herbaria: the National Herbarium Netherlands, the Harvard University Herbaria, the Herbarium of Kunming Institute of Botany of the Chinese Academy of Sciences, Smithsonian National Museum of Natural History, US National Herbarium, the VNM Herbarium, and the Herbarium of Kyushu University. Additional fruit specimens came from our field collections from southern China between 2015 and 2020. We confirmed the nomenclatures and corrected synonyms based on the International Plant Names Index (2019, http:// ipni. org). Fruit dissection followed the protocol of our previous morphometric study [26], and the images of fruit longitudinal section were captured with a Canon SLR camera (EOS M3, Tokyo, Japan). The fruit image standardization and processing was performed in Adobe Photoshop CS 5.1 [26]. Fruit morphometric dimensions' estimation Assuming each fruit to be a perfect rotating body, we employed the Pappus-Guldinus Theorem to reconstruct 3D fruit shapes from the 2D fruit images [17,26] and to obtain six fruit dimensions S s , S r , S p , V s , V r and V p , i.e., the coverage (S) and the volume (V) of seed space (s), receptacle (r), and pericarp (p) respectively. To estimate these six fruit dimensions, 12 parameters of each fruit were collected using Image J 1.51 h [34]: the internal curve length of pericarp (L p ), internal curve length of receptacle (L r ), curve length of seed space (L s ), and the distances from longitudinal axis to the centroid of L p , L r and L s , namely r p , r r , and r s , respectively (Fig. 1b & e); the section area of pericarp (A p ), receptacle (A r ), and seed space (A s ), the distances from the longitudinal axis to the centroid of A p , A r , and A s , namely R p , R r and R s , respectively ( Fig. 1c & f ). The surface area of the seed space (S s ), the coverage by pericarp (S p ), the coverage by receptacle (S r ), and volume of pericarp (V p ), receptacle (V r ) and seed space (V s ) were determined by the Pappus-Guldinus Theorem [26]. The fruit husk volume was defined as V h = V p + V r . Fruit type identification and fruit morphometric analysis All 168 species were classified into 138 AC-and 30 ERtype species based on the species average seed surface coverage by pericarp and receptacle (Table 1) [27]. The marked difference between pericarp and receptacle coverage (Table 1) of 23 species with a single fruit sample indicated a clear fruit type distinction. The morphometric analysis was performed in R 4.1.0 [35]. We hypothesized a morphometric trade-off and mechanical distinction between AC-and ER-type species: smaller-seeded AC species were mainly enclosed by pericarp which represents lower mechanical defense, whereas larger-seeded ER species were mainly enclosed by receptacle representing better mechanical defense. To test this hypothesis, we first applied two linear mixed models (package lme 4 [36]) with Gaussian-distributed errors to examine the variation in mechanical defense (fruit husk volume divided by seed surface area) and seed volume between the two fruit types by employing all 168 species. Model one set fruit type as the fixed factor, species as random factors; model two only included species as the random factors. We selected the best model by comparing two models by Akaike information criterion (AIC). Then, we employed the standard major axis (MA) regression with package smatr 3.4.3 [37] to compare the seed enclosure by pericarp and receptacle within each fruit type and analyze intra-and inter-specific allometry across selected pairs of dimensions after logarithmic transformation. Interspecific variation was examined on 168 species represented by 2,613 fruit samples. Intraspecific variation was examined on 61 species with eight or more fruit samples each, including 11 ER-type and 50 AC-type species with a total of 2,254 fruit samples. Partitioned Bayesian and Maximum Likelihood (ML) analyses were then conducted on the three concatenated data sets. Bayesian inference was performed in MrBayes v.3.2.7 [40]. According to the Akaike Information Criterion (AIC) values obtained using ModelFinder [41], the best-fit models of nucleotide substitution for the atpB-rbcL, psbA-trnH, matK, rbcL, trnL-trnF and nrITS datasets were determined to be TIM3 + I, TIM3 + I, TVM, GTR, TPM1uf and GTR + G, respectively. Two independent runs with one cold and three incrementally heated Monte Carlo Markov chains (MCMCs) were run for 50,000,000 generations, with trees sampled every 500th generation. Model parameters were unlinked across partitions. We discarded the first 2,500 trees out of the 10,001 trees as burn-ins and used the remaining trees to build a 50% majority rule consensus tree. Maximum likelihood analyses were performed using raxmlGUI 2.0 [42]. A separate General Time Reversible + Gamma model (GTR + G) of nucleotide substitution was specified for each data partition, and 500 independent searches were conducted. Support values for nodes in the phylogenetic tree were estimated across 1,000 pseudoreplicates using the GTRGAMMA model and mapped thereafter onto the best-scoring tree from the 500 independent searches. Finally, FigTree v1.4.4 (http:// tree. bio. ed. ac. uk/ softw are/ figtr ee) was applied to visualize the phylogenetic trees. Based on the phylogeny reconstructed from the concatenated data sets, we ran RASP v.4 [43,44] and Mesquite v.3.61 [45] to infer the ancestral states and evolution of the fruit type and the six fruit morphometric dimensions (S p , S r , V P , V R , V H and V S, after log-transformation), respectively. The phylogenetic signals of the fruit types and six fruit morphometric measures were also estimated in RASP v.4. For comparative purpose, these analyses were also conducted based on the phylogenetic tree of 64 Lithocarpus species that was reconstructed by Yang et al. [30] using cpDNA and nrITS data. Thermal distribution analysis The morphometrics of both fruit types was found positively related to the mean annual temperature of 91 Chinese Lithocarpus species in our previous study [27]. To identify the thermal distribution differences among lineages with AC-and ER-type species, we retrieved the present-day distribution ranges and location description of 72 species from the Plants of the World Online website (https:// powo. scien ce. kew. org/), then applied Google Earth (https:// www. google. com/ earth/) to obtain the geographic coordinates for each locality. To better match with the land size of Southeast Asian countries, the location resolution in China was set to provincial level. As the morphometrics of both AC and ER species were positively related to mean annual temperature according to our previous study [27], we applied Raster v3.4.10 [46] to obtain mean annual temperature (MAT; in degree Celsius) from WorldClim 2.0 (https:// www. world clim. org/). The median annual temperature among distribution locations were calculated for each species, and the thermal distribution was categorized as follows: (i) MAT > 22 °C, tropical; (ii) MAT in between 19 °C and 22 °C, subtropical; (iii) MAT in between 11 °C and 19 °C, warm temperate to correspond to the warmth-index based definition of our target regions [47].	2023-04-29T13:40:58.974Z	2023-04-29T00:00:00.000	{ "year": 2023, "sha1": "0eba848b5759e8345d3f309b6e61111aa0928e92", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Springer", "pdf_hash": "0eba848b5759e8345d3f309b6e61111aa0928e92", "s2fieldsofstudy": [ "Biology", "Environmental Science" ], "extfieldsofstudy": [ "Medicine" ] }
253523013	pes2o/s2orc	v3-fos-license	Exploiting Contrastive Learning and Numerical Evidence for Improving Confusing Legal Judgment Prediction Given the fact description text of a legal case, legal judgment prediction (LJP) aims to predict the case’s charge, law article and penalty term. A core problem of LJP is how to distinguish confusing legal cases, where only subtle text differences exist. Previous studies fail to distinguish different classiﬁcation errors with a standard cross-entropy classiﬁcation loss, and ignore the numbers in the fact description for predicting the term of penalty. To tackle these issues, in this work, ﬁrst, we propose a moco-based supervised contrastive learning to learn distinguishable representations, and explore the best strategy to construct positive example pairs to beneﬁt all three subtasks of LJP simultaneously. Second, in order to exploit the numbers in legal cases for predicting the penalty terms of certain cases, we further enhance the representation of the fact description with extracted crime amounts which are encoded by a pre-trained numeracy model. Extensive experiments on public benchmarks show that the proposed method achieves new state-of-the-art results, especially on confusing legal cases. Ablation studies also demonstrate the effectiveness of each component. Introduction Artificial intelligence (AI) technologies have recently been applied to the legal domain to help legal practitioners reduce heavy and repeat work in many tasks, such as legal judgment prediction (Cui et al., 2022;Feng et al., 2022a;Zhong et al., 2020a), legal information extraction (Cardellino et al., 2017;Yao et al., 2022), legal summarization (Bhattacharya et al., 2019), and legal question answering (Zhong et al., 2020b;Martinez-Gil, 2021). Legal judgment prediction (LJP) is one of the most attractive research topics among these tasks. Given the fact description text of a legal case, legal judgment prediction (LJP) is aimed at predicting its charge, law article, and penalty term (Aletras et al., 2016;Zhong et al., 2018;Yang et al., 2019;Chalkidis et al., 2019). One core problem hindering the performance of LJP from being satisfying is confusing legal cases, which have subtle text differences, but with totally different charges, law articles and penalty terms. A strand of studies on this problem has been conducted. Hu et al. (2018) manually annotate discriminative legal attributes for confusing charges and incorporate this explicit knowledge into an attention-based multi-task learning judgment prediction framework. Xu et al. (2020) explore to extract distinguish knowledge from similar law articles with a graph-based method. NeurJudge (Yue et al., 2021) utilizes the results of intermediate subtasks to separate the fact description into different circumstances and exploits them to make the predictions of other subtasks. However, we argue that there exist two drawbacks to these studies. First, they all use a standard cross-entropy classification loss, which cannot distinguish different mistaken classification errors. For example, the error of classifying the charge Theft into Theft's confusing charge Robbery is the same as classifying Theft into a not confusing charge Rape. The model needs to be punished more if it classifies a charge into its corresponding confusing charges. Second, the numbers in fact descriptions are crucial evidence for predicting the penalty terms of certain cases, such as financial cases. However, these numbers are processed equally with other words of the fact descriptions. It is hard for the model to deduce the total crime amount, and make correct penalty term predictions based on the scattered numbers. To tackle these issues, in this work, we improve confusing legal judgment prediction via contrastive learning and numeracy evidence. In order to deal with the drawback of the cross-entropy classification loss, we propose to use supervised contrastive learning (Khosla et al., 2020;Gunel et al., 2020) to pull fact representations from the same class closer and push apart fact representations from confusing charges. The use of supervised contrastive learning has been shown effective in many domains (Suresh and Ong, 2021;. However, it is non-trivial to apply the original supervised contrastive learning to LJP. The reasons are two-fold. First, the number of charge classes is much larger than the number studied in (Gunel et al., 2020), which increases the difficulty to find sufficient negative examples in the mini-batches. Second, how to apply the original single-task supervised contrastive learning to the three subtasks of LJP remains explorations, especially when the subtasks labels are contradictory. The word "contradictory" here means that instances with the same charge labels may have different law labels or penalty term labels. If we pair the training instances with the same charge labels as positive examples, the resulting learned shared features will benefit the charge prediction task but will degrade the performance of the other two tasks. To tackle these challenges, we modify the momentum contrast (MoCo) unsupervised representation learning (He et al., 2020) into the supervised learning setting, and explore the best way to construct positive examples, which can benefit all the three subtasks of LJP simultaneously. Second, in order to provide direct numerical evidence for the penalty term prediction, we need to extract the total crime amount from the fact descriptions and fuse the crime amount number into the penalty prediction model. The first challenge here is computing the total crime amount. Because the numbers scatter in the fact descriptions, only some numbers are part of the crime amount while the rest are not. To deduce the total crime amount of the legal case, we formalize the number extraction as a named entity recognition task, where the recognized numbers make up the final crime amount. Then, we consider fusing the crime amount number into the penalty prediction model while preserving the numeracy of the crime amount, which is achieved by a pre-trained number encoder model DICE (Sundararaman et al., 2020). We conduct extensive experiments on public benchmarks (i.e., CAIL-small and CAIL-big). The results show that the proposed method achieves new state-of-the-art results, giving up to a 1.6 F1 score improvement for confusing legal cases and a 3.73 F1 score improvement for numerically sensi-tive legal cases. Ablation studies also demonstrate the effectiveness of each component. Related Work Legal Judgment Prediction. In recent years, with the increasing availability of public benchmark datasets (Xiao et al., 2018a;Ma et al., 2021;Feng et al., 2022b) and the development of deep learning, LJP has become one of the hottest topics in legal artificial intelligence (Yang et al., 2019;Zhong et al., 2020a;Cui et al., 2022;Feng et al., 2022a). Various deep learning techniques have been exploited for LJP, such as attention (Luo et al., 2017;Ma et al., 2021), graph neural network (Xu et al., 2020), multi-task learning (Zhong et al., 2018;Hu et al., 2018;Yue et al., 2021) and knowledge injection (Luo et al., 2017;Hu et al., 2018;Xu et al., 2020;Gan et al., 2021). Our work focuses on confusing legal judgment prediction, which is a difficulty in LJP. To solve this challenge, Hu et al. (2018) manually annotate discriminative legal attributes for confusing charges. Xu et al. (2020) explore to extract distinguish knowledge from similar law articles using a graph-based method. Neur-Judge (Yue et al., 2021) utilizes the results of intermediate subtasks to separate the fact description into different circumstances and exploits them to make the predictions of other subtasks. Different from previous studies, our work propose to use supervised contrastive learning to better exploit the label information and addition extracted numerical evidence for penalty term prediction. Contrastive Learning. The goal of contrastive learning (Chopra et al., 2005) is to learn distinguishing representations so that similar examples are closer in the feature space while dissimilar examples are far away in the feature space. Contrastive learning has been widely discussed in computer vision, especially in self-supervised settings (Wu et al., 2018;He et al., 2020). Recently, several studies have extended contrastive learning to supervised settings (Gunel et al., 2020;Khosla et al., 2020;Suresh and Ong, 2021;, where examples belonging to the same label in the mini-batch are regarded as positive examples to compute additional contrastive losses. In this work, we explore how to apply this supervised contrastive learning, which is originally designed for single tasks, for the three subtasks of legal judgment prediction. 3 Method A Multi-Task Learning Framework for Legal Judgment Prediction Let f = {s 1 , s 2 , ..., s N } denotes the fact description of a legal case, where sentence s i = {w 1 , w 2 , ..., w M } contains M words, and N is the number of sentence. Given a fact description f , the LJP task aims at predicting its charge y c ∈ C, law article y l ∈ L and term of penalty y t ∈ T. While previous studies have designed various neural architectures for LJP, these models can be boiled down to the following multi-task learning framework. The overall multi-task learning framework of LJP is shown in Figure 1. Specifically, first, a shared fact encoder is used to encode f into basic legal document representations. The shared fact encoder SharedFactEncoder could be Long-Short Term Memory Network (LSTM; (Hochreiter and Schmidhuber, 1997)), Gated Recurrent Unit (GRU; (Cho et al., 2014)) or pre-trained langugae models, e.g., BERT (Devlin et al., 2019). Second, in order to learn task-specific representations for each subtask, three private encoders are designed to transform the shared representation H f into specific task representations. For example, NeurJudge (Yue et al., 2021) utilizes the results of intermediate subtasks to separate the shared fact representations into different circumstances to improve the predictions of other subtasks. This process is formalized as follows. It is worth noting that the subtasks do not share the parameters of these private encoders. Third, based on these task-specific representations, different classification heads are used to compute the losses for the three tasks. where MLP is multi-layer perceptron and CrossEntropy is the cross-entropy classification loss, respectively. The training objective is the sum of each task's loss as follows: MoCo-based Supervised Contrastive Learning To deal with the drawback of cross-entropy classification loss, supervised contrastive learning (Khosla et al., 2020;Gunel et al., 2020) can be a solution to learn distinguishable representations for confusing legal judgment prediction. Given a batch instance I and an example with index i, the standard supervised contrastive learning views the current minibatch as a dictionary, and looks up the positive key e p that query e i matches best as follows: However, this standard supervised contrastive learning can not be directly applied to LJP. The reasons are two-fold. First, the class numbers (e.g., 119 classes for charge prediction and 103 classes for law article prediction) in LJP are much larger than the number studied in (Gunel et al., 2020), which increases the difficulty of finding sufficient positive examples in the mini-batch. Second, because an instance in LJP has multi labels, how to construct positive examples pairs in LJP remains explorations. To tackle these challenges, we propose MoCobased supervised contrastive learning, and explore the best way to construct positive examples so that it can benefit the three subtasks of LJP simultaneously First, we propose MoCo-based supervised contrastive learning to overcome the difficulty of finding sufficient positive examples in LJP, which use the momentum update queue (He et al., 2020) as the look-up dictionary. Specifically, we maintain one feature queue Q to store sample features of interest, and one label queue L to store sample labels. The queue size can be much larger than a typical mini-batch size, thus can provide sufficient samples for computing the contrastive loss. Given Q and L, for each example i in the mini-batch I, according to the labels in L, we can select positive and negative samples from Q to compute the supervised contrastive loss as follows: e i is the query feature encoded by a query encoder whose parameters are denoted as θ q . The features in Q are encoded by a momentum key encoder whose parameters θ k are momentum updated as follows: where m ∈ [0, 1) is a momentum coefficient. Note that θ k are fixed during the back-propagation. Next, based on the above MoCo-based supervised contrastive learning, we explore two strategies to perform multi-task supervised contrastive learning for LJP. Strategy I A straightforward strategy is to compute a contrastive loss for each task, and then sum them into one loss. Formally, three feature queues, i.e., Q c , Q l and Q t , are used to store task-specific feature, i.e., H c , H l and H t . Three label queues, i.e., L c , L l and L t are used to store task labels, i.e., y c , y l and y t . Based on these denoted queues, we use Eq. 10 compute contrastive loss for each task and the weighted sum of them is denoted as the final loss: The final training objective of Strategy I is defined by: Strategy II When closely examining the above supervised contrastive loss, we can find a severe contradictory phenomenon: summing the three contrastive losses together into one loss may not improve the overall performance. The reason is that while L i sup will improve the performance of task i, it may have negative effect on the other two tasks. For example, L c sup view instances with the same charge labels as positive examples, however these instances may have different law labels or penalty term labels. As a result, L c sup will force the shared encoder SharedF actEncoder to learn features that benefit charge prediction, but these shared features will degrade the performance of the other two tasks. To solve this problem, we propose to view the instances whose three subtask labels are all the same as positive examples, and impose MoCo-based contrastive loss on the shared features H f . Specifically, we use a feature queue Q B to store the shared features H f , and three label queues L c , L l and L t to store the task labels. Then the positive samples set for sample i is denoted as P Based on this strategy, we can use Eq. 10 to compute the contrastive loss, denote as L B sup . Compared with Strategy I, Strategy II is able to overcome the contradictory phenomenon and improve the performance of all three subtasks. The final training objective of Strategy II is defined by: Numerical Evidence Extraction as Named Entity Recognition The numbers in legal cases are crucial evidence for predicting the penalty terms of specific charges, such as financial cases. However, the numbers are distributed randomly throughout the fact description and are processed the same as other words in earlier studies. Thus it is difficult for the model to directly deduce the accurate total crime amount and predict correct penalty terms based on the scattered numbers. In this work, we propose to provide direct numerical evidence for the penalty term prediction. Towards this goal, first, we need a model to annotate the total crime amount for each instance in the LJP dataset. We formalize this crime amount annotation as a named entity recognition (NER) task, and the sum of the recognized numbers is regarded as the final crime amount. The reason for this formalization is that recognizing which numbers are part of the crime amount is a much simpler Fact Description The court found that at xxx, the defendants A and B stole a black rectangular wallet and a white cell phone (Metu M4 brand, worth $900) from the victim xxx, while he was asleep. The wallet contained cash of $2,600. After Being arrested, the two defendants returned the aforementioned cell phone and cash of $2,000 to the victim. Crime Amount: $3,500 Figure 2: Numerical evidence extraction as named entity recognition. task than directly computing the total crime amount from the fact description. Specifically, we train the crime amount tagging model on the LAC dataset 1 . Given an instance (f, m) in LAC, where f and m denote fact description and crime amount respectively, we first convert the instance into the BIO NER format. The concrete format conversion process is as follows: we use a 0-1 package algorithm to find a set of sentences in f , where the sum of their numbers equals to the crime amount m. Then the numbers in the selected sentences are labeled as named entities. The converted dataset is named LAC-NER, based on which we train the state-of-art BERT-CRF as the NER model. Now, each instance in the LJP dataset can obtain a crime amount label m annotated by the well-trained NER model, denoted as (f, y c , y l , y t , m). Figure 2 shows an example of extracting crime amount from the fact description. Second, we consider fusing the crime amount number into the penalty prediction model while preserving the numeracy of the crime amount. To achieve this goal, we propose to pre-train a number encoder model using DICE (Sundararaman et al., 2020). Given a pair of numbers, the cosine similarity of their DICE embeddings has a linear relationship with respect to their actual distance. Specifically, given a training data where (x i , y i ) represents a pair of numbers, we use the following training objective to optimize the DICE encoder: where x i and y i are contextual representations of the two numbers, which are obtained by a LSTM model. Given a training instance (f, y c , y l , y t , m), its crime amount m is encoded by the pre-trained DICE and then is fused into the penalty prediction head. Specifically, we modify Eq. (7) into the following formulas: (17) where [; ] denotes a concatenation operation. Datasets To evaluate the effect of our method, we conduct experiments on two public benchmark datasets (i.e., CAIL-small and CAIL-big), from the Chinese AI and Law challenge (CAIL2018) (Xiao et al., 2018b). Each instance in both datasets contains one fact description, one applicable law article, one charge, and one term of penalty. To give a fair comparison, we use the code released by (Xu et al., 2020) to process the data. All models are trained on the same dataset. Table 1 shows the detailed statistics of the datasets. Implementation Details Following (Xu et al., 2020), we use the THULAC 2 tool to segment Chinese into words. The word embedding layer in the neural network is initialized by pre-train word embeddings provided by (Zhong et al., 2018 the parameters, and the learning rate is set to 1e-3. We train the DICE model for 100 epochs 3 . For training, we use the Adam optimizer and set the learning rate to 1e-3. The batch size is set to 128. We train the model for 16 epochs, and select the best model on the validation set for testing. In contrastive learning, the MoCo queue size and the temperate t is set to 65536 and 0.07, respectively. More details about the hyper-parameters can refer to Baselines We implement our method based on the following non-pretraining and pre-training models: • HARNN (Yang et al., 2016): an RNN-based neural network with a hierarchical attention mechanism for document classification. • LADAN (Xu et al., 2020): LADAN distinguishes confusing law articles by extracting distinguishable features from similar law articles using a graph-based method. • NeurJudge+ (Yue et al., 2021):NeurJudge+ utilizes the results of intermediate subtasks to separate the fact description into different representations and exploits them to make the predictions of other subtasks. We also evaluate the effects of the proposed method on pre-trained language models. • BERT (Devlin et al., 2019): BERT is a pretrained language model, which has achieved results for numerous tasks. As the evaluated datasets are in Chinese, we use bert-base-chinese 4 to conduct experiments. • CrimeBERT (Zhong et al., 2019): Crime-BERT is initialized by BERT, then is further pre-trained on crime data. Development Experiments To empirically evaluate which strategies is better for performing multi-task supervised contrastive learning for LJP, we conduct development experiments on CAIL-small. The adopted backbone is LADAN. The results are listed in Table 4. From LADAN on the three tasks. However, the gains of Strategy I is much smaller than the gains of Strategy II, which verifies our hypothesis that directly imposing contrastive learning loss L i sup for each task will improve the performance of task i, but it has negative effect on the other two tasks. Second, we also explore the effect of combining these two strategies, i.e., using L cl + L B as the supervised contrastive loss. As seen, the improvement of this combination is not significant. Thus, in the rest experiments, we take Strategy II as the supervised contrastive learning loss if not stated otherwise. The final method combined superviced contrastive learning and numerical evidence is denoted as NSCL. Main Results To evaluate the effectiveness of the proposed method, we argument each baseline with the proposed two methods (i.e., NSCL) and conduct experiments on CAIL-small and CAIL-big. The results are listed in Table 2 and Table 3, from which we have the following observations. First, the proposed method can improve all the baselines and achieve new state-of-the-art results on the two datasets. Specifically, on CAIL-small, the absolute improvements can be up to a 1.15 F1 score on the charge prediction, a 1.57 F1 score on the law article prediction and a 1.67 F1 score on the term of penalty prediction, respectively. Second, on CAIL-big, the gains are relatively small, giving an absolute 0.54 F1 score improvement on the charge prediction, an absolute 1.16 F1 score improvement on the law article prediction, and an absolute 0.53 F1 score improvement on the term of penalty prediction, respectively. Third, we observe that on CAIL-big, NeurJudge+ gives worse performance than other baselines. We suppose the reason is that the complex neural networks of NeurJudge+ cause its overfitting on CAIL-big. We also conduct experiments on BERT and its obtains an absolute 2.55 F1 score improvement on confusing charges, which shows that the mocobased supervised contrastive learning is effective for learning distinguishable representations for confusing charges. Effect of Numerical Evidence for Number-Sensitive Legal Cases To investigate the effect of numerical evidence for predicting the term of penalty, we conduct ablative experiments. As shown in Table 6 Figure 3, compared with LADAN MTL , the learned features (e.g., the features of charge Pick a Quarrel) of Strategy II are more compact. However, the features of charge Intentional Injury with Strategy I are less compact than the corresponding learned features of Strategy II, which demonstrates the advantage of Strategy II. Table 8 shows two cases to demonstrate the effect of the proposed method. As shown in the first case, LADAN MTL wrongly predicts the case's charge, which should be Intentional Injury, into its confusing charge Pick a Quarrel. However, with the proposed contrastive learning, this error is corrected. The second case demonstrates the effect of numerical evidence. With the extracted crime amount of $20,800, the model is able to correctly predict the term of penalty. Case Study In this work, firstly, we explore to use the supervised contrastive learning to learn distinguishable representations for confusing legal judgment predictions. Secondly, we provide direct numerical evidence for predicting the term of penalty. The extraction of the crime amounts is formalized as a named entity recognition task, and the extracted number evidence is fused into the representation of the fact description with a pre-trained numeracy model. Extensive experiments on public benchmark datasets show that the proposed method can improve existing LJP models, giving new stateof-the-art results. Ablation studies further demonstrate the effectiveness of the proposed method on confusing and number-sensitive legal cases.	2022-11-16T10:47:42.683Z	2022-01-01T00:00:00.000	{ "year": 2022, "sha1": "2da9c6819872a69f26e80cab7f7c177a31e62678", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "2da9c6819872a69f26e80cab7f7c177a31e62678", "s2fieldsofstudy": [ "Law", "Computer Science" ], "extfieldsofstudy": [ "Computer Science" ] }
67799529	pes2o/s2orc	v3-fos-license	INSTITUTIONAL THEORY FOR EXPLAINING CORRUPTION : AN EMPIRICAL STUDY ON PUBLIC SECTOR ORGANIZATIONS IN CHINA AND INDONESIA Many researches on corruption examined macro factors such decentralization, political democracy, press freedom, and economic freedom, as shown by Lecuna (2012), Alexeef and Habodazzova, (2012) and Goel and Nelson (2005). However, there are limited studies on corruption that examine this topic from organizational approach. The main purpose of this paper is to investigate existing institutional theories describing corrupt behaviour in Asian public sector organizations. A total of 171 questionnaires were distributed to public service officers who were currently enrolled as accounting postgraduate students in both China and Indonesia. The results support the institutional theoretical model used to explain corruption in public sector organizations. However, cultural differences in democracy was not a significant factor on respondent’s perception concerning corruptions in both of countries. Introduction Asia is a potential region for conducting research on corruption issues (Luo, 2002).Transparency International (2013) showed high-corruption level among public sector organizations in 34 countries in Asia.As compared to other Asian Countries, Singapore, Hongkong SAR, and Japan have relatively higher Corruption Perception Index (CPI) with score of 86, 75, 74 respectively.Rank below these three countries is United Arab Emirates, Qatar, Buthan, Taiwan, Brunei, Korea Selatan and Malaysia that have CPI ranking between 50 and 70.Countries which have scored less than 50, indicate corruption is a serious problem.Past studies on corruption have focused macro factors such as decentralization, political democracy, press freedom, economy freedom and fiscal decentralization that could affect corruption (e.g., Lecuna, 2012; Alexeef and Habodazzova, 2012; Goel and Nelson, 2005).Furthermore, limited studies have examined corruption on organizational level.The work of Luo (2002), Pillay and Kluvers (2014) are some of the examples that examined corruption in organizational level.Luo (2005) argued corruption research that used organizational approach is vital for many reasons.First, organization is the place where corruption may take place.Second, studying corruption among organizations may lead us to understand what drive corruption at organizational level.Third, organization is the primary lead to understand corruption level of a country.Fourth, corruption studies might significantly contribute to improve organization well being.Corruption often slows down organization performance and firms have to pay more for the damages caused by corrupt practices. The Global Economic Crime Survey conducted by Pricewaterhouse Coopers (2011) showed that assets misappropriations, accounting fraud, bribery and corruption are considered the most fraudulent practices found in public sector organizations.Hence the research objective of this study is to provide empirical evidence on public sector organizations' involvement in corruption within the institutional framework.The model is derived from Luo model (2005) which incorporates institutional theories as the grand theory (DiMaggio and Powell, 1983;Eisenhardt, 1988;Tolbert and Zucker, 1996;Scott, 2001). Institutional model believes that corruption at organizational level is caused by lack of support from task environment, poor comprehension of the regulations as well as execution and practices of these regulations.Other aspects are weak commitment to eradicate corruption, lack of transparency of institutional environment and, the complexity of administration system (Luo, 2005;Pillay and Kluvers, 2014).Pillay and Kluvers (2014) provided empirical evidence of corruption at public sector organizations in South Africa which is a democratic and developing country.They managed to describe corruption at organizational level by incorporating Luo's Model in their study. Besides the motives for corruption, studies conducted in some high corrupt level countries considered culture as a variable that is significantly correlate to corruption.For examples, Treisman (2000), and Pellegrini and Gerlagh (2008) have identified culture as a factor that contribute toward corruption. As for the two countries in this study (based on CPI 2013 issued by Transparency International), China scored 39 which is slightly better than Indonesia with a score of 32.These scores reflected a high-level of corruption in these two countries. The general aim of this research is to understand corruption at public sector organizations from organizational perspective.The specific aim is to examine the influence of task environment and institutional environment concern for corruption, and the culture of democracy toward corruption in China and Indonesia. Institutional theory Institutional theory has been discussed in many research and literatures (DiMaggio and Powell, 1983; Tolbert and Zucker,1996;Scott, 2001).However, the rationalization of corruption research has shifted from the competitive marketplace to the state and professions. As for isomorphic process, DiMaggio and Powell (1983) had identified three mechanisms namely: coercive, mimetic and normative that could influence organizations guest for change.Coercive isomorphism describes organizational change as a result of a political decisions introduced by the authority.In public sector, an organization often must implement new regulation(s) initiated by the government.Mimetic isomorphism refers to environment uncertainty and ambiguous goals that lead organizations to imitate others.In normative isomorphism, organizations and professions are subjected to change as a result of pressure from peers.Tolbert and Zucker (1996) explained that based on individuals' interests, they would accept and follow social norms unquestioningly, without any critical reflection or resistance.For instance, corrupt environment would lead individuals to behave corruptly as they considered it as a common norm. Corruption Corruption can be defined broadly or narrowly depending on the focus of study.Generally, corruption is defined as a behaviour, which deviates from the norm or violation of the rules.The motivation is personal gain by using his/her position.While the broadest sense, corruption is a deviant behaviour against individual formal responsibilities in various institutions / organizations (not just the government or the public sector) for the sake of personal gain (Luo, 2002).Corruption may be characterized based on the following nature: Corruption is perceptual.The behaviour of an individual related to corruption due to wrong perception.This entails community perception that corruption is common. Corruption is contextual.Corrupt behavior is influenced by ideology, paradigm, culture and other inherent corruption contexts.Politics not only affect how one understands and defines corruption, but it also generates a social behavior such as corruption itself. Corruption is power-related.To perpetuate corruption, a corrupt person should be in a strong position in a government or an organization. Corruption is illegal or norm-deviated.Corruption is an illegal act that is characterized by the unauthorized money transfer in which the aim is to gain a personal advantage.Violation of the regulation is one characteristic of corruption.Hence, the government should promulgate easy-to-understand laws relating to corruption for a clear-cut between corruption gift-giving, which is common in some culture. Corruption is intentional.The motivation for personal gain is attached to the connotation of corruption. Corruption's mode of expression is usually covert.Corruption tends to be hidden or informal, making it difficult to detect.This study will use the generic definition of corruption, in the context of examining corruption in public sector organizations.Acts of corruption in this study also referred to theft of assets (asset misappropriation), tax fraud, accounting fraud, bribery and money laundering (Pricewaterhouse Coopers, 2011).Luo (2005) explained that in the institutional model, the task environment and institutional environment will affect individuals in an organization to perform fraudulent acts (malfeasant behaviour).Malfeasant behaviour could led to the development of lack of focus and deterrent outcomes.This could result in the organization weak and unable to respond to environment change. Organizational anti-corruption became mechanism described by Luo (as a mean to prevent corrupt practices) incorporates the elements of organizational culture, organizational structure and compliance system.Organizational culture is the tradition where decision-making is morality-based, while organizational structure is a structure that aims on detecting and correcting any fraudulent act occurs within the organization.Compliance systems, on the other hand, are built to prevent corrupt practices through the development of anti-corruption programs and codes of conduct within an organization. Corruption in China and Indonesia CPI ranks countries based on their level of corruptions.The index indicates perception of respondents toward corruption based on a scale between 0-100.Zero score indicates highest level of corruption, while a score of 100 indicates the lowest corruption index.Based on CPI year 2013, China scored 39, which is relatively higher than Indonesia with a score of 32.This implies that China is slightly less-corrupt than Indonesia.However, both countries are the same range of score between 30-40, thus, corruption level is still considered high in both countries. China has become one of the world economy giants and attracts large foreign direct investment.However, it continues to have a high-level of corruption.CPI issued by Transparency International (2013) ranked China in 87 position from total 177 countries with a CPI of 39. President Xi Jinping is determined to eradicate corruption and one of his political priority is to combat corruption in China.In 2013, China Government has expelled 17 senior officials (positions equivalent to ministry-assistant and governor-assistant), conducted investigation on 197.000 corruption cases, and has sentenced up to 182.000 government employees who were involved in corrupt practices (Pei, 2014). Hypothesis development DiMaggio and Powell (1983) argued that organization would change correspondingly to coercive, mimetic and normative processes.Coercive process explains that external pressure such as from the government, will direct organizational to change.Mimetic process describes that environment uncertainty and ambiguous goals that would also guide the organization for change.Luo (2005) explained that task environment consists of information, external resources or conditions that might affect achievement of a strategy. The concentration of powers in the government and weak regulatory oversight allow government officials to intervene policies and gain access to resources.These conditions resulted in the opportunities for the business community to co-opt with government officials with motive of achieving individual gains. Institutional environment consists of three elements namely: transparency, fairness and complexity.Transparency is the degree of openness and ease in understanding the applicable rules.Luo (2005) found that ambiguous rules provide opportunities for government officials to engage in corrupt practices and exploit weaknesses of these rules.Fairness is described as rules that can be enforced and implemented fairly.Complexity, on the other hand is a system of rules and socio-cultural environments that are difficult to understand, subsequently, triggers people to commit corrupt practices (Luo, 2005; Pillay and Kluvers, 2014). The hypotheses (H) for this study are stated as the follow: H1: Task environment affects the occurrence of corruption H2: Institutional environment affects the occurrence of corruption Culture of Democracy Evaluation of corruption level across countries based on empirical research is challenging because the definition of corruption is fluid influenced by different cultures and also difficult to detect corruption as it is a closed nature (Pellegrini dan Gerlagh, 2008).Hence, current research tried to compare respondents' perception towards high-level corruption countries with different cultures and democracy level. China which is ruled by single political party (Communist Party).The Economist Intelligence Unit (2007) classified china as an authoritarian regime based on its democracy index.On the other hand, Indonesia is categorized as a flawed democracy since it has multiple political parties.Democracy index was built based on five categories; election process and pluralism, civil freedom, government function, political participation, and political cultures.Some researches argued that level of democracy correlates negatively with corruption (See Hill, 2003; Chowdhury, 2004; Bohara et al, 2004; and Pellegrini and Gerlagh, 2008).On the other hand, Treisman (2000) argued that level of democracy has no significant impact on corruption.Based on explanation above, the following hypothesis is being proposed. H3: Level of democracy affects perception of corruption. Methodology Data presented in this paper were collected by distributing 210 questionnaires to public service officers who were pursuing postgraduate study in accounting at Faculty of Economics and Business, Universitas Jenderal Soedirman (UNSOED), Indonesia, and College of Management, Hebei University (HBU), China.They were chosen based on their work experiences in public sector organizations with a minimum of one year working experience.Out of the 210 responses, 39 questionnaires were not usable due to incomplete items.Only 171 questionnaires representing a response rate of 81,42% were valid for further analysis. Research variables The variables used in this study consist of task environment and institutional environment as independent variables, as well as deterrent outcomes to measure the impact of corruption among the public sector institutions.The research model was developed from previous studies conducted by Luo (2005) and Pillay and Kluvers (2014). The questionnaires were translated into Indonesian and Chinese language which were reviewed by language experts to avoid linguistic ambiguities.Seven items represent task environment, eight items for institutional environment, and two items for deterrent outcomes were adopted from previous studies.All items were measured using Likert scale, ranging from ‗1' representing Strongly Disagree, and a score of ‗5' representing Strongly Agree. Validity and realibility of the questionnaires Reliability testing (Table 1) for each variables showed Cronbach's Alpha of more than 0.60.It gave the indication of an acceptable internal consistency.Pearson Product Moment was used to confirm validity of the instrument in this research.Table 2 showed each items in questionnaires has r-value > 0.126 (rtable ). Descriptive statistics A total of 171 responses were received, 89 of the respondents were from China and 82 were from Indonesia.Table 3 showed that slightly more than half (57.9%) of the respondents were females while 42,1% were male respondents.Table 4 categorized the respondents based on job hierarchy: Low management public service officers (63.7%), middle management (24%), top management (4.1%), and others (8.2%).5 showed that asset misappropriations was the most frequent type of corrupt practices committed by respondents in China (47.2%) and Indonesia (32.9%).This result was consistent with the survey conducted by Pricewaterhouse Coopers (2011) among public sector organizations across countries.Bribery was also a common type of corruption committed by respondents in China (37.1 %).Luo (2002) argued that corruption in Asia came from gift-giving culture (known as ‗guanxi' in China; ‗wa' in Japan; and ‗inhwa' in Korean), and an act for keeping relationship within business environment.Under corruption motive, gift-giving culture between individuals was orientated to achieve personal gains.It is a norm-deviated motives applied in business practice in many countries. Hypothesis testing The results from subsequent analysis were discussed below.Table 6 provided the results from regression analysis to test the regression model developed for this study.Unstandardized regression coefficient (B), standardized regression coefficient or beta (β), the intercepts, multiple correlation coefficients (R), and the coefficient of determinations ( ) for the model were presented.The next step was to examine the model .The value for the regression model was 0.540 and, it was statistically significant at the p < 0.05 level.These results implied that all the independent variables had accounted for 54% of the variation in the model.The remaining variation 46% could be explained by unobserved independent variables. Based on the information on Table 4, task environment has t-value 2.246 > t-table 1.653, and significant at 0.05 level (p>0.026).Hypothesis 1 that stated task environment affects the occurrence of corruption was supported statistically.Institutional environment has t-value 6.492 > t-table 1.653, and significant at 0.05 level (p>0.000).Hypothesis 2 that stated institutional environment affects the occurrence of corruption was also supported statistically. Hypothesis 3, examined the distinct perception toward factors (task environment and institutional environment) which affect occurrence of corruption based on different democracy-culture among countries.Table 7, presented homogeneity of variances test p-value > 0.05.This means that there is similarity between respondent groups.Therefore, Analysis of Variance (ANOVA) was used to examine the differences between groups.ANOVA test shown that p-value > 0.05 (Table 6).Both groups have similar perceived factors (task environment and institutional environment) which trigger corruption.Hence, Hypothesis 3 could not be supported statistically.This study has also provided evidence on the factors that drive corruption in China and Indonesia.DiMaggio and Powell (1983) argued that organizations would respond accordingly to coercive, mimetic and normative process.The mimetic process explains that external pressure such as control of the task environment, regulations, inconsistent structure, concentration of power in a particular group, and coercive process (institutional environment) such as transparency, fairness, complexity of institutions could influence corruption level in the context of public sector organizations.Hypothesis 3 could not be supported statistically.It could be explained that culture of democracy has no correlation with occurrence of corruption in China and Indonesia.This result is consistent with Treisman (2000) who argued that level of democracy has no significant impact on corruption level. Conclusions Task environment and institutional environment are significant factors influencing corruption in public sector organizations.Institutional theory can be used to explain the occurrence of corruption at the organizational level.However, culture of democracy has less influence on corruption in China and Indonesia.Treisman (2000) argued that level of democracy has no significant impact on corruption level. This study contributes to corporate governance research.It is aimed to explain corruption behaviour that could be used as fundamentals in designing a management control system in public sector organizations especially in countries with highcorruption level such as China and Indonesia.The limitation of this study is that survey respondents do not represent public service officers in each department / divisions.Future research could involve respondents from wider geographical area and across departments / divisions, which in turn could enhance generalization of the results. This research hypothesized that task environment and institutional environment affect the occurrence of corruption (Hypothesis 1 and 2).The multiple regression results showed the independent variables (task and institutional environment), have a positive correlation with corruption.These results provided empirical support for institutional models proposed byLuo (2005), Pillay and Kluvers (2014) developed these models further by taking into account the background of countries with relatively high level of corruption. Table 4 . Job Hierarchy Table 5 . Types of Corruption Table 7 . Homogenity of Variances Test Table 8 . Analysis of Variance	2018-12-29T13:24:07.176Z	2015-01-01T00:00:00.000	{ "year": 2015, "sha1": "50090e9b93ac516e375c02392da00602b1cb548a", "oa_license": "CCBYNC", "oa_url": "https://virtusinterpress.org/spip.php?action=telecharger&arg=6344&hash=1cc175d3532f1633392a756aa71add927ec16c00", "oa_status": "HYBRID", "pdf_src": "ScienceParseMerged", "pdf_hash": "50090e9b93ac516e375c02392da00602b1cb548a", "s2fieldsofstudy": [ "Political Science" ], "extfieldsofstudy": [ "Business" ] }
5305290	pes2o/s2orc	v3-fos-license	Demystifying Consciousness with Mysticism? Cognitive Science and Mystical Traditions The article considers whether, and how, current scientific studies of consciousness might benefit from insights of mystical traditions. Although considerable effort has been expanded towards introducing mysticism into mainstream cognitive science, the topic is still controversial, not least because of the multifariousness of meaning associated with the term (from " illogical thinking " through " visions " and " raptures " to " paranormal " and " psychopathological phenomena "). In the context of the present article, mysticism is defined as a set of practices, beliefs, values etc. developed within a given religious tradition to help the practitioner realize the experiential and existential transformations associated with mystical experiences, i.e. experiences characterized by the breakdown of the subject-object dichotomy. It is then examined in which areas mysticism so defined might provide beneficial for consciousness studies; broadly, three such areas are identified: phenomenological research (mysticism as a repository of unique experiential material and practical know-how for rigorous phenomenological analyses), the problem of the self (mysticism as a repository of experiential-existential insights into one's fundamental selflessness), and the so-called hard problem of consciousness (mysticism as a unique experiential-existential answer to the mind-body problem). It is contended that, contrary to popular belief, cognitive science could benefit from insights and practices found in mystical traditions, especially by way of grounding its findings in the lived experience and thereby (potentially) demystifying some of its self-imposed abstract conundrums. INTRODUCTION 1 The main goal of this article is to consider whether current studies of consciousness, conducted under the aegis of cognitive science, might benefit from insights of mystical traditions.It is our contention that the answer to this question is in the affirmative, and we will try to substantiate this claim by briefly sketching why, where, and how this exchange between cognitive science and mysticism might take place.Before proceeding, however, it should be noted that this paper makes no pretence of being a comprehensive study of the suggested "points of contact", but is merely a preliminary sketch intended to provide food for additional thought (no more, no less). Although considerable effort has been expanded towards introducing the insights of mystical traditions into mainstream cognitive science (cf.[1][2][3]) the topic is still likely to raise many a sceptical eyebrow.Since the term "mysticism" is frequently used as a synonym for "religious experience, mythology, miracles, schizophrenia, hallucinations, trances, altered states of consciousness, alleged psychic powers such as levitation, visions, parapsychology, and in general anything considered irrational, unintelligible, or occult" (e.g.[4]), some of the readers with a more "conventionally scientific bent" might feel ill at ease -and rightly so! -with the prospect of introducing such seemingly obscure topics into what is believed to be a rigorous scientific discipline.To alleviate these worries, however, we would like to point out that this ordinary or "folk" understanding of the term is in stark contrast to how it is normally used in academic circles, particularly in religious studies and philosophy of religion, where it typically refers to a particular set of trans-confessional and trans-cultural experiences that are said to possess a unique set of phenomenological characteristics and are capable of instigating a profound existential transformation in the life of a mystic 2 .In order to get a better understanding of what mysticism (in the proper sense of the word) stands for, it is therefore necessary to get a better understanding of what mystical experience stands for, i.e. to see whether it is possible to elucidate its main phenomenological features. We will try to do this in two steps.Firstly, it is important to consider what mystical experiences are not.Some unusual experiences are frequently associated with mystical experiences, but should not be confused with them; these include: visions, auditions, locutions, trances, ecstasies, paranormal phenomena (telepathy, precognition, clairvoyance etc.) [5][6][7].Although these experiences can and do occur on the "mystical path", they do not qualify as mystical in the strict sense.Different mystics tend to interpret such "accompanying experiences" differently: most frequently, they are perceived as hindrances and thus as something to be avoided or at least ignored; occasionally, however, they are depicted as helpful, but tricky and potentially dangerous guides on one's spiritual journey.Either way, they are normally delineated from mystical experiences proper.The words of Grace Jantzen on the status of visions, auditions etc. in Christian mysticism are instructive in this regard: Some of them, like John of the Cross or the author of The Cloud of Unknowing give grave warnings against all such experiences, advocating that we should treat them all as demonic in origin and saying caustically that they make people behave 'like sheep with the brain disease'.Others, like Teresa of Avila and Julian of Norwich, are grateful for the visions they have had, and recognize in them the source of much spiritual teaching.Yet even those who are most affirmative of the experiences they have had never see them as essential to spiritual growth, never advocate that others should try to have them too, and certainly never see them as central to what union with God is about [4; p.70]. But if mystical experiences are not to be confused with visions, auditions etc., then what are they?Several portrayals of mystical experiences have been proffered, and although they differ in particularities, they tend to agree in generalities.The most prominent characteristic of "mystical experience proper" seems to be the breakdown of the subject-object dichotomy, i.e. of the sense of my being separated from the world.This breakdown, where both "the self" (interiority) and "the world" (exteriority) are extinguished or transcended, is normally associated with the experience of oneness and/or nothingness, and entails a radical transformation of one's state and manner of being.For lack of space, I am unable to go into a more detailed phenomenological analysis of mystical experiences (see [8;Ch.1,8;Ch.3]for a more detailed account), but suffice it to say that the term seems to be covering a whole spectrum of experiences distinguished by how this subject-object breakdown is realized.On the one end of the spectrum, there are experiences of absolute nothingness/oneness, i.e. experiences emptied of all phenomenological content (sensations, thoughts, volitions, emotions etc.) in which nothing but pure oneness/nothingness is present; and on the other end of the spectrum we find experiences where this nothingness/oneness is present in and through phenomenological content.Between these two extremes lie experiences in which nothingness/oneness is experientially/existentially realized to a lesser or greater degree.I will refer to the first type of experience as the (experience of) transcendental (introvertive) Oneness-Nothingness (TON; corresponding approximately to Forman's Pure Consciousness Event), and to the second type as the (experience of) lived (extrovertive) Oneness-Nothingness (LON; corresponding approximately to Forman's Unitive Mystical State), cf.[9,10].It seems that most mystical traditions (contra Stace, pro Forman) put greater value on LON, seeing it as the pinnacle of their spiritual/religious/existential striving.Now that we have a general (albeit admittedly very sketchy) understanding of mystical experiences it is time to consider how they are related to mysticism in general.In the context of the present article, mysticism will be understood as the general platform where mystical experiences are developed, i.e. as a set of different practices, beliefs, values etc. (characteristic of a religious tradition in which the whole process takes place) that help the practitioner realize experiential and existential transformations associated with mystical experiences, cf.[11].Although individual practices, beliefs etc. may differ from one religiocultural context to another, they bring about the same type of experience.Particularly important, and in need of special mention in this context, are meditative/contemplative practices that are considered to play a particularly important role in the overall process. POINTS OF CONTACT Armed with the basic understanding of mysticism and mystical experience, we are now in a position to move on to the central part of our discussion and try to provide a tentative outline of the possibilities for establishing a potentially fruitful and mutually enriching dialogue between mysticism and cognitive science in the field of consciousness studies.This outline will consist of two major parts.In the first part, we will try to elucidate how experiential insights and practical know-how gained and developed by mystical traditions might contribute to current phenomenological research.In the second part, we will then go on to suggest how mysticism might help scientists confront two metaphysical conundrums that have been bedevilling modern cognitive science: the problem of the (unified) self and the socalled hard problem of consciousness.It will be argued that mysticism may provide interesting, if challenging and perhaps even controversial contributions to the study of consciousness, ones that should not be dismissed on a priori grounds but deserve full attention. PHENOMENOLOGICAL RESEARCH The first area where "mystical lore" might prove of great value is phenomenological research.In recent years, the importance of phenomenology in cognitive science has steadily increased, and it has now become generally acknowledged that phenomenology (pace cognitivism) is indispensible for the study of consciousness: "For consciousness is essentially an interior phenomenon, something we experience as subjectivity.Thus if we were not able to identify the subjective phenomena of consciousness directly, that is, subjectively, we would have no way to know which externally observable phenomena were relevant to what phenomena of consciousness, or in what ways" [3].There are at least two possible ways in which mysticism could contribute to phenomenological research. EXPERIENTIAL ASPECT Mystical accounts (written and oral) abound in descriptions of unique experiential (i.e.mystical) states that are (i) normally not present in our ordinary lives and (ii) exhibit surprising phenomenological similarities across cultural and religious traditions.It would seem, then, that they provide precious material for the study of the utmost recesses of consciousness that are unreachable for most.But why should this, aside from sheer intellectual curiosity, matter?What, if anything, consequential can these experiences contribute to our understanding of the normal waking consciousness?There seem to be at least two good reasons for pursuing this type of research.The first reason is best exemplified by drawing a parallel with pathological experiences, i.e. experiences that accompany specific physiological or psychological abnormalities.It is an established fact that the scalpel of a disease, albeit usually limited to very few, can be extremely useful in revealing the normal functioning of our mind-body; and it seems plausible to assume that extraordinary experiential states that are not abnormal (there are actually good grounds for calling them hypernormal) would be equally if not even more revealing in that sense.In words of Robert Forman: From the pathology of a very few we have learned a great deal about the relationship of one side of the brain to the other, of two kinds of knowing, of information storage and retrieval, of impulse control etc. Indeed it is common practice to take data about a few unusual individuals and generalize it to the many.Here again we are studying the data of a few.But rather than the pathological, we will be studying people … who are not 'pathological' but unusually self-actualized [1; pp.363-364]. Another reason for the study of mystical experiences relates to the possibility of their being a more "fundamental" mode of experiencing, i.e. a mode that underlies our everyday (waking) consciousness.This is probably best seen in TON (also called "pure consciousness [event]"), a state of consciousness that is characterized by absolute absence of all phenomenological content.Just as biologists try to get a better understanding of a complex biological phenomenon (e.g. a living organism) by looking at its simplest form (e.g.E. Coli), so cognitive scientists might learn a lot about consciousness by looking at its most rudimentary representatives.And since TON, unlike our everyday consciousness, which is a "an enormously complex stew of thoughts, feelings, sensations, wants, snatches of song, pains, drives, daydreams", is characterized by a state of absolute stillness in which one "neither thinks nor perceives any mental or sensory content", and is therefore completely "perception-and thought-free" [1; pp.360-361], it seems to be the perfect candidate for the job.Moreover, mystics claim that mystical experiences lift an (experiential) veil and reveal that our ordinary (dual) way of perceiving things is actually secondary and derivative, i.e. superimposed on a more rudimentary (non-dual) experiential mode.If this were true -and it is a hypothesis that cannot be discarded on a priori grounds -it would mean that studying mystical experiences (in all their guises) provides crucial insights into "the ground, structure and dynamics of consciousness" [3; p.190]. However, not only mystical experiences as such, but also accompanying experiences (visions, trances, raptures etc.) can prove immensely valuable.Brian Lancaster has argued convincingly that the study of Jewish, Taoist and Buddhist mysticism might provide useful insights into the nature of everyday consciousness, as these traditions contain rich phenomenological descriptions and practical means of experientially accessing what is normally referred to as "preconscious" or "preattentive" cognitive processes: Research in cognitive neuroscience and depth psychology indicates that normal, mundane, consciousness arises from a complex stage of information processing during which diverse associations to the stimulus (be it sensory, object, a memory or a thought) are accessed preconsciously.... Put simply, mystical practice seems to entail a shift in the "leading edge" of consciousness such that elements previously obscured (preconscious) enter the clarity of consciousness [2; p.253]. Moreover, texts such as the Buddhist Abhidharma contain detailed phenomenological accounts of different conscious states and could therefore help us map the vast and rugged terrain of consciousness.Also, several authors (see e.g.[12,13]) have pointed out phenomenological similarities between states encountered on a mystical journey and in certain psychopathological states (esp.psychosis).Since mystics seem to be able to successfully master phenomena that overwhelm psychotics (hallucinations, ego-death etc.) a better understanding of processes involved in meditative/contemplative life can have not only theoretical (i.e.providing a clearer understanding of how such phenomena occur), but also practical implications (i.e.illuminating ways and methods of preventing and/or alleviating such occurrences in psychotic patients). PRACTICAL ASPECT This last remark brings us to the next domain in which phenomenological research could be coupled, and thereby enriched, with insights from mystical traditions.Several authors have stressed the need for a skilful application of improved practical methods that would enable a more rigorous and systematic approach to the study of human experience.Shear and Jevning, for example, point to a "significant asymmetry" present in current neurophysiological approaches to consciousness: For while their objective side employs sophisticated scientific methodologies, capable of isolating and evaluating variables completely outside of ken of ordinary sense perception, their subjective side typically uses mere everyday sorts of introspection, capable of isolating only ordinary internal phenomena such as sense perception, imagining and verbal thought … The need for systematic first-person methodologies here is thus starkly apparent" [3; p.109]. Similarly, Francisco Varela calls out for "a systematic exploration of the only link between mind and consciousness that seems both obvious and natural: the structure of human experience itself", and emphasizes the need for investigating "the concrete possibilities of a disciplined examination of experience" [14; p.330, 14; p.335]. Rich and multifarious meditative/contemplative practices developed in different mystical traditions seem to be particularly useful in this regard as they provide "a repository of contemplative and phenomenological expertise" for obtaining "precise and detailed firstperson accounts of experience".Contemplative mental training "cultivates a capacity for sustained, attentive awareness of the moment-to-moment flux of experience" and could thereby improve our overall understanding of the phenomenological domain [15; p.216, 15; pp.228-229].Note that the diversity of approaches in different traditions is by no means a hindrance, but an added value, for it is possible that a given tradition "may have gleaned some valuable knowledge or developed some practice that is not found elsewhere" [16; pp.498-499].In other words, although meditative/contemplative practices may induce phenomenologically identical experiences, they may differ significantly in their inner dynamics, i.e. in ways how they bring these experiences about, and each tradition could therefore offer unique insights into the structure and dynamics of consciousness. METAPHYSICAL DILEMMAS In addition to phenomenological research, mysticism could also engage in a fruitful exchange with cognitive science in addressing certain theoretical puzzles that have persistently bedevilled consciousness studies.Although several such problems exist, we will focus on two that are particularly pressing, i.e. the problem of the (unified) self and the hard problem of consciousness. THE PROBLEM OF THE (UNIFIED) SELF It has long been recognized that the notion of a unified, discrete, (semi-)autonomous entity called "the self" is all but unproblematic: On the one hand, even a cursory attention to experience shows us that our experience is always changing and, furthermore, is always dependent on a particular situation.… Yet most of us are convinced of our identities: we have a personality, memories and recollections, and plans and anticipations, which seem to come together in a coherent point of view, a center from which we survey the world, the ground on which we stand [17; p.59]. The currently predominant attitude in cognitive science seems to be that this elusive "ego"/"self" is a beneficial illusion, a useful construct with no independent existence.In words of Daniel Dennett: But the strangest and most wonderful constructions in the whole animal world are the amazing, intricate constructions made by the primate, Homo sapiens.Each normal individual of this species makes a self.Out of its brain it spins a web of words and deeds, and, like the other creatures, it does not have to know what it is doing; it just does it [18; p.416]. The idea is that mental life in its entirety consists of nothing but sub-personal mental processes, and that the sense of self is merely a useful superimposition on this array of unconscious events.To quote Dennett again: So far as I can see, however, every cognitivist theory currently defended or envisaged, functionalist or not, is a theory of the sub-personal level.It is not at all clear to me, indeed, how a psychological theory -as distinct from a philosophical theory -could fail to be a sub-personal theory [19; pp.153-154]. The main problem with this view, however, is that it seems to contradict our everyday experience: although there might be good scientific reasons to claim that there are no egos/selves (i.e. that all there exists are unconscious mental [neural?] processes), there seem to be even more persuasive phenomenological reasons that such entities do in fact exist.In other words, despite the fact that recent scientific studies seem to indicate that there are no unified selves, I still have the seemingly indelible feeling of being such a self my-self. Mystical traditions based on meditative/contemplative practices concur with scientific claims about the non-existence of unified selves, but do so on different grounds: they claim that "an untrained mind is inevitably deluded over the real nature of mind and consciousness" and that it takes strenuous mental discipline for the "elements previously obscured (preconscious) [to] enter the clarity of consciousness" [2; p.249, 2; p.253].In other words, mystical traditions claim that it is not only possible to think (reflect) on the non-existence of the self, but to actually make it a living experience.TON and LON are examples of conscious ego-less states, i.e. experiential states not referring/belonging to any self, and thus provide a phenomenological counterpart to scientific findings with potentially valuable insights into the true nature and origin of the (sense of) self (how and why it emerges, is it possible to live without it etc.).This is the reason why Varela, Thompson and Rosch felt the question of the self to be "the meeting ground" of (cognitive) science, philosophy and meditative/contemplative traditions: "[A]ll reflective traditions in human history … have challenged the naïve sense of the self" [17; p.59].And it is our contention that they have done so in different, yet mutually enlightening ways. THE HARD PROBLEM Another important area where mystical traditions could prove of value is the so-called "hard problem of consciousness".Chalmers explains: The really hard problem of consciousness is the problem of experience.When we think and perceive, there is a whir of informationprocessing, but there is also a subjective aspect.… This subjective aspect is experience.… It is undeniable that some organisms are subjects of experience.But the question of how it is that these systems are subjects of experience is perplexing.Why is it that when our cognitive systems engage in visual and auditory information-processing, we have visual or auditory experience: the quality of deep blue, the sensation of middle C [20; p.201]? At the bottom of the hard problem lies the notorious "explanatory gap" [21] between subjective, phenomenal and objective, physiological aspects of consciousness.Several solutions have been proposed throughout the years, but none of them met with unanimous support and approval.One of the suggestions maintains that the problem cannot be solved by a set of conceptual/theoretical "fixes", but demands a radical (existential) reorientation of our attitude towards it.In other words, the hard problem is not so much an intellectual, as it is an experiential-existential question: it is dependent on a specific (dualistic) manner of experiencing.To put it more precisely: From Descartes on, the guiding question in Western philosophy has been whether body and mind are one or two distinct substances (properties, levels of description etc.) and what the ontological relationship between them is.… We are suggesting that Descartes' conclusion that he was a thinking thing was the product of his question, and that question was a product of specific practicesthose of disembodied, unmindful [i.e.rational, abstract] reflection.… [But] theoretical reflection need not be mindless and disembodied … the mind-body relation or modality is not simply fixed and given but can be fundamentally changed [17; p.28]. In this view, the hard question is a practical question: by applying a set of skills/methods that can change our manner of experiencing it is possible to change ("answer away") the nature of the problem: the problem remains problematic only as long as it is experientially persuasive, as long as our way of experiencing presents it as a problem."[T]he mind-body issue is not simply a theoretical speculation but is originally a practical, lived experience, involving the mustering of one's whole mind and body.The theoretical is only a reflection on this lived experience" [17; p.28]. It was mentioned that mystical traditions, by cultivating a specific set of practical know-how, can induce radical and long-lasting transformations in our general way of being and experiencing.This is especially prominent in LON where subjectivity and objectivity are experienced non-dualistically, i.e. as "oneness in duality, duality in oneness".In this state, the mind and the body are said to be working as an integrated whole, so the hard problem does not even arise.This, however, introduces an intriguing turn into our discussion.We started off by fearing that mysticism might be something too obscure, too out-of-worldly (not rooted in solid, verifiable facts) to be of interest to cognitive science; now, in the face of the hard problem, it seems that it is cognitive science that might be too abstract, too out-of-worldly (not rooted in concrete, lived experience) and thus incapable of realizing that some of the problems it struggles with are (perhaps) the creations of (abstract, disembodied) premises it endorses, premises that are derivative on and secondary to the (non-dual, integrated) lived experience.Could it be that the greatest gift that mysticism could give to cognitive science is to save it from some of its own metaphysical spectres that haunt it -and thus help demystify it?REMARKS 1 A 1 more in-depth analysis of interrelations between cognitive science and mysticism is given in [8].	2016-07-05T18:43:41.079Z	2013-10-31T00:00:00.000	{ "year": 2013, "sha1": "f6a6a4df2a852bb8085cae01b274e7d0e77a5243", "oa_license": "CCBY", "oa_url": "https://doi.org/10.7906/indecs.11.4.4", "oa_status": "GOLD", "pdf_src": "Grobid", "pdf_hash": "9240d537bc0700dc0e5f0d5405e501d0f39a49ba", "s2fieldsofstudy": [ "Philosophy", "Psychology" ], "extfieldsofstudy": [ "Psychology" ] }
247031005	pes2o/s2orc	v3-fos-license	Applicability of rpoB Gene for PCR-RFlP based Discrimination of Bifidobacterial Species Isolated from human and Animal sources Bifidobacteria are widely used as probiotics for their application in the development of functional food and prophylactic therapy. This has necessitated the development of a molecular approach for the genera to be widely identified up to species and subspecies level. In the current study, PCR-RFLP of the partial RNA polymerase β-subunit ( rpoB ) gene fragment was evaluated for differential identification of Bifidobacterium species. The rpoB gene partial sequences of 575 bp were amplified from 93 previously identified isolates collected from various sources of human and animal origin along with 12 standard reference strains. The PCR amplified products were digested with three restriction endonucleases HhaI, HinfI and BanI separately. Dendrograms constructed from the patterns of HhaI, were found to be more discriminatory and successfully differentiated all the twelve species and also at sub-species level in between B. longum subsp. longum and B. longum subsp. infantis . However, B. adolescentis and B. pseudocatenulatum group clusters were not separated and represented by one group. The groups were further discriminated by HinfI restriction digestion. A separate combination thereof may be used for inferring the classification of bifidobacterial species targeted on rpoB PCR-RFLP analysis. To our knowledge, this work is the first report based on use of rpoB PCR-RFLP for discrimination of the isolates of genus Bifidobacterium and also provides insights into specific advantages of this method over hsp60 PCR-RFLP in differentiating B. longum subsp. longum and B. longum subsp. infantis. iNtRODUCtiON 5][6][7] Many species were recognized and classified under this genus after it was designated as an independent taxon by Orla-Jensen. 8,91][12] The classification is important in the food and pharmaceutical industries as some of developed products require specific strains.Thus, the immense contributions to human health 13 imposed for precise taxonomic identification and molecular based discrimination of genus Bifidobacterium. Traditional techniques for identification of bifidobacterial species have limited application due to low reproducible and indecisive results. 14he current understanding on the complexity and diversity within the genus Bifidobacterium has targeted the use of sophisticated molecular techniques for the differentiation of species and subspecies.Stackebrandt and Ludwig 15 reported that DNA hybridization >70% between two species possessed more than 97% 16S rRNA sequence similarity.However, 16S-based classification of genus Bifidobacterium is sometimes not discriminative enough to allow differentiation between certain species. 16The rate of evolutionary substitution in protein-encoding genes has higher importance and provides improved resolution than those in 16S rRNA genes. 17,18So, modern taxonomic approaches classified bifidobacterial phylogeny based on these highly conserved protein coding genes like tuf, atpD, recA, dnaK. 19In microbial ecological studies, the use of a single copy gene with increased resolution has highlighted enhanced phylogenetic relationships and diversity measurements at subspecies level in comparison to 16S rRNA gene. 20Among these conserved macromolecules, rpoB gene encoding the RNA polymerase β-subunit has come up into view as one of the major genes which is correlated with distinct phylogenetic analysis and identification of firmly related bacterial species. 21Adekambi et al. 22 proposed entire rpoB gene sequencing as a suitable approach to DNA-DNA hybridization technique for differentiation of bacterial genus and species.In recent decades, PCR-RFLP technique of rpoB gene has been implemented for species identification in genera viz.Legionella, 23 Mycobacterium, 24 Cronobacter, 25 Leptospira, 26 Lactic acid bacteria 27 etc.Kim et al. 28 reported that rpoB gene sequences are appropriate molecular markers for differentiation and determining the relationships of bifidobacterial species.Mainly this gene is responsible for rifaximin resistance in Bifidobacterium species. 29So far, no information is available with reference to the context of PCR-RFLP of rpoB gene in genus Bifidobacterium.Therefore, in the current study we have tried to build up a PCR-RFLP based identification strategy targeting partial rpoB gene sequence of bifidobacterial isolates, isolated from different sources of human and animal origin. Bifidobacterial isolates and standards used in the study Previously identified 93 bifidobacterial isolates comprised of 59 isolates from human sources (milk samples and faeces) and 34 isolates from animal origin (rumen samples, chicken faeces) as reported in our earlier study of hsp60 PCR-RFLP analysis 30 were included in this work.Standard strains representing 12 different Bifidobacterium species were procured from Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ, Germany) were included for validation, as listed in Table 1.Standard strains and our isolated 93 isolates were cultured and revived in M-58 medium whenever required, and stored supplemented with 10% glycerol (v/v) at -80 °C. Genomic DNA extraction Genomic DNA extraction was carried out from the cultures following the protocols of Brookman and Nicholson, 31 with some modifications.Bacterial strains were grown in Bifidobacterium medium (DSMZ-specification) and conditions as described in our previous study. 30After incubation, the bacterial pellets were collected by centrifugation at 10,000 rpm for15 min, and to it 800 µL of CTAB buffer (Cetyl trimethylammonium bromide) was added and mixed thoroughly.The suspension was kept for 1 h at 70°C and then mixed thoroughly.Then around 500 µL of chloroform: isoamyl alcohol (24:1) was added to the suspended cells and blended to form a white emulsion.In the subsequent stage centrifugation was done at 10,000 rpm for 20 min.The aqueous layer was moved and DNA was precipitated with addition of 300 µL of isopropanol and centrifuged at 10,000 rpm for 10 min).At that point 500 µL of 70% ethanol was added to the white pellet, mixed and afterward kept at 60°C for 10 min in an incubator. Air dried pellet was dissolved in 60 µL TE buffer [10 mmol Tris-HCl and 1mM EDTA (pH 8.0)].The quality of the isolated DNA was checked by 0.8% agarose gel electrophoresis and quantified by Nanodrop spectrophotometer (ND-1000; V3.5.2, Nanodrop Technology, Cambridge, UK).For working conditions, measurement was trailed by weakening DNA in Milli Q water.For working conditions, quantification was followed by diluting DNA in Milli Q water. Amplification of rpoB gene A m p l i f i c a t i o n o f r p o B g e n e f ra g m e n t w i t h p r i m e rs B i l o n -r p o B -F -5'-AGACCGACAGCTTCGATTGG-3'; Bilon-rpoB-R 5'-AACACGATGGCGGACTGCTT-3' (Sigma, Aldrich) were carried out as mentioned by Makino et al. 32 The composition of PCR mixture (50 μL) was 10X buffer (10 mM Tris-HCl (pH 8.3), 1.5 mM MgCl2 (Genei, Bangalore), 50 mM KCl, 200 μM each dNTP (Genei, Bangalore), template DNA (50 ng), primer (20 pmol each), Taq DNA polymerase (1 unit Sigma, Aldrich) and the volume was made up with sterile PCR grade water.The reaction was subjected to 35 cycles of amplification: initial denaturation 95°C for 8 min, denaturation 95°C for 45 sec, annealing 57°C for 30 sec, extension 72°C for 1 min and this was completed by final extension of 72°C for 10 min.Amplified PCR products were separated on 1.2% agarose gel (Sigma, Aldrich) containing 10 µg/mL ethidium bromide in 1X TBE electrophoresis buffer (10.8 g/L Trizma base (Sigma, Aldrich), 5.5 g/L Boric acid (SRL Ltd., Mumbai), 2.0 ml/L 0.5 M EDTA at 100 V. 100 bp DNA ladder (Promega, #G2101) was used as a molecular base size marker.Photographs of the amplified products were captured using Syngene G-Box, UK gel documentation system. PCR-RFLP and dendrogram construction Appropriate restriction enzymes were identified using different accessible nucleotide sequences from NCBI database through Cleaver software (http://cleaver.sourceforge.net/)so as to evaluate different restriction polymorphisms.HhaI (GCG↓C), HinfI (G↓ANTC) and BanI (G↓GYRCC) were selected as appropriate enzymes which were expected to give up distinct and sufficient fragment sizes to produce species specific restriction profiles.Restriction digestion was performed in a reaction volume of 20 µL having HhaI (1unit, Gene Mark; #E143) 0.05 µL; 10X Y buffer 2 µL; BSA 2 µL (0.1%); rpoB gene amplified PCR product (1 µg B. merycicum Rumen of cattle DSM-6492 approx).For HinfI and BanI the restriction digestion mixture was prepared having 0.1 µL HinfI (Takara, Shiga, Japan; #1238A) and BanI (1unit, Thermo Scientific; #ER1001), 2 µL 10X H buffer and the rpoB gene amplified PCR product (approximately 1 µg.).A total volume of 20 µL was adjusted with Milli Q water (sterile) for the above reactions and incubated for 2 h at 50°C in HhaI digestion and 37°C for HinfI and BanI digestion respectively.After the incubation period completion, the digested PCR products were separated using 3% agarose gel (1X TBE) for 4 h supplemented with ethidium bromide at a final concentration of 10 µg/mL and photographs of the patterns were visualized under gel documentation system.50 bp DNA ladder (Thermo Fisher Scientific; #SM0371) was used as a molecular marker for estimation of migration distance of individual bands of DNA fragments.RFLP restriction patterns generated from enzyme digestion of all the isolated bifidobacterial isolates and reference standard strains of DSMZ were analysed. To estimate the group specific positions of the isolates, and their similarity with standard reference strains, fingerprinting profiles generated with each enzyme (HhaI, HinfI and BanI) were evaluated with the restriction profiles of 12 reference cultures.Based on the RFLP fingerprints, binary matrixes were generated using Gel Quest software for each enzyme separately.The restriction fragments which are shorter than 40 bp were not included in this analysis.The data matrices were transferred to MEGA6 software 33 including the standards to estimate distances and the Unweighted Pair Group Method with Arithmetic averages (UPGMA) was implemented for cluster analysis. Amplification and PCR-RFLP of rpoB gene The DNA extracted from the standard cultures amplified with rpoB gene primers (Bilon-rpoB-F and Bilon-rpoB-R) produced amplicons of 575 bp size in all the isolates.The rpoB amplification of bifidobacterial isolates has been depicted in Fig. 1 animalis species in Group-III and 13 isolates were clustered under B. adolescentis in Group-V.Group-IV contained 12 isolates which did not cluster with any of the standard strain.It was observed that none of our isolated species grouped with B. thermophilum.Grouping of two to three species in a cluster was found in HinfI restriction digestion.To some extent HinfI restriction endonuclease does not provide sufficient discriminatory power for differentiation of species and subspecies.Most of the species can be identified by using a single enzyme, HhaI except for a few.But genetically it may be possible that close subspecies and species can be distinguished by rpoB PCR-RFLP with the combination of these two enzymes.BanI based fingerprinting patterns of our isolates with standard cultures yielded 5 different clusters (Fig. 2. DisCUssiON In the investigation, attempts have been made to demonstrate the efficacy of rpoB PCR-RFLP method in discrimination of closely related Bifidobacterium up to subspecies level.Baffoni et al. 34 reported that B. longum subsp.infantis and B. longum subsp.longum were not identified individually in other genes like hsp60 PCR-RFLP.But in rpoB PCR-RFLP with HhaI enzyme clear differentiation in between B. longum subsp.infantis and B. longum subsp.longum was observed (Fig. 2).In another study it was difficult to differentiate in between both the species of B. longum using 16S rRNA based species specific and group specific primers. 35akata et al. 36 39 In multigenic approach, 40 rpoB gene can be a gene of choice for providing detailed molecular analysis due to its easy accessibility and high resolving nucleotide-based molecular typing system. 28Ventura et al. 41 reported that single gene trees may not provide adequate phylogenetic relationships because of horizontal gene transfer events, inappropriate mutation rates and variable recombination rates.On the other hand, Vos et al. 42 reported that single-copy protein-encoding genes are of benefits in higher sampling efficiency, error-correction and easy survey of homologous recombination rate and mutations effects.But Ventura et al. 17 reported that multigene concatenation approach comprising of conserved orthologous proteins to be more discriminatory for differentiation of bifidobacterial species.However, rpoB gene sequences present improved phylogenetic resolution over the 16S rRNA gene. 21o, rpoB gene can be used as a sequence of choice in case of ambiguity when other gene are taken into account for identification. CONClUsiONs PCR-RFLP based on rpoB amplified partial gene product can be successfully applied to differentiate in between different species of genus Bifidobacterium as well as at sub-species level.The study also indicates that among the restriction enzymes HhaI, was found to be more discriminatory in differentiating different species of the genus as well as at sub species level.In addition to that whenever a single restriction enzyme fails to differentiate in between different species of the genus additional enzymes thereof may be used for inferring the classification of bifidobacterial species targeted on rpoB PCR-RFLP analysis. Fig. 1 . Fig. 1.Agarose gel photographs showing rpoB gene amplification of bifidobacterial isolates (M-100 bp DNA size ladder; Lane 1-19 respective isolates) and PCR-RFLP fingerprinting patterns of standard cultures with HhaI, HinfI and BanI are shown in Fig. 2, 3, 4 respectively.Dendrogram Analysis Three dendrograms in respect to each enzyme (HhaI, HinfI and BanI) were constructed on the basis of digested amplified rpoB PCR products.HhaI restriction enzyme based fingerprinting patterns of the isolates with standard cultures produced 10 different clusters (Fig. 5).The study conducted on the constructed dendrogram showed 13 isolates clustering with B. bifidum in group-I.Nine isolates showed similarity pattern with B. animalis in group-II.Three isolates grouped together with B. catenulatum in group-III whereas a total of 15 isolates were clustered under B. longum subsp.infantis in group-IV.Group-V contained 4 isolates which showed similarity outline with B. longum subsp.longum.The digestion performed with standard strains depicted the clear differentiation between B. longum subsp.longum and B. longum subsp.infantis.Eleven isolates with similar RFLP patterns clustered with B. breve in group-VI.Five isolates grouped to B. pullorum in group-VII.In group-IX both B. adolescentis and B. pseudocatenulatum were grouped together.Eight isolates clustered with B. merycicum in Group-X.Group-VIII contained same 12 isolates as obtained in a separate group in hsp60 which were identified as B. ruminantium in our previous study. 30None of our isolates grouped with the restriction patterns of B. boum and B. thermophilum included from the standard isolates list.Cluster analysis of HinfI based fingerprinting patterns yielded 5 different clusters (Fig. 6).Dendrogram showed a total of 29 isolates clustering with B. bifidum, B. catenulatum, B. pullorum, B. merycicum, B. boum and B. pseudocatenulatum in Group-I.Thirty isolates grouped mutually with B. breve, B. longum subsp.longum and B. longum subsp.infantis having same restriction patterns in Group-II.Nine isolates showed a same similarity pattern with B. 7 ) in the dendrogram.The study conducted on the constructed dendrogram showed 36 isolates clustered with B. catenulatum, B. adolescentis, B. Table 1 . Standard strains of the Bifidobacterium species included in this study Table 2 . Group wise estimated band patterns obtained from rpoB PCR-RFLP with respect to each restriction enzyme (approximate DNA fragment size ± 5bp) of the isolated strains 43e authors also presented hsp60 based results which provided clear differentiation between these closely related groups.This substantiated the results of Jian et al.43and the present study with respect to discrimination of B. pseudocatenulatum and B. catenulatum species in different groups.Matsuki et al. 35 observed that it is very difficult to differentiate B. adolescentis, B. catenulatum and B. pseudocatenulatum based on the conventional sugar fermentation pattern which puzzled the identification between B. catenulatum group and B. adolescentis.Kim et al. 28 reported 351 bp partial rpoB gene sequences to differentiate species of the genus Bifidobacterium which proved to be a suitable marker and based on this B. pseudocatenulatum and B. catenulatum were evidently discriminated.But the authors didn't state their discrimination with regard to B. longum subsp.longum and B. longum subsp.infantis.Targeted pyrosequencing method of rpoB amplicons (351 bp) revealed the low number of Bifidobacterium species with major incidence of B. longum and B. breve species in the antibiotictreated infants. 28reported that rpoB and hsp60 gene sequences based phylogenetic trees showed more discriminative patterns than the 16S rRNA gene tree.Their finding was especially discernible among related species such as B. pseudocatenulatum and B. catenulatum, B. saeculare, B. gallinarum and B. pullorum and B. pseudolongum subsp.pseudolongum and B. pseudolongum subsp.globosum.	2022-02-23T16:29:20.057Z	2022-02-21T00:00:00.000	{ "year": 2022, "sha1": "2310b695f25fa4e65862d092c4c54e8897d3154a", "oa_license": "CCBY", "oa_url": "https://microbiologyjournal.org/download/55176/", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "5a425cd6b61d1eb78805e2b79c2bc75535d6ebea", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [] }
229394363	pes2o/s2orc	v3-fos-license	Strategies to enhance graft performance in cord blood transplantation Umbilical cord blood (CB) serves as one of the alternative donor sources for hematopoietic stem cell transplantation (HSCT) in those patients lacking a matched related donor. However, the relatively high incidence of engraftment failure and/or delayed engraftment following cord blood transplantation (CBT) has been a major obstacle against its success. The usage of two CB units has made CBT possible in adults when a single CB unit with an adequate cell number is lacking, but it failed to demonstrate any beneficial effect in children and adolescents. In recent years, unmanipulated haploidentical graft has been replacing CB owing to its rapid availability along with low incidence of acute and chronic graft-versus-host disease due to the usage of post-transplant cyclophosphamide. Given this situation, overcoming the engraftment related issues is crucial in order for CBT to remain a realistic alternative of allogeneic HSCT. This review discusses the engraftment-related issues in CBT and recent advances to overcome those, highlighting the role of conditioning, ex vivo CB stem cell expansion, and cotransplantation of mesenchymal stem/stromal cells. INTRODUCTION The first successful cord blood transplantation (CBT) dates back to 1988 when a boy with Fanconi anemia received cord blood (CB) from his human leukocyte antigen (HLA)-identical sibling [1]. This success proved the notion that hematopoietic stem cells (HSCs) residing in human CB is enough to repopulate human bone marrow following an immunoablative or myeloablative conditioning. This pioneering achievement was followed by the first successful CBT using unrelated CB, delivered from the New York Blood Center, in an adult with chronic myelogenous leukemia in 1996 [2]. Since then, over 700,000 CB units have been donated for public use, and tens of thousands of banked CB have been used for alternative donor hematopoietic stem cell Improving graft performance in CBT transplantation (HSCT) in patients with various malignant and non-malignant disorders lacking a matched related or matched unrelated donor [3]. However, the limited cell dose in a CB unit has often led to delayed engraftment or failed primary engraftment resulting in high incidence of infectious complications and unacceptably high non-relapse mortality [4][5][6][7][8]. Nonetheless, due to the rapid availability and less stringent HLA matching requirements, CB has become a useful alternative HSC source for patients who need urgent allogeneic HSCT [9,10]. Typically, the rate of neutrophil engraftment after CBT is in the range of 70% to 90% with a significant delay compared to that after other types of HSCT [4,11]. In a previous report by Yoo et al. [12], the rates of neutrophil and platelet recovery were approximately 90% and 75%, respectively, with the overall survival rate of only 47.5% following CBT in Korean children and adolescents. Given that the delayed engraftment or primary engraftment failure has been the biggest hurdle to a successful CBT, strategies aiming at enhancing the graft performance are urgently needed in order for CBT to remain a realistic alternative of allogeneic HSCT. Herein the engraftment-related issues in CBT and recent advances in promoting engraftment after CBT have been discussed. Undoubtedly, selecting the best available CB unit that is well matched and has an adequate number of cells is the most important strategy for a successful CBT. In addition, avoiding CB units where preformed host donor specific anti-HLA antibodies exist is also a common practice to prevent graft failure or delayed engraftment [13][14][15]. With these strategies having been repeatedly stressed numerously, other strategies aiming at enhancing engraftment will be highlighted, which include modifying conditioning regimens, ex vivo CB stem cell expansion, and cotransplantation of third party mesenchymal stem/stromal cells (MSCs). This review mainly focuses on pediatric recipients who are more likely to benefit from CBT, in the hope of implicating some of them in adults as well. HAPLOIDENTICAL TRANSPLANTATION VERSUS CBT According to the European Group for Blood and Marrow Transplantation (EBMT) activity survey report in 2017, the number of unrelated CBT has been decreasing while that of haploidentical transplantation has been rapidly increasing in recent years [16]. The situation is similar in Korea Blood and Marrow Transplantation Nurses Society). These trends are largely attributed to recent advances in haploidentical transplantation using post-transplant cyclophosphamide, which has proved effective in preventing severe graftversus-host disease (GVHD) without the need for ex vivo T-cell depletion [17][18][19]. In addition, a haploidentical donor is usually rapidly available, and an adequate cell number is guaranteed in most cases. According to a recent retrospective study by the Eurocord-EBMT, the rates of overall survival, leukemia-free survival, and GVHD-free/relapse-free survival (GRFS) were significantly lower after CBT than those after haploidentical transplantation in adult patients with leukemia [20]. A more recent prospective randomized study by Sanz et al. [21] also revealed lower rates of disease-free survival and GRFS in adult CB recipients. These findings explain the general preference of physicians for a haploidentical transplantation over a CBT in adults when a matched related or matched unrelated donor is lacking. However, the situation is somewhat different in children. According to recent studies including a meta-analysis, the outcome measures for the two transplant modalities for childhood leukemia were comparable [22,23]. Given that CBT has advantages such as a low rate of chronic GVHD and a greater graft-versus-leukemia/graft-versus-malignancy (GVL/GVM) effect compared to other transplant modalities [24,25], establishing effective strategies to overcome its current limitations is expected to revitalize the use of CB especially for pediatric patients. CB ENGRAFTMENT AND SURVIVAL IN PEDIATRIC MALIGNANCY In a recent Japanese nationwide retrospective study comparing the donor sources in childhood acute leukemia, both neutrophil and platelet recovery after unrelated CBT were significantly delayed compared to those after unrelated bone marrow transplantation (BMT) [26]. (approximately 95%). However, CBT, whether single or double, showed an evident delay in platelet engraftment, and a substantial proportion of patients failed to achieve it. The inferiority of double CBT in that study is not very surprising since an earlier study already revealed no beneficial effect of double CBT in children and adolescents [28]. As noted in these studies, both neutrophil and platelet engraftment are delayed after CBT for childhood malignancy; the primary graft failure rate appears to be decreasing, while a substantial proportion of patients still fail to achieve platelet engraftment. Nevertheless, overall and leukemia-free survivals after pediatric CBT are generally comparable to those after matched unrelated donor transplantations, and it would be desirable to avoid double CBT unless the cell dose of a unit is unacceptable for transplantation. In patients with malignant diseases, there are many factors affecting the survival rates other than engraftment (e.g., disease status at transplantation, aggressiveness of the underlying disease, comorbid conditions). Therefore, the effectiveness of engraftment promoting strategies to improve survival rates is relatively uncertain, especially considering a relatively lower rate of primary engraftment failure in recent years than that in the past. Nonetheless, there are substantial evi-dences regarding the positive impact of total body irradiation (TBI) on the outcome of CBT in both children and adults. Nakasone et al. [29] showed that the rate of neutrophil engraftment was almost 20% higher in adult CB recipients following TBI-based myeloablative conditioning compared to that without TBI. Eapen et al. [30] compared an antithymocyte globulin (ATG)-free TBI-based regimen, named TCF (12 Gy TBI, cyclophosphamide 120 mg/m 2 , fludarabine 75 mg/m 2 ), with other regimens for childhood acute leukemia. The showed that TCF regimen, whether as part of a clinical trial or not, resulted in superior survival as compared to other non-trial chemotherapy-or TBI-based regimens, suggesting that a specific conditioning regimen could affect the outcome of CBT for childhood malignancy [30]. However, they failed to prove its positive impact on engraftment. In our institution, we have been using an ATG-free modified TCF regimen, containing 10 Gy TBI, for childhood acute leukemia over a decade. All patients were engrafted successfully using this regimen; notably, their event-free and overall survival reached to almost 80% (unpublished data). The main reason for using ATG in conditioning is in vivo depletion of host T-cells in order to prevent graft rejection. However, due to its long half-life and highly variable exposure between individuals caused by variables influencing the pharmacokinetic profiles, it also affects graft T-cells to a variable degree and can have a negative impact on engraftment [31]. Although ATG has been associated with a lower incidence of GVHD, it also decreases GVL/GVM effect and delays immune reconstitution [32,33]. Therefore, reducing its dose or even omitting it may be a reasonable choice given that an adequate cell dose and fairly good HLA matching are ensured. CB ENGRAFTMENT AND SURVIVAL IN NON-MALIGNANT DISORDERS In patients with non-malignant disorders, engraftment issues have been more problematic and the outcomes of CBT in those patients have largely been dependent on successful engraftment [34][35][36][37]. Unlike patients with malignant diseases who unexceptionally receive repetitive cytotoxic and immunosuppressive chemotherapy prior to HSCT, most of those with non-malignant disorders are not exposed to such treatments before transplantation. Therefore, they are capable of exerting relatively stronger alloimmune response against the graft despite a conditioning using similar intensity. This is also true for patients with severe aplastic anemia (SAA) whose marrow cellularity has already decreased, thus requiring a Improving graft performance in CBT less intensive conditioning regimen. As mentioned earlier, the first successful CBT was performed using CB from a matched related sibling donor [1]. However, a matched related CB or a fully matched unrelated CB unit is hardly available in clinical settings. Therefore, CB is mostly unrelated and partially mismatched in real practice. Nevertheless, matched related CBT has highly been successful in treating acquired and inherited bone marrow failure syndromes (BMFS) [35,38]. According to a study by the Eurocord and EBMT including 20 patients with acquired BMFS and 97 with inherited BMFS who were treated with matched related CBT, the cumulative incidence of neutrophil recovery by day 60 was 84% for acquired BMFS and 90% for inherited BMFS resulting in > 80% long-term survival rates [38]. In contrast, an earlier report by the Eurocord-Netcord and EBMT on the unrelated CBT in Fanconi anemia showed a disappointing outcome [34]. The neutrophil recovery rate by day 60 was only 60%, and the 3-year overall survival was 40%. In this study, patients who received a nucleated cell dose greater than 4.9 × 10 7 per kg of recipients' body weight showed a superior survival rate to those with a lower cell dose, which suggested that a higher cell dose is needed for the non-malignant disorders than that required in malignant diseases (2.5-3 × 10 7 /kg). In a subsequent report in 2011, the Eurocord and EBMT compared the related and unrelated CBT for hereditary BMFS except for Fanconi anemia [35]. The engraftment rate of unrelated CB was only 55%, and there was substantial difference in 3-year survival between related and unrelated CBT (95% vs. 61%), again proving the necessity of a higher cell dose in unrelated CBT for these non-malignant disorders. In a concurrent study by the Eurocord and EBMT reported on the unrelated CBT in SAA including mainly pediatric patients, the rates of neutrophil and platelet engraftment were only 51% and 37%, respectively, resulting in very poor overall survival of 38% [36]. According to this study, infection (38%) and poor graft performance (32%) were the predominant causes of death followed by GVHD (7%). In 2014, the Korean CB Registry data on the unrelated CBT for pediatric non-malignant disorders was published [37]. The neutrophil engraftment rate was 85% in patients with acquired SAA and only 37% in patients with inherited BMFS, leading to the overall survival of 54% in the former and 37% in the latter. Because of consistent poor outcomes, CB has not been considered a good stem cell source for treating non-malignant disorders. However, a recent Japanese study showed that a certain conditioning regimen was capable of produc-ing an excellent outcome of CBT in pediatric patients with SAA [39]. They omitted ATG and used low-dose TBI (2 Gy), fludarabine (125 to 175 mg/m 2 ), and either cyclophosphamide (50 to 100 mg/kg) or melphalan (80 to 120 mg/m 2 ) for the conditioning before transplantation. This combination led to 100% engraftment and survival in contrast to the poor engraftment and survival with other regimens. More recently, a French nationwide prospective study on the unrelated CBT in patients with idiopathic refractory SAA was published [40]. The eligibility criteria required one or two CB units, containing separately or together more than 4 × 10 7 frozen nu- EX VIVO EXPANSION OF CB STEM CELLS Ex vivo expansion of CB stem cells is one of the most attractive strategies to enhance graft performance. A number of methods for ex vivo expansion, such as Notch-mediated expansion [43][44][45], MSC coculture [46,47], nicotinamide [48][49][50], StemRegenin-1 [51,52], copper chelation [53,54], chrysin [55], and UM171 [56,57], have been explored and some of them have already been tested in clinical trials (Table 1). Ideally, expanded stem cells should maintain their quality by in-hibiting hematopoietic differentiation, have to be safe and feasible for clinical application, must result in enhanced engraftment, and ultimately should lead to an improved survival of the recipients. Notch family members are widespread by hematopoietic cells, and Notch signaling in combination with cytokines induces a robust increase in marrow repopulating ability of CB cells [43]. The first clinical trial with CB cells that have undergone Notch-mediated expansion was conducted in recipients of double CBT, and one of the two units was infused after ex vivo expansion [44]. The mean fold increase of nucleated cells was > 500 times and that of CD34+ cells was > 160 times, leading to a faster neutrophil engraftment compared to that in the historical controls (median reduction in time to neutrophil recovery, 10 days). This is the first report that demonstrated the rapid myeloid engraftment achieved using ex vivo expanded cells in human. The safety of Notch ligand-expanded progenitor cells was also demonstrated by a subsequent phase I study for patients with acute myeloid leukemia following intensive chemotherapy [45]. Nicotinamide is a small molecule which is known to inhibit the differentiation of hematopoietic stem and progenitor cells expanded by cytokine-induced ex vivo cultures and to enhance their functionality [48,49]. Recently, a phase I/II study assessing the safety and efficacy of CB grafts that were expanded ex vivo in the presence of nicotinamide and transplanted after myeloablative conditioning as a stand-alone graft was reported [50]. Copper chelation by tetraethylenepentamine in combination with various cytokines inhibits stem cell differentiation transiently during the expansion period [53]. A cohort-controlled study using these expanded cells was recently reported [54]. Only a small portion of a CB unit was used for expansion in this trial, and the larger fraction containing nucleated cells at least greater than 1 × 10 7 /kg served as the unmanipulated fraction. The study presented an encouraging outcome showing a better engraftment kinetics and a significantly better 100-day survival in recipients as compared to that in the historical controls. UM171, an HSC self-renewal agonist, has been shown to expand CB stem cells and enhance multi-lineage blood cell reconstitution in mice [56]. Very recently, a single-arm, phase I/II study of single UM171-expanded CBT was reported in patients with hematologic malignancies [57]. In this trial, no graft failure occurred in 22 patients who received single UM171-expanded CBT. The median time to neutrophil and platelet recovery was 18 and 42 days, respectively, without any unexpected adverse event. As noted by a number of clinical trials, ex vivo CB expansion strategies are obviously encouraging in terms of facilitating engraftment following CBT. However, their impact on survival outcomes is uncertain due to the lack of large-scale long-term follow-up study on these measures. Further optimization of these strategies is expected to produce better outcomes of CBT in the near future. (Fig. 1) [80]. Therefore, special precautions need to be taken while developing MSC-based cell therapies. COTRANSPLANTATION OF MSCS WITH CB Improving the quality of life as well as survival of recipients is the ultimate goal of HSCT. Thus, strategies aiming at controlling severe GVHD and preventing regimen-related toxicities and relapse are also as important as those aiming at facilitating engraftment in order to achieve the goal. Considering the safety and potential benefits of MSCs, the establishment of optimal treatment strategies with MSCs might pave the way for substantial improvement in the outcomes of not only CBT but also other types of allogeneic HSCT. CONCLUSION Although a minority of pediatric CB recipients with hematologic malignancy suffer from primary graft failure, their survival appears to be comparable to that of haploidentical or matched unrelated donor recipients and even to that of matched related donor recipients in recent years. However, a delay in both myeloid and platelet engraftment is not likely to be overcome by double CB, which requires effective strategies to enhance engraftment. Engraftment failure has been CONFLICTS OF INTEREST No potential conflict of interest relevant to this article was reported. Drafting the work or revising: KHY. ACKNOWLEDGMENTS Final approval of the manuscript: KHY.	2020-12-03T09:06:52.380Z	2020-11-26T00:00:00.000	{ "year": 2020, "sha1": "23076bbefa9d2627b04a12600ffcf3940383e276", "oa_license": "CCBYNC", "oa_url": "https://www.pfmjournal.org/upload/pdf/pfm-2020-00149.pdf", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "be3973ddb78a359bed80edc0f1752c7b83ce8682", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
222009649	pes2o/s2orc	v3-fos-license	Predictors of COVID-19 severity: a systematic review and meta-analysis [version 1; peer review: 2 approved] Background: The unpredictability of the progression of coronavirus disease 2019 (COVID-19) may be attributed to the low precision of the Open Peer Review Introduction The coronavirus disease 2019 (COVID-19) pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is a global crisis across health, economic, and educational dimensions 1,2 . The disease has spread rapidly, can cause severe illness, and is characterized by a high mortality rate in certain groups. Mortality is particularly high in the absence of proven effective standard management measures 3 . One of the problems with the management of this disease is the absence of standardized methods for diagnosis and the inability to estimate prognosis based on clinical features. Certain reports have shown that poor prognostic prediction has correlated with high mortality among patients with COVID-19 4,5 . Among patients with similar clinical characteristics and with similar treatment regiments, there may be a diversity in clinical outcomes 6 . Therefore, the development and use of an accurate predictor for COVID-19 prognosis will be beneficial for the clinical management of patients with COVID-19, and will help reduce the mortality rate. Successful implementation of such a prediction mechanism could have a large public health impact. Better understanding of clinical progression could also improve public health messaging, particularly as many individuals may consider COVID-19 to not be severe. Prognostic tools for the prediction of COVID-19 severity in patients have been in development since January 2020. At least nine studies proposed the use of prognostic tools for the prediction of COVID-19 severity [7][8][9][10][11][12][13][14][15] . However, a recent systematic review and critical appraisal study evaluated the accuracy of these tools using prediction model risk of bias assessment tool (PROBAST) and reported a high risk of bias 16 . The establishment of a prediction model for the estimation of disease prognosis may help health workers segregate patients according to prediction status. However, the high risk of bias in these prediction tools might lead to inaccurate prediction of COVID-19 severity. A comprehensive study of the identification of risk factors that might play a significant role in determining the severity of patients with COVID-19 is necessary. We performed a systematic review and meta-analysis to assess the risk factors associated with poor clinical outcomes among patients with COVID-19. To the best of our knowledge, this is the first meta-analysis to assess the comprehensive risk factors that might affect the severity of COVID-19 in patients. The results of our study might serve as preliminary data for the compilation or improvement of the scoring system in the prediction of COVID-19 severity. Study design We performed a systematic review and meta-analysis to evaluate potential risk factors that might influence the severity of COVID-19. These risk factors include comorbidities, clinical manifestations, and laboratory findings. Accordingly, we searched the relevant studies from major scientific websites and databases to collect the data of interest, and determined the association and effect estimates by calculating the combined odds ratio (OR) and 95% confidence intervals (95% CI). The protocols for the systematic review and meta-analysis were similar to those used in previous studies 17-23 , as well as to those recommended by Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) 24 . Eligibility criteria Studies were included in this review if they met the following inclusion criteria: (1) assessed the clinical manifestations and laboratory findings of patients with mild to severe COVID-19; (2) provided adequate data for the calculation of OR and 95% CI. Review articles, articles with non-standard data presentation, and duplicate publications were excluded. Search strategy and data extraction Major scientific databases (PubMed, Embase, Cochrane, and Web of Science) were searched for articles as of April 5, 2020. A comprehensive initial search was performed to identify the potential predictors, and a final search was performed to identify the relevant papers that could be included in the meta-analysis. We used the keywords adapted from medical subject headings: ["COVID-19" or "Coronavirus disease-19" or "SARS-CoV-2"] and ["mild" or "severe" or "prognosis" or "clinical outcome"] and ["clinical manifestation" or "morbidity" or "laboratory findings"]. Only studies written in English were included. If a duplicate publication was found, the article with the larger sample size was included. We also searched for relevant studies from the reference lists in the articles. During data extraction, the following information of interest was extracted: (1) first author name; (2) publication year; (3) sample size of mild and severe cases, (4) clinical manifestations, (5) morbidities, and (6) laboratory findings. Data extraction was performed by two independent investigators (JKF and MI) using a pilot form. Assessment of the methodological quality Before inclusion in the meta-analysis, the methodological quality of the articles was assessed using the New Castle-Ottawa scale (NOS). NOS scores range from 0 to 9 and consider three items: selection of patients (4 points), comparability of the groups (2 points), and ascertainment of exposure (3 points). Each study was interpreted to be of low quality (for scores ≤ 4), moderate quality (for scores between 5-6), or high quality (for scores ≥ 7) 25 . Articles with moderate to high quality were included in the analysis. The study assessment was conducted by two independent investigators (MI and YP) using a pilot form. The discrepancies between the findings of the two investigators were solved by consulting with another investigator (JKF). Statistical analysis The significant risk factors that might govern the severity of COVID-19 were determined by the calculation of a pooled OR and 95% CI. The significance of the pooled ORs was determined using the Z test (p<0.05 was considered statistically significant). Prior to identification of the significant risk factors, data were evaluated for heterogeneity and potential publication bias. The heterogeneity among included studies was evaluated using the Q test. If heterogeneity existed (p<0.10), a random effect model was adopted; if not, a fixed effect model was adopted. Egger's test and a funnel plot were used to assess the reporting or publication bias (p<0.05 was considered statistically significant). Furthermore, we performed a moderator analysis to identify the independent predictors of poor clinical outcomes among patients with COVID-19. The data were analyzed using Review Manager version 5.3 (Revman Cochrane, London, UK). To prevent analytical errors, statistical analysis was performed by two authors (JKF and MI). The cumulative calculation was presented in a forest plot. Eligible studies Our searches yielded 6,209 potentially relevant studies, of which 6,170 studies were excluded after assessment of the titles and abstracts. Subsequently, further review of the complete texts was performed for 39 potential studies. In the full text review, we excluded 20 studies because they were reviews articles (n = 9), inadequacy of data for the calculation of OR and 95% CI (n = 7), and poor quality (n = 4). Eventually, 19 papers were included in our meta-analysis 26-42 The paper selection process adopted in our study is summarized in Figure 1, and the characteristics of studies included in our analysis are outlined in Table 1. Risk factors of severe COVID-19 We found that eight comorbidities, 19 clinical manifestations, and 35 laboratory parameters were available for the metaanalysis (Table 2 and Table 3 (Figure 2A-D). Source of heterogeneity Heterogeneity was detected in the data of chronic kidney disease, cerebrovascular disease, cardiovascular disease, diabetes mellitus, hypertension, and malignancy among the comorbid factors analyzed. Therefore, we used the random effect model to analyze the data. The fixed effect model was used to analyze the data on chronic liver disease and chronic respiratory disease, as there was no evidence of heterogeneity. For clinical manifestations, the data on fever, cough, sore throat, dyspnea, diarrhea, anorexia, fatigue, temperature >38°C, respiratory rate, and diastolic blood pressure were analyzed using the random effect model while the rest of clinical manifestation data were analyzed using the fixed effect model. Among laboratory parameters, evidence of heterogeneity was found in count of WBC, neutrophil, monocyte, lymphocyte, platelet, CD4, and CD8; the presence of lymphocytopenia and anemia; the levels of AST, ALT, total bilirubin, albumin, aPTT, PTT, serum creatinine, BUN, Hs-Troponin I, creatine kinase, IL-6, Hs-CRP, glucose, D-dimer, sodium, potassium, lactate dehydrogenase, and procalcitonin; elevated CRP; and ESR. Accordingly, the data were analyzed using the random effect model. The data for the remaining parameters were analyzed using the fixed effect model. Potential publication bias We used Egger's test to assess the potential publication bias. Our cumulative calculation revealed that reporting or publication bias (p<0.05) existed with respect to chronic liver disease, expectoration, myalgia, abdominal pain, heart rate, leukocytosis, elevated ESR, and elevated IL-6 levels. Discussion Our data suggest that comorbidities, such as chronic respiratory disease, cardiovascular disease, diabetes, and hypertension, were associated with a higher risk of severe COVID-19, among which, hypertension was the strongest risk factor. These results are consistent with those of previous meta-analyses 43,44 that indicated that chronic respiratory disease, cardiovascular disease, diabetes, and hypertension are significantly associated with higher COVID-19 mortality. Hypertension and diabetes are also associated with higher mortality among patients with dengue fever, West Nile virus infection, Zika virus infection, and yellow fever 45 . To date, no study has reported details of the primary mechanism underlying the association between severe COVID-19 and comorbid factors. However, immune responses might be the most crucial factor underlying this association. Patients with comorbidities such as cardiovascular disease, chronic respiratory disease, hypertension, and diabetes were observed to have a lower immunity status than healthy individuals 46-48 . Since COVID-19 primarily affects the respiratory tract 49 , patients with chronic respiratory diseases might be at a higher risk of contracting severe COVID-19. In addition, endothelial dysfunction might also play a pivotal role 50 . COVID-19 is a novel disease, and the immune response of this disease is not completely understood. Our data suggest that elevated leukocyte and neutrophil levels and reduced lymphocyte levels are associated with severe COVID-19. In other viral infections, such as influenza, elevated leukocyte and neutrophil levels serve as important predictors of disease severity 51 . The role of leukocytes in the pathogenesis of COVID-19 is conflicting. In most cases, viral infections have been observed to cause leukopenia 52 . Furthermore, a study also reported that leukopenia was observed at a significantly higher frequency among COVID-19 patients than among non-COVID-19 patients 53 . However, in our present study, we did not compare COVID-19 and non-COVID-19 patients. The major factor that seemed to affect our findings was the occurrence of cytokine storm in patients. In COVID-19, there is an immune system overreaction, which results in a cytokine storm. In this condition, leukocytes might be over-activated, which might lead to the release of high levels of cytokines 54 . Consistent with our data, a study has confirmed that cytokine storm is significantly associated with severe COVID-19 55 . The theory underlying the role of neutrophils in COVID-19, as reported in our study, remains unclear. The speculations might be attributed to the involvement of neutrophil extracellular traps (NETs). While no study has assessed the precise role of NETs in COVID-19 pathogenesis, certain researchers speculate that SARS-CoV-2 might stimulate neutrophils to produce NETs, similar to several other viral pathogens 56 . Furthermore, this might lead to neutrophil infiltration in pulmonary capillaries, organ damage, and the development of acute respiratory distress syndrome 57 . Low lymphocyte levels were observed in patients with severe COVID-19 compared with those with mild COVID-19. In the context of the immunological mechanism, our results might be contradictory. Lymphocyte subsets are known to play an important role in the action against bacterial, viral, fungal, and parasitic infections 58 ; therefore, the levels of circulating lymphocytes should increase. The immunological response in COVID-19 is unique and remains unclear. However, certain propositions might help describe our findings. First, coronaviruses infect human cells through ACE2 receptors 59 . Since ACE2 receptors are also expressed by lymphocytes 60 , the coronaviruses may enter lymphocytes and induce apoptosis. Second, the feedback mechanism between pro-inflammatory cytokines (such as IL-6) and lymphocytes might also explain our results. A study revealed that elevation in the levels of pro-inflammatory cytokines correlated with reduction in the levels of lymphocytes 61 . Moreover, our findings also confirmed the significant elevation in the levels of IL-6. Third, ACE2 receptors are expressed by cells from various organs, including the thymus and spleen 62 . As coronaviruses infect human cells through the ACE2 receptors, the spleen and thymus might also be damaged in patients with COVID-19, which would lead to lower levels of lymphocyte production. Fourth, lymphocyte proliferation requires a balanced metabolism, and metabolic disorders such as hyperlactic acidemia have been reported to disturb lymphocyte proliferation 63 . Hyperlactic acidemia has been observed in patients with severe COVID-19 64 . The studies included in this systematic review also suggest that the levels of D-dimer were significantly higher in patients with severe COVID-19. Coagulation in patients with COVID-19 has been a major concern, and the lack of reliable data and meta-analyses prevents a holistic comparison. Certain infectious diseases that cause abnormal coagulation have been associated with poor clinical outcomes 65 . The theory behind this mechanism is not understood clearly. It is widely known that ACE2 receptors are important for the infection of host cells by SARS-CoV-2, and ACE2 receptors are expressed in various cells in the human body, including endothelial cells 66 . Consequently, a massive inflammatory reaction may occur in endothelial cells owing to SARS-CoV-2 infection 67 , which may lead to increased coagulation, disseminated intravascular coagulation 68 , and increased fibrin degradation 69 . High fibrin degradation leads to elevated levels of fibrinogen and D-dimer 70 , which might also explain the occurrence of venous thromboembolism in critical patients of COVID-19 71 . In addition, a study with a short follow-up period also reported the existence of a dynamic correlation between the D-dimer levels and the severity of COVID-19 72 . Furthermore, pulmonary embolism and deep vein thrombosis were also observed in patients with severe COVID-19 73,74 , which suggests that D-dimer might play a prominent role in governing the severity of COVID-19 patients. We also observed that inflammatory markers, including elevated levels of CRP, ESR, and IL-6, were found both in patients with severe and mild COVID-19, with a significant increase detected in patients with severe COVID-19. Other variables associated with adverse outcomes, such as ferritin, lactate dehydrogenase, and procalcitonin levels, were found to be elevated predominantly in patients with severe COVID-19. Our findings were consistent with those of a previous meta-analysis 75 , and indicated that high levels of CRP, lactate dehydrogenase, and ESR were associated with adverse outcomes in COVID-19. Another meta-analysis had also confirmed that elevated levels of IL-6 were observed in patients with COVID-19 who exhibited poor clinical outcomes 76 . Therefore, the levels of CRP, ESR, IL-6, ferritin, procalcitonin, and lactate dehydrogenase can serve as potential markers for the evaluation of COVID-19 prognosis. The high mortality rate and treatment failure in patients with COVID-19 can be attributed to the fact that COVID-19 affects multiple organs, including the lung, heart, kidney, and liver 77 . Our data suggest that elevated levels of urea and creatinine, and not chronic kidney disease, were associated with severe COVID-19, which indicates that acute inflammation might be caused by SARS-CoV-2 infection. Previous meta-analyses have also reported findings consistent with our results 78,79 . Moreover, anatomical studies have reported significant renal inflammation in patients with severe COVID-19 75,80,81 . There might be two mechanisms by which SARS-CoV-2 induces renal inflammation. First, SARS-CoV-2 might directly infect renal tubular epithelial cells and podocytes through ACE2 receptors, which facilitates the targeted infection of certain cells by the virus. Consequently, acute tubular necrosis, podocytopathy, microangiopathy, and collapsing glomerulopathy might occur owing to the massive inflammation in renal tubular epithelial cells and podocytes 82,83 . Second, the binding between SARS-CoV-2 and ACE2 receptors might activate angiotensin II and induce cytokine production, which may lead to hypercoagulopathy and microangiopathy, and eventually cause renal hypoxia 84,85 . Conversely, with respect to liver function, we observed that the levels of liver enzymes were higher in patients with severe COVID-19. Previous studies in this context have elucidated that ACE2 receptors are highly expressed in bile duct cells; therefore, infection of these cells by coronaviruses might lead to abnormalities in the levels of liver enzymes 86 . However, a recent anatomical study on liver biopsy specimens from patients with severe COVID-19 revealed that moderate microvascular steatosis and mild lobular and portal activities were observed 87 . These data suggest that it cannot be determined clearly whether the elevated levels of liver enzymes in patients with severe COVID-19 are caused by direct infection or by drug-induced liver injury. Therefore, further studies are required to elucidate the precise mechanism underlying the elevation of liver enzymes levels in patients with severe COVID-19. 75 . In our study, we included all comorbidities, clinical manifestations, and laboratory characteristics. Additionally, compared to previous studies, this study has a larger sample size; the data on 1,934 patients with mild and 1,644 patients with severe COVID-19 treated across 19 hospitals were retrieved. However, this study also has certain limitations. Certain crucial factors that might play an important role in the pathogenesis of COVID-19, including secondary infection, treatment, and immunological status were not controlled for. Our current findings should be interpreted with caution because the majority of studies included were cross-sectional, and the samples corresponding to the data analyzed originated only in China. Longitudinal studies may reveal more long-term impacts of SARS-CoV-2 infection 92 . Conclusion COVID-19 is an emergent infectious disease, and the major problem associated with it is the unknown pattern of disease development. We identified 34 factors that are associated with severe COVID-19. This might improve our understanding of COVID-19 progression and provide baseline data to compile or improve the prediction models for the estimation of COVID-19 prognosis. Data availability Underlying data All data underlying the results are available as part of the article and no additional source data are required.	2020-09-10T10:10:59.582Z	2020-09-09T00:00:00.000	{ "year": 2020, "sha1": "ea11ba394511fc13613adadc672220d67896c6dc", "oa_license": "CCBY", "oa_url": "https://f1000research.com/articles/9-1107/v1/pdf", "oa_status": "GOLD", "pdf_src": "MergedPDFExtraction", "pdf_hash": "84a11b2c03cae7294efc560297c45dd00e2bcd3b", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
11189601	pes2o/s2orc	v3-fos-license	Elevation of plasma basic fibroblast growth factor after nocturnal hypoxic events in patients with obstructive sleep apnea syndrome Obstructive sleep apnea syndrome (OSAS) is associated with recurrent nocturnal hypoxia during sleep; this hypoxia has been implicated in the pathogenesis of cardiovascular complication. However, a useful soluble factor that is sensitively correlated with OSAS severity for the diagnosis remains unidentified. We hypothesized that systemic levels of basic fibroblast growth factor (bFGF), a hypoxia-induced cytokine, were affected by nocturnal hypoxemia in OSAS patients, and we assessed whether the degree of change in the plasma bFGF concentrations before and after nocturnal hypoxia is correlated with the severity of OSAS. Thirty subjects who had suspected OSAS and had been investigated by nocturnal polysomnography (PSG) were enrolled. Plasma bFGF and vascular endothelial growth factor (VEGF) concentrations the night before PSG and the next morning were measured by sandwich enzyme-linked immunosorbent assay. Correlations between the changes in these factors and hypoxia-associated parameters for OSAS severity were analyzed. Patients with OSAS had significantly elevated levels of plasma bFGF but not VEGF and hemoglobin after rising. The degree of change in bFGF concentrations after nocturnal apnea episodes was significantly correlated with diagnostic parameters for OSAS severity. The change in plasma bFGF levels is associated with the degree of hypoxic state in OSAS patients, implying that bFGF might be a useful soluble factor for evaluating OSAS severity. Background Obstructive sleep apnea syndrome (OSAS), which is a common public health issue affecting as much as 4% of the adult population, is associated with recurrent hypoxia during sleep (Young et al. 1993;Strollo and Robers 1996). Patients with OSAS are exposed to decreased oxygen saturation by repeated episodes of apnea and hypopnea. The oxygen desaturation events result in the development or exacerbation of cardiovascular, cerebrovascular, and metabolic diseases (Peppard et al. 2000;Bradley and Floras 2009;Nishibayashi et al. 2008;Vgontzas et al. 2000;Marin et al. 2005). Hypoxia and the subsequent tissue ischemia are major pathophysiological regulators of angiogenesis (Lavie and Lavie 2009). Increased angiogenesis rates and the elevation of hemoglobin (Hb) levels in response to hypoxia are part of an adaptive response aimed at achieving increased delivery of oxygen and nutrients to the tissues (Bunn and Poyton 1996;Bicknell and Harris 2004). On a per-cell basis in hypoxic tissues, the transcriptional response of mammalian cells to hypoxia is largely mediated by hypoxia-inducible factor-1 (HIF-1) (Wang et al. 1995). HIF-1 is a basic helix-loop-helix transcription factor composed of a HIF-1β and a HIF-1α, the production of which is strongly upregulated under hypoxic conditions (Acker and Plate 2003). Vascular endothelial growth factor (VEGF) is a soluble factor that regulates multiple functions of endothelial cells (Forsythe et al. 1996). Expression of VEGF is rapidly induced by HIF-1 -mediated transcriptional activation in response to low oxygen levels in vivo as well as in vitro (Acker and Plate 2003;Forsythe et al. 1996). Previous studies have reported that levels of VEGF in the peripheral blood are elevated in patients with OSAS (Imagawa et al. 2001;Schulz et al. 2002;Lavie et al. 2002;Gozal et al. 2002). Basic fibroblast growth factor (bFGF) is also known to be strongly upregulated by HIF-1 in response to hypoxic conditions (Calvani et al. 2006). It is recently reported that bFGF-dependent induction of HIF-1α expression results in the formation of an HIF-1α-bFGF positive feedback amplification pathway under hypoxic conditions in human umbilical vascular endothelial cells and rat cardiac microvascular endothelial cells in vitro, although VEGF, unlike bFGF, does not enhance HIF-1 gene expression (Calvani et al. 2006;Li et al. 2002). The existence of an HIF-1α-bFGF autocrine loop suggests that the systemic level of bFGF, but not VEGF, is likely to be strongly correlated with the degree of hypoxic state in OSAS patients. However, thus far no report has demonstrated the relationship between the severity of OSAS and levels of bFGF in patients with the disease. Furthermore, a definitive diagnostic marker that is correlated with the severity of OSAS remains to be identified (Arnardottir et al. 2009). Here, we measured the plasma concentrations of bFGF, VEGF, and Hb in subjects at the night before sleep and in the next morning immediately after rising, and we verified the relationship between the degree of change in the plasma bFGF concentrations before and after nocturnally recurrent apnea episodes and the severity of OSAS. Characteristics of subjects The characteristics and severity of OSAS in all subjects are summarized in Table 1. The mean ages of the two groups were similar, even though body mass index (BMI) significantly increased in the OSAS group. Apnea-hypopnea index (AHI), oxygen saturation with pulse oximetry (SpO 2 ) < 90% (% total sleep time; TST), 4% oxygen desaturation index (ODI), median and minimal SpO 2 , as evaluation parameters of OSAS severity determined by polysomnography (PSG), significantly differed between the groups. All of these values were markedly higher or lower in the OSAS group than in the control group. The concentrations of VEGF and bFGF in plasma samples at night and the next morning (at 8 p.m. and at 6 a.m., respectively) were measured by sandwich enzyme-linked immunosorbent assay (ELISA) ( Table 1). Unexpectedly, the plasma levels of VEGF and Hb did not differ between the groups, regardless of the morning and the night. However, the plasma bFGF levels in OSAS patients were significantly higher than those in control subjects only in the morning but not at night. Changes in VEGF, bFGF, and Hb after nocturnal apnea events To assess changes in the concentrations of plasma VEGF, bFGF, and Hb after exposure to nocturnal hypoxia, we quantified these soluble factors at the night before PSG and the next morning after PSG in control and OSAS groups ( Figure 1). In control subjects, the concentrations of all the factors were similar before and after sleep. In contrast, the level of plasma bFGF in the OSAS group was significantly greater after nocturnal hypoxic events than before, although there were no significant changes in VEGF or Hb (Figure 1). The elevation of plasma bFGF concentrations in OSAS patients was also confirmed in regression plots of bFGF values measured before the nigh PSG and the morning after PSG against AHI ( Figure 2). These results imply that plasma levels of bFGF, unlike those of VEGF and Hb, were affected by frequent hypoxic exposure through the night in OSAS patients. Correlation between overnight change in VEGF and bFGF and OSAS severity To evaluate the usefulness of the changes in plasma VEGF and bFGF levels to predict OSAS severity, we next investigated the correlation between the overnight changes in VEGF and bFGF, which were defined as All participants were categorized into two groups by their OSAS severity (control, AHI ≤ 15; and OSAS, 15 < AHI). Concentrations of plasma VEGF, bFGF, and Hb the night before PSG and the next morning were measured (night and morning, respectively). Data are presented as mean values ± standard deviations. Statistical analysis for evaluation parameters among the categories was performed by the Student t-test. AHI apnea-hypopnea index, bFGF basic fibroblast growth factor, BMI body mass index, SpO 2 < 90% (% TST) percentage of total sleep time with oxygen saturation below 90%, Hb hemoglobin, ODI oxygen desaturation index, SpO 2 oxygen saturation with pulse oximetry, VEGF vascular endothelial growth factor. ΔVEGF and ΔbFGF, respectively, and the hypoxiaassociated parameters for severity of OSAS. Statistical analysis using the Pearson's correlation coefficient revealed that ΔbFGF, but not ΔVEGF, was positively and significantly correlated with AHI, SpO 2 < 90% (% TST) and 4% ODI, but neither age nor BMI (Table 2). ΔbFGF was also negatively and significantly correlated with median and minimal SpO 2 . These results suggest that the change in plasma bFGF might closely reflect the degree of hypoxia during sleep, i.e., the severity of OSAS. Discussion Production of the cytokines VEGF and bFGF is well known to be induced by hypoxia in a HIF-1-dependent manner (Forsythe et al. 1996;Calvani et al. 2006). We therefore hypothesized that plasma levels of these cytokines would be affected by nocturnal hypoxemia in OSAS patients. We actually indicate that levels of plasma bFGF, but not VEGF or Hb, are greater in the morning than at the night before in patients with OSAS. Notably, the change in bFGF, but not VEGF, concentration after nocturnal apnea episodes was correlated with AHI, SpO 2 < 90% (% TST), 4% ODI, median and minimal SpO 2 . Although we could not demonstrate the role of bFGF in OSAS patients in this study, and could not completely exclude the potential cofounders that might have distributed inequally between control and study groups, these results raised the possibility that bFGF might be useful soluble factor for evaluating the severity of OSAS. Previous studies have reported that the concentrations of various soluble factors, including VEGF, Hb, Creactive protein, and some inflammatory cytokines in the serum or plasma, are correlated with AHI (Imagawa et al. 2001(Imagawa et al. , 2004Schulz et al. 2002;Lavie et al. 2002;Gozal et al. 2002;Ohga et al. 2003;Entzian et al. 1996;Ryan et al. 2005;Yokoe et al. 2003). In particular, VEGF has been regarded as a potential marker of OSAS (Lavie et al. 2002). However, the feasibility of using VEGF has been controversial, partly because there is an Figure 2 Association between bFGF and AHI before and after nocturnal hypoxia. Concentrations of plasma bFGF measured the night before PSG and the morning after PSG were plotted against AHI. (Valipour et al. 2004;Peled et al. 2007). Additionally, the specificity of C-reactive protein and inflammatory cytokines (e.g. tumor necrosis factor-α and interleukin-6) for OSAS seems to be low because of the high basal levels of these substances in patients with inflammatory disease, autoimmune disease, and many other diseases (Sharma et al. 2008;Dinarello 1991;Feldmann and Maini 2001;Ishihara and Hirano 2002). This is despite the finding that comprehensive measurement of multiple cytokines has shown high levels of agreement with the AHI and may help in the diagnosis of OSAS severity (Li et al. 2009). age-dependent increase in systemic levels of VEGF in OSAS patients In contrast, only a few reports have demonstrated correlations between levels of bFGF and some diseases, not including OSAS; positive correlations have been found with severe limb ischemia and specific kinds of tumors (Rohovsky et al. 1996;Fujita et al. 1996;Clarke et al. 1998). We hypothesized that bFGF was strongly associated with the hypoxic condition in patients with OSAS, as suggested by the formation of a bFGF-HIF-1α amplification loop (Calvani et al. 2006). In the patients, vascular endothelial cells are readily speculated to be the major source of bFGF during nocturnal hypoxia, which is a situation similar to that of upregulation of bFGF in vascular endothelial cells after the hypoxic exposure in vitro (Calvani et al. 2006;Li et al. 2002). bFGF is shown to mediate a strong angiogenic activity through induction of proliferation of vascular endothelial cells as well as migration of inflammatory cells, perithelial cells, and smooth muscle cells (Calvani et al. 2006;Li et al. 2002;Ruel and Sellke 2003). Therefore, bFGF is one of the feasible target molecules for therapy for ischemic heart diseases including chronic myocardial ischemia, coronary sclerosis, and myocardial infarction (Ruel and Sellke 2003;Hammond and McKirnan 2001). In fact, bFGF is reported to have a therapeutic effect in clinical trials in coronary angiogenesis and in animal models of an ischemic heart disease and critical limb ischemia (Nakajima et al. 2004;Hammond and McKirnan 2001;Simons et al. 2000). On the other hand, bFGF has a hypotensive activity (Cuevas et al. 1991). Taken together, the elevation of systemic levels of bFGF after hypoxic exposure in OSAS patients is primarily supposed to have a pathophysiological role for protection of exacerbated complications of OSAS, e.g. coronary sclerosis, hypertension, ischemic heart diseases. We observed that the change of bFGF concentrations reflected the hypoxic condition in patients with OSAS during sleep periods. Moreover, our results showed the significant relevance between the change in bFGF after nocturnal hypoxic events and diagnostic parameters for OSAS severity, suggesting that it would be worth validating the clinical usefulness of bFGF. In light of our relatively small sample size, further studies of larger numbers of subjects, including OSAS patients treated by nasal continuous positive air pressure, are required to validate the reliability of bFGF as a diagnostic marker of OSAS in the clinical arena. Conclusions We suggested that levels of plasma bFGF were greater in the morning than at the night before in patients with OSAS. The change in bFGF concentration after nocturnal apnea episodes was associated with the degree of hypoxic state in OSAS patients. These results raise the possibility that bFGF might be useful soluble factor for evaluating the severity of OSAS. Subjects Thirty male subjects who were suspected OSAS at the Division of Sleep Medicine participated in this study at the Tsukuba University Hospital. Male subjects with coronary artery disease, hypertension, chronic obstructive pulmonary disease, and inflammatory diseases, and female subjects were excluded in this study, because the basal levels of their plasma bFGF and VEGF might be affected by inflammation, female sex hormones, and menopausal status. All subjects with suspected OSAS were enrolled in this study and were categorized by their AHI values (control, AHI < 15; and OSAS, 15 < AHI) ( Table 1). All subjects provided written informed consent after receiving a full explanation of the procedures. This study was approved by the University of Tsukuba institutional review committee and was performed in accordance with the recommendations found in the Helsinki Declaration. Correlations between the overnight changes (before and after PSG) in plasma VEGF and bFGF concentrations (ΔVEGF and ΔbFGF, respectively) and hypoxiaassociated parameters of OSAS severity were analyzed by using the Pearson's correlation coefficient. Coefficients of correlation are represented by rs and statistical probabilities by P values in parentheses. AHI apnea-hypopnea index, BMI body mass index, bFGF basic fibroblast growth factor, SpO 2 < 90% (% TST) percentage of total sleep time with oxygen saturation below 90%, ODI oxygen desaturation index, SpO 2 oxygen saturation with pulse oximetry, rs Pearson's coefficient of correlation, VEGF vascular endothelial growth factor. BMI was calculated as the ratio of weight (kg) to height (m) squared. PSG The sleep and respiratory evaluations of all participants were manually scored by specialists at the Division of Sleep Medicine, Tsukuba University Hospital. All subjects underwent a standard overnight PSG with computerized recording by the Alice-5 system (Respironics; Pittsburgh, PA). Recordings included an electroencephalogram (at positions C3/A2, C4/A1, O1/A2, and O2/A1 of the International 10-20 System), electrooculogram, submental electromyogram, left and right tibialis anterior electromyogram, thoracic-abdominal effort, oral/ nasal airflow (thermistor-and pressure-based flow measurement), SpO 2 , and body position. Scoring of sleep stage and respiratory events were based on the criteria of the American Academy of Sleep Medicine (Iber et al. 2007). Apnea was defined as the cessation of inspiration for no less than 10 seconds. Hypopnea was defined as a reduction in airflow by no less than 30%, with a decrease in SpO 2 by 4% or more, for at least 10 seconds in the presence of thoracoabdominal ventilatory effort. Obstructive apnea was defined as the absence of airflow in the presence of ribcage or abdominal excursions. The AHI was calculated as the total number of apneic and hypopneic events per hour of sleep periods. All such events were counted, irrespective of the occurrence of arousal that was defined according the criteria of the American Sleep Disorders Association (1992). TST was defined as the time from the first to last recorded sleep periods, excluding wakefulness. The 4% ODI was defined as the total number of 4% desaturation of oxyhemoglobin divided by TST and was expressed as the number of events per hour. The percentage of the TST with SpO 2 below 90% (SpO 2 < 90% (% TST)) was also measured. An AHI of more than 15, irrespective of the presence of sleep-related symptoms (i.e. snoring, witnessed apnea, excessive daytime sleepiness), was considered as diagnostic of OSAS. Measurement of VEGF and bFGF in plasma At the night before the PSG and in the next morning after the subject arose (at 8 p.m. and at 6 a.m., respectively), peripheral blood was carefully collected into tubes with buffered citrate. Plasma samples were prepared by centrifugation at 1500 rpm for 10 minutes at 20°C followed by collection of the supernatants. The supernatants were frozen at −80°C immediately after the collection in order to avoid degradation of soluble factors in them. The concentrations of plasma VEGF and bFGF were determined by sandwich ELISA with quantitative ELISA kits (R&D, Minneapolis, MN). The change in the plasma VEGF and bFGF concentrations from the night of the PSG to the next morning after rising was described as ΔVEGF and ΔbFGF, respectively. Statistical analysis Statistical analysis for evaluation parameters among categories of subjects (age, BMI, AHI, SpO 2 < 90% (% TST), 4% ODI, median and minimal SpO 2 , and plasma VEGF, bFGF, and Hb concentrations the night before PSG and the next morning) was performed by the Student t-test. Statistical analysis for concentrations of plasma VEGF, bFGF, and Hb before and after sleep was performed by the paired t-test. P < 0.05 is regarded as a statistically significant difference. Statistical power analysis was performed by a G*Power 3 program (Heinrich-Heine-Universität). The correlation between ΔVEGF and ΔbFGF, and AHI, SpO 2 < 90% (% TST), 4% ODI, median and minimal SpO 2 in each subject was evaluated with the Pearson's correlation coefficient.	2018-04-03T00:22:29.547Z	2013-06-13T00:00:00.000	{ "year": 2013, "sha1": "1b5149b27b7ae4b034e665dd8b116725ca61df84", "oa_license": "CCBY", "oa_url": "https://springerplus.springeropen.com/track/pdf/10.1186/2193-1801-2-260", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "642e9e8de269104100031e5c800754fcc8d830ef", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
228099911	pes2o/s2orc	v3-fos-license	Typhoid fever infection – Antibiotic resistance and vaccination strategies: A narrative review Travel and Infectious Disease Typhoid fever is a bacterial infection caused by the Gram-negative bacterium Salmonella enterica subspecies enterica serovar Typhi ( S . Typhi), prevalent in many low- and middle-income countries. In high-income terri-tories, typhoid fever is predominantly travel-related, consequent to travel in typhoid-endemic regions; however, data show that the level of typhoid vaccination in travellers is low. Successful management of typhoid fever using antibiotics is becoming increasingly difficult due to drug resistance; emerging resistance has spread geograph- ically due to factors such as increasing travel connectivity, affecting those in endemic regions and travellers alike. This review provides an overview of: the epidemiology and diagnosis of typhoid fever; the emergence of drug- resistant typhoid strains in the endemic setting; drug resistance observed in travellers; vaccines currently available to prevent typhoid fever; vaccine recommendations for people living in typhoid-endemic regions; strategies for the introduction of typhoid vaccines and stakeholders in vaccination programmes; and travel recommendations for a selection of destinations with a medium or high incidence of typhoid fever. Introduction Typhoid is a bacterial infection caused by the Gram-negative bacterium Salmonella enterica subspecies enterica serovar Typhi (S. Typhi). Typhoid fever is usually contracted by ingestion of food or water contaminated by faecal or urinary carriers excreting S. Typhi [1]. The predominant symptom of infection is high fever, with other symptoms including nausea, abdominal pain and abnormal bowel movements [2]. Once prevalent worldwide, improvements in the provision of clean water and sewerage systems has led to a dramatic decrease in the incidence of typhoid fever with the burden of disease now predominantly residing in low-and middle-income countries where sanitary conditions may be poor [1]. In developed countries, typhoid fever is a predominantly travelassociated disease [3], impacting travelling populations such as tourists, military personnel, temporary workers, or travellers visiting friends or relatives (VFR) in endemic areas, with risk varying by the geographical region visited [3][4][5][6], the duration of travel, integration with local cultures, traveller concurrent diseases or medications. Successful management of typhoid fever using antibiotics is becoming increasingly difficult due to drug resistance [7,8]; emerging resistance has spread geographically due to factors such as increasing travel connectivity, affecting those in endemic regions and travellers alike. The aim of this article is to provide an overview of: the epidemiology and diagnosis of typhoid fever; the emergence of drug-resistant typhoid strains in the endemic setting; drug resistance observed in travellers; vaccines currently available to prevent typhoid fever; vaccine recommendations for people living in typhoid-endemic regions; strategies for the introduction of typhoid vaccines and stakeholders in vaccination programmes; and travel recommendations for a selection of destinations with a medium or high incidence of typhoid fever. Along with paratyphoid fever (caused by Salmonella enterica subspecies serovar Paratyphi A, Paratyphi B and Paratyphi C), typhoid fever is a type of enteric fever [18]. S. Typhi is estimated to cause 76.3% (95% CI 71.8-80.5) of cases of enteric fever globally in 2017, with an estimated mean all-age global case fatality of 0.95% (95% CI 0.54-1.53), with higher fatality among children and older adults, and among those living in low income countries [16]. During the course of infection, carriage can be split into three periods convalescence, temporary and chronic [19]. As many as 10% of convalescent carriers shed bacilli in faeces for up to 3 months post-infection [1]. Following resolution of the disease, approximately 1-4% of typhoid patients progress to the carrier state, excreting the bacilli for more than 1 year [1]. Typhoid fever-causing Salmonella have no known environmental reservoir, but it is thought that the chronic asymptomatic carrier is responsible for continued maintenance of the bacteria within human populations [19]. The carrier state has not been extensively studied as the majority of chronic carriers in endemic settings are asymptomatic [1] and a substantial proportion may have had no clinical history of typhoid [19]. However, it is known that chronic carriage is more common in women and increases with age [20]. It is likely that breach of the intestinal epithelial barrier, evasion of early innate immune-mediated killing, and localisation to the biliary tract and gallbladder are requirements for development of the carrier state [19]. Approximately 90% of carriers present with gallstones [21]. Mechanistic studies have shown that S. Typhi forms biofilms on gallstones [22], the formation of which is proposed to aid in their survival and persistence during this phase of the transmission cycle [23]. Recent outbreaks, such as the ongoing outbreak in Zimbabwe (active since 2017), demonstrate how poor sanitation and overcrowding are drivers of seasonal outbreaks of typhoid fever and other waterborne diseases [24]. In addition to water-related and socio-economic risks, improper food handling is a risk factor [25,26], underlying the importance of detecting index cases to allow education (and vaccine strategies) to be appropriately targeted. Typhoid fever is extremely uncommon in countries that have access to treated water supplies and sanitation systems that remove human waste [27]. In non-endemic regions, typhoid is a predominantly travel-associated disease [28][29][30][31][32][33]. In 2016, 22 countries from the EU reported over 1100 cases (typhoid and paratyphoid combined), with France, Italy and the UK accounting for approximately 70% of those cases [33]. Of those with available information, 82.5% were travel-related, with India and Pakistan accounting for the majority of cases [33]. In the USA, 79% (240/303) of typhoid fever patients travelled or lived outside of the country in the 30 days prior to illness [31]. Previously it was estimated that the incidence of typhoid among travellers to a selection of low-and middle-income countries was ≤1 case per 100,000 travellers, except for those who visited Nepal or India where the case rate was 7.9 or 27-81 cases per 100,000 travellers, respectively [3]. The most common reason to travel abroad for patients who subsequently contracted typhoid fever was to visit friends or relatives [29,31,33]. The literature suggests that VFR are at higher risk of contracting travel-related illness, compared with other groups of travellers, due to factors including an increased likelihood of travelling for longer periods of time, travel to rural destinations and being more likely to drink untreated water [34]. Additionally, as VFR may co-habit with their friends and family, they may have greater exposure to asymptomatic carriers in the endemic population (and possibly experience lower levels of sanitation in the home setting), contributing to their increased risk of typhoid fever compared with other travellers. Another group of travellers potentially at risk of typhoid are military personnel deployed to typhoid-endemic regions. A review of the typhoid fever cases reported in the US military (1998-2011) revealed 205 cases of typhoid fever among military service members, for an incidence of 1.09 per 100,000 person-years, consistent with the incidence of typhoid in the general US population during that same time period, the majority of which was travel-related [35]. Increases in deployment of military personnel to countries with medium or high incidence estimates could be expected to increase the risk of typhoid fever, however this has not been shown, in part due to deployed troops not experiencing the living conditions of the local population [35]. The risk of acquiring typhoid increases with the duration of stay. Correspondingly, in a study of travellers from Greece to typhoidendemic areas (such as sub-Saharan Africa and Southeast Asia), it was shown that typhoid vaccine administration is statistically associated with the duration of stay [36] as travellers look to mitigate their risks. Approximately 20% of travellers who stayed <1 month underwent vaccination compared with nearly 35% who stayed ≥6 months [36]. However, data show that typhoid fever is a risk even for short-term travel -5% (31/626) of travellers from the USA with typhoid fever reported that their travel outside of the USA lasted ≤1 week compared with 60% (376/626) whose travel lasted ≤6 weeks [37]. Therefore, behaviours and expected exposure to typhoid fever at the destination should be considered, not only the length of travel. Travellers who acquire typhoid fever are rarely immunised. Previous data of travellers who developed typhoid fever show that just 4% (36/ 1027) reported having received a typhoid vaccination at any point during the previous 5 years preceding travel [37]. Data from the US military revealed that 26% (53/205) of military personnel had a documented S. Typhi vaccination within 2 years of typhoid fever diagnosis [35]. More recent data of young Canadian travellers with typhoid fever revealed that 0% (0/39) had a record of typhoid vaccination [30]. This low rate of vaccination is therefore one of the key factors that contribute to the occurrence of typhoid infection in travellers [30,37]. Diagnosis of typhoid fever Isolation of S. Typhi from blood is the current gold standard for determining typhoid fever infection [38]. Bacterial culture from bone marrow is more sensitive but is difficult to obtain, invasive and is impractical for routine use [38]. Although blood cultures remain the gold standard for diagnosis, this method has poor sensitivity [38]. In addition, culturing takes at least 48 h to generate results, and optimal usage is hampered by lack of healthcare infrastructure and adequately trained laboratory personnel in resource-limited countries where typhoid fever is common [38,39]. The Widal test is a serological test that detects agglutinating antibodies against the O and H antigens [38]. It is widely used due to its simplicity and low cost [38] but has low sensitivity [40] and is highly operator dependent [41], with values varying considerably between geographical areas [1]. In addition, the background level of antibodies in a normal healthy population within a typhoid-endemic region means that proper interpretation of the Widal test results entails collection of sera from two visits (spaced 10-14 days apart) and requires each country to determine the appropriate antibody titre with which to diagnose typhoid [39,42], since there is no universal titre cutoff to define the disease. With the emergence of antibiotic resistance (see Section 4, "Overview of the treatment of typhoid fever and the emergence of drug-resistant typhoid strains"), the Widal test is inadequate as it does not provide susceptibility results [43]. Nucleic acid amplification tests, including conventional polymerase chain reaction (PCR), nested, multiplex and real-time PCR, have been developed for the detection of S. Typhi DNA in blood [39]. The main challenges preventing widespread use in low-resource settings are the high costs associated with this technology [39]. There are several commercially available typhoid rapid antibody tests that can generate results in as little as 2 min [44] allowing prompt treatment with antimicrobials for those with a positive result. One qualitative test, Typhidot®, uses pre-dotted antigen strips to detect the presence of immunoglobulin (Ig)M and -G antibodies to an outer membrane protein [44]. A semi-quantitative colourimetric test (IDL TUBEX® TF) relies on visual and subjective examination of colour reactions to detect anti-O:9 antibody titres [44]. However, in comparative studies of performance utilising blood culture as the comparator, this type of testing can perform poorly [44]. A recent meta-analysis reported an average sensitivity of 78% (95% CI 71-85; the ability to identify true positives) and specificity of 87% (95% CI 82-91; the ability to identify true negatives) for TUBEX [38]. Analysis of all Typhidot variants as a group showed an average sensitivity of 84% (95% CI 73-91) and specificity of 79% (95% CI 70-87) [38]. A recent study conducted in Bangladeshi patients revealed that the sensitivity of typhoid rapid antibody tests may be as low 60.2% (95% CI 49.3-71.2) and 59.6% (95% CI 50.1-69.3), respectively [45]. These results emphasise the difficulties of laboratory diagnosis of typhoid. At present there is no gold standard for the detection of chronic typhoid carriers. Standard practice has been to detect typhoid carriage through serial analysis of stool and urine samples (using culturing), however this is logistically challenging and suffers from low sensitivity [19]. The presenting signs and symptoms of typhoid fever are shared with other febrile illnesses (such as malaria, dengue or other arbovirus fevers) [46], and thus differentiating S. Typhi infection from other sources of fever in endemic areas is a diagnostic challenge [38]. In most endemic areas (at least in Africa), the only means for diagnosis of a febrile patient is the Widal test for typhoid fever or the thick blood smear for malaria and, as has been shown, the Widal test performs poorly. The lack of diagnosis (because tests are not ordered), misdiagnosis (due to the use of insensitive methods), poor surveillance and inadequate healthcare infrastructure are common problems in resource-limited settings. New strategies that are specific, sensitive, scaleable and cost-effective are required to correctly identify both acute sufferers and chronic carriers of typhoid, in order to stop expansion of the disease and determine the real global burden of S. Typhi. For patients in countries where typhoid is not endemic, determination of the patient's travel history is crucial [1]. Overview of the treatment of typhoid fever and the emergence of drug-resistant typhoid strains The treatment of typhoid fever normally consists of antibiotics [1,11]; early initiation of effective antimicrobial therapy has been shown to shorten the duration of the illness and reduce the risk of complications and death [47]. Because of the high risk of morbidity and mortality if left untreated [48], clinicians may administer antibiotics to patients in the absence of a confirmed diagnosis (on the clinical suspicion of typhoid fever). However, large surveillance studies from Asia and Africa indicate that only 1-4% of people with suspected typhoid actually have culture-confirmed typhoid [49], which suggests that there might often be substantial overtreatment with unnecessary antibiotics. One of the effects of empiric prescribing of antimicrobials has been an increase in selective pressure on S. Typhi [49,50]. Since 1948, when the efficacy of chloramphenicol to treat typhoid was discovered, there has been a pattern of antibiotic usage and resultant development of resistance to antimicrobial therapies [7]. Subsequently, antimicrobial resistance has become a major threat to the treatment of typhoid with increasing levels of treatment failure [7,8] (Fig. 2). Multidrug resistance (MDR) is historically used to describe combined resistance to the first-line antibiotics chloramphenicol, co-trimoxazole (trimethoprim-sulfamethoxazole) and ampicillin [2]. S. Typhi can harbour complex MDR elements, either on self-transmissible plasmids carrying a cassette of antimicrobial resistance genes [51], or integrated into the chromosome (which is more common than previously thought) [52]. While antibiotic selection maintains resistance genes on the plasmid, there also appears to be competition between plasmids encoding the same resistance phenotype [53]. MDR S. Typhi is now considered endemic in many developing countries, especially in areas of South and Southeast Asia [2], mediated by the dissemination of the specific H58 lineage across Asian and African countries [8]. As with the disease incidence, antimicrobial agent susceptibility patterns vary geographically [54]. High incidences of MDR S. Typhi are found in areas with a high burden of typhoid, particularly in children aged under 15 years [55]. There is a paucity of data on the geographical distribution, incidence and phylogenetics of MDR S. Typhi in sub-Saharan Africa [55], however the H58 clade of S. Typhi is associated with the MDR phenotype [56] and with much of the typhoid occurring in the last decade in East and Southern Africa [54]. Longitudinal studies show that the proportion of MDR strains decrease over time as clinicians respond to resistance and use alternative drugs [57]. Correspondingly, MDR S. Typhi is on the decline in South and Southeast Asia, because of the reduced usage of these first-line drugs in this region [54]. In response to the development and spread of MDR S. Typhi, the use of fluoroquinolones, (ciprofloxacin, ofloxacin, fleroxacin and pefloxacin) became widely accepted as an alternative to treat typhoid fever [2]. However reports of decreased susceptibility to fluoroquinolone soon followed both in endemic areas and in travellers returning from such areas [2]. In areas with a high prevalence of both MDR and fluoroquinolone resistance, azithromycin (an azalide antimicrobial) and extended-spectrum cephalosporins (e.g. ceftriaxone) tend to be used for treatment [2]. The first outbreak of an extremely drug resistant (XDR) H58 clone harbouring resistance to not only the three first-line drugs (chloramphenicol, co-trimoxazole and ampicillin), but also fluoroquinolones, and extended-spectrum cephalosporins, began in Sindh province, Pakistan in 2016 [43,58,59]. All of the XDR isolates belonged to the H58 clade and it is thought that the plasmid conferring resistance originated in Escherichia coli and was acquired by an MDR H58-endemic S. Typhi clone in Pakistan [43]. This outbreak resulted in 5372 XDR S. Typhi cases reported from 2016 to 2018 [60]. First reported in June 2019, several areas of Sindh province are again suffering from an outbreak of typhoid, with two child deaths and illness in more than 150 people already recorded. Thus far, five of 51 typhoid cases have been reported as XDR, with results of several other cases awaiting. Health authorities are planning to launch a mass vaccination drive to gain control of this current outbreak [61]. Where patients experience resistance to all firstand second-line drugs, the carbapenems (e.g. imipenem, meropenem and ertapenem) and tigecycline are considered potential alternatives [2]. In addition to the difficulties of ensuring patients receive effective treatment, there are also cost implications associated with resistance: the cost of therapy for resistant cases can be approximately 70% greater than the cost of sensitive typhoid cases, driven mostly by an increase in physician and nursing care [62]. Even with appropriate treatment, the risk of relapse remains, with approximately 5-10% of immunocompetent patients relapsing [63]. Typhoid relapse cases have been reported in typhoid-endemic countries and in travellers returning from those regions [48,63,64]. In one small study from India, patients with drug-resistant typhoid who initially received ineffective therapy had a higher relapse rate following effective treatment compared with those infected with pan-sensitive strains [63]. For the treatment of chronic carriers, eradication has been achieved with some success using ampicillin or amoxicillin, sometimes combined with probenecid or co-trimoxazole (dependent upon the susceptibility of the strain) [2]. A small study of 12 chronic S. Typhi carriers revealed a 92% cure rate with a 4-week regimen of fluoroquinolones (ciprofloxacin orally twice a day for 28 days) [65]. At present, there are few data on this subgroup and more research is necessary to determine the best approach for treatment. Drug-resistance in travellers Resistance to antibiotics has been observed in travellers from nonendemic regions with typhoid fever, with patterns reflecting the conditions in typhoid-endemic countries. Case reports of travellers returning to Spain from Guatemala [66], and to Germany from Iraq [67], have documented infections of an extended-spectrum beta lactamase-producing S. Typhi strain. Data gathered in the USA from the National Typhoid and Paratyphoid Fever Surveillance System and National Antibiotic Resistance Monitoring System revealed that antimicrobial resistance was common [68]. Among S. Typhi isolates, 314 (29%) were susceptible to all clinically relevant antimicrobial agents, 750 (69%) were resistant to nalidixic acid or had decreased susceptibility to ciprofloxacin, 127 (12%) were MDR, and 108 (10%) were both nalidixic acid resistant/decreased susceptibility to ciprofloxacin and MDR [68]. Isolates resistant to nalidixic acid (a synthetic quinolone that is a marker of decreased susceptibility or resistance to fluroquinolones) originated mainly in southern Asia whilst MDR cases originated from southern Asia and Africa [68]. The afore-mentioned typhoid fever outbreak of 2016-2018 in Sindh province resulted in five travel-related cases in children visiting relatives in Pakistan and returning to the USA [69], and one in the UK [43]. In April 2019, a ceftriaxone-resistant S. Typhi case was detected in a pregnant woman returning to Denmark from a family visit to Sindh province [70]. In August 2019, the first case of XDR typhoid was reported in Australia [71]. Collectively, the occurrence of resistant cases necessitates heightened vigilance and re-consideration of treatment strategies for those who have visited a region that has high levels of antimicrobial resistance. S. Typhi is quickly able to acquire new resistance mechanisms. As long-distance travel becomes more accessible and migration between endemic and non-endemic regions increases, access to information regarding local and regional susceptibility is important to guide empirical treatment [10]. Consequently, the landscape of resistance is dynamic. The World Health Organization (WHO) has recommended that the surveillance of typhoid fever be strengthened, including surveillance to monitor known resistance, detect new and emerging resistance, and mitigate its spread [72]. Vaccines available for typhoid To stop the expansion of typhoid, interventions should mostly focus on prevention. Along with improvements in antibiotic stewardship, public sanitation, availability of clean drinking water, safe food handling practices and public health education, vaccination in endemic areas is one strategy to prevent typhoid fever. The WHO recommends programmatic use of typhoid vaccines for the control of typhoid fever [73]. Ty21a vaccine The Ty21a vaccine was developed by chemical-induced mutagenesis of the S. Typhi Ty21 strain, resulting in a galE mutant and the inability to express the Vi polysaccharide antigen [109]. Inactivation of the galE gene generates a lack of uridine-diphosphate (UDP)-galactose-4-epimerase [110], which is responsible for the conversion of UDP glucose into UDP galactose. Consequently, there is intracellular accumulation of galactose derivatives and subsequent bacterial lysis, thus eliminating virulence of the vaccine strain [110,111]. Lipopolysaccharide synthesis is preserved [109]. S. Typhi is an intracellular bacterium and thus needs T celldependent immunological response for resolution [112]. Vaccination with Ty21a activates a broad immune response, with both humoral and systemic cell-mediated immune responses. The serum antibody response to Ty21a is a rise in serum IgG antibody against the O polysaccharide of S. Typhi [113]. Data have shown that the rate of seroconversion of IgG anti-O antibodies increases as more doses are administered within a period of 7 days [88] and that the humoral response is dose-dependent [114]. Following oral immunisation, mucosal antibody responses have been reported. Vaccination studies in healthy adult volunteers have revealed that oral Ty21a immunisation results in an increase in the concentration of IgG anti-lipopolysaccharide [115,116] as well as an increase in levels of faecal IgA (total and/or specific) [115][116][117]. Data have also shown that after oral vaccination antibody-secreting cells that produce specific IgA to S. Typhi O-polysaccharide bear homing receptors which commit them to migrate to the intestinal mucosa [118]. Cell-mediated responses include contributions from CD4 + helper T cells and CD8 + cytotoxic T cells [27]. Highest level of recommendation: 2b • Adults and children ≥2 years of age [75] Highest level of recommendation: 4 • Adults and children ≥6 months to ≤45 years of age [77] Highest level of recommendation: 4 • Children of all age groups from 6 months to 12 years of age [86] Safety Post-marketing surveillance [90] Most common AEs reported in the USA, • No serious or unusual side effects in either children or adults Post-marketing surveillance [90] Most common AEs reported in the USA, July 1990-June 2002& • Fever (n = 11), headache (n = 11), dizziness (n = 9), rash (n = 9), urticaria (n = 9), myalgia (n = 6), pain (n = 6), abdominal pain (n = 6), pruritis (n = 6), injection site pain (n = 5) Human challenge study [85] Typbar-TCV vaccination arm: adults, n = 41 [87]. The justification for the four-dose regimen employed in the USA and Canada originates from a large randomised field study of Chilean school children, which assessed the comparative efficacy of a two-, three-or four-dose regimen of Ty21a [81]. The incidence of typhoid fever in recipients of two or three doses was significantly higher than in those who received four doses (48% vs 40%, p < 0.002) [81]. A subsequent analysis of vaccine efficacy, using data from the two-dose group as standard, reported efficacy of 13% (95% CI 0-33; years 1-3) for those who received three doses and 48% (95% CI 29-62; years 1-3) for those who received four doses [88]. The effectiveness of public health vaccination programmes can depend on both the vaccinated and unvaccinated population. Herd protection is the reduction in infection or disease in the non-immunised proportion of the population as a result of vaccination in another proportion of the population [119]. Evidence that Ty21a can provide indirect herd immunity comes from a field trial in Santiago, Chile, that used one or two doses, during which the incidence of typhoid fever in the control group fell during the follow-up period [102]. The incidence rate in the randomised placebo group in the first year of surveillance was 227 cases per 100,000 school children [102]. The incidence in the second year fell to 139 cases per 100,000 school children [102]. The data suggest that large-scale vaccination appears to cause indirect herd immunity in non-vaccinated subjects [88]. Data regarding co-administration with malaria prophylaxis were determined using the previously developed liquid formulation, rather than the capsule formulation in current usage, and suggested that the vaccine can be administered concurrently with malaria prophylaxis using atovaquone/proguanil [120]. Current manufacturer recommendations state that vaccination should be completed prior to malaria prophylaxis, with an interval of at least 3 days between the last dose of vaccine and the start of malaria prophylaxis [74]. Ty21a can be administered to immunologically stable HIV-positive individuals (CD4% >25% for children aged <5 years or CD4 count ≥200 cells/mm 3 if aged ≥5 years) [73,121]. Ty21a is not recommended for individuals with a known depression of cell-mediated immunity [73]. Ty21a may be given concurrently with the yellow fever vaccine, CVD 103-HgR cholera vaccine and oral polio vaccine [74]. No data are available regarding the interaction with other live attenuated vaccines. Due to the potential inhibition of the growth of the vaccine organisms and potential attenuation of the immune response, vaccination with Ty21a should be postponed during, and for at least 3 days before and after, antibiotic or antibacterial sulphonamide treatment [74]. For longer-acting antibiotics, eg azithromycin, a longer postponement should be considered [74]. At present there are no data on the safety and immunogenicity of any of the typhoid vaccines in pregnant or lactating women [122]. Generally, live vaccines, and therefore Ty21a, should not be given during pregnancy due to theoretical safety concerns [122], however the manufacturer suggests that it could be administered in cases of increased risk of infection [74]. The efficacy trials of the three-or four-dose alternate-day regimen did not specifically include safety monitoring. However, safety of the Ty21a vaccine can be assessed from post-marketing surveillance data. Post-marketing surveillance data collected in the USA through the Vaccine Adverse Event Reporting System (July 1990 to June 2002; fourdose regimen) revealed 345 reports of adverse events (AEs) related to the administration of Ty21a alone or in combination with other vaccines [90]. The rate of AEs or serious AEs was 9.7 or 0.59 per 100,000 doses, respectively and AEs included diarrhoea, nausea, fever and abdominal • Licensed in the USA in 1994 [73] • Licensed in over 100 countries [95] • Recommended by the WHO since 2008 for the control of typhoid in endemic and epidemic settings [73] • Licensed in India in 2013 [96] and Nepal [97] • Attained WHO prequalification status in 2018 [97] • Licensed in India in 2008 [73] Δ Injection site pain was statistically significant, p = 0.0499 [85]. Abbreviations: CI, confidence interval; AE, adverse event; S. Typhi, Salmonella enterica subspecies enterica serovar Typhi; WHO, World Health Organization. a The four dose regimen was registered in the USA and Canada in 1987 as a result of the study by Ferreccio and colleagues [81]. b Note that information from three additional studies has been excluded as the formulation, dose regimen or time interval between doses does not reflect those in current recommendations. Black and colleagues evaluated the efficacy of one or two doses of Ty21a in a randomised placebo-controlled field study in >80,000 children in Santiago, Chile [102]. Simanjuntak and colleagues evaluated three doses at weekly intervals [98]. Wahdan [103,106]. Typherix has been withdrawn in some countries [107,108]. f This can be used from 3 years of age according to technical data, however as the user has to be able to swallow the whole capsules it is recommended from 5 years of age. g Adverse events for the vaccine when given alone (not in combination with any other vaccine). h This monograph includes data from 2 clinical trials [84,100] as well as data held on file by Sanofi Pasteur. pain [90]. Of 38 million immunisations administered between 1990 and 2000, the total number of spontaneously reported adverse drug reactions was 0.001954% (i.e. ~20 in every million immunisations will result in an adverse drug reaction) [123]. During the early pre-licensure field trials in Chile, there was a considerable level of paratyphoid fever due to S. Paratyphoid B, allowing estimation of cross-protection. The Ty21a vaccine has also been shown to confer moderate protection against S. Paratyphi B (efficacy 49%) [89]. A randomised double-blind field trial of 20,543 subjects in Plaju, Indonesia (one dose per week for 3 weeks) suggests that Ty21a does not offer cross-protection against S. Paratyphi A [98]. Vi polysaccharide vaccine ViPS vaccines are based on the purified capsular polysaccharide, S. Typhi Vi antigen [11]. The Vi antigen of S. Typhi is well recognised as a factor for virulence and as an antigen that confers immunity against the typhoid fever [124]. The ViPS induces T cell-independent immune responses against S. Typhi, thus resulting in a lack of prolonged protection [97]. Evidence of the efficacy of ViPS vaccines comes from studies that utilised vaccines from two different manufacturers, TYPHIM Vi® (Table 1) and Typherix® (GlaxoSmithKline) vaccines. Typherix vaccine is no longer commercialised in some countries and may be unavailable for use [107,108]. Early pre-licensure randomised studies demonstrated that ViPS vaccines offer a moderate level of protection. A Nepalese study of TYPHIM Vi in subjects aged 5-44 years reported a total efficacy (blood culture proved and clinically suspected) of 75% [82]. A double-blind randomised trial of TYPHIM Vi in South African school children aged 5-15 years reported an efficacy of 55% [84]. A meta-analysis of four individually randomised clinical trials (including the two studies mentioned above plus two additional trials in China utilising locally produced ViPS) demonstrated efficacy of 69% (95% CI [63][64][65][66][67][68][69][70][71][72][73][74] in year 1 (three trials) and 59% (95% CI in year 2 (four trials) [11]. An outbreak of typhoid fever in previously vaccinated French soldiers stationed in Cote d'Ivoire revealed that an interval of >3 years from the time of vaccination may result in reduced immunity [126]. Hypo-responsiveness or immunotolerance to re-vaccination is the inability of an individual to mount an immune response equal or greater than the immune response induced by primary vaccination [127]. This phenomenon has been described for polysaccharide vaccines [127]. Roggelin and colleagues assessed the influence of previous vaccinations with a ViPS (not more than 5 years ago) on post-vaccination antibody concentrations [128]. Thirty-six of the participants had been vaccinated once, nine had been vaccinated at least twice. There was no evidence of immunotolerance with multiple versus primary vaccination using a ViPS vaccine. However this was a small study, and the subgroup of those who had been previously vaccinated once showed the highest post-vaccination geometric mean antibody concentration compared with those who had been vaccinated twice before or received their primary vaccination [128]. A post-licensure cluster randomised trial in Kolkata, India, indicated that Typherix offers herd immunity. Among unvaccinated members of the vaccine clusters (geographical units that served as units of randomisation), the level of protection was 44% (95% CI 2-69, model based on 25,083 subjects) [103]. ViPS vaccines, because of their poor immunogenic and T cell-independent properties, are not suitable for immunisation of infants [112]. ViPS vaccines are safe and recommended for HIV-infected individuals [73,121], however elicitation of protective antibodies is directly correlated with the levels of CD4 + T cells and thus the proportion of individuals with protective antibody concentrations will be lower than in healthy controls [129]. Antibody responses are severely impaired in those with CD4 <200 cells/μL [129]. ViPS vaccines can be co-administered with other vaccines, including inactivated or attenuated, which is relevant for international travellers and routine childhood vaccination schedules [73]. As previously mentioned, there are no data on the safety of typhoid vaccines in pregnant women [122] and thus the manufacturer of TYPHIM Vi recommends that it should only be administered if clearly required and following risk assessment [75]. Post-marketing surveillance data collected in the USA through the Vaccine Adverse Event Reporting System (July 1990 to June 2002) revealed 321 reports of adverse events related to the administration of TYPHIM Vi alone or in combination with other vaccines [90]. The rate of AEs or serious AEs was 4.5 or 0.34 per 100,000 doses, respectively [90]. Common AEs included fever, headache and dizziness [90]. Serious side effects were rare [90]. Neither S. Paratyphi A or B express the Vi capsular antigen therefore there is no biological plausibility for cross-protection [27]. Theoretically, ViPS should offer protection against S. Paratyphi C, which does express the Vi capsular polysaccharide, although this is a rare cause of enteric fever and no data from field trials have reported protection [27]. Typhoid conjugate vaccines TCVs combine the Vi-polysaccharide capsule with a protein carrier [85]. For example, each dose of PedaTyph contains 5 μg of Vi polysaccharide of S. Typhi conjugated to 5 μg of tetanus toxoid [91], a recombinant inactive form of tetanus toxin produced by Clostridium tetani [97]. Covalently conjugating ViPS to carrier proteins overcomes the limitation of the ViPS vaccine by changing the immune response from T cell-independent to T cell-dependent, enabling immunisation even in young children [97]. Several conjugate subunit vaccines combining ViPS with another protein antigen (typically inactive forms of bacterial exotoxins) are under investigation [97]. At present, hyporesponsiveness is not thought to be a feature of repeated vaccination with conjugate vaccines [127]. Limited data are currently availablelarge field studies are required and further investigations are being conducted by the Typhoid Vaccine Acceleration Consortium [130]. Since TCVs are less well established for typhoid vaccination, data are comparatively limited. A Phase 2b randomised controlled trial showed that the efficacy of Typbar-TCV was 54.6% in a controlled human infection model of typhoid fever in adult volunteers aged from 18 to 60 years; a post hoc analysis of these data revealed that if alternative diagnostic criteria were applied, such as fever of 38.0 • C or higher followed by bacteraemia, the estimated efficacy would be 87.1% [85]. In a report of two trials (randomised controlled and an open-label trial), adverse events for Typbar-TCV were uncommon [92]. Fever was the most common, with 4.3% (14/340) of subjects experiencing this AE [92]. Post-marketing surveillance data reported fever, pain and swelling in 1-10% of vaccinees in any age group (no serious AEs were reported) [131]. Whether Typbar-TCV offers herd immunity is unknown at present, however herd immunity has been observed with other conjugate vaccines due to mucosal immunity and will need to be addressed in future studies. The efficacy of PedaTyph was determined in a single cluster openlabel randomised control trial in 950 children aged 6 months to 12 years: year 1 efficacy was reported to be 100% (95% CI 97.6-100) [86]. A study of 400 children to assess the safety and immunogenicity of PedaTyph reported non-severe AEs in 17% of children with one dose, recoverable within 48 h [91]. At present, there is no evidence of interference with other vaccines (e.g. the trivalent measles, mumps and rubella vaccine) [122]. Although there are no data to this effect, like the ViPS vaccine, it is not biologically plausible for cross-protection to be induced against S. paratyphoid A and B. Theoretically, cross-protection against S. paratyphoid C is possible. Other typhoid conjugate vaccines There are several other typhoid vaccines available, or in the development pipeline, that utilise a variety of carrier proteins. Vi-rEPA, Vi polysaccharide bound to the recombinant exoprotein of Pseudomonas aeruginosa, showed efficacy of >90% in a Phase 3 study of Vietnamese children aged 2-5 years [132] and will be available in China soon [133]. Vi-CRM 197 , Vi polysaccharide conjugated with CRM 197 (a non-toxic mutant of diphtheria toxin), has completed Phase 2 trials in infants, children and adults [134] and is in clinical development [135]. Vi-DT vaccine, Vi polysaccharide conjugated to diphtheria toxoid, is currently undergoing assessment in a Phase 2 trial in children ages 3-23 months [136]. A Phase 3 trial in children and adults is due to complete in January 2020 [137]. In a Phase 1 study in adults aged 18-40 years old and children aged 2-5 years old [138], and in Phase 2 studies in children aged 6 to ≤24 months [139,140], the Vi-DT vaccine has been shown to be well tolerated and immunogenic. In summary, there are several vaccines available for the prevention of typhoid, the choice of which will include factors such as the local availability and age of the intended recipient. Vaccination strategies A combined approach to typhoid health security based on public health that includes the use of typhoid vaccines, improvements in sanitation and safe water supply is necessary [141]. Vaccination has the potential to decrease the use of antibiotics, limit the emergence of resistant S. Typhi strains and create a herd immunity. Consequently, it should be offered to those who reside in endemic regions as well as to travellers to destinations where antibiotic-resistant strains of S. Typhi are prevalent [142]. As it is difficult to obtain the output from monitoring and reporting information in real time, and because typhoid fever may not have the highest priority due to competition from other diseases with higher morbidity/mortality, vaccination should be offered irrespective of the intensity of other control strategies [142]. Feedback from regions and countries on the implementation of the Strategic Advisory Group of Experts recommendations on typhoid identified that, as the incidence of S. Typhi declines, there has been a concomitant rise in the incidence of S. Paratyphi [143]. This increases the uncertainty of the value of typhoid vaccination in the absence of a vaccine against S. Paratyphi. In this context, vaccines that offer cross-protection against S. Paratyphi (Ty21a for example) may be considered to be of greater value. Global recommendations for vaccination in endemic areas The WHO currently recommend vaccination, using Ty21a, ViPS or TCV, to control endemic typhoid fever and outbreaks. The WHO recommends TCV in all ages due to its improved immunological properties, suitability for use in younger children and expected longer duration of protection [73]. Use of ViPS vaccine in individuals aged ≥2 years, and Ty21a vaccine for individuals aged >6 years, is also suitable [73]. Strategies for the introduction of typhoid vaccines Routine typhoid fever vaccination for public health use has been very limited in the past, even in endemic areas, hampered by the lack of an effective vaccine for young children and relatively short duration of protective efficacy [144][145][146]. Typbar-TCV attained WHO pre-qualification in 2017 so the current landscape of typhoid vaccine usage is set to change (see Section 6.1.2, "Stakeholders in typhoid vaccine programmes") [147]. The introduction of TCVs should be prioritised in countries with the highest burden of typhoid disease or a high burden of antimicrobial resistant S. Typhi [73]. Data from Breiman and colleagues showed a high disease burden in densely populated urban slums compared with a low incidence in rural areas, supporting a typhoid immunisation strategy with a geographically and environmentally targeted approach [148]. Studies are ongoing to demonstrate whether the programmed introduction of an effective typhoid vaccine into countries with high burden of disease or significant antimicrobial resistance could have a dramatic impact, protecting children from infection and reducing antimicrobial usage [149]. Health-economics and cost-effectiveness studies suggest that vaccination (using a TCV) is likely highly cost-effective in high-burden settings (>100 cases per 100,000 population) [49]. An age-structured transmission and cost-effectiveness model that simulated multiple vaccination strategies with a TCV showed that TCVs would be highly cost-effective in low-income countries in settings of moderate typhoid incidence (50 cases/100,000 annually) [150]. In countries with epidemiological evidence of high incidence in welldefined subpopulations, a vaccination strategy based on risk assessment should be considered [142]. An evaluation of the cost-effectiveness of ViPS vaccination against typhoid in multiple Asian sites demonstrated that a vaccination programme targeting children would be "very cost-effective" (e.g. costs per disability-adjusted life-years averted less than per-capita gross national income), but programmes that also target adults are less cost-effective (albeit due to the lower incidence in adults than children) [151]. Recent cost-effectiveness analysis in five endemic, low-and middle-income settings showed that routine vaccination with TCVs, as well as one-time catch-up campaigns, would be cost-effective in most settings [152]. A recent study has modelled the predicted impact of TCVs on antimicrobial resistance, using the relative fitness of the resistant strain(s), prevalence of chronic carriers, and rates of recovery without treatment as variables. Notably, herd immunity was not included in the model. The study found that increasing vaccination coverage would decrease the total number of antimicrobial-resistant typhoid infections but not affect the proportion of cases that were antimicrobial resistant [153]. Further evaluation in real-world typhoid-endemic settings is necessary to confirm the model's predictive results. However, in low-resource settings (where typhoid is endemic), there is little chance of being able to confirm these results. As such, the likely impact of TCVs on antimicrobial resistance remains to be determined. As asymptomatic carriers are thought to be responsible for maintenance of S. Typhi in the human population, a strategy for reducing this reservoir could be of interest. Stratification of carriers by age could support the vaccination strategy in endemic countries. Interestingly, concomitant administration of Ty21a and ViPS has also been explored [154]. The enhanced immune response observed suggests that usage in this manner should elicit higher protective efficacy, although this is still to be investigated. The authors suggest that concomitant use should be encouraged for those at significant risk [154]. Recent findings from Bhutta and colleagues, studying surveillance data from several countries, documented substantial reductions in typhoid fever burden over recent decades which may be associated with improvements in economic conditions, education and environmental health [155]. It is likely that public health programmes of vaccination could also impact the burden of disease, as has been seen with other pathogens. However, it is possible that differing patterns of urbanisation, and resultant population densities, could impact on the success of vaccination strategies. Like typhoid, cholera is a bacterium transmitted by faecally contaminated water and it has a high burden of disease, predominantly in areas that lack access to safe water [156]. In 2011, the first low-cost oral cholera vaccine obtained WHO pre-qualification status and efforts began to stockpile the vaccine for use in endemic and epidemic settings [157]. There were operational challenges that hampered the subsequent vaccination campaigns, including regulatory hurdles, cold chain logistics, and vaccine coverage and uptake, however from 2011 to 2015, 4.8 million doses of cholera vaccine were administered globally [157]. Whilst it may be too early to assess the impact of these vaccinations, surveillance data from Malawi has shown that cholera outbreaks were absent in vaccinated high-risk areas (despite a national outbreak during the surveillance period), suggesting the suitability of this approach [158]. Vaccines have been used with great effect for other endemic diseases, such as polio. Since 1988, when the sustained use of polio vaccines began, the number of countries with endemic polio has reduced from 125 to just two in 2015 (Afghanistan and Pakistan) [93]. Concerted and sustained efforts concerning typhoid vaccinations might similarly result in substantial reductions in the decades to come. Stakeholders in typhoid vaccine programmes Several global activities support the introduction of vaccines, with efforts ongoing to make new typhoid vaccines available and accessible to those that need them. As part of their vaccine investment strategy, Gavi, the Vaccine Alliance, prioritised typhoid vaccines in 2008 but did not make a financial commitment at that time [147]. All vaccines supported by Gavi are supplied through the United Nations procurement process and so pre-qualification is necessary for purchase and subsequent introduction into Gavi-eligible countries [159]. Following the WHO pre-qualification of the TCV vaccine Typbar-TCV in 2017 [147], Gavi approved US$85 million in funding to support eligible countries to introduce TCVs [160]. At present, there are 58 countries eligible for Gavi fundingto be eligible, a country's gross national income per capita must be ≤US $1580 over the previous 3 years [161]. A country's Gavi-eligibility status may have a large influence on a country's vaccine introduction decision and the introduction of TCVs may be more difficult in Gavi-ineligible countries [162]. The Coalition against Typhoid was created with the aim of preventing typhoid among vulnerable populations through research, education and advocacy [163]. The Typhoid Vaccine Acceleration Consortium aim to facilitate the introduction of TCVs into Gavi-eligible countries to reduce the burden of typhoid using an integrated, proactive approach [164]. Through the efforts of these groups and others, progress towards the reduction in typhoid fever is being made. Recommendations for travellers to high-risk areas For individuals in non-endemic areas, typhoid fever impacts travelling populations such as tourists, military personnel, or travellers VFR in endemic areas. Travel destination is the main factor for assessing the risk of acquiring typhoid fever [36]. Other factors include area of stay, duration of stay, purpose of travel and type of accommodation [36]. There are also risk factors intrinsic to the traveller, including age and diagnosed/undiagnosed underlying medical conditions [36]. In those travellers with impaired immunity (e.g. HIV), vaccination is less effective [129]. There is relatively little evidence of the effect of vaccinations in preventing typhoid in travellers who have visited typhoid-endemic areas. Data from 4 years of US national surveillance reported by Mahon and colleagues showed moderate protection (80%) from typhoid fever in vaccinated US travellers who travelled to southern Asia [166]. As the vaccination information gathered from reports did not specify the vaccine administered, the effectiveness of each individual vaccine could not be estimated. In the UK, between 2007 and 2012, >99% of typhoid vaccines prescribed were ViPS vaccines (TYPHIM Vi, Typherix, Hep-atyrix® or VIVAXIM®) [167]. In travellers from England (children ≥2 years and adults) the overall effectiveness of typhoid ViPS vaccines was 65% (95% CI 53-73) [167]. In Canada, the Committee to Advise on Tropical Medicine and Travel suggest that typhoid vaccine (Ty21a or ViPS) be used for Canadian travellers visiting South Asia [168]. In the UK, the National Travel Health Network and Centre generally recommend vaccination for most travellers to typhoid-endemic countries where there is a "medium" disease incidence and the access to improved sanitation is <80% [169]. In the USA, guidelines were updated in 2015, and continue to recommend vaccination for travellers to certain countries, close contacts of chronic carriers, and certain laboratory workers [170]. At present, recommendations worldwide are not harmonised and can be generic or tailored for risk groups (Table 2). Experts must assess vaccination of travellers based on current country-specific recommendations and travel characteristics. Regardless of the incidence rate of typhoid in individual countries, European countries have different vaccination recommendations for travellers to those countries, with the need for vaccination based on travel duration, travel destination, circumstances of travelling or a combination of these factors. Interviews with staff members from migrant resource centres have revealed that, paradoxically, migrants may believe that they are at lower risk than non-VFR travellers [34]. This is in part due to a generalised perception of immunity that they hold [34]. Efforts should be increased to target this sub-group of travellers. As typhoid vaccines do not offer complete protection, safe water, sanitation and hygiene interventions are critical to preventing the spread of typhoid, both in endemic settings and among travellers [189]. A multidisciplinary strategy of public health based on personal protection and infrastructure interventions is needed. Conclusion Resistance has become a major threat to the treatment of typhoid, leading to treatment failure and subsequent changes in antimicrobial policy. The increasing prevalence of MDR S. Typhi strains is an important factor in the development of vaccination strategies for the prevention of typhoid infection in high-risk populations. The use of TCVs in endemic regions may be the best defence against MDR S. Typhi [54]. There is a need for worldwide surveillance and access to information on resistant strains, parallel to improvements in antibiotic usage and stewardship. Country-level decision-making and programme planning are critical for local uptake and sustainability of vaccination strategies [122]. National decisions on the preferred vaccination strategy should be based on an analysis of the disease burden and risk factors for transmission, availability and quality of surveillance data, affordability, and operational feasibility. In parallel to vaccination, efforts to improve sanitation should continue. For travellers, harmonisation of recommendations is likely to be of benefit. In particular, greater effort should be made to encourage vaccination for those visiting family and friends. Funding source The manuscript was funded by Emergent BioSolutions. The study sponsors had no role in the writing of the manuscript or the decision to submit the manuscript for publication. Declaration of competing interest The authors declare no conflicts of interest.	2020-12-10T09:06:11.712Z	2020-12-07T00:00:00.000	{ "year": 2020, "sha1": "1505f33eeedd64af19ade14ee74987ba0cce4e64", "oa_license": "CCBY", "oa_url": "https://doi.org/10.1016/j.tmaid.2020.101946", "oa_status": "HYBRID", "pdf_src": "Adhoc", "pdf_hash": "684f8403261d434b9c1b602d067581a7e637149d", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
248976730	pes2o/s2orc	v3-fos-license	Modeling and Experimental Studies on Polymer Melting and Flow in Injection Molding Injection molding, in addition to extrusion, is the most important technology in the polymer processing industry. When modeling injection molding, the global approach is necessary to take into account the solid polymer transport, polymer melting and the polymer melt flow. The model of polymer melting is fundamental for the development of such a global injection molding model. In the paper, the state-of-the-art of modeling and experimentation of the flow and melting in injection molding machines has been presented and discussed. It has been concluded that the existing mathematical models have no strong experimental basis. Therefore, experimentation of the polymer flow and melting in the injection molding machine has been performed, and the effect of processing conditions: the screw speed, the plasticating stroke and the back pressure on the process course has been investigated. Starving in the beginning sections of the screw has been observed, which was not presented in the literature so far. The novel concepts of injection molding modeling have been discussed. Introduction At present, the polymer processing designing is aided by computer process simulations. Modeling allows you to predict the course of the process on the basis of processing conditions, i.e., material characteristics, process operating parameters and machine geometry data. An immense number of different types of polymeric materials and the variety of their applications require the use of various processing methods, among which extrusion and injection molding play a special role. Extrusion is the most mass technology in the polymer processing industry since more than half of all plastics are extruded. Extrusion is also fundamental for compounding, i.e., mixing, pelletizing, filling, reinforcing, etc. Injection molding is suited well for mass production of parts of complex shapes and precise dimensions. About one-third of all thermoplastics are injection molded. Extrusion is the process of continuous forming of polymer products, the so-called profiles, e.g., pipes, rods, plates, films, etc. Extrusion consists of the continuous melting of a polymer in the plasticating unit of the extruder and then pushing (extruding) this melted material, under the pressure developed in the plasticating unit, through the openings of the forming tool that is called the die. Melting occurs as a result of the polymer heating up by the extruder heating system and as a result of energy dissipation in the extruded material. Injection molding is a cyclic process of forming polymer products called moldings. It consists of cyclic melting of a polymer in the plasticating unit of the injection molding machine and then injecting the melted polymer, under the pressure developed in the plasticating unit, into the injection mold, shaping the product. Similar to extrusion, melting takes place here as a result of the polymer heating up by the heating system and as a result of energy dissipation in the material. When modeling extrusion and injection molding, the global approach is necessary to take into account the solid polymer transport, polymer melting and the polymer melt flow. The model of polymer melting is fundamental for the development of such global models. When modeling the polymer melt flow in injection molds, the input data are basically unknown, for example, the polymer temperature. Thus, a global model of the injection molding process is needed that would describe both the flow in the injection molding machine and in the mold. The process output parameters from the plasticating unit would be the input parameters for the mold. The first modeling approaches for screw processing of polymers were made for extrusion. The first analysis of flow in single screw machines were performed by Rowell and Finlayson [2] and later by Carley et al. [3,4], who described the drag flow and pressure flow for an isothermal Newtonian fluid. Solid conveying was modeled first by Darnell and Mol [5], who described material transport and pressure development. The first melting studies in the single screw extruder were carried out by Maddock and Street [6,7], who used the technique of pulling out the screw from an extruder. They analyzed the cross-sections of the polymer removed from the screw and observed the melting mechanism. According to this mechanism, the so-called contiguous solid melting (CSM), a melt layer was formed between the barrel and the solid material, which was scrapped off by the transverse flow in the screw and accumulated at the active screw flight. The solid was gradually diminished by the effects of heat conduction from the barrel and viscous dissipation in the polymer melt ( Figure 2). Thus, two steps of melting were distinguished, the so-called pre-melting zone (or delay zone), corresponding to the formation of the polymer melt layer, and the melting zone (or plasticating zone), corresponding to the accumulation of the polymer melt at the active screw flight. Tadmor et al. [9][10][11] did a similar experiment and proposed the first model of polymer melting in extrusion, which allowed for the elaboration of the first computer model of polymer extrusion EXTRUD [12]. This polymer melting model was based on the computation of the velocity and temperature distribution in the polymer melt film and temperature distribution in the polymer solid bed ( Figure 2). Then, a mass balance in the polymer melt film and the polymer solid, as well as an energy balance at the interface melt film/solid bed, were carried out. Finally, the melting rate was predicted. Later, more detailed models were proposed for solid conveying [13,14] and for the pre-melting zone (or delay zone) [15]. Thus, the extrusion model has been substantially improved. Studies of polymer melting performed by other researchers confirmed the Tadmor model. The issue of modeling the extrusion has been presented in books, for example, by White and Potente [23], Rauwendaal [26], Agassant et al. [28], and has been reviewed in papers by Arrifin et al. [32], Wilczyński et al. [33], Teixeira et al. [34], Malik et al. [35], Wilczyński et al. [36], Hyvärinen et al. [37], Lewandowski et al. [38] and Marschik et al. [39]. Tadmor et al. [9][10][11] did a similar experiment and proposed the first model of polymer melting in extrusion, which allowed for the elaboration of the first computer model of polymer extrusion EXTRUD [12]. This polymer melting model was based on the computation of the velocity and temperature distribution in the polymer melt film and temperature distribution in the polymer solid bed ( Figure 2). Then, a mass balance in the polymer melt film and the polymer solid, as well as an energy balance at the interface melt film/solid bed, were carried out. Finally, the melting rate was predicted. Later, more detailed models were proposed for solid conveying [13,14] and for the pre-melting zone (or delay zone) [15]. Thus, the extrusion model has been substantially improved. Studies of polymer melting performed by other researchers confirmed the Tadmor model. The flood fed single screw extrusion was widely studied and modeled; however, little information was available on the starve fed extrusion. The first studies were carried out by Lopez-Latorre and McKelvey [40] and later by Isherwood et al. [41], Strand et al. [42] and Thompson et al. [43]. Recently, Wilczyński et al. [8,44] performed extensive experimental studies on the starve fed extrusion and proposed the mechanism and model of polymer melting. Then, they built the first computer model of this process, SSEM-Starve [45]. Two stages of melting have been distinguished. In the partially filled section of the screw, the polymer granules were collected at the active screw flight and were melted by conduction, mainly. In the fully filled section, the unmolten solid particles were suspended in the earlier molten polymer, and melting progressed by heat dissipation (Figure 3). This model was later extended to the non-conventional screw configurations [46,47] and to the extrusion of polyblends and composites [48][49][50]. The flood fed single screw extrusion was widely studied and modeled; however, little information was available on the starve fed extrusion. The first studies were carried out by Lopez-Latorre and McKelvey [40] and later by Isherwood et al. [41], Strand et al. [42] and Thompson et al. [43]. Recently, Wilczyński et al. [8,44] performed extensive experimental studies on the starve fed extrusion and proposed the mechanism and model of polymer melting. Then, they built the first computer model of this process, SSEM-Starve [45]. Two stages of melting have been distinguished. In the partially filled section of the screw, the polymer granules were collected at the active screw flight and were melted by conduction, mainly. In the fully filled section, the unmolten solid particles were suspended in the earlier molten polymer, and melting progressed by heat dissipation ( Figure 3). This model was later extended to the non-conventional screw configurations [46,47] and to the extrusion of polyblends and composites [48][49][50]. The polymer flow in the classical flood fed extrusion in comparison with the polymer flow in the starve fed extrusion is depicted in Figure 4. It is worth noting that the difference in the flow rate (extrusion throughput) was very small here, about 2%. However, the filling of the screw was significantly different. A detailed discussion of this issue has been presented in [51] based on the extensive process simulations and experimentations. The research on melting in twin-screw extruders was much more limited. It is worth noting here that twin screw extrusion is performed with metered feeding, mainly, thus with clear starving. The studies were performed mainly for co-rotating twin-screw extruders, both by experimentation, by Bawiskar and White [52], Todd [53], Sakai [54] and Gogos [55][56][57], and by modeling. Potente and Melish [58], as well as Bawiskar and White [59], developed the models based on the classical Tadmor model [20] while assuming the gradual forming of the molten layer from the barrel toward the screw. Bawiskar and White [59] described the formation of two stratified layers of polymer melt in contact with the barrel and solid pellets in contact with the colder screw. Potente and Melisch [58] considered the melting of particles uniformly suspended in the polymer melt. Vergnes et al. [60,61], as well as Zhu et al. [62], developed dispersive models based on the analysis of the flow of the solid/melt mixture of an equivalent viscosity. Based on these investigations, the global models of the co-rotating twin screw extrusion were built [34,[63][64][65][66][67][68]. This issue was reviewed and discussed by Teixeira [69]. Melting in counter-rotating twin-screw extruders is much less known. The first observations were presented by Janssen [70]. White et al. [71,72] noticed that melting occurs here more rapidly than in co-rotating extruders. Wilczyński and White [73] revealed the The polymer flow in the classical flood fed extrusion in comparison with the polymer flow in the starve fed extrusion is depicted in Figure 4. It is worth noting that the difference in the flow rate (extrusion throughput) was very small here, about 2%. However, the filling of the screw was significantly different. A detailed discussion of this issue has been presented in [51] based on the extensive process simulations and experimentations. The flood fed single screw extrusion was widely studied and modeled; however, little information was available on the starve fed extrusion. The first studies were carried out by Lopez-Latorre and McKelvey [40] and later by Isherwood et al. [41], Strand et al. [42] and Thompson et al. [43]. Recently, Wilczyński et al. [8,44] performed extensive experimental studies on the starve fed extrusion and proposed the mechanism and model of polymer melting. Then, they built the first computer model of this process, SSEM-Starve [45]. Two stages of melting have been distinguished. In the partially filled section of the screw, the polymer granules were collected at the active screw flight and were melted by conduction, mainly. In the fully filled section, the unmolten solid particles were suspended in the earlier molten polymer, and melting progressed by heat dissipation ( Figure 3). This model was later extended to the non-conventional screw configurations [46,47] and to the extrusion of polyblends and composites [48][49][50]. The polymer flow in the classical flood fed extrusion in comparison with the polymer flow in the starve fed extrusion is depicted in Figure 4. It is worth noting that the difference in the flow rate (extrusion throughput) was very small here, about 2%. However, the filling of the screw was significantly different. A detailed discussion of this issue has been presented in [51] based on the extensive process simulations and experimentations. The research on melting in twin-screw extruders was much more limited. It is worth noting here that twin screw extrusion is performed with metered feeding, mainly, thus with clear starving. The studies were performed mainly for co-rotating twin-screw extruders, both by experimentation, by Bawiskar and White [52], Todd [53], Sakai [54] and Gogos [55][56][57], and by modeling. Potente and Melish [58], as well as Bawiskar and White [59], developed the models based on the classical Tadmor model [20] while assuming the gradual forming of the molten layer from the barrel toward the screw. Bawiskar and White [59] described the formation of two stratified layers of polymer melt in contact with the barrel and solid pellets in contact with the colder screw. Potente and Melisch [58] considered the melting of particles uniformly suspended in the polymer melt. Vergnes et al. [60,61], as well as Zhu et al. [62], developed dispersive models based on the analysis of the flow of the solid/melt mixture of an equivalent viscosity. Based on these investigations, the global models of the co-rotating twin screw extrusion were built [34,[63][64][65][66][67][68]. This issue was reviewed and discussed by Teixeira [69]. Melting in counter-rotating twin-screw extruders is much less known. The first observations were presented by Janssen [70]. White et al. [71,72] noticed that melting occurs here more rapidly than in co-rotating extruders. Wilczyński and White [73] revealed the The research on melting in twin-screw extruders was much more limited. It is worth noting here that twin screw extrusion is performed with metered feeding, mainly, thus with clear starving. The studies were performed mainly for co-rotating twin-screw extruders, both by experimentation, by Bawiskar and White [52], Todd [53], Sakai [54] and Gogos [55][56][57], and by modeling. Potente and Melish [58], as well as Bawiskar and White [59], developed the models based on the classical Tadmor model [20] while assuming the gradual forming of the molten layer from the barrel toward the screw. Bawiskar and White [59] described the formation of two stratified layers of polymer melt in contact with the barrel and solid pellets in contact with the colder screw. Potente and Melisch [58] considered the melting of particles uniformly suspended in the polymer melt. Vergnes et al. [60,61], as well as Zhu et al. [62], developed dispersive models based on the analysis of the flow of the solid/melt mixture of an equivalent viscosity. Based on these investigations, the global models of the co-rotating twin screw extrusion were built [34,[63][64][65][66][67][68]. This issue was reviewed and discussed by Teixeira [69]. Melting in counter-rotating twin-screw extruders is much less known. The first observations were presented by Janssen [70]. White et al. [71,72] noticed that melting occurs here more rapidly than in co-rotating extruders. Wilczyński and White [73] revealed the mechanism of melting in counter-rotating twin-screw extruders. They observed that melting was initiated both between the screws and at the barrel. The melting between the screws was induced by frictional work on the granules by the calendering stresses between the screws. The melting at the barrel was initiated by the barrel temperature higher than the melting point and was propagated by the viscous dissipation heating of the melt film. These observations ( Figure 5) allowed us to develop models of melting in both of those regions [74]. Further studies of melting were reported by Wang and Min [75,76] as well as by Wilczyński et al. [77]. Based on these studies, the global models of the counter-rotating twin screw extrusion were built [78][79][80][81]. mechanism of melting in counter-rotating twin-screw extruders. They observed that melting was initiated both between the screws and at the barrel. The melting between the screws was induced by frictional work on the granules by the calendering stresses between the screws. The melting at the barrel was initiated by the barrel temperature higher than the melting point and was propagated by the viscous dissipation heating of the melt film. These observations ( Figure 5) allowed us to develop models of melting in both of those regions [74]. Further studies of melting were reported by Wang and Min [75,76] as well as by Wilczyński et al. [77]. Based on these studies, the global models of the counter-rotating twin screw extrusion were built [78][79][80][81]. The methods of investigating and modeling the polymer flow and melting in the extrusion process were adopted to modeling of these in injection molding. However, while many books and review papers have been devoted to the modeling of extrusion, there has been much less to injection molding. These were mainly limited to modeling the flow in injection molds without considering the polymer flow and melting in the plasticating unit, e.g., by Manzione et al. [82], Kennedy and Zheng [83], Kamal et al. [84], Osswald et al. [85] and Wang et al. [86]. Other books have been devoted to the mold design, e.g., by Menges et al. [87], Rees [88], Unger [89], Mennig and Stoeckhert [90], Kazmer [91], Turng and Chen [92], Beaumont [93] and Catoen and Rees [94], or to the injection molding machines by Johannaber [95]. An important review paper on modeling and optimization of injection molding was presented by Fernandes et al. [96]. At present, the important software for simulating the flow in injection molds are MOLDFLOW [97], Moldex3D [98] and CADMOULD [99], by example recently used in [100,101]. Simulations of the polymer flow in injection molds were also performed using the CFD software COMSOL Multiphysics [102] with an example of [103]. The first experimental studies of melting in injection molding machines were carried out by Donovan et al. [104] using the "screw pulling out technique". They revealed that the screw recharge process was the transient plasticating extrusion which gradually approached the equilibrium extrusion behavior as the screw rotated. If the time of screw rotation was a high fraction of the total cycle time, the plasticating behavior was similar to the extrusion behavior, but if the time of screw rotation was a small fraction of the total cycle time, the plasticating behavior was substantially different. The reports were also presented on the experimental study of the solid bed transport and melting in the screw channel of the injection molding machines [105][106][107] with the use of "transparent windows" made in the barrel to observe the polymer flow. Important experimental studies on polymer melting in injection molding machines were carried out by Gao et al. [105]. They developed the visual system to observe the polymer flow in the reciprocating screw injection molding machine. The effects of pro- The methods of investigating and modeling the polymer flow and melting in the extrusion process were adopted to modeling of these in injection molding. However, while many books and review papers have been devoted to the modeling of extrusion, there has been much less to injection molding. These were mainly limited to modeling the flow in injection molds without considering the polymer flow and melting in the plasticating unit, e.g., by Manzione et al. [82], Kennedy and Zheng [83], Kamal et al. [84], Osswald et al. [85] and Wang et al. [86]. Other books have been devoted to the mold design, e.g., by Menges et al. [87], Rees [88], Unger [89], Mennig and Stoeckhert [90], Kazmer [91], Turng and Chen [92], Beaumont [93] and Catoen and Rees [94], or to the injection molding machines by Johannaber [95]. An important review paper on modeling and optimization of injection molding was presented by Fernandes et al. [96]. At present, the important software for simulating the flow in injection molds are MOLDFLOW [97], Moldex3D [98] and CADMOULD [99], by example recently used in [100,101]. Simulations of the polymer flow in injection molds were also performed using the CFD software COMSOL Multiphysics [102] with an example of [103]. The first experimental studies of melting in injection molding machines were carried out by Donovan et al. [104] using the "screw pulling out technique". They revealed that the screw recharge process was the transient plasticating extrusion which gradually approached the equilibrium extrusion behavior as the screw rotated. If the time of screw rotation was a high fraction of the total cycle time, the plasticating behavior was similar to the extrusion behavior, but if the time of screw rotation was a small fraction of the total cycle time, the plasticating behavior was substantially different. The reports were also presented on the experimental study of the solid bed transport and melting in the screw channel of the injection molding machines [105][106][107] with the use of "transparent windows" made in the barrel to observe the polymer flow. Important experimental studies on polymer melting in injection molding machines were carried out by Gao et al. [105]. They developed the visual system to observe the polymer flow in the reciprocating screw injection molding machine. The effects of processing conditions on the polymer melting were studied, and it was concluded that the modeling of the plasticating phase cannot be treated simply as an extrusion process. Different barrel temperatures, plastication strokes, screw rotation speeds, and back pressures were used to observe their effects on the melting behavior. It was observed that there was more polymer melted at a lower screw speed than at a higher speed. As the plastication stroke was the same, a lower speed resulted in longer plastication time, while a higher speed resulted in shorter plastication time. It was also seen that for the case of the lower plastication stroke, the melting was faster, while for the case of the higher plastication stroke, the polymer melting length was longer. With a small injection stroke, the injection volume was small in comparison to the amount of material enclosed in the screw channel. As a result, the residence time was comparatively long, and hence melting was primarily due to heat conduction. Conversely, with a long injection stroke, the injection volume was large in comparison with the screw channel volume, and the melting was primarily due to the shearing. It is known that the back pressure affects the melting, temperature distribution, and pressure distribution in the screw channel. Different back pressures result in different pressure profiles in the screw channel and different plastication times. It was observed that an increase in back pressure was advantageous to the melting. A higher back pressure also led to a larger left-hand melt pool size than a lower back pressure. An increase in the dwelling time results in more heat being conducted and consequently melts more polymer. A comparison between the barrel temperature and the temperature at the interface of the polymer and barrel showed that the polymer-barrel interface temperature was higher than the barrel temperature, even more than 20 • C. This resulted from the combination effects of shear heating at the polymer-barrel interface and the barrel heat loss to the environment. It is worth noting that in injection molding modeling, the polymer temperature at the wall is commonly assumed to be the same as the barrel temperature, which is not an accurate reflection of reality. Modeling of plasticization during injection molding has been a complex problem, particularly for the solid conveying section. This can be further complicated by a breakup of the solid bed due to the reciprocation motion in the process. The break-up of the solid bed can introduce inhomogeneities in the moldings. The solid bed break-up in a reciprocating screw was experimentally studied using a transparent barrel system designed by Jin et al. [106]. It was observed that the break-up occurred in the latter melting stage, where the solid bed strength was reduced because of the continuous reducing the solid bed thickness, where also the melt pressure was high. The onset of break-up was dependent on the process conditions. It was observed that the solid bed broke easier at the higher screw rotation speed and that the longer injection stroke, the higher backpressure and the longer dwell time. The break-up typically started by forming a slit or slits that were nearly perpendicular to the screw flight. It is worth noting that the properly designed barrier screw may be effectively used to prevent solid-bed break-up. Other visual studies were performed by Pham et al. [107]. The results confirmed the usual view of a contiguous solid bed melting process. It was observed that the solid bed width increased with the screw rotation speed, as predicted by a classical theory. The solid bed break-up was seen at every screw speed. Conductive melting from the flight side of the solid bed was also viewed. The solid bed velocity was distinct from the screw velocity. Donovan [108,109] was the first who proposed the model for polymer melting in the injection molding process. This was a heuristic model which required experimental evaluation of an empirical parameter specific to a particular polymer over the studied range of operating conditions. This model was based on the steady-state extrusion model [110]. Lipshitz et al. [111] proposed the model for melting, which was based on the detailed physical mechanisms taking place in the reciprocating screw injection molding machine. This model permitted the computation of the solid bed profile as a function of time during the injection cycle. It consisted of the dynamic extrusion melting model for the rotation period, the transient heat conduction model with a phase transition for the screw rest period, and the model for the drifting of the beginning of melting during the injection cycle. Several papers have been published on the modeling of plasticating in injection molding machines. Rauwendaal [112,113] studied the effects of axial screw movement on solids and melt conveying in reciprocating extruders. Dormeier and Panreck [114] briefly discussed the melting process when the screw stopped in the reciprocating extruder. Rao [115] developed a relatively simple model for describing the plasticating process in a reciprocating injection molding screw by combining the melting model for extrusion of Tadmor (20) with the conduction melting model of Donovan (108). Potente [116] presented the mathematical approach to simulate the polymer melting in injection molding. However, this model did not take into account the solid conveying and transient sections and used some special modeling empirical constants. Later, he calculated the power requirement of the plasticating systems of the injection molding machines and the extruders [117]. Bereaux et al. [118] developed a simple model of throughput and pressure development for single screw extruders and injection molding machines. The model is based on viewing the entire screw simply as a pump, conveying a solid fraction and a molten fraction. This concept results from that during most of the melting process, a conveying of the solid material side by side with the melt takes place. This two-phase conveying can be drawn to the one-phase flow model of a melt pump by assuming that the pressure gradient in the screw channel is solely due to melt flow, but the flow rate at any cross-section can be considered as the sum of the solid bed flow rate and the melt pool flow rate. At that time, no studies existed for modeling the transient melting process in the feeding stage (with rotating screw), the injection stage and the whole cycle of injection molding. Yung et al. [119][120][121] were the first who considered the whole cycle of the process, which comprised of three stages: feeding (screw rotating and moving backward), stopping (no screw movement) and injecting (screw moving forward without rotation) and developed the transient models for the melting process in these. Another comprehensive model of plasticization in injection molding, which reflects well the process dynamics, was developed by Steller and Iwko [122,123]. This model takes into account the periodical action of the screw, to-and-from screw motion with controlled stroke and static and dynamic melting. However, this model does not take into account the existence of a delay zone, uses a simplified melting model, and the equations of energy and momentum are solved analytically. More details of this model were published in [124][125][126]. A few years ago, a novel model for the plasticization process of injection molding was proposed by Fernandes et al. [127]. In this study, the software developed by Gaspar-Cunha [128] for modeling the extrusion process was modified to take into account the backward movement of the screw, the presence of a non-return valve and the heat conduction during the idle time. The model used the Tadmor approach [20] for temperature computation during the plasticating and 3D transient equation of heat conduction for the polymer in the injection chamber during the idle times. The computation of the pressure profile was done by force and torque balances on a differential screw channel element [20]. The forces included friction between the barrel and the solid bed, friction due to the contact of the solids with the screw root and screw walls and normal reactions and forces due to the pressure gradient. The model was used to study the effect of some important process parameters, such as barrel temperature profile, screw speed, backpressure and flow rate during injection and cycle times. An experimental assessment of the computational results was also done. Recently, a comprehensive model of injection molding has been presented by Iwko et al. [129]. The main assumptions of this model are the following: • The process is quasi-steady; • The polymer melting contains two phases: the static melting at the stationary screw and the dynamic melting at the rotating screw with its axial backward movement; • The screw retreat is computed by the equality of the calculated pressure in the front of the screw and the back pressure; • Three zones of the plasticating system are distinguished: the feed port and the solid conveying zone, the delay zone, as well as the melting and melt conveying zone. The starting point for modeling was the model of steady state extrusion, which is similar to the classical extrusion model of Tadmor [20]. However, in contrast to the steady conditions characteristic of extrusion, the lengths and positions of dynamical zones change in time within the injection cycle. So as to describe these changes, it was assumed that two coupled states are valid during the cycle: at the end of screw rotation, i.e., at the beginning of static melting, and at the beginning of screw rotation, i.e., at the beginning of dynamic melting. The model makes it possible to predict the most important process characteristics, such as the solid bed profile, the pressure and temperature distributions, the mass flow rate, the power consumption and the screw torque. The model was verified experimentally for five thermoplastic polymers during the injection molding with different back pressures, screw speeds, dwell times and barrel temperatures. The important output process parameters, such as the temperature and pressure profiles, the power demand by the screw, the torque on the screw and the screw rotation time, were measured. It was found that the model correctly determines the dynamics of the plasticization process under the changes in the most important input parameters. It follows from the above literature data that the development of an adequate and comprehensive model of the melting process in injection molding has not been fully completed so far. It also results from the literature [36] that there exist several commercial software programs for the extrusion process, such as EXTRUD [20], SPR [130], NEXTRUCAD [131], REX [132,133], PASS [134], SSEM [135][136][137], or the research computer models developed by Fukase et al. [138], Zavadsky and Karnis [139], Vincelette et al. [140] and Amellal and Lafleur [141]. However, there is probably one computer program-PSI-available for the analysis of the injection molding process [142]. In summarizing, there are several models of injection molding (plasticating unit) describing the process with different levels of complexity, accuracy and scope of the description of process phenomena. These models were developed mainly based on the experiments and observations performed for the extrusion process and were verified experimentally by selected process parameters measurements, such as pressure and temperature profile, power consumption, etc. However, these were not related to the melting rate or solid bed profile. It is worth noting that most of the research on the polymer flow in extrusion and injection molding is related to polymeric materials in the form of granules. Studies on the polymer melting and flow behavior of polymer powders instead of granules are very limited [143,144]. An early model of Donovan [108] was based on the injection molding studies with the "screw pulling-out technique"; however, the photographs of the screws removed from the barrel were not presented. Thus, there is an obvious lack of experimental studies on the polymer flow and melting in injection molding machines showing the filling of the screw and transport mechanism, especially in the beginning sections of the screw. The reason is that it is difficult to perform such an experiment, and it is especially difficult to quickly remove the screw from the injection molding machine in order to prevent the material from melting and to maintain the structure (state) of the material corresponding to the moment of stopping the machine. Therefore, in this review paper, the experiments have been performed to evaluate the existing concepts of process modeling and to clear the physical basis of modeling. In the experiment, the "screw pulling-out technique" was applied to observe the polymer transport and melting. Based on these experiments, novel concepts of modeling have been proposed. In general, the existing models of the injection molding process (plasticating unit) discussed above differ from the extrusion models in that they involve the static and dynamic phases of melting (stationary and rotating screw) with an axial screw movement. However, it is assumed that the screw is fully filled with a material as in the flood fed extrusion (Figure 2), which is inconsistent with our observation [36] where starvation is clearly seen as in the starve fed extrusion (Figure 3). Material and Process Data The polystyrene Polystyrol 454H (manufactured by BASF, Ludwigshafen, Germany) was used in the research. The basic material data were as follows: the specific density ρ = 1.05 g/cm 3 The injection molding machine FORMOplast 235/80 (manufactured by PONAR, Zywiec, Poland) was applied with the clamping force F clamp = 80 T. A conventional screw of the diameter D = 45 mm, the length L = 540 mm and the screw pitch t = 36 mm was used. The lengths of the screw sections were equal to L F = 305 mm (the feeding sectrion), L C = 135 mm (the compression section), L M = 100 mm (the metering section). The depths of the screw channel in the sections were equal to H F = 10 mm (the feeding section), H M = 3.3 mm (the metering section) and the compression ratio, which is the ratio of the screw channel depth in the feeding section to the screw channel depth in the metering section was equal to CR = H F /H M = 3. Different processing conditions have been applied to study their effect on the polymer flow as well as melting in the injection molding machine. The screw rotation speeds were equal to N = 100 rpm and 300 rpm, the plastication strokes: h plast = 1 D and 3 D, and the back pressures (hydraulic): P back = 0 MPa, 2 MPa and 3 MPa. The temperature was set at T I = 190 • C, T II = 200 • C, T III = 220 • C in the barrel sections, and T nozzle = 230 • C in the nozzle. Results The results of the experiments are depicted in Figures 6-19. An important observation is that in all the cases under study, the screw channel was not fully filled in the beginning section of the screw, and starvation was clearly seen. A yellow mark means the region of starving. clearly seen as in the starve fed extrusion (Figure 3). Material and Process Data The polystyrene Polystyrol 454H (manufactured by BASF, Ludwigshafen, Germany) was used in the research. The basic material data were as follows: the specific density ρ = 1.05 g/cm 3 The injection molding machine FORMOplast 235/80 (manufactured by PONAR, Żywiec, Poland) was applied with the clamping force Fclamp = 80 T. A conventional screw of the diameter D = 45 mm, the length L = 540 mm and the screw pitch t = 36 mm was used. The lengths of the screw sections were equal to LF = 305 mm (the feeding sectrion), LC = 135 mm (the compression section), LM = 100 mm (the metering section). The depths of the screw channel in the sections were equal to HF = 10 mm (the feeding section), HM = 3,3 mm (the metering section) and the compression ratio, which is the ratio of the screw channel depth in the feeding section to the screw channel depth in the metering section was equal to CR = HF/HM = 3. Different processing conditions have been applied to study their effect on the polymer flow as well as melting in the injection molding machine. The screw rotation speeds were equal to N = 100 rpm and 300 rpm, the plastication strokes: hplast = 1 D and 3 D, and the back pressures (hydraulic): Pback = 0 MPa, 2 MPa and 3 MPa. The temperature was set at TI = 190 °C, TII = 200 °C, TIII = 220 °C in the barrel sections, and Tnozzle = 230 °C in the nozzle. Results The results of the experiments are depicted in Figures 6-19. An important observation is that in all the cases under study, the screw channel was not fully filled in the beginning section of the screw, and starvation was clearly seen. A yellow mark means the region of starving. An effect of the screw speed (N = 100 rpm and 300 rpm) on the polymer flow in the injection molding process at the constant plastication stroke and the constant back pressure is shown in Figures 6-10. It is seen that when screw speed increases, the melting slows, which results from the shorter plastication time. It is also seen that the starvation decreases since more polymer is delivered in front of the screw tip. An effect of the plastication stroke (hplast = 1 D, hplast = 3 D) on the polymer flow in the injection molding process at the constant screw speed and the constant back pressure is shown in Figures 11-15. It is seen that when the plastication stroke increases, the starvation increases, too. An effect of the plastication stroke (hplast = 1 D, hplast = 3 D) on the polymer flow in the injection molding process at the constant screw speed and the constant back pressure is shown in Figures 11-15. It is seen that when the plastication stroke increases, the starvation increases, too. An effect of the plastication stroke (hplast = 1 D, hplast = 3 D) on the polymer flow in the injection molding process at the constant screw speed and the constant back pressure is shown in Figures 11-15. It is seen that when the plastication stroke increases, the starvation increases, too. An effect of the back pressure (Pback = 0 MPa, Pback = 2 MPa, Pback = 3 MPa) on the polymer flow in the injection molding process at the constant screw speed and the constant plastication stroke is shown in Figures 16-19. It is seen that when the back pressure increases, the starvation decreases. It can be concluded from the presented experiment that starvation appears in the injection molding process when the screw moves forward, injecting the polymer into the mold. This is dependent on the screw speed, the plastication stroke and the back pressure. The starvation increases when the screw speed and the back pressure decrease, and the plastication stroke increases. It is also clearly seen that melting slows down when screw speed increases which is consistent with Gao's observation [105]. Future Concepts of Injection Molding Process Modeling There are distinct differences between the polymer flow and melting in injection An effect of the screw speed (N = 100 rpm and 300 rpm) on the polymer flow in the injection molding process at the constant plastication stroke and the constant back pressure is shown in Figures 6-10. It is seen that when screw speed increases, the melting slows, which results from the shorter plastication time. It is also seen that the starvation decreases since more polymer is delivered in front of the screw tip. An effect of the plastication stroke (h plast = 1 D, h plast = 3 D) on the polymer flow in the injection molding process at the constant screw speed and the constant back pressure is shown in Figures 11-15. It is seen that when the plastication stroke increases, the starvation increases, too. An effect of the back pressure (P back = 0 MPa, P back = 2 MPa, P back = 3 MPa) on the polymer flow in the injection molding process at the constant screw speed and the constant plastication stroke is shown in Figures 16-19. It is seen that when the back pressure increases, the starvation decreases. It can be concluded from the presented experiment that starvation appears in the injection molding process when the screw moves forward, injecting the polymer into the mold. This is dependent on the screw speed, the plastication stroke and the back pressure. The starvation increases when the screw speed and the back pressure decrease, and the plastication stroke increases. It is also clearly seen that melting slows down when screw speed increases which is consistent with Gao's observation [105]. Future Concepts of Injection Molding Process Modeling There are distinct differences between the polymer flow and melting in injection molding machines and extrusion machines (reciprocating machines vs. non-reciprocating machines). In the typical cycle of injection molding, the screw starts the process in the forward position in the barrel. It rotates (screw recharge) and conveys the polymer forward, and the pressure is developed before the screw tip until the desired volume of the molded part is reached. The screw is then idled in the back position while the polymer previously injected into the mold is being cooled. After cooling the part, the mold opens, and the part is ejected. Then, the mold closes, and the screw moves forward by hydraulic pressure, injecting the newly melted portion of the polymer collected at the tip of the screw into the empty mold, and then starvation appears. The valve, such as the check ring, prevents the back flow during injection. The injection molding cycle is composed of three stages: feeding (screw rotating and moving backward), stopping (no screw movement) and injecting (screw moving forward without rotation). The injection molding process is performed both in the plasticating unit of the injection molding machine and in the mold. To some extent, it is similar to extrusion; however, extrusion is a continuous process in which the extruder cooperates with the die continuously, without any interruptions. Injection molding is a cyclic process, and the cooperation of the injection molding machine (plasticating unit) with the mold is not continuous. However, the flow in the plasticating unit influences the flow in the mold, similarly to the extrusion, where the flow in the extruder (plasticating unit) influences the flow in the die. In the extrusion and injection molding process, the process output parameters in the plasticizing unit are the process inputs for the flow in the die and in the mold, respectively. The conclusion is that the correct modeling of the injection molding process, and the correct modeling of the flow in the injection mold, would require including the modeling of flow in the plasticating unit into the global model of the process. It results from the presented experiment that when injecting the polymer, the starvation appears, and this should be included in the process modeling. This can be done by adopting the concepts developed by Wilczyński et al. [8,36,44] for starve fed single screw extrusion. In general, the models of solid conveying in the classical single screw extrusion with gravitational (flood) feeding are based on the work of Darnell and Mol [5], that is, on the assumption that the polymer granules are transported as a solid bed without relative movement of the granules. However, some experiments, for example, performed by Fang et al. [145], indicated that this cannot be valid any longer since a relative movement of the individual granules in the screw channel is clearly seen. Thus, in order to model the solid transport in the single screw extruder, the discrete element method (DEM) was applied. This was first performed by Potente and Pohl [146], who studied the polymer flow in the hopper region in the extruder. Later, important and fundamental studies were carried out by Moysey and Thompson [147][148][149], who performed the first computations for com-pacting granules and the pressure/throughput relation in the feed zone of the single screw extruder [150]. Further studies of this issue were performed by Schöppner et al. [151,152], who proposed the model which enabled the computation of solid conveying with considering the pressure build-up and the filling degree. It can be concluded that the discrete element method (DEM), which is a useful and powerful tool for modeling the solid conveying in single screw extruders, could also be applied to modeling this in the injection molding process. A good example of the discrete element method software is EDEM [153]. Similarly, as the model of Darnell and Mol [5] for solid conveying, the model of Tadmor and Klein [20] for polymer melting was usually the basis for modeling the extrusion process, as well as the injection molding. However, the models of melting presented so far were mainly based on the a priori assumed melting mechanism. Instead of the melting mechanism assumed, the polymer melting in single screw (non-reciprocating) extruders can be modeled by solving the conservation equations of mass, momentum and energy. This concept was presented first by Viriyayuthakorn and Kassahun [154], who proposed the 3D model without assuming any particular melting mechanism. The phase change problem was solved using the functional dependence of the specific heat on the temperature. The solution of equations of momentum and energy provided the solid/melt distribution, which was determined by the temperature distribution. Syrjala [155] carried out the 2D simulation of the polymer melting without any melting mechanism assumed. However, in both cases, the computations were not validated experimentally. Later, Altinkaynak et al. [156] carried out the experimental and theoretical research on modeling of polymer melting using this approach. The two-phase solid/melt polymer flow was studied with the Cross-WLF rheological model, which allowed us to consider the solid material as a high-viscous fluid while the molten material as a low-viscous fluid. Hopmann et al. [157] solved the equations of momentum and energy using the finite volume method (FVM) with the Carreau rheological model. Kazmer et al. [158] used this approach to modeling the polymer melting in barrier screws, as well as Lewandowski and Wilczyński in conventional screws [159]. This approach can be probably also applied for modeling the polymer melting in the injection molding process, e.g., using the Ansys Polyflow software [160]. It is worth noting that when modeling the injection molding process, the three separated phases of the process have to be considered, i.e., the feeding phase, the idle phase and the injecting phase, and the concepts presented above should be applied separately to these. The promising concept would be the coupled DEM/CFD modeling. Recently, EDEM and Ansys have developed an open-source, two-way coupling between EDEM and Ansys Fluent to provide a unique capability to model particle/fluid flows. Conclusions The issue of modeling of polymer injection molding has been reviewed and discussed, which includes both the plasticating unit and the injection mold. It has been noted that a comprehensive (global) concept of modeling of injection molding might be useful for simulating the polymer flow in injection molds, and the global model of the process might be considered for simulating the polymer flow in the plasticating unit as well as in the mold. The output parameters of the plasticating unit computations would be the input parameters for the mold flow computations. It has also been concluded that the existing mathematical models of injection molding have no strong experimental basis. Thus, experimentation of the polymer flow and melting in the injection molding machine have been performed, and the effect of processing conditions: screw speed, plasticating stroke and back pressure on the process course has been studied. Starvation has been discovered, which was not presented in the literature so far, and which may considerably influence the process modeling. The novel concepts of injection molding modeling have been discussed. The DEM/CFD concept has been indicated as promising. Author Contributions: Conceptualization, K.W. and K.J.W.; methodology, K.W. and K.J.W.; validation, K.B.; writing-original draft preparation, K.W. and K.J.W.; writing-review and editing, K.W.; visualization, K.B. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The data presented in this study are available on request from the corresponding author. Conflicts of Interest: The authors declare no conflict of interest.	2022-05-23T15:13:25.491Z	2022-05-01T00:00:00.000	{ "year": 2022, "sha1": "66b9ff7c93e0d930c0bd89dc60a7e5e2d2197152", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2073-4360/14/10/2106/pdf?version=1653127466", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "977bc22715cb358fa5ab9ed0a1aad8f721ec10bc", "s2fieldsofstudy": [ "Engineering", "Materials Science" ], "extfieldsofstudy": [ "Medicine" ] }
35460023	pes2o/s2orc	v3-fos-license	In-Plane Optical Beam Collimation Using a Three-Dimensional Curved MEMS Mirror The collimation of free-space light propagating in-plane with respect to the substrate is an important performance factor in optical microelectromechanical systems (MEMS). This is usually carried out by integrating micro lenses into the system, which increases the cost of fabrication/assembly in addition to limiting the wavelength working range of the system imposed by the dispersion characteristic of the lenses. In this work we demonstrate optical fiber light collimation using a silicon micromachined three-dimensional curved mirror. Sensitivity to micromachining and fiber alignment tolerance is shown to be low enough by restricting the ratio between the mirror focal length and the optical beam Rayleigh range below 5. The three-dimensional curvature of the mirror is designed to be astigmatic and controlled by a process combining deep, reactive ion etching and isotropic etching of silicon. The effect of the micromachining surface roughness on the collimated beam profile is investigated using a Fourier optics approach for different values of root-mean-squared (RMS) roughness and correlation length. The isotropic etching step of the structure is characterized and optimized for the optical-grade surface requirement. The experimental optical results show a beam-waist ratio of about 4.25 and a corresponding 12-dB improvement in diffraction loss, in good agreement with theory. This type of micromirror can be monolithically integrated into lensless microoptoelectromechanical systems (MOEMS), improving their performance in many different applications. Introduction Optical microelectromechanical systems (MEMS) technology has attracted great attention over the past couple of decades because of its reduced size, light weight and low cost [1]. There are two main architectures in the optical MEMS, namely in-plane architecture [2], where the light propagates from one component to another parallel to the substrate, and out-of-plane architecture [3], where the light hits the optical component either perpendicular to or with inclination on the substrate. For many applications, such as in optical telecommunication [1], optical coherence tomography [4] and on-chip sensing [5], the light source is connected to the optical MEMS device through a single-mode optical fiber, where the optical beam output from the fiber behaves as a Gaussian beam [2]. In this case, the propagation can be associated with beam size expansion before detection, leading to optical losses. This is even more serious in optical MEMS due to the size limit of the optical components [6,7]. Several Reflecting curved micromirrors are achromatic and can provide much a wider spectral response, but they need special attention during fabrication to obtain the curved surface. The common non-planar micro surfaces fabrication techniques are gray-tone mask [12], excimer laser [13], Reactive Ion Etching (RIE) lag effect [14] and photo resist (PR) reflow [15,16]. On one hand, non-silicon curved micromirrors were reported using a polymer dispensing and sucking technique [17], residual internal material stress resulting from deposition of gold on polysilicon for the purpose of light focusing [18], trapping of gas bubbles during melting a stack of small borosilicate glass tubes under a nitrogen atmosphere and further grinding and polishing for atomic studies [19] and deep silicon etching and PR reflow targeting optical interconnects [20]. On the other hand, silicon curved micromirrors fabricated on the wafer top surface were reported using isotropic chemical etching for the sake of optical detection of single atom [21], selective polishing method on the top of MEMS tunable vertical-cavity surface-emitting laser [22] and ion beam irradiation and electrochemical etching for atomic studies as well as optical interconnects [23]. The principal axis of the aforementioned micromirrors is oriented out-of-plane with the respect to the wafer substrate. This rendered the micromirror incompatible with silicon micro-optical bench systems where the light is propagating in-plane with respect to the substrate. Three-dimensional (3-D) micro optical bench systems requiring further assembly or mounting steps after fabrication were introduced in the literature. The most common is to use rotational assembly to create micro-optical subsystems that process free-space beams travelling above the surface of the chip [24]. Non-monolithically integrated mechanical mounting systems for connecting and aligning optical components on a micro optical bench (OB) were also reported [25,26]. This is, however, not compatible with the monolithic integration efforts for the microoptoelectromechanical systems (MOEMS) [27][28][29][30]. In this work, we demonstrate optical beam collimation and propagation loss reduction using a monolithic micromachined curved mirror with an in-plane principal axis, which is compatible with silicon micro-optical bench technology [31]. The paper is organized in the following manner. In Section 2, a theoretical study is carried out for the possibility of Gaussian beam collimation using curved surfaces exhibiting microscale focal lengths, i.e., not so large compared with the incident Gaussian beam Rayleigh range. The design of astigmatic micromirror curvatures is related to incidence angle of the incident Gaussian beam in order to generate a stigmatic collimated beam. The effect of the surface roughness of the micromirror is analyzed in Section 3. Then, the fabrication steps Reflecting curved micromirrors are achromatic and can provide much a wider spectral response, but they need special attention during fabrication to obtain the curved surface. The common non-planar micro surfaces fabrication techniques are gray-tone mask [12], excimer laser [13], Reactive Ion Etching (RIE) lag effect [14] and photo resist (PR) reflow [15,16]. On one hand, non-silicon curved micromirrors were reported using a polymer dispensing and sucking technique [17], residual internal material stress resulting from deposition of gold on polysilicon for the purpose of light focusing [18], trapping of gas bubbles during melting a stack of small borosilicate glass tubes under a nitrogen atmosphere and further grinding and polishing for atomic studies [19] and deep silicon etching and PR reflow targeting optical interconnects [20]. On the other hand, silicon curved micromirrors fabricated on the wafer top surface were reported using isotropic chemical etching for the sake of optical detection of single atom [21], selective polishing method on the top of MEMS tunable vertical-cavity surface-emitting laser [22] and ion beam irradiation and electrochemical etching for atomic studies as well as optical interconnects [23]. The principal axis of the aforementioned micromirrors is oriented out-of-plane with the respect to the wafer substrate. This rendered the micromirror incompatible with silicon micro-optical bench systems where the light is propagating in-plane with respect to the substrate. Three-dimensional (3-D) micro optical bench systems requiring further assembly or mounting steps after fabrication were introduced in the literature. The most common is to use rotational assembly to create micro-optical subsystems that process free-space beams travelling above the surface of the chip [24]. Non-monolithically integrated mechanical mounting systems for connecting and aligning optical components on a micro optical bench (OB) were also reported [25,26]. This is, however, not compatible with the monolithic integration efforts for the microoptoelectromechanical systems (MOEMS) [27][28][29][30]. In this work, we demonstrate optical beam collimation and propagation loss reduction using a monolithic micromachined curved mirror with an in-plane principal axis, which is compatible with silicon micro-optical bench technology [31]. The paper is organized in the following manner. In Section 2, a theoretical study is carried out for the possibility of Gaussian beam collimation using curved surfaces exhibiting microscale focal lengths, i.e., not so large compared with the incident Gaussian beam Rayleigh range. The design of astigmatic micromirror curvatures is related to incidence angle of the incident Gaussian beam in order to generate a stigmatic collimated beam. The effect of the surface roughness of the micromirror is analyzed in Section 3. Then, the fabrication steps of the micromirror and the resulting structure are presented in Section 4. Finally, optical measurements are presented and discussed in Section 5 using the introduced curved micromirror for single-mode fiber output collimation and propagation loss reduction where the fiber axis lies in-plane with the substrate. Theoretical Analysis of Optical Beam Collimation Consider the incidence of a Gaussian beam on a curved micromirror as shown in Figure 2. The parameters of the reflected beam are related to the incident beam by: where w in and w out are the min waist radii for the incident and reflected beams, respectively, d in and d out are the distances between the beam waist location and the mirror surface at the point of incidence for the incident and reflected beams, respectively, f is the focal length of the mirror and z o is the Rayleigh range of the incident beam. The beam-waist ratio w out /w in is denoted by G c and represents the collimation gain. The dependences of the beam-waist ratio and the ratio d out /f on the ratio d in /f for different ratios of f /z o are shown in Figure 3. The beam-waist ratio has a maximum value occurring when the input distance and the focal length are equal. The maximum beam-waist ratio is given by: The variation of the beam-waist ratio around d in /f = 1 is symmetric. The variation of the ratio d out /f possess odd symmetry around the point (d in /f = 1, d out /f = 1). The output beam waist location doesn't change with the input beam Rayleigh range when the input beam waist is located at the focus of the mirror. Negative values of d out /f are obtained when d in /f < 1, which means the output beam waist is located virtually behind the mirror and the beam is diverging after reflection. The opposite case occurs when d in /f >1 and the beam is reflected in a converging state. The output beam waist may have its waist located just at the mirror surface for a single value of d in /f when z o /f = 2 and for two value of d in /f when z o /f = 0.5; one time for a very small value of d in /f and the second time for a d in /f that is slightly smaller than unity. of the micromirror and the resulting structure are presented in Section 4. Finally, optical measurements are presented and discussed in Section 5 using the introduced curved micromirror for single-mode fiber output collimation and propagation loss reduction where the fiber axis lies in-plane with the substrate. Theoretical Analysis of Optical Beam Collimation Consider the incidence of a Gaussian beam on a curved micromirror as shown in Figure 2. The parameters of the reflected beam are related to the incident beam by: where win and wout are the min waist radii for the incident and reflected beams, respectively, din and dout are the distances between the beam waist location and the mirror surface at the point of incidence for the incident and reflected beams, respectively, f is the focal length of the mirror and zo is the Rayleigh range of the incident beam. The beam-waist ratio wout/win is denoted by Gc and represents the collimation gain. The dependences of the beam-waist ratio and the ratio dout/f on the ratio din/f for different ratios of f/zo are shown in Figure 3. The beam-waist ratio has a maximum value occurring when the input distance and the focal length are equal. The maximum beam-waist ratio is given by: The microfabrication process tolerance may result in a variation of the curved micromirror radius of curvature, which affects the obtainable beam's beam-waist ratio. The impact depends on the gain sensitivity to the curved surface focal length. The corresponding change is determined by: For a given percentage change in the focal length, the gain sensitivity becomes very high when the ratio z o /f is very small. As depicted in Figure 4a, the beam-waist ratio is less sensitive to the focal length variation when z o /f is larger than 0.2. The output beam waist location is, however, very sensitive to the variations as shown in Figure 4b. In the case of z o /f > 0.2, the fabrication tolerance impact on the output beam waist location can be compensated by active axial alignment. The variation of the beam-waist ratio around din/f = 1 is symmetric. The variation of the ratio dout/f possess odd symmetry around the point (din/f = 1, dout/f = 1). The output beam waist location doesn't change with the input beam Rayleigh range when the input beam waist is located at the focus of the mirror. Negative values of dout/f are obtained when din/f < 1, which means the output beam waist is located virtually behind the mirror and the beam is diverging after reflection. The opposite case occurs when din/f >1 and the beam is reflected in a converging state. The output beam waist may have its waist located just at the mirror surface for a single value of din/f when zo/f = 2 and for two value of din/f when zo/f = 0.5; one time for a very small value of din/f and the second time for a din/f that is slightly smaller than unity. The microfabrication process tolerance may result in a variation of the curved micromirror radius of curvature, which affects the obtainable beam's beam-waist ratio. The impact depends on the gain sensitivity to the curved surface focal length. The corresponding change is determined by: For a given percentage change in the focal length, the gain sensitivity becomes very high when the ratio zo/f is very small. As depicted in Figure 4a, the beam-waist ratio is less sensitive to the focal length variation when zo/f is larger than 0.2. The output beam waist location is, however, very sensitive to the variations as shown in Figure 4b. In the case of zo/f > 0.2, the fabrication tolerance impact on the output beam waist location can be compensated by active axial alignment. The inclined incidence of the beam on the mirror in a tangential plane, while being normal to the sagittal plane, has the effect of splitting the focal length as well as the input ratio din/f of the mirror each into two different values: where the subscripts "ip" and "op" are used for the in-plane and out-of-plane directions, The inclined incidence of the beam on the mirror in a tangential plane, while being normal to the sagittal plane, has the effect of splitting the focal length as well as the input ratio d in /f of the mirror each into two different values: where the subscripts "ip" and "op" are used for the in-plane and out-of-plane directions, respectively, and R is the radius of curvature of the mirror in the indicated plane. The inclined incidence has the effect of effectively increasing the out-of-plane focal length of the curved surface while at the same time decreasing its in-plane focal length, and therefore, a stigmatic inclined curved surface should have non-equal radii of curvature in the two orthogonal planes. As will be shown in the fabrication section, the out-of-plane plane radius of curvature can be limited to 100 µm. Fortunately, increasing angle of incidence compensates for this limit. For instance, focal length matching occurs at incidence angles θ inc = 0 • , 45 • and 60 • for R op /R ip = 1, 0.5 and 0.25 respectively. Away from the stigmatic beam generation angle, the reflected beam exhibits an elliptical cross section as well different beam waist location in the two orthogonal planes. This can be of particular interest in beam shaping/matching applications. Effect of Surface Roughness The effect of the surface roughness expected from the micromachining of the 3-D curved surface on the collimated optical beam profile is investigated in this section. For this purpose, the overall phase transformation of the 3-D mirror is divided into the phase curvature responsible for the collimation of the beam, which is already considered in Section 2, and a random phase due to the surface roughness. The phase curvature corresponding to the curvature of the mirror surface is given by: where f is the equivalent focal length of the mirror. The random phase is given by: where z n = f(x,y) is the random height variation of the surface due to the surface roughness. In our analysis, f(x,y) is assumed a random rough surface that has a Gaussian height distribution function and Gaussian autocovariance functions (in both xand y-direction). The surface is assumed to have an RMS height σ rms and assumed to be isotropic in the sense that the correlation length L c in the xand y-direction are assumed equal. The simulation procedure is carried out using the Fourier optics approach as follows [32]. The field at the mirror surface, denoted by E in (x,y,d in ), is multiplied by the phase transformation function and the new output field is denoted by E o (x,y,d in ) : A fast Fourier transform (FFT) is applied to get this output field in the spatial frequency domain: The field is propagated a distance d out by phase multiplication in the spatial frequency domain: where k z is the axial components of the wave vector. Finally, the output field profile after propagating the distance d out is obtained by inverse Fourier transform: A simulation study was carried out to analyze the effect of the surface roughness of the etched mirror on the collimated beam. The effect is evaluated by calculating the coupling efficiency (overlap integral) between the resulting and the ideal beam. The radius of curvature of the mirror in the in-plane direction is assumed 300 µm, while the out-of-plane radius of curvature is 150 µm, similar to the value obtained practically as will be shown in the next section. The incident beam has a minimum waist radius of 5 µm, a wavelength of 1550 nm and located at the focal plane of the mirror in a 45-degree incidence orientation. The RMS roughness σ rms is assumed in the range of 0 to λ/10. Three values of the correlation were assumed: 5λ, 10λ and 20λ. The resulting coupling efficiency is depicted in Figure 5a. Since the roughness generation is a stochastic process, the simulation was repeated 20 times for each point and the average was taken. The coupling efficiency decreases with the increase of the RMS value of the roughness, as expected. It reaches about 75% for the case of L c = 10λ and σ rms = λ/10. If we would like to maintain at least 95% of the coupling efficiency, then σ rms should be less than 0.04λ, 0.06λ and 0.1λ for L c = 5λ, 10λ and L c = 20λ, respectively. Example resulting beam profiles for the case of σ rms = 0.1λ are shown in Figure 5b. The x-axis is normalized to the waist of the resulting beam profile in case of The loss in efficiency is resulting from the asymmetry in the beam profile in addition to the widening of the profiles out of the ±4w limit due to the surface roughness. in a 45-degree incidence orientation. The RMS roughness σrms is assumed in the range of 0 to λ/10. Three values of the correlation were assumed: 5λ, 10λ and 20λ. The resulting coupling efficiency is depicted in Figure 5a. Since the roughness generation is a stochastic process, the simulation was repeated 20 times for each point and the average was taken. The coupling efficiency decreases with the increase of the RMS value of the roughness, as expected. It reaches about 75% for the case of Lc = 10λ and σrms = λ/10. If we would like to maintain at least 95% of the coupling efficiency, then σrms should be less than 0.04λ, 0.06λ and 0.1λ for Lc = 5λ, 10λ and Lc = 20λ, respectively. Example resulting beam profiles for the case of σrms = 0.1λ are shown in Figure 5b. The x-axis is normalized to the waist of the resulting beam profile in case of The loss in efficiency is resulting from the asymmetry in the beam profile in addition to the widening of the profiles out of the ±4w limit due to the surface roughness. Silicon Micromirror Fabrication The optical axis of the target 3-D curved micromirror lies in-plane with respect to the wafer substrate to collimate the optical beam generated from single-mode optical fibers located horizontally on the wafer substrate or any other light source integrated in the system. It enables the use of the fiber-mirror configuration to replace the lensed fiber as previously shown in Figure 1d. The fabrication of the micromirror was carried out into six main steps [33]. First the definition of the in-plane profile of the micromirror with a 300-μm radius of curvature was performed using standard photolithography (see top view in Figure 6a). The lithographic process ends with a patterned SiO2 mask layer for the following etching. Second, anisotropic deep reactive ion etching of the silicon was carried out, ending with a deeply etched cylindrical surface as shown in Figure 6b [34]. By this anisotropic etching step, the central line of the out-of-plane curvature (principal axis) is defined. The axis depth with respect to the wafer top surface was chosen to be large enough that optical fiber can be inserted and aligned with micromirror. Then, side wall protection was carried out using a Teflon-like layer to prevent sidewall etching from top and ensure the following isotropic etching starts at the mirror principal axis as shown in Figure 6c. The protection step was followed by a long isotropic etching step using SF6 plasma to achieve the desired out-of-plane profile of the micromirror as shown in Figure 6d, in a similar way to that used to fabricate micro fluidic channels reported in [35]. The out-of-plane radius of curvature of the micromirror surface is about 150 μm. Achieving larger radii of curvatures requires deeper etching, which may result in a fragile wafer. The protective layer was removed in the fifth step as shown in Figure 6e using a high-temperature oxygen plasma ashing process. As will be shown below, the resulting surface roughness was about 22 nm RMS. Therefore, the surface was post-processed for optical quality requirement by smoothing Silicon Micromirror Fabrication The optical axis of the target 3-D curved micromirror lies in-plane with respect to the wafer substrate to collimate the optical beam generated from single-mode optical fibers located horizontally on the wafer substrate or any other light source integrated in the system. It enables the use of the fiber-mirror configuration to replace the lensed fiber as previously shown in Figure 1d. The fabrication of the micromirror was carried out into six main steps [33]. First the definition of the in-plane profile of the micromirror with a 300-µm radius of curvature was performed using standard photolithography (see top view in Figure 6a). The lithographic process ends with a patterned SiO 2 mask layer for the following etching. Second, anisotropic deep reactive ion etching of the silicon was carried out, ending with a deeply etched cylindrical surface as shown in Figure 6b [34]. By this anisotropic etching step, the central line of the out-of-plane curvature (principal axis) is defined. The axis depth with respect to the wafer top surface was chosen to be large enough that optical fiber can be inserted and aligned with micromirror. Then, side wall protection was carried out using a Teflon-like layer to prevent sidewall etching from top and ensure the following isotropic etching starts at the mirror principal axis as shown in Figure 6c. The protection step was followed by a long isotropic etching step using SF 6 plasma to achieve the desired out-of-plane profile of the micromirror as shown in Figure 6d, in a similar way to that used to fabricate micro fluidic channels reported in [35]. The out-of-plane radius of curvature of the micromirror surface is about 150 µm. Achieving larger radii of curvatures requires deeper etching, which may result in a fragile wafer. The protective layer was removed in the fifth step as shown in Figure 6e using a high-temperature oxygen plasma ashing process. As will be shown below, the resulting surface roughness was about 22 nm RMS. Therefore, the surface was post-processed for optical quality requirement by smoothing and Aluminum metallization as shown in Figure 6f. Top and tilted views of the fabricated micromirror after step 5 are shown in Figure 7a,b, recorded using a scanning electron microscope (SEM). and Aluminum metallization as shown in Figure 6f. Top and tilted views of the fabricated micromirror after step 5 are shown in Figure 7a,b, recorded using a scanning electron microscope (SEM). More than one effect was encountered regarding the isotropic etching of silicon using SF6. First, a significant dependence of the etch rate on the trench width was observed, as shown in Figure 8. The etch rate is normalized with respect to the etch rate of the largest trench width. The data markers represent the measured normalized data while the solid line is a logarithmic fitting. This kind of logarithmic behavior is well-known for a diffusion-limited etching process [14]. The etch rate for a 10 and Aluminum metallization as shown in Figure 6f. Top and tilted views of the fabricated micromirror after step 5 are shown in Figure 7a,b, recorded using a scanning electron microscope (SEM). More than one effect was encountered regarding the isotropic etching of silicon using SF6. First, a significant dependence of the etch rate on the trench width was observed, as shown in Figure 8. The etch rate is normalized with respect to the etch rate of the largest trench width. The data markers represent the measured normalized data while the solid line is a logarithmic fitting. This kind of More than one effect was encountered regarding the isotropic etching of silicon using SF 6 . First, a significant dependence of the etch rate on the trench width was observed, as shown in Figure 8. The etch rate is normalized with respect to the etch rate of the largest trench width. The data markers represent the measured normalized data while the solid line is a logarithmic fitting. This kind of logarithmic behavior is well-known for a diffusion-limited etching process [14]. The etch rate for a 10 µm trench width is about one fifth the rate for a 500 µm trench width. The second observation is the correlation between the mask opening width and the isotropic etching roughness as shown in Figure 9. The smaller the mask opening is, the higher the roughness. Considerable roughness can be observed in the smallest opening by inspecting the SEM images with the naked eye, while the roughness in the largest opening is much less, but still observable. The atomic force microscope (AFM) was used in order to get a quantitative measurement for the roughness of the largest opening. The top and 3-D tilted views of the surface topology, obtained using the AFM on an area of 10 µm by 10 µm, are shown in Figure 10a,b respectively. The measured roughness has a peak of 319 nm, an average of 16 nm and an RMS 22 nm. The lag effect as well as the surface roughness of the isotropic etching roughness can be interpreted knowing that a diffusion process governs the transport of the etching radicals from the plasma, where it is created, to the substrate, where chemical etching occurs. Due to this diffusion process, a lower amount of etchants is received in thinner trenches. This directly relates to the lag effect. At the same time, when the amount of etchants is not enough, a rough surface results from the etching process because the surface is not overwhelmed by the etchants. correlation between the mask opening width and the isotropic etching roughness as shown in Figure 9. The smaller the mask opening is, the higher the roughness. Considerable roughness can be observed in the smallest opening by inspecting the SEM images with the naked eye, while the roughness in the largest opening is much less, but still observable. The atomic force microscope (AFM) was used in order to get a quantitative measurement for the roughness of the largest opening. The top and 3-D tilted views of the surface topology, obtained using the AFM on an area of 10 μm by 10 μm, are shown in Figure 10a,b respectively. The measured roughness has a peak of 319 nm, an average of 16 nm and an RMS 22 nm. The lag effect as well as the surface roughness of the isotropic etching roughness can be interpreted knowing that a diffusion process governs the transport of the etching radicals from the plasma, where it is created, to the substrate, where chemical etching occurs. Due to this diffusion process, a lower amount of etchants is received in thinner trenches. This directly relates to the lag effect. At the same time, when the amount of etchants is not enough, a rough surface results from the etching process because the surface is not overwhelmed by the etchants. correlation between the mask opening width and the isotropic etching roughness as shown in Figure 9. The smaller the mask opening is, the higher the roughness. Considerable roughness can be observed in the smallest opening by inspecting the SEM images with the naked eye, while the roughness in the largest opening is much less, but still observable. The atomic force microscope (AFM) was used in order to get a quantitative measurement for the roughness of the largest opening. The top and 3-D tilted views of the surface topology, obtained using the AFM on an area of 10 μm by 10 μm, are shown in Figure 10a,b respectively. The measured roughness has a peak of 319 nm, an average of 16 nm and an RMS 22 nm. The lag effect as well as the surface roughness of the isotropic etching roughness can be interpreted knowing that a diffusion process governs the transport of the etching radicals from the plasma, where it is created, to the substrate, where chemical etching occurs. Due to this diffusion process, a lower amount of etchants is received in thinner trenches. This directly relates to the lag effect. At the same time, when the amount of etchants is not enough, a rough surface results from the etching process because the surface is not overwhelmed by the etchants. Measurement Results and Discussion In this section, the manufactured 3-D curved micromirror is utilized for collimating the output beam of single-mode fibers and propagation loss reduction thereof. Consider the arrangement shown in Figure 11. A single-mode optical fiber is inserted on the silicon substrate such that its optical axis is parallel to the silicon substrate and tilted with respect to the mirror principal axis. For the sake of optical spot characterization, the reflected beam is captured in the far field on a scanning-slit beam profiler. The observed beam ellipticity, defined by the ratio of the spot size in the in-plane direction to the out-of-plane direction, is adjusted to be close to unity (about 1.05) by letting the incidence angle of the beam on the mirror be about 45°. The axial distance between the optical fiber and the mirror was adjusted such that the fiber tip is located at the micromirror focal plane by minimizing the observed output beam diameter at the far field. The collimated output beam spot diameter was measured at different locations away from the micromirror and compared to the measurements of the optical fiber output beam without using the micromirror. In the case of using a standard single-mode fiber with a core radius of 4.5 μm fed from 1550 nm laser source, a reduction in the divergence angle of the beam by a factor of 2 was achieved by the micromirror. The output beam has a minimum waist radius of about 10 μm, which is a typical value for many optical MEMS applications. A typical captured beam profile at one location d is shown in Figure 12a. The profile was fitted to a Gaussian profile with average root mean square errors smaller than 1% and 1.5% in the x-and y-directions respectively as shown in Figure 12b,c. This is an indication of the good performance offered by the fabricated micromirror, using the presented method, in terms of its phase front transformation function. This experiment was repeated with a special single-mode fiber with a core radius of 2 μm working at a 675 nm wavelength. The special Measurement Results and Discussion In this section, the manufactured 3-D curved micromirror is utilized for collimating the output beam of single-mode fibers and propagation loss reduction thereof. Consider the arrangement shown in Figure 11. Measurement Results and Discussion In this section, the manufactured 3-D curved micromirror is utilized for collimating the output beam of single-mode fibers and propagation loss reduction thereof. Consider the arrangement shown in Figure 11. A single-mode optical fiber is inserted on the silicon substrate such that its optical axis is parallel to the silicon substrate and tilted with respect to the mirror principal axis. For the sake of optical spot characterization, the reflected beam is captured in the far field on a scanning-slit beam profiler. The observed beam ellipticity, defined by the ratio of the spot size in the in-plane direction to the out-of-plane direction, is adjusted to be close to unity (about 1.05) by letting the incidence angle of the beam on the mirror be about 45°. The axial distance between the optical fiber and the mirror was adjusted such that the fiber tip is located at the micromirror focal plane by minimizing the observed output beam diameter at the far field. The collimated output beam spot diameter was measured at different locations away from the micromirror and compared to the measurements of the optical fiber output beam without using the micromirror. In the case of using a standard single-mode fiber with a core radius of 4.5 μm fed from 1550 nm laser source, a reduction in the divergence angle of the beam by a factor of 2 was achieved by the micromirror. The output beam has a minimum waist radius of about 10 μm, which is a typical value for many optical MEMS applications. A typical captured beam profile at one location d is shown in Figure 12a. The profile was fitted to a Gaussian profile with average root mean square errors smaller than 1% and 1.5% in the x-and y-directions respectively as shown in Figure 12b,c. This is an indication of the good performance offered by the fabricated micromirror, using the presented method, in terms of its phase front transformation function. This experiment was repeated with a special single-mode fiber with a core radius of 2 μm working at a 675 nm wavelength. The special A single-mode optical fiber is inserted on the silicon substrate such that its optical axis is parallel to the silicon substrate and tilted with respect to the mirror principal axis. For the sake of optical spot characterization, the reflected beam is captured in the far field on a scanning-slit beam profiler. The observed beam ellipticity, defined by the ratio of the spot size in the in-plane direction to the out-of-plane direction, is adjusted to be close to unity (about 1.05) by letting the incidence angle of the beam on the mirror be about 45 • . The axial distance between the optical fiber and the mirror was adjusted such that the fiber tip is located at the micromirror focal plane by minimizing the observed output beam diameter at the far field. The collimated output beam spot diameter was measured at different locations away from the micromirror and compared to the measurements of the optical fiber output beam without using the micromirror. In the case of using a standard single-mode fiber with a core radius of 4.5 µm fed from 1550 nm laser source, a reduction in the divergence angle of the beam by a factor of 2 was achieved by the micromirror. The output beam has a minimum waist radius of about 10 µm, which is a typical value for many optical MEMS applications. A typical captured beam profile at one location d is shown in Figure 12a. The profile was fitted to a Gaussian profile with average root mean square errors smaller than 1% and 1.5% in the xand ydirections respectively as shown in Figure 12b,c. This is an indication of the good performance offered by the fabricated micromirror, using the presented method, in terms of its phase front transformation function. This experiment was repeated with a special single-mode fiber with a core radius of 2 µm working at a 675 nm wavelength. The special fiber is positioned at the same location used for the standard one because of the constant focal length of the mirror independent of the wavelength value. A reduction in the divergence angle of the beam by a factor of 4.25 was achieved. The resulting output beam has a minimum waist radius of about 10 µm as well. This visible beam will be used hereinafter for evaluating the propagation loss reduction offered by the micromirror. fiber is positioned at the same location used for the standard one because of the constant focal length of the mirror independent of the wavelength value. A reduction in the divergence angle of the beam by a factor of 4.25 was achieved. The resulting output beam has a minimum waist radius of about 10 μm as well. This visible beam will be used hereinafter for evaluating the propagation loss reduction offered by the micromirror. The collimation of the beam by the micromirror was also evaluated by measuring the detected power in free space with a limited-aperture detector as shown in Figure 13. Theoretically, the transmitted power in terms of the system aperture radius a and the beam spot radius at the detector is given by [36]: The power collected by a detector with 3.5 mm aperture radius is shown in Figure 14a. The power was measured at different distance d in the far field away from the beam waist. The measurements were carried out one time for the collimated beam by the micromirror, denoted by Pc, and another time for beam originally emitted by the single-mode fiber, denoted by Po. The experimental data are depicted using markers while the theoretical data are depicted using lines. The power is normalized with respect to the initially maximum power. The measured power clearly starts to fall when the beam diameter starts to exceed the detector aperture as given by Equation (9). The micromirror significantly reduces the propagation losses with respect to the original fiber output. The detected power from the micromirror has a slower roll-off and drops to half its maximum value 25 cm far from the micromirror compared to less than 8 cm without using the micromirror. The ratio between the two detected powers is depicted in Figure 14b, where the improvement reaches about 11-12 dB. Indeed, in the far field, the ratio between the detected powers is given by: The collimation of the beam by the micromirror was also evaluated by measuring the detected power in free space with a limited-aperture detector as shown in Figure 13. fiber is positioned at the same location used for the standard one because of the constant focal length of the mirror independent of the wavelength value. A reduction in the divergence angle of the beam by a factor of 4.25 was achieved. The resulting output beam has a minimum waist radius of about 10 μm as well. This visible beam will be used hereinafter for evaluating the propagation loss reduction offered by the micromirror. The collimation of the beam by the micromirror was also evaluated by measuring the detected power in free space with a limited-aperture detector as shown in Figure 13. Theoretically, the transmitted power in terms of the system aperture radius a and the beam spot radius at the detector is given by [36]: The power collected by a detector with 3.5 mm aperture radius is shown in Figure 14a. The power was measured at different distance d in the far field away from the beam waist. The measurements were carried out one time for the collimated beam by the micromirror, denoted by Pc, and another time for beam originally emitted by the single-mode fiber, denoted by Po. The experimental data are depicted using markers while the theoretical data are depicted using lines. The power is normalized with respect to the initially maximum power. The measured power clearly starts to fall when the beam diameter starts to exceed the detector aperture as given by Equation (9). The micromirror significantly reduces the propagation losses with respect to the original fiber output. The detected power from the micromirror has a slower roll-off and drops to half its maximum value 25 cm far from the micromirror compared to less than 8 cm without using the micromirror. The ratio between the two detected powers is depicted in Figure 14b, where the improvement reaches about 11-12 dB. Indeed, in the far field, the ratio between the detected powers is given by: Theoretically, the transmitted power in terms of the system aperture radius a and the beam spot radius at the detector is given by [36]: The power collected by a detector with 3.5 mm aperture radius is shown in Figure 14a. The power was measured at different distance d in the far field away from the beam waist. The measurements were carried out one time for the collimated beam by the micromirror, denoted by P c , and another time for beam originally emitted by the single-mode fiber, denoted by P o . The experimental data are depicted using markers while the theoretical data are depicted using lines. The power is normalized with respect to the initially maximum power. The measured power clearly starts to fall when the beam diameter starts to exceed the detector aperture as given by Equation (9). The micromirror significantly reduces the propagation losses with respect to the original fiber output. The detected power from the micromirror has a slower roll-off and drops to half its maximum value 25 cm far from the micromirror compared to less than 8 cm without using the micromirror. The ratio between the two detected powers is depicted in Figure 14b, where the improvement reaches about 11-12 dB. Indeed, in the far field, the ratio between the detected powers is given by: where the beam spot radius in the far field was replaced by wd/z o = d/θ div . The maximum improvement is achieved when the spot radius becomes much larger than the detector aperture. In this case, Taylor expansion of the exponential terms can be applied to second order and Equation (16) becomes: The maximum power gain due to the usage of the collimating mirror is given by the beam-waist ratio squared. For the fabricated micromirror and using the single-mode fiber at 675 nm, the power gain is G p = (4.25) 2 = 18 that is about 12.5 dB, in good agreement with the measured data. This value is independent of the specific sizes of the beam spot and the detector aperture, as long as significant truncation loss is encountered. (16) where the beam spot radius in the far field was replaced by wd/zo = d/θdiv. The maximum improvement is achieved when the spot radius becomes much larger than the detector aperture. In this case, Taylor expansion of the exponential terms can be applied to second order and Equation (16) becomes: Figure 14. (a) The normalized power collected by the detector; (b) diffraction loss reduction in dB using the collimating micromirror. The measured data is given in markers when the theoretical one is given in lines. The maximum power gain due to the usage of the collimating mirror is given by the beam-waist ratio squared. For the fabricated micromirror and using the single-mode fiber at 675 nm, the power gain is Gp = (4.25) 2 = 18 that is about 12.5 dB, in good agreement with the measured data. This value is independent of the specific sizes of the beam spot and the detector aperture, as long as significant truncation loss is encountered. Conclusions Optical beam collimation was analyzed and successfully carried out using a micro-reflector with a three-dimensional curved surface. The surface was etched in silicon by a technique combining deep reactive ion etching and isotropic etching technologies. The produced surface is astigmatic with an out-of-plane radius of curvature that is about half the in-plane radius of curvature. Having the incident beam in-plane and inclined by 45° with respect to the principal axis, the reflected beam is kept stigmatic with about a 4.25-fold reduction in the beam expansion angle in free space and about 12-dB reduction in propagation losses. The fibre-mirror configuration may serve as a potential replacement for the lensed fibers widely used in the MOEMS system. This replacement has the advantage of producing monolithically integrated systems with a wider-band spectral response. Author Contributions: Y.M.S. carried out the theoretical analysis, studied the roughness effect using optical simulations, fabricated the structure, performed the experiments and wrote the manuscript. D.A. participated in the design of the optical setup and the method of roughness simulation. B.S. participated in the idea and design of the fabrication steps. T.B. participated in the idea, revised the paper and supervised the overall work. Conflicts of Interest: The authors declare no conflict of interest. Figure 14. (a) The normalized power collected by the detector; (b) diffraction loss reduction in dB using the collimating micromirror. The measured data is given in markers when the theoretical one is given in lines. Conclusions Optical beam collimation was analyzed and successfully carried out using a micro-reflector with a three-dimensional curved surface. The surface was etched in silicon by a technique combining deep reactive ion etching and isotropic etching technologies. The produced surface is astigmatic with an out-of-plane radius of curvature that is about half the in-plane radius of curvature. Having the incident beam in-plane and inclined by 45 • with respect to the principal axis, the reflected beam is kept stigmatic with about a 4.25-fold reduction in the beam expansion angle in free space and about 12-dB reduction in propagation losses. The fibre-mirror configuration may serve as a potential replacement for the lensed fibers widely used in the MOEMS system. This replacement has the advantage of producing monolithically integrated systems with a wider-band spectral response. Author Contributions: Y.M.S. carried out the theoretical analysis, studied the roughness effect using optical simulations, fabricated the structure, performed the experiments and wrote the manuscript. D.A. participated in the design of the optical setup and the method of roughness simulation. B.S. participated in the idea and design of the fabrication steps. T.B. participated in the idea, revised the paper and supervised the overall work. Conflicts of Interest: The authors declare no conflict of interest.	2017-07-30T02:31:33.519Z	2017-04-25T00:00:00.000	{ "year": 2017, "sha1": "c2cb3c78535226a776deda6cb35efd937d3e2575", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2072-666X/8/5/134/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "c2cb3c78535226a776deda6cb35efd937d3e2575", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Materials Science" ] }
52846873	pes2o/s2orc	v3-fos-license	Late Diagnosis of HIV Infection in Children : Prevalence and Outcome Background: Late diagnosis of HIV infection is an important cause of death in children. Objectives: To determine the prevalence of late diagnosed HIV infection in children, describe the socio-demographic characteristics and to analyze outcome of these children. Methods: From January 2015 to October 2016, we carried out a prospective analytical study in the pediatric departments of University Teaching Hospital of the Brazzaville. Late diagnosed HIV children were selected for this work. Data analysis was performed in univariate and multivariate with Epi Info 7.2.1. Results: Of the 6058 hospitalized children, 103 (1.7%) were selected, 57.3% were boys; the median age was 21.9 months (IQR, 17.8 76.7 months). Children of low socio-economic status accounted for 68.0%, those motherless: 43.7%. None of the children were tested for HIV before hospitalization. Mothers had a low education level in 60.2% of cases and were unaware of prevention of mother-to-child transmission (PMTCT): 60.3%. Children mostly showed signs of stages 4 (49.5%), and 3 (31.1%) of HIV infection, immunodeficiency was severe for 68.0% of children. Children discharged from the hospital accounted for 62.1% of which 15.53% against medical advice. The case fatality rate was 37.9%. The risk factors for death in univariate analysis were: age < 12 months (OR = 8.66), maternal death (OR = 17.93), severe malnutrition (OR = 66.07), clinical stages 4 (OR = 66.07) and severe immunodeficiency (OR = 17.37). The main pathologies responsible for death were respiratory infections (38.5%) and diarrheal diseases (30.8%). Conclusion: Improvement of PMTCT program effectiveness, universal access to early detection and antiretroviral therapy for infants are needed to reduce the number of late diagnosed HIV-children and therefore HIV-related morbidity and How to cite this paper: Oko, A.P.G., Olandzobo, A.G., Ekouya-Bowassa, G., Ndjobo, M.I.C., Ollandzobo, L., Pandzou-Guembo, N., Lombet, L., Poathy, J.P.Y., Missambou-Mandilou, S.V., Mbika-Cardorelle, A. and Moyen, G.M. (2017) Late Diagnosis of HIV Infection in Children: Prevalence and Outcome. Open Journal of Pediatrics, 7, 331-344. https://doi.org/10.4236/ojped.2017.74038 Received: November 12, 2017 Accepted: December 12, 2017 Published: December 15, 2017 Copyright © 2017 by authors and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/ Open Access Introduction Human Immunodeficiency Virus (HIV) infection described for the first time in children in the 80 s [1], has had different epidemiological evolution depending on the regions and countries.In high income countries, after a critical period, the epidemiological situation has changed drastically: mother-to-child transmission of HIV infection has almost been eliminated through an effective prevention policy, morbidity and mortality have been significantly reduced by the early and wide use of high-activity antiretroviral therapy (HAART) in children [2] [3].In low and middle income countries, especially those located in sub-Saharan Africa, despite the successes in the response to HIV infection, it still contributes significantly to the childhood morbidity and mortality [4] [5]; In 2016, UNAIDS estimated that 2.1 million children under 15 years were living with HIV, among whom 160,000 new cases and nearly 90% of these children were in sub-Saharan Africa. Of the estimated 1.0 million people who died of AIDS-related illnesses in 2016, 120,000 of them were children under 15 years of age and the vast majority of these deaths occur in sub-Saharan Africa [5].One of the main causes of HIV-related morbidity and mortality still high in Africa is late diagnosis and consequently late initiation of antiretroviral therapy (ART) [6].Indeed several studies highlight the survival benefit of early initiation of antiretroviral therapy for HIV-infected infants especially when ART was initiated before the clinical signs and symptoms of HIV infection [6] [7] [8].And based on these data, the WHO has changed its treatment recommendations, calling for treatment of all infected infants under 24 months of age, irrespective of clinical stage and CD4 count; To improve early access to pediatric antiretroviral therapy, WHO recommends that countries with high incidence should ensure scaling up of program services prevention of mother-to-child transmission (PMTCT) of HIV infection and routine early detection infant [9].Despite these recommendations, the number of HIV-children diagnosed late in hospital remains high in Africa [10] [11].The proportion of late diagnosed HIV-infected children is also a good indicator of the effectiveness of HIV-related services in a country, especially prevention and screening.In Congo, child data are rare, in 2016, UNAIDS estimated that 6000 children living with HIV, 1100 newly infected children, and HIV infection is one of the main causes of death in children with nearly 1000 deaths [12] [13].Furthermore, the prevalence of HIV infection remains high among pregnant women, the knowledge of the PMTCT program by them is poor and coverage of PMTCT services is low [12] [14].Additionally, early detection was only performed in 3% of PMTCT in-fants and for the moment early detection is not yet routinely offered to all infants, as currently recommended by WHO [9] [15].All these data indicate potential deterioration of the epidemiological situation of HIV infection in children, including an increase in the number of late diagnosed HIV-infected children.In this work, we sought in era of the PMTCT, routine early diagnosis in infants and HAART, to determine the prevalence of late diagnosed HIV infection in children, to describe socio-demographic characteristics and to analyze outcome of late diagnosed HIV-children. Setting The study was carried out in the pediatric departments of the University Teaching Hospital (UTH) of Brazzaville.Brazzaville is the capital of the Republic of Congo and the UTH, the national referral hospital.The population of Brazzaville was estimated at 1,373,382 at the last national census (2009), HIV prevalence is 3.1% in adults and 6000 children live with HIV [12].PMTCT activities started in 2003 and are being carried out to date as a project [16].The coverage of early detection of PMTCT infants is only 3%.Pediatric care of HIV infection has been free since 2008, but only 23% of infected children are supported [12] [17]. Study Design and Patients We carried out a prospective analytical study between January 2015 and July 2016. All children aged from 1 month to 17 years hospitalized in the pediatric wards of the UTH of Brazzaville and presenting suggestive manifestations of the HIV infection according to the clinical definition of WHO cases [18], have undergone a HIV rapid antibody diagnostic test.In case of positive test, a polymerase chain reaction (PCR)-DNA test was performed in children under 18 months of age or a confirmatory second rapid antibody diagnostic test in those aged 18 months and over.Pre-and post-test counseling was performed in all screened children.Children meeting the criteria for late diagnosed HIV infection as defined for the study [19] were included systematically and consecutively. Data Collection We recorded all data on a standardized and structured form.The data included two types of variables: variables relating to children and those for parents/caregivers.For children, we recorded: age, gender, pathological history suggestive of HIV infection, nutritional status as assessed by WHO standards [20], clinical stage of disease and immune status according to WHO clinical and immunological classifications [21], the cause of hospitalization and evolution.For mothers/caregivers: age, vital status, educational level, socio-economic status, HIV knowledge, knowledge of PMTCT, health status. The sources of information were the caregiver of the child, the child himself Open Journal of Pediatrics when he was able to express, the attending physician and medical records. Statistical Analysis The data was processed and analyzed with Epi info 7.2.1 software.Quantitative variables were expressed in median and interquartile range (IQR) and qualitative variables in percent.The numbers of each variable were also specified.The percentages were compared with the independence Chi-2 test or the Fischer test (when at least one of the theoretical numbers was less than 5) and the odds ratio (OR).In the multivariate logistic regression model, we included only variables significantly associated with death from the univariate analysis (p ≤ 0.20).The significance level was set at 5% and the 95% confidence interval (CI). Ethical Considerations We obtained informed consent from parents/caregivers for interviews.The study was conducted in compliance with the Helsinki Declaration [23].The study was approved by the National Ethics Committee. Description of the Population From January 2015 to July 2016, 6058 children were hospitalized to UTH, among them 103 (1.7%) HIV-infected children, all of whom meet the criteria for late The socio-demographic, clinical and evolutionary characteristics of the population are detailed in Table 1. Knowledge and Practices HIV knowledge assessed in all caregivers was considered good in 30 (29.1%) of them, acceptable in 50 (48.6%)and poor in 23 (22.3%).The existence of the PMTCT program was known by 23 (39.7%) of the 58 living mothers.Twenty mothers (34.5%)/58 reported having been tested for HIV during pregnancy; for 10 mothers the test result was negative in the first trimester, 3 mothers did not know the test result and for the 7 others the test result was positive.5 out of 7 HIV-positive mothers had initiated antiretroviral therapy, which was subsequently discontinued for a variety of reasons (denial of the disease, fear of stigma and discrimination, fear that the spouse discovers her status and stock-outs of antiretroviral), and 2 others did not received antiretroviral therapy. None of the children were tested for HIV before hospitalization, even those whose mothers were diagnosed with HIV during pregnancy.Open Journal of Pediatrics Causes of Hospitalization The main causes of hospitalization were bronchopulmonary infections, which accounted for 51.5% of cases (including pulmonary tuberculosis) and diarrheal diseases (39.8%) (Figure 2). Outcome Among the 103 children identified, 48 (46.6%) were discharged with ART, sixteen children (15.53%) went out against medical advice.for 39 children (37.9%), the evolution was towards a death.The median duration of hospital stay was 9 days (IQR, 6 -14 days). The risk factors for death was identified only in univariate analysis, the multivariate analysis showed no risk factors independently associated with death (Table 2). Discussion This work allowed us to determine the number of late diagnosed HIV-infected children in the UTH of Brazzaville and to analyze mortality.The prevalence of 1.7% noted in this work is lower than those reported by MOYEN (3%) and MBIKA (2.5%) respectively in 1993 and 1998 [24] [25], but close to that reported by M'PEMBA (1.9%) in 2003 and Dicko-Traoré (1.8%) in Mali in 2013 [11] [26]. The first three studies were also conducted at the UTH of Brazzaville, but unlike the present study, they were conducted before the implementation of the PMTCT program and the introduction of free ART in the Congo [16] [17].Notwithstanding methodological differences, these hospital data show that the prevalence of HIV infection discovered late during hospitalization has remained relatively stable over the past decade, despite the various control measures implemented in the Congo.In addition, the prevalence observed in this work is likely underestimated because only children with suggestive symptomatology of HIV infection according to WHO criteria [18] were tested; these WHO criteria do not identify all infected children because of their low sensitivity and specificity [27] [28] [29].Routine HIV testing of all hospitalized children with unknown HIV status, as recommended by WHO [9] [15], would allow to have the actual prevalence.Kankassa found a HIV prevalence (29.2%) significantly higher than that observed in this study, after routine screening of almost all (87.4%) children hospitalized at UTH of Lusaka [29]. The often late and advanced diagnosis of HIV infection in children is a reality in many low and middle income countries [11] [30] [31] [32], the main reasons for the low impact of control measures on the prevalence of late diagnosed HIV infection are the insufficient coverage of PMTCT services, the lack of an effective early HIV testing policy for infants (PMTCT beneficiary or not), denial of the disease, fear of discrimination and stigma of people living with HIV.In our study more than 60% of the mothers were unaware of the PMTCT program, only 34.5% were tested for HIV during pregnancy and those who knew their HIV-positive status were either in denial or fear of stigma and discrimination, to the point of refusing treatment.Kambourou et al. still in the Congo had found such a large proportion (53.2%) of mothers who did not know about the PMTCT program [14].despite WHO recommendations on early and routine screening of infants, practices of child health professionals are not appropriate because none of the children in this study have been detected before.Other African countries, such as Congo, have difficulties in implementing and scaling up HIV services [33] [34] [35].To remedy this situation, of course, all of these contributing factors must be addressed, but early diagnosis of HIV in infants must be a priority and full use of new technologies and service delivery strategies.Thus, HIV-related services can be combined with other child-care offerings: Counseling and early detection of HIV infection in infants can be routinely offered at birth, at vaccination sites, weighing, at the level of consultation and hospitalization services to increase the number of children detected early.The expansion HIV services in the commu-Open Journal of Pediatrics nity is also an effective way to reach populations, particularly those without access to the health system, as demonstrated by the SEARCH (Sustainable East Africa Research in Community Health) project [36].To meet these challenges will no doubt require investment in key areas such as training and support for child health professionals, improvement of laboratory facilities and referral networks, and community mobilization. HIV infection in this work was often early-onset, 36.9% of children less than 1 year old and the majority less than 2 years old at the time of diagnosis.The high incidence of early clinical form of HIV infection in children in Africa is a known fact [37] [38] and requires the implementation of an effective screening strategy to identify these infected infants prior disease progression.The antenatal and per-partum contamination associated with the very advanced state of infection in the mothers, as suggested by the high maternal mortality rate in our study, the multiple deficiencies in the mother and in the child are the main reasons for this high frequency [39] [40] [41]. The pathologies responsible for hospitalization and death were dominated by respiratory infections and diarrheal diseases, as already reported by other authors [11] [42] [43].Most of these diseases are preventable by hygiene measures, vaccination and especially cotrimoxazole chemoprophylaxis. The mortality rate observed in this study is particularly high, other authors consulted reported similar mortality rates [43] [44].The severity of the clinical symptomatology following the late diagnosis explains this high mortality rate.The Children with HIV Early Antiretroviral Therapy (CHER) trial showed that initiation of antiretroviral therapy before the symptoms and signs of HIV infection reduced mortality by 75% [7], several others studies have also shown the benefit early initiation of ART in reducing mortality [6] [8] [45] [46].In response, WHO now recommends initiating ART as soon as possible after the diagnosis of HIV infection in children and routinely in HIV-infected infants under 2 years of age [9]. The Congo has adopted these recommendations, but is still slow to ensure that these recommendations are scaled up. The risk factors for death identified in univariate analysis in this work are iden- Caution is however needed when interpreting these results, since sample size was small and likely insufficient to detect significant differences among deceased group and survivals group. Limitations This work presents some limitations, the first concerns the site of the study, the fact that it was performed in a single hospital makes it difficult to extrapolate the Open Journal of Pediatrics results to the whole country, however, the UTH is the center where the majority of late diagnosed HIV-children are hospitalized.The second is related to the study population, it would have been interesting to compare children diagnosed late to those diagnosed early to better identify the causes behind the late diagnosis.Routine screening of all hospitalized children would have made it possible to have the actual prevalence.And the small size of our sample did not allow identify risk factors for death in multivariate analysis. Implications Despite its limitations, this study is the first on this subject since the implementation of PMTCT and setting up free healthcare for HIV in Congo and our findings draw the attention of health authorities and health professionals to the importance of late diagnosis of HIV infection in HIV-related morbidity and mortality in Brazzaville and probably in many cities in sub-Saharan Africa, to emphasize the value of early diagnosis and treatment of HIV infection in children for reducing mortality, and finally to describe the factors that contribute to late diagnosis. Conclusion Late diagnosed HIV infection remains common and causes heavy mortality in children in Brazzaville.Congo should build capacity in its PMTCT program, but also implement an effective strategy for routine early detection of HIV infection and treatment of infected children to reduce prevalence, improve diagnosis of infection HIV and therefore reduce HIV-related morbidity and mortality in children. HIV status: HIV infection was defined as a positive HIV rapid antibody diagnostic and PCR tests results for Children under 18 months and as two positive HIV rapid antibody diagnosis tests results for those aged 18 months and over.The diagnosis of HIV infection was considered late when it was made after the onset of the signs of the disease in children aged 5 and under, and in those over 5 when the signs of the disease corresponded to clinical stage 3 or 4 or when advanced or severe immunodeficiency [20] [21].Nutritional status as assessed by WHO standards [20]: we assessed the children's nutritional status using z-scores of weight-for-length (or weight-for-height).Children with values below −2 z-scores (below −3 z-scores) of the reference population were considered moderately malnourished (severely malnourished), those with values between −2 and 2 z-scores normal, and the values above 2 z-scores (above 3 z-scores) indicated overweigh (obesity).Socio-economic status (SES) was assessed using the National Congolese Center for Statistics and Economic Studies (CNSEE) classification [22], the socio-economic level of family was considered low when family expenses for meals were estimated less than 523 FCFA (African Financial Community Franc. 1 Dollar = 550 FCFA) per person per day, middle between 523 and 1685 FCFA and high between 1685 and 5055 FCFA. Figure 2 . Figure 2. Causes of hospitalization of HIV children at University Teaching Hospital of Brazzaville during the study. tical to those reported by other authors[47] [48], most of whom are directly or indirectly related to the severity of the clinical symptomatology.With regard to maternal death, it increases the vulnerability of these children because of precarity, lack of love, and maternal attention and sometimes lack of care.No risk factors independently associated with death were found in the multivariate analysis. Table 1 . Socio-demographic and clinical characteristics of HIV children and parents. Table 2 . Risk factors of death.	2018-10-17T13:40:45.007Z	2017-10-27T00:00:00.000	{ "year": 2017, "sha1": "4831c0c69bf2579a54cc90f352cff945ab106ba4", "oa_license": "CCBY", "oa_url": "http://www.scirp.org/journal/PaperDownload.aspx?paperID=81087", "oa_status": "GOLD", "pdf_src": "ScienceParseMerged", "pdf_hash": "4831c0c69bf2579a54cc90f352cff945ab106ba4", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
264468138	pes2o/s2orc	v3-fos-license	Premotor cortex is hypoactive during sustained vowel production in individuals with Parkinson’s disease and hypophonia Introduction Hypophonia is a common feature of Parkinson’s disease (PD); however, the contribution of motor cortical activity to reduced phonatory scaling in PD is still not clear. Methods In this study, we employed a sustained vowel production task during functional magnetic resonance imaging to compare brain activity between individuals with PD and hypophonia and an older healthy control (OHC) group. Results When comparing vowel production versus rest, the PD group showed fewer regions with significant BOLD activity compared to OHCs. Within the motor cortices, both OHC and PD groups showed bilateral activation of the laryngeal/phonatory area (LPA) of the primary motor cortex as well as activation of the supplementary motor area. The OHC group also recruited additional activity in the bilateral trunk motor area and right dorsal premotor cortex (PMd). A voxel-wise comparison of PD and HC groups showed that activity in right PMd was significantly lower in the PD group compared to OHC (p < 0.001, uncorrected). Right PMd activity was positively correlated with maximum phonation time in the PD group and negatively correlated with perceptual severity ratings of loudness and pitch. Discussion Our findings suggest that hypoactivation of PMd may be associated with abnormal phonatory control in PD. Introduction The majority of individuals with Parkinson's disease (PD) experience adverse changes to their speech and voice at some point throughout the disease course (Sapir, 2014).The constellation of motor speech symptoms in PD, referred to collectively as "hypokinetic dysarthria, " include monopitch, monoloudness, reduced stress, imprecise consonants, Manes et al. 10.3389/fnhum.2023.1250114Frontiers in Human Neuroscience 02 frontiersin.orginappropriate silences, short rushes of speech, harsh voice, breathy voice, low pitch, and variable rate.Among the most prevalent changes to the parkinsonian voice is the development of hypophonia -a condition characterized by reduced loudness or "soft speech" (Duffy, 2019).The physiology of hypophonia includes deficits in both laryngeal function and respiratory support for speech breathing (Solomon and Hixon, 1993;Huber et al., 2003;Hammer et al., 2013;Huber and Darling-White, 2017).However, at the cortical level, it is not clear how reduced vocal intensity relates to changes in the activity of the motor cortices (i.e., primary motor cortex, premotor cortex, and SMA). The reduction of vocal intensity in hypophonia appears to mirror the hypokinesia (reduced movement amplitude) observed in PD limb movements, and may reflect scaling deficits seen in PD (Sapir, 2014).Within the framework of the classic rate model of PD, hypokinetic and bradykinetic movements are purported to arise from reduced thalamocortical excitation of the motor cortices following the degeneration of dopaminergic cells in the substantia nigra pars compacta and subsequent dysregulation of the cortico-basal ganglia pathways.Thus, if the reduced vocal amplitude observed in hypophonia is indeed hypokinetic in nature, it should follow that motor cortical activity is hypoactive during phonation.However, reduced voice intensity in PD does not appear to have a consistent or robust response to dopaminergic therapy (Daniels et al., 1996;Kompoliti et al., 2000;Skodda et al., 2010;Fabbri et al., 2017).Moreover, existing neuroimaging studies of speech production in PD have reported mixed findings with respect to motor cortical activation, making it difficult to discern whether reduced vocal intensity is explicitly linked to reduced activity in the motor cortices (Pinto et al., 2004;Rektorova et al., 2007;Pinto et al., 2011;Arnold et al., 2014;Baumann et al., 2018;Narayana et al., 2020).The variability across studies may be in part due to differences in the clinical presentation of the participants (e.g., whether participants presented with voice symptoms at the time of the study).In addition, some of these findings may be specific to the speech paradigms used (e.g., overt vs. covert speech paradigms; connected vs. single word production). Employing a task that is specifically phonatory in nature, such as sustained phonation of a vowel, may help to establish whether hypophonia is related to hypoactivation in the motor cortices.The existing imaging literature on PD speech has focused more broadly on speech production using overt sentence reading (Rektorova et al., 2007;Arnold et al., 2014), covert sentence reading (Baumann et al., 2018), and word production (Pinto et al., 2011) tasks.However, speech production tasks in healthy adults have been shown to recruit additional regions of the cortex that are not found during sustained phonation alone (Ozdemir et al., 2006).While sentence production tasks may provide a more global picture of hypokinetic dysarthria in PD and any associated changes in functional communication, they may not be sufficient for differentiating the physiological mechanisms of voice dysfunction from changes in articulation.Using a phonatory task, such as sustained vowel production, can help to disentangle mechanisms of hypophonia from other PD speech characteristics and provide clearer insight into which neural systems are driving changes in voice output. In healthy adults, fMRI studies of vowel production/phonation have shown activity in primary motor cortex (Ozdemir et al., 2006;Soros et al., 2006;Brown et al., 2008;Grabski et al., 2013;Correia et al., 2020;Eichert et al., 2020;Brown et al., 2021), premotor cortex (Brown et al., 2008;Grabski et al., 2013), and supplementary motor area (SMA) (Soros et al., 2006;Brown et al., 2008;Grabski et al., 2013).Within the primary motor cortex, fMRI activity during phonation is most commonly reported in the bilateral laryngeal/ phonatory area (LPA), also referred to as the 'larynx motor cortex' .This functional area is located towards the middle/inferior region of the precentral gyrus and often includes two peaks of activation -the dorsolateral LPA and the ventromedial LPA (Brown et al., 2008;Bouchard et al., 2013;Belyk and Brown, 2017;Eichert et al., 2020;Belyk et al., 2021).In addition to the LPA, some fMRI studies of phonation have also reported activity in the trunk motor area, located on the superior aspect of the precentral gyrus (Guenther, 2016;Correia et al., 2020).For example, Correia et al. (2020) found additional bilateral activation in the trunk motor area when comparing voiced versus voiceless utterances.The authors proposed that activation of the trunk motor area is linked to the use of trunk muscles for respiratory control during phonation.Within the premotor cortex, it has been proposed that dorsal aspect (PMd), may be more involved with the coordination of pitch-related vocalizations and prosody, while the ventral aspect (PMv) may be more involved with the coordination of the phonetic or syllabic components of speech (Hickok et al., 2023).This model further suggests that the region of PMd responsible for coordinating vocal pitch provides input to the dorsal LPA (Hickok et al., 2023).Finally, SMA is involved with the initiation of voicing during phonation, as well as the initiation of speech production more broadly (Ziegler et al., 1997;Guenther, 2016;Jonas, 2019). The purpose of this study was to (1) characterize the BOLD responses of older healthy controls (OHC) and PD patients with hypophonia during a sustained vowel task, (2) determine whether there are group differences in motor cortical activity between PD patients with hypophonia and OHC participants during sustained vowel production, and (3) determine the strength of the association between motor cortical activity and acoustic and perceptual measures of phonation.To accomplish this, we recruited a group of PD participants with hypophonia and a group of age-matched OHC participants to perform a sustained vowel production task while undergoing fMRI.In order to obtain high fidelity measures of voice intensity, participants also performed the same vowel production task outside of the scanner in a simulated MRI environment (Manes et al., 2021).We hypothesized individuals with PD and hypophonia would have significantly lower activity in the LPA, SMA, and premotor cortex compared to age-matched healthy controls and that our acoustic and perceptual measures of phonation would correlate BOLD activity in areas of hypoactivation. Participants We recruited 15 OHC participants and 15 participants with PD to undergo fMRI scanning during vowel production.All participants provided informed consent for the study procedures, in accordance with Northwestern University's guidelines.Of the initial 15 OHC and 15 PD participants recruited, 1 OHC and 2 participants with PD were excluded from the final analysis due to excessive motion or extracranial susceptibility artifacts in the fMRI data (for quality 10.3389/fnhum.2023.1250114Frontiers in Human Neuroscience 03 frontiersin.orgassurance criteria, refer to Section 2.6).The remaining 14 OHC and 13 PD participants were included in the final fMRI analysis.All participants were right-handed, native English speakers between 40 and 80 years old.All participants had normal cognition, scoring ≥26 on the Montreal Cognitive Assessment (MoCA; Nasreddine et al., 2005) or ≥ 18 on the MoCA-Blind (if screened over the phone).Hearing thresholds for each participant were < 35 dB HL when assessed using bilateral pure tone average at 0.5, 1, and 2 kHz.Healthy control participants reported no history of speech, language, hearing, or neurological disorders.Participants in the PD group reported no history of speech, language, hearing, or neurological disorders aside from PD.There were no statistically significant group differences in age, sex, cognition, or hearing threshold (Table 1). All PD participants were initially judged to have hypophonia by their referring movement disorders neurologist or by an MDS-UPDRS certified researcher trained to identify reduced voice intensity.The presence of hypophonia was later confirmed through perceptual ratings by three expert raters (2 speech-language pathologists and 1 research assistant) not involved in data collection (Section 2.3.2).None of the PD participants had completed the Lee Silverman Voice Treatment (LSVT) program (Ramig et al., 2001b) within the past 2 years.PD disease severity ranged from mild to moderate (Hoehn and Yahr Stage I-III), and all PD participants were being treated with antiparkinsonian medication.PD characteristics, including MDS-UPDRS ratings (Goetz et al., 2008), levodopa equivalent daily dose (Tomlinson et al., 2010), Hoehn and Yahr stage (Hoehn and Yahr, 1967;Goetz et al., 2004), and PD subtype (Stebbins et al., 2013) are reported in Table 2. Motor testing for the MDS-UPDRS Part III was conducted after 12-h overnight withdrawal from antiparkinsonian medications.fMRI testing was also completed after 12-h medication withdrawal in order to maximize the disease-state. Sustained vowel production task During the task, participants were presented with either a "+" symbol (Rest) or "Ah" (Vowel Production) block.During the "Ah" blocks, subjects were instructed to produce an /a/ vowel for approximately 3-5 s at their normal conversational loudness and repeat for the duration of the block.This self-paced paradigm was designed so that participants would rely on internal, rather than external cueing mechanisms for the initiation of each utterance.In-scanner recordings were used to verify task compliance and to measure the duration and number of vowels produced by the participant. Acquisition and analysis of voice measures 2.3.1. Collection of voice samples Outside of the MRI scanner, we collected samples of maximum phonation time as well as voice sound pressure level (SPL) samples.Voice samples were recorded using a head mounted, unidirectional microphone (Shure SM10A) positioned 3 cm from the lower lip.The microphone was channeled through a pre-amplifier (ART Project Series USB Dual Pre) and then relayed onto a laptop computer for recording in Audacity (sampled at 44.1 kHz).For measures of maximum phonation time, participants were instructed to sustain an /a/ vowel for as long as possible in their normal conversational loudness.The productions were recorded in Audacity and manually marked for voice onset/offset and the maximum phonation time was reported in seconds.For measures of voice SPL, we used recordings collected outside of the scanner in which participants performed the sustained vowel production task.To capture performance that would be ecologically similar to our fMRI task, participants completed the sustained vowel task in a mock MRI scanner with fMRI acoustic noise presented over headphones at 90 dB.Recordings for the sustained vowel task were also collected with the participants seated upright and without the presence of fMRI noise.We calculated the difference in SPL between the Mock Scanner + Noise and Upright recording conditions to correct for group differences in the effects of the scanning environment on voice SPL (Manes et al., 2021).The order of fMRI and speech testing was counterbalanced to control for the effects of practice and vocal fatigue. Inside of the MRI scanner, we collected acoustic recordings to monitor task compliance and to record the timing and duration of self-paced vowel productions.An MRI compatible microphone was mounted to the head coil and positioned ~1 cm from the lower lip.The audio signal was digitized using a Measurement Computing USB-1608G A/D board and processed in MATLAB.Using the volume trigger from the scanner, a template of the scanner noise was made during the time when the MR signal was reaching equilibrium and no patient sounds were made.Each subsequent time period had the template noise signature subtracted followed by temporal and frequency filtering to remove any scanner noise from the auditory signal (Sen and Parrish, 2007).The final result was output as a .wavfile with time = 0 synchronized with the start of the experiment. Perceptual ratings of PD voice severity Twelve of our thirteen PD participants received perceptual ratings of voice quality as part of a previous behavioral study (Manes et al., 2021) . Perceptual ratings were performed using the Consensus Auditory-Perceptual Evaluation of Voice (CAPE-V; Kempster et al., 2009), which uses a 100 mm analogue scale to rate voice deviance in 5 perceptual domains (roughness, breathiness, strain, pitch, and loudness).Three expert raters listened to recordings of the participants producing /a/ and rated the perceived deviance for each perceptual domain. Acoustic analysis of voice measures (outside the scanner) Maximum phonation time was analyzed using a one-way ANCOVA, with group defined as a fixed factor and age and hearing threshold defined as covariates.To examine the effect of group on voice SPL in our fMRI-analogous environment (Mock Scanner + MRI Noise), we fit a mixed effects (multi-level) models using the lmer command in the R package lme4 (version 1.1-21; Bates et al., 2015).In this model, random effects characterized the degree of difference across individual intercepts and slopes, while the fixed effects estimates were drawn from the average of these individual intercepts and slopes.Estimated p values were obtained via Kenward-Roger's degrees of freedom method using the lmerTest package (version 3.1-0; Kuznetsova et al., 2017).We modeled the effect of group (OHC vs. PD) on vowel intensity, with fixed effects estimates of group, block, and their interactions.We coded factors using effects (simple) coding (Clopper, 2013).Group was coded as −0.5 for OHC and + 0.5 for PD participants.We treated block as a numeric variable to account for possible vocal decay across blocks, with the first block coded as zero.To account for individual differences in vowel intensity and vocal decay across blocks, we also included random intercepts and random linear slopes for each subject within the model.Random effects were estimated with an unstructured covariance matrix and restricted maximum likelihood estimation, using the Nelder-Mead optimization function. Our previous study (Manes et al., 2021) suggested that our PD participants increased voice SPL to a greater degree than controls when laying supine and listening to MRI noise.Thus, to account for inter-subject differences in Lombard responses during fMRI, we calculated the difference in voice SPL between the "Mock Scanner + MRI Noise" (described above) and the "Upright" conditions in which participants were recorded while seated upright outside of the mock scanner (Manes et al., 2021).This "Lombard response" measure was entered as a covariate in the fMRI analysis. Acoustic analysis of task performance (inside the scanner) Acoustic recordings from the MRI-compatible microphone were used to assess the durations and number of vowels produced in each task block.The onset and offset of each vowel production was manually labelled in Audacity. 1 To test for group differences in task performance, we compared the mean vowel duration between PD and OHC groups using a two-sample t-test.We also compared the total number of vowels produced between PD and OHC groups using a two-sample t-test. fMRI data acquisition Imaging data were collected on a Siemens 3 T PRISMA-FIT MRI scanner using a 64-channel head coil.T1-weighted anatomical scans were collected in the sagittal plane using an MPRAGE GRAPPA sequence at a voxel resolution of 0.8 mm 3 (TR = 2000 ms, TE = 2.99 ms, TI = 1,010 ms, flip angle = 8°, FOV = 256 mm, inversion time = 1,010 ms, BW = 240 Hz/pixel).T2-weighted anatomical scans were collected in the sagittal plane using a T2 SPACE sequence at a voxel resolution of 0.8mm 3 (TR = 2,500 ms, TE = 566 ms, flip angle = 120°, FOV = 256 mm, BW = 710 Hz/pixel).For the fMRI scans, we employed a continuous scanning protocol.BOLD T2-weighted functional scans were collected in 56 interleaved slices using a multiband acceleration factor of 2 and voxel size of 2mm 3 (TR = 2000 ms, TE = 25 ms, flip angle = 80°, FOV = 208 mm).The sustained vowel production task was presented in the magnet room on a NordicNeuroLab LCD monitor using E-Prime and viewed through an angled mirror mounted to the head coil. fMRI experimental procedure All testing was performed at Northwestern University's Center for Translational Imaging.During the scanning session, participants lay 1 https://audacityteam.org/ supine in the MRI scanner while wearing foam earplugs for hearing protection.An angled mirror was affixed to the head coil so that participants could view the stimulus prompt and an MRI-compatible microphone was affixed to the head coil to record the onset and offset of voicing.Prior to beginning the task, the participants were told that they would see either an "Ah" or "+" prompt on the screen.We instructed participants that whenever the "Ah" prompt appeared on the screen, they should produce an /a/ vowel for approximately 3-5 s at their normal conversational loudness and repeat at their own pace until the "Ah" prompt disappeared.We instructed the participants that whenever the "+" prompt appeared on the screen, they should stay silent and look at the crosshair.In order to limit movement of the head and jaw when vocalizing, we also instructed participants to keep their mouths slightly open through the duration of the task.After the instructions were given, the participants performed the sustained vowel production task during fMRI scanning.Total task time was 10 min, including ten vowel production blocks (30s each) and ten rest blocks (30s each). fMRI data preprocessing All fMRI data was processed and analyzed using SPM12 and AFNI tools.To be included in the statistical analysis, we required at least 7 min (210 time points) of usable data for each fMRI scan, with a framewise displacement (FD) < 0.5 mm and root-mean square signal change (DVARS) < 5%.Functional images underwent initial despiking and realignment to the reference frame (first time point).During realignment, 6 motion parameters were extracted.Additional motion correction was applied using the ART Repair Toolbox version 5b in SPM12.2Corrupted volumes were 'repaired' using linear interpolation if the framewise displacement (FD) was greater than 0.5 mm.Functional images were then smoothed to a full-width-half maximum of 6 mm.Subject-level analysis of BOLD fMRI data was conducted in subject-space using a general linear model (GLM) in SPM12 before being normalized to MNI space for group-level statistical analysis.To perform the first-level GLM, we extracted the timing and duration of each self-paced vowel production using noise-attenuated, in-scanner microphone recordings.We then analyzed the data in a block design, with the block onset starting at the initiation of the first vocalization and ending at the end of the last vocalization.The subject-level GLM included the block timing and 6 motion parameters, with global normalization scaling applied.The first-level statistical results (Ah vs. Rest) were then normalized to MNI space using a multi-stage procedure that also accounts for geometric distortions due to magnetic field inhomogeneities (Jezzard and Balaban, 1995;Chambers et al., 2015).First, we co-registered the T2-weighted scans to the T1-weighted scan and then rigidly co-registered both T1-and T2-weighted scans to the first volume of the participant's functional scan.Next, using a similar approach to Chambers et al. (2015), we nonlinearly warped the fMRI reference volume to the co-registered T2-weighted scan to estimate the distortions due to the magnetic field inhomogeneities (Warp 1 ).Using AFNI's auto_ warp.pycommand, we then normalized the co-registered T1-weighted scan to the MNI 2009c symmetric template (Fonov et al., 2009) using combined affine and nonlinear warp transformations (Warp 2 ).Finally, we combined Warp 1 and Warp 2 to calculate a single nonlinear warp transformation from fMRI space to MNI space.The combined transformation was then applied to the fMRI statistical maps in order to perform grouplevel statistical analysis in MNI space. fMRI analysis of task-related activations Group-level statistical analysis was conducted using SPM12.The second-level general linear model included the covariates of age, hearing threshold, SPL (within group), and Lombard response.Within each group, we generated task activation maps (Ah vs. Rest).We then calculated our between-group contrasts (OHC vs. PD).For all contrasts, we report results using a voxel-wise height threshold of p < 0.001, uncorrected and cluster extent threshold of p < 0.05, uncorrected (58 mm 3 ).We also report cluster-level FDR-corrected value of ps for all contrasts. Perceptual ratings of PD voice severity Deviance ratings across the five CAPE-V perceptual domains are reported in Table 3. Higher scores indicate more deviant voice characteristics.Table 3 suggests mild to moderate severity of voice characteristics of the PD group. Acoustic voice measures (outside the scanner) A one-way ANCOVA revealed a significant main effect of group for maximum phonation time (F(27) = 5.604, p = 0.027).Maximum phonation time was significantly shorter in the PD group compared to OHC (difference = 4.239, p = 0.027; Figure 1A). Table 4 reports the linear mixed effects model for voice SPL in the mock MRI scanner.Although the mean SPL of the PD group was 4.712 dB lower than the OHC group (Figure 1B), our model output did not indicate a significant effect of PD on voice SPL (β = −3.29,SE = 2.26, p = 0.158).There was no significant effect of block, nor was there a significant groupblock interaction. Task performance inside the scanner Due to the self-paced design of the sustained vowel task, we recorded the duration and number of vowels produced by each subject during the vowel production blocks.On average, participants in the OHC group sustained /a/ vowels for a mean duration of 3.82 s, while the mean duration for participants in the PD group was 2.85 s (Figure 2A).Despite the trend for shorter vowels in the PD group, there was no statistically significant group difference in vowel duration (t = 1.920, p = 0.066).Participants in the PD group produced significantly more vowels than the OHC group (OHC mean: 62.8 vowels, PD mean: 83.2 vowels, t = −3.71,p = 0.026; Figure 2B).Thus, overall, the vowels produced by the PD group tended to be of higher frequency and shorter duration.The total duration of vowel productions across the entire fMRI task was not statistically different between OHC and PD groups (OHC mean: 215.8 s, PD mean: 212.1 s, t = 0.335, p = 0.741). Bold activity during sustained vowel production Table 5 reports the anatomical labels, size, coordinates, and t-statistics of all clusters meeting statistical significance.When comparing ' Ah' versus Rest conditions in the OHC group, BOLD activation in the motor cortices was found bilaterally in the LPA of the primary motor cortex (M1) and in the trunk motor area of M1, as well Comparison of acoustic measures between OHC and PD groups (outside the scanner).Bar graphs depict (A) mean maximum phonation times, and (B) the mean sound pressure levels collected when the sustained vowel production was performed in the mock MRI scanner.Error bars represent 95% confidence intervals.Whisker lines reflect 1.5 times the interquartile range.Dots represent data points for individual participants.as right SMA and right dorsal premotor cortex (PMd; p < 0.001, uncorrected).Activation was also found bilaterally in the cerebellum (lobule VI), caudate, inferior frontal gyrus, and precuneus.In the right hemisphere, BOLD activity was found in superior temporal gyrus, cuneus, medial frontal gyrus, and insula.In the left hemisphere, additional BOLD activations were found in the inferior occipital gyrus and postcentral gyrus.Fewer areas of BOLD activation were found in the PD group when comparing ' Ah' versus Rest conditions.In the motor cortices, BOLD activity for the PD group was found in the bilateral LPA region of M1 and right SMA; however, there was no BOLD activity in the trunk motor area of M1 or in the right PMd (p < 0.001, uncorrected).Additional BOLD activations were found in the right cerebellum (lobule VI), right cuneus, left superior temporal gyrus, and left inferior occipital gyrus.Motor cortical activation maps for OHC and PD groups are shown in Figure 3 top and middle rows. Comparison of BOLD activity between OHC and PD groups A vowel-wise comparison of OHC and PD groups showed that BOLD activity in right PMd was significantly lower in the PD group (p < 0.001, uncorrected; Figure 3 bottom row).However, when applying a more conservative cluster-level statistical threshold (p = 0.05, FDR-corrected), this finding did not reach statistical significance.There were no regions in which BOLD activity was higher in PD compared to OHC (Table 4). Correlation of right PMd with voice measures We examined behavioral correlations with right PMd by first extracting the mean beta values of right PMd for each participant.For the 12 PD participants who had CAPE-V ratings of dysphonia severity, we correlated right PMd activity with expert ratings of deviancy in the following domains: loudness, pitch, roughness, breathiness, and strain.Correlations were subjected to false-discovery rate (FDR) correction using the Bejanamini and Hochberg linear step-up procedure (Benjamini and Hochberg, 1995).Results of the correlation analyses indicated that right PMd beta values were negatively correlated with severity ratings in the loudness domain (r = −0.669,p uncorr = 0.017, p FDR = 0.043) and pitch domain (r = −0.731,p uncorr = 0.007, p FDR = 0.035) the CAPE-V.Right PMd beta values were not correlated with ratings of severity in the roughness (r = 0.215, p uncorr = 0.503, p FDR = 0.629), breathiness (r = −0.434,p uncorr = 0.159, p FDR = 0.265), or strain (r = 0.037, p uncorr = 0.908, p FDR = 0.908) domains of the CAPE-V. Within both the PD and OHC groups, we also correlated right PMd beta values with the out-of-scanner acoustic measures of maximum phonation time, and voice SPL.There was a positive correlation between right PMd beta values and maximum phonation time in the PD group (r = 0.612, p uncorr = 0.026, p FDR = 0.052), but not the OHC group (r = 0.181, p uncorr = 0.536, p FDR = 0.536).There was no correlation between right PMd beta values and voice SPL in either the PD group (r = −0.409,p uncorr = 0.166, p FDR = 0.166) or OHC group (r = 0.364, p uncorr = 0.201, p FDR = 0.40).Scatter plots depicting the behavioral correlations for OHC and PD groups are depicted in Figure 4. We further correlated right PMd values with measures of in-scanner task performance.For the OHC group, there was an inverse correlation between right PMd beta values and vowel duration (r = −0.574,p uncorr = 0.032, p FDR = 0.064) and a marginal positive correlation between right PMd beta values and number of vowels produced (r = 0.505, p uncorr = 0.066, p FDR = 0.066).For the PD group, PMd beta values did not correlate with either vowel duration (r = 0.262, p uncorr = 0.388, p FDR = 0.600) or number of vowels produced (r = −0.161,p uncorr = 0.600, p FDR = 0.600). Discussion The present study has four main findings.First, our acoustic measures showed that the PD group had reduced maximum phonation time compared to OHCs, and that the PD group tended to produce more vowels of shorter duration when completing the selfpaced sustained vowel task.Second, individuals in the PD group recruited fewer brain regions during the sustained vowel production task compared to the OHC group.Most notably, activations in the bilateral trunk motor area and right PMd were present in the OHC group, but not in PD.Third, we found that BOLD activity in right PMd was significantly lower in the PD group compared to OHC.Fourth, we observed that lower BOLD activity in right PMd was associated with shorter maximum phonation time as well as higher CAPE-V ratings of loudness and pitch deviance in the PD group. Acoustic measures of voice collected both inside and outside of the MRI scanner were indicative of reduced respiratory-phonatory support in the PD group.Outside of the scanner, we observed significantly lower maximum phonation times in the PD group compared to OHC.The maximum phonation time reported in the OHC group is consistent with published norms for older healthy adults (Maslan et al., 2011).In addition, the observation of lower maximum phonation time in PD supports previous studies (Yuceturk et al., 2002;Midi et al., 2008;Bauer et al., 2011) and may reflect reduced respiratory support for phonation and/or reduced glottal efficiency in PD.When participants performed the sustained vowel production task in a simulated fMRI environment, SPL values were higher than those typically observed during sustained vowel recording due to the effects of laying supine and listening to 90 dB acoustic noise (Manes et al., 2021).The magnitude of the SPL difference between the groups (−4.712 dB SPL) was comparable to the difference of −4.734 dB SPL reported in our previous paper; however, unlike our previous paper, the mixed effects model did not yield a significant effect of group on SPL.The lack of a significant group effect on SPL (p = 0.158) is in part due to the reduced power of fewer overall observations compared to our prior study.Furthermore, the previously reported group effects were smallest in the Mock Scanner + MRI noise condition, reflecting a group*condition interaction effect on voice SPL.While the SPL data from the Mock Scanner + MRI Noise condition were deemed to be the most ecologically similar to the fMRI task, on their own they were unable to capture group differences in voice intensity.When participants performed the sustained vowel production task in the scanner, we found that individuals in the PD group tended to produce more vowels of shorter duration compared to the OHC group.The shorter vowel duration in PD is likely the result of the decreased respiratory-phonatory support, in line with decreased maximum phonation time observed in this group.Individuals with PD may be self-selecting shorter phonation times to conserve respiratory/phonatory effort.Studies of speech breathing have found that individuals with PD have a faster resting tidal breathing rate and spend less time producing speech within a breath group (Solomon and Hixon, 1993).The observed duration and number of vowels produced by each group suggest that the performance of self-paced vowel productions is not identical between OHC and PD groups.However, given that there was no significant group difference in the BOLD activation in primary motor and premotor cortices during sustained vowel production (Ah vs. Rest).BOLD activation maps show areas of significant activity within the older healthy control group (top), significant activity within the PD group (middle), and significant group differences in activity between the older healthy control group and Parkinson's disease group (bottom); voxel-wise p < 0.001, uncorrected, cluster p < 0.05, uncorrected.Slice numbers reflect y coordinates in MNI space.Color bars correspond to t-statistics. 10.3389/fnhum.2023.1250114 Frontiers in Human Neuroscience 10 frontiersin.orgtotal time spent vocalizing during the task, the tradeoff between vowel length and number of vowels produced likely balances out the overall level of BOLD activation within each block. With respect to motor cortical activity, SMA activity was present in both OHC and PD groups.However, there were differences in the recruitment of primary motor and premotor cortical areas during sustained vowel production.In the OHC group, we found activity bilaterally in two clusters along the precentral gyrus: (1) the dorsal LPA, which is purported to control the intrinsic laryngeal muscles (Brown et al., 2008;Simonyan, 2014;Correia et al., 2020), and (2) the trunk motor area, which is involved in volitional breathing (Ramsay et al., 1993;Takai et al., 2010) as well as phonation (Olthoff et al., 2008;Correia et al., 2020).By contrast, BOLD activation maps for the PD group showed activity in the bilateral dorsal LPA, but no activity in the trunk motor area.This absence of observed activity in the trunk motor area may reflect a reduced cortical drive for engaging the muscles of the chest wall during speech breathing (Hovestadt et al., 1989;Solomon and Hixon, 1993;Sabate et al., 1996).Reduced respiratory-phonatory support from the expiratory muscles of the trunk would be consistent with our finding of reduced maximum phonation time in PD.However, this would need to be examined in greater detail using more sophisticated measures of respiratory physiology. Activity in the right PMd was found in the OHC group, but not in the PD group when performing within-group ' Ah' versus Rest contrasts.Further, the group comparison of BOLD activity in OHC vs. PD contrast demonstrated that the PD group had significantly lower activity in right PMd compared to OHC.Hypoactivity of right PMd has not been previously reported during speech production in PD.However, a study by Narayana et al. (2020) reported that PD participants with hypokinetic dysarthria had less activity in left PMd compared to typical speakers when performing a connected speech task during both PET and fMRI.The authors also noted that while both the PD group and control group recruited bilateral PMd during the fMRI connected speech task, the right hemisphere cluster was much smaller in the PD group.Moreover, while the control group recruited both left and right PMd when performing the speech task during PET, the PD group recruited only left PMd.Although hypoactivation in right PMd has not been previously reported in PD during speaking tasks, our finding is in line with prior work examining the neural correlates of voice treatment in PD.Narayana et al. (2010) found that PD participants had activation in right PMd only after completing the LSVT program.Further, the activity in right PMd was positively correlated with increased voice intensity (SPL) following successful voice treatment (Narayana et al., 2010).As the participants in our PD group had not completed the LSVT program within the past 2 years, the absence of right PMd activity in our study is consistent with the findings observed in the pre-treatment group.It also suggests that reduced activity in PMd is a potential biomarker for therapeutic intervention. While the precise role of right PMd in speech production is unclear, there is growing evidence that PMd plays a key role in phonatory control.Indeed, neuroimaging studies have reported activation of the right PMd during singing (Saito et al., 2006;Wilson et al., 2011;Belyk et al., 2021), humming (Hickok et al., 2003), and vowel production (Loucks et al., 2007), as well as during volitional exhalation (Loucks et al., 2007).Furthermore, our right PMd cluster appears to overlap with the 'dorsal precentral speech area' (dPCSA), which has been proposed by Hickok et al. (2023) to serve as a region for coordinating pitch-related vocalizations, such as song and prosody.Hickok et al. (2023) propose that the dPCSA (PMd) is part of a hierarchical system responsible for planning and coordinating pitchrelated vocalizations via projections to the dorsal LPA.Indeed, evidence from younger healthy adults suggests that right PMd is one of several areas which exhibit functional connectivity with the LPA during voice production (Simonyan et al., 2009).In the present study, we found no differences in LPA activity between or OHC and PD groups, suggesting that phonatory symptoms in PD may be driven more by changes in the preparatory input to LPA rather than changes in the LPA itself. In addition to its proposed role in phonatory control, it is also worth considering the potential role of right PMd in self-perception during phonation.A number of functional imaging studies have shown that PMd is active during both speech production and perception tasks (Skipper et al., 2005;Meister et al., 2007;Callan et al., 2010;Tremblay and Small, 2011;Glanz Iljina et al., 2018) -although we note that some studies refer to this area as "superior PMv" rather than PMd.Moreover, cortical stimulation studies have shown that premotor cortex is critical for the self-perception of motion or 'motor awareness' (Desmurget and Sirigu, 2009;Bolognini et al., 2016;Fornia et al., 2020;Marotta et al., 2021).For example, disruption of right PMd activity via cathodal transcranial direct current stimulation (tDCS) impaired the participants' ability to reliably estimate their own motor performance (Bolognini et al., 2016).Given that people with PD have poorer self-awareness of their own voice intensity (Ho et al., 1999(Ho et al., , 2000) ) and abnormal monitoring of their own sensory feedback during phonation (Liu et al., 2012;Mollaei et al., 2016), it is possible that the hypoactivation we observed in right PMd reflects an impaired motor awareness for phonation in PD. Our correlation analyses showed that PD participants with lower levels of right PMd activity had more deviant perceptual ratings of voice loudness and pitch on the CAPE-V.The inverse correlation between right PMd and loudness deviance is in line with the findings of Narayana et al. (2010), showing that right PMd activity was positively correlated with increased voice intensity following successful voice treatment.Engaging right PMd during overt speech and vowel production tasks may reflect more purposeful or effortful phonation to compensate for disease-related impairments in automated scaling of voice intensity.The correlation of PMd with pitch deviance is in line with the hypothesis that PMd is involved in the coordination of vocal pitch (Hickok et al., 2023).The fact that our right PMd cluster was correlated with both loudness and pitch deviation measures could point to a slightly broader impairment of coordinating the laryngeal musculature for phonatory control in PD.For example, engaging right PMd during phonation may help to (Ramig et al., 2001a) and PD (Ma et al., 2020), the need for premotor activity may reflect an increased effort required for generating or maintaining vocal fold tension during phonation.In the case of PD, dysregulation of the cortico-basal ganglia motor loop may impair the ability to effectively engage PMd for increasing vocal fold tension, resulting in hypophonia as well as changes in pitch modulation. In addition to our perceptual measures, we found that activity in right PMd was positively correlated with maximum phonation time in the PD group.If our right PMd cluster is involved in the modulation of laryngeal muscle activity, its association with longer maximum phonation times may be a function of increased glottal efficiency.Impaired glottal closure is associated with difficulty sustaining phonation as well as reduced voice intensity in PD (Hanson et al., 1984).It is possible that engaging right PMd can help to achieve more glottic closure for increasing phonation time as well as increasing loudness.An alternative interpretation is that right PMd is involved with generating and maintaining a respiratory drive for sustained phonation in PD.The evidence for this is somewhat tempered by the lack of a significant correlation between in-scanner vowel duration and PMd activity in the PD group.However, we also note that participants were instructed to produce vowel durations of 2-3 s in the scanner and that these shorter vowel production times do not necessarily reflect the full range of respiratory-phonatory capacity that is captured by maximum phonation times.Given that the correlation between PMd and maximum phonation time was marginal after applying FDR correction, the effect should be interpreted with caution until replicated in a larger sample. In contrast to the PD participants, right PMd activity in the OHC group was not correlated with maximum phonation time.Rather, right PMd activity in OHCs was inversely correlated with vowel duration and marginally correlated with the number of vowels produced during the sustained phonation task.We note that neither of these correlations were statistically significant upon FDR correction; however, the observation suggests a trend for right PMd activity to be higher for OHC participants who self-selected to produce shorter, more frequent vowels.It may be the case that OHCs utilized right PMd for initiating, rather than sustaining, the vowel productions.If so, this would be in contrast to the PD group, who showed no relationship between right PMd activity and either vowel duration or number of vowels produced.Given the self-paced nature of the sustained vowel production task, we expected that movement initiation would be driven primarily by SMA rather than the lateral premotor cortex (Lu et al., 2012).Still, it is possible that PMd played a role in vowel initiation or received downstream modulation by SMA for initiating vowel production. Interestingly, activity in our right PMd cluster was not correlated with SPL measures taken from our mock scanner recordings.While we sought to make our mock scanner environment as comparable to the fMRI environment as possible (Manes et al., 2021), and counterbalanced the order of testing, we cannot rule out the possibility that the participants behaved differently in our mock scanner and fMRI sessions.The fact that our right PMd cluster was inversely correlated with perceptual ratings of loudness deviance suggests that activity in right PMd is indeed related to perceived changes in loudness in PD.However, it will be important to replicate this finding in future studies using additional acoustic and perceptual measures. While the present study points to a potential role of hypoactive right PMd in PD hypophonia, the findings should be interpreted cautiously, as the group differences in right PMd were not statistically significant when correcting for multiple comparisons.Further research will be needed to confirm whether hypoactivation of right PMd is associated with respiratory-phonatory symptoms in PD hypophonia.To fully understand the contributions of right PMd to phonatory control in PD, it will be important to investigate whether PMd activity scales with increasing respiratory-phonatory effort.Collecting more direct assessments of respiratory physiology during a phonatory fMRI task would help to assess whether there is a link between hypoactivity in the PMd and changes in respiratory drive or timing for phonation.Correlating PMd activity with measures of prosody in PD will also be a useful extension of the present research in light of the proposed role of PMd in pitch-related vocal control.In addition, it will be important to ascertain what role right PMd plays within the broader networks controlling vocalization.For example, future studies will be needed to investigate whether right PMd serves as a preparatory area for modulating the dorsal LPA, consistent with the model proposed by Hickok et al. (2023), or whether the influence of PMd on phonation in PD may be independent of the dorsal LPA.Investigating whether PMd is modulated by input from SMA will also provide insight into the possible role of PMd for coordinating voice initiation. Collectively, our findings suggest that PD hypophonia is associated with functional changes in the right premotor cortex.Hypoactivation of right PMd may be related to decreased respiratory-phonatory support for pitch and loudness regulation in PD hypophonia.JM contributed to the selection and interpretation of voice measures.AR contributed to the collection and analysis of perceptual voice ratings.JM, EH, and AK contributed to the pre-processing of fMRI data.DC contributed to the editing, style, and structure of the manuscript.All authors contributed to the article and approved the submitted version. FIGURE 2 FIGURE 2Comparison of self-paced task performance between OHC and PD groups (inside the scanner).Box plots depict (A) mean vowel durations, and (B) the mean number of vowels produced when performing the sustained vowel task in the MRI scanner.Whisker lines reflect 1.5 times the interquartile range.Dots represent data points for individual participants. FIGURE 4 FIGURE 4 Correlation of right PMd activity values with voice measures.Scatter plots depict the correlation of right PMd beta values with (A) CAPE-V ratings of loudness severity, (B) CAPE-V ratings of pitch severity, (C) voice sound pressure level during sustained vowel production in a mock MRI scanner, and (D) maximum phonation time, (E) mean vowel duration in the scanner, and (F) total number of vowels produced in the scanner.Lines represent the linear fit of data points within each group.Data from the OHC group are plotted in green and data from the PD group are plotted in red. TABLE 1 Demographic features for the OHC and PD groups. TABLE 2 Parkinson's disease characteristics for PD participants. TABLE 3 Perceptual ratings of PD voice severity on the CAPE-V. TABLE 4 Output of linear mixed effects model for group effects on voice SPL.	2023-10-26T15:38:55.975Z	2023-10-23T00:00:00.000	{ "year": 2023, "sha1": "a7ef204126a6ced6e7cd14ad08ca0e177f26105c", "oa_license": "CCBY", "oa_url": "https://www.frontiersin.org/articles/10.3389/fnhum.2023.1250114/pdf?isPublishedV2=False", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "cbab067a98c348848bc8d0fdd3a7aa7182ed8732", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [] }
213838667	pes2o/s2orc	v3-fos-license	A Genome-Wide Screen for Genes Affecting Spontaneous Direct-Repeat Recombination in Saccharomyces cerevisiae Homologous recombination is an important mechanism for genome integrity maintenance, and several homologous recombination genes are mutated in various cancers and cancer-prone syndromes. However, since in some cases homologous recombination can lead to mutagenic outcomes, this pathway must be tightly regulated, and mitotic hyper-recombination is a hallmark of genomic instability. We performed two screens in Saccharomyces cerevisiae for genes that, when deleted, cause hyper-recombination between direct repeats. One was performed with the classical patch and replica-plating method. The other was performed with a high-throughput replica-pinning technique that was designed to detect low-frequency events. This approach allowed us to validate the high-throughput replica-pinning methodology independently of the replicative aging context in which it was developed. Furthermore, by combining the two approaches, we were able to identify and validate 35 genes whose deletion causes elevated spontaneous direct-repeat recombination. Among these are mismatch repair genes, the Sgs1-Top3-Rmi1 complex, the RNase H2 complex, genes involved in the oxidative stress response, and a number of other DNA replication, repair and recombination genes. Since several of our hits are evolutionarily conserved, and repeated elements constitute a significant fraction of mammalian genomes, our work might be relevant for understanding genome integrity maintenance in humans. homologous pairing and strand invasion, with the help of Rad54, Rad55, and Rad57. After copying the homologous template, recombination intermediates are resolved with the help of nucleases and helicases, and the HR machinery is disassembled (Symington et al., 2014). While HR is important for genome integrity, excessive or unregulated recombination in mitotic cells can be deleterious. Indeed, even though HR is generally considered an error-free DNA repair pathway, outcomes of HR can be mutagenic. For instance, recombination between ectopic homolog sequences can lead to gross chromosomal rearrangements (Heyer 2015). Mutations and chromosomal aberrations can be the outcome of recombination between slightly divergent DNA sequences, a process termed "homeologous recombination" (Spies and Fishel 2015). Allelic recombination between homologous chromosomes can lead to loss of heterozygosity (LOH) (Aguilera and García-Muse 2013). Finally, the copying of the homologous template occurs at lower fidelity than is typical for replicative DNA polymerases, resulting in mutagenesis (McVey et al., 2016). For these reasons, the HR process must be tightly controlled, and spontaneous hyper-recombination in mitotic cells is a hallmark of genomic instability (Aguilera and García-Muse 2013;Heyer 2015). Pioneering mutagenesis-based screens led to the identification of hyper-recombination mutants (Aguilera and Klein 1988;Keil and McWilliams 1993). Subsequently, several systematic screens were performed with the yeast knockout (YKO) collection to identify genes whose deletion results in a spontaneous hyper-recombinant phenotype. In particular, Alvaro et al. screened an indirect phenotype, namely elevated spontaneous Rad52 focus formation in diploid cells, which led to the identification of hyper-recombinant as well as recombination-defective mutants (Alvaro et al., 2007). A second screen for elevated Rad52 foci in haploid cells identified additional candidate recombination genes (Styles et al., 2016), although the recombination rates of these were not assessed directly. A distinct screen of the YKO collection measured elevated spontaneous LOH events in diploid cells, which arise through recombination between homologous chromosomes or as a consequence of chromosome loss (Andersen et al., 2008). Here we describe two systematic genome-scale screens measuring spontaneous recombination in haploid cells, since the sister chromatid is generally a preferred template for mitotic recombination relative to the homologous chromosome, both in yeast and mammalian cells (Johnson and Jasin 2000;Kadyk and Hartwell 1992). We use a direct-repeat recombination assay (Smith and Rothstein 1999), because recombination between direct repeats can have a significant impact on the stability of mammalian genomes, where tandem and interspersed repeated elements, such as LINEs and SINEs, are very abundant (George and Alani 2012;López-Flores and Garrido-Ramos 2012). Recombination rate screens were performed both with the classical patch and replica-plating method and with our recently developed high-throughput replica-pinning technique, which was designed for high-throughput screens involving low-frequency events (Novarina et al., 2020). High-throughput replica-pinning is based on the concept that, by robotically pinning an array of yeast strains many times in parallel, several independent colonies per strain can be analyzed at the same time, giving a semi-quantitative estimate of the rate at which a specific low-frequency event occurs in each strain. We used both approaches to screen the YKO collection with the direct-repeat recombination assay. Bioinformatic analysis and direct comparison of the two screens confirmed the effectiveness of the highthroughput replica-pinning methodology. Together, we identified and validated 35 genes whose deletion results in elevated spontaneous direct-repeat recombination, many of which have homologs or functional counterparts in humans. MATERIALS AND METHODS Yeast strains and growth conditions Standard yeast media and growth conditions were used (Sherman 2002;Treco and Lundblad 2001). All yeast strains used in this study are derivatives of the BY4741 genetic background (Brachmann et al., 1998) and are listed in Supporting Information, Table S1. Each YKO strain carrying the recombination reporter was streaked for single colonies on SD-ura. Single colonies were then streaked in a 1 cm · 1 cm patch on YPD, incubated at 30°for 24 h, and then replica-plated to SD-leu to detect recombination events as papillae on the patch. RDY9 (wild-type) and RDY13 (elg1Δ::kanMX; positive control (Bellaoui et al., 2003;Ben-Aroya et al., 2003)) were included on each plate. The papillae on SD-leu were scored by visual inspection relative to the control strains, yielding 195 positives (Table S2). The 195 positives were tested in a fluctuation test of 5 independent cultures, and those with a recombination rate of at least 2x10 25 (approximately twofold greater than that of RDY9) were identified (43 strains; Table S2). Positives from the first fluctuation tests (except slm3Δ and pex13Δ, where rates could not be determined due to the large numbers of 'jackpot' cultures where all colonies had a recombination event) were assayed further, again with 5 cultures per fluctuation test. Thirty-three gene deletion mutants displayed a statistically supported increase in recombination rate (Table S2, Figure 1D), using a one-sided Student's t-test with a cutoff of P = 0.05. Fluctuation tests of spontaneous recombination rates Fluctuation tests as designed by Luria and Delbrück (Luria and Delbrück 1943) were performed by transferring entire single colonies from YPD plates to 4 ml of YPD liquid medium. Cultures were grown at 30°to saturation. 100 ml of a 10 5 -fold dilution were plated on a fully supplemented SD plate and 200 ml of a 10 2 -fold dilution were plated on an SD-leu plate. Colonies were counted after incubation at 30°for 3 days. The number of recombinant (leu+) colonies per 10 7 viable cells was calculated, and the median value was used to determine the recombination rate by the method of the median (Lea and Coulson 1949). High-throughput replica pinning screen High-throughput manipulation of high-density yeast arrays was performed with the RoToR-HDA pinning robot (Singer Instruments). The MATa yeast deletion collection (EUROSCARF) was arrayed in 1536 format (each strain in quadruplicate). The leu2DEcoRI-URA3-leu2DBstEII marker to measure direct-repeat recombination (Smith and Rothstein 1999) was introduced into the deletion collection through synthetic genetic array (SGA) methodology (Kuzmin et al., 2016) using the JOY90 query strain. The procedure was performed twice in parallel to generate two sets of the yeast deletion collection containing the leu2 direct-repeat recombination reporter. Each plate of each set was then pinned onto six YPD+G418 plates (48 replicates per strain in total), incubated for one day at 30°and then scanned with a flatbed scanner. Subsequently, each plate was pinned onto SD-leu solid medium and incubated for two days at 30°to select recombination events. Finally, all plates were re-pinned on SD-leu solid medium and incubated for one day at 30°before scanning. Colony area measurement was performed using the ImageJ software package (Schneider et al., 2012) and the ScreenMill Colony Measurement Engine plugin (Dittmar et al., 2010), to assess colony circularity and size in pixels. Colony data were filtered to exclude artifacts by requiring a colony circularity score greater than 0.8. Colonies with a pixel area greater than 50% of the mean pixel area were scored for strains pinned to YPD+G418. Following replica-pinning to SD-leu, colonies were scored if the pixel area was greater than 10% of the mean pixel area for the same strain on YPD+G418. For each deletion strain, the ratio of recombinants (colonies on SD-leu) to total colonies (colonies on YPD+G418) is the recombinant frequency (Table S3). Strains where fewer than 10 colonies grew on YPD+G418 were removed from consideration, as were the 73 YKO collection strains carrying an additional msh3 mutation (Lehner et al., 2007). The final filtered data are presented in Table S4. Gene Ontology enrichment analysis and functional annotation GO term analysis was performed using the GO term finder tool (http://go.princeton.edu/) using a P-value cutoff of 0.01 and applying Bonferroni correction, querying biological process enrichment for each gene set. GO term enrichment results were further processed with REViGO (Supek et al., 2011) using the "Medium (0.7)" term similarity filter and simRel score as the semantic similarity measure. Terms with a frequency greater than 15% in the REViGO output were eliminated as too general. Gene lists used for the GO enrichment analyses are in Table 1, and the lists of enriched GO terms obtained are provided in Table S6. Human orthologs in Table 3 were identified using YeastMine (https://yeastmine.yeastgenome.org/ yeastmine; accessed June 25, 2019). Protein-protein interactions were identified using GeneMania (https://genemania.org/; (Warde-Farley et al., 2010)), inputting the 35 validated hyper-rec genes, and selecting only physical interactions, zero resultant genes, and equal weighting by network. Network edges were reduced to a single width and nodes were annotated manually using gene ontology from the Saccharomyces Genome Database (https://www.yeastgenome.org). Network annotations were made with the Python implementation of Spatial Analysis of Functional Enrichment (SAFE) ( (Baryshnikova 2016); https://github.com/baryshnikova-lab/safepy). The yeast genetic interaction similarity network and its functional domain annotations were obtained from . The genetic Figure 1 A genome-wide patching and replica plating screen for mutants with increased direct-repeat recombination. (A) The leu2 direct-repeat recombination assay. Spontaneous recombination between two leu2 heteroalleles, either through gene conversion or intra-chromosomal single strand annealing (SSA), yields a functional LEU2 gene. (B) Schematic representation of the screen based on patching and replica plating. The leu2 direct-repeat recombination cassette was introduced into the yeast deletion collection (YKO) by crossing the collection with a query strain containing the cassette. Haploid strains containing each gene deletion and the recombination cassette were isolated using SGA methodology. Each strain was patched on rich medium and replicaplated to selective medium, where hyper-recombinant mutants form papillae on the surface of the patch. Recombination rates were measured for positives from the patch assay using fluctuation tests. (C) Example plates from the patch assay. Each plate bears a negative control (wild type) and a positive control (elg1Δ). Two positive hits from the screen (rad4Δ, ydl162cΔ) are shown. (D) Recombination rates are plotted for the validated positives from the patch screen, alongside the wild-type strain. Each data point is from an independent fluctuation test, with n $ 3 for each strain. The vertical bars indicate the mean recombination rate for each strain. (E) The top 10 statistically supported GO terms enriched in the hits from the patch assay screen are shown, with the -fold enrichment for each term. Data availability Strains are available upon request. Table S1 lists all the strains used in this study. Table S2 contains the fluctuation test data from the patch screen. Table S3 contains the raw high-throughput replica pinning screen data. Table S4 contains the filtered pinning screen data. Table S5 contains the fluctuation test data from the pinning screen. A genetic screen for elevated spontaneous directrepeat recombination The leu2 direct-repeat recombination assay (Smith and Rothstein 1999) can detect both intra-chromosomal and sister chromatid recombination events ( Figure 1A). Two nonfunctional leu2 heteroalleles are separated by a 5.3 kb region containing the URA3 marker. Reconstitution of a functional LEU2 allele can occur either via gene conversion (either inter-or intra-chromatid), which maintains the URA3 marker, or via intra-chromosomal single strand annealing (SSA), where the URA3 marker and one of the leu2 repeats are lost (Symington et al., 2014). Both recombination events can be selected on media lacking leucine. We used this assay to systematically screen the yeast knockout (YKO) collection for genes whose deletion results in hyper-recombination between direct repeats ( Figure 1B). We introduced the leu2 direct-repeat recombination reporter into the YKO collection via synthetic n■ Table 1 Hyper-recombination genes from the patch assay and pinning assay screens genetic array (SGA) technology (Kuzmin et al., 2016). Each of the 4500 obtained strains was then patched on non-selective plates and replica-plated to plates lacking leucine to detect spontaneous recombination events as papillae on the replica-plated patches ( Figure 1C). We included a wild-type control and a hyper-recombinant elg1Δ control (Bellaoui et al., 2003;Ben-Aroya et al., 2003) on every plate for reference. The recombination rates for 195 putative hyper-rec mutants identified by replica-plating (Table S2) were measured by a fluctuation test. Strains with a recombination rate greater than 2x10 25 (approximately twofold of the wild-type rate; 38 strains) were assayed in triplicate (or more). Thirty-three gene deletion mutant strains with a statistically supported increase in direct-repeat recombination rate relative to the wild-type control were identified ( Figure 1D, Table S2, Table 1). The genes identified showed a high degree of enrichment for GO terms reflecting roles in DNA replication and repair ( Figure 1E). A high-throughput screen for altered spontaneous direct-repeat recombination We recently developed a high-throughput replica-pinning method to detect low-frequency events, and validated the scheme in a genomescale mutation frequency screen (Novarina et al., 2020). To complement the data obtained with the classical screening approach, and to test our new methodology independently of the replicative aging context in which it was developed, we applied it to detect changes in spontaneous direct-repeat recombination ( Figure 2A). We again introduced the leu2 direct-repeat recombination reporter into the YKO collection. The collection was then amplified by parallel highthroughput replica-pinning to yield 48 colonies per gene deletion strain. After one day of growth, all colonies were replica-pinned (twice, in series) to media lacking leucine to select for recombination events. Recombinant frequencies (a proxy for the spontaneous recombination rate) were calculated for each strain of the collection ( Figure 2B, Table S3, Table S4). As a reference, recombinant frequencies for the wild type (46%) and for a recombination-deficient rad54Δ strain (21%) obtained in a pilot replica-pinning experiment of 3000 colonies are indicated. In the screen itself, where 48 colonies were assessed, the wild type (his3Δ::kanMX) had a recombinant frequency of 56%. Notably, a group of strains from the YKO collection carry an additional mutation in the mismatch repair gene MSH3 (Lehner et al., 2007). Given the elevated spontaneous recombination rates of several mismatch repair-deficient strains ( Figure 1D), we suspected that these msh3 strains would display increased recombinant frequencies, independently of the identity of the intended gene deletion. Indeed, the distribution of recombinant frequencies for msh3 strains (median: 74%) is shifted toward higher values compared to the overall distribution of the YKO collection (median: 60%) ( Figure 2B). The 73 msh3 strains were excluded from further analysis. To explore the overall quality of the high-throughput replicapinning screen and to determine a cutoff in an unbiased manner, we performed Cutoff Linked to Interaction Knowledge (CLIK) analysis (Dittmar et al., 2013). The CLIK algorithm identified an enrichment of highly interacting genes at the top and at the bottom of our gene list (ranked according to recombinant frequency), confirming the overall high quality of our screen, and indicating that we were able to detect both hyper-and hypo-recombinogenic mutants ( Figure 2C). The cutoff indicated by CLIK corresponds to a recombinant frequency of 87% for the hyper-recombination strains (75 genes; Table 1), and of 33% for the recombination-deficient strains (122 genes; Table 2). Hyper-recombination genes: We assessed the functions of the 75 hyper-recombination genes identified by our high-throughput screen ( Figure 2D). As with the genes identified in the patch screen, the genes identified in the pinning screen were enriched for DNA replication and repair functions. Most importantly, at the very top of our hyper-recombination gene list (with 96-100% recombinant frequency), 11 out of 13 genes were identified in the patch screen and validated by fluctuation analysis (Table S2). We tested the two additional genes, CSM1 and NUP170, by fluctuation analysis, and found that both had a statistically supported increase in recombination rate ( Figure 2E and Table S5). Eighteen validated hyperrecombination genes from the patch screen were not identified in the pinning screen, and so are false negatives. Although we have not validated the weaker hits from the pinning screen (those with recombinant frequencies between 87% and 96%), four genes in this range were validated as part of the patch screen (APN1, RMI1, YLR235C, and RNH201), 9 caused elevated levels of Rad52 foci when deleted (APN1, NFI1, RMI1, POL32, RNH201, DDC1, HST3, MFT1, and YJR124C) (Alvaro et al., 2007;Styles et al., 2016), and 3 are annotated as 'mitotic recombination increased' (RMI1, DDC1, and HST3; Saccharomyces Genome Database). Together these data suggest that additional bona fide hyper-recombination genes were identified in the pinning screen. Hypo-recombination genes: By contrast to the replica-plating screen, the pinning screen detected mutants with reduced recombinant frequency, with 122 genes identified ( Table 2). The genes identified were functionally diverse, with no gene ontology (GO) processes enriched. Only 19 nonessential genes are annotated as having reduced recombination as either null or hypomorphic alleles in the Saccharomyces genome database (SGD; accessed January 11, 2020 via YeastMine). Of these, three genes (RAD52, LRP1, and THP1) were detected in the pinning screen. In addition, other members of the RAD52 epistasis group important for effective homologous recombination (RAD50, RAD54 and RAD55) displayed a recombinant frequency lower than 33%, and RAD51 was just above the cutoff (Table S3). Thus, our high-throughput replica-pinning approach detects mutants with very low recombinant frequencies. More generally, this observation suggests that if the pinning procedure is properly calibrated, a high-throughput replica-pinning screen is able not only to detect mutants with increased rates of a specific low-frequency event (in this case direct-repeat recombination), but also mutants with reduced rates of the same low-frequency event. Validated hyper-recombination genes identified in both screens: We compared the genes identified in the pinning screen with those identified in the patch screen, revealing 15 genes that were identified in both screens, a statistically supported enrichment ( Figure 3A; hypergeometric P = 1.2x10 221 ). Combining the results of the two screens, we validated 35 genes whose deletion results in elevated spontaneous direct-repeat recombination ( Table 3). Analysis of the group of 35 hyper-rec genes revealed 68 pairwise proteinprotein interactions ( Figure 3B), with many cases where several (if not all) members of the particular protein complex were identified. We found that 29 of the hyper-rec genes had at least one human ortholog (Table 3), indicating a high degree of conservation across the 35 validated genes. To assess the functional properties of the 35 gene hyper-rec set, we applied spatial analysis of functional enrichment (SAFE) (Baryshnikova 2016) to determine if any regions of the functional genetic interaction similarity yeast cell map are over-represented for the hyper-rec gene set ( Figure 3C). We found a statistically supported over-representation of the hyper-rec genes in the DNA replication and repair neighborhood of the genetic interaction cell map, highlighting the importance of accurate DNA synthesis in suppressing recombination. Finally, we compared the validated hyper-rec genes to relevant functional genomic instability datasets (Saccharomyces Genome Database annotation, (Alvaro et al., 2007;Hendry et al., 2015;Stirling et al., 2011;Styles et al., 2016); Figure 3D). Eight of our hyper-rec genes (HTA2, MSH6, YER188W, ABZ2, PMS1, MSH2, DFG16, and VMA11) were not identified in these datasets, indicating that our screens identified uncharacterized recombination genes. HTA2, MSH6, PMS1, MSH2 have recombination phenotypes reported (see Discussion). Thus, we identify four genes without a characterized role in preventing recombination: YER188W, ABZ2, DFG16, and VMA11. To infer gene function for the four genes lacking a characterized role in suppressing recombination, we again applied SAFE analysis (Baryshnikova 2016) to annotate the functional genetic interaction similarity yeast cell map to identify any regions that are enriched for genetic interactions with each of the four genes ( Figure 4). Of particular interest, the mitochondrial functional neighborhood is enriched for negative genetic interactions with YER188W Figure 2 A high-throughput replica-pinning screen for genes controlling direct-repeat recombination. (A) Schematic representation of the screen based on high-throughput replica-pinning. The leu2 direct-repeat recombination cassette was introduced into the yeast deletion collection as in Figure 1B. The resulting strains were amplified by parallel high-throughput replica pinning and subsequently replica-pinned to media lacking leucine to select for recombination events. Recombinant frequencies were calculated for each strain of the YKO collection. (B) Recombinant frequency distribution for the YKO collection (MSH3 strains) and for the msh3 strains in the collection. Recombinant frequencies for a wild-type and for a recombination-defective rad54Δ strain derived from a pilot experiment are indicated by the dashed lines. (C) Interaction densities determined by CLIK analysis are plotted as a two-dimensional heatmap. The cutoffs established by CLIK analysis for hyper-recombination (hyper-rec) and recombination-defective (hypo-rec) genes are shown in the insets. (D) The statistically supported GO terms enriched in the hits from the pinning assay screen are shown, with the enrichment for each term. (E) Recombination rates from fluctuation tests of csm1Δ and nup170Δ are plotted. Each data point is from an independent fluctuation test, with n = 3 for each strain. The vertical bars indicate the mean recombination rate for each strain and the wild-type data from Figure 1D are plotted for comparison. (Figure 4), suggesting that deletion of YER188W confers sensitivity to mitochondrial dysfunction. Analysis of DFG16 revealed enrichments for positive interactions in the RIM signaling neighborhood, which is expected (Barwell et al., 2005), but also for negative interactions in the DNA replication region of the map (Figure 4), indicating that DFG16 is important for fitness when DNA replication is compromised. Analysis of VMA11 revealed enrichment in the vesicle trafficking neighborhood, typical of vacuolar ATPase subunit genes, and analysis of ABZ2 revealed little (Figure 4). We conclude that functional analysis suggests mechanisms by which loss of YER188W (oxidative stress) or DFG16 (genome integrity) results in hyper-recombination. DISCUSSION We report here the first systematic, genome-wide approach to identify genes that affect direct-repeat recombination. By combining the classical patch and replica-plate method and our new replica-pinning approach, we identified many genes already implicated in homologous recombination, as well as genes with no previous connection to recombination. We failed to identify several genes known to suppress direct-repeat recombination-e.g., SRS2 and HPR1 (Aguilera and Klein 1988)-but this is not surprising, since most, if not all, genome-wide screens have false negatives. Here we briefly discuss the functions of the genes and complexes identified in the screens and subsequently validated by fluctuation analysis. Mismatch repair MLH1, MSH2, MSH6 and PMS1 are evolutionarily conserved genes involved in mismatch repair (MMR), a pathway that detects and corrects nucleotide mismatches in double-strand DNA (Spies and Fishel 2015). An anti-recombinogenic role for these four MMR genes in yeast has been previously described: specifically, MMR proteins are important to prevent homeologous recombination and SSA between slightly divergent sequences, via mismatch recognition and heteroduplex rejection (Datta et al., 1996;Nicholson et al., 2000;Spies and Fishel 2015;Sugawara et al., 2004). The role for MMR in preventing homeologous recombination is conserved also in mammalian cells (de Wind et al., 1995;Elliott and Jasin 2001;Spies and Fishel 2015). It is worth noting that the presence of sequence differences between the two leu2 alleles in the leu2 direct-repeat assay is essential to genetically detect recombination events. Therefore, it is reasonable that this assay should detect genes involved in suppressing homeologous recombination. n■ Sgs1-Top3-Rmi1 complex The evolutionarily conserved helicase-topoisomerase complex Sgs1-Top3-Rmi1 is involved in DSB resection and in dissolution of recombination intermediates (Symington et al., 2014). Consistent with previous observations (Chang et al., 2005), our screen identified all three members of the complex, together with YLR235C, a dubious ORF that overlaps the TOP3 gene. The Sgs1-Top3-Rmi1 complex dissolves double Holliday junction structures to prevent crossover formation (Cejka et al., 2010). The same role has been reported for BLM helicase, the human Sgs1 homolog mutated in the genome stability disorder Bloom syndrome (Wu et al., 2006;Yang et al., 2010). Furthermore, several genetic studies indicate that the antirecombinogenic activity of Sgs1-Top3-Rmi1 cooperates with MMR proteins in heteroduplex rejection to prevent homeologous recombination (Chakraborty et al., 2016;Goldfarb and Alani 2005;Myung et al., 2001;Spell and Jinks-Robertson 2004;Sugawara et al., 2004). MGS1 In our screen we also identified MGS1, the homolog of the WRN-interacting protein WRNIP1. Mgs1 displays DNA-dependent ATPase and DNA strand annealing activities. Deletion of MGS1 causes hyper-recombination, including elevated direct-repeat recombination (Hishida et al., 2001). It seems that Mgs1 promotes faithful DNA replication by regulating Pold, and promoting replication fork restart after stalling (Branzei et al., 2002;Saugar et al., 2012). The absence of Mgs1 could result in increased replication fork collapse, leading to the formation of recombinogenic DSBs (Branzei et al., 2002). Similar roles have been suggested for WRNIP1 in mammalian cells (Leuzzi et al., 2016;Tsurimoto et al., 2005). RNase H2 complex RNH201 encodes the evolutionarily conserved catalytic subunit of RNase H2, while the two non-catalytic subunits are encoded by RNH202 and RNH203 genes. This enzyme cleaves the RNA moiety in RNA-DNA hybrids originating from Okazaki fragments, co-transcriptional R-loops, and ribonucleotide incorporation by replicative polymerases (Cerritelli and Crouch 2009). Deletion of any of the three subunits in yeast inactivates the whole complex. Human RNase H2 genes are mutated in Aicardi-Goutières syndrome, a severe neurological disorder (Crow et al., 2006). Inactivation of yeast RNase H2 causes elevated LOH, ectopic recombination and direct-repeat recombination (Conover et al., 2015;Potenski et al., 2014), mostly dependent on Top1 activity. What is the recombinogenic intermediate accumulated in the absence of RNase H2? It has been suggested that Top1-dependent cleavage at the ribonucleotide site creates a nick that can be further converted into a recombinogenic DSB (Potenski et al., 2014). Recent genetic studies indicate that, while in the case of LOH events hyper-recombination is caused by Top1-dependent processing of single ribonucleotides incorporated by leading strand polymerases and/or by accumulation of recombinogenic R-loops (Conover et al., 2015;Cornelio et al., 2017;Keskin et al., 2014;O'Connell et al., 2015), elevated direct-repeat recombination results instead from Top1-dependent cleavage of stretches of ribonucleotides, resulting from defective R-loop removal or Okazaki fragment processing in the absence of RNase H2 (Epshtein et al., 2016). In line with this model, we also detected elevated direct-repeat recombination rate in the absence of the Thp2 member of the THO complex, which functions at the interface between transcription and mRNA export to prevent R-loop accumulation (Chavez et al., 2000;Huertas and Aguilera 2003), DST1, which encodes a transcription elongation factor and is anti-recombinogenic (Owiti et al., 2017), and the flap endonuclease encoded by RAD27, which is involved in Okazaki fragment processing (Balakrishnan and Bambara 2013) (Table 3). Furthermore, deletion of the dubious ORF YDL162C, also identified in our screen, likely affects the expression level of neighboring CDC9, an essential gene encoding DNA Ligase I, involved in Okazaki fragment processing and ligation after ribonucleotide removal from DNA. Together, available data suggest that different modes n■ leading to accumulation of RNA-DNA hybrids or unprocessed Okazaki fragments result in hyper-recombination. RRM3 The RRM3 gene, encoding a 59 to 39 DNA helicase, was initially identified because its absence causes hyper-recombination between endogenous tandem-repeated sequences (such as the rDNA locus and the CUP1 genes) (Keil and McWilliams 1993). The Rrm3 helicase travels with the replication fork and facilitates replication through genomic sites containing protein-DNA complexes that, in its absence, cause replication fork stalling and breakage. Such Rrm3-dependent sites include the rDNA, telomeres, tRNA genes, inactive replication origins, centromeres, and the silent mating-type loci (Azvolinsky et al., 2006;Ivessa et al., 2003Ivessa et al., , 2000Schmidt and Kolodner 2004;Torres et al., 2004). Intriguingly, a tRNA gene is located about 350 bp upstream the chromosomal location of the leu2 direct-repeat recombination marker. Increased replication fork pausing in the absence of Rrm3 could cause recombinogenic DSBs, explaining the elevated direct-repeat recombination we observe in the rrm3Δ strain. Oxidative stress response genes YAP1 and SKN7 encode two transcription factors important for the activation of the cellular response to oxidative stress (Morano et al., 2012). The glutathione peroxidase encoded by HYR1 has a major role in activating Yap1 in response to oxidative stress (Delaunay et al., 2002). TSA1 is a Yap1 and Skn7 target and encodes a peroxiredoxin that scavenges endogenous hydrogen peroxide (Wong et al., 2004). Deletion of TSA1 causes hyper-recombination between inverted repeats (Huang and Kolodner 2005), and oxidative stress response genes (including TSA1, SKN7 and YAP1) are synthetic sick or lethal with HR mutants (Pan et al., 2006;Yi et al., 2016). A likely explanation for the elevated direct-repeat recombination we measured in strains defective for the oxidative stress response, therefore, is that oxidative DNA damage generates replication blocking lesions and/or replication-associated DSBs, both of which are processed by the HR pathway (Huang and Kolodner 2005). An alternative explanation could be that extensive oxidative DNA damage results in the saturation of the mismatch-binding step of MMR, compromising MMR-dependent heteroduplex rejection, resulting in increased homeologous recombination (Hum and Jinks-Robertson 2018;Spies and Fishel 2015). Other DNA repair genes APN1 encodes the main apurinic/apyrimidinic (AP) endonuclease involved in yeast base excision repair (BER). Removal of endogenous alkylating damage can generate abasic sites, which are mostly processed by Apn1 (Boiteux and Guillet 2004;Popoff et al., 1990;Xiao and Samson 1993). In the absence of APN1, abasic sites accumulate, which can hamper DNA replication. The recombination pathway is involved in the repair and/or bypass of these abasic sites, as suggested by the genetic interactions between the BER and the HR pathways (Boiteux and Guillet 2004;Swanson et al., 1999;Vance and Wilson 2001). The APN1 gene is adjacent to RAD27, and therefore it is also possible that the hyper-recombination phenotype of apn1Δ is due to a "neighbouring-gene effect" on RAD27, as was reported in the case of telomere length alteration (Ben-Shitrit et al., 2012). HTA2, which encodes one copy of histone H2A, is of course important for appropriate nucleosome assembly. Reducing histone levels by deleting one H3-H4 gene pair or by partial depletion of H4 increases recombination (Clemente-Ruiz and Prado 2009;Liang et al., 2012;Prado and Aguilera 2005), and it is likely that reducing HTA2 gene dosage also does so. Since histone depletion results in diverse chromatin defects, the exact mechanisms by which recombination is induced are elusive. RAD4 encodes a key factor of nucleotide excision repair (NER), and is involved in direct recognition and binding of DNA damage (Prakash and Prakash 2000), while RAD6 is a key gene controlling the post replication repair (PRR) DNA damage tolerance pathway (Ulrich 2005). Genetic studies suggest that BER, NER, PRR and HR Figure 4 Spatial analysis of functional enrichment for four hyper-rec genes. The genetic interactions of each of the indicated genes was tested for enrichments in the functional neighborhoods of the yeast genetic interaction similarity network. The overlay indicates a subset of functional domains as annotated on Figure 3C. Nodes with statistically supported enrichments (Neighborhood enrichment P , 0.05) are colored, black for negative genetic interactions and red for positive genetic interactions. can redundantly process spontaneous DNA lesions, and inactivation of one pathway shifts the burden on the others. This mechanism could explain why deletion of RAD4 or RAD6 causes a modest increase in spontaneous direct-repeat recombination (Swanson et al., 1999). CSM1 encodes a nucleolar protein that serves as a kinetochore organizer to promote chromosome segregation in meiosis, and is involved in localization and silencing of rDNA and telomeres in mitotic cells (Poon and Mekhail 2011). Interestingly, Csm1 is important to inhibit homologous recombination at the rDNA locus and other repeated sequences (Burrack et al., 2013;Huang et al., 2006;Mekhail et al., 2008). The nuclear pore complex has an intimate connection to recombination, in that some DSBs move to and are likely repaired at the NPC (Freudenreich and Su 2016). The NPC gene NUP170 has not been directly implicated in DSB repair, but is important for chromosome segregation (Kerscher et al., 2001). The unknowns (YER188W, ABZ2, DFG16, and VMA11) Unexpectedly, the top hyper-rec gene identified in our screen is VMA11, which encodes a subunit of the evolutionarily conserved vacuolar H + -ATPase (V-ATPase), important for vacuole acidification and cellular pH regulation (Hirata et al., 1997;Kane 2006;Umemoto et al., 1991). VMA11 involvement in genome maintenance is suggested by the sensitivity of a vma11Δ strain to several genotoxic agents, namely doxorubicin, ionizing radiation, cisplatin and oxidative stress (Thorpe et al., 2004;Xia et al., 2007). V-ATPase defects in yeast result in endogenous oxidative stress and defective Fe/S cluster biogenesis as a consequence of mitochondrial depolarization (Hughes and Gottschling 2012;Milgrom et al., 2007;Veatch et al., 2009). Of note, several DNA replication and repair factors are Fe/S cluster proteins (Veatch et al., 2009;Zhang 2014). Therefore, the hyper-recombination phenotype of vma11Δ could be due to increased spontaneous DNA damage, caused by elevated endogenous oxidative stress and/or by defective DNA replication and repair as a consequence of compromised Fe/S cluster biogenesis. However, VMA11 was not detected in screens for increased Rad52 foci (Alvaro et al., 2007;Styles et al., 2016), or in a screen for increased DNA damage checkpoint activation (Hendry et al., 2015), suggesting that spontaneous DNA damage might not accumulate to high levels in vma11Δ. ABZ2 encodes an enzyme involved in folate biosynthesis (Botet et al., 2007). Folate deficiency and the resulting compromise of nucleotide synthesis could promote recombination, although yeast culture media are rich in folate, and the ABZ2 genetic interaction profile reveals no similarity to nucleotide biosynthesis genes (Usaj et al., 2017). DFG16 encodes a predicted transmembrane protein involved in pH sensing (Barwell et al., 2005). Interestingly, SAFE analysis indicates a role for DFG16 in DNA replication and/or DNA repair, in addition to the expected role in pH signaling. There is currently little insight into the function of YER188W. SAFE analysis indicates a possible role in mitochondrial function, however a protein product of YER188W has not been detected to date in either mass spectrometry or GFP fusion protein analyses (Breker et al., 2014;Ho et al., 2018;Huh et al., 2003). In summary, despite direct-repeat recombination having been studied for decades, with our combined screening approach we were able to identify novel genes that affect this process, several of which are evolutionarily conserved. Since repeated sequences are abundant in mammalian genomes, our findings might be important for future studies on recombination and genome integrity in human cells. In addition, our high-throughput screening approach will likely be useful to study other cellular processes that occur at low frequency.	2020-02-13T09:13:15.969Z	2020-02-11T00:00:00.000	{ "year": 2020, "sha1": "5e70e44a6f3f3c66fea0bcca1ea4e9a2b4ceecf6", "oa_license": "CCBY", "oa_url": "https://academic.oup.com/g3journal/article-pdf/10/6/1853/37124426/g3journal1853.pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "c16c182969599ee917ef63de4949b162d36ae822", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [ "Biology" ] }
268251564	pes2o/s2orc	v3-fos-license	A metagenomic investigation of phytoplasma diversity in Australian vegetable growing regions Abstract In this study, metagenomic sequence data was used to investigate the phytoplasma taxonomic diversity in vegetable-growing regions across Australia. Metagenomic sequencing was performed on 195 phytoplasma-positive samples, originating either from historic collections (n=46) or during collection efforts between January 2015 and June 2022 (n=149). The sampled hosts were classified as crop (n=155), weed (n=24), ornamental (n=7), native plant (n=6), and insect (n=3) species. Most samples came from Queensland (n=78), followed by Western Australia (n=46), the Northern Territory (n=32), New South Wales (n=17), and Victoria (n=10). Of the 195 draft phytoplasma genomes, 178 met our genome criteria for comparison using an average nucleotide identity approach. Ten distinct phytoplasma species were identified and could be classified within the 16SrII, 16SrXII (PCR only), 16SrXXV, and 16SrXXXVIII phytoplasma groups, which have all previously been recorded in Australia. The most commonly detected phytoplasma taxa in this study were species and subspecies classified within the 16SrII group (n=153), followed by strains within the 16SrXXXVIII group (‘Ca. Phytoplasma stylosanthis’; n=6). Several geographic- and host-range expansions were reported, as well as mixed phytoplasma infections of 16SrII taxa and ‘Ca. Phytoplasma stylosanthis’. Additionally, six previously unrecorded 16SrII taxa were identified, including five putative subspecies of ‘Ca. Phytoplasma australasiaticum’ and a new putative 16SrII species. PCR and sequencing of the 16S rRNA gene was a suitable triage tool for preliminary phytoplasma detection. Metagenomic sequencing, however, allowed for higher-resolution identification of the phytoplasmas, including mixed infections, than was afforded by only direct Sanger sequencing of the 16S rRNA gene. Since the metagenomic approach theoretically obtains sequences of all organisms in a sample, this approach was useful to confirm the host family, genus, and/or species. In addition to improving our understanding of the phytoplasma species that affect crop production in Australia, the study also significantly expands the genomic sequence data available in public sequence repositories to contribute to phytoplasma molecular epidemiology studies, revision of taxonomy, and improved diagnostics. Abstract In this study, metagenomic sequence data was used to investigate the phytoplasma taxonomic diversity in vegetable-growing regions across Australia.Metagenomic sequencing was performed on 195 phytoplasma-positive samples, originating either from historic collections (n=46) or during collection efforts between January 2015 and June 2022 (n=149).The sampled hosts were classified as crop (n=155), weed (n=24), ornamental (n=7), native plant (n=6), and insect (n=3) species.Most samples came from Queensland (n=78), followed by Western Australia (n=46), the Northern Territory (n=32), New South Wales (n=17), and Victoria (n=10).Of the 195 draft phytoplasma genomes, 178 met our genome criteria for comparison using an average nucleotide identity approach.Ten distinct phytoplasma species were identified and could be classified within the 16SrII, 16SrXII (PCR only), 16SrXXV, and 16SrXXXVIII phytoplasma groups, which have all previously been recorded in Australia.The most commonly detected phytoplasma taxa in this study were species and subspecies classified within the 16SrII group (n=153), followed by strains within the 16SrXXXVIII group ('Ca.Phytoplasma stylosanthis'; n=6).Several geographic-and host-range expansions were reported, as well as mixed phytoplasma infections of 16SrII taxa and 'Ca.Phytoplasma stylosanthis'.Additionally, six previously unrecorded 16SrII taxa were identified, including five putative subspecies of 'Ca.Phytoplasma australasiaticum' and a new putative 16SrII species.PCR and sequencing of the 16S rRNA gene was a suitable triage tool for preliminary phytoplasma detection.Metagenomic sequencing, however, allowed for higher-resolution identification of the phytoplasmas, including mixed infections, than was afforded by only direct Sanger sequencing of the 16S rRNA gene.Since the metagenomic approach theoretically obtains sequences of all organisms in a sample, this approach was useful to confirm the host family, genus, and/or species.In addition to improving our understanding of the phytoplasma species that affect crop production in Australia, the study also significantly expands the genomic sequence data available in public sequence repositories to contribute to phytoplasma molecular epidemiology studies, revision of taxonomy, and improved diagnostics. INTRODUCTION Phytoplasmas are a diverse monophyletic clade of unculturable bacteria in the class Mollicutes within the provisional genus 'Candidatus Phytoplasma' [1].They are phloem-limited plant pathogenic bacteria that are transmitted by phloem-feeding hemipteran insects [2].Diseases associated with phytoplasma infections have been described in over 700 plant hosts globally from agriculturally and horticulturally important crops, ornamental plants, weeds, and native plants [3][4][5]. In Australia, phytoplasma-like symptoms were first recorded in the 1900s from tomato (Tomato Big Bud, TBB) [6], lucerne (Lucerne Witches'-Broom) [7], and pasture legumes (Legume Little Leaf) [8].All were thought to be vector-transmitted viruses until mycoplasma-like structures were identified by electron microscopy in fabaceous plants showing symptoms of little leaves and spindled stems [9].By the 1990s, molecular detections of phytoplasmas were being made globally [10], including in Australia [11], by PCR amplification of the 16S rRNA gene region.The molecular, PCR-based techniques facilitated the screening of many more plants for phytoplasma infections than was previously possible by microscopy and serological techniques.These PCR-based techniques involved restriction fragment length polymorphism (RFLP) analysis and/or sequencing for phytoplasma characterization and identification.Phytoplasma surveys were done in Australia using these molecular-based approaches and a large diversity of phytoplasma taxa affecting grains, legumes, fruit and vegetable crops, ornamentals, native plants, weeds and putative phytoplasma vectors were uncovered [12][13][14][15][16][17][18].To date, twelve phytoplasma 16 Sr groups and five 'Candidatus Phytoplasma' representatives have been described from Australia [19][20][21][22][23][24][25].Members within the 16SrII group are the most commonly detected phytoplasma in Australia, affecting a large number of host species [12,13,16,26].However, members within the 16SrXII group have been described as the most economically important based on their association with yield-reducing diseases in high-value crops such as grapevines and strawberries [26][27][28].Competent vector species remain to be confirmed for many of the phytoplasma taxa present in Australia, although Orosius argentatus has been shown to transmit diseases associated with 16SrII phytoplasmas [8,9,29,30]. Phytoplasma diversity and taxonomic analyses, including studies in Australia, have largely relied on the 16S rRNA gene sequence, which only offers low-resolution analyses of inter-and intraspecies diversity [1,12,31,32].Diversity assessments and species delimitation studies have also involved higher-resolution analyses of three to nine additional housekeeping genes for multilocus sequence typing (MLST) and analysis (MLSA) [24,33,34].With the decreasing cost of high-throughput sequencing (HTS) and increasing sequence data outputs, as well as the advancements in bioinformatic tools for metagenomic data assessments, phytoplasma genomes have increasingly been used to understand their taxonomy [31] and biology [35].Obtaining draft or complete phytoplasma genomes allows for higher-resolution analyses than those based on one or a few genes for diversity analyses [36].Additionally, by applying genome-based species delimitation thresholds and criteria specified for culturable bacteria, taxonomic boundaries between phytoplasma strains can be identified [36,37]. The aim of this study was to assess and update the species and genetic diversity of phytoplasmas in vegetable growing regions in the various states and territories of Australia using whole-genome-based approaches.To this end, (i) plants displaying phytoplasma associated symptoms and insects were collected from vegetable growing regions in Australia and screened by PCR for the presence of phytoplasma infections, (ii) DNA from key historical phytoplasma strains from previous phytoplasma surveys Impact Statement Phytoplasmas are unculturable, plant pathogenic bacteria that infect and impact yield of many agriculturally important plant species.In this study, 16S rRNA gene detection and sequencing for triaging was coupled with metagenomic sequencing to determine the diversity of phytoplasma taxa and associated diseases in Australian vegetable growing regions.It is the first study to use metagenomic analysis to improve the understanding of phytoplasma diversity in Australia.None of the phytoplasma taxa that were detected were exotic to Australia, but host-and geographic-range expansions were recorded for some.Since the metagenomic approach obtains DNA sequences from all organisms in a sample, the identification of plant and insect host families, genera, or species was possible by DNA barcode analysis when they were undetermined based on morphology.This study has analysed the largest number of phytoplasma whole genomes to date (n=195) and significantly contributes to the available sequence data for these bacteria.The sequence and metadata provided in this study offer an improved understanding of the phytoplasma taxa present in the different states and territories of Australia, and contributes to improving phytoplasma taxonomy, molecular epidemiology, and diagnostics. in Australia were obtained, and (iii) a metagenomic approach was used to obtain draft phytoplasma genome assemblies to be used for genome-based investigations into the phytoplasma taxa infecting the samples, and to identify the host when the plant or insect host identity was inconclusive based on morphology.The results of this study demonstrate the applications, benefits, and challenges of applying metagenomic sequencing to phytoplasma diversity analyses. Sample collection, total nucleic acid extraction, plant tissue preservation, and DNA quantification Plant samples with phytoplasma-associated symptoms, including little leaf, yellowing, phyllody, and/or stunting that were collected Australia-wide between March 2019 and June 2022 were sent to the laboratory in Melbourne, Victoria (VIC) for analysis (Table S1).Total nucleic acid was extracted from samples using an iodixanol-based phytoplasma enrichment procedure [38] or a modified CTAB-DNeasy protocol without RNase treatment [39] (Table S1 and S2).Petioles, whole leaves or leaf veins were used in total nucleic acid extractions for most samples.Phloem scrapings were sampled for woody material (e.g.Melaleuca spp.and Vitaceae spp.).When possible, a subsample of the plant material was freeze dried for at least 72 h at −50 °C in individually labelled screw cap tubes using the FreeZone 2.5 Liter Benchtop Freeze Dry System (Labconco, MO, USA) and deposited in the Victorian plant pathology herbarium (VPRI) (Table S1).Insect samples were supplied as DNA extracts (Table S1) and had been collected by suction trapping and sweep netting in Jennings, New South Wales (NSW) and Palmerston, Northern Territory (NT), respectively.Insect collections were done in these areas as plants displaying phytoplasma-associated symptoms were present nearby.Phytoplasma-positive samples collected prior to 2019 that are held at The Northern Territory Department of Industry, Tourism and Trade (NT DITT), Darwin, NT, Australia and Department of Agriculture and Fisheries, Mareeba, Queensland (QLD), Australia were also supplied as total nucleic acid extracts (Table S1). DNA quantity was estimated using a Qubit 2.0 fluorometer (Thermo Fisher Scientific, MA, USA) with the Qubit 1X dsDNA HS Assay Kit (Thermo Fisher Scientific, MA, USA).All DNA samples were stored at −20 °C. Phytoplasma screening and preliminary identification by universal phytoplasma 16Sr PCR, and Sanger sequencing Screening for PCR inhibitors was done using PCR primers for the generic amplification of the bacterial 16S rRNA gene [40].A nested PCR assay using P1/P7 and R16F2n/m23sr primer pairs were used to screen recent samples for phytoplasma infection and to confirm phytoplasma presence in total nucleic acid extracts from samples collected prior to 2019 [41].These primers all bind to the phytoplasma 16S rRNA gene, apart from the P7 primer, which binds to the 5′ region of the 23S rRNA gene.The R16F2n/m23sr amplicon of some samples were cloned and screened according to [24] when a poor Sanger sequencing quality was observed in the forward and/or reverse read (Table S1) [24].All PCR amplicons were visualized by electrophoresis through 1 % agarose gels stained with SYBR Safe DNA gel stain (Thermo Fisher Scientific, MA, USA).PCR amplicons of the expected size were purified and directly Sanger sequenced (Macrogen, Seoul, South Korea).The identities of the Sanger sequenced PCR amplicons were determined by blastn analysis [42] at the NCBI (https://www.ncbi.nlm.nih.gov/, last accessed 30 July 2022).During the blastn analyses, the top hit with the 16S rRNA gene of a 'Ca.Phytoplasma' species reference strain was used to determine the identity of the sample investigated based on the top bit score, percent identity, and e-value, as well as considering the query coverage. Library preparation and sequencing Libraries were prepared with fragment sizes between 300 and 500 bp by following the manufacturer's protocols of the NEXTFLEX Rapid XP DNA-Seq Kit (PerkinElmer, MA, USA) with the Unique Dual Index (UDI) barcodes (PerkinElmer, MA, USA) or the Nextera DNA Flex Library Preparation Kit (Illumina, CA, USA) with the IDT for Illumina Nextera DNA Unique Dual (UD) Indexes (Illumina, CA, USA) (Table S1).The libraries were pooled and size-selected prior to sequencing according to [38] and sequenced with Illumina platforms including the MiSeq (2×250 bp), HiSeq2000 (2×150 bp), NovaSeq 6000 on an SP flow cell (2×250 bp), or NovaSeq 6000 on an S1 flow cell (2×150 bp).All the library pools sequenced on the NovaSeq 6000 platform were treated with the Illumina Free Adapter Blocking Reagent (Illumina, CA, USA) according to the manufacturer protocols prior to HTS to mitigate aberrant sequencing results caused by the presence of free adapters. The phytoplasma genome sequences of 25 phytoplasma-positive samples have been used in previous studies [24,37,38].The genome sequences and metadata associated with these samples were included in this study, however, they were gathered during the sample collection period of this study (Table S2). Read quality filtering, metagenomic assembly, and identification of phytoplasma-derived contigs and gene annotations Illumina read filtering and adapter trimming for each sample was done using FastP [43], removing reads shorter than 50 bp and with a Phred quality score (Q score) below Q20.The trimmed reads were used in a metagenomic assembly pipeline according to [38], which implements metaSPAdes version 3.15.2[44,45].Phytoplasma-derived contigs were identified and retrieved using blast+v2.11.0 [42] and a custom grep script, respectively.Contigs shorter than 500 bp were removed using the reformat.sh script implemented in the BBMap v.38.61b software suite [46].The phytoplasma genomes were analysed in metaQUAST [47] to estimate the genome N50 values.Protein coding, tRNA, and rRNA genes were annotated and counted using Prokka [48], specifying RNAmmer for 5S, 16S, and 23S rRNA gene annotations [49]. The phytoplasma genomes obtained in this study were analysed differently based on the total genome size recovered to mitigate spurious results related to highly incomplete and fragmented genomes.The number of tRNA gene sequences annotated were used to estimate the completeness of the phytoplasma assemblies and, therefore, their suitability for the downstream genomebased analyses.Phytoplasma genomes that were larger than 300,000 bp and which encoded 13 or more tRNA genes were used in further whole-genome-based assessments to classify the phytoplasma strains (Table S1).For genomes smaller than 300,000 bp and/or encoding fewer than 13 tRNA genes, only the 16S rRNA gene was used to classify the phytoplasma strain within a 16 Sr group (Table S1). Whole-genome analyses Whole-genome comparisons were performed for all the phytoplasma genomes by average nucleotide identity (ANI) analysis using the ANI with MUMmer (ANIm) algorithm in pyani version 0.2.10 [50].The pyani heatmap output was manually overlaid with blastn results to characterize clusters, which did not contain a representative genome. For further investigation into phytoplasma samples that did not clearly cluster with reference sequences in the initial ANI analysis, the coverage of aligned genomic segments in each pairwise comparison was analysed along with publicly available phytoplasma reference genomes using pyani version 0.20.10.These values are referred to as the alignment fraction (AF). In cases where a historic strain was previously classified using in vitro RFLP but appeared to be a different species based on the maximum-likelihood tree, the recognition sites for 14 of the 17 restriction enzymes used in phytoplasma subgroup classification were visualized in Geneious Prime version 2022.2.2 (https://www.geneious.com/prime/).The three restriction enzymes that were not used as they are unavailable on Geneious Prime version 2022.2.2 are: BfaI, BstUI (ThaI), and SspI. Identification of mixed phytoplasma infections using whole-genome analyses Samples which demonstrated high ANI values (>90 % ANI) with more than one representative phytoplasma genome were considered to have a mixed phytoplasma infection of the phytoplasma taxa for which genome sequence data was available.The 16S rRNA gene sequence analyses, including those of the cloned PCR amplicons, as well as the number of annotated tRNA genes were revisited for the mixed infection samples identified during the blastn analyses. Identification of unknown host species The host family, genus, or species of some plant hosts and two insect hosts were unconfirmed based on external morphology prior to total nucleic acid extraction (Table S1).To confirm the identities of the insects to the species-level, contigs encoding cytochrome c oxidase subunit 1 (coI) were analysed [51].To confirm the family, genus, and/or species of the unknown plant hosts, the contigs encoding maturase K (matK) and ribulose-bisphosphate carboxylase (rbcL) genes were analysed [52,53].These barcoding genes were obtained from the metagenome assemblies using a custom grep script to identify the genes and their identities from the blastn results.A publicly available accession was listed as the top blastn hit when it produced the highest bit-score, lowest e-value, and highest percentage identity.In cases where multiple species in the same genus were considered the top blastn hit, only the genus was recorded for the sample and the species names were considered undetermined.In cases where multiple genera were listed as top blastn hits, the family of these samples were recorded, and the genus and species names were considered undetermined.The detection location of host species listed as top hits were determined based on searches performed at the Australasian Virtual Herbarium website (https://avh.chah.org.au/, last accessed March 2023), with hosts that are not known to be present in Australia removed from the list. Phytoplasma-positive sample information A total of 195 samples were collected between 1998 and 2022 and were either confirmed or suspected to be infected with phytoplasma based on previous PCR-based analyses (sequence similarity and RFLP) or disease symptoms.These samples were subsequently confirmed to be positive for phytoplasma by PCR and direct Sanger sequencing of the amplicon generated using the universal phytoplasma 16S rRNA primers.Most of the samples were collected from QLD (n=78), followed by Western Australia (WA, n=46), NT (n=32), NSW (n=29), and VIC (n=10) (Tables 1 and S1). Phytoplasma samples analysed in this study were collected from all the states and territories of Australia, apart from the Australian Capital Territory (ACT), South Australia (SA), and Tasmania (TAS) (Tables 1 and S1).No samples were collected between 2019 and 2022 in SA due to Covid-19 travel restrictions.There was also an absence of plants showing typical phytoplasma symptoms in TAS during the 2019-2022 collection period (Callum R. Wilson, personal communication), which corresponds to previous observations of low phytoplasma prevalence for the state [54]. An asterisk () in the sample name(s) column indicates samples for which the host identity was determined using DNA barcode analysis. Metagenomic sequencing output and phytoplasma genome information After metagenomic HTS, 178 of the 195 total samples passed the phytoplasma genome criteria for further comparative genomic analyses in this study (Table S1).The Illumina sequence data output for these 178 samples ranged from 0.31 Gb (sample BAWM-193a-F1) to 32.60 Gb (sample BAWM-354A) with an average output of 4.76 Gb per metagenomic library (Table S1).The phytoplasma genome sizes of these 178 samples ranged between 321 651 bp (sample BAWM-201) and 1 488 020 bp (sample BAWM-255), with an average genome size of 632,634 bp for of all 178 phytoplasma samples (Table S1).Of the 178 draft genomes, an average of 28 tRNA gene sequences were recovered per genome and ranged between 13 tRNA genes (sample BAWM-198) to 61 tRNA genes (sample BAWM-255) (Table S1).The most tRNA genes recovered from a complete phytoplasma genome to date is 35 from the genome of 'Ca.Phytoplasma australiense' strain NZSb11 [57], indicating that the phytoplasma genome data of 13 of the 178 samples that had more than 35 tRNA genes annotated may represent tRNAs of more than one phytoplasma species in the sample (Table S1).An average of two phytoplasma rRNA genes could be annotated from these 13 genomes, with a max of five rRNA genes (sample BAWM-307) and none obtained from sample BAWM-350 (Table S1).To date, two identical or nearly identical 16S rRNA genes are known to be encoded per phytoplasma genome [1,36] indicating that sample BAWM-307 potentially harbours a mixed phytoplasma infection. Phytoplasma genome sequences that were <300 000 bp were recovered for 17 samples in six different host families (Table S1).These genome sequences ranged in size from 1784 bp (sample BAWM-004) to 289 060 bp (sample BAWM-184), with an average size of 114 181 bp (Table S1).The average data output for these 17 samples was 5.07 Gb, with a range of 1.28 Gb (sample BAWM-173) to 22.53 Gb (sample BAWM-189) (Table S1).An average of six tRNA genes could be retrieved from these 17 phytoplasma genomes (range 0 tRNA genes for samples BAWM-003, BAWM-004, BAWM-083, BAWM-216, and BAWM-233 to 21 tRNA genes for BAWM-183).No rRNA genes were retrieved from seven of the 17 phytoplasma genomes.These samples were not used in further genomic-based analyses as these results indicate poor-quality genomes from which limited information can be obtained, including the 16S rRNA gene-based correlation of taxon identification prior to and after metagenomic HTS [58]. Subspecies of 'Ca. Phytoplasma australasiaticum' Based on whole-genome ANI analyses (Fig. 1; Table S1), 160 of the 178 samples (ca.90 %) used in further genome-based analyses clustered at >96 % ANI solely with representative genome sequences of 16SrII phytoplasmas.The majority of these samples classified within the 16SrII phytoplasma group were identified as 'Ca.Phytoplasma australasiaticum' subspecies, including of 'Ca.Phytoplasma australasiaticum subsp.australasiaticum' (n=67), 'Ca.Phytoplasma australasiaticum subsp.ipomoeae' (n=51), and strains identified as a new 'Ca.Phytoplasma australasiaticum' subspecies (n=12, referred to as 'Ca.Phytoplasma australasiaticum' taxon 1) (Fig. 1).When the 12 16S rRNA sequences of the 'Ca.Phytoplasma australasiaticum' taxon 1 samples extracted from the genomic sequences were queried further, the historic samples were classified as 'TBB' (i.e.'Ca.Phytoplasma australasiaticum') in the NT DITT phytoplasma database and their 16S rRNA genes shared the highest nucleotide sequence similarity and coverage with the 'Ca.Phytoplasma australasiaticum subsp.australasiaticum' representative strain PR08 (99.92 % nucleotide sequence identity, 100 % coverage) in blastn analyses (Table 1).Additionally, these sequences could only be differentiated from 'Ca.Phytoplasma australasiaticum subsp.australasiaticum' and 'Ca.Phytoplasma australasiaticum subsp.ipomoeae' during in silico RFLP analysis by the HaeIII restriction enzyme (Fig. S1).Therefore, the identification and subsequent characterization of this 'Ca.Phytoplasma australasiaticum' subspecies was likely missed in previous analyses as the HaeIII restriction enzyme was infrequently used during in vitro RFLP analyses [13,14,16,59].These results illustrate the low resolution of the RFLP of the 16S rRNA gene sequence to delimit separate phytoplasma taxa compared to species and subspecies characterization that is possible using the genome ANI, which has also been emphasized in previous studies [36]. Together, these results suggest that the previously unrecorded taxa, 'Ca.Phytoplasma australasiaticum' taxon 1 to taxon 5, are endemic to Australia.This is supported by 'Ca.Phytoplasma australasiaticum' taxon 1 being present, but misclassified, in a historic sample collected in 2004 (Table S1) but also because these phytoplasma taxa have not been detected in any other country to date based on the 16S rRNA sequences and the limited number of 16SrII phytoplasma genomes that are publicly available.Using the genome-sequence data obtained for these strains in this study, further analyses are required and could be done to confirm whether these five new taxa are truly distinct subspecies of 'Ca.Phytoplasma australasiaticum' and not artefacts generated during metagenomic sequencing and analyses [36,37]. A putatively new species within the 16SrII phytoplasma group The phytoplasma strains obtained from samples BAWM-167 and BAWM-339 shared 100 % ANI and >80 % AF with each other and approximately 94 % ANI and <80 % AF with any phytoplasma genomes used in this study, including the closely related subspecies of 'Ca.Phytoplasma australasiaticum' (Fig. 1b, c).These results indicate that the phytoplasma strains from samples BAWM-167 and BAWM-339 may represent a novel 'Ca.Phytoplasma' species within the 16SrII group.The 16S rRNA gene analyses support the divergence of these two strains compared to other previously described 16SrII phytoplasmas.The 16S rRNA genes of BAWM-167 and BAWM-339 had 99.92 % sequence similarity to the reference sequence of 'Ca.Phytoplasma australasiaticum subsp.australasiaticum' (100 % coverage, 2180-2211 total score) (Table S1).Further phylogenomic analyses are required to confirm whether these three taxa are truly distinct subspecies and not due to artefacts generated during metagenomic sequencing and analyses.However, the identification of this putative novel species from two distinct hosts (Medicago sativa and Ipomoea sp.) across a large geographic separation may provide positive support for the existence of this species. This species could be endemic to Australia because these two detections are the first time this species has been detected globally based on the 16S rRNA sequences and they were detected in two geographically distinct areas in Australia (Darwin, NT and Dandaragan, WA) in different hosts and years. Other phytoplasma taxa identified in this study, which have previously been recorded in Australia Strains of the 16SrII species, 'Ca.Phytoplasma fabacearum' , formed the third largest cluster (n=16) in the ANI analyses, while the cluster containing the group 16SrXXXVIII species 'Ca.Phytoplasma stylosanthis' was the fifth largest (n=6) (Fig. 1).The single WaLL phytoplasma (group 16SrII) strain from BAWM-227, showed between 90 and 95 % ANI with four 16SrII phytoplasma The previous identification of BAWM-227 as a WaLL phytoplasma based on RFLP in the nt DITT phytoplasma database was used as no fulllength 16S rRNA gene exists for this taxon [13].However, the 16S rRNA gene of this phytoplasma shared 99.65 % nucleotide identity with that of 'Ca.Phytoplasma asiaticum'.†Mixed infection was identified for these samples by Sanger sequencing cloned 16S rRNA PCR amplicons.species previously determined to be closely related to each other [37], namely 'Ca.Phytoplasma citri' , 'Ca.Phytoplasma asiaticum' , 'Ca.Phytoplasma gossypii' , and 'Ca.Phytoplasma crotalariae' (Fig. 1).The two 'Ca.Phytoplasma bonamiae' strains (group 16SrII) formed their own cluster with 100 % ANI between them in the pairwise analysis and shared the next highest ANI with strains of 'Ca.Phytoplasma fabacearum' (ANI of <94 %, Fig. 1).The two 'Ca.Phytoplasma melaleucae' strains (group 16SrXV) formed their own cluster with 100 % ANI between them in the pairwise analysis (Fig. 1).The ViLL phytoplasma strains (16 Sr group unassigned) clustered with each other, but with a pairwise ANI of ca.97 % to each other. The blastn analyses of the 16S rRNA genes of all these taxa supported the ANI results (Fig. S1A), and the WaLL phytoplasma was characterized based on RFLP in previous analyses (NT DITT record; Table 2) [13].Further, the ANI results of WaLL and ViLL phytoplasmas suggest that these taxa could be described as two novel 'Ca.Phytoplasma' species (ANI <95 % with any other genome available for described phytoplasma species).Future work is required to determine whether the WaLL and ViLL phytoplasmas meet the updated requirements for the description of novel 'Ca.Phytoplasma' species [31] and whether the two ViLL strains represent two individual subspecies.Additionally, the competent insect vector species of WaLL and ViLL taxon remain to be determined. The identification of mixed phytoplasma infections Close analysis of the ANI heat map revealed evidence of mixed phytoplasma infections (Fig. 1), where several samples showed a high ANI with two representative genomes.Samples BAWM-044, BAWM-186, and BAWM-316 had mixed infections comprising of 'Ca.Phytoplasma fabacearum' and 'Ca.Phytoplasma australasiaticum subsp.ipomoeae' .blastn analyses of the 16S rRNA genes obtained for these samples identified the presence of 'Ca.Phytoplasma australasiaticum subsp.ipomoeae' but failed to identify 'Ca.Phytoplasma fabacearum' (Table S1). When revisiting the number of tRNA gene sequences annotated from these 12 samples identified to contain mixed phytoplasma infections, more than 35 tRNA gene sequences were identified from all of these samples apart from sample BAWM-307 from which 32 tRNA genes were obtained (Table S1).Further, only one sample for which no mixed phytoplasma infection was identified using the ANI approach encoded more than 35 tRNA genes (BAWM-311, n=40 tRNA genes).These results highlight that the metagenomic sequencing, assembly, and tRNA annotation approach used in this study can sufficiently resolve the distinct tRNA genes encoded by each phytoplasma species and, thus, support the utility of using the tRNA count as an indicator of mixed phytoplasma infections in a sample in addition to genome completeness criteria that has been proposed previously [60,61]. The results of species identification based on the 16S rRNA gene and whole-genome comparisons emphasize several important implications of these approaches to phytoplasma identification.Firstly, the 16S rRNA gene sequences obtained by either method often only represented one of the phytoplasma taxa involved in the mixed infection.It is likely that this arose due to differences in the titres of the multiple phytoplasma taxa in the sample, with only the one gene sequence being obtained by direct Sanger sequencing or more genomic sequence data obtained from the phytoplasma present at the higher titre.Alternatively, the multiple 16S rRNA gene sequences of closely related phytoplasma taxa obtained in a sample may have been missed during the process of obtaining the consensus sequence from both Sanger sequencing and metagenomic HTS data. Phytoplasma identifications made for samples that could not be used in ANI analyses For the samples for which insufficient phytoplasma data was obtained for ANI analyses, the 16S rRNA sequences shared high sequence similarity with 'Ca.Phytoplasma australasiaticum subsp.australasiaticum' (n=9 sequences), 'Ca.Phytoplasma australasiaticum subsp.ipomoeae' (n=7 sequences), 'Ca.Phytoplasma australiense' (n=1 sequence, BAWM-189), and the ViLL phytoplasma (n=1, BAWM-337) (Table S1).These results highlight the pitfall of applying metagenomic-based approaches to identify the species diversity of phytoplasmas obtained from a diversity of host species.Specifically, it may be difficult to get sufficient data from hosts that harbour low titre infections for comparison with other taxa from metagenome sequencing.Whereas PCR of the 16S rRNA gene enriches for these regions of interest that provide taxonomically informative information [62], albeit at a low taxonomic resolution [37].This highlights the need to sequence additional genomic regions in diversity studies, and/or to have pre-sequencing enrichment tools for phytoplasma cells or DNA to improve genome sequence retrieval and assembly for genomic analysis relevant for applications such as taxonomy [38,[63][64][65] Host and geographic ranges of the phytoplasma taxa identified in this study Identification of unknown hosts Eight weed samples were either unknown or not confidently identified to family, genus or species based on visual identification.The combination of two barcodes, matK and rbcL, extracted from metagenomic data as well as occurrences of the identified species in the geographic region of collection, as recorded in the Australasian Virtual Herbarium, were used to indicate the host plant species (Table 3). Two plants visually identified as members of the Lamiaceae were identified as Asteraceae members, including Chromolaena odorata (BAWM-316) and an undetermined species likely in the Chromolaena or Praxelis genera (BAWM-321).(Tables 3 and S1).One weed species (sample BAWM-319) was also visually identified as a member of the Lamiaceae but was identified as a member of the Malvaceae instead based on the DNA barcodes (Sida sp., BAWM-319) (Table S1).The host species of BAWM-319 is likely Sida rhombifolia as this species is present in QLD where the sample was collected, whereas Sida fallax reports were made from WA in the Australasian Virtual Herbarium records.Species in the Sida genus have been reported as a host in Australia previously [12].The results for these three samples are consistent with other phytoplasma detections made in Australia, where members in the Lamiaceae are not known to host phytoplasmas in Australia but where several Asteraceae and Malvaceae species have been recorded [27]. Samples BAWM-037, BAWM-057, BAWM-079, and BAWM-302 were all recorded as 'unknown weed' species based on morphological observations (Tables 3 and S1).Samples BAWM-057 and BAWM-302 were subsequently identified to the genus-level based on BLASTn of the matK and rbcL gene sequences (both Solanum spp., Solanaceae).The host of BAWM-079 was also identified as a Solanum sp.(Solanaceae) by the host DNA barcode analyses (Table 3, either Solanum nigrum or Solanum rostratum).However, it is likely that the host species of BAWM-079 is Solanum nigrum as this plant species has a wide geographic distribution in Australia based on the Australasian Virtual Herbarium, including in the NT where BAWM-079 was sampled, and due to the high blastn results for this gene with Solanum nigrum (228 total score, 5E-60 e-value, 99.2 % identity; GenBank accession number: M588530).Solanum rostratum is not present in the NT according to the Australasian Virtual Herbarium.However, the rbcL gene sequence had higher blastn scores with Solanum rostratum and it was, therefore, recorded in Table 3.Based on blastn of the matK and rbcL genes obtained from the metagenomic data, the host of BAWM-037 was identified as a Sida sp.(Malvaceae). The host species of BAWM-037 could not be determined due to the inconsistencies between the top hit species listed for the two DNA barcodes (Table 3). Two insect hosts were identified as Orosius sp.based on their external morphology (samples BAWM-342B and BAWM-343A). Using sequence analyses of their coI gene, both samples were identified as Orosius argentatus.Based on several studies, Orosius argentatus is a known phytoplasma vector in Australia and is detected across a broad geographic range in Australia [9,29,30,66]. These sequence-supported identifications of plant or insect hosts at the species-, genus-, or family-level when they were not known based on visual inspection highlight the added benefit of a metagenomic-based approach to investigating phytoplasma diversity and their host associations.However, the host species listed using this approach are considered preliminary indications of the host taxa sampled, especially when (i) the nucleotide identities of the DNA barcodes were not identical to those of voucher specimens on the NCBI, despite the nucleotide identities being above 90 % in all cases in this study (Table 3), (ii) recording species-level identifications, and (iii) considering that some barcodes may be missing for the species under investigation but for which they are available for a closely related species [67,68].This is due to the limitations of the available and well-validated plant DNA barcodes in the public databases. Of the 24 phytoplasma-positive weed samples, 16 were unidentified to the genus-or species-level based on morphology or DNA barcode analysis (Tables 3 and 4).4). Orosius species, including Orosius argentatus, have been identified as confirmed or putative vectors of diseases thought to be associated with 'Ca.Phytoplasma australasiaticum subsp.australasiaticum' and 'Ca.Phytoplasma stylosanthis' , such as tomato big bud, tobacco little leaf, and legume little leaf diseases [15,27,29].The detection of 'Ca.Phytoplasma australasiaticum subsp.australasiaticum' and 'Ca.Phytoplasma stylosanthis' in Orosius argentatus in this study (Tables 1 and S1) is therefore consistent with the results of previous collection efforts in Australia [15] and it may be that this leafhopper species is a vector of several phytoplasma taxa.However, the detection of phytoplasma strains from the total nucleic extracts from insect whole bodies or Continued subsampled body sections, as done in this study, does not provide definitive evidence of vector competence and further transmission trials are required.A comprehensive list of insect species that serve as competent vectors of phytoplasma diseases of vegetable crops in Australia remains to be determined. The observation that 'Ca.Phytoplasma australasiaticum subsp.australasiaticum' and 'Ca.Phytoplasma australasiaticum subsp.ipomoeae' were the most frequently detected, the most widespread geographically, and detected from the most host families and species in this study is not surprising (Figs 1-3).These two species have historically been detected from a broad range of hosts in Australia, including many crop species [27].'Ca.Phytoplasma australasiaticum subsp.australasiaticum' and 'Ca.Phytoplasma australasiaticum subsp.ipomoeae' are major threats to crop production in Australia as they are commonly associated with crop species in many vegetable growing areas around the country, and disease incidences can be high [72].Additionally, the symptoms they are associated with directly affect crop yield (e.g.phyllody).Further research is required to determine whether these two subspecies have distinct host ranges, symptomologies, or vector species for a better understanding of their biology and how to mitigate outbreaks of diseases associated with them. This taxon may present a moderate threat to crop production in Australia, with the potential to affect Solanaceae hosts in particular.This is due to this taxon being detected from several crop hosts displaying symptoms that directly affect crop yield but also due to the large geographic range of the detections made in this study (from the NT, QLD, and VIC).Additionally, the detection of 'Ca.Phytoplasma australasiaticum' taxon 1 was likely missed in previous RFLP-based analyses done to assess taxon diversity in Australia and may, therefore, have a broader host and geographic range than what is reported in this study.Further research is required to investigate the prevalence and vector(s) of this taxon. The ANI results indicated that these taxa were distinct and are likely novel subspecies of 'Ca.Phytoplasma australasiaticum' (Fig. 1), representing the first time these taxa have been identified in Australia and likely globally.These novel detections are potentially due to these identifications being missed previously due to the high sequence similarity of their 16S rRNA gene to other described 16SrII phytoplasma taxa (>99 % nucleotide sequence similarity; Table S1).Additionally, these novel detections may not have been made in the past as four of these novel phytoplasma taxa were detected from either weed species or from phytoplasma crop host species that are first records of phytoplasma hosts for Australia (Table 4). The potential threat of these phytoplasma taxa remains to be determined as only a few detections in plant hosts were made in this study and their detections were likely missed in previous low resolution (RFLP-based) analyses (Table S1).Additionally, no vector species have been identified for these taxa.These results highlight the importance of sampling weed species in and around cropping areas, as well as collecting diverse species of symptomatic hosts in an area.These taxa need to be assessed further to determine whether they are truly distinct subspecies, which can be done using further comparative and phylogenomic assessments in future [36,37]. These results highlight that Fabaceae crops across a broad geographic range in Australia are at a high risk of losses due to infection by 'Ca.Phytoplasma fabacearum' , although some Asteraceae, Cucurbitaceae, and Solanaceae hosts might also be at risk.This is also supported by reports of high incidence of phytoplasma diseases in Australia likely attributed, in part, to 'Ca.Phytoplasma fabacearum' [72]. Historic samples: 'Ca.Phytoplasma bonamiae' and the Waltheria Little leaf phytoplasma (16SrII) Strains of 'Ca.Phytoplasma bonamiae' (n=2) identified from Bonnamia pannosa and the WaLL phytoplasma (n=1) identified from a Waltheria sp. were only identified from the historically collected samples analysed in this study from QLD and the NT, respectively (Figs 2 and 3b, c; Table S1).No new host species or geographic range expansions are, therefore, reported for these taxa.'Ca.Phytoplasma bonamiae' was associated with little leaf symptoms in both samples, while the WaLL phytoplasma was associated with both little leaf and witches'-broom symptoms.No insect vector species have been identified for these phytoplasma taxa. This study provides a full-length sequence of the 16S rRNA gene as well as genomic data for the WaLL phytoplasma for the first time.This sequence data identified this phytoplasma as a member of the 16SrII group (Figs 1 and S1; Table S1), which confirms previous reports based on nucleotide analysis of regions within the 16S rRNA gene [13].Additionally, 16 Sr rRNA and ANI sequence analysis showed that the 'Ca.Phytoplasma bonamiae' and WaLL phytoplasmas were close relatives of the 'Ca. Phytoplasma fabacearum' strains (Figs 1 and 2) [13].However, since both these phytoplasma taxa are infrequently detected in crop plants, and since 'Ca.Phytoplasma bonamiae' has only been detected from the Australian native plant Bonamia pannosa (based on 14 samples in the NT DITT database and [15,27]), they are both unlikely to pose a major threat to crop production in Australia. It is likely that the WaLL phytoplasma strain can be described as a novel species as it shares less than 96 % ANI with other phytoplasma species (Fig. 1).Eight WaLL strains are in the NT DITT database and future work can investigate whether these WaLL phytoplasma strains further fulfil the updated guidelines to be described as a novel 'Ca.Phytoplasma' species should more sequence data be made available for them [31]. 'Ca.Phytoplasma planchoniae' (16SrII) 'Ca.Phytoplasma planchoniae' was detected from a Planchonia careya host sampled in QLD (Figs 2 and 3b) that displayed little leaves and witches'-broom symptoms (Fig. 3c).'Ca.Phytoplasma planchoniae' has previously been detected in Australia and has only been associated with native plant Planchonia careya in far north QLD [73].Due to its narrow host range in a non-crop species and its restricted geographic range, 'Ca.Phytoplasma planchoniae' is unlikely to pose a major threat to crop production in Australia. Potentially new 16SrII species Strains of the potentially new 16SrII species were detected from an Ipomoea sp.(Convolvulaceae, sample BAWM-339) and from a Medicago sativa sample (Fabaceae, sample BAWM-167) from the NT and WA, respectively (Figs. 2 and 3b; Table S1).Both hosts showed symptoms of little leaf and witches'-broom, but the Medicago sativa sample BAWM-167 also showed symptoms of yellowing (Fig. 3c and Table S1).While further investigations are required to determine whether these two strains belong to a novel 'Candidatus Phytoplasma' species within the 16SrII group, support for their delimitation include the observation that (i) they produced ANI and AF values below the within-species threshold (<95 % and <80 %) with already described 'Candidatus Phytoplasma' species, and (ii) more than one strain of this potential species was identified from distinct hosts from different areas in Australia.It remains to be determined what threat to crop production in Australia this taxon presents as only these two strains were identified in this study with the competent or putative vector species unknown. Since 'Ca.Phytoplasma stylosanthis' had previously only been reported in NT, QLD, and NSW [13,18,74], the Solanum tuberosum sample represents both a host and geographic range expansion for this phytoplasma (Table S1) as described previously (sample VPRI 43683 [24]).Since this phytoplasma has been identified from a broad range of crop species across a large geographic area, 'Ca.Phytoplasma stylosanthis' has the potential to be associated with reductions of economically important crops such as Carica papaya [15,27]. Group 16SrXXV Phytoplasma samples -'Ca.Phytoplasma melaleucae' 'Ca.Phytoplasma melaleucae' was detected in QLD (n=1; BAWM-155) and WA (n=1; BAWM-354) (Fig. 2) and were only detected as single infections from Melaleuca spp.(Myrtaceae) in Australian regions above the Tropic of Capricorn (Tables 1 and S1).This is the first report of 'Ca.Phytoplasma melaleucae' for WA and the furthest west occurrence of this phytoplasma.Prior to the present study, this phytoplasma was only reported from far north QLD and the Western Province of Papua New Guinea in Melaleuca spp., with one case reported for Synsepalum dulcificum (Sapotaceae) [16,37].The two samples analysed in this study displayed little leaf and witches'-broom symptoms (Fig. 3c and Table S1), which is consistent with previous detections [16].At present, this phytoplasma is unlikely to pose a major threat, if any, to vegetable crop production in Australia due to its restricted host range to non-crop hosts. Phytoplasma 16Sr group unassigned -vigna little leaf phytoplasma The ViLL phytoplasma was detected in the NT and WA (n=1 per state/territory; Fig. 2).Sample BAWM-245 represents a host and geographic range expansion for the ViLL phytoplasma, being detected for the first time in WA and in a Catharanthus roseus sample (Apocynaceae; Table S1; [27]).A second host expansion for this phytoplasma and new phytoplasma host for Australia was Momordica charantia (Cucurbitaceae), detected in the NT (sample BAWM-336) where there was a high incidence of disease (70-80 % of crop affected, in-field observation by S. Bond).Prior to this study, this phytoplasma taxon was only reported in Australia from within or near Katherine and Darwin in the NT [13,15,74].Both samples showed little leaf symptoms, however symptoms were recorded (Fig. 3a).The most symptom types (n=12) were recorded for the plants classified within the Solanaceae, followed by those in the Asteraceae and Fabaceae plant families (n=5 each).No symptoms were provided for 12 samples in this study (Fig. 3a; Table S1). No additional symptoms were observed for the mixed infection compared to samples where only a single phytoplasma taxon was observed (Fig. 3c; Table S1).While these associations need to be investigated more thoroughly as only a few host species overlapped between the single and mixed infection samples in this study, these observations have been reported previously [74]. The association of all plant symptoms with phytoplasma infection is, however, difficult to disentangle as phytoplasmas remain to be cultured and, as such, Koch's postulates cannot be fulfilled.Abiotic factors, herbicide treatments, insect damage, the presence of other microbes, viruses, or a combination of these factors could also contribute to the symptoms that are presented by the plant hosts [79,80]. CONCLUSIONS In this study, phytoplasma-infected crop and non-crop hosts from historic collections and contemporary collections (2015 to 2022) from vegetable growing regions around Australia were metagenomically sequenced to identify the crop-infecting phytoplasma taxa and potential alternative hosts.A total of 15 distinct phytoplasma taxa were identified from the metagenomic data obtained for these hosts (Figs 1 and 2).'Ca.Phytoplasma australasiaticum' subspecies and 'Ca.Phytoplasma stylosanthis' were two of the most frequently detected taxa identified, and from the broadest range of hosts and locations sampled across Australia (Fig. 2).Additionally, six previously undescribed phytoplasma taxa were identified from the samples analysed in this study, namely: 'Ca.Phytoplasma australasiaticum' taxon 1 to 5, and a potentially new 16SrII phytoplasma species.A few phytoplasma taxa were infrequently detected in this study, with some only associated with diseases in non-crop plants and, therefore, likely pose a low threat to crop production in Australia (e.g.'Ca.Phytoplasma melaleucae').Five different phytoplasma mixed infections were also identified (Figs 1 and 2).An updated list of phytoplasma 16 Sr groups, species, subspecies, and unclassified taxa present in Australian vegetable growing areas, as well as the prevalence and combinations of mixed phytoplasma infections was therefore provided by this study.A list of symptoms per host (Fig. 3a) and per phytoplasma taxon (Fig. 3c) are also provided and, with the previous literature, will aid in-field detections of phytoplasma associated disease in crop and non-crop plant hosts (Table S1). PCR of the 16S rRNA gene using universal nested phytoplasma primers combined with direct Sanger sequencing was sufficient as a triage tool to screen and provide a preliminary identification of the phytoplasma taxon present in every sample analysed in this study.However, it lacked the taxonomic resolution afforded by the ANI analysis of draft metagenomic-assembled phytoplasma genomes (Fig. S1), emphasizing results from other studies [36].Additionally, the PCR-based approach often failed to accurately identify mixed infections (Table S1), which has been reported previously for Sanger sequencing of the PCR amplicon obtained directly from a sample [74].The metagenomic-based approach employed in this study based on whole-genome ANI, however, was able to resolve strains to the subspecies-level and could identify the presence of a mixed phytoplasma infection in a single sample.An additional benefit of using the metagenomic approach during phytoplasma collection was that it allowed for host taxa to be identified through the use of genetic barcodes present in the metagenomic dataset when they were unable to be resolved to the family-, genus-, or species-level based on morphology (Table 3).Together, these results provided more informative data with a more precise assessment of the prevalence and host range of phytoplasmas in vegetable growing regions in Australia compared to previous studies, which could only use RFLP or sequence analysis of the 16S rRNA gene [12,13,15,74].The results presented in this study highlight the benefits of combining metadata (host, location, date, etc.) and metagenomic sequencing for phytoplasma diversity assessments and to understand their epidemiology. Sufficient phytoplasma genomic data was obtained for 178 (12 mixed infections) of the 195 symptomatic samples for genome based sequence analyses and to be submitted to public sequence repositories (excluding mixed infection samples).The dataset presented here is the largest contribution of phytoplasma genome sequences from a single study to date, increasing the number of publicly available sequences from 47 [81] to a total of 213 (when excluding samples with mixed phytoplasma infections).The incomplete and draft phytoplasma genomes sequenced in this study have significantly increased the taxon sampling of subclade II, which is one of three subclades described in [82] (Fig. S1).The work presented here was possible due to the ever-decreasing cost of HTS and the increased volume of sequence data generated.The phytoplasma genome data obtained in this study can be used in future research to improve phytoplasma taxonomy and diagnostics, and will assist in genomic epidemiology analyses.The reliable genome sequence assemblies will also serve as a resource from which genes involved in symptomology and host/vector interactions can be investigated when combined with the appropriate metadata, as well as comparative and functional analyses. Together, this genome resource will contribute significantly to the knowledge of phytoplasma biology, ecology, and can be used to inform management practises to help mitigate or prevent losses associated with major phytoplasma outbreaks in Australia. Fig. 1 . Fig. 1.Whole-genome comparisons for phytoplasma genome sequence data obtained for 178 samples.(a) ANI heatmap, generated by pyani version 0.2.10 using the ANIm algorithm, for all strains sequenced in this study alongside representative and publicly available genomes.Some clusters are highlighted using brackets.(b) ANI percentages and (c) alignment fractions (AF) in each pairwise comparison of samples that did not cluster with representative genomes in Fig. 1a.The genomes of representative strains and publicly available are shaded in grey.See colour gradient representing the percent identities in the heatmaps of (a) and (b) or the AF per genome in (c). Fig. 2 . Fig. 2. Map of Australia showing the number of phytoplasma-positive samples collected per state or territory, with pie charts illustrating the proportions of ANI identified phytoplasma taxa identified per state or territory (see key below for descriptions of colour-coding).The scale on the right indicates the number of samples collected for each state or territory, with the number in brackets indicating the total number of ANI-identified samples per location within the map area.Abbreviations: ACT, Australian Capital Territory; NSW, New South Wales; NT, Northern Territory; QLD, Queensland; TAS, Tasmania; VIC, Victoria; WA, Western Australia. Fig. 3 . Fig. 3. Bar graphs indicating the relative abundances of (a) symptom types recorded for each plant host family analysed in this study (n=176 samples); (b) the ANI-identified phytoplasma taxa per plant or insect host family analysed in this study (n=178 samples); and (c) the symptom types recorded for each ANI-identified phytoplasma taxon analysed in this study (n=176 samples).Numbers in the bar graphs indicate the total number of samples.Colour legends are shown above each graph. Table 1 . Summary of all host species, genera, and families investigated in this study and associated metadata organized according to their sampling location in Australia (state/territory and closest town/city).Metadata recorded included the sample names, sampling years, and the phytoplasma taxa identified based on blastn or ANI analyses performed in this study.Original sample names, if provided, are recorded alongside the corresponding 'BAWM' name in TableS1 Table 2 . Summary of blastn top hits of the 16S rRNA gene sequences obtained for the samples investigated in this study, including the putative phytoplasma taxon identified, the number of samples with this result, and the range of percent of nucleotide identities shared with the top hit Table 3 . Summary of samples for which initial host identifications were unresolved to the family-, genus-or species-level based on visual inspections and for which additional gene regions obtained from metagenomic data were used to determine the host identity.The genes used for plant host identifications included the maturase K (matK) and ribulose-bisphosphate carboxylase (rbcL) genes, and the cytochrome C oxidase subunit 1 (coI) gene was used for insect identification.The e-value, percent identity, and bitscores of the top blastn hit(s) for the sample are provided to illustrate the support for the gene-based host identification.na=Not applicable Table 4 . List of phytoplasma hosts investigated in this study, characterized based on whether the host was recorded previously as a phytoplasma host/ putative vector or not, and whether these hosts were classified as crop (C), insects (I), native plant (NP), ornamental (O), or weed (W) in this study.A list of the phytoplasma taxa that were identified are listed for the respective host	2024-03-07T06:16:46.797Z	2024-03-01T00:00:00.000	{ "year": 2024, "sha1": "5e48de058a3ba2a084b9b7d5edfdd0d16423a20a", "oa_license": "CCBY", "oa_url": null, "oa_status": null, "pdf_src": "PubMedCentral", "pdf_hash": "2caf4ae2f5946b04a1c200065bf30ce26ac1bf3d", "s2fieldsofstudy": [ "Environmental Science", "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
247011296	pes2o/s2orc	v3-fos-license	Can Social Robots Effectively Elicit Curiosity in STEM Topics from K-1 Students During Oral Assessments? This paper presents the results of a pilot study that introduces social robots into kindergarten and first-grade classroom tasks. This study aims to understand 1) how effective social robots are in administering educational activities and assessments, and 2) if these interactions with social robots can serve as a gateway into learning about robotics and STEM for young children. We administered a commonly-used assessment (GFTA3) of speech production using a social robot and compared the quality of recorded responses to those obtained with a human assessor. In a comparison done between 40 children, we found no significant differences in the student responses between the two conditions over the three metrics used: word repetition accuracy, number of times additional help was needed, and similarity of prosody to the assessor. We also found that interactions with the robot were successfully able to stimulate curiosity in robotics, and therefore STEM, from a large number of the 164 student participants. I. INTRODUCTION Social robots have proven to be an effective aid to early childhood language acquisition [1]. Their welcoming designs and expressive movements make them engaging for children to speak and play with [2]. Several studies have shown that children experience psychological and educational benefits by spending time with social robots ( [3], [4]) including students' improved academic and social outcomes and student engagement. The recent success of these devices with children may also imply that their introduction to educational settings will make children more curious about them and, more broadly, robotics and STEM. Previous studies such as [5] and [6] show the effectiveness of robots as teaching tools in STEM lessons through collected user surveys, and [7] introduces the design for a conversational robotic interface that engages primary school children in STEM discussions. However, further work is needed to explore how robots can be effectively integrated into existing lessons and tasks for young children rather than creating new robot-centered lessons. In addition, more free-response feedback is needed in order to determine what aspects of human-robot interaction pique young children's curiosity in robotics and could be used to motivate them to later pursue STEM education. Inspired by previous work using voice-enabled social robots to encourage children's oral language development through storytelling ( [8], [9]), we apply social robots to the critical task of conducting assessments of language and early literacy-related skills in children. Such assessments are particularly important for kindergarten and first-grade children, as they are beginning to learn to read, and any oral language delays should be addressed before they negatively impact early literacy development. Children's early language abilities are also indicators of their later aptitude in other subjects including writing, math, and science. Public schools typically do not have sufficient support from speech-language pathologists, reading specialists, or literacy coaches to conduct detailed, dialect-appropriate assessments of every child's oral language abilities on a routine basis, depriving some children in need of extra language support or intervention from such resources. A social robot enabled with automatic speech recognition (ASR) technology may be used to conduct some of these intensive assessments with children without the need for a trained language specialist. The presence of the social robot during these assessments can also become a seamless way to introduce children to robotics and STEM. However, before these systems are implemented, it must be determined whether or not the presence of a robot has a profound effect on the children's speech. A person's tone, prosody, and word choice can vary with whomever they are speaking. Therefore, further investigation is needed to verify whether or not children's speech productions differ significantly in the presence of a voice-enabled social robot, compared to a human oral assessment administrator, in a way that might hinder obtaining accurate assessments of speech production. In this work, we investigate 1) how effective social robots are in the direct assessment of children's speech production (phonological processing skills), and 2) if social robots can be used to promote young children's curiosity in robotics and STEM during necessary oral language assessments. In order to investigate these questions, we conduct common oral language assessments with kindergarten and first grade children in two scenarios: In one case, the assessments are administered by a human assessor as typically conducted. In the other case, assessments are administered by, JIBO 1 , a social robot designed for interacting with children (pictured in Figure 1), while the human assessor watches and intervenes only as necessary. We selected children of these grade levels because they are at the critical age range for language development and literacy acquisition ( [10], [11]). To identify any significant differences in speech production between children in the two cases that may affect assessment scores, we investigate the students' changes in behavior, prosody, and accuracy of word repetition during the assessments across the two scenarios for both grade levels. We additionally documented the children's responses to interacting with JIBO that indicate curiosity in robotics or may be used to further interest in STEM-related topics. A. Participants A sample of 164 kindergarten and first-grade students, consisting of 53 kindergarteners and 111 first graders from a Southern California elementary school, were recruited for the study. Selection criteria included parental consent and completion of all tasks in the study. All of the students were English speaking with some reporting additional language exposure (most often Spanish) at home. B. Recording Interactions with each student were recorded with a Logitech C920 Webcam microphone with a sampling rate of 48kHz. Each student was approximately two feet from the test administrator (either JIBO or the human assessor), and the microphone was placed equidistantly between them. Recordings took place in an empty office at the school site during school hours to simulate a realistic environment in which JIBO could be used. C. Experiment Data collection followed the protocol described in ( [12]- [14]). The children's speech samples were collected using Goldman-Fristoe Test of Articulation-3 (GFTA-3) [15] sounds in sentences protocol. Each student was first read a story appropriate for their grade level. The kindergarten students were read a story containing 20 simple sentences and the first-grade students were read a story containing 15 sentences mixing both simple and compound structures. As a fictitious example (the actual assessment material is under copyright restrictions), a sentence such as, "The cat was fat, fuzzy, and orange," may have appeared in the story for first-graders. After the story was read once in its entirety, the story was repeated sentence by sentence, and the student was asked to repeat each sentence back to the test administerer. Ten kindergarten students and ten first-grade students were administered the test by a human assessor, and the others were administered the test by JIBO. Both gave the same prompts to the children. In giving each prompt, JIBO played a corresponding audio file containing a recording of an adult woman whose voice was pitch shifted up to match the pitch of a child and slowed down. For sessions in which the child was given the prompts by JIBO, a human assessor was present and intervened if the child was unable to complete the task with only the social robot's instructions. In sessions in which the child was given the prompt by a human assessor, they gave the prompt to the child a second time if they were unable to repeat the sentence after hearing it once. Any questions or comments that the child made referring to JIBO before, during, or after the assessment were documented and categorized by their different characteristics. D. Assessment Quality Evaluation The 10 kindergarteners and 10 first-graders who completed the assessment with only the human assessor were compared to a group of 10 kindergarteners and 10 first-graders who completed the same assessment administered by the social robot. The robot-assessed children in the comparison group were selected from the total participant pool on the basis that they were assessed in the same time period and overseen by same human assessor as the students in the human-assessed children. The following metrics were calculated for comparison between the two conditions: 1) Accuracy: each sentence that the students were asked to repeat was scored out of the total number of words in the sentence. One point was given for each word that the child repeated correctly in the correct order. 2) Need for additional prompting: we noted each time that a child required additional prompting to repeat a sentence and noted the reason for the additional prompting. If the child was silent for at least 3 seconds after given the prompt, the reason was noted as "reticence." If the child forgot the prompt or did not understand what to say, the reason is given as "needs repetition." If the child said something unrelated to the prompt then the reason was listed as "distracted." 3) Pitch correlation: An additional question is whether students attempt to follow the administrator's pitch and rhythm during the assessment. While not directly related to speech production assessment quality, differences in how the children incorporate the human and the robot's prosody into their repetition of the prompt may indicate how engaged the children were with the assessor. For each prompt, the pitch contour of both the administerer (JIBO or the human assessor) uttering the prompt and the child's voice while repeating it were extracted using Praat's [16] pitch tracking algorithm with parameters set to best measure a child's pitch. The Praat pitch contours were manually corrected if any errors occurred in the pitch tracking algorithm. Then the Pearson correlation coefficient between the child's pitch contour and administrator's (either the robot or human assessor) pitch contour was calculated and taken as a measure of the perceptual similarity of the two as in [17]. Sentence word accuracy scores were scored manually by the researchers who are experienced in evaluating children's speech. Reasons for the child needing additional prompting were also noted manually following the above criteria. E. Student Responses The students were allowed to interact with the robot and speak freely before, during, and after the assessment. Any questions or comments made by students about JIBO that may be used to grow interest in robotics or STEM were characterized by the following five categories: 1) Mechanical: Referring to how robots work or asking about fundamentals of robotics or other STEM concepts 2) Functional: Referring to why JIBO looks and acts as it does or asking about engineering design choices 3) Relational: Relating JIBO to something that the student has seen previously 4) Personifying: Assigning human characteristics to JIBO 5) Hypothetical: Exploratory questions about JIBO or related STEM concepts Questions/comments could by characterized by more than one category. Some examples of children's comments are given in Table I. Category Examples Mechanical "You program him to do that?", "He didn't get rusty" Functional "Why is JIBO have one eye?", "Why he dance like this?" Relational "I have a robot at my home", "I know robot [that] is more strong" Personifying "Now, JIBO is looking at me", "You miss your mommy, JIBO or not?" Hypothetical "But what if you flip JIBO backwards?", "Can JIBO do a handstand?" III. RESULTS In this section, we present the results of the experiment. Table II shows the percent word accuracy of the repetition task for children in each group. The maximum score, minimum score, and standard deviation for the group are also given. Table III gives the number of times students in the two groups required additional prompting to be able to complete the exercise and the reason for the need for additional prompting. C. Prosody Average correlation between the assessor and students' pitch contours when uttering the same prompt are shown in Table IV for each group. The percentage of repetitions spoken by students that are significantly correlated in pitch contour to the prompt given by the assessor are also given in this D. Scientific Curiosity In total, 43 kindergarteners and 101 first-graders were assessed using JIBO. Table V shows the percentage of these children in each grade level who made a comment or question in the given category. In total, approximately 45% of the students made comments or questions about JIBO pertaining to at least one of the categories. A. Evaluation of Speech Assessment Quality As shown in Table II, in the repetition task, the students' performance was not significantly affected by the presence of the social robot, JIBO. The table shows that when repeating words to the examiner, neither the kindergarten students (p > 0.1) nor the first-grade students (p > 0.05) are significantly worse at repeating words with JIBO than they are with a human. Although the mean sentence score was about 2% higher when the test was administered by a human assessor for both the kindergarten and first-grade students, this number is not statistically significant (p > 0.05). There is however a significantly larger standard deviation in the sessions administered by JIBO (note the differences between the minimum and maximum sentence scores). Table III shows that kindergarteners are more reticent when the test is administered by JIBO than by a human assessor. However, this reticence typically only occurs at the beginning of the session and disappears as the students are exposed more to the social robot. A possible explanation for this finding is that at this early stage of the activity, students may have been adapting to the activity. This trend in reticence does not extend to the first-grade students. Most of the first-grade students did need sentences to be repeated more often by the human facilitator present when the test was administered by JIBO. It is worth noting that the first-grade student who needed intervention most often when using JIBO predominately spoke a language other than English at home. The prompts that needed repetition were most often those containing longer sentences with compound structures. Table IV implies that the pitch contours of the students' repetition of the prompt more closely match those of the prompt given by the human assessor. This means that while repeating sentences said by the assessor, the students' changes in pitch and rhythm matched the human assessor more closely than the robot. This may be due to the robot having less natural sounding prosody, making it more difficult to match. However, these correlations are not significant, and we would need a larger sample size to definitively state that the use of a robot affects whether or not the students follow the speech prosody of the human assessor more. It is important to note that the correlations for first graders are lower than those of the kindergarteners. This is likely due to the fact that the prompts read to the first graders were longer and contained more complex sentence structures, requiring the students to turn their attention more towards remembering the words than matching the prosody given. B. Motivating Scientific Curiosity Almost half of the students made inquiries as to how robots work, why they are designed in the way that they are, and other ideas that are conducive to furthering children's interest in STEM. The kindergarteners most commonly asked questions related to how JIBO is able to move, speak, and show images. That is, their questions typically asked for explanations to observed phenomenon. This implies that these interactions can be used to motivate kindergarten-level lessons on more visual concepts in robotics and STEM like motors, cameras, and programming fundamentals. The first-graders also asked a large number of questions of this nature, albeit fewer on average than the kindergarteners, while also asking personifying and hypothetical questions that delve past just what they can directly observe in the robot, such as, "Where does JIBO sleep?" and "Can you make him do [a different type of movement]?" These questions may be used to motivate further interest into engineering concepts such as biological inspiration for design and user-centered design. We also note that only the first graders related JIBO to other robots that they had seen previously. This may be because the younger students have had less exposure to other instances of robots. We observed that students became excited to participate in the assessment when they learned that they would be working with a robot. This experience was, for most, if not all, of the students, their first interaction with a social robot. From the students' excitement and frequency of questions asked, we believe that JIBO can be used to prime young students for robotic and STEM-centered lessons, possibly motivating them to further pursue them in secondary and post-secondary education. We believe that many of the students who did not make comments or questions about JIBO were simply shy, as our work shows that even the students working with only the human assessor still displayed a large amount of reticence throughout the assessment. Further interactions with the robot to build familiarity may reduce that reticence over time and lead to additional questions from students in the future. V. LIMITATIONS Comparing our study with others on robot-directed speech ( [18], [19]) using a power analysis, we noted that that our results on children's human-directed and robot-directed speech are underpowered. However, our findings demonstrate strong tendencies and suggest directions for future research. In addition, variability might have been reduced if the same children had completed the assessment under both assessment conditions (with or without JIBO), giving different, equally difficult prompts in each condition. However, in the current study, we had only a short duration in each session to work with the children, making it infeasible to give one child both assessment conditions within a session. We instead balanced the pools for each condition as effectively as possible within the constraints. VI. CONCLUSIONS Our pilot results show no detriment in the students' speech production performances as a result of working with the social robot. Students were sometimes more hesitant to talk to the robot at first but became more willing over the course of the session. The time spent with the social robot also appears to have elicited questions from the students that can be used to grow their interest in this STEM domain. Future work includes measuring students' curiosity and speech differences in a longitudinal study to investigate how these factors change with time. VII. ACKNOWLEDGEMENTS This work is supported in part by the NSF.	2022-02-22T06:47:21.935Z	2022-02-19T00:00:00.000	{ "year": 2022, "sha1": "8a4cf653bc4cae3f60af940a038ac08e9c1de6e8", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "8a4cf653bc4cae3f60af940a038ac08e9c1de6e8", "s2fieldsofstudy": [ "Engineering", "Education", "Computer Science" ], "extfieldsofstudy": [ "Computer Science", "Engineering" ] }
221573206	pes2o/s2orc	v3-fos-license	Temporomandibular disorder, body pain and systemic diseases: assessing their associations in adolescents Abstract Painful temporomandibular disorders (TMD) in children and adolescents may impact negatively the individual´s life. The presence of comorbidities associated with TMD tends to increase the persistence of pain and to facilitate its chronification. Objective To investigate the presence of other painful conditions and systemic diseases and their association with painful TMD. Methodology In this cross-sectional study, 690 adolescents aged between 12-14 years old were evaluated through questionnaires and clinical examinations. Results Painful TMD was found in 16.2% of the sample, with a significant association with bronchitis (OR= 2.5; p=0.003) and asthma (OR=3.1; p=0.013), reported by the parents/legal guardians of the participants. Adolescents with regional and widespread pain were 2.7 (95% CI: 1.65-4.55) and 3.6 (95% CI: 1.29-10.14) more likely to also present painful TMD. Painful TMD was associated with a higher number of body pain sites in the last 12 months (4.26 vs. 2.90; p<0.001), as well as a higher number of systemic diseases (1.48 vs. 1.18; p=0.048), when compared to adolescents without painful TMD. Conclusion The findings of this study point out the importance of considering the presence of comorbid conditions in the diagnosis and management of painful TMD in adolescents. A multidisciplinary approach would contribute to better control of painful TMD and decrease its chronification risk. Introduction Adolescence is a complex phase of life when individuals experience intense emotional, cognitive, social, physical, and hormonal transformations. It is also the period in which the behavioral traits are defined. 1 Occurrence of pain at this stage can be associated with disability and a predictive factor for pain in adulthood. 2 Anxiety, depression, reduced quality of life, and school absenteeism are conditions frequently present in adolescents with somatic pain. 3 These experiences represent significant risk factors for the development of chronic pain in the later stages of life. Therefore, a better understanding of pain features in this age group is fundamental to define more efficient treatment protocols as well as to prevent future health problems. 2,4 Temporomandibular disorders (TMD) are defined as a group of musculoskeletal and neuromuscular conditions involving temporomandibular joints (TMJs), masticatory muscles, and associated structures. Pain is the most frequent symptom and can affect both muscles or joints. 5 Although TMD has been more investigated in adults than in adolescents, recent studies have pointed to an increase in its prevalence among the latter, with rates ranging from 7.3% and 30.4%. 6 The complexity of TMD is enhanced by its association with other painful conditions as headaches, neck pain, and back pain. 7 Moreover, the emotional and psychosocial aspects can influence the painful experience related to TMD. 8,9 Previous studies investigated the prevalence of headache, abdominal pain, low back pain and TMD in children and adolescents, [10][11][12] in adolescents". So far, two previous papers were published, presenting part of the collected data. 16,17 Each of them, although based on the same sample, had specific aims, and consequently, different instruments and evaluations were applied. For this reason, it is recommended to calculate the sample size separately for each of the primary outcomes by applying a Bonferroni correction to adjust the significance level. Thus, for the sample size calculation, we considered the prevalence obtained in the pilot study, a statistical power to refute H0 of 80%, and alpha of 0.007, adopted after Bonferroni correction, according to our goals. 18 Accordingly, the sample should consist of approximately 120 individuals with painful TMD and 120 individuals free of painful TMD. Also, we added 25% due to possible losses during the data collection, thus totaling a minimum of 300 participants in the study. The investigation of the association between systemic diseases and painful TMD was an exploratory analysis. Assessment instruments The family economic classification was determined according to the updated Brazilian Classification days. The body areas were grouped in regional pain (neck/shoulder) and widespread pain (upper back, elbows, wrists/hands, lower back, hips/thighs, knees and ankles/feet). 23 Parents or legal guardians of adolescents responded a questionnaire about the current medical condition, considering a previous medical diagnosis. This information was collected and evaluated in a list, which was presented to legal guardians, containing the most frequent diseases reported among Brazilian youth. 24 These diseases were grouped as follows : 1) Respiratory diseases (bronchitis, rhinitis, asthma, sinusitis, pulmonary emphysema, and pneumonia); 2) Gastrointestinal diseases (gastro-esophageal reflux; irritable bowel syndrome, constipation, gastritis/ ulcer, and colitis); 3) Endocrine diseases (diabetes, hypothyroidism, hyperthyroidism, and hypoglycemia); of having painful TMD compared with individuals free of these respiratory conditions ( Table 2). Adolescents reporting regional pain were 2.7 (95% CI: 1.65-4.55) more likely to present painful TMD when compared to those who did not report regional pain. When the areas of regional pain were (Table 3). Moreover, adolescents presenting painful TMD reported a significantly higher number of body pain areas in the last 12 months compared with adolescents with no painful TMD (p<0.001). They also reported a significantly higher number of systemic diseases (p=0.048) ( Table 4). Discussion TMD has been associated with several painful and non-painful conditions in adults. 14,15,23 Additionally, it has been demonstrated that the presence of comorbidity tends to increase the severity and the disability related to the involved conditions, as well as make worse their prognostic. 7 Because there is a lack of information regarding the associations between TMD and other conditions among adolescents, our study is justified. Our main findings are: (1) The prevalence rate of painful TMD (16.2%) found in the present sample is in accordance with previous studies that have shown rates varying from 7.3% to 30.4% in adolescents 6 . The instrument used for the TMD assessment and the age of the sample can contribute to the variation observed across previous studies. We did not find significant differences regarding the presence of TMD pain between genders, or the pubertal developmental stages. Previous studies have shown that painful TMD is associated with the presence of pain in other parts of the body. 26,27 Our findings agree with the literature confirming the association between painful TMD and regional and widespread pain. The association between painful TMD and the persistent regional pain can be related to the neural inputs from the superior There is a lack in the literature regarding the association between systemic diseases and TMD in adolescents. In adults, previous studies showed an association between the presence of TMD and some specific systemic diseases, such as irritable bowel syndrome, 32 heart diseases, 13 and autoimmune diseases, such as rheumatic disease. 15 In our sample, asthma and bronchitis were significantly associated with painful TMD. We highlight that although we found a significant association between respiratory conditions and the presence of painful TMD, the number of participants affected by both was small. We found 18 adolescents with painful TMD reporting previous diagnosis (informed by their parents or legal guardians) of bronchitis, and eight reporting asthma. Although our results should be interpreted with caution and confirmed in future studies, they are aligned with a previously conducted study, demonstrating BRAIDO GV, CAMPI LB, JORDANI PC, FERNANDES G J Appl Oral Sci. 2020;28:e20190608 8/10 a positive correlation between TMD and respiratory diseases in children. 33 The presence of some degree of inflammation in the respiratory system and in the organism overall seems to be associated with bronchitis and asthma. 34 Our study has some limitations. Firstly, since it was a cross-sectional study, causation cannot be inferred. We assessed the presence of systemic diseases through the report of the adolescent's parents or legal guardian, so no definitive diagnosis was provided, neither the severity of the diseases. Therefore, the gold standard diagnostic tool, which would be an accurate medical examination and assessment conducted within the last 12 months, 37 was not possible. However, the sample had an adequate size, as described in the results. Another limitation was regarding the type of school. Most of the sample were enrolled in public schools, due to a low participation response of students from private schools, who did not return the consent form signed by their parents/legal guardians. Finally, it is important to point out that the presence of pain in adolescents is often associated with anxiety and depression symptoms. 8,9 Given this, future studies should address the interaction of these psychological factors with painful TMD, and the associated variables found in this study. The study's strengths are related to instruments and sampling. Translated and validated instruments with easy interpretation for adolescents were applied to evaluate the conditions studied. For TMD classification we applied the RDC/TMD, and the researcher who	2020-09-10T10:08:09.775Z	2020-09-07T00:00:00.000	{ "year": 2020, "sha1": "6f89637a8d1d0ce69dba3ce8079228d0d08108b8", "oa_license": "CCBY", "oa_url": "http://www.scielo.br/j/jaos/a/9BcxhTf9p3sHJxTwxPzdspQ/?format=pdf&lang=en", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "a32170e1d3f59293e144a3cfd62836a2ccecb16b", "s2fieldsofstudy": [ "Medicine", "Psychology" ], "extfieldsofstudy": [ "Medicine" ] }
53982961	pes2o/s2orc	v3-fos-license	ASSESSMENT OF THE AGRONOMIC PERFORMANCE OF TWO VARIETIES OF SOYBEAN AS INFLUENCE BY TIME OF INTRODUCTION OF MAIZE AND CROPPING PATTERN Field experiments were conducted to assess the effect of time of introduction of maize and cropping pattern on two varieties of soybean in a soybean/maize intercrop in 2007 and 2008 seasons at the University of Agriculture Makurdi. Three periods of introduction of maize were evaluated: (i) planting at the same time with soybean;(ii) introduction of maize two weeks after planting soybean and (iii) introduction of maize four weeks after planting soybean. There were five cropping patterns viz: (i) soybean variety Samsoy-2 planted sole (ii) soybean variety TGX 1448-2E planted sole (iii) maize variety DMR-ESR-Y planted sole (iv) Samsoy -2 intercrop with maize and (v) TGX1448-2E intercrop with maize. The experiment was a split plot laid on complete randomized block design replicated three times. Results obtained revealed that there was no significant effect of time of introduction of maize on any soybean parameter observed. Yield and yield components of soybean indicated significant effect of cropping pattern (P<0.01) with number of pods per plant and grain yield. Grain yield and number of pods per plant decreased in intercrop with soybean variety Samsoy-2 having the highest number of pods per plant and grain yield. The result on maize indicated that maize planted sole was significantly (P<0.01) taller than intercropped maize, there was significant (P<0.01) effect of time of introduction of maize and cropping pattern on maize plant height, ear length, ear diameter, survival and barrenness percentage and grain yield. The Land Equivalent ratio (LER) values indicated that higher yield advantages were obtained from maize introduced two weeks after planting soybean and maize intercropped with soybean variety Samsoy-2 (1.35) than with soybean variety TGX1448-2E (1.12). INTRODUCTION Southern Guinea Savannah of Nigeria is a major producer of cereals and legumes. The ecological characteristics of the zone have been described by Agboola, (1979) and Agabi, et al.,(1995) as good for production of cereals, legume and some tree crops. Soybean (Glycine max. L.) is chiefly grown in Benue State as cash crop by small scale farm holders who grow it sole or in mixture simultaneously with cereals. The crop has become an integral component of the traditional cropping systems of this agro ecological zone due to its beneficial effect on sustainability and nutritious food ( Henriet et al., 1997 ). Maize ( Zea mays L.) is grown almost in all parts of the country except where rainfall becomes a limiting factor; it is increasingly being accepted as a major source of food and cash income among its predominantly small holder producers in Nigeria (NARP, 1994). Although the bulk of soybean production takes place in crop mixture, most of the cultivars were developed and tested under the sole cropping condition. Studies on evaluation of genotype for the adaptability to intercropping have indicated differential responses of LTD, 2015 genotype to cropping systems (Biabani et al., 2008;Mudita et al., 2008). Intercropping is the traditional farming system of the tropics, in this system, the component crops in the mixture are not always sown at the same date, and the crops may be sown at the same time or different times depending on the farmer's preferences (Francis, 1978;Remison, 1982;Ofori and Stern, 1987). The performance of a crop in mixture might vary depending on when it is sown relative to the companion crop (Baker, 1978;Zhang and Li, 2003). IITA (1989) recommended proper adjustment of time of sowing and spacing and plant types so as to minimize competition for light to enhance productivity, that the growth habit and plant architecture must be considered when deciding to defer planting of any component crop. The productivity of legumes in legumes/cereals intercropping is reported to be low (Ennin et al., 2002). A number of measures have been recommended for achieving increase in legume productivity in intercropping among which are -Identification of shade tolerant varieties (Francis, 1986), identifying the best suitable time of sowing the component crop in the intercropping (Singh and Ajeigbe, 2002) and choice of suitable companion crop in the intercropping (Olufajo, 1995). Recent reviews emphasize the need for research on how to minimize competition through tolerant varieties for intercropping (Biabani et al.,.2008;Mudita et al.,2008) and time of sowing component crop in the intercropping (Singh and Ajeigbe, 2002;Olufajo and Singh, 2002). The objective of this study was to investigate the performance of two varieties of soybean as influenced by time of introduction of maize and intercropping pattern. MATERIALS AND METHODS Field experiments were conducted during the 2007 and 2008 cropping seasons to investigate the influence of date of planting and time of introduction of maize in a soybean/maize intercropping system at the Teaching and Research farm of the University of Agriculture, Makurdi (7.41 o N ; 8.28 o E ) which falls within the Southern Guinea Savannah agro-ecological zone of Nigeria. The experimental design was a 3x5 split plot laid in a randomized complete block design replicated three times. Three times of introduction of maize (planting soybean and maize simultaneously, maize introduced two weeks after planting soybean and maize introduced four weeks after planting soybean) and five cropping patterns (soybean variety-Samsoy-2 sole, soybean variety-TGX 1448-2E sole, maize variety DMR-ESR-Y planted sole, maize variety intercrop with Samsoy-2, maize intercrop with TGX1448-2E).Maize variety DMR-ESR-Y and soybean variety TGX1448-2E and Samsoy-2 were obtained from National Cereal Research Institute sub-station Gboko. Each experimental unit (plot size) measured 5mx3m with four ridges of 5m long spaced 0.75m, site clearing and ridge making was done manually, spacing for sole maize was recommended spacing of 0.75m x 0.5m at 2 plants per stand giving a population of approximately 53,333 plants per hectare. Intercrop maize was spaced 0.75m x 0.5m at one plant per stand using the semi additive mixture as stated by Fisher (1977) giving a population of approximately 26,666 plants per hectare. Spacing for soybean was 0.75m x 0.05m at one plant per stand giving a population of approximately 266,666 plants per hectare. Maize seeds were sown (4 seeds/hill) on the side of the ridges and thinned to two plants per stand for sole and one plant per stand for intercrop, soybean was sown on top of the ridges with the seeds drilled, which were thinned to one plant per stand both for sole and intercrop. Fertilizer was applied based on recommended fertilizer rates for Benue State as followsmaize sole-90kgN/ha, 45kgP 2 Kalu, 1993). Data collected on maize were plant height, number of days to 50% tasseling and silking, numbers of leaves per plant, leaf area, numbers of ears per plant, ear diameter, ear length, percent survival, percent barrenness, weight of 100 seeds and net yield. Data collected on soybean were -plant height, number of days to 50% flowering, number of primary branches, number of pods per plant, number of seeds per pod, leaf area, weight of 100 seeds and net yield. Both crops were harvested when they were fully matured and dried. All the data collected were subjected to analysis of variance using GENSTAT (2003) statistical package. Land Equivalent Ratio was computed as stated by Mead and Willey (1980). Competitive Ratio was computed as stated by Willey and Rao (1980) Soybean- The result on soybean showed that time of introduction of maize (T) had no significant effect on any vegetative flowering traits of soybean (Table 1). Leaf area and days to 50% flowering of soybean were significantly affected by cropping pattern, while no significant effect was observed on soybean plant height and number of branches. Generally, there was an increase in leaf area and days to 50% flowering for soybean intercropped. There was also variation in the growth components among the varieties of soybean Table 2 showed results on yield and yield component of soybean. Time of introduction of maize had no significant effects on all the parameters quantified. However, there were significant effects of cropping pattern on number of pods/plant and grain yield. Both number of pods/plant and grain yield decreased in intercrop, there was significant differences between sole planted and intercropped soybean and also among the varieties. Results of mean effects of different times of introduction and cropping patterns on the vegetative and flowering data of maize are as presented in table 3. There were significant differences in the vegetative and flowering traits of maize for the different time of introduction of maize. Results also showed significant differences between four weeks of introduction (T 3 ) and the other times of introduction (T 1 and T 2 ) for plant height (which decreased with delayed introduction of maize). For days to 50% tasseling, simultaneous planting of maize and soybean (T 1 ) and introduction of maize at two weeks (T 2 ) were not significantly different but showed significant lower values compared with introduction of maize at four weeks (T 3 ). For days to 50% silking, two weeks of introduction of maize showed higher values with no significant differences among the various times of introduction. Number of leaves per plant showed the same trend as days to 50% silking. There were no significant effects of cropping pattern on 50% tasseling and silking as well as number of leaves per plant. International Letters of Natural Sciences Vol. 31 Cropping pattern significantly affected plant height, there were highly significant differences between maize planted sole (CP 3 ) and intercropped maize (CP 4 and CP 5 ) for plant height.( Maize planted sole grew taller than intercropped maize). Furthermore, there were highly interaction effects of time of introduction and cropping pattern (T x CP) on plant height and number of leaves per plant (Table 4). Plant height both for sole and intercrop maize decreased with delayed time of introduction. However, for number of leaves per plant, maize planted sole did not significantly differed from intercropped maize ILNS Volume 31 Results of yield and yield component of maize as influenced by time of introduction of maize and cropping pattern are as shown in table 5.There were significant differences in the ear diameter, ear length, percent barrenness, percent survival, 100 -seed weight and grain yield for the different time of introduction of maize. The result showed that these parameters (excerpt percent barrenness which increased) decreased with delayed maize introduction. Significant differences between four weeks of maize introduction (T 3 ) and other times of introduction of maize (T 1 and T 2 ) for percent survival, percent barrenness and grain yield were also observed, while significant differences existed between the various times of introduction of maize for ear length and 100 -seed weight. Cropping pattern significantly affected ear diameter, ear length, percent survival, percent barrenness and grain yield. Maize sole (CP 3 ) significantly differed from intercropped maize (CP 4 and CP 5 ) for ear length, ear diameter, percent survival and grain yield (all of which were lower in intercropped maize ) but percent barrenness increased in intercropping while no significant differences were observed for 100 -seed weight. Significant interaction effects occurred between time of introduction and cropping pattern (T x CP) for percent survival and percent barrenness. T 1 = maize planted at the same time with soybean, T 2 = maize introduce two weeks after planting T 3 = maize introduce four weeks after planting soybean. CP 3 = maize planted sole, CP 4 = maize intercropped with samsoy-2, CP 5 = maize intercropped with TGX 1448-2E Land Equivalent Ratios (LER) and Competitive Ratio (CR) The LERs and CRs of maize/soybean intercropping are as presented in Table 5. generally, LER values were greater than unity (>1.00), however, introduction of maize two weeks after planting produced highest LER values (1.30 and 1.28) Highest CR value (0.66) of maize/soybean intercropping was obtained when maize was introduced two weeks after planting soybean. Generally, CR values decreased with delayed maize introduction, In the current study, sole maize tasseled and silked at the same period with intercropped maize does not agree with the report of Ugen and Wien (1986) that sole maize tasseled and matured earlier than intercropped maize. The significant reduction in ear length and ear diameter of maize with intercropping and delayed introduction could be attributed to inter-specific competition and variation in the weather conditions especially in terms of solar radiation, humidity and temperature as the season progressed since these weather factors have a profound influence in the growth and ILNS Volume 31 development of maize. Similar findings have been reported by Elmore and Jacobs (1984) and Enyi (1973). The much reduction in maize grain yield in intercropped could be as a result of the semi additive population used, which translates to half maize population in cereal/legume intercropping in order to reduce excessive shading (Fisher, 1977). Soybean plants grown in mixture were taller than those grown sole. These could be reported significant yield differences among cultivars of the same maturity classes and growth habit and observed that some were better adopted to shading/ intercropping environment than others this they believed was as a result of higher concentration of chlorophyll in the leaves after experienced shading/intercropping. In the current study, highest LER and CR values were obtained from two weeks delayed maize introduction perhaps because delayed maize introduction helped the soybean plant to withstand competition while at the same time not significantly affecting the maize plant. CONCLUSION LER characterizes the performance of an intercrop by giving the relative land area under sole crop required to produce yields achieved in intercropping. The LER values were all above unity The LER values indicated that higher yield advantages were obtained from maize introduced after two weeks of planting soybean. However, soybean variety samsoy-2 with highest number of pods per plant and seed yield was the most desirable variety for intercropping with maize, it can be concluded that in Makurdi, a location within the Guinea savanna agro-ecological zone of Nigeria, for higher yield, maize should be introduced two weeks after planting soybean using the soybean variety samsoy-2. It is however suggested that further investigation be conducted across different locations in the Guinea savanna agro ecological zone of Nigeria. International Letters of Natural Sciences Vol. 31	2019-04-24T13:13:46.145Z	2015-01-04T00:00:00.000	{ "year": 2015, "sha1": "39b07ccc5ddc8e56998874d23d2783e3d1f2fc76", "oa_license": "CCBY", "oa_url": "https://www.scipress.com/ILNS.31.36.pdf", "oa_status": "HYBRID", "pdf_src": "ScienceParsePlus", "pdf_hash": "0800a089b2560bb987e74a77576a81ce6a94fb51", "s2fieldsofstudy": [ "Agricultural and Food Sciences" ], "extfieldsofstudy": [ "Biology" ] }
6037238	pes2o/s2orc	v3-fos-license	Topping-off technique prevents aggravation of degeneration of adjacent segment fusion revealed by retrospective and finite element biomechanical analysis Aim The aim of this study was to evaluate the effect of the Topping-off technique in preventing the aggravation of degeneration caused by adjacent segment fusion. Methods Clinical parameters of patients who underwent L5-S1 posterior lumbar interbody fusion + interspinous process at L4-L5 (PLIF + ISP) with the Wallis system (Topping-off group) were compared retrospectively with those of patients who underwent solely PLIF. Pre- and post-operative x-ray measurements, visual analogue scale (VAS) scores, and Japanese Orthopaedic Association (JOA) scores were assessed in all subjects. Normal L1-S1 lumbosacral finite element models were established in accordance with the two types of surgery in our study, respectively. Virtual loading was added to assess the motility, disc pressure, and facet joint stress of L4-L5. Results There were 22 and 23 valid cases included in the Topping-off and PLIF groups. No degeneration was observed in either group. Both VAS and JOA scores improved significantly post-operatively (P < 0.01). The intervertebral angle and lumbar lordosis of L4-L5 were both significantly increased (t = −2.89 and −2.68, P < 0.05 in the Topping-off group and t = −2.25 and −2.15, P < 0.05 in the PLIF group). In the Topping-off group, x-ray in dynamic position showed no significant difference in the angulation or distance of the anterior movement of the L4-L5 segment. The angle of hyper-extension and distance of the posterior movement of L4 were significantly decreased. In the PLIF group, both hyper-flexion and hyper-extension and posterior movement were increased significantly. In finite element analysis, displacement of the L4 vertebral body, pressure of the annulus fibrosus and nucleus pulposus, and stress of the bilateral facet joint were less in the Topping-off group under loads of anterior flexion and posterior extension. Facet joint stress on the left side of the L4-L5 segment was also less in the Topping-off group under left flexion loads. Conclusion Short-term efficacy and safety between Topping-off and PLIF were similar, whilst the Topping-off technique could restrict the hyper-extension movement of adjacent segments, prevent back and forth movement of proximal vertebrae, and decrease loads of intervertebral disc and facet joints. Introduction Posterior lumbar intervertebral fusion (PLIF) is the conventional treatment for lumbar degeneration entailing however a series of complications, amid which acceleration of adjacent segment degeneration (ASD) remains the major problem. Some scholars suggest that prophylaxis application of an interspinous process (ISP) or spacer in fusion segments could decelerate this process [1,2]. The Topping-off technique is a newly developed surgical technique, which combines rigid fusion with an interspinous process device in the adjacent segment to prevent ASD. Nonetheless, there are few reports on Topping-off surgery and its rationality, and the indications remain unclear. Finite element analysis was first introduced in spine biomechanics in the 1970s with a wide application ever since, as the methodology is minimally impacted by other parameters [3]. In the current study, we have compared the outcomes, safety, motility, and interspinous space between patients who underwent Topping-off and conventional PLIF techniques. Also, we used a finite element model to assess the motility, disc pressure, and facet joint stress under virtual loads in silico. We therefore aim to reveal the role of the Topping-off technique in preventing the aggravation of ASD. General information Patients between Jan 2008 and Mar 2010 diagnosed with lumbar disc herniation or lumbar spinal stenosis in our centre were included. Informed consents were obtained from all patients, and the study was approved by the local ethical committee (Peking University People's Hospital Institutional Review Board). The Topping-off technique was suggested to patients with mild L4-L5 disc degeneration (e.g. MODIC1) not meeting the criteria for internal fixation fusion as a supplement to conventional PLIF. Choice of surgery was at patient's discretion. The Topping-off group consisted of 22 patients who underwent L5-S1 PLIF + L4-L5 ISP (Wallis), among which there were 14 men and 8 women with an average age of 44.5 years (21 to 64 years). The PLIF group consisted of 23 patients who underwent L5-S1 PLIF solely, among which there were 11 men and 12 women with an average age of 40 years (12 to 77 years). Pre-operative standing lumbar anteroposterior and lateral, flexion, and hyperextension x-ray examinations were performed. All patients were evaluated pre-operatively for visual analogue scale (VAS) and Japanese Orthopaedic Association (JOA) scores. Device used and surgical technique The Wallis system was provided by Abbott Spine (Bordeaux, France). The device has notches that fit the physiological shape of the spinous processes. It also has a flat band for best spread of constraints in contact with the bone to ensure the rigidity of the unstable segment, whilst preserving its mobility. The design and materials minimize the need for bony resection and avoid any concentration of constraint on the bone. Pedicle screws (XIA) were provided by Stryker Spine (Cestas, France). The cross-connectors were provided by Stryker Spine (XIA) and Moss Miami (DePuy Spine, Sarl, Switzerland). Fusion cages were provided by Stryker PEEK, Synthes, and Plivios. Choice of internal fixation was according to intraoperative calibrations. Posterior L5-S1 partial laminectomy, decompression discectomy, implantation of the PEEK fusion device, and bilateral L5-S1 pedicle fixation were performed in all patients. Autologous bone grafting was applied in the surgeries, in which too much bony structures were resected and stability could be compromised. Patients in the Topping-off group were additionally implanted with Wallis at L4-L5 as the interspinous spacer fixation after L5-S1 decompression fusion. All surgeries were performed by the same surgeon. Patients were bedded for 1 week post-operatively and were discharged with lumbar supports. Follow-ups Patients were followed at post-operative months 1, 3, 6, and 12 and annually thereafter with standing lumbar anteroposterior and lateral, flexion, and hyper-extension xray examinations. VAS and JOA evaluations were also performed at the last follow-up. Radiological measurement All measurements were conducted by a blinded orthopaedist using the GE Healthcare Centricity RIS CE V2.0 software. The anterior height (aDH) and posterior height (pDH) of the L4-L5 disc space, sagittal diameter of the L4 vertebral body (W), L4-L5 intervertebral angle (α), and lordosis of the upper endplates of L1-S1 (β) were measured on lateral x-ray images. Flexion (θ + ) and hyperextension angles (θ − ) of the L4-L5 intervertebral space and the anterior and posterior slipping distance of the posterior edge of L4 (AO and RO, respectively) were measured on flexion and hyper-extension x-ray images. All length parameters were expressed in relative values to avoid magnification bias. Criteria for adjacent segment degeneration Degeneration was defined radiologically as follows: 1) percentage of anterior and posterior lumbar slipping of >25% [4], 2) dynamic angulation of the interspinous space of >10° [5], 3) complete loss of the interspinous space [6], and 4) recurrence and aggravation of back pain post-operatively with no positive radiological findings and with exclusion of surgery-related complications [2]. Statistical analysis Data were presented as mean ± standard deviation and were processed with SPSS ver. 21.0 software. The twotailed Student's t-test was used for comparisons between the two groups. A P value of <0.05 was accepted as statistically significant. Finite element model Serial thin-section CT images of L1-S1 of a healthy male volunteer were analysed using Mimics, Geomagic, and Ansys. A three-dimensional finite element model of L1-S1, which consisted of 459,340 units and 661,938 nodes, was constructed. After setting the boundary condition, unit type, and dividing finite element mesh, the S1 vertebra was constrained and loading was added. Five hundred newtons of vertical compressive loading was added and distributed evenly to each node of the endplate surface of the L1 vertebra. Ten newton metres of flexion, extension, left flexion, and right rotation torque loading was added, and average rigidity at all directions was calculated. Results were validated by reviewing previous reports. After exclusion of the supraspinous ligament, interspinous ligament, yellow ligament, part of the spinous process, part of the lamina, part of the nucleus in the gap, and part of the annulus of L5/S1, pedicle screws, connecting rods, and a single cage were implanted to obtain the PLIF model ( Figure 1 (1~2)). The Topping-off model was constructed based on the PLIF origin by further excluding the supraspinous ligament, interspinous ligament, and part of the spinous process of L4-L5 and by implanting the Wallis ISP system (Figure 1 (3~4)). Relative L4 displacement, pressure on the annulus and nucleus, and stress on the bilateral facet joints of both models were measured four times in flexion, extension, left flexion, and right rotation status, and the mean value was calculated. In order to compare our model to the others, we did a thorough literature search by means of PubMed/MEDLINE under the search term "finite element" in combination with all keywords relevant to our subject of interest. Only full text published in English or Chinese were selected. Studies that covered all our parameters were eventually selected. Results Twenty-two patients in the Topping-off group were followed for a mean duration of 24.8 months (12-40 months). Twenty-three patients in the PLIF group were followed with a mean duration of 23.7 months (12-38 months). There was no statistical significance in followup time (P > 0.05). Pre-operative VAS score of 7.6 ± 1.2 dropped to 1.9 ± 0.9 post-operatively in the Topping-off group, and the Pre-operative VAS score of 6.8 ± 1.2 dropped to 1.9 ± 0.9 post-operatively in the PLIF group, and the difference also reached statistical significance (t = 26.9, P < 0.01). Better improvement in VAS score was noted in the Topping-off group over the PLIF group (P < 0.01). Pre-operative JOA score of 12.7 ± 2.9 improved to 22.5 ± 3.5 post-operatively in the Topping-off group, and the difference reached statistical significance (t = −14.6, P < 0.01). Pre-operative JOA score of 13.7 ± 1.9 improved to 23.2 ± 2.8 post-operatively in the PLIF group, and the difference also reached statistical significance (t = −16.9, P < 0.01). The improvement of JOA score did not differ significantly between the two groups. No patient presented symptoms indicative of ASD or implant rupture or loosening. Standing lateral x-ray showed that both anterior (aDH/W) and posterior (pDH/W) disc height of L4-L5 increased in both groups, but the difference did not reach statistical significance (P > 0.05). Compared with pre-operative status, both groups showed a significant change in α post-operatively (t = −2.89, P < 0.05 in the Topping-off group; t = −2.25, P < 0.05 in the PLIF group). Compared with pre-operative status, both groups also showed a significant increase in β post-operatively (t = −2.68, P < 0.05 in the Topping-off group; t = −2.15, P < 0.05 in the PLIF group). The parameters measured on lateral x-ray are summarized in Table 1. Hyper-flexion and hyper-extension x-ray showed that at last follow-up, the θ + angle of L4-L5 and anterior movement of the posterior edge of L4 were not significantly changed compared with pre-operative status in the Topping-off group (P > 0.05). The θ − angle and posterior movement of the posterior edge of L4 were significantly decreased (t = 5.57 and 4.75, respectively, P < 0.01). The parameters measured on hyper-flexion and hyper-extension x-ray are summarized in Table 2. Compared with pre-operative status, the PLIF group showed a significant increase in motility angle post-operatively both at hyper-flexion and hyper-extension positions (t = −5.77, P < 0.05 and t = −5.01, P < 0.01, respectively). The PLIF group also showed a significant distance increase in both anterior and posterior L4 movements post-operatively (t = 6.69 and t = −24, respectively, P < 0.01). The rigidity parameters we acquired in our finite element model in comparison with other reports are summarized in Table 3. Upon flexion-extension loading, there was less relative displacement of L4 in the Topping-off group than in the PLIF group (P < 0.05). However, upon flexor and rotation loading, relative L4 displacement was not significantly different between the two groups ( Table 4). Upon flexion-extension loading, there was less pressure on the annulus and nucleus and less stress on the bilateral facet joint of the L4-L5 section in the Topping-off group than in the PLIF group (P < 0.05), whilst the difference was no longer statistically significant upon flexor and rotation loading (Table 5). Upon left flexion, there was less stress on the left facet joint in the Topping-off group (P < 0.05), whilst the difference lost statistical significance upon other types of loading (Table 6). Discussion Degeneration of the interspinous disc is thought to be a physiological process, and spine fusion is believed to accelerate the process of ASD. The incidence of ASD was 5.2% to 18.5% based on symptoms and was up to 100% according to radiological evaluation [5]. Current risk factors for acceleration of degeneration included preoperative degenerative presentation, long segment fusion, low fusion, and sagittal imbalance. Upper adjacent segments were more prone to degenerate than lower ones, due to the compensatory increase of postfusion adjacent motility and increased loading onto the disc and facet joints [5,11]. ISP systems, like Wallis or Coflex, can constrain lumbar hyper-flexion and hyper-extension, decrease pressure on the disc and facet joints, and thereby at the mean time reserve part of motility in the corresponding segments [12,13]. The Topping-off technique is therefore introduced by implanting ISP in the adjacent segments for protection. Nonetheless, prophylaxis ISP implantation using the Topping-off technique remains disputable. In a prospective cohort study, Korovessis et al. suggest that the Topping-off technique can alter the natural disease course of ASD and thereby reduce the incidence of the disease [2]. We suggest that the Toppingoff technique has the following advantages [1]: 1) Topping-off can avoid fusion of segments with preexisting degeneration and minimize fusion length, 2) Topping-off can protect the adjacent segments in the case of long segmental fusion and can offer transition to the unaffected segments, and 3) implantation of ISP can lower the complexity for potential second surgery. Virginie et al. have combined a carcass Wallis surgery model with finite element biomechanics and have revealed that Wallis, as an ISP, can reduce the lumbar flexion-extension motility and disc pressure of adjacent segments, but may increase stress of the corresponding spinal processes [14]. There have been a number of retrospective and carcass studies investigating ISP devices, yet there has been no report integrating a retrospective study with finite element analysis. Finite element analysis works as a virtual biomechanical tool and warrants mechanical and clinical validation. In the current study, the retrospective analysis covered symptomology and radiology and evaluated the trend of post-operational alteration of the adjacent segments. The finite element model, on the other hand, demonstrated the cumulative change of the adjacent segmental motility and stress of adjacent segments. Both modalities showed similar trends of alteration over time. The finite element model also revealed the change of stress of the adjacent segment and in part interpreted the outcomes from the retrospective analysis in a biomechanical perspective. In the retrospective part, the short-term outcome and safety between Topping-off and PLIF were similar, as revealed by post-operative indices and radiologic measurements. No degeneration of the adjacent segment was seen in any patient. The flexion-extension x-ray image revealed that ISP implantation restricted the segment extension and kept the anterior displacement of the adjacent segment similar to pre-operative status. Whereas the PLIF group did not meet the radiologic criteria for ASD, the motility of the adjacent segment and L4 slipping distance were both increased. As the Topping-off technique could prevent relative displacement, it could potentially reduce the risk of degeneration of the implanted segments. In the finite element analysis, less relative displacement of L4 vertebrae was noticed in the Topping-off group under flexion-extension loading, indicating that Topping-off not only reserved part of the motility of the adjacent segment but also restricted the extension of the spine via intraspinous cushion and restricted the flexion via ligation, thus increasing the stability of the adjacent segment. There was also less pressure of the annulus and nucleus and less facet joint stress upon L3/L4 in the Topping-off group under flexion-extension loading. This indicates that Toppingoff can change the stress conduction of the spine during flexion and extension, and ISP and the spinal process share the loadings, in particular the loading onto the disc and facet. Of note, our study has provided a convincing advantage of implicating the Topping-off technique in addition to conventional PLIF, as the outcome in terms of VAS score in our Topping-off cohort is statistically better than that in PLIF. Although not adjusted or confirmed in a larger population, this implies a potential advantage of the novel technique over the conventional one in the selected patient group. These two techniques, rather than in the dispute of replacing one another, may cater for different subgroups of patients who were once lacking multiple modalities. Nonetheless, our study has limitations. First, the retrospective nature of this study brings about inevitable bias, which warrants trials in the future to validate the results. Second, the sample size in the current study is still limited with relatively short follow-up time. The long-term outcome of the Topping-off technique is still awaited. Third, our study solely aimed at a single segment with ISP implantation, and study on long segment fusion with ISP is warranted. Though the finite element model of the L1-S1 segment has been established in the current study, more integral models of the lumbar region are needed for the study of lumbar biomechanics. Like most finite element studies, we did not simulate the muscular tissue. This could lead to deviation of the results. Also, due to the inherent limitation of the Wallis system, the effect of Topping-off may differ from clinical outcomes. Our results therefore are limited to profile the outcome of Wallis as an ISP.	2016-05-04T20:20:58.661Z	2015-01-28T00:00:00.000	{ "year": 2015, "sha1": "40535c4ce5806849e81d7efe689a273e2b0439da", "oa_license": "CCBY", "oa_url": "https://josr-online.biomedcentral.com/track/pdf/10.1186/s13018-014-0142-z", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "40535c4ce5806849e81d7efe689a273e2b0439da", "s2fieldsofstudy": [ "Engineering", "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
56001266	pes2o/s2orc	v3-fos-license	The evaluation and enhancement of quality, environmental protection and seaport safety by using FAHP The evaluation and enhancement of business processes in any organization in an uncertain environment presents one of the main requirements of ISO 9000:2008 and has a key effect on competitive advantage and long-term sustainability. The aim of this paper can be defined as the identification and discussion of some of the most important business processes of seaports and the performances of business processes and their key performance indicators (KPIs). The complexity and importance of the treated problem call for analytic methods rather than intuitive decisions. The existing decision variables of the considered problem are described by linguistic expressions which are modelled by triangular fuzzy numbers (TFNs). In this paper, the modified fuzzy extended analytic hierarchy process (FAHP) is proposed. The assessment of the relative importance of each pair of performances and their key performance indicators are stated as a fuzzy group decision-making problem. By using the modified fuzzy extended analytic hierarchy process, the fuzzy rank of business processes of a seaport is obtained. The model is tested through an illustrative example with real-life data, where the obtained data suggest measures which should enhance business strategy and improve key performance indicators. The future improvement is based on benchmark and knowledge sharing. Introduction Changes in the business world -above all in the domain of politics, the economy and environment -demand the con-tinuous improvement of business processes that are provided by strategic managers with the goal of increasing business in a seaport.A quality management system (QMS) conforming to ISO 9001:2008 should be considered as an important additional step, in terms of quality, because ISO 9001 also takes into account economic, financial, design and development aspects and introduces a management review for measurement and analysis of a process with the aim of improving performances (Poli et al., 2012).However, this important issue forces every organization to start either with ISO 9000 or total quality management (TQM) as a business strategy (Sedani and Lakhe, 2011).There are various ways in which an enterprise can claim that its QMS meets the requirements of ISO 9001 (Ali, 2014). Lately, economics, geography and operations research including risk management (John et al., 2014) have dominated seaport research.The impact of multidisciplinarity and interdisciplinarity has significantly increased, and management studies have also substantially increased since the 2000s (Woo et al., 2011). Seaport operations may be described with many uncertainties, so lately there have been many papers that deal with risk management models (John et al., 2014) and metrics, proposed and numerically implemented to assess the overall performance of large systems, during natural disasters and their recovery and resilience (Shafieezadeh and Burden, 2014).This is due to the fact that much of the available data associated with port operations require a flexible but robust approach to the handling and updating of existing information with new data.As risk management activities are ori-Published by Copernicus Publications on behalf of the European Geosciences Union.D. Tadic et al.: The evaluation and enhancement of quality, environmental protection and seaport safety ented toward safety, port safety evaluation (Pak et al., 2015) is the first step in overall safety enhancement.After quality management certification, the determining of performances of business processes is based on predefined critical success factors (Oakland, 2004). Motivation for this research comes from the fact that there are no research papers that treat seaports in the context of the process approach -an assessment of business processes' performances and their improvement which may be obtained by the application of different methods.These methods are developed on a different mathematical and logical framework.According to results of good practice, it is known that it is almost impossible to enhance all business processes at the same time, keeping in mind overall complexity and definite resources (costs, time, human resources, etc.).Enhancement activities are based on the already defined priority of business processes.Business processes may be described by different performances.Different preferences of business processes may vary depending on the needs of the seaport's business strategy, international standards related to seaport services and change of the environment, etc.According to the results of seaport good practice, the key performance indicators (KPIs) of business processes may be defined.It may be assumed that the defined performances and their KPIs do not have the same relative importance.In compliance with this assumption, it can be said that priorities of business processes can be stated as a multi-criteria optimization problem. The wider objective of this research may be interpreted as an integration of the process approach, the management of business processes and multi-criteria optimization methods.The mentioned integration includes (a) the presentation of a seaport as a network of unrelated business processes so the overall success of the business processes may be assessed on the level of predefined criteria; (b) the assessment of business processes by fuzzy analytic hierarchy process (FAHP); (c) the definition of management initiatives which should lead to the improvement of business success.The order of taking management initiatives is based on the obtained rank of business processes. This paper continues to investigate beyond the existing seaport literature and intends to consider all the significant performances, which have not had much attention in the management of business processes.The main contribution of the proposed model is that by its application, the fuzzy rank of business processes and the degree of belief that one business process is better than another with respect to each business' performance can be obtained. As the environment changes rapidly or becomes uncertain, the values of some performances and their relative importance become difficult or impossible to quantify.All existing uncertainties can be adequately described by linguistic expressions which are modelled by applying the fuzzy sets theory (Klir and Folger, 1988;Zimmermann, 2001).The fuzzy sets theory resembles human reasoning in its use of approximate information and uncertainty to generate decisions.It may be suggested that the fuzzy approach to treating uncertainties in real-wold applications has numerous advantages when compared to other approaches such as applying the probability theory and the rough set theory.The appropriate technique for determining the rank of business processes with respect to all identified performances in a seaport is FAHP.It is assumed that FAHP is closer to human thinking that the relative importance of KPIs and performances of business processes are assigned according to a pairwise comparison matrix (Saaty, 2008).The issue of evaluation of the relative importance of performances and their KPIs may be based on the FAHP framework (Chan and Kumar, 2007;Kaya and Kahraman, 2011;Tadic et al., 2015;Hsu, 2012;Kuo et al., 1999).It is assumed that all experts do not have equal importance, so in this paper the aggregation of the individual opinions of experts is performed by using fuzzy ordered weighted averaging (FOWA) (Merigo and Casanovas, 2008) (by comparison to Aleksic et al., 2013;Tadic et al., 2015). The main difference between FAHP and the other methods in the literature is its modelling of a decision problem in a holistic manner.This method offers a number of benefits because (1) the other multi-criteria methods experience difficulty in capturing uncertain and imprecise judgments of experts and (2) FAHP is an efficient tool for handling the fuzziness of the data involved in deciding the preferences or assessment of different decision variables.In this paper, a new approach for handling pairwise comparison based on trapezoidal fuzzy numbers (Wu et al., 2004) is used. In the literature, there are many developed approaches for handling FAHP.The use of the developed approach (Chang, 1996) does not involve cumbersome mathematical operations, and it has the ability to capture the vagueness of the human thinking style.Wang et al. (2008) have shown that the extent analysis method cannot estimate the true weights from a fuzzy comparison matrix and has led to quite a number of misapplications.There are many differences between traditional FAHP (Chang, 1996) and the FAHP which is proposed in Wu et al. (2004).Firstly, fuzzy numbers can extend the range of a crisp comparison matrix of the AHP method.Secondly, in the proposed method, the weights of the criteria and preferences of an alternative under each criterion are derived from the fuzzy preference rations; thus, the developed approach allows for a more reasonable description of the decision-making process and reflects the thinking style of a human. Analysis of performances, key performances indicators and business processes in a seaport The product of seaports belongs to a generic product categories called a service (ISO 2000(ISO :2007)).The performances, KPIs of performances and the business processes will be further discussed.The considered performances of business processes are quality, environmental protection and seaport safety.These performances can be decomposed into various other KPIs, which are described in the following subsections. Quality Quality is defined as the degree to which a set of inherent characteristics fulfil requirements (ISO 9000:2005), so fuzzy sets may be used when it needs to be assessed (Yaqiong et al., 2011).This performance of a business process has a high impact on customers, income and indirectly on the long-term sustainability of the seaport.The KPIs affecting this performance can be determined through the use of literature data and as a result of good practice.KPIs of the quality (Tadic et al., 2013) of seaport services are derived from ISO 9001:2008 and Resolution 10011 and are described as follows. Quality of the seaport services (Q1) A seaport usually defines this KPI through the satisfaction and loyalty (ISO 10002:2014;ISO 10003:2007) of customers.It is supported by the quality of the internal customeroriented activities of the seaport and customers' perception of these activities. Average number of customers (Q2) This KPI is very important for overall profit, the local community and company image.The impact on the local community is important since customers satisfy their needs in a seaport by using the infrastructure around it (hotel services, banking services, shops, etc.). Average number of vessels in the queue (Q3) As a seaport is customer oriented, this number should be as low as possible so the satisfaction of vessel owners and passengers will be increased.Additionally, this performance is important for different organizational units of the seaport such as repair services or services for loading and unloading vessels.This should be assessed in communication with the different services in the seaport that define approach positions and anchoring places of vessels. Pilotage and operation of the vessel (Q4) This performance is important from the perspective of vessel owners, customers and seaport management.All of them always require the minimum time needed for placing vessels in the limited seaport infrastructure.This should lead to overall cost minimization. Environmental protection In seaports worldwide, many accidents may occur, leading to the pollution of the environment and biodiversity change.This is further linked to the decrease of business effectiveness in a seaport and, in the worst case scenario, it can lead to total stoppage of provision of seaport services.It is important that maritime transport operates in a safe, secure and environmentally friendly way, so the EU has introduced legislation under Port State Control Directive 95/21.Besides this, ISO 14001:2004 sets out the criteria for an environmental management system (EMS) so that, in compliance with its demands, KPIs that describe environmental protection can be measured in terms of the following. Quality of air (E1) The level of air quality is important from the perspective of public health and change of biodiversity.It is determined and should be assessed by the level of smoke, dust and harmful gasses present.According to the evidence data, around 95.75 % CO 2 , 22 % nitrogen oxides and 0.6 % sulfur oxides are emitted into the air.All of these could lead to greenhouse effects and damage of the ozone layer. Water quality (E2) This KPI is related to the protection of sea biodiversity, tourists and the local community.The level of water quality depends on the presence of microbiological, mechanical and chemical substances which are discharged by vessels entering the seaport. Noise (E3) This is significant from the perspective of customers.The other interested party is seaport management since noise represents a source of pollution.The increase in noise level may lead to change in biodiversity and to the minimization of profits since it reduces the satisfaction of customers and other stakeholders. Hazardous substances (E4) Hazardous substances may be generated in the majority of technical processes in a seaport and they potentially represent the most dangerous pollution source for the environment.The management of hazardous substances is a very im- portant task for a seaport when keeping biodiversity, public health and long-term sustainability in mind. Seaport safety This significant performance has to meet legislative demands and has a serious impact on seaport competitiveness.Different accidents could occur in ports, causing extensive loss of life, damage to vessels and cargo, serious water pollution and changes in biodiversity.The KPIs of seaport safety may be defined based on the literature (Pak et al., 2015;Trbojevic and Carr, 2000).Based on the literature review (Pak et al., 2015) and evidence data of Montenegro seaports, the following KPIs are identified as the most significant. Vessel safety (S1) This KPI is related to the number of accidents caused, for example, by the collision of vessels in the port and docks, unmooring from the dock and capsizing of small boats.This KPI may be assessed by taking into account (Trbojevic and Carr, 2000) size, type, age, crew, maneuverability, pilotage requirements and escorting requirements of the vessel. Traffic volume (S2) A traffic-related factor may be seen as the "volume of traffic inside a port" (Pak et al., 2015).While assessing this KPI, a comprehensive database of port accidents may be used. Weather, sea, and channel conditions (S3) This KPI deals with (1) weather conditions, such as wind speed, sea state and visibility (Balmat et al., 2009) and (2) channel conditions including the perspectives of depth, complexity and width (Pak et al., 2015). Other safety factors (S4) Many factors impact the safety of the port and thus they may be addressed as one joint KPI and taken into account.These factors may include, for example, fire safety, in-port communication, terrorist attacks and natural disasters. The model for evaluation of seaport business processes The proposed evaluation procedure using FAHP is presented in Fig. 1. The evaluation procedure should be delivered by the expert team, which is comprised of the owner, main manager, local government and the operational management of the seaport.Formally, this expert team is presented by a set of indices {1, . . ., e, . . ., E}.The index for an expert is denoted as e, and E is the total number of experts.The members of the expert team have different influences in the considered decision-making process.The importance of experts, w e , e = 1, . . ., E, should be determined with respect to the results of good practice. Experts and operational managers use the predefined linguistic expressions, which are modelled by triangular fuzzy numbers (TFNs).The shape of the membership functions can be obtained based on one's experience, the subjective belief of decision makers and their knowledge.Jointly used shapes of triangular function offer a good compromise between descriptive power and computational simplicity. The total number of KPIs under performance k, k = 1, . . ., K is denoted as J k ; j is the index for KPI j , j = 1, . . ., J k . The fuzzy rating of the relative importance of each pair of performances and their KPIs is described by each expert and presented by TFN W e kk = (l e kk , m e kk u e kk ), k = 1, . . ., K and V e jj = (l e jj , m e jj u e jj ), j = 1, . . ., J k .The aggregation of individual opinions into a group consensus is calculated by the performed fuzzy ordered weighted averaging operator (Merigo and Casanovas, 2008).The aggregated value of the considered variables is Similarly, the aggregated value of the relative importance of each pair of KPIs under the identified performance is determined. Fuzzy pairwise comparison matrices of the relative importance of performance and the relative importance under performance and preference of business processes of each KPI are stated.It is necessary to determine the coefficient of consistency to reflect the consistency of the decision makers' judgement during the evaluation phase by using the eigenvector method (Saaty, 2008).The eigenvector method represents a natural measure for inconsistency and it is used extensively in the literature and in this paper.It is worth mentioning that all relevant consistency indices (C.I.'s) should be equal to or below the threshold of 0.1.The weights vector of performances, weights vector of KPIs under each performance and the preference vector of business processes with respect to each KPI are determined by FAHP, which is developed in Wu et al. (2004). The developed procedure is illustrated on the example of determination of the performances' weights vector in compliance with Eqs. ( 2) and (3): and The weight of performance, k = 1,. . ., K, is calculated as The weight of KPI j , v k j = (l k j , m k j , u k j ), j = 1, . . ., J k , k = 1, . . ., K and the preference of business process i p k ij = (l j i , m j i , u j i ), i = 1, . . ., I are calculated in a similar way to Eqs. ( 2) and ( 3). The reference model of an organization (in this case a seaport) may be seen as a general model which can be used to gain other forms of models (Spiegel and Caulliraux, 2012).In compliance with this, an organization may be viewed as a network of interrelated processes that are focused towards achieving organizational goals (Oakland, 2004).The defining of seaport business processes is based on the process approach (ISO 9000:2008) and the assessment of seaport operational management (quality manager, environmental manager and security manager).The identified business processes are presented by the set of indices {1, . . ., i, . . ., I }.The total number of treated business processes is I and i, i = 1, . . ., I is the index of the business process.The assessment of the relative preference value of each pair of business processes is achieved by group consensus. The ranking of business processes is performed according to the overall index of preference.The preference index of business process i, i = 1, . . ., I under performance k can be calculated as The overall preference index of each business process is described by a TFN.The overall preference index of business process i, i = 1, . . ., I can be calculated as The rank of business processes corresponds to the rank of TFNs which are described by the overall indices' preferences. The ranking of the TFNs p i , i = 1, . . ., I and the calculation of the degree of belief that other business processes can be better than the business process placed first in the rank are based on the method for comparison of fuzzy numbers (Bass and Kwakernaak, 1977;Dubois and Prade, 1979).4 Application of FAHP in business processes' ranking The proposed model was tested in the Kotor seaport, located in a region protected under national legislation.In recent years, the seaport has been certified with ISO 9001:2008 andISO 14001:2004.This seaport is a relatively small port so this fact is taken into account during the definition of the reference model of the organization. In literature concerning business process management, processes of seaport services represent the processes of realization (Arsovski, 2013).The number and type of business processes in a seaport is defined with respect to the American Productivity and Quality Center (APQC) Process Classification Framework (PCF) and process owner's opinion (Table 1).A short description of the selected business processes in ports is further discussed.This process covers a set of activities to be implemented under the common goals of the process (responsibility for each activity, resources, timelines and desired outputs from each activity in terms of the characteristics of services and processes).This corresponds to the "Plan for and align supply chain resources" process which is defined in the APQC specification. Technology management of service providing (p = 2) This process covers standard procedures for vessel access to the port, pilotage, maintenance, port transportation, disembarking and cleaning, among others. Maintenance of infrastructure (p = 3) This process covers maintenance procedures of docks, cranes, warehouses and roads, as well as other transport manipulating systems.This corresponds to the "Manage logistics and warehousing" process (defined in APQC).This process is defined in compliance with the APQC specification and is important from the perspective of seaport sustainability.The effectiveness of this business process is important for the management of the port and the local and state administration. Activities in the seaport (p = 5) This is a complex business process where many different activities are defined and realized according to APQC and literature data (Medison, 2005).These activities are material purchase, service delivery to seaport customers, marketing and service sale, management of customer demands, management of information technology and knowledge, management of financial resources and management of external relations. Business processes' ranking on real-life data The notation used is provided in Table 2. Based on the internal policy of the treated seaport, the expert team is adjoined with different specific weights (Table 3). The elements of the constructed fuzzy pairwise matrices are defuzzified and the consistency of the fuzzy pairwise matrices is determined.This is determined by comparison with Torfi et al. (2010). The weights of sub-criteria under quality performance are v 1 1 = (0.19, 0.28, 0.38), v 1 2 = (0.19, 0.29, 0.41), v 1 3 = (0.09, 0.14, 0.25) and v 1 4 = (0.21, 0.28, 0.41).The overall preference index of each business process was calculated by using this procedure.The rank of business processes with respect to all identified evaluation criteria and their weights and the degree of belief that a business process can be placed at first place in the rank were calculated and presented in Table 7. Discussion According to the final score, business process p = 2 is preferred because it has the highest priority.According to the calculated degree of belief, it may be assumed that all identified processes are significant for the seaport and, at the same time, it can be suggested that the management team has defined an adequate reference model for an organization.In the last place in the rank is business process p = 4.In the treated seaport, occupational health and environmental protection based on the OHSAS 18001 standard has been introduced recently.Some activities related to occupational health and environmental protection are delegated to employees that have not been part of the management team.From this fact, it can be concluded that the management team has not given a full commitment to the new demands and does not have enough knowledge; therefore, the assessment was obtained through previous experiences.In the course of determining the appropriate actions for performance enhancement within each identified business process, it is necessary to present the sensitivity of each business process with respect to the KPIs and the main performances (Figs. 2 and 3). Business process p = 1 is the most sensitive with respect to quality performance.Since customers represent the end users of seaport services, a low level of quality of the treated business process will decrease profits.KPIs that generate the highest impact within this performance are Q2 and Q3.Management initiatives which could lead to the enhancement of the denoted KPIs are application quality methods (for example, quality function deployment, Define-Measure-Analyse-Improve-and-Control, and cost-benefit analyses). Business process p = 2 is the most sensitive with respect to environmental protection.In relation to the conducted activities during this process's realization (maintenance of vessels, port transportation, cleaning, garbage and hazardous substance disposal, etc.), the quality of air and water could be decreased, and the generation of noise and leaking of hazardous substances could be increased.It may be concluded that all KPIs are almost equally important.Management initiatives that should lead to KPI values' enhancement should cover activities of the definition of procedures that are based on international standards and directives.Other ac- tivities could be oriented to the training of employees.It is worth mentioning that business process p = 2 is very sensitive to safety performance of the seaport.In that manner, KPI S2 has greatest impact on this performance. Business process p = 3 is the most sensitive in terms of environmental protection performance.Endangering the environment occurs during the implementation of maintenance dock activities and cranes, as well as other transport manipulating systems, warehouses, roads, etc.The most significant KPIs in the scope of this analysis are E2 and E3.The management initiatives that should lead to the KPI values' enhancement correspond to process p = 2. From the data in Fig. 2 and Tables 4-6, it can be concluded that the business process p = 4 has an almost equal impact on all three treated performances.Enhancement of this business process can be achieved by applying different procedures, which should lead to the increase of KPI values' emphasizing of safety performance.These procedures should be in compliance with international standards and directives. When business process p = 5 is analysed, quality performance has the most significant impact in terms of sensitivity.Most of the activities generated by this process are customer oriented; therefore, low performances in this process could lead to a decrease in competitiveness and a bad image of the port.Enhancement of this process with respect to qual-ity performance may be achieved by applying the measures for enhancement of KPI Q2. Research implications By comparing papers which propose a model for evaluating business processes under uncertainties, certain differences could be noted, which are further described.This analysis, at the same time, shows the advantages of the proposed model. In this paper, it is assumed that determination of the relative importance of the performance of business processes and the relative preference of KPIs of performances and priority of business processes are more reliable when obtained using pairwise comparisons than when they are directly obtained.This is because it is easier to make a comparison between two criteria than make to an overall weight assignment.The fuzzy pairwise comparison matrices of the relative importance of performances, their KPIs and preference of business processes were constructed.The weights vectors of performances and of KPIs under the performance and preferences vector of business processes for each KPI are calculated by using the method developed in Wu et al. (2004).This can be denoted as the main difference between this paper and others (Tadic et al., 2013;Pak et al., 2015;Kaya and Kahraman, 2011;Hsu, 2012). The overall index of the preferences for each business process is described by TFNs.According to fuzzy algebra rules, values of the overall index of a preference are not TFNs, but it is possible to express approximated values of fuzzy operations as TFNs (Kwong and Bai, 2003).Therefore, according to the overall index of a preference, the ranking order of all business processes can be determined and the most important one from among a set of treated business processes can be selected.The degree of belief that any business process can be the business process associated with the highest value of the overall index of a preference can be determined.The priority of management initiatives that should lead to the enhancement of business processes should be based on the rank of business processes and the calculated degrees of belief.The introduced modifications in determining the priority of management measures represents the main difference and, at the same time, the advantage of the proposed model compared to the proposed FAHP methods which can be seen in the literature. Conclusion and future work Seaport management practices show that the evaluation and enhancement of business processes represent the most relevant issues for competitiveness and sustainability.The definition of an enhancement strategy should be based on the rank of the business processes.The main performances and their KPIs are determined in compliance with the process approach and ISO 20000-1:2000.A large number of decision variables demonstrating the complexities are involved in the ranking of business processes.It is assumed that the application of analytic methods in determining the rank of business processes is better than applying intuitive decision-making methods.It may be suggested that each solution obtained in an exact way is less encumbered by the subjective views of decision makers, which could make it more accurate. A fuzzy AHP has been proposed.Uncertainties in the relative importance of each pair of performances and their KPIs and the preference of business processes with respect to each identified KPI are described by predefined linguistic expressions which are modelled by using the fuzzy sets theory.The fuzzy approach is easy to understand, flexible and tolerant to imprecise data.These linguistic expressions are modelled by TFNs. The evaluation of the relative importance of business process performances and their KPIs is based on knowledge, the experience of the seaport decision makers, the needs of local government and other stakeholders.Applying fuzzy group decision-making in determining these decision variables can be considered as one of the contributions of this paper.The main contribution of this paper may be seen as an application of the proposed FAHP with the goal of obtaining the fuzzy rank of business processes and the degree of belief that a business process can be placed in first place.With respect to the fuzzy rank and degrees of belief, it is possible to rationalize the expenditure of time, money and other resources.Also, a good scheduling of management initiatives' orders could increase the efficiency of the enhancement strategy.This can be considered as the main contribution of the proposed FAHP, which was tested with real-life data; the obtained results were presented. The main advantages of the proposed FAHP are related to the fact that it does not involve cumbersome mathematical operations, and it could be easily employed within seaports which operate in an uncertain environment.The proposed FAHP can be easily extended to the analysis of other management decision problems in different research areas.The general limitation of the model is the need for well-structured business processes and the comprehensible definition of their performances. Finally, it is clear that further research could cover a more detailed decomposition of business processes, an increased number of performances and their KPIs, and connection of the business processes of the treated seaport with business processes of other seaports in similar regions. Data availability The setting of the problems and calculations are given in the Appendix.Additional explanations regarding the research in this paper may be obtained upon request by writing to the authors. Figure 1 . Figure 1.The evaluation procedure of seaport business processes by FAHP. 4. 1 Selected business processes in ports 4.1.1Planning and service monitoring (p = 1) Figure 2 . Figure 2. Sensitivity of each business process with respect to the KPIs. Figure 3 . Figure 3. Sensitivity of each business process with respect to the performances. Table 1 . Identified business processes in the seaport. Table 3 . Specific weights of expert team. Table 4 . Preference indices of business processes and their rank under quality performance. Table 5 . Preference indices of business processes and their rank under environmental protection performance. Table 6 . Preference indices of business processes and their rank under safety performance. Table 7 . The overall preference index.	2018-12-10T23:32:49.066Z	2017-02-23T00:00:00.000	{ "year": 2017, "sha1": "7b17a75cb01eef87bdd17fafffb25c2bd244705f", "oa_license": "CCBY", "oa_url": "https://www.nat-hazards-earth-syst-sci.net/17/261/2017/nhess-17-261-2017.pdf", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "1b7cc6f375aea950dedc5e0b80c17a35352eaa3d", "s2fieldsofstudy": [ "Environmental Science", "Business", "Engineering" ], "extfieldsofstudy": [ "Engineering" ] }
54572248	pes2o/s2orc	v3-fos-license	Analysis of L-type amino acid transporter in canine hepatocellular carcinoma Analysis of L-type amino acid transport expression of hepatocellular carcinoma cells (HCCs) of the dog was performed. The leucine transport activity of canine HCCs was 0.628 ± 0.018 nmol/mg protein/min. The inhibitor of LAT 2-aminobicyclo[2.2.1]heptane-2-carboxylic acid (BCH) reduced 90% of the activity at 1 mM. The deduced amino acid sequences of canine LAT2, LAT3 and LAT4 were well conserved in mammalians, exhibiting 89, 88 and 77% homology, respectively. RT-PCR revealed distinct LAT1 expression compared with normal hepatocytes. Western blotting analysis confirmed the potent LAT1 expression in canine HCCs but not hepatocytes, and real-time RT-PCR analysis indicated that canine HCCs possessed 28 times higher LAT1 expression than hepatocytes. These results indicated that the leucine transport activity of canine HCCs was due to LAT1. Many amino acid transport systems have been distinguished based on differences in their substrate-selectivity, ion-dependence, pH sensitivity, kinetics and regulatory properties by using membrane vesicle preparations or cultured cells [5,6]. Among them, system L is an amino acid transport system that mediates sodium-independent transport of neutral amino acids. It was first described in Ehrlich ascites tumor cells as a sodium-independent transport system for neutral amino acids that is specifically inhibited by a bicyclic amino acid, 2-aminobicyclo[2.2.1]heptane-2-carboxylic acid (BCH) [28,29,37]. Later, 2 subtypes for system L with distinct characteristics in substrate affinity, selectivity, and transport properties were reported and named systems L1 and L2 [37]. To date, 4 isoforms of system L have been identified at the molecular level. The first cloned transporters with system L activity were LAT1 and LAT2, which are members of the SLC (solute carrier) 7 family of transporters [17,21,34,36]. They mediate sodium-independent amino acid exchange and recognize a wide range of large neutral amino acids as substrates [35], expanding to small neutral amino acids in the case of LAT2 [30,31,33]. These proteins form heteromeric complexes via a disulfide bond with the heavy chain of 4F2 antigen (4F2hc, SLC3A2), a single transmembrane domain protein essential for the functional expression of LAT1 and LAT2. Later, other transporters were identified by expression cloning [1,3], and it was reported that they exhibited Naindependent, BCH-sensitive neutral amino acid transport activity. In spite of possessing the properties of system L, these transporters (LAT3 and LAT4) were structurally different from system L1 (LAT1 and LAT2), and additionally, they did not require co-expression with 4F2hc to elicit transport activity at the plasma membrane. LAT3 and LAT4 transport activities show specific properties: (i) preferential substrate specificity [6], (ii) complex kinetics [6], which are compatible with 2 simultaneous apparent affinities and (iii) inactivation by the thiol reagent N-ethylmaleimide [25]. Northern blot analysis indicated the tissue distribution of system L2 was different between LAT3 and LAT4. Potent expression of LAT3 was found in the placenta and kidney, and that of LAT4 was found in the pancreas and skeletal muscle [1,3]. Previously, we reported the cDNA sequence and distribution of canine LAT1 in the healthy dog and in several tumor cells [27]. The accumulating data indicated the close relationship between tumor malignancy and LAT1 expressions in human tumor tissues [15,16]. Notably, the expression of LAT1 has been described to be a significant factor indicating a poor outcome in various human cancers [2,12,14,22,32]. Hepatocellular carcinoma is the most common primary hepatic tumor in dogs and arises from the uncontrolled proliferation of hepatocytes. Viral infections have been associated with hepatocellular carcinoma in humans, but no causal link with hepatocellular carcinoma in dogs has yet been established. While Fukumoto et al. reported the LAT1 expression in the canine mammary gland tumor and melanoma in the veterinary clinic [9][10][11], little information is available concerning LAT expression in hepatocellular carcinoma cells (HCCs) of dog. In this study, we examined functional analysis of LAT activity, determined the molecular structure of canine LAT3 and LAT4 and investigated the expression of the 4 isotypes of LATs by RT-PCR and Western blot analysis. Animal and cells samples: All experiments were performed according to the guidelines of the Laboratory Animal Care Committee of Azabu University, and were in compli- ance with the Fundamental Guideline for Proper Conduct of Animal Experiment and Related Activities in Academic Research Institutions. A tumor was removed from an 11-year female Yorkshire Terrier with hepatocellular carcinoma and diagnosed histologically. Tissues samples were fixed in 20% buffered formalin, embedded in paraffin, sectioned at 4 µm and stained with hematoxylin and eosin (HE).A cell line was established after transplantation to a nude mouse, cultured through >100 passages and maintained in Eagle's MEM (Gibco, Carlsbad, CA, U.S.A.) supplemented with 10% fetal bovine serum, 6.6 mg/l spectinomycin and 2 mM glutamine in a 60 mm dish (BioCoat Collagen I 60 mm Dish, Corning, Tokyo, Japan) 37°C in a 5% CO 2 humidified incubator. Tissue samples were also obtained from a healthy 5-year-old male Shiba dog. Determination of the cDNA sequence of canine LATs: Determination of the cDNA sequence was carried out as described previously [26]. Total RNA was isolated from tissues of the healthy dog using an RNA extraction solution (Isogen, Nippon Gene, Tokyo, Japan). The primers used in this study were prepared according to the conserved sequence region between the human and mouse. The DDBJ accession numbers of the human and mouse LATs used in this study were as follows: human LAT2, AF171669; mouse LAT2, AF171668; human LAT3, AB103033; mouse LAT3, AB103034; human LAT4, AB120364; mouse LAT4, AB120363). RT-PCR amplification was performed by employing a SuperScript III first-Strand Synthesis System kit (Invitrogen, Carlsbad, CA, U.S.A.) with Hot start Ex Taq DNA polymerase (Takara Bio, Kyoto, Japan). The band was excised from the agarose gel and purified using a Wizard SV gel clean-up system (Promega, Tokyo, Japan). The extracted and purified DNA was cloned into a pCR II-TOPO cloning vector (Invitrogen) and sequenced with a BigDye Terminator kit ver.3.1 Cycle Sequencing Kit (Applied Biosystems, Carlsbad, CA, U.S.A.). The nucleotide sequence obtained exhibited high similarities (>85%) to human LATs cDNA sequences. In order to determine the 3′ and 5′ regions of cDNA, RACE methods were carried out using a SMARTer RACE cDNA Amplification Kit (Clontech, Mountain View, CA, U.S.A.) and a set of canine LATs sequences obtained. Measurement of leucine transport activity: Radioactive ( 3 H-) L-leucine was purchased from American Radiolabeled Chemicals (St. Louis, MO, U.S.A.). Na-independent leucine uptake was measured as described elsewhere [20]. Simply, all buffers containing Na were replaced with N-methyl-Dglucamine. The cells were plated 5 × 10 5 cells/6-well plate 24 hr before the experiment. After washing the cells, the medium containing 10 µM leucine with a radioisotope was added and incubated at 37°C. To study the effect of BCH, the inhibitor of LAT, was included in an incubation medium containing radiolabeled leucine. Uptake was terminated by washing with ice-cold phosphate-buffered saline. After solubilizing the cells with 1% SDS, the radioactivity was measured with a liquid scintillation counter, and the protein content was determined by the Micro BCA method. RT-PCR analysis of the 4 isotypes of LAT expression in hepatocarcinoma and normal hepatocytes: In order to examine the expression of LATs, we performed RT-PCR using the primers specific to each LAT (Table 1). RT-PCR conditions were as follows: 25 cycles of three steps, 94°C for 15 sec, 68°C for 15 sec and 72°C for 20 sec. Real-time RT-PCR evaluation of LAT1: The housekeeping gene RP19 was used as a reference gene. The primer set for LAT1 used for real-time PCR was designed and purchased from Takara (Table 1). Real-time quantitative PCR was performed using a Thermal Cycler Dice ® Real Time System II (Takara). Samples (final volume of 25 µl) were run in duplicate and contained the following: X1 SYBR ® Premix Ex Taq™ II (Takara), 1 µl 10 µM of each primer and 2 µl cDNA template. Amplification was conducted using the following protocol: initial denaturation phase at 95°C for 10 sec and then 40 cycles at 95°C for 5 sec for denaturation and 60°C for 20 sec for annealing and extension. The qRT-PCR results are presented as the gene expression of the target gene (LAT1) relative to that of the housekeeping gene (RP19), and LAT1 gene expression levels were obtained using the standard method of quantification. Western blot analysis of LAT: The anti-canine LAT1 serum was prepared as described previously [26]. The cell membrane of the tissues for Western blot analysis was prepared as reported by Denker et al. [7]. In brief, the cells were homogenized at 4°C in the buffer containing 0.1 M KCl, 5 mM Na 2 HPO 4, pH 7.5, 0.75 mM Na-EGTA, pH 7.5, 1 mM DDT, 5 mM MgCl 2 , 200 µg/ml phenylmethylsulfonyl fluoride and 4 µg/ml leupeptin. Homogenates were centrifuged for 10 min to remove debris. The 1 ml supernatant was laid Fig. 6. Amino acid sequences of the four isotypes of canine LATs were compared. Multiple sequence alignments were performed using the GENETYX program (ver. 10). Asterisks and dots indicate identical residues and conservative substitutions, respectively. over a 5 ml sucrose solution containing 0.8 mM sucrose and 2 mM Na-EGTA, and was centrifuged at 32,000 g for 40 min. Protein concentration of the pellet was determined by the BCA method and was used for Western blot analysis. The membranes were simply solubilized, electrophoresed into 12% polyacrylamide gels and immunoblotted with a chemiluminescence autoradiograph. The membrane was then treated with a primary antibody against the C-terminus of canine LAT1, followed by a secondary antibody (anti-rabbit IgG (H+L) goat IgG Fab' HRP, ×100,000, Seikagaku Corp., Tokyo, Japan). The LAT1 protein was detected with an ECL Plus chemiluminescence detection system (GE Healthcare Bioscience, Chalfont, UK) and exposed to x-ray film. RESULTS Histologic sections from the specimen of hepatocellular carcinoma from the dog are shown in Fig. 1 (A, B). The tumor was composed of 2 dissimilar forms, hepatocellular carcinoma and cholangiocellular carcinoma. The tumor cells contained various-sized cord-like structures and conspicuous nucleoli. Histologic sections of a nude mouse with a transplanted tumor are also shown in Fig. 1 (C, D). Mitosis was observed, but cord-like structures disappeared. The tumor cytoplasm was abundant and eosinophilic. The leucine transport activity of canine HCCs is indicated in Fig. 2. The transport activity was 0.628 ± 0.018 nmol/ mg protein/min, while in the presence of 1 mM BCH, the specific inhibitor of carnitine transporter, the transport activity was reduced by 90%. Figures 2-4 show the deduced amino acid sequences of canine LAT2-LAT4 and a comparison with those of other mammalians. Canine LAT2 was 1 amino acid longer than that of the mouse, and 3 amino acids shorter than that of the human (Fig. 8). Canine LAT3 was 1 amino acid shorter than that of the human. Mouse LAT3 possessed 5 and 14 amino acid insertions in the same location compared with those of the dog and human (Fig. 4). Canine LAT4 consisted of the same number of amino acid resides as human LAT4, while mouse LAT4 possessed 4 amino acid insertions (Fig. 5). Canine LAT2, LAT3 and LAT4 exhibited 89, 88% and 77% homology among mammalians, respectively. The homology of canine LAT1 and LAT2 was 54%, and LAT3 and LAT4 showed 58% identity (data not shown). On the other hand, significant homology was not observed among LAT1-LAT4 (Fig. 6). Figure 7 shows the RT-PCR analysis of LAT expression in HCCs and hepatocytes from the healthy dog. A potent signal was observed only in LAT1 in the hepatocarcinoma, while LAT1 was not observed in hepatocytes. Signals of LAT3 (weakly) and LAT4 were detected in hepatocytes, while LAT3 and LAT4 were not observed in HCCs. As a control experiment, the expressions of LAT2 and LAT4 were confirmed in the kidney, and that of LAT3 was confirmed in the pancreas (Fig. 7B). Detailed evaluation of LAT1 expression in the hepatocarcinoma and healthy liver was investigated by real-time RT-PCR analysis. Expression of LAT1 in the hepatocarcinoma was 28 times higher than that in the healthy liver (Fig. 8A). Western blotting analysis using anti-LAT1 serum confirmed the certain signal in the hepatocellular carcinoma (Fig. 8B). RT-PCR analysis indicated no LAT3 and LAT4 expression in canine HCCs, while hepatocytes possessed weak expression. DISCUSSION In this study, we investigated the leucine transport activity and confirmed potent BCH-sensitive, Na-independent transport activity in canine HCCs. To evaluate the contribution of 4 LAT isoforms, we determined the cDNA sequences of LAT2, LAT3 and LAT4, and the expression of the 4 isotypes of LATs was examined by RT-PCR analysis. Distinct LAT1 expression in canine HCCs was observed, while other signals were either weak or not detected. Real-time RT-PCR evaluation of LAT1 indicated that the expression of LAT1 was 28 times higher in HCCs than that in hepatocytes. Western blotting analysis using anti-LAT1 serum supported these results. Therefore, potent BCH-sensitive, Na-independent leucine transport activity occurred due to LAT1. LAT1 has been closely associated with cancerous or proliferative cells, and previous studies have shown it to be highly expressed in proliferating tissues, many tumor cell lines and primary human tumors [18,24,38]. The expression of LAT1 has been described to be a significant factor indicating a poor outcome in various human cancers [12,16,22,33]. Cancer cells require considerable nourishment to meet their demands for rapid growth and cell division. Therefore, upregulation of amino acid transporters, especially the system L amino acid transporters that supply large neutral amino acids including many essential amino acids, has been reported [8]. Notably, downregulation of LAT1 by silencing LAT1 shRNA introduction or BCH treatment suppressed invasion and migration of cholangiocarcinoma [13]. Hepatocellular carcinoma is the most common primary hepatic tumor in dogs. Canine hepatocellular carcinoma arises from the uncontrolled proliferation of hepatocytes. The pathogenesis of hepatocellular carcinoma is a multistep process involving sequential events, such as chronic inflammation, hyperplasia, dysplasia, and ultimately, malignant transformation [4]. Several epigenetic and genetic alterations are involved in hepatocellular carcinoma, which ultimately lead to alterations of molecular pathways. Upregulation of LAT1 was reported to be associated with a poor prognosis in malignant tumors [19,23]. Potent expression of LAT1 may involve the alternation of amino acid metabolism in hepatocellular carcinoma. Therefore, alteration of LAT1 function could possibly be a therapeutic target for canine hepatocellular carcinoma. Although we confirmed that LAT1 played an important role in leucine transport in canine HCC in this study, only one cell line was investigated. We will next investigate more cell lines of canine hepatocellular carcinoma or other tumor cell lines. The data clarified in this paper may facilitate the study on understanding of the pathophysiology of canine hepatocellular carcinoma.	2018-04-03T02:27:40.955Z	2015-01-29T00:00:00.000	{ "year": 2015, "sha1": "a69822e0e40758f3404ff18fe6d7d16ec0fb2a94", "oa_license": "CCBYNCND", "oa_url": "https://www.jstage.jst.go.jp/article/jvms/77/5/77_14-0392/_pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "a69822e0e40758f3404ff18fe6d7d16ec0fb2a94", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [ "Biology", "Medicine" ] }
234371448	pes2o/s2orc	v3-fos-license	An Analysis on Fragmentation, Fetishism and Liquidity in Travel Narratives Literary tastes and genres are getting wider and plethora of new trends in literature makes it vibrant and sensible. In the present decades travel literature; memoirs and nature writing become very much prevalent in the society. Starting from the grand old period to till now the term Travel and its meaning was largely modulated and understood in different dimensions. At the Age of Chaucer, Sir John Mandeville’s travel memoirs was attained a huge popularity among people. It was circulated from 1357 to 1371 with box office hit and it was orinally published in the language of french and later it was translated to many languages. Describing the landscapes, unknown terrians and endless beauty of nature creates a fascination among people. And it helps the reader to viscerally able to witness a new world through words. Beyond that there are many literary genres formulated under the umbrella term ‘Travel’. They are Adventure literature, Journals, emergence of new literary theories such as wanderism, etc., slave travel narratives and so on. Especially slave trade narratives attained a huge sensation among people, mostly it’s an autobiographical accounts of enslaved black people from african continent and sold across the world for doing hard labour. The first slave narrative was wriited in 1789 by Gustavus Vassa Equiano as a memoir. This slave narratives was largely a recollection of his troubleful life. These type of works hepful in providing the authentic truth about the turmoils faced by the coloured people as well as they used writing as a tool to liberate themselves from the racial injustice and discrimination that thrust upon them. In other dimension even before globalization and modernization begins, Travel was very much instrumental in developing an intellectual way of life style from gaining knowledge about various food crops, currencies, medicines, rulers’s lifestlye, architecture, artforms, culture, cuisines and so on which was largely understood through taking travel voyages. The travel and its curiousity begins in the time of greeks and romans who undergoes travel for leisure purposes. And after the time moves on travel turned to have specific motto’s to seek, explore new places with rich resources and to captivate them and later it was devloped as an ideology called OPEN ACCESS Literary tastes and genres are getting wider and plethora of new trends in literature makes it vibrant and sensible. In the present decades travel literature; memoirs and nature writing become very much prevalent in the society. Starting from the grand old period to till now the term Travel and its meaning was largely modulated and understood in different dimensions. At the Age of Chaucer, Sir John Mandeville's travel memoirs was attained a huge popularity among people. It was circulated from 1357 to 1371 with box office hit and it was orinally published in the language of french and later it was translated to many languages. Describing the landscapes, unknown terrians and endless beauty of nature creates a fascination among people. And it helps the reader to viscerally able to witness a new world through words. Beyond that there are many literary genres formulated under the umbrella term 'Travel'. They are Adventure literature, Journals, emergence of new literary theories such as wanderism, etc., slave travel narratives and so on. Especially slave trade narratives attained a huge sensation among people, mostly it's an autobiographical accounts of enslaved black people from african continent and sold across the world for doing hard labour. The first slave narrative was wriited in 1789 by Gustavus Vassa Equiano as a memoir. This slave narratives was largely a recollection of his troubleful life. These type of works hepful in providing the authentic truth about the turmoils faced by the coloured people as well as they used writing as a tool to liberate themselves from the racial injustice and discrimination that thrust upon them. In other dimension even before globalization and modernization begins, Travel was very much instrumental in developing an intellectual way of life style from gaining knowledge about various food crops, currencies, medicines, rulers's lifestlye, architecture, artforms, culture, cuisines and so on which was largely understood through taking travel voyages. The travel and its curiousity begins in the time of greeks and romans who undergoes travel for leisure purposes. And after the time moves on travel turned to have specific motto's to seek, explore new places with rich resources and to captivate them and later it was devloped as an ideology called OPEN ACCESS Volume: 9 And in the present milieu Travel was done for mutiple reasons Everyone travel for to learn; to escape; to discover; to experience the life to the fulllest and add some meaning into it. The significance of travel was largely understood by the people through the current mass media culture but unfortunately it understood in a wrong perception and turned out to be a mere and meek fashion for exploring new lanscapes and without any insights and intentions. Undergoing a compelling journey for seeking public attention through social media now turned out to be a recent fashion on travel. Especially in India lots of youngsters start travelling to northern regions especially to Jammu, Leh and Ladakh regions usually in motorcycles. Hitch hiking also prevalent among them. And narrating their journey's via social media's creates a living for them too. But while scrutinizing their travel and its insights most of them did it for a mere replication from their friends and without any specfic purposes. According to the famous saying Any job done without a want finally results as waste of time and energy. Likewise beyond the flavours of explorarion and love for traveling majority of people perceives in a false notion of potryaing their bourgeiose culture and it's tendencies. As a strong upholder of of materilaistic beliefs and imposing it all on the different streams of life. Focusing on the face value and blindly ignoring the utility of any things and rejecting all notions of logicality and reasoning can be defined as a bourgeoise or our present mass media culture. It fetishes all irrespective of gender, things and so on. For example Che guevara, Periyar, Jesus, Tamil Poet Bharahiar and so on are largely turned out to be a fashion icon beyond understanding their crucial contribution to the world. Likewise Travel and exploration was seen in a perspective of fashionble asset not as a tool for enriching oneself. A small minded exploration with economic supremacy for attaining social supremacy in the name of Art, Culture, Politics, Religion and so on can be called as Bourgeoise lifestyle or Fetishism. Consciously or unconsciuosly people are getting trapped and shackle down in the nothingness of Bourgeoise culture and it results in the production of endless replicas. The sheer orginality of people was largely reduced into a single monopolistic fabric of humaness. Eevn the world turns it's notion and ideology into Socialistic and capitalistic way, the prominence of Marxism is largely questioned as well as criticised. And it was considered a false and unjust ideology and turned out to be practically impossible in the neo-liberalised society where globalization and modernization puts it's stance firmer. And the emergence of single monopolies can be seen across the various nations as a result of it. In India Reliance Group's telecommunication brand Jio, U.S Apple, Britain's Standard Oil corporation and so on. These are some of the world MNC's which invest in various fields and earning enormous power and has the decisive role in maintaining the power and politics of the respective nations in one way or other way. In reference to the The Hindu newspaper an article was published in the last week (April 25,2020) about the 'Joining of Gaints', The Facebook decision to invest forty three thousand crores of money for a 9.9% stake in Reliance Industries especially in Jio platform which marks a focus on the Indian retail market. But in other way it produces a risk of an emergence of strong monopoly in India which shatters the life of million small scale retailers. The strategies and paradigims of Capitalism was strongly influenced and intermingles in our way of lifestyle starting from Clothings, food habits, fetishising the westernized way of interpreting art forms, movies, literary style and copying it and following in our indigenous soil. It directly or indirectly develops a class tension between the people. The famous American Marxist Philosopher Fredric Jmaeson in his phenomenal work, PostModernism states that the present post-modern condition results in the production of commodification tendencies among people and each and every things in the society was valued based on it, this provides a greater danger for pursuing truth, knowledge and historicity from the past. ''Post modernism is not the cultural dominant of a wholly new social order.... but only the reflex and the concomitant of yet another systemic modification of capitalism itself'' (27). According to him even the ideas are merely commodified and quantified in the post modern era. Here the researcher focused this idea with the modern travel adventurous journey undertook by the people only for commdifying themselves than understanding the elements and the importance of travel. The real historicity and the significance of travel was largely blurred and it was turned as a particle of mass culture. Starting from travel to all other things exist in the society and it's perception was largely modulated by the idea of post modernity. Another crucial factor exist in the scoiety was our new depthlessness regarding our lifestyle and it's conceptions. Development of false consciousness, lack of thought, our psychic experiences and culture was largely understood with materialstic pursuits results in the fragmentation of our ideas into a menial level. Clarity in thought process and ideological development or conceptual formation of any substance can be done only by removing the pre-conceived notions with the refernece to our affinity and false perception towards the Capitalism.	2020-12-24T09:12:12.467Z	2020-12-22T00:00:00.000	{ "year": 2020, "sha1": "9ed24f6e978142b707f84e783a843a28cf502a24", "oa_license": null, "oa_url": "https://doi.org/10.34293/english.v9is1-dec2020.3615", "oa_status": "GOLD", "pdf_src": "MergedPDFExtraction", "pdf_hash": "f41aa180255ad100cc625b78a34826b061fc6c08", "s2fieldsofstudy": [ "History" ], "extfieldsofstudy": [ "Economics" ] }
119271934	pes2o/s2orc	v3-fos-license	Zero energy Majorana modes in spin ladders and a possible realization of the Kitaev model We show that in double-chain Mott insulators (ladders), disordered alternating ionic potentials may locally destroy coherence of magnetic excitations and lead to the appearance of spontaneously dimerized islands inside the Haldane spin-liquid phase. We argue that a boundary between the dimerized and Haldane phases of a spin-1/2 ladder supports a localized zero-energy Majorana fermion mode. Based on these findings we suggest a realization of a generalized Kitaev model where Majorana fermions can propagate in more than one dimension. INTRODUCTION A possible existence of Majorana zero modes in certain condensed matter systems has been a subject of much discussion in recent literature. The interest is caused by a strong nonlocality of Majorana zero modes which rises hopes of possible applications in such areas as quantum computation [1] and electron teleportation [2]. It has been argued that stable Majorana zero energy modes (MZEM) may appear on surfaces [3] and in vortex cores of certain superconductors with a nontrivial topological invariant (this includes spinless p x + ip y ones) [4], [5]. As far as magnetic systems are concerned, the existence of MZEM has been suggested only for a model system (the Kitaev model) [6] which material realization has not yet been achieved. This paper presents two main results. First, we show that localized MZEM appear on a boundary between spontaneously dimerized (SD) and magnetically disordered phases in systems of two coupled spin-1/2 chains (spin ladders). As was demonstrated in [7], [8], the SD phase may exist in spin ladders with sufficiently strong four-spin interaction. There are reasons to believe [10] that such interactions are present, for instance, in spin ladder material CaCu 2 O 3 . Second, we suggest that a balance between competing spin-disordered and SD phases can be shifted by application of a staggered ionic potential which effectively enhances the four-spin interaction. This may be important for applications since realization of such potentials is quite feasible by chemistry means. Based on these findings we suggest a particular realization of a generalized Kitaev model where MZEM are located on a manifold of arbitrary dimension. As in the original Kitaev model, these dispersionless modes facilitate propagation of Majorana fermions with nonzero dispersion so that such propagation indeed becomes becomes akin to teleportation as described in [2]. The idea that an alternating ionic potential can lead to SD was expressed in [11], [12] where the authors studied a transition from the Mott to a conventional ionic band insulator (IBI) in one dimension. It was found that since the Mott and IBI states differ under a local parity inversion, the transition driven by a change in the strength of the alternating ionic potential proceeds through an intermediate spin liquid phase characterized by a spontaneous dimerization. In the present paper we generalize these ideas for the case of (i) disordered ionic potentials, (ii) the systems consisting of two half filled chains coupled together. In the second part of the paper we construct a generalized Kitaev model. STAGGERED IONIC POTENTIAL IN A SINGLE HUBBARD CHAIN Let us briefly recall the single Hubbard chain case. At half filling such chain has a spectral gap in the charge sector and gapless excitations in the spin sector. At low energies the spin sector decouples from the charge one and can be considered separetly. For reasons of convenience we will use the continuous approximation for the entire Hubbard model. This is legitimate if the Mott-Hubbard gap is much smaller than the bandwidth. The effective Euclidian Lagrangian of the Hubbard model in this case is where Φ c,s are charge and spin bosonic fields, v c,s are their velocities, K is the Luttinger parameter and G is the Umklapp coupling constant. All these parameters are related to the bare parameters of the Hubbard model. The cosine term in (2) is relevant for K < 1 (repulsion) and opens a gap (Mott-Hubbard gap ∆ MH ) in the charge sector. The spin sector remains gapless and the correlation functions of the spin fields display power-law decay at T = 0. We omitted in (3) a marginally irrelevant term contributed by backscattering of the electrons. Let us consider the influence of a staggered ionic potential. Using the well known textbook bosonization formulae [13], [14] for the staggered charge density we obtain the following contribution of such potential to the Lagrangian density: where V (π, x) is an envelope made of the Fourier harmonics of the potential concentrated around wave vector π, and ρ stag (x) is the continuum limit part of the staggered density (−1) n C + nσ C nσ . Since the charge sector has a gap and the vacuum average of sin √ 2πΦ c is zero, the effect of the staggered potential shows up only in the second order of perturbation theory in V (2k F = π) [11]: where Φ sR;L are chiral components of the scalar field Φ s , λ ∼ V 2 and α is a short-distance cutoff of the bosonic theory. More precisely, the coupling λ is expressed as a convolution of the disorder average of the potentials with the correlation function of the charge fields: and in principle is coordinate dependent. To provide the convergence of (6) in time domain we need an exponential decay of the dynamical correlation function which occurs only if the charge gap is nonzero. To estimate the average coupling we may assume that In the most interesting case when the disorder correlation (5) is marginally relevant and competes with the marginally irrelevant interaction of the same form present in the spin sector of the original Hubbard model. Therefore such interaction will open a spin gap only when λ exceeds some critical value related to the aforementioned marginally irrelevant backscattering term. Refs. [11], [12] consider only a uniform potential, and one could get an impression that the onset of the gapped dimerized phase would require doubling of the unit cell across the whole chain. However, from the above example it follows that, to achieve the desired effect, it is sufficient to break translational symmetry only locally. DOUBLE CHAINS. THE CASE WHEN CHARGE GAPS >> THAN THE SPIN ONES Now we consider a system of two coupled half-filled chains. For simplicity one may think about them as Hubbard chains even though precise details of the charge sector are not important. It was shown in [7], [8] that at half filling, when the low-energy dynamics in the spin sector of a single chain coincides with the one for the spin S=1/2 Heisenberg model, the continuum limit description of the ladder is given by the model of four Majorana fermions with the Lagrangian density where the Majorana fields χ R , χ L in the path integral are real Grassmann numbers and parameters v, m t , m s depend on the bare interactions. This description works well when the interchain exchange is much smaller than the exchange along the ladder. Then parameters m t , m s (masses of the triplet and the singlet Majorana fermions) are proportional to the linear combinations of the conventional J ⊥ and the four-spin J cycle exchange integrals [8], [9]: (A, B are numerical coefficients). The spectra of the triplet and singlet Majorana fermions are ǫ t,s (k) = (vk) 2 + m 2 t,s ; they do not depend on the signs of m t,s . However, since the spin operators are nonlocal in terms of the fermions, their correlation functions do depend on these signs. In particular, if m t m s < 0 the ladder is in the spin-disordered (Haldane) state characterized by coherent magnetic excitations with spectral gap \|m t \|. If, on the other hand, m t m s > 0 the ground state is spontaneously dimerized (SD) , that is the energy density acquires a nonzero staggered component. Away from quantum critical points where some of the masses are zero, the spin excitations have spectral gaps but become incoherent [8]. In most experimental realizations of spin ladders the interchain exchange is either equal or larger than the exchange along the ladder. In that case the singlet gap is large and the corresponding excitation has no influence on the physical properties. However, in such systems as CaCu 2 O 3 where the rung exchange is much smaller than the leg one the situation is different and the singlet mode (though appearing only in the magnetic continua) visibly affects the dynamical correlation functions [10]. One may expect that by arranging a proper coordinate dependence of the interchain interactions one can create a situation when some or all of the Majorana masses change sign. This will lead to creation of Majorana zero modes [15]. In [16] it was shown that this does happen when one puts a single vacancy on one of the chains; then all Majorana masses change sign. The result of such change, as it might be expected, is a creation of local spin 1/2 localized at the vacancy. Here we would like to consider a different possibility, namely, when only an odd number of Majorana modes change sign. This happens either on a border between SD and magnetically disordered state or, if the SU(2) symmetry in the spin sector is lost, by a change of sign in one the components of the Majorana triplet. The latter can be achieved by applying an easy axis anisotropy. To be definite we will consider the SD case. We suggest that from the practical point of view a local change of m s can be achieved by applying staggered ionic potentials acting on the staggered charge densities of both chains: where V i (x) is a slow function of x = na 0 . In order to achieve the desired effect that only m s changes sign we need V 1 V 2 < 0. In the opposite case V 1 V 2 > 0 the sign change will occur in the triplet mass m t . For simplicity we consider only the former possibiity. To perform actual calculations we adopt the following hierarchy of energy scales 4t ≫ V, ∆ MH ≫ t ⊥ , where ∆ MH is the Mott-Hubbard gap for the individual chains, V is the interchain density-density interaction and t ⊥ is the interchain tunneling. With such assumptions the high energy degrees of freedom are the charge modes which can be described by the following model: where fields Φ c1 , Φ c2 are the charge fields of individual chains, V being proportional to the interchain densitydensity forward scattering. Passing in (10) to the continuum limit and bosonizing the resulting expression we obtain: where ǫ ± are the symmetric and antisymmetric combinations of the chain spin-dimerization operators [17]. In the Majorana representation (8) these operators are expressed as linear combinations of products of four order or disorder parameters of the corresponding Ising models (each noninteracting Majorana fermion is equivalent to the quantum Ising model; see [7] and also [13] for details): At small distances the following fusion rules take place: where κ i = χ i R χ i L . Using fusion rule (14) and keeping only the most relevant terms we obtain in the second order in the potentials the following shift of all masses: Naturally, C 12 vanishes in the absence of interchain interactions. Since in the spin-liquid state the masses of triplet and singlet Majorana fermions have different signs, such shift may lead to a change in sign of one of the masses. In the uniform case this would push the system to a spontaneously dimerized phase via Quantum Critical Point [8]. If the ionic potential is coordinate dependent the sign change will occur locally. For the Hubbard model where the interchain coupling is antiferromagnetic only the singlet mass changes sign. Then on phase boundaries between SD and disordered phases a single Majorana mode has zero energy states, as it occurs with fermionic states located on the boundary of quasi-1D p-wave superconductor [3]. Another direct analogy is a boundary between ordered and disordered states in the quantum Ising model. A purely one-dimensional case (single ladder) will display disorder effects studied in [18] and [16] where the case of random potential was considered. It was shown that when such boundaries are randomly distributed with a final concentration n i they give a singular contrubution to the specific heat: We remark that the models considered in [18], [16] dealt with the situation when the number of Majorana zero modes on each domain wall was four. In that case, as we have already mentioned, each domain wall carries local spin 1/2. HOW TO CONSTRUCT THE KITAEV MODEL A much more interesting situation emerges when the number of modes is odd like in the case we consider. Let us consider the above ladder model with large areas of the dimerized phase separated by large areas of the Haldane phase, such that the wave functions of the zero modes from different domain walls do not overlap. Now let us put these ladders on top of a superparamagnet whose magnetic susceptibility has a peak at (π, 0) wave vector. So the strongest magnetic fluctuations in such a material are smooth along the ladder direction and oscillate in the direction perpendicular to the ladders. Then they strongly couple to the the operator representing the difference between magnetization densities of two legs of each ladder [7]. Integrating over the bulk magnetic fluctuations we obtain the effective interaction between Majorana modes from different ladders: where y is a location of the ladder in the transverse direction, x, x ′ are the coordinates along the ladder and N a is the a-th component of the staggered magnetization of the paramagnet. Since the paramagnet is disordered, this interaction is short ranged. To get the Kitaev model we have to arrange the positions of domain walls in such a pattern that a wall on one ladder interacts only with one other wall. Such pattern does not need to be regular though disorder will lead to localization of the propagating Majorana modes. As an example we may consider a pattern where each ladder contains alternating SD and Haldane phases of alternating lengths a and b. The ladders are shifted by amount c < a in the y direction (see Fig.1). All lengths are assumed to be much larger than the domain wall width so that the wave functions of MZEMs on different domain walls do not overlap. Since the correlation function J ab is short range, we can neglect all interactions between domain walls except between the ones located on neighboring ladders. Hence we get H = (19) y,a=1,2,3 where x c coordinates mark positions of domain walls. For each wall one has to specify whether it holds a zero mode with right or left chirality and replace χ 0 operator with its zero mode component: where the sign depends on the sign of δm s = m s (∞) − m s (−∞), γ(x c , y) is a zero energy Majorana fermion mode satisfying the Clifford algebra anticommutation relations, N is the normalization factor, the dots stand for contributions of higher energy modes. Model (19) is a particular type of the Kitaev model fermionized by means of the Jordan-Wigner transformation, as in [19]. The latter method, when the original spin model is treated as a model of coupled chains and Jordan-Wigner transformation is done for each chain, represents certain advantages since it allows to avoid a cumbersome gauge fixing used in the original paper [1]. The most important feature of the arrangement depicted on Fig. 1 is that each MZEM in Hamiltonian (19) interacts only with one nearest neighbor exactly as in the original Kitaev model. In this way the products of neighboring zero Majorana modes γ(x, y)γ(x, y + 1) are integrals of motion and can be replaced by constants. Thus the four fermion term in (19) becomes effectively a hopping term for Majorana fermions χ a which now can propagate in the transverse direction. The latter fermions have spectral gaps, but such sector also exists in the Kitaev model. To make sure that MZEMs are protected, we consider the operators which couple these modes to local perturbations. The dangerous ones are those which can couple MZEMs located at different positions. Neither K operator considered above nor the staggered components of magnetization can do this since they contain massive modes and their correlations falls off exponentially. This leaves the staggered energy density operators ǫ ± which are products of four order ( disorder) Ising model operators (13). The antisymmetric staggered energy density ǫ − contains operators σ a (a=1,2,3) which have no ground state average and therefore vanishes. Operator ǫ + has a nonzero ground state average affected by MZEMs and therefore may couple to it. The coupling originates from the change in µ 0 ; the operators from the triplet sector do not experience any change at the boundary. Using Eqs.(83,84) from [16] we get (x < 0 is in the SD phase): This function is depicted on Fig.2. Here m s stands for the asymptotic value of the singlet gap far from the boundary. As we see, at the boundary where the MZEM is located the symmetric staggered energy density vanishes: ǫ + (x) ∼ \|x − x c \| 1/8 . This fact renders the interaction with MZEM numerically weak. CONCLUSIONS We have demonstrated that zero energy Majorana modes (MZEMs) robust against external perturbations exist on a boundary between the Haldane and SD phases of spin S=1/2 ladders. We have also suggested a practical way to manufacture systems with such phase boundaries. Namely, the effective interactions in spin ladders can be manipulated by purely chemical means through staggered ionic potentials. There are two ways MZEMs from different domain walls interact between each other: by direct overlap and via interaction of a particular Fourier component of the magnetization (17). The former interaction leads to formation of an impurity band inside of the spin gap with associated singularity in the specific heat (16). However, by maintaining a proper distance between the phase boundaries one can make this overlap small. For this case we have suggested a particular arrangement when the interactions between MZEMs from different ladders create a network similar to the one existing in the Kitaev model in its gapped phase. We are grateful to B. L. Altshuler and V. Gritsev for valuable discussions. A. M. T. is grateful to Abdus Salam ICTP for kind hospitality and also acknowledges a support from the US DOE, Basic Energy Sciences, Material Sciences and Engineering Division. A. N. N. is supported by grants GNSF-ST09/4-447 and IZ74Z0-128058/1.	2011-08-04T15:04:04.000Z	2011-05-29T00:00:00.000	{ "year": 2011, "sha1": "aedd8158c3f6e3f1c6a914f46e2127cd474dfbfe", "oa_license": null, "oa_url": "http://arxiv.org/pdf/1105.5835", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "aedd8158c3f6e3f1c6a914f46e2127cd474dfbfe", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics" ] }
120422294	pes2o/s2orc	v3-fos-license	A STUDY ON THE MIXED CONVECTION BOUNDARY LAYER FLOW AND HEAT TRANSFER OVER A VERTICAL SLENDER CYLINDER Rahmat ELLAHI a,b*, Arshad RIAZ b, Saeid ABBASBANDY c, Tasawar HAYAT d,e, and Kambiz VAFAI a a Department of Mechanical Engineering, Bourns Hall, University of California, Riverside, Cal., USA b Department of Mathematics, Faculty of Basic and Applied Science, International Islamic University, Islamabad, Pakistan c Department of Mathematics, Science and Research Branch, Islamic Azad University, Tehran, Iran d Department of Mathematics, Quaid-i-Azam University, Islamabad, Pakistan e Department of Mathematics, Faculty of Science, King Abdulaziz University, Jeddah, Saudi Arabia Introduction The mixed convection flows under boundary layer analysis are of fundamental theoretical and practical interest.Heat exchangers, electric transformers, refrigeration coils, solar collectors, study of movement of natural gas, oil, water through oil reservoirs and nuclear reactors are few examples in this direction.The flows of non-Newtonian fluids through a porous medium [1][2][3][4][5][6] are also quite prevalent in nature.Examples of these applications are filtration processes, biomechanics, packed bed, ceramic geothermal engineering, insulation systems, ceramic processing, chromatography, and geophysical phenomena.Over the last several decades a large number of research work related to the mixed convection flow of a viscous and incompressible fluid over a vertical flat plate was conducted by a number of investigators.For instance, a critical review of mixed convection flows has been presented by Pop and Ingham [7]. Mahmood and Merkin [8] have presented dual similarity solutions for the axisymmetric mixed convection boundary layer flow along a vertical cylinder in the case of opposing directions only.Later on Ridha [9] has reported that when the buoyancy force acts in the direction of flow then the dual solutions for aiding flow also exist.Ishak et al. [10] have also reported a numerical simulation by Keller-box method for boundary layer flow and heat transfer over a vertical slender cylinder.In this study, the authors concluded that for aiding/assisting flow dual solutions exist but for opposing flow there are either dual solutions, unique solutions or no solution exist. The purpose of the present investigation is to revisit the problem formulated in references [7][8][9][10] for the analytic solutions.The main stream velocity and wall surface temperature depend upon the axial distance along the cylinder surface.The homotopy analysis method (HAM) [11] has been applied to obtain series solutions of velocity and temperature fields by making choices of two base functions.Convergence of the obtained series solutions is taken care properly.Furthermore, the skin friction and Nusselt number are analyzed.Comparison with the previous relevant studies is made.To the best of our knowledge, the series solutions for this particular model have not been presented in the past.It is worth mentioning that in both cases, the dual solutions do not exist, therefore, the comparison with existing results reveal that our series solutions are valid for l > 0 (heated cylinder) and for l < 0 (cooled cylinder) as well.It is shown that like several existing studies [12][13][14][15][16][17][18], the HAM in the present paper is accepted as an elegant tool for effective solutions for a number of complicated fluid problems. Problem formulation Here, we analyze the convection flow and heat transfer characteristic along a vertical permeable slender cylinder with radius a.We choose the cylindrical co-ordinates (x, r) such that x-and r-axes are along the cylinder surface (vertically) and radial directions, respectively.Symmetric nature of the flow is assumed with respect to the transverse co-ordinate.Furthermore, cylinder is kept in an incompressible viscous fluid of uniform ambient temperature T 4 and constant density r 4 .Using continuity, momentum, and energy equations, one obtains the following boundary layer equations [19,20]: In the equations, u and w indicate the velocity components, U(x) -the mainstream velocity, T -the fluid temperature, g -the gravitational acceleration, a -the thermal diffusivity, b -the thermal expansion coefficient, and n -the kinematic viscosity.The subjected boundary conditions are [10]: where V(>0) and V(<0) correspond to the injection and suction velocities, respectively.Further, U(x) and temperature of the cylinder surface T w (x) are: in which R denotes characteristic length and DT is the characteristic temperature.Note that DT > 0 corresponds to a heated surface and DT < 0 for a cooled surface.Defining where the stream function y can be defined by u = r -1 ¶y/ ¶r and w = -r -1 ¶y/ ¶x.Moreover prime denotes differentiation with respect to h.Invoking eqs.( 7) and ( 8), eq. ( 1) is identically satisfied, and eqs.( 2) and (3) yield: and the curvature parameter g and V are: , where f 0 = f(0).It is noted that f 0 < 0 is for mass injection and f 0 > 0 is for mass suction.The buoyancy or mixed convection parameter l is: Here l > 0 and l < 0 are for aiding flow (heated cylinder) and for opposing flow (cooled cylinder), respectively.For l = 0, one has pure forced convection flow without buoyancy force. The boundary conditions (4) and ( 5) reduce in the form: Now the skin friction coefficient (C f ) and the local Nusselt number (Nu x ) are: where ] and the skin friction t w and the heat transfer from the plate q w can be written as: , in which m and k are the dynamic viscosity and thermal conductivity, respectively.Making use by eq. ( 7) one obtains: 1 . . Solution of the problem Here we offer two types of series solutions by choosing exponential and rational bases. Exponential bases According to eqs. ( 9) and ( 10) and the boundary conditions (11), we write the solution in the form: where a 0 , a q,m , and b q,m are coefficients to be determined and z is a spatial-scale parameter.By rule of solution expression denoted by eqs.( 14) and ( 15) and the boundary conditions (11), it is natural to choose: ) as the initial approximation to f(h) and q(h), respectively.We use the method of higher order differential mapping, [21] to choose the auxiliary linear operators L 1 and L 2 by: (20) where C i , i = 1, 2, ..., are constants.This choice of L 1 and L 2 is motivated by eqs.( 14) and ( 15), respectively, and from boundary conditions (11), we have C 2 = C 4 = 0. From ( 9) and ( 10) we define non-linear operators as: and then construct the homotopy: where h 1 ¹ 0 and h 2 ¹ 0 are the convergence-control parameters [22], H 1 (h) and H 2 (h) are auxiliary functions.Setting H i [f(h, z; p), y(h, z; p)] = 0, for i = 1, 2, we have the following zero-order deformation problems: ( subject to conditions: ( , ; ) in which pÎ[0, 1] is an embedding parameter.When the parameter p increases from 0 to 1, the solution f(h, z; p) varies from f 0 (h) to f(h) and the solution y(h, z; p) varies from q 0 (h) to q(h).If these continuous variations are smooth enough, the Maclaurin's series with respect to p can be constructed for f(h, z; p) and y(h, z; p), respectively, and further, if these series are convergent at p = 1, we have: To calculate f m (h), we now differentiate eqs.( 25) and ( 26) and related conditions m times with respect to p, then set p = 0, and finally divide by m!.The resulting m th -order deformation problems are: , where R 1,m (h) and R 2,m (h) are given by: where prime denotes differentiation with respect to h and: , The general solution of eqs.(29) and (30) are: 3 e (32) where C i for i = 1, ... 5 are constants, $ ( ) f m h and $ ( ) q h m are particular solutions of eqs.( 29) and (30), respectively. For simplicity, here we take H 1 (h) = H 2 (h) = 1.According to the rule of solution expression denoted by eqs.( 14) and ( 15), C 2 = C 4 = 0.The other unknowns are governed by: and according to our algorithm, the other boundary conditions are fulfilled.In this way, we derive f m (h) and q m (h) for m = 1, 2, 3,..., successively for every z. At the N th -order approximation, we have the analytic solution of eqs.( 9) and ( 10), namely: For simplicity, here we take h 1 = h 2 = h.The auxiliary parameter h is useful to adjust the convergence region of the series (34) in the homotopy analysis solution.By plotting h-curve, it is straightforward to choose an appropriate range for h which ensures the convergence of the solution series.As pointed out by Liao [11], the appropriate region for h is a horizontal line segment. Rational bases Invoking eqs.( 9) and ( 10) and the boundary conditions (11), one can write: where d 0 , d q,m , and e q,m are coefficients to be determined.According to the rule of solution expression denoted by eqs.( 35) and (36) and the boundary conditions (11), the initial approximations of f(h) and q(h) are selected as: The auxiliary linear operators L 1 and L 2 are : where D i , i = 1, 2, ..., 5 are constants.This choice of L 1 and L 2 is motivated by eqs.( 35) and (36), respectively, and from boundary conditions (11), we have D 2 = D 4 = 0.In this case, the non-linear operator N i [f, y], the homotopy H i [f, y], R i,m (h) for i = 1, 2, and the zero-order deformation equations, the m th -order deformation equations are designed as in the previous case without the parameter z. The general solution of m th -order deformation equations here are: where D i for i = 1, ... , 5 are constants, $ ( ) f m h and $ ( ) q h m are particular solutions of m th -order deformation equations. By rule of solution expressions denoted by ( 35) and ( 36) and from m th -order deformation equations, the auxiliary functions H 1 (h) and H 2 (h) are chosen in the form: It is found that when x 1 < 4 and x 2 < 3 the term log(1 + h) appears in the solution expression of f m (h) and q m (h), which disobeys the rule of solution expression denoted by ( 35) and (36), respectively.In addition, when x 1 > 4 and x 2 > 3 we omit some terms in solution expression.This uniquely determines the corresponding auxiliary functions: According to ( 35) and (36), D 2 = D 4 = 0.The other unknowns are governed by: and according to our algorithm, the other boundary conditions are satisfied.In this way, we derive f m (h) and q m (h) for m = 1, 2, 3,..., successively.Like previous case, we will take here Numerical results and discussion We use the widely applied symbolic computation software MATHEMATICA to solve eqs.( 29) and (30).By means of the so-called h-curve, it is straightforward to choose an appropriate range for h which ensures the convergence of the solution series.As pointed out by Liao [11], the appropriate region for h is a horizontal line segment.We can investigate the influence of h on the convergence of ¢¢ f ( ) 0 and -¢ q ( ) 0 , by plotting the curve of it vs. h, as shown in figs.1-4 (for l > 0) and figs.5-8 (for l < 0) are the examples of two cases.By considering the h-curve we can obtain the reasonable interval for h in each case.Our computations show that in first case it is better to we choose z ³ 2. The so-called homotopy-Pade technique (see [11]) is employed, which greatly accelerates the convergence.The [r, s] homotopy-Pade approximations of ¢¢ f ( ) 0 (or C f ), and -¢ q ( ) 0 (or Nu x ) in eq. ( 13), according to eqs. ( 27) and (28) are formulated by: respectively.In many cases, the [r, r] homotopy-Pade approximation does not depend upon the auxiliary parameter h.To verify the accuracy of HAM, comparisons of ¢¢ f ( ) 0 and -¢ q ( ) 0 with those reported by Ramachandran et al. [23], Hassanien and Gorla [24], Lok et al. [25], and Ishak et al. [10] (upper branch values) are given in tabs. 1 and 2 for l = 1 and tabs.3 and 4 for l = -1 when f 0 = g = 0 and different values of Pr, respectively. Conclusions Here we have applied the homotopy analysis method which has been proven to be successful in tackling non-linear problems to compute the influence of suction/injection on the mixed convection flows along a vertical cylinder.It is interesting to note that when g = 0 then one recovers the flat plate case [26].The problem reduces to the case of impermeable cylinder for f 0 = 0 [27].The case of arbitrary surface temperature can also be recovered by g = f 0 = 0 [23].It is further revealed that usage of rational base is easier, because it has one auxiliary parameter less than the exponential case (z).It is worth mentioning that in both cases, the dual solutions do not obtain, therefore, the comparison with existing results reveal that our series solutions are valid for all values l.	2017-10-21T12:26:17.879Z	2014-01-01T00:00:00.000	{ "year": 2014, "sha1": "67865edc5599ef45b6d19ebcd1a673a4b8e0baa5", "oa_license": null, "oa_url": "http://www.doiserbia.nb.rs/ft.aspx?id=0354-98361200097E", "oa_status": "GOLD", "pdf_src": "ScienceParseMerged", "pdf_hash": "67865edc5599ef45b6d19ebcd1a673a4b8e0baa5", "s2fieldsofstudy": [ "Engineering" ], "extfieldsofstudy": [ "Mathematics" ] }
55776142	pes2o/s2orc	v3-fos-license	The effects of job satisfaction, employee commitment, workplace friendship and team culture on service recovery performance Article history: Received January 5, 2016 Received in revised format August 2, 2016 Accepted September 27, 2016 Available online September 28, 2016 The existing literature has called for more studies to be conducted on how human resource activities affect service recovery performance. This study therefore ascertains the effects of Job Satisfaction, employee Commitment, Workplace Friendship and Team Culture on Service Recovery Performance. The survey research design was used in this study. The participants were frontline employees from the various service sectors in Ghana. The convenience sampling was used as the sampling technique. A total of 372 responses were used in the final analysis. The scale items were adapted from the existing literature. Confirmatory factor analysis was used to assess the fit of the model. Multiple linear regression was used to test the hypotheses. The findings indicate that Job Satisfaction, Employee Commitment, Workplace Friendship and Team Culture significantly exerts positive influence on Service Recovery Performance of frontline employees. The findings from the study imply that there are several antecedents to Service Recovery Performance. Team Culture, Workplace Commitment, and Employee Commitment can influence Job Satisfaction which in turn will affect Service Recovery Performance resulting in customer satisfaction and retention. © 2016 Growing Science Ltd. All rights reserved. Introduction The contribution of the services sector to the Gross Domestic Product (GPD) of economies of the world cannot be underestimated.The service sector offers employment to a greater number of people globally.In spite of this, the service sector continues to face difficulties in satisfying customers or meeting the expectations of customers.There are always delays in delivery, informational failure, rudeness on the part of staff, response failure etc. (Singhal et al., 2013).Most of the time these failures are caused by frontline staff who are arguably least paid by service organizations.Furthermore, when service failure occurs, the same frontline staff are expected to resolve the problem.The successful resolution of service failures can create a competitive advantage for service firms and lead to customer loyalty (Hinson, 2012).Darby (1999) noted that front line staff influences formation of customers' expectations and experiences to the extent that they are inseparable from the actual service.This suggests that the performance of frontline employees in service delivery and recovering service failure is very essential for service firms.Consequently, the need to investigate the workplace practices that have the potential of motivating service firms' employees to successfully recover service failure is relevant (Karatepe & Vatankhah, 2015).Human resource management outcomes that are performance driven are key to ensure service recovery performance (Davidson et al., 2011).The existing literature has shown that human resource management outcomes such as selective staffing, teamwork, empowerment and training drive employees to meet job requirements and performance (Solnet et al., 2010).Similarly, Ng et al. (2011) noted that job satisfaction drives job performance.Again, Karatepe and Vatankhah (2015) found career opportunities, rewards, selective staffing, empowerment, teamwork, job security and training as the key performance-driven factors affecting service recovery performance in the airline industry.In this view, it can be argued that human resource activities and workplace activities can improve service recovery performance. From the extant literature, it appears studies on the effects of job satisfaction, employee commitment, workplace place friendship and team culture on service recovery performance are very minimal.However, these are key components of human resource activities that influence employee performance (Hinson & Kuada, 2014).Again, it appears that service recovery performance has received limited attention from scholars in developing countries like Ghana (Boateng & Agyemang, 2015).Meanwhile, human resource activities such as team culture and job satisfaction can be context dependent (Katzell et al., 1992).In view of this gap in literature, this current study investigates the effect of employee commitment, job satisfaction, workplace friendship and team culture on service recovery performance of frontline employees in service organizations in Ghana.The rest of the paper is divided into five sections.Section one presents literature review and hypotheses while section two captures the methodology employed.Section three focuses on findings of the study while section four presents discussions and conclusions.The last section focuses on the implications of the study. Service Recovery Performance According to Zemke and Schaaf (1990) the word "recovery" within a service context was derived from British Airway's "putting the customer first" campaign.Service recovery is defined as a firm's attempt to overcome the negative effect of a failure or breakdown.In recent times, Zeithaml et al. (2010 p. 213) have defined service recovery as "actions taken by an organization in response to a service failure".The importance of service recovery is well documented in the marketing literature.Zeithaml and Bitner (2000) believe that while it is not possible for organizations to prevent all problems, they can, at least, try to recover from them.As far back as 1992, Berry and Parasuraman posited that firms should consider failure as an opportunity to satisfy customers better rather than a problem that creates dissatisfied customers.Heskett et al. (1990) referred to service organizations that provide effective response to service failures as "breakthrough" organizations.Such firms, according to Hinson and Kuada (2014) will reap success in the long-term.It is therefore critical that organizations pay attention to service recovery performance of employees.Liao (2007) defined employee service recovery performance (SRP) as "the behaviors in which customer service employees who directly handle customer complaints engage to recover customer satisfaction and loyalty after service failures."He identified three key features of SRP.First, SRP should focus on frontline employees who should be seen as boundary spanners for the service firm.Although boundary spanners may not necessarily be responsible for the service failures, most of the recovery activities fall within the prerogative of frontline employees.Second, SRP refers to specific types of employee behaviors and should be distinguished from its results.SRP refers to the words and actions of the employees in handling customer complaints.On the other hand, results of SRP focuses on how SRP influences customer perceptual, affective and behavioral intentions such as customer satisfaction and repurchase intentions.Third, SRP is a multidimensional construct.Previous studies have investigated various types of service recovery efforts after service failures including apology, empathy, explanation, compensation, repair, replacement, and refund (Sparks & McColl-Kennedy, 2001;Zeithaml et al., 2009;Hinson, 2012;Hinson & Kuada, 2014).This study conceptualizes service recovery performance as the words and behaviors of frontline service employees aimed at resolving customer complaints as a result of service failures with the ultimate aim of recovering the service and thereby ensuring customer satisfaction and retention.Boshoff and Allen (2000) investigated the potential impact of organizational variables on frontline service recovery performance.They hypothesized that job perceptions and employee attitudes have a major impact on frontline staff service recovery performance.Three factors were identified as being directly affected by employee attitudes and which consequently influence a firm's SRP and these include employee commitment to service excellence, an organization's customer service orientation and employee rewards for service excellence.Jenkins et al. (1982) postulated that the lack of employee commitment accounts for poor quality improvement efforts.Employee commitment is believed to be a function of top management commitment.In other words, lack of commitment on the part of management to the objective of service excellence, is directly proportionate to employee commitment to service recovery performance.Branson (1991) identified managerial actions as the most important variable in influencing employee behavior.By implication therefore managers are expected to lead from the front and demonstrate the behavior they require of employees.While it is acknowledged that top management usually places a distance between them and frontline staff, it is still prudent for management to demonstrate strong commitment to service excellence, if employees are to emulate their example.Zeithaml et al. (2009) supported this view and indicated that unless management are fully committed to SRP, the desired objective would not be achieved.However, studies on the relationship between employee commitment and performance indicate mixed results.Boshoff and Allen (2000) believed that the inconsistencies could be attributed to the measurement of different types of commitment.In spite of these disagreements, recent studies have shown that, in a service environment, employee commitment positively influence service quality, as they tend to dedicate more time, more energy and more of their skills and talents to the firm for which they are employed compared to their uncommitted counterparts (Liao, 2007).The study therefore hypothesizes that; H1: A high level of employee commitment towards the service organization will pose a favorable influence on SRP. Job Satisfaction and Service Recovery Performance For several decades, the link between job satisfaction and performance has received a lot of scholarly attentions.As far back as the 1920s, the Hawthorne Studies were among the pacesetters in the empirical investigation of this relationship (Katzell et al. 1992).In spite of the plethora of literature on the subject, the debate is still alive with regards to the nature and causal direction of the relationship (Boshoff & Allen, 2000).The popular proposition (especially among practitioners) held is that satisfied employees outperform the dissatisfied ones.Several empirical studies, however have demonstrated a weak relationship between the two variables (Hess Jr. et al., 2003;McCollough et al., 2000).Evidence also exists pointing to the fact that, rather, employees who do well enjoy their jobs.In service oriented firms, according to Hinson (2012), job satisfaction drives customer-oriented behavior.This finding is supported by the view that providers derive both extrinsically (in the form of financial benefits) and intrinsically (job satisfaction) as a result of satisfaction of the customer.Also, service providers who are in a happy mood are more likely to engage in prosocial behavior.Further, Zeithaml et al. (2009) suggested that people in a happy mood are more likely to offer help and be considerate.The findings are subsequently linked to job satisfaction indicating that satisfied employees are more prone to be in a happy mood and therefore more disposed to behaving in a considerate manner towards others.Roger et al. (1994) earlier pointed out that it is highly unlikely for boundary spanning employees to deliver superlative service when they are unhappy in their work.Their conclusion that there exists a positive relationship between job satisfaction and good customer care has however been proved to be weak, empirically.Bagozzi (1980) argued it is rather easier for service performance to cause job satisfaction as employees compare expected intrinsic and extrinsic rewards with the actual rewards received and forms favorable or unfavorable emotions.On the contrary, he believed that it is more difficult to justify the link between job satisfaction and performance.Despite this assertion, this study takes the conventional view and models SRP as an outcome of job satisfaction.Therefore, this study hypothesizes that; H2: Employee job satisfaction will pose a favorable influence on service recovery performance. Workplace Friendship and Service Recovery Performance The concept of workplace friendship still faces definitional problems.Lin (2010) defines workplace friendship as "the close human relationship that employees who work in an organization build up".Earlier, Song (2006) took inspiration from Hays (1988) and Jehn and Shah (1997) and defined workplace friendship as "a voluntary interdependent relationship in the workplace between two persons to facilitate social-emotional goals of the participants with different types or degrees of intimacy, companionship, mutual assistance and amiable relationship."While acknowledging the plethora of definitions, it is clear that workplace friendship can either hinder or help the formal organizational goals (Nielsen et al., 2000) including service recovery performance. Majority of adult people spend a large part of their lives at work and therefore friendship between coworkers usually foster.Such friendship may be linked to critical work-related performances including organizational commitment, job involvement and job satisfaction (Riordan & Griffeth, 1995).Workplace friendship is found to benefit both the individual worker and the organization (Rawlins, 1992).From the individual standpoint, workplace friendship provides emotional advantages (Kram & Isabella, 1985), improved job performance (Zetlin, 1991), enhanced career development and working environment (Yager, 1997).From the firm's viewpoint, favorable outcome can result from favorable workplace friendship (Morrison, 2004;Jehn & Shah, 1997) including encouraging employee retention (Riordan & Grifeth, 1995), organizational productivity and institutional capacity (Crabtree, 2004;Ellingwood, 2001) In spite of these favorable impact of workplace friendship, Song (2006) identified several reasons why these positive effects have been overlooked.These were listed as the negative effects of workplace including gossip, sexual harassment, fleeting loyalty to the organization, favoritism and organizational politics.Sias et al. (2004) argued that despite the negative impact of workplace friendship, the subject is still important, not just because of its benefits to both the individual and the firm, but more importantly because of its association with job performance.By implication, workplace friendship is expected to correlate with service recovery performance. Many reasons have been adduced for the study of work place friendship (Nielsen et al., 2000).The key reason is the relationship between workplace performance and work-related performance.According to Riordan and Griffeth (1995), there is a positive correlation between friendship, job satisfaction and performance resulting in less employee job turnover.Neilson et al. (2000) also found a close relationship between workplace friendship and organizational performance.Indeed, workplace friendship is considered a significant factor in the study of employee relationship in the workplace (Lin, 2010).It is still debatable whether workplace friendship influences job performance or the other way round as mixed results have been produced (Nielsen et al., 2000;Lin, 2010).This study hypothesizes that: H3: Workplace friendship positively influences service recovery performance. Team Culture and Service Recovery Performance According to Boshoff and Allen (2000), the everyday working environment in which employees function heavily influences their behavior and performance.One of the variables that moderate job performance is the role of knowledge sharing within a team.Team work has been found to impact job performance in a positive way within several environments (Andrews, 1995).While it cannot be denied that superlative service will result periodically from the initiative of one person, it is usually as the result of the activities of several people working together (Berry, 1995) Ultimately, the boundary spanners (frontline staff) are responsible for interfacing with customers, but the full support of the backroom staff is required to ensure a smooth service encounter situation.Zeithaml et al. (2009) explained that customers were only exposed to a limited part of the service delivery process (banking services, for example).While not every employee of the organization deals directly with end users, each task is vitally important in ensuring service excellence (Lovelock & Wrtiz 2004).Each employee serves a particular customer, whether internal customer (a fellow employee) or an external one (end users).Once employees understand that each of them serves a customer and accountable to them, the task they perform becomes more meaningful. An effective team culture glues employees together and motivates team members to contribute for the common good.As Berry (1995) pointed out, the team culture should help support and motivate members to share knowledge and appreciate the contributions of each individual in the achievement of primary objectives.The management and services marketing literature provide evidence that knowledge sharing in teamwork promotes performance (Kuada & Hinson, 2012).The study hypothesized that: H4: Teamwork culture will influence employee SRP through the moderating role of knowledge sharing Overview of the Service Sector and Service Failure in Ghana Ghana's services sector is reported to be the leading contributor to Gross Domestic Product (GDP).According to Ghana Statistical Service Report (2014), the services sector contributed more than 50% of GDP, with manufacturing and agriculture recording a distant 28.57% and 22.0% respectively.Thus, the services sector's contribution to the economy of Ghana totals the contributions of all the other sectors combined.The report indicated that the service sector exceeded its 2012 target of 7.7% by 1.1% points to register an impressive growth rate of 8.8%.The major areas of growth included the Information and Communication sub sector (24.7%),Finance and Insurance subsectors (23.2%), Hotels and Restaurants (13.6%), Business Services (13.5%) and Transport and Storage (11.4%).As Kotler and Keller (2006) indicated, the growth of the services sector serves as a good barometer for the development of a country. Developed economies therefore tend to be more service-oriented compared to their developing country counterparts.On the employment front, Otoo et al. (2009) reported that the services sector in Ghana employs about 20% of the total workforce representing one-fifth of the working population.Thus, the relevance of the services sector cannot be underestimated and therefore serves as focal point for this research.In spite of the importance of services, the extant literature on service failure and recovery suggests that a majority of the studies were conducted in developed countries (Amankwah-Amoah & Debrah, 2010).Nevertheless, the scanty literature on the subject of service failure in Ghana indicates service failure remains critical in the demise of most organizations in Ghana. Research Design and Sampling Design This study used the survey research design.This was to enable the collection of large data and generalize the findings to the entire population.They were selected from the service sectors in Ghana.In all 372 respondents were used in this study.The respondents of the study consisted of the frontline employees from the service sector who consistently interact with customers.The respondents were made of 58.9 females and 41.1% males.Most (49.6%) of the respondents have worked for 1-3 years while 34.2% have worked for 4-6 years.Additionally, 8.8 % have worked for 7-9 years while 7.4 % have worked for 10 years and above.The sample was selected using the convenience sampling technique.This decision was taken because of lack of sample frame.The sampling technique also made it easy for data collection as the participants were easily available. Measures The scale items for this study were all adapted from the extant literature.The scale items measuring Employee Commitment were derived from Chang and Chang (2009) while that of Job Satisfaction (JB) were adapted from Scarpello and Campbell (1983).Also, scale items measuring Workplace Friendship (WF) were borrowed from Neilsen et al. (2000); items measuring Team Culture were adapted from Stashevsky and Koslowsky (2006).SRP was measured using scale items from Boshoff and Allen (2000).Table 1 contains all the items that were used to measure the constructs.All the items were measured on a five-point likert scale; 1=strongly disagree, 2=disagree, 3= neutral, 4= agree and 5= strongly agree.This was to avoid ambiguity in the scale. Reliability and Validity of Constructs The reliability and validity of the constructs were measured using the factor loadings, average variance extracted (AVE), discriminant validity and composite reliability.As shown in Table 1 the factor loadings were all high and there were no cross loadings.Furthermore, the composite validity values for all the constructs were above the recommended threshold of 0.6 (Hair et al., 2006).Additionally, the squared correlations of the constructs were less than the AVEs (see Table 2).This indicates that convergent validity and discriminant validity have been achieved (Bagozzi & Yi, 2012). Results The study used multiple linear regression analysis to predict employees' service recovery performance from job performance, employee commitment, workplace friendship and team culture.This method has been found to be a robust method for predicting an outcome from two or more predictor variables (Hair et al., 2006).All the predictor variables were significant (see Table 3).Employee commitment made the highest (β=.168) contribution to the model followed by team culture (β=.152).Furthermore, workplace friendship made the third highest contribution (β=.145) while job satisfaction made the least (β=.144) (see table 4). Table 3 Model fit Predictors: Employee commitment, Job Satisfaction, Workplace Friendship, Team Culture The results suggest that employee commitment explains 17% of the variance in employee service recovery performance while team culture explains 15%.Similarly, workplace friendship explains 15% of the variance in employee service recovery performance while job satisfaction explains 14% of the variance in employee service recovery performance.The four predictor model accounted for 47% of the variance in service recovery performance, F(6, 366) = 56.359,p< .001,R 2 = .47(see Table 5). Discussions and conclusions The objective of this study was to assess the antecedents of Service Recovery Performance (SRP).The study found employee commitment, job satisfaction, team culture and workplace friendship as essential factors for service recovery performance.Satisfying dissatisfied customers will require commitment from employees; it will require employees who care about the future development of the company and are willing to serve the company.Additionally, employees who are emotionally attached to companies are able to satisfy complaining customers by putting in much effort to ensure that they resolve their problems.Furthermore, the study noted that in order for employees to recover service failure and satisfy dissatisfied customers, the employees themselves must first be satisfied.They tend to do better in service recovery when they are given the freedom to use their own discretion to resolve customers' complaints.Additionally, resolving customers' complaints and turning them to loyal customers will require employees who are satisfied with the working conditions of their companies.This finding is consistent with Boshoff and Allen (2000) who noted that employee commitment has a significant effect on service recovery performance.Employees who found themselves in an enabling working environment tend to perform; they give their best because they are assured of steady employment (Karatepe & Vatankhah, 2015).Furthermore, the findings of this study corroborates that of Andrews' (1995) observation that team culture is a key determinant of service recovery performance.This is probably because when employees work in teams, there are always people around to discuss and resolve customers' problems.Again team culture has effect on service recovery performance probably because members of the team have a sense of participation and possess a fine spirit which encourages members to help each other to resolve problems. In addition, the study found workplace friendship as determinant of service recovery performance. Employees having the opportunity to get to know each other, encourages free flow of information.This enables employees to acquire knowledge from colleagues which help them to deal with customer complaints (Yager, 1997).However, workplace friendship might not necessarily lead to service recovery performance; it should be backed by the willingness of the "friends" to share their professional and personal experiences.Again, they should be willing to share new methodology for performing task and valuable knowledge with others.This can be achieved by promoting socialization among employees (Neilsen et al., 2000).Employees' socialization can influence them to express their opinions and ideas towards during company's meetings.Finally, this study found that job satisfaction has a significant and positive effect on service recovery performance.This is not surprising as Hinson (2012) noted that job satisfaction drives customer-oriented behavior.This suggests that for employees to recover service failure successfully, the right working conditions must be created and communicated to them.This will involve creating a cordial relationship among employees and developing policies that offer steady employment to employees.An important finding from this study is establishment of causality between the other factors and Service Recovery Performance which was not clearly established in other studies (Nielson et al., 2000;Boshoff & Allen, 2000) 6. Implications of the study The implications of these findings are that, first, employees must be made to feel that they have a promising future in the organization; they can attain a position they are aspiring for.Employees must be satisfied before they can also satisfy the customers.The cliché that satisfied customers are more likely to be retained applies to employees as well.In other words, satisfied employees are more likely to be retained which positively correlates with customer retention.This will mean that organizations need to practice internal marketing.Organization need to identify the needs and wants of employees and satisfy them. Second, organizations must encourage teamwork or team culture by giving tasks that demand efforts from different employees within the organization.Team culture can be promoted by rewarding teams instead of individuals.The general notion is that employees who get along well tend to work well together.Firms must therefore purposely select team members on the basis of their compatibility and possessing the necessary traits required in team relations. Third, workplace friendship should be encouraged in organizations.Employees should be made to see each other as a true friend.However, workplace friendship for its own sake will not necessarily promote service recovery performance.Organizations should gear workplace friendship towards meeting organizational goals.Organizing social events and creating an environment where families of employees regularly meet can promote friendship among them.Practical examples include the establishment of schools and other social amenities for employees and their families.Indeed, workplace friendship can help create effective teams and promote job satisfaction resulting in positive service recovery performance. One of the limitations of the study is that it employed a convenience sampling technique which makes it difficult for the findings to be generalized.Generalization of the findings must therefore be done cautiously.Future studies may employ a probability sampling technique to address this limitation.Again, future studies should explore other human resource activities that improves service recovery performance.Furthermore, future studies may assess service recovery performance from customers' perspectives since in reality the employee are likely to be bias in assess their performance. I care about the future development of this company 0.70 I feel I have a promising future if I stay with this company 0.74 I will pass on my working experience to new staff 0.69 I am emotionally attached to this company 0.69 Job Satisfaction CR = .87;AVE = .52I am satisfied with the freedom to use my own judgment 0.63 I am satisfied to try my own methods of doing the job 0.60 I am satisfied with the way my co-workers get along with each other 0.69 I am satisfied with the way my job provides for steady employment 0.76 I am satisfied with the way the company policies are put into practice.0.74 I am satisfied with the working conditions of my company 0.71 Workplace Friendship CR = .79;AVE = .50I have the opportunity to get to know my co-workers 0.67 I am able to work with my co-workers to collectively solve problems 0In my organization, there is always someone to address work problems 0.69 My team supports knowledge and technical information sharing 0.79 This company coordinates teamwork through team leaders 0.74 This company's employees are flexible about meeting to discuss problems and coordinate teamwork 0.76 This company teams possess a fine spirit 0.73 Members of this team have a sense of participation 0.73 Service Recovery Performance CR = .85;AVE = .53Considering all the things I do, I handle dissatisfied customers quite well 0.70 I don't mind dealing with complaining customers 0.73 No customer I deal with leaves with problems unresolved 0.75 Satisfying complaining customers is a great thrill to me 0.74 Complaining customers I have dealt with in the past are among today's most loyal customers 0.70 Table 2 Squared correlations and AVEs Table 5 Coefficients a	2018-12-05T19:45:10.590Z	2016-11-01T00:00:00.000	{ "year": 2016, "sha1": "3537b11d050b0fb3b607d47d1ea08bddf9a47423", "oa_license": "CCBY", "oa_url": "https://doi.org/10.5267/j.msl.2016.9.002", "oa_status": "GOLD", "pdf_src": "MergedPDFExtraction", "pdf_hash": "3537b11d050b0fb3b607d47d1ea08bddf9a47423", "s2fieldsofstudy": [ "Business" ], "extfieldsofstudy": [ "Psychology" ] }
252896172	pes2o/s2orc	v3-fos-license	Unambiguous identification of α-Gal epitopes in intact monoclonal antibodies by NMR spectroscopy ABSTRACT The α-Gal epitope consisting of the terminal trisaccharide Galα1,3Galβ1,4GlcNAc exposed on cell or protein surfaces can cause severe immune reactions, such as hypersensitivity reactions, in humans. This epitope is also called the xenotransplantation epitope because it is one of the main reasons for the rejection of non-human organ transplants by the human innate immune response. Recombinant therapeutic proteins expressed in murine cell lines may contain α-Gal epitopes, and therefore their absence or presence needs to be tightly monitored to minimize any undesired adverse effects. The analytical identification of α-Gal epitopes in glycoproteins using the common standard techniques based on liquid chromatography and mass spectrometry is challenging, mainly due to the isobaricity of hexose stereoisomers. Here, we present a straightforward NMR approach to detect the presence of α-Gal in biotherapeutics based on a quick screen with sensitive 1H-1H TOCSY spectra followed by a confirmation using 1H-13C HSQC spectra. Abbreviations: α-Gal: α1,3-linked galactose; AGC: automatic gain control; CHO: Chinese hamster ovary; CE: capillary electrophoreses coupled to mass spectrometry; COSY: correlation spectroscopy; DSS: 2,2-dimethyl-2-silapentane-5-sulfonate; DTT: dithiothreitol; GlcNAc: N-acetyl glusomamine; HCD: higher-energy collisional dissociation; HMBC: heteronuclear multiple-bond correlation; HPLC: high-performance liquid chromatography; HSQC: heteronuclear single-quantum corre; LacNAc: N-acetyl lactosamine; mAb: monoclonal antibody; MS: mass spectrometry; NMR: nuclear magnetic resonance; NOESY: 2D) nuclear Overhauser spectroscopy; PEG: polyethylenglycol; pH: observed pH meter reading without correction for isotope effects; PTM: post-translational modification; TCEP: tris(2-carboxyethyl) phosphine hydrochloride; TOCSY: total correlation spectroscopy; xCGE-LIF: multiplex capillary gel electrophoresis with laser-induced fluorescence detection. Introduction Comprising ~1% of all immunoglobulins, human antibodies that recognize the α-Gal epitope consisting of the terminal carbohydrate Galα1,3Galβ1,4GlcNAc 1,2 ( Figure 1) are the most abundant. The α-Gal epitope, sometimes also called the Galili epitope, is abundantly expressed on glycoconjugates of most mammals. However, α-Gal is absent in glycoconjugates of humans, apes and Old World monkeys, because they lack an active glycosyltransferase, the UDP-Gal: β-galactosyl α1,3galactosyltransferase, EC 2.4.1.151 (α1,3GT). This loss was estimated to have occurred less than 28 million years ago in common ancestors. 1 Instead, these latter mammals produce large quantities of anti-Gal antibodies that specifically recognize the α-Gal epitope. It is hypothesized that these antibodies are part of an innate immune response against zoonotic viruses and protozoa, e.g., Trypanosoma, Plasmodium or Leishmania. 3 In humans, these anti-Gal antibodies are responsible for the rejection of non-human organ transplants containing the α-Gal epitope, e.g., organs grown in pigs, which induce complement destruction of the organ within minutes to hours. 1 The absence of immunogenic epitopes, which can cause undesired immune reactions like hypersensitivity reactions, 4 is highly preferred for therapeutic proteins. However, the presence of immunogenic epitopes cannot be totally avoided in all cases, and therefore the exact nature and quantity of these epitopes must be tightly monitored. Many monoclonal antibodies (mAbs) are produced in non-primate mammalian cell lines, which contain the machinery to express α-Gal epitopes and do typically express them. 5 In particular, antibodies produced by mouse cell lines such as NS0 or SP2/0 often contain α-Gal epitopes. Despite reports that Chinese hamster ovary (CHO) cell lines typically do not express the α1,3GT, and thus lack α-Gal epitopes, 1 certain CHO cells produce α-Gal epitopes under certain conditions. 6,7 In addition to the immunogenicity, the in vivo half-life of antibodies containing α-Gal epitopes in the human body is significantly reduced. 8 Confirming the absence or presence of α-Gal epitopes in therapeutic proteins by the classical proteomics approach of mass spectrometry (MS) coupled to high-performance liquid chromatography (HPLC) is analytically challenging because the difference between an N-glycan containing Galα1,3Gal and one lacking it lies in the mass difference of one or several hexose moieties (n × 162 Da), which may be galactose, mannose or glucose. More importantly, the key characteristic for immunogenicity is the α1,3 linkage between the terminal galactose and the underlying galactose. 9 One possibility to confirm Chemical and symbolic structures of the α-Gal epitope and the investigated trisaccharides. (a) chemical structure of the α-Gal epitope Galα1,3Galβ1,4GlcNAc that can be attached to any underlying glycan or glycolipid. The symbol presentation of the same trisaccharide is shown below. (b) structure of the investigated α-Gal trisaccharide Galα1,3Galβ1,4GlcNAc with a free reducing end containing a mixture of α-and β-anomers of the first GlcNAc moiety. (c) structure of the studied xeno antigen type 1 trisaccharide Galα1,3Galβ1,3GlcNAc with a free reducing end containing a mixture of α-and β-anomers. Figure 2. Chemical shift assignments of the α-Gal antigen trisaccharide (Galα1,3Galβ1,4GlcNAc) and the xeno antigen type 1 trisaccharide (Galα1,3Galβ1,3GlcNAc) measured either in D 2 O (blue) or in 7 M urea-d 4 /D 2 O at 298 K and pH 7.4 (red). (a) Overlay of two 1 H-13 C HSQC spectra of Galα1,3Galβ1,4GlcNAc measured either in D 2 O (blue, with assignment) or in 7 M urea-d 4 /D 2 O pH* 7.4 (red). The spectra were recorded with multiplicity-editing leading to opposite sign for CH 2 groups (green and purple). The spectra were measured with 32 transients, a recycle delay of 1.5 sec and 2048 × 512 points resulting in a total measurement time of 7.5 hours. A schematic presentation of the trisaccharide structure in its α and β anomeric form is shown on the top. At the bottom the anomeric region of the 1 H-13 C HSQC spectrum is shown. Few 1 H-13 C three-bond correlations are visible as well (without label) in particular between H2-C3 and H3-C2 of Gal2 and Galα3. (b) Overlay of two 1 H-13 C HSQC spectra of related xeno antigen type 1 trisaccharide measured either in D 2 O (blue, with assignment) or in 7 M urea-d 4 /D 2 O pH* 7.4 (red). The spectra were recorded with the same parameters as the α-Gal antigen trisaccharide. Asterisks indicate signals of a minor impurity. Its identity and the chemical shift assignments are provided in the Supporting Information ( Figure S2 and Table S2). this is the analysis by porous graphitized carbon-liquid chromatography coupled to tandem MS (PGC-MS 2 ). 10 However, the sample preparation and data analysis are very elaborate. Treatment with α1,3Gal-specific glycosidases followed by MS is another possibility, but this method is indirect. Orthogonal methods are desirable for cross-validation of the current analytical methods for α-Gal detection as a critical quality attribute of biopharmaceuticals. Nuclear magnetic resonance (NMR) spectroscopy is a promising tool that has been of limited use in the context of biotherapeutics due to the large molecular sizes, in particular of mAbs, mAbconjugates or Fc-chimera. This obstacle has been overcome by investigating fully denatured proteins, whose backbone tumbles similarly to a peptide or a small protein, resulting in sufficiently sharp NMR signals of all entities in the protein including post-translational modifications (PTMs) e.g., glycans. 11 1 H-13 C chemical shift correlations that are unique for a certain PTM, such as a certain glyco-epitope, are suitable for the unambiguous detection, identification and quantification of a modification using a single 2D NMR experiment. [11][12][13][14][15][16][17] With the help of a few oligosaccharides, it was possible to assign the chemical shift correlations of the major N-glycans in a variety of mAbs and other glycoproteins, allowing also quantification. [15][16][17] Sequence-specific assignment is lost when denaturing conditions are applied, but this is considered acceptable for confirming the presence and exact nature of a certain PTM. Here, we aimed at finding unique NMR signals that would identify α-Gal epitopes in therapeutic proteins and at confirming the previously reported presence of α-Gal in the biotherapeutic mAb cetuximab, which was approved by the US Food and Drug Administration in 2004 for the treatment of head and neck cancer and metastatic colorectal cancer. 18 Cetuximab is produced in the mouse myeloma NS0 cell line and was reported to contain α-Gal. [19][20][21][22] The presence of α-Gal can cause severe side effects in some patients. 21,23,24 A biobetter of cetuximab expressed in CHO cells was better tolerated by patients. 25 Nevertheless, even CHO cells have been reported to express α-Gal epitopes under certain conditions. 6 As a proof of concept, we analyzed commercially available oligosaccharides containing Galα1,3Gal to find characteristic chemical shift correlations that are unique for Galα1,3Gal and screened a variety of biopharmaceuticals under denaturing conditions for those characteristic signals. The NMR signature of α-Gal epitopes To explore the NMR signature of the α-Gal epitope and in particular characteristic NMR correlations in denatured glycoproteins, we investigated the two related trisaccharides Galα1,3Galβ1,4GlcNAc ( Figure 1b) and Galα1,3Galβ1,3GlcNAc ( Figure 1c) using 2D NMR spectroscopy. The first trisaccharide, also called xenoantigen type 2 or linear blood group B type 2 trisaccharide, represents the α-Gal epitope that is recognized by the anti-Gal antibodies. The second trisaccharide is a rare epitope called xenoantigen type 1 or linear blood group type 1 trisaccharide that is structurally similar, but differs at the linkage between Gal and GlcNAc. Complete 1 H and 13 C chemical shift assignments for both carbohydrates, including their α-and β-anomeric form at the reducing end ( Figure 2 and Table 1), were obtained using standard 1 H-1 H and 1 H-13 C correlation spectra both in plain D 2 O and 7 M urea-d 4 in D 2 O (pH* 7.4, uncorrected). Denaturing conditions were used for optimal comparison, with spectra of denatured glycoproteins measured under the same conditions. In contrast to other biomolecules, carbohydrates are fairly flexible and, with few exceptions, do not form stable secondary structures. [26][27][28] Therefore, the chemical shifts of uncharged oligosaccharides should not depend very much on the solvent conditions. Neither should there be large differences between small oligosaccharides and glycans attached to denatured proteins. Surprisingly, we observed small, but significant deviations of the resonances between the two conditions for both oligosaccharides ( Figure 2). Many resonances showed deviations between half to three line-widths, either in the 1 H, 13 C or both dimensions. The largest 1 H deviations were Table 1. Observed chemical shifts of the α-Gal epitope trisaccharide (Galα1,3Galβ1,4GlcNAc) dissolved in 7 M urea-d 4 pH* 7.4 (indicated by U) and in plain D 2 O (indicated by D). The data were measured at 298 K and are referenced to internal DSS. For comparison, previously reported values of Galα1,3Galβ1,4GlcNAcβ-linker 29 The chemical shifts depend on the anomeric form at the GlcNAc at the reducing end. (α) stands for GlcNAcα at the reducing end and (β) stands for GlcNAcβ at the reducing end. b The reported values showed an offset. For better comparison a correction of +0.015 ppm was added to the 1 H values from the publication and +1.5 ppm was added to the 13 C values c The reported trisaccharide had a linker (O-(CH2) 3 -NH 2 ) at the reducing end observed for the two H2 resonances of the terminal Gal3 and the connecting Gal2. For the use of unique chemical shift correlations as evidence for the presence of a certain glycoepitope in a glycoprotein, it is important that the correlations match exactly. Therefore, the measurements in 7 M urea-d 4 in D 2 O (pH* 7.4) are crucial for the two α-Gal trisaccharides in order to obtain good matches to denatured glycoproteins. Although the signals measured in D 2 O are all very close to the signals detected in 7 M urea, which allows hypothesizing about the presence of a certain glyco-epitope, an exact match is needed to obtain reliable information. The chemical shift changes seem even more severe in the 2D 1 H-1 H TOCSY spectrum ( Supplementary Fig. S1). One characteristic correlation for the α-Gal epitope is the C1-H1 signal of the terminal Galα1,3 (Gal3) moiety at 98.3 ppm/5.16 ppm (Figure 2), which is quite isolated. The chemical shift correlations of this anomeric 1 H resonance of the terminal Gal are also visible and isolated in the 2D 1 H-1 H TOCSY spectrum ( Supplementary Fig. S1). The 13 C chemical shift of C3 of the connected second Gal residue at ~80 ppm ( Figure 2), quite downfield shifted due to the substitution at O3, is unique too. All 1 H-13 C correlations of the terminal Galα1,3 moiety are basically identical between the two trisaccharides. The signals of the connecting Gal2 (Galβ1,4 or Galβ1,3) occur at similar positions but are not identical as expected. Furthermore, a minor impurity of Galβ1,3Galβ1,3GlcNAc was detected in the sample of Galβα1,3Galβ1,3GlcNAc at an abundance of ~15% (Supplementary Figure S2 and Table S2). The assignment is reported in the Supplementary Table S1. Detection of α-Gal in therapeutic monoclonal antibodies In order to recognize a certain glyco-epitope by NMR spectroscopy in denatured proteins, chemical shift correlations unique for this particular epitope must be identified. Any overlap with other common protein and glycan signals must be avoided. 1 H-13 C HSQC spectra of denatured mAbs, including rituximab, adalimumab, denosumab, trastuzumab and cetuximab, and other proteins were inspected for the typical chemical shift correlations of Galα1,3. All mAbs except cetuximab lacked Galα1,3 signals. However, the 1 H-13 C HSQC spectrum of cetuximab contained an anomeric signal (98.3 ppm/5.16 ppm) matching that of Galα1,3 ( Figure 3). Other 1 H-13 C correlations confirmed this finding. In particular, the isolated signal of C5-H5 at 73.8/4.20 ppm of the terminal Galα1,3 is characteristic (red, Figure 3a,b), together with the signals of C3-H3, C4-H4 and C5-H5 (80.2/3.78 ppm, 67.7/4.15 ppm and 78.0/3.72 ppm) correlations of the central Gal2 of Galα1,3Galβ1,4GlcNAc (magenta, Figure 3a). These four signals are already very indicative of the Galili epitope. The signals of the connecting Gal2 fit much better to the spectra of the α-Gal epitope (Figure 3a) than to the xeno antigen type 1 epitope (Figure 3b). This agrees with the known structures of N-glycans that contain Galα1,3Galβ1,4GlcNAc ( Figure 4a). The NMR spectra confirm the presence of the highest abundant glycans previously reported for cetuximab, 19,20 as shown in Figure 4a. Galα1,3 was reported to be part of various glycan types, with A2Ga2F and A2Sg1Ga1F as the most abundant ones, 19 which agrees with the NMR signals (Figure 4b). Other abundant glycans of the Fc part, such as A2G0F, A2G1F and A2G2F with terminal GlcNAc5 and GlcNAc5ʹ (attached to Man4/4ʹ) or Gal6 (attached to GlcNAc5/5ʹ), were confirmed by NMR spectroscopy as well. Isolated signals of terminal Gal (Gal6t), for example, the C2-H2 correlation at 73.7/3.59 ppm and of terminal GlcNAc (GlcNAct) in particular the C2-H2 and C5-H5 correlations at 58.1/3.73 ppm and 78.7/3.46 ppm, respectively, are clearly present (Figure 4b left). Smaller signals of high mannose glycans are also visible (not labeled in Figure 4). A few small signals of terminal Man4 and Man4ʹ are visible as well (Figure 4b bottom left, labeled Man4t and Man4ʹt). The presence of the α-Gal epitope in cetuximab was also confirmed by 1 H-1 H correlations in a 2D TOCSY experiment that show the same pattern as in a comparable spectrum of the α-Gal trisaccharide ( Figure 5 a,b, red label). Although the H1 resonance partially overlaps with H1 of Man4, strong characteristic H1-H2, H1-H3 and H1-H4 correlations of Galα1,3 are visible (Figure 5b). Even these four 1 H chemical shifts seem characteristic because a search in the carbohydrate chemical shift database of Glycosciences.de using GlycoNMRSearch 30 revealed as top hit a Galα1,3Gal-containing glycan ( Supplementary Fig. S3a). Although it should get even clearer when adding the 13 C chemical shift of C1, the coverage of the database is lower for 13 C, resulting in three top hits containing αlinked Gal, but not Galα1,3Gal-containing glycans (Fig. S3b). The mAbs rituximab (Rituxan, MabThera®), adalimumab (Humira®), denosumab (Prolia®) and trastuzumab (Herceptin®) did not show the characteristic signals of α-Gal in 2D TOCSY spectra ( Figure 5 c-f), and neither does the Fc-fusion biotherapeutic etanercept (Enbrel®) nor fetuin from fetal bovine serum ( Figure 5 g,h). 2D 1 H-1 H TOCSY correlations have the tremendous advantage that they can be recorded within a short time (~1 hour) because the sensitivity of 1 H is high and the natural abundance is nearly 100%, in contrast to experiments involving 13 C with a natural abundance of only 1.1%. A semiquantitative analysis of the terminal glycan moieties based on the cross-peak volumes in the 1 H-13 C HSQC spectrum of cetuximab is presented in Figure 6a. The individual signals used for each terminal moiety are indicated in Fig. S4. The approach is semiquantitative because the applied recycle delay of 1.4 sec was not long enough for fully quantitative measurements, but a compromise to obtain spectra with a sufficient sensitivity in a feasible time. A comparable analysis of the MS data of the same batch is summarized in Figure 6b. The detailed HPLC-MS data are provided in the Supporting Information as an Excel table. Both methods found terminal GlcNAc moieties as the most abundant terminal monosaccharides, 39.1% by NMR and 30.0% by MS. The amount of α-Gal epitopes reported by terminal Galα1,3, with 21.4% detected by NMR spectroscopy and 29.5% by MS, is in a similar range, but deviates slightly. The amount of terminal Man is comparable with both methods as well. There is an apparent disagreement between both approaches for terminal Gal (β1,3-linked to GlcNAc) and terminal sialic acids. NMR did not detect any sialic acids. The sialic acids were lost during sample preparation, as buffer exchange to ddH 2 O leads to a significant drop in pH, which causes sialic acid to be cleaved and subsequently lost during dialysis. Interestingly, the amount of Neu5Ac +Neu5Gc detected with MS added on top of the terminal Gal amount results approximately in the amount of Gal detected by NMR. This agrees with the assumption that the terminal sialic acid was removed in the NMR sample, exposing then terminal Gal. The quantity of terminal α-Gal epitopes and terminal GlcNAcs is roughly the same as MS, but with NMR spectroscopy the signals of the terminal GlcNAc are clearly stronger than the signals of the α-Gal epitope (Fig. S4), suggesting a bias in the quantification by MS. Comparable histograms of terminal sugars were determined for middle-up glycoprofiling, 19 CEX-MS and CE-MS, 20 as well as for xCGE-LIF-based glycoprofiling of cleaved fluorescently labeled N-glycans (Figure 6c-f). All these methods critically depend on the correct assignment of observed species to individual N-glycan structures. Typically, a certain cutoff is applied and lower abundant glycan species are ignored. However, many low-abundance glycans, especially highly branched ones, can still contribute significantly to such histograms of terminal sugar moieties. The agreement among the different methods is still improvable, illustrating the importance of orthogonal methods for cross-validation. How sensitive is the 1 H-1 H TOCSY experiment for detecting α-Gal? It was not entirely unexpected that α-Gal could be easily detected in cetuximab, an NS0-expressed therapeutic mAb in which the N-glycosylation within the Fd contains 78% α-Gal. 19 However, the use of murine cell lines to express antibody therapeutics is now rare, so that αGal does not seem to be an issue. CHO cell lines, currently the most popular expression system for mAbs, were reported to lack expression of α1,3GT and thus do normally not produce α-Gal epitopes. 31 Is that really correct? There are few reports that under certain conditions, CHO cells can produce α-Gal epitopes. 6 Abatacept (ORENCIA®) was reported to contain very low amounts of α-Gal, namely 0.37 nmol/mg protein, 6 corresponding to an average of 3.4% α-Gal epitope per Fc-fusion dimer. Detecting α-Gal epitopes at such a low concentration is a challenge with conventional techniques because, in the case of abatacept, this corresponds to only ~0.2% of the mass of total N-linked glycans. 32 The distribution of those α-Gal epitopes over several N-glycan forms is the real challenge because N-linked glycans with abundance below 0.1% could often not be properly analyzed. 32 Here, we present a 1 H-1 H TOCSY spectrum of abatacept under denaturing conditions (Figure 7a). The characteristic pattern of α-Gal correlations is visible. The H4-H1 correlation is isolated and clearly visible. The H2-H1 and H3-H1 correlations are partially overlapping with other signals of the H1 resonance of Man4. All three α-Gal correlations match exactly the corresponding signals of cetuximab (Figure 7b). Slices through the 2D spectrum illustrate the signal to noise of the α-Gal signals (Figure 7a). The signal intensity of the αGal H4-H1 correlation is 83× smaller than the signal of the H2-H1 correlation of Man4. In the case of cetuximab, the intensity ratio between αGal H4-H1 and Man4 H2-H1 was 1.2. In a rough estimation, the amount of α-Gal in abatacept is ~100× smaller than in cetuximab (per N-glycan). Based on the reported quantification for cetuximab containing 2.62 μmol α-Gal/μmol IgG (see Materials and Methods), we calculated an amount of ~26 nmol α-Gal/μmol abatacept (dimer). This rough estimate is in good agreement with Bosques et al. reporting ~34 nmol α-Gal/μmol abatacept. 6 The 2D spectrum was recorded on an ordinary 600 MHz spectrometer with a four-channel room temperature probe within 24 hours, but a spectrometer with a cryoprobe that typically gains at least a factor of 2.8 in sensitivity would be able to measure a comparable spectrum in 3 hours. 33 Discussion We aimed at establishing an NMR spectroscopy approach as an orthogonal method to HPLC-MS techniques. Although Galα1,3Gal-containing oligosaccharides and glycopeptides have been investigated by NMR spectroscopy before, 29,[34][35][36] the chemical shift assignments were not measured under denaturing conditions and were thus not fully comparable to the spectra of denatured glycoproteins. The published NMR data Figure 7. The CHO cell-expressed therapeutic mAb abatacept contains α-Gal epitopes. A 2D 1 H-1 H TOCSY spectrum of abatacept under denaturing conditions shows three characteristic signals of α-Gal (red arrows). The spectrum was measured using a mixing time of 80 ms, 64 scans, a recycle delay of 1.5 sec, 2048 × 700 points and spectral widths of 12.9 ppm × 8.3 ppm with a duration of nearly 23 hours 45 min. To illustrate the signal-to-noise ratio a horizontal and a vertical slice are shown on the top and left taken at the dashed red lines in the spectrum. On the right the same spectrum is shown on top of a comparable spectrum of cetuximab (red contours), which illustrates that the α-Gal signals of abatacept match exactly the positions in the spectrum of cetuximab. The H2/H1 correlation of Man4 fits as well, which is then used as reference for estimating the amount of α-Gal in both biotherapeutics. Please note that the vertical red dashed line does not go through the maximum of the Man4 H2/H1 correlation, therefore the real signal intensity of this Man4 signal is larger than indicated in the 1D slice. Signals of maltose glycation products are indicated by a dark magenta MG label and magenta dashed lines. Signals of free maltose likely released due to reversible glycation are indicated by an olive green M label and olive green dashed lines. were sometimes contradictory or incomplete, and characteristic or unique NMR correlations in the context of biotherapeutics had not been investigated so far. HPLC-MS-and CE-MS-based methods are very sensitive and can reveal the distribution of N-glycosylation of intact mAbs, 20 mAb subunits, 37 glycopeptides 38 or the released glycan. 10,39 However, of these methods, only PGC-MS 2 can confirm the type of linkage and the exact identity of the terminal hexoses, and therefore unambiguously identify the Galα1,3Gal epitope. Nevertheless, one of the central questions for the safety of biotherapeutics is as follows: Are there relevant amounts of immunogenic epitopes detectable? In this context, the exact position or the nature of the underlying glycan is less relevant. In the case of cetuximab, the large amount of unique NMR correlations of α-Gal epitopes present in 1 H-13 C and 1 H-1 H spectra can robustly and unambiguously identify this epitope. As NMR spectroscopy is not as sensitive as MS, the presence of all those unique correlations indicates a significant amount of the immunogenic epitope. NMR spectroscopy is per se a quantitative method; therefore, exact quantification is achievable using appropriate experimental parameters like sufficiently long recycle delays. 13 Our semiquantitative analysis of the terminal monosaccharides by NMR spectroscopy and MS revealed a striking agreement for the α-Gal abundance of ~25% of all terminal glycan moieties. With the α-Gal epitope, we have further emphasized that NMR spectroscopy is suited to unambiguously confirm the presence of a certain glyco-epitope with the exact linkages and stereochemistry within the context of a glycoprotein. In addition, there is a good chance to identify even unknown glyco-epitopes with NMR spectroscopy. In contrast to 1 H-13 C HSQC spectra, 2D 1 H-1 H TOCSY spectra have the advantage that they are quite sensitive and can be measured typically in less than an hour because only sensitive 1 H nuclei at nearly 100% abundance are involved. 1 H-13 C correlations involving 13 C nuclei that are only present at 1.1% natural abundance are by a factor of 90 less sensitive. 2D 1 H-1 H TOCSY spectra are suited for fast screening. Nevertheless, the 1 H-13 C HSQC spectrum shows a higher dispersion of signals and is thus more reliable. In addition, it is more quantitative than 1 H-1 H correlations due to the more uniform one-bond coupling constant 1 J CH . For α-Gal in mAbs, extracting relative abundances might be sufficient, for example the ratio between terminal Galα1,3 and terminal Galβ1,3 moieties. Although accurate quantification is possible with 1 H-13 C HSQC spectra of mAbs, as was reported for the quantification of glycan species 16,17 and of methionine sulfoxide, 13 it required long recycle delays and long measurement times of typically 1-3 days. If semiquantitative results are sufficient, e.g., distinguishing an α-Gal content of <1%, ~50%, ~100%, ~200% per chain, standard parameters for 1 H-13 C HSQC spectra can be used, resulting in typical measurement times of 10-40 hours [12][13][14][15] depending on the sensitivity of the NMR spectrometer and the sample concentration. The 1 H-1 H TOCSY spectrum recorded on an ordinary 600 MHz spectrometer was so sensitive that even very low amounts of α-Gal epitopes could be detected in abatacept, an Fc-fusion therapeutic expressed in CHO cells. We estimated the α-Gal content of ~26 nmol α-Gal/μmol abatacept (dimer), which is in good agreement with a previous report of ~34 nmol α-Gal/μmol abatacept. 6 This corresponds to an α-Gal epitope for 3.4% of the Fc-fusion dimers or ~0.2% of the mass of total N-linked glycans. 32 Although our intention was always to use NMR spectroscopy as a complementary tool to MS, it did not escape our attention that in this particular case of α-Gal epitopes a 2D 1 H-1 H TOCSY recorded in <1 hour for detecting the presence or absence of α-Gal in a protein preparation might be quite competitive to MS. Oligosaccharides For a D 2 O sample, 2.5 mg of the trisaccharide Galα1,3Galβ1,3GlcNAc (Elicityl, catalog number GLY074-1) was dissolved in 500 μL D 2 O resulting in an uncorrected pH (pH) of 6.7. For measurements under denaturing conditions, the same amount was dissolved in 500 μL D 2 O followed by lyophilization to remove exchangeable protons. The solid residue was dissolved in a 7 M urea-d 4 (98 atom%D, ARMAR Chemicals) solution in D 2 O followed by adjusting the pH to 7.4 (uncorrected pH) using deuterated hydrochloric acid (DCl) in D 2 O (ARMAR Chemicals). For the trisaccharide Galα1,3Galβ1,4GlcNAc (Carbosynth, catalog number OL06495), 1 mg was dissolved in 500 μL D 2 O and another 1 mg was used for the sample with 7 M urea using the same procedure as described above. Etanercept For etanercept (Enbrel® Amgen/Pfizer; formerly distributed by Immunex/Wyeth; Batch Nr.: 1015854, Exp Date.: 11/11), 1 mL formulation (50 mg) was dialyzed against ddH 2 O (pH 7.0) using a SpectraPor 3.5 kDa cutoff membrane for 5 hours and was then lyophilized. The lyophilized powder was dissolved in 1.5 mL 7 M urea-d 4 (98 atom%D, ARMAR Chemicals) solution in D 2 O leading to a concentration of 33 mg/mL. For reducing the disulfide bonds, ~4 mg TCEP was added and the solution was placed at 60°C for 15 min. Aliquots of 500 μL were used for NMR measurements. The pH* was adjusted to 7.4 using NaOD (Armar Chemicals). Adalimumab and rituximab The NMR sample preparation of adalimumab (Humira®, AbbVie) and rituximab (MabThera, Roche) was described previously by Hinterholzer et al. 12 Fetuin For Fetuin from fetal bovine serum (Sigma F3385), 32 mg was dissolved in 650 μL of a 7 M urea-d 4 (98 atom%D, ARMAR Chemicals) solution in D 2 O. Approximately 2 mg TCEP was added, and after an incubation for 15 min at 65°C the pH was adjusted to 7.4 using NaOD (Armar Chemicals). Abatacept Abatacept (ORENCIA®, Bristol Myers Squibb; Lot. OE61132, exp. 08/2012) 60 mg in 2.4 mL formulation buffer was dialyzed twice against 4 L ddH 2 O using a SpectraPor 3.5 kDa cutoff membrane overnight. The resulting sample was lyophilized and dissolved in 650 μL of a 7 M urea-d 4 (98 atom%D, ARMAR Chemicals) solution in D 2 O. Deuterated DTT-d 10 (Cambridge Isotope Laboratories) was added to a concentration of 15 mM, and after an incubation for 15 min at 60°C, the pH was adjusted to 7.4 using NaOD (Armar Chemicals). Mass spectrometry Cetuximab (Erbitux®, Lot. 208480, exp. 09/2019, Merck KgaA) was rebuffered to 150 mmol.L −1 ammonium acetate using 50 kDa cutoff filters (Sartorius Vivaspin 500 from Sigma-Aldrich). Afterward, the samples were denatured and reduced to 3 mol.L −1 guanidine hydrochloride 10 mmol.L −1 tris (2-carboxyethyl)phosphine in 150 mmol.L −1 ammonium acetate at 70°C for 15 minutes. Followed by alkylation of the thiol groups with the addition of a final concentration of 20 mmol.L −1 iodoacetamide and incubation at 22°C for 30 minutes. Next, the samples were again rebuffered to 150 mmol.L −1 ammonium acetate using 3 kDa cutoff filters (Amicon Ultra from Sigma-Aldrich), and lastly digested by addition of 1:20 (w/w) trypsin (porcine pancreas, from Promega, catalog number V5111) and incubation at 37°C for 3 hours. Chromatographic separation of 100 ng sample was carried out on a Thermo Scientific™ UltiMate™ 3000 RSLCnano System using a Thermo Scientific™ Acclaim™ PepMap™ 100 C18 reversed phase HPLC column (500 × 0.075 mm i.d., d p 3 µm). For the separation 0.1% aqueous formic acid (solvent A) and 0.1% formic acid in acetonitrile (solvent B) were pumped at a flow rate of 300 nL.min −1 in the following order: 1.0% B for 5.0 min, a linear gradient from 1.0% to 40.0% B in 80.0 min, flushing at 90.0% B for 10 min and column re-equilibration at 1.0% B for 25 min. The column temperature was kept constant at 50°C. The nanoHPLC system was hyphenated to a Thermo Scientific™ Q Exactive™ Plus Hybrid Quadrupole-Orbitrap™ mass spectrometer via a Thermo Scientific™ Nanospray Flex™ ion source. The source was equipped with a SilicaTip emitter with 360 µm o.d., 20 µm i.d. and a tip i.d. of 10 µm purchased from CoAnn Technologies Inc. (Richland, WA, USA). The spray voltage was set to 1.5 kV, the S-lens RF level to 55.0 and the capillary temperature to 320°C. Each scan cycle consisted of a full scan at a scan range of m/z 350-2,000 and a resolution setting of 70,000 at m/z 200, followed by 5 datadependent higher-energy collisional dissociation (HCD) scans using a 2.0 m/z isolation window for precursor isolation and 28% normalized collision energy for fragmentation with a resolution setting of 17,500 at m/z 200 for data acquisition. For the full scan, the automatic gain control (AGC) target was set to 3e6 charges with a maximum injection time of 100 ms; for the HCD scans, the AGC target was 1e5 charges with a maximum injection time of 100 ms. Already fragmented precursor ions were excluded for 30 seconds. The sample was measured in five technical replicates. Data acquisition was conducted using Thermo Scientific™ Chromeleon™ 7.2 CDS, data evaluation and relative quantification of N-glycosylation with Thermo Scientific™ BioPharma Finder TM software version 4.0. Glycoprofiling of Released N-Glycans of cetuximab by xCGE-LIF Released N-glycans were prepared using a glyXprep™ kit and analyzed on a xCGE-LIF-based glyXboxCE™ system (glyXera, Magdeburg, Germany), according to Cajic et al. 40 For migration time alignment, crucial for glycan peak annotation via migration time matching with database entries of glyXbaseCE™ (glyXera), 1.0 μL of sample was mixed with 1.0 μL of 2 nd NormMiX™ (glyXera) and 1.0 μL prediluted GeneScan™ 500 LIZ® Size Standard. The mixture was combined with 6.0 μL glyXinject (glyXera) and subjected to xCGE-LIF analysis. For xCGE-LIF measurements, the samples were electrokinetically injected and separated with a running voltage of 15 kV for 40 min. The generated glycan data was analyzed with the glyco-analysis software glyXtoolCE™ (glyXera), which automatically performed migration time alignment, raw data smoothing, peak picking, relative quantification, and peak/ structure annotation. NMR spectroscopy Unless stated otherwise, spectra were either recorded on a 600 MHz Bruker Avance III HD spectrometer equipped with a 1 H/ 13 C/ 15 N/ 31 P quadruple-resonance room temperature probe at 298 K. Few spectra were recorded on 500 MHz or 700 MHz Bruker Avance III HD spectrometer equipped with cryogenic 1 H/ 13 C/ 15 N triple-resonance probes (TCI) at 298 K. For all the NMR measurements, standard 5 mm NMR tubes (ARMAR, Type 5TA) with a sample volume of 500 μL were used. The oligosaccharides were assigned using the following 2D experiments: 1 H-13 C HSQC, 1 H-13 C HMBC (hmbcgpndqf), 1 H-1 H TOCSY, 1 H-1 H COSY (cosygpppqf), 1 H-13 C HMQC-COSY, 11 1 H-1 H ROESY and 1 H-15 N HSQC. Data were measured and processed using Topspin 3.5/3.6.1 (Bruker). Resonance assignments were obtained using the software Sparky 3.114 (T. D. Goddard and D. G. Kneller, SPARKY 3, University of California, San Francisco, USA). For referencing, 2,2-dimethyl-2-silapentane-5-sulfonic acid (DSS) (ARMAR Chemicals) was added to the samples after measuring all other spectra. A 1D 1 H experiment was performed for referencing the proton chemical shift. The carbon and nitrogen dimensions were referenced according to the IUPAC-IUB recommended chemical shifts referencing ratios of 0.251449530 ( 13 C) and 0.101329118 ( 15 N). 41 Estimated quantification of α-Gal content of abatacept The 2D 1 H-1 H TOCSY spectrum of abatacept under denaturing conditions measured using a mixing time of 80 ms, 64 scans, a recycle delay of 1.5 sec, 2048 × 700 points and spectral widths of 12.9 ppm × 8.3 ppm with a duration of nearly 23 hours 45 min was used for an estimation of α-Gal content. The ratio between the signal intensity of the H4/H1 correlation of α-Gal and the H2/H1 correlation of Man4 was 1:83 in abatacept. In contrast, the same ratio was 1.2:1 in cetuximab, so that the α-Gal content of abatacept was ~100× smaller than in cetuximab. Ayoub et al. reported that 78% of the N-glycans of the Fd part of cetuximab contained at least one α-Gal epitope. 19 Considering the reported N-glycan profile, on average, 1.31 equivalents of α-Gal are present on each N-glycan of the Fd part, resulting in 2.62 equivalents α-Gal per IgG, i.e., 2.62 μmol/μmol. Based on the ~100× smaller α-Gal content in abatacept, a content of 26 nmol/μmol Fc-fusion dimer is calculated. Using an average molecular weight of 92300 g/mol (containing glycans) and the reported content of 370 pmol/mg abatacept reported by Bosques et al. 6 results in 34 nmol α-Gal /μmol abatacept.	2022-10-15T06:17:44.996Z	2022-10-14T00:00:00.000	{ "year": 2022, "sha1": "6d15ae755e37ccee96b76d8fb35bc6cd2f5ac7cb", "oa_license": "CCBYNC", "oa_url": "https://doi.org/10.1080/19420862.2022.2132977", "oa_status": "GOLD", "pdf_src": "TaylorAndFrancis", "pdf_hash": "d4bc2fdca98ca304cd62a49d197432e59dc428bc", "s2fieldsofstudy": [ "Chemistry", "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
16959814	pes2o/s2orc	v3-fos-license	Excitations in correlated superfluids near a continuous transition into a supersolid We study a superfluid on a lattice close to a transition into a supersolid phase and show that a uniform superflow in the homogeneous superfluid can drive the roton gap to zero. This leads to supersolid order around the vortex core in the superfluid, with the size of the modulated pattern around the core being related to the bulk superfluid density and roton gap. We also study the electronic tunneling density of states for a uniform superconductor near a phase transition into a supersolid phase. Implications are considered for strongly correlated superconductors. The theoretical idea that 4 He could become a supersolid [1] at low temperature has motivated experimental studies of defect excitations in the solid phase [2], as well as a theoretical investigation of vortices in a superfluid near a first-order transition to a supersolid [3], as indirect windows into the proposed supersolid phase. Recent experimental hints [4] for a non-classical rotational inertia in solid 4 He have revived interest in this field. The copper oxide superconductors (SC) present a different situation, where the primary interest is in the uniform superconducting state, but it has long been recognized that the underdoped regime is plagued by various competing phases [5] involving spin, charge, and current ordering. Such competing phases can significantly affect the excitations in the superconducting state if there is a continuous transition between the uniform SC and these ordered phases. This is viewed as one possible cause for the anomalous excitation spectra of the underdoped cuprates. However, continuous phase transitions between phases with different order parameters are unusual and have only recently begun to be explored [6]. Further, recent tunneling experiments [7] in highly underdoped Ca 2−x Na x CuO 2 Cl 2 (T c ≈ 10 − 20K) suggest that the ground state in this regime may be a supersolid rather than an insulator. Motivated by this, we consider here a continuous transition between a superfluid (superconductor) and a supersolid, on a lattice, and explore some qualitative consequences for excitations in the uniform superfluid (superconducting) state near such a transition. The following are our main results. (i) A continuous transition into a commensurate supersolid phase is known to arise, with increasing interactions, from condensation of rotons when the roton gap vanishes. Here, we show that alternatively a uniform current flow in the superfluid can also induce supersolid order by driving the roton gap to zero. (ii) This current-driven collapse of the roton gap results in a supersolid pattern emerging around the vortex core in the uniform superfluid as shown in Fig. 1. The length scale over which this modulation is significant is R v ∼ D s /E rot , where D s is the bulk superfluid stiffness and E rot is the roton gap in the homogeneous superfluid. (iii) We use microscopic calculations and Ginzburg- Landau (GL) theory to obtain the long distance profiles of the modulated and uniform components of the density and superfluid order parameter around the vortex, as shown in Fig. 1. (iv) Finally, we ask whether tunneling into a uniform SC near a transition into a supersolid phase can probe the low energy roton excitations. Within a slave boson approach, we show that a tunneling electron can excite the bosonic condensate modes in the SC. In SCs with a small superfluid density, this leads to inelastic secondary peaks in the tunneling spectrum as shown in Fig. 2. Coherence factors however suppress contributions from wavevectors corresponding to the roton minimum. We conclude with possible implications for correlated SCs such as the doped cuprate materials. Microscopic model and Landau theory: Consider a model of hard-core bosons with nearest-neighbor repulsion on the triangular lattice, For large J ⊥ /J z ≡ ∆ the ground state of this Hamiltonian is a uniform superfluid. In the interaction dominated regime, for ∆ < ∼ 0.2, quantum Monte Carlo simulations [8] have shown that the uniform superfluid is unstable to density-wave modulations at wavevectors ±Q ≡ (±4π/3, 0), leading to a supersolid. Through the standard mapping ( , this Hamiltonian is equivalent to an S = 1/2 XXZ spin model with ferromagnetic in-plane interaction J ⊥ and an antiferromagnetic out-of-plane interaction J z . Spin-wave calculations on this XXZ model [9], valid at large-S, show that the phase transition into the supersolid is caused by the vanishing of the roton gap at ±Q, which leads to roton condensation. They also capture the precise structure of the supersolid phase of model (1) which is now known from direct numerics for S = 1/2 [8]. Further, both techniques agree that the transition from the superfluid to the supersolid is continuous which is also indicated from a GL analysis outlined in Melko et al. [8]. Since we are interested in the consequences of a continuous superfluid to supersolid transition, we focus on this specific model. We will use (semi)classical analyses of the XXZ model, supplemented by GL theory, as a reliable guide to the physics of model (1). Keeping the relevant wavevectors q = 0, ±Q to describe the superfluid to supersolid transition, the superfluid order, ψ(r), and the deviation of the density from half-filling, δn(r) = n(r) − 1/2, can be expressed as The complex numbers ψ 0 , ψ ±Q , ρ and the real number m are, thus, order parameters in the GL theory [8,10,11]. For the model (1), ψ 0 is nonzero in, both, the superfluid and the supersolid, while ψ ±Q , m, ρ are nonzero only in the supersolid. Since we are interested in a hard-spin formulation here, it is convenient to work with the phase and amplitude of the superfluid order parameters ψ 0 , ψ ±Q . Terms such as ψ 2 Q ψ * −Q ψ * 0 and ψ * Q ψ * −Q ψ 2 0 in the GL functional lock the phases of these different components, so that ψ 0 = \|ψ 0 \|e iϕ , ψ ±Q = \|ψ ±Q \|e iϕ are determined by three amplitudes and a single phase variable ϕ. The amplitudes are constrained by the \|ψ 0 \| 2 +\|ψ Q \| 2 +\|ψ −Q \| 2 +m 2 +2\|ρ\| 2 = 1, which fixes the spin length (to unity). The where we have only displayed terms relevant to our analysis below. These terms respect the particle-hole and lattice symmetries of the Hamiltonian in Eq. (1), and could be, in principle, derived using functional integral methods as outlined for a related model in Ref. [12]. Here we treat the coefficients as phenomenological parameters. In Eq. (2), α ρ ≡ a(∆ − ∆ c0 ) measures the distance to the superfluid-supersolid transition point with a > 0. Spin-wave theory and Monte Carlo calculations indicate w ρ < 0, which prefers ρ to be real in the supersolid. In f c , the term λ 1 couples the modulated solid order and the superflow current ̺ s ∇ϕ. The λ 2,3,4 terms induce effective "magnetic fields" upon the different order parameters. Current driven collapse of the roton gap: For superfluids close to a supersolid instability, external perturbations that suppresses the kinetic energy relative to the interaction energy can drive the system into the supersolid phase. Consider a uniform superflow, say in the x direction, which introduces a phase difference between neighboring sites, δ = φ j − φ i . The phase gradient effectively suppresses the nearest-neighbor kinetic energy from ∆ to ∆ cos(φ i − φ j ). Carrying through the calculation of spin-wave fluctuations around such a state with superflow, we find a spin-wave dispersion with a roton gap ω(Q) = 1.5 6∆(∆ − 1/2), wherẽ ∆ ≡ ∆[2 cos(δ/2) + cos δ]/3, reduced by the current flow. The point where the roton gap collapses is thus shifted upward to ∆ c ≃ ∆ c0 [1 + (δ/2) 2 ], where in the last step we assumed δ ≪ 1. Recalling the current density J ∼ sin δ ≈ δ, we conclude the boost of ∆ c is quadratic in the supercurrent density close to the transition, and the superflow induces supersolid ordering. In the GL approach, superflow-induced supersolid ordering can be described phenomenologically by the coupling term λ 1 between the supercurrent and the supersolid order parameter ρ in Eq.(5). This term effectively shifts α ρ to α ′ ρ = α ρ − λ 1 J 2 s , where the supercurrent J s = ̺ s \|∇ϕ\|. We thus see that one can induce a transition from a superfluid (α ′ ρ > 0) into a supersolid (α ′ ρ < 0) by a superflow which leads to a sufficiently large \|∇ϕ\|. The "critical current density" required to obtain the supersolid is J c = α ρ /λ 1 . This is analogous to superflow induced spin-density modulation [13] and gradient couplings inducing competing phases [14] considered earlier only within GL theory. We next turn to implications of this for a vortex in the superfluid phase. Vortex in the superfluid phase: For a superfluid vortex, the current density increases upon approaching the core as J s = ̺ s /r. The above result then indicates that the superflow would cause a supersolid pattern to be stabilized in a region with characteristic radius R v = ̺ s λ 1 /α ρ around the core. It is clear from the GL functional that α ρ ∼ E 2 rot , since the roton gap determines the distance to the transition. We thus identify R v ∼ ̺ s /E rot . We have confirmed this by microscopic calculations where we have evaluated the roton gap, the superfluid stiffness D s (which has energy units and can be viewed as ̺ s /m * with an effective mass m * ) and the "critical current" using spin-wave theory. To study, further, the interplay of the various order parameters near a vortex, we numerically studied model (1) in the presence of an orbital magnetic field, by replacing t in Eq. (1) with t ij = t exp(i i j A · dl). We considered an L×L lattice with periodic boundary conditions, and a vector potential corresponding to a uniform magnetic field with the total flux chosen to be one flux quantum. The ground state then corresponds to a single vortex within the cell. Transforming to an effective spin model, and treating the spins as classical unit vectors, we found the ground state spin configuration {S r } using a simulated annealing algorithm. The ground state corresponds to a single vortex with its core on a site of the lattice. The local order parameters at each r, i.e. ρ, m, ψ 0 , and ψ ±Q , were extracted from {S r } by coarse graining over six nearest neighbor sites. The inset to Fig. 1 shows the map of the boson density δn(r) around a vortex for ∆ = 0.505 -it is clear from this that supersolid order is induced in the vicinity of the vortex core as expected from earlier arguments. Each data point in the plotted profiles in Fig. 1 represents the angular average taken at a fixed distance from the core. It is apparent that as the superflow induces a nonzero ρ, the order parameter \|ψ 0 \| is suppressed. At the same time, \|ψ ±Q \| and m are generated but they are insignificant except very near the vortex core. We explain these features below using GL theory. Tunneling: We next analyze electron tunneling into the uniform SC to see if it probes the low-energy roton excitation when the system is on the brink of becoming a supersolid. We use a slave boson formulation, writing the electron operator as c σ (r, τ ) = f σ (r, τ )b(r, τ ) where f σ is a spinful fermion (spinon) and b is a charged boson. Assuming that gauge fluctuations and interactions are innocuous in the superconducting state, the electron Green function can be factorized as G c σ (r, τ ) = G f σ (r, τ )G b (r, τ ). In mean-field theory, when the boson is condensed, we can replace G b (r, τ ) ∼ b 2 , so that G c ∼ G f , their ratio being the condensate density b 2 . The electron tunneling spectrum thus reflects the tunneling density of states (TDOS) associated with the spinon spectrum. Going beyond mean field theory, we assume that the bosons are described by an interacting Hamiltonian such as (1) and evaluate the boson Green function using spinwave theory. Combining this with the spinon Green function, for a model with spinon pairing, leads to the TDOS Here u 2 k , v 2 k are the usual superconducting coherence factors associated with spinon pairing, while γ 2 q is the coherence factor associated with the Bogoliubov transformation for diagonalizing the boson model [15]. The first term, which dominates at large-S, reflects the mean field result where the condensate does not fluctuate. The second term represents inelastic processes where tunneling inserts a spinon and excites condensate fluctuations. To show the effects of condensate fluctuations in a nodal SC on the triangular lattice, we plot in Fig. 2 the TDOS N c (ω) obtained for f -wave triplet pairing of spinons [16]. The two main features of the spectrum in Fig. (2) are: (i) At low energy the nodal spectrum is nearly unaffected by the condensate fluctuations. (ii) At higher energies, condensate fluctuations lead to a broad secondary peak. This arises from q regions where there is a large density of states for condensate excitations. These remain true even in the presence of low-energy rotons since the γ 2 q coherence factor tends to reduce the contribution to the TDOS coming from near the roton wavevector Q. The energy difference between "elastic" and "inelastic" peaks scales with the superfluid stiffness. Experimental implications: Our results for the superflow induced supersolid, the vortex core size scaling and the existence of secondary tunneling peaks rely mainly on the existence of a continuous superfluid-supersolid transition and a small superfluid density, largely independent of the detailed microscopics. Here we consider some implications for the cuprate superconductors. Tunneling measurements on near-optimal superconducting Bi 2 Sr 2 CaCu 2 O 8+δ (T c = 89K) in a B = 5T magnetic field see a modulated TDOS around the vortex core [17]. Although different explanations have been proposed for this pattern (e.g., Refs. [20,21,22]), it is possible that over the large region (with radius ∼ 75Å) in which the modulated pattern appears the system may be better viewed as a supersolid (which appears to be seen at low doping and B = 0 in a related cuprate [7]). If this supersolid has a charge modulation, the roton in this system at B = 0 could be observed using inelastic X-ray scattering. We expect that the roton energy should scale as E rot ∼ D s /R v where D s is the bulk superfluid stiffness and R v is the radius of the region around the vortex core exhibiting a checkerboard modulation. Further, the observed modulation region around the vortex core should grow with underdoping upon approaching the supersolid. Both predictions could be tested experimentally. We have also shown that condensate fluctuations can lead to inelastic secondary peaks at higher energies, while leaving the low energy V-shaped spectrum unaffected. This is broadly consistent with tunneling data in the cuprates where regions with a small superfluid density and broad secondary tunneling peaks (so-called "large gap" regions) coexist with "small gap" regions of larger superfluid density and single peaks [18]. However, the fact that the high energy structure carries information about the ordering wavevector in the cuprates [19] does not appear to be a feature of the model we have studied. This may need coupling to disorder which can simultaneously lead to low superfluid density and supersolid order in certain regions and also to scattering of quasiparticles off the disorder induced supersolid. Further, one needs to address possible bond-order instabilities [20,23,24]. We hope our work will also stimulate experiments to study supersolids and vortices [25] using cold atoms on optical lattices, and vortices in superfluid 4 He near freezing pressure. We thank J.C. Davis, T. Hanaguri, H. Hoffman, H.-Y. Kee, Y.-B. Kim, Y.-J. Kim, S. Sachdev, A. Vishwanath, and Z. Wang, for useful discussions and correspondence. This work was supported by a startup grant and a Connaught grant from the University of Toronto.	2017-04-13T00:21:51.594Z	2005-12-30T00:00:00.000	{ "year": 2005, "sha1": "ba2f32be4426e99e21c1731d8f96f4cfe31b6cfc", "oa_license": null, "oa_url": "http://arxiv.org/pdf/cond-mat/0512717", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "ba2f32be4426e99e21c1731d8f96f4cfe31b6cfc", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics", "Medicine" ] }
269478737	pes2o/s2orc	v3-fos-license	The impact of apicobasal ridges on dental load-bearing capacity in aquatic-feeding predatory amniotes Non-technical Summary Complex enamel ridges have evolved multiple times on the teeth of unrelated aquatic predators, including extinct marine reptiles, toothed whales, crocodilians, and aquatic-feeding dinosaurs. Their appearance in such a wide range of groups suggests that they are a specialized structure adapted to perform specific functions in the capture and/or processing of prey, although these functions are unknown. This study used computer modeling to apply bite force simulations to a set of digital tooth models in order to identify whether the ridges strengthened the tooth. These models enabled us to visualize how bite force stress is distributed around smooth teeth compared with ridged teeth, including a range of ridge types. Our results suggested that the ridges do not strengthen the tooth crown overall, indicating that they may instead serve another role in prey handling. Abstract Apicobasal ridges are longitudinal ridges of enamel that are particularly common in several clades of aquatic-feeding predatory amniotes, including Plesiosauria, Ichthyosauria, Mosasauridae, Crocodylia, and Spinosauridae, as well as some early members of Cetacea. Although the repeated evolution of these dental ridges in unrelated clades suggests an adaptive benefit, their primary function in feeding is debated. Hypothesized functions range from increasing tooth strength to improving prey puncture or removal efficiency, but these have never been quantitatively tested. This study utilizes finite element analysis (FEA) to assess the impact of apicobasal ridges upon tooth crown strength in aquatic-feeding amniotes. Drawing on morphometric data from fossilized tooth crowns, a set of digital models was constructed to calculate the performance of smooth and ridged tooth variants under simulated bite force loadings. The similarities in overall stress distribution patterns across models of the same tooth shape, regardless of the presence or morphology of ridges, indicate that apicobasal ridges have little impact on stress reduction within the tooth crown. Ultimately, these findings suggest that apicobasal ridges have a minimal role in improving crown strength and form a framework for future research into the remaining hypotheses. Introduction Geological time has borne witness to many reinvasions of the aquatic environment by lineages of terrestrial tetrapods (Fish 2016).These reinvasions were dependent upon a variety of evolutionary shifts in morphological and behavioral traits (Fish 2016;Houssaye and Fish 2016).The return to aquatic environments placed immense selective pressure on the terrestrial Bauplan, resulting in widespread convergence of morphological traits between secondarily aquatic tetrapods (Howell 1930;Braun and Reif 1985;Thewissen and Nummela 2008;Houssaye and Fish 2016).Although recent research into aquatic-feeding predators has focused on the dietary significance of tooth shape (Lukeneder and Zverkov 2020;Sulcova et al. 2020;Fischer et al. 2022), little is known about the function of crown surface morphology.This gap is particularly evident in the case of apicobasal ridges, longitudinal enamel ridges present in several clades of secondarily aquatic amniotes (McCurry et al. 2019) (Fig. 1).The structural makeup of these ridges varies between taxa; cross-sectional analysis by McCurry et al. (2019) demonstrates that in plesiosaurs, the ridges do not extend into the dentine, resulting in thicker enamel at their peaks, which may be due to differential growth rates between enamel types (Sander 1999(Sander , 2000)).In other clades, including spinosaurids, odontocetes, and crocodilians, the intersection between dentine and enamel extends parallel to the outer surface of the ridges (McCurry et al. 2019).In ichthyosaurs, the ridges mirror an internal dentine folding known as plicidentine, which begins within the crown and extends toward the root (Maxwell et al. 2012;McCurry et al. 2019).These disparate structural morphotypes support a convergent result, implying distinct amelogenetic pathways to an analogous trait (Sander 1999;McCurry et al. 2019). Apicobasal ridge morphologies vary in shape, size, and arrangement.The abundance of morphological variation, even within some lower-ranked taxa, suggests that-like crown shape (Massare 1987)-ridge morphologies are related to diet rather than clade, although species specificity does occur (Benson et al. 2013;Zverkov et al. 2018).One example of this species specificity is the common characterization of plesiosaur teeth by widely spaced, distinct ridges on a smooth surface, with an often trihedral tooth shape (Benson et al. 2013;Zverkov et al. 2018). Some specimens with ridges around the entire crown circumference, such as some Late Cretaceous material previously attributed to Polyptychodon, display density that differs between tooth faces, with lingual ridges clustering more closely than buccal ridges (Madzia 2016).In taxa with trihedral or subtrihedral crown cross sections, ridges may be present on only one or two of the faces, typically excluding the buccal face (Fischer et al. 2015).These taxa also often trend toward an increased clustering of ridges on the lingual surface (Taylor 1992;Buffetaut 2013;Richter et al. 2013). The ridges themselves can be high or low relief, varying within individual teeth as well as between species, but remain visually distinct from smooth enamel (Buffetaut 2013).Coverage of the apicobasal length is also variable; across taxa, ridges taper off or anastomose anywhere from a few millimeters above the base to directly below the apex (Massare 1987;Young et al. 2014;Madzia 2016). The strength hypothesis is among the most frequently cited hypotheses aiming to explain the function of apicobasal ridges in aquatic feeding strategies.This hypothesis focuses upon the potential for the ridges to strengthen the tooth crown (Preuschoft et al. 1974;Vaeth et al. 1985;Young and Kardong 1996;Schulp 2005;Young et al. 2014;Zverkov et al. 2018), possibly by channeling pressure toward its base and the "valleys" between ridges (Preuschoft et al. 1974;Rieppel and Labhardt 1979;Sander 1999) or by increasing the second moment of area (I ), an engineering principle that quantifies the rigidity of a shape and would enable the ridges to act comparably to corrugations in a metal sheet by increasing the load-bearing surface area (Khalid et al., 2004).It should be noted, however, that the latter comparison is somewhat confounded by the addition of dentine below the enamel ridges, increasing I, while metal sheets generally bear no additional material below their corrugations.As the strength hypothesis is only one of several plausible hypotheses, the remainder of which include improving grip on slippery prey and enhancing puncture efficiency (McCurry et al. 2019;Crofts et al. 2020), testing is required to better understand the physical capabilities of the ridges and any strength advantages they may have conferred. Here we aim to assess the impact of apicobasal ridges on the ability of the tooth crown to resist various loading conditions, through the application of finite element analysis (FEA).We compare von Mises stress responses across a range of typical apicobasal ridge morphologies, analyzing the influence of ridge shape, size, and arrangement on crown strength.Furthermore, we use comparative morphometrics to evaluate morphological variation in apicobasal ridges and determine whether morphologies are clustered by phylogeny.Although ridge morphology is relatively well documented in plesiosaurs (Benson et al. 2013;Fischer et al. 2015;Madzia 2016;Zverkov et al. 2018), our knowledge of trends in other taxa, such as odontocetes, ichthyosaurs, and mosasaurs, is currently limited and may be expanded through this study. Specimen Selection and Imaging Forty-eight individual tooth specimens were selected to represent six major clades of aquatic-feeding amniotes (Plesiosauria, Ichthyosauria, Crocodylia, Mosasauridae, Spinosauridae, and Odontoceti) (Appendix 1).Specimen selection was informed primarily by preservation, excluding worn and damaged specimens for which a cross section could not be measured.As only one set of specimens (ROM 12809 Tooth A-F) was originally found in partial articulation, all specimens in the dataset have been treated as isolated teeth.We were unable to assign positions within the jaw for each tooth due to the prevalence of homodont dentition in marine reptiles (Ciampaglio et al. 2005).While position along the toothrow affects the amount of force experienced by a crown (Bourke et al. 2008;Cohen et al. 2020), it is less likely to affect potential directions of force (lateral and vertical).As this study compares responses to directional force, jaw position is unlikely to greatly affect the results.Specimens were instead selected to portray a variety of ridge and crown morphologies. The specimens were drawn from the collections of four museums.High-resolution microcomputed tomography and neutron imaging were utilized by previous researchers to digitize the specimens (McCurry et al. 2019).This process was completed using either Nikon Metrology's combined 225/450 kV microfocus X-ray and computed tomography (CT) walk-in vault system at Chesapeake Testing in Belcamp, Md., U.S.A., or the DINGO neutron beamline at Australia's Nuclear Science and Technology Organisation (ANSTO) in Lucas Heights, NSW, Australia (McCurry et al. 2019).The teeth were then digitally isolated from the CT scan and exported as three-dimensional surface models in STL (standard triangular language) format by previous researchers (McCurry et al. 2019). Data Collection The three-dimensional models were orientated within a Cartesian coordinate system and measured in Rhino v. 6 (McNeel 2019). Crown height was recorded from the enamel dentine junction to the apex of the crown (Fig. 2, Appendix 2).Where the apex was incomplete, measurement concluded at the highest point of the crown to avoid overestimation of crown height, although we acknowledge this would also result in a slight underestimation within some specimens.Crown width was recorded at thewidest point of the crown for both the labiolingual and mesiodistal orientations (Fig. 2, Appendix 2).For mold fossil specimens that lacked a three-dimensional crown, width was recorded only for the visible orientation. Ridge height and width measurements were taken for 10 ridges per tooth across a lateral cross section of the crown at 50% of its height (Appendix 3).The apicobasal span of the ridges was recorded as a percentage of the total crown height (Appendix 2).If a cross section bore more than 10 ridges, 10 were randomly selected for data collection.If fewer than 10 ridges were identified, the specimen was excluded from the ridge measurement section of the study.The overall count of ridges around the circumference of the cross section were recorded, regardless of whether the tooth was included in the measurement section, to gain a holistic understanding of ridge count across all taxa.Carinae were not classified as apicobasal ridges for the purposes of this study, as they are commonly accepted to be separate structures differing in abundance and relief (Young et al. 2014;Hendrickx et al. 2015;McCurry et al. 2019).All linear measurements were log transformed in R v. 4.0.4(R Core Team 2021) to conform to a normal distribution and to visualize morphological differences between specimens that vary considerably in the analyzed traits. Analysis of Ridge Morphology Linear Regression Models.A range of morphological variables measured on the specimens were fit to linear regression models to investigate the dental diversity of aquatic-feeding amniotes and identify taxon-specific traits.The use of linear regression models allows an evaluation of variance in all relationships of potential interest and provides insight into isometry or allometry within the scaling mechanics of the ridges.The compared variables included relative apicobasal ridge height and width, crown height and width, and ridge count.Relative ridge measurements were included to ensure the exclusion of size bias and were calculated by dividing each tooth's average ridge height or width by the crown height or width respectively.The labiolingual and mesiodistal crown width measurements were averaged to provide a representative singular width for the regression models.All models utilizing count data were analyzed using a generalized linear model with a quasi-Poisson distribution to accommodate this type of data and variance dispersion.All regression models were conducted and plotted in R v. 4.0.4(R Core Team 2021). Taxonomic Variance in Ridge Morphology.To determine any significant differences between the average ridge data of the six aquatic-feeding amniote groups examined in this study, univariate Kruskal-Wallis tests were performed in R v. 4.0.4(R Core Team 2021).These variance tests were run on data for relative crown height, relative crown width, and ridge count across all six groups.All R scripts are available in the Supplementary Material. FEA FEA is a numerical technique enabling the digital simulation of directional force across mechanical components of complex structures (Clough 1990;Dumont et al. 2005).In FEA software, the visualization of stress and strain is directed by the application of load cases and constraints in contextually relevant positions across a model (Dumont et al. 2005).Our study conducted FEA upon digital tooth crown models to determine the impact of apicobasal ridges upon overall crown strength.These tests utilized von Mises stress distributions to compare model performance under the simulation of various loading conditions.Von Mises stress is often incorporated into biological model analyses to establish potential failure points under a ductile model of fracture or a model in which deformation precedes breakage (Dumont et al. 2009;Rajabizadeh et al. 2021).Although we acknowledge that enamel and dentine conform to a brittle pattern of fracture, thus making von Mises stress unsuitable to predict their failure points, it does offer a thorough visualization of stress distribution for these materials (Neves et al. 2015) and is frequently used as such in biomechanical analyses of teeth (Whitenack et al. 2011;Palci et al. 2021;Rajabizadeh et al. 2021;Pollock et al. 2022).Thus, to enable a comparison between our results and similar biological studies, we have utilized von Mises stress to visualize our stress distribution.All FEA was completed using Strand7 v. 2.4.6.* Model Creation.The tooth crown and apicobasal ridge data were used as a foundation for the construction of a digital set of seven three-dimensional tooth crown models in Rhino v. 6 (Fig. 3; see Supplementary Material).To reflect the scope of tooth crown morphologies within the dataset, the models were based on two of the most widespread crown shapes in aquatic-feeding dentition (conical and recurved).Common variations in ridge size and shape were re-created on conical models, while variations in ridge arrangement were re-created on recurved models to replicate the range documented in recurved teeth (Taylor 1992;Buffetaut 2013;Richter et al. 2013). The Average Ridges model represents the average ridge measurements, the height and width of which were applied to all models except the Large Ridges model, which instead bears both wider and taller ridges matching the maximum recorded dimensions.The Rounded Ridges model exemplifies ridges with a rounded cross section.The Lingual Clustering model bears ridges only on the recurved lingual face, while the Symmetrical Ridges model represents recurved teeth with ridges around the entire circumference. Although many fossil specimens have ridges of varying lengths along the crown, all ridges in our models ended at 72% of the crown height (the proportional average) to ensure standardization.Control models without ridges were created for both tooth shapes to enable a comparison between identical teeth with and without ridges (Fig. 3).All models were designed to fit the average crown height and width to eliminate size bias.Although ridge shape can manifest as triangular or rounded in cross section, depending on taxon (McCurry et al. 2019), triangular ridges were used as the default ridge shape due to the challenges in modeling rounded ridges. Meshing and Scaling.Each model was imported as an STL surface mesh into Strand7 v. 2.4.6.The models were cleaned and automatically remeshed to a new surface mesh of three-noded triangular plate elements.This methodology reduces warping and allows the accurate application of geometry in complex areas (Dumont et al. 2005).The refined surface mesh was then used as a base to create a solid mesh of four-noded tetrahedral elements, and the plate elements were deleted upon completion.All models were scaled to the same surface area, as previous research indicates that scaling to surface area will return a consistent stress energy across models of different sizes and surface areas (Dumont et al. 2009). Material Properties.As dentine comprises the bulk of the crown mass and aids in the dissipation of force, the mechanical properties of dentine are applied to the mesh, in line with previous studies (Jansen van Vuuren et al. 2016).These parameters were utilized for both the Young's modulus (E = 18,000 MPa) and Poisson's ratio (v = 0.31) (Bensonne et al. 2014).It is possible that the applied properties do not accurately represent the enamel structure of the crown, due to the differing compressive and tensile strength of enamel and dentine (Milewski 2005).This is exacerbated by the taxonomic differences in the thickness and arrangement of enamel, from the thickened ridges in plesiosaurs to the plicidentine of ichthyosaurs (McCurry et al. 2019), as these compositional disparities may have biomechanical effects due to the higher force resistance of dentine and higher wear resistance in enamel (Chun et al. 2014).Although a recent study found that the material properties of dentine and enamel did not differ enough to affect von Mises stress distributions in reptile mandibles when enamel thickness was altered (Herbst et al. 2021), it is possible that an effect would have been present in a smaller area localized only upon the tooth.While the internal homogeneity of our models is, as such, a potential limitation, the study makes an assumption similar to that of Dumont et al. (2009) due to the difficulties of accurately representing and modeling the range of internal heterogeneity.Accordingly, we predict that the effects of internal tissue properties would be consistent across all models and treat tooth material properties as homogenous, as in recent literature (Palci et al. 2021;Pollock et al. 2022). Constraints and Load Cases.For the application of realistic constraints, each tooth model was fixed along the degrees of translation and three degrees of rotation by the external nodes across the entire base.This corresponds to the lack of axial displacement that would have been present in the tooth socket. Load cases were comprised of three primary directional pressures, each simulating a direction of stress that could have been experienced during prey capture, incapacitation, or processing.The primary three load cases were Bite (−40 N per node on the z-axis), Shake (40 N per node on the y-axis), and Pull (−40 N per node on the x-axis) (Fig. 4).Two further load cases called Bite and Shake and Bite and Pull combined the prior individual cases to analyze the effects of multiple simultaneous forces.In each of the five load cases, the force loads were applied to the topmost five nodes of each model's apex.While we acknowledge that confining force application to the apex does not encompass a full representation of conditions where the tooth is embedded in prey, it does allow us to model the effects of force applied through a hard surface, such as when the tip of a tooth contacts bone, where any additional strength conferred by ridges would be most relevant. Bite force data are challenging to attain, even for extant species (Koc et al. 2010), and are highly correlated to body size (Erickson et al. 2012), preventing simple estimation for the isolated teeth in this study.Thus, the above loading units were assigned to standardize the applied bite force between models.As this study seeks to compare only the relative performances of models, the standardization of loads without precise bite force data is appropriate to provide an accurate result (Dumont et al. 2009). The Bite load case was informed by the expected pressures of a typical vertical bite, involving the forceful closure of the mandible documented in a range of vertebrate predators, including crocodilians (Araújo and Polcyn 2013).The Shake load case aimed to test the response of teeth to the lateral-shake feeding strategy observed in crocodilians, proposed for the pliosauroid Rhomaleosaurus due to similarities with crocodilian cranial morphology (Taylor 1992).Similarly, the Pull load case was implemented to test the models' response to the backward pulling movements noted in modern crocodilian feeding (Gallagher et al. 2018).Each of the Bite, Shake, and Pull load cases were applied and assessed on each model separately to obtain an understanding of how the models responded to these pressures individually.However, the inclusion of the Bite and Shake and the Bite and Pull load cases ensured an evaluation of the models' performance under more biologically accurate conditions.FEA Contour Maps.Each solid mesh was solved as a linear static model, calculating brick stress for both internal and external bricks.The results were displayed as von Mises stress distribution maps, in which concentrations of warm colors indicated highstress areas, while concentrations of cool colors indicated lowstress areas.Statistical tests of these maps were omitted from this study due to inaccurate conclusions resulting from differences in brick size within and between models. Crown and Ridge Measurements From the collected measurements, the average tooth crown height was 32 mm, while the average crown width was calculated as 18.136 mm mesiodistally and 17.917 mm labiolingually.The average absolute apicobasal ridge height and width were calculated as 0.136 mm and 0.663 mm, respectively, and the average ridge extends along approximately 72% of the height of the tooth crown.The average tooth bore 26 apicobasal ridges that reached at least 50% of its height.These measurements were utilized to create the seven theoretical digital tooth models representing the scope of common tooth morphologies in extinct aquaticfeeding amniotes. Analysis of Ridge Morphology Linear Regression Models.Regression analysis showed a positive correlation between the average ridge height and crown height (p < 0.001, R 2 = 0.26) (Fig. 5A) and between the average ridge width and crown width (p < 0.001, R 2 = 0.43) (Fig. 5B).The A negative correlation was found between average relative ridge width and ridge count (p = 0.003, t = −3.114).However, there was no significant correlation found between relative average ridge height and ridge count (p = 0.067, t = −1.883). Taxonomic Variance in Ridge Morphology.The univariate Kruskal-Wallis tests returned no statistically significant difference between the average relative ridge height (chi-squared = 7.897, p = 0.162, df = 5) or the average relative ridge width (chi-squared = 6.241, p = 0.284, df = 5) of the six taxonomic groups examined within the dataset.Furthermore, no statistically significant difference was found between the average ridge counts for these six groups (chi-squared = 1.691, p = 0.89, df = 5). FEA Contour Maps The contour maps (Figs.6-10) show a near-identical distribution of von Mises stress across models of the same tooth shape.Although ridged models show a slightly increased stress distribution within the ridges themselves, the interjacent smooth enamel retains a similar vertical stress gradient across the crown throughout both smooth and ridged models, as well as across all ridge morphologies.As this overall gradient does not differ, it does not appear that the slightly heightened stress in ridges relieves stress within the crown. Bite Load Case.Under the Bite load case, conical models exhibit high stress concentrations around the apex with an even gradient of stress reduction down the crown (Fig. 6A-D).This longitudinal stress reduction occurs at an equal rate in all conical models.Some slightly higher stress values appear to be concentrated on the ridges, presenting more prominently around the middle of the Large Ridges model (Fig. 6C). In recurved models, the higher stress values are located along the lingual face of the crown despite the application of force in the middle of the apex (Fig. 6E-G).This trend occurs in all recurved models under the Bite load case.The overall distribution of stress again appears almost identical across all recurved models, with most of the stress concentrated in the lingual tip of the apex and decreasing toward the buccal base, with a slightly increased stress in the ridges.Shake Load Case.Under the Shake load case, conical models display a bifurcated stress distribution in which stress is concentrated down the lingual and buccal faces (Fig. 7A-D).Despite the application of force toward the buccal surface, the stress distribution is approximately equal between buccal and lingual faces.Stress reduces gradually from the apex to the base, and the overall stress gradients are similar between ridged and smooth models, although again the ridges present with slightly higher values toward the middle of the tooth. Recurved models display a similar bifurcation of stress that is spread with a slight skew toward the buccal face due to the recurved shape (Fig. 7E-G).As in the conical models, most of the stress is centered on the apex.The stress gradually reduces toward the base, with a small increase in stress particularly around the lingual edge of the base. Bite and Shake Load Case.Under the Bite and Shake load case, conical models again demonstrate a bifurcated stress distribution, with stress reducing toward the base, but with a skew toward the buccal face (Fig. 8A-D).This skew aligns with the direction of applied force (toward the apical buccal surface).Again, the ridges Impact of apicobasal ridges on tooth strength display slightly higher stress, although the overall gradient of distribution is identical between models. The recurved models all display a bifurcation of stress with a greater buccal skew, again in line with the direction of applied force (Fig. 8E-G).Although the spread of high stress along the crown height is visibly reduced in contrast to the conical models, it is similar between all of the recurved models, with an extremely slight increase in stress on the lingual ridges. Pull Load Case.Under the Pull load case, the stress distribution in conical models presents as a reoriented version of the distribution in conical models under the Shake load case, due to the symmetry of the models (Fig. 9A-D).Both ridged and smooth models display a bifurcated distribution in which stress is concentrated equally down the mesial and distal faces, with a slightly increased stress in the ridges. The bifurcated stress distribution also presents in the recurved models (Fig. 9E-G).As in the conical models, the recurved models show an equal spread of stress down the mesial and distal faces of the crown, with most of the stress centered on the apex.The stress gradually reduces toward the base, although the slight increase in stress in the ridges is more noticeable under this load case than in the Shake case.There are otherwise no substantial differences between the smooth and ridged models within the recurved crown set. Bite and Pull Load Case.Under the Bite and Pull load case, conical models again demonstrate a bifurcated stress distribution, with stress reducing toward the base, but with a skew in the direction of the applied force (toward the mesial surface) and a slight increase of stress in the ridges (Fig. 10A-D). The recurved models all display a comparable bifurcation of stress with a mesial skew (Fig. 10E-G).The pattern of slightly higher stress in ridges is more pronounced on the mesial face under the Bite and Pull load case (Fig. 10F-G), in line with the direction of force. Patterns in Ridge Morphology Linear Regression Models.The positive correlation between apicobasal ridge measurements and tooth crown size indicates a slight trend toward increasing ridge height and width with increasing crown size (Fig. 5).Larger crowns are more likely to develop larger ridges in absolute terms; however, as the regression slope (b) is less than 1 in both models, the relationship between ridge size and crown size is negatively allometric.This suggests that ridges do not develop proportionally to crown size and may be under different selective pressures compared with crown size. The significant relationship between the relative average ridge width and ridge count (p < 0.001, t = 3.114) is likely a consequence of spatial constraints.In occupying less space, thinner ridges have a capacity to become densely packed and greater in number, whereas wider ridges limit the available enamel and restrict ridge abundance. Taxonomic Variance in Ridge Morphology.The lack of statistically significant differences in the average apicobasal ridge measurements between groups suggests an absence of taxon-specific ridge morphologies among the measured characteristics.This supports previous hypotheses claiming that phylogeny bears little to no influence upon ridge morphology (Benson et al. 2013;Zverkov et al. 2018) within the context of ridge height, width, and count. Our data do provide insight into the breadth of ridge morphologies displayed throughout the studied taxa.Some plesiosaurs have developed low-abundance, high-relief, and lingually clustered ridges (e.g., NHM 2680; NHM 46239 B), while others are characterized by high counts of finer ridges (e.g., ROM 71030; ROM 71033).Crocodilian ridge morphologies are among the most variable, ranging from the high-relief ridges displayed in genera such as Deinosuchus (USNM 5351) and Purussaurus (USNM 205338) to the rounded, low-relief morphologies present in Bottosaurus (USNM 508536).However, one Bottosaurus specimen exhibits one of the most distinctive morphologies of the dataset, in its "rippled" and apically anastomosed ridges (USNM 540758).The low, robust crown shape of such teeth has been hypothesized to assist in the crushing of hard-shelled prey items in Bottosaurus harlani (Cossette and Brochu 2018), and although our Bottosaurus specimens have not been identified to species level, the comparable morphology suggests an equivalent function.Although purportedly molluscivorous mosasaurs such as Globidens and Carinodens developed similarly robust and ridged teeth (Schulp 2005), the mosasaurs in this dataset-Mosasaurus missouriensis (USNM 4910) and Tylosaurus proriger (USNM 3885)-bear a recurved dentition to process meat, such as fish (Konishi et al. 2014), with broad, rounded, and low-relief ridges.The ichthyosaurs are the only group in which lingual clustering is universally absent (Appendix 3).Ichthyosaur ridge morphology is also highly conserved, consisting of dense, parallel, and rounded ridges that extend an equal length along the tooth crown (although the exact length differs between specimens).The few spinosaurid specimens in the dataset present wider, low-relief ridges (Appendix 3), but descriptions of various spinosaurid dentitions suggest a diversity of ridge morphologies (Buffetaut 2013).Odontocetes bear some of the most unique ridge morphologies in the dataset, ranging from discontinuous ridges that present as elongated bumps (USNM 25210 Tooth 2) to basally or exceptionally anastomosed ridges (USNM 25210 Tooth 1; USNM 23546) (Appendix 3).Although some rarer ridge morphologies may be considered potentially taxonomically informative, such as the rugose anastomosis in odontocetes, a majority of basic ridge morphologies occur across many clades.Specific combinations of ridge morphologies, such as low-density, high-relief, and lingually clustered ridges in some pliosaurids (e.g., NHM 2680; NHM 46239 B), may provide more taxonomic information than the individual traits analyzed here, but this is beyond the scope of this paper. FEA The similarities between von Mises stress distributions suggest that, overall, models of the same crown shape exhibit no substantial difference in their response to force beside a slightly increased stress localized within ridges .The otherwise comparable gradient of stress across the crown indicates that apicobasal ridges have little impact on overall tooth crown strength in these models.While, as mentioned, these results cannot account for internal heterogeneity between enamel and dentine material properties, they demonstrate the lack of influence of the ridges' external morphology on stress distribution.Furthermore, the general consistency of stress levels across the range of ridge morphologies implies that ridge shape, size, and arrangement do not contribute to a substantial increase in crown strength.While the mechanical cause of the slightly higher stress within ridges is unclear (although it may be attributable to the smaller surface area), the interjacent smooth enamel shows a rate of vertical stress decrease equal to that of smooth models.This similarity between gradients suggests that the ridges do not act to channel stress away from the crown or alleviate its loading in any form. The overall Bite distribution patterns recorded in our models (both recurved and conical) resemble those documented in comparably shaped teeth in a recent study (Pollock et al. 2022).Similarly, the bifurcated stress patterns across all teeth in our Shake load case reflect the results under those authors' equivalent Pull load case (Pollock et al. 2022), although the highest concentration of stress occurs closer to the base of their models due to the difference in point loading location.Our results further align with those in a previous study on von Mises stress distributions in smooth snake fangs (Rajabizadeh et al. 2021).When recurved models were subjected to vertical loading (equivalent to the Bite load case), the study documented a similar bifurcation of stress with localization on the lingual face of the crown.A lingual localization of stress also appears on the finite element map of a tooth from the extant shark Sphyrna mokarran (Whitenack et al. 2011).Although the crown of S. mokarran bears a higher degree of labiolingual compression, its overall curvature resembles that of the three recurved models and is thus potentially subject to similar forces under longitudinal load cases.As none of the specimens used in these studies bear visible apicobasal ridges, the similarity between distribution patterns suggests that stress distribution is predominantly a product of crown shape rather than the influence of apicobasal ridges. Relationships between Stress Distribution and Ridge Locations. The results of the finite element modeling deliver additional insight into the relationships between stress distribution and the position of ridges.If enhanced tooth strength were indeed a significant selective pressure, the centralization of stress upon the apex should lead to a greater need for strength around this region.Apicobasal ridges that increase in prominence toward the apex have been described, but only rarely, such as in the mosasaur Globidens phosphaticus (Bardet et al. 2005) or in teeth from Bottosaurus (USNM 540758) and Gavialosuchus (USNM 244431) in the current dataset.Overall, our morphometrics indicate that the average apicobasal ridge ends at approximately 72% of the height of the tooth crown (Appendix 2).This assessment may even be a slight overestimation, as ridges that ended below the cross section at 50% of the crown height were not able to be included in the sampling pool, and some taxa have developed a number of ridges that disappear beneath this point, including plesiosaurs (NHM 430179; NHM 46239 B), odontocetes (USNM 11962), and crocodilians (USNM 205338).However, as most ridges fail to reach the apex, it is uncertain whether they would increase the second moment of area sufficiently to reduce the chance of buckling. While the stress concentrations generally do not coincide with the locations of ridges, one exception exists in the recurved tooth shape.McCurry et al. (2019) suggest that the lingual clustering observed in many recurved teeth could act to dissipate stress created along the lingual face as a product of the recurved shape.When subjected to the Bite load case in this study, all recurved models do exhibit a concentration of stress toward the lingual face (Fig. 6E-G).This pattern conflicts with the results of the conical models, which display a uniform decrease of stress under the same Bite load case (Fig. 6A-D).Although this could indicate that apicobasal ridges are lingually clustered in recurved forms to alleviate the increased stress, in congruence with the hypothesis of McCurry et al. (2019), the lack of stress reduction in ridged models alludes to another purpose. As this study investigated only the most common tooth crown shapes in aquatic-feeding amniotes, future research could be conducted into the relationships between stress distribution and ridge locations on less common tooth shapes.Such shapes may consist of the multicusped form found in several extinct odontocetes, including Zygorhiza kochii (USNM 11962) and Aetiocetus cotylalveus (USNM 25210 Tooth 1) or the flattened dentition of the crocodilian Bottosaurus (USNM 540758). Evaluating the Strength Hypothesis The strength hypothesis has been widely referenced in the literature on aquatic-feeding tetrapods (Rieppel and Labhardt 1979;Schulp 2005;Young et al. 2014;Zverkov et al. 2018).However, these claims are based on mechanical theory rather than a quantitative assessment of relevant traits.In Varanus niloticus, an extant varanid bearing apicobasal ridges, the dietary transition from insectivory to molluscivory that occurs at adulthood is associated with the appearance of ridges on the tooth crown (Rieppel and Labhardt 1979).This ontogenetic shift has been interpreted as an adaptation to strengthen the teeth in response to a more durable prey source, with the ridges channeling pressure down the crown's surface (Rieppel and Labhardt 1979).As V. niloticus is known to feed predominantly upon aquatic mollusks as well as crabs and amphibians (Dalhuijsen et al. 2014), it is more probable that the ridges evolved to address a challenge in aquatic feeding, such as reduced grip (Chomba and M'Simuko 2013), rather than durophagy-a conclusion that is equally applicable to the taxa studied here. Alternative Hypotheses of Ridge Function With the support of previous research (Sander 1999(Sander , 2000;;McCurry et al. 2019), our results indicate that the improvement of crown strength was unlikely to be a primary function of apicobasal ridges in aquatic-feeding amniotes.Additionally, our findings may shed light on other hypotheses that focus upon the ridges interacting with the prey item, rather than augmenting the crown. Our data dispute the probability of the puncture hypothesis, which argues that apicobasal ridges may have conferred an advantage in predation by acting as a cutting edge, improving the crown's ability to puncture and quickly entrap prey items (McCurry et al. 2019).While ridges around the apex would provide the most assistance during prey puncture (Vaeth et al. 1985;McCurry et al. 2019), our dataset found that most ridges fail to reach the upper quarter of the crown. Supporting many past findings (Buffetaut 2013;Richter et al. 2013;Young et al. 2014), a variety of tooth specimens in the current dataset display an increased prominence or clustering of apicobasal ridges on the lingual surface of the crown, including the pliosaurids Kronosaurus queenslandicus (KKF 0534) and Liopleurodon ferox (NHM 2680), an indeterminate Spinosaurus species (ROM 64659), and the crocodilian Deinosuchus rugosus (USNM 5351) (Appendix 3).The occurrence of clustering upon the lingual face coincides with the region that contacts maximally with prey items (McCurry et al. 2019).This in turn indicates that ridges may have been adapted primarily for prey interaction (McCurry et al. 2019), as serrated or cutting edges in carnivore teeth are commonly located on areas that contact the substrate to aid in processing (D'Amore 2009). The array of characteristics benefiting grip efficiency in aquatic feeders, such as spiricules on the toes of sea eagles (Chomba and M'Simuko 2013) and sharp, elongate teeth in sharks (Bergman et al. 2017), implies a substantial selective pressure for enhanced grasping mechanisms in these predators.The evolution of the forelimbs into flippers or flipper-like appendages in many secondarily aquatic taxa restricts these clades to grasping with only their jaws.As such, any adaptation for grip improvement would be localized to the cranial region and, due to its direct contact with prey items, would almost certainly involve the dentition.Previous physical testing on the scales of male sea snakes established that increased rugosity resulted in maximized friction and grip (Avolio et al. 2006).As this aided the male to maintain contact with the female during copulation in the slippery medium of water (Avolio et al. 2006), the addition of apicobasal ridgesparticularly rugose morphologies such as in Squalodon cheek teeth (USNM 23546)-could equally assist aquatic predators to grip struggling prey items in the same medium.This is further supported by the development of rugose crown apices in some crocodyliform and ichthyosaur taxa (Massare 1987;Fischer et al. 2011), which may demonstrate a selection for convergent grip morphologies in areas without ridges. Impact of apicobasal ridges on tooth strength However, a function in prey handling does not parsimoniously account for ridge configurations that remain symmetrical across the entire crown circumference.Such configurations may instead provide evidence for the removal hypothesis, which proposes that apicobasal ridges functioned like anti-stick grooves in a knife, reducing suction to allow the removal of prey caught in the teeth (McCurry et al. 2019).Wright et al. (1979) argued that the need for removal efficiency would be greater on the mesial face of the tooth due to its initial contact with prey, which would provide an advantage in ridges evolving on regions of the tooth with reduced prey contact in comparison to the lingual surface. Conclusions This study finds that within typical aquatic-feeding amniote dentition, external apicobasal ridges appear to have little impact upon the ability of the tooth to resist biologically relevant loading.The overall distributions of von Mises stress in finite element maps are highly similar between smooth and ridged tooth crown models of identical shape, differing only in a slightly increased stress localized within ridges, while retaining an equal rate of decreasing stress on areas of smooth enamel.These parallel gradients suggest that the higher-stress ridges do not alleviate crown stress.Furthermore, this trend of similarity persists across all tested ridge morphologies, demonstrating that variation in ridge size, shape, and arrangement around the crown is unlikely to influence the tooth's resistance to failure. The factors shaping tooth morphology may be related to the selective pressures imposed by predation, such as the need for increased grip on slippery prey, as the variance in intra-taxon ridge morphology supports previous hypotheses that dental ornamentation is more affected by the functional requirements of diet than by phylogenetic grouping. Our results offer a foundation for future research into other prominent hypotheses of apicobasal ridge function, including improved removal efficiency and enhanced grip.Although our data on tooth crown and ridge morphometrics indicate varying support for these hypotheses across the range of ridge morphologies, further testing is required to explore their biomechanical nuances and return quantifiable evidence.Nevertheless, the absence of stress resistance in our models and the prevalence of lingual clustering in fossil specimens supports the concept that apicobasal ridges evolved for direct contact with prey items, rather than to augment crown robustness.Impact of apicobasal ridges on tooth strength Figure 2 . Figure 2. Standard measurements recorded for fossilized tooth specimens.A, Crown height and position of lateral cross section at 50% of the height (dotted line).B, Crown width measurements (both mesiodistal and labiolingual).C, Apicobasal ridge height and apicobasal ridge width within a cross-section outline (not to scale).Demonstrated on a Bottosaurus sp.tooth crown (USNM 508536). Figure 4 . Figure 4. Finite element solid mesh models demonstrating directions of load cases and constraints.All models are Average Ridges model.A, Bite load case from the crown's apex (arrow) and constraint nodes (in pink and white).B, Shake load case from the crown's apex (arrow) and constraint nodes (in pink and white).C, Pull load case from the crown's apex (arrow) and constraint nodes (in pink and white).D, Distribution of constraint nodes across the base (in pink and white). Figure 5. Linear regression models for apicobasal ridge measurements and tooth crown size in aquatic-feeding amniotes.A, Relative average apicobasal ridge height and crown height.B, Relative average apicobasal ridge width and crown width.Predicted slope of isometry falls outside the visualized datasets. 352 Ailie S. MacKenzie et al. regression slope (b) is less than 1 in both the height (b = 0.094) and width (b = 0.285) regressions.Both averages were calculated from relative measurements. Figure 6 . Figure 6.Finite element analysis (FEA) maps for the Bite load case.Models depict von Mises stress distribution for: A, No Ridges (Conical) model; B, Average Ridges model; C, Large Ridges model; D, Rounded Ridges model; E, No Ridges (Recurved) model; F, Symmetrical Ridges model; G, Lingual Clustering model.All models are depicted in occlusal view.High stress is represented by warmer colors, while low stress is represented by cooler colors.Areas of white denote stress values higher than 1.5 × 10 1 .+V indicates vertical load force has been applied on all models. Figure 7 . Figure 7. Finite element analysis (FEA) maps for the Shake load case.Models depict von Mises stress distribution for: A, No Ridges (Conical) model; B, Average Ridges model; C, Large Ridges model; D, Rounded Ridges model; E, No Ridges (Recurved) model; F, Symmetrical Ridges model; G, Lingual Clustering model.All models are shown in occlusal view.Areas of white denote stress higher than 2 × 10 2 .Arrows indicate direction of applied load force on all models. Figure 8 . Figure 8. Finite element analysis (FEA) maps for the Bite and Shake load case.Models depict von Mises stress distribution for: A, No Ridges (Conical) model; B, Average Ridges model; C, Large Ridges model; D, Rounded Ridges model; E, No Ridges (Recurved) model; F, Symmetrical Ridges model; G, Lingual Clustering model.All models are shown in occlusal view.Areas of white denote stress higher than 2 × 10 2 .Arrows indicate direction of applied load force on all models.+V indicates vertical load force has been applied on all models. Figure 9 . Figure 9. Finite element analysis (FEA) maps for the Pull load case.Models depict von Mises stress distribution for: A, No Ridges (Conical) model; B, Average Ridges model; C, Large Ridges model; D, Rounded Ridges model; E, No Ridges (Recurved) model; F, Symmetrical Ridges model; G, Lingual Clustering model.All models are shown in occlusal view.Areas of white denote stress higher than 2 × 10 2 .Arrows indicate direction of applied load force on all models. Figure 10 . Figure 10.Finite element analysis (FEA) maps for the Bite and Pull load case.Models depict von Mises stress distribution for: A, No Ridges (Conical) model; B, Average Ridges model; C, Large Ridges model; D, Rounded Ridges model; E, No Ridges (Recurved) model; F, Symmetrical Ridges model; G, Lingual Clustering model.All models are shown in occlusal view.Areas of white denote stress higher than 2 × 10 2 .Arrows indicate direction of applied load force on all models.+V indicates vertical load force has been applied on all models.	2024-05-01T15:09:44.840Z	2024-04-29T00:00:00.000	{ "year": 2024, "sha1": "d2cc9c26208f434fe075a8e9de203dbc26ba9213", "oa_license": "CCBY", "oa_url": "https://www.cambridge.org/core/services/aop-cambridge-core/content/view/E18E61EF6DAC60EFE0218B4F520F9305/S0094837324000101a.pdf/div-class-title-the-impact-of-apicobasal-ridges-on-dental-load-bearing-capacity-in-aquatic-feeding-predatory-amniotes-div.pdf", "oa_status": "HYBRID", "pdf_src": "BioOne", "pdf_hash": "9bd4c991b4d0760fe0dcc38b992dcc33910e79f3", "s2fieldsofstudy": [ "Biology", "Environmental Science" ], "extfieldsofstudy": [] }
126599795	pes2o/s2orc	v3-fos-license	Efficient and Secure Mobile Cloud Networking Mobile cloud computing is a very strong candidate for the title"Next Generation Network"which empowers mobile users with extended mobility, service continuity and superior performance. Users can expect to execute their jobs faster, with lower battery consumption and affordable prices; however this is not always the case. Various mobile applications have been developed to take advantage of this new technology, but each application has its own requirements. Several mobile cloud architectures have been proposed but none was suitable for all mobile applications which resulted in lower customer satisfaction. In addition to that, the absence of a valid business model to motivate investors hindered its deployment on production scale. This dissertation proposes a new mobile cloud architecture which positions the mobile operator at the core of this technology equipped with a revenue-making business model. This architecture, named OCMCA (Operator Centric Mobile Cloud Architecture), connects the user from one side and the Cloud Service Provider (CSP) from the other and hosts a cloud within its network. The OCMCA/user connection can utilize multicast channels leading to a much cheaper service for the users and more revenues, lower congestion and rejection rates for the operator. and focus their capital on creating a competitive edge against other products and services. Even though many companies (especially major ones) are reluctant to migrate their data to the cloud [1], non-corporate end users are easily adopting this new trend. Weak privacy and business models are two main aspects delaying wide adoption of computation outsourcing [3]and hindering an investment boom until having all these concerns tackled. With its mobility, reduced latency and increased bandwidth, mobile networks are becoming the network of choice for many corporate and non-corporate users to connect to Internet. Mobile devices are increasingly popular and currently constituting 40% of Internet accessing devices [4].Cloud-based mobile applications are expected to reach 9.5$ billion by end of 2014 [4]. Cloud and mobile computing are performing a successful and very promising tag team which is estimated to dominate the storage and processing traffic over the Internet. Mobile cloud computing is a very strong candidate for the title "The Next Generation Network" which empowers mobile users with extended mobility, service continuity and superior performance. Its main concept lays in offloading a job from a resource-limited mobile device to be executed at the cloud and then receiving the result once done. Users can expect to execute their jobs faster, with lower battery consumption and affordable prices; however this is not always the case. Various mobile applications have been developed to take advantage of this new technology, but each application has its own requirements. Several mobile cloud architectures have been proposed in literature but none was suitable for all mobile cloud applications and this leads to low customer satisfaction for all the customers using the applications not suited by the implemented architecture. Having a mobile cloud architecture suitable for all applications is simply a must, but it is not enough. A valid business model that motivates investors is a vital factor in making any technology reach deployment phase. This fact emphasizes the need for a technology capable of satisfying the requirements of both major players (users and investors) who control the dynamics of market strategy. This thesis targets at proposing competitive mobile cloud architecture based on an interesting business model which motivates investors, especially mobile operators. This architecture should be future-proof by being compatible with horizontal federation, the destiny of maturing industries. This thesis also targets at proposing a multi-layer private authentication and single-sign-on protocol in addition to various privacy applications on the proposed mobile cloud architecture. In the following, the main objectives, challenges and our approaches for achieving this goal are described. Objectives and Workflow Mobile cloud computing is a very promising technology that provides its users a wide range of services and its operators elevated user traffic. This trend is putting the infrastructure providers (mobile operators) face to face with critical problems such as: demanding QoS (Quality of Service), investment optimization, user privacy, scalability etc. We are able to tackle these problems in four interdependent and incremental objectives: Ensuring scalability through federation. Future-proof mobile cloud architecture should consider its scalability, especially with the forecasted increase in traffic. Two aspects should be considered to ensure system-wide scalability which are: Access Network scalability. If the current 5G research initiative succeeded in achieving its targets, access network scalability can be ensured for undetermined period. This aspect is considered outside the scope of this manuscript. "a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources that can be rapidly provisioned and released with minimal management effort or service provider interaction." [5]. "a parallel and distributed computing system consisting of a collection of inter-connected and virtualized computers that are dynamically provisioned and presented as one or more unified computing resources based on Service-Level Agreements (SLA) established through negotiation between the service provider and consumers." [6] [7]. "a large pool of easily usable and accessible virtualized resources (such as hardware, development platforms and/or services). These resources can be dynamically reconfigured to adjust to a variable load (scale), allowing also for an optimum resource utilization. This pool of resources is typically exploited by a pay-per-use model in which guarantees are offered by the Infrastructure Provider by means of customized Service Level Agreements." [7] [8]. "hardware-based service offering compute, network, and storage capacity where: Hardware management is highly abstracted from the buyer, buyers incur infrastructure costs as variable OPEX, and infrastructure capacity is highly elastic." [7] [9]. We believe that cloud computing can be better described by two complementary definitions, one conveying the user's perspective and the other conveying the provider's. So we define cloud computing to be: Online, infinite-like, easily provisioned and pay-as-you-go resource pool that are bound by SLAs. Pool of resource under the management of a centralized entity that is capable of dynamically provisioning it. The user can offload her job (computation, storage, etc.) to the cloud to decrease the utilization at her local devices and to get additional functionalities she doesn't have locally. The advantages and disadvantages of cloud computing as described in [10] are shown in table 1.1. The cloud architecture stack is composed of three main layers [11]: SaaS (Software as a Service), PaaS (Platform as a Service) and IaaS (Infrastructure as a Service) as shown in figure 1.2. Additional service layers are continuously proposed to add supporting features, such as PasS [12] and HuaaS [11]. IaaS is the only presented layer, in this introduction, because of its importance to intercloud federation which is thoroughly discussed in chapter 4. It is the bottom layer at the cloud stack and nearest to the hardware. It enables on-demand provisioning of servers [7]. IaaS offers two services: VRS (Virtual Resource Set) and PRS (Physical Resource Set) [13]. PRS is hardware dependant and creates an abstraction layer to be used by VRS. VRS is hardware independent and used to monitor virtualized applications. One of the most dominant VRS monitor technologies is called hypervisor or VMM (Virtual Machine Monitor) where the virtualized applications are user virtual machines [7]. VIM (Virtual Infrastructure Manager), regardless of its underlying VMM layer, provides the tools for scheduling and managing VMs across multiple physical layers [14]. VRS, PRS, VIM and VMM are shown in figure 1.2. Cloud Manager Layer is still needed to completely define the cloud. It provides cloudlike interfaces and higher-level functionalities for authentication, identity management, contextualization and VM disk image management [16]. The cloud manager layer contains also the federation manager [15] which performs the following abstracted operations: The first operation is selecting a cloud provider (foreign) having its IdP (Identity Provider) trusted by the client's (the cloud experiencing saturation and interested in federation) IdP, then establishing a secure connection after successful authentication. The second operation is transferring the VMs (Virtual Machine) and extending the hypervisor. Mobile cloud computing Mobile cloud has been considered by [17] to be a collection of mobile devices within the same vicinity all having interest in processing the same data. In this case, the processing cost (battery power and CPU cycles) will be divided on the participating devices and hopefully fulfilling mobile cloud's goals. Mobile cloud is controversial since we are expecting one architecture to be adequate for all mobile applications, always decreasing power consumption in mobile devices and most importantly cheap. No architecture found in literature was able to satisfy the above expectation, and none was standardized. Mobile cloud computing has been understood differently by the research community and this explains the deep difference in defining this technology and designing its architectures. To give a generalized definition which includes all points of view and architectures, it should be abstract and doesn't specify detailed features. For this reason, we define mobile cloud computing as: "A technology which allows the user to access cloud services through mobile devices". Although EPS offers considerably low latency and high traffic rates, current standardization efforts are attempting to enhance mobile networks' performance to be able to cope with the increasing user demand. These attempts are mainly focusing on RAN (Radio Access Network) with minor modification on the core resulting in future proof services. In addition to its elevated QoS performance, EPS has the following benefits [19]: Reduced cost per bit. Increased service provisioning. Flexible use of existing and new frequency bands. Simplified architecture and open interfaces. Reasonable terminal power consumption. Some of its functions are [21]: o Contains the authentication information which will be used to authenticate the user during EPS-AKA (Authentication and Key Agreement) as will be shown in the coming section. In figure 1.3, the Red line represents normal data traffic between the mobile user and the destination (other mobile user, Web server, etc.). In case the two communicating parties are connected to the same operator, then the traffic will bounce at PGW without reaching PDN. The Blue line describes the path followed by user plane traffic of the services offered by the operator. The green line describes the path followed by the user generated control plane (signaling) traffic. EPS AKA A telecom operator offers restricted services to its customers and roaming users from other operators (if a roaming agreement exists with the user's home network). The operator has to filter out illegitimate users, and be able to bill those who benefit from the offered services. The procedure for identifying and authenticating users is shown in figure 1.4 [28]. In UMTS and its predecessors, after authentication, the user can only assure "that he is connected to a serving network that is authorized by the user's HN to provide him services; this includes the guarantee that this authorization is recent" [29]. Security network authentication was considered unnecessary in UMTS, since there was an assumption of mutual trust among UMTS operators [30]. In EPS this assumption is considered not valid, thus serving network authentication was added to ensure that the serving network's identity is really confounded with what it is claiming. In other words, the user is connecting to the same network which its home operator has generated the AV (Authentication Vector) to. The protocol responsible for identification and authentication in 3GPP mobile technologies is called AKA (Authentication and Key Agreement). More information on 3GPP EPS AKA is presented next. EPS identifies subscribers permanently using a unique identifier called IMSI (International Mobile Subscriber Identity) [30]. Size of IMSI is 15- Capturing a user's permanent identifier transmitted over the air channel can be used to detect the user's current position in addition to user tracking and other privacy breaching attacks. 3GPP has tried to overcome these attacks by introducing a new temporary identifier, GUTI (Globally Unique Temporary UE Identity), to provide an unambiguous identification of the UE. GUTI is only relevant in the serving MME's area, thus it is sent by the network over a nonaccess stratum layer connection when confidentiality and integrity protected. A user's GUTI is not expected to be constant over a long period of time, since it can then be used in user tracking instead of IMSI. GUTI might be updated in an attach accept message, tracking area update accept message or GUTI reallocation command. This modification is presented in using GUTI instead of previously used TMSI (Temporary Mobile Subscriber Identity). As mentioned above, GUTI is only transmitted when the NAS layer connection is confidentiality and integrity protected. Confidentiality and integrity protection is enabled only after key sharing which is a late step in the access control mechanism. Conceptual steps in access control are: identification, authentication, authorization, key sharing, and finally enabling secure access. It can be seen from the discussed steps that identification occurs before establishing a secure connection, thus user identities are transmitted in plain text. Confidential identification is a very expensive task when compared to confidential data exchange in a security established connection, thus a compromise has to be made between additional costs resulting from ID hiding and high level of privacy. This dilemma faced the designers of GSM, UMTS, and finally LTE. Mobile Authentication The user is identified, authenticated and allowed to use the services 1. -MME: NAS Attach Request (IMSI) A user interested in connecting to a network, if the user has no previous temporary identifier, he has to identify himself by transmitting his permanent identifier (IMSI) in a NAS attach request. If the user has a temporary identifier from a previous connection, he can send his GUTI\|\|LAI/RAI. The S-MME will contact the MME serving the LAI sent by the user, if it succeeded in retrieving GUTI/IMSI couplet it proceeds to step 2, else it requests the user to send his permanent identifier. 2. S--HSS: Authentication Info Request (IMSI, SNID) S-MME retrieves MCC\|\|MNC from IMSI and route the request to H-HSS concatenated with the serving network's ID (IMSI\|\|SNID). It then concatenates its ID to the request send by the user and forwards it to the corresponding HSS. AUTN) The H-HSS fetches for IMSI/K/SQN triplet in its database. A random variable named AV is then sent back to S-MME. MME forwards a challenge towards UE containing RAND and AUTN. UE verifies AUTN to authenticate legitimacy of the serving network. If the request is legitimate RES is generated in addition to the confidentiality and integrity keys. 5. -MME: Authentication Reply (RES) UE replies to S-MME's challenge with RES. S-MME compares RES (from UE) and XRES (from H-HSS), if they are equal then the user is authenticated, keys will be derived, NAS security context will be established and S-MME sends eNB the needed keys to establish AS security context. Organization of this Report The rest of this thesis is organized as follows. Chapter 2 proposes a resource optimizing mobile cloud architecture named OCMCA and rank this architecture against others found in literature to prove its superior performance. OCMCA's business model is also described in this chapter which is based on cloud federation. Chapter 3 proposes an authentication and singlesign-on protocol named EC-AKA3 which creates a trust context between the mobile user and her cloud provider passing through the mobile operator. Other privacy applications of OCMCA are presented in this chapter. Chapter 4 proves mathematically that cloud federation is financially feasible by itself in addition to its critical contribution to OCMCA's business model. This chapter also shows that "distance" is an important selection criterion to maintain the feasibility and profitability of cloud federation and proposes a federation manager to enforce this selection and it is not likely to change in the foreseen future. Mobile devices, either computation/storagelimited (slim client) or power-limited (Smartphone), need a technology that helps in decreasing power consumption, delay and user's cost without compromising the privacy, availability and mobility offered by mobile networks. Mobile cloud computing is on-demand, dynamic and self-provisioned outsourcing of IT resources from a centralized service provider over a mobile access network. It is a technology that offloads resource-intensive applications from resource-limited mobile devices to be processed "somewhere else". This technology aims to decrease mobile's power consumption, allow complex applications to be managed from mobile devices and most importantly keep the expenses within the user's cost budget. Many mobile applications are multicast-based in nature (such as: Google cloud messaging platform [31], multimedia applications and same-content delivery applications like: Google play, Apple store, etc.) but its physical implementation and mobile network's characteristics transform its communication into a group of unicast messages leading to unnecessary congestion, additional delay and more expensive fees. None of the mobile cloud architectures found in literature exploited this weakness but rather adopted it to become an inherited mobile cloud limitation. In order to solve this problem and make mobile cloud computing more agreeable and efficient for both customers and operators, network core adaptations should be considered. In this chapter we discuss mobile cloud from telecommunication perspective by proposing an innovative architecture and interesting business model. This new architecture keeps the mobile operator at the center of mobile cloud computing and offers revenue-making business model that motivates operators to invest in this technology. The rest of this chapter is organized as follows: In section 2.2 we survey existing mobile cloud architectures and applications found in literature. Section 2.3 presents the architecture requirements that are used to measure architecture performance and evaluate its suitability for various mobile cloud applications. It also presents the simulation environment, configuration and results which are analyzed to rank the architectures. Section 2.4 presents our proposed architecture which benefits from the mobile network to offer superior services as shown in its performance comparison with all other architectures. In this section, we also present our proposed architecture's business model which draws the guidelines to offer high profitability and penetration rates. Section 2.5 summarizes this chapter. State-of-the-Art Mobile cloud computing is relatively a new technology having lots of potential applications but no standardized architecture until nowadays. This motivated researchers to innovate different architectures each trying to emphasize on a certain requirement (such as mobility, power consumption, etc.) and select the optimal compromise for the others. In this section, we are going to survey existing mobile cloud architectures and applications found in literature. Mobile Cloud Architectures As Connecting through Wi-Fi interface is also possible but it creates many concerns regarding mobility such as: network availability, handover, etc. An example on the above concerns is shown in the following use case: The user is accessing her CSP using a mobile device over a Wi- This architecture proposes installing cloudlet servers in high density areas (such as: coffee shops, malls, etc.) collocated with Wi-Fi hotspots [33]. The user connects to the cloudlet server and offloads its job to be processed. It contacts the user's cloud service provider to retrieve the user's environment through federation. Starting this point, future jobs will be processed at the cloudlet until getting disconnected (user leaves the hotspot coverage). This architecture is shown in figure 2.3. CloneCloud is an application layer solution, while all other architectures discussed above are physical layer and try to answer different questions (such as: how to connect to the cloud network? is user mobility ensured? how to ensure user confidentiality?, etc.). CloneCloud is complementary to the other architectures, thus will not be included in the performance comparison. Cuckoo [35], various Cloudlet versions [36][37], Spectra [38], Chroma [39], Hyrax [40], MMPI (Mobile Message Passing Interface) framework [41], MobiCloud [42] and MAUI [43] are offloading mechanisms and frameworks for dynamic selection of code partitions (pieces of the program) to be executed remotely similar to CloneCloud so will not be included in the performance comparison. Mobile cloud Applications Computation-intensive mobile applications are the best candidates for being transformed into mobile cloud applications. It has been shown in [44] Application 1 [4]: A foreign traveler tries to understand a street sign by capturing the image using his mobile, extracting the text using an OCR application and finally translating it to an understandable language. Another scenario for the same application was discussed in [17], where a foreign tourist is visiting a museum in South Korea. He is not able to understand an interesting exhibit written in Korean, so he captures the exhibit's image and tries to translate it using an OCR application. Since his mobile device is not able to process the captured image (due to limited or expensive resources such as RAM, Swap space and Internet connection), the application tries to scan nearby devices. Interested nearby devices create an ad hoc network and a virtual mobile cloud to process the image cooperatively. The extracted text is then translated to English. Natural language processing: Is a useful tool for travelers to be able to communicate with locals. Application 2 [45]: Text-to-speech is an application allowing a mobile user having a file to be read to locals Crowd computing: Is a method allowing different video recordings captured by different mobile devices to construct a single video covering an entire event [40]. Application 3 (Lost child) [46]: A five years old John, is attending a parade with his parent in Manhattan. John goes missing and his parents report the incident to the Police, who send out an alert via a message to all mobile phones within two miles radius, requesting them to upload the parade images they have, to a server that only the police have access to. John is spotted in some images and his position was located, and he was reunited with his parents. This application requires high privacy as will be shown in section 3.3.2. Privacy is discussed in section 3.3.1. The participation invitation in this application is multicast/broadcast by nature. Application 4 (Disaster relief) [46]: Electronic maps become useless after a disaster, thus hindering disaster relief teams from performing rescue operations efficiently. Local citizens are asked to use their mobile phones to photograph disaster sites, and upload it to a central server [4]. The collected images are used to create a panoramic view of the sites, thus facilitating the navigation of the relief teams. This application requires high privacy similar to application 3. The participation invitation in this application is multicast/broadcast by nature. Sharing GPS/Internet data [4]:Instead of reading common data from internet or GPS by multiple users, one user can download the data and share it with interested nearby devices through local-area or peer-to-peer networks. This application group helps in decreasing the cost and delay [40] resulting Application 5[4]: Scans using Bluetooth for a co-located device which has a recent GPS reading. Instead of using the expensive GPS connection the needed information can be retrieved from co-located devices. The performance of this application augments in dense events such as: party, football match etc. Other format of this application is: "Traffic Lights Detector for Blind Navigation" [48]. The requestor's communication with other participants is multicast/broadcast by nature. from downloading online information. Application 6 [17]: Instead of downloading a P2P file from the internet over an expensive interface (GPRS, UMTS, LTE etc), a mobile user scans using Bluetooth for a nearby device which has downloaded the needed file and retrieve it over a less expensive interface (Bluetooth). The requestor's communication with other participants is multicast/broadcast by nature. Application 7: We propose a new application which fits under this group. Bicycle fans gather yearly to watch "tour de France", a 23-day racing event [49], taking place mainly in France and other nearby countries. Fans and racers are interested in knowing the instantaneous detailed ranking of racers with additional information (the time difference between the first racer and other contestants, velocity etc.). Each user needs to receive the data on his device, which is transmitted to all application users. The transmitted data is similar to all the users and uses multicast/broadcast by nature. Crowdsensing: Are used to share timestamped sensor readings, such as GPS, accelerometer, light sensor, microphone, thermometer, clock, and compass. Application 8 [4]: Queries the users located in 1 mile radius to get the average temperature of nodes within a mile. The requestor's communication with other participants is multicast/broadcast by nature. Multimedia search [4]: Mobile devices store many types of multimedia content such as videos, photos, and music which can be shared by other users. Application 10 [40]: Multimedia files can be searched in the contents of nearby mobile devices. The query in this application is multicast/broadcast by nature. Social networking [4]:Sharing user content with friends on social media facilitates automatic sharing and P2P multimedia access, thus reducing the need for huge servers to manage this amount of data [40]. Application 11 [40]: Integrating Facebook with mobile cloud to share the active files using mobile interfaces. Crowdsourcing: Is outsourcing tasks, used to be executed by machines or employees, into an external set of people [50][51]. Application 12: "Social search engine" is one of the most popular crowdsourcing application theme which focuses on answering context-related questions using human-help (crowd) instead of or in complementary with search engines [52] [53]. Chacha [54] and Aardvark [55] are two popular crowdsourcing applications that have similar objectives, but for detailed discussion, application 12 will represent Chacha. This application requires high privacy as shown in section 3.3.1. Application 13: "Crowdsourced location-based service" is a method to fetch the recommendations about certain location based categories posted by people with taste and interest similar to the requester. Foursquared [56] is a popular application offering crowdsourced location based service. This application requires high privacy as shown in section 3.3.1. Mobile augmented reality [57] [58]:Is the creation of an information-intensive virtual world confounded with the physical environment allowing the user to "display related information, to pose and resolve queries and to collaborate with other people" [57]. Mobile augmented reality (MAR) is running AR applications in the mobile environment Application 14 [59]: "Google glasses" is a wearable computing platform which allows the person wearing those glasses to view the real word in addition to supplementary requested information. It can be used also to execute certain jobs such as sending messages, translating etc. Wearable computing [57][58]:Accessing information generated (sensor) or received (downloaded) using mini-electronic devices having very limited battery, storage and computation resources. [69]. These applications cannot be profiled, due to the variety of proposed solutions and optimization techniques, thus will not be included in the simulation results. Architectures evaluation In this section, we are going to prove that none of the architectures found in the literature is suitable for all applications and none can be selected as optimal. To do so, we start by showing the application requirements which is used to evaluate the performance of each architecture. We then profile each application into numerical values (application configuration) which is used in the simulation. The network is then configured into numerical values and finally the performance of all the architectures is shown and analyzed. Architecture Requirements MCC (Mobile Cloud Computing) was developed to respond to mobile devices' needs for a technology that helps in decreasing power consumption, delay and user's cost without compromising the privacy, availability and mobility offered by mobile networks, as shown in section 2.1. When designing a mobile cloud architecture several, requirements should be taken into consideration which represents the aim behind using MCC. Similarly, these requirements are used as metrics to evaluate the performance and compare the studied architectures. The requirements are divided into quantifiable and non-quantifiable as shown next: Quantifiable requirements: can be calculated in numerical metrics. The quantifiable requirements are: • • Cost: Financial cost due to network usage. Cost is calculated in terms of "financial units" relative to the mobile fees. An architecture better suits the application if it achieves lower cost. • • Delay: Transmission, propagation and processing delay. In case of in-house processing, it is the time between starting the execution and finishing the job. In case of offloading, it is the time between sending the first offloaded bit till receiving the last reply bit. Delay is calculated in milliseconds (ms). An architecture better suits the application if it achieves lower delay. Non-quantifiable requirements: cannot be calculated in numerical metrics, but will be represented by subjective values. The non-quantifiable requirements are: • • Privacy: Privacy of data during transmission and processing. Privacy is considered high, if no private personal data is stored or processed in nontrusted devices. It is considered medium if private personal data (such as the temperature read by a mobile's sensor) is stored or processed in non-trusted devices, and considered low if private personal data is stored or processed in non trusted devices and can be correlated to the user's id. A cloudlet server is considered non-trusted since it is not under the supervision of a trusted provider. An architecture better suits the application if it achieves higher privacy. Architectures get disqualified if it is not able to satisfy the privacy requirements of an application. • • Mobility: The expected distance to be covered by a user before disconnecting from a service. Cloudlet servers can be accessed through local APs (Access Points) only, then this very short range hinders the device's online mobility especially that we don't expect LAN coverage to be continuous. We consider that Cloudlet offers very low mobility. In virtual cloud, the user has more mobility freedom since not bounded to the fixed position of the AP, but bounded to the group forming the virtual cloud he is connected. We consider this architecture offers medium mobility. The user in "Cloud computing with mobile terminals" has the freedom to move in the entire region covered by his or any another operator (using roaming). We consider this architecture offers high mobility. Since mobility is one of MCC's requirements, we consider that a studied architecture better suits an application if it achieves higher mobility. o Cloudlet: This architecture requires the investment of random business owners in deploying physical devices (cloudlet servers) even with the absence of a valid business model. Cloudlets will face low utilization rates, over-investment and high rejection rates due to the traffic commuting. As residential areas are usually apart from business areas, the cloudlet servers in one area will be severely under-utilized throughout the period the users have commuted to other locations. This requires overinvestment to offer acceptable services in different areas. If some areas are not well equipped, customer rejection rates will be elevated. This architecture is considered to have low scalability. • • Multicast-capable: Boolean value specifying whether the studied architecture is capable of handling multicast traffic efficiently (not transforming it into bulk of unicast traffic). Since the majority of the top ranked mobile applications are multicast-based, having a multicast-capable mobile cloud architecture is of critical importance. Application configuration In this section, we profile the mobile cloud applications into numerical values which are used, in the simulation shown later, as application configuration. Each presented application is profiled using the following parameters: Uploaded data: Size of the uploaded data which will be processed. Unit is byte. Processing: The processing needed in terms of processing units to generate the response. Unit is "processing unit". Downloaded data: Size of the processed data downloaded. Unit is byte. To better simulate the architectures, each application is replaced by its profile and a sample job will be executed. Application profiling is shown in table 2.2. Parameter Value Mobile device (UE) processing speed 1000 unit per second Cloud processing speed (100X faster than UE) "Assumed" (depends on the used UE and VM) 100,000 unit per second Cloudlet Server processing speed [73] 5,000 unit per second UE's power cost for sending 1 bit over Wi-Fi 1 power units UE's power cost for sending 1 bit over LTE [72] 23 power units UE's power cost while waiting for 1 second [73] 30,000,000 power units UE's power cost while computing over 1 second [73] 80,000,000 power units UE's cost for sending 1 bit over Wi-Fi to cloudlet 1 financial unit UE's cost for sending 1 bit over LTE (recorded in Lebanon on 1 -100 financial units 2014, these scores are country/operator related). Some mobile operators offer "lump sum" fees for their local users and elevated expenses for roaming users which are represented by the following range. (1 for local user 100 for roaming user) Wi-Fi bandwidth (upload and download) 100 Mbps The abstract "power cost" values are used to make the simulation "device-independent". To give an insight about real values, HP iPAQ PDA with a 400-MHz Intel XScale processor has the following scores: Simulation Results In It is shown in figure 2.7 that "Virtual cloud computing provider" has been disqualified in different applications (App. 3, 4, 11, 12 and 13) since failed to meet applications' privacy constraints (failed to offer the privacy level required by an application). For the remaining applications, this architecture overcomes "Cloud computing with mobile terminals" in delay and power consumption (for App. 2, 5, 6, 7 and 10) and fails to compete in the same requirements for applications 1, 8, 9, 14 and 15. We can conclude that this architecture suits applications 2, 5, 6, 7 and 10 more than "Cloud computing with mobile terminals". These results will be analyzed thoroughly in section 2.3.5. Table 2.8 studies the third architecture "Cloudlet" against various applications. It is shown in figure 2.8 that "Cloudlet" achieves low scores in privacy, mobility, scalability and delay but relatively good scores in power consumption. Result Analysis In this section, we deduce the strengths and weaknesses of each architecture based on the simulation results. We then rank the best performing architectures across requirements and prove that no mobile cloud architecture is suitable for all applications. We start by comparing the architectures. As for power consumption "Cloudlet" overcomes "Cloud computing with mobile terminals" in applications 2, 5, 6, 7 and 10 which are the applications with "low processing". We deduce that "cloudlet" suits applications with "low processing" and we confirm our previous deduction that "Cloud computing with mobile terminals" suits applications with "high processing". Virtual cloud computing provider vs. Cloudlet When comparing the delay results of these two architectures, we find that "Virtual cloud computing provider" overcomes "Cloudlet" for all applications except applications 8 and 9. These two are the only applications having high processing (100,000 processing units) and very small uploaded data (1KB). As for power consumption "Cloudlet" overcomes "Virtual cloud computing provider" in all applications. We deduce that "cloudlet" results in lower power consumption than "Virtual cloud computing provider". Architecture Ranking In order to rank the cloud architectures for each application, many grading schemes can be used. We are going to use five different schemes just to show that for different preferences no mobile cloud architectures can be considered optimal. The grading schemes are: The first grading scheme orders architectures based on delay. The architectures are ranked in table 2.9, where smaller rank is better. The second grading scheme orders architectures based on power consumption. The architectures are ranked in table 2.10, where smaller rank is better. The third grading scheme orders architectures based on cost. The architectures are ranked in table 2.11, where smaller rank is better. The fourth grading scheme orders architectures based on privacy and mobility. The architectures are ranked in table 2.12, where smaller rank is better. The fifth grading scheme orders architectures based on scalability. The architectures are ranked in table 2.13, where smaller rank is better. "Cloud computing with mobile terminals" becomes more competitive against other architectures in delay as the processing becomes more complex. Oppositely, "Virtual cloud computing provider" becomes more competitive against other architectures in its delay results as the processing becomes lighter. "Cloudlet" architecture offers higher delay for all applications, which is considered a drawback. "Cloud computing with mobile terminals" becomes more competitive against other architectures in power consumption as processing becomes more complex. Oppositely, "Cloudlet" becomes more competitive against other architectures in power consumption as the processed data becomes lighter."Virtual cloud computing provider" architecture offers higher power consumption for all applications, which is considered a drawback. Virtual cloud computing provider "Cloud computing with mobile terminals" achieves best cost performance for local users across all applications except (1, 2 and 15), but worst performance for roaming users across all applications. "Cloudlet" has good performance with applications 5,6,7,8,9,10 and 14 while "Virtual cloud computing provider" performs better with applications 1, 2 and 15. Similarly, it is most private to execute a job in-house, followed by a trusted service provider (Cloud computing with mobile terminals), then a collection of users share the same interest in the processed job (Virtual cloud computing provider) and finally the least privacy offering architecture is "Cloudlet" where data is processed in a server not under the monitoring of a trusted party. "Cloud computing with mobile terminals" and "Virtual cloud computing provider" are considered to have higher scalability than "Cloudlet" which requires investment by third parties even without a valid business model. For "Cloudlet" to be a well deployed mobile cloud architecture, huge number of servers should be implemented to offer suitable coverage. Based on the shown comparisons and grading schemes, we can deduce that no architecture best suits all applications across all requirements. These comparisons are studied from user's perspective i.e. to have lower cost, power consumption and delay in addition to higher mobility, scalability and privacy. Although the mobile operator can impose the greatest influence in implementing MCC, architecture comparison and selection, in literature and throughout this chapter, were focusing on users' needs while neglecting the operator's concerns. "Cloudlet" and "Virtual cloud computing provider", for example, eliminate the use of the mobile operator. "Cloud computing with mobile terminals" significantly increase the traffic leading to higher rejection rates and lower quality for real-time applications, such as voice calls, which are the operator's major profit. The problem in all existing mobile cloud architectures (previously surveyed) is that they discard the financial incentive behind MCC and fail to take advantage of the major traffic type The proposed architecture The decision makers in the technology industry are biased by financial and economic motivations which make their decisions, in many cases, technology unfair. An adopted solution should be technologically and financially feasible and introduces a revenue-making business model. Even a good solution doesn't make it to implementation phase if it discards these motivations, especially if a better solution exists and focuses on promoting economic incentives? The slow progress in deploying IPv6 is a very well-known example on economic effects and market statics that goes even beyond major technology players who are pushing toward this migration. Another example is Fujitsu which has developed a data transfer protocol that overcomes "UDP" and "TCP's" speed by 30 times [74]. Although Fujitsu has achieved impressive breakthrough, they are not expecting it to replace "TCP" or "UDP" because of the absence of a valid business model that persuade the huge investment needed to perform this change (new TCP/IP stacks, APIs, application upgrades, etc.). We have shown in a previous section that the majority of the top ranked mobile applications are multicast-based, and none of the mobile cloud architectures found in literature was designed to take advantage of this mobile traffic characteristic. The weakness in existing architectures is that they tried to meet MCC requirements through network edge solutions (focusing on network edge entities without any modification to the network core) by either trying to eliminate the use of mobile networks (such as "Cloudlet" and "Virtual cloud computing provider") or abstracting it ("Cloud computing with mobile terminals") and in both cases failing to exploit its benefits. At the physical layer of current mobile technology implementations (LTE and UMTS), network layer multicast message is translated into a group of unicast messages (we will call this operation "unicast bundling"). Let BW(n) be the Bandwidth consumed by sending the same data to n hosts, then: From network edge perspective, the transmitted message might be 100% efficient (i.e. sending a multicast message to n hosts), but from network core perspective and the system as a whole the efficiency is (1/n). The current physical layer implementation in mobile technologies does not suit applications using multicast/broadcast communications. We deduce that abstracting the mobile network as "Cloud computing with mobile terminals" proposed is definitely not a future proof solution. We believe that utilizing multicast is not a feature anymore, but a necessity to overcome the drawbacks generated from "unicast bundling". Network edge solutions are not able to force the core network to forward multicast traffic without utilizing "unicast bundling". In order to solve this problem and make mobile cloud more agreeable and efficient, network core adaptations should be considered. When designing our architecture, we seriously took into consideration users' and operators' concerns since we are completely aware that this architecture can make it to production only if it overcomes its competitors in performance and offers a valid business model that convinces investors and decision makers. We also designed this architecture to handle separately both types of traffic (unicast and multicast) to achieve elevated revenues and optimized utilization of resources. In the coming section we propose our architecture titled OCMCA (Operator-Centric Mobile Cloud Architecture) as a response for the previously discussed concerns and a lesson learnt from previous architectures' mistakes. Architecture Description The lessons learnt from section 2.3, corresponds to users' concerns, are summarized as follows: Delay: "Cloud computing with mobile terminals" suits applications requiring complex processing because it is equipped with powerful cloud servers. Oppositely, "Virtual cloud computing provider" suits light applications because processing is executed close to the mobile device. Our architecture uses powerful cloud servers, but these servers should be closer to the mobile device. Power consumption: Similarly "Cloud computing with mobile terminals" suits applications requiring complex processing and "Virtual cloud computing provider" suits light applications. Our architecture offloads the job into powerful cloud servers thus will be suitable for complex applications. By decreasing delay as shown above, the power consumed by the mobile device waiting for the reply will be decreased. Privacy: Executing a job in a trusted environment has higher privacy level. Our architecture positions the cloud servers in a trusted environment to ensure privacy. Mobility: The operator's network coverage and handover mechanisms make a connected user (connected to his services through operator's network) benefit from high mobility. Our architecture uses the mobile network to connect between cloud servers and users. Scalability: Statistical multiplexing proves that unified resources are more scalable, and this is why "Cloudlet" was not. Our architecture uses cloud servers accessible by any connected user (connected to the mobile operator's network) and connected with high bandwidth and low latency link to the operator's core network which makes it scalable. Its scalability can be increased through federation. Our architecture proposes the allocation of a cloud (called "Cloud server" throughout the remaining of this chapter) hosted by the mobile operator (in the mobile operator's premises or in a trusted network as shown in figure 2.5). "Cloud server's" location within the operator network is responsible for the following features: Close to the mobile core in order to eliminate the "delay" resulting from accessing the internet. The decreased distance can be seen in figure 2.5. Connected to the MME allowing it to differentiate between unicast and "multicast" traffic which facilitates separate handling of these two traffic types. Under the supervision of the mobile operator or a trusted party. The operator is also trusted by the user, since an SLA is already signed, thus the "Cloud server" is considered a "private" environment. The cloud is accessible by all connected users, since a direct connection exists with the core network. This feature is responsible for high user "mobility" and participates in increasing the "scalability" of this architecture. The positioning of the "Cloud server" is graphically presented in figure 2.5. When the "Cloud server" receives a request it checks if the data should be unicast or multicast. In case of multicast, "Cloud server" notifies MME which in turn requests the P-GW through S-GW to add the participant to a multicast group. When the user is ready to receive the multicast traffic, the "Cloud server" forwards the data to BM-SC which in turn forwards it back to MBMS GW which is responsible for targeting the needed eNBs. If the operator is providing multicast services (such as mobile TV), then MBMS GW and BM-SC are already installed and could be utilized by our architecture without the need for additional investment. Enabling multicast in LTE allows the local cloud, located within the mobile operator's premises, to transmit data efficiently to the users of one or more cells. Unicast based applications will use the normal "user dedicated channels" to send their unicast data to a specific user. The full configuration of our proposed architecture is presented in figure 2.6. Cloud Server has been carefully positioned, in direct connection with P-GW using the standardized "SGi" interface, to require minimum standardization and development effort. Using this interface, makes the development and integration of "Cloud server" much easier ("SGi" User's Processing Environment For a user to be able to execute her tasks at the "Cloud server" she has to be subscribed with either the mobile operator's cloud or any other cloud provider. In the first case, the user's environment is created and maintained at the "Cloud Server" so enjoying all the features mentioned in the previous section is possible. In the second case, the user's environment is located at the CSP which the user is subscribed to. The user environment has to be offloaded securely to the "Cloud server" which will be able then to offer its services in CSP's place. Offloading user environment is not possible without a trust relationship and financial motivation for the CSP. The most suitable method to establish monitored and accountable trust relationship between cloud providers is "cloud federation". Federation will be discussed in details in section 4.4. Since users are usually connected with the same operator for long durations, we propose that the user's CSP should federate resources at the mobile operator's "Cloud server" and then offload the user's applications and environment settings. In this case the CSP will be able to offer its clients a faster and cheaper service while the mobile operator will be able to serve not just its clients, but any user subscribed with different cloud providers. Enabling federation in OCMCA will lead to higher penetration rates since mobile operators don't need to enter in competition with mature cloud providers to gain market shares, while on the contrary cloud providers will be interested in creating a partnership with the mobile operator due to mutual profit opportunities. In this case, all computation-intensive processing is implemented within the mobile operator's network and the Terminal generated data are offloaded to the local cloud without the need to access the internet (which decreases the cost per bit of the transmitted data). Performance evaluation In this section, we present the simulation results of OCMCA using the same network configuration, application configuration and architecture requirements presented in section 2.3.3 in order to rank our architecture compared to the ones found in the literature. The simulation code is also developed using C# and the results are used to rank the performance by OCMCA. The results are shown in table 2.14. Business Model We are going now to calculate the theoretical cost of application 7 using our proposed architecture and the three architectures discussed in section III. "Tour De France" is in its epic, racers are achieving close times and fans are surrounding the tracks of the race. The fans are viewing the race and keeping an eye on the ranking and timing of each racer, using a mobile cloud application. Let the cost of: 1 bit transmitted from a mobile phone to the "Serving GW" be noted as "A". 1 bit transmitted from the "Serving GW" to a mobile phone over a unicast channel be noted as "B". 1 bit transmitted from the "Serving GW" to a mobile phone over a multicast much cheaper than unicast. 1 bit transmitted from a mobile to another over a Wi-Fi interface be noted as "C". Let the size of the request to get periodically the scores be "rq" bits, and the size of the retrieved data be "rp" bits sent every period T (we will consider the update rate to be once every minute). Let the average number of fans and riders within a Wi-Fi range using application 7 be N. Let the average number of fans and riders within a cell be M. The cost of application 7 over one hour period per user is shown in table 2.16. In "Virtual cloud computing provider" architecture, one user will download the information from internet (downloader) and broadcast it to the remaining users (sharer) using Wi-Fi connection who in turn will resend the data in broadcast. Thus users are noted as "downloader" and "sharer". rq × A + 60 × rq × (B) "Sharer" in Virtual cloud computing provider rq × C + 60 × rp × C + N× 60 × rp × C "Downloader" in Virtual cloud computing provider rq × A + 60 × rp × B + 60 × rp × C Cloudlet rq × C + 60 × rp × C rq × A is the cost to join a certain event. 60 × rq × (B/m) is the cost of receiving 60 updates in one hour over a multicast channel, considering that the scores are transmitted every minute. 60 × rq × (B) is the cost receiving 60 updates in one hour over a unicast channel. 60 × rp × C is the cost of broadcasting 60 updates to nearby users. N × 60 × rp × C is the cost of receiving broadcast updates from nearby users. We are going now to compare the architectures based on financial cost. Wi-Fi is free (C=0), and multicast traffic is m times cheaper than unicast (where m is fixed by the operator). The cost will be in this case: This architecture has good coverage and achieves good profits, but it is not scalable. The number of connected users is limited to "sub-carriers" available at this sector/cell. It also increases the customer rejection rate since the "subcarriers" will be already reserved. Higher customer rejection decreases potential profits and customer satisfaction. Although this architecture achieves low average user cost, but in reality only one user (Downloader) is paying full tariff and the remaining users (sharer) are getting advantage of it. The Downloader's motivation should be studied more for this architecture to implantable. Cloudlet 0 0 Although this architecture seems very attractive for users, it creates both mobility and coverage issues; since most of "Tour De France's" race tracks are not covered by Wi-Fi signals. This architecture can only serve a small portion of the fans. OCMCA B/m M×B/m Usually M > m > 1, this architecture decreases the user cost and increases the operator's profit without affecting potential profits and decreasing customer satisfaction. This win-win situation is definitely interesting for operators. Chapter summary We have shown in this chapter, that OCMCA achieves excellent scores in the majority of architecture requirements and a superior performance against all competitive architectures. It also achieves a business model that motivates investors due to the very promising revenues and limited investment needed (for instance: only software upgrade is needed to run OCMCA if IMS and MBMS servers are already installed at the operator's premises). Since OCMCA can offer IMS as one of its SaaS services, other operator functionalities can be migrated to this cloud such as billing (also known as Intelligent Network), VAS, etc. This opens a huge door on the different models a mobile operator can be such as: Outsourced Services: The telecom vendor or any other CSP can offer VAS services to be rented by small operators that aren't able to deploy all the services in their network. Shallow Operators: An operator can outsource all its core services from a CSP and just maintain its radio network. Virtual Operator: An operator can outsource all its core services from a CSP and rent its radio network from other operators. This result in zero investment needed to join the mobile market. No Operator: A mobile network can be developed from HeNBs (Home eNBs) connected over the internet to a cloud-hosted core network which eliminates the need for fixed line operators. The proposed operator models are all possible, but further research is needed to decide which model is capable of surviving till the deployment phase.	2019-04-23T13:32:53.592Z	2014-12-18T00:00:00.000	{ "year": 2021, "sha1": "c0c1e296ef836c9d037f2d952220c2321f82e9da", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "ed34627c9496618310a1a2acc5450095605fe38d", "s2fieldsofstudy": [], "extfieldsofstudy": [ "Computer Science", "Geography" ] }
207961921	pes2o/s2orc	v3-fos-license	An Internet-Based Therapeutic Tool for American Indian/Alaska Native Adults With Posttraumatic Stress Disorder: User Testing and Developmental Feasibility Study Background Posttraumatic stress disorder (PTSD) is a major public health concern among American Indian and Alaska Native populations. Primary care clinics are often the first point of contact for American Indian and Alaska Natives seeking health care and are feasible locations for trauma-focused interventions. Objective Web-based therapeutic interventions have the potential to reduce PTSD symptoms by offering psychoeducation and symptom self-management tools. We investigated the feasibility of a culturally adapted Web-based therapeutic intervention in two American Indian and Alaska Native–serving primary care sites. We developed and tested a self-guided Web-based therapeutic intervention aimed at improving knowledge and awareness of, and provision of guidance, support, and symptom-management for, PTSD symptoms. Methods A community-based participatory research process was used to refine adaptations to the veteran’s administration’s PTSD Coach Online, to develop new content, and to guide and interpret the results of the feasibility pilot. This process resulted in a 16-guide intervention “Health is Our Tradition: Balance and Harmony after Trauma” website. The feasibility pilot included 24 American Indian and Alaska Natives aged 18 years and older who scored positive on a primary care PTSD screener. Enrolled participants completed a demographic questionnaire, an experience with technology questionnaire, and baseline behavioral health measures. Once measures were complete, research staff described weekly text messages, minimum study expectations for website use, and demonstrated how to use the website. Feasibility measures included self-reported website use, ratings of satisfaction and perceived effectiveness, and website metrics. Feasibility of obtaining measures for an effectiveness trial was also assessed to include behavioral health symptoms and service utilization through self-report instruments and electronic health record queries. Self-reported measures were collected at enrollment and at 6 and 12 weeks post enrollment. Electronic health records were collected from 12 months before study enrollment to 3 months following study enrollment. Changes between enrollment and follow-up were examined with paired t tests, analysis of variance or logistic regression, or the Wilcoxon signed rank test for nonnormally distributed data. Results The culturally adapted website and associated text message reminders were perceived as satisfactory and effective by participants with no differences by age or gender. The majority of participants (86%, 19/24) reported use of the website at 6 weeks and nearly all (91%, 20/22) at 12 weeks. At 6 weeks, 55% (12/22) of participants reported using the website at the recommended intensity (at least three times weekly), dropping to 36% (8/22) at 12 weeks. Participant use of modules varied from 8% (2/24) to 100% (24/24), with guide completion rates being greater for guides that were only psychoeducational in nature compared with guides that were interactive. There were no significant changes in patterns of diagnoses, screening, medications, or service utilization during exposure to the website. Conclusions “Health is Our Tradition: Balance and Harmony after Trauma” shows promise for an effectiveness pilot. Introduction Background American Indian (AI) and Alaska Native (AN) people experience disproportionally higher rates of acute, chronic, and intergenerational trauma than their non-AI/AN counterparts [1][2][3][4], with adverse impacts on physical and behavioral health [1,[5][6][7][8][9]. As the frequency of traumatic events is significantly elevated, there is an increased likelihood that an AI/AN individual will experience multiple traumas compared with national samples [2,[10][11][12]13]. Multiple lifetime traumas have additive risks to developing behavioral health disorders such as posttraumatic stress disorder (PTSD). AI/AN people have low rates of access to behavioral health services, especially those that are culturally responsive. Primary care clinics are often the first point of contact for AI/AN people seeking health care and are feasible locations for trauma-focused interventions [14][15][16]. Given the multitude of medical (eg, chronic pain, irritable bowel syndrome, and autoimmune disorders) [17][18][19] and behavioral (eg, depression, substance use disorders, eating disorders, somatization disorders, personality disorders, and PTSD) [20][21][22][23] health problems that occur following traumatic events, clinicians face multiple challenges in appropriately addressing these conditions and their comorbidities. This is especially critical as some research has suggested that the quality of medical outcomes for individuals with a history of trauma is related to their behavioral health outcomes [24,25]. Although behavioral health outcomes may be particularly important in ensuring good medical outcomes for individuals with a history of trauma, psychotherapeutic treatments for these conditions often involve the use of complex, manualized treatment that many behavioral health clinicians have not been trained to deliver (eg, cognitive processing therapy for PTSD and dialectical behavioral therapy for borderline personality disorder) [26,27]. Not all patient education programs are effective in improving health knowledge and treatment outcomes [28]. Research suggests that, to be effective, patient education programs must be intensive, focused on developing specific skills, and fostering a stronger sense of self-efficacy [28]. Health information technologies (HIT) offer innovative solutions to address patient access to critical health information and to reinforce key components of patient treatment plans. HITs offer multiple communication (eg, websites, apps for mobile phones, and clinical decision support systems) to improve the quality of service delivery [29][30][31], with interactive HIT shown to increase patient knowledge and related health outcomes [32,33] and reduce hospitalizations and health care costs [34]. Web-based therapeutic interventions (WBTIs) are self-or clinician-guided programs that are developed with the aim of positively improving knowledge, awareness, support, and treatment for health problems at low cost. WBTIs are a promising intervention modality for behavioral health treatment support among AI/AN and other indigenous people who have barriers to effective treatment because of access to care as WBTIs can be culturally tailored to address specific clinical issues and the unique characteristics of patient populations. Objectives We undertook a feasibility pilot to develop and test a self-guided, WBTI aimed at improving knowledge and awareness of, and provision of guidance, support, and symptom-management for, PTSD symptoms. This manuscript describes our examination of the website-based intervention's feasibility in 2 AI/AN-serving primary care sites-the Cherokee Nation Health Services (CNHS) and Southcentral Foundation (SCF). Feasibility was assessed by website metrics, use reported by participants, and participant ratings of satisfaction and perceived effectiveness. We also assessed the feasibility of querying data from the electronic health record 12 months prior and 3 months after enrollment and obtaining self-reported PTSD and other behavioral health symptoms at enrollment, at 6 weeks (intervention midpoint), and at 12 weeks. Finally, patterns of change in behavioral health symptoms, diagnoses, and service utilization served as an additional feasibility indicator for an effectiveness pilot. Steering Committee An 11-person steering committee guided each step of the cultural adaptation of the website and associated text messages. A community-based participatory research process was used to refine adaptations and content, and to guide and interpret the results of the feasibility test of the resultant intervention ( Figure 1). The steering committee was composed of individuals from the 2 tribal health study sites and the University of Colorado Denver (UCD). Members of the steering committee had multiple roles within their respective organizations. For instance, 5 individuals were clinical providers (3 at SCF, 2 at CNHS, and 1 at UCD), 6 individuals were community members at their study locations (4 at SCF and 2 at CNHS), 1 individual from SCF held an administrative leader role, and 6 individuals had extensive AI/AN health research backgrounds (4 at SCF and 2 at UCD). The committee included the cross-site Principal Investigator and study coordinator from UCD, and a site Principal Investigator at SCF and CNHS. Under advisement of the Steering Committee, we sought broad community feedback on the intervention content at both SCF and CNHS in 2 phases and verified the intervention materials in a second round of qualitative feedback using methods previously implemented by the research team [35]. Following development and AI/AN cultural adaptation of the website, the feasibility pilot was conducted at CNHS and SCF as described further. Setting and Participants The project settings included 2 of the largest AI/AN-serving primary care facilities in the United States: CNHS and SCF. The Cherokee Nation operates a network of 8 health centers and 1 hospital in the tribe's jurisdictional boundaries, serving more than 130,000 AI patients. CNHS operates the W.W. Hastings Hospital in Tahlequah, Oklahoma, which provides primary care and behavioral health services. SCF is an AN nonprofit health care organization that provides a wide range of health services to over 65,000 AN/AIs from 231 federally recognized tribes in the Anchorage Service Unit. The Anchorage Service Unit is a geographical area extending from the Canadian border on the east to the entire Aleutian Chain and Pribilof Islands on the west, although most SCF patients live near Anchorage, Alaska. SCF operates the Anchorage Native Primary Care Center, which provides primary care and behavioral health services. Each site provides both primary care and behavioral health specialty services and has behavioral health staff embedded within primary care who screen for depression and substance use. The Centers for American Indian and Alaska Native Health at the University of Colorado-Anschutz Medical Campus provided overall coordination between sites. Before data collection, the Alaska Area, Cherokee Nation, and Colorado Multiple Institutional Review Boards approved study procedures. SCF provided tribal approval. Cultural Adaptation Website content was designed to align with national guidelines and local CNHS and SCF preferences for behavioral health interventions. To develop website and text message content, we first completed 2 cycles of 1-hour, semistructured interviews with key stakeholders (providers and clinical administrators) and 2-hour focus groups with providers/administrators and patients in each setting. Leaders and providers were concerned that exposure to website content would be triggering for participants and were concerned about additional health service utilization. Qualitative feedback indicated that a website intervention for PTSD would be welcome with several qualifications, including the emphasis should be on people in context of their community and families, patient modules should be accessible on a website rather than exclusively through mobile technology, and all website information should be consistent with existing clinical recommendations from primary care providers. The adapted content consisted of patient education on the website and the communication of weekly tips via text messaging. On the basis of qualitative feedback, the research team revised the PTSD Coach Online [36,37], originally developed by the Veterans Administration to benefit those coping with stress following a traumatic event. The revision process included consultation with our Steering Committee and SCF's Family Wellness Warriors Initiative (FWWI) [38], an initiative that focuses on providing services for healing from trauma and abuse. The resultant website entitled Health is Our Tradition: Balance and Harmony after Trauma included patient education and interactive healing activities (Table 1) [39]. The Health is Our Tradition: Balance and Harmony after Trauma website included 16 sections referred to as guides [39]. Web page content was either educational only or educational with an interactive activity (eg, drag and drop), an observation/listening activity (eg, audio recorded relaxation exercises), or a written activity (eg, developing a plan for coping with trauma reminders). Video testimonials from AI/AN people who had experienced traumatic events were included in each guide. Web pages used text, programmed interactions, animations, and English language close-captioned videos to present the guide content. Select guides included an audio icon to click for listening to narration of the screen contents. In conjunction with the website, we developed strategic patient communications sent via weekly text-messaged tips ( Table 2). Following the initial intervention adaptations, we conducted a second cycle of qualitative feedback with information used to further refine the adapted website. A professional website development organization was engaged to develop, test, and deploy the adapted website, whereas the research team planned for testing feasibility of its use in the participating primary care clinics ( Figure 2). How to get back in harmony Is there one best way to heal from trauma? The answer is no, there are many different things to try. Some may work better for you than others. Go to the guide called How to get back in harmony in the website to find out more. Week 3 Be in the moment What is a grounding activity and how can it help you feel more safe and stable while healing from trauma? This information and more can be found in the website guide called Be in the moment. Week 5 Identify your values and goals We are looking forward to your midpoint check in visit. Until then, remember to be gentle with yourself. Learning new ways to help yourself heal from trauma can be difficult. Give yourself permission to heal at your own pace and on your own terms. Week 6 Discover your story Feeling like you are alone on your healing journey? Throughout the website, there are real people telling their own stories of harm, healing, and their return to wellness. These stories may help you feel less alone with yours. Week 7 Change feelings by changing thoughts Do you feel stuck with negative thoughts about your past and future? Learn more about ways to change your perspective in the Change feelings by changing thoughts guide. Week 8 Plan something fun Doing something you enjoy can help you feel better. Get some great ideas for planning and doing a fun activity in Plan something fun. Week 9 Relax through breathing; Relax through visualization; Relax your body Feeling frazzled, stressed, or worried? Try a relaxation activity from any of the following guides: Relax through visualization, Relax your breathing, or Relax your body. Week 10 Express your feelings and wants Is there a situation where you feel your voice has been silenced? Working through the activities in the guide called Expressing your feelings and wants may give you a way to break the silence. Week 11 N/A We are looking forward to seeing you for your final research visit when we can celebrate the progress you have made toward balance and harmony. Recruitment and Intervention Procedures Research staff used lobby-based recruitment at both sites to recruit participants for the intervention. Staff recruited and enrolled 12 participants, at each site, as they waited for their scheduled appointments with a primary care provider. Recruitment occurred in winter 2016 through spring 2017. Patients expressing interest at the lobby recruitment table were screened by trained staff for study eligibility and completed the informed consent process with eligible individuals. Inclusion criteria included ≥18 years of age, of AI/AN heritage, diagnosis of one or more chronic conditions (ie, heart disease, stroke, diabetes, kidney disease, arthritis, osteoporosis, cancer, asthma, depression, and chronic obstructive pulmonary disease), and 2 or more visits to the primary care clinic in the past 12 months. Exclusionary criteria included lack of access to a text message-capable mobile phone and a computer with an internet connection as well as cellular phone service for the 3 months following enrollment. Ineligible participants were thanked for their participation with a US $10 gift card and returned to the waiting room for their provider appointment. After completing their primary care provider visit, participants met with a licensed, PhD level psychologist (CNHS) or a Master's level Behavioral Health Consultant (SCF) to complete the Primary Care-PTSD Screen (PC-PTSD) [40] and an assessment assuring the participant was not currently in crisis. Any participants who were in crisis were counseled by the licensed psychologist/Behavioral Health Consultant and referred to additional behavioral health services, if needed. These participants and those who scored negative on the PC-PTSD were thanked for their participation with a US $10 gift card. Participants who were not in crisis and scored positive on the PC-PTSD [39] completed a demographic questionnaire, an experience with technology questionnaire, and the following baseline behavioral health measures: Alcohol Use Disorders Identification Test (AUDIT) [41], Drug Abuse Screening Tool (DAST) [42], AI Symptom Inventory, PTSD Checklist-Civilian Version (PCL-C) [43], Patient Health Questionnaire Depression Scale (PHQ-9) [44], and Patient Health Questionnaire-Somatic, Anxiety, and Depressive Symptoms [45]. Once measures were complete, research staff described the weekly tips (text messages) and the website (Figure 3), including minimum study expectations for its use, and demonstrated how to use the website. Participants were asked to log in and use the website 5-10 min a minimum of 3 times a week over the 12-week intervention period. This expectation was a Steering Committee's recommendation based on clinical judgment as well as Phase 1 interview and focus group data. Finally, research staff scheduled the 6-and 12-week follow-up visits and thanked participants with a US $60 gift card. A weekly tip via text message was sent by researchers to participants during the 12-week intervention ( Table 2). Text message verbiage was devised to reinforce website use, complement website content, and remind participants about follow-up visits. Research staff also called each participant to remind them of follow-up appointments. At each follow-up visit (6 weeks and 12 weeks), participants repeated the experience with technology questionnaire, all baseline measures, as well as a use, satisfaction, and perceived effectiveness questionnaire. Upon completion of each follow-up visit, research staff thanked participants and provided them with a $50 gift card. Finally, a qualified staff member, approved at each site to access electronic health records, did so after the last participant completed the 12-week follow-up. Measures from the electronic health record query were collected for the periods extending to 12 months before study enrollment and 3 months following study enrollment. Demographic Questionnaire At baseline, data were collected on participants' self-reported demographic characteristics. Racial/ethnic heritage was asked using a check all that apply list with the following response categories: AN, AI, white/European-American non-Hispanic, Filipino, other Asian, black/African American non-Hispanic, Hispanic any race, and other. Individuals were asked if they were male or female. Income was assessed by household yearly income before taxes with the following response categories: (check all that apply) with associated response categories: Some high school, but did not graduate, high school graduate or General Educational Development (GED), some college, 2-year college graduate (eg, community college), 4-year college graduate, some graduate school or postgraduate degree, and trade school. Number in household was determined using the question How many people (including yourself) currently live in the home where you live? with the following response categories: 1, 2-4, and 5 or more. Employment status was assessed using the question Which of the following best describes your current status? (check all that apply) with the following response categories: employed full-time, employed part-time, in the military, unemployed and looking for work, unemployed and not looking for work, retired, student, homemaker, disabled or too ill to work, living off the land/a subsistence lifestyle (hunting, fishing, and berry gathering), and other. Use, Satisfaction, and Perceived Effectiveness Questionnaire At 6 weeks and 12 weeks, participants completed 12 multiple choice and 2 open-ended questions with no time frame specified. Use was assessed by asking a frequency question About how often did you access the Health is Our Tradition: Balance and Harmony After Trauma website? with response categories: more than once daily, once daily, 2-3 times a week, once a week, less than once a week, and never. Website use duration was assessed by the survey question On average, how much time did you spend using the website once you accessed it? with response categories: never went on the website, 5 min or less, at least 10 min, at least 30 min, at least 1 hour, more than 1 hour, and can't recall. Barriers to use were assessed by asking Did you experience any of the following difficulties using the website? with response categories: I did not have access to a computer, my computer was not working, I forgot my password, I could not play the videos, I could not play the audio, I could not print, and other problems. Satisfaction was assessed by asking, Overall, how satisfied are you with your experiences with Health is Our Tradition: Balance and Harmony after Trauma? Responses to this item were: never went on the website, quite dissatisfied, indifferent or mildly satisfied, mostly satisfied, or very satisfied. Participants were asked if they would recommend the website to a friend with response categories: yes, definitely; yes, I think so; no, I don't think so; and no, definitely not. Participants were asked if they would use the website in the future in the same manner. The first perceived effectiveness item was If you completed at least one website Guide, did it help you deal more effectively with your traumatic experience and its consequences? Response categories to this item were: yes, it helped a great deal; yes, it helped; no, it really didn't help; and no, it seemed to make things worse. Perceived helpfulness of the text messages and testimonial videos were assessed in the same manner. Whether the testimonial videos were upsetting was also rated as: no, not upsetting; yes, slightly upsetting; yes, moderately upsetting; and yes, very upsetting. Website Metrics Employing industry-standard website analytic tools [46], we tracked visit patterns, frequency of use, and session duration. Types of interaction with each section (guide) in the website were measured (eg, advanced through an entire guide, played a video, entered personal data into an activity, completed a drag and drop activity, and completed a relaxation meditation) per individual. Experience With Technology Questionnaire This 16-item questionnaire, developed by the research team, assesses perceived skill and confidence using computers and peripherals as well as aspects of using the internet such as navigating between pages, installing software such as a flash player, managing popups, and downloading, saving, or printing files from the internet. Items assessing skills used the following stems Can you… or Do you… with binary (no/yes) response categories. Items assessing confidence of technology use had the following response categories: not confident at all; I usually need help; it takes me a while, but I can manage; and confident. Posttraumatic Stress Disorder Screening The PC-PTSD scale consists of 4 binary (no/yes) items assessing PTSD symptoms and included an introductory sentence to cue respondents to lifetime traumatic events. A positive screen is indicated and has been validated in primary care settings, when the sum of positive responses is 3 or higher [39]. Posttraumatic Stress Disorder Checklist-Civilian Version The PCL-C assesses key symptoms of PTSD via 17 items on a 5-point rating scale: (1) not at all; (2) a little bit; (3) moderately; (4) quite a bit; or (5) extremely. The PCL-C is applied generally to any traumatic event with respondents considering how much they have been bothered by the PTSD symptom in the last month. Higher scores indicated more severe symptomatology. Alcohol Use Disorders Identification Test The AUDIT is a 10-item screener for alcohol use in the past year. Higher scores indicated higher levels of alcohol use and problems related to alcohol abuse. Possible responses included never (0), monthly (1), 2-4 times per month (2), 2-3 times a week (3), or 4 or more times a week (4). Patient Health Questionnaire The PHQ measures the severity of somatic, depression, and anxiety symptoms in the past 2 weeks. All items used rating scales. Somatic symptom severity was derived from the first 15 items (PHQ-15). Possible responses to these were not bothered (0), bothered a little (1), or bothered a lot (2). PHQ-15 scores of 5, 10, and 15 represented cut-points for low, medium, and high somatic symptom severity, respectively. The next 9 items (PHQ-9) measured depression symptoms with a 4-point rating scale: not at all (0), several days (1), more than half the days (2), nearly every day (3). These were followed by binary (no/yes) questions and 3-point ratings of frequency of anxiety symptoms: not at all (0), several days (1), and more than half the days (2). A final item assessed the severity of somatic, anxiety, and depression symptoms with possible responses of not difficult at all (0), somewhat difficult (1), very difficult (2), or extremely difficult (3). Higher scores indicated more severe anxiety/depression symptomatology. Drug Abuse Screening Tool The DAST is a 10-item screener for drug use in the past 12 months. Items were binary (no/yes); higher scores indicated higher levels of problems related to drug abuse. American Indian Symptom Inventory The American Indian Symptom Inventory was developed for this study [5], specifically for use with AI/AN people. Drawing on items from such common measures as the Symptom Checklist-90 [47], as well as items suggested by focus groups as highly relevant to psychological distress among AI/ANs, this 50-item measure assesses a wide range of symptomatology prefaced by the statement, "Here is a list of problems people may have. How much have any of these problems bothered or upset you during the past month, including today...Not at all (0), Some (1), or A lot (2)." Electronic Health Records Electronic health record data collected included date of birth to calculate age; any depression and substance use screening scores; diagnoses of behavioral and trauma-related physical health disorders (eg, International Classification of Diseases-10 codes); prescribed behavioral health medications; and number of visits to primary care, emergency room/urgent care, inpatient, other clinical/ambulatory, and behavioral health clinics. Data Analysis Using SAS version 9.4 (SAS Institute, Inc), analysts created dummy variables for check all that apply questions (race, education, and employment) and examined frequencies, means, and standard deviations for these demographic responses as well as sex, number in household, and income. Chi-square, Fisher exact, or t tests were calculated to compare participant demographic characteristics by site. Changes in use, feasibility, and perceived effectiveness between the 6-week and 12-week follow-ups were examined with paired t tests. To examine the potential relationship of the intervention with service utilization, analysts compared clinic visit frequency from the health record data in the 12 months before the intervention (divided by 4 for comparison) and the 3 months after it with paired t tests. Patterns of symptomatology over the 12-week intervention was explored through one-way analysis of variances for repeated measures or repeated measures logistic regression tests. The Wilcoxon signed rank test for nonparametric data was used to test significance for nonnormally distributed data. Participant Demographics A total of 24 participants were enrolled in the feasibility pilot at baseline (Table 3). Their mean age was 49 (SD 14) years, and 71% (17/24) were women. Participant data from SCF (n=12) and CNHS (n=12) were compared. The 2 samples were similar in most regards, except CNHS participants were more likely to have attained a 4-year college degree or greater (P=.04; Fisher exact Test), and be employed full time or part time (Χ 2 1 =4.2; P=.04). One individual at each site was lost to follow-up before the 6-week measures resulting in a retention rate of 92% (22/24). Intervention Use, Satisfaction, and Perceived Effectiveness The majority of participants 86% (19/24) reported any use of the website at 6 weeks, and nearly all 91% (20/22) used it at 12 weeks. However, participants did not use the website with the intensity requested by the researchers. Twelve of 22 participants reported that they used the website at the recommended intensity (a minimum of 3 times per week) at 6 weeks. At 12 weeks, use at recommended intensity dropped to 36% (8/22) ( Table 4). Twenty of 22 participants reported reading the weekly tip text messages at 6 weeks. Of these 20 participants, 75% (15/20) considered them as at least moderately helpful. Similarly, 96% (21/22) reported reading the messages at 12 weeks, and 91% (19/22) considered them at least moderately helpful. At both 6 and 12 weeks, the majority of participants were mostly or very satisfied with the website, would recommend it to a friend in need, and reported that the tools in the website helped them cope with their trauma and its consequences. Testimonial videos were viewed by 77% (17/22) at 6 weeks and 82% (18/22) reported viewing at 12 weeks. In all, 77% (13/17) and 83% (15/18) reported feeling that the testimonials were at least moderately helpful at 6 and 12 weeks, respectively. Out of 17 participants, 4 (24%) noted that testimonial videos were moderately to very upsetting at 6 weeks, a rate that dropped to 0% at 12 weeks (t 14 =2.36; P=.03). Out of 22 participants, 20 (91%) reported at each time point that they would use the website again in the future. Difficulties in using the website included loss of the website address and/or password and slow internet connection speeds. One participant reported computer access/performance issues. A research team member spoke with the participant in the first week of the intervention and was able to address these issues, allowing this participant to use the website. Website Usage A total of 158 unique website sessions were recorded in the 12-week study period. The average duration of a session per participant was 16 min. Although all guides in the website were completed at least once, some were completed by more participants ( Table 5). The 4 guides labeled for first time visitors were completed at rates ranging from 42% (10/24 participants; What can you do for yourself while healing?) to 100% (24/24 participants; What is trauma?). All 4 first time visitors guides were only psychoeducational in nature and did not include interactive activities. Among those for returning visitors, guides that did not include written exercises (Be in the moment, Discover your story, Notice your thoughts and feelings, Relax through breathing, Relax through visualization, and Relax your body) were completed by a minimum of 3 out of 24 participants (13%) and a maximum of 18 out of 24 participants (75%; Be in the moment). Guides that included written exercises (Change feelings by changing thoughts, Change negative thinking patterns, Cope with trauma reminders, Develop healthy relationships, Express your feelings and wants, and Identify your values and goals) were completed less frequently. The least frequently completed guide was Develop healthy relationships. Out of 24, 7 participants (29%) completed it. The most frequently completed guide was Discover your story, which was completed by 11 out of 24 participants (49%). Overall, completion rates among guides that were only psychoeducational in nature were greater (t 18 =−2.90; P<.01) than those that were interactive (guided relaxation, checklists/click or drag activities, and written exercises). There were no significant differences in completion rates relative to the content length of the guide. Feasibility of Measures and Indicators of Effectiveness Active participants at both sites were able to complete all self-report measures at baseline, 6 weeks and 12 weeks. Analyses of patterns of symptoms found reductions in self-reported PTSD, depression/anxiety/panic, physical symptoms related to PTSD, and problematic alcohol use between baseline and the 12-week follow-up. Depression symptom severity decreased between baseline and the 12-week follow-up visit. Both sites were also able to query all data elements from electronic health records. There were no significant changes in patterns of diagnoses, screening, medications, or service utilization during exposure to the website. Principal Findings A culturally adapted website for AI/AN people reporting symptoms of PTSD was developed and assessed for feasibility in 2 large AI/AN primary care settings. The website was well received with no difference in use, satisfaction, perceived effectiveness, or technical skills needed for use by age or gender. However, only half of the participants reported using the website at the recommended intensity at the 6-week follow-up and just over one-third did so at the 12-week follow-up (Table 4). Although website usage was only moderate and decreased across the study period, participants reported that the website was helpful in coping with their PTSD symptoms and related problems, and some self-reported symptoms decreased between enrollment and 12 weeks. Leader and provider concerns voiced during Phase 1 interviews and focus groups that participants may be unduly triggered by the website content, clinically decompensate, or have other negative impacts from using the website on their own were not supported by the data. Collecting measures via self-report and electronic health record queries was successful, and a larger effectiveness trial is also warranted and appears feasible. Our participant attrition of 8% (2/24) is notably lower than the sizable attrition rates reported in self-administered internet interventions for depression, anxiety, and PTSD [48,49]. The weekly text-messaged tips may have contributed to retention and intensity of use by reminding participants about the study, their follow-up appointments, and highlighting content that may have sparked their interest. Although internet connectivity and speed were concerns for this population, very few participants reported these items as barriers to website access and usage. Participant website usage occurred throughout the 12-week intervention period with psychoeducational and shorter interactive activities being completed by more participants compared with worksheet-based activities requiring data input by participants. These findings align with preliminary evaluation of the PTSD Coach Online app, where participants indicated psychoeducation and self-management components as moderately helpful with symptom self-management as the most useful functions of the intervention [37]. Limitations As a small pilot study with a convenience sample intended to assess feasibility, we cannot determine whether the culturally adapted intervention is effective in reducing posttraumatic and related symptoms. Key limitations of this study include its small sample size, lack of a control group, and lack of clinician-administered outcome measures. Calculations of website use via Google Analytics was limited as we could not link individuals to specific usage data owing to confidentiality and feasibility constraints. Text message reminder delivery, reading, and potential impacts on website usage were not electronically tracked. Finally, in the review of the AUDIT and DAST measures, we note that these measures assessed past year use rather than current use. Thus, we recommend that these measures be edited or reconsidered in a future effectiveness study. Although not many participants found the video testimonials upsetting, additional exploration of the degree of upset and content provoking upset is indicated. Intensity of use required for maximum therapeutic benefit should be assessed as well as potential order effects. The last guide was specifically developed, given feedback from Phase 1 participants about how trauma impacts important relationships, yet viewed by only 2 participants. Future research may be enhanced by adding clinician support to compare the effectiveness of the self-management WBTI alone compared with clinician-supported use. Additional usage log data collection and analysis methods may also be employed to assess more detailed activity of participant website usage behavior. Assessment of the effectiveness of text message reminders to reinforce website usage and effectiveness is warranted. Conclusions Despite these challenges and limitations, the results indicate Web-based therapeutic support of AI/AN adults with posttraumatic symptoms is feasible and warrant a large-scale randomized control trial to assess its potential effectiveness. The website appears to be a promising intervention given feasibility testing in 2 large, real-world AI/AN health systems. The intervention could be easily implemented by behavioral health staff integrated into primary care settings and was satisfactory to and viewed as helpful by AI/AN adults in this study.	2019-09-17T03:08:21.291Z	2019-11-13T00:00:00.000	{ "year": 2019, "sha1": "1d5ba9ef44fb1da079deb5d54cf661be8d86f003", "oa_license": "CCBY", "oa_url": "https://formative.jmir.org/2019/4/e13682/PDF", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "d7a55bf58eed40f83d1b9a331abe1826ed0d1f1a", "s2fieldsofstudy": [ "Medicine", "Psychology" ], "extfieldsofstudy": [] }
251911696	pes2o/s2orc	v3-fos-license	Stage-dependent prognostic shift in mismatch repair-deficient tumors: Assessing patient outcomes in stage II and III colon cancer Introduction Deficient mismatch repair (dMMR) or high microsatellite instability (MSI-H) is associated with an improved prognosis in colon cancer stage II but poor prognosis in stage IV colon cancer. The clinical significance of dMMR in colon cancer stage III is not established. Methods Tissue microarrays (TMAs) from 544 patients with colon cancer stage II and III with clinicopathological and survival data were stained for mismatch repair (MMR) proteins, CD3, CD8, and programmed death ligand-1 (PD-L1), and programmed death ligand- 1 (PD-L1). Patient outcomes were reviewed. Results In stage III colon cancer, dMMR was a marker of poor disease-free survival (DFS) (Kaplan–Meier, mean survival in months: dMMR: 28.76 (95% CI 18.46–39.05) vs. pMMR 40.91 (37.20–44.63), p=0.014, multivariate Cox regression: hazard ratio (HR) 4.17 (95% CI 2.02–8.61), p<0.001). In stage II colon cancer, there was a tendency toward improved DFS for dMMR patients (dMMR: 57.14 (95% CI 54.66–59.62) vs. pMMR 53.54 (95% CI 51.48–55.60), p=0.015, multivariate Cox regression HR 0.24 (95% CI 0.06-1.04), p=0.057). CD3, CD8, and PD-L1 expression was not associated with prognosis of dMMR patients. Multivariate Cox regression analysis showed a significant interaction between the MMR phenotype and stage (p=0.001). Conclusion dMMR is associated with an improved prognosis in stage II colon cancer but is no longer associated with a favorable prognosis in stage III colon cancer. Introduction Deficient mismatch repair (dMMR) or high microsatellite instability (MSI-H) is a marker of improved prognosis in stage II colon cancer (1). This phenomenon has been attributed to the beneficial immune response associated with dMMR tumors (2). However, in stage IV colorectal cancer (CRC), several studies demonstrate that a dMMR phenotype is a marker of poor prognosis (1,3,4). In stage III, MMR phenotyping is currently only used to detect Lynch syndrome and the effect of dMMR on prognosis is controversial (5). The reason for the contrasting prognostic impact of the MMR phenotype in different stages of colon cancer is not established. A dMMR phenotype gives rise to a high mutational burden, particularly insertion and deletions (indels), which again leads to an increased expression of neoantigens and higher density of tumor-infiltrating lymphocytes (TILs) (6, 7). A high infiltration of CD3+ (pan Tcell marker) and CD8+ (cytotoxic T-cell marker) cells is viewed as a manifestation of a favorable immune response and a beneficial prognostic marker in colon cancer (8). Tumors might evade their immunosuppressive environment by exploiting existing immune checkpoints that in normal physiology are used to maintain self-tolerance. Programmed death-ligand 1 (PD-L1) is a transmembrane protein typically expressed on hemopoietic and antigen-expressing cells. During immune regulation, PD-L1 binds to programmed death-1 receptor (PD-1) on cytotoxic T cells and other immune cells. It has been suggested that increased PD-L1 expression in tumor cells represents an escape from immune surveillance, allowing for the spread of tumor cells and decreased cancer-specific survival in dMMR tumors (9). Few studies have described the prognosis of dMMR in stage III colon cancer. Improving the prognostication of dMMR colon cancer is highly warranted to identify patients at risk of colon cancer relapse that might benefit from immunotherapy. The main goal of this study was to analyze the prognosis of dMMR stage III colon cancer and study the prognostic interplay between MMR status, tumor cell PD-L1 expression, and density of CD3+ and CD8+ lymphocytes. Patient cohorts The study cohorts have been described previously (10). It includes tissue from the primary tumors of all included patients, 544 in total, with colon cancer stage II and III from two clinical studies. The 276 patients from the Norwegian Gastrointestinal Cancer Group (NGICG) material were included between 1993 and 1996 (11). They were randomized to receive adjuvant chemotherapy with fluorouracil/levamisole after surgery or to surgery only. The remaining 268 patients were included from the The Haraldsplass Deaconess Hospital (HDH)-material; a population-based cohort with colon cancer patients recruited from 2007 to 2011 (12,13). The patients from the NGICG and HDH series have similar clinicopathological characteristics, as shown in Table 1. None of the patients received immunotherapy or radiotherapy. The molecular biomarkers analyzed in this study were not used for treatment selection. Median follow-up time was 5 years for both the HDH cohort and the NGICG cohort. Tissue microarrays and immunohistochemistry Immunohistochemistry (IHC) was performed on tissue microarrays (TMAs) containing tissue cores from primary Scoring of tissue microarrays 2.3.1 CD3 and CD8 The densities of CD3+ and CD8+ T lymphocytes were both assessed within the tumor margins (intraepithelial TILs) and stroma <100 mm from the tumor margin (stromal lymphocytes). In cases of intratumoral heterogeneity, the cylinder with the highest density of lymphocytes was included in the analysis. Two of the authors (MPM and KEH) scored the TMAs. The scores were semiquantitative and based on the density of positive T cells. We divided the intraepithelial lymphocytes into four groups based on the density of positive lymphocytes: 0 = no positive cells, 1= low, 2 = high, and 3 = very high density. For the stromal lymphocytes, we observed that most cases had a low density of lymphocytes, and few cases had a very high density. The score was therefore made from 0 to 100 to detect variations in the lower segment. For presentation in Kaplan-Meier plots, the stromal scores were divided into four groups based on score quartiles. A total TIL density score was computed for multivariate analyses. For the stromal scores and the intraepithelial scores to contribute equally to the combined score, the stromal scores were divided into four groups based on quartile scores and the given score 0-3. The sum of the intraepithelial CD3+ TILs (0-3), intraepithelial CD8+ TILs (0-3), stromal CD3+ TILs (0-3) and stromal CD8+ TILs (0-3) were used as the total TIL score (TIL density score, 0-12). Programmed death ligand-1 PD-L1 expression was assessed by authors KEH and MPM as recommended by the Agilent Dako Interpretation Manual (originally made for NSCLCs). This scoring method is compatible with the tumor proportion score (14). PD-L1 expression was divided into three groups based on the percentage of PD-L1-positive tumor cells: <1%: no PD-L1 expression. 1%-49%: positive PD-L1 expression, ≥50%: highly positive PD-L1 expression. Positive staining was defined as partial or complete cell membrane staining of any intensity that was perceived distinct from cytoplasmic staining. Cytoplasmic staining was excluded from the scoring. Only viable tumor cells were scored. Immune cells, normal cells, and necrotic cells were excluded from the scoring. Staining in tumor-associated immune cells was recorded separately. Mismatch repair protein expression scoring Negative MMR protein staining was defined as <5% positive tumor cells in the presence of positive staining in internal positive control cells in the same tissue core (normal colon epithelium or stromal cells). The MMR protein staining was nuclear. Cores with negative staining for both tumor cells and internal control cells generated a missing staining result. MMRdeficient cases (dMMR) were defined as cases with negative staining for MLH1 + PMS2, MSH2 + MSH6, MSH6 alone, or PMS2 alone. Cases with positive MMR protein staining were defined as MMR proficient (pMMR). We have validated the MMR protein staining in whole tissue sections (10). Statistics Disease-free survival (DFS) was defined as time from surgery until recurrence. Overall survival (OS) was defined as time from surgery until death of any cause. Two-sided p-values are reported and values <0.05 were considered statistically significant. P-values were not adjusted for multiple testing. Cox regression models were fitted using the enter method with the clinically relevant factors included in the analyses. The MMR phenotype-by-stage interaction was tested by adding a crossproduct term of indicator variables for the MMR phenotype and stage to the Cox regression model. The assumption of proportional hazards was assessed by the inspection of log-log HR (hazard ratio) and HR of Cox regression models stratified on the categorical variables and Schoenfeld residuals (continuous variables). Differences in means was tested using the T test (for normally distributed variables) or Mann-Whitney U-test. Median follow-up time was calculated by the reverse Kaplan-Meier method. Statistical analyses were performed using IBM SPSS Statistics for Windows (v25.0) Ethics Study protocols were approved by The Regional Committee for Medical Research Ethics of Western Norway and the Data Inspectorate for National Registries (REK 1992-55.92, REK 15666). All patients signed informed consent. Missing staining results were due to technical issues: insufficient staining in cells used as internal positive controls (MMR proteins), necrosis, tissue core detached from slide during staining, or few tumor cells available for scoring. Goodness-of-fit analyses have been made for all markers [Supplementary and (10)] and showed no statistically significant differences in the distribution of clinicopathological variables between the original study cohort and patients with valid staining results. The number of examined lymph nodes and number of positive lymph nodes for stage III dMMR and pMMR tumors were assessed from pathology reports and were found not to be different in dMMR patients versus pMMR. For stage III patients in the HDH cohort, the mean number of examined lymph nodes was 15.92 (SD 3.4) for patients with dMMR versus 16.89 (SD 5.4) for patients with pMMR (p=0.53). The mean number of positive lymph nodes for stage III was 2.38 (SD 1.9) for dMMR tumors and 3.14 (SD 2.3) for patients with pMMR (p=0.23). The mean number of examined lymph nodes for stage III colon cancer in the NGICG cohort was 8.00 (SD 4.6) for patients with dMMR versus 9.73 (SD 6.8) for patients with pMMR (p=0.68),. The mean number of positive lymph nodes for stage III was 3.67 (SD 3.2) for dMMR tumors and 3.30 (SD 4.1) for patients with pMMR (p=0.48). Mismatch repair proteins and prognosis MMR deficiency was significantly associated with lower stage, CDX2 negativity, right-sided cancer, mucinous-or signet-ring histology, PD-L1 positivity, and a high density of CD3+ and CD8+ TILs, both stromal and intraepithelial ( (Figure 2). These data show that dMMR is a marker of poor prognosis in stage III colon cancer but a marker of improved prognosis in stage II. Effect of tumor-infiltrating lymphocytes High intraepithelial CD3+ and CD8+ scores and high stromal CD3+ and CD8 + TIL scores were all individually significantly associated with a lower stage, CDX2 negativity, right-sided cancer, the dMMR phenotype, and PD-L1 positivity ( Table 2 and Supplementary Table 1). In addition, there was a slightly higher median infiltration of stromal CD3+ TILs in patients aged 67 years or older compared to younger patients (median density 25.0 vs. 20.0, p =0.005, Supplementary Table 1). The association between TIL density and age was not observed for CD3+ intraepithelial TILs or CD8+ TILs. TILs scores were all strongly associated with each other (p <0.001 for all comparisons, Supplementary Table 1). For pMMR tumors, the total TIL density score was significantly lower in stage III, compared to stage II ( Figure 3C). In contrast, dMMR tumors had a high total TIL score for both stage II and III. The total TIL score was associated with survival for pMMR tumors but not for dMMR tumors (Figures 3A, B). These data suggest that the prognostic difference between dMMR stage II vs. stage III tumors is not explained by a difference in the density of TILs. Prognostic shift of deficient mismatch repair cancers in multivariate Cox regression model Cox regression analysis models were made for both DFS and OS. The univariate Cox regression model for all variables assessed separately is found in the supplementary material (Supplementary Table 4). To determine whether the prognostic effect of MMR depends on the stage, our multivariate Cox regression models included a stage*MMR interaction term. In addition, they were run as split models layered by stage to assess the effect of MMR in stage II and stage III separately ( Kaplan-Meier curves comparing disease-free survival (DFS) and overall survival between dMMR and pMMR cases for stage II and stage III colon cancer. P-values calculated by the log-rank test. patients, there is a tendency toward an improved prognosis for patients with dMMR for DFS (HR 0.24 (95% CI 0.06-1.04) p=0.057) but not for OS (HR 0.80 (95% CI 0.37-1.73), p=0.571). As the indications for administering adjuvant chemotherapy and surgical methods changed between the time of including patients in the two different studies, we adjusted for these differences by adding the variable "cohort" in our model. Decreasing the density of TILs and low CDX2 expression were negative prognostic markers for DFS but not for OS. PD-L1 expression was not associated with prognosis. Our multivariate Cox regression models support that the prognostic impact of dMMR differs between stage II and III. Discussion Our study demonstrates that the prognostic effect of dMMR differs between colon cancer stage II and stage III. In this study, dMMR was a poor prognostic marker in colon cancer stage III, both for DFS and OS. In stage II, dMMR was a marker of improved prognosis for DFS but not for OS. This stagedependent difference remains significant when adjusting for the influence of TIL density, PD-L1 expression, CDX2 expression, and administration of adjuvant chemotherapy. The favorable prognosis associated with dMMR in stage II colon cancer is supported by several studies (15). According to current treatment guidelines, dMMR stage II colon cancer represents a low-risk group of patients who do not need adjuvant chemotherapy. Still, high-risk T4 stage II patients should be considered for adjuvant chemotherapy regardless of the MMR status because of their increased risk of relapse (5). In our study, dMMR in stage II was a marker of improved DFS but not OS. Stage II colon cancer patients have a good prognosis, and many patients are diagnosed at an older age. We therefore believe that DFS is a better reflection of the biology of dMMR stage II tumors than OS. Decreasing percentages of dMMR colon tumors are seen with advancing tumor stage (16). Still, there is increasing evidence that dMMR cancers that do progress to stage IV represent tumors with a more aggressive biology compared to pMMR cancers. For stage IV colorectal cancer, several studies including the consensus molecular subtypes (CMS) classification report that dMMR is a poor prognostic marker (3,17,18). Assessing the prognosis of dMMR stage III patients might help us understand at what point in the tumorigenesis that dMMR goes from a favorable to an unfavorable biomarker. However, results from studies assessing the prognosis of stage III dMMR colon cancers are discrepant. Some report that the dMMR status conveys an improved prognosis in stage II and III analyzed together (19)(20)(21)(22), and some report dMMR as a marker of improved prognosis in stage III specifically (1,20,21,23). Other studies report that the prognostic impact of dMMR in localized colon cancer differs between stage II and stage III. In a retrospective single-center cohort consisting of 1,250 patients, Mohan et al. demonstrated a worse DFS for stage III patients with dMMR compared to stage III pMMR patients. In stage I and II, dMMR patients had an improved DFS. Their results indicate a prognostic switch in dMMR colon cancer in line with our findings (24). In other studies, there is a favorable prognostic impact of the dMMR phenotype in stage II but the prognostic effect of dMMR is less prominent or lost in stage III dMMR colon cancer. In a single-center consecutive population-based cohort, the dMMR status conveyed an improved outcome in stage II colorectal cancer but did not impact survival in stage III (25). Klingbiel et al. studied 1,254 patients from the PETACC trial. They found that the positive prognostic effect of dMMR was stronger for stage II patients than for stage III patients. In addition, they reported a statistically significant interaction between the stage and the MMR status (16). Cohen et al. proposed that the impact of dMMR on the prognosis of stage III patients depends on the extent of the lymph node metastasis. In the report by Cohen, dMMR was a positive prognostic factor in stage II patients. In stage III, dMMR was positive prognostic factor in N1 disease but not in N2. In fact, the N2 dMMR group had a worse survival than the N2 pMMR group the first two years after adjuvant treatment but similar long-term survival (26). The improved prognosis of dMMR patients in stage II has been largely explained by a beneficial immune response. The density of tumor-infiltrating CD3+ and CD8+ lymphocytes has been recognized as a prognostic marker in colon cancer (5) and highly correlated with dMMR (27). In our study, TIL density was a strong prognostic marker in the pMMR subgroup. There was no significant difference in the density of TILs between stage II and III dMMR tumors. Other studies report that TIL density impacts the prognosis of both pMMR and dMMR CRC (28)(29)(30). Having a limited number of dMMR tumors with a low density of TILs, our series might be too small to assess the prognostic effect of TIL density in dMMR tumors separately. The prognostic effect of PD-L1-expression in colon cancer is not fully established (31). Rosenbaum et al. demonstrated reduced cancer specific survival in dMMR patients with positive PD-L1 expression (9). Other studies report a better prognosis for patients with positive PD-L1 expression, especially in the pMMR subgroup (32). We were not able to demonstrate a prognostic impact of PD-L1 expression in any subgroups. Limitations to this study mainly include the use of TMAs and the lack of a validation cohort. The use of TMA can produce inadequate results if the utilized marker is heterogeneously expressed. The density of TILs is considered a robust marker (8). Although the immunoscore was originally developed for assessment in whole tissue sections, strong correlations between the density of TILs observed in TMAs versus whole tissue sections has been reported in several studies, supporting the use of TMAs to evaluate TIL densities (7,33,34). PD-L1 can be heterogeneously expressed, and it was assessed in TMAs only in our study. Still, by setting a low threshold for PD-L1-positive cases (1% positive tumor cells), we believe that we have detected the cases with clinically relevant PD-L1-expression. In this study, the MMR phenotype was assessed by immunohistochemistry, using an MMR panel with four proteins to increase sensitivity (35). This method has a high sensitivity for the detection of dMMR and high concordance with results from PCR-based MSI testing (36). Although MMR proteins are considered to be homogenously expressed with either complete positive or complete negative staining of tumor cells, heterogenous immunohistochemical staining patterns have been reported (37). We therefore validated our MMR protein immunohistochemistry in whole tissue sections in our previous study (10). The number of published articles specifically evaluating the prognosis of stage III dMMR colon cancer is low. Therefore, it is our opinion that this study adds important information about prognostication in this group of patients. In previous literature on this subject, the prognosis of localized colon cancer is often assessed in stage II and III as one group. However, the present study shows that the prognostic impact of dMMR depends on the stage. Our study would benefit from an external validation, and we hope that future studies will further assess this issue. Having a limited number of patients did not allow for subgroup analysis to assess how the prognostic effect of dMMR affects stage III patients with N1 disease versus patients with N2 disease as discussed in the article by Cohen et al. As there was no significant difference between the number of lymph nodes examined or the number of positive lymph nodes between patients with dMMR and patients with pMMR in our cohort, we do not think that the difference in prognosis between dMMR and pMMR in stage III can be explained by a difference in the severity of lymph node metastasis. We have not performed investigations allowing for separating Lynch syndrome from sporadic dMMR. The cohorts in this study represent different time periods. The patients have received different adjuvant chemotherapy regimens and may have been subjected to different surgical regimes, as reflected in the low number of examined lymph nodes in the NGICG cohort. The distribution of other clinicopathological variables is similar between the two cohorts. We have adjusted statistically for the difference between cohorts by including the variable "cohort" in our multivariate Cox regression models. The reason for the prognostic shift of dMMR colon cancer during tumor progression needs to be studied further. Tumorigenesis, including dMMR colon cancer, are driven by both genetic and epigenetic changes in the colonic epithelium and interactions with the tumor microenvironment. dMMR colon tumors are hypermutated and associated with features that, in isolation, make a poor prognosis. We propose that the prognosis of dMMR tumors depends on the balance between the influence of favorable tumor-host interactions and unfavorable genetic and epigenetic alterations. When the favorable interaction between tumor and the microenvironment is revoked or decreased, the unfavorable influence of poor prognostic markers associated with dMMR tumors will dominate the diagnosis. This is illustrated in our recently published study, where we describe the interplay between the MMR status, tumor grade, and expression of cell maturation marker CDX2 (caudal type homeobox 2). The loss of CDX2 expression and low tumor grade in isolation are regarded as poor prognostic features. We show that there is a large overlap between dMMR and the loss of CDX2 expression and between dMMR and a high tumor grade. Still, the poor prognostic effect of these markers is restricted to the pMMR group (10). Our current study advances our previous findings by assessing the prognosis of dMMR stage II versus stage III while adjusting for the prognostic impact of the most established immune markers. To our knowledge, this is the first study to show that the prognostic impact of dMMR differs between stage II and III colon cancer also when adjusting for the impact of TIL density and PD-L1 expression. Treatment with anti-PD-1/PD-L1 antibodies yield large tumor responses in metastatic dMMR colon cancer (38). More recently, the striking effects of neoadjuvant immunotherapy have been demonstrated for early-stage dMMR colon cancer (39)(40)(41). There are several ongoing studies on the effects of adjuvant (chemo)immunotherapy in early-stage colon cancer (42). As treatment with immunotherapy is expensive and associated with side effects (43), the expected prognostic outcome in dMMR colon cancer needs to be established before treatment guidelines can be proposed. We believe that the clinical picture for dMMR stage III tumors is far more complex than for dMMR stage II. More comprehensive prospective studies are needed to corroborate the results in our study. Still, this study supports further analyses of the effect of adjuvant immunotherapy in this group of patients. This article describes the prognostic shift in dMMR colon cancer. We conclude that despite being a marker of improved prognosis in stage II, dMMR is not associated with a favorable prognosis in stage III colon cancer. Data availability statement Data will be made available by reasonable request. Requests to access the datasets should be directed to Olav.dahl@helsebergen.no. Ethics statement The studies involving human participants were reviewed and approved by The Regional Committee for Medical Research Ethics of Western Norway. All patients have signed written, informed consents for study participation. Author contributions KEH: conceptualization, writingoriginal draft and editing, CD3 IHC staining, MMR-score, statistical analyses. NBR: CD3 IHC staining, MMR-score, manuscript revision LS: database editing, manuscript revision. OD: project administration, funding acquisition, acquiring of patient materials, manuscript editing. MPM: project management, manuscript editing, conceptualization. All authors contributed to the article and approved the submitted version. Funding For this study, the authors received funding from Health Region West, Norway, University of Bergen, Norway, and the Norwegian Cancer Society.	2022-08-30T13:55:01.352Z	2022-08-30T00:00:00.000	{ "year": 2022, "sha1": "b200496422355f7fda3a22e5691693e462beccfa", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Frontier", "pdf_hash": "b200496422355f7fda3a22e5691693e462beccfa", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [] }
14865065	pes2o/s2orc	v3-fos-license	Antitumor activity of miR-34a in peritoneal mesothelioma relies on c-MET and AXL inhibition: persistent activation of ERK and AKT signaling as a possible cytoprotective mechanism Background The value of microRNAs (miRNAs) as novel targets for cancer therapy is now widely recognized. However, no information is currently available on the expression/functional role of miRNAs in diffuse malignant peritoneal mesothelioma (DMPM), a rapidly lethal disease, poorly responsive to conventional treatments, for which the development of new therapeutic strategies is urgently needed. Here, we evaluated the expression and biological effects of miR-34a—one of the most widely deregulated miRNAs in cancer and for which a lipid-formulated mimic is already clinically available—in a large cohort of DMPM clinical samples and a unique collection of in house-developed preclinical models, with the aim to assess the potential of a miR-34a-based approach for disease treatment. Methods miR-34a expression was determined by qRT-PCR in 45 DMPM and 7 normal peritoneum specimens as well as in 5 DMPM cell lines. Following transfection with miR-34a mimic, the effects on DMPM cell phenotype, in terms of proliferative potential, apoptotic rate, invasion ability, and cell cycle distribution, were assessed. In addition, three subcutaneous and orthotopic DMPM xenograft models were used to examine the effect of miR-34a on tumorigenicity. The expression of miRNA targets and the activation status of relevant pathways were investigated by western blot. Results miR-34a was found to be down-regulated in DMPM clinical specimens and cell lines compared to normal peritoneal samples. miR-34a reconstitution in DMPM cells significantly inhibited proliferation and tumorigenicity, induced an apoptotic response, and declined invasion ability, mainly through the down-regulation of c-MET and AXL and the interference with the activation of downstream signaling. Interestingly, a persistent activation of ERK1/2 and AKT in miR-34a-reconstituted cells was found to counteract the antiproliferative and proapoptotic effects of miRNA, yet not affecting its anti-invasive activity. Conclusions Our preclinical data showing impressive inhibitory effects induced by miR-34a on DMPM cell proliferation, invasion, and growth in immunodeficient mice strongly suggest the potential clinical utility of a miR-34a-replacement therapy for the treatment of such a still incurable disease. On the other hand, we provide the first evidence of a potential cytoprotective/resistance mechanism that may arise towards miRNA-based therapies through the persistent activation of RTK downstream signaling. Electronic supplementary material The online version of this article (doi:10.1186/s13045-016-0387-6) contains supplementary material, which is available to authorized users. Background Diffuse malignant peritoneal mesothelioma (DMPM) is an uncommon though locally aggressive tumor that develops from mesothelial cells lining the peritoneal cavity [1]. DMPM prognosis is dismal and standard therapy, including palliative surgery, systemic/intraperitoneal chemotherapy, and abdominal irradiation, showed to be ineffective, with a median survival of about 1 year [1]. Currently, the most effective treatment is a loco-regional approach combining aggressive cytoreductive surgery (CRS) with hyperthermic intraperitoneal chemotherapy (HIPEC), which significantly extended survival in selected series of patients [1]. However, for recurrent patients and for those who are not eligible to CRS+HIPEC, the prognosis remains severe due to the lack of alternative treatment options [2]. Considering the rarity of the disease and the unavailability of experimental models, the biology of DMPM is still largely unknown. It is anticipated that advances in the knowledge of the mechanisms responsible for the biological aggressiveness and the relative chemoresistance of DMPM will allow the identification of relevant targets for the development of novel therapeutic strategies. MicroRNAs (miRNAs) are single-stranded endogenous evolutionary conserved, non-coding RNA molecules acting as post-transcriptional regulators of gene expression [3]. Deregulated miRNA expression and/or function have been observed in a variety of human solid and hematological tumors and have been causatively linked to the pathogenesis of cancer [4]. Depending on their expression levels, cellular context, and target functions, miRNAs can act as oncogenes or tumor suppressors [4] and may represent novel targets or tools for cancer therapy [5]. MiR-34a is one of the most widely studied miRNAs in cancer. Its expression has been found to be decreased in a variety of human tumors [5] due to DNA copy number loss or epigenetic silencing through aberrant CpG methylation [6,7]. Results of studies where the expression of miR-34a was manipulated in human tumor experimental models clearly showed that the miRNA acts as a tumor suppressor by regulating highly relevant processes such as proliferation, cell cycle, apoptosis, invasion, and metastasis [8]. Reinforced expression of miR-34a has been found to positively modulate drug response in cancer cells [8]. In addition, a liposomal nanoparticle-formulated synthetic miR-34 (MRX34) recently entered a phase I clinical study for patients with different tumor types [9]. No information concerning the expression and/or the functional role of miRNAs in DMPM is currently available in the literature. Based on the knowledge that several receptor tyrosine kinases (RTKs) are validated targets of miR-34a [8] and our previous results indicating that activation of downstream RTK signaling, in terms of phosphorylation/ overexpression of extracellular signal regulated kinase ½ (ERK1/2), AKT, and mTOR, is present in a considerable fraction of DMPM clinical specimens [10], we proposed to investigate the possible relevance of miR-34a in the disease, with the final aim to develop novel therapeutic strategies. Here, we report that miR-34a is down-regulated in DMPM clinical specimens and demonstrate that miR-34a replacement in a unique collection of in-house-developed human DMPM experimental models [11][12][13] inhibits cell proliferation and invasion and impairs tumor growth formation in SCID mice, mainly as a consequence of c-MET and AXL inhibition. These findings identified miR-34a-AXL and -c-MET axes as promising therapeutic targets for DMPM. Moreover, we provide evidence of persistent activation of ERK1/2 and AKT as a possible cytoprotective mechanism to RTK inhibition by miR-34a. Clinical samples Forty-five DMPM specimens classified as epitheliod (40), sarcomatoid (1), and biphasic (4) from patients treated with CRS+HIPEC at the Fondazione IRCCS Istituto Nazionale dei Tumori, Milan (INT) from October 1997 to February 2013, and 7 normal peritoneum specimens from patients who underwent surgery for non-oncologic disease were available for miR-34a expression analysis. This study was approved by the Institutional Review Board and Ethical Committee and each patient provided written informed consent to donate to INT the leftover tissue after diagnostic and clinical procedures. Cell transfection Mimic pre-miR-34a precursor (miR-34a) and mimic negative control (Neg) were purchased as Pre-miR™ miRNA precursor molecules (Thermo Fisher Scientific, Monza, Italy). Knockdown of AXL and c-MET was performed using specific siRNAs (siAXL and siMET; ON-TARGET plus SMART pool) and, as a control, a siRNA with a nonsense/scrambled sequence (siNeg, ON-TARGET plus non-Targeting Pool) (Dharmacon, CO, USA) was used. Cells were transfected for 24 h with 20 pM miR-34a or Neg, or 100 nM siAXL, siMET, or siNeg, using Lipofectamine® RNAiMAX Transfection Reagent (Thermo Fisher Scientific) with Opti-MEM I (Gibco, NY, USA) according to the manufacturer's instructions. RNA extraction, cDNA synthesis, and qRT-PCR Quantification of miR-34a expression levels was assessed by qRT-PCR. Total RNA was isolated using the miRNeasy Mini Kit (QIAGEN, Hilden, Germany) and 1 μg of RNA was reverse transcribed by miScript II RT Kit (QIAGEN). Mature miRNA expression was assayed by miScript Primer Assays specific for miR-34a (MS00003318) and normalized on SNORD48 (MS00007511) (QIAGEN). Quantitative RT-PCR was conducted using miScript SYBR Green PCR Kit (QIAGEN). The reaction was carried out in a 96-well PCR plate at 95°C for 15 min followed by 40 cycles of 94°C for 15 s, 55°C for 30 s, and 70°C for 30 s and a dissociation step to distinguish specific from non-specific amplification products. Each sample was analyzed in triplicate. Amplifications were run on the 7900HT Fast Real-Time PCR System (Applied Biosystem). Data were analyzed by SDS 2.2.2 software (Applied Biosystems) and reported as -ΔCt, that is the difference between the Ct of the target gene and the Ct of the housekeeping gene (where Ct is the threshold cycle), or as relative quantity (RQ) or -ΔΔCt with respect to a calibrator sample (i.e., negative control transfected cells) according to the 2 −ΔΔCt method. Cell growth assay To assess the effect of miR-34a restoration on cell proliferation, DMPM cells were transfected with Neg or miR-34a as described above. At different intervals from transfection, cells were trypsinized and counted in a particle counter (Beckman Coulter, Cassina de' Pecchi, Italy). Results were expressed as percent variation in the number of miR-34a-transfected cells compared with Negtransfected cells. Immunoblotting analyses Cell lysates were fractionated by SDS-PAGE, transferred to nitrocellulose membranes, and probed with specific antibodies, as described in [14]. Cells were lysed and western blot was performed using the following primary antibodies Secondary antibodies used were conjugated to horseradish peroxidase (GE Healthcare, Little Chalfont, UK). Immunostained bands were detected by chemoluminescence method (ECL, GE Healthcare). In many experiments, membranes were stripped and reblotted with a second antibody. Moreover, membranes were cropped to allow simultaneous incubation of different primary antibodies on the same samples. For the preparation of figures, we cropped the original western blot to generate the appropriate figure panels with the relevant lanes. This cropped image was then subjected to uniform image enhancement of contrast and brightness. Molecular weights were determined using the Precision Plus Protein™ Standard (Bio-Rad, Segrate, Italy), which yields a colorimetric image only and has been removed from the chemoluminescent blot image. Drugs The AKT-1/2 inhibitor trifluoroacetate salt hydrate (A6730, Sigma Chemical Company) and the MEK inhibitor CI-1040 (PD184352, Selleck Chemicals, Houston, TX, USA) were dissolved and diluted in DMSO. Final concentration of DMSO in cell cultures never exceeded 0.5%. The antiproliferative activity was evaluated by cell counting at different times after exposure of miR34areconstituted MesoII cells to drug concentrations able to inhibit cell proliferation by 20% (IC 20 ). Apoptosis detection and cell cycle analysis At different time points after transfection with miRNAs mimic or siRNAs, floating and adherent cells were harvested and processed for apoptosis evaluation by TUNEL assay according to manufacturer's instructions (Roche, Mannheim, Germany) and for cell cycle [15]. For cell cycle, cells were fixed in 70% ethanol 96 h after transfection, stained in phosphate-buffered saline (PBS) containing 10 μg/ml propidium iodide (PI; Sigma Chemical Company), and RNase A (66 U/ml; Sigma Chemical Company) for 18 h and analyzed by FACScan flow cytometer (Becton Dickinson, Mountain View, CA, USA). Senescence-associated β-galactosidase staining Cells were transfected with miR-34a or Neg for 24 h. Samples were washed in PBS 72 h after transfection and processed for senescence-associated β-galactosidase (SA-β-Gal) staining. Cells were fixed for 5 min (room temperature) in 2% formaldehyde/0.2% glutaraldehyde, washed and incubated overnight at 37°C (no CO 2 ) with fresh solution as previously described [15]. At least 300 cells were examined, and the results were expressed as percentage of SA-β-Gal positive cells over the whole population. Transwell invasion assay Invasion assay was performed 72 h after transfection using a 24-well Boyden chamber with 8-mm pore size filter in the inset chambers (Costar, Corning Inc., NY, USA). The Transwell membranes were previously coated with 3.47 μg Matrigel/well (BD Biosciences) and dried for 30 min. Cells were suspended in 300 μL serum-free medium and seeded into the insert chambers. After 24 h of incubation at 37°C in 5% CO 2 , cells that migrated into the bottom chamber containing 1 ml of serum-free medium were fixed in 95% ethanol, stained with a solution of 0.4% sulforhodamine B in 0.1% acetic acid, counted under an inverted microscope, and then photographed. Antibody arrays and ELISA Cells were seeded at 2 × 10 4 cells/dish in complete medium and transfected with Neg or miR-34a for 24 h before serum starvation for 72 h. Conditioned media were then harvested and clarified by centrifugation at 13,000 rpm for 15 min. Cells were trypsinized, counted, and lysed for assaying protein content. Supernatant aliquots were used to assess angiogenesis-related protein content by Antibody Arrays (R&D System, SPACE Import Export, Milan, Italy) according to manufacturer's instructions. The ELISA kit for Maspin (human Maspin "Super X" ELISA Kit, Antigenix America, Huntington Station, NY, USA) was used according to the manufacturer's instructions for quantitative analysis. In vivo experiments All experimental protocols were approved by the Ethics Committee for Animal Experimentation of INT. Experiments were performed using 8-week-old female SCID mice (Charles River, Calco, Italy). Each group contained five to six mice. Cells were transfected with miR-34a or Neg for 24 h, as described above, and then inoculated subcutaneously or intraperitoneally after the analysis of the transfection efficiency by qRT-PCR. Subcutaneous tumor models STO, MesoII, and MP8 cells were injected subcutaneously into the right flank (1-1.2 × 10 7 cells/mouse). Inoculated animals were inspected daily to establish the time of tumor onset. Tumor growth was measured every 2 to 3 days using a Vernier caliper ( Table 1). The subcutaneous tumor volume was calculated as follows: TV (mm 3 ) = d 2 × D/2 where d and D are the shortest and the longest diameter, respectively. Volume inhibition percentage (TVI%) in tumors derived from miR-34a-over Neg-transfected cells was calculated as follows: TVI% = 100 − (mean miR-34a TV/mean Neg TV × 100). Proteins were obtained as described previously [16] from frozen s.c. tumors derived from two additional mice sacrified at different time points. Briefly, samples were pulverized by Mikro-Dismembrator II (B. Brown Biotech International, Melsungen, Germany) and suspended in lysis buffer supplemented with protease and phosphatase inhibitors. Proteins were processed as described [16]. Intraperitoneal (orthotopic) tumor models STO and MP8 cells were injected into the peritoneal cavity (10 7 and 2.5 × 10 7 cells/mouse, respectively). Animals were monitored and weighed daily and sacrificed at different times from cell injection ( Table 2). A careful necropsy was performed to evaluate the take rate and spread of mesothelioma cells in the abdominal cavity. Solid masses were gently detached from organs and abdominal walls, removed, and weighed for calculating the percentage of tumor weight inhibition (TWI %) in mice inoculated with miR-34a-over Neg-transfected cells. Statistical analyses If not otherwise specified, in vitro data are presented as mean values ± SD from at least three independent experiments. Statistical analysis of the data was performed by two-tailed Student's t test. For in vivo data, two-tailed Student's t and Fisher's exact test were used to compare tumor volumes/weights and tumor takes, respectively. Patient survival analysis was performed using Cox proportional regression model [17]. p values <0.05 were considered statistically significant. miR-34a is down-regulated in DMPM clinical samples and cell lines We first evaluated miR-34a expression by qRT-PCR in 45 DMPM and 7 normal peritoneum specimens as well as in 5 unique cell lines established in our laboratory from clinical samples of epithelioid (STO, MP4, MesoII, MP8) and biphasic (MP115) DMPM. Results indicated that miR-34a abundance is significantly reduced in DMPM compared to normal tissues (Fig. 1). Consistently, miR-34a expression was found down-regulated in all DMPM cell lines, thus indicating an oncosuppressive function of the miRNA also in this disease. No significant difference in miR-34a expression was observed as a function of demographic and clinicopathologic characteristics, including gender, histologic subtype, and peritoneal cancer index [18] (data not shown). In addition, at 5 years of follow-up, miR-34a expression did not significantly affect the probability of disease-free survival of DMPM patients (high expressing versus low expressing-categorized on the basis of the median miR-34a expression value-36 versus 20%; hazard ratio, 1.85; 95% confidence interval, 0.86-4.01; p = 0.11). Overall, such findings suggest a role for miR-34a as a possible therapeutic target rather than a prognostic/predictive biomarker in DMPM. miR-34a reconstitution variably affects DMPM cell growth and apoptosis To functionally assess the possible role of miR-34a as a novel therapeutic target in DMPM, we transiently transfected cells with miR-34a synthetic mimic and miRNA negative control. As assessed by qRT-PCR, marked increase in miRNA abundance was consistently observed in all cell lines at 24 h from transfection (Additional file 1: Figure S1) and was still maintained, although to a lesser extent, at 168 h (Additional file 1: Figure S1 and data not shown). miR-34a reconstitution significantly inhibited the proliferation of four out of five DMPM cell lines in a time-dependent manner, though with a different kinetics (Fig. 2a). Specifically, a more rapid cell growth decline was observed in STO and MP4 cells (~80% inhibition at 96 and 168 h, respectively), whereas, in MesoII and MP8 cells, the same level of inhibition was recorded at later time points (192 and 216 h, respectively) (Fig. 2a). Conversely, ectopic expression of miR-34a only induced a weak inhibition (~30%) of MP115 cell growth, which was almost constant until the end of the experiment (Fig. 2a). The variable antiproliferative effects consequent to miR-34a reconstitution in the different DMPM cell lines was paralleled by a different kinetics of apoptosis induction, as detected by TUNEL assay (Fig. 2b). Specifically, a significant enhancement in the percentage of apoptotic cells was already appreciable at 96 h upon transfection of STO and MP4 cells, whereas the apoptotic response was induced at a later time point (168 h) in MP8 and MesoII cells (Fig. 2b). No induction of apoptotic cell death was observed in MP115 cells until 216 h after miR-34a reconstitution (Fig. 2b). To investigate whether validated targets of miR-34a were modulated by its synthetic mimic in DMPM cells, we assessed protein expression levels of c-MET, AXL, and CDK6 considering their established role in the control of cell proliferation and apoptosis in different tumor types [8,[19][20][21]. A marked down-modulation of the three proteins was consistently observed in all DMPM cell lines (Fig. 2c), regardless of the effects induced by miR-34a reconstitution on cell growth and apoptosis. ERK1/2 and AKT activation as a possible cytoprotective mechanism following miR-34a reconstitution Based on the evidence that the activation of downstream RTK signaling pathways, including PI3K/AKT and RAF/ (Fig. 2c). A reduced expression of phospho-ERK1/2 was also observed in miR-34a-reconstituted MP4 and MP8 cells at the latter time point (Fig. 2c). Consistent with the delayed antiproliferative and proapoptotic response following miR-34a reconstitution, a decline in the expression levels of phospho-ERK1/2 and phospho-AKT was observed in MesoII cells at only 168 h (Fig. 2c, d), whereas no decrease in the abundance of the two phopho-proteins was found in the less sensitive MP115 cell line at either time point (Fig. 2c, d). To assess whether activation of AKT or MAPK/ERK1/2 signaling pathways, which is a well-known mechanism of resistance to RTK inhibitors [23][24][25][26], could also represent a cytoprotective mechanism to the oncosuppessive effects of miR-34a, we exposed miRNA-reconstituted MesoII cells to subtoxic concentrations of small-molecule AKT (A6730) and MEK1 (CI-1040) inhibitors (Fig. 3a). Interestingly, inactivating ERK1/2 or impeding the reactivation of AKT only slightly affected MesoII response to miR-34a, whereas the concomitant blockade of the two pathways, made the sensitivity profile of MesoII cells superimposable on that of the inherently sensitive STO cell line (Fig. 3b). Such a growth inhibitory effect was paralleled by an earlier onset of apoptosis, as detected by caspase-3 cleavage (CPP32) at 96 h (Fig. 3a). To better characterize the cytostatic-rather than cytotoxic-effect observed in MP115 cells following miR-34a reconstitution, we assessed cell cycle distribution by flow cytometry. A cell accumulation in the G1-phase, which was paralleled by an enhanced fraction of senescence-associated β-galactosidase-positive cells, was observed (Fig. 4a, b). Such a senescence-like phenotype, which was not appreciable in the other DMPM cell models ( Fig. 4b and data not shown), could be related to the marked phospho-AKT accumulation observed in miR-34a-reconstituted MP115 cells, according to previous evidence indicating that constitutively active AKT induces senescence in human endothelial cells and human fibroblasts [27,28]. miR-34a oncosuppressive activities mainly rely on c-MET and AXL inhibition To corroborate the hypothesis that miR-34a oncosuppressive functions mainly rely on the down-regulation of c-MET and AXL, we performed siRNA-based phenocopy experiments (Additional file 2: Figure S2). When transfected into STO cells, which do not inherently express AXL, siMET was able to recapitulate the effects induced by miR-34a reconstitution, in terms of cell growth inhibition (Fig. 5a), apoptosis induction (Fig. 5b), impairment of invasive capability (Fig. 5c), and inactivation of both ERK1/2 and AKT pathways (Additional file 2: Figure S2). In MesoII cells, siMET did not appreciably affect cell proliferation, apoptosis, or invasion (Fig. 5a-c). Conversely, siAXL reduced MesoII proliferation, although a cell growth inhibition comparable to that induced by miR-34a reconstitution was only observed following combined silencing of c-MET and AXL (Fig. 5a). Interestingly, siAXL alone phenocopied the effects of miR-34a, in terms of apoptosis and invasion (Fig. 5b, c), suggesting a main role of this RTK in mediating the oncosuppressive effects of miR-34a in MesoII cells. Moreover, AXL silencing did not inhibit AKT and ERK1/2 signaling pathways similarly to miR-34a reconstitution (Additional file 2: Figure S2). The decreased EGFR abundance observed in miR-34areconstituted STO and MesoII cells (Additional file 3: Figure S3a), together with preliminary evidence indicating that the RTK is expressed/activated in DMPM clinical specimens [10], prompted us to investigate a possible role for EGFR down-regulation in sustaining the miR-34ainduced cell phenotype. However, siEGFR failed to affect cell growth and invasion capability in both cell lines (Additional file 3: Figure S3b). miR-34a reconstitution inhibits DMPM cell invasion and impairs the secretion of angiogenesis-related factors Enforced expression of miR-34a significantly inhibited invasion of all DMPM cell models, as detected in a matrigelbased assay at 72 h after transfection (Fig. 6a), likely as a consequence of c-MET or AXL down-regulation. c-MET and AXL signaling pathways are indeed known to affect cell motility and invasion primarily through the activation of mitogen-activated protein kinase (MAPK) [29][30][31][32]. However, since a decreased invasive potential was observed also in MesoII and MP115 cells in spite of MAPK iperactivation (Fig. 2c), the status of focal adhesion kinase (FAK), known to mediate cell migration and anchorage-independent growth downstream of RTKs [29][30][31][32][33], was assessed. Interestingly, miR-34a ectopic expression consistently reduced FAK posphorylation at Y576/577 in STO, MesoII, and MP115 cell models, suggesting that miR-34a-induced inhibition of cell invasion can occur regardless of MAPK and AKT activation (Fig. 6b). Since AXL and c-MET inhibition by drugs and monoclonal antibodies was shown to induce antiangiogenic effects in tumors [34,35], we investigated whether miR-34a reconstitution was able to affect the production/release of angiogenesis-related proteins by DMPM cells. We found 's t test). c, d Effects of miR-34a on validated miRNA targets and RTK downstream signaling cascades as assessed by western blot analysis at 72, 96 (c) and 168 (d) h after cell transfection with Ctrl, Neg, or miR-34a. Actin was used to confirm equal protein loading. A representative experiment of three was reported. Cropped images of selected proteins are shown that miR-34a reconstitution in MesoII cells impaired the secretion of angiogenesis-related molecules (Fig. 7). Indeed, antibody array results obtained in conditioned medium of miR-34a-reconstituted cells (at 72 h following transfection) showed a reduced secretion of the urokinase-type plasminogen activator (uPA), which plays a major role in promoting angiogenesis [36], together with and increased release of maspin, a member of the serine protease inhibitor (serpin) superfamily, which exerts antiangiogenic effects through the inhibition of both the growth and migration of endothelial cells [37,38] (Fig. 7a). Western blot and ELISA experiments carried out on the same conditioned media confirmed a reduced expression of uPA precursor (Fig. 7b) and an increased abundance of maspin (Fig. 7c). miR-34a reconstitution inhibits tumor formation in SCID mice To investigate whether miR-34a reconstitution affected DMPM formation in vivo, we subcutaneously (STO, MesoII, MP8) and intraperitoneally (STO, MP8) inoculated miR-34a mimic-or miRNA negative control-transfected cells into SCID mice. Results indicate that miR-34a consistently impaired the growth of all s.c. xenograft models, with maximum tumor volume inhibitions ranging from 57 to 98% (Fig. 8 and Table 1). In addition, an appreciably delayed tumor onset was observed for MesoII and MP8 cell models Fig. 3 Inhibition of AKT and ERK1/2 signaling pathways increases the antiproliferative and proapoptotic effects of miR-34a. MesoII cells were transfected with Ctrl, Neg, or miR-34a for 24 h and successively exposed to vehicle (DMSO, unt) or low concentrations (corresponding to IC 20 values) of A6730 (AKT inhibitor, 5 μM) and/or CI-1040 (MEK inhibitor, 3 μM). a Expression and phosphorylation status of AKT and ERK1/2 and amount of cleaved CPP32 at 72 h after drug treatment were assessed by western blot. Actin was used to confirm equal protein loading. A representative experiment of three was reported. The panel shows cropped blots. b The effects of AKT and MEK inhibitors on the growth of miR-34a-transfected cells were assessed by cell counting. Data are expressed as percentage of the proliferation of miR-34a-versus Neg-transfected cells. The cell growth inhibition curve of STO cells after enforced expression of miR-34a was reported for comparison. Means ± SD values of three independent experiments are shown (Table 1). Western blot carried out in tumors collected from additional mice sacrificed at different time points after DMPM cell inoculum indicated a decreased expression of c-MET in all xenograft models and AXL in MesoII and MP8 models (Fig. 8b), in accordance with in vitro results. As regards i.p. xenograft models, at necropsy, all control mice showed a large tumor mass at the site of cell injection mainly invading the peritoneum wall and widespread small nodules in the peritoneum and attached to the diaphragm, liver, and bowel. Conversely, only one mouse out of the five mice receiving miR-34a-reconstituted STO cells developed small tumor nodules in the abdominal cavity (Table 2). miR-34a ectopic expression did not influence the take of MP8 cells but markedly reduced their growth, as indicated by a significantly reduced tumor weight (Table 2). Discussion No information is currently available on the expression and functional role of miRNAs in DMPM. Here, we demonstrated that miR-34a is down-regulated in a large series of DMPM clinical samples and in a unique panel of cell lines, established from DMPM patients in our laboratories, compared to normal peritoneum specimens. We also illustrated that miR-34a exerts oncosuppressive functions in our tumor models, consistent to what previously observed in a variety of human tumor types [5,[39][40][41]. Indeed, miR-34a reconstitution impaired proliferation and induced an apoptotic response in DMPM cell lines, although at a variable extent and with different kinetics, mainly through the down-regulation of c-MET and AXL and the interference with the activation of downstream signaling. Interestingly, results also indicated that a transient or persistent activation of ERK1/2 and AKT can delay or prevent the cytotoxic and proapoptotic effects of miR-34a reconstitution, as observed in MesoII and MP115 cells, respectively. Noteworthy, DMPM cell feedback to AXL and c-MET down-regulation induced by miR-34a reconstitution is to directly activate ERK1/2 and AKT survival signaling cascades rather than up-regulate the expression levels of the receptors, thus ensuring a more prompt counter-response. Such findings provide the first evidence that tumor cells can exploit a wellknown mechanism of resistance to RTK inhibitors-i.e., the activation of RTK downstream signaling [23][24][25][26]-to counteract the antiproliferative/proapoptotic effects of miR-34a. However, such a mechanism was not found to protect DMPM cells from the anti-invasive effect of the miRNA. Noteworthy, the cytoprotective mechanism based on ERK1/2 and AKTactivation was mainly evident in the DMPM cell line MP115, derived from a biphasic subtype tumor. Such subtype is known to be more aggressive and associated with a reduced patient survival compared to the epithelioid [42], although differences in specific relevant biological properties between the two DMPM subtypes are currently unknown. In addition, the delayed pro-apoptotic and cytotoxic effects observed in MesoII cells following miR-34a reconstitution are consistent with the finding that, unlike other epithelioid cell lines, they carry a mutant p53 [11]. A novel mechanism of miR-34a-dependent AKT inhibition has been recently proposed by Wang et al. [43]. In this study, miR-34a is reported to inhibit Bmi-1 by targeting c-Myc in gastric cancer cells, resulting in a PTEN-dependent reduction of phospho-AKT. The observation that in DMPM cell lines more susceptible to the cytotoxic effects of miR-34a (STO, MP4, MP8), a decrease in phospho-AKT abundance is observed early after miRNA reconstitution would suggest the possibility that the above-described mechanism is also operating in our models. However, results of phenocopy experiments showing that siRNA-mediated silencing of c-MET and AXL was able to decrease AKT activation in sensitive cells (STO) but not in those less susceptible to miR-34a (MesoII) would suggest that the main mechanism controlling AKT phosphorylation status relies on RTK activity. Interestingly, miR-34a induced a remarkable antitumor activity in the three cell lines (STO, MesoII and MP8) able to generate tumors following xenotransplantation into immunodeficient mice. Although to a different extent, miR-34a reconstitution significantly reduced the growth of the three s.c. xenograft models. Highly relevant to the disease, miRNA ectopic expression also impaired the growth of STO and MP8 orthotopic xenografts, which properly recapitulate the dissemination pattern in the peritoneal cavity of human DMPM [11,12], thus representing improved models to investigate novel therapeutic approaches. Specifically, miR-34a significantly inhibited the take of STO cells, with only one mouse developing small tumor nodules in the abdominal cavity. Although the miRNA did not influence the take of MP8 cells, a significantly reduced tumor growth was observed. Unfortunately, the inability of MP115-the only biphasic DMPM model in our panel-to grow in vivo prevented us to assess whether the in vitro cytostatic effect consequent to miR-34a reconstitution, which was paralleled by the induction of a senescence-like phenotype possibly sustained by ATK activation, may result or not in tumor growth impairment. However, the significantly reduced invasive potential induced by miR-34a in DMPM cell lines through the inhibition of FAK signaling could primarily contribute to the antitumor effect observed in the xenograft models. Moreover, the occurrence of miR-34a-induced inhibition of cell invasion in the absence of appreciable antiproliferative and proapoptotic effects that we observed in MP115 is not surprising since the same phenotype has been previously reported by Li et al. [44] for miR-34a reconstituted HepG2 hepatocellular carcinoma cells. Interestingly, our evidence indicating that miR-34a ectopic expression impairs the secretion of angiogenesisrelated factors by MesoII cells strongly suggests that the antitumor effect observed in both s.c. and ortothopic xenograft models can also rely on miRNA-induced modification of tumor microenvironment, making it less favorable to tumor growth. In summary, the impressive inhibitory effects induced by miR-34a on DMPM cell proliferation, invasion, and growth in immunodeficient mice suggest a possible utility of the clinically available miR-34a as novel therapeutic option for DMPM patients who are not eligible for or relapse after CRS+HIPEC. In addition, the evidence that miR-34a reconstitution positively modulates the activity of antitumor drugs in experimental models of different human tumor types [8,[45][46][47] highlights the possibility that the miR-34a mimic could have an important role also in combined strategies for treating DMPM patients. Conclusions DMPM is a rapidly fatal tumor with scanty therapeutic options. Here, we demonstrated for the first time that Silencing AXL and c-MET phenocopies miR-34a effects. STO and MesoII cells were treated with transfection reagent (Ctrl), siNeg (siRNA with a nonsense/scrambled sequence) or AXL-and c-MET-directed siRNA (siAXL, siMET) for 24 hours. a Effect of AXL and/or c-MET knockdown on cell growth, as detected by cell counting at different times after transfection. The antiproliferative effect induced by miR-34a reconstitution is reported for comparative purposes. b Induction of apoptosis at 96 h after transfection, as assessed by TUNEL assay (p < 0.01 by Student's t test). Data are expressed as percentage (mean ± SD) of the proliferation of siAXL-/siMET-versus siNeg-transfected cells. c Effect of RTK-siRNA on DMPM cell matrix degrading/invasive activities. Cells were silenced and, after 72 h, subjected to Matrigel invasion assay in serum-free medium. The number of invading cells per field is reported. Histogram bars represent mean values ± SD of at least three independent experiments (p < 0.001 by Student's t test) Fig. 6 miR-34a inhibits the invasion of DMPM cells. Cells were transfected with Ctrl, Neg, or miR-34a for 24 h. a Cells were subjected to Matrigel invasion assay in serum-free medium 72 h after transfection. Top: the number of invading cells per field is reported. Histogram bars represent mean values ± SD of three independent experiments (p < 0.01; p < 0.001 by Student's t test). Bottom: micrographs from one experiment representative of three. Original magnification, × 40. b Whole-cell lysates were analyzed by western blot with anti-phospho-FAK. Protein extraction was performed 96 h after transfection with Ctrl, Neg, or miR-34a. Vinculin was used to confirm equal protein loading. A representative experiment of three was reported. Cropped blots are shown Fig. 7 Effect of miR-34a restoration on the secretion of angiogenesis-related factors. a Secreted angiogenesis-related molecules were examined by antibody arrays in conditioned medium obtained from Neg-or miR-34a-transfected MesoII cells grown for 72 h in the absence of serum. C+, internal standards. The same conditioned media were used to assess the expression of b uPa precursor by western blot and c maspin protein by ELISA. Each bar represents mean ± SD of triplicate samples from a representative experiment (p < 0.01 by Student's t test) reconstitution of miR-34a in relevant models of the disease induced a significant antitumor effect, which mainly relied on c-MET and AXL down-regulation and impairment of their downstream signaling. In vivo results were complemented by in vitro data showing significant antiproliferative, proapoptotic, and anti-invasive activities. Taken together, our results provide evidence that (i) c-MET and AXL signaling pathways are critical determinants of DMPM cell survival, growth, and invasiveness and that miR-34a reconstitution can impair all these functions and (ii) persistent activation of AKT and ERK1/2 downstream signaling pathways represents a cytoprotective mechanism against miRNA-induced proapoptotic effects, though not preventing its anti-invasive activity, which instead mainly relies on FAK inhibition. Overall, our preclinical data form a solid foundation that could promote the clinical translation of clinically available miR-34 mimic for the treatment of a still incurable disease such as DMPM and, on the other hand, provide the first evidence of a possible cytoprotective/resistance mechanism that may arise towards miRNAbased therapeutics. Additional files Additional file 1: Figure S1. Expression of miR-34a upon restoration in DMPM cells. Cells were transfected with either Neg or miR-34a for 24 h.	2019-03-17T13:06:12.319Z	2017-07-01T00:00:00.000	{ "year": 2017, "sha1": "b193dcf8e6f46e02486d8991de7db4d54b6133fc", "oa_license": "CCBY", "oa_url": "https://jhoonline.biomedcentral.com/track/pdf/10.1186/s13045-016-0387-6", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "503158a7dc91e96fb5e4b18ef15d4539291cd3f6", "s2fieldsofstudy": [ "Biology", "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
118239667	pes2o/s2orc	v3-fos-license	Degeneracy and Consistency Condition of Berry phases: Gap Closing under the Twist We have discussed a consistency condition of Berry phases defined by a local gauge twist and spatial symmetries of the many body system. It imposes a non trivial gap closing condition under the gauge twist in both finite- and infinite-size systems. It also implies a necessary condition for the gapped and unique ground state. As for the simplest case, it predicts an inevitable gap closing in the Heisenberg chain of half integer spins. Its relation to the Lieb-Schultz-Mattis theorem is discussed based on the symmetries of the twisted Hamiltonian. The discussion is also extended to the (approximately) degenerated multiplet and fermion cases. It restricts the number of the states in the low energy cluster of the spectrum by the filling of the fermions. Constraints by the reflection symmetry are also discussed. We have discussed a consistency condition of Berry phases defined by a local gauge twist and spatial symmetries of the many body system. It imposes a non trivial gap closing condition under the gauge twist in both finite-and infinite-size systems. It also implies a necessary condition for the gapped and unique ground state. As for the simplest case, it predicts an inevitable gap closing in the Heisenberg chain of half integer spins. Its relation to the Lieb-Schultz-Mattis theorem is discussed based on the symmetries of the twisted Hamiltonian. The discussion is also extended to the (approximately) degenerated multiplet and fermion cases. It restricts the number of the states in the low energy cluster of the spectrum by the filling of the fermions. Constraints by the reflection symmetry are also discussed. For the last decade, quantum criticality with gapless excitations have been focused in its relation to a quantum phase transition. Generically speaking, the gapless phase is only realized by a fine tuning of the parameters of the quantum hamiltonian. Some machinery to protect the gap closing of the quantum mechanical system is necessary. Symmetries as a spatial translation can be one of the reasons which plays an important role in the Fermi liquids. Today we have other several machinery for the gap closing in a generic situation. One is the Nambu-Goldstone mechanism. When a continuous symmetry is broken spontaneously, there can be a gapless excitation by slowly varying its local order parameter which is responsible for the symmetry breaking. It is closely related to the Lieb-Schultz-Mattis(LSM) argument[1, 2, 3, 4, 5, 6, 7] to prove existence of the gapless excitations in a half-odd-integer spin chains. This LSM argument is justified in restricted situations. It allows existence of a gapped phase which is known as the Haldane phase in the integer spin chains. This is a typical quantum liquid where strong quantum fluctuations prevent from forming an ordered phase. The other mechanism to protect the gapless excitations is an appearance of the edge states such as the one in quantum Hall states and Haldane phases [8,9,10,11,12,13,14,15,16]. There is a fundamental class of physical phases with intrinsic energy gap as topological insulators. They are gapped quantum liquids in a bulk without any fundamental symmetry breaking. However the system has a geometrical (topological) perturbation such as the existence of the boundaries, low energy modes (quasi-particle) as the generic edge states appear and the phase becomes gapless when the boundaries are infinitely separated. The existence of the energy gap is also closely related to the degeneracy of the ground state. When the discrete symmetry is broken such as for the Heisenberg spins with ising anisotropy or charge ordered states, the symmetry broken states split into a low energy cluster with small energy splitting in a finite periodic system. The exact degeneracy is only realized in the infinite size limit. For the topological insulators without any symmetry breaking, there can be additional degeneracy as the topological degeneracy which characterizes the non trivial quantum liquids [17,18]. As for the topological insulators, quantum order parameters by the many body Berry phases, associated with the local gauge twists, have been proposed and its validity is justified for several concrete models [19,20,21]. Using this Berry phase and its symmetry property, we prove that there is an inevitable degeneracy during the twist as for some classes of Hamiltonian which is also related to the LSM theorem. Our discussions are valid not only in infinite-size systems but also in finite-size systems, which can be applicable for various correlated electron systems. Let us define a local order parameter γ ij at a link (ij) by the Berry phase [22] [19], where φ is a parameter of the local U (1) twist on the link. It is gauge dependent but is well defined up to modulo 2π. Further it is quantized if the ground state is invariant under some anti-unitary operation. As for the Heisenberg model with generic connectivity, the Hamiltonian with the local twist is given as We should note that the Berry phase is only well defined unless the energy gap between the ground state and the first excited states does not vanish during the twist. Since the twist on the the link can be modified by the local gauge transformation, it gives some constraints for the Berry phase distribution. For simplicity, let us first discuss the one dimensional nearest neighbor antiferromagnetic Heisenberg model of generic spins on a finite lattice. We discuss within the subspace of fixed S z tot = S z j since it commutes with H i,j (φ). Performing a local gauge transformation at the site j as which is single valued in the parameter space, one has an important relation between the two different hamiltonians (which are gauge equivalent), H j,j+1 and H j−1,j as H j,j+1 = U † j H j−1,j U j . Correspondingly the states are mapped into each other as \|ψ j,j+1 = U † j \|ψ j−1,j , where \|ψ j−1,j and \|ψ j,j+1 are two different ground states of H j−1,j and H j,j+1 , respectively. Note that, if the state \|ψ j−1,j is gauge fixed as a single valued in the parameter space, it is also true for the state \|ψ j,j+1 . Assuming that the ground state is unique and gapped during the twisting, this simple relation gives a constraint for the quantum local order parameters γ ij 's Since the expectation value of the hermitian operator S z j is independent of the gauge transformation, we have dropped the label of the wave function to specify the position of the twist. The time reversal invariance of the state \|ψ implies ψ\|S z i \|ψ = − ( ψ\|S z i \|ψ ) * = 0. Then, one obtains Since the present 1D Heisenberg model has a translational symmetry, the Berry phases as the quantum order parameters should also respect this translational symmetry, that is, γ ij should be independent of the link (ij) in mod 2π unless they are well defined. However this is impossible for the half integer spins due to the above constraint γ j−1,j ≡ γ j,j+1 + π, (mod2π). It implies that there is a gap closing under the local twist, as for the antiferromagnetic Heisenberg chains with half-integer spins on a finite lattice. Although our results can be applicable for arbitrary half integer spins, similar conclusions for the S = 1/2 case are also obtained from different analysis [7,23,24]. As for the integer spin Heisenberg model, the above constraint does not forbid uniform distribution of the Berry phase γ ij . Actually the uniform π Berry phase is realized in a S = 1 case [19,21,25]. The idea here is extended to the general cases without any difficulty. Let us consider sets of some generic twists for several links labeled by "In". Assuming the uniqueness of the ground state even under the twisting, one can define the Berry phase γ In . Next let us perform a set of the local gauge transformation within the area A. Then we have a new set of twists labeled by "Out". Under this generic setup, one has a constraint between the two corresponding Berry phases γ In and γ Out as Translational symmetry For a spin ladder as an example of translational invariant case, we take A as a unit layer and define Berry phase by twisting the links on the left boundary of the area A simultaneously (see FIG.1). Due to the translational symmetry, we have a constraint for the Berry phases as γ In ≡ γ Out (mod 2π) assuming that the ground state is gapped under the twisting. Since the total S z is conserved, we have where \|A\| is a number of sites in the unit layer and m is the average magnetization (N is a number of the unit layers). Note that the translational invariance of the magnetization at arbitrary φ is guaranteed by the fact that S z total commutes with a unitary transformation which spreads the flux in a transitionally invariant way. Then we have a necessary condition for the unique and gapped ground state as It is a condition to have a magnetization plateau [26]. By considering a non-Abelian Berry phase [19], this argument can be extended to the case with (approximate) degeneracy. Now let us assume the low energy spectrum near the ground states forms a multiplet Ψ = (\|ψ 1 , · · · , \|ψ M ) with M eigenstates, \|ψ i (i = 1, · · · , M ). Here, we assume that they are in the subspace of the same S z tot and the energy gap above the multiplet is stable. Then as for the Berry phase γ = dφ Tr Ψ † ∂ φ Ψ, we have a relation γ In = γ Out + 2πM ν based on the same assumption. Therefore as for the translational invariant system, we have a consistency requirement 2πM ν ≡ 0 (mod 2π) assuming the gap even under the twist. Then as for ν = i∈A (S i − m) = p q with mutually co-prime p and q case, M has to be a multiple of q, that is, the low energy spectrum has to form a cluster of qℓ states (ℓ: integer). This situation naturally occurs with discrete symmetry breaking [27] or topological degeneracy. As for the S = 1/2 Heisenberg ladder with n legs, the discussion here predicts a gap closing under the twist when n is odd [28,29]. It allows to have an energy gap above the low energy multiplet composed of two states even with the S = 1/2 system, which is realized for the spin tube [30]. The present argument also gives a consistent description for ferrimagnets discussed by the effective field theory and the LSM argument [31]. Consider a Heisenberg spin chain with different spin quantum numbers as S j = S 1 for j = 1 (mod M ) and S j = S 2 for others. Taking the unit layer to include these M spins, the gauge transformation yields the relation of the Berry phase as γ 1 = γ 2 + 2π(S 1 + (M − 1)S 2 ). The system must have the gapless excitations or the ground state degeneracy if (S 1 + (M − 1)S 2 ) ∈ Z, and the system can have a unique and gapped ground state if (S 1 + (M − 1)S 2 ) ∈ Z. We shall now discuss the connection to the LSM theorem. Our argument suggests that there is at least one level-crossing point during the local twist if Berry phases cannot be arranged in a compatible way with the translational symmetry. Indeed, one can rigorously show the degeneracy of the ground state at φ = π. For simplicity, we consider the first example, namely the halfinteger Heisenberg spin chain with length L in a zero magnetic field. We introduce the following two symmetry operations. One is U j (φ)T , where T is the operation for the one-step translation along the chain. The Hamiltonian H j−1,j is invariant under this operation: The other is the spin flip operation F defined by F S z j F = −S z j and F S ± j F = S ∓ j for any j. The Hamiltonian has this symmetry if and only if φ = 0 or π, i.e., F H j−1,j (φ = 0/π)F = H j−1,j . At φ = π, there is a hidden algebraic relation: {U j (π)T, F } = 0 where {, } denotes the anticommutator. This can be shown by the fact that F U j (π)F = U j (π)e 2πiS z j = −U j (π) since we consider the half-oddinteger spin chains. From the anticommutation relation, there exist at least two ground states labeled by the quantum number associated to F at φ = π. We call two of them \|G(π, +1) and \|G(π, −1) ≡ (U j (π)T )\|G(π, +1) , where F \|G(π, η) = η\|G(π, η) . This degeneracy is not restricted to the ground state. From our argument, it is obvious that every energy level is at least doubly degenerate at φ = π and can be distinguished by the eigenvalue of F . An extension to the above argument to other systems with translational symmetry can be done in a straightforward way by replacing U j (φ)T with U A (φ)T A , where U A (φ) = j∈A U j (φ) and T A is the translation of the unit layer. To discuss the relation between this degeneracy and the LSM theorem, it is useful to introduce the translationally invariant Hamiltonian with the twist φ asH(φ) ≡ U † (φ)H L,1 (φ)U (φ), where U (φ) = L j=1 U j (−jφ/L). The level-crossing at φ = π suggests that one of the excited state ofH(0) = H L,1 (0) is smoothly connected to the ground state ofH(2π). Since H L,1 (2π) = H L,1 (0), the ground state ofH(2π) is given by U † (2π)\|G(0) , where \|G(0) is the ground state of H(0). Finally, along the same lines as the LSM argument, one can show that U † (2π)\|G(0) is orthogonal to \|G(0) and the energy difference between them is O(1/L) using the translational and F symmetry. Another comment is that, in the case of one dimensional spin chains with open boundary condition, the local gauge twists are always gauged away. Using the gauge transformation of the string type, we obtain the Berry phase as γ j,j+1 = 2 j i=1 S i π assuming the energy gap. It is consistent with the generic VBS state [21]. As for the S = 1 spin chain with open boundaries, the Haldane gap corresponds to the energy gap above the Kennedy triplet [13]. Then the Berry phase of the low energy cluster below the Haldane gap, which includes contribution of the edge states, gives vanishing Berry phase. It should be distinguished from the translationally invariant case without edge [19,25]. Further this present argument is also applicable for systems with charge degrees of freedom. Let us consider a fermion model with conserved particle number n tot = i n i such as spinless fermions with interaction H = ij (tc † i c j + h.c. + V ij n i n j ), where n i = c † i c i and c i is a fermion annihilation operator at site i. The U (1) gauge twist against the charge degree of freedom is introduced by replacing a hopping at the special link (ij) as e iφ c † i c j and the U (1) local gauge transformation is given by Then the transformation property of the Berry phase under the gauge transformation leads the relation φ\|n i \|φ dφ as the case of spins and A is a unit layer. Following the same argument as the spin case, the translational symmetry gives a requirement for the gap under the twist as where ρ is an average particle number per unit layer and N is a number of the unit layers. It has also a non-Abelian extension for the degenerate multiplet which is just a repetition of the spin case. It requires that, when the filling is ρ = p/q with mutual co-prime p and q, there exists a multiplet of M states in the low energy spectrum to form a cluster, which is separated from the else under the twist as M = qℓ, ℓ = 1, 2, · · · , ρ = p/q. [27,32,33] Reflection symmetry We may further apply the present argument for the generic symmetry, such as a reflection symmetry. Consistency between the possible Berry phases and the reflection symmetry of the physical system requires some restriction. Let us consider a reflection symmetric system consisting of two subsystems R and L which are mirror images of one another. We first choose a set of sites A where we perform a gauge transformation. A itself is chosen to be reflection symmetric. See FIG.2(a) for the simplest example. Then we define the Berry phase γ L by twisting some links on the boundary of A: ∂A. Note that the self-reflection-symmetric bond, the mirror image of this bond is itself, is not twisted. We denote this Berry phase as γ L , and the symmetric partner of γ L as γ R . Our gauge transformation of the Berry phase results in γ L = −γ R + 2π j∈A (S j −m j ). The relations −γ R ≡ γ R , mod 2π andm j = 0 hold if the time reversal symmetry is also present. In such case, we obtain that γ L = γ R + 2π A S j . Since the reflection symmetry of the physical structure implies γ L ≡ γ R , mod 2π, when the following case, we predicts a level crossing during the twisting. Our argument can be extended without any difficulty for generic reflection symmetric models including even threedimensional ones. Numerical results showing the pattern of level crossings are given in Fig.2(b). Note that there is the ground state degeneracy at φ = π. Similarly to the translational symmetric case, this degeneracy can also be explained by mutually anticommuting symmetry operations U A (π)R and F if there is no magnetic field and j∈A S j ∈ Z, where R denotes the reflection. Similar arguments can be applied to the molecular magnets with Dzyaloshinsky-Moriya(DM) interactions. We can also apply this argument for the Majumdar-Ghosh model of length 4n + 2 (n ∈ N) with a periodic boundary condition [34]. It gives a gap closing under the twist, which is consistent with the doubly degenerate ground state at φ = 0. HK was supported by the Japan Society for the Promotion of Science. YH was supported by Grants-in-Aid	2008-03-21T15:37:16.000Z	2008-03-21T00:00:00.000	{ "year": 2008, "sha1": "bfa6600e6635d5618e4e66daded5830906657557", "oa_license": "CCBYNCND", "oa_url": "https://tsukuba.repo.nii.ac.jp/record/16902/files/PRB_78-5.pdf", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "bfa6600e6635d5618e4e66daded5830906657557", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics" ] }
203405645	pes2o/s2orc	v3-fos-license	Orostachys japonicus exerts antipancreatic cancer activity through induction of apoptosis and cell cycle arrest in PANC‐1 cells Abstract Targeted therapy at the molecular level is important for pancreatic cancer treatment. This study looked over the anticancer activity of Orostachys japonicus in a human pancreatic cancer cell line, PANC‐1. An ethyl acetate fraction containing quercetin, kaempferol, and flavonol glycosides from O. japonicus (OJE) exhibited significant anticancer activity against the PANC‐1. OJE activated caspase‐3, caspase‐8, and caspase‐9, leading to the induction of both intrinsic and extrinsic apoptosis pathways. It also inhibited cyclin D1, cyclin B1, and cyclin‐dependent kinase 4, representing cell cycle arrest at both G1/S and G2/M phases. In addition, OJE phosphorylated MAPKs such as p38, JNK, and ERK, which are important upstream signaling factors in apoptosis and arrest of cell cycle inducing system. In conclusion, OJE effectively exerted antipancreatic cancer activity via induction of apoptosis directed by both intrinsic and extrinsic pathways and arrest of cell cycle regulated at both G1/S and G2/M stages, which is activated by MAPKs, p38, JNK, and ERK. Apoptosis is an important pathway that is triggered by both pathological and physiological conditions. Dysfunction of apoptosis leads to the development of various cancers in humans, whereas regulation of apoptosis can avoid the same. Accumulating evidence suggests that apoptosis or cell cycle arrest is closely related to anticancer therapy. In apoptosis, biochemical events lead to distinct changes in the cell, including shrinkage, blebbing, DNA fragmentation, and chromatin condensation (Kaufmann & Earnshaw, 2000) through intrinsic and extrinsic stimuli (Call, Eckhardt, & Camidge, 2008;Kacar et al., 2017;Olechowska-jarząb, Ptak-belowska, & Brzozowski, 2016). The intrinsic pathway functions in response to cell damage and cellular stress stimuli. Following the reception of these signals, the ratio of Bcl-2 family members in mitochondrial outer membrane is altered, affecting cytochrome c release and resulting in activation of procaspase-9 (Antonsson, 2004;Jiang & Wang, 2004). The extrinsic pathway is triggered by extracellular stimuli through death receptors at the cell surface, and its activation leads to the activation of procaspase-8. The subsequent activation of intrinsic and extrinsic pathway initiates the caspase cascade, causing the activation of effector caspase-3 (Ashkenazi, 2008;Wong, 2011). Changes in the cellular components during the cell cycle depend on the inhibition or induction of apoptosis in cancer cells (Evan & Vousden, 2001). The cell cycle consists of four phases: G1 (Gap1), S (synthesis), G2 (Gap2) (collectively known as interphase), and M (mitosis). There are also three checkpoints: G1/S, G2/M, and metaphase. These checkpoints regulate the proper functioning of the cell cycle. CDKs are key mediators of cell cycle checkpoints and regulate the cell cycle (Williams & Stoeber, 2012). The MAPK family comprises p38 mitogen-activated protein kinase (p38), c-Jun N-terminal kinase (JNK), and extracellular signal-regulated kinase (ERK; Lunghi et al., 2017). Studies on cell death have shown that MAPKs are involved in the induction of apoptosis and are formed by various stimuli (Chen, Meyer, & Tan, 1996;Sebolt-leopold & Herrera, 2004 \| Reagents All reagents for cell culture and Western blotting were of the highest quality or analytical grade available. \| HPLC analysis HPLC was performed using an Agilent 1100 series system according to the protocol of the manufacturer. \| Cell line and culture The PANC-1 and CAPAN-1 human pancreatic cancer cell lines were purchased from the Korean Cell Line Bank and cultured similarly as described in the previous study . PANC-1 cells were cultured in DMEM, and CAPAN-1 cells were cultured in RPMI containing 10% heat-inactivated FBS, penicillin, and streptomycin. Both cells were incubated in a cell culture dish and maintained in a humidified atmosphere containing 5% CO 2 and 95% air at 37°C. \| MTS assay Inhibition of cell growth was assessed using a CellTiter 96 AQueous One Solution Cell Proliferation Assay Kit (Promega) according to the manufacturer's manual. MTS assay was carried out similarly as described previously (Ryu et al., 2018). \| Cell cycle analysis Cell cycle phases were assessed by staining DNA fragments with PI using a Cell Cycle Phase Determination Kit (Cayman Chemical) according to the manufacturer's protocol. Cell cycle analysis was conducted similarly as performed previously (Ryu et al., 2018). \| Western blotting analysis The cells were mixed with different concentrations of OJE and harvested using a cell scraper and then resuspended on ice for 30 min, followed by removal of cell debris by centrifugation at 10,000 g for 10 min. Protein concentrations were measured using the bicinchoninic acid (BCA) protein assay (Pierce). Protein samples were separated on 10%-15% SDS-polyacrylamide gels through electrophoresis (Bio-Rad) and transferred onto a polyvinylidene difluoride (PVDF) membrane. The membrane was incubated for 2 hr at room temperature with 1:5,000 dilutions of the secondary antibody (horseradish peroxidase-conjugated goat anti-rabbit IgG) and washed in phosphate-buffered saline with Tween-20 (PBST) thrice. \| Effects of OJE on cell proliferation Cell viability was analyzed to examine the effects of OJE by treating PANC-1 cells with varying concentrations of OJE (6.25, 12.5, 25, 50, and 100 μg/ml) for 12, 24, and 48 hr. The effect of treatment with OJE on the proliferation of PANC-1 cells was measured using the MTS assay. OJE significantly inhibited the proliferation of PANC-1 cells in a concentration-dependent manner (Figure 3). \| Effects of OJE on apoptosis in PANC-1 cells To examine whether the antipancreatic cancer effects of OJE were due to the induction of apoptosis, PANC-1 cells were treated with OJE and apoptotic cells were observed by staining with DAPI and annexin V/PI. The results of DAPI staining demonstrated that OJE induced apoptosis of PANC-1 cells in a dose-related manner, indicated by chromatin condensation and nuclear fragmentation in several PANC-1 cells (Figure 4). In flow cytometry results, the lower left quadrants in Figure 5a depict a viable normal cell group without damage. The lower right exhibits cells undergoing early apoptosis with phosphatidylserine (PS), a marker of early apoptosis, externalized to the outer layer of the membrane and DNA, another marker of late apoptosis, which was not stained by PI. The upper right displays late apoptosis. The total apoptotic rate (79.86%) of treated cells was greater than that (33.24%) of the control. Overall, these results represent that OJE treatment increased the total apoptotic rate in a dose-dependent manner (Figure 5a,b,c). \| Effects of OJE on regulation of apoptosisrelated proteins in PANC-1 cells Activation of caspases in cancer cells is closely associated with apoptosis. Hence, we determined whether OJE activated caspases in PANC-1 cells. Western blotting showed that OJE effectively activated procaspases; decreased the levels of procaspase-3, procaspase-8, and procaspase-9; and increased the levels of cleaved caspase-3, cleaved caspase-9, and cytochrome c in a concentration-dependent manner, compared to the control in PANC-1 cells (Figure 6a,b). On the contrary, procaspase-3 activated by OJE was completely inhibited by Z-VAD-FMK (20 μM) in PANC-1 cells ( Figure 6c). This result indicates that apoptosis induction in PANC-1 cells by OJE correlates with the cascade-type activation of caspases. \| Effects of OJE on cell cycle arrest in PANC-1 cells The cell cycle distribution of PANC-1 cells was analyzed by flow cytometry. As exhibited in Figure 7, the population of control cells in the sub-G1 phase was 1.57%, while the population of cells administered with 50 μg/ml of OJE was 32.89%. Hence, the population of cells treated with OJE increased in a concentration-dependent manner in the sub-G1 phase. Simultaneously, the population of the G1 and the G2/M phase cells decreased in OJE-treated PANC-1 cells, compared to the control (Table 1). In addition, the protein levels of CDK4, cyclin D1, and cyclin B1 measured by Western blotting decreased in a dose-related manner (Figure 8). The decreased expression of CDK4 and cyclin D1 is believed to correlate with the decrease of cells in the G1 phase, and the decreased expression of cyclin B1 is thought to correlate with the decrease of cells in the G2/M phase. OJE (6.25,12.5,25,50, and 100 μg/ml) for 12, 24, and 48 hr. The data are expressed as mean ± SD (n = 3). p < .05, p < .001 versus control \| Effects of OJE on the regulation of MAPKs in PANC-1 cells We performed Western blotting to analyze the level of MAPK, which is known to affect upstream signaling pathways. As shown in Figure 9a cell surface, and the treatment of PI enables detection of early and late apoptotic cells (Mahassni & Al-reemi, 2013). Flow cytometry indicated significant concentration-dependent increase in apoptotic cell death and the sequential events observed during early and late apoptosis in PANC-1 cells, especially at higher concentrations of OJE ( Figure 5). In apoptosis, the intrinsic and extrinsic pathways are categorized based on the origin of the initiating signal (inside or outside the cell). In our study, PANC-1 cells treated with OJE raised the levels of cleaved caspase-3 and cleaved caspase-9, indicating the intrinsic apoptosis pathway, and lowered the level of procaspase-8 in a doserelated manner, indicating extrinsic apoptosis ( Figure 6). Activation of caspase-9, caspase-8, and caspase-3 in PANC-1 cells suggests that apoptosis was elicited via the mitochondria-dependent-and extrinsic mediator-dependent-caspase cascade pathways. \| D ISCUSS I ON Because cancer involves abnormal cell growth and proliferation, arrest of cell cycle is another effective marker of anticancer activity (Singh et al., 2004). To observe the stages of the cell cycle, PANC-1 cells were marked with PI and quantitated using flow cytometry. The number of OJE-treated cells in the sub-G1 phase of the cell cycle was F I G U R E 6 Effects of OJE on the expression of apoptosis-related proteins in PANC-1 cells. OJE decreased the level of procaspase-8 and procaspase-3, and increased the level of cleaved caspase-3 in PANC-1 cells in a dose-dependent manner (a). OJE decreased the level of procaspase-9 and increased the level of cytochrome c and cleaved procaspase-9 in PANC-1 cells in a dose-dependent manner (b). Z-VAD-FMK decreased the level of OJE-induced cleaved procaspase-3 in PANC-1 cells (c). Control cells or cells were treated with 0.1% DMSO, 12.5, 25, and 50 μg/ml of OJE for 12 hr. The expression of indicated proteins was examined by Western blotting. The data are expressed as mean ± SD (n = 3). p < .05, ** p < .001 versus control, ## p < .001 versus OJE-treated PANC-1 PANC-1 cells. The major mechanism of arrest of cell cycle is the regulation of the activity of protein kinase family, also known as CDKcyclin complexes (Shackelford, Kaufmann, & Paules, 1999). Cyclin D1 is essential for inducing advancement into the G1 phase. The CDK4/cyclin D1 complex is required at the beginning of the early G1/S phase (Figure 8a). The CDK1/cyclin B1 complex leads to the advancement from G2 to M stage of the cell cycle. CDK1 is mainly activated in association with cyclin B1 during the G2/M phase. The expression of cyclin B1 was reduced when the cells were treated with OJE ( Figure 8b). In addition, the MAPK pathway, including p38, JNK, and ERK, has been recognized in the process proliferation, differentiation, survival, and death of cells (Wada & Penninger, 2004). ERK is associated with both cell proliferation and cell survival. JNK and p38 are activated in response to cellular stress and cellular damage (Prasad, Vaid, & Katiyar, 2012). OJE elevated the phosphorylation of p38, JNK, and ERK in a dose-related manner (Figure 9a,b,c). F I G U R E 9 Effects of OJE on phosphorylation of MAPKs in PANC-1 cells. OJE increased p38 phosphorylation in PANC-1 cells in a dose-dependent manner (a). OJE increased JNK phosphorylation in PANC-1 cells in a dose-dependent manner (b). OJE increased ERK phosphorylation in PANC-1 cells in a dose-dependent manner (c). U0126 decreased OJE-induced ERK phosphorylation in PANC-1 cells (d). In summary, OJE effectively induces apoptosis directed by intrinsic and extrinsic caspase cascade pathway, arrest of cell cycle regulated at G1/S and G2/M stage, and activation of MAPK upstream signaling pathways, including p38, JNK, and ERK in PANC-1 cells. Therefore, these results suggest that flavonoid-rich OJE harboring quercetin, kaempferol, and flavonol glycosides such as afzelin, astragalin, quercitrin, and isoquercitrin from O. japonicus can be utilized as potential anticancer agents for patients with pancreatic cancer. CO N FLI C T O F I NTE R E S T The authors declare no conflict of interest.	2019-09-19T09:10:39.799Z	2019-09-13T00:00:00.000	{ "year": 2019, "sha1": "5567c650550f457bcbded7b10fbf1b3ca6e039b5", "oa_license": "CCBY", "oa_url": "https://onlinelibrary.wiley.com/doi/pdfdirect/10.1002/fsn3.1207", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "ee57f7ee27bc77fe388e9a64e687fbb12bffe788", "s2fieldsofstudy": [ "Biology", "Chemistry" ], "extfieldsofstudy": [ "Chemistry", "Medicine" ] }
202701908	pes2o/s2orc	v3-fos-license	Serologic Evidence of Exposure to Highly Pathogenic Avian Influenza H5 Viruses in Migratory Shorebirds, Australia Highly pathogenic avian influenza (HPAI) H5Nx viruses of the goose/Guangdong/96 lineage continue to cause outbreaks in poultry and wild birds globally. Shorebirds, known reservoirs of avian influenza viruses, migrate from Siberia to Australia along the East-Asian-Australasian Flyway. We examined whether migrating shorebirds spending nonbreeding seasons in Australia were exposed to HPAI H5 viruses. We compared those findings with those for a resident duck species. We screened >1,500 blood samples for nucleoprotein antibodies and tested positive samples for specific antibodies against 7 HPAI H5 virus antigens and 2 low pathogenicity avian influenza H5 virus antigens. We demonstrated the presence of hemagglutinin inhibitory antibodies against HPAI H5 virus clade 2.3.4.4 in the red-necked stint (Calidris ruficolis). We did not find hemagglutinin inhibitory antibodies in resident Pacific black ducks (Anas superciliosa). Our study highlights the potential role of long-distance migratory shorebirds in intercontinental spread of HPAI H5 viruses. is that shorebirds play a limited role in AIV epidemiology and long-distance dispersal, explaining the absence of gs/ GD lineage HPAI H5 viruses on the continent of Australia. In contrast to shorebirds, waterfowl in Australia are largely nomadic species that do not migrate outside the Australian-Papuan zone (23). We examined whether long-distance shorebird migrants were exposed to gs/GD lineage viruses. We used the red-necked stint (Calidris ruficolis), which uses Australia as a nonbreeding area, as a model migratory species. The red-necked stint has known stopover locations in East and Southeast Asia, where HPAI virus is endemic. We contrasted findings from red-necked stints with those from the resident Pacific black duck (Anas superciliosa), a nonmigratory dabbling duck believed to be a natural reservoir for LPAI virus in Australia. Ethics Statement We received study approval from Deakin University Species and Sample Collection We targeted mixed flocks of shorebirds for capture with cannon nets as part of a long-term ringing scheme. Since 2011, these birds also have been used for avian influenza surveillance (24). Red-necked stints consistently are captured in large numbers during October-March each year, predominantly in the state of Victoria. We also opportunistically collected samples from Western Australia, Northern Territory, and Queensland as part of ringing expeditions. Samples from these locations are not central to the longterm avian influenza surveillance project. Because the rednecked stint is in Australia during October-March, we analyzed and reported data for this species by using the austral summer season. We captured resident Pacific black ducks by using either baited funnel walk-in traps (25) or mist nets. We deployed walk-in traps on shorelines and baited them with a seed mix. We set these traps before dawn and operated them during the day; at night, we left traps open so birds could enter and leave freely. To capture waterbirds at night, we erected mist nets on poles above the water surface. We collected most samples from the state of Victoria but also collected samples from South Australia and New South Wales. After capture, we individually banded all birds with a metal ring with a unique identifier and collected <200 μL of blood from the brachial vein by using the Microvette 200 Z (Sarstedt, https://www.sarstedt.com) capillary blood collection system. We released all birds after banding and collecting blood samples. We stored blood samples at 4°C-8°C until we separated serum by centrifugation 12-24 hours after sampling. We collected 1,531 serum samples from red-necked stints and 394 serum samples from Pacific black ducks for this study. General AIV Immunity We screened serum samples for nucleoprotein (NP) antibodies to ascertain general AIV seroprevalence. We assessed NP antibodies by using a commercially available ELISA, MultiS-Screen Avian Influenza Virus Antibody Test Kit (IDEXX, https://www.idexx.com), following the manufacturer's recommendations, where a sample/negative (S/N) ratio of <0.5 indicates a positive result. We considered S/N ratios of 0.5-0.6 inconclusive, although this ratio has been demonstrated to correspond to antibody presence in wild birds (26,27). We calculated seroprevalence and 95% CI by using the bioconf() function of the Hmisc package in R 3.5.1 (https://www.r-project.org). Hemagglutinin Inhibition Assay After NP antibody screening, we assayed positive and inconclusive serum samples for H5 antibodies by using a hemagglutinin inhibition (HI) assay with 1% vol/vol chicken erythrocytes. We selected 7 contemporary HPAI H5 viruses from different gs/GD lineage clades and 2 LPAI H5 viruses endemic to Australia as antigens (Table). We could only test up to 8 antigens per sample because we could collect only a small volume of serum from rednecked stints; for some samples, we could only test against 4 relevant viruses. We selected representative H5 viral lineages because of their known spatial and temporal distribution and availability of reference viral antigens, such as those selected by the World Health Organization (WHO) as candidate vaccine viruses (CVVs; http://www.who.int/influenza/vaccines/virus/candidates_reagents/a_h5n1/en/) for pandemic preparedness. WHO's CVVs are 6:2 recombinant viruses on an A/Puerto Rico/8/1934(H1N1)(PR8) backbone with the multibasic cleavage site removed. The 2 LPAI H5 viruses from Australia were gamma-irradiated antigens. We conducted a hemagglutinin assay on selected antigens to determine virus titer, which we then added to HI plates at a dilution of 4 hemagglutinin units. We treated all NPpositive ELISA field samples with a Vibrio cholerae receptor-destroying enzyme (RDE II; Denka Seiken Co., https://denka-seiken.com), then inactivated samples with 1.5% sodium citrate. We raised control antiserum against all virus antigens, except the LPAI viruses A/duck/Victoria/0305-2/2012(H5N3) and A/wild bird/Queensland/P17-14428-30-01/2017(H5N1), in 6-18-month-old ferrets. In brief, we inoculated ferrets intranasally with 1 mL of virus; at 14 days postexposure, we boosted ferrets by intramuscular delivery of a concentrated dose of the same virus into the hind leg; and at 21 days postexposure, we collected a terminal blood sample. We monitored ferrets' weights, temperatures, and clinical signs throughout. We used antibodies for all 7 H5 viruses in each assay to measure both homologous titers and cross reaction; we also ran antibodies without virus to assess nonspecific agglutination. We serially diluted all serum samples across assay plates, starting with a titer of 1:20, and calculated specificity of antigen-antibody agglutination (Appendix Table 1, http:// wwwnc.cdc.gov/EID/article/25/10/19-0699-App1.pdf). Population Immunity to AIVs During 2011-2018, we collected 1,531 serum samples from red-necked stints, ≈200 samples per year, most from Victoria. Overall, 20% of red-necked stints were seropositive for NP antibodies, with variations among collection years and locations ( Figure 1, panel A; Appendix Table 2). We collected 394 blood samples from Pacific black ducks during 2011-2018. Temporal structure of the data for this species was more variable, with few samples collected during 2015-2017 (Appendix Table 3). We only collected samples from the southeastern states of Australia. Overall, ≈55% of Pacific black ducks sampled were seropositive for NP antibodies. We experienced some variation across sampling events, but average seropositivity was similar across locations ( Figure 1, panel B). Differences in Exposure to HPAI H5 Virus in Migratory and Resident Birds We assayed 307 NP ELISA-positive or -inconclusive serum samples from red-necked stints and 240 from Pacific black ducks for antibodies against H5 viruses by HI assay (Appendix Tables 2, 3). Of HI-positive serum samples, ≈12% were inconclusive by NP ELISA. Because of the small volume of serum collected from red-necked stints, we could assay only 33 serum samples for <4 antigens each (Appendix Table 2). Nonetheless, 23 red-necked stint serum samples contained detectable HI antibodies against >1 of the 7 HPAI H5 virus antigens tested (1.5%, 95% CI 1.0%-2.3%) (Figure 2 panel A). We detected HI antibodies against antigens belonging to clade 2.3.4-derived lineages, Table 2). Overall, HI titers were low; 9/23 serum samples had an HI titer of 20 and 14/23 an HI titer of 40. One serum sample had HI antibodies against the LPAI H5 virus A/duck/ Victoria/0305-2/2012(H5N3) (Figure 2, panel A). Overall, no red-necked stint samples were positive for both HPAI and LPAI virus antigens. Discussion Despite intercontinental spread of gs/GD lineage HPAI H5Nx viruses from Asia to Europe, Africa, and North America, we have no evidence that incursions of these viruses have occurred in Australia. A leading hypothesis for the lack of incursion is the absence of Anseriformes birds migrating between Asia and Australia (23,29). However, millions of shorebirds that are reservoirs for AIV migrate from Siberia to Australia, with stopover sites along the coast of East Asia (15)(16)(17)29). We demonstrated that these intercontinental migratory birds have been exposed to gs/GD lineage HPAI H5Nx viruses and have the potential to bring these viruses into Australia. The absence of HI antibodies against gs/GD lineage HPAI H5Nx viruses in a widespread and abundant Anseriformes birds in Australia and the lack of detection Table 4, https://wwwnc.cdc. gov/EID/article/25/10/19-0699-App1.pdf). B) For Pacific black duck, year represents calendar year. White indicates untested serum samples; gray indicates a titer <20, the starting titer for this assay; blue indicates hemagglutinin inhibition (HI) antibodies, and shades vary depending on HI titer . Sample numbers are ordered by collection year and sequentially from left to right in the order in which individual birds were caught. Antigens used in this study are on the y-axis, and abbreviated with relevant clade information; full strain names are available in the Table. NT, no titer. Greater detail on positive samples appears in Appendix Figure 1. during ongoing surveillance activities (12) suggest that a virus incursion has not occurred yet. Overall, red-necked stints we sampled had low prevalence (≈20%) of NP antibodies, and 1.5% of all serum samples contained HI antibodies against gs/GD lineage HPAI H5Nx virus antigen. We detected the highest seroprevalence of gs/GD lineage HPAI H5Nx virus HI antibodies, 4.5% of all serum samples collected, during the 2016-17 austral summer. A previous study in northwestern Australia during 1992Australia during -2009 showed that the red-necked stint and other members of the Scolopacidae family had higher AI virus seroprevalence than other shorebird species tested. Furthermore, H5 HI antibodies were common; 31/260 NP ELISA-positive serum samples had HI titers against HPAI H5N1 virus clade 1 A/chicken/Vietnam/8/2004 (21). Similarly, serum samples from ducks during this period also had HI antibodies against this clade but not HPAI H5N1 clade 2 viruses (21,30). One explanation for the lack of evidence for circulation of HPAI H5N1 virus clade 1 during this time is that exposure to endemic H5 virus strains in Australia produces HI antibodies with broad serologic cross-reactivity (24,31). We found no evidence of cross-reactivity in control antibodies (Appendix Table 1) or cross-reaction in any positive serum samples, including no cross-reactivity between LPAI and HPAI virus antigens. Furthermore, the clades we detected HI antibodies against, 2.3.4.2 and particularly 2.3.4.4, are antigenically distant from previously circulating H5 viruses (32), so LPAI virus crossreactivity is unlikely. Long-distance migratory shorebirds captured in Australia could have been exposed to HPAI H5 virus in the northern hemisphere. Indeed, a red-necked stint tested positive for HPAI H5N6 virus in Hong Kong in 2016 on its southward migration (pers. comm.), strengthening evidence of gs/GD H5Nx virus exposure in this species. Studies of ducks in Europe and Mongolia provide further perspective. Gilbert et al. demonstrated the presence of HI antibodies against gs/GD lineage HPAI H5N1 virus in waterfowl in Mongolia. These birds had higher serologic reactivity to HPAI H5 virus than to LPAI H5 virus antigens. That study found limited or no evidence of exposure to HPAI virus antigens in a small representation of waterfowl in Europe (31). However, Gilbert The prevalence of HI antibodies we detected in rednecked stints during the 2016-17 season were comparable to those reported in ducks in Europe, even though red-necked stints have a much lower seroprevalence of AIV in general. Some studies suggest that long-lived avian species, such as swans and seabirds, retain HI antibodies over the course of many years, which could enable expansion of antibody breadth, increasing the number of subtypes against which these birds have antibodies over time (35,36). An additive effect could explain why mute swans maintained high rates of HPAI H5N8 virus HI antibodies after AIV outbreaks in the United Kingdom (34). In contrast, ducks are believed to have relatively poor immune memory and to retain HI antibodies only briefly (37,38). The expected antibody longevity patterns in shorebirds such as the red-necked stint is unknown, but given the relatively high prevalence of HI antibodies, particularly during 2016-17, we hypothesize that shorebirds retain antibodies longer than ducks. We saw generally low HI titers in serum samples from red-necked stints; 82% of serum samples with detectable HI antibodies had titers <40. Gilbert et al. also reported low titers and hypothesized that tested waterfowl were exposed months or years previously (31). Alternatively, the antigens might not have matched the antibodies tested. Waterfowl species comprise the bulk of avian species sampled in most surveillance schemes for avian influenza and are sampled heavily for H5Nx viruses (33,39). Shorebirds are central to the ecology of AIV (13) but are tested rarely beyond the study from Delaware Bay, NJ, USA, and infection prevalence is much lower (18)(19)(20)(21)(22) than in waterfowl (13). However, virology and serology data from Delaware Bay suggest both ruddy turnstones (Arenaria interpres) and red knots (Calidris canutus) are exposed to a large diversity of hemagglutinin subtypes, and <36% of birds tested had neutralizing antibodies against multiple subtypes, demonstrating host competency (40). Furthermore, migratory shorebirds have been implicated in the long-distance movement of LPAI viruses (41). Experimental studies have shown limited morbidity and mortality rates associated with infection of some 2.3.4.4 subclades in ducks, demonstrating their ability to act as migratory vectors for these viruses (11,(42)(43)(44)(45). Our understanding of infection and pathogenesis of HPAI H5Nx virus in shorebirds is extremely limited. Experimental exposure of dunlins (Calidris alpina) and ruddy turnstones to HPAI H5N1 virus clade 2.2 resulted in contrasting outcomes (46,47). Immunologically naive dunlins showed clinical signs of infection, and 19/20 birds receiving high or mild doses of the virus died. Birds inoculated with low doses did not get infected (46). Ruddy turnstones, in contrast, were not immunologically naive, none died, and birds infected with LPAI and HPAI had similar patterns of viral shedding (47). The authors attribute the contrasting results between dunlins and ruddy turnstones to the immunological status of birds, suggesting that cross-immunity played a key role in limiting clinical disease (47,48). In these experiments, infection rates were the same between birds that were seropositive (subtype unknown) before capture, were first exposed to a LPAI H5 virus strain, or were first exposed to an H7 virus strain, suggesting both homosubtypic and heterosubtypic immunity could play a role in protection (47,48). However, in ducks, phylogenetic distance between hemagglutinin subtypes plays a role in the degree of the heterosubtypic protective response (49), and closely related hemagglutinin subtypes likely drive protection. Red knots were more susceptible to acquiring HPAI H5N1 virus, especially clade 2.2.1, and shed higher viral titers before the onset of clinical disease during the migratory phase because of increased plasma corticosterone (50). The authors saw no difference in susceptibility to disease between birds in premigration, fueling, or migratory phases and suggested that, assuming no effect of subclinical exposure on the likelihood of migratory takeoff, red knots could spread HPAI H5 virus through migration (50). These studies demonstrate shorebirds could be exposed, survive infection, and potentially disperse HPAI H5 virus over long distances during their migratory phase. In conclusion, we demonstrated that the long-distance migratory red-necked stint, which spends nonbreeding seasons in Australia, has been exposed to HPAI H5 virus clade 2.3.4.4. We did not detect antibodies against this or other HPAI viruses in our sample of resident Pacific black ducks, suggesting exposure has not occurred in Australia. However, our study highlights the potential for migratory shorebirds to spread HPAI H5 viruses, which should inform future avian influenza surveillance.	2019-09-17T02:58:39.550Z	2019-10-01T00:00:00.000	{ "year": 2019, "sha1": "c88e079be68eb39cbad2b8245e94a897de0664df", "oa_license": "CCBY", "oa_url": "https://wwwnc.cdc.gov/eid/article/25/10/pdfs/19-0699.pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "93954322f17bd47b0ec0cfec805f9e0403669fa3", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [ "Biology", "Medicine" ] }
259832336	pes2o/s2orc	v3-fos-license	Shark teeth zinc isotope values document intrapopulation foraging differences related to ontogeny and sex Trophic ecology and resource use are challenging to discern in migratory marine species, including sharks. However, effective management and conservation strategies depend on understanding these life history details. Here we investigate whether dental enameloid zinc isotope (δ66Znen) values can be used to infer intrapopulation differences in foraging ecology by comparing δ66Znen with same-tooth collagen carbon and nitrogen (δ13Ccoll, δ15Ncoll) values from critically endangered sand tiger sharks (Carcharias taurus) from Delaware Bay (USA). We document ontogeny and sex-related isotopic differences indicating distinct diet and habitat use at the time of tooth formation. Adult females have the most distinct isotopic niche, likely feeding on higher trophic level prey in a distinct habitat. This multi-proxy approach characterises an animal’s isotopic niche in greater detail than traditional isotope analysis alone and shows that δ66Znen analysis can highlight intrapopulation dietary variability thereby informing conservation management and, due to good δ66Znen fossil tooth preservation, palaeoecological reconstructions. T he lineage of the Elasmobranchii originated in the Devonian Period 1,2 , and its members have since played major roles in marine ecosystems occupying trophic positions from primary consumers to apex predators 3 . In modern marine food webs, most sharks play the roles of apex and mesopredators, and their removal is likely to result in large-scale, but still uncertain, ecological consequences [4][5][6][7] . Many shark species experience global population declines, attributed to fishing, habitat degradation, climate change and pollution [8][9][10] . Conservation efforts will depend on a better understanding of extant shark ecology but could also be informed by how ancient species adapted to comparable environmental and climatic pressures in the past. Stable isotope analysis can be a powerful tool to decipher an animal's life and migration history, including dietary and habitat preferences. As such, stable isotope studies are becoming more widespread in shark ecology, especially given the difficulty of directly monitoring individuals or populations 11 . Bulk carbon and nitrogen isotope values (δ 13 C, δ 15 N) analysed in an animal's organic tissues, including bone and tooth collagen, are the most applied dietary and trophic level proxies [12][13][14] . Generally, both δ 13 C and δ 15 N values increase with trophic level due to the preferential loss of the lighter isotopes 12 C and 14 N in respiration and urea, respectively 15,16 . However, δ 13 C values behave more conservatively, increasing~1‰, or not at all, per trophic level compared to δ 15 N values, which increase on average by 3.4‰ per trophic level in most tissues [17][18][19] . Thus, δ 13 C values are more commonly applied to identify the primary producer(s) in a food web 20,21 , whereas δ 15 N values are used to estimate trophic levels. The application of bulk δ 15 N values as a trophic level proxy relies on the assumption of predictable, constant nitrogen isotopic fractionation factors between the diet and the examined consumer tissue. Yet, physiological and environmental factors can also influence diet-tissue fractionation factors shifting them from their predicted values [22][23][24][25][26] . In addition, the traditional bulk collagen δ 13 C and δ 15 N method used in an archaeological and palaeontological context is not applicable to 'old' fossil material (>100,000 years) as it is limited by the degree of protein preservation 27 . In recent years, other non-traditional dietary/ trophic isotopic proxies have emerged that can complement the traditional δ 15 N proxy to provide additional otherwise inaccessible information, as well as act as a substitute in ancient sample material with poor or no collagen preservation [28][29][30][31][32][33][34] . Among these methods, zinc isotope ratios ( 66 Zn/ 64 Zn expressed as δ 66 Zn) were recognised as a particularly promising trophic level proxy in mammals 29,35 and fish 33 . Pristine, i.e., diet-controlled δ 66 Zn values can be preserved in shark tooth enameloid for millions of years, making it an effective tool to investigate deep-time shark ecology and evolution 33 . Zinc is supplied into vertebrate tissues through diet 36,37 and undergoes mass-dependent fractionation within an organism [38][39][40] . Zinc intra-organism isotopic fractionation is tissue-specific, depending on the Zn coordination environment [38][39][40] . Heavier Zn is typically concentrated in stiffer bonds, i.e. heavier Zn preferentially binds to ligands with a stronger electronegativity (oxygen>nitrogen>sulphur) 38,39,41 . This preference results in a depletion of 66 Zn in organs and muscles relative to the animal's diet leading to successively lower δ 66 Zn values as trophic levels increase. For ecological, archaeological, and palaeontological studies, δ 66 Zn is commonly measured in the mineral phase of skeletal tissues (bioapatite) of bones and teeth [29][30][31][32][33]35 . Notably, there appears to be a predictable offset in δ 66 Zn values when analysing different bioapatite tissues of the same individual, with bone/tooth dentine having values that are on average 0.2‰ higher compared to tooth enamel/enameloid 33,35 . Per trophic level, bioapatite δ 66 Zn values decrease by approximately 0.4‰ in both marine and terrestrial mammals, although the exact trophic level fractionation factors have yet to be determined experimentally 32,35 . Dissolved zinc is highly depleted in the surface ocean with typical concentrations of 0.01-0.5 nmol kg −1 , while deep water usually exhibits higher concentrations, e.g. up to 10 nmol kg −1 in the North Pacific 42 . Despite this nutrient-like Zn distribution, which is a result of biological uptake, there is no consistent covariation between dissolved Zn concentration and isotopic composition 42 . The shallow upper ocean exhibits a large δ 66 Zn variation for dissolved Zn ranging from −1.1 to +0.9‰ 43,44 . This variability and its causes are still the subject of debate (reviewed by refs. 42,45 ). In contrast, dissolved δ 66 Zn values from deep water are near homogenous at approximately +0.45‰ 42,43,46,47 . Marine Zn sources, mainly stemming from rivers and aeolian dust, are also relatively homogenous with δ 66 Zn values around +0.33‰ 48 , close to the average upper continental crust value of +0.3‰ 45,49 . Few studies have investigated zinc isotope variability in marine vertebrates 29,32,33 . These initial results suggest that, within the same tissues analysed (e.g. bone, enameloid), zinc isotope values have a similar range across geographic locations within the same taxa/trophic levels, implying little variation in marine Zn isotope food web baselines, perhaps even on geological timescales 33 . However, large-scale marine food web baseline data derived from primary producers are currently non-existent, which limits any detailed discussions of marine Zn isotope baselines at this time. Notably, δ 66 Zn and δ 15 N are measured in different components of skeletal tissues, bioapatite and collagen respectively, and are metabolically and environmentally independent of each other. As such, a combined analysis of both trophic level proxies offers the potential of gaining additional ecological information, crossverification of results and correction for environmental and/or physiological factors influencing either proxy 50 . The inclusion of δ 66 Zn may also provide a more direct comparability of marine species trophic ecology between spatially and temporally distinct locations than possible by traditional isotope analysis alone. While it remains unknown for δ 66 Zn values, both bulk δ 13 C and δ 15 N values are known to be strongly affected by marine baseline variability 26,51,52 . Here, we test, whether enameloid δ 66 Zn (δ 66 Zn en ) values can be used to investigate foraging differences within a single marine predator population. We analyse tooth dentine collagen δ 13 C and δ 15 N (δ 13 C coll , δ 15 N coll ) values with δ 66 Zn en values of the same tooth from 53 sand tiger shark (Carcharias taurus) individuals that were caught and released in Delaware Bay (New Jersey, USA). Declines in abundance are observed globally for C. taurus 53 . This species is particularly vulnerable to overexploitation due to slow growth, late maturation, and low reproductive output 54,55 . Subsequently, C. taurus is now listed as "Critically Endangered" on the International Union for Conservation of Nature (IUCN) Red List of Threatened Species 56 . Understanding this species feeding and habitat requirements is essential to facilitate conservation and management efforts. Carcharias taurus is a large coastal shark that primarily feeds on teleosts and smaller elasmobranchs 57,58 . This species is the ideal subject for a study such as this; Carcharias taurus is the only extant member of a genus who's now extinct members, along with analogue Odontaspididae (e.g. Striatolamia), were globally distributed and abundant throughout the Cenozoic, dating back to the Cretaceous period [59][60][61] . Carcharias is thus a key genus for palaeontological and conservation palaeobiological studies. In addition, while C. taurus is wide-ranging in its distribution, extant populations are isolated and require conservation efforts that are tailored to the population-specific regional ecology 62 . Carcharias taurus is a migratory species and previous studies suggested differences in the timing and routes of migration between juvenile and adult individuals as well as males versus females within the western Atlantic population 55,63 . Yet, little is known about migration, sex, and size-related differences in their foraging ecology. This multi-proxy isotope study investigates the isotopic variability and niches within this population and associated differences in resource use related to sex, ontogeny and migration patterns. Finally, we discuss the implications of our δ 66 Zn en results for ecological and palaeobiological studies investigating intrapopulation dietary variability. Results and discussion Modern ecological studies often rely exclusively on δ 15 N values from organic tissues to establish trophic levels in marine predators, especially as foraging ecology typically cannot be directly observed. Despite its merits, metabolic and environmental factors may compromise δ 15 N trophic level interpretations [22][23][24][25][26] . In addition, organic substances are commonly not available for fossil studies. Here, we measure δ 66 Zn en , an emerging trophic level indicator preservable in mineralised tissues over geological timescales 33 , together with bulk dental δ 13 C coll and δ 15 N coll values in western Atlantic C. taurus individuals. All three dietary indicators vary substantially with significant correlations with body size, and, thereby, ontogeny ( Fig. 1a-c). The heterogeneous δ 66 Zn en values and their correspondence to body size, and δ 13 C coll and δ 15 N coll values (Fig. 2g, h) indicates that the diet variability of this C. taurus population is well represented by its δ 66 Zn en values. These results demonstrate that zinc isotopes record diet with high fidelity and serve as a trophic level indicator even on an intrapopulation level. Isotope compositions of Carcharias taurus teeth. There is a large isotopic variability among the teeth of different individuals for all three isotopes measured. Collagen δ 13 C values have a range of 2.7‰, from −15.3 to −12.6‰, with a mean of −14.0 ± 0.6‰ (n = 53). Collagen δ 15 N values have a range of 3.6‰, from +13.3 to +16.9‰, with a mean of +14.9 ± 0.8‰ (n = 53). Enameloid δ 66 Zn values have a range of 0.64 ‰, from −0.49 to +0.15‰, with a mean of −0.02 ± 0.13‰ (n = 54). All three isotopes show a significant variation in body size, with relationships decreasing from δ 13 C coll to δ 66 Zn en to δ 15 N coll (Fig. 1a-c). Physiological effects that could affect all three isotope systems for all studied individuals, leading to a correlation with body size, may be related to growth rate or body size-related differences in isotopic fractionation factors. Other physiological effects could only impact individuals or certain groups within the population (e.g. gestating females). Differences in body size may lead to lower δ 15 N trophic fractionation factors in larger individuals compared to smaller ones 64 , but are unlikely to be as pronounced as our results, and while growth rate may impact δ 15 N values for certain tissues, δ 13 C values should not be affected 65 . In addition, neither of these physiological effects appear to influence δ 66 Zn values 30,40,66 . Rather than physiological effects, the correlation of all three isotopes with body size within the western Atlantic C. taurus population must reflect ontogenetic niche shifts related to changes in diet composition and trophic level and/or foraging habitat, and associated variation in isotopic food web baselines. There are noticeable differences in this population's isotopic composition between juvenile females, juvenile males, adult females, and adult males. The four individuals, labelled as adult females, based on their size 67,68 , are most distinct in their mean δ 13 C coll , δ 15 N coll and δ 66 Zn en values (Figs. 1, 2, Table 1 and Supplementary Table 1). All three isotopes have a relationship with total length in females but this pattern varies for males ( Fig. 1). Similarly, δ 13 C coll values show a better relationship with δ 15 N coll and δ 66 Zn en in females than in males (Fig. 3). Enameloid δ 66 Zn values show a significant relationship with collagen δ 15 N values in females but not in males (Fig. 3). Isotopic niche regions (N R ) differ for all four C. taurus groups (i.e. juvenile females, juvenile males, adult females, adult males, Fig. 2). Adult females have the lowest probability to be found in the isotopic niche of other groups (in all cases <20%, Fig. 4). Juvenile females and adult males have the highest probability to be found in each other's isotopic niche region (Fig. 4). Scatter plots include linear regression fits with 95% confidence intervals (grey shaded area) for all data points shown in the respective panel (in all cases, the number of analysed individuals is n = 53). Carcharias taurus individuals are separated by sex (female/male) and total length (juvenile/adult). Plots were produced using the R package nicheROVER 97 . The isotopic niche region is projected as ten random ellipses for each C. taurus group and isotope pair. Error bars in d, g, h represent analytical uncertainty (1 SD) derived from standard replicate analyses. Chronology of tooth formation and incorporated isotopic signals. We document sex and ontogeny-related intraspecific δ 13 C coll , δ 15 N coll and δ 66 Zn en isotope differences within the C. taurus population (Figs. 1, 2), reflecting a variable foraging ecology at the time of tooth formation. As the North Atlantic C. taurus population undertakes large-scale sex and size-segregated seasonal migrations 55,63 , accurate interpretations of tooth isotopic compositions depend on reconstructing the timing of tooth mineralisation. The C. taurus population examined here only spends a short time in Delaware Bay (June-September, October) 55 ; therefore, the chemical composition of their teeth does not reflect the dietary uptake in late summer Delaware Bay 61 . Sharks continuously grow their teeth in a conveyor belt system and the timing of tooth formation from below the epithelial tissue to the outermost, functional position was previously quantified as~260 days for captive Leopard Sharks Triakis semifasciata 13 . Further, this captive study determined the incorporation rate of dietary nutrients into dentine collagen is 32-83 days 13 . While the T. semifasciata has different life history traits and evolutionary history from C. taurus, this captive study provides an order of magnitude for the timing delay and formation rate to interpret meaningful ecological data from isotopic results. All C. taurus individuals and their teeth in this study were collected within a 15-day span and we expect the teeth featured here to represent a similar formation timing given each individual was only sampled once, all teeth come from the functional position, from a single population and all teeth within a row form at the same time 69 . Given the sampling date and timing for a tooth to reach the functional position via the tooth 'conveyor belt' system, we estimate a formation date (i.e. most lingual position, below the epithelium) in December 2011 (e.g. 260 days prior to August 30, 2012, is December 14, 2011). In addition, the isotopic incorporation rate of dental collagen means dietary information is integrated into organic carbon and nitrogen from September or November 2011 (e.g. 32-83 days prior to December 14, 2011). The mineralisation rate of enameloid and incorporation rate of dietary zinc is not characterised for sharks, and we expect our calculated timing of organic nutrient incorporation to be a conservative estimate. Diet and foraging habitat of adult males and juveniles. Carcharias taurus individuals convene in Delaware Bay in the summer months, but there are variable migratory behaviours during other seasons, likely related to sex and reproductive status. Most movement observations of C. taurus are from acoustic tracking, which is confined to the coastal ocean and embayments where receivers are located to detect tagged animals 63,70,71 . From these studies, we know that generally, C. taurus departs the Delaware Bay and Mid-Atlantic region in September and October 63,71 , migrates south in nearshore waters (<15 km from shore), and overwinters off the North Carolina and Florida coast 70,71 . The isotope composition from adult males and juveniles of both sexes in this study likely represents integrated dietary signals after leaving Delaware Bay on the route south from nearshore waters off the coasts of Delaware, Maryland, Virginia and North Carolina 55,63 . Adult males have the largest variability in δ 66 Zn en values spanning from +0.09 to −0.49‰ (n = 13). The expanded isotopic niche of adult males (and higher δ 66 Zn en variability, Fig. 2) suggests a dietary shift associated with ontogeny beyond gape-limited predation alone, which would be associated with increased trophic level. Observations from tagging data indicate that larger males dive deeper 55 where they may have access to prey unavailable to juveniles. The ability of larger-sized individuals within a population to capture higher trophic level prey or forage in deeper depths, thereby affecting their isotopic composition, was also observed for other marine vertebrates 72 . Stomach content analyses of C. taurus from the Eastern Cape, South Africa revealed a broadening of dietary niche with total length; larger individuals consumed a higher taxonomic diversity that included larger prey species 58 . Similarly, South-west Atlantic C. taurus individuals also consume larger benthic elasmobranchs as they reach a total length >190 cm 73 . The larger isotopic niche and more variable δ 66 Zn en values suggest that adult males (Figs. 1i, 2), while opportunistic predators, are indeed able to feed upon a wider range of prey than juveniles, including from different habitats and higher trophic levels. Adult males and juvenile females have a larger overlap of isotopic niches, whereas juvenile males differ from these two groups (Fig. 4). We hypothesise this difference is due to size since adult males are more similar in size to juvenile females in our sample set. In this study, 33% of the juvenile females and 74% of the juvenile males are <180 cm in total length. Adult males sampled were 190 to 251 cm in total length. In South Africa, a few small demersal and benthic fish were important in the diet of the smallest C. taurus individuals 58 . If smaller individuals are more restricted in terms of prey size, this may explain the lower δ 66 Zn en variability in juvenile males compared to adult individuals, which can feed on a wider range of prey sizes and thus trophic levels (Fig. 1i). Hence, differences in an isotopic niche between juvenile sexes may be primarily related to the size of the here examined individuals rather than sexual segregation among juveniles. While δ 66 Zn en variability is low for juvenile males, juveniles of both sexes have a wider variability in δ 13 C coll values compared to adult individuals of the same sex, which could indicate a greater reliance of juveniles on both pelagic and demersal resources. In addition, the lowest δ 13 C coll values, <−14.5‰, are only found in juveniles of both sexes that are <180 cm in total length (Fig. 1a, d, g). This result is in line with a recent study featuring blood plasma δ 13 C and δ 15 N values from mainly smaller juvenile C. taurus from the same population 74 . In it, the authors noted that juveniles with a fork length smaller than 121 cm relied almost equally on pelagic fish and demersal omnivores (fish and squid), whereas juveniles with a larger fork length relied more heavily on demersal omnivores. In shelf environments, pelagic phytoplankton has typically lower δ 13 C values compared to benthic primary producers 75 . Pelagic prey for juvenile C. taurus sampled in Great South Bay, New York, USA also had lower δ 13 C and δ 15 N values compared to demersal prey species 74 . Thus we tentatively interpret lower and more varied δ 13 C values in smaller juveniles (mostly male juveniles) as a greater reliance on lower δ 13 C value pelagic prey than in larger juveniles and adults. However, we cannot exclude the possibility that some variation may also relate to differences in migration timing with ontogeny 55 , where small juveniles may forage in areas with a different baseline. Diet and foraging habitat of adult females. Four individuals, which are the only adult females analysed, are distinct from the other individuals in terms of their isotopic composition for all three proxies (δ 13 C coll , δ 15 N coll and δ 66 Zn en ) and their isotopic niche ( Fig. 2 and Table 1). The likelihood of juveniles of both sexes and adult males to overlap with the isotopic niche of analysed adult females is, in all cases, less than 2% (Fig. 4). Higher δ 13 C coll and δ 15 N coll and lower δ 66 Zn en values coupled with low isotopic variability indicate that adult females generally fed at a higher trophic level and perhaps a narrower prey spectrum than juveniles and adult males. Although the number of adult female individuals is low (n = 4), these individuals are so distinct in terms of their isotopic composition that this trait is unlikely a sample size artefact alone, rather indicating differences in dietary resources. Other than diet, physiological and/or environmental effects may also contribute to the observed isotopic differences. The total length of the four adult females indicates maturity, although it is unclear whether they were gestating at the time of tooth formation. Pregnancy is one potential physiological effect that may cause intraspecific isotopic variability, especially considering adult females are most distinct from the other groups in terms of their isotopic composition for all three proxies (δ 13 C coll , δ 15 N coll , δ 66 Zn en ) and hence their isotopic niche (Figs. 2, 4 and Table 1). Previous studies focused on pregnant sharks of other species found either no differences in δ 13 C and δ 15 N values 76 or lower δ 15 N values during pregnancy 77 . In other vertebrate groups, pregnancy results in lower δ 13 C and δ 15 N values 23,24 . Pregnancy is thus unlikely to cause the higher δ 13 C coll and δ 15 N coll values observed in the adult C. taurus individuals (Fig. 1). The physiological effects of pregnancy on δ 66 Zn en remain largely unexplored, but to date, sex, age and body mass have no effect on δ 66 Zn values in terrestrial mammals, including humans 30,40,66 . While pregnancy may not impact isotopic fractionation, it could cause isotopic distinction with changes in foraging behaviour. For example, captive adult, female C. taurus increased their food consumption significantly during a reproductive year, which may help females prepare for investing nutrition for the large, well-developed pups of this K-selected species 78 . This physiological demand may require that adult females need to access alternative food resources than the rest of the population. There are also behavioural changes in migration noted for adult female C. taurus. Pop-off satellite archival tags revealed that adult females left Delaware Bay and headed 100's km offshore to the edge of the continental shelf 55 . Species distribution models predicted that adult females were likely to occur in waters further offshore than the rest of the population in autumn 63 . It has been hypothesised that this offshore movement of adult females may be related to increased energetic demands for females in preparation for or recovery from pupping 55,63 . Adult females were also observed in North Carolina during summer 79 , suggesting that not all adult females migrate annually to Delaware Bay, perhaps linked to their reproductive cycle. Pregnancy was never determined via ultrasound in female C. taurus individuals while residing in Delaware Bay, but there is evidence of alternate, biennial, or triennial reproductive cycles with a "resting stage" and gestational stage leading to individuals migrating north or remaining south of Cape Hatteras in summers, respectively 54,[80][81][82] . It is possible that the isotopic distinctiveness observed in adult females (Fig. 4) is a result of different foraging habitats due to migration further offshore in autumn after leaving Delaware Bay 63 or proximity to Cape Hatteras during their reproductive cycle 79,80 . Movement and diet inferences. Baseline δ 13 C and δ 15 N values in coastal food webs are typically higher compared to those offshore 75,83,84 . For example, δ 13 C and δ 15 N values of particulate matter decrease along a transect from the US eastern continental shelf in the Western Atlantic to pelagic waters 85 . For C. taurus δ 13 C coll values, this trend seems inconsistent with the proposed foraging site of adult females further offshore east of New Jersey when compared to adult males and juveniles with lower δ 13 C coll values supposedly remaining in neritic to coastal waters. However, Northwest Atlantic δ 13 C spatial isoscapes derived from sampling of shelf animals such as winter skates (Leucoraja ocellata) and loggerhead sea turtles (Caretta caretta) document low δ 13 C variability along the shelf from Cape Hatters and Cape Cod 86,87 . Whereas little skate (Leucoraja erinacea) muscle tissue δ 13 C values document higher values to the northeast of Delaware Bay towards Cape Cod and lower values to the south towards Cape Hatters 87 . This δ 13 C spatial pattern in L. erinacea is more consistent with the patterns observed in C. taurus if adult females were foraging NE of Delaware Bay and adult males and juveniles in neritic waters enroute to Cape Hatteras. The temporal lag between tooth formation and sampling provides insights into the animal's past diet, but also complicates ecological inferences where habitat occupation is unknown. For example, if adult females foraged in waters offshore of North Carolina 79,80 , regional isotopic baselines may vary compared to the foraging sites of juveniles and adult males. We are unaware of any documented significant δ 13 C variability between the area south of Cape Hatteras and the northern neritic waters up to Delaware Bay 52,86 except for some lower values for particulate matter in neritic waters of the Eastern Shore of Virginia 85 . Due to the novelty of δ 66 Zn as a marine trophic level proxy, baseline variability remains poorly understood, but previous studies cautiously suggest a potentially more homogenous marine baseline for δ 66 Zn than for δ 13 C and δ 15 N, although this remains to be tested 32,33 . Differences in δ 13 C coll , δ 15 N coll and δ 66 Zn en values among sex and ontogenetic stage may also relate to a greater reliance on benthic/demersal prey by large females in lower latitude water. There is also the possibility that these four larger females consumed other elasmobranchs, which are too large to be consumed by smaller C. taurus 58,73,74 . Zinc isotopes as an intrapopulation diet proxy. The multiisotope proxy approach combining δ 13 C coll , δ 15 N coll and δ 66 Zn en values documents the distinct ontogenetic isotopic niches and sexual segregation within the Northwest Atlantic C. taurus population more confidently than in the case of a traditional isotope analysis alone. Despite the novelty of zinc isotope compositions as a dietary proxy, and the therewith associated interpretational uncertainties, C. taurus δ 66 Zn en values document ontogenetic and sexual differences in foraging ecology. Therefore, the inclusion of δ 66 Zn en values provides another independent isotope proxy to decipher intraspecific trophic level and niche variability. This is particularly important as bulk δ 13 C coll and δ 15 N coll are likely influenced to some degree also by baseline variation in distinct intraspecific habitats within the population at the time of tooth formation. Notably, δ 66 Zn en is environmentally and metabolically independent from the traditional light isotopes (δ 13 C coll and δ 15 N coll ), yet δ 66 Zn en values show the same distinctiveness in terms of the isotopic composition of adult females as observed for the traditional isotopes. Zinc isotope analysis of shark tooth enameloid thus allows to reconstruct not only interspecific dietary differences but also intraspecific ecological variability using both modern and fossil shark teeth 33 . Indeed, pristine dietary zinc isotope values can be preserved in fossil shark enameloid over millions of years 33 . As such δ 66 Zn analysis can enable cross-disciplinary research with an application for modern ecological research questions, including conservation science, and palaeontology, including palaeoecology, as well as the relatively recently named field of conservation palaeobiology, aiming to increase and utilise our understanding of past ecosystems to respond to modern and future conservation challenges 88 . Uncertainties remain regarding the movement and habitat use (and therefore isotope baseline) and prey consumption of the imperilled North Atlantic C. taurus population. Combining δ 66 Zn en values with compound-specific nitrogen isotope analysis of amino acids from dentine collagen (and other tissues) and tagging data could further eliminate uncertainties in baseline variation and allow more detailed reconstructions of habitat resource usage of mobile predators in future studies. Such reconstructions may be particularly valuable when targeting shark teeth, in a multi-proxy approach, across multiple rows with different tooth formation time, potentially allowing a detailed reconstruction of an individual's migration route and resource uptake along the way 14 . In any case, zinc isotope analysis is a promising tool, especially combined with other methods to decipher the resource use of extant and extinct mobile marine animals, and as such, has great potential to inform preservation efforts for endangered species, including C. taurus. Finally, δ 66 Zn analysis has the potential to enable reconstructing populationspecific diet variability in extinct species too, especially related to shark ontogeny, as shark tooth size scales allometrically with body size 89 . . After sharks were caught, biological data (i.e. sex, fork length, total length and clasper measurements) were recorded, dart tags inserted, and 1-2 teeth in the functional position were extracted (i.e. representing the oldest teeth in the jaw) before sharks were released again. Teeth that were loose were prioritised for sampling. We analysed teeth from 22 females (18 juvenile and 4 adult) and 32 males (19 juvenile and 13 adult) ranging in total length from 143 to 272 cm. Methods Dental collagen carbon and nitrogen isotope analysis. To analyse the stable isotopic compositions of organic carbon (δ 13 C) and nitrogen (δ 15 N), dentine collagen was isolated from 53 shark teeth. Powdered dentine samples were collected using a Dremel on low speed with a 300-micron diamond-tipped bit. Samples were demineralised using chilled 0.1 M HCl, then rinsed five times with deionised water, before being freeze-dried overnight 90 . Samples were weighed to 0.4-0.5 mg in 3 × 5 mm tin capsules. All collagen samples were measured for δ 13 C and δ 15 N values using a Costech 4010 Elemental Analyser coupled to a Delta V Plus continuous flow isotope ratio mass spectrometer with a Conflo IV in the Stable Isotope Ecosystem Lab (SIELO) of the University of California Merced. All data were corrected for linearity and drift using a suite of calibrated reference materials: USGS 40 (n = 19), USGS 41a (n = 10) and costech acetanilide (n = 10). Long term standard deviation for the instrument is 0.1‰ for δ 13 C and 0.2‰ for δ 15 N values. Enameloid zinc isotope analysis. Powdering of samples, zinc purification, and zinc isotope measurements were performed at the Max Planck Institute for Evolutionary Anthropology (Leipzig, Germany) following the protocol as described in ref. 33 . Zinc isotopes were measured from tooth enameloid of 54 sand tiger individuals (i.e. including one individual for which δ 13 C and δ 15 N values could not be analysed). All teeth were cleaned by ultrasonication in ultrapure water (Milli-Q water) for 5 min and dried in a drying chamber at 50°C overnight. Enameloid samples were abraded from the top surface using a dental drill. The enameloid powder were dissolved in closed perfluoroalkoxy vials with 1 ml 1 M HCl on a hotplate for 1 h at 120°C and then evaporated. The residue was then dissolved in 1 ml 1.5 M HBr and placed in an ultrasonic bath for 30 min. Zinc purification was performed in two steps, following the modified ion exchange method adapted from ref. 91 , described in ref. 35 , and always included a chemistry blank and reference standard (NIST SRM 1400) to monitor contamination and Zn elution. One ml of AG-1×8 resin (100-200 mesh) was placed in 10 ml hydrophobic interaction columns (Macro-Prep® Methyl HIC). The resin was then cleaned twice with 5 ml 3 % HNO 3 followed by 5 ml ultrapure water and subsequently conditioned with 3 ml 1.5 M HBr. Following sample loading, 2 ml HBr were added for matrix residue elution, after which Zn was eluted with 5 ml HNO 3 . After performing the column chromatography twice, the solution was evaporated for 13 h at 100°C and the residue re-dissolved in 1 ml 3% HNO 3 for analysis. Zinc isotope measurements were performed using a Thermo Fisher Neptune MC-ICP-MS. Instrumental mass fractionation was corrected by Cu doping following the protocol of Maréchal et al. 92 and Toutain et al. 93 . The in-house reference material Zn Alfa Aesar-MPI was used for standard bracketing. All δ 66 Zn values are expressed relative to the JMC-Lyon standard material (mass-dependent Alfa Aesar-MPI offset of +0.27‰ for δ 66 Zn 35,94 ). Zinc concentrations in the respective samples were estimated following a protocol adapted from one used for Sr by ref. 95 , applying a regression equation based on the Zn signal intensity (V) of three solutions with known Zn concentrations (150, 300 and 600 ppb). The δ 66 Zn measurement uncertainties were estimated from standard replicate analyses and ranged between ±0.02‰ and ±0.04‰ (2 SD). Samples were typically measured at least twice with mean analytical repeatability of 0.01‰ (1 SD, n = 53). Reference material NIST SRM 1400 was prepared and analysed alongside the samples and had δ 66 Zn values (0.92 ± 0.02‰ 1 SD, n = 6) as reported elsewhere 32,96 . Reference materials and samples show a typical Zn mass-dependent isotopic fractionation, i.e. the absence of isobaric interferences, as the δ 66 Zn vs. δ 67 Zn and δ 66 Zn vs. δ 68 Zn values fall onto lines with slopes close to the theoretic mass fractionation values of 1.5 and 2, respectively (Supplementary Data 1). Statistics and reproducibility. Sharks were sorted into four groups separated by sex (female/male) and total length (juvenile/adult). Females and males are considered adults when the total length reaches 220 and 190 cm, respectively 67,68 . The isotopic tracers δ 13 C coll , δ 15 N coll and δ 66 Zn en were used to quantify the isotopic niche size and overlap among the groups within the Delaware Bay sand tiger shark population using the R package nicheROVER 97 . Using nicheROVER we estimated the isotopic niche per group in a three-dimensional (δ 13 C coll , δ 15 N coll and δ 66 Zn en ) space. NicheROVER uses a probabilistic method to calculate our three-dimensional isotopic niche regions (N R ) using a Bayesian framework. Niche overlap was calculated from 1000 Monte Carlo draws with an alpha = 0.95 probability level and is defined as the probability of finding an individual from one group in the N R of another group 97 . Reporting summary. Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Data availability All data generated during this study are included in this published article (Supplementary Data 1).	2023-07-13T06:17:31.293Z	2023-07-11T00:00:00.000	{ "year": 2023, "sha1": "83c759046b5c95d31c4cd1fd555fe0c67dbf84a1", "oa_license": "CCBY", "oa_url": "https://www.nature.com/articles/s42003-023-05085-6.pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "c7fff28c45906ff8d1746f7b959bffa491899995", "s2fieldsofstudy": [ "Environmental Science", "Geography" ], "extfieldsofstudy": [ "Medicine" ] }
81617902	pes2o/s2orc	v3-fos-license	Refractory HHV-6 encephalitis in a patient treated with anti-CD20 monoclonal antibody therapy Human herpes virus 6 (HHV-6) is an infectious cause of encephalitis, rarely identified following allogeneic haematopoietic stem cell transplantation (alloHSCT). The typical symptoms are memory loss, deteriorating conscious state and seizures, which reflect limbic pathology and may be demonstrated by MRI. To date only 9 cases reported in non-alloHSCT patients [1-3]. This report describes the clinical, radiological and pathological features of a patient with HHV6 encephalitis who was treated with chlorambucil and obinutuzumab (a CD-20 monoclonal antibody) for chronic lymphocytic leukaemia (CLL), where HHV-6 was detected in cerebrospinal fluid (CSF) and brain tissue via nucleic acid detection. The case highlights the importance of considering HHV-6 as a cause of encephalitis in patients undergoing low or intermediate immunosuppression, of particular relevance given the introduction of B cell therapies for multiple sclerosis [4]. Introduction Human herpes virus 6 (HHV-6) is an infectious cause of encephalitis, rarely identified following allogeneic haematopoietic stem cell transplantation (alloHSCT). The typical symptoms are memory loss, deteriorating conscious state and seizures, which reflect limbic pathology and may be demonstrated by MRI. To date only 9 cases reported in non-alloHSCT patients [1][2][3]. This report describes the clinical, radiological and pathological features of a patient with HHV-6 encephalitis who was treated with chlorambucil and obinutuzumab (a CD-20 monoclonal antibody) for chronic lymphocytic leukaemia (CLL), where HHV-6 was detected in cerebrospinal fluid (CSF) and brain tissue via nucleic acid detection. The case highlights the importance of considering HHV-6 as a cause of encephalitis in patients undergoing low or intermediate immunosuppression, of particular relevance given the introduction of B cell therapies for multiple sclerosis [4]. Case Report The patient is a 72 year old farmer who was receiving chemotherapy with chlorambucil for CLL. Three months prior to presentation he had adjuvant treatment with 3 cycles of Obinutuzumab; a monoclonal CD20 antibody dosed monthly, as part of a clinical trial. Obinutuzumab was ceased due to systemic side effects including fever, sweats, weight loss and malaise, and single agent therapy with Chlorambucil was continued. Despite cessation of Obinutuzumab, the symptoms continued, and were accompanied by apathy, generalised weakness and muscular atrophy, deteriorating cognition and occasional visual hallucinations. There was no headache. During an admission to hospital for investigation of the above symptoms, two right upper limb focal motor seizures occurred with subsequent, secondary generalisation and status epilepsy requiring intensive care. Magnetic resonance imaging (MRI) showed areas of T2 and FLAIR high signal within the superior left frontal lobe and right parafalcine region, without gadolinium enhancement or restricted diffusion ( Figure 1). An EEG demonstrated poorly developed background rhythms occasionally reaching 8Hz, symmetric throughout the recording and no focal seizure activity after treatment. A lumbar puncture showed a raised CSF protein of 2.46 g/L (0.15 -0.45 g/L), normal glucose and a CSF pleocytosis of 165 x10^5/L mononuclear cells. Overall the clinical picture was consistent with encephalitis. Lymphoproliferative, infective and paraneoplastic/autoimmune causes were considered. Initial CSF culture and viral PCR for HSV, VZV and enterovirus were negative. John Cunningham Virus (JCV) PCR was repetitively negative on serial CSFs. Flow cytometry and cytology excluded infiltration with CLL but showed an extraordinary predominance of CD4 T-lymphocytes. Screening for paraneoplastic antibodies (anti-neuronal antibodies including VGKC, NMDA-R) in CSF and serum was negative. The patient deteriorated over the next two weeks with rapidly progressive cognitive impairment to the point of becoming aphasic, with recurrent visual hallucinations and intermittent focal motor seizures of both upper limbs independently. A brain biopsy of the left frontal lesion was undertaken ( Figure 2). The specimen included grey and white matter. Findings were unusual, with a predominance of small, reactive and perivascular CD4+ lymphocytes in the absence of a specific pathology. Demyelination and necrosis were absent and axons and neurons were preserved. There was neither neuronophagia, viral inclusions nor bizarre astrocytes. CD8 T-lymphocytes were occasionally found in the parenchyma and there were rare B-lymphocytes. Biopsy effectively excluded lymphoma, with viral infection or paraneoplastic disease the likely differential aetiology and a trial of intravenous steroids was administered without effect. CSF was checked for HHV6 and HHV8 using PCR on two occasions. Both CSF specimens were positive for HHV6. Brain tissue was also positive for HHV6 by PCR. Treatment with foscarnet was commenced, and subsequently changed to ganciclovir due to acute kidney injury. Antiviral agents were continued for a period of 21 days and anticonvulsants tapered to maintenance doses. Serial MRI scans showed resolution of the T2-FLAIR high signal change involving the posterior frontal lobes bilaterally. A left caudate nucleus lesion had developed in the interim. The patient improved cognitively and gradually regained sufficient strength to walk and was discharged 6 months after the original presentation. Serial CSF examinations demonstrated that anti-viral treatment suppressed HHV-6 to undetectable levels, but withdrawal of treatment resulted in recurrent encephalitis with clinical deterioration and the re-appearance of HHV-6 DNA in CSF. As such, he remains on maintenance treatment with oral valganciclovir. Discussion HHV-6 reactivation is common among patients post alloHSCT and is known to cause limbic encephalitis, but rare in patients who are immunocompetent or have received alternate treatments for between higher HHV-6 serum viral load and development of encephalitis. While there is a paucity of evidence for HHV-6 infection of neuronal cells, the virus is known to display tropism for astrocytes in the hippocampus [5]. It is hypothesised that the ability of HHV-6 to evade host immune responses in the context of moderate intensity immunosuppression (including monoclonal antibodies) and associated viral down-regulation of cytokines including IL-2 may be adequate to allow viral reactivation resulting in infection of microglia, astrocytes and oligodendrocytes. It is therefore proposed that the CNS may act as a reservoir for HHV-6 infection, even in the absence of viral replication in immunocompromised individuals. This would explain the recurrence of CSF PCR positivity seen in our patient repeatedly after weaning anti-viral medication. Outcomes in AHSCT cases are invariably poor, with a reported mortality rate of roughly 25% of patients. Of the 'immunocompetent' haematological malignancies. To date only 9 cases have been reported in the English literature of patients who have not had preceding AHSCT or solid organ transplant. Development of HHV6 encephalitis in the context of anti-CD20 mAb use is particularly relevant given the recent launch of Ocrelizumab, a humanised monoclonal antibody against CD20, for the use in multiple sclerosis [4]. Awareness of rare, infective causes of new MRI lesions in patients treated with anti-CD20 therapy is vital for neurologists planning to initiate this treatment in an MS patient cohort. The driving factors for HHV6 reactivation remain poorly understood, and there is additional debate as to whether HHV-6 encephalitis results from viral reactivation in immunocompetent patients or primary infection. In alloHSCT reactivation of HHV-6 is observed in 30-70% of patients between weeks 2-4 [5]. Encephalitis develops concomitant to peak serum replication, with a correlation cases described in the literature, all patients survived following antiviral treatment, with the majority experiencing a protracted recovery and mild ongoing neurocognitive deficits [1][2][3]. No controlled trial to evaluate the efficacy of anti-viral agents has been performed, although infectious disease guidelines including Clinical Practice Guideline by the Infectious Disease Society of America and the European Conference of Infections in Leukaemia all recommend foscarnet or ganciclovir as first-line therapies for HHV-6 encephalitis. Duration of therapy is unclear but principles of 21 days therapy as per HSV encephalitis are typically prescribed, or until serial CSF PCR testing is negative. Our patient relapsed multiple times on attempted weaning of anti-viral medications, and continued to respond to re-treatment, creating challenges in anti-infective treatment in a condition where CNS viral clearance appears transient. This case of HHV6 encephalitis following moderate intensity immunosuppression should serve as a reminder to clinicians to consider testing for this virus in other patients presenting with similar clinical features beyond patients post AHSCT, particularly in MS patients where use of anti-CD20 therapy is likely to grow in the near future.	2019-03-18T14:03:24.521Z	2018-01-01T00:00:00.000	{ "year": 2018, "sha1": "5d674a30cd94fc87a745090d59299f27203eb194", "oa_license": "CCBY", "oa_url": "https://www.oatext.com/pdf/CCRR-4-423.pdf", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "23b5a4492f1236126c3ac991a293e61f559f4490", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
211725172	pes2o/s2orc	v3-fos-license	Effect of Stem Cells, Ascorbic Acid and SERCA1a Gene Transfected Stem Cells in Experimentally Induced Type I Diabetic Myopathy Background and Objectives Sarco/endoplasmic reticulum Ca2＋-ATPase (SERCA) inhibition was proved in streptozotocin (STZ)-diabetic rats. The present study aimed at investigating and comparing the therapeutic effect of bone marrow mesenchymal stem cells (BMMSCs), BMMSCs combined with ascorbic acid (AA) and SERCA1a gene transfected BMMSCs in induced type I diabetic myopathy of male albino rat. Methods and Results 54 rats were divided into donor group of 6 rats for isolation, propagation and characterization of BMMSCs and SERCA1a transfected BMMSCs, groups I∼V 48 rats. Group I of 8 control rats, group II (Diabetic) of 10 rats given STZ 50 mg/kg intraperitoneal, group III (BMMSCs) of 10 rats given STZ and BMMSCs intravenous (IV), group IV (BMMSCs and AA) of 10 rats given STZ, BMMSCs IV and AA 500 mg/kg and group V (SERCA 1a transfected BMMSCs) of 10 rats given STZ and SERCA1a transfected BMMSCs IV. The rats were sacrificed after 8 weeks. Gastrocnemius specimens were subjected to biochemical, histological, morphometric and statistical studies. Diabetic rats revealed inflammatory and degenerative muscle changes, a significant increase in blood glucose level, mean DNA fragmentation and mean MDA values and a significant decrease in mean GSH and catalase values, area of pale nuclei, area% of CD105 and CD34 ＋ve cells, SERCA1a protein and gene values. The morphological changes regressed by therapy. In group III significant decrease in DNA fragmentation and MDA, significant increase in GSH and catalase, significant increase in the mean area of pale nuclei, area % of CD105 and CD34 ＋ve cells versus diabetic group. In group IV, same findings as group III versus diabetic and BMMSCs groups. In group V, same findings as group IV versus diabetic and treated groups. Western blot and PCR proved a mean value of SERCA1a protein and gene comparable to the control group. Mean calcium concentration values revealed a significant increase in the diabetic group, in BMMSCs and AA group versus control and SERCA1a group. Conclusions SERCA1a transfected BMMSCs proved a definite therapeutic effect, more remarkable than BMMSCs combined with AA. This effect was evidenced histologically and confirmed by significant changes in the biochemical tests indicating oxidative stress, muscle calcium concentration, morphometric parameters and PCR values of SERCA1a. Introduction Impairment of energy metabolism, muscle weakness, attenuated proliferation and differentiation of skeletal muscle stem cells (SCs) were reported in type 1 diabetes (T1D). Diabetic myopathy, a less studied complication of diabetes, exhibits the clinical observations characterized by a less muscle mass, muscle weakness and a reduced physical functional capacity (1). Altered cellular Ca 2＋ handling is evident and a fundamental disorder in streptozotocin (STZ) induced diabetes. Sarco/endoplasmic reticulum Ca 2＋ -ATPase (SERCA) defective homeostasis was proved in STZ-diabetic rats (2). SERCAs are 110 kDa integral membrane proteins responsible for maintaining cytosolic Ca 2＋ and initiating muscular relaxation through the adenosine triphosphate (ATP)dependent translocation of Ca 2＋ into the sarcoplasmic lumen against a concentration gradient thus regulating muscle contraction (3). SERCA1a was recorded to be concerned with regulation of contraction in the skeletal muscle (4). In diabetes, accumulation of reactive oxygen species (ROS) interferes with normal tissue and organ function. Antioxidants as ascorbic acid (AA) may shift the balance back to antioxidant defense that may protect diabetic tissues from stress by affecting oxidant and antioxidant biomarkers that can be useful in the management of diabetic complications (5). Oyarce et al. (6) assumed induction of SCs differentiation by AA. Bone-marrow-derived mesenchymal stem cells (BMMSCs) are currently considered among the best candidates in the field of regenerative medicine. The recent advances in the genetic modification of these cells have created new prospects for cell-based therapies. The therapeutic potential of engineered stem cells should ideally be validated in animals due to practical and ethical restrictions (7). The present study aimed at investigating the hypothesis of possible therapeutic effect of SERCA1a gene transfected BMMSCs on induced type 1 diabetic myopathy in male albino rat and comparing it to that of non-transfected BMMSCs individually or combined with AA. Drugs Streptozotocin (STZ): was supplied by Sigma Company (St. Louis, Mo, USA) in a powder form as 1 g vial. The required dose was weighed using a digital scale and dissolved in citrate buffer. Ascorbic acid (AA): was supplied by Development of Chemical Industries (Giza, Egypt) in a 1 g Vitacid C effer-vescent tablet form. The required doses were weighed using a digital scale and dissolved in water. Experimental design This study was conducted on 54 adult male albino rats with average body weight 200 grams, housed according to the ethical guidelines of treatment of animals and were divided into the following. Donor group It included 6 rats used for BMMSCs isolation, culture, phenotyping and labeling. In addition, a part of Gastrocnemius muscle was homogenized for SERCA1a preparation. Group I (control group): It included 8 rats that corresponded to and were sacrificed with the experimental groups II, III, IV and V respectively. The first 2 rats received a single intraperitoneal (IP) injection of 0.5 ml citrate buffer and left for 8 weeks without therapy. The second 2 rats received a single IP injection of 0.5 ml citrate buffer. 4 weeks later, it was given 0.5 ml of phosphate buffered saline (PBS) intravenously via the tail vein twice with 24 hours interval and left for another 4 weeks without therapy. The third 2 rats received citrate buffer and PBS as for the second rat together with daily oral AA dissolved in 0.5 ml water using a syringe without a needle for 4 weeks. The fourth 2 rats received citrate buffer and PBS as for the 2 nd and 3 rd rats, in addition to 1.5 μl lipofectamine (used for lipid-mediated transfection) (8) via the tail vein twice with 24 hours interval. Group II (diabetic group): It included 10 rats. Diabetes was induced by a single IP injection of STZ at a dose of 50 mg/kg body weight (9) dissolved in 0.5 ml citrate buffer for each rat. Three days following STZ injection, diabetes was confirmed by measuring the blood glucose level. The animals were considered diabetic if their blood glucose level was higher than 200 mg/dl (10). The rats were left for 8 weeks without therapy. Group III (BMMSCs group): It included 10 rats, diabetes was induced and confirmed as in group II. Four weeks following STZ injection, 0.5 ml (1/2×10 6 cells) of cultured and labeled BMMSCs (11) suspended in PBS were injected in the tail vein twice with 24 hours interval (12). The rats were left for another 4 weeks without therapy. Group IV (BMMSCs and ascorbic acid group): It included 10 rats, diabetes was induced and confirmed as in group II. Four weeks following STZ injection, 0.5 ml of cultured and labeled BMMSCs were suspended and injected as in group III. Combined daily oral administration of 500 mg/kg of AA (13) dissolved in 0.5 ml of water was performed for each rat for another 4 weeks using a syringe without a needle. Group V (SERCA1a transfected BMMSCs): It included 10 rats, diabetes was induced and confirmed as in group II. Four weeks following STZ injection, 0.5 ml of cultured, labeled and SERCA1a transfected BMMSCs were isolated, suspended and injected as in group III. Lipofectamine 3000 transfection (8) of cultured and labeled BMMSCs by SERCA1a (4) was performed. The rats were left for another 4 weeks without therapy. In vitro studies Isolation and propagation of BMMSCs from rats (14): Under sterile conditions, the femur and tibia of rats were excised, and all connective tissue attached to bones was removed with special attention. Bone marrow was harvested by flushing the tibiae and femurs with Dulbecco's modified Eagle's medium (DMEM), (GIBCO/BRL) supplemented with 10% fetal bovine serum (FBS) (GIBCO/ BRL). Nucleated cells were isolated with a density gradient [Ficoll/Paque (Pharmacia)] and resuspended in complete culture medium supplemented with 1% penicillin-streptomycin (GIBCO/BRL). Cells were incubated at 37 o C in 5% humidified CO2 for 12∼14 days as primary culture or upon formation of large colonies. When large colonies developed (80∼90% confluence), cultures were washed twice with PBS and cells were trypsinized with 0.25% trypsin in 1 millimolar EDTA (GIBCO/BRL) for 5 minutes at 37 o C. After centrifugation (at 2400 rpm for 20 minutes), cells were resuspended with serum-supplemented medium and incubated in 50 cm 2 culture flask Falcon). The resulting cultures were referred to as firstpassage cultures. On day 14, the adherent colonies of cells were trypsinized, and counted. Characterization of BMMSCs: Adherent cells (at the end of 3rd passage) were trypsinized and adjusted to 1×10 6 cells/ml determined by hemocytometer and incubated with 10 μl of monoclonal antibodies: CD34 and CD105 (Beckman coulter, USA) at 4 o C in the dark. After 20 minutes incubation, 2 ml of PBS containing 2% fraction crystallizable solution (FCS) were added to each tube of monoclonal treated cells. The mixtures were then centrifuged for 5 minutes at 2500 rpm followed by discarding the supernatant and resuspending cells in 500 μl PBS containing 2% FCS. Cell analysis was performed using CYTOMICS FC 500 Flow Cytometer (Beckman coulter, FL, USA) and analyzed using CXP Software version 2.2 (12). Immunohistochemical characterization was performed by using streptavidin immunoperoxidase techni-que for CD105 and CD34 (15). Labeling of BMMSCs with PKH26 dye (16): Nontransfected BMMSCs cells were harvested during the 4th passage and were labeled with PKH26 fluorescent linker dye to be examined by fluorescent microscope. Lipofectamine3000 transfection (8) and characterization of transfected cells (18): The cells were plated at a confluency of 50% (500,000 cells/plate) one day prior to transfection. 1.5 μl Lipofectamine 3000 reagent (Invitrogen) was used for transfection of 1.5 μg SERCA1a gene labeled with green fluorescent protein (GFP) using serum-free DMEM. GFP directly fuses to selected locations on SERCA1a. Culture media were changed 4 hours after transfection and replaced with DMEM/10% FBS without antibiotics. To assess transfection efficiency, cells labeled with GFP are first visualized via fluorescence microscopy for qualitative assessment of protein expression, morphology and viability. Cells were then prepared for flow cytometry by aspirating the medium and replacing it with 250 μl of a 7：3 mixture of TrypLE reagent: 1× DPBS. Cells were incubated at 37°C for 10 min and then pipetted up and down to ensure single cells for flow cytometry analysis. The cells were then allowed to proliferate and differentiate in DMEM/2% FBS. Labeling of transfected cells with feridex (19): Transfected BMMSCs cells were harvested during the 4th passage and were labeled with Feridex (25 microgram Fe/ ml, Feridex, Berlex Laboratories) in culture medium for 24 hours with 375 nanogram/ml polylysine added 1 hour before cell incubation (evidence of long term overexpression). Animal studies Serological study: Before sacrifice tail vein blood samples were collected at the end of the 8 th week for blood glucose level estimation. The animals were sacrificed by cervical dislocation (20). Gastrocnemius specimens were obtained from the hindlimbs and subjected to: Biochemical study: Half the specimens were kept in the deep freezer for homogenization. 1 gm of muscle was homogenized in 10 ml normal saline by using homogenizer (Ortoalresa, Spain). The homogenates were centrifuged at 1000 ×g for 15 minutes. The supernatant was collected in epindorff tubes that were kept in the deep freezer (at −20°C). DNA fragmentation (21), reduced glutathione (GSH), catalase and malondialdehyde (MDA) (22) were measured in tissue homogenates by using biodiagnostic colorimetric assay kits (Biodiagnostic, Cairo, Egypt). The calcium ion (Ca 2＋ ) concentration was assessed after treatment by colorimetric methods according to the kit instructions. The analytical kits were supplied by Quimica Clinica Aplicada SA (Amposta, Spain) (23). Histological study The 2 nd half of the specimens were fixed in 10% formol saline for 24 hours. Paraffin blocks were prepared and 5 μm thick sections were stained with Hematoxylin and eosin (H&E) (12) and Prussian blue (Pb) stain to demonstrate the injected transfected cells labeled with iron oxide (10). CD105 (559286 Ab, BD Biosciences, San Jose, California, USA) immunostaining, was used for detecting endogenous and exogenous undifferentiated MSCs (24) and CD34 (SAB4300690 Ab, Sigma-Aldrich Chemie Corporation laboratories, Taufkirchen, Germany) immunostaining, for detecting progenitor cells (25) 0.1 ml diluted goat polyclonal 1ry antibodies were applied to sections for 60 minutes. Tonsil sections were considered ＋ve control and the reaction is membranous. Morphometric study Using Leica Qwin 500 LTD (Cambridge, UK) computer assisted image analysis system, assessment of the area (μ2) of pale nuclei of muscle fibers was performed in H&E stained sections using interactive measurements menu. The area percent (%) of CD105 ＋ve and CD34 ＋ve cells was measured using binary mode. The measurements were done in 10 high power fields. Western blot analysis (26): Gastrocnemius tissue was homogenized with a Polytron Homogenizer. Protein content in the samples was measured and adjusted for equal loading. Next, 20 g protein was resolved on a 4∼20% Tris-glycine polyacrylamide gel (Novex, San Diego, CA). After the gel was electrotransferred to a nitrocellulose membrane, the blot was cut in two pieces at a 70-kd marker band. The upper part of the blot was incubated with a polyclonal antibody directed against SERCA1a and the lower part with a mouse monoclonal actin antibody (Affinity BioReagents, Golden, CO). The two parts of the blot were exposed to secondary antibodies carrying horseradish peroxidase (Amersham, Arlington Heights, IL) to drive a chemiluminescence reaction (Amersham, Little Chalfont, U.K.), which was captured on film. Signals on the films were digitized on a 600-dots per inch scanner. Band densities were quantified with the Scion Image Software for PC (Scion, Frederick, MD). The SERCA1a signals were normalized to actin. SERCA1a protein levels were compared in control, diabetic and treated groups. Quantitative polymerase chain reaction (qPCR): Evaluation of formalin-fixed paraffin-embedded (FFPE) skeletal muscle specimens by qPCR was performed (27). Reverse Transcription is carried out with the SuperScript First-Strand Synthesis System for reverse transcriptase (RT)-PCR. The following procedure is based on Invitrogen's protocol. The following ribonucleic acid (RNA)/primer (5'-CCT CCA CTT CCT CAT CCT CTA-3') (3'-A GCC TTG ATG GAC CTC CCT CCA CCA GTG GTG GTG GTG GTG GTG ATT CTC GAG TGA C-5') (mixture was prepared in each tube: 5 μg total RNA and 3 μl random hexamers). The samples were incubated at 65 o C for 5 min and then on ice for at least 1 min. Reaction master mixture was prepared for each reaction then added to the RNA/primer mixture, mix briefly, and then placed at room temperature for 2 minutes. 1 μl (50 units) of Super-Script II RT was added to each tube, mixed and incubated at 25 o C for 10 min. The tubes were incubated at 42 o C for 50 min, heat inactivated at 70 o C for 15 min, and then chilled on ice. 1 μl RNAase H was added and incubated at 37 o C for 20 min. The 1st strand complementary deoxyribonucleic acid (cDNA) was stored at −20 o C until use for real-time PCR. The primer concentrations were normalized, gene-specific and reverse primer pair were mixed. Each primer (forward or reverse) concentration in the mixture is 5 pmol/μl. The PCR program was set up on ABI Prism standard deviation score (SDS) 7000. A copy of the setup file was saved and all PCR cycles were deleted (used for later dissociation curve analysis Statistical analysis Any significant ANOVA was followed by Bonferroni post-hoc test to detect which pairs of groups caused the Results General observations included lethargy and polyurea noticed in diabetic rats that improved in treated groups. Death of 3 diabetic rats was recorded which were compensated. In vitro data Characterization of BMMSCs and transfected BMMSCs: BMMSCs appeared mostly spindle in shape, immunostaining showed ＋ve membranous brownish reaction for CD105 (Fig. 1a) and −ve for CD34 (Fig. 1b). Immunophenotyping of BMMSCs by flow cytometry showed high number of cells ＋ve for CD105 (Fig. 1c). Transfected cells appeared mostly as spindle fluroscent labeled cells (Fig. 1d). Immunophenotyping of transfected BMMSCs by flow cytometry showed high number of GFP labeled cells (Fig. 1e). Histological phenotyping: Sections in the skeletal muscle of all control rats showed fibers exhibiting peripheral nuclei and acidophilic sarcoplasm (Fig. 2a), regular trans- verse striations, pale nuclei and flat satellite cells close to the sarcolemma (Fig. 2b). Sections of diabetic rats revealed multiple atypical widely separated fibers, few disrupted fibers, dark nuclei and few central nuclei (Fig. 2c). Multiple disrupted widely separated atypical fibers with no obvious striations, few central pale nuclei and multiple macrophages in between were evident in other fields (Fig. 2d). Sections of rats injected by BMMSCs recruited flat cells among organized fibers, some separated fibers and congested vessels (Fig. 3a). Few pale peripheral nuclei, few pale central nuclei and striations in some parts of the sarcoplasm were evident (Fig. 3b). Sections of rats injected by BMMSCs and received oral AA showed some fibers with rows of central oval nuclei, multiple flat cells and some mononuclear cells between the fibers (Fig. 3c). Fibers appeared with rows of central pale nuclei, minimal separation of myofibrils, striations in multiple areas of the sarcoplasm by close observation (Fig. 3d). Sections in group V showed regular fibers with rows of central nuclei, other fibers with peripheral nuclei and congested vessels (Fig. 3e). Obvious striations, central pale nuclei, few flat dark cells at the center of some fibers and focal separation of myofibrils were noticed by close observation (Fig. 3f). Animal data In fluorescent labeled sections belonging to group III multiple PKH26 fluorescent labeled cells were evident among the fibers (Fig. 4a), while in group IV labeled cells were fewer (Fig. 4b). In group V feridex labeled transfected cells appeared as few Pb ＋ve spindle cells among the fibers (Fig. 4c) and few GFP fluorescent labelled cells were evident among the muscle fibers (Fig. 4d), and. The count was 8/field in group III, 5/field in group IV and 4/field in group V. In CD105 immunostained sections, −ve immunostaining was found among the fibers and in vessels in group I (Fig. 5a). In group II some ＋ve spindle cells were seen in the CT near vessels and near obviously vacuolated fibers in few fields (Fig. 5b). Group III demonstrated multiple ＋ve cells among the fibers and few in the CT in multiple fields (Fig. 5c). Group IV showed multiple ＋ve cells in the CT and few on the fibers in some fields (Fig. 5d). Group V showed some ＋ve cells in vessels and few on fibers in some fields (Fig. 5e). In CD34 immunostained sections, −ve immunostaining was seen among the fibers and in vessels. Group II demonstrated few ＋ve flat cells in the CT and in the lining of few vessels (Fig. 6a). Group III showed some ＋ve cells at the periphery and at the center of fibers (Fig. 6b). Group IV recruited multiple ＋ve cells at the periphery and at the center of some fibers, in the CT and in the lining of vessels (Fig. 6c). Group V demonstrated numerous ＋ve cells at the periphery of some fibers (Fig. 6d). Morphometric results (Table 1b): As regards the mean area of pale nuclei a significant (p≤0.05) decrease was detected in group II compared to control and treated groups and in group III versus control and the other treated groups. Concerning mean area% of CD105 ＋ve cells a significant (p≤0.05) increase was found in group III compared to diabetic and other treated groups, in group IV versus diabetic and SERCA1a transfected BMMSCs groups and in the latter group compared to diabetic group. While mean area% of CD34 ＋ve cells was significantly (p≤0.05) increased in SERCA1a transfected BMMSCs group compared to diabetic and other treated groups, in BMMSCs and AA group versus BMMSCs and diabetic groups and in BMMSCs group versus diabetic group. Western blot values ( Fig. 7a and 7b): The mean SERCA1a protein values±SD were (1.91±0.32) in group I, (0.31±0.05) in group II, (0.55±0.03) in group III, (0.72±0.09) in group IV and (1.51±0.24) in group V. The previous values indicated a significant (p≤0.05) increase in groups I and V compared to the other groups and in groups III and IV compared to group II. Mean qPCR values (Fig. 7c): The mean SERCA1a gene values in μg±SD were (1.53±0.30) in group I, (0.17± 0.02) in group II, (0.39±0.05) in group III, (0.65±0.13) in group IV and (1.15±0.14) in group V. The previous values indicated a significant (p≤0.05) increase in groups I and V compared to the other groups and in groups III and IV compared to group II. Discussion The findings of the present study pointed to improvement of the skeletal muscle morphological changes that were induced by T1D, confirmed by morphometric quantitative data. In addition, amelioration of the biochemical parameters of oxidative stress, skeletal muscle calcium concentration, and the values of PCR concerning SER-CA1a gene content was recorded. These findings denoted therapeutic effects of the transfected SCs and the nontransfected SCs together with AA that are going to be compared and discussed. In the present study single dose of STZ has been administered to rats followed by 8 weeks interval before sacrifice, to evoke a model of diabetic myopathy. The aim was to establish a progressive myopathy, in order to provide the optimal time course and interrelationships between the serological and morphological events underlying diabetic myopathy in rat. A significant increase was reported in the mean blood glucose values in group II compared to the control and treated groups. In agreement, it was reported that blood glucose fluctuations have higher risk in diabetic chronic complications. This was confirmed by Li et al. (28). Mean DNA fragmentation and mean MDA values revealed a significant increase in the diabetic group compared to all other groups. On the other hand, mean GSH and catalase values recorded a significant decrease, indicating oxidative stress. Going with, it was proved that oxidative stress facilitates increased cellular stress levels and initiates DNA fragmentation to induce apoptosis (29). Group II recruited multiple atypical widely separated fibers with dark nuclei, multiple macrophages between and disrupted fibers. The previously mentioned results denoted inflammatory changes progressing to degenerative changes. In accordance, it was stated that inflammation plays a considerable role in the progression of genetic diseases affecting the skeletal muscle and in defective muscle regeneration with atrophy and wasting (30). In group II, few CD105 ＋ve cells were seen near obviously vaculated fibers confirmed by a significant decrease in the mean area% of CD105 ＋ve cells. It was stated that CD105 is the marker of stromal cells and it was added that diabetes negatively impacts on functional properties of tissue resident SCs (31). In group II, few CD34 ＋ve cells were seen between some fibers and in the lining of few vessels, confirmed by a significant decrease in the mean area% of CD34 ＋ve cells. In support, it was recorded that interstitial progenitor cells are CD34 ＋ve (25). It was added that skeletal muscle progenitor cells are multipotent and give rise to myoblasts and myotubes. Unfortunately, advanced glycation end products exert negative regulation on myogenesis by skeletal muscle progenitor cells (2). As regards PCR values, a significant decrease in SERCA1a gene was recorded in group II, which can be related to oxidative stress. In agreement, it was confirmed that SERCA1a expression decreased in a mouse model of diabetes (32). Group III (BMMSCs Group) recorded significant decrease in DNA fragmentation and MDA, significant increase in GSH and catalase versus diabetic group. Concomitantly, phenotyping showed flat cells among organized and some separated fibers, congested vessels, few peripheral pale nuclei, few central pale nuclei and striations in some parts of the sarcoplasm A significant increase in the mean area of pale nuclei versus diabetic group was proved. The above mentioned results denoted regression of induced diabetic inflammatory and degenerative changes by BMMSCs therapy. In addition, it can be proposed that the flat cells found on fibers are stem/progenitor cells. In accordance, the homing of transplanted MSCs was found to be associated with anti-inflammatory effect on cytokines and regenerative potency (33). Multiple fluorescent labeled cells were found among the fibers. The multiplicity of labeled cells in group III can be referred to some residual tissue changes. While, the significant increase in the mean area % of CD105 ＋ve cells recorded due to BMMSCs therapy can be explained by that, culture expanded MSCs enhance the efficacy of endogenous SCs. A significant increase in the mean area% of CD34 was reported in group III compared to the diabetic group. This outcome was concomitant with the record of powerful chemoattraction of stem/progenitor cells during organ damage. In support Souidi et al. (34) reported that the stromal cells act as guardians for endothelial progenitors by reducing their immunogenicity. Diabetic rats injected by BMMSCs and administrated AA (Group IV) recorded significant decrease in DNA fragmentation and MDA, significant increase in GSH and catalase versus diabetic and BMMSCs groups. Histological phenotyping demonstrated fibers with rows of central pale nuclei, minimal separation of myofibrils, striations in multiple areas of the sarcoplasm and multiple flat cells. A significant increase in the mean area of pale nuclei versus diabetic and BMMSCs groups was reported. These results indicated regression of degenerative changes and obvious regeneration versus diabetic and BMMSCs groups, that can be referred to AA antioxidant effect, in agreement with Masola et al. (5). Characterization of the cultured cells used in groups III and IV by immunohistochemistry and immunophenotyping proved that a large number of the cells were CD105 ＋ve. Fluorescent labeled cells were fewer than group III, indicating less resident tissue changes. Few CD105 ＋ve cells were seen on the fibers, suggesting more pronounced transdifferentiation of SCs. While multiple CD34 ＋ve cells were evident at the periphery and at the center of some fibers, confirmed by a significant increase in mean area% of CD34 ＋ve cells versus BMMSCs group. In support, it was documented that supplementation of AA profoundly rescued muscle resident and nonresident MSCs stemness that is usually altered by oxidative stress with consequent better muscle regeneration (35). A significant increase in the PCR value of SERCA1a gene was reported in groups III and IV compared to group II, concomitantly it was stated that antioxidants, increased expression of myogenic specification genes in muscle progenitors (36). In group V (SERCA1a transfected BMMSCs group), a significant decrease in DNA fragmentation and MDA, a significant increase in GSH and catalase were found versus diabetic and the other treated groups. Histological phenotyping revealed regular fibers with rows of central pale nuclei, other fibers with peripheral pale nuclei, obvious striations, focal separation of myofibrils few flat dark cells and congested vessels. These findings indicated minimal residual inflammatory changes compared to the other treated groups. De Jesus et al. (37) related reduced muscle inflammation to increased expression of SERCA and functional improvements. Characterization of the cultured gene transfected BMMSCs by GFP and immunophenotyping proved transfection of a large number of cells. Few GFP labeled, few Pb ＋ve and few CD105 ＋ve undifferentiated cells were detected on the muscle fibers. While, the area % of CD34 ＋ve cells recorded a significant increase compared to the other groups, which may indicate the predominance of progenitors in this group versus CD105 undifferentiated MSCs. These changes can be explained by more pronounced differentiation of MSCs in SERCA group. Going with, it was confirmed that intracellular Ca＋＋ homeostasis is cruicial for cell proliferation, migration, differentiation and consequently for regeneration, repair and function of skeletal muscle (38). In all SCs treated groups a significant decrease was reported in the mean blood glucose values compared to group II, this effect can be related to that MSCs improve diabetic glycemic control. Additionally, in support to SERCA1a transfected BMMSCs therapy, it was proved that SERCA activation in diabetes resulted in lowered blood glucose level (39). Concerning the mean calcium concentration values, a significant increase was found in the diabetic group compared to all other groups, indicating defective Ca homeo-stasis, in BMMSCs group versus control and the other treated groups, in BMMSCs and AA group versus control and SERCA1a group, indicating re-establishment of Ca homeostasis. In addition, PCR proved a mean value of SERCA1a gene in group V comparable to the control group. In agreement, it was documented that excitation-contraction coupling and initiation of contraction of skeletal muscle fiber is fired by an increase in cytosolic Ca 2＋ . Mutations in genes encoding SERCA, lead to clinical disorders that manifest as myotonia, muscle weakness, paralysis or muscle wasting. Consequently, established homeostasis keeps histophysiology (40). It can be concluded that T1D induced skeletal muscle injury in the form of inflammatory changes that progressed into degenerative changes. SERCA1a transfected BMMSCs proved a definite therapeutic effect on diabetic myopathy, more remarkable than BMMSCs combined with AA. This effect was evidenced by regression of induced inflammatory and degenerative changes. The latter changes, in addition to the therapeutic effect were confirmed by significant changes in the blood glucose level, the biochemical tests indicating oxidative stress, muscle calcium concentration, morphometric parameters and PCR values of SERCA1a.	2020-03-03T14:04:35.234Z	2020-02-29T00:00:00.000	{ "year": 2020, "sha1": "68f35d4e3f3fd2a6c21ffd21d14e24c06aaf1688", "oa_license": "CCBYNC", "oa_url": "https://www.ijstemcell.com/journal/download_pdf.php?doi=10.15283/ijsc18066", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "1a7e53910fe9ca8004e331240bc09ba87ad82f5e", "s2fieldsofstudy": [ "Biology", "Medicine" ], "extfieldsofstudy": [ "Chemistry", "Medicine" ] }
245199991	pes2o/s2orc	v3-fos-license	INFLUENCE OF CHEMICAL COMPOSITION AND THE PROCESSING CONDITIONS ON STRUCTURE, THERMAL STABILITY AND MECHANICAL PROPERTIES OF RAPIDLY COOLED Al-TM-RE-BASED ALLOYS The results of t studies are directed to development of new competitive amorphous and nanocrystalline alloys as well as to improvement of technology of their manufacturing. The physical and technological aspects of interrelations between the conditions for production of rapidly quenched alloys, formation of different structural and phase states, and their properties are discussed. The influence of the chemical composition of alloys and the conditions of their quenching on the glassforming ability, phase composition, and the structure of the rapidly cooled specimens is investigated; the regularities of the effect of alloying elements concentration on the structural features of the Al75– 87(Ni,Co,B/Ga)8–20Gd1Y4 alloys obtained by superfast quenching from the liquid state are established. The thermal stability of the rapidly quenched ribbons with an amorphous structure is investigated and the temperature ranges of phase transformations at continuous heating and under isothermal conditions are found. The strength characteristics of the ribbons as a function of the content and nature of alloying elements as well as the melt cooling rate are determined. The methods of obtaining both Al-based bulk nanocrystalline composites with the shapes of rods and plates with thickness of 0.5–3.5 mm and metal matrix hardening coatings are worked out. Introduction Technologies that are based on the quenching from liquid state (QLS) methods allow at extreme non-equilibrium conditions to realize specific (including amorphous) structural phase states in metallic alloys which provide the unique combinations of strength, electrical, magnetic, corrosion, and other characteristics which are unattainable by using of the traditional metallurgical technologies. It is well known in the scientific community that priority in this field which is naturally united the fundamental investigations with the effective material science and technological developments belongs to Ukrainian scientist Prof. I.S. Miroshnichenko which centenary was recently celebrated by the domestic scientific community. In the late 1950s, Prof. Miroshnichenko was one of the first researchers who developed the techniques of the ultrarapid quenching of metallic melts, obtained metallic glasses, and began the systematic studies of rapidly quenched materials [1]. His monograph "Quenching from liquid state" [2] written in the early 1980s remains very popular until now among the scientists working in this field. The studies of Prof. Miroshnichenko focused on the analysis of the formation of supersaturated solid solutions in the alloy systems with eutectic and peritectic reactions [3]. In particular, he performed a detailed analysis of the structural features formed in binary Al-based alloys during melt quenching and showed for the first time that the degree of solid solution saturation depended on the rate of melt cooling. Prof. Miroshnichenko foresaw the effective perspectives of the materials obtained by the consolidation of the rapidly quenched lightweight Al-based powders or flakes. From the analysis of the publications [4][5][6] he concluded that the most important problem in the processing using hot pressing, extrusion techniques involving the heat treatments is the achievement of the highest strengthening of the consolidated products. In particular, for the supersaturated solid solutions formed in powders and flakes during melt quenching the highest strengthening occurs when the heat treatment results in the formation of the intermetallic phases particles with the optimal sizes, shapes, and distribution in the matrix. The Al-based alloys which combine low specific weight and high strength are widely used as structural materials in the industry, especially as lightweight components in aircraft and aerospace constructions [7]. The growing importance of environmental problems and energy saving, as well as technical progress, require new materials with enhanced levels of both mechanical properties and thermal stability. A solution to this problem may be realized by using new strengthening mechanisms because the traditional ones such as solid solution and dispersive hardening, grain refinement, deformation strengthening, and use of fibres result in Al-based alloys in the maximum value of yield stress of about 600 MPa [7]. Subsequent progress in enhancement of strength of Al-based alloys was reached by obtaining materials with extremely disordered (amorphous), quasiand nanocrystalline structures which formed in the course of melt quenching or by controlled crystallization of amorphous phases. Among these structures which are formed presumably in the Al-RE-TM the partially crystalline structures composed of Al nanocrystals embedded into the amorphous matrix have the highest level of strength equals 1560 MPa [8]. Despite the high level of the strength of the Al-RE-TM alloys with amorphous and nanocomposite structures, their applications are limited by the low sizes of the glassy samples (≤ 1 mm in one dimension) [9] and the high cost of the RE metals. To resolve these problems in the present study the attention was focused on the preferable use of the TM. The important characteristic of the alloys with both amorphous and nanocomposite structures are the temperature and time ranges of their exploitation without deterioration of their properties which depends on the alloy chemical composition and the conditions of their fabrication [10,11]. Their determination requires research of the dynamics and kinetics of structural and phase transformations at the heating of the samples with nonequilibrium structures. In these studies, the features of the nanocluster structure of the amorphous phase were accounted which in turn were related to the structure of the melts [12]. Thus, the aim of this work was studies of the influence of the alloying TM such as Ni as well as its partial replacement with Co on the structure of the rapidly quenched Al75-83(Ni,Co,B)12-20Gd1Y4 alloys, the thermal stability of amorphous phases, changes of microhardness caused by nanophase composite formation as well as working out of methods of the casting of bulk rods and plates with nanocomposite structures and obtaining metal-matrix strengthening coatings. Ingots of the alloys studied were prepared from pure components (Al -99.99 wt.%, Ni, Co, B, Ga, Gd and Y with ≥ 99.95 wt.% purity) by electric arc melting in a purified Ar atmosphere. RE were introduced into alloy as the preliminary prepared master alloys with compositions Al3Y and Al3Gd. Rapidly quenched ribbons (30-60 μm thick and 10 mm width) were produced by single roller melt-spinning technique in a helium atmosphere. The chemical composition of the as-prepared ribbons was controlled by Xray fluorescence analysis. The structural characterization of the as-prepared and heat-treated ribbons was carried out by X-ray diffraction (XRD) employing monochromatic Mo K and filtered Co K radiations in a standard DRON-3M diffractometer. The average Al grain sizes in nanophase composites were estimated from the full width at half maximum of the Bragg 31 reflections using the deconvoluted XRD patterns. The average size of nanocrystals L (= λ/Bcos(θ) with λ being the wavelength and θ the scattering angle) and the average nearest-neighbour distance, R = 0.615λ/sin(θ) were calculated from the X-ray data [13]. The thermal stability and crystallization kinetics was studied at constant rate heating (5-40 K/min) conditions by differential scanning calorimetry (DSC 404 F1 NETZSCH). The characteristic temperatures corresponding to the onset crystallization (Tona) and to the maximum rates of crystallization stages (Ti) were determined from the DSC scans with an accuracy of ± 1 K. These data were used for calculations of the activation energies of crystallization stages using the well-known the Kissinger method [13]. The microhardness (Hµ) measurements of the as-cast and heat-treated samples were performed using standard PMT-3 microhardness tester calibrated with a NaCl monocrystal. The measurements were carried out on the flat areas presumably contact surfaces of 15 specimens at a load of 0.49 N (50 gf) during 10 s. The Hµ values for each specimen were determined by averaging of 10 indents with regular shape which gave the standard deviation of the data about ±1.5%. 3. Results and discussion 3.1. Effect of the TM concentration on structure and thermal stability of Al95-x(Ni,Co/Ga)xGd1Y4 (x = 8, 14 (12), 20) rapidly quenched samples As mentioned above, until now the majority of studies were devoted to investigations of relatively low-alloyed Al-TM-RE alloys containing 85-87 at.% Al. These studies (e.g., [11,15,16]) showed that the alloys Al86-87(Ni,Co)8(Gd,Y)5--6 obtained by the melt-spinning processing have amorphous structure according to the XRD data. It should be noted that the profiles of the first diffuse broad halo at the XRD patterns have the 'shoulders' at large scattering angles side. The analysis showed that this shoulder corresponds to the average interatomic distance of Al-TM pair such indicating the presence of the clusters. Another feature of the diffraction patterns glasses with low concentrations of TM was the presence of 'pre-peaks' at relatively low scattering angles (about 10 o in MoKα). The interatomic distance estimated from the Ehrenfest relation corresponding to the pre-peak is about 0.53-0.56 nm. As it has been shown in [17] the pre-peak is caused by formation of the Al-RE clusters in which rare earth atoms are not in direct contact. At the same time, it has been established that amorphous phases may be obtained in the melt-spun samples of Al-based alloys with enhanced content of transition metals up to 15 at.% (Al81Ni15Y4 and Al75Ni15Y10) [18] and to 20 at.% (Al75(Ni,Co,В)20Gd1Y4) [19]. In order to study the effects of the chemical compositions on the structure of the rapidly quenched samples a series of the melt-spun ribbons with thickness 35-45 μm from the number of the Al-based alloys listed in Table 1 was obtained in the present study. As can be seen from Fig.1 XRD patterns of all samples did not contain any sign of crystallinity. In turn, the DSC scans of the samples are typical for metallic glasses crystallizing via several studies (Fig. 2). Analysis of the main broad maximum profiles shows that the increase of TM content in the alloys investigated results in enhancement of intensity of the high angle shoulder (Fig. 1). It lends support the above suggestion that this shoulder caused out Al-TM atomic interactions. In contrast, the relative intensity of the pre-peaks has no systematic dependency on the changes in concentration. In general, the structural studies of the rapidly quenched samples show the clear tendency of the AlNiCoGdY melts to phase separation. In order to estimate thermal stability of amorphous structure in the rapidly quenched samples the DSC thermograms were measured under constant rate heating. The results of these studies at 20 K/min for the Al87Ni8Gd1Y4, Al81Ni14Gd1Y4, Al75Ni20Gd1Y4 and Al75Ni16Co4Gd1Y4 are presented in Fig. 2. From the DSC scans the values of the onset crystallization temperatures (Tons) as well as the temperastures corresponing to the maximum rates of transformations (Tp) at each crystallization stage were determined and listed in Table 1. As it is expected the increase of Ni content from 8 at.% to 14 at.% and to 20 at.% results in enhancement of the onset crystallization temperature of amorphous phase from 463 K to 578 K and to 629 K, respectively. The partial replacement of Ni with Ga and with Co in low-alloyed glasses leads to the decrease of Tons while thermal stability of the highly-alloyed glasses increases (up to 658-669 K) due to additions of 4 at.% Co and B and Ga at the expence of Ni (Table 1). Structural analysis of the melt-spun samples heated up above the first crystallization stage showed that the main reason of the enhanced thermal stability of the medium-and highly-alloyed glasses was change of crystallization mechanism. While thermal stability of the low alloyed glasses is determined by the single-phase primary crystallization (formation of nanocrystals of pure Al + residual amorphous matrix) (Fig. 3a, pattern 1), simultaneous formation of several phases (Al and Al-rich binary and ternary intermetallic compounds) takes place at the first crystallization stage of the glasses with enhanced content of TM, Fig. 3(b-d), Table 2. The diffraction peaks of these phases are broadened indicating small (nanoscale) sizes of crystals. The absence of the amorphous halo on the diffraction patterns after the first crystallization stage of the medium and highly alloyed glasses supposes the eutectic crystallization mechanism. It follows from the X-ray data that second and third peaks at the DSC scan of the Al87Ni8Gd1Y4 glass (Fig. 2) are caused by residual amorphous matrix crystallization (Fig. 3a), while second peaks at DSC scans of medium and highly alloyed glasses related to the crystalline phases transformations (grain growth of the phases formed at the first crystallization stage and increasing of the ternary intermetallic phases content, Table 2). It should be also noted here that first crystallization stage in the medium-alloyed glasses takes place in very low temperature range (scan 2 in Fig. 2) and there is a small endothermic effect before the crystallization peak (not shown). This minimum of heat flow is caused by the increase of the specific heat due the glass transition in amorphous phase which is not observed in both low-and highly alloyed glasses. Both the relatively enhanced the onset crystallization temperatures (583-587 K) and the activation energies of crystallization (424-473 K) of the Al81-83(Ni,Co)14-12Gd1Y4 compared with those of low alloyed glasses (Table 2) make the medium alloyed glasses attractive candidates for engineering applications. The more detailed analysis of the phase compositions of the heat treated melt-spun samples was performed with using of the PowderCell 2.4 and FullProf2009 Software [20]. The results are summarized in Table 2. It is evident that the increase of the TM content results in the decrease of the fraction of Al nanocrystals and to complication of the phase composition. It is interesting to note that partial replacement of Ni with Co results in formation Al9Co2 phase and in suppression of formation of Al9Ni3RE ternary intermetallic compound. It indicates that the phase transformations in the melt-spun samples are governed mainly by interactions of the transition metals with Al. Influence of transition metals content on microhardness of rapidly quenched samples As it is known the strength is the most important property of structural alloys. For its characterization the values of microhardness (Hμ) which are directly proportional to the yield strength [21] were used in the present study. The measurements showed that the increase of the TM content led to the enhancement of the Hμ from 2.6-2.8 GPa in the ascast glasses with 87 at.% to 4.6-6.0 GPa in the glasses with 75 at.% (Table 3). It is interesting to note that partial replacement of Ni by Ga results in softening of the low alloyed glass and in hardening of the highly alloyed one. The opposite effects on hardness are observed due to partial replacement of Ni with Co in medium and highly alloyed glasses (Table 3). Formation of crystalline phases in amorphous glasses results in essential hardening ( Table 3, Fig. 4) by from 50 to 100%. In part, microhardness of the heat treated glasses is increased up to 10 GPa which corresponds to the extremely high value of yield stress about 3 GPa [21]. The main contribution into the increase of the hardness is caused by small sizes of crystals of Al and intermetallic compounds. The coarsening of the grain structure results in the decrease of hardness down to values of the as-cast samples (Fig. 4). Thus, the very high values of microhardness found in the heat-treated samples with fine-grained structure of the medium-and highly alloyed Al-based glasses indicate that 35 these alloys may be used as the basis for the development high-strength light-weight alloys with the properties superior to the commercial alloys. Priority of the authors in the field of development these alloys id protected by the Ukrainian patent [21]. Accounting these results the attempt to obtain bulk samples of highly alloyed alloys by ejection of melt into wedge-shape ((2.5-3.5)5 mm 2 copper mold. It allows obtaining in single sample the set of structures formed at different cooling rates (q) which were calculated from the empirical relationship [22] q = 10 + 989/D where D is thickness (in mm) of the wedge. XRD studies showed that all plates had crystalline structure (Fig. 5) without any observable differences in the parts of samples with different thickness. Analysis of the XRD patterns revealed the multi-phase structure of the plates in which the the ternary Al19Ni5RE2 dominated. This phase with orthorhombic unit cell (a = 0.4078 nm, b = 1.599 nm, c = 2.698 nm) is formed at the final stages of transformation of low alloyed amorphous alloys [11,23]. Table 4. Alloy composition Hμ, GPa Microhardness of the rapidly solidified plates is in the range between 5.0 (Al75Ni16Co3B1Gd1Y4) and 6.0 GPa (Al75Ni18Ga2Gd1Y4) ( Table 4), i.e. lower than that found in the glassy melt-spun ribbons with the same composition ( Table 3). The most probable reason of this difference is relatively coarse grained structure of the plates (Fig. 6) which consists of at least three structural elements with different morphology. In contrast to the samples with amorphous structure heat treatment of the bulk plates does not appreciably change their hardness (Table 4). In view that engineering applications of high-strength structural Al-based alloys require materials in bulk forms while maximum size of the cast billets with amorphous structure achieved until now is about 1 mm, the most perspective way to realize the high level of mechanical properties of disordered and partially ordered (nanocrystalline) structures is consolidation of the rapid quenching products. Recommendations concerning development and practical application of new nanocrystalline high-strength light-weight Al-based alloys In the frames of these investigations the conclusion has been drawn that consolidation of rapidly quenched ribbons and flakes is required to obtain bulk billets with amorphous and nanocomposite structures. Because, the supercooled liquid region (the difference between the onset crystallization and glass transition temperatures) in amorphous Al based alloys is small (≤ 20 K [9]) the thermomechanical processes in the course of consolidation will inevitably lead to crystallization. In fact, the temperatures required for full consolidation by extrusion [24-28], by spark plasma sintering [29,30] or by forging [31] as a rule exceed 673 K, and it results in formation of microstructures which consist of crystals of Al and Al(TM,RE) intermetallic compounds. In the other hand, it is known from the literature [32-34] that maximum of microhardness is observed in the samples with nanocomposite structure (nanocrystals of Al embedded in the residual amorphous matrix), while formation of the intermetallic compounds results in softening. In order to achieve the maximum density at lower temperatures and retaining of nanocomposite structure in Al-based alloys the l methods of severe plastic deformation (SPD) were successfully applied [35,36,37]. The experiments carried out allowed to obtain bulk nanostructured samples of Al86Ni6Co2Gd6 alloy at 523 K using twist extrusion [37], Al89Gd7Ni2Fe1 alloy at 473 K by equal channel angle extrusion [36] and Al90Fe5Nd5 alloy even at room temperature using high pressure torsion [35]. As it was shown above, the nanoscale crystals of intermetallic compounds which formed simultaneously with Al nanocrystals may contribute to hardening of nanophase composites. The comparison of the thermal stability of the investigated alloys at isothermal annealing with temperatures required for full consolidation by methods of equal channel angle extrusion (473 K [36]) and by twist extrusion (523 K [37]) shows that these SPD techniques allow to retain the nanocomposite structure in the majority of new amorphous alloys. One of the new approaches for fabrication of light-weight high-strength ductile bulk alloys is using of amorphous Al-based alloys as reinforcement particles in metal-matrix composites (MMC) [38]. For example, addition of 10 and 20% of amorphous Al85Ni10La5 powder to cryomilled 5083 Al results in enhancement of the ultimate strength of the extruded bars from 681 to 813 and 906 MPa, respectively [38]. The extruded bars had limited fracture deformation (3.3 and 2.4 % in MMC with 10 and 20%, respectively), however, subsequent swaging of 10% MMC led to essential enlargement of plasticity at compression (up to 22.5%) combined with relatively high yield stress (729 MPa) [39]. Another way to increase of plasticity of MMC was proposed in [40]. In this study the 37 ductile nanocrystalline powder of Al94Fe4Mo0,6V1,1Zr0,3 alloy was added to high strength amorphous Al84Ni10Ce6 powder and the fully consolidated bulk samples were obtained by forging at 648-773 K. It was established that the MMC samples consolidated at 648-673 K had the high compression strength (1.2 GPa) and plasticity 1 %, while the bulk sample consolidated at 723 K had very high strength (1.34 GPa) without ductility. We have assumed that the increase of the volume fraction of ductile material will allow to obtain MMC with high strength and acceptable level of plasticity. For this purpose the alloys which retain the high strength in the fully crystalline state such as Al83-75(Ni,Co)12-20Gd1Y4 are available. For fabrication MMC we used the embrittled amorphous ribbons and flakes of medium (Al81Ni14Gd1Y4) and highly (Al75Ni20Gd1Y4) alloyed alloys. Amorphous Al81Ni14Gd1Y4 ribbon was annealed at 663 K in an argon atmosphere, milled and separated into 10-60 µm fractions. The hardness of annealed ribbon was 5.5 GPa. The powder was mixed with 1-20 µm fraction of Al powder in the ratio to 3:7, respectively, and the mixture was used for fabrication of MMC as protective coating on the surface of the hypoeutectic soft AlSi alloy by the cold gas-dynamic spraying (https://reym.prom.ua/p739290046-apparat-gazodinamicheskogo-napyleniya. html). This technology used now for renewal of surface of products, hardening and for corrosion protection allowed forming coatings as Al-based MMC (Figs. 7, 8). The formation of the MMC was performed at relatively low temperatures at which nanoscale microstructure of reinforcing Al81Ni14Gd1Y4 powder was retained. The coatings are dense because they form as a result of precipitation of the mixture of powder moving with supersonic speed (500…1000 m/s). The microhardness of the coating ≤ 40 μm was 2.7 GPa, i.e. by about 2.5 times higher than Hμ of AlSi substrate (1.1 GPa). The reached values of MMC microhardness are not record, partially due to using of fully crystallized Al81Ni14Gd1Y4 powder but not the only embrittled. In general, gas-dynamic impact on nanostructure of reinforcing powder, possible diffusion of its components into matrix AlSi material, structure and phase composition of MMC were not investigated in this study. Nevertheless, the analysis and preliminary experiments carried out for fabrication nanostructured MMC on metallic surfaces indicate that the treatment which unites the nanocrystallization of amorphous phase and consolidation processes is a perspective way for production of bulk light-weight and high strength Al-based materials. Conclusions It is established that the increase of the TM content from 8 to 20 at.% essentially enhances thermal stability of rapidly quenched amorphous Al87-75(Ni,Co)8-20(Gd,Y)5 alloys which is caused out by the change of mechanism of the first crystallization stage from the single phase (formation of Al nanocrystals in the Al87Ni8Gd1Y4 alloy) to the multi phase one (simultaneous formation of nanoscale crystals of Al and intermetallic compounds in the Al75Ni16Co4Gd1Y4 alloy). The specific feature of the DSC scans of the firstly prepared in this study medium alloyed Al81Ni14Gd1Y4 and Al81Ni10Co4Gd1Y4 alloys is the presence of the glass transition and sharp first crystallization peak such indicating the increase of the rate of crystallization in the range (583 -587 K) above Tons. The activation energy of the first crystallization stage in the Al81-83(Ni, Co)14-12Gd1Y4 amorphous alloys are relatively high (424-473 kJ/mol) and even exceed that for the Al75Ni20Gd1Y4 glass. The Al81(Ni, Co)14Gd1Y4 and Al75(Ni, Co)20Gd1Y4 alloys possess essentially enhanced microhardness (by 50% and 100%, respectively) in comparison with Hμ low alloyed Al87Ni8Gd1Y4 glass in the as-prepared and in heat-treated states. In new medium alloyed glasses formation of nanocomposite structure results in relatively small increase in microhardness from 5.1 -5.4 to 7.6 -7.7 GPa. In the highly alloyed glasses the highest microhardness (9,7 GPa) is achieved after the second crystallization stage. Coarsening of crystalline structures at final stages of transformations results in the essential decrease of microhardness. The presented results allow to formulate the following recommendations concerning the development of practical application of new nanocrystalline high-strength Al-based alloys (composites): for obtaining of bulk rod and plates consolidated from amorphous melt-spun ribbons the alloys Al86Ni6Co2Gd3Y2Tb1 and Al86(Ni,Co)8(Gd,Y)6 [41] may be efficiently used due to their high glass forming ability, thermal stability and hardness; for obtaining of metal-matrix composites by the cold gas-dynamic spraying on surfaces structural light alloys the medium-and highly-alloyed Al81-75(Ni,Co)14-20Gd1Y4 alloys may be used due to their high microhardness in the nanocrystalline states and high stability of the properties in wide temperature ranges; the proposed method for fabrication of the Al-based nanostructured metal-matrix composites with high density is energy saving and may be used for development of new metal-polymer composites.	2021-12-16T18:02:51.217Z	2021-09-07T00:00:00.000	{ "year": 2021, "sha1": "6484015f4b09e386a531e65b4db2d8d718b17661", "oa_license": null, "oa_url": "http://jphe.dnu.dp.ua/index.php/jphe/article/download/124/117", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "af4e48bfca2d069d70c0202932f75ca584f26008", "s2fieldsofstudy": [ "Engineering", "Materials Science" ], "extfieldsofstudy": [] }
244912874	pes2o/s2orc	v3-fos-license	Spatio-temporal variability of climatic parameters across different altitudes of North- Western Himalaya Climate change impact varies across different altitudinal ranges and demands local specific management strategies for water resource and farming system management. The present study analyses spacio-temporal climate parameters across different altitudes of Himachal Pradesh a hilly state of India. Analysis shows that annually, minimum temperature has significantly decreased by -0.09°C at altitude I (350 400 m) while maximum temperature has significantly increased by 0.05°C at altitudes I and II (14001500 m) and decreased significantly by -0.08°C at altitude III (20002100 m). Higher regions Altitude – IV (2900-3000 m) received lowest rainfall (746.1 mm) with 30.2 % variation. Seasonal rainfall variability was higher in post monsoon (102 174%) and least in monsoon (21 57%). Annual rainfall at altitude I is strongly irregular (PCI 20.1 to 22.3), followed by altitude – IV (PCI 15-25); altitude – II irregular (PCI 15-20) and altitude – III moderate to irregular (PCI 12 -19) rainfall. Seasonal Index values for four altitudes fall between 0.91-0.96 revealed that rainfall is irregular and markedly seasonal with longer drier season. Higher wavelet powers in altitude I and II after 2005 suggests frequency of extreme rainfall occurrence had increased. Climate change is inevitable; however, it is imperative to known the variability of climate change so that necessary measures can be taken at local levels for adapting to such changes. Indian economy is largely agricultural based. Agriculture is a climate sensitive sector and any change in climate will have direct or indirect impact on agricultural productivity by affecting the crop yield, water availability and associated drought stress, pest-diseases etc affecting economy of the country. Cropping intensity and cropping area are also greatly influenced by the climatic variability (Kotera et al. 2014). Extreme temperature and rainfall variability have significant negative impacts on crop production. It was observed that temperature have more significant positive or negative impact on crop production as compared to rainfall (Lobell et al. 2011). Other impacts of climate change include higher runoff from the fields and increased soil erosion due to more frequent intense rainfall events. Piao et al. (2010) opined that the impact of climate change plays a pivotal role for future agricultural planning and crop production in the world, including India. North Western (NW) Himalayan region of India covers states of Jammu and Kashmir, Himachal Pradesh and Uttrakhand. About 1.6 degree of temperature rise in the NW Himalayan region was reported in the last century by Bhutiyani et al. (2007). Significant increasing trend -1 of 0.06 °C year on annual mean maximum temperatures and -1 0.02 °C year on annual mean temperatures for the 1951-2010 time period were observed for Himachal Pradesh (Rathore et al. 2013). Increasing trend over the central India region for the period 1950-2000 was also observed for number of rainfall events per year with precipitation greater than 100 mm whereas number of moderate rainfall events had shown decreasing trend in the same period (Goswami et al. 2006). Decreased number of rainfall events but increasingly common high rainfall events was reported by Dourte et al. (2012). The impacts of such extreme events in hills will leads to migration of species towards higher altitudes, loss/drying of traditional water sources, changes in planting time and dates, reduced crop yields and increased vulnerability of farming communities to climatic changes. The direction of change and the rate of increase or decrease in the annual and seasonal trend is revealed through the slope (Choudhury et al. 2012). Precipitation concentration index (PCI) Long-term total variability in the amount of rainfall was assessed using Precipitation Concentration Index (Oliver, 1980). It is taken as an indicator of rainfall concentration and rainfall erosivity (Michiels et al. 1992). PCI was calculated on an annual and seasonal scale as follows: th where p is the monthly precipitation in i month . i Where n is the no. of the months in a season. PCI was computed on five yearly basis. PCI values are interpreted differently by different studies (Table 2). Seasonality index Seasonality comparison for different altitudes in a state or over different locations in a larger region is possible only when there is some quantitative tool for accounting for rainfall regimes (Adejuwon, 2012). Rainfall regimes are important to know the degree of variability in monthly rainfall throughout the year (Walsh and Lawer, 1981). Seasonality Index (SI) is a function of mean monthly and annual rainfall and was estimated using the following formula: Where X is the mean rainfall of month n and R is the n mean annual rainfall. The SI values ranges from zero (if all the months have equal rainfall) to 1.83 (if all the rainfall occurs in one month). Different rainfall regimes with respect to SI values (Kanellopoulou, 2002) is given in Table 3. Change point analysis Change-point analysis is a powerful tool for detecting a change in the time series. Abrupt changes occurring in the climatic records are evaluated using non-parametric Pettitt's test, developed by Pettitt (1979). The Pettitt's test affectability to breaks in the middle of any time series make it the most commonly used test for change point detection (Winingaard et al. 2003). The exact time of the change can be easily obtained specific interventions can be advocated for adaptation under changing climatic conditions. Pawar et al. (2019) have analysed trends of rainfall and temperature at Shimla and Dharmshala. In this paper we aim to analyse the temperature and rainfall trend and pattern over different altitude of Himachal Pradesh on monthly, seasonal and annual timescale. Study area Himachal Pradesh is located between 30°22'40" N to 33°12'40" N latitudes and 75°45' 55" E to 79°04' 20" E longitudes. The state has a hilly terrain and complex topography with the altitude ranging from 350 meters to 7000 meters above sea level. The annual rainfall of the state varies from 2909 -3800 mm and temperature ranges from -4°C (minimum) to 42°C (maximum). Based on its altitude, precipitation and varied cropping patterns, the state is divided into four agro climatic zones : Low-hill subtropical zone, midhill subhumid zone, high-hill temperate wet zone and high hill temperate dry zone. One station from each of the four agroclimatic zones was selected for the study. Data source Different stations located at varying altitudes were selected for study (Table 1). Monthly temperature (minimum and maximum) and rainfall data of selected stations of Himachal Pradesh for years 1980-2015 were collected from India Meteorological Department (IMD), Regional Research Station Shimla. Four Principal seasons observed in the state are winter (December, January, February), pre monsoon (March, April, May), monsoon (June, July, August, September) and post monsoon (October, November). Trend analysis Monthly, annual and seasonal trends of temperature and rainfall were computed using non-parametric Mann-Kendall (MK) statistic (Mann, 1945;Kendall, 1975). This MK test is recommended by the World Meteorological Organization and has been widely used to test for trends in hydrological and meteorological data, including precipitation, temperature and runoff (Li et al. 2008;Zang and Liu, 2013). The change in trend per unit time was calculated using non parametric Sen's slope estimator (Sen, 1968). The Sen's slope estimator is computed by choosing the median of the slopes of all lines through pairs of points as follows: (Jana et al. 2017;Dhorde and Zarenistanak, 2013;Kang and Yusof, 2012;Gao et al. 2011). Spatial temperature and rainfall distribution Analysis of temperature revels that annual minimum 0 and maximum temperature varied from 4.5 -15.6 C and 15.9 0 -30.0 C respectively. Minimum temperature was in Altitude-IV with 19.6 % variation and maximum temperature was in Altitude-I with 6.3 % variation. The annual rainfall varied from 746.1 -1398 mm with an average of 1075.8 mm. Higher region (Altitude -IV) received lowest rainfall (746.1 mm) with 30.2 % variation and Altitude -II received highest (1398 mm) rainfall with 17.9 % variation. For minimum temperature, maximum temperature and rainfall higher variation was 36.9 %, 13.7 % and 42.0 % respectively in altitude-III (Table 4). altitudes ranging from -4.0 -4.0 C. Mean maximum 0 temperature was higher in the month of June (22.7 -33.9 C) in 0 altitude II, III and IV whereas, it was higher (38.1 C) in altitude -I in May. Variation for minimum temperature was higher (-125.4 %) for altitude -III in January. Maximum temperature variability was higher (53.0 %) in January for altitude -IV. Mean monthly rainfall was higher in July in altitude -II (352.3 mm) and altitude -III (184.6 mm). Rainfall variability was higher (240.1 %) in November at altitude -I (Table 4). Seasonal temperature analysis implies that average 0 minimum temperature was higher (11.4 -21.7 C) in monsoon 0 and least (-2.6 -3.7 C) in winter season. Average maximum 0 temperature was higher (22.2 -34.8 C) in monsoon season. Variation of minimum temperature was higher (-317.8 %) in winter and for maximum temperature it was also higher (36.6 %) in winter in altitude -III. Seasonal rainfall variability was higher in post monsoon (102.6 -174.6 %) and least in monsoon (21 -57 %) (Table 4). Monthly trends Minimum temperature has significantly decreasing trend during July, August and December by 0.09°C, 0.08°C 0 and 0.12 C respectively for the altitude I. Significant decreasing trend in minimum temperature (0.04°C) was found for December in altitude II. Minimum temperature of October and November had significant increasing trend of 0.06°Cand 0.05°C respectively for Altitude III (Table 5).. Significant increase in maximum temperature by 0.12°C, 0.1°C and 0.04°C was observed during the months of March, May and October, respectively for lowest Altitude I. However, significant increase in maximum temperature by 0.1°C, 0.06°C, 0.06°C, 0.05°C, 0.06°C and 0.1°C was observed during March, April, May, October, November and December, respectively in Altitude -IV. No significant trend in rainfall was observed for Altitude I and II (Table 5). At altitude -III, rainfall showed significant increase by 2.36 mm and 1.64 mm during the months of January and June, respectively. Significant increase in rainfall by 2.22 mm in the month of January was also observed for the altitude -IV. Seasonal and annual trends Magnitude of seasonal temperature and rainfall trend was analyzed and is presented in Table 5. Maximum temperature showed a significant increasing trend in most of the seasons. Minimum temperature had a decreasing trend in most of the season. However, significant decrease was (-0.051 0 C) was observed for altitude -IV during winter season. Monsoon rainfall in altitude-IV had a significant increasing trend. Decreasing trend of winter rainfall in lower altitudes (altitude -I and II) and its increasing trend in higher altitudes (altitude -III and IV) are cause of concern and requires contingency planning. At lower altitude, which is agricultural production areas, rainfall trends show increased concentration in monsoon and its decrease in other all seasons. Such changes on one hand will hamper winter cropping and on the other creates flood like situation during rainy season. Annual trend shows significant increase in maximum temperature in lower altitudes. Seasonal rainfall concentration Pattern of rainfall distribution was assessed using PCI on 5 yearly bases. Annual and seasonal PCI values varied from lower than 10 in the monsoon season to higher than 30 in the post-monsoon season (Fig 2). Annual rainfall at altitude I is strongly irregular (PCI 20.14 to 22.37), followed by altitude -IV with PCI moderate to strongly irregular (PCI 15-25); altitude -II irregular and altitude -III moderate to irregular rainfall. This shows that annual rainfall is unevenly distributed among the months and is concentrated more in one third of the year. Seasonal PCI showed mixed rainfall distribution in pre-monsoon, post-monsoon and winter rainfall with PCI values varying from 9 to 46. This shows highly irregular and unpredictable nature of rainfall during the seasons. This uncertainly in rainfall affects the pre-kharif and rabi crop sowing and management (Jana et al. 2017). Monsoon season shows relatively stable and uniform rainfall concentration with PCI in a range of 9-15. Highest PCI values were for post monsoon season in all altitudes with highest PCI values in altitude -I. Seasonality index Seasonal Index values for four altitudes fall between 0.91-0.96 (Fig. 2). Decadal trend of rainfall is decreasing for all altitudes except for altitudes I and II for the decade of 1980-89 which had an increasing trend. Over all decreasing trend agrees with the fact that monsoon rainfall is decreasing and on the other hand pre-and post-monsoon rainfall is increasing which in fact reduces the rainfall concentration disparity. This suggests that there a relative shifting in rainfall pattern with passing time with inconsistency in rainfall. Similar observations were made for the state of Kerala by Nair et al. (2014), Bundelkhand by Jana et al. 2017 and Thomas and Vol. 21,No. 3 Prasannakumar, 2016 who had reported mixed but slight decreasing trend in the time series. Change point analysis Annual series of rainfall was analyzed using Pettitt's test for change detection analysis. Time series of annual rainfall with the change points is presented in Fig. 3. The change point analysis on time series of annual rainfall indicated that significant change points in all the altitudes was in year 2014 from where above average rainfall was observed. The significant change points in the series observed during 2014 may be attributed to the growing commercial activities in the region (Jaiswal et al. 2015). Besides 2014, change point was also found in 1998 (showing decreased rainfall), 1983 (showing increased rainfall) and 2009 (showing decreased rainfall) for altitude-II, altitude-I and altitude-III, respectively. Change point in 1990's was also observed for Agra district, India (Jana et al. 2016) Rainfall periodicity Wavelet analysis for annual rainfall periodicity is given in Fig. 4 for all the four altitudes. It was found that periodicity of rainfall in altitude -I had decreased and remained almost similar in other altitudes. Higher wavelet powers in altitude -I and II after 2005 suggests frequency of extreme rainfall occurrence had increased. Dash et al. (2012) while analysing rainfall characteristics over nine agrometeorological divisions also found that heavy intensity rainfall spells are increasing. In altitude III the periodicity of such extreme events has also increased from 7 -24 years (during 1980 to 1995) to 4 -24 years (during 2006 to 2015). A periodicity of 2-8 years was found in annual and seasonal rainfall patterns in Kerala state, India (Thomas and Prasannakumar, 2016). CONCLUSIONS The study reveals that annual variability was more for minimum temperature (19.6%) in higher altitude and for maximum temperature in lower altitude. Winter season showed maximum variation for both minimum and maximum temperature in altitude -III. Monsoon rainfall is the major contributor to the annual rainfall and has higher consistency compared to other seasons. Rainfall variability was higher in post-monsoon (102 -145 %). At lower altitude, which is agricultural production areas, rainfall trends show increased concentration in monsoon and its decrease in other seasons. Maximum PCI values also confirm that annual rainfall amount is highly variable and is concentrated in the 2-3 months of the monsoon season. PCI values were higher for post-monsoon in altitude -I. High seasonal index values (0.91-0.96) revealed that rainfall is irregular and markedly seasonal with longer drier season. Higher wavelet powers in altitude -I and II after 2005 suggests frequency of extreme rainfall occurrence had increased.	2021-12-07T16:07:44.527Z	2021-11-10T00:00:00.000	{ "year": 2021, "sha1": "8c0eada6a7549743bc0cb873546f2112e3eb85ba", "oa_license": "CCBY", "oa_url": "https://journal.agrimetassociation.org/index.php/jam/article/download/252/198", "oa_status": "HYBRID", "pdf_src": "Anansi", "pdf_hash": "f3b3d50ee105616d99f9978dcf2d2424fe2fc09d", "s2fieldsofstudy": [ "Environmental Science" ], "extfieldsofstudy": [] }
118651126	pes2o/s2orc	v3-fos-license	Suzaku Observations of Three FeLoBAL QSOs, SDSS J0943+5417, J1352+4239, and J1723+5553 We present Suzaku observations of three iron low-ionization broad absorption line quasars (FeLoBALs). We detect J1723+5553 (3\sigma) in the observed 2--10 keV band, and constrain its intrinsic nh column density to nh>6\times10^{23} \cmsq by modeling its X-ray hardness ratio. We study the broadband spectral index, aox, between the X-ray and UV bands by combining the X-ray measurements and the UV flux extrapolated from 2MASS magnitudes, assuming a range of intrinsic column densities, and then comparing the aox values for the three FeLoBALs with those from a large sample of normal quasars. We find that the FeLoBALs are consistent with the spectral energy distribution (SED) of normal quasars if the intrinsic nh column densities are nh>7\times10^{23} \cmsq for J0943+5417, nh>2\times10^{24} \cmsq for J1352+4293, and 6\times10^{23}<\nh<3\times10^{24} \cmsq for J1723+5553. At these large intrinsic column densities, the optical depth from Thomson scattering can reach \sim6, which will significantly modulate the UV flux. Our results suggest that the X-ray absorbing material is located at a different place from the UV absorbing wind, likely between the X-ray and UV emitting regions. We find a significant kinetic feedback efficiency for FeLoBALs, indicating that the outflows are an important feedback mechanism in quasars. Introduction Quasars and active galactic nuclei (AGN) are ubiquitous phenomena wherever we look in the universe. The subset of quasars known as broad absorption line quasars (BALQSOs) are characterized by blue-shifted, rest-frame UV absorption troughs due to gas outflow. Even before the first definitive survey in Weymann et al. (1991), BALQSOs provided a unique glimpse into the central engines of AGN, where the nature of the intrinsic absorption has important implications for the structure of the central engine. To classify a quasar as a BALQSO, the traditional definition of Weymann et al. (1991) requires the absorption troughs to be at least 2,000 km s −1 wide. Recent studies have relaxed this, e.g., Trump et al. (2006), to include quasars with absorption troughs over 1,000 km s −1 in width. Regardless of which definition is used, BALQSOs are further divided into those that exhibit absorption troughs from only high-ionization species (HiBALs) and those with low-ionization species (LoBALs). Most BALQSOs are HiBALs, and most LoBALs also have high-ionization troughs in their spectra (Weymann et al. 1991;Trump et al. 2006). LoBALs are further classified by the low-ionization species they exhibit: in this paper, the objects surveyed have strong iron absorption, and are therefore iron LoBALs (FeLoBALs). This type of BALQSO is rare, and comprises only 1.5-2.1% of the entire quasar population (Dai et al. 2010b), while LoBALs and BALQSOs make up ∼4-7% (Dai et al. 2010b) and ∼20-40% Shankar et al. 2008a;Ganguly & Brotherton 2008;Knigge et al. 2008;Maddox et al. 2008;Allen et al. 2010), respectively, depending on the threshold of the trough width. Focusing on FeLoBALs is meaningful because of their unique characteristics and the bearing they have on the broader picture: either as an evolutionary link in the normal AGN lifetime, or as a geometric interpretation of a uniform class of objects. Besides the low-ionization absorption lines, the optical continua for LoBALs are more reddened than HiBALs and normal QSOs, suggestive of stronger dust absorption, perhaps from large quantities of dust in the vicinity of the central engine (Sprayberry & Foltz 1992). LoBALs in particular are more apparently X-ray weak than normal BALQSOs (Green et al. 2001;Gallagher et al. 2002a). There is a large portion, ∼80%, not detected in X-rays, along with lower than expected values for the UV to X-ray luminosity ratio, α ox 1 (α ox lower than expected by ∼ −0.9), that leads most previous studies, e.g. Green et al. (2001); Gallagher et al. (2002a), to conclude that there is extreme X-ray absorption in LoBALs. Streblyanska et al. (2010), however, found that LoBALs in general have lower column densities (N H < 10 22 cm −2 ) than HiBALs. This conclusion may have been influenced by their X-ray bright sample, required for sufficient quality to perform X-ray spectral analyses. Low-ionization BALQSOs with broad iron absorption lines (FeLoBALs) represent an extreme category. They have the reddest continuum spectra, complex UV spectra, and possibly largest X-ray absorption column densities. There are a handful of FeLoBALs already studied in X-rays, with only a few detections. Almost all were observed using the sensitive X-ray telescope, Chandra. The nearby FeLoBAL Mrk 231 (z = 0.042) was also detected by XMM-Newton and BeppoSAX (Turner & Kraemer 2003;Braito et al. 2004) in addition to Chandra (Gallagher et al. 2002b), which allowed for the most in-depth studies of an FeLoBAL to date. All three studies find the 0.4-10 keV X-ray emission to be mostly reflected and scattered into our line of sight, and the direct X-ray emission is mostly absorbed. Using the PDS detector on board BeppoSAX, Braito et al. (2004) detect the direct X-ray emission of Mrk 231 at 3σ in the 15-60 keV band, which allows the authors to constrain an intrinsic column density N H ∼ 2 × 10 24 cm −2 . Another recent study focused on two FeLoBALs is Rogerson et al. (2011), where the authors calculated α ox upper limits for both objects, comparing them with a large sample of normal AGN (Steffen et al. 2006). Other studies focused on subsets of quasars, but had FeLoBALs contained within their scope. Quite a few of these focused on radio-detected objects. In one study, Urrutia et al. (2005) selected 12 luminous red quasars, with the criteria that they were also detected by FIRST (Becker et al. 1995), and used standard aperture photometry (for details see § 2.2) to conduct X-ray analyses. They found that all of their objects are X-ray absorbed to some degree; in addition, they found a slightly steeper spectral slope than that for normal quasars, with an unweighted mean of Γ = 2.2 ± 0.4 and a high-energy region weighted mean of Γ = 2.1 ± 0.5. Brotherton et al. (2005) selected five radio-loud BALQSOs, three of which are LoBALs, and two of those LoBALs exhibited iron features. The three LoBALs were found to exhibit the steepest α ox values in the study. Miller et al. (2009) also selected 21 radio-loud BALQSOs, one of which is classified as an FeLoBAL. The authors proposed that the X-ray properties are explained by a model that consists of X-ray emission in the disk/corona and linked to the radio-jet. Almost all of these studies calculated intrinsic column density for these objects, most of which are in the 10 22 -10 24 cm −2 range. These will be discussed in detail in § 5.2. In this study, we will use our X-ray detections and non-detections to constrain X-ray absorption column densities. One of our objects is the first X-ray detection (at the 3σ level) by Suzaku of an FeLoBAL. It is not understood why FeLoBALs possess unique physical characteristics; evolution and geometry are the contending theories. A geometric interpretation (e.g., Elvis 2000) proposes that the broad absorption line region of a BALQSO simply arises from looking straight down a narrow outflow, while normal AGN and quasars are the same object viewed from different angles. A recent discovery of a LoBAL transitioning from an FeLoBAL within a few years rest frame favors special lines of sight, because of the short time scale (Hall et al. 2010). On the other hand, Lípari et al. (2009) proposed that BALQSOs, and FeLoBALs especially, are early progenitors of normal AGN and quasars, at the very end of an extreme starburst period and in the process of blowing out material from type II SNe that is Fe II rich. This evolutionary theory is partially supported by the higher fraction of LoBALs at high IR luminosities (e.g., Farrah et al. 2007;Urrutia et al. 2009;Dai et al. 2010b). A recent Spitzer spectral survey of six FeLoBALs (Farrah et al. 2010) showed significant signatures of dust and PAH emissions. Spectral modelings in the UV regime also support a large covering fraction of the wind for FeLoBALs (e.g., Casebeer et al. 2008). Other evidence, such as the decreasing fraction as a function of radio luminosity, is consistent with a geometric interpretation (Shankar et al. 2008a;Dai et al. 2010b). Therefore, the population of LoBALs and FeLoBALs could also be a combination of both evolution and geometry (Dai et al. 2010b). Another interesting question is how FeLoBALs contribute to the quasar feedback process. AGN feedback is widely used in galaxy formation models (e.g., Granato et al. 2004;Scannapieco & Oh 2004;Hopkins et al. 2005;Shankar et al. 2008b) to explain the co-evolution of AGN and their host galaxies. Feedback can reproduce such phenomena as the M-σ relation (e.g., Graham et al. 2011;Gültekin et al. 2009), star formation rates (Silk & Nusser 2010), and even the shortfall in the halo baryon fraction (Silk & Nusser 2010). Whether FeLoBALs are an evolutionary stage or a geometric interpretation of AGN, studying their feedback efficiency allows us to add to our limited picture of their physical characteristics. In this paper we present new Suzaku observations of three FeLoBALs, SDSS objects J094317.59+541705.1, J135246.37+423923.5, and J172341.08+555340.5 (hereafter J0943+5417, J1352+4239, and J1723+5553, respectively). The targets are classified as FeLoBALs by Trump et al. (2006), and are the brightest (K s < 14.4 mag) of the 2 Micron All Sky Survey (2MASS ) selected BALQSOs . J0943+5417 is a radio-intermediate quasar detected with a peak flux of 1.71 mJy/beam and RMS noise of 0.145 mJy/beam by Faint Images of the Radio Sky at Twenty-Centimeters (FIRST). None of these objects are detected at the flux limit in the NRAO/VLA Sky Survey (Condon et al. 1998). J0943+5317 and J1352+4239 are typical FeLoBALs with extremely complex rest frame UV spectra, whereas J0943+5417 has absorption troughs so wide that they overlap and completely suppress the UV continuum. We extrapolate our UV data from 2MASS, which samples the rest frame optical bands, and then correct to the rest frame UV. The complexity of the observed optical/rest frame UV spectra makes it extremely difficult to model the continuum. This is addressed in § 3. Since the targets all have redshifts z ∼ 2, the XIS on-board Suzaku can probe the rest frame energy up to 30 keV, where the rest frame 6-30 keV (observed 2-10 keV) emission is only moderately affected by absorption even if N H ∼ 10 24 cm −2 , allowing us to detect the FeLoBALs if they are Compton thin. We study the relationship between X-ray and UV luminosities for these three FeLoBALs, and compare our values to a large sample of optically selected normal quasars. Our Suzaku observations and data reduction are described in §2, followed by calculations of 2500Å luminosities in §3. Section 4 presents results, followed by discussion in §5. Throughout the paper we assume a cosmology of H 0 =70 km s −1 Mpc −1 , Ω M =0.3, Ω Λ =0.7, and k=0. Suzaku Observations We observed the three FeLoBALs in 2009 with the Suzaku X-ray observatory (Mitsuda et al. 2007), with exposure times ranging between 32-36 ks for each object. Table 1 lists the observation log. During each observation, the standard data mode was used with the pointing observation mode. The three working X-ray Imaging Spectrometer (XIS) cameras (XIS0, XIS1, XIS3, Koyama et al. 2007) were operated in both the 3x3 and 5x5 photon counting modes with a minimum time resolution of 8 seconds. Suzaku operates in a low-earth orbit, 550 km above the earth, completing one period in 96 minutes. The background in our observations is low and stable, without flares, consistent with most Suzaku observations (Yamaguchi et al. 2006). Suzaku Data Reduction Starting with the cleaned event files, we visually inspect the data using DS9 2 to check for any anomalies and ensure that our data extraction is completely contained within the Suzaku 2 An astronomical imaging and data visualization application (Joye & Mandel 2003) field. Using the standard aperture photometry, the half-power diameter region around each source is obtained using the Suzaku Technical Description 3 , along with a corresponding background annulus. The background annuli are chosen individually for each mode (3x3, 5x5) to maximize the background area while keeping the inner radius large enough not to be contaminated by the source and the outer radius still contained within the field. The effective area for the background annulus is measured from the mean of the inner and outer radii of the background region. We then use the High Energy Astrophysics Science Archive Research Center (HEASARC) tool XSELECT in conjunction with these regions to filter the energy to the observed full (0.2-10 keV), soft (0.2-2 keV), and hard (2-10 keV) bands, extracting the counts and exposure times. We first combine the 3x3 and 5x5 mode data, then measure the source and background counts for the three CCDs, XIS0, XIS1, and XIS3, separately. We note that XIS1 is a back-illuminated instrument, which has a higher background and sensitivity at lower energy ranges. The count rates and uncertainties for XIS0 and XIS3, which are both front illuminated (FI) instruments, are found separately and then combined using the Bayesian estimation method to give a total FI count rate and error for the count rate. The final count rates and uncertainties are listed in Table 2. We first examine the full band, and do not detect the sources; then we examine the soft band, with the same result. In the hard band, we do not detect the sources in most situations. We marginally detect SDSS J1723+5553 (2.6σ and 1.6σ) in both of the FI CCDs in the observed 2-10 keV band. Combining the data from the two FI CCDs, we detect J1723+5553 at the 3σ significant level ( Table 2). The FI CCDs have comparable effective areas, and lower background rates at 2-10 keV than the BI CCD, which means they are more efficient at detecting faint sources in this energy band. Thus the non-detection of the source in the BI CCD (XIS1) is expected, given the low X-ray flux. Therefore, we conclude that the detection of J1723+5553 in the 2-10 keV band is real. We do not detect the other objects in any of the detectors. Finally, we use the HEASARC tool PIMMS to convert the count rates into fluxes. Since the intrinsic column density of the quasars is unknown, we use a range of 10 20 ≤ N H ≤ 10 25 cm −2 to calculate the unabsorbed, intrinsic fluxes. For the galactic column density, the Dickey & Lockman (1990) average N H values are used (1.3 × 10 20 cm −2 , 1.21 × 10 20 cm −2 , and 3.08 × 10 20 cm −2 , for J09843+5417, J1352+4239, and J1723+5553, respectively). The photon index is taken to be Γ=1.9 (e.g., Reeves & Turner 2000;Dai et al. 2004;Saez et al. 2008), equivalent to an energy index α=0.9. The unabsorbed fluxes are calculated from the BI and FI rates separately using PIMMS and then the upper limits are combined using the Bayesian estimation method. We use just the FI rates in calculations for J1723+5553. Next, we estimate the monochromatic 2 keV (rest frame) luminosity using the unabsorbed fluxes (or flux limits) in the observed 0.2-2 and 2-10 keV bands. Using the power-law model, the flux, F , is given by, where f 0 is a constant found analytically for the 2-10 and 0.2-2 keV ranges. From the simple relation in Equation 1, we interpolate the value for the monochromatic flux density at the observed energy E X = 2 keV /(1 + z) using, We find the 2 keV rest frame luminosity (l 2keV ) from the corresponding observed fluxes using, where ν, obs = ν/(1 + z) and D L is the luminosity distance. Since f ν,obs is calculated for a range of intrinsic absorption, we obtain a range of monochromatic luminosities at 2 keV (Table 3). The 2500Å Luminosity We calculate the rest frame 2500Å luminosity, l 2500 , by extrapolating the observed 2MASS magnitudes (probing rest frame 4000-7000Å for z ∼ 2 objects) using the mean quasar SED of Richards et al. (2006). Since all three objects are at z ∼ 2, the 2MASS bands sample the rest-frame optical bands of the FeLoBALs, where there is less spectral Note. -The 3σ upper limits are reported here, and the 3σ hardband detection of J1723+5553. The count rates are within the half-power diameter region of Suzaku, and are scaled to full power after the conversion to fluxes. The units are cnt s −1 . complexity than the rest-frame UV bands. The Sloan Digital Sky Survey (SDSS ) magnitudes sample the rest frame UV of the FeLoBALs, where it is extremely difficult to model the continuum due to the presence of complicated absorption troughs and dust extinction. Although the observed 2MASS wavelengths are further from the rest frame UV than the observed SDSS wavelengths, we do not have to rely on a difficult to estimate model of the continuum. The observed optical spectra of J0943+5417 is almost completely suppressed (see Figure 1), which demonstrates why we use the 2MASS data that circumvents reliance on complex corrections. We first correct for Galactic extinction caused by the Milky Way for the 2MASS magnitudes. This extinction is given for the u-band in the SDSS quasar catalog for each object (Schneider et al. 2007). From that value, we obtain the extinction in the 2MASS bands using the Milky Way extinction curve, where the values of color excesses, E(λ − V ), and R V = 3.08 are taken from Pei (1992). After correcting for Galactic extinction, we calculate the observed flux densities at the central J, H, and K s wavelengths, and obtain the rest frame monochromatic luminosities at optical bands (at ∼6,900Å) using Equation 3. We correct for the intrinsic dust extinction in these quasars assuming E(B − V ) = 0.077 with an SMC-like extinction curve, found by Reichard et al. (2003) to be a good approximation for LoBALs. Since FeLoBALs have higher dust extinctions than LoBALs, we possibly under-correct the intrinsic dust extinction for our targets. Finally, we extrapolate the rest-frame optical luminosities to l 2500 using the mean quasar SED of Richards et al. (2006). The parameters and intermediate results used in all these calculations are listed in Table 4. The results obtained from J, H, and K s bands are consistent within 20% for J0943+5417 and J1352+4239, suggesting no significant complexity in the corresponding spectra. For J1723+5553, the 2500Å luminosities extrapolated from J and H bands are consistent within 20%; however, the value extrapolated from the K s band is twice as large. It is possible that the observed K s band is contaminated by a strong H α emission line. For all three objects, we use the average of the l 2500 values extrapolated from the J, H, and K bands, in our following calculations. Results After calculating the intrinsic monochromatic luminosities at 2500Å and 2 keV rest frame, we measure the broadband spectral index between the UV and X-ray bands, α ox = 0.3838 × log (l 2keV /l 2500 ), and then compare our α ox values to those of the large sample of Table 3. 2 keV rest frame Monochromatic flux and luminosity. Intrinsic Absorption and α ox We first study the relation between α ox and the intrinsic absorption in Figure 2. Each sub-figure corresponds to one FeLoBAL, and shows the entire input range for intrinsic column density. The values calculated from the 2-10 keV (hard) spectrum are depicted in black, and the 0.2-2 keV (soft) spectrum are depicted by red, filled symbols. Where monochromatic luminosities at 2 keV are upper limits due to the non-detection of a source, α ox is also only an upper limit, depicted in all Figures by arrows. Included are solid lines for the expected α ox ∼ l 2500 relation, as found by Steffen et al. (2006), and dashed lines to depict the expected scatter. At 2 keV the intrinsic luminosity and its uncertainty are dependent on the column density within the FeLoBAL itself, which causes intrinsic absorption that must be corrected for. We use a range 10 20 ≤ N H ≤ 10 25 cm −2 for intrinsic column densities since the true quantity is unknown. This gives a different value of intrinsic luminosity, and therefore α ox , for each column density. The luminosities at 2500Å are extrapolated from the rest frame optical luminosities, using a normal quasar SED from Richards et al. (2006), where we have corrected both the intrinsic extinction and Galactic extinction due to dust as discussed in § 3. We find the uncertainty in the luminosities to be ∼20%. In the two non-detected FeLoBALs (J0943+5417 and J1352+4239), for column densities up to 1 × 10 23 cm −2 , the soft limit is more stringent for α ox , although the difference between the two limits diverges by ≤ 0.2. At higher column densities, the hard spectrum is the appropriate limit to use. The divergence between the soft and hard limits is increasingly larger with increasing column density. J1723+5553 is a different story: only at N H ≥ 6×10 23 cm −2 are the upper limits from the soft X-ray spectrum consistent with the 3σ detection in the hard X-ray spectrum. The 3σ detection and its uncertainties for α ox calculated from the hard X-rays are ≥0.3 above the upper limits given by the soft X-ray energy range for column densities < 6×10 23 cm −2 . Thus, the X-ray hardness ratio of J1723+5553 constrains the intrinsic column density to N H ≥ 6 × 10 23 cm −2 . Figure 3 shows the correlation between l 2500 and α ox , with the best-fit linear regression for normal AGN found by Steffen et al. (2006) plotted. This relation is given by α ox = (−0.137 ± 0.008)log(l 2500 ) + (2.638 ± 0.240) (Steffen et al. 2006). The solid line shows the expected α ox , and the scatter is depicted by the dashed lines. Column density increases as α ox becomes less negative. The upper limits on α ox from the soft X-ray spectrum are shifted slightly to the right of the hard X-ray upper limits in Figure 3 for ease of viewing. The upper limits plotted for J0943+5417 and J1352+4239 are consistent with the normal AGN in the scatter of ±0.24 for N H ≥ 4 × 10 23 and N H ≥ 8 × 10 23 cm −2 , respectively. This range falls in the column density region (N H ≥ 1 × 10 23 ) where the hard spectrum provides the appropriate upper limit. For J1723+5553, combining the constraint from X-ray hardness ratio (N H > 6 × 10 23 cm −2 ), the values of α ox fall within the expected scatter if the intrinsic absorption is in the range of 6 × 10 23 ≤ N H ≤ 3 × 10 24 cm −2 . Figure 4 shows the correlation between l 2keV and α ox , again with the best-fit linear regression and expected scatter, using the relation, α ox = (−0.077±0.015)log(l 2keV )+(0.492± 0.387), from Steffen et al. (2006). Column density increases from left to right, corresponding to rising intrinsic luminosity. For column densities N H ≥ 7 × 10 23 cm −2 for J0943+5417 and N H ≥ 2 × 10 24 cm −2 for J1352+4239, α ox is consistent with the normal AGN sample. The upper limits given by the soft spectrum fall within (and above) the scatter in the high column density region, where the hard spectrum gives the better constraint. Our 3σ detection of J1723+5553 indicates that for column densities N H ≤ 8×10 24 cm −2 , J1723+5553 is consistent with the SEDs of normal AGNs. Column Densities and Intrinsic X-ray Luminosities We find significantly high column densities for all three FeLoBALs studied, 6 × 10 23 ≤ N H ≤ 3 × 10 24 cm −2 for J1723+5553, N H ≥ 7 × 10 23 cm −2 for J0943+5417, and N H ≥ 2 × 10 24 cm −2 for J1352+4239. Only a small sample of FeLoBALs has been studied in Xrays previously. We summarize all the measurements in Table 5, which is divided into two subsamples: those that are radio loud, and those that are not. The radio loud objects tend to have lower column densities, which could be due to geometry or evolution. The X-ray emission may be linked to the jets as well as from the corona right above the central portion of the accretion disk (e.g., Miller et al. 2009). The BAL wind (see Gallagher & Everett (2007) for a detailed illustration), would then only absorb X-ray emission from the accretion disk region. We do not know for certain what complicated structure dependence for Xray emission may exist in a radio loud FeLoBAL, but the column density results are still important and thus we include them for completeness. Our three FeLoBALs are radio quiet, and our constraints are consistent with the subsample that is not known to be radio loud. These objects in general are found to have N H on the order of a few 10 23 cm −2 or higher. Notably, the most recent study by Rogerson et al. (2011) found column densities for J2215-0045 and J0300+0048 to be N H ∼ 10 24 , with J2215-0045 just above the range we calculate for J1723+5553, but still consistent with our N H for the two non-detections. The most recent calculation of column density for MrK 231 by Braito et al. (2004), using BeppSAX, found N H ∼ 2 × 10 24 cm −2 , in agreement with all three of our objects. Green et al. (2001) cite lower column densities (N H ≥ 6.5 × 10 22 cm −2 ), but they also noted that this is an average expected column density for a HiBAL, and that N H for a LoBAL would be higher. The fact that our 3σ detection in J1723+5553 is in the observed 2-10 keV (rest frame 6-30 keV) hard band rather than the soft X-ray band is significant, as this is expected from highly absorbed X-ray emission from the corona above the SMBH. The non-detection of the source in the observed 0.2-2 keV (rest frame 0.6-6 keV) band is consistent with our interpretation that we detect the direct X-ray emission in the observed 2-10 keV band, as high column densities will suppress the soft X-rays more than the hard X-rays. This is further supported by the fact that none of these objects is detected, even at the 3σ level of significance, for the broadband observed 0.2-10 keV (rest frame 0.6-30 keV) energy range. If the observed 2-10 keV spectrum is reflected or scattered into our line of sight, the X-ray spectrum will be a flat power-law, and we would also detect the source in the observed 0.2-2 keV band, which has a larger collecting area for Suzaku. Combining the measurements from this paper and those from Braito et al. (2004); Rogerson et al. (2011), it is most likely that the intrinsic N H column density absorbing the X-ray emission in a radio-quiet FeLoBAL is in the range from ∼ 10 24 up to 10 25 cm −2 . This is about 1-2 orders of magnitude higher than the X-ray absorption range in BALQSOs (N H ∼ 10 22 − 10 24 cm −2 , e.g., Gallagher et al. 2006). Thomson Scattering Implications Such high column densities are interesting physically, as they are either in or on the cusp of a regime where the Thomson scattering cross-section, σ T = 6.65 × 10 −25 cm 2 , will be large enough to scatter incident photons. Although we probe the rest frame quasar spectra in the rest frame 6-30 keV (observed 2-10 keV) band, the Klein-Nishina correction to the Thomson scattering cross section is still negligible. For J1723+5553, using the expected α ox from the α ox -l 2kev relation, the optical depth is τ ∼ 1. For J0934+5417 and J1352+4239, the expected α ox value from the α ox -l 2keV relation gives an intrinsic N H ∼ 8.5 × 10 24 cm −2 , which indicates a Thomson scattering optical depth of τ ∼ 6. Such a large optical depth will also significantly modulate the UV flux (∼400 times dimmer); however, these objects are already more luminous in the UV than most of the AGN in the Steffen et al. (2006) study. As another check, we calculate the black hole mass of the quasars assuming they are emitting at 1/3 of the Eddington luminosity (e.g., Shankar et al. 2010). We extrapolate our rest frame luminosities to 5100Å and assume a bolometric correction of 10.33 for 5100Å (Richards et al. 2006) before using Equation 5 to find the black hole masses. For the values reported here, all three FeLoBALs have M BH ∼ 1.4 × 10 8 M ⊙ . Even at the lower end of the column density constraints, N H ≥ 6 × 10 23 cm −2 , the optical depth would Urrutia et al. (2005) Note. -Previous studies of FeLoBALs in the X-ray. We have added column density constraints when the study provided them. a Radio moderate. b Green et al. (2001) note that this should represent an average HiBAL, and would be higher for a LoBAL. c Sometimes classified as a mini-BAL. be τ ∼ .4, which would dim the intrinsic luminosity by one and a half times. At the column density corresponding to the expected α ox for J0943+5417 and J1352+4239, ∼ 8.5 × 10 24 cm −2 (from the α ox -l 2kev relation, and still within the limits for J1723+5553), the optical depth of τ ∼ 6 would mean the intrinsic luminosity would be increased by 400 times, giving a black hole mass of M BH ∼ 3.2 × 10 12 M ⊙ . Such a large mass is not a physical possibility; therefore, it is unlikely that these objects are super-luminous in the UV, and we only see the tiny fraction of Thomson scattered emission. Since the X-ray emission region is expected to be smaller than the UV emission region from either variability (e.g., Chartas et al. 2001) or quasar microlensing (e.g., Dai et al. 2003Dai et al. , 2010aPooley et al. 2007;Morgan et al. 2008;Chartas et al. 2009) arguments, it is instead possible that the X-ray absorbing material is located between the X-ray and UV emission region and only covers the X-ray emitting region. This is consistent with the disk wind models of Gallagher & Everett (2007) and Murray et al. (1995), where an essential component of the model is the Compton thick shielding gas between the X-ray and UV emission, protecting the disk wind from being ionized by the X-ray emission. Rogerson et al. (2011) also reached a similar conclusion, where the Thomson scattering optical depth was argued to have τ ≥ 3 (≥ 20 times dimmer). In this paper, the constraint is likely more stringent with τ ∼ 6 (∼ 400 times dimmer). Aoki (2010) analyzed the UV spectrum of J1723+5553, and found the N H column density in the UV wind was N H ≥ 5 × 10 17 cm −2 using the curve of growth method with the unresolved Balmer absorption lines. Aoki (2010) pointed out that the covering fraction cannot be determined due to unresolved absorption lines, and assumed a covering fraction of 1. The column density would be higher than the given limit if the covering fraction is significantly less than 1. The limit derived by Aoki (2010) is seven orders of magnitude less than, but still consistent with, our X-ray column densities. If the X-ray and UV absorbing gasses are located in different regions, it can still be incorporated into both existing geometrical interpretations of FeLoBALs (e.g., Murray et al. 1995;Gallagher & Everett 2007) or evolutionary models (e.g., Fabian 1999). However, for evolutionary models the UV emission should come from the photosphere of the gas/dust cloud rather than the accretion disk. Such models have been simulated in recent studies (e.g., Casebeer et al. 2008), and support that FeLoBALs may be an evolutionary stage in the development of normal quasars. Another explanation is that FeLoBALs have intrinsic SEDs different from normal quasars. They could be extremely X-ray weak compared to normal quasars, as in the case of the narrow-line quasar PHL 1811 (Leighly et al. 2007), or the Narrow-Line Seyfert 1 galaxy WPVS 007 (Grupe et al. 2008). This would nullify our N H constraints for J0934+5417 and J1352+4239, since they are obtained by assuming a normal quasar SED. However, this would not explain J1723+5553, which we detect in the observed 2-10 keV band, but not in the ob-served 0.2-2 keV band. If J1723+5553 were intrinsically X-ray weak, and X-ray unabsorbed, we would expect to see a flat-line power-law slope for quasars in the hard and the soft X-ray bands. This is not consistent with our observations. In addition, the N H constraint from the X-ray hardness ratio is still valid for N H ≥ 6 × 10 23 cm −2 with the Thomson scattering optical depth τ ≥ 0.4, which could only marginally scatter the UV emission if the true N H column density is close to the lower boundary. More X-ray detections of FeLoBALs are needed so that we are not restricted by using the broadband SED constraints. It is also possible that quasar host galaxies contribute a fraction of rest frame 4000-7000Å emission; however, the contamination at this band is usually small. In addition, we correct the intrinsic dust extinction for FeLoBALs using the average dust extinction for LoBALs, which will underestimate the intrinsic flux and give less steep α ox values. A priori, it is unclear what the net effect of these two possibilities will be on the broadband flux; IR spectroscopy is needed to help resolve these issues. AGN Kinetic Feedback We calculate the kinetic feedback efficiency, ǫ k =Ė k /L Bol , of the X-ray absorber in FeLoBALs using the location, luminosity, and N H column density constraints obtained in this paper. In particular,Ė k is the kinetic feedback power,Ė k =Ṁv 2 /2 = 4πµm p f c rN H v 3 (e.g., Moe et al. 2009), where µ is the mean molecular weight, m p is the proton mass, f c , r, N H and v are the covering fraction, location, column density, and velocity of the wind, respectively, and L Bol is the bolometric luminosity of the quasar. If the SEDs of FeLoBALs are consistent with those of normal quasars, we obtain a reasonable N H ∼ 3 ×10 24 cm −2 ; this is the upper end of the column density range for our 3σ detection, and entirely consistent with our column density lower limits for the non-detections. We locate the wind between the UV and X-ray emission regions, ∼ 40r g and r g = GM/c 2 , using the microlensing constraints of Dai et al. (2010a). For the covering fraction, we use the intrinsic FeLoBAL population range of 1.5% and 2.1%, depending on which model (Dai et al. 2010b) is used, and assume the quasars are emitting at a typical 0.3 Eddington limit (Shankar et al. 2010). We are left with one uncertainty, the velocity of the X-ray absorber. The velocity of the BAL wind is usually measured in the UV wind, which can reach up to ∼ 0.1c. Since the X-ray absorbing wind is located at smaller radii than the UV absorber, its velocity can be higher than the wind velocity measured from the UV spectrum. The only velocity measurement of the X-ray absorber is from the blue-shifted X-ray absorption lines detected in a few gravitationally lensed BALQSOs (Chartas et al. 2002(Chartas et al. , 2007, and this velocity can reach 0.3-0.8c. Since the X-ray absorption lines are mostly detected in mini-BALs, we could possibly be looking through the edge of the wind, where the wind can be by fully accelerated. Thus, we consider the X-ray absorption line as providing an upper limit, and assume our wind velocity is 0.1-0.3c, between the estimates from the two methods. We find the feedback efficiency, ǫ k , for FeLoBALs is in the range (outflow velocity dependent) of either 0.2%-4.8% for a covering fraction of 1.5%, or 0.3%-6.9% for a covering fraction of 2.1%. To reach the minimum of 5% required to explain the co-evolution between black holes and host galaxies (e.g., Silk & Rees 1998;Granato et al. 2004;Hopkins et al. 2005), with the other variables given here, we need a column density of N H ≥ 8 × 10 25 cm −2 or N H ≥ 6 × 10 25 cm −2 for low velocity wind with the different covering fractions. These are consistent with our lower limits on column density for the two non-detections, but out of reach for J1723+5553. However, for the high end of the conservative wind velocity range, only N H ≥ 3×10 24 and N H ≥ 2×10 24 are needed for f c =1.5% and 2.1%, respectively. Both of these column densities are consistent with the results for J1723+5553, and thus it is likely that FeLoBALs can contribute to AGN feedback. This becomes even more likely when models like Hopkins & Elvis (2010) are considered. They describe a "two-stage" feedback process, where a weak wind from the central engine energizes a hot, diffuse interstellar medium, which is then amplified when it it hits instabilities in a cold cloud within the host. This amplification outside of the central engine requires feedback efficiency as small as 0.5%. The BAL wind in FeLoBALs, therefore, is a promising candidate for the feedback process responsible for the co-evolution between black holes and host galaxies. More measurements of N H are needed to further quantify how important the contributions are from FeLoBALs and BALQSOs to kinetic feedback efficiency.	2011-05-29T20:33:25.000Z	2010-11-18T00:00:00.000	{ "year": 2010, "sha1": "80c2881338705aeb882c23a819bb000adb2163c3", "oa_license": null, "oa_url": "http://arxiv.org/pdf/1011.4327", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "80c2881338705aeb882c23a819bb000adb2163c3", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics" ] }
251317303	pes2o/s2orc	v3-fos-license	Tartaric acid as a novel additive for approaching high-performance capacity retention in zinc-ion battery Among the rechargeable batteries, aqueous zinc-ion batteries (ZIBs), due to their safety, low cost, eco-friendly, and simplicity in construction, have received much attentions. One of the most critical parts of the battery technology is the electrolyte additives, which have been less studied against their essential roles. To develop the quality of these batteries, specific parameters such as economics, easy design, significant time duration, high electrical discharge, fast charge/discharge rate, acceptable power/ energy density, and acceptable cycle efficiency are essential. In this report, is focused on the aqueous solution of some white crystalline organic acids as novel electrolyte additives such as succinic, tartaric, citric, maleic, and/or acetic acids as battery over-voltage reducing agents to modify the electrical performance of the ZIBs. For instance, significant characteristics of tartaric acid as specially selected electrolyte additive to the ZIBs, exhibit an excellent capacity up to 374 mAh g−1 with acceptable rate capability and high-capacity retention as large as 91.0% after 7200 cycles. To investigate the battery behavior and propose the probable mechanism behind this phenomenon, some analytical methods are utilized. www.nature.com/scientificreports/ Current research in battery technologies is focused on the exploration of the anode/cathode materials and liquid-free electrolyte materials during the optimization of the electrode/electrolyte interface 20 . Many outstanding types of research in this aspect have been done, such as the widening of the electrochemical window of electrolyte 21 , electrostatic shell protection of the electrolyte on the anode side, and cathode/electrolyte ion balance control 22 . But, in new research on ARBs, additives are being used as a low-cost solution and a practical, stable, and durable way to solve the zinc ion battery problem 17 . However, recent studies on the ZIBs have been majorly focused on cathode materials and Zn ion intercalation mechanisms 23 . For instance, to solve the issues of the Mnbased cathode, several strategies have been proposed, such as surface coating 24 , microstructure construction 25 , doping 26,27 , electrolyte regulation 22 , etc. Fluorine doping is considered one of the newest methods for fabricating a stable and reversible framework of the Mn-based cathode, which has been widely used in electrode materials of electrochemical energy storage 28 . However, one part of the battery, which possessed less attention during the last decades, is the cathodic/anodic over-voltage reducing species that are often considered as the electrolyte additives to the battery's electrolyte medium to enhance the cell's performance 29 . To overcome the cell's over-voltage and have significant improvement in the battery's figures of merit, such as the reversibility, durability, and capacity, some electrolyte like tartaric, maleic, succinic, and citric acid are adopted as the electrolytes additives to access the mentioned goal. This process also is considered as one of the most cost-effective and effective methods for of the improvement the battery's performance 29 . To solve these problems, herein, we modified the performance of the reversible aqueous Zn/MnO 2 battery using some white crystalline organic acids as battery over-voltage reducing agents. In this system, briefly, the γ-MnO 2 cathode was highly reversible and stable in a ZnSO 4 aqueous electrolyte with the aqueous solutions of white crystalline organic acids such as tartaric acid as effective electrolyte additives. The motivation behind this study, therefore, creates a more efficient, reliable, environmentally friendly, low cost, high enough electrical energy, higher power densities, more improved reversibility, and longer cycle life, compared to the generally commercialized ZIBs. Because of the safety, low cost, and better stability of each anode and cathode inside the aqueous solution, mid acidic pH was selected 31 . The pH 4.0 ± 0.1 condition has been chosen as optimum pH using HCl or NaOH standard solutions with 0.1 mol L −1 concentration, due to its moderate acidic conditions and the minimum anode corrosion, which were promoted especially at high acidic electrolyte solutions 14 . Cathode and electrolyte preparation. The γ-MnO 2 cathode was prepared by mixing synthesized γ-MnO 2 powder, carbon black, and polyvinylidene fluoride at the optimum weight ratio of 7.0:2.0:1.0 with total mass the same as 10.00 ± 0.1 mg cm -2 inside 5.0 mL of N-methyl pyrrolidone as solvent inside a Cole-Parmer watch glass (12/pk, Universal Scientific Industrial (Shanghai Co., Ltd. Company, China). These compounds were then mixed and homogenized using a stainless steel spatula to provide a pasty slurry mixture. This composite www.nature.com/scientificreports/ was then coated (immobilized) onto one side of the carbon fiber sheet handy using the spatula at room temperature. To have a regular and reproducible coating, the paste composite was scrapped vigorously long the carbon fiber sheet at least three times. After preparing the cathode support, it was put inside the oven at 140 °C for 8.0 h inside the air atmosphere to evaporate the solvent, dry the reagent and provide robust connectivity between the carbon-based cathode support and the introduced composite material. The electrolyte for this test was the mixture of 3.5 mL of triple-distilled water and 1.0 mL ZnSO 4 solution (2.0 mol L −1 ), along with individual introduction of the aqueous solutions (0.5 mL, 1.0 mol L −1 ) of different white crystalline organic acids such as tartaric, maleic acid, citric acid, and succinic acids to the electrolyte solution. Procedure. A modified Zn/MnO 2 battery system was assembled using the γ-MnO 2 -modified electrode as the cathode, a glass fiber sheet as the separator, and the Zn foil as the anode electrode. The electrolyte mixture of 3.5 mL triply distilled water and 1.0 mL ZnSO 4 solution (2.0 mol L −1 ) and 0.5 mL aqueous solution of tartaric acid as additive with 1.0 mol L −1 concentration. The analyses were evaluated at the selected pH value. The OCP test was also measured by the electro analyzer. The characterization process was also based on analytical methods such as "Brunauer-Emmett-Teller, BET, " electrochemical impedance spectroscopy, scanning electron microscopy, cyclic voltammetry, X-ray photoelectron spectroscopy, and Tafel analysis were utilized. Results and discussion Among the rechargeable battery, the aqueous ZIB, due to possessing different advantageous such as safety, cost-effective, availability, eco-friendly, and simplicity in construction, has received lots of attention. These characteristics have made the ZIB to be considered an appropriate candidate for various applications 3 . For this purpose, many scientists worldwide have focused on the ZIB features to enhance its capacitive performance. Initial attention have been focused on some general aspects of the battery technology, such as electrolyte type, and concentrations, electrode geometry and modified species on the electrode surface. Fortunately, for least two last decades, these features have significantly improved in the capacity, lifetime, electrical stability, and durability of the ZIB 6 . However, less attention has been applied to the enfluential role of physicochemical phenomena, affecting the over-voltage of the electrode half-reactions. These attentions have been focused on different phenomena 5, 6 . For instance, electrode morphology is vital to change the electrode system pathways towards less positive/ negative anodic/cathodic half-reactions, respectively. These factors also limited the polarization challenge via inter-cell ohmic potential (IR) reduction by controlling the electrode surface area dimension 5 . These phenomena consequently prevent the ZIBs from any high current flow of shock. However, these features are besides the promotion of the electrical conductivity of the electrode materials, controlling their morphology and lowering the inter-electrode distance 5,6 . At these conditions, therefore, great applications of depolarizers are limited to neither the kinetic polarization nor reducing the electrolyte ohmic resistance 5 . However, to develop these systems, scientists have intensely focused on the inert ionic species, nanostructures, different surfactants (to adjust the electrolyte surface tension), etc. 5,6 . Nevertheless, to the bests of knowledge, no reports have been published on the phenomena, during smartly and reversibly playing with the electrode active surface area. For this purpose, in this research, fantastic features of the ZIBs have been reported by focusing on different materials, as evaluated in detail in the following sections. It should be noted that, because of better chemical stability (i.e., minimum anode corrosion) of each anode and cathode electrode inside the aqueous solution, moderate acidic pH condition (4.0 ± 0.1) 14,22 has been selected as the optimum value. Approach the smartly and reversibly over-voltage reducing agents. To introduce smartly and reversibly over-voltage reducing agents, characteristics such as partial adsorption properties, moderate solubility inside the aqueous electrolyte solution, and preferential electrode's surface intercalation properties seemed necessary. To access these reagent(s), white crystalline organic acids were considered as initial candidates. www.nature.com/scientificreports/ According to our preliminary tests, different white crystalline organic acids, as electrolyte additives, were evaluated on the OCP (open circuit potential) response, according to the recommended procedure. Based on the results, significant enhancement (~ 20%) was observed in the OCP, compared to the control system (ZIB electrolyte) as well as other water-soluble organic compounds such as EDTA, glucose, etc. (Fig. 2). This phenomenon provided a significant impact on the battery's OCP. As shown, maximum OCP (vs. Zn) was observed for our available 4-6-C carboxylic acids such as tartaric acid, maleic acid, and citric acids. Consequently, the aqueous solution of tartaric acid with a 1.0 molar concentration was selected for future analyses. According to the results ) Fig. 3(, effects of the tartaric acid as the external electrolyte additive on some EIS (electrochemical impedance spectroscopy) parameters such as R s (solution resistance), R dl (double-layer resistance), and C dl (double-layer capacitance) have been evaluated based on the Nyquist plots (Figs. 3A,B). Based on the formulas reported in Ref. 32 , the estimated parameter's values were shown in the insets of Fig. 3. The coordinate unit is based on the ohm (multiple in hundred zʹ). As clearly shown, significant reductions were observed in the R s and R ct of the electrolyte solution during the introduction of tartaric acid as an additive. These reductions from one side were attributed to the decrease in the IR drop (ohmic potential) of the cell, and from the other side, exhibited a significant increase in the K app (apparent rate constant) value 33 , which was in the reverse relationship with the C dl 25 . Therefore, higher ion conductivity occurred. So, to evaluation of tartaric acid addition in the battery electrolyte, the ionic conductivity was determined by σ = L A×R s , where L was the electrode distance and A the area of stainless-steel electrode, and R s as electrolyte resistance (it should be noted that to better observation of tartaric acid effect, this experiment was performed in the 2.5 cm distance between electrodes, and 1 cm 2 surface area of electrodes). According to the bulk resistance in the presence and absence of tartaric acid, the ionic conductivity was obtained at 62.5 and 11 mS/cm, respectively. According to the obtained results, significant increase in ionic conductivity of battery electrolytes were observed in the presence of tartaric acid. This enhancement resulted in a decrease in the resonance frequency and, consequently enhancement in the C dl value. All these effects led to the significant reduction in the cathodic over-potential of the ZIBs and, consequently promotion of the electromotive force (E emf ) of the battery from + 1.44 to + 1.75 V (vs. Zn). In addition, the higher stability of the battery at high cycles (up to at least This evidence, therefore, indicated thats tartaric acid, as the electrolyte additive, did not affect the redox reactions in the battery cell. For more confidence, the two plateau in the charge/discharge curves (Fig. 4B) approved the cycling performance of the battery and also showed significant improvements compared to the previous reported researches 14,16 . To compare the effect of the tested electrolyte additives, the battery's capacity was measured in the presence and absence of tartaric acid as an additive (Fig. 4C), which illustrated that; the capacity was significantly high enough compared to the lack of any additives. As exhibited, when using additive in the battery electrolyte, it was observed that the capacity vs. cycles were more improved compared to the absence of the additive that exhibited no sharp drop in the capacity curve. As mentioned, the presence of additives significantly improved the cyclability and cycle life of the battery. Based on these results, these additives not only reduced the irreversible capacity, but also protected the cathode material from the overcharge. These results were beside their influenced some physical properties of the electrolyte, such as ionic conductivity, viscosity, and wettability to the separator. Adding tartaric acid as an additive to the electrolyte greatly improved the delivery of a higher capacity versus that tested without a tartaric acid additive. Decent capacities of 374 mAh g -1 were delivered for cells at the current rate of 1C, which was majorly (35%) higher than previously reported articles 2 . Figure 4D also shows the Zn/MnO 2 battery with the tartaric acid as an additive that revealed an excellent rate capability, achieving high capacities of 327 and 270 mAh g −1 at 2C and 5C, respectively. The Zn/γ-MnO 2 battery with tartaric acid as additive, therefore, exhibited excellent long-cycle stability, with high-capacity retention. Figure 4C indicated that the adopted additive used in this system was very promising for approaching high-performance ZIBs. This phenomenon caused to have an average operating voltage at around + 1.46 V (vs. Zn) and reversible capacity up to 340.0 mAh g −1 (MnO 2 ) at C/5 (during a 5-h discharge) in the initial two cycles. Compared to other reports 2 , it was considered as a remarkable improvement. The CV curves reveal two peaks at around 1.30 and 1.60 V vs. Zn 2+ /Zn. The two consistent peaks at 1.3 and 1.4 V vs. Zn 2+ /Zn showed the Zn-insertion into the γ-MnO 2 host and the consequent reduction of Mn (IV) to the Mn (III)/Mn (II) states, which agrees with Ref. [14][15][16][17][18][19][20] . Similarly, the appearance of a peak and a shoulder at around + 1.60 and + 1.70 V vs. Zn 2+ /Zn, respectively, corresponded to Zn-extraction from the γ-MnO 2 cathode as the Mn (III)/Mn (II) states undergo oxidation to the Mn (IV) state. This evidence was probably pointed to www.nature.com/scientificreports/ the proposed behavior (a mechanism) of the adopted additive during the promotion of the efficiency of the modified ZIB. According to the other previously reported and published references [14][15][16][17][18] , partially, all types of MnO 2 often fade their capacity sharply, and it is a traditional phenomenon in many aqueous rechargeable MnO 2 batteries. According to the obtained results in Fig. 4B, it seems that in the absence of tartaric acid, Mn 2+ is dissolved, and the γ-MnO 2 structure was changed. On the other hand, the existence of tartaric acid can suppress this process and fix the Mn 2+ in the γ-MnO 2 , and consequently, the interface of the cathode convert to quasi changeless phase. However, to further evaluate the ZIB's behavior before and after introduction with the tartaric acid with 1.0 mol L −1 concentration, the correlation between the logarithm of the electrical current and the potential window (between 1.0 and 2.0 V, vs. the Zn) was plotted based on the applied potential at high possible potential range using the Tafel equation (Fig. 5). Correlation between electrical current and the electrical potential, before (Fig. 5A) and after (Fig. 5B) introduction of tartaric acid again indicated that the adopted additive used in this system was very promising for high-performance and capacity retention. As a result, capacity retention (Fig. 5C) of the battery after 7200 cycles at a rate of 5C was estimated to be 91.0%. Based on the slope, intercept of the linear correlation, net current (i 0 ) as well as transfer coefficient (α) of the electrode system in the presence of the electrolyte additive, were found as 5.62 × 10 -5 A and 0.56, respectively, Whereas, these values were found as 3.57 × 10 -5 A and 0.27, respectively, in the absence of the electrolyte additive, under similar conditions. Compared to the values estimated in the absence of any additives, these results again exhibited the influential role of tartaric acid for lowering the over-potential of the electrochemical reaction during the promotion of the capability, capacity as well as net electrical potential of the modified ZIB. Enhancement effect of the i 0 (5.62 × 10 -5 A) and α (0.56) also revealed the promotion of the efficiency of the ZIB during the additive introduction. It seems that tartaric acid, by increasing in ionic conductivity of battery electrolyte, leads to the enhancement of electrical current in contrast to the absence of tartaric acid. On the other hand, it can be concluded that tartaric acid can play a vital role in decreasing transfer coefficient (α) for charge transfer and intercalation of zinc ions into γ-MnO 2 layer structure. It seems that tartaric acid can be adsorbed onto the surface of the electrode and prepare a suitable energy active site for intercalation and de-intercalation for ions. Also, it should be noted that MnO 2 is a ceramic Nanoparticle that cannot have a chemical reaction in an acidic medium. However, it should be noted that, due to the obstacle of infrared or Raman spectroscopy in the observation of physical interaction, no significant changes were observed in the infrared and/or Raman spectra ea of tartaric acid. On the other hand, MnO 2 is well known as an inorganic ceramic. In addition, according to much literatures, MnO 2 cannot has any chemical reaction(s). So, except changing in the electrode surface voltage as illustrated in Fig. 5A,B as well as the BET test, it was impossible for us to express some other proofs. On the Proposed mechanism. Since γ-MnO 2 is a ceramic and the ceramics are chemically and electrochemically less reactive 34 , compared to other types of materials, as well as based on the results obtained from Tafel plot, electrochemical impedance spectroscopy, and capacity retention tests, some experiments were used to confirm (propose) the following probable mechanism. At first, BET analysis was performed, and various electrolyte additives were used (1.0 mol L −1 ) under similar conditions, as clearly pointed out in Fig. 6A,B during the electrical charging steps. Tartaric acid and maleic acid showed the highest storage and adsorption isotherms on the γ-MnO 2 cathode surface. Whereas, no significant changes were observed between the storage H 2 and adsorbed N 2 during the discharging process under similar conditions. No significant change(s) was observed on the XPS spectrum and the XRD patterns (Fig. 1A) in the presence and absence of tartaric acid. Especially, the XPS, due to different reasons, such as the very thin surface thickness (depth) of MnO 2 on the cathode surface, the X-ray photoelectron spectroscopy (XPS) spectrum was very noisy and not-reproducible. Consequently, it was impossible to test the morphology of the synthesized MnO 2 by other spectroscopic methods such as XPS. However, at the initial stage, it seemed that the interaction between Zn 2+ and tartaric acid, was considered the responsible factor, but, according to the similarity of the XRD patterns and the XPS spectra in the presence and absence of tartaric acid as the electrolyte additive as well as at both charges and discharge conditions, pointed to the presence of physical interaction such as surface interaction, which was in good agreement with the BET analyses (Fig. 6). Based on the results, the sequential adsorption/desorption of tartaric acid on the surface cathode, ultimately on the polar functional groups like hydroxyl (−OH) and carboxylic acid (−COOH), resulted in the significant enhancement of the double-layer capacitance, while providing a suitable active site for the perching the Zn 2+ for the intercalation and electrochemical reduction of MnO 2 . These results showed the reversible and smart adsorption/desorption of tartaric acid on the surface of the modified cathode. Therefore, it can be concluded that, these acids were adsorbed on the γ-MnO 2 edges in reversible Langmuir isotherm mode. Consequently, it is the cause of extended retention capacity, and make a suitable site in terms of energy and surface area. It should be mentioned that, in addition to the BET test, EIS, linear polarization (Tafel plot) as well as CV were performed, which can be used as an acceptable proof (proposed behavior) for this claim. According to the EIS spectrum, the charge transfer of cathode almost was modified as large as around tenfold. On the other hand, enhancement in double layer capacitance after tartaric acid addition resulted in accessing to a significant value as large as ~ 500-fold. This significant change is considered as another proof for indirectly demonstrating the reversible adsorption of tartaric acid to the cathode interface. In addition to this phenomenon, the transfer coefficient of MnO 2 as a unique criterion of charge transfer in the interface of electrode/electrolyte was increased approximately twofold (0.56, 0.27 in the presence and absence of tartaric acid, respectively). All these obtained results affirmed the physisorption of tartaric acid onto the electrode surface in this report. Conclusions We fabricated a reversible aqueous Zn/MnO 2 battery. The modified γ-MnO 2 cathode was highly reversible and stable in a ZnSO 4 aqueous electrolyte during using tartaric acid as the electrolyte additive that played the role in reducing the over-voltage. This process, therefore exhibited an excellent capacity as large as 374 mAh g −1 , rate capability, and high-capacity retention of 91.0% after 7200 cycles. Some comparisons of electrochemical performances between this research and previously reported researches are based on the Zn/MnO 2 battery (Table 1). Based on this comparison, it was concluded that, we introduced novel types of electrolyte additives to modify the Zn/MnO 2 battery, which efficiently improved the electrochemical behavior of the battery and brought a longer battery lifetime, significantly. The motivation behind this study results in creating a more efficient, reliable, environmentally friendly, low cost, high enough electrical energy, higher power densities, more improved reversibility, and longer cycle life, compared to the general ZIBs.	2022-08-05T06:17:39.690Z	2022-08-03T00:00:00.000	{ "year": 2022, "sha1": "e85bdccf3a7d68ce1b59f0655d59a18b921a157a", "oa_license": "CCBY", "oa_url": "https://www.nature.com/articles/s41598-022-13897-5.pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "f3c86937235f276296ec74ed0a52e0f36ab0c127", "s2fieldsofstudy": [ "Materials Science", "Chemistry", "Engineering" ], "extfieldsofstudy": [ "Medicine" ] }
219424607	pes2o/s2orc	v3-fos-license	National Examination as a Discourse: Perspective on Socio-Anthropology of Education This research is intended to examine the reasons underlying the variety of discourse in shaping social perceptions of the importance of the implementation of national examinations which are applied as national education standards. The study was conducted through the sociology approach to Education, making it possible to make observations into social situations related to the development of Education. This study found that, the components of the national examination became a confusion in its implementation. This applies because of several factors that underlie the emergence of the national exam quality policy namely, first; Educational competency that contains a variety of material, values, etc. seems to be ignored in the implementation of the National Examination because in practice it contains some material that is grouped as tested material, secondly, there is a standard graduation rate that must be pursued in order to pass without disregarding ethics and morals that were conceived in the nature of national education. INTRODUCTION Simply put, the study of educational socio-anthropology is related to scientific analysis of the process of social interaction associated with educational activities, both in the family, the socio-cultural life of the community and at the objective level of social relations that compiles the total construction of the reality of education in a country [9]. Up to this understanding all forms of sociological insight and knowledge in order to dissect the body of national education need to be discussed so that the learning processes take place remain relevant to the needs of the nation [5]. Here, the criteria for schools as formal educational institutions provide new ideas in the implementation of a system that contains a policy of national examinations held at the end of each semester at the end of the level or at the end of the school year. [8] Furthermore, the dimensions of the national examination are included as a forum for evaluating the learning process which is grouped into several subjects that we know as the Natural Sciences, Social Sciences, and Languages group [12]. This gives an understanding of the differences in knowledge in which capacities are narrower and which is superior. If attention is focused on understanding the form of the knowledge group, then attention and learning at school will only be of direct interest in the main topics of discussion that include groups of Natural Sciences, Social Sciences and Languages [14]. Nevertheless, the idea in the national examination needs to be reviewed in its implementation considering the energy given in the learning process contains different competencies in each field [16]. The word national examination or abbreviated as National Examination is a system of evaluating basic and secondary education standards nationally and the equality of the quality of local education level carried out by the education assessment center [3], transparent and systematic to get satisfying results. This is in line with research conducted by Nurul Hidayah, that the exam that was intended to measure graduation standards with the achievement of the target value that has been set, in fact is not actually able to increase and motivate graduation but rather as a venue for various interests alone [12]. Therefore, the study of professional education sociology explains that a school must have a clear output, namely, first graduation of students in the education process, the second component of the school as a system in the implementation of education in schools [6]. In addition, the factors that influence the results of the national examination include internal factors namely, firstly related to the physical and spiritual conditions of students which include physiological aspects and psychological aspects, secondly external factors related to environmental conditions around students which include social environment and non-social environment [ 6] The phenomenon of the national examination in the city of Baubau in practice causes some imbalances, the expected process of education as a mouth of science seems to be broken with various ceremonial activities in the national examination process. For this reason, this study discusses "the underlying reasons for the variety of discourses in shaping social perception as a national examination standard". RESEARCH METHODS The study was conducted with a descriptive qualitative model, in which the collection of sample data was determined through a purposive sample through selected respondents, which was then conducted by conducting indepth interviews with evaluations of the implementation of national examinations where the potential for public duping and planned fraud [7] In addition, the sample is assisted through the snowball sampling method to enable more facts to be obtained in the field [2]. The location of this research is in the city area of Baubau which is focused on high school-equivalent schools. RESULTS AND DISCUSSION National exams become a new problem in the world of education [4]. The existence of quality standards at the education level both at central and regional levels, so far the problem of the UN has become a special scourge that must be faced by students not on the issue of standards or learning competency but the problem of addressing the UN in the scientific family. In this case the competence of science is different, each element as well as a matter of talent and interest, let alone involving national standards with quality assessment and this is a different matter. Scientific competence seems to be the main assessment to determine whether students pass or not. [11] Though the education program is based on the general objectives of teaching which are derived from three sources: the community, students, and fields of study. Includes broad concepts such as forming humans, making human development, human personality, human responsibility, and so on. This general goal involves philosophical and ethical considerations derived from the expectations of society such as what is contained in the philosophy of the nation, the goals of national education, the nature of educational institutions, religious values, ideology and so on [10]. Internal Factors Another analysis also revealed that the basic educational goals of each school were to educate children to become useful members of the community. However, education at school is often less relevant to people's lives. The curriculum is mostly centered on fields of study that are arranged logically and systematically and whether or not know its relationship with the daily lives of students. What learners learn seems to only fulfil the interests of the school for exams, not to help the totality of students to live more effectively in society [15]. From this it appears that the difference in the dimensions of national education was conceived through the development of ideas in order to create conditions that contain the quality of education so that quality in the lives of students in the community is not seen from the success of the national exam with the maximum value standard set by the government as the success of the education process through the national exam. If this happens then education is valued as a ceremonial activity measured by numbers and carried out annually as a standard for achieving the quality of education in Indonesia. This picture shows our immaturity views the quality of education is only limited to tests and numbers, educational achievement today must be seen into several aspects that influence one of them related to the learning model applied to the learning process in schools. Therefore, the manipulative UN system is certainly not in favour of any school because it is not realistic. In fact, it is realized or not the weakness of the implementation of the UN has given impetus to the growth of immoral acts in the world of education, including forced and helpless actions by educators or teachers where in part they have helped answer the statement items on the questions. Educators or teachers are busy while students are relaxed. Even this dedication is a process of duping that makes students become lazy and not independent. In fact, the quality of national education is not always measured by numbers, but is prioritized by the importance of the quality of personality, morals, character, and behaviour as expected in the concept of national education goals, which is to create Indonesian people who have knowledge and character and also as a concept of ideals sublime development of national education. The UN is expected to encourage various elements to uphold and realize the noble ideals of national education. Evaluation of the implementation of the UN which tends to have the potential to carry out "public dupes" and commit planned fraud even though it is very dangerous for the future generation of the nation, should be immediately abandoned. Students will be smart, personality, independent, skilled, moral, and virtuous only if the educational climate allows students to gain knowledge that promotes honesty, discipline, hard work, tolerance, and transparency in various relationships in society. Between knowledge (general and religion) must relate to students' expected behaviour. So that our form of education does not become mere discourse. External Factors The idea of national exams carried out today brings social inequality for students and the school [13], where the role of schools as educators seems to be set aside and only pursue high graduation targets to meet graduation standards set by the government and measured as mere quoata. This means that, in principle, this national examination standard is only used as a standard or a prerequisite for the success of education in an area that is measured in a respectful manner regardless of the methods adopted by educational practitioners in this case schools and their learning devices. This condition, becomes an overlap between the concept of education and the implementation of national examinations which actually leads to the complexity of understanding the concept of education itself, educational products are humanizing humans in various aspects of social life. Estuary education today has no role at all in the humanizing aspect of human character so that the dominance of education is only seen how high the graduation rate obtained by schools and government quantitatively regardless of the development and direction of education of students in shaping their abilities. The point is that the implementation of education does not provide a concrete picture of the future that will be reached after graduating from high school-equivalent and how the community's needs and locality potential in an area. The role of education is currently limited to the academic routine that is undertaken as a must in everyday educational activities, school as a place to study is made a shield to manage learning devices which ends only to pursue high Advances in Social Science, Education and Humanities Research, volume 436 graduation rates so that the good name of the school is not tarnished. This condition is a picture that the implementation of the national exam does bring various imbalances in our world of education. The education dimension does not provide a solution to the potential of students and the potential of the region as a basis for the development of our education and its relationship with aspects of the development of students themselves. Based on the description above, it can be seen that the description of the implementation of the national exam in the city of Baubau brings various academic and social inequalities so that it becomes a ceremonial activity carried out by the school as an annual routine and results in erratic social upheavals in shaping the nation's character. This condition, becomes very lame in determining the content of learning that is oriented only to the group of knowledge being examined without seeing the various academic potential in students. CONCLUSION The implementation of national exams in the city of Baubau became various targets in the form of: Classification of knowledge groups, only trapped in tests and figures, forming a-moral actions, public fooling, destruction of the younger generation and only a mere discourse in the social world.	2020-05-21T00:10:48.414Z	2020-05-04T00:00:00.000	{ "year": 2020, "sha1": "74be42ef154bd3afb316daf13ff99d0f0aa2c55e", "oa_license": "CCBYNC", "oa_url": "https://www.atlantis-press.com/article/125939552.pdf", "oa_status": "GOLD", "pdf_src": "Adhoc", "pdf_hash": "9ffe9ba079204062dc965bb71229a376af4d5cd6", "s2fieldsofstudy": [ "Education", "Sociology" ], "extfieldsofstudy": [ "Sociology" ] }
245207558	pes2o/s2orc	v3-fos-license	Functional Profile of Older Adults Hospitalized in Convalescence Units of the National Network of Integrated Continuous Care of Portugal: A Longitudinal Study Aim: To evaluate the evolution of the functional profile of older adults admitted to a health unit in Portugal; to relate the functional profile of these individuals with age, sex, education level and emotional state; and to evaluate the probability of the degree of dependence as a function of age and sex. Methods: longitudinal, retrospective study with a sample of 59,013 older adults admitted to convalescence units of the National Network of Integrated Continuous Care of Portugal. Results: In the first 75 days of hospitalization, activities of daily living, mobility and cognitive state improved, but there was a decline after 75 days of hospitalization. The ability to perform instrumental activities of daily living improved in the first 15 days of hospitalization, stabilized until 45 days and then began to worsen. Women had a higher probability of having a severe/complete dependence three years earlier than men (88 years to 91 years). A higher education level and stable emotional state were protective factors against functional decline. Conclusions: The functional profile of older adults improved during the length of stay recommended for hospitalization in convalescence units (30 days). It is critical for health systems to adopt strategies to prevent declines in the emotional state of frail individuals. Introduction With the increasing aging of the population and the increase in life expectancy, it is necessary to pay greater attention to the health of older adults. The European Pathway Association states that for care pathways to be successful, they must obey a set of principles, among which the following stand out: (a) the definition of clear care goals, based on scientific evidence, the best clinical practices and the expectations and characteristics of the person being cared for; (b) the facilitation of communication among all those involved; (c) effective coordination; (d) the correct monitoring and evaluation of results; and (e) the identification of the appropriate resources for the individual and the clinical situation [1]. In this sense, a care pathway should be based on both the integration of care and the continuity of care [2]. The integration of care is considered by the World Health Organization (WHO) to be the result of multifaceted efforts made to promote integration, with benefits for people [3]. It has been considered an international priority in health policy and health management research [4]. Recent data confirm that integrated care models have had benefits in improving the health-related quality of life and functionality of people with multimorbidity and frailty [5]; in reducing hospitalization and readmission rates [6]; in reducing polypharmacy [7]; and in improving patient satisfaction, perceived quality of care and access to services [8]. In Portugal, the National Network of Integrated Continuous Care (RNCCI, acronym in Portuguese) was created in 2006 through Decree Law no. 101/2006, which is intended for people who, regardless of age, are functionally dependent. The RNCCI was developed as an integrated Health and Social Security response that mobilizes the public, private and social sectors. Its intervention objectives are the rehabilitation, readaptation and reintegration of frail individuals who no longer require acute hospital care [9]. The RNCCI is focused on community outreach services and includes hospitals, health centers, district and local social security services, the Solidarity Network and local authorities. There are several types of care provided in the RNCCI, which includes inpatient units (convalescence units; mediumterm and rehabilitation units; long-term and maintenance units; and level 1 pediatric integrated inpatient units) and outpatient units (mental health integrated continuous care units; day and autonomy promotion units; pediatric outpatient units and integrated continuous care teams) [9]. Convalescence Units (CUs) are intended for individuals with a potentially recoverable transitory loss of autonomy, and their purpose is clinical stabilization and functional rehabilitation. These units have their own facilities and are connected to an acute care hospital to ensure health care 24 h a day over a total length of hospitalization of 30 days [9]. As the purpose of CUs is functional rehabilitation, it seems clear to us the importance of assessing the functionality of patients to better evaluate health outcomes. The WHO developed the International Classification of Functioning, Disability and Health (ICF) to standardize the international assessment of functioning and disabilities related to the healthdisease process, taking into account the body's structures and functions and environmental factors [10]. Studies indicate that the evaluation of functionality is crucial in care models for older adults [11,12]. The performance of self-care behavior is considered extremely important for the individual person and for the health system, due to the benefits associated with it [13]. The gains from the development of self-care health behavior are related to reduced risk of complications, healthcare expenditures, hospital readmission rates, increased satisfaction with care, feelings of responsibility, control, independence, and autonomy, adoption of effective coping strategies, improved well-being, functional capacity, quality of life, symptom control, and pain [14][15][16]. In the care process, we seek to assess dependence in self-care skills (mobility, basic and instrumental activities of daily living, and cognitive status) [17,18], with the purpose of maintaining life, healthy functioning and personal development. The self-care deficit presents itself in degradé, as it can fluctuate in different levels, from mild to complete/severe [15]. The WHO report for the 2021-2030 decade emphasizes the importance of implementing actions that improve the functional ability of older adults, presenting four areas of action for this purpose: a) changing the way we think, feel and act toward age and aging; b) ensuring that communities foster the abilities of older people; c) providing integrated person-centered care and services that meet the needs of older people; and d) providing access to long-term care for older people who need it [19]. One of the strategies defined in the same report to accelerate the implementation of functional ability is to strengthen data, research and innovation [19]. Thus, the aim of the present study is to assess the functional trajectory of older adults hospitalized in CUs of the RNCCI of Portugal. For this purpose, the following objectives were defined: (1) to evaluate the evolution of the functional profile of older adults hospitalized in CUs; (2) to relate the functional profile of these individuals with age, sex, education level and emotional state; and (3) to evaluate the probability of the degree of dependence as a function of age and sex. Study Type and Sample This was a longitudinal, retrospective study with a sample of 59,013 older adults aged 65 or older hospitalized in health units belonging to CUs of the RNCCI of Portugal. Instrument To evaluate the functional profile, variables of the International Classification of Functioning, Disability and Health (ICF), which contains the ICF components: Body Functions (Mental Functions) and Activities and Participation (Mobility, Self-care, Communication, Domestic Life, Main life areas), were used. The ICF items were transformed into a Likert scale so that they could be analyzed (no problem = 1; mild or moderate problem = 2; severe problem = 3; complete problem = 4). The global Cronbach's α = 0.951 is obtained, which means excellent internal consistency. Data Collection The data were collected from the records made by health workers and entered in the RNCCI portal. To trace the sociodemographic profile, data related to the first evaluation of each hospitalization episode occurring between 1 January 2010, and 27 February 2017, were selected. Subsequently, the evolution of the functional profile was evaluated through the analysis of biweekly evaluations of older adults hospitalized in CUs who were targets of a set of structured professional interventions (objective 1). To meet the second objective, which was to relate the functional profile with the variables age, sex, education level and emotional state, principal component analysis of each of the instrument domains (mobility, ADL, IADL, and cognitive state), was combined with cluster analysis. This procedure presented the advantage of reducing the number of input variables in the cluster analysis, thus helping to simplify the characterization of the upstream clusters [20]. To meet the third objective, the probabilities of each class as a function of age and by sex were evaluated. Statistical Analysis To analyze the four components of self-care capacity, over the days of hospitalization, a longitudinal analysis was performed, based on parametric tests (One-way ANOVA and t-Student test) (objective 1). The exploratory analysis of clusters was performed using the hierarchical method (Analyze Classify Hierarchical Cluster). Since it is big data, it was necessary to perform a random partition of the database, to create a sub-sample with approximately 20% of the data, so that it was possible to process the information by SPSS. When obtaining the agglomeration coefficients (Ward's method), a graphic projection was performed, of the highest (last 30), to visualize their distances, where it was possible to verify that the best solution resided in the retention of 3 clusters. Consecutively, the cluster analysis was carried out using the non-hierarchical optimization method available in IBM SPSS: K-means (objetive 2). Ordinal regression with the probit link function was used to assess whether age and sex had a significant effect on the probabilities related to the type of dependence (objective 3). The link function was chosen based on the frequency distribution criteria of the classes of the dependent variable "degree of dependence" [21]. Ethical Procedures The study was conducted in accordance with the guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of Scientific Research in the Areas of Human Health and Welfare of the University of Évora (report number, 17036; date of approval, 26 April 2017). Sociodemographic and Clinical Characteristics The mean age of the sample was 78.93 years (SD = 7.28), with an age range of 65 to 109 years. The majority of older adults admitted to CUs were in the age group of 75 to 84 years (47.2%), followed by 65 to 74 years (29.0%), and last, people aged 85 or older (23.8%). Most of the sample was female (61.5%), had a partner (44.7%), had less than 6 years of schooling (60.4%), and had an unskilled professional level at working age (70.3%). Table 1 provides the sociodemographic characterization of the sample. In general terms, the three clusters were quite distinct, with the following configuration, presented in Figure 2: (1) Cluster 1: Older adults with a higher degree of dependence (severe/complete selfcare deficit); (2) Cluster 2: Older adults with an intermediate degree of dependence (moderate selfcare deficit); (3) Cluster 3: Older adults with a lower degree of dependence (mild self-care deficit). Dependence Clusters The nonhierarchical exploratory method of cluster grouping was used, yielding the following partition: Cluster 1-29.6% (n = 11,248); Cluster 2-52.1% (n = 19,785); and In general terms, the three clusters were quite distinct, with the following configuration, presented in Figure 2: (1) Cluster 1: Older adults with a higher degree of dependence (severe/complete selfcare deficit); (2) Cluster 2: Older adults with an intermediate degree of dependence (moderate selfcare deficit); (3) Cluster 3: Older adults with a lower degree of dependence (mild self-care deficit). Figure 3 shows the differences among the three clusters and the variables sex, age group, education, sad/depressed emotional state and anxious emotional state. (1) Cluster 1 (severe/complete dependence) is composed of a higher percentage of males, aged 85 years or older, older adults who did not attend school and who have been feeling depressed and anxious for a long time; (2) Cluster 2 (moderate dependence) encompasses a greater percentage of females, aged between 65 and 84 years, with 1 to 6 years of education and who have felt sad or anxious for a short time; (3) Cluster 3 (mild dependence) is predominantly composed of males aged 65 to 74 years, with more years of schooling (7 or more) and who feel depressed or anxious for a short time. J. Pers. Med. 2021, 11, x FOR PEER REVIEW 6 of 12 Figure 3 shows the differences among the three clusters and the variables sex, age group, education, sad/depressed emotional state and anxious emotional state. (1) Cluster 1 (severe/complete dependence) is composed of a higher percentage of males, aged 85 years or older, older adults who did not attend school and who have been feeling depressed and anxious for a long time; (2) Cluster 2 (moderate dependence) encompasses a greater percentage of females, aged between 65 and 84 years, with 1 to 6 years of education and who have felt sad or anxious for a short time; (3) Cluster 3 (mild dependence) is predominantly composed of males aged 65 to 74 years, with more years of schooling (7 or more) and who feel depressed or anxious for a short time. Degree of Dependence as a Function of Age and Sex The assumption of the slope homogeneity model was validated ( Degree of Dependence as a Function of Age and Sex The assumption of the slope homogeneity model was validated (χ 2 (2) = 4.531; p = 0.104). The model was considered highly significant (χ 2 (2) = 274.822; p < 0.001), although the effect size was small (R 2 MF = 0.096; R 2 N = 0.132; R 2 CS = 0.129). In the ordinal regression model, the link function "Probit" was adopted ( N (0, 1)), because this function is recommended when the latent variable presents a normal distribution. The coefficients and statistical significance of the adjusted ordinal model are shown in Table 2. The results obtained suggest that with advancing age, the probability of observing higher-order classes, i.e., the probability of observing a higher degree of dependence, increases (b Age = 0.043; p < 0.001). Regarding sex, the results obtained suggest a higher probability of observing a higher dependence type in women than in men (b Sex(Female) = 0.101; p < 0.001); however, this difference is moderate. The evaluation of the probabilities of each of the classes as a function of age and by sex is shown in Figure 4. The analysis revealed that a) for both women and men, the probability of a dependence profile of "mild" is always inferior to any of the other two profile types; b) for women, starting at 88 years of age, the dependence profile that most likely occurs is "severe/complete"; for men, the same occurs, but starting at 91 years of age (three years later); c) by year of age, the ratio of the probability of observing profiles of lower dependence, compared to the probability of observing profiles of greater dependence, decreases by 4.2% (1-exp(-0.043)); and d) the odds ratio of lower dependence (mild) relative to higher dependence (severe/complete) decreases by 9.6% (1-exp(-0.101)) from males to females. types; b) for women, starting at 88 years of age, the dependence profile that most likely occurs is "severe/complete"; for men, the same occurs, but starting at 91 years of age (three years later); c) by year of age, the ratio of the probability of observing profiles of lower dependence, compared to the probability of observing profiles of greater dependence, decreases by 4.2% (1-exp(-0.043)); and d) the odds ratio of lower dependence (mild) relative to higher dependence (severe/complete) decreases by 9.6% (1-exp(-0.101)) from males to females. Discussion This study evaluated the evolution of the functional profile of older adults hospitalized in CUs of the RNCCI, related it to the variables discussed below and evaluated the probability of the degree of dependence as a function of age and sex. Regarding the evolution of self-care components throughout hospitalization, in the first 75 days of hospitalization, ADLs, mobility and cognitive state improved, but there was a decline after 75 days of hospitalization. In contrast, the ability to perform IADLs improved in the first 15 days of hospitalization, stabilized up to 45 days, and worsened thereafter. Considering that IADLs include tasks such as preparing meals, washing clothes, and housework, activities that are not performed by patients during hospitalization, it would be expected that with the lack of practice, the ability to perform these tasks Discussion This study evaluated the evolution of the functional profile of older adults hospitalized in CUs of the RNCCI, related it to the variables discussed below and evaluated the probability of the degree of dependence as a function of age and sex. Regarding the evolution of self-care components throughout hospitalization, in the first 75 days of hospitalization, ADLs, mobility and cognitive state improved, but there was a decline after 75 days of hospitalization. In contrast, the ability to perform IADLs improved in the first 15 days of hospitalization, stabilized up to 45 days, and worsened thereafter. Considering that IADLs include tasks such as preparing meals, washing clothes, and housework, activities that are not performed by patients during hospitalization, it would be expected that with the lack of practice, the ability to perform these tasks would begin to decrease earlier than would ADLs, mobility or cognitive state because the tasks related to these components are maintained throughout hospitalization. However, at 75 days of hospitalization, the ability to perform these activities also begins to decline. One study reports that 30 to 60% of older adults experience a functional decline in ADL during hospitalization [22]. Low mobility during hospitalization and functional and physical changes can lead to functional deficits in ADLs and cognitive impairments [23]. However, 80% of cases are preventable with effective health care [24,25]. Importantly, individuals who are hospitalized for more than 30 days in CUs are those with less potential for rehabilitation or without family support to provide continuity of rehabilitation at home; the length of hospitalization recommended for individuals referred to CUs is 30 days, which may explain these results. Thus, it is concluded that most individuals stay for the recommended time and experience reductions in acquired deficits in all dimensions. Regarding age, as would be expected, most individuals older than 85 years are in the severe/complete dependence cluster, the age group of 65 and 84 years is mostly in the moderate dependence cluster, and most individuals in the age group of 65 to 74 years old are in the mild dependence cluster. Other authors have also concluded that the older the age, the worse the functionality [26]. When analyzing the clusters by sex, we found that there was a higher percentage of males in the severe/complete dependence cluster and in the mild dependence cluster and a higher percentage of females in the moderate dependence cluster. However, when analyzing the probability of the degree of dependence as a function of age and sex, for women, starting at 88 years of age, the type of dependence profile that most likely occurs is "severe/complete", while for men, the same occurs only after 91 years of age (three years later). This finding should be interpreted with caution because studies are conflicting, as some report that women have a higher degree of dependence than do men and others report that there are no differences. A literature review revealed that the incidence of functional disability was identical between sexes [27]. However, the results from two Portuguese studies with older adults indicated that the functional profile is worse in women than in men [26], regardless of age [28]. Another study that analyzed data from Spain, Portugal and Italy found that compared with men, women after 65-70 years of age had a higher risk of suffering from severe functional limitations. The same study added that in the age group of 75-80 years of age, women were 3.3% more likely than men to have severe functional limitations and that in the age group of 80 years or older, this probability increased to 15.5% [29]. The results from another study indicated that there are significant differences between sexes regarding the probability of occurrence of disability, with a higher probability for women in the following countries: United States, Korea, Southern Europe, Mexico and China. However, these differences do not exist in Northern, Central and Eastern Europe, England, and Israel. The authors conclude that gender inequality in society at the macro level is significantly associated with the probability of women developing disabilities [30]. A European study concluded that there are differences between sexes in relation to some important health indicators and that these differences are higher in Southern and Eastern Europe than in Western and Northern Europe. The same study warns that the presence of a sex difference in health cannot be considered a universal factor because, depending on the indicator and the country, the difference tended to increase, decrease or even reverse with age [31]. It should be noted that neither of the two previous studies included Portugal. With regard to education, several studies are consistent in indicating that the lower the educational level, the worse the functional profile and the greater the degree of dependence [26,28,32] and that a lower education level is associated with worse physical and mental health outcomes [33][34][35]. In addition, a study showed that the impact of multimorbidity on ADLs was three times higher at the lowest education level than at the highest education level [36]. Another study added that after 65 years of age, the average probability of severe functional limitations for individuals with a low education level increased to more than 40% at age 80 years and older, while it was 26% in the higher education category [29]. The results of the present study are in line with those described in the literature, with a predominance of individuals who did not attend school in the severe/complete dependence cluster; those who studied for 1 to 6 years were predominant in the moderate dependence cluster, and those with a higher education level (7 or more years) were mostly in the mild dependence cluster. Thus, it is confirmed that in individuals admitted to CUs, the higher the educational level is, the lower the likelihood of dependence; additionally, education is extremely important for health and for active and healthy aging. Regarding the analysis of emotional state, in the severe/complete dependence cluster, there was a predominance of individuals who felt depressed and anxious for a long time, and in the moderate dependence and mild dependence clusters, there was a predominance of older adults who felt sad or anxious for a short time. In fact, in a literature review, depression was considered one of the main risk factors for functional disability in older adults [27]. Another recent study that analyzed several combinations of multimorbidity concluded that those that included depression were the only consistent predictors of disability [37]. Another study followed the same direction, with the results indicating that depression and/or cognitive deficit was associated with a substantially greater potential disability than combinations composed exclusively of somatic diseases [38]. The strengths of the present study include the relatively large sample size and the heterogeneity of the participants. The limitations of the present study include the lack of an analysis of the time period after 2017 and the inclusion of only some of the determinants identified in the literature as influencing functional ability (e.g., emotional state). Conclusions The recommended length of hospitalization in CUs of the RNCCI of Portugal is 30 days, although some individuals remain longer. During this period, there was a significant improvement in the functional profile of older adults hospitalized in these units, and there was a decline in those who remained beyond 75 days. More studies are needed to understand the reason for the observed decline after this cutoff point. Women hospitalized in CUs were more likely to have a severe degree of/complete dependence three years earlier than men (88 years compared to 91 years of age, respectively). The higher the education level and the better the emotional state was, the lower the degree of dependence. This study confirms the importance of the education level and emotional state in functional ability in older adults. It is necessary that future studies evaluate the effectiveness of interventions that seek to prevent declines in the emotional state of individuals so that they are applied in clinical practice when a situation occurs that threatens the independence of such individuals, as is the case of people referred to CUs. In addition, the importance of strategies that promote literacy in the population is reinforced. Conflicts of Interest: The authors declare no conflict of interest.	2021-12-16T17:11:53.222Z	2021-12-01T00:00:00.000	{ "year": 2021, "sha1": "33882015fb53a606f695c30e20b2c82d5e2d91ad", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2075-4426/11/12/1350/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "8e5388e8a2af88467d9f6c4c3a4b3e97b06324d9", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
212532002	pes2o/s2orc	v3-fos-license	Optimal Inventory Strategies for Pharmaceutical Products Incorporating Carbon Emissions Health sector is a significant source of carbon emissions around the world. Today healthcare industries are looking for solutions to reduce carbon emissions associated with their operations. Operational adjustments, such as modifications in inventory strategies, have proven to be an effective way to decrease emissions. This paper offers optimal inventory strategies for healthcare system to achieve target customer service level while minimizing system total cost and carbon emissions. Introduction A healthcare supply chain (HSC) consists of pharmaceutical wholesalers, pharmaceutical companies, hospitals, medical surgical distributors, etc. A stated one-third of hospital decision makers have belief their hospital supply chain is working at greatest point doing work well, but the other two thirds are discovering themselves one foot in fee-for-service and the other in value-based giving back of money. Many of these hospitals are not implementing the suitable supply chain management strategies. This is because most hospital administrators and pharmacy managers are doctors with expert knowledge in medicine and are not supply chain professionals. Unlike other industries, hospital administrators and pharmacy managers have to manage very complicated distribution networks and inventory management problems without proper guidance on efficient practices. Hence, given the high costs, constraints, and perishability of pharmaceuticals, more study is necessary to help health care managers in setting optimal HSC strategies. Recently few numbers of papers designed in the field of healthcare supply chain and inventory management (see [1][2][3][4]). All steps from the supply of raw materials to the finished products can be covered in a supply chain connecting raw material suppliers, manufacturers, retailers, and the customer/ hospital. Multi-echelon inventory management is the management of inventory and coordination of the distribution process at more than one level of a supply chain network. Kim [5] presented an explanation of an integrated supply chain management system developed to specifically address issues related to pharmaceuticals in the health care sector. Meijboom and Obel [6] addressed supply chain coordination for a pharmaceutical company with a multi-location and multi-stage operations structure. Amir et al. [7] developed a generalized network oligopoly model for HSC competition that takes into account product perishability, brand differentiation, and discard costs. Lead time is an essential factor in any supply chain and inventory system. Liao & Shyu [8] devised a probabilistic inventory model in which lead time was considered as a decision variable. Several researchers extended this approach in integrated inventory models for lead time reduction (see [9][10][11][12][13][14]). Glock [13] studied alternative methods for controlling lead time and their impact on the safety stock and the expected total costs of a continuous review inventory control system. Moua et al. [14] introduced a more realistic lead time crashing cost in Glock [13]'s model. They proposed a modified integrated inventory model by adding the transportation time as a decision variable and assumed that there were two different safety stocks. They numerically showed that their modified model was more realistic than Glock [13] model. Without considering carbon emissions in inventory management, the decision objective is usually set as the total cost minimization or the total profit maximization. However, taking emission reduction into account is likely to change the optimal solution and incurs additional cost. The health sector is a vital direct source of carbon emissions. It is also a highly heterogeneous segment that has seen both rapid increases and decreases in direct emissions from different sources over the last decade. A modification in inventory strategies is an effective way to decrease emissions so that carbon emission issues in inventory management have attracted attention in literature recently (see [15][16][17][18][19]). This research explores the impact of both lead time and carbon emissions on the inventory decisions of a pharmaceutical company and hospital supply chain system. The paper assumes that the lead time for the first shipment consists of production time, setup and transportation time (nonproductive time), while the lead time for the rest of the shipments is only transportation time, and that lead time can be reduced by shortening nonproductive time and transportation time. The proposed study also adds the carbon emissions from transportation and storage into the account. The intention of our investigation is to obtain the optimal decisions (order quantity, production rate, nonproductive time, transportation time and total number of deliveries) to achieve hospital's CSL while minimizing expected total cost, and to examine the effects of carbon emissions in the proposed healthcare system. Notations and Assumptions To develop the proposed model, we adopt the following notations and assumptions which are similar to those used in Uthayakumar & Priyan [1]. I. For the Hospital Hospital's expected total cost The hospital places an order after every Q units, therefore for expected ordering cycle time of Q/D, the expected ordering cost per unit time can be expressed by . AD Q Similar to Glock [13], we assume that the nonproductive time t s consists of m mutually independent components t sr , i.e. To gain the transportation time crashing cost C(t T ) for the shipments 1,2,3,....n, Similar to the assumptions made by Mou [14], we assume that the l th component of the nonproductive time t s is the setup time. Thus, the transportation time t T is composed of m-1 mutually independent components, i.e. Let t T j be the length of the transportation time with the components 1,2,3,...j crashed to their minimum duration, then t T j can be expressed as .m-1. Hence, by adopting the same technique used by Mou [14], the transportation time crashing cost C(t T ) for the shipments 2,3,...n ( ) ( ) Thus the lead time crashing cost per unit time can be derived by In addition, the reduction of lead time leads to a lower demand uncertainty, which may decrease the safety stock and the stock-out loss. Since for the shipments 2,3,...n it might be beneficial to avoid holding additional safety stock by decreasing the reorder points. Consequently, Similar to Mou [14], we consider that there are two different safety stocks and reorder points for the first shipment and the shipments 2,3,...n. Then, the hospital's expected holding cost per unit time is The hospital's freshness index of a pharmaceutical product (perishable product) may be pretentious by time, temperature, dampness, refrigeration, etc. In practice, most pharmaceutical products have their own expiration dates. Feng et al. stated that health-conscious consumers prefer a perishable product with a longer remaining shelf life to a shorter one and they assumed that the customer's freshness index is age-dependent. That is, it begins with 1 at time 0, and then gradually degrades closer to 0 as the expiration date of the product approaches. Consequently, similar to Feng et al., we assume that the age-dependent freshness at time t is linearly decreasing from 1 at the beginning to 0 at the maximum lifetime m as follows: Subsequently the expired finished products of the hospital for the ordering cycle Q D can be detected by ε since the hospital's safety stock inventories are sufficient to replace these expired finished products. Then similar to the existing models, we consider the reorder point r DL k L σ = + where kis the safety factor. Hence, hospital's expected total cost per unit time is composed of ordering cost, holding cost, disposal cost and crashing cost, is expressed by We assume in view of realistic healthcare system that the hospital has set a target service level in terms of the fill rate corresponding to the proportion of all product demands to be satisfied directly from available stock. Therefore, the Service Level Constraint (SLC) imposes a limit on the proportion of demands that are not met from stock. This can be mathematically expressed as Pharmaceutical company's expected total cost Once the hospital orders a lot size of Q units, the company produces the products in a lot size of nQ units with the setup cost S in each production cycle of length nQ D with constant production rate P units per unit time, and the hospital will receive the supply in n lots each of size Q units. The first lot size of Q units is ready for shipment after time Q P just after the start of the production. During the production run time, Then, the company's expected holding cost per unit time Let g 0 denote fuel consumption of the unloaded (empty) vehicle and g 1 denote unit fuel consumption factor if the supplier's transportation vehicle is loaded with goods. This research considers g 0 +g 1 Q captures amount of fuel consumption for a one-way delivery from supplier to buyer. We assume that backhaul is not used and return vehicles are empty, and that g 0 is amount of the fuel consumption for the return trip. For one inventory replenishment, the total amount of fuel consumption 2g 0 +g 1 Q. Hence, supplier's average transportation cost per unit of time is In this model, we measure the total amount of carbon emissions from transportation and storage. Thus, the associated amount of carbon emissions is designed by In the above equation, Π 1 (Q,P,n) and Π 2 (Q,P,n) are the amounts of carbon emissions from transportation and storage, respectively. Then θ 1 and θ 2 are carbon emission factors for fuel and electricity. In Π 2 (Q,P,n), w 0 and w 1 are fixed and unit variable electricity consumption for storage. Accordingly, the company's carbon emissions cost per unit time is derived by Thus, the total cost function per unit time for the pharmaceutical company comprising of setup cost, holding cost, transportation cost and carbon emission cost, is expressed by Then the joint expected total cost per unit time for the proposed pharmaceutical company and hospital integrated supply chain system is Now our problem involves finding the optimal (Q,P,n,t s ,t T ) in a production cycle that minimize the joint expected total cost, JETC(Q,P,n,t s ,t T ) while satisfying SLC. That is, the primal problems of the proposed model is given as Solution Procedure The problems formulated in the previous section appear as constrained non-linear programming problems. To solve these problems, we temporarily ignore the SLC and relax the integer requirement on n, then we have to minimize joint expected total cost, JETC(Q,P,n,t s ,t T ) Proposition 1: For fixed P,n,t s ϵ[t s i , t s i-1 ] and t T ϵ[t s i , t s i-1 ], JETC(Q,P,n,t s ,t T ) is convex in Q. Proof: Taking the first and second partial derivatives of JETC(Q,P,n,t s ,t T ) with respect to Q we have ( ) 1 2 , , , , and ( ) This completes the proof of Proposition 1. , , , , Proof: Taking the second partial derivatives of JETC(Q,P,n,t s ,t T ) with respect to t s and t T respectively, we obtain the following Hessian matrix It is clear that the Hessian matrix H is negative definite. Hence, JETC(Q,P,n,t s ,t T ) is a concave function with respect to t s and t T for fixed values of Q, P, n. This completes the proof of Proposition 3. This completes the proof of Proposition 4. Remark 1: The search for the optimal number of deliveries, is reduced to find a local minimum. According to the Result 1, for fixed of P,n,t s ϵ[t s i , t s i-1 ] and t T ϵ[t s j , t s j-1 ], when the SLC is ignored, Eq. (3) gives optimal value of Q such that the joint expected total cost is minimum. Now, the SLC is taken into consideration and if for Q= Q H then Q H is the local minimum of JETC(Q,P,n,t s ,t T ) and SLC is inactive. Otherwise, optimal value of Q should be at least equal to which is greater than Q H so that specified level of service can be achieved at minimum joint expected total cost and Q SLC is the local minimum of JETC(Q,P,n,t s ,t T ). So, for fixed of P,n,t s ϵ[t s i , t s i-1 ] and t T ϵ[t s j , t s j-1 ] the optimal Q is given by max{Q H, Q SLC }. Based on the convexity and concavity behavior of objective function with respect to the decision variables the following algorithm is designed to find the optimal values of the decision variables. Algorithm i. Conclusion In this paper, we designed a new algorithm to achieve the target customer service level while minimizing system cost of the proposed healthcare supply chain system. We provided an optimal inventory policy through mathematical model to modify the operators in order to reduce carbon emissions. This research explores the impact of both lead time and carbon emissions on the inventory decisions of a pharmaceutical company and hospital supply chain system.	2020-03-07T10:55:54.829Z	2019-05-28T00:00:00.000	{ "year": 2019, "sha1": "8844374d98be5fd059611c3afaa32ccd78049ea3", "oa_license": "CCBY", "oa_url": "http://juniperpublishers.com/ctbeb/pdf/CTBEB.MS.ID.556008.pdf", "oa_status": "GOLD", "pdf_src": "MergedPDFExtraction", "pdf_hash": "8844374d98be5fd059611c3afaa32ccd78049ea3", "s2fieldsofstudy": [ "Environmental Science", "Medicine", "Engineering" ], "extfieldsofstudy": [] }
237371659	pes2o/s2orc	v3-fos-license	Seeing Implicit Neural Representations as Fourier Series Implicit Neural Representations (INR) use multilayer perceptrons to represent high-frequency functions in low-dimensional problem domains. Recently these representations achieved state-of-the-art results on tasks related to complex 3D objects and scenes. A core problem is the representation of highly detailed signals, which is tackled using networks with periodic activation functions (SIRENs) or applying Fourier mappings to the input. This work analyzes the connection between the two methods and shows that a Fourier mapped perceptron is structurally like one hidden layer SIREN. Furthermore, we identify the relationship between the previously proposed Fourier mapping and the general d-dimensional Fourier series, leading to an integer lattice mapping. Moreover, we modify a progressive training strategy to work on arbitrary Fourier mappings and show that it improves the generalization of the interpolation task. Lastly, we compare the different mappings on the image regression and novel view synthesis tasks. We confirm the previous finding that the main contributor to the mapping performance is the size of the embedding and standard deviation of its elements. Introduction INR is a novel field of research in which the traditional discrete signal representation (e.g., images as discrete grids of pixels, 3D shapes as voxel grids or meshes) are replaced with continuous functions that map the input domain of the signal (e.g., coordinates of a specific pixel in the image) to a representation of color, occupancy or density at the input location. However, these functions typically are not analytically tractable; INRs approximate those functions with fully connected neural networks (also called multi-layer perceptrons (MLPs)). INRs are not coupled to the spatial resolution (e.g., voxel Figure 1: Visualization of SIREN [29], Gauss ReLU [32] and our method (INT SIN). The top row is the first period and bottom row is the second period, which shows our method enforcing periodicity. size in a 3D scene) and theoretically have infinite resolution. Therefore, these representations are naturally suited in applications with high-dimensional signals and heavy memory consumption. Also, since they are differentiable, they are suitable for gradient-based optimization and machine learning. In addition, the application of INRs for images [10,31], volume density [19], and occupancy [17] enhanced the performance on various tasks such as shape representation [4,5,8,7,12,18,25], texture synthesis [10,22], and shape inference from images [16,15]. However, early architectures lacked accuracy in highfrequency details. Sitzmann et al. [29] proposed SIRENs, which could represent high frequencies. They argued that sinusoidal activations work better than ReLU networks because ReLU networks are piecewise linear, and their second derivative is zero. As a result, they are incapable of modeling data contained in higher-order derivatives of signals. However, a concurrent work [19] proposed positional encoding, which also enabled the networks to learn high-frequency information. The positional encoding uses a heuristic sinusoidal mapping to input coordinates before passing them through a ReLU network. They did followup work [32] exploring the general Fourier mapping and explaining why it worked using a Neural Tangent Kernel (NTK) framework [11]. They found out that the Fourier mapping transforms the NTK into a shift-invariant kernel. And modifying the mapping parameters enables tuning the NTK's spectrum, therefore controlling the range of frequencies the network can learn. They also showed that a random Fourier mapping with low standard deviation learns only low frequencies of the signal. On the contrary, a high standard deviation lets the network learn high frequencies only, which leads to over-fitting. They recommended a linear search to find the optimal value of the standard deviation for the corresponding task. They also showed that increasing the number of parameters in the mapping improves the performance constantly. However, we wonder what the difference between SIRENs and Fourier mapping is? Will the performance be saturated when we continue to increase the mapping parameters? Is there a way to avoid over-fitting when training networks using Fourier mapping? And is random Fourier mapping the optimal mapping? To answer these questions, we explored the mathematical connection between Fourier mappings and SIRENs and showed that a Fourier mapped perceptron is structurally like a one hidden layer SIREN. However, in the SIREN case, the mapping is trainable, and it is represented in the amplitudephase form instead of the sine-cosine form in the case of Fourier mappings. Also, we looked at the functions we want to learn, and we observed that they have a limited input domain (e.g., the height and width of an image), and their values are defined on a finite set. Hence, we can assume that they are continuous and periodic over their input bound, which satisfies all the requirements to represent them with a Fourier series. Furthermore, we determined the d-dimensional Fourier series's trigonometric form and showed that it is precisely a single perceptron with an integer lattice mapping applied to its inputs. The weights of that perceptron are the Fourier series coefficients. As the Fourier series can theoretically represent any periodic signal, this perceptron can represent any periodic signal if it has an infinite number of frequencies in its mapping. However, in practice, the Fourier series coefficients are finite, and we can get them by sampling the signal at the Nyquist rate (twice the bandwidth) and applying a fast Fourier transform (FFT) to the signal. Thus, the number of Fourier coefficients is the theoretical upper bound of the number of parameters needed in the mapping. Moreover, we modified the progressive training strategy of [13], where we train the lower frequencies in the initial training phase and gradually add the higher frequency components as the training progresses. As a result, we show that our Progressive Training strategy avoids the problem of over-fitting. Finally, we tested our proposed Integer Lattice mapping in the image regression and novel view synthesis tasks. We found out that the main contributor to the mapping performance is the number of parameters and the standard deviation, as was shown in [32]. In summary, we offer the following contributions: • We introduce an integer Fourier mapping and prove that a perceptron with this mapping is equivalent to a Fourier series. • We explore the mathematical connection between Fourier mappings and SIRENs and show that a Fourier mapped perceptron is structurally like a one hidden layer SIREN. • We show that the integer mapping forces periodicity of the network output. • We modify the progressive training strategy of [13] and show that it improves the generalization of the interpolation task. • We compare the different mappings on the image regression and novel view synthesis tasks and verify the previous findings of [32] that the main contributor to the mapping performance is the number of elements and standard deviation. In early architectures, there was a lack of accuracy in fine details of signals. Mildenhall et al. [19] proposed positional encodings to tackle this problem, then Tancik et al. [32] further explored positional encodings in an NTK framework, showing that mapping input coordinates to a representation close to the actual Fourier representation before passing them to the MLP lead to a good representation of the high-frequency details. Furthermore, they showed that random Fourier mappings achieved superior results than if one takes the simple positional encoding. Sitzmann et al. [29] also attempted to solve the problem of getting high-frequency details. They proposed SIRENs and demonstrated that SIRENs are suited for representing complex signals and their derivatives. In both solutions, they used a variant of Fourier neural networks (FNN) for the first layer of the MLP. FNN are neural networks that use either sine or cosine activations to get their features [14]. The first attempt to build an FNN was by [6]. They proposed a one-layer hidden neural network with a cosine squasher activation function and showed if they hand-wire certain weights, it will represent a Fourier series. Silvescu [28] proposed a network that did not resemble a standard feedforward neural network. However, they used a cosine activation function to get the features. Liu et al. [14] introduced the general form for Fourier neural networks in a feedforward manner. They also proposed a strategy to initialize the frequencies of the embedding, which helped for convergence. Our work will show another way to initialize the embedding, which results in a neural network that is precisely a Fourier series. Integer lattice mapping This section explains how a perceptron with an integer lattice Fourier mapping applied to its inputs is equivalent to a Fourier series. First, we present the Fourier mapped perceptron equation and then link it to the Fourier series's general equation. The fundamental building block of any neural network is the perceptron, and it is defined as Here y ∈ R dout is the perceptron's output, g(·) is the activation function (usually non-linear), x ∈ R din is the input, W ∈ R dout×din is the weight matrix, and b ∈ R dout is the bias vector. Now, if we let g(·) to be the identity function and apply a Fourier mapping to the input we get where γ(x) is the Fourier mapping defined as W ∈ R dout×2m , B ∈ R m×din is the Fourier mapping matrix, and m is the number of frequencies. Equation 2 is the general equation of a Fourier mapped perceptron, and we will relate it to the Fourier series's general equation. A Fourier series is a weighted sum of sines and cosines with incrementally increasing frequencies that can reconstruct any periodic function when its number of terms goes to infinity. In applications that use coordinate-based MLPs, the functions we want to learn are not periodic. However, their inputs are naturally bounded (e.g., height and width of an image). Accordingly, it doesn't harm if we assume that the input is periodic over its input's bounds to represent it as a Fourier series. We will explain later why this assumption has many advantages. A function f : where • is the Hadamard product. As it is plausible to normalize the inputs to their bounds, we assume that each variable's period is 1. The Fourier series expansion of function (4) with p = 1 d is defined by [24]: where c n are the Fourier series coefficients, and they are calculated by: For real-valued functions, it holds that c n = c * -n where c * n is the conjugate of c n . Using Euler's formula and mathematical induction we showed that equation (5) can be written as (see supplementary material): And if we write equation (7) in vector form, we get where a B = (a n ) n∈B , and b B = (b n ) n∈B . Now, if we compare 2 and 9, we find similarities. We see that , is the concatenation of all possible permutations of n. For practicality we limit B to where N will be called the frequency of the mapping, In practice, we can find the Fourier series coefficients by sampling the function uniformly with a frequency higher than the Nyquist frequency and apply a Fast Fourier Transform (FFT) on the sampled signal. The resulting FFT coefficients are the Fourier series coefficients multiplied by the number of the sampled points. And in theory, if we initialize the weights with the Fourier series coefficients, our network should give the training target at iteration 0. SIRENs and Fourier mapping comparison In this section we want show that a Fourier mapped perceptron is structurally like a SIREN with one hidden layer. If we evaluate W · γ(x) in equation (2), using (3) and combine the Sine and Cosine terms, we get: where φ := (π/2, . . . , π/2, 0, . . . , 0) T ∈ R 2m and C := (B, B) T (see supplementary material). Here we see that C is acting as the weight matrix applied to the input, φ is like the first bias vector and sin(·) is the activation function. Hence, the initial Fourier mapping can be represented by an extra initial SIREN layer, with the difference that B and φ are trainable in the SIREN case. This finding closes the bridge between Fourier frequency mappings and sinusoidal activation functions which have recently attracted a lot of attention. Progressive training Chen-Hsuan et al. [13] introduced a training strategy for coarse-to-fine registration for NeRFs which they called BARF. Their idea is to mask out the positional encoding's high-frequency activations at the start of training and gradually allow them during training. Their work showed how to use this strategy on positional encodings only to improve camera registration. In our work, we will show how to run this strategy on an arbitrary Fourier mapping and show that it improves generalization of the interpolation task. We weigh the frequencies of γ as follows: where w α B is the element wise application of the function w α (z) on the vector of Norms of B on the input dimension: where w α (z) is defined as: Here, α ∈ [0, max((\|\|B i \|\| din ) i∈{1,...,m} )] is a parameter which is linearly increased during training. This strategy forces the network to train the low frequencies at the start of training, ensuring that the network will produce smooth outputs. Later, when high-frequency activations are allowed, the low-frequency components are trained, and the network can focus on the left details. This strategy should reduce the effect of overfitting, which was introduced by Tancik et al. [32] when using mappings with large standard deviations. Pruning The standard way of using equation (10) is by defining a value N and taking the whole set B N . High-dimensional tasks lead to high memory consumption, and it is not clear whether this subset of Z d brings best performance. We, therefore, propose a way to select a more appropriate subset through data pruning. A pruning pr(N, M ) is done as follows: Assume we have N, M ∈ N with M >> N and \|B N \| = n, \|B M \| = m. We train a perceptron with an integer mapping given by B M . After training we define D such that D contains only those elements of B M where the respective weights are greater than a margin, that is chosen to yield \|D\| = n. While B N and D now have the same size, we believe that D will yield better performance because it contains the most relevant frequencies of the signal we want to reconstruct. Integer lattice mapping applied to MLPs Although we showed in section 3.1 that we can represent any bounded input function with only one Fourier mapped perceptron, in practice, these networks can become very wide to give high performance. As a result, the number of calculations will increase. To compromise between performance and speed, one can add depth and reduce the width of the network. First, it is natural that using MLPs rather than perceptrons increases the performance. However, it remains unclear why our proposed integer mapping should perform better than competing mappings for multilayer networks. One could argue that if a mapping gives the perceptron a high representation power, it will also provide a high representation power to the MLP and vice versa. First, however, we should verify this claim with experiments. In addition, we remind the reader that a periodic function has integer frequencies. And because our assumption that the signal we want to reconstruct is periodic, it will have only integer frequencies. Also, the activation functions we are using only introduce integer frequencies when applied to a periodic function, as shown for the 1D case in the supplementary material. With this, we reduce the search space for frequencies from R to Z, which could make the optimization easier as the search space is more compact and approachable. Weight initialization and progressive training In this section, we want to prove our mathematical claims by experiments. First, we will show that the derivation of the integer mapping indeed represents the Fourier series. Secondly, we want to check whether progressive training helps with generalization. We conducted our experiments on the image regression task. This task aims to make a neural network memorize an image by predicting the color at each pixel location. We use ten images with a resolution of 512 × 512, which can be found in the supplementary material, and report the mean peak signal-to-noise ratio (PSNR). We divide the image into train and test sets, where we use every second pixel for training and take the complete image for testing. We utilize 3 Fourier-mapped perceptrons with N = 128 (Nyquist frequency), one for each image channel. We normalize the input (x) to have an interval between [0,1] in both width (x) and height (y) dimensions. In this experiment, we made an ablation: With and without weight initialization using the normalized FFT coefficients of the image's training pixels, with and without the progressive training scheme explained in section 3.3. For progressive training, α was linearly increased from 0 to its maximum value at 75% of training iterations. In training, we only make an update step after we accumulate the gradients of the whole image. We did not study learning schedules in this work, and the reader is encouraged to try different schedules. Figure 2 shows a visualization of one of the images, and figure 3 shows the training progress, where the solid line is the mean PSNR and the shaded area shows the standard deviation. As can be deducted from figure 3, one can see that the train PSNR starts at an optimum at the start of training when we use weight initialization (WI), and we don't use progressive training (PT). This fact underlines our claim that a perceptron with an integer lattice mapping is indeed a Fourier series. Note that in case both WI and PT are used, the train PSNR is not optimal at the start because the PT masks out high-frequency activations. We can also see from figure 3 that whenever we use progressive training, it always shows a higher test PSNR, which certifies that progressive training helps with generalization. Lastly, when we did not employ both PT and WI, the perceptron overfits to the training pixels, and this can be seen quantitatively with a very low test PSNR (red line in figure 3) and qualitatively with grid-like artifacts (in figure 2a)). Perceptron experiments In this experiment, we want to compare the representation power of the different mappings in the single perceptron case. We conducted our experiments in the same setting as in section 4.1, where we used progressive training and did not use weight initialization. In the integer mapping, we increased N 's value from 4 until half the training image dimension (Nyquist frequency) and calculated all possible permutations B N , as discussed in section 3.1. For the Gaussian mapping, we sample m = \|B N \| parameters from a Gaussian distribution with a standard deviation of 10 (which was the best value for this task in our experiments). Also, we test a one-layer SIREN with one hidden layer having the same size m. Finally, we adopt the positional encoding (PE) scheme from [19] and limit its values to N . Figure 4 shows our experiments' results on the train and test pixels, respectively. Figure 5 shows the networks' outputs trained on one of the images. At low N values (figure 5a), we see that the Gaussian mapped perceptrons do not work because the number of sampled frequencies is low, so there is a low chance that samples will be near the image's critical frequencies. On the other hand, the integer mapped perceptrons give a blurry image because they can only learn low frequencies. The SIREN performs relatively well in this case, and we think this is because SIRENs naturally inherit a learnable Fourier mapping that is not restricted to the initial sampling, as de-Activ. scribed in section 3.2. PE can only produce horizontal and vertical lines because it has diagonal frequencies (only one non-zero frequency is allowed), and this effect is persistent at any value of N . As N increases, SIREN, Gauss, and integer mapping performance increase giving similar performance around N = 16 (figure 5b). For high values of N , we see that in figure 5c, the integer lattice mapping of the Fourier coefficients outperforms the competing mappings, clearly displaying more details in the reconstruction. On the other hand, the PSNR of the SIREN and the Gaussian mapped perceptrons saturates. We think this is because both mappings rely on sampling the frequencies. Although we can get many of the critical frequencies of the image with sampling, it is improbable to get all of them simultaneously. Even the trainability of the SIREN mapping did not help in this case. MLP experiments Our theory for integer mapping assumes an underlying function that is periodic. However, it is not clear that we will end up with a periodic function if we go the other way, using an integer mapping. In this experiment, we want to check if applying an integer mapping forces periodicity. Secondly, we want to validate our claim (in section 3.5) that if a mapping gives the perceptron a high representation power, it will also give a high representation power to the MLP and vice versa. We compared ReLU networks with integer, Gaussian, PE, and pruned integer mapping (section 3.4). We also compared SIRENS with no mapping (extra layer), integer, pruned mapping. We made a grid search of the parameters N= [8,16,32], depth=[0, 2, 4, 6] (depth=0 represents a perceptron), and fixed the width to 32. For the pruned mapping, we used a pr (N, 128). And for the Gaussian mapping, we had two settings. The first one had a standard deviation of 10 (σ 10 ), which had the best performance in the perceptron experiments. In the second one, we set the standard deviation the same as the pruned integer mapping's standard deviation (σ pr ) to check its effect. Tables 1 and 2 show the mean train and test PSNRs respectively. Figure 1 shows a visualization of the network's outputs at N = 16, depth = 4 and width = 32 for the first period and next period in the height and width directions (f ([x + 1, y + 1])). And we see that the integer mapping forces the network's underlying function to be periodic unlike the SIREN and ReLU network with Gauss mapping, which proves our first hypothesis. From the table 1 we see that if a mapping at d = 0 gives the highest PSNR, this does not mean that it will give the highest PSNR for d > 0 and vice versa. One clear exam- ple at N = 32 is the Gauss σ pr , where it has a PSNR of 13.01 dB at d = 0, which is lower than integer mapping (19.84 dB), but has the highest PSNR at d = [4,6]. This result disproves our initial assumption that if a mapping gives the perceptron a high representation power, it will also give a high representation power to the MLP. We see also that the pruned integer mapping has comparable results with the Gauss σ pr , and this shows that the main contributor to the performance is the mappings' standard deviation. From the tables, we can also observe some trends. First, networks with sine activations and large mappings collapse during perform worse than Relu networks. Second, the integer mapping usually gives the best test PSNR, demonstrating its effectiveness in the MLP case. Third, the pruned integer mapping shows consistently better train PSNR than the normal integer mapping at d > 0. We believe this is because pruned mapping has a higher standard deviation. Finally, the PE is worse in every case because we cannot easily control the standard deviation, and it has very few parameters. Novel view synthesis experiments This section wants to see if our findings in the image regression task transfer to the novel view synthesis (NVS) task. In NVS, we are given a set of 2D images of a scene, and we try to find its 3D representation. With this represen-tation, one can render images from new viewpoints. In contrast to the 2D experiments, the inputs are (x, y, z) coordinates that are mapped to a 4-dimensional output, the RGBvalues, and a volume density. For this experiment, a simplified version of the official NeRF [19] is used, where the view dependency and hierarchical sampling are removed. Here, we experiment with the input mappings used in section 4.3. Unless otherwise stated, we adopt the settings from the image regression task. We set the network width to be 64. As the mapping size increases exponentially, we do our experiments with lower frequencies than in the 2D case. Specifically, we used the integer mapping on four frequencies. The frequencies of our mapping were limited to the maximum network size which we could fit on NVIDIA GTX-2080Ti. The pruning is given by pr (4,8). We conduct our experiments on the bulldozer scene, which is commonly used for Nerf experiments. For training, we used a batch size of 128, 50.000 epochs and a learning rate of 5 × 10 −4 . As seen in Table 3, in the perceptron case (d = 0), SIREN provides the best performance, which aligns with our image regression results at low values of N . We observe that the pruned mapping increases the performance compared to normal mapping for both Relu and sinusoidal activation. This increase in performance is because pruned mapping has a higher standard deviation than normal mapping. Qualitative improvements of the pruning can be seen in Figure 6. Gauss gives comparable results to pruned integer mapping because they have the same standard deviation. These findings align with our conclusions from image regression experiments. However, due to memory limitations, we could not test a perceptron with frequencies higher than 8, which was superior in image regression. Conclusion In this work, we identified a relationship between the Fourier mapping and the general d-dimensional Fourier series, which led to the integer lattice mapping. We also showed that this mapping forces periodicity of the neu-ral network underlying function. From experiments, we showed that one perceptron with frequencies equal to the Nyquist rate of the signal is enough to reconstruct it. Furthermore, we showed that the progressive training strategy improves the generalization of the interpolation task. Lastly, we confirmed the previous findings that the main contributor to the mapping performance is its size and the standard deviation of its elements.	2021-09-02T01:15:59.982Z	2021-09-01T00:00:00.000	{ "year": 2021, "sha1": "3fac7e69f463ed64821fa29f7053f5651a7f70ce", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "3fac7e69f463ed64821fa29f7053f5651a7f70ce", "s2fieldsofstudy": [ "Computer Science" ], "extfieldsofstudy": [ "Computer Science" ] }
219845975	pes2o/s2orc	v3-fos-license	THE EFFECT OF FINANCIAL LITERATION ON FINANCIAL PERFORMANCE IN COOPERATIVES OF STATE EMPLOYEES IN BANDAR LAMPUNG (Financial Management Behavior Of Cooperative Managers As Intervening Variables) Allegedly, the reason for the number of cooperatives not actively running is because there are still many financial problems. A person's ability to use money depends very much on the knowledge they have, the better the financial literacy that someone has, the better the financial management behavior of that person. This kind of thing, then, will have an impact on one's financial performance. Therefore, studies regarding this matter are very important to do. In this study, the total sample was 50. all the data obtained were analyzed using path analysis. Path analysis is used to analyze the pattern of relationships between variables in order to determine the direct and indirect effects of a set of independent variables on the dependent variable. Based on the results of data processing, it can be seen that the value of the direct effect of the variable behavior of financial management on financial performance is equal to 0.681. Whereas, the indirect effect of variables x1, x2 and x3 on financial performance through financial management behavior is 0.0051 (0.367 x 0.480 x 0.029). by comparing the two values, it can be proved that the influence of knowledge variables (X1), understanding (X2), and application (X3) on financial performance is directly smaller than their influence on financial performance (Z) through financial management behavior (Y), so it can be concluded that the behavior of financial management is an intervening factor for financial literacy with financial performance. Someone who has good financial literacy, coupled with proper financial management behavior, can be said as prosperous in terms of finances. By having good skills in managing finances, a cooperative manager will be free from financial problems so that the cooperative's financial performance will be achieved optimally. Keyword :financial literacy, financial management behavior, cooperative financial performance A. INTRODUCTION In an increasingly rapid economic life, the role of cooperatives is still considered very important in helping the people's economy. The cooperative is a forum held by government institutions for the benefit of its members. It can be seen from the efforts made by the government in providing guidance, convenience, and protection to cooperatives. Over time, the development of cooperatives in Indonesia continues to grow. it is indicated by BPS data from year to year which continues to increase. Individual needs and increasingly complex financial products demand that cooperative managers must have adequate financial literacy. Individuals need basic financial knowledge and the ability to manage financial resources effectively for better welfare and development of cooperatives. With good and proper financial management coupled with good financial literacy, it is expected that all members of the cooperative can get the maximum benefit from the money they have today which, in the end, can be beneficial for improving financial performance. Based on the background above, then I focused this research on investigating the effect of financial literacy on the financial performance of civil servant cooperatives in Bandar Lampung city by using financial management behavior as an intervening variable. B. REVIEW OF LITERATURE financial understanding as financial knowledge and the ability to apply it (knowledge and ability) which includes basic knowledge about personal finance (basic personal finance), knowledge about money management (cash management), knowledge of credit and debt, knowledge of savings and investment as well as knowledge about risks 2 .A person's ability to use money depends very much on the knowledge they have, the better the financial literacy that a person has, the better the financial management behavior of that person, which, in turn, will affect the financial performance (Perry and Morris, 2005). Financial literacy and financial knowledge are important assets in a company, especially for small businesses. Having financial literacy means that an individual must have confidence in using his financial knowledge to make a financial decision. Financial literacy is not only lampung-terancam-dibubarkan. 2018 measured by how someone has information about finance but how he can apply it appropriately. Financial literacy is defined as financial knowledge and its application in daily life. The meaning of financial knowledge is that individuals have knowledge about the use of financial products obtained through education or experience 3 . The application of financial knowledge is defined as one's ability and one's confidence in using their knowledge in using financial products. The Republic of Indonesia Financial Services Authority (OJK) is an Education Implementation Agency that has a role in improving public finances. OJK categorizes the level of Indonesian financial literacy into four: 1) Well literate, means that an individual has knowledge of financial service institutions and financial services products, has confidence in selected financial products, and has skills in using financial products and services 2) Sufficient literate, means that an individual has knowledge and beliefs about financial service institutions and financial products and services; including features, benefits and risks, as well as rights and obligations related to financial products and services. 3) Less literate, means that an individual only has knowledge and understanding of financial service institutions, as well as financial products and services. 4) Non literate, means that an individual does not have the knowledge and beliefs about financial services institutions, financial products and services. In addition, someone in this category also does not have the skills in using financial products and services. Based on the explanation above, I conclude that financial literacy is a way of helping in providing an understanding of financial management and opportunities to achieve good financial performance in the future. In other words, financial literacy can be used as a tool that needs to be improved by a person or individual if they want to have good financial performance 6 . This study aims to examine the effect of financial literacy on Likert Scale financial performance with financial management behavior as an intervening variable. This research is categorized into descriptive analysis, namely research carried out to more accurately describe the characteristics of businesses to determine the frequency of occurrence of something or the relationship of something else and provide symptoms as well as facts or events in a systematic and accurate manner regarding the natures of a particular population or area. This research was conducted in Bandar Lampung in 7 Civil Servants Cooperatives; KPRI Saptawa, KPRI BetikHati, KPRI TigaSehat, KPRI BetikGawi, KPRI Kosuma SMPN 10, KPRI Intan Sejahtera IAIN Lampung and KPRI RagomGawi. Data sources used in the study are primary data sources. The primary data obtained were obtained through surveys where the instrument used was a questionnaire. Questions answered on a Likert scale. C. RESULT AND DISCUSSION Instrument testing in this study was continued by using the partial least square (PLS) method. PLS is an alternative method of analysis with structural Equation Modeling (SEM) based on variance. The tool used is a version 3 smartPLS program, specifically designed to estimate structural equations on a variance basis 7 . Influence of knowledge on financial management behavior H0: knowledge does not influence financial management behavior H1: knowledge influences financial management behavior From the calculation above, the t-statistic value of knowledge (X1) is 3.110 greater than the value of t table (1.6778) and the value of p value is 0.002 (smaller than 0.05) which means that Ho is rejected and H1 is accepted. In other words, knowledge influences financial management behavior. Influence of understanding on financial management behavior H0: understanding does not affect financial management behavior H2: understanding influences financial management behavior From the results of the calculation, the value of t-statistic understanding (X2) is 3.460 (greater than the value of t table = 1.6778) and the p value is 0.001 (smaller than 0.005) which means that H0 is rejected and H2 is accepted. In other words, understanding influences financial management behavior. H0: application does not influence financial management behavior H3: application influences financial management behavior From the results of the calculation above obtained t-statistics application value (X3) of 0.196 (smaller than the value of t table = 1.6778) and p value of 0.844 (greater than 0.005) which means that H0 is accepted and H3 is rejected. In other words, implementation has no effect on financial management behavior. intervening variable, effect of financial literacy on financial performance through conformity of financial management behavior. H0 :financial literacy does not affect financial performance through conformity of financial management behavior H4 :financial literacy affects financial performance through conformity of financial management behavior Based on the results of processing the data above, it can be seen that the direct effect value between the variables of financial management behavior on financial performance is 0.681 while the indirect effect between variables x1, x2 and x3 on financial performance through financial management behavior is 0.0051 (0.367 x 0.480 x 0.029). by comparing the two values, it can be proved that the influence of knowledge variables (X1), understanding (X2), and application (X3) on financial performance is directly smaller than their influence on financial performance (Z) through financial management behavior (Y) Thus, it can be concluded that financial management behavior is an intervening factor between financial literacy and financial performance 8 . Goodness of fit Evaluation This analysis is used to determine whether the model constructed is good enough to explain the phenomenon being studied. From the analysis of the goodness of fit model, the value of q square is: This indicates that from the statistical calculation, the model is good. Because the variation of the variables involved in the model has been able to explain the variables that affect financial performance by 76.20%, the remaining 23.80% is explained by other factors that also affect inancial performance, including the standard error. The results of this study aim to provide answers to the formulation of the problems that have been stated in the previous chapter. Based on the results of hypothesis testing using the above methods, several issues can be explained as follows: Effect of knowledge on financial management behavior. Based on the results of hypothesis testing, H0 is rejected. It means that knowledge influences financial management behavior. this result is in accordance with the theory put forward by Remund (2010) which says that knowledge is one of the common and, at the same time, a must-have aspect in the context of financial literacy. In order to manage finances, one must have a strong knowledge of finance. The better the knowledge possessed about financial literacy, the better it is for a cooperative administrator to manage cooperative finance. Based on the results of the study, it is known that the level of educations of civil servant cooperative administrators in Bandar Lampung are mostly scholars (with different educational backgrounds). Not all administrators are graduates from a financial background, but basically they are given debriefing about financial literacy in the form of workshops, seminars and classes on cooperative finance, hence, it can be said that their knowledge of finance is good. This is in line with the results of the hypothesis 1 which indicates that knowledge influences the behavior of financial management of civil servant cooperatives in Bandar Lampung. Effect of understanding on financial management behavior Hypothesis test results show that H0 is rejected and H2 is accepted, meaning that understanding influences financial management behavior. This is in line with the theory put forward by Budiono (2012), which says that someone who has a basic personal finance has an understanding of a financial system, such as the calculation of interest, opportunity costs, time value, liquidity and so forth. In other words, the more understanding a cooperative administrator has about financial literacy, the better the cooperative financial management behavior. Based on this research, financial literacy measured on cooperative financial management behavior is related to understanding of the exchange rate, features of financial services, financial recording, attitudes in issuing finance which includes understanding of compiling the income budget that will be received by each member and management of the cooperative. The better the understanding of the cooperative management regarding financial literacy, the better the financial management behavior that has an impact on the results of cooperative financial management. this finding is in line with the results of the second hypothesis test where understanding influences the financial management behavior of civil servant cooperatives in Bandar Lampung. Effect of application on financial management behavior The results of the third hypothesis test show that H0 is accepted and H3 is rejected, meaning that the application does not affect the behavior of financial management. This is in accordance with the category stated by the Financial Services Authority (OJK), where the officials of the civil service cooperative in Bandar Lampung fall into the category of Less literate (the administrators only have the knowledge and understanding of financial services institutions, financial products and services) 9 . This means that the management has good human resources equipped with knowledge and understanding, while, in practice, the cooperative management does not have financial literacy skills in financial management. The only resources available are knowledge and understanding gained from workshops, and seminars, so that the application in financial literacy does not affect the behavior of financial management. Another thing that also contributes as a cause is the fact that some of their educational background does not come from finance. In addition, the recruitment of cooperative administrators is not based on educational background (financial) but from the position and personal closeness between members of the board. In financial management behavior, not only knowledge and understanding of financial literacy are needed, but also the need for application or skills in financial management. The results of hypothesis testing indicate that the application does not affect the behavior of financial management. The effect of financial literacy on financial performance through intervening variables of financial management behavior The results of the fourth hypothesis test show that H0 is rejected and H4 is accepted, meaning that financial literacy affects financial performance through conformity of financial management behavior. According to Zimmerere and Scarborough (2008), the behavior of financial management is the process of predicting, Liya, Dimas, and Diah The effect of Financial Literation….. collecting, issuing, investing and planning cash that is needed by companies or individuals so that the economy can run smoothly 10 . Success in managing cooperative finances is strongly influenced by personal behavior in the cooperative management, knowledge or understanding of management, lifestyle and financial goals (short, medium and long-term goals of the family). Justin (2010) revealed that someone who has good financial literacy plus with behavior in managing finances appropriately, can be said as prosperous in financial terms. Thus, to be prosperous, financial literacy must lead to good financial management behavior for someone because with good skills in managing finances, it will avoid a financial problem. In the context of cooperatives, financial performance will be achieved maximally if these conditions can be met. This finding is in accordance with the results of hypothesis testing which show that financial literacy influences financial performance through the conformity of financial management behavior in civil servant cooperatives in Bandar Lampung. D. CONCLUSION From the results of the discussion in the previous chapter, the following conclusions are obtained: 1. In order to be able to manage finances, one must have a strong knowledge of finance, the better the management's knowledge about finance, the better the financial management of cooperatives. 2. Someone who has basic personal finance will also have an understanding of a financial system, such as calculating interest, opportunity costs, time value, liquidity, etc., meaning that the more understanding the cooperative management has about financial literacy, the better the cooperative financial management behavior. 3. Administrators of civil servant cooperatives in Bandar Lampung fall into the category of Less literate, which means they only have knowledge and understanding of financial service institutions, as well as financial products and services without having financial literacy skills in financial management. In addition, some of them did not come from the financial sector, and the recruitment of cooperative management changes was also not based on educational background (financial) but was seen from the position and personal closeness between the administrators. 4. Someone who has good financial literacy coupled with proper financial management behavior can be said as financially prosperous. Thus, to be prosperous, financial literacy must lead to good financial management behavior for someone. In the context of cooperatives, if the above conditions are met, the management of the cooperative will be able to avoid financial problems, which, in turn, make the cooperative's financial performance runs optimally.	2020-05-21T00:15:44.299Z	2019-09-12T00:00:00.000	{ "year": 2019, "sha1": "5ec8282c9f6620debd5faf09803ca00a2b2d81db", "oa_license": "CCBYSA", "oa_url": "https://e-journal.metrouniv.ac.id/index.php/FINANSIA/article/download/1441/1310", "oa_status": "HYBRID", "pdf_src": "Anansi", "pdf_hash": "6d7eb7c91a8b971624a5653cd7f3856a079764d5", "s2fieldsofstudy": [ "Business" ], "extfieldsofstudy": [ "Business" ] }
259097559	pes2o/s2orc	v3-fos-license	Development and Evaluation of a Pragmatic Measure of Adherence to Dialectical Behavior Therapy: The DBT Adherence Checklist for Individual Therapy This paper presents two studies conducted to develop and evaluate a new pragmatic measure of therapist adherence to Dialectical Behavior Therapy (DBT): the DBT Adherence Checklist for Individual Therapy (DBT AC-I). Study 1 used item response analysis to select items from the gold standard DBT Adherence Coding Scale (DBT ACS) using archival data from 1271 DBT sessions. Items were then iteratively refined based on feedback from 33 target end-users to ensure relevance, usability, and understandability. Study 2 examined the psychometric properties of the DBT AC-I as a therapist self-report and observer-rated measure in 100 sessions from 50 therapist-client dyads, while also evaluating predictors of therapist accuracy in self-rated adherence. When used as a therapist self-report measure, concordance between therapist and observer ratings was at least moderate (AC1 ≥ 0.41) for all DBT AC-I items but overall concordance (ICC = 0.09) as well as convergent (r = 0.05) and criterion validity (AUC = 0.54) with the DBT ACS were poor. Higher therapist accuracy was predicted by greater DBT knowledge and adherence as well as more severe client suicidal ideation. When used by trained observers, the DBT AC-I had excellent interrater reliability (ICC = 0.93), convergent validity (r = 0.90), and criterion validity (AUC = 0.94). While therapists’ self-rated adherence on the DBT AC-I should not be assumed to reflect their actual adherence, some therapists may self-rate accurately. The DBT AC-I offers an effective and relatively efficient method of evaluating adherence to DBT when used by trained observers. Supplementary Information The online version contains supplementary material available at 10.1007/s10488-023-01274-x. Dialectical Behavior Therapy (DBT; Linehan, 1993) is an evidence-based psychotherapy (EBP) that is primarily used to treat borderline personality disorder (BPD; Storebø et al., 2020) and self-injurious behaviors (SIB;DeCou et al., 2019). Substantial efforts have been made to implement DBT around the world at the clinic, system, state, and national levels (e.g., Carmel et al., 2014;DuBose et al., 2019;Flynn et al., 2020;Herschell et al., 2014). Encouragingly, research has found that leading DBT training models result in high rates of adoption and sustainability of DBT in diverse healthcare settings (e.g., Navarro-Haro et al., 2019;Swales et al., 2012). However, little is known about the degree to which DBT is delivered with fidelity once adopted, which is a critical aspect of quality assurance and indicator of implementation success (Proctor et al., 2009). Therapist adherence is a core component of fidelity and refers to the degree to which therapists deliver the treatment procedures as intended (Perepletchikova et al., 2007). In DBT, higher therapist adherence leads to better outcomes for clients (Harned et al., 2022), making it a particularly important target for implementation and quality assurance efforts. A primary barrier to evaluating the impact of DBT implementation efforts on therapist adherence has been the lack of a pragmatic method of measuring adherence. As is typical for adherence measures more broadly (Schoenwald & Garland, 2013), the gold standard measure of adherence to DBT utilizes direct observational methods in which trained coders rate recorded therapy sessions. The DBT Adherence Coding Scale (DBT ACS; Linehan & Korslund, 2003) is a comprehensive, observational measure that assesses the degree to which therapists sufficiently deliver each DBT strategy in a session. The DBT ACS has excellent inter-rater reliability, strong discriminant validity, and generates dependable adherence scores (Harned et al., 2021a, b). The DBT ACS is also effective in predicting reductions in key client outcomes, including suicide attempts, psychiatric hospitalizations, and treatment dropout (Harned et al., 2022). Although the DBT ACS is a highly effective method of measuring adherence to DBT, it is not practical for use as a quality assurance tool in routine care. As is true with any observational adherence measure (Schoenwald et al., 2011), the DBT ACS requires considerable resources to use (e.g., time, money, equipment) and can be burdensome for both therapists and clients (e.g., due to having to record sessions). Additional barriers specific to the DBT ACS include its length and complexity (66 items coded on a 6-point Likert scale), the extensive training required to become a reliable coder, access restrictions that limit its availability to approved coders, and the high cost associated with its use. As a result, the DBT ACS has only been used for high stakes purposes, including treatment outcome research and therapist certification. This has left therapists, programs, and systems that have invested in the implementation of DBT with no way to evaluate if the treatment is being delivered in a manner consistent with the evidence-based intervention. This is particularly concerning given a recent study finding a high rate of non-adherent sessions (48%) among therapists trained in a system-level DBT initiative (Harned et al., 2021a, b). Thus, there is a critical need to develop pragmatic measures of DBT adherence that can be used to guide quality assurance efforts in diverse practice contexts. Developing pragmatic measures of adherence is challenging because of the tension that exists between effectiveness (psychometric rigor) and efficiency (feasibility for use in routine care), both of which are difficult to achieve in a single instrument (Schoenwald et al., 2011). To date, most efforts to resolve this tension have focused on developing self-report measures for therapists to rate their own delivery of a treatment (Schoenwald & Garland, 2013). While therapist self-report measures are typically feasible (e.g., quick, inexpensive, nonintrusive), most studies have found low concordance between therapist self-report and observer ratings due to therapists tending to overestimate their adherence (e.g., Brookman-Frazee et al., 2021;Brosan et al., 2008;Hogue et al., 2015;Hurlburt et al., 2010;Martino et al., 2009). However, higher concordance has been found for some EBPs (e.g., Brookman-Frazee et al., 2021;Hogue et al., 2015) and when rating the presence/absence of techniques rather than frequency or competence (e.g., Martino et al., 2009). In addition, certain therapist characteristics (e.g., higher confidence, less professional experience) and session factors (e.g., ability to carry out planned activities) may improve the accuracy of self-ratings (Brookman-Frazee et al., 2021;Loades & Myles, 2016). Taken together, these findings suggest that therapist self-report offers an efficient method of measuring adherence but requires evaluation to determine whether and under which conditions it may be effective. Given the challenges associated with therapist self-report, another option for improving the efficiency of adherence measurement without sacrificing effectiveness is to simplify and streamline gold standard observational measures. Potential modifications to observational measures that may improve efficiency include converting them to a checklist format, simplifying the rating scale, and reducing the number of items (Schoenwald et al., 2011). For the DBT ACS, this could involve condensing the measure to the most critical DBT strategies, collapsing the 6-point rating scale to a binary scale, and using a checklist format that includes behavioral anchors that describe adherent and non-adherent delivery of each strategy. This paper reports the results of two studies that describe the development and evaluation of a new pragmatic measure of adherence to DBT: the DBT Adherence Checklist for Individual Therapy (DBT AC-I). Consistent with recommendations to use well-established observational measures as the foundation for developing pragmatic adherence measures (Schoenwald et al., 2011), the DBT AC-I was derived from the DBT ACS as this approach was expected to be more likely to yield a measure with strong psychometric properties and construct validity (Hogue et al., 2015). Overall, the aim of this project was to develop a measure that: (a) was brief and easy to use, (b) could be completed as a therapist self-report or observer-rated measure, (c) had strong psychometric properties, (d) could be used for quality assurance purposes in routine practice, and (e) would be free and publicly available. Study 1 focused on developing the DBT AC-I and an accompanying training manual based on item response analysis of archival DBT ACS data followed by iterative feedback from target end-users. Study 2 evaluated: (1) the reliability and validity of the DBT AC-I when used as a therapist self-report measure, (2) therapist-, client-, and session-level predictors of therapist accuracy in rating their own adherence on the DBT AC-I, and (3) the reliability and validity of the DBT AC-I when used by trained observers. Participants DBT Therapists (n = 27) Therapists were recruited from July to October 2019 by posting information about the study on listservs (email groups) for mental health professionals, including a DBT-specific listserv. Interested individuals were directed to complete an online screening questionnaire to determine eligibility. Inclusion criteria were: (1) currently delivering DBT individual therapy in a routine practice setting (defined as any setting other than an academic medical or research setting); (2) completed at least some training in DBT (defined as at least one of the following: 2 + days of workshops, a graduate level course in DBT, 6 + hours of online training, or reading any of Linehan's DBT manuals); (3) licensed or working under the supervision of a licensed mental health provider; and (4) age 18 + . Individuals were excluded if they: (1) had insufficient English proficiency; (2) did not reside in the United States; and/or (3) conducted professional training in DBT and/or were trained in the DBT ACS. See Table 1 for therapist characteristics. Overall, 215 individuals completed the online therapist eligibility screen, 136 (63.3%) of whom reported meeting the eligibility criteria. Reasons for exclusion were: (1) did not reside in the United States (n = 16), (2) worked in an academic medical or research setting (n = 9), (3) did not provide DBT individual therapy (n = 4), (4) was unlicensed without a supervisor (n = 2), and (5) conducted professional training in DBT and/or was trained in the DBT ACS (n = 48). For each iteration of the formative evaluation, 6-12 unique therapists were recruited from the pool of eligible therapists with the goal of achieving a sample with varying degrees, practice settings, and levels of training in DBT. Of the 136 eligible therapists, 39 were invited to participate of whom 27 (69.2%) enrolled in the study. The remainder did not respond to recruitment emails (n = 9), declined participation (n = 2), or were no longer eligible to participate (n = 1). DBT Experts (n = 6) DBT experts were recruited from the professional networks of the research team. Experts could participate in multiple iterations of the formative evaluation. Inclusion criteria were: (1) conducts professional training in DBT for a recognized DBT training company, and/or (2) is a reliable coder on the DBT ACS. Procedures All study procedures were approved by the VA Puget Sound Institutional Review Board. Initial Scale Development Original Measure Items were derived from the DBT ACS (Linehan & Korslund, 2003) that includes 66 items across 12 subscales that reflect the strategies used in DBT as speci- fied in the treatment manual (Linehan, 1993). Not all strategies (items) are required in each session, and the measure includes "if/then" rules for when they must be used. Each item is rated on a scale where 0 = not used/not necessary, 1-3 = below adherence, 4 = minimum threshold for adherence, and 5 = adherent with high sufficiency. The DBT ACS is scored by averaging all non-zero items to create a computed global score where 4.0 and higher is considered "adherent." Data Sources An archival dataset of DBT ACS data from six DBT clinical trials was used for analyses. Across the six trials, trained observers used the DBT ACS to rate 84 therapists during 1,271 DBT individual therapy sessions with 292 clients. In addition, 78 sessions were coded by the gold standard rater (Dr. K. Korslund) for the purpose of evaluating interrater reliability according to the procedures of each trial. The dataset included both efficacy and effectiveness trials of DBT with adults and adolescents from a variety of client populations. A detailed description of this archival dataset can be found in Harned et al., (2021a, b). Item Selection The first step in reducing the pool of DBT ACS items was to remove items with poor pooled interrater reliability (IRR < 0.60; Cicchetti, 1994) and limited variability (Hinkin, 1998). The remaining items were then analyzed using item response theory (IRT) to identify the items that were most informative regarding the DBT ACS computed global score. IRT was selected because it improves upon classical test theory and is widely used in clinical research to shorten measures (Reise & Waller, 2009). Item response models are built around the assumption of unidimensionality. Given that the DBT ACS has multiple subscales, to accommodate unidimensionality we conducted exploratory and confirmatory factor analyses (EFA and CFA) on the remaining items. The sample was randomly split into two near equal sized sub-samples. EFA was conducted on the first subsample (i.e., developmental sample) to identify the factor structures and CFA was implemented on the second subsample (i.e., validation sample) to confirm the factor structures. Items were identified as loading within a factor (i.e., loadings ≥ 0.30) or non-loaders. We then fit separate IRT graded response models (GRMs) for all the loading items within each factor as well as for the collection of non-loading items. Unidimensionality was assessed based on the eigenvalues of polychoric correlations. Candidate items for inclusion were selected based on: (1) the slope parameters using a threshold of > 0.5 (An & Yung, 2014) to indicate the item was informative regarding overall adherence, and (2) item information curves (IIC), test information curves (TIC), and item characteristic curves (ICC) to determine how well the item discriminated between therapists at different levels of adherence. When these fit indices were comparable among items, we selected items that were viewed as important in defining DBT and maximized coverage of content across the 12 DBT ACS subscales. Feedback on the candidate items was obtained in the first phase of the formative evaluation (see below) before the items were finalized. A second IRT model was conducted with the final set of items to examine overall model fit as well as item slope and threshold parameters. Rating Scale Selection Analyses were also conducted to evaluate whether the 6-level DBT ACS rating scale could be simplified to a binary scale. To that end, a total score was computed by dichotomizing each candidate item as nonadherent (0 = 1, 2, or 3) or adherent (1 = 0, 4, or 5) and summing them. To determine whether there was a significant loss in discrimination with the binary scale, the Area Under the Curve (AUC) was used to quantify the ability of this total score (sum of binary items) to correctly identify adherent versus non-adherent sessions according to the original DBT ACS computed global score. We set a benchmark for using the binary scale as achieving at least excellent discrimination (AUC ≥ 0.90) as well as being significantly correlated with the DBT ACS computed global score. Formative Evaluation Consistent with best practices from product development models, the DBT AC-I and the accompanying manual were tested and iteratively revised in a formative evaluation. Revisions were based on feedback from the target audience (DBT clinicians) and DBT experts with the goal of ensuring relevance, usability, and understandability. Expert Interviews In the first phase, feedback was obtained from DBT experts via individual interviews about the relevance and breadth of the candidate items (i.e., whether the items included the most critical strategies of DBT), item wording and clarity, and the rating scale. This expert feedback was used to finalize the items that were included in the measure. Experts were paid $40 to complete an interview. Therapist Surveys In the second phase, DBT therapists provided feedback via an online survey that asked them to: (1) rate each item in terms of how easy it was to understand on a scale ranging from 1 (not at all) to 5 (extremely), and (2) suggest changes for each item. The measure was then iteratively revised based on therapist feedback until adequate understandability was achieved, which was defined as at least a mean rating of 4 ("very" easy to understand) out of 5 for each DBT AC-I item. Therapists were paid $30 to complete an online survey. Think-Aloud Interviews The third phase involved conducting think-aloud interviews with DBT therapists and experts to evaluate the measure's usability and effectiveness in enabling participants to think like expert adherence coders. Think-aloud interviews are a method of capturing participant thought processes when engaging with an instrument (Wolcott & Lobczowski, 2021). Participants were instructed to complete the measure about a recent session of their own (therapists) or of a supervisee (experts) while continuously verbalizing their thought processes out loud (e.g., about why they decided to rate an item as adherent or non-adherent). During the think-aloud process, the interviewer transcribed participants' thoughts and did not speak except to prompt them to continue talking if they were silent for more than 20 s. Participants were then asked to provide qualitative feedback on the measure and the manual as well as to complete the 10-item System Usability Scale (SUS; Brooke, 1996), a widely used measure of usability that has excellent psychometric properties (Bangor et al., 2008). All SUS items were modified by changing the generic term "the system" to refer specifically to the DBT AC-I (e.g., "I thought the DBT Adherence Checklist was easy to use"). Items are rated on a 5-point scale from 1 (strongly disagree) to 5 (strongly agree) and used to generate a total score ranging from 0 (negative) to 100 (positive) where scores of 68 or higher indicate above average usability (Bangor et al., 2008). Therapists were paid $30 and experts were paid $40 to complete the think-aloud interviews. Expert review and finalization After additional revisions were made, a final round of written feedback was obtained from DBT experts before the measure and manual were finalized for validation in Study 2. Experts were paid $40 to provide this final review. Initial Scale Development Of the 66 DBT ACS items, 17 were removed due to having poor pooled interrater reliability (IRR < 0.60) and 4 were removed due to having extremely low variability (coefficient of variation < 5%). The remaining pool of 45 items was analyzed using IRT. Based on the IRT results, 24 items were initially selected as candidates for inclusion in the measure that showed strong overall model fit. Specifically, the 24 items explained 90.6% and 88.3% of the total information for the latent trait of adherence compared to the 45 items included in the IRT and all 66 DBT ACS items, respectively. To evaluate whether the continuous (0-5) rating scale could be collapsed to a binary (0-1) scale, a revised total score was created by dichotomizing and summing the 24 candidate items. Analyses indicated that this revised total score had excellent discrimination between adherent versus non-adherent sessions (AUC = 0.90, SE = 0.01) and was significantly correlated with the original DBT ACS computed global score (r = 0.70, p < 0.0001). Thus, a binary scale was selected for the measure to reduce complexity. Each candidate item was then written using a common format with behavioral anchors describing adherent (1) and non-adherent (0) use of the strategy based on the definitions provided in the DBT ACS. Formative Evaluation The initial 24-item measure was then iteratively refined based on therapist and expert feedback. In total, four iterations were completed that included expert interviews (n = 6), therapist surveys (n = 21), think-aloud interviews (n = 6 therapists and 3 experts), and final review by experts (n = 2). In this process, 4 items were added and 3 items were removed to improve construct and content validity, item wording was refined to enhance clarity, and the training manual was developed. The final version of the measure included 25 items rated on a binary scale that were each rated as "very" to "extremely" easy to understand (M's = 4.1-4.9 out of 5). In addition, the measure was rated as highly usable on the SUS (M = 80.0, SD = 8.6). Fit of the Final Scale Given that items were added and removed during the formative evaluation process, an IRT model was re-run on the final set of 25 items. Overall, the model showed a good fit with our data. Specifically, the 25 items explained 92.1 and 89.4% of the total information for the latent trait of adherence compared to the 45 items included in the IRT and all 66 DBT ACS items, respectively. All but one item had a slope parameter ≥ 0.5, indicating the items provided important information about the overall construct of adherence (see Supplemental Table 1 for all parameter estimates). The one item with a slope < 0.5 (informal exposure) was viewed as important from a content validity perspective and was therefore retained despite its low informativeness. To confirm the use of a binary rating scale, the AUC was re-computed using the revised total score (sum of the final 25 dichotomized items). This total score yielded excellent discrimination between adherent and non-adherent sessions (AUC = 0.91, SE = 0.01) and was highly correlated with the DBT ACS computed global score (r = 0.72, p < 0.0001). Participants Therapists (n = 50) Therapists were recruited from January 2020 to February 2021 by posting information about the study on listservs for mental health professionals, including a DBT-specific listserv, emailing therapists who self-identified as delivering DBT in the Psychology Today online database, and emailing therapists who had previously attended DBT workshops conducted by Behavioral Tech, LLC. Interested individuals were directed to complete an online screening questionnaire to determine eligibility. Inclusion criteria were the same as in Study 1 plus therapists were required to: (1) be able to video-record and share therapy sessions with the study team and (2) have one DBT client (18 +) who was able to attend treatment for the duration of the study period and consented to having their sessions recorded and shared with the study team. Individuals were excluded if they: (1) had insufficient English proficiency; (2) had already participated in Study 1; (3) were trained in the DBT ACS; and/or (4) did not reside in the United States. Eligible therapists who were invited to participate in the study were sent an Information Statement as well as a packet of information to provide to a study-eligible client. Clients (n = 50) Therapists were asked to recruit a studyeligible client from their caseload and to have interested clients complete a permission to contact form. Once therapists returned the signed permission to contact form, a member of the study team called the client to describe study procedures and answer any questions. Interested clients were sent an Information Statement and HIPAA authorization form to complete and return to the research team. Upon receipt of the signed consent materials, clients were compensated $30 for their time. Procedures All study procedures were approved by the VA Puget Sound Institutional Review Board. Study Assessments Data collection occurred from February 2020 to July 2021 and involved having therapists complete four online surveys over an approximately 4-month period. Therapists were sent a link to access the baseline survey after their client returned their consent materials. Therapists who completed the baseline assessment were considered enrolled in the study. Upon enrollment, therapists were emailed electronic copies of the DBT AC-I and its accompanying manual and asked to review them prior to using the measure to rate their first session. Each therapist was also mailed an encrypted USB key on which to save and return their video-recorded therapy sessions to the research team. Therapists were asked to video-record the first two sessions with their study client that occurred following their baseline assessment and to complete a brief online survey after each session that included the DBT AC-I. Therapists were asked to complete the DBT AC-I as close to the end of the session as possible and were told they could initially complete it on paper if needed prior to entering it into the online survey. Therapists also completed a follow-up survey 3 months after the second session. Therapists were paid $30 for completion of the baseline survey, $30 for completion of each of the two session recordings and associated surveys, and $40 for completion of the follow up survey. Participant Flow and Retention Overall, 253 individuals completed the online eligibility screen, 173 (68.4%) of whom met the eligibility criteria. Reasons for exclusion were: (1) did not reside in the United States (n = 29), (2) participated in Study 1 (n = 16), (3) worked in an academic medical or research setting (n = 8), (4) did not have an eligible client (n = 8), (5) did not provide DBT individual therapy (n = 3), (6) was unlicensed without a supervisor (n = 1), (7) was trained in the DBT ACS (n = 1), and (8) was unable to video-record sessions (n = 1). Thirteen individuals were deemed potentially eligible and in need of additional follow up if there were insufficient eligible therapists. Therapists were selected to participate based on the goal of achieving a sample with a variety of degrees, practice settings, and levels of training in DBT. Of the 173 eligible therapists, 133 were invited to participate of whom 53 (39.8%) enrolled in the study. The remainder did not respond to recruitment emails (n = 15), failed to return completed permission to contact forms (n = 20), declined participation (n = 23), could not get agency approval to participate (n = 6), did not have an eligible client (n = 6), were unable to participate due to COVID pandemic-related changes to their work environment (n = 6), or did not complete the baseline assessment (n = 4). Of the 53 enrolled clinicians, 50 completed all the study procedures and are included in the analyses. Observational Coding Raters (n = 3) were selected for this study who had previously been trained to reliability on the DBT ACS at the level of the subscale by Dr. K. Korslund, the gold standard rater. Prior to commencing coding for the present study, all raters coded a calibration session to ensure they had not drifted from the gold standard rater. During the present study, a total of 100 sessions (2 per therapist/ client dyad) were coded using the DBT ACS. Two raters were assigned to code each therapist (1 per session) to minimize the risk of rater bias. In addition, reliability checks were periodically conducted during the study for each rater using a random selection of 10% of all coded sessions (n = 10). Reliability was evaluated by comparing the rater's scores to those of the gold standard rater. Interrater reliability was excellent for the DBT ACS computed global score (ICC = 0.98, 95% CI = 0.90-0.99). Therapist-Level Measures Therapist Characteristics At baseline, therapists provided information about their demographics, professional characteristics, DBT program, DBT training/experience, and self-efficacy in delivering DBT (ranging from 1 'not at all comfortable' to 4 'very comfortable'). DBT Knowledge A 46-item multiple choice test with 4 response choices per item was developed by study investigators to assess knowledge of DBT strategies and skills as well as ability to apply knowledge in hypothetical clinical scenarios. Items were developed to map onto the therapist strategies assessed in the DBT ACS and multiple iterations were piloted with DBT and non-DBT therapists during Study 1 before finalizing the measure for use in this study. The score used for analysis was the proportion of items correct. The test had excellent discriminant validity between DBT therapists in the present study (n = 50, 81.5% correct) and therapists with no training in DBT (n = 49, 35.4% correct), t (97) = 23.3, p < 0.001, d = 4.68. Client-Level Measures Client Demographics and DBT Treatment Therapists reported on their clients' demographics and current DBT treatment modes at baseline. Psychiatric Functioning Therapists reported their clients' current psychiatric diagnoses at baseline based on DSM-5 diagnostic codes. Therapists also rated their clients' global functioning via the Global Assessment Scale (GAS; Endicott et al., 1976), a 0-100 scale of the overall severity of illness where higher scores indicate better functioning. Self-Injurious Thoughts and Behaviors Therapists completed the informant report version of the Columbia Suicide Severity Rating Scale (C-SSRS; Posner et al., 2011) to assess the number of times their client attempted suicide (actual, aborted, or interrupted attempts) and engaged in non-suicidal self-injury (NSSI) in the 3 months prior to the baseline assessment. The C-SSRS also assessed the severity of their clients' suicidal ideation (SI) in the past 3 months ranging from 0 (none) to 5 (active SI with specific plan and intent). Session-Level Measures Therapist Self-Reported Adherence The 25-item DBT AC-I was used to assess therapists' self-reported adherence to DBT in two consecutive individual therapy sessions. A total score was calculated by summing all items (range = 0-25). In addition, therapists completed items assessing: (1) when they completed the DBT AC-I in relation to the session; (2) whether they read the DBT AC-I manual and/or referred to it while rating the session; (3) how long it took to complete the DBT AC-I, and (4) how helpful the DBT AC-I manual was as a training method (ranging from 0 'not at all' to 5 'extremely'). At the Session 2 assessment, therapists also completed the SUS (Brooke, 1996; see Study 1 for a description). Observer-Rated Adherence The DBT ACS (Linehan & Korslund, 2003) was used to code each therapy session (see Study 1 for a description). To enable comparison of the observer and therapist ratings, the DBT ACS items that make up the briefer DBT AC-I were recoded to a binary scale to match the scale of the DBT AC-I. Specifically, DBT ACS ratings of 0, 4, and 5 were recoded to 1 (adherent) and ratings of 1, 2, and 3 were recoded to 0 (non-adherent). Session Targets After each session, therapists completed a brief survey to assess their clients' target behaviors since the last session. The 11-item Borderline Symptom List-Behavioral Items (BSL-BI; Bohus et al., 2009) assessed a variety of impulsive behaviors common in BPD (e.g., selfinjury, substance use, binge-eating) on a scale from 0 (not at all) to 4 (daily or more often). The 23-item version of the Therapist Interview (TI; Chalker et al., 2015) assessed three types of challenging client behaviors common in BPD that are based on the DBT concept of therapyinterfering behaviors: (1) interpersonal negativity (e.g., "behaved in an inflexible or defiant manner"), (2) avoidant/disengaged (e.g., "missed session without calling"), and (3) behavioral dysregulation (e.g., "arrived at sessions under the influence of drugs or alcohol"). All items were rated on a 5-point scale from 0 (behavior did not occur) to 4 (behavior occurred and seriously interfered). A single item also assessed therapists' overall level of stress and/ or burnout in treating the client from 1 (none at all) to 5 (extreme). Data Analysis Aim 1 evaluated the psychometric properties of the therapist-rated version of the DBT AC-I. A confirmatory factor analysis (CFA) was tested to determine if the DBT AC-I items conformed to a single dimension consistent with the scoring of the DBT ACS. To determine goodness of fit, the following indices and thresholds were used: Root Mean Square Error of Approximation (RMSEA < 0.06; Hu & Bentler, 1999), Comparative Fit Index (CFI > 0.95; Byrne, 2006), Normed Fit Index (NFI > 0.90; Bentler & Bonett, 1980), and Nonnormed Fit Index (NNFI > 0.90; Bentler & Bonett, 1980). Therapist-observer concordance was evaluated at the item-level using the Agreement Corrected 1 (AC 1 ) statistic, which is recommended as an alternative to kappa when rater agreement and trait prevalence are both high (Gwet, 2008). The AC 1 statistic resolves the well-known paradox in which kappa appears small when rater agreement is high (Cicchetti & Feinstein, 1990). The AC 1 statistic is interpreted using the same criteria for interpreting kappas where less than 0.20 is poor, 0.21-0.40 is fair, 0.41-0.60 is moderate, 0.61-0.80 is good, and 0.81-1.00 is very good (Altman, 1999). We set a benchmark for retaining items as achieving at least moderate therapist-observer agreement (AC 1 ≥ 0.41). To provide an index of overall therapist-observer concordance, an ICC was computed for the DBT AC-I total score using the randomeffects estimate with two raters pooled (ICC [2, 2]; Shrout & Fleiss, 1979). According to Cicchetti's (1994) criteria for interpreting ICCs, less than 0.40 is poor, 0.40-0.59 is fair, 0.60-0.74 is good, and 0.75-1.00 is excellent. Convergent validity was evaluated by examining the correlation between the DBT AC-I total score and the DBT ACS computed global score. Criterion validity was evaluated by determining how well the DBT AC-I identified adherent versus non-adherent sessions as defined by the gold standard DBT ACS (i.e., computed global score ≥ 4.0). All possible cut-off scores were examined in terms of sensitivity (i.e., the percent of true positives that are above the cutoff) and specificity (i.e., the percent of true negatives that are below the cut-off). As the cut-off score varies, the locus (1-specificity, sensitivity) yields a receiver operating characteristic (ROC) curve that displays the performance of all possible cut-off scores and the optimal score was selected by the max-min approach that simultaneously maximizes the probability of predicting a true-positive or a true-negative (Gallop et al., 2003). The AUC was calculated to quantify the ability of the DBT AC-I to correctly classify adherent versus non-adherent sessions and an AUC ≥ 0.90 indicates excellent criterion validity. Content validity was evaluated by examining whether the items included in the DBT AC-I adequately represented the strategies most often in need of improvement among DBT therapists in routine practice. Specifically, the 66 items in the DBT ACS were examined to determine if those with high rates of non-adherence (defined as ≥ 25% of sessions) were included in the DBT AC-I. Aim 2 sought to identify therapist-, client-, and sessionlevel predictors of therapist accuracy in rating their adherence to DBT. Therapist factors included demographics (sex, age, degree), DBT experience (workshop days, years providing DBT, self-efficacy), and DBT knowledge. Client factors included psychiatric diagnoses (BPD, total diagnoses), SIB history (C-SSRS), and functioning (GAS score). Session factors included session type (telehealth vs. in-person), target behaviors (BSL-BI, TI), and observer-rated adherence (DBT ACS). Therapist accuracy was calculated as the percentage of concordant items between therapist and observer ratings. One subject with an extreme outlying low value was capped at the value of the next closest subject through winsorization. Mixed effects models were used to accommodate the nested nature of the data (sessions within client-therapist dyads). Initial univariate mixed effects models were conducted to examine the effect of each predictor on therapist accuracy. A final multivariate mixed effects model was run that included all predictors found to be significant in the univariate models. For Aim 3, the structural, convergent, and criterion validity of the observer-rated DBT AC-I were evaluated using the same analytic approaches as described in Aim 1. Interrater reliability (ICC) was evaluated between observers and the gold standard rater of the DBT ACS using a random subset of 10% of all coded sessions (n = 10). Aim 1. Psychometric Properties of the DBT AC-I as a Therapist Self-Report Measure Therapists completed the DBT AC-I immediately after the session (n = 31, 31.0%), the same day as the session but not immediately after (n = 48, 48.0%), and one or more days after the session (n = 21, 21.0%). The median time to complete the DBT AC-I was 20.0 min. Prior to completing the checklist, most therapists read the DBT AC-I manual completely (n = 43, 86.0%) and the remainder read the manual partially (n = 6, 12.0%) or not at all (n = 1, 2.0%). Therapists rated the manual as 'very helpful' on average (M = 3.9 out of 4, SD = 1.0) and the DBT AC-I as having above average usability on the SUS (M = 71.8, SD = 10.8). Structural Validity A CFA with one dimension that included all the items yielded an acceptable fit: RMSEA = 0.059, CFI = 0.958, NFI = 0.935, NNFI = 0.821. This indicates that the use of a total summed score adequately represents the data. Therapist-Observer Concordance Descriptive data for the therapist and observer ratings for each DBT AC-I item and the total score are shown in Table 3. Table 4 reports the frequencies of concordant and discordant ratings, agreement rates, and AC 1 statistics for each DBT AC-I item. The average agreement rate across all items was 84.2% (range = 63.0-98.0%, SD = 0.10). All 25 items had at least moderate therapist-observer concordance (AC 1 ≥ 0.41), which we considered the minimum acceptable threshold of agreement. Concordance was very good for 14 items (56%), good for 7 items (28%), and moderate for 4 items (16%). Therapists accurately rated themselves as adherent on 81.9% of items and non-adherent on 2.4% of items. Therapists overestimated their adherence on 8.8% of items and underestimated their adherence on 7.0% of items. Given that therapists over-and under-estimated their adherence at similar rates, these discordant ratings canceled themselves out such that the average DBT AC-I total score did not significantly differ between therapists and observers (M difference = 0.46, range = − 8-14, SD = 3.52, t (99) = 1.31, p = 0.19, d = 0.13). However, therapist-observer concordance for the DBT AC-I total score was poor (ICC = 0.09, 95% CI − 0.35-0.39) due to the variability in agreement rates across therapists. Convergent Validity The therapist-rated DBT AC-I total score was not significantly correlated with the observer-rated DBT ACS computed global score (r = 0.05, p = 0.64), indicating it had poor convergent validity with the gold standard measure of adherence to DBT. Table 4 Concordance between therapist and observer ratings on the DBT AC-I items "True positive" = rated as adherent by both therapists and observers "True negative" = rated as non-adherent by both therapists and observers "False positive" = rated as adherent by therapists and non-adherent by observers "False negative" = rated as non-adherent by therapists and adherent by observers Content Validity Four DBT ACS items met the threshold for having a high rate of non-adherence (≥ 25% of sessions), of which 3 were included in the 25-item DBT AC-I (generalizes learning, dialectical synthesis, activates new behavior) and 1 was not (troubleshooting). In DBT, troubleshooting is required whenever a commitment is obtained. Overall, the contingent strategies of asking for and troubleshooting a commitment were below adherence in 75.0% of the sessions, suggesting that adding these strategies to the DBT AC-I would improve its content validity as a measure designed to help therapists identify and improve areas of non-adherence. Aim 2. Predictors of Therapist Accuracy of Self-Rated Adherence to DBT Therapist accuracy (i.e., the percentage of concordant items between therapists and observers) ranged from 68 to 100% across therapists (M = 84.4%, SD = 0.08). Table 5 reports the results of univariate mixed effects models predicting therapist accuracy. Therapist Factors Therapists with greater knowledge of DBT were significantly more accurate in rating their own adherence (p = 0.03). Therapist sex, age, degree, number of DBT workshop days, years of DBT experience, and self-efficacy did not significantly predict accuracy. Client Factors Therapists were significantly more accurate when rating sessions conducted with clients with more severe suicidal ideation (p = 0.02). Client diagnoses, SIB history, and functioning did not significantly predict therapist accuracy. Session Factors Therapists with higher observer-rated adherence on the DBT ACS were significantly more accurate in rating their own adherence (p < 0.0001). Session type and target behaviors did not significantly predict therapist accuracy. Combined Model A multivariate MEM that included the three significant predictors from the individual models was significant (Likelihood χ 2 (2) = 8.01, p = 0.019). In this model, only greater observer-rated adherence remained a significant predictor of higher therapist accuracy (B = 0.277, SE = 0.050, p < 0.0001). Therapist accuracy was not significantly related to knowledge of DBT (p = 0.76) or the severity of the client's suicidal ideation (p = 0.11). Aim 3. Psychometric Properties of the DBT AC-I as an Observer-Rated Measure All analyses utilized the observer ratings on the DBT ACS items that make up the briefer DBT AC-I. Each DBT ACS item was recoded to a binary scale to match the scale of the DBT AC-I. We examined the psychometric properties of two potential versions of the observer-rated measure: (1) a 25-item version that was identical to the DBT AC-I completed by therapists, and (2) a 26-item version that added the commitment/troubleshooting item identified as important to improve the content validity of the measure (see Aim 1 results above). Interrater Reliability The average item-level agreement rate between observers and the gold standard rater was 94.8% for the 25-item version and 95.0% for the 26-item version. Inter-rater reliability was excellent for the DBT AC-I total score for both the 25-item version (ICC = 0.93, 95% CI 0.71-0.98) and the 26-item version (ICC = 0.91, 95% CI 0.64-0.98). Structural Validity CFAs with one dimension that included all the items indicated adequate fit for both the 25-item version (RMSEA = 0.06, CFI = 0.96 NFI = 0.90, NNFI = 0.93) and the 26-item version (RMSEA = 0.09, CFI = 0.95, NFI = 0.86, NNFI = 0.91). Although the RMSEA and NFI fit indexes were slightly above/below the recommended thresholds for one or both versions, this is quite common in smaller samples (n ≤ 100; Taasoobshirazi & Wang, 2016). Thus, the use of a total summed score appears to adequately represent the data. Convergent Validity The DBT AC-I total score was significantly correlated with the DBT ACS computed global score for the 25-item (r = 0.89, p < 0.001) and the 26-item version (r = 0.90, p < 0.001). Criterion Validity Both versions of the observer-rated DBT AC-I were used to predict adherent versus non-adherent sessions according to the DBT ACS computed global score. For the 25-item version, the AUC was 0.93 and maximum accuracy was achieved at a cut-off score of 23 (85% overall agreement; 95% CI 76.5-91.4%; sensitivity = 89.3%; specificity = 79.6%). For the 26-item version, the AUC was 0.94 and the optimal cut-off score was 23 (86% overall agreement; 95% CI 77.6-92.1%; sensitivity = 91.1%, specificity = 79.6%). This indicates that the 26-item version increases the ability to identify adherent sessions to an excellent level (sensitivity ≥ 0.90), whereas the 25-item version is slightly below this threshold. Discussion These studies aimed to develop and test a pragmatic measure of adherence to DBT individual therapy that would enable therapists, programs, and systems to evaluate and improve the quality of services they deliver. In Study 1, the DBT AC-I was empirically derived from the gold standard DBT ACS using a large archival database to identify critical items and an optimal rating scale. To maximize construct validity, content for the items and an accompanying manual was written to be consistent with the definitions used in the DBT ACS. The measure and manual were then iteratively revised based on feedback from DBT therapists and experts until high usability and understandability were achieved. In Study 2, the psychometrics of the DBT AC-I as a therapist self-report and an observer-rated measure were evaluated among 50 therapist-client dyads engaged in DBT in routine practice settings. Results have important implications for the implementation and practice of DBT. When used by therapists to rate their own adherence, the DBT AC-I was efficient (median time to complete was 20 min) and rated as highly usable. Concordance between therapist and observer ratings was at least moderate for all items and very good for more than half the items. Agreement tended to be lower for items that were more frequently non-adherent (e.g., chain analysis, activating new behavior, modeling dialectical thinking) and higher for items with the highest rates of adherence (e.g., warm engagement, validation, reinforcement). This suggests that therapists may be more accurate in evaluating their own adherence for strategies that they are more proficient in delivering. Alternatively, given prior research indicating therapists tend to overestimate their adherence (e.g., Brookman-Frazee et al., 2021;Hogue et al., 2015;Hurlburt et al., 2010), agreement may be higher when items are adherent due to therapist rating bias. Notably, therapists in the present study overestimated and underestimated their adherence at similar rates. As a result, the average DBT AC-I total scores generated by therapists and observers did not significantly differ even though there was poor overall therapist-observer concordance. Additionally, the therapist-rated DBT AC-I total score had poor convergent validity with the gold standard DBT ACS and was unable to distinguish between adherent and non-adherent sessions. These findings are consistent with the larger literature indicating that therapist self-report is an efficient method of evaluating adherence but often has limited effectiveness due to therapists' difficulty accurately self-rating their adherence (e.g., Brosan et al., 2008;Hogue et al., 2015;Hurlburt et al., 2010;Martino et al., 2009). Importantly, therapist-observer concordance varied considerably across therapists with overall accuracy rates ranging from 68 to 100%. Higher therapist accuracy was predicted by greater DBT knowledge and observer-rated adherence. More knowledgeable and adherent therapists may be more able to deliver the strategies they intended to carry out and self-reflect on their performance, both of which predict greater accuracy in self-reported adherence in other EBPs (Brookman-Frazee et al., 2021). Therapists were also more accurate when rating sessions with clients who had more severe suicidal ideation. DBT sessions often have a clearer structure when clients' suicide risk increases (e.g., certain strategies and protocols are required), and therapists have been found to rate their adherence more accurately when delivering EBPs with more structured content (Brookman-Frazee et al., 2021). Overall, these findings indicate that the DBT AC-I may be both efficient and effective as a selfreport measure of adherence for some but not all therapists. Given that there is no way to easily predict which therapists are likely to be accurate raters, DBT AC-I scores generated by therapist self-report should not be assumed to be reliable. Future research may be able to develop algorithms to improve the accuracy of therapists' self-rated adherence on the DBT AC-I by adjusting scores based on key predictors. When rated by trained observers, the DBT AC-I had excellent interrater reliability as well as strong convergent and criterion validity with the gold standard DBT ACS. Of the two versions that were evaluated, the 26-item DBT AC-I is recommended given its improved content validity and excellent ability to identify adherent sessions (sensitivity) compared to the 25-item version. Notably, these strong psychometric properties were achieved despite eliminating more than 60% of the DBT ACS items (from 66 to 26) and simplifying the rating scale (from a 6-point to a binary scale). In addition, the DBT AC-I is free and publicly available (whereas the DBT ACS is not), further reducing burden for users. Altogether, the DBT AC-I offers a more efficient and comparably effective alternative to the gold standard DBT ACS when used by trained observers. Nonetheless, any method that relies on direct observation of sessions by trained raters is likely to be infeasible in some settings (e.g., agencies that do not have the equipment needed to record sessions or access to trained observers). Limitations Results must be interpreted in the context of several limitations. First, the participants in these studies may not be representative of DBT therapists and clients in routine practice settings more broadly. To maximize generalizability, we recruited therapists with a range of degrees, disciplines, training backgrounds, and practice settings and did not use inclusion/exclusion criteria for clients except to ensure they were not minors. However, therapists who enrolled in the study were typically highly trained, experienced in DBT, and worked primarily in private practice settings. In addition, therapists may have been particularly motivated to learn about and improve their adherence to DBT. Clients were moderately impaired on average, had relatively low rates of recent self-injurious behavior, and consented to have their sessions reviewed by the study team, which may not be typical of DBT clients more broadly. It is also unknown if the results would vary if the DBT AC-I were used to rate sessions with adolescent clients, although prior research has found comparable reliability and validity of the DBT ACS in adult and adolescent samples (Harned et al., 2021a, b). Additional research with larger and more diverse therapist and client samples is needed to replicate and extend the present findings. Other limitations include that most sessions were conducted via telehealth due to the COVID-19 pandemic and it is possible that this impacted results; however, session type did not predict therapist accuracy in self-rated adherence. Additionally, to enable evaluation of convergent and criterion validity with the gold standard measure, the trained observers completed the full DBT ACS rather than just the items included in the DBT AC-I. It is possible that ratings may have differed if only the DBT AC-I items were completed. The use of a binary rating scale (adherent vs. non-adherent) for the DBT AC-I also limits the ability of the measure to evaluate therapist competence compared to the continuous scale of the DBT ACS. Finally, this study focused only on developing a pragmatic measure of adherence to DBT individual therapy. Future research is needed to develop measures of adherence for the other modes of DBT (e.g., group therapy). Implications and Next Steps The lack of a pragmatic adherence measure has been a significant barrier to DBT implementation and quality assurance efforts. Without such a measure, there has been no way to evaluate if DBT is being delivered as intended to the clients who need it. This is particularly concerning given that therapists in routine practice often deliver DBT nonadherently according to this study (44% of sessions) and prior research (48% of sessions; Harned et al., 2021a, b). Moreover, many clients who receive DBT are at high risk for suicide and therapist adherence is an important mechanism by which suicidal behavior is reduced (Harned et al., 2022). The DBT AC-I represents a significant advance towards enabling stakeholders to evaluate if the treatment being provided is consistent with DBT as it is defined in the treatment manual. The final 26-item DBT AC-I has several notable strengths, including being brief, easy to use, freely available (at www. dbtad heren ce. com), and having excellent construct and content validity. However, other indices of effectiveness (interrater reliability, convergent and criterion validity) vary depending on the characteristics of the rater, indicating a need for caution in how the measure is used. Given that therapists overall had difficulty generating reliable and valid scores on the DBT AC-I, therapists' self-rated adherence on the DBT AC-I should not be assumed to reflect their actual adherence to DBT. These findings suggest that the brief training provided to therapists in this study (i.e., reading the DBT AC-I manual) is likely to be insufficient for many therapists to complete the DBT AC-I accurately. Consequently, we have since created a set of expert-rated mock sessions for therapists to use for training and calibration purposes (available at www. dbtad heren ce. com). We also made minor revisions to the measure and manual prior to making it publicly available with the goal of increasing the reliability of therapist ratings (e.g., clarifying definitions and giving more examples for items with lower agreement rates). An important next step for research will be to evaluate whether these revisions and training methods improve the effectiveness of the DBT AC-I as a therapist self-report measure. Until then, it is recommended that the DBT AC-I be used for lower stakes purposes (e.g., quality improvement, supervision and training), but not as a formal assessment of adherence, when completed by therapists (and others) whose reliability has not been established. In contrast, when completed by trained observers the DBT AC-I had excellent psychometric properties but was more burdensome than therapist self-report, reflecting the tension between effectiveness and efficiency that is often encountered in adherence measurement (Schoenwald et al., 2011). Future research is needed to identify methods for assessing adherence to DBT that are effective and less resource-intensive than direct observation of sessions. For example, recent studies have found that behavioral rehearsal (Becker-Haimes et al., 2022) and review of clinical worksheets (Wiltsey Stirman et al., 2021) hold promise as indirect methods of evaluating therapist adherence. Thus, while the DBT AC-I represents an important advance in DBT practice and research, additional work to synthesize the tension between efficiency and effectiveness is needed.	2023-06-08T06:16:43.331Z	2023-06-07T00:00:00.000	{ "year": 2023, "sha1": "758b977a006304d9eca6fa398b81cab7f74c1767", "oa_license": null, "oa_url": "https://link.springer.com/content/pdf/10.1007/s10488-023-01274-x.pdf", "oa_status": "BRONZE", "pdf_src": "PubMedCentral", "pdf_hash": "8870e94ad018ea7cd1ab880e3327809a8785f705", "s2fieldsofstudy": [ "Psychology" ], "extfieldsofstudy": [ "Medicine" ] }
250131266	pes2o/s2orc	v3-fos-license	A Review of Researches on the Influence of Feeling Trusted on Employees' Work Performance : In recent years, trust-related research has been well-known by scholars, and feeling trusted is an important part of trust-related fields. Trust and feeling trusted are reflected in the different perspectives of the relationship between leaders and employees. However, research on feeling trusted is relatively inadequate. Previous studies have mentioned that feeling trusted has a positive effect on employee performance, but with the deepening of research, scholars have proposed that the impact of feeling trusted on employee performance exists as a double-edged sword. Through the establishment of different mechanisms, from different perspectives such as appropriateness theory, social exchange theory, and resource conservation theory, we have found that feeling trusted has positive and negative effects on job performance. This article summarizes the impact of perceived trust on performance through a review of previous studies, and proposes possible future research directions. Introduction In recent years, trust research has been extensive, and trust is very important for all employees in an organization. In a diversified organizational structure, if the trustee does not have a long-term cooperative relationship, correspondingly, in management research, trust-related topics have also been paid more and more attention. Trust is an important part of workplace relationships. Feeling trusted and trust are two related but not the same concepts [1]. Employees' perception of being trusted by leaders in an organization is a subjective consideration of the leaders' willingness to take risks in their actions. And the leader's trust in employees is to point out the leader's willingness to take risks for the employees' behavior from an objective perspective. Therefore, it is necessary to fully understand the mechanism and root causes of perceived trust in employee performance. Concept of feeling trusted In the workplace, employees' perception of being trusted by the leader refers to the employee's perception of the extent to which the leader is willing to take risks for his own behavior, representing the employee's subjective feelings, and having an important impact on the employee's work. For example, employees perceive that their superiors have given him a very important task, which is directly related to the future development of the enterprise, which is a perception of trust. Since only the superior really knows whether he really trusts a subordinate, any superior trust information learned from the subordinate's perspective should be regarded as "feeling trusted" [2]. At present, the research on feeling trusted is still in the preliminary stage. Existing studies have found that subordinates' feeling trusted not only improves their subordinate's organizational selfesteem and work performance, but also improves organizational responsibility norms and organizational performance. At present, a small number of scholars have raised questions about this. For example, Wang verifies that being trusted can increase employee loyalty to leaders by improving employee satisfaction with leaders from the perspective of social exchange theory [3]. Dimensions and measurement of feeling trusted Trust, as behavioral intention, is often characterized from two aspects: Reliance, which refers to the willingness to rely on the skills, knowledge and abilities of others; the other is disclosure or open meeting (Disclosure), which refers to the willingness to share sensitivity with others Information, this information disclosure often contains a lot of sensitive information in work and private life. A higher level of trust in subordinates is often accompanied by higher levels of dependence and information disclosure. Therefore, employees' feeling trusted can also be obtained from two aspects: one is the superior's perception of dependence on their subordinates, and the other is the superior's perception of their subordinates' information disclosure. The current research on feeling trusted is still immature, and the scales used include the 10-item trust measurement scale of Gillespie. This scale has been proven effective by many studies and has good psychological measurement characteristics. In the study, feeling trusted was measured based on the reliance of superiors on subordinates (Reliance) and the disclosure of information from superiors to subordinates (Disclosure). The high dependence of supervisors on their subordinates is that they are willing to let them do and work. Related decisions; at the same time, information disclosure is usually closely related to personal information disclosure, likes, and high-quality relationships. In addition, Chen et al. adopted a perceptual trust-based scale in the Experimental Research Center. For example, "Your boss asks you to be important to him/her." Play a role at work", "Your boss relies on your judgment on work-related issues and believes that it is not necessary to supervise your work", etc. Concept of work performance Work performance research has long been a hot spot in academic circles, but the definition of performance concept has not been unified yet. Bernarding stated that the so-called performance refers to the performance. It has been a hundred years since the inception of performance research. However, the core issue is controversial. Is performance a result or behavior? This question is the key factor in defining job performance. Work performance inquiry has always been a key part of human resource management research. Work performance is a pertinent perception of work completion. However, due to the differences in various types of information of the research objects in the empirical process, there are also certain differences in corporate organizational structures, salary standards, and management regulations. Therefore, there are also obvious corporate performance indicators. Dimensions and measurement of work performance Common job performance includes task performance, peripheral performance, and adaptive performance. All of these have greatly enriched the scope of work performance research, and can effectively fit the rapidly changing organizational environment, and can provide important assistance to the achievement of organizational goals in a true sense. Similarly, it can also ensure that organizational behavior research and human resource management research obtain new analytical approaches. Contextual performance is a behavioral performance that "does not directly point to production or service activities" but constitutes the social and psychological environment of the organization. Contextual performance using the scale of Van (1996), including two dimensions. Among them, there are 7 questions in the interpersonal promotion dimension, such as "When a colleague succeeds, the subordinate will say yes"; the work dedication dimension has a total of 8 questions, such as "In order to complete the work on time, the subordinate will work overtime". Adaptive performance is a new concept proposed in response to the needs of rapid changes in the organization. Overview of the influence of feeling trusted on work performance The behavior of employees depends not only on how much they trust management, but also how much they feel trusted by management-these are two different concepts. Therefore, some studies have made up for the theoretical black box of perceived trust on job performance by examining whether and how employees' perception of their degree of trust affects organizational performance. Perceived as the other side of trust, can and when will it have an impact on employee performance? Through research, some scholars have opened the black box of perceived trust in work performance. Salamon et al. [4] believe that when employees believe that they are trusted by management, they may be more responsible for their work and pay more attention to the realization of organizational goals. At the same time, higher responsibilities will improve job performance. Bayer [5] proposed that perception of trust has different effects on employees, which can bring reputation to employees, and can also bring pressure, which in turn affects employees' emotional exhaustion and ultimately affects work performance. When an individual perceives being trusted by the leader, it means the leader's recognition of his abilities and the improvement of his status; but at the same time, his own tasks and responsibilities may be greater, which will bring pressure to employees and ultimately affect their work performance. Conclusion Studies have shown that employees' feeling trusted ultimately has different effects on performance by influencing employees' behavioral responses. Different from the impact of trust on employees, employees' feeling trusted starts from a psychological perspective that is more biased towards the employees themselves, and can more directly and effectively derive the effect of trust on individual attitudes and behaviors. At present, the research on feeling trusted is still in its infancy, and there are not many researches on perception and behavior of employees. Therefore, future research can start from feeling trusted, and explore the impact of employees' subjective feeling trusted on their attitudes, behaviors, and even social status.	2022-06-30T15:24:04.467Z	2022-01-01T00:00:00.000	{ "year": 2022, "sha1": "0e34f826809b6846557b5814be3e1a6a146d8ea9", "oa_license": "CCBY", "oa_url": "http://www.clausiuspress.com/conferences/LNEMSS/MSIED%202022/Y0994.pdf", "oa_status": "HYBRID", "pdf_src": "Anansi", "pdf_hash": "d2aefa7f93407009b3d1ad84870a5ad584d16bfe", "s2fieldsofstudy": [ "Business" ], "extfieldsofstudy": [] }
55423698	pes2o/s2orc	v3-fos-license	Impacts of simulated herbivory on volatile organic compound emission profiles from coniferous plants The largest global source of volatile organic compounds (VOCs) in the atmosphere is from biogenic emissions. Plant stressors associated with a changing environment can alter both the quantity and composition of the compounds that are emitted. This study investigated the effects of one global change stressor, increased herbivory, on plant emissions from five different coniferous species: bristlecone pine (Pinus aristata), blue spruce (Picea pungens), western redcedar (Thuja plicata), grand fir (Abies grandis), and Douglas-fir (Pseudotsuga menziesii). Herbivory was simulated in the laboratory via exogenous application of methyl jasmonate (MeJA), a herbivory proxy. Gas-phase species were measured continuously with a gas chromatograph coupled to a mass spectrometer and flame ionization detector (GC–MS–FID). Stress responses varied between the different plant types and even between experiments using the same set of saplings. The compounds most frequently impacted by the stress treatment were alpha-pinene, beta-pinene, 1,8cineol, beta-myrcene, terpinolene, limonene, and the cymene isomers. Individual compounds within a single experiment often exhibited a different response to the treatment from one another. Introduction The largest global source of volatile organic compounds (VOCs) in the atmosphere is emissions from vegetation (Guenther et al., 2000(Guenther et al., , 2012)).These biogenic VOCs (BVOCs) oxidize in the atmosphere and can contribute significantly to the formation of secondary pollutants such as ozone and secondary organic aerosol (SOA) (Atkinson, 2000;Ehn et al., 2014;Hamilton et al., 2009;Kroll and Seinfeld, 2008), and thus play a key role in Earth's climate (Carslaw et al., 2010).Plants emit a wide range of organic compounds that will be classified here structurally into three categories: small oxygenated VOCs (OVOCs), terpenoids (isoprene, monoterpenes, sesquiterpenes, and their oxygenated derivatives), and aromatics (Herrmann and Weaver, 1999;Kesselmeier and Staudt, 1999).The regulation of BVOC emissions depends on both physiological and physicochemical controls that vary both between plant species and between different compounds produced within a single tree (Niinemets et al., 2004).Some BVOCs are constitutive, meaning they are continuously synthesized and emitted by the plant while being regulated by the physiological and physicochemical mechanisms described above.Constitutive emissions can be either de novo or pooled depending on the absence or presence of storage structures.A single plant can emit both de novo and pooled emissions simultaneously (Loreto et al., 2000).In contrast to constitutive emissions, some BVOC emissions are inducible, meaning they are only synthesized and emitted when the plant is exposed to an abiotic or biotic stress that initiates their production.These stress-induced emission rates can make up a significant amount of total plant BVOC emissions (Blande et al., 2007;Brilli et al., 2009;Staudt and Lhoutellier, 2007).They also can increase or decrease the secondary organic aerosol formation potential of the BVOC emissions depending on the types of VOCs that are induced (Mentel et al., 2013). C. L. Faiola et al.: Impacts of simulated herbivory on VOC emission profiles Plant stress can significantly alter the BVOC emission profile both by inducing emissions of additional compounds and by changing the emissions of constitutive compounds (Arneth and Niinemets, 2010).This is an important consideration because different VOCs, even within the same class of compounds, can vary by orders of magnitude in their chemical reactivity (Atkinson and Arey, 1998).A variety of stress exposure studies have been performed investigating BVOC emission changes due to ozone exposure (Heiden et al., 1999;Vuorinen et al., 2004), salt stress (Loreto and Delfine, 2000;Teuber et al., 2008), increased CO 2 (Calfapietra et al., 2009;Constable et al., 1999), enhanced radiation (Harley et al., 1996), drought and/or high temperatures (Kleist et al., 2012;Niinemets, 2010;Niinemets et al., 2010), herbivory (Achotegui-Castells et al., 2013;Copolovici et al., 2011;Engelberth et al., 2004), and pathogen attack (Jansen et al., 2009a;Toome et al., 2010).A thorough review on this topic was presented by Peñuelas and Staudt (2010).Despite the numerous studies investigating this topic, most of these stress influences on BVOC emission rates are still not understood well enough to be included in the models used to develop emissions inventories (Guenther et al., 2012).This is in large part the result of two main factors: (1) the absence of enough quantitative experimental data to generate useful algorithms and (2) the large variability in stress response between trees and even between different compounds emitted by the same tree (Peñuelas and Staudt, 2010, and references therein). Generally, a plant's response to stress depends on the longevity and severity of the stress exposure.Under mild to moderate abiotic stress, biochemical defense pathways are activated that induce and/or increase BVOC emissions, a response that protects the plant from both oxidative and thermal stress (Loreto and Schnitzler, 2010).However, the stress response changes for different types of compounds depending on the physicochemical properties of the compound.For example, emissions of small OVOCs (e.g., methanol, acetaldehyde, and acetone) are closely related to stomatal conductance whereas terpenes are not (Niinemets et al., 2004).Terpenes are hydrocarbons that can diffuse out of the plants into the atmosphere directly through the plant's membranes (Fall and Monson, 1992;Loreto et al., 1996).Consequently, stomatal conductance has no impact on the regulation of terpene emissions because of their chemical properties.In contrast, OVOCs cannot diffuse directly through plant membranes and easily dissolve in aqueous solutions, which further hinders volatilization.Thus, the effects of drought and/or heat stress impact OVOC emissions and terpene emissions differently because plants have evolved mechanisms to deal with these stressors by controlling their stomata.This stressor increases OVOC emissions in the short-term, but after prolonged exposure to the stressor, plants close their stomata to conserve water and a resulting drop in OVOC emissions occurs (Filella et al., 2007;Graus et al., 2013).This same threshold effect was not observed for terpene foliar con-centrations and terpene emissions from Mediterranean tree species and C4 crops (Blanch et al., 2009;Graus et al., 2013).However, other studies have demonstrated that under severe enough drought stress, monoterpene emissions also begin to decrease (Ormeno et al., 2007;Simpraga et al., 2011).Presumably, at some extreme, the plant shuts down metabolic activity and terpene pools, if present, are depleted. One important stressor in future climates will be the increased number of plant-eating pests, leading to increased herbivory (Bale et al., 2002).Plants have evolved to respond to herbivory stress by emitting BVOCs as a defense, using them for communication with other plants, and to signal natural predators of the herbivores (Engelberth et al., 2004).It is well established that herbivory can increase monoterpene, sesquiterpene, and small OVOC emission rates and substantially alter the BVOC profile (Achotegui-Castells et al., 2013;Hu et al., 2008;Laothawornkitkul et al., 2008;Semiz et al., 2012).The presence of herbivore infestation can increase BVOC emissions by 4-to 20-fold (Amin et al., 2012(Amin et al., , 2013;;Berg et al., 2013), and this response can last for several weeks (Priemé et al., 2000).These results suggest that herbivore stress could have a substantial impact on SOA formation in forest environments in the future.However, the number of plants studied using quantitative analytical techniques to measure compound-specific BVOC emission rates is not representative of all the major BVOC emitters in different environments.Furthermore, within the pool of plants that have been studied, large variation has been observed in responses.Emissions of different compounds from the same plant exhibit different temporal responses to herbivory stress (Copolovici et al., 2011).Additionally, the plant stress response varies depending on the type of biotic stress and/or the type of plant; other studies have shown increases in total terpene emission rates after herbivory exposure with no change in the VOC profile (Jansen et al., 2009b;Priemé et al., 2000) or different responses of the same plant to pathogen vs. herbivory stress (Vuorinen et al., 2007).Finally, extrapolating these results to natural environments is further complicated where simultaneous exposure to multiple stressors is likely the rule rather than the exception; multiple abiotic and biotic stressors can interact to significantly alter the plant's response relative to any single stressor (Holopainen and Gershenzon, 2010;Trowbridge et al., 2014;Winter et al., 2012). This study adds to our knowledge of climate change stress impacts on BVOC emission rates by quantitatively investigating the impacts of an herbivore treatment on the VOC profile and emission rates from five different coniferous tree species that have not been the focus of other herbivory studies.This study was a component of a project that investigated the effects of herbivory stress on the composition of biogenic SOA generated from BVOC emissions (Faiola et al., 2014b).Published data on this topic is extremely limited, so one goal of this work was to identify key tree species that could produce a large herbivore-treatment effect on SOA composition.The herbivore treatment was an exogenous application of the plant hormone methyl jasmonate (MeJA).MeJA is a compound that plants use in nature to warn neighboring plants about the presence of herbivores; when plants are exposed to this compound, their emissions respond in a manner similar to if they were being attacked (Martin et al., 2003).This response is not plant species specific and allows even plants of different species to communicate with one another (Farmer and Ryan, 1990).The plant species used in this study are native to temperate coniferous forests in the mountainous regions of the western United States and Canada. Responses to the simulated herbivory stress varied between tree types.Additionally, responses also varied between experiments using the same group of trees within a single tree species, and for different compounds within the same experiment.These results reinforce the necessity to obtain quantitative, compound-specific stress response measurements on a survey of representative trees in an area before stress-induced emissions can be integrated into biogenic emissions model inventories.We also identify a list of VOCs that showed similar stress responses across experiments and could significantly affect atmospheric chemical processes in future scenarios where increased herbivory is present. Experimental approach This research is a component of a larger project investigating plant stress impacts on biogenic secondary organic aerosol formation, using the Washington State University's Biogenic Aerosol Formation Facility.This facility is a dual chamber system with two separate Teflon ® FEP bags: one a dynamic plant emission enclosure where sapling trees are stored and the other an aerosol growth chamber.This dual chamber system uses emissions from living vegetation as a precursor VOC source for SOA generation.The objective of this paper is to present impacts of plant stress on the BVOC emission profile from the subset of experiments where continuous gas-phase measurements were available from the plant chamber.Analysis of the impacts of the stress treatment on the composition of subsequently formed SOA are presented in a separate paper (Faiola et al., 2014b). Tree description and treatment Experiments were performed with saplings from five different coniferous species: bristlecone pine (Pinus aristata), blue spruce (Picea pungens), western redcedar (Thuja plicata), grand fir (Abies grandis), and Douglas-fir (Pseudotsuga menziesii).Pinus aristata and Picea pungens are found in the Rocky Mountains of Colorado.Thuja plicata, Abies grandis, and Pseudotsuga menziesii have wider latitudinal ranges and are found in the northern Rockies of the United States and Canada as well as the western mountain ranges of North America from Alaska to California.Emphasis in the experimental design was on the diversity of representative tree species included, with the goal of identifying species that responded strongly to stress treatment in ways that might affect SOA composition.This emphasis limited the number of replications that were possible. Saplings were 1-3 years of age at the time of the experiments, and were purchased from the University of Idaho Forestry Nursery.Plants were cared for by greenhouse staff to ensure consistent watering and fertilization.They were stored outside of the greenhouse to be closer to their natural environmental conditions and prevent unnatural plant emission behavior that could occur within greenhouse conditions.This also meant the plants could have been exposed to natural stressors (e.g., heat or herbivory).These natural stressors were not controlled but would be representative of conditions encountered by the plants in nature because it is likely that exposure to multiple stressors is the rule rather than the exception in a forest environment (Holopainen and Gershenzon, 2010).Plant specimens were transported from the greenhouse to the laboratory plant chamber at least 2 days before treatment in order to capture a baseline VOC profile.Plants required 24-36 h to acclimate to the plant chamber after transportation.A summary of experiments is provided in Table 1. Treatments using MeJA or jasmonic acid have been used to simulate herbivory response in plants (Filella et al., 2006;Rodriguez-Saona et al., 2001) and can change the terpene emission profile (Martin et al., 2003).The stress treatment used in these experiments was a foliar application of 200 mL of 10 mM MeJA solution in 18.2 M purity water, based on previously reported methods (Martin et al., 2003).Negative control experiments were performed with each tree species, but only two (one from Pinus aristata and one from Picea pungens) were performed while the GC-MS-FID was in operation.The negative control treatment was a foliar application of 200 mL of 18.2 M purity water. Description of plant chamber and analytical instrumentation Three to nine individual saplings were stored in the 0.9 m × 0.9 m × 0.9 m plant enclosure for each experiment; the number depended on the size and age of the trees.6890/5973 GC-MS-FID, DB-5MS column) with a time resolution of ∼ 70 min.This instrument was equipped with a custom-built pre-concentration system described previously by Faiola et al. (2012Faiola et al. ( , 2014a)).The pre-concentration unit traps analytes on the Tenax © GR adsorbent and uses thermodesorption to inject compounds into the GC system.The FID is essentially a "carbon counter", meaning that the current produced from the detector is a function of the number of carbons in the molecule.Consequently, if the structure of the molecule is known, the concentration may be quantified using the effective carbon number concept with an upper-limit instrumental error of ±10 % (Faiola et al., 2012).Identifications of the following compounds could be made based on retention times determined using commercial standards: 3-carene, terpinolene, limonene, alphapinene, beta-pinene, alpha-terpinene, beta-myrcene, and ocymene.Molecular structures of other peaks were determined by interpreting the mass spectra acquired with the MS detector along with retention indices for monoterpenes.Integrated peak areas from the FID were converted to emission rates using Eq. ( 1): Here, E is the emission rate normalized to plant biomass in units of µg-C g −1 h −1 , A a and A s are the integrated FID peak areas of the analyte and internal standard, respectively, χ s is the mixing ratio of the internal standard (ppbV), N a and N s are the effective carbon numbers of the analyte and internal standard, respectively, M a is the analyte molar mass of carbon (g-C mol −1 ), F is the molar flow through the plant enclosure (mol-air h −1 ), 1000 is a conversion factor to obtain the appropriate units, and B is the biomass of needles in the plant enclosure (g).Effective carbon numbers were estimated using the effective carbon number concept (Faiola et al., 2012;Sternberg et al., 1962).Biomass was estimated by collecting and weighing a subset of needles from each tree after they were removed from the plant chamber.Needles were dried for a minimum of 24 h in an oven before weighing; furthermore, dry needle weight was scaled up to the tree level by estimating the number of needles on each tree.The GC-MS-FID used in this study was optimized to quantify monoterpenes.It can also quantitatively analyze aromatic emissions of a similar size.These emissions are dependent on temperature and were temperature normalized to 303 K using Eq. ( 2) (Guenther et al., 1993): where E(T ) is the measured emission rate at a measured temperature (T ), and E s is the standardized basal emission rate (BER) at standard temperature (T s ).The activity adjustment factor, β (K −1 ), was calculated for each experiment using measured emission rates between the post-acclimation period and treatment application.The number of points varied from experiment to experiment, but included a minimum of 24 h of measurements.Activity adjustment factors were calculated for terpenes and terpenoid aromatics separately because their chemical structures are slightly different and thus their chemical properties are expected to also differ.Results of these calculations are summarized in Table 2.The activity adjustment factors calculated here ranged from 0.15 to 0.59 K −1 , with most values ranging from 0.15 to 0.26 K −1 . Where a relationship between temperature and emission rate was observed and an activity adjustment factor could be calculated, nearly all values calculated for the terpenes were consistent with the ranges previously reported for coniferous tree species by Helmig et al. (2013) and Ortega et al. (2008) (0.08 to 0.28 K −1 ) (0.00 to 0.23 K −1 ).The one exception was the activity adjustment factor calculated for Pseudotsuga menziesii, which was much higher than any of the others, but which also had the highest temperature/emission rate (ER) correlation observed from any experiment (r 2 = 0.91 for monoterpenes and r 2 = 0.89 for aromatics).No aromatic compounds were observed above detection limit during the pre-treatment period for experiment PP-E1, so no activity adjustment factor could be calculated.Additionally, there was no relationship between temperature and emission rate during the pre-treatment period for the Abies grandis experiment.In this case, the average activity adjustment factor from the other experiments was used to temperature normalize the emissions for the Abies grandis experiment (excluding the apparent outlier from Pseudotsuga menziesii). In addition to monoterpenoids, this analytical system could detect and identify isoprene and some small OVOCs.However, these compounds had low breakthrough volumes for the Tenax © adsorbent used, and so they were not quanti- tatively captured on the adsorbent trap.Thus, absolute emission rates are not reported for those compounds.Instead, the relative measured value could be analyzed to look at trends in changing emissions from day to day.Where used, these emissions were normalized to their maximum measured emission rate and presented as a unitless value. Calculating atmospheric reactivity of BVOC emissions One potential impact of stress-induced changes in the monoterpenoid profile is on the oxidative reactivity of the BVOC emissions.To evaluate this, it is necessary to isolate the impact of the changing terpenoid profile on reactivity and exclude any impacts from changes to absolute emission rates.To do this, the sum total monoterpenoid mixing ratio was normalized to 1 ppbV and the mixing ratio of each individual monoterpenoid was calculated from the relative terpenoid contribution.This reactivity will be referred to as the concentration-normalized reactivity of the BVOC emission profile.The total mixing ratio value of 1 ppbV was selected as a reasonable approximation of summertime afternoon monoterpene mixing ratios in the canopy of a forest environment (Bryan et al., 2012;Nölscher et al., 2012).The compounds used in the reactivity calculations and their corresponding OH and O 3 rate constants are presented in Table 3. Reaction rate constants were obtained from experimental results in the literature where available (Atkinson et al., 1990;Calvert et al., 2000;Corchnoy and Atkinson, 1990;Gai et al., 2013;Reissell et al., 2001; United States Environmental Protection Agency, 2014) or were calculated using the method described in Calvert et al. (2000).Ring strain was ignored for the ozone reaction rate constants.Concentration-normalized OH and O 3 reactivity of plant BVOC emission profiles were calculated from the sum of the individual BVOC reactivities, which were calculated as the product of the reaction rate constant and the normalized mixing ratio.The resulting total OH and O 3 reactivity is the inverse of the OH and O 3 lifetime. Only those compounds listed in Table 3 were included in the calculation.This list includes all the major VOCs that were identified in these experiments.ting.Then, the stress response from each tree type is described separately, including changes to the daily average monoterpenoid profiles and temporal trends in absolute emission rates.A summary of the main compounds that were affected by the stress treatment from each tree is presented.Finally, the concentration-normalized OH and O 3 reactivity are presented to investigate the impact of changing the BVOC profile before and after stress treatment. Pre-treatment monoterpene profiles Monoterpenoids were the dominant biogenic emissions that were quantitatively measured from each tree type in this study.These compounds have been the focus of numerous field measurements using the same species used in these experiments.Figure 1 summarizes the pre-treatment monoterpene profile for each experiment in this study.Values are presented as the percent of total monoterpenoid emission rates for each experiment.The same results are provided in absolute emission rates in Table 4.The profiles were calculated using all data from the end of the acclimation period until immediately before the stress treatment was applied.This time period varied from experiment to experiment but always included a minimum of 24 h of measurements.In total, 32 monoterpenoid chemical species were observed prior to treatment, including two oxygenated monoterpenes, camphor, and 1,8-cineol.Minor constituents were summed for inclusion in the profile.This group includes the following compounds: santene, 2-bornene, alpha-fenchene, 2,4-thujadiene, beta-terpinene, 2-carene, alpha-phellandrene, alpha-terpinene, gamma-terpinene, alpha-thujene, the aromatic cymenene isomers, acetophenone, two unidentified monoterpenes, and four unidentified aromatic compounds.Together, this category accounted for < 10 % of all pretreatment monoterpenoid emissions.Toluene was also measured during some experiments but was not a major component and was not included in this analysis. The pre-treatment monoterpene profile varied between the tree species (Fig. 1).However, despite differences in their distribution, the same seven compounds made up greater than 75 % of all monoterpene emissions from all trees: alphapinene, limonene, 3-carene, beta-pinene, beta-myrcene, camphene, and beta-phellandrene.For the two sets of Picea pungens experiments, the pre-treatment profiles were substantially different, even though the same four saplings were used in each of the three experiments.Picea pungens emissions in May (PP-E1) were dominated by alpha-pinene and limonene, while in July (PP-E2 and PP-C) they were dominated by limonene and beta-myrcene.Each of these profiles were consistent with previous measurements made in a field setting.The Picea pungens monoterpene profile presented by Helmig et al. (2013) The left axis shows the proportion of each compound emitted as a percent of total monoterpenoids.The diamonds associated with the right axis show the average pre-treatment basal emission rate (BER) of total monoterpenes normalized to a temperature of 303 K in units of µg-C g −1 h −1 .The x axis label is the experiment ID (Table 1). The average BER was calculating using all data from the end of the acclimation period until immediately before the stress treatment was applied (> 24 h of measurements).The error bars represent the standard deviation of the averaged value. 1,8-cineol in the July experiments vs. the May experiment, which Helmig et al. (2013) also described.The Picea pungens monoterpenoid BER in this study ranged from 0.29 to 0.81 µg-C g −1 h −1 (0.32-0.92 µg g −1 h −1 ).Previous reports ranged from < 0.10 to 1.45 µg g −1 h −1 throughout the year, and during the months of May-July (the time period when our experiments were performed) the reported BER range was 0.87-1.45µg g −1 h −1 (Helmig et al., 2013).Thus, the Picea pungens BER in our experiments was on the lower end of what has been reported from Picea pungens in the field.The monoterpenoid profile of the Rocky Mountain bristlecone pine (Pinus aristata) has not been previously reported to our knowledge.A profile of the Great Basin bristlecone pine (Pinus longaeva) was presented by Helmig et al. (2013), and is used here for comparison.Both profiles were dominated by 3-carene, alpha-pinene, and beta-pinene.Within this study, the two Pinus aristata experiments exhibited nearly identical pre-treatment monoterpene emission profiles.These measurements were taken within 2 weeks of one another.The Pinus aristata monoterpenoid BER was 0.62-0.75µg-C g −1 h −1 (0.70-0.85 µg g −1 h −1 ), which is on the higher Table 4. Summary of the temperature-normalized pre-treatment emission rates for the dominant compound emissions.Units are emission rates in µg-C g −1 h −1 normalized to 303 K.A dash indicates the compound was not detected and bdl indicates the compound was detected but it was below the calculated detection limit for quantification (detection limit = 0.003 µg-C g −1 h −1 ).The average sum basal emission rate (BER) is provided at the bottom of the table for each experiment; σ denotes the standard deviation of the measurements used to calculate the pre-treatment average.The Abies grandis, Pseudotsuga menziesii, and Thuja plicata monoterpene profiles each differed from what has been reported previously.The profile from Abies grandis in this study was dominated by beta-pinene, but no beta-pinene was observed by Ortega et al. (2008).This difference could be explained by natural genotypic variation because Ortega et al. (2008) also observed natural variation in the constitutive BVOC profiles between plants of the same tree species.However, the Abies grandis monoterpenoid pre-treatment BER measured in our experiment was 12.67 µg-C g −1 h −1 , substantially higher than any other pre-treatment monoterpenoid BER observed in this study and more than an order of magnitude greater than that reported by Ortega et al. (2008) for the same tree species.These high emission rates could suggest the Abies grandis saplings were likely exhibiting a stress response prior to treatment. PP-E1 For Pseudotsuga menziesii, the dominant monoterpene emission measured in this study was beta-phellandrene (40 % of all monoterpenoid emissions).Helmig et al. (2013) observed alpha-pinene and beta-pinene comprising more than 50 % of all Pseudotsuga menziesii monoterpenoid emissions throughout an entire year of measurements, which was consistent with the profile presented in Geron et al. (2000).However, Ortega et al. (2008) observed variability in Pseudotsuga menziesii monoterpene profiles in the field, reporting that limonene and camphene were the dominant emissions during one set of measurements, while sabinene and alphapinene were for another.Furthermore, beta-pinene emissions were measured for one reported BVOC profile by Ortega et al. (2008) but not for the other.Thus, the pre-treatment profile in this laboratory study could still be representative of a natural baseline condition.The pre-treatment Pseudotsuga menziesii BER measured in our laboratory chamber was 3.66 µg-C g −1 h −1 .This was the second-highest observed BER value prior to treatment, and is consistent with previous reports where values as high as 3.40 µg-C g −1 h −1 were measured from Pseudotsuga menziesii branch enclosures by Ortega et al. (2008).However, our laboratory experiment was conducted in September when seasonal reports of emissions have shown decreasing emission trends.For example, the highest BER reported in the field by Helmig et al. (2013) was 2.51 µg-C g −1 h −1 in June, but they reported that by September the monoterpenoid BER had dropped back down to 0.12 µg-C g −1 h −1 .Thus, the BERs in our experiment were at the upper range of what would be expected in the natural environment from Pseudotsuga menziesii at this time of year. Thuja plicata monoterpenoid emissions in this study were dominated by beta-pinene, camphene, and betaphellandrene, whereas Ortega et al. (2008) found that 61 % of all monoterpenoid emissions were composed of the oxygenated compounds alpha-and beta-thujone.We did not observe any thujone emissions throughout the measurement period.The monoterpenoid pre-treatment BER from Thuja plicata was the lowest we observed from any species at 0.28 µg-C g −1 h −1 .This was consistent with the Thuja plicata BER reported by Ortega et al. (2008) Blue spruce (Picea pungens) Three experiments were performed using Picea pungens saplings, two with MeJA treatments and one negative control. All three experiments were performed using the same four saplings, and the negative control experiment was performed the week prior to the July MeJA-treatment experiment.The two MeJA-treatment experiments did not produce consistent results.To illustrate this, a plot of the total monoterpenoid BER vs. elapsed time since treatment is shown in Fig. 2. The first treatment experiment performed in May exhibited a clear stress response where monoterpene emissions increased from 0.29 ± 0.2 to 23.27 ± 2.15 µg-C g −1 h −1 .This represents an 80-fold increase after treatment.Emissions remained elevated above pre-treatment values over the next 50 h.In stark contrast, the monoterpene emissions from the July MeJA experiment did not demonstrate a significantly different response to stress than did the negative control. There was a small increase in emissions for both PP-N and PP-E2 on the day of treatment.The short lived, slight emissions increase observed in these experiments could possibly be the result of an abiotic surface adsorption disruption effect; water displaces organic molecules previously adsorbed to the needle surfaces and produces a burst in measured emissions.This phenomenon has been observed in a natural forest environment where bursts of VOC emission were observed following rain (in a natural forest setting) or water application (in a laboratory setting) (Faiola et al., 2014a;Greenberg et al., 2012;Warneke et al., 1999).This would suggest that there was no significant stress treatment effect and that the small increase in some emissions observed on the treatment day could be a function of the treatment method itself rather than an actual stress response.The difference in these results was also apparent when the complete BVOC profiles were examined (Fig. 3).These values are the average daytime emissions (6:00 to 18:00 LT).To simplify the presentation, BVOCs that individually constituted less than 1 % of all monoterpenoid emissions were summed and presented in the "Other" category.The pretreatment aromatic emissions for the PP-E1 experiment were too low to calculate an aromatic activity adjustment factor, so the activity adjustment factor for aromatics calculated from PP-E2 data was used to normalize aromatic emission rates for both experiments. In PP-E1, the maximum stress response for all classes of compounds was observed the day after treatment (day +1).The highest-emitted monoterpene before treatment was alpha-pinene (> 40 % of all MT emissions, Fig. 1).After treatment, limonene, beta-myrcene, and 1,8-cineol dominated the emission profile.Limonene and beta-myrcene were constitutive emissions that were stimulated more than other constitutive emissions after treatment.In addition to enhancing constitutive emissions, the stress treatment also induced many new monoterpenoid emissions, including alphaphellandrene, alpha-terpinene, 1,8-cineol, ocimene, gammaterpinene, and terpinolene.Some of these induced compounds did not contribute significantly to the overall posttreatment emissions and were thus lumped into the other category, but they are worth noting because they were only observed after treatment had been applied.Specifically, 1,8cineol and ocimene were emitted at rates well over 2 orders of magnitude higher than the detection limit after treatment, i.e., above the 80-fold increase in total emissions, which suggests these emissions were truly induced and not just emitted at rates below the detection limit prior to treatment.Negligible amounts of aromatic compounds were observed before treatment.After treatment, even though aromatics still made up a small relative proportion of overall emissions, the aromatic emissions (predominantly p-cymene) increased significantly to 0.5 µg-C g −1 h −1 , which was similar to the pre-treatment sum monoterpenoid BERs for many of the tree species presented in Fig. 1.Emissions of all classes of compounds began to decrease again within 48 h after treatment, but still remained elevated relative to pre-treatment values when measurements ceased. In contrast to the May experiment, in the July Picea pungens experiment the monoterpenoid average profile did not significantly change after treatment (Fig. 3).This could be due to seasonal differences in the sensitivity of Picea pungens to herbivore treatment.This has been observed in other coniferous plant species.For example, monoterpene synthesis in Pinus sylvestris is more responsive to plant stressors during the spring when shoots are actively growing (Bäck et al., 2005).In the Picea pungens experiment presented here, there were small increases in terpinolene and ocimene emissions on the day of treatment, but they quickly returned to pre-treatment levels.Furthermore, results from the May experiment suggested that 1,8-cineol was a stress-induced compound that was only observed after treatment, but this same compound constituted a significant proportion of the pretreatment BVOC emission profile in the July experiment.This could be a natural seasonal effect; field measurements have demonstrated seasonal changes in 1,8-cineol emission rates from Picea pungens (Helmig et al., 2013).However, it is also possible that the 1,8-cineol emission rate fluctuations observed in the field were due to the presence of some natural stressor.Thus, the pre-treatment profile for the July experiment could indicate that the trees' metabolic stress pathways had been activated prior to experimental treatment.This hypothesis is further supported by the higher percentage of beta-myrcene and limonene emissions present in the July pre-treatment profile that more closely resemble the posttreatment stress profile from the May experiment.This combined with the low emission rate values could suggest that the trees had been exposed to an external stressor for an adequate length of time to cause the plant to begin shutting down metabolic processes.If this was the case, the application of an additional stress treatment did not produce a stress response under those conditions. Averaging emission rates over each day provides a clean picture of the overall VOC profiles, but any patterned variability that may occur through the day would be hidden by this approach.Another way to investigate changing VOC profiles is to compare the emission rate data for different compounds to evaluate their covariance.If paired compounds co-vary, then their relative emissions are consistent over time.If their correlation is weaker, it suggests that the profile is changing, possibly due to differences in the factors regulating the compounds' emissions. Constitutive emissions co-varied throughout the negative control experiment (PP-N).Emission rates of beta-myrcene, alpha-pinene, and beta-phellandrene were plotted against limonene emissions and shown in Fig. 4. Limonene was used as the basis for comparison because it was the dominant constitutively emitted compound (Fig. 1).Measurements from the first 36 h while the plants were acclimat- ing to the plant chamber were excluded from the analysis. Correlations between these three constitutively emitted compounds and limonene were high with r 2 values ranging from 0.87 to 0.98.This was also true for the other compounds' emissions, with emission rate correlation coefficients with limonene ranging between 0.85 and 0.96.Camphor was the exception; the correlation between camphor and limonene emissions was 0.35. In the May MeJA experiment (PP-E1), the dominant pretreatment constitutive emission was alpha-pinene but after treatment, the major emissions were limonene, betamyrcene, and 1,8-cineol (Fig. 3).For this experiment, it was informative to look at both the time series of emission rates as well as the covariance between emission rates of difference compounds.A time series of the emission rates after treatment for a subset of the compounds is shown in Fig. 5. Immediately after treatment on 15 May, 2013 at 11:40 LT, alpha-pinene was still the dominant terpene emitted.However, emissions of limonene and beta-myrcene began to increase quickly and had exceeded alpha-pinene emissions by later that evening.Emissions of 1,8-cineol did not begin to increase until 17:00 LT.After that, they continued to increase and surpassed alpha-pinene emissions early the following morning.Beta-phellandrene is also shown on the figure to provide an example of a less dominant emission trend.It immediately began to increase after treatment but never exceeded alpha-pinene emissions.The emission trends of betamyrcene, limonene, 1,8-cineol, and beta-phellandrene are in contrast to the trend in alpha-pinene emission rates.Alphapinene was not impacted by the treatment and maintained a stable emission rate throughout the evening while emission rates of other compounds steadily increased. Western redcedar (Thuja plicata) The VOC daily profiles for the Thuja plicata MeJA experiment are summarized in Fig. 6.For this experiment, nine small saplings were kept in the plant chamber for 6 days before applying treatment, and were removed from the chamber the day after treatment.However, for this group of plants there was an exceptionally strong emission response that continued to increase throughout the night following treatment.Consequently, day +1/2 has been included on the chart to capture peak emission response, and refers to the nighttime period that occurred half a day after treatment application.The pre-treatment and post-treatment profiles were plotted separately due to the drastic increase in the emission rate; monoterpene BER increased from an average value of 0.28 ± 0.02 µg-C g −1 h −1 on days −6 to −4 to a maximum average value of 11.88 ± 0.18 µg-C g −1 h −1 during the evening after treatment.This is a 42-fold increase in monoterpenoid BER.Terpinolene, beta-myrcene, and the cymene isomers increased most substantially and dominated the monoterpene profile after treatment. The post-treatment temporal emissions trends for the Thuja plicata experiment exhibited a pattern that was not observed for other trees species.Figure 7 shows the monoterpenoid BER time series immediately following treatment.In Fig. 7, the treatment was applied on 22 September at 08:30 LT, and emissions of all compounds began to increase by 13:00 LT the same day.The emissions of nearly all compounds continued to rise or stabilized at an elevated emission rate for the remainder of the measurement period until 23 September at 05:00 LT when measurements were stopped.However, beta-pinene did not follow this trend; instead, betapinene emissions immediately increased after treatment, but began to decrease a few hours later, starting at 15:00 LT on the treatment day.It was the only compound to exhibit this emission pattern. Terpinolene also demonstrated a slightly different emission pattern from most other monoterpenes.This is evident from the linear regression results presented in Table 5. Terpinolene reached a maximum emission rate on the evening of the treatment day at 20:30 LT (Fig. 7).Afterwards it began to decrease slowly.The only other compound to exhibit this emission trend was ocimene, which had a linear regression correlation with terpinolene emissions of 0.86.Most other compounds continued to increase throughout the night.Thus, most compound emission rates were highly correlated with limonene emissions, which exhibited this continually increasing emission trend.Ten compounds were highly correlated with limonene emissions with r 2 > 0.90 (Table 5). Beta-phellandrene and gamma-terpinene were well correlated with both limonene and terpinolene with r 2 ≥ 0.80.Their emission rates stabilized more quickly than most other compounds during the night.They were best correlated with one another with an r 2 = 0.96.This could suggest four different types of emission responses (1) quick increase followed by a slow decrease within 10 h of treatment similar to terpinolene, (2) quick increase followed by a rapid decrease similar to beta-pinene, (3) long-term increase throughout the night similar to limonene, and (4) an increase followed by stabilization within ∼ 12 h of treatment similar to beta-phellandrene. Monoterpenoid BER values for Thuja plicata were the lowest pre-treatment emissions that were measured from all trees in this study.After treatment had been applied, monoterpenoid BERs increased to the third-highest emission rates measured throughout the experiments.This suggests that stress exposure in natural environments could turn normally low-emitting trees into high emitters that could contribute substantially to the net ecosystem BVOC flux.This should be considered in future experimental designs, where it may be tempting to limit tree species representation to only the known highest BVOC emitters in a region because there may be some tree species that are only high emitters under stressed conditions. Douglas-fir (Pseudotsuga menziesii) The daily average VOC emission profile from Pseudotsuga menziesii is shown in Fig. 8.Some of the minor constituents (< 1 % of BER) have been grouped together within the "Other" category to simplify the presentation.For this experiment, 2 days of measurements were collected prior to treatment after plants had acclimated to the chamber.Following treatment, BVOC emission rates were monitored for another 4 days.Absolute monoterpenoid BERs approximately doubled on the day of treatment.They increased from 3.66 ± 0.88 to 7.34 ± 1.04 µg-C g −1 h −1 .Emissions then remained 34 % higher, on average, than baseline emissions for the following 4 days.Aromatics (predominantly o-cymene) comprised more than 10 % of the total Pseudotsuga menziesii VOC emissions even before treatment, and thus could be significant contributors to SOA formation in natural forest environments.Emissions of alphapinene, beta-pinene, and 3-carene increased most after treatment relative to the other constitutive monoterpenes.Alphapinene emissions increased by ∼ 100 %, beta-pinene emissions by ∼ 570 %, and 3-carene emissions by ∼ 640 %.This effect was sustained until measurements ceased 4 days after treatment.One of these stress-enhanced compounds, betapinene, co-varied with the dominant constitutive emission, beta-phellandrene, prior to treatment (r 2 = 0.89), but was de-coupled from beta-phellandrene emissions after treatment (r 2 = 0.48).However, nearly all other compounds continued to co-vary with beta-phellandrene emissions from day +1 to day +4 after treatment.Emissions from beta-myrcene, the cymene isomers, alpha-pinene, limonene, ocimene, and terpinolene all had linear regression results of r 2 > 0.90 vs. betaphellandrene.3-Carene emissions did not co-vary with any other compound emissions. The overall stress response exhibited by Pseudotsuga menziesii was not as dramatic as the 80-fold increase observed during experiment PP-E1 or the 42-fold increase observed during experiment TP-E.There was also no single stressenhanced compound that completely dominated the posttreatment emission profile as terpinolene did during experiment TP-E.Despite all this, the three most stress-enhanced compounds (alpha-pinene, beta-pinene, and 3-carene) did contribute significantly to the overall BVOC emissions during this experiment, which were substantial.Pre-treatment, the monoterpenoid BERs for Pseudotsuga menziesii were the second-highest pre-treatment values measured in this study (Fig. 1), with a daytime average pre-treatment monoterpenoid BER of 3.39 ± 0.01 µg-C g −1 h −1 .The daytime average post-treatment BER was 5.46 ± 0.37 µg-C g −1 h −1 .This is only a modest increase in overall emission rates relative to some of the other experiments.However, of the 2.06 µg-C g −1 h −1 total increase in BER, 1.75 µg-C g −1 h −1 was due to the increase in just the three most stress-enhanced compounds: alpha-pinene, beta-pinene, and 3-carene (85 % of the total increase).The post-treatment average BER of these three compounds was 2.48 ±0.15 µg-C g −1 h −1 , 73 % of the total monoterpenoid pre-treatment BER.Thus, these stimulated monoterpenes can significantly contribute to total BVOC emissions.This is important because different monoterpenes have widely varying chemical reactivity and SOA formation potential (Atkinson and Arey, 1998;Griffin et al., 1999). Grand fir (Abies grandis) As shown in Fig. 1, the pre-treatment monoterpene BER for the grand fir experiment was greater than for any other experiment, and was much greater than what had been previously reported elsewhere.This suggests that these trees had been exposed to some unknown external stress while being stored outdoors prior to use.To investigate this, we examined the entire BVOC profile throughout the measurement period (Fig. 9).All monoterpenoid emissions steadily decreased from day −2 to day 0. It is possible that the trees were still acclimating to the plant chamber on day −2, but they should have been well acclimated by day −1 because trees take 12-36 h to acclimate to the plant chamber (having been transported to the chamber on day −3).The observed steady decrease from day to day could be indicative of the hypothesized unknown stress effect waning once the trees were brought into the laboratory.Laboratory notes on tree appearance for this experiment indicate that the trees had a number of dry, orange-red needles when they were transported on 23 June 2013.Another note from 28 June 2013 described large clumps of needles dropping from the trees at the slightest touch during watering.The trees were kept well watered at the greenhouse and in the laboratory chamber and outdoor temperatures were normal for the area, so we do not believe that the needle damage was the consequence of drought or temperature stress.However, this possibility cannot be ruled out completely.Alternatively, the observed effects may have been the result of an unseen herbivore or pathogen that was not detected prior to the experiment.Despite the possible presence of an uncontrolled stressor, the experimental MeJA stress treatment did still have a small effect on BVOC emission rates and profile (Fig. 9).This effect was not immediate; emissions continued their decreasing trend on day 0, but then increased slightly on day +1.The BVOC profile was altered both by the induction of emissions of new compounds and by the alteration of the distribution of constitutive emissions.1,8-Cineol and, to a much lesser extent, p-allylanisole were induced.The former is an oxygenated monoterpene and the latter is a phenylpropanoid produced from the shikhimic acid pathway ( , 2006).These emissions were not observed until 6 h after treatment for 1,8-cineol and 22 h after treatment for pallylanisole.Small OVOCs and unidentified compounds exhibited maximum emissions the day following stress treatment and may also have been induced by the stress treatment.Similar to the other stress-induced and stress-enhanced compounds, they exhibited a delayed response in emissions.These small OVOCs include alcohols, ketones, and aldehydes that have less than eight carbon atoms including small 5carbon to 6-carbon OVOCs produced from the lipoxygenase biochemical pathway (Connor et al., 2008;Maffei, 2010). The constitutive monoterpene emission profile also changed.For the first 3 days, the terpene profile was dominated by beta-pinene, beta-phellandrene, and alpha-pinene, and their relative contribution to total emissions did not vary significantly.After the MeJA treatment, beta-pinene emissions continued to decrease as they had been for the previous 3 days, but limonene, beta-myrcene, beta-phellandrene, terpinolene, and alpha-pinene all increased.Increases in these compounds were observed 6 h after treatment, similar to when the induced compound, 1,8-cineol, was first observed.Prior to treatment, constitutive emissions of alpha-pinene, limonene, and terpinolene all co-varied with the dominant constitutive emission, beta-pinene, with all r 2 values greater than 0.90 (Fig. 10,left).Two separate bursts in emissions occurred 24 h apart from one another that produced the threehighest points on the plots (two measurements during one burst and one measurement during the other burst).With those points removed, alpha-pinene and limonene were still well correlated with beta-pinene with r 2 values of 0.97 and 0.89, respectively.The terpinolene r 2 reduced to 0.52 when the two emission bursts were excluded.Other major constitutive emissions also co-varied with beta-pinene prior to treatment but were not shown in the figure; camphene, betaphellandrene, p-cymene, and beta-myrcene also co-varied with beta-pinene prior to treatment with r 2 values ranging from 0.94 to 0.99.However, after treatment, beta-pinene no longer co-varied with alpha-pinene, limonene, or terpinolene with r 2 values of 0.53, 0.25, and 0.12, respectively (Fig. 10, right).Thus, even with the emission bursts removed pre-treatment, all r 2 values decreased relative to the posttreatment correlations.Furthermore, all of the other most highly enhanced constitutive compounds except for betaphellandrene were well correlated with limonene after treatment with r 2 values > 0.80 (not shown).The MeJA stress treatment de-coupled the dominant constitutive emissions from beta-pinene, which was not enhanced by the stress, while most of the compounds enhanced by the treatment continued to co-vary.1,8-Cineol, the induced emission, was not well correlated with the most enhanced constitutive emission, limonene (r 2 = 0.18). Bristlecone pine (Pinus aristata) A time series of the summed monoterpenoid BERs are presented in Fig. 11.There was a large spike in emissions immediately following the MeJA treatment where monoterpenoid emissions increased from 0.54 to 12.52 µg-C g −1 h −1 .The negative control experiment also demonstrated a slight increase in emissions, but to a much lesser extent than the MeJA experiment; monoterpenoid emissions increased from 0.81 to 2.68 µg-C g −1 h −1 .The emissions increase was short lived for both experiments and the emissions trend started to reverse within just a few hours following treatment. The monoterpene profiles for the days before (day −1) and after (day +1) treatment are shown in Fig. 12.The total emissions were slightly reduced for the MeJA experiment on the day following treatment, but not substantially so, and the monoterpenoid profile did not change.The negative control BER and emission profile were similar before and after spraying the trees with water. Major monoterpene emissions were plotted against the emission rates of the dominant monoterpene throughout these experiments, 3-carene, in Fig. 13.Both the negative control and MeJA experiment demonstrated high correlations (r 2 > 0.9) for all monoterpene emissions relative to 3-carene.Beta-pinene, beta-phellandrene, and terpinolene are shown in the figure for illustration, and this was also true for alphapinene, o-cymene, p-cymene, limonene, camphene, betamyrcene, and m-cymene.This indicates that the monoterpene profile did not change substantially during either experiment. Summary of emission rate changes A summary of the change in emission rates after stress treatment for some of the key compounds is summarized for each experiment where a plant stress response was observed (Fig. 14).Note the difference in the y axis scale for each experiment because the overall change in emission rates varied between plant types.For the Thuja plicata experiment, the delta value was calculated from the day +1/2 post-treatment value minus the baseline daily average from day −4 to day −6.This is a conservative estimate of emissions changes because all emissions decreased during the 2 days prior to treatment (days −1 and −2) but these lower emission values were not used in the calculation.For the Picea pungens experiment, the delta BER was calculated by subtracting the average daily value on day −1 from day +1.The maximum response was observed on day +1 and day −2 was excluded because the plants may have still been acclimating to the chamber.For the Pseudotsuga menziesii experiments, the delta BER was calculated by subtracting the average daily values on day −2 and day −1 from the average daily values on days +1 to +4.For the Abies grandis experiment, the delta BER was calculated as the difference between day 0 and day +1.The compounds that were most impacted by the stress treatment were highly variable between tree types.In the Thuja plicata experiment, the two monoterpenes that increased most were terpinolene and beta-myrcene.The emissions of these compounds increased by a combined 7.04 µg-C g −1 h −1 .This represents just over 80 % of the total increase in monoterpene BER with terpinolene alone contributing to just over 60 % of the total increase.The cymene isomers also exhibited a significant emission increase.The only other experiment where all three cymene isomers were measured was in the Pseudotsuga menziesii experiment.In this case, all cymene isomers increased, but to a lesser extent than during the Thuja plicata experiment.The most stress-enhanced compounds in the Pseudotsuga menziesii experiment were alpha-pinene, beta-pinene and 3-carene.1,8-Cineol was identified as an important stress-enhanced or stress-stimulated compound in the Picea pungens and Abies grandis experiments but was never emitted from the second experiment (PP-E2), the results could indicate there is some breaking point where the plants simply do not respond to an additional stressor.These results would be in stark contrast to the Abies grandis stress response.The Abies grandis results suggest that despite the possible presence of an unknown stress prior to treatment, the simulated herbivory stress still caused additional changes to the emission profile.Thus, the presence of one stressor does not necessarily prevent a tree from responding to another stressor at the same time, and it is possible the effects of the two stressors could be additive.The response of the Thuja plicata emissions to the stress treatment can also provide valuable insight. Even though the pre-treatment emissions from the Thuja plicata plants were the lowest we measured from all the exper-iments, the post-treatment emission rates were substantial.This suggests that even naturally low-emitting species that would not contribute significantly to total forest BVOC flux under baseline conditions could be major sources of BVOC emissions under stressed conditions in a changing climate.Consequently, future surveys of BVOC emitters should not be limited to only the highest BVOC emitters in a region because this could change as global change stressors intensify.Finally, the near lack of any long-term response from Pinus aristata could indicate that some trees are more resistant to certain types of stress exposure than others.On the other hand, it is possible that, like Picea pungens, the Pinus aristata could demonstrate a completely different stress response depending on the season.The Pinus aristata experiments were conducted in May when pre-treatment emissions were low and the plants may have still been coming out of winter dormancy.This could have contributed to their apparent resistance to the treatment. Implications for BVOC atmospheric reactivity The MeJA stress treatment significantly changed the BVOC profile in many of the experiments.As discussed in the previous section, the specific compounds that were impacted by the treatment were highly variable between the different plant types.Consequently, the overall implications for atmospheric reactivity for the different plant types was also highly variable because different monoterpenoids have widely varying atmospheric reactivity (see Table 3).The pre-and posttreatment BVOC profile for each experiment was used to calculate the concentration-normalized OH and O 3 reactivity by normalizing the relative contribution of each monoterpenoid to a sum monoterpenoid mixing ratio of 1 ppbV.The goal was to isolate the impact on reactivity due to changes in the BVOC profile only.Thus, the focus of this analysis was to investigate the change to the concentration-normalized oxidant reactivity value rather than the absolute pre-and post- treatment values; the reactivity results are presented in Table 6. For all experiments where a change in concentrationnormalized reactivity was observed, the O 3 reactivity was more significantly affected than the OH reactivity.The three experiments that demonstrated the largest changes were TP-E, PP-E1, and AG-E.For each of these experiments, the stress-induced changes to the BVOC profile increased both the OH and O 3 concentration-normalized reactivity.The normalized OH reactivity of the Thuja plicata emission profile (TP-E) approximately doubled with an increase from 2.21 to 4.57 s −1 (106.8 % increase).This corresponds to a decrease in OH lifetime from 0.45 to 0.22 s.The normalized O 3 reactivity increased by nearly an order of magnitude from 3.53 × 10 −6 s −1 to 30.3 × 10 −6 s −1 (758.4 % increase).This corresponds to a decrease in O 3 lifetime from 3.3 days to 9.2 h.This is primarily due to the large increase in the relative amount of terpinolene, which has a high ozone reaction rate constant relative to most other monoterpenoids (Table 3).The normalized OH reactivity of the Picea pungens emission profile during the first experiment (PP-E1) increased from 2.43 to 3.50 s −1 (44 % increase).This corresponds to a decrease in the OH lifetime from 0.41 to 0.29 s.The normalized O 3 reactivity increased from 2.99 × 10 −6 s −1 to 10.7 × 10 −6 s −1 (257.9 % increase) corresponding to a decrease in O 3 lifetime from 3.9 days to 1.1 days.The normalized OH reactivity of the Abies grandis emissions increased by a small amount from 2.43 to 2.74 s −1 (12.8 % increase) corresponding to a decrease in OH lifetime from 0.41 s to 0.36 s.However, the normalized O 3 reactivity significantly increased from 3.46 × 10 −6 s −1 to 7.40 × 10 −6 s −1 (113.9 % increase) corresponding to a decrease in O 3 lifetime from 3.3 days to 1.6 days. The Pinus aristata experiments (PA-C and PA-E) demonstrated very little change to the BVOC profile (see Sect. 3.6).For the negative control experiment (PA-C), the concentration-normalized reactivity results were consistent with no BVOC profile change; a 0 % change was observed for OH reactivity and a 0.4 % change was observed for O 3 reactivity.The normalized OH reactivity increased slightly af-ter treatment during the PA-E experiment with an increase of 8.8 %.However, the PA-E normalized O 3 reactivity increased significantly by 69.6 % after MeJA treatment despite only minor changes to the BVOC profile (see Fig. 12).These results demonstrate that even small changes to the BVOC profile can have significant impacts on the overall atmospheric reactivity of the BVOC emissions. Concentration-normalized reactivity of emissions from Pseudotsuga menziesii decreased slightly after treatment.The normalized OH reactivity decreased from 2.75 to 2.44 s −1 (decrease of 11.3 %) corresponding to a small increase in OH lifetime from 0.36 to 0.40 s.The normalized O 3 reactivity decreased from 3.37 × 10 −6 s −1 to 2.49 × 10 −6 s −1 (decrease of 26.1 %) corresponding to an increase in O 3 lifetime from 3.4 days to 4.6 days.This was due to an increase in the relative amount of beta-pinene and 3carene emissions.Both of these compounds have reduced oxidant reactivity relative to other monoterpenoid compounds emitted in higher amounts prior to treatment (Table 3). Conclusions While many uncertainties remain regarding the impacts of herbivory stress on plant BVOC emissions, it is clear that plant responses are highly variable.Emissions of different compounds were impacted by the stress treatment for different tree types.The compounds that tended to be most affected by the stress treatment were alpha-pinene, betapinene, beta-myrcene, 3-carene, limonene, 1,8-cineol, terpinolene, and the cymene isomers.Aromatic cymenes sometimes contributed significantly to the emission profile pretreatment (i.e.Pseudotsuga menziesii), and often increased significantly post-treatment.These aromatic compounds are often not considered to be major precursors of biogenic SOA, but the emission rates observed in these experiments suggest they could be significant contributors to SOA formation in forests. Four possible plant herbivory response patterns were observed in these experiments: (2) after long-term exposure to one stressor, plant emissions decrease overall and do not respond to additional stressors; (3) alternatively, multiple stressors can be additive, perhaps if the second stressor is applied before the first stressor depletes terpene pools and initiates metabolic shutdown; and (4) herbivory stress could turn naturally low-emitting plants in a region to high emitters that would need to be considered in future climate scenarios with increased herbivory.Stress-induced changes to the BVOC emission profile can result in significant changes to the concentration-normalized oxidant reactivity of plant emissions in the atmosphere.Increases in reactivity as high as 758.4 % with O 3 and 106.8 % with OH were observed during the Thuja plicata experiment (TP-E).Furthermore, even small changes to the BVOC profile during the Pinus aristata MeJA experiment (PA-E) increased O 3 reactivity by 69.6 %.These results highlight the importance of making quantitative, compound-specific BVOC emission rate measurements to understand the potential impact of stress-induced emissions on atmospheric chemistry.Changes in the oxidant reactivity of BVOC emissions have significant implications for the production of pollutants like ozone and secondary organic aerosol in forest environments. Many questions still need to be addressed before stress impacts on BVOC emissions can be incorporated into emissions models.Future research needs to address the seasonality influence on plant susceptibility to herbivory stress.Additionally, the interaction between multiple stressors needs to be addressed because in the natural environment it is likely that plants are being exposed to multiple stressors more often than a single stressor in isolation.A broad survey of plant types should be used in these experiments to investigate which plants could become dominant BVOC emitters under future climate scenarios.Finally, all of these questions need to be asked regarding other types of plant stress including drought, thermal stress, ozone stress, and using different types of real herbivores and pathogens. Figure 2 . Figure 2. A summary of monoterpenoid emissions from all three Picea pungens experiment.The only experiment to exhibit a clear stress effect on monoterpenoid emission rates following treatment was the first methyl jasmonate (MeJA) experiment performed in May (PP-E1). Figure 3 . Figure 3. Summary of monoterpenoid profile for the two Picea pungens methyl jasmonate (MeJA) experiments.The x axis denotes the day relative to treatment, where treatment was performed on day 0. The y axis is the monoterpenoid (MT) basal emission rate normalized to 303 K. Results from the methyl jasmonate (MeJA) experiment performed in May are presented in the top plot and the results from the MeJA experiment performed in July are presented in the bottom plot.Note the difference in y axis scale for the top plot vs. the bottom plot.The inset in the top plot is provided as an enlargement of the profiles for days −2, −1, and 0 for experiment PP-E1. Figure 6 . Figure 6.Emission profile of emissions from Thuja plicata during methyl jasmonate (MeJA) experiment TP-E.The x axis denotes the day relative to treatment application.The y axis shows the monoterpenoid BER normalized to 303 K.Note the drastic scale change between the pre-and post-treatment y axes.The insert shows an enlarged view of the first 6 days to allow better visualization of the pre-treatment period. Figure 7 . Figure 7. Time series of monoterpene emission rates from Thuja plicata.The x axis shows the elapsed time since treatment application in hours.Alternating shaded and unshaded regions demonstrate when the light above the plant enclosure was turned off and on, respectively. Figure 8 . Figure 8. Douglas-fir VOC profile.The x axis denotes the day relative to treatment application.The y axis is the monoterpenoid basal emission rate normalized to 303 K. Figure 9 . Figure9.Grand fir BVOC profile.The x axis denotes the day relative to treatment application.The top panel summarizes the monoterpenoid emissions where the y axis is the monoterpenoid basal emission rate normalized to 303 K.The bottom panel summarizes the emissions of small oxy-VOCs and other unidentified compounds where the y axis is the fraction of the emission rate relative to the maximum measured value. 10.Figure 11 . Figure 11.Results from two Pinus aristata experiments.Shown above is the time series of the sum monoterpenoid basal emission rates normalized to 303 K as a function of elapsed time since treatment application for the methyl jasmonate (MeJA) experiment (PA-E) and the negative control experiment (PA-C). Figure 12 .Figure 13 . Figure12.The Pinus aristata BVOC profile the day before treatment and the day after treatment for both the methyl jasmonate (MeJA) experiment (PA-E) and the negative control experiment (PA-C).The x axis denotes the day relative to treatment application.The y axis shows the monoterpenoid basal emission rate normalized to 303 K.The left two bars illustrate the BVOC profiles from the methyl jasmonate (MeJA) experiment and the right two bars illustrate the BVOC profiles from the negative control experiment. Figure 14 . Figure 14.A summary of the change in basal emission rates after stress treatment application for some key compounds for each experiment where a stress response was observed. (1) plant susceptibility to her-et al.: Impacts of simulated herbivory on VOC emission profiles bivory stress changes seasonally; Table 2 . Summary of activity adjustment factors for total monoterpenes and total aromatics that were calculated from pre-treatment emissions.Dashes indicate that no relationship could be established between temperature and emission rate for that experiment. Faiola et al.: Impacts of simulated herbivory on VOC emission profiles had higher contributions from alpha-pinene in spring, but decreased in August and September in a manner similar to what we observed in July.Furthermore, we observed an increase in the contribution of is Picea pungens stress experiment 2, PP-N is Picea pungens negative control, PA-E is Pinus aristata stress experiment, PA-N is Pinus aristata negative control, AG-E is Abies grandis stress experiment, and PM-E is Pseudotsuga menziesii stress experiment.The two shaded boxes denote the paired stress/negative control experiments that were performed consecutively with the same set of saplings. www.biogeosciences.net/12/527/2015/ Biogeosciences, 12, 527-547, 2015 536 C. L. Faiola et al.: Impacts of simulated herbivory on VOC emission profiles Figure5.Post-treatment emission rates for five monoterpenoid species during the PP-E1 experiment.The x axis denotes the elapsed time since treatment application in hours.Alternating shaded and unshaded regions demonstrate when the light above the plant enclosure was turned off and on, respectively. Table 5 . Results of linear regression correlation analysis (r 2 ) between all monoterpenoid emission rates (ERs) vs. terpinolene emission rates and limonene emission rates during experiment TP-E. Table 6 . Summary of the BVOC pre-treatment (PreT) and post-treatment (PostT) concentration-normalized OH reactivity (rOH) and concentration-normalized O 3 reactivity (rO 3 ) at 298 ± 2 K. Reactivity values are presented in units of s −1 .The σ is the standard deviation of the averaged measurements.The percent difference between the pre-treatment and post-treatment values is also shown.	2018-12-12T11:52:17.745Z	2015-01-28T00:00:00.000	{ "year": 2015, "sha1": "c3ffc0151e51ce64cdbee020303acf54b6880a6a", "oa_license": "CCBY", "oa_url": "https://bg.copernicus.org/articles/12/527/2015/bg-12-527-2015.pdf", "oa_status": "GOLD", "pdf_src": "ScienceParseMerged", "pdf_hash": "c3ffc0151e51ce64cdbee020303acf54b6880a6a", "s2fieldsofstudy": [ "Environmental Science" ], "extfieldsofstudy": [ "Chemistry" ] }
225312235	pes2o/s2orc	v3-fos-license	EXTRACTION AND CHARACTERISATION OF CELLULOSE FROM THE RESIDUE OF OIL PALM EMPTY FRUIT BUNCH-XYLAN EXTRACTION Cellulosic solid residual, which is the by-product of oil palm empty fruit bunch (OPEFB)-xylan extraction process, was evaluated in terms of cellulose recovery by eco-friendly approach. The OPEFB-fibre was initially treated with alkaline for xylan extraction. The resulting solid residue was dried before treated with mild concentration of bleaching agents of 20% (v/v) formic acid and 5% (v/v) hydrogen peroxide at 85°C for the extraction of cellulose. The α-cellulose content of 84.16±0.4% was achieved. Fourier transform infrared spectroscopy (FTIR) analysis showed that the products obtained have the standard cellulose structure and functional group. The crystalline nature of the extracted cellulose, as proven by X-ray diffractometer (XRD) has crystallinity index and crystalline size of 54.4% and 5.5 nm, respectively, whereas the thermogravimetric analysis (TGA) indicates the sample purity in terms of thermal stability at 370°C by comparing the results to commercially available cellulose. Meanwhile, scanning electron microscope (SEM) observation showed that the extracted cellulose morphology is in the form of separated fibrils. In terms of impurities, trace elemental analysis showed the presence of phosphorus (P) as the major element but the quantity of fluorine (F), chlorine (Cl), bromine (Br), iodine (I) and sulphur (S) detected on the cellulose was insignificant. INTRODUCTION Oil palm empty fruit bunch (OPEFB) is the most abundant biomass generated by the mills after extracting the oil from fresh fruit bunch (FFB). In 2018, 485 palm oil mills in Malaysia processed 118.9 million tonnes of FFB generating approximately 9.16 million tonnes (dry weight) of OPEFB (Kushairi et al., 2019). The abundant quantity of OPEFB generated in the oil palm mills provides huge resources for conversion into value-added products, as the OPEFB is inexhaustible, renewable, biodegradable, recyclable and a derivatisable biopolymer. Lignocellulosic material (LCM) in biomass waste composed of about 37.1% cellulose, 39.9% hemicellulose, 18.6% lignin and 3.1% extractives (Megashah et al., 2018). Each of these lignocellulosic components can be extracted and exploited for the production of fine chemicals, especially from cellulose and hemicellulose. In previous study, a method was developed to produce the xylooligosaccharides (XOS) from OPEFB fibre at laboratory scale (Noorshamsiana et al., 2015). In this method, hemicellulose was extracted from OPEFB-fibre using alkaline solution and precipitated into xylan as starting material for the production of XOS using enzymatic method. XOS is one of the most promising prebiotic products which is extensively studied and proven to EXTRACTION AND CHARACTERISATION OF CELLULOSE FROM THE RESIDUE OF OIL PALM EMPTY FRUIT BUNCH-XYLAN EXTRACTION stimulate microflora growth in the gastrointestinal tract (Manderson et al., 2005;Parajó et al., 2004). In fact, XOS gains much attention as it possesses many health benefits, particularly in prebiotic activities besides being food additive and health products (Gupta et al., 2012;Seema and Arun, 2011;Moure et al., 2006;Alonso et al., 2003). The solid residue generated from xylan extraction using alkaline solution has yet to be explored. The solid residue is in fact a LCM which consists of cellulose, lignin, as well as some traces of hemicellulose. Cellulose is another value-added product from natural fibre and it is useful in food and pharmaceutical applications. Cellulose is the most abundant polysaccharide, comprising about 44%-64% of plant materials (Mansor et al., 2019;Nazir et al., 2013;Dhillon et al., 2013;Nawirska and Kwasniewska, 2005;Wanrosli et al., 2004;McKendry, 2002) and OPEFB-fibre contains about 40%-65% of cellulose (Megashah et al., 2018;Zakaria et al., 2015;Adela et al., 2014;Chang et al., 2014;Ching and Ng, 2014;Khalil et al., 2008;Law et al., 2007). Cellulose is a natural long-chain polymer that indirectly contributes to the human food cycle. This polymer has versatile uses in many industries, such as veterinary food, pulp and paper, and fibre for clothing. In addition, cellulose has a semi-synthetic derivative which is extensively used in pharmaceutical and cosmetic industries. Cellulose can be extracted from natural fibres via chemical and mechanical methods. Acidified sodium chlorite (NaClO 2 + H 3 O + ) is commonly used as the standard reagent for cellulose delignification and extraction from wood materials (Szymańska-Chargot et al., 2017;Hook et al., 2015;Brendel et al., 2000;Leavitt and Danzer, 1993;Loader et al., 1997;Wise et al., 1946). Chlorite (ClO 2 -) produces chlorine radical which is reactive and fragments the LCM into highly toxic organochlorine. The cellulose extraction from OPEFB-fibre involves pretreatment with toluene/methanol (2:1, v/v) (Sun et al., 1999) and ethanol/benzene (1:2) (Fahma et al., 2010), with further delignification using chlorite to prepare microcrystalline cellulose (Rosnah et al., 2009). However, the existing procedures produce low yields as well as neither environment-friendly nor energy efficient. Therefore, more focus should be placed on eco-friendly methods in cellulose extraction by utilising green treatment that reduces or eliminates hazardous chemicals, using safer solvent and safer reaction conditions. Organic acids such as acetic acid and formic acid are ecofriendly reagents, less corrosive and effective for the pre-treatment of lignocellulosic biomass, and they provide a more stable medium in the aqueous phase for monosaccharide than sulphuric acid (Marzialetti et al., 2011). A mixture of hexane/ethanol and water is used to remove wax from the rice husk, followed by autoclaving and ultrasound treatments to delignify and purify cellulose using hydrogen peroxide and acetic acid (Simone et al., 2012). The cellulose extraction from OPEFB using hydrogen peroxide with ultrasound-assisted alkali extraction produced cellulose with 49% yield (Nazir et al., 2012). The use of low concentrations of formic acid (20%, v/v) and 30% hydrogen peroxide (10%, v/v) at 85°C to extract cellulose from OPEFB produced 64% (w/w) cellulose, which is the highest ever reported (Nazir et al., 2013). The development of a novel process that uses ionic liquid followed by alkaline method to extract cellulose from OPEFB fibre resulted in 93% α-cellulose recovery (Norzita and Lani, 2014). From the literature, most studies only focused on cellulose extraction from raw OPEFB fibre. This research focused on cellulose recovery from the by-product of hemicellulose extraction from OPEFB. This study aims to extract cellulose from the by-product of hemicellulose extraction using green method in order to solve the disposal problem of the process as well as to create another value added products from the OPEFB fibre. Thus, the use of bleaching agents in mild concentration to recover cellulose from the residue of OPEFB-xylan extraction is out of concern for environmental protection. The characterisation of the product obtained was carried out by Fourier transform infrared spectroscopy (FTIR), X-ray diffractometry (XRD), thermogravimetric (TG), elemental analyser and scanning electron microscope (SEM). Materials OPEFB-fibres were obtained from Biomass Plant, Biomass Technology Unit, MPOB-UKM in Bangi, Selangor, Malaysia. The treated and dried OPEFB fibre was ground and stored in the cold room prior to use. The hemicellulose extracted from OPEFB fibres using alkaline solution and precipitated into xylan was the starting material for the production of XOS using enzymatic method (Noorshamsiana et al., 2015). The resulting solid residue from the xylan extraction process was further treated for cellulose recovery. The cellulose standard derived from spruce with >99% purity (FLUKA) was used as reference. All chemicals and solvents used are of reagent grade and were used as received. Cellulose Recovery from the Solid Residue of Xylan Extraction Cellulose extraction was adapted and modified from the method developed for OPEFB fibres (Nazir et al., 2013). The solid residue was dried in the oven at 50°C to a constant weight before extracting. Two grams of solid residue were soaked in a mixture of 100 ml 20% formic acid (CH 2 O 2 ) and 100 ml 5% hydrogen peroxide (H 2 O 2 ). The mixture was then placed in the shaking water bath (WiseBath, Daihan Scientific; 45 litre) at 85°C for 2 hr. After hydrolysis, the samples were filtered with grade 4 filter paper (Whatman, England) and the supernatant washed with 10% CH 2 O 2 before being washed with distilled water several times. The supernatant was resuspended with 10% H 2 O 2 and 10% NaOH at pH 11 and 60°C for 90 min. The product obtained was then filtered, washed with distilled water and dried in oven at 60°C before being characterised. The α-cellulose content was determined by TAPPI T203 method. FTIR The infrared spectrum of the extracted cellulose was recorded using Perkin-Elmer Spectrum Frontier FTIR Spectrometer with wavelength in the range of 4000-650 cm -1 . The spectrum was compared with standard cellulose. XRD The phase behaviour of standard cellulose and extracted cellulose was elucidated by using an X-ray diffractometer (D8-Advance Bruker-AXS Germany). The peak measurement was conducted by taking 2Ɵ value between 5° and 60° with 0.04° min -1 step size. The crystallinity index (C Ir ) was calculated based on Segal method (Segal et al., 1959) as follows: Equation (1) where I 002 is the maximum peak intensity of crystalline fractions located at 2Ɵ=22.0°-22.5°, whereas I am is the low peak intensity in the amorphous region at 2Ɵ=16°, which corresponds to the crystallographic plane. Crystallite size (ε) is used to indicate the quality of the crystals. It was determined using the Scherrer's equation (Patterson, 1939) as follows: where k is the constant of 0.91, λ is the X-ray wavelength, β is the full width at half maximum (FWHM) peak intensity and Ɵ is the Bragg's angle. TGA TGA of the standard cellulose and extracted cellulose were performed using a Perkin Elmer Pyris 1 thermogravimetric analyser. About 10 mg to 20 mg of each sample was analysed continuously from 30°C to 600°C at 20°C min -1 heating rate. The analyser was purged with nitrogen, followed by 20 ml min -1 oxygen flow to prevent any thermo-oxidative degradation. SEM Morphological studies of the extracted cellulose and standard cellulose were carried out with a SEM, Hitachi S-3400 N. A piece of dry sample was put on the SEM sample lens, then the lenses were placed in a vacuum chamber to avoid obstruction and contamination by other particles. The lenses helped to direct electrons towards the sample. The electron imprint was converted into three-dimensional image. Elemental analysis was performed using an energy-dispersive X-ray (EDX) coupled with SEM. When electron beam penetrated the sample, the produced X-ray was collected, analysed and detected by silicon drift detector (SDD) and the X-ray map recorded. Elemental Composition Analysis The elements in the cellulose samples were analysed using CHN628 Series Elemental Determinator. The instrument is used to determine nitrogen, carbon/nitrogen, and carbon/hydrogen/ nitrogen composition in organic matrices by utilising combustion technique. Trace Element Content Cellulose samples were analysed for nitrogen, sulphur and halogen content using NSX-2100 Automatic Combustion Unit (Model AQF-2100 series). Eco-friendly Extraction of Cellulose from the Solid Residue of OPEFB-xylan Extraction Process The solid residue generated from OPEFB-xylan extraction was treated with mild bleaching agents of 20% (v/v) formic acid and 5% (v/v) hydrogen peroxide at 85°C, which yielded 82.69±3.9% cellulose with α-cellulose content of 84.16±0.4%. Formic acid was used to remove hemicellulose, leaving the solid residue to have lesser hemicellulose with high cellulose content (Goldmann et al., 2017;Li et al., 2015). Hence, more cellulose was obtained in the treated product. The major contributor of cellulose extraction was formic acid because it is very stable in aqueous medium (Nazir et al., 2013). A stable formic acid provided an alternative for the interaction and dissolution of non-crystalline part in cellulose. This helped to remove the non-lignin parts, hemicellulose and lignin by converting the hemicellulose into salt and lignin was converted to carboxylic acid. Hemicellulose and lignin were removed as water-soluble fraction while the insoluble crystalline cellulose was left suspended in the aqueous medium. A study by Li et al. (2015) found that formic acid is useful to remove hemicellulose, swells cellulose fibres, and releases cellulose nanocrystals (CNC). The findings showed that formic acid does increase the amount of hemicellulose extracted from biomass fibre, leaving behind the cellulose component. Furthermore, it was found that more cellulose can be recovered using formic acid at low temperature. The low temperature of 85°C favours hemicellulose hydrolysis rather than cellulose hydrolysis, resulting in more cellulose to be retained in the solid residue (Goldmann et al., 2017). It was also observed that cellulose recovery decreased with the increase in hydrogen peroxide concentration. This is due to the presence of H + ions that help to strengthen the microfibrillated lignin-cellulose structure, stiffness and stability (Lindman et al., 2010). Chemical Characterisation FTIR. The extracted cellulose was analysed with FTIR and compared with standard cellulose (FLUKA), to determine the composition of functional group and confirm its chemical structure. FTIR analysis showed that the spectrum of the OPEFB-based cellulose extracted using eco-friendly reagents was identical to that of standard cellulose, suggesting the presence of similar backbone and functional groups in both samples (Figure 1). The O-H bond which indicates the presence of -OH groups on the glucose unit is characterised by the broad band found in the range of 3200-3400 cm -1 (Norzita and Lani, 2014;Moran et al., 2008;Khalil et al., 2001). In this study, the O-H bond of the extracted cellulose and standard cellulose was found at 3327.84 cm -1 and 3330.34 cm -1 , respectively. The medium peak at 2891.61 cm -1 for extracted cellulose and 2893.33 cm -1 for standard cellulose reflects the C-H bonds. Meanwhile, the peak at 1023.84 cm -1 for extracted cellulose and 1052.33 cm -1 for standard cellulose is assigned to the C-O-C linkage of the anhydrous repeating units. The peaks at 1735 cm -1 and 1375 cm -1 , which represent wax and hemicellulose, respectively were not present in the extracted cellulose, indicating both components were removed during bleaching. From literature, the FTIR spectral peaks representing lignin at 1248 cm -1 and 1037 cm -1 were not observed in the extracted cellulose (Ching and Ng, 2014;Nazir et al., 2013;Simone et al., 2012). It was observed that the extracted cellulose had lower transmittance than standard cellulose. This may be due to the compound that exists in the extracted cellulose causes an increase in absorptivity, which leads to lower transmittance percentage. The extracted cellulose has similar structure and functional groups to that of standard cellulose, thus, confirming that the product was cellulose. XRD. The XRD patterns of the extracted cellulose and standard cellulose are shown in Figure 2. The diffractogram (diffraction pattern) shows the presence of crystalline and amorphous fractions in both samples. Extracted cellulose It is observed from Figure 2 that the standard cellulose has diffraction peaks around 2Ɵ=20°-24° while the extracted cellulose has diffraction peaks at 2Ɵ=18°-24°. The standard cellulose C Ir and ε were 71.9% and 6.0 nm, respectively, whereas for the extracted cellulose, they were 54.4% and 5.5 nm, respectively. For comparison purposes, the crystallinity indexes of cellulose derived from native OPEFB fibres in the previous studies were 82.4% (Megashah et al., 2018), 70% (Nazir et al., 2013), 69% (Jonoobi et al., 2011), 63% (Norzita and Lani, 2014) and 55% (Fahma et al., 2010). In general, the structures of the natural cellulose fibres are classified as type I, II, III and IV. It is observed that the standard cellulose is classified as cellulose type I as it has no doublet in the main peak of 2Ɵ=23° (Norzita and Lani, 2014;Bhatnagar and Sain, 2005). Meanwhile, the cellulose derived from the solid residue of OPEFB-xylan extraction using mild bleaching agents is categorised as type II. This is based on the crystalline peak that appears as a doublet at 2Ɵ=20° and 2Ɵ=22°, demonstrating the transformation of native cellulose from cellulose type I to cellulose type II. TGA. The thermal behaviour and thermal stability of the samples obtained were studied using TGA to determine the changes in mass when temperature is increased over time. Figures 3a and 3b show the thermogram of the extracted cellulose and standard cellulose, respectively. It was observed that both cellulose samples exhibited single step degradation in the temperature range of 200°C-400°C, indicating that the degradation was due to one main component. Weight loss was observed at around 100°C, due to the loss of moisture in both samples. Looking at Table 1, it was observed that the standard cellulose was degraded by half at higher temperature and has higher maximum thermal decomposition compared to the extracted cellulose. The findings suggest that standard cellulose has better thermal stability than extracted cellulose. This is because the thermally unstable component in the extracted cellulose contains celluloses from different sources. The standard cellulose (FLUKA) (purity >99%) is isolated from spruce wood, while the extracted cellulose is from the empty fruit bunch fibres. In addition, the cellulose extraction for these two samples are different, whereby, the standard cellulose is extracted using chemical method, whereas the cellulose in this study was extracted using eco-friendly reagents. Also, crystallinity is important in determining thermal stability. Lower crystalline region in the extracted cellulose results in lower thermal stability (Megashah et al., 2018;Nascimento et al., 2015;Poletto et al., 2014). SEM. Surface morphology of the solid residue from the OPEFB-xylan extraction is depicted in Figure 4, whereas the surface morphology of the extracted cellulose and standard cellulose are depicted in Figures 5 and 6, respectively. Figure 4a shows that the solid residue of OPEFB xylan extraction is a lignocellulosic composite with compact fibril packing, having average diameter of 93.07±2.78 μm, suggesting incomplete removal of lignin and hemicellulose (Megashah et al., 2018;Nazir et al., 2013). Complete microfibrillation with microfibrils in the fibre bundle with diameter of 6.29±0.17 µm is observed in Figure 4b. The extraction of hemicellulose (xylan resource) from native OPEFB fibre via alkaline treatment generates solid residue consisting of cellulose, lignin while the remaining hemicellulose results in compact fibrillated structure. From Figure 5a, the microfibrils of the extracted cellulose have average diameter of 8.90±0.39 μm, indicating the disruption of lignocellulosic structure. The fibre bundles were separated into individual cellulose microfibres, indicating that lignin and hemicellulose were removed through bleaching (Megashah et al., 2018;Sofla et al., 2016;Zeng et al., 2014). As depicted in Figure 5b, there was an individual cellulose microfibre with a diameter of 6.31±0.30. Figures 6a and 6b show incomplete microfibrillation as compared to solid residue and extracted cellulose morphologies due to the presence of non-fibrillated structure. There are uniform microdimensional crystal bundles and smooth surface, indicating that it was in the crystalline form. This supports the XRD results that standard cellulose has higher crystallinity index than extracted cellulose. The smooth surface indicates there is no inorganic materials or impurities on the standard cellulose. Table 2, oxygen (O) was detected as the main element in both samples besides carbon (C) and hydrogen (H). Nitrogen (N) was also detected but the value was insignificant. No sulphur (S) was found in both samples. The C, H, O, and N composition in the extracted cellulose was similar to standard cellulose. Chemical compositions analysis. As shown in Trace elements analysis. The existence of impurities in samples indicates the quality of the product. The quantity of trace elements and halogen content in both samples is tabulated in Table 3. From the findings, no bromine was detected in both samples. The concentration of chlorine in the extracted cellulose (0.21%) was comparable to the standard cellulose (0.22%). It is deemed negligible as the amount is below 1.0%. Fluorine (F) (0.01%), iodine (I) (0.13%), and S (0.04%) were also detected but in insignificant quantity. Concentration of phosphorus (P) was highest in both extracted and standard celluloses, at 1.20% and 1.40% respectively. P comes from soil and fertiliser. Fertiliser is used in the oil palm plantation when the available P in the soil is less than 15 mg kg -1 Afandi et al., 2016;Goh and Rolf, 2003;Adiwiganda, 2002;Rankine and Fairhust, 1999). P is essential for plant growth and also important for root growth in the early growth stage of the plant. Therefore, the presence of P in the OPEFB-based cellulose is higher than other elements. CONCLUSION Cellulose with α-cellulose content and crystallinity index of 84.16±0.4% and 54.4%, respectively was successfully recovered from the OPEFB-xylan extraction residue using green method of 20% formic acid and 5% hydrogen peroxide. The results from FTIR, TG and elemental analyses showed that the extracted cellulose has similar chemical properties and characteristics of standard cellulose. With the recovery of all OPEFB components, disposal issue can be solved, as well as creating value-added products from OPEFB-fibre.	2020-09-10T10:07:51.372Z	2020-09-03T00:00:00.000	{ "year": 2020, "sha1": "6663df2522c4ec3a70e44cd85368c8c1548087f6", "oa_license": null, "oa_url": "http://jopr.mpob.gov.my/wp-content/uploads/2020/12/joprv32dec2020-noorshamsiana.pdf", "oa_status": "GOLD", "pdf_src": "ScienceParsePlus", "pdf_hash": "a593782309d0861f23b4d291fb2875f7951c9dc3", "s2fieldsofstudy": [ "Agricultural And Food Sciences", "Materials Science" ], "extfieldsofstudy": [ "Chemistry" ] }
239053685	pes2o/s2orc	v3-fos-license	Do News Actually “Find Me”? Using Digital Behavioral Data to Study the News-Finds-Me Phenomenon Research on news exposure has shown that while political knowledge and interest largely determine the degree of active engagement with online news, some people are generally less willing to invest into actively staying informed. Instead, these people report to pursue a passive mode of relying on specific sources, such as social media, based on the belief that “news finds me” (NFM). Notably, the three dimensions of NFM—feeling informed, relying on peers, and not actively seeking news—combine intentions and perceptions related to news use. Understanding NFM perceptions, hence, requires an analytical distinction between active and passive modes of news use as well as reliable measures of (different types of) news exposure. We contribute to this field by combining a survey, tracked web-browsing data, and tracked Facebook data to investigate the relationship between NFM perceptions and exposure to online news, also taking into account political knowledge and interest as traditional predictors of active news use. Our results show that both political knowledge and interest are associated with more news exposure via web browsers and that political knowledge—but not political interest—is also associated with more news in people’s Facebook feeds. Compared with the NFM dimensions, political knowledge and interest are stronger predictors of online news exposure in our study. Taken together, the novel combination of Facebook and web tracking data provides evidence that online news exposure is shaped by a confluence of traditional factors and more diffuse interpersonal processes. People around the world are increasingly consuming news online (Newman et al., 2020). The quantity and diversity of online news sources create a "high-choice media environment" (van Aelst et al., 2017), in which users can come across news in a variety of ways. One key factor that drives differences in online news exposure is the level of activity people engage in to stay informed. While various studies have found political interest and knowledge to be strong predictors of active news use (e.g., Lecheler & de Vreese, 2017;Strömbäck & Shehata, 2019), recent research has demonstrated the prevalence of a more passive mode of news use, showing that some people are less willing to actively invest into staying informed about current news and political affairs (e.g., Kim et al., 2013;Lee & Xenos, 2020;Leonhard et al., 2020). Several studies have shown that a substantial amount of media users pursue a more passive mode of mostly relying on friends and exposure on social media for receiving relevant news (Boczkowski et al., 2018;Song et al., 2020). This mode of "being informed without actively seeking the news" , p. 1606 has been called the "news-finds-me perception" (Gil de Zúñiga et al., 2017). The news-finds-me (NFM) perception describes "the extent to which individuals believe they can indirectly stay informed about public affairs [ . . . ] through general Internet use, information received from peers, and connections within online social networks" (Gil de Zúñiga et al., 2017, p. 107). In Germany, where the study we present in this article was conducted, 70% of online users report to find their news primarily online, 22% report to use Facebook for getting news (Newman et al., 2020), and 49% of online users hold high NFM perceptions . In general, pronounced NFM perceptions have been found to be 1033820S MSXXX10.1177/20563051211033820Social Media <span class="symbol" cstyle="Mathematical">+</span> SocietyHaim et al. research-article20212021 1 University of Leipzig, Germany 2 GESIS-Leibniz Institute for the Social Sciences, Germany associated with higher levels of social media use and lower levels of political interest and knowledge (Gil de Zúñiga & Diehl, 2019;Leonhard et al., 2020). Regarding the elements that define an NFM perception, a recent study has suggested to differentiate the concept into three dimensions, namely, feeling informed, relying on peers, and not actively seeking for news (Song et al., 2020). This differentiation highlights different stages of media use and also has implications for causal assumptions in that, for example, "not actively seeking" points to a priori predispositions for news use, whereas "feeling informed" rather describes a post hoc rationalization of previous or typical behavior. Subsequently, this raises questions about how these "passive" approaches to news use relate to more "active" media-use predictors traditionally discussed in research on news use. Hence, building on these different predictors identified in prior research, the goal of our study is to assess how political interest and knowledge as traditional predictors of active news use compare to NFM perceptions as indicators of more passive news use when it comes to explaining actual online news exposure. Considering the limitations of self-reports on media use (Araujo et al., 2017;Prior, 2009;Scharkow, 2016), we look at actual news exposure as measured by tracking data to answer the three general research questions: (1) How do political interest and knowledge relate to observed news exposure in web-browsing histories and Facebook feeds? (2) How does NFM perception and its dimensions relate to observed news exposure in web-browsing histories and Facebook feeds? (3) How do political interest and knowledge (i.e., drivers of active news use) and NFM perception and its dimensions (i.e., perceived reliance on passive news use) interact? To answer these questions, we use a unique dataset combining passively observed tracking and self-report data from a sample of German internet users. Specifically, we use a multi-method combination of an online survey and a passive recording of both web browsing and Facebook usage to investigate the relationship between NFM perception and its dimensions, political interest and knowledge, and actual online news exposure. With this study, we seek to add to the existing NFM literature by making use of data on actual news exposure as assessed by tracking data while also contrasting the role of established determinants of active news use (political interest and knowledge) with attitudes underlying a more passive mode of news use (NFM perception). Active and Passive Online News Use Previous research on both general and online news use has shown repeatedly that political interest and knowledge are two very important determinants (Lecheler & de Vreese, 2017;Strömbäck et al., 2012;Strömbäck & Shehata, 2019). When it comes to these dimensions of "political sophistication," empirical evidence suggests that political interest can be considered a rather stable predisposition (Kruikemeier & Shehata, 2017;Lecheler & de Vreese, 2017), which has been found to be particularly relevant for offline news use (Boulianne, 2011), whereas political knowledge, which is typically highly correlated with political interest, has been shown to not only influence subsequent news use but also benefit from it (Lecheler & de Vreese, 2017;Möller & de Vreese, 2019). Social networking sites (SNS) have been considered a special case for online news exposure. That is, while SNS use has been shown to positively affect online political participation (Dimitrova et al., 2014), political knowledge does not seem to profit accordingly from their use (Dimitrova et al., 2014;van Erkel & van Aelst, 2021). SNS have, therefore, been criticized for having a potentially detrimental effect on the quantity and diversity of online news exposure. However, a recent study using web-browsing data found that the use of SNS and other intermediaries is associated with an increased exposure to online news, in terms of both quantity and source diversity (Scharkow et al., 2020). This supports earlier indications based on different methodological approaches (Fletcher & Nielsen, 2018;. Results like these suggest that the use of SNS might be especially beneficial with regard to news use for people who tend to less actively seek out and follow news online as SNS use facilitates incidental exposure. However, researchers have also argued that there might be a "Matthew Effect," meaning that the positive effect of social media use is more pronounced for individuals who already consume a comparably high amount of news and, accordingly, also show higher political interest and knowledge (e.g., Kümpel, 2020). For example, knowledge about the topic of a given story has also been shown to affect intentions of reading the full story (Karnowski et al., 2017). Within social media, incidental news exposure is affected to a substantial degree by algorithmic selection, which is largely based on the reading, commenting, liking, and sharing of news by users (Boczkowski et al., 2018). At the same time, incidental news exposure might also be a delayed result of actively following media outlets in the past (Boczkowski et al., 2018). In other words, a brief period of active news use, such as clicking the "follow" (or "like") button for a news outlet, likely affects subsequent passive news use as the content posted by the respective news outlets will be included in the user's feed. Importantly, while more passive users less regularly click the "follow" button for news outlets than more active news consumers (Leonhard et al., 2020), more news could also be included in one's social media feeds because of friends' engagement with news outlets (Boczkowski et al., 2018). Driven by the perceived "shareworthiness" (Karnowski et al., 2021;Trilling et al., 2017) of news items, this diffusion of news via SNS "is a continuous process involving journalists, users, intermediaries, and algorithms" (Brosius et al., 2019, p. 133), which equally affects and is affected by a user's surrounding social network. News on SNS are, thus, brought to one's attention through the "interactive features and affordances of various social media platforms, particularly those which reduce attentional effort" (Oeldorf-Hirsch & Srinivasan, 2021, p. 11). It has repeatedly been shown that especially networks of likeminded contacts within SNS create fruitful grounds for further engagement with news items (commenting, liking, sharing) (cf. Barberá et al., 2015;Mitchelstein et al., 2020), which, in turn, benefits the diffusion of news within the network and may, thus, create or amplify the perception that "news finds me." NFM Perception In previous studies, the use of SNS has been shown to be positively correlated with a high NFM perception . Notably, NFM perception neither depicts a motivation nor does it point to individual interest in public affairs. People with high NFM perceptions could easily be interested in news and public affairs, but do not see a need to actively search for news. As such, high NFM perceptions are widespread as findings from a multitude of countries indicate. In the studied countries, more than half of all online users hold high NFM perceptions, for example, in Italy (59%), Ukraine (71%), Taiwan (72%), or Spain (85%) , p. 1612); slightly lower rates of online users displaying high NFM perceptions have been measured in the United Kingdom (39%), Japan (39%), New Zealand (42%), and Germany (49%). NFM perceptions are also more common among adults aged 35 years and younger , women, and individuals reporting lower incomes (Gil de Zúñiga & Diehl, 2019). Moreover, so-called "news avoiders" have been shown to internalize NFM perceptions (Toff & Nielsen, 2018), and higher NFM perception has also been found to correlate with lower levels of political interest (Gil de Zúñiga & Diehl, 2019;Park & Kaye, 2020) as well as lower levels of political knowledge (Gil de Zúñiga & Diehl, 2019). Conversely, a recent survey among German online news users has shown passive SNS news use to be associated with an overestimation of one's own political knowledge (Leonhard et al., 2020), while people with higher NFM perceptions have also expressed the perception to be surrounded by politically knowledgeable and interested peers on SNS (Song et al., 2020). NFM perception has been considered both a dependent and an independent construct in previous research on online news use. Typically measured through surveys by means of a multi-item construct (Gil de Zúñiga et al., 2017, p. 112), the items in the NFM scale ask respondents to judge (1) how well they feel informed, (2) how much they rely on peers for information, and (3) how regularly they actively seek for information. There are versions of the NFM scale with a total of three (e.g., Park & Kaye, 2020), four (e.g., Gil de Zúñiga & Diehl, 2019;Gil de Zúñiga et al., 2017, 2020, or six (e.g., Song et al., 2020) items, and the items are answered on either 7-or 10-point Likert-type scales. Typically, the items have then been aggregated into a mean score to look at a "monotonic relationship with other variables" (Song et al., 2020, p. 65). However, this one-dimensional aggregation may not fit the conceptional ambiguity of the construct. That is, the three dimensions address not only different aspects but also different stages of media use as they refer to attitudes that can be predictors or outcomes of news use (or both). Song and colleagues (2020) have, hence, pointed out that separately considering the three dimensions of NFM perception-feeling informed, relying on peers, and not actively seeking-allows for more nuanced insights into respective dynamics. Specifically, the authors argue that not actively seeking information could be a consequence of the perception of feeling informed because people might rely on their peers for that. This additional nuance is also supported by their confirmatory factor analysis, which highlights the threefold dimensionality of the NFM construct. Yet, also these relationships have to be interpreted against the backdrop that relying on surveys for capturing online news exposure can only explain parts of the story (Kümpel et al., 2015;Prior, 2009). That is, several studies have shown that self-report data on media use tend to be unreliable, especially if respondents are asked about types of media use that are socially (un)desirable or that they only very infrequently engage with (Araujo et al., 2017;Guess, 2015;Prior, 2009;Scharkow, 2016); two attributes that may also apply to online news use. Conversely, the digital trace data that are more commonly used in the growing field of computational social science often lack important information about the individuals, such as beliefs, attitudes, or perceptions. A way to address these limitations is to combine these two types of data , which is also what we did in this study. Hypotheses In contrast to active news use, which has been shown repeatedly to be driven primarily by political interest and knowledge, NFM perception presumes a passive online news-use behavior. To test the implications of this perception empirically, our study relies on a combination of passive measurements of news exposure on both websites and Facebook with self-report data from an online survey. To answer our first guiding research question, we look at the long-standing assumption of active news use being driven by political interest and knowledge. Specifically, we expect that people with higher levels of political interest and knowledge are exposed to more news content in both web-browsing histories and Facebook feeds. H1a. The amount of news in web-browsing histories is positively associated with political interest and political knowledge. H1b. The amount of news in Facebook feeds is positively associated with political interest and political knowledge. Pertaining to the second guiding research question, previous research has shown that SNS use is associated with increased online news exposure and that SNS use also leads to news exposure from a higher number of different news sources (Scharkow et al., 2020). As NFM perception is associated with higher levels of social media use and a stronger reliance on social connections (Gil de Zúñiga & Diehl, 2019), it can be assumed that people with higher NFM perception also profit from more active SNS use as news are more likely to "find them" via these platforms. This should particularly be true for the NFM dimension "relying on peers" that directly addresses the importance of peers on social media. Conversely, we expect people with higher NFM perception who particularly report to "not actively seek" news (i.e., the second NFM dimension) to have smaller amounts of news sites in their web-browsing histories. In addition, we expect people who report to "feel informed" (i.e., the third NFM dimension) to be exposed to higher amounts of news in both their web-browsing histories and their Facebook feeds. H2a. The amount of news in web-browsing histories is negatively associated with the NFM perception of "not actively seeking." H2b. The amount of news in Facebook feeds is positively associated with the NFM perception of "relying on peers." H2c. The amounts of news in both web-browsing histories and Facebook feeds are positively associated with the NFM perception of "feeling informed." Conceptualizing NFM perception as a passive mode of news use naturally raises the question of how these perceived or actual individual-level tendencies relate to well-established predictors of more active news use. Regarding the third guiding research question on the interplay of active and passive modes of online news use, then, we expect the relationship between the NFM dimension of "not actively seeking" and news exposure in web browsing to be moderated by political interest and knowledge. That is, while a perception of not actively seeking news is expected to relate to less news in web browsing (H2a), people with higher political interest and knowledge are also expected to receive more news via SNS (H1b). In line with findings indicating that social media use yields higher variation in online news (Scharkow et al., 2020), we expect people who are exposed to more news in their Facebook feeds to also click on some of these news which should, ultimately, also increase the amounts of news in web-browsing histories. In addition to that, the concept of news diffusion suggests that one's Facebook feed is partly determined by actively following news outlets, which, in turn, can be expected to be driven by political interest and knowledge, but partly also by a passive observation of peers and their engagement with news. Hence, we assume the relationship between "relying on peers" and the amount of news on Facebook (H2b) to be moderated by political interest and knowledge. That is, a more active SNS use should only yield more news in one's Facebook feed if a person shows a certain degree of political interest and knowledge and might, hence, have been actively subscribing to news outlets in the past or is surrounded by politically interested peers. Owing to homophily in social networks, individuals with high political sophistication can be expected to have a higher share of politically interested and knowledgeable individuals in their social networks (Gil de Zúñiga & Diehl, 2019). As a consequence, a perceived reliance on peers should only translate into more actual news exposure on SNS among politically sophisticated individuals. H3a. Political interest and knowledge moderate the relationship between NFM perception of "not actively seeking" and news exposure in web-browsing histories. H3b. Political interest and knowledge moderate the relationship between the NFM perception of "relying on peers" and news exposure in Facebook feeds. Method We employed a multi-method combination of an online survey and a passive recording (i.e., observation) of both web browsing and Facebook usage behavior. All of the data we used originate from a non-probability panel of roughly 2,000 German internet users who agreed to use a tracking software that hooks into the web browsers on their desktop computers and/or smartphones. The panel is managed by a German market research company which invited panelists to an online survey, in which they were asked to also install a plugin for their desktop browser to unobtrusively capture public Facebook posts (Haim & Nienierza, 2019). A replication data set 1 and all analyses scripts for this study are available via the Open Science Framework (OSF): https://osf.io/6xrej/. Web-Browsing Data We used web-browsing data from March through May 2019 collected by the market research company in accordance with the ICC/ESOMAR International Code of Marketing and Social Research Practice. These data contain all logged visits from the respondents' web browser on the domain level (e.g., facebook.com, nytimes.com) along with a timestamp and an anonymous user ID. Hence, no individual pages (e.g., news articles) are identifiable in this dataset. We then coded domains as news based on whether they are listed within the German Audit Bureau of Circulation's (IVW) traffic rankings for news 2 and/or included in previous research with similar data (Scharkow et al., 2020). Overall, a total of N = 14,296,676 domain visits was tracked across N = 1,121 panelists (also see the "Survey Measures" section) over the 3 months. Of these visits, n = 995,989 (7%) account for Facebook visits and n = 395,401 (3%) account for visits to news websites. These shares are very much in line with comparable numbers from similar recent studies (e.g., Cardenal et al., 2019;Scharkow et al., 2020). Survey Measures For this study, we use data from an online survey conducted in March 2019. A total of 1,121 panelists for whom we also have web tracking data completed this questionnaire (also see Online Appendix, Section 1, Table A1). Forty-nine percent of these respondents were female and the age ranged from 16 to 67 years (M = 45.3, SD = 12.75). In the survey, we asked participants whether they have a personal Facebook account (80% indicated that they did) and those who reported to have one were invited to install a browser plug-in that captures the public posts in their Facebook news feed (see the "Facebook Data" section). In line with Gil de Zúñiga and colleagues (2017), we also asked about education, collapsing 13 education categories into four categories reflecting the German educational system: lower secondary education or less (26%), secondary education (30%), postsecondary education (24%), and tertiary education (20%). In addition to education, we also assessed income (a scale consisting of 13 income brackets; Md = "between €2,000 and <€2,500") to control for these variables in our models. 3 Comparing age, gender, and education with the most recent German census data from 2017 indicated a slightly younger and more highly educated sample, which can be considered typical for online surveys (also see Online Appendix, Section 1, Table A2). Participants were also asked about their general political interest on a 4-point Likert-type scale from 1 (not at all interested) to 4 (very interested), averaging at M = 2.9 (SD = 0.85). Moreover, we posed three questions asking respondents to assign three pictures of leading politicians to their party and two questions about the German political system to estimate their political knowledge. The index of political knowledge was constructed by summing up correct answers, ranging from 0 to 5 (M = 2.9, SD = 1.60). For NFM perception, we used the same six items as Song and colleagues (2020) who added two items ( "I am up-to-date and informed about public affairs news, even when I do not actively seek news myself," "I do not have to actively seek news because when important public affairs break, they will get to me through social media") to the original set of four statements by Gil de Zúñiga and colleagues (2017; "I rely on my friends to tell me what's important when news happens," "I can be well informed even when I do not actively follow the news," "I do not worry about keeping up with the news because I know news will find me," and "I rely on information from my friends based on what they like or follow through social media"). We used the German translations provided by the authors of the original questionnaire and five-point Likerttype scales, ranging from 1 (strongly disagree) to 5 (strongly agree). The full list of items, together with their descriptive statistics, can be found in the Online Appendix (Section 2, Table A5). In concordance with Song and colleagues (2020), we employed confirmatory factor analyses to compare NFM perception as single four-item (n = 1,104, χ 2 = 105.21, df = 2, While none of the models shows an excellent fit, similar to Song and colleagues (2020), the three-factor solution fits our data best (for details, see Section 2 in the Online Appendix). Hence, we used the factor scores for the three NFM dimensions as separate predictors in our analyses. Facebook Data The n = 895 respondents who reported having a personal Facebook account were asked whether they would be willing to install a browser plug-in that unobtrusively collects public posts from their Facebook news feed. The plug-in was available for Firefox and Chrome on desktop computers and respondents could install it through the respective official plug-in stores (Haim & Nienierza, 2019). Participants received an additional incentive of €5 for installing the browser plug-in (i.e., in addition to the compensation they received from the market research company for participating in the web tracking panel and the online survey). They were shown a brief informed consent describing the browser plugin and the purpose of data collection in the online survey and could access an extended data-privacy information page hosted on an institutional website via a provided link. This procedure and the content of both the informed consent and privacy information page were based on work by Sloan and colleagues (2020) who combined survey and Twitter data in their study and are described in detail in a paper by Breuer et al. (2021). The plugin continuously collected public posts if they appeared on participants' desktop screens while browsing their Facebook news feed. Accordingly, the posts were likely seen by the users, but not necessarily clicked on. Hence, similar to the web-browsing data, these data can be seen to capture exposure rather than actual consumption. Participants were also able to log in to a project website for the browser plug-in to see and even delete any collected Facebook data. Notably, however, no participant made use of this option. Of the n = 895 Facebook users in our sample, 48% (n = 426) agreed to install the browser plug-in, and 27% (n = 246) actually installed and used the plug-in. 4 Notably, this multi-step selection process entails risks of systematic biases. Indeed, our plug-in/Facebook subsample includes slightly higher shares of male participants than the original sample, is slightly more educated, and shows slightly higher political sophistication; however, all of these differences are very small (for details, see Section 1 in the Online Appendix, Table A1). Of these n = 246 users, 60% (n = 147) used the plug-in for a Chrome browser, while the remaining n = 99 installed the Firefox plug-in. Installed plug-ins automatically prompted study participants to end the observation period by manually uninstalling the plug-in, starting two weeks after installation. Our time frame under investigation, thus, ranges from the beginning of the survey (March 5, 2019) until the date of the last data point in the plug-in data (May 30, 2019). In total, n = 37,820 sessions (i.e., Facebook feeds) were recorded with a total of n = 739,710 public posts. To identify news posts, the Facebook accounts behind all public posts included in our data were extracted and matched manually with the news domains in the web-browsing data (e.g., facebook.com/derspiegel and www.spiegel.de), yielding a total of n = 28,727 (4%) identified public news posts. Results We tested our hypotheses through multiple regression analyses to probe the relationships between political knowledge and interest, the three NFM dimensions, sociodemographic information from the survey, and behavioral measures from both Facebook and web tracking. All dependent variablesnews exposure in web histories (H1a, H2a, H2c, and H3a) and news exposure in Facebook posts (H1b, H2b, H2c, and H3b)-are heavily skewed; that is, in line with prior findings, few people saw a lot of news content, whereas the majority did not. As likelihood-ratio tests indicated the presence of overdispersion, we used negative binomial regression models instead of Poisson regressions to test our hypotheses. 5 We included the overall number of public Facebook posts (H1b, H2b, H2c, and H3b) and general website visits (H1a, H2a, H2c, and H3a) in the respective models to control for the total amount of tracked internet and Facebook use which varied considerably. 6 The first model for web-browsing histories (Table 1) tests H1a, H2a, and parts of H2c by regressing the amount of news visits in web-browsing histories on the three NFM dimensions, political knowledge and interest, and the control variables. Consistent with our first hypothesis (H1a), both political knowledge and interest are positive and significant predictors of the amount of news website visits. Also, in line with our assumptions (H2a), the regression coefficient for the NFM dimension "not actively seeking" is negative and emerges as a significant predictor of the amount of news visits. In contrast, the NFM dimension of "feeling informed," despite showing relationships in the expected direction, is not a significant predictor of news exposure in this model (H2c). While most of the control variables did not add significantly to the model, the total number of websites tracked (i.e., general internet use) is a strong predictor of the amount of news visits. 7 The first model for news in Facebook feeds (Table 1) tests H1b, H2b, and one part of H2c by regressing the amount of public Facebook feed news posts on the three NFM dimensions, political interest and knowledge, and the control variables. Notably, here, only political knowledge is a positive and significant predictor of news exposure, whereas political interest is not significantly associated with the number of public news posts in the Facebook feed, thus providing only partial support for our assumption (H1b). Moreover, no NFM dimension is a significant predictor of news exposure via public Facebook posts, thus providing no support for this hypothesis (H2b). As such, also the NFM dimension of "feeling informed" did not show any significant relationship with the amounts of news posts in people's Facebook feeds (H2c). Again, the control variable of total number of public posts tracked (i.e., general Facebook use) is the strongest predictor in this model. In addition, also tertiary education (positive) as well as household income (negative) show relationships with news exposure in participants' Facebook feeds. 8 In additional regression models for the two outcome variables, we found almost no indication for moderating effects of political knowledge and interest. That is, neither political knowledge nor interest generally moderate the relationship between the NFM perception of "relying on peers" and news exposure in Facebook feeds (H3b; Model 2 in Table 1). A small relationship only appears for political knowledge when omitting political interest from the model (Online Appendix, Section 3, Table A10, Model K-2). Similarly, neither political knowledge nor interest moderate the relationship between the NFM perception of "not actively seeking" and news exposure in web-browsing histories (H3a; Model 2 in Table 1). Again, a small association exists between political interest and news exposure when omitting political knowledge from the model (Online Appendix, Section 3, Table A9, Model I-2). Discussion This study combines survey data with digital trace data to compare the predictive power of NFM perceptions with political knowledge and interest for online news exposure via web browsing and Facebook. In contrast to studies relying solely on survey data, our results are more robust toward social desirability (as consuming news is typically deemed socially desirable) and a potential confounding 9 between NFM and news-use measures. By differentiating NFM perception into previously suggested dimensions (Song et al., 2020) this study also takes into account the conceptual ambiguity of considering various stages of media use in that "not actively seeking," arguably, addresses a priori predispositions regarding news use, whereas "feeling informed" might actually describe a post hoc rationalization of behaviors. Overall, our findings indicate that both political knowledge and interest are associated with more news exposure in web browsing and that political knowledge-but not political interest-is also associated with more news exposure via Facebook. Overall, this is very much in line with prior findings on the relevance of these dimensions for online news use (Karnowski et al., 2017;Lecheler & de Vreese, 2017;Strömbäck & Shehata, 2019). It also further emphasizes the importance of political knowledge and interest as predictors of exposure to news online. In contrast, however, our findings indicate that neither web-browsing histories nor Facebook feeds differ much across varying levels of NFM perception. The only exception is that the perception of "not actively seeking" for news is associated with visiting fewer news websites in our sample. As in previous studies using a singular/unidimensional NFM perception (Gil de Zúñiga & Diehl, 2019), "not actively seeking" also correlates negatively with political knowledge (Pearson's r = −.20, p < .001, df = 1,090). Considering "not actively seeking" an a priori predisposition toward news use and political knowledge both a predictor of news use and a result of news use, this finding suggests that this NFM dimension likely captures a more general attitude toward online news use rather than a perception that "news finds me." Importantly, neither political knowledge nor interest generally interacts with individual NFM dimensions with respect to the amounts of online news exposure, both in web browsing and on Facebook. Instead, differences in the amounts of online news exposure were generally better explained by differences in political knowledge and interest as well as general internet or Facebook use. Despite the absence of strong associations between NFM perceptions and online news exposure, our findings provide various suggestions for possible future directions in this stream of research. Specifically, the results of our confirmatory factor analysis for the NFM items and the varying internal relationships between the three dimensions (Online Appendix, Section 2, Figure A8) highlight important conceptual differences, especially between "not actively seeking" on one hand and both "feeling informed" and "relying on peers" on the other. The former might be interpreted as an a priori predisposition regarding news use, whereas the other Note. Dependent variables are the raw counts of tracked news website visits and captured news posts in users' Facebook feeds. Unstandardized coefficients (standard errors) from negative binomial regression models. "Primary education" is the reference category for education. p < .05. p < .01. **p < .001. two dimensions presumably describe post hoc rationalizations of past or typical behavior. What follows from this is that NFM perception can be (and has been) conceptualized as either an independent or a dependent variable, calling for a differentiation between the different dimensions. The discrepancy between these dimensions also shows in the regression models. As such, a general belief to "indirectly stay informed about public affairs [ . . . ] through general internet use, information received from peers, and connections within online social networks" (Gil de Zúñiga et al., 2017, p. 107) does not translate into a mere increase in news exposure through either general internet or social media use. Instead, and in line with recent NFM research (Song et al., 2020), these findings highlight the need for a more differentiated view on the individual dimensions of NFM perception, which, ideally, also require multi-method investigations. Among the three NFM dimensions, relying on peers shows the weakest relationships with both dependent variables in all of our regression models (Table 1). Not actively seeking only shows relevant (negative) relationships with the amount of news websites in web-browsing histories in the present study. While this relationship seems intuitive, the lack of a similar relationship with news on Facebook raises concerns about the applicability of this dimension to the broader concept of a passive NFM perception. In a similar vein, feeling informed also only shows (positive) relationships with the amount of news websites in web-browsing histories but not with the amount of news in Facebook feeds. This particularly highlights the importance of observational or, ideally, multimethod studies to investigate (online) news use. While these indications need to be interpreted with caution given the limitations of the present study (more on this below), they echo a news environment driven by "shareworthiness" (Trilling et al., 2017) and a multifarious diffusion of information (Brosius et al., 2019). In such environments, news outlets facing fragmented audiences tend to address social media users separately (e.g., following different priorities of issues; Haim, 2019), yet with an engaging tone to have users spread their news throughout individual networks. A single news item, then, takes various routes in various contexts to reach users in various situations-in short, it diffuses throughout differently structured social networks. In contrast to a universally valid explanation, such an environment promotes a multitude of potential access routes to news for people (and vice versa). The absence of relevant relationships with the dimension "relying on peers" vis-à-vis small relationships of the other NFM dimensions with news only in web browsing rather than with news in Facebook feeds seem to support this notion. Instead, the long-standing indicators for active news use, political knowledge and interest, show clear associations with news exposure in both web browsing and Facebook feeds. Taken together, it seems more likely that online news use is shaped by a confluence of traditional factors and more diffuse interpersonal processes. A passive NFM perception might, therefore, help to further elaborate on traditional reasons for active news use but not so much to explain actual news exposure. With regard to explaining actual online news exposure, the conceptual ambiguity between the NFM dimensions may be the reason for inconclusive or mixed findings. Of course, this multi-method study also comes with limitations. As is typical for self-selected samples, opt-in biases need to be considered. Particularly, the subsample of 246 Facebook users is arguably even more biased due to self-selection than our starting sample of 1,121 web-tracking panelists (which, notably, already differs in several sociodemographic dimensions from national census statistics; see Online Appendix, Section 1). While respective biases need to be considered with great care (Jürgens et al., 2020;Sen et al., 2019), the consent rate of 22% for the use of the browser plug-in is in line with a previous study (Haim & Nienierza, 2019). 10 Moreover, technical limitations make it impossible to track Facebook use within mobile apps. Consequently, our Facebook data are limited to desktop usage. When asked in the survey, respondents reported their share of Facebook use through desktop computers to be 46% on average. Importantly, this share is higher (i.e., 54%) among those participants who also installed the Facebook-tracking browser plug-in. Moreover, the data we have only capture news exposure, meaning that it is not clear whether or to what degree participants actually read and processed the respective news items. Finally, we broadly conceptualized and measured "news" content only at the domain level. Especially Facebook posts and websites of commercial broadcasters contain considerable shares of nonnews content, such as entertainment TV shows. Future research will have to derive even more fine-grained measures of news exposure by also looking at the article content. Despite these limitations, this study adds to existing knowledge about NFM perceptions and online news exposure, as it shows that other factors, such as education, political interest, and political knowledge, are better predictors of observed (instead of self-reported) exposure to online news. Notably, this is true both for news exposure via Facebook news feeds as well as through web browsing. Beyond that, the study also demonstrates the added empirical value of distinguishing the dimensions of NFM perception over combining them into one single factor. 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. Funding The author(s) received no financial support for the research, authorship, and/or publication of this article. Supplemental Material Supplemental material for this article is available online. Notes 1. For data-privacy reasons, we cannot share the raw tracking data but only aggregated data files. 2. German Audit Bureau of Circulation (IVW; https://www. ivw.eu/nachrichten/digital) is one of two major providers in Germany of neutral and comparable audience metrics for advertisers. Aside from other categories, it maintains one of the country's largest datasets on news-traffic rankings, which includes the websites of all print outlets as well as a large share of other news outlets (Hasebrink, 2006). 3. Slightly deviating from the set of control variables included in the study by Gil de Zúñiga et al. (2017), we omitted "race," as it is not commonly included in German survey studies. 4. The discrepancy between those who agreed to install (48%) and those who actually installed (27%) the plugin, to a large extent, likely arises from technological restrictions, such as that the plug-in was available only for Firefox and Chrome and did not work with smartphones. 5. The main results are similar in ordinary least squares (OLS) regressions, yet with more uncertainty associated with coefficients and standard errors. 6. To address potential multicollinearity among independent variables (see Online Appendix, Section 1, Figure A3), we also present all models in versions with either only political knowledge or only political interest in Online Appendix (Section 3). With the exception of some minor changes in the size of some of the coefficients, however, the outcomes across these different models are stable. 7. This also holds when only including political knowledge or interest (Online Appendix, Section 3, Table A9). 8. Again, these findings hold when only including either political knowledge or interest (Online Appendix, Section 3, Table A10). 9. This potential confounding between NFM and news-use measures relates to the fact that the two measures are likely to influence each other, as people who report to rarely use news might, for example, rationalize that they rely more on their friends for staying informed. 10. As a further robustness check, we estimated both regression models for the news website visits (i.e., Models 1 and 2 in Table 1) also for the subsample of Facebook plug-in users (Online Appendix, Section 3, Table A11).	2021-09-29T15:25:21.453Z	2021-07-01T00:00:00.000	{ "year": 2021, "sha1": "b011ccb50bb517f745d9fb3709e3895152fb87e6", "oa_license": "CCBYNC", "oa_url": "https://journals.sagepub.com/doi/pdf/10.1177/20563051211033820", "oa_status": "GOLD", "pdf_src": "Sage", "pdf_hash": "c60434fdc3d856ea2367f61e939b0311e283e1ff", "s2fieldsofstudy": [ "Political Science", "Sociology" ], "extfieldsofstudy": [] }
18457850	pes2o/s2orc	v3-fos-license	Foxp3 Interacts with c-Rel to Mediate NF-κB Repression Expression of the lineage-specific DNA-binding factor Foxp3 controls the development and function of naturally occurring regulatory T cells. Foxp3 has been shown to interact with a multitude of transcriptional regulators including NFAT, NF-κB (p65), Runx1 and RORγt, as well as the histone modification enzymes TIP60, HDAC7 and HDAC9. The sum of these interactions is believed to cause the change in the transcriptional program of regulatory T cells. Here we show that Foxp3 directly or as part of a multimeric complex engages with the NF-κB component c-Rel. We demonstrate that the N-terminal region of Foxp3 is required for the binding of c-Rel, but not NFAT. Conversely, deletion of the forkhead domain causes a loss of interaction with NFAT, but not c-Rel. Our findings are of particular interest, as c-Rel is crucial for the induction of Foxp3 in regulatory T cells during thymic development, but has to be repressed in mature regulatory T cells to maintain their suppressive phenotype. Introduction The NF-kB family of transcription factors is coordinating the expression of a wide variety of genes [1]. The expression of its family member c-Rel is restricted to the haematopoietic and lymphoid cell lineages. In T cells, it is an important mediator of pro-inflammatory gene activation [1]. T cell receptor (TCR) signaling in combination with CD28 co-stimulation leads to the activation of c-Rel, resulting in the up-regulation of a variety of pro-inflammatory cytokines such as IL-6, IL-12, IL-15 and IFN-c [2]. Furthermore, c-Rel plays a key role in the transcriptional activation of IL-2 [3,4]. Indeed, one of the defining features of c-Rel deficient mutant mice is the severe abrogation of IL-2 expression in the T cell compartment [5]. Somewhat surprisingly, c-Rel also appears to play a central role in the development of regulatory T (T R ) cells [6][7][8][9][10][11][12]. TCR signaling in thymic T cells results in the formation of a c-Rel enhanceosome at the Foxp3 promoter. Foxp3 is exclusively expressed in regulatory T (T R ) cells and is considered to be a master regulator of T R cell function [13,14]. It was initially thought to be a transcriptional repressor [15], but it is now generally believed to work as both a repressor and an activator [16][17][18]. It plays a key role in the development and function of T R cells [19][20][21]. FOXP3 mutants in human patients lack functional, thymically derived CD4 + CD25 + regulatory T R cells and develop a severe systemic immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX) [22]. While lack of Foxp3 leads to a deficiency in functional T R cells, ectopic expression of the gene in conventional T H cells leads to a gain of T R cell phenotype [19,20]. The fact that the c-Rel is involved in both, the proinflammatory transcriptional program and the opening of the Foxp3 locus required for T R cells development appears to be paradoxical. Indeed, over-expression of c-Rel in T R cells handicaps their suppressive activity and function [7]. This suggests that whilst c-Rel is required for the development of T R cells through activation of Foxp3 transcription, its pro-inflammatory activities have to be repressed in mature T R cells. The fact that Foxp3 exerts its function by interacting with a variety of transcription factors such as NFAT [18,23], Runx1 [24] and RORct [25,26], all of which play important roles in the development and function of conventional T H cells, makes it a prime candidate for the suppression of c-Rel. Here we show that Foxp3 interacts with c-Rel and thereby represses c-Rel mediated NF-kB activation. Foxp3 inhibits c-Rel mediated NF-kB activation To assess the role of Foxp3 in the suppression of c-Rel, we transiently transfected 293ET cells with plasmids encoding (i) an NF-kB-reporter m3p-firluc[NF-kB], (ii) an internal control reporter pRL-TK, (iii) FLAG-tagged c-Rel (m5p[FLAG-c-Rel]) or p65 (m5p[FLAG-p65]) and (iv) increasing amounts of HAtagged Foxp3 (P8[HA-Foxp3]) or GFP as a control gene (P8[control]). The resulting activity of firefly luciferase from the NF-kB-reporter was normalized to the activity of renilla luciferase arising from the control reporter. 48 h after transfection with m5p[FLAG-p65] an ,8-fold induction of NF-kB activation was observed, compared to cells transfected with P8[control] (Fig. 1A). Transfection with m5p[FLAG-c-Rel] resulted in a ,2-fold induction of NF-kB activation (Fig. 1B). To investigate the effect of Foxp3 on NF-kB activation, we co-transfected increasing amounts of P8[HA-Foxp3] or P8[control]. p65-mediated NF-kB activation was only mildly affected by co-expression of Foxp3 (Fig. 1A), whereas c-Rel-mediated NF-kB activation was severely inhibited (Fig. 1B). The difference in susceptibility to Foxp3mediated repression remained even when we lowered the amount of m5p[FLAG-p65] used for the transfection (Fig. S1). While Foxp3 appears to affect both p65 and c-Rel mediated NF-kB activation, the effect on the latter is substantially more marked. Foxp3 interacts with c-Rel Next, we investigated whether the observed Foxp3-mediated inhibition of NF-kB activity might be due to Foxp3 binding to c-Rel. Using LUMIER assays with lysates prepared from 293ET cells transiently transfected with m5p[FLAG-Foxp3] and m5p[lucc-Rel], we were able to detect an interaction between Foxp3 and c-Rel, which was more marked in nuclear than cytoplasmic extracts (Fig. 1C). Removal of the FKH domain (Foxp3 DFKH ) led to a substantial increase in the interaction detected in crude lysates, which could be counteracted by the addition of a nuclear localization signal (Foxp3 DFKH+NLS ) to Foxp3 DFKH (Fig. 1D). Together the data suggest that Foxp3, directly or as part of a multimeric complex interacts with c-Rel in a FKH independent fashion. Indeed, deletion of the FKH domain leads to an increase in the amount of protein in the cytoplasm and thus appears to facilitate the interaction between the proteins. Regions of Foxp3 required for interaction with c-Rel To further characterize the interaction between Foxp3 and c-Rel we performed a mutational analysis of Foxp3, assessing the domains required for the interaction between the two proteins. As before, we used LUMIER assays with total cell lysates prepared from 293ET cells transiently transfected with the various Foxp3 mutants (m5p[FLAG-Foxp3 mut ]) and m5p[luc-c-Rel]. As we had already established that the FKH domain was dispensable for binding to c-Rel, we based the various deletion and point mutants on Foxp3 DFKH (Fig. 2). Deletion of exon 1 (Foxp3 DFKH: De1 ) was sufficient to almost entirely abrogate c-Rel binding (Fig. 2). Deletion of exon 6, which contains a coiled-coil (Foxp3 DFKH: De6 ), abolished binding altogether, suggesting that dimerization of Foxp3 is required. In contrast, individual deletions of exons 2 or 3 only partially reduced binding. No interaction between Foxp3 DFKH: De5 and c-Rel could be detected, but this particular mutant was expressed poorly (Fig. 2, arrow) making an interpretation of this result all but impossible. Deletion of exon 8 (Foxp3 DFKH: De8 ), which has been reported to be involved in Runx1 binding [24], also led to a loss of interaction between c-Rel and Foxp3. The transactivation domain of c-Rel is required for interaction with Foxp3 c-Rel, like p65, belongs to the NF-kB family of proteins, which are defined by highly conserved N-terminal DNA-binding domains (REL homology domains; RHD) [1]. They are sequestered in the cytoplasm as homo-or heterodimers through association with IkB, which binds to their IPT (Ig-like-plexinstranscription factors) domains [27,28]. The C-termini of c-Rel and p65 contain non-conserved transactivation domains, which interact with various components of the basal transcription apparatus [28,29]. To map the regions of c-Rel required for Foxp3 binding, we performed LUMIER assays with lysates prepared from 293ET cells transiently transfected with m5p[FLAG-Foxp3 DFKH ] and various c-Rel mutants (m5p[luc-c-Rel mut ]). Deletion of the Nterminal 167 amino acids (c-Rel 118-588 ) did not affect the interaction of c-Rel with Foxp3 DFKH , indicating that the RHD is not required for the interaction (Fig. 3). However, if the deletions extended to the IPT domain (c-Rel 221-588 and c-Rel 278-588 ) the interaction with Foxp3 DFKH was abolished. Likewise, a C-terminal deletion of the last 63 amino acids of c-Rel (c-Rel 1-525 ) was sufficient to severely compromise binding to Foxp3 DFKH and further deletion of the transactivation domain (c-Rel 1-399 and c-Rel 1-287 ) completely abolished the interaction between the two proteins ( Fig. 3). This suggest, that c-Rel binds to Foxp3 via its Cterminal transactivation domain and the loss of binding in the absence of the entire RHD and IPT indicates that dimerization of c-Rel is required. c-Rel binding to Foxp3 is not required for its interaction with NFAT Foxp3 has been described to form a complex with NFAT via its FKH domain [18]. Mutations that were predicted to disrupt the interaction between the two proteins, interfered with the ability of Foxp3 to inhibit the expression of several cytokines, including IL-2, and to up-regulate T R cell associated markers such as CD25 and CTLA-4 [18]. However, an N-terminal deletion mutant of Foxp3 was also incapable of performing these functions, suggesting that binding of further factors might be required [18]. It remained unclear whether this is facilitated through interaction with a different region of NFAT or whether this is independent of NFAT [18]. To test whether the N-terminal region of Foxp3 is involved in NFAT binding, we performed a mutational analysis of Foxp3 assessing the domains required for this interaction. As previously shown [18], we found that deletion of the FKH domain (Foxp3 DFKH ) abrogated NFAT binding (Fig. 4). In contrast, deletion of exon 1 (Foxp3 De1 ) had no effect on the interaction between the two proteins. As was the case for c-Rel binding, the deletion of the coiled-coil of Foxp3 (Foxp3 De6 ) had a marked effect on NFAT binding. Individual deletions of all other exons led only to a partial reduction in the interaction between Foxp3 and NFAT (Fig. 4). In agreement with previous studies [18], this shows that the FKH is required for NFAT binding, whereas the N-terminal region is dispensable for complex formation between the two proteins. Furthermore, this indicates that the interaction of Foxp3 with c-Rel can occur in the absence of its interaction with NFAT and vice versa. The N-terminal region of Foxp3 is required for NF-kB repression Our data suggests that the N-terminal region plays a crucial role in c-Rel binding, but is not required for NFAT interaction, which binds to the FKH domain. Thus, we investigated the role of these domains in repressing c-Rel mediated NF-kB activity. We transiently transfected 293ET cells with the NF-kB reporter plasmid and the relevant Foxp3 mutant, as described above. Deletion of exon 1 (Foxp3 De1 ), which disrupts c-Rel binding (Fig. 2), led to a marked, albeit not complete, reduction in the repression of c-Rel induced NF-kB activation (Fig. 5A). Disruption of NFAT binding through the deletion of the FKH domain (Foxp3 DFKH ) had only a minor effect on the potency of the repression of c-Rel mediated NF-kB activation (Fig. 5B). c-Rel causes a shuttling of Foxp3 DFKH into the nucleus We were intrigued to find that Foxp3 DFKH , despite having lost its nuclear localization [30], was still capable of regulating c-Rel mediated NF-kB activation. As Foxp3 is a DNA binding factor, we would have expected that nuclear localization is a prerequisite for its activity as a transcriptional regulator [31]. We hypothesized that either Foxp3 DFKH retains c-Rel in the cytoplasm or c-Rel shuttles Foxp3 DFKH into the nucleus. To test this, we transfected 293ET cells with plasmids encoding c-Rel (m5p[c-Rel]) and a GFP-tagged Foxp3 DFKH (m5p[GFP-Foxp3 DFKH ]) and analyzed the sub-cellular localization of GFP-Foxp3 DFKH by confocal microscopy. In 293ET cells a large proportion of ectopically expressed c-Rel localizes to the nucleus in the absence of any kind of stimulation (data not shown). Whilst in the absence of c-Rel, GFP-Foxp3 DFKH was predominantly found in the cytoplasm (Fig. 5C), co-expression of c-Rel caused a redistribution of GFP-Foxp3 DFKH into the nucleus (Fig. 5D). Notably, deletion of either the RHD (c-Rel 168-588 ) or transactivation domain (c-Rel 1-287 ) of c-Rel led to a loss of this redistribution (Fig. 5E). The former causes a loss of nuclear translocation of c-Rel [1], whereas the latter leads to a loss of Foxp3 binding (Fig. 3). Wild type Foxp3 is already nuclear by itself and co-transfection of c-Rel didn't change this (data not shown). Discussion Here we show that Foxp3 either directly or as part of a larger multimeric complex [32] interacts with c-Rel. We found that deletion of the N-terminal region of Foxp3 led to a loss of inhibition of c-Rel-mediated NF-kB activity. Furthermore, we demonstrate that this region is required for binding of the transactivation domain of c-Rel, which interacts with components of the basal transcription apparatus such as TATA-binding proteins, p300 and cyclic-AMP-response element (CREB)-binding protein (CBP) [33]. An attractive interpretation would be that Foxp3 competes with such factors for binding, which would result in an inhibition of transcription. However, simple competition for interaction with c-Rel is unlikely to be the entire explanation, as while Foxp3 DFKH is still capable of binding c-Rel, it has lost the ability to inhibit IL-2 expression [18]. In the case of IL-2 and several other Foxp3 target genes, NFAT is an essential component of transcriptional control [18]. In the absence of Foxp3, both c-Rel and NFAT activate IL-2 transcription [34], whereas in the presence of Foxp3 both appear to be involved in its repression. Foxp3 De1 binds NFAT, but has lost its ability to interact with c-Rel. Conversely Foxp3 DFKH interacts with c-Rel, but not NFAT. This demonstrates that either factor can bind to Foxp3 without the other being present, suggesting that in some cases the individual interactions might be sufficient for the regulation of target gene transcription. However, we observed an increase in the binding of c-Rel to Foxp3 when we deleted its FKH and thereby interfered with NFAT binding. However, as this also leads to a repositioning of Foxp3 into the cytoplasm, an experimental bias favoring an interaction with c-Rel might be introduced. An alternative interpretation would be that under physiological conditions c-Rel and NFAT directly or indirectly compete for binding to Foxp3. While we found that c-Rel can shuttle Foxp3 DFKH into the nucleus, this might not always be sufficient to reconstitute its function. Indeed, it has already been shown that the FKH is crucial for nuclear localization, DNA binding as well as interaction with NFAT [18]. We consider the shuttling of Foxp3 DFKH by c-Rel to be an experimental peculiarity. Nevertheless, it points towards a potential role of c-Rel in transporting Foxp3 towards particular target genes, which warrants further investigations in future. Foxp3 and c-Rel predominantly exist as dimers. In the case of Foxp3, this is mediated by the coiled-coil [35]. Point mutations found in IPEX patients, which lie in critical residues (DK249 or DE250) within the coiled-coil, are defective in dimerization and lead to a loss of Foxp3 function [35]. For c-Rel, the dimerization is dependent on the presence of the IPT domain [36]. We show that disruption of either the coiled-coil in Foxp3 or IPT in c-Rel leads to a loss of complex formation between the proteins. This suggests We found that while the FKH domain of Foxp3 is not necessary for c-Rel binding, deletion of exon 8, lying immediately upstream of the FKH is. This region of Foxp3 is required for its interaction with the transcription factor Runx1, which also has been implicated in IL-2 regulation [24]. However, we cannot exclude that the progressive deletion of the C-terminus of Foxp3 leads to a loss of structural integrity per se, rather than invoking a requirement for cooperative binding of Runx1 and c-Rel. It has been proposed that Foxp3 forms high molecular weight, multimeric transcriptional complexes [32]. From this, and other studies [18,23,24,26,37], it is becoming increasingly clear that Foxp3 works by hijacking already resident transcriptional mechanisms that are important for T cell activation. In particular the N-terminal region has been shown to interact with many factors including RORct, RORa [25,26], Tip60, HDAC7/9 [37] and Eos [38]. It might well be that the composition of these complexes differs depending on the developmental stage and activation status of the cell, as well as the target gene. In some cases, the presence of Foxp3 may lead to a switch from gene activation to repression, and in others it may lead to enhanced transcription. This might explain how a single factor, Foxp3, can redirect the transcriptional program from that of a proinflammatory T H cell to that of a suppressive T R cell -two diametrically opposed immunological T cell functions. Cells 293ET human embryonic kidney cells [39] were grown in IMDM supplemented with 10% FCS 50 mM b-mercaptoethanol and 50 mg/ml gentamicin. Eighteen hours prior to transfection, 293ET cells were seeded into plates at a density of 1.5610 5 cells/ cm 2 . Transfections were performed using standard calcium phosphate precipitation. NF-kB reporter assays 293ET cells were seeded and eighteen hours later the cells were transfected with 50 ng of m3p-firluc[NF-kB], 20 ng of pRL-TK control vector and the indicated plasmids. The transfected cells were analyzed 48 h later using the Dual Luciferase Reporter Assay system (Promega) following the manufacturer's instructions. Preparation of cytoplasmic and nuclear extracts Preparation of 293ET extracts was performed on ice and all centrifugations were performed at 4uC. Crude lysates were prepared by addition of lysis buffer (20 mM Tris, 150 mM NaCl, 5 mM EDTA, 0.1% Triton X-100, 10% glycerol, 5 mg/ml leupeptin, 2 mg/ml aprotinin, 1 mM benzamidine, 1 mM PMSF) and incubation for 15 minutes. The extracts were cleared by centrifugation at 16000 xg for 15 minutes. Where cytoplasmic and nuclear extracts were required, cells were collected, washed with PBS and lysed in hypotonic buffer (10 mM HEPES pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 5 mg/ml leupeptin, 2 mg/ml aprotinin, 1 mM benzamidine, 1 mM PMSF) for 5 minutes. NP-40 was added to a final concentration of 0.6%, the cells were vortexed briefly and nuclei were collected by centrifugation at 16000 xg for 1 minute. The nuclear pellet was vortexed for 15 minutes in cold hypertonic buffer (20 mM HEPES pH 7.9, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 5 mg/ml leupeptin, 2 mg/ml aprotinin, 1 mM benzamidine, 1 mM PMSF) to release nuclear proteins. Following centrifugation at 16000 xg for 15 minutes, the supernatant was collected and where necessary, the protein concentration was quantified using Bradford (BIO-RAD). Unless used directly, the extracts were stored at -70uC. Luminescence-based mammalian interactome mapping (LUMIER) 293ET cells were co-transfected with a FLAG-tagged construct and a Renilla Luciferase-tagged construct. Extracts were incubated with anti-FLAG M2 Affinity Gel (Sigma) for 2 h, rotating at 4uC. Unbound protein was removed by washing 5 times with 20 mM Tris, 150 mM NaCl, 5 mM EDTA, 0.1% Triton X-100, 10% glycerol and centrifugation at 6000 xg for 30 seconds. Bound proteins were eluted from the affinity gel by incubating with 150 mg/ml FLAG peptide (Sigma) for 30 minutes, rotating at 4uC. A sample of the lysate (input) and a sample of the gel eluate (immunoprecipitated) were measured for Renilla Luciferase activity using the Renilla Luciferase Assay Kit (Promega) as described in the manufacturer's protocol. The results are shown as the ratio of immunoprecipitated versus input Renilla Luciferase activity. Western blots Extracts were separated on NuPAGE 4-12% Bis-Tris gels (Invitrogen) and blotted onto nitrocellulose membrane (Whatman) as described by the manufacturer. The membrane was blocked for 1 h at room temperature with 3% milk and incubated at 4uC overnight with one of the following antibodies: 1:1000 of rabbit anti-c-Rel (sc-71, Santa Cruz Biotechnology), 1:5000 of mouse anti-FLAG M2 (Sigma), 1:1000 of mouse anti-HA.11 (Covance Research Products). The membrane was washed and incubated with anti-mouse or anti-rabbit HRP-conjugated secondary antibody (DakoCytomation) for 2 h at room temperature. HRP was detected by incubation of the membrane with ECL Detection Reagents (Amersham Biosciences). Translocation of Foxp3 DFKH 293ET cells were co-transfected with m5p[GFP-Foxp3 DFKH ] in the presence or absence of m5p[c-Rel]. Twenty-four hours later, 1610 4 cells were transferred into Lab-Tec chambered coverslips (Nalge Nunc International). After an additional 24 h, the subcellular localization was assessed by observing for cytosolic and nuclear presence of GFP-Foxp3 DFKH by using a BioRad Radiance 2000 confocal microscope.	2018-04-03T01:33:33.549Z	2011-04-07T00:00:00.000	{ "year": 2011, "sha1": "32e43ef7b9ab8de5ec97482c8ced473ae2e98064", "oa_license": "CCBY", "oa_url": "https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0018670&type=printable", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "32e43ef7b9ab8de5ec97482c8ced473ae2e98064", "s2fieldsofstudy": [ "Biology", "Medicine" ], "extfieldsofstudy": [ "Biology", "Medicine" ] }
246339528	pes2o/s2orc	v3-fos-license	Long-memory recursive prediction error method for identification of continuous-time fractional models This paper deals with recursive continuous-time system identification using fractional-order models. Long-memory recursive prediction error method is proposed for recursive estimation of all parameters of fractional-order models. When differentiation orders are assumed known, least squares and prediction error methods, being direct extensions to fractional-order models of the classic methods used for integer-order models, are compared to our new method, the long-memory recursive prediction error method. Given the long-memory property of fractional models, Monte Carlo simulations prove the efficiency of our proposed algorithm. Then, when the differentiation orders are unknown, two-stage algorithms are necessary for both parameter and differentiation-order estimation. The performances of the new proposed recursive algorithm are studied through Monte Carlo simulations. Finally, the proposed algorithm is validated on a biological example where heat transfers in lungs are modeled by using thermal two-port network formalism with fractional models. Introduction Fractional calculus was explored in the twentieth century as a way to model diffusion phenomena for different geometries [32,33]. It has been shown that diffusion phenomena can be modeled by using fractionalorder transfer functions with differentiation orders that are multiples of 0.5. In electro-chemistry, charge diffusion in acid batteries is governed by Randles model, which involve a half-order fractional integrator [38]. Studies have also shown a half-order fractional integrator model for semi-infinite thermal systems [1]. These physical phenomena of diffusion have led to the theoretical conception of constant-phase elements [16,30] that may be used as building blocks for circuit models. Constant phase elements may be used to model intestine tissue behavior [9], cardiac tissue [21], porous films [4] and lung impedance behavior [14,44]. New challenges appear on system identification: [46] provides new paradigms and challenges in system identification such as broader types of uncertainties, networked systems or even data explosion; [35] proposes kernel methods; [20] gives new kernel-based regularization methods; [2] proposes to estimate time delay with sampled limit cycle in frequency domain, etc. There is a vast literature concerning the analysis and identification of discrete-time models [19]. However, note that such type of models may lead to parameters that lack physical meaning. The sampling zero problem [48] related to discretization should not be neglected either. As a model with physical meaning is sought, a continuous-time identification is better suited. The main challenge with adapting system identification for continuous-time models from sampled data relies on the differentiation approximations for the information matrix [10]. When dealing with fractional models, this information matrix will contain approximations of fractional-order differentiations from experimental data. Since the 1990s, substantial contributions have been made for fractional-order system identification by estimating the coefficients of a fractional-order transfer function. In continuous-time identification, methods relying on state variable filters, least squares and instrumental variables were introduced in [3,31,42]. These techniques have been applied for identifying thermal diffusion on an aluminum rod [24,42], charge diffusion lithium-ion batteries [8], muscle relaxation [40], heat transfers in biological system [44]. Most of the fractional system identification methods work offline, which means that the whole data of an experiment are used to provide parameter estimation. In some cases, online parameter estimation is needed such as in open-heart surgery where lung thermal online system identification will prevent pulmonary cell dying. The most famous online identification method is the recursive least squares. Recursive prediction error methods are also available in the literature [19]. Padilla [34] provides a recursive version of LSSVF and RIVC for integer-order systems. By starting from Le Lay's method and online estimation concepts, a method to estimate coefficients of a fractionalorder transfer function was proposed in [6]. Fractionalorder models may also require the estimation of one or several differentiation orders, which has already been studied in the offline cases by means of gradient descent for single-input single-output fractional systems [42] or even in the multiple-input single-output fractional systems [26,45]. An extension of Djouambi's identification method has also been proposed to identify the fractional orders by using a scheme similar to gradient descent in [13]. The aim of this paper is to first provide efficient recursive all parameter estimation methods, namely both the coefficients and the differentiation orders. First of all, when the differentiation orders are assumed known, a new method called long-memory recursive prediction error method is proposed and compared to existing methods through Monte Carlo simulation analysis. A specific aspect of fractional differential equation modeling is the determination of the differentiation orders. In system identification of classic rational models, as the model order remains unchanged, only the coefficients are estimated. When estimating both coefficients and differentiation orders of fractional models, this task is not trivial: Indeed, in an online recursive algorithm, the model order changes at each iteration. Therefore, the long-memory prediction error method (LMRPEM) is extended for estimating both parameters and differentiation orders. Monte Carlo simulation analysis is again proposed to analyze the statistical properties of our new method. Finally, the new LMR-PEM algorithm is applied for identifying heat transfers on a human bronchus. Real experiments for modeling heat transfers in sheep lung were previously proposed in [44] for system identification by using black box models through Havriliak-Negami functions. It is now proposed to use a more realistic model, which is closer to the physics that governs heat transfers in human lungs. By using thermal two-port network formalism [22], which is none other than the solving of the heat equation on a unidirectional medium, and fractional-order impedance approximations [7], heat transfers in lungs of a human main bronchus can be modeled as an equivalent T network (see Fig. 5). All analytical developments for heat transfer modeling are detailed in [7]. The paper is organized as follows: Sect. 2 describes the problem formulation for recursive fractional system identification, Sect. 3 presents the recursive estimation methods for coefficient estimation of fractional systems, and then, Sect. 4 introduces a recursive estimation method for all-parameter estimation by combining both coefficient and fractional differentiation-order estimation. Section 5 presents an application to thermal transfer in human lungs, and conclusions are drawn in Sect. 6. Mathematical background Many definitions exist for fractional differentiations [11,15,39]. Grünwald [12] and Letnikov [17] independently developed a definition for an arbitrary-order differentiation in terms of a convergent series: where . stands for the floor operator, p is the differential operator p = d dt , and α j is Newton's binomial coefficient generalized to real numbers: where is Euler's Gamma function. The slow convergence of this binomial justifies why fractional-order operators are well suited for long-memory behavior modeling, as the whole past of the function is required for its definition. By replacing h by the sampling time T s , an implementable approximation of the fractional differentiation is obtained: Note that the error terms are proportional to the sampling time [36,Section 7.4]. Consequently, the smaller the sampling time, the lesser the approximation errors. 1 If one considers a case with null initial conditions, the Laplace transform of the fractional differentiation may lead to a simple expression [36,Section 2.8.4]: For a fractional single-input single-output (SISO) model, a fractional differential equation relates the input u(t) to the output y(t): where (a i , b j ) are real numbers and α 1 < α 2 < . . . < α N and β 0 < β 1 < . . . < β M are non-integer positive numbers. The Laplace transform of (5) gives the following fractional-order transfer function model: Definition 1 A SISO system G(s) (6) is commensurate of order ν, if all differentiation orders are integer multiples of ν: where all j = β j ν and i = α i ν are integer numbers. Stability of fractional-order systems has been analyzed in different contexts. The most well-known stability criterion was established by Matignon [25] and allows to check the stability of a commensurate-order system through the location of its s ν -poles. The original theorem was established for commensurate orders ν, for 0 < ν < 1, but was extended for orders between 1 and 2 in [29]. Orders with commensurate orders beyond 2 are unstable (see [23]). Further extensions have been developed in order to check stability of noncommensurate systems [37]. Theorem 1 Let S be a commensurate transfer function and ν its commensurate order. G(s) = Q ν (s) P ν (s) is BIBOstable if and only if: and, for every pole s k (P ν (s k ) = 0): Problem formulation Let be an input u(t) and its noise-free output y(t) related by equation (6), where A(s) and B(s) are assumed coprime and the system stable. Data are collected at regular samples from t = 0 to the current time t = kT s with the sampling time T s . Furthermore, the number of data samples is assumed to be large enough so that the convergence of the estimated parameters to the true ones is guaranteed. Moreover, the quasistationary input signal {u(t), 0 ≤ t ≤ kT s } applied to the system is persistently exciting and gives rise to an output signal {y(t), 0 ≤ t ≤ kT s }. The output data are corrupted by an additive white measurement noise ξ(t) due to experiment and sensor imperfections. The noise is assumed normally distributed with a zero mean and variance, σ 2 , considered at discrete instants. The complete system equation can be written as: where y * (t) is the noisy sampled value of the unobserved noise free output y(t). When using model (6), the parameter vector, θ , is defined as where ρ is the vector of N + M + 1 coefficients: and μ is the vector of N + M +1 differentiation orders: The total number of parameters to be estimated is then 2×(N + M +1). If N or M are too high, the complexity may be too high and nonlinear optimization algorithms may fail to converge to a global minima. In order to reduce the number of parameters, a commensurate-order model (7) can be used, where the differentiation order vector μ is reduced to a single parameter ν. The parameter vector θ is then reduced to N + M + 1 coefficients as in (12) added up of the commensurate order μ = ν. Coefficient estimation of fractional-order systems In this section, all differentiation orders are assumed known; therefore, only the coefficients are estimated. The parameter vectorθ boils down to: 3.1 Recursive least squares with state variable filter (RLSSVF) The estimated output can be expressed in a linear form as: where the regression vector is as follows A suitable error function is defined as: where the parameter vectorθ minimizes the quadratic error function: which can be put under the form of an optimization problem: It is well known that the least squares are biased in the presence of noise, and therefore, state variable filters (SVFs) can be used to obtain a filtered input u f (t) and output y f (t). This filtering allows to reduce the noise influence on the estimation and is normally a lowpass filter tuned to emphasize a particular bandwidth. A typical choice is the low-pass Poisson filter: The filter should be designed by respecting the following [3]: where · is the ceil operator. It has been shown in [5] that the cutoff frequency ω f may have an influence on the convergence rate of the algorithm, as well as sampling time and input. This frequency is typically chosen to be somewhat larger to the system bandwidth [34]. This means that a priori estimation or tuning of the system bandwidth may be required to get better results. The error function (t) is now replaced by its filtered version f (t): where the filtered regression vector φ * . (23) and the filtered input u f (t) and output y * f (t) are defined as: Inspired by the recursive least squares [19, chap. 11], the RLSSVF algorithm is obtained for fractional-order systems: where γ is a forgetting factor that can give more or less weight to the most recent past of signals and F is the covariance matrix. Remark: the forgetting factor γ may actually be a function of time and be adaptive as identification is running [41]. In this study, it will be taken as a constant and equal to unity. Recursive prediction error method (RPEM) By following the prediction error method, a suitable error function is given by the output error: where the estimated outputŷ(t) is computed as: The quadratic error function defined in (18) is minimized. By taking the gradient of the criterion: with its error sensitivity is computed by: where and Inspired by the recursive least squares [19, chap. 11], the RPEM algorithm is obtained for fractional-order systems: where γ ρ is a refining gain analogous to the step in a gradient descent. This forgetting factor γ ρ may actually be adapted during the identification or be a function of time. However, the determination of an adequate γ ρ may prove to be non-trivial, as stated in [18]. For the present work, it will be taken as a constant. Contribution with long-memory recursive prediction error method (LMRPEM) Fractional-order systems have a natural long-memory behavior as the whole past of a signal differentiation is needed, and therefore, an extended error with long memory is more suited: This error signal˜ will include errors at all instants from t = 0 to the current time t = kT s and will be increased by one data-point per iteration. Consequently, the resulting gradientψ ρ (kT s ) is now a matrix: where k indicates the present iteration. By introducing the extended measured outputỸ * (kT s ) and the extended estimated outputỸ (kT s ) as: the long-memory recursive prediction error method (LMRPEM) is proposed for fractional-order systems: where again, γ ρ is a refining gain analogous to the step in a gradient descent. As previously stated in Sect. 3.2, γ ρ will be taken as a constant. True system description Let us consider the following true system model: with K = 1, ζ = −0.5, ω 0 = 0.5rad/s and ν = 0.5. Such a fractional model is called a second kind fractional-order system [23] and has a resonant peak around ω r = 0.45rad/s (see Fig. 1). In fact, this is a typical kind of model structure that well represents diffusive phenomenon such as thermal diffusion [43]. Moreover, such a model structure is similar to the one that models diffusive phenomenon in lungs (see the human bronchus heat transfer application in Sect. 5). The model structure is chosen such as the true one (39) and is expressed by: with differentiation orders assumed known and fixed just like the true system (39), where ν = 0.5. The aim is to estimate the system coefficients and to see the statistical properties of the proposed LMRPEM algorithm through Monte Carlo analysis, with N exp = 100 runs and different signal-to-noise (SNR) levels. The input signal is a pseudo-random binary sequence oscillating between − 5 and 5 and containing 1000 values with a sampling time T s = 0.01s. All simulations Figure 2 shows the input used for simulations, as well as the noise-free output and the noisy output with a SNR = 15 dB. All simulations have been carried out with the CRONE toolbox developed in MATLAB , which is dedicated to fractional calculus, fractional system simulation and system identification with fractional models. 2 Coefficient estimation with known differentiation orders From Fig. 1, the true system is resonant in the frequency range [1, 10]rad/s. Therefore, the Poisson filter (20) is set with a cutoff frequency of ω f = 10rad/s and as max(β M , α N ) = 1, N f = 2. Note that in a realistic scenario such information may not be available and an empirical estimation of the cutoff frequency ω f and the order N f will be required. The Monte Carlo simulation results for 100 runs are summarized in Table 1. The RLSSVF method lacks robustness with respect to measurement noise. The lesser the SNR is, the greater the estimation bias gets and the greater the standard deviations become. Moreover, the RLSSVF is not efficient as the parameters are not correctly estimated for any noise level. An improvement can be obtained by adjusting the cutoff frequency of the filter; however, it should be noted that a frequency ω f extremely close to the system bandwidth would deteriorate the estimation in a practical scenario. The RPEM method (γ ρ was set to 0.01) is more consistent than the RLSSVF method as the bias is less present. However, the lesser the SNR, the greater the bias. The true value of the parameters is always included in the range determined by the standard deviation. On the other hand, a rough factor of 3 can be observed between the standard deviations of SNR = 20 dB and 10 dB. The LMRPEM method (γ ρ was set to 0.01) has the most consistent estimation as the estimation has no bias and the estimation standard deviations (and consequently the estimation variances) are the lowest whatever the noise level. This method has significantly reduced the parameter deviations. Commensurate order with unknown differentiation orders In practice, the differentiation orders are not always known a priori. The commensurate-order influence is evaluated by computing the cost function, defined as the 2 -norm of the normalized output error: where the long-memory error˜ (N T s ,θ) is defined as in (34): from relations (36) and (37). Figure 3 shows the criterion J versus the commensurate order obtained by using LMRPEM method for different values of the commensurate order in order to approximate the true system with a noise-to-signal ratio N S R = −20 dB. As expected, the criterion minimum is obtained at ν = 0.5, which is the true commensurate order. Indeed, when the model is in the same model class as the true system, the minimum of the cost function is found at ν = 0.5 and equals − 20 dB which exactly corresponds to the NSR. Other values of ν lead to non-negligible modeling errors as the coefficient estimation would not converge to the true parameters. For example, at ν = 1, the cost function is around − 10.01 dB, which corresponds to a modeling error around 10 dB. This simulation result motivates to estimate the fractional differentiation orders, as they may considerably influence the estimation. Differentiation-order estimation and all-parameter system identification In this section, the differentiation orders are assumed unknown, as it is often the case in practice. Thus, it is helpful to consider differentiation-order estimation along with coefficient estimation. As the differentiation orders are not expressed in a linear way in (6) or (7), it is necessary to recursively estimate these parameters with a suitable error function. As the LMRPEM method has given the best estimation results in the pre- vious section for coefficient estimation, it is proposed to use this method for the differentiation-order estimation. Then, all parameters are estimated for both, the coefficients and the differentiation-order estimation, through a two-stage algorithm. Differentiation-order estimation In a general context, a vector of differentiation orders could be determined and when dealing with a commensurate model, a single differentiation order is estimated. By following the prediction error method, a suitable error function is given by the output error: and differentiation orders may be recursively adjusted by adding a correction term to the previous iteration: where Recall that the LMRPEM method has given the best results for estimating the coefficients for fractionalorder systems. Consequently, an extended error is used to take into account the long-memory behavior of the system: (46) which turns the error sensitivity into a matrix: For commensurate-order estimation, the gradient boils down to: The error sensitivity function is defined by: where the commensurate-order output sensitivity ∂ŷ ∂ν is expressed by: As time simulation of such a function is non-trivial, the gradient will be numerically computed and the LMR-PEM will be used for differentiation-order estimation: 4.2 All-parameter system identification with a hybrid two-stage algorithm The best previous method for coefficient estimation, namely the LMRPEM, is combined with differentiation-order estimation, through another LMRPEM. This two-stage algorithm, called LMRPEM-2, is used for all-parameter estimation of a fractional-order model. 38) (b) Differentiation order estimation Assuming the coefficient ρ(k) known, compute ν(k) from equation (51) As stated by Young in [47], the tuning of the forgetting factor γ in recursive PEM methods is non-trivial and influences the convergence of the parameters. Consequently, the LMRPEM-2 algorithm 1 requires the tuning of two forgetting factors (one for the coefficients and one for the differentiation orders). Simulation results for recursive all-parameter estimation The same system (39) as detailed in Sect. 3.4.1 is used with different noise levels: ∞ without noise, 20 dB, and 15 dB. The tuning of the γ gains was performed empirically and is given in Table 2. The parameter evolution for an SNR = 20 dB is plotted in Fig. 4 for Algorithm 1. On the commensurateorder estimation, hybrid LMRPEM-2 Algorithm 1 exhibits some oscillations at the beginning. However, the hybrid LMRPEM-2 algorithm tends to the true commensurate order ν = 0.5. Monte Carlo simulation analysis is provided in Table 3. Three levels of noise are considered, and for each level, 100 runs have been carried out. Note that all runs have converged to stable models. In noiseless conditions, the hybrid LMRPEM-2 Algorithm 1 has converged to the true commensurate order, and therefore, the coefficient estimations are pretty accurate in mean. As the noise level increases, hybrid LMRPEM-2 Algorithm 1 presents consistent estimates, with low variance and without bias. The present work shows parameter estimation within a finite data length. In practice, realistic applications may require much longer datasets. One main issue regarding longer datasets lies on the computation time. Grünwald-Letnikov's differentiation is longer to calculate at each iteration when the dataset gets long enough, as the whole past of the signal is used for its computation. If the dataset gets even longer, the computation time will definitely be affected. Therefore, an extremely long dataset would require an extremely long calculation time by the end of the simulation, which would be impossible to implement in a real-time scenario. By allowing some approximation errors, the past of the function used may be limited to a memory length L, as stated by the short memory principle [36]. Application: human bronchus heat transfer system identification Medical and surgery applications need online biological parameter estimation: Time can be highly critical, and more especially, accurate real-time estimation is required. Real experiments for modeling heat transfers in sheep lung were previously proposed in [44] for system identification by using black box model through Havriliak-Negami functions. It is now proposed to use a more realistic model for heat transfers in human lungs. By using thermal two-port network formalism [22], which is none other than the solving of the heat equation on a unidirectional medium, and fractional-order impedance approximations [7], heat transfer in lungs of a human main bronchus can be modeled as an equivalent T network (see Fig. 5). All model developments are detailed in [7]. The bronchus is modeled as an air cylinder with an intermediate length of L = 0.0236 m [28] and radius r ≈ 1 mm [27]. Impedance approximations are given by: and For an insulated output case (φ out = 0), the transfer function relating the output temperature to the heat flux The input signal is a pseudo-random binary sequence oscillating between − 1 and 1 and containing 2000 values with a sampling time T s = 1s. The initialization is θ T = [2, 5, 5, 10] and ν = 0.35. The noise was set to SNR = 20 dB. The input/output data are shown in Fig. 6. This model is set as the true system (54): Hybrid LMRPEM-2 Algorithm 1 is used to estimate the parameters as it is the most efficient algorithm for all-parameter estimation (see Sect. 4.3). γ ρ and γ ν are the same as given in Table 2. The recursive coefficient and differentiation-order estimations are shown in Fig. 7. The commensurate-order convergence is slow but Fig. 7 Parameter estimation for bronchus application very accurate toward the system true commensurate order ν = 0.5. The coefficient convergence exhibits fairly slow convergence rate, especially for a 2 ; however, it tends to its true value toward the end of the simulation. Therefore, when the dataset gets long enough, the convergence to the true parameters is improved. Finally, at the last iteration, the estimated model is expressed as the following: G(s) = 2.519s 0.502 + 1.013 12.66s 1.005 + 2.601s 0.502 + 1 , an estimated model that has well converged to the true parameters such as predicted by the hybrid LMRPEM-2 Algorithm 1. Note that the coefficients have correctly converged to their true values only after the convergence of the commensurate order. Consequently, well estimating the differentiation orders is necessary so that the coefficients well converge, and therefore, physical parameters can be extracted. Conclusions and perspectives A new method called LMRPEM-2 is proposed for allparameter recursive system identification of continuous-time fractional-order models. Least squares (RLSSVF) and prediction error method (RPEM) were compared to a new variant of the RPEM, called LMRPEM. The latter method takes into account the long-memory behavior of fractional-order systems. RLSSVF method showed poor performances, as it provides high bias as noise increases and implies a priori estimation of the system bandwidth. Other methods provided more accurate coefficient estimation as they present no bias. The statistical properties of the LMRPEM have proven to produce the lesser parameter variance as compared to the other methods. Then, for all-parameter estimation, especially for differentiation order estimation, a new hybrid algorithm is introduced: the hybrid LMRPEM-2 for both coefficient and differentiation-order estimation. The LMRPEM-2 algorithm proved to be consistent: Parameter estimations are without bias with very low variance. Also, the longer the dataset, the better the convergence to the true parameters as fractional systems have long-memory behavior. Finally, a biological example was proposed: Heat transfers in lungs are modeled on a human main bronchus by using thermal two-port network formalism with fractional models. Applying the hybrid LMRPEM-2 method has provided consistent estimates without bias and with very low variance. Research perspectives may include further studies regarding a method to tune the forgetting factor γ in a methodical way. The issue regarding computation time with long datasets should also be studied by taking into account the short memory principle as well as the frequency content of the analyzed signals. Real experiments on sheep will be carried out at the Bordeaux University Hospital and IHU Liryc Institute for online system identification of lung heat transfers. Author contributions J.-F. Duhé has developed the recursive algorithms with S. Victor. He also made the simulations. S. Victor, P. Melchior, Y. Abdelmounen and F. Roubertie have contributed to the biological and physiological results. Funding Not applicable. Data availability Enquiries about data availability should be directed to the authors.	2022-01-28T16:55:31.463Z	2022-01-24T00:00:00.000	{ "year": 2022, "sha1": "21fc4692583f862697c2e22d76494ff7e5d3b3aa", "oa_license": "CCBY", "oa_url": "https://oskar-bordeaux.fr/bitstream/20.500.12278/170342/3/IMS_ND_2022_Victor.pdf", "oa_status": "GREEN", "pdf_src": "SpringerNature", "pdf_hash": "241d0b126949179b46f778bcfffd569d04beec08", "s2fieldsofstudy": [ "Mathematics", "Engineering" ], "extfieldsofstudy": [] }
225174933	pes2o/s2orc	v3-fos-license	Warmer winters causes an increase of chlorophyll-a concentration in deeper layers: the opposite role of convection and self-shading on the example of the Black Sea Winter vertical entrainment of deep waters determines not only the amount of nutrients in the upper layers, but also the light conditions in it, through the self-shading mechanism. In this paper, we use Bio-Argo data to demonstrate significant differences in the vertical distribution of chlorophyll-a concentration (Chl) in the Black Sea between a year with cold winter (2017) and a year with warm winter (2016). Stronger vertical entrainment of nutrient-rich waters from deeper isopycnal layers 10 in cold 2017 caused an increase of Chl in winter up to 0.6-0.7 mg/m compared to a warm winter of 2016, when Chl was only 0.4-0.5 mg/m. Further, during almost the whole year from February to October Chl in the upper 0-40 m layer of cold 2017 year was on 0.1-0.2 mg/m higher than in 2016. This rise of Chl in 2017 led to an increase in light attenuation due to the selfshading effect. In contrast, in warm 2016 with a lower amount of nutrients light attenuation decreased and the irradiance reached deeper isopycnals layers with a higher amount of nutrients. As a result, in warm 2016 the subsurface chlorophyll 15 maximum deepens and the values of Chl in 40-60 m layers were significantly higher than in 2017. The maximum positive difference in this layer (0.5 mg/m) was observed during a summer seasonal peak of irradiance due to the largest increase of light attenuation in the summer of 2017. As a result, the column-averaged yearly values of Chl in warm 2016 and cold 2017 were comparable. However, in the year with intense winter mixing upper layers are more productive, while in the year with low winter vertical mixing, subsurface chlorophyll maximum widens and reaches deeper layers. These results show that the 20 observed long-term warming may lead to the continuous deepening of the subsurface chlorophyll maximum in the ocean. Introduction Convective mixing is one of the most important mechanisms supplying nutrients in the euphotic layer in mid-and high latitudes (see e.g. Sorokin, 2002;Williams & Follows, 2003, Severin et al., 2014. With the rise of stratification and irradiance vertically entrained nutrients during winter are further consumed by phytoplankton, which causes the early-spring bloom in the upper 25 layers of the ocean (Sverdrup, 1953;Sorokin, 2002). After the bloom, part of the nutrients in organic form sinks out to the nitroclyne and another part regenerates, which can fuel the phytoplankton bloom in the warm period of the year (Williams & Follows, 2003). https://doi.org/10.5194/bg-2020-366 Preprint. Discussion started: 15 October 2020 c Author(s) 2020. CC BY 4.0 License. Particularly, in the Black Sea according to Lebedeva & Vostokov (1984), Karl & Knauer (1991) only a small fraction (~10%) of particulate flux is exported to deep anoxic part of the sea, while the largest part of nutrients (90%) recirculates in the upper 30 layer (Oguz et al. 1999). The intensity of winter convection in the Black Sea, impact on the phytoplankton concentration in both cold and warm period of the year, as well as on its yearly-averaged characteristics. The most intense winter-early spring bloom of diatoms (Mashtakova, 1985;Sorokin, 2002;Mikaelyan et al., 2018) and the following early-summer bloom of coccolithophores (Mikaelyan et al., 2015;Silkin et al., 2014Silkin et al., , 2019 in the Black Sea are observed after cold winters and are both related to the entrained in winter nutrients. The biomodelling study by Kubryakova et al. (2018) shows that the peak 35 values of the subsurface phytoplankton maximum in summer in the Black Sea also depend on the winter convection. Satellite data demonstrated that the variability of surface concentration of chlorophyll-а (Chl) on interannual time scales is correlated with winter sea surface temperature Finenko et al., 2014). Moreover, long-term analysis of in-situ data (Mikaelyan et al., 2018) showed that winter temperature significantly affects the taxonomic composition and seasonal succession of phytoplankton in the Black Sea throughout the whole warm period of the year. 40 A general feature of Chl vertical distribution in the Black Sea is the deepening of Chl peak during the warm period of a year and a formation of a so-called deep chlorophyll-a maximum (DCM) (Sorokin, 1983;Vedernikov & Demidov, 1993), similarly as in the other areas of the World Ocean at the same latitudes (see e.g. Mignot et al., 2014). According to the classical point of view, phytoplankton occupies intermediate layers between the nutricline, situated in deeper layers, and illuminated layers near the surface (Cullen, 2015). The nutrients entrainment in the euphotic layer in the warm period is strongly related to the short-45 period mixing events (Williams & Follows, 2003) associated e.g. with storms (Iverson et al., 1974;Zhang et al., 2014;Kubryakov, Zatsepin, et al., 2019), wind-driven or dynamic upwelling (McGillicuddy et al., 1998;Mikaelyan et al., 2020). Such physical processes provide vertical fluxes of nutrients in the lower part of the euphotic zone, which is one of the important reasons for the phytoplankton growth in its subsurface maximum in summer (Jonhson et al., 2010;Cullen, 2015). Another reason for the formation of DCM is the summer rise of solar radiation, which leads to the widening of the euphotic 50 layer, and deepening of the productive zone (Letelier et al., 2004;Mignot et al., 2014;Lavigne et al., 2015). At the same time, high values of irradiance may cause photoinhibition and decrease of Chl near the surface (Platt et al., 1982) caused by several effects including non-photochemical quenching, photoinhibition, and photoadaptation (Falkowski & Raven, 2013). The latter is partly associated with the increase of Chl content per cell (MacIntyre et al., 2002), documented for the Black Sea in (Finenko et al., 2002(Finenko et al., , 2005Churilova et al., 2019). 55 Intense phytoplankton bloom causes strong light attenuation, as photosynthetic pigments strongly absorb the light (Morel, 1991). This effect known as self-shading can significantly impact on the phytoplankton growth in deeper layers, where the highest amount of nutrients is observed throughout the year (Shigesada & Okubo, 1981). The amount of Chl and related water clarity largely control the depth of the euphotic zone (Shigesada & Okubo, 1981;Morel, 1991) and, therefore, impact on the position of DCM (Letelier et al., 2004;Leach et al., 2018). The self-shading mechanism is one of the important factors driving 60 the phytoplankton variability and taxonomic composition in the most productive waters, e.g. lakes, rivers, and coastal waters (Morel, 1991;Leach et al., 2018;Churilova et al., 2020). On seasonal time scales, Letelier et al. (2004) have shown that the https://doi.org/10.5194/bg-2020-366 Preprint. Discussion started: 15 October 2020 c Author(s) 2020. CC BY 4.0 License. winter bloom of phytoplankton in the tropical Pacific Ocean leads to the additional shoaling of the euphotic zone, impacting on the development of phytoplankton in the deep layers. At the same time, the impact of self-shading on the interannual variability of the vertical distribution of Chl in the open ocean and, particularly, in the central Black Sea is mostly overlooked. 65 Recently-deployed Bio-Argo buoys give a possibility to obtain continuous measurements of Chl and light characteristics on the interannual time scales with high vertical and time resolution (Claustre et al., 2010;Mignot et al., 2014). In our study, we use the measurements of these buoys in the Black Sea to investigate the effect of winter mixing and irradiance on the vertical distribution of Chl in two years with contrasting winter conditionswarm in 2016 and cold in 2017. The Black Sea bio-optical properties have several unique features, related mainly to the impact of large river inflow. Rivers 70 cause its strong haline stratification and bring a vast amount of organic matter (Vladimirov et al., 1997;Mankovsky et al., 2010), which reduces the transparency of its water. As a result, the diffuse attenuation coefficient in the Black Sea is high Churilova et al., 2019) and the euphotic layer is only about 50 m (Vedernikov & Demidov, 1993). Therefore, the DCM in the Black Sea is relatively shallow occupying the upper 15-50 m. The variability of the thickness, depth, and shape of summer DCM in the Black Sea was investigated in detail by Finenko et al. (2005), Krivenko (2010). Below 75 the productive layer, the concentration of the nutrients began to rise sharply and the upper border of the nitroclyne in the Black Sea is situated approximately 50 m depth (Konovalov et al., 2005). A strong vertical gradient of salinity suppresses the vertical mixing and the mixed layer depth (MLD) in the Black Sea in winter on average does not exceed 40-50 m. The spatial and temporal variability in winter MLD is strongly related to the vertical displacement of the pycno-halocline (Kubryakov, Belokopytov, et al., 2019). Due to the strong haline stratification, 80 the position of chemical layers in the Black Sea is tightly coupled to certain isopycnals and the variations of their concentration in density coordinates are significantly less than in z-coordinates (Konovalov et al., 2005;Tuğrul et al., 2014). The variations of the isopycnals are rather intense and are mainly caused by the dynamic forcing, which is represented by the large-scale cyclonic Rim Current and the abundance of mesoscale eddies (Korotaev et al., 2003). During winter convection, cold waters do not penetrate through the pycno-halocline and form the cold intermediate layer with temperature (T<8°C) situated at 50-85 150 m depth, which further can be observed during the whole year (Staneva & Stanev, 1997;Korotaev et al., 2014). However, continuously observed warming causes the disappearance of the cold intermediate layer, which in the recent period is formed only in a very cold year, such as 2017 (Stanev et al., 2019). Long-term rise of temperature observed globally (Behrenfeld et al., 2016) and in the Black Sea (Ginzburg et al., 2004;Belokopytov, 2018;Stanev et al., 2019) increase the stratification in the upper layers, which may impact significantly on the 90 convection mixing and related vertical nutrient fluxes. Such changes are expected to cause a decrease in overall biological productivity in the ocean (Behrenfeld et al., 2016). In our study, we show that the effect of self-shading may partly compensate for this decrease and will lead to the widening of the productive layer in the ocean. Data and methods The study is based on the bio-optical and hydrological measurements of Bio-Argo buoys in the deep part of the Black Sea in 95 2016 and 2017. These years have contrasting winter conditions with very cold winter documented in 2017 by (Stanev et al., 2019, Capet et al., 2020 and warm winter in 2016. We use the data of buoys #6901866 and #7900591, as they provide the simultaneous continuous measurements of both Chl and photosynthetic active radiation (PAR) in these years. The measurements of both these buoys were made in the deep part of the Black Sea with depths of more than 1000 m. The buoys were located in different areas of the basin (Fig. 1). Buoy #6901866 was moving in the cyclonic direction over the 100 continental slope of the basin with the Rim Current. Buoy #7900591 was situated in the east-central part of the basin. As we will show below, despite these differences in the geographical position both buoys show similar results concerning the paper topic. Data on the Chl and downwelling instantaneous irradiance (Ed, μmol photons m -2 s -1 ) integrated over 400-700 nm were analyzed. Chl (mg/m 3 ) was retrieved from the chlorophyll fluorescence sensor of the Wetlabs ECO Triplet Puck. The product 110 "CHL_ADJUSTED" downloaded from ftp://ftp.ifremer.fr/ifremer/argo (Schmechtig et al., 2015) was used. This product includes the correction on non-photochemical quenching (Roesler et al., 2017) and the correction on the contribution of fluorescence by non-algal organic matter (Xing et al., 2017). As it is shown in (Xing et al., 2017li) the latter correction is very important for the Black Sea waters, which contains a huge amount of dissolved organic matter , and the https://doi.org/10.5194/bg-2020-366 Preprint. Discussion started: 15 October 2020 c Author(s) 2020. CC BY 4.0 License. adjusted product is about twice lower than non-adjusted. The description of the used protocol is given at 115 http://www.argodatamgt.org/Documentation. The Bio-Argo buoys provided data on salinity and temperature at the time of the bio-optical measurements. All the data was visually quality-controlled for consistency with known T, S-distribution in the Black Sea. Eventually, we find no outliers in the data of these buoys. MLD was estimated using a potential density threshold of 0.007 kg/m 3 . This criterion was justified by 125 Kubryakov, Belokopytov, et al. (2019) on the base of the composite analysis of the long-term data for the Black Sea in 1985- 2017. Bio-Argo buoy data have a high time resolution (from 1 to 5 days) and vertical resolution (1 m), is regular and is publicly available at http://doi.org/10.17882/42182 or can be downloaded from the DAC (such as Coriolis). Unfortunately, the optical-based nitrate measurements of Bio-Argo buoys in the Black Sea are poorly consistent with 130 information on nitrates distribution known from numerous in-situ studies. Particularly Bio-Argo buoys show the persistent presence of more than 3 µM nitrates in the upper layer of the Black Sea throughout the year (see fig. S1), which is not consistent with 0.5 µM documented in many previous studies (Konovalov & Murray, 2001;Tuğrul et al., 2015). A possible reason for this is the complex optical characteristics of the Black Sea with a lot of dissolved organic matter, e.t.c. (see e.g. . That is why this data was not used for this study. 135 Physical characteristics in years with cold and warm winter The results of this study are based on a comparative analysis of the physical and bio-optical characteristics of the Black Sea in 2016 and 2017, which were characterized by different winter thermal conditions. In 2016, the winter was significantly warmer than in cold 2017 ( Fig. 2a, b). The temperature in the upper 70-meter layer from January to March 2016 did not fall below 140 8.5°C (Fig. 2a), and its column-averaged value varied from 8.5 to 9.0°C ( (Fig. 2c). In the strongly stratified Black Sea, the location of nutricline is tightly coupled to certain isopycnals as it is shown in many chemical studies (Tuğrul et al., 1992;Konovalov et al., 2005). For example, the concentration of nitrates begins to gradually increase below the isopycnal of 1014 kg/m 3 (see Supplementary Fig. S1), and rises more sharply below the isopycnal of 1014.4 kg/m 3 where the upper part of the nutricline is located (Konovalov & Murray, 2001). The deeper isopycnals the winter convection reaches, the more new nutrients will be 160 entrained in the upper layers. The tight relation between density and the position of chemical elements (see Konovalov et al., 2005) suggests that the density of the upper mixed layer can be used as a proxy for the qualitative estimation of the vertically entrained nutrients in winter (Kubryakova et al., 2018). isopycnals layers depends not only on the vertical mixing, but also on the vertical uplift of isopycnals during the seasonal winter intensification of cyclonic circulation in the Black Sea (Titov, 2004). Usually, the intensity of the large-scale circulation in the basin is observed in the cold years (Blatov et al., 1984;, as both of them are driven by the same atmospheric patternsintensification of northeast winds from Eurasia . The rise of cyclonic 175 circulation and uplift of the pycno-halocline on the opposite decreases MLD (Kubryakov, Belokopytov, et al., 2019). That is why, in the Black Sea the MLD is not correlated with sea surface temperature (Titov, 2004), but strongly depends on dynamic forcing, such as eddies, large-scale circulation. Thereby, in the Black Sea density of the mixed layer rather than its depth reflect the intensity of the entrainment of new nutrients in the upper layers. In any way, in the considered case, both the depth and density of the upper mixed layer were larger in the cold winter of 2017. 180 To conclude, the above analysis is used to argue that the vertical entrainment of nutrients from deep isopycnals layers was more intense in the cold winter of 2017 than in the warm winter of 2016. The increase of nutrient concentration in the Black Sea in the cold years by in-situ data was previously documented by Tuğrul et al. (2015). Bio-optical characteristics in years with cold and warm winter The winter entrainment of nutrients in the upper layer caused the intense growth of phytoplankton, which is reflected in the 185 increase in the Chl in February-March. The average monthly data of two buoys show that in February of cold 2017 Chl in the upper 30-meter layer reach 0.7 mg/m 3 (Fig. 4b) In the same period of 2016 Chl was only about 0.4-0.5 mg/m 3 (Fig. 4a). Column-averaged Chl in 0-70 m layer in 2017 was 0.4 mg/m 3 (Fig. 3b, black dashed line), which was on 0.1 mg/m 3 higher than in 2016 (Fig. 3b, black solid line). productivity not only during winter-spring bloom of phytoplankton, but also in the following months in agreement with the results of previous investigators (Burenkov et al., 2011;Finenko et al., 2014;Mikaelyan et al., 2015Mikaelyan et al., , 2018Kubryakova et al., 2018). In particular, this effect is a probable reason for the unusually strong coccolithophore bloom in May-July 2017, which was documented on the base of satellite and Bio-Argo measurements by Kubryakov, Mikaelyan, et al. (2019). increasing the nutrient fluxes in the euphotic layer (Mikaelyan et al., 2017(Mikaelyan et al., , 2020Kubryakov, Zatsepin et al., 2019). (Fig. 4d). This maximum was investigated in detail in (Kubryakov, Mikaelyan, et al., 2019), where it was related to the release of dissolved organic matter (DOM) after enormously strong early-summer coccolithophore blooms formed due 235 to intense winter convection in 2017. In comparison, in 2016, the values of Kd in 0-40 m in February-October was only about 0.1 m -1 (Fig. 4c). The diagram of Kd difference shown in (Fig. 5b) is very similar to that for Chl (Fig. 5a) The rise of Kd led to a decrease in PAR in the subsurface layers and shallowing of the euphotic zone. The euphotic zone is marked in Fig. 4a-d as the isolume Ed=3 μmol of photons m-2 s-1 (or 0.08 mmol photons m -2 day -1 ). This isolume was chosen because in the Black Sea the position of the upper border of relatively high Chl values follow this isolume, which indicate that this value of PAR can play a role of the compensational irradiance in the basin (Kubryakov et al., 2021 (manuscript in press)). This isolume was located at 40-45 m in cold 2017, which was on more than 10 m shallower than in warm 2016, when it was 245 located at 55-65 m (see their comparison in Fig. 3bmagenta lines). This indicates that in cold 2017, the growth of phytoplankton in the deep layers was limited by the low light conditions. The decrease of Chl in the upper layers in warm 2016 led to a stronger penetration of PAR in the subsurface layers of the sea. The monthly diagram of the vertical distribution of PAR excess in 2016 compared to 2017 is shown in Fig. 5c. A significant increase of PAR in warm 2016 by more than 1 µmol photons m -2 s -1 was observed in the entire 0-60 m layer throughout the 250 whole year. In the 10-30 m layer, the values of PAR difference during the early spring bloom in February-April and autumn bloom in October-November reached 20 µmol photons m -2 s -1 (Fig. 5c). In summer, these differences were greatest and exceeded 200 µmol photons m -2 s -1 in the 0-25 m layer and were more than 10 µmol photons m -2 s -1 in the 25-40-m layer. Taking the value of compensational irradiance in the Black Sea as 3 µmol photons m -2 s -1 , such a PAR increase caused a significant widening of the euphotic zone. 255 As a result of PAR penetration, phytoplankton growth was increased by light in the deepest layers (40-65 m) below the maximum winter MLD (45 m). The penetration of light into this layer in warm 2016 due to the decrease of self-shading caused the growth of phytoplankton directly in the nitroclyne without mandatory mixing. A similar process is observed, in particular, in subtropical regions, where winter cooling is almost absent. The above-discussed features of the differences in vertical Chl distribution in cold and warm year are well seen in its yearly-260 averaged profiles (Fig. 7). (Fig. 4a, Fig. 6a, c). At this time the column-averaged Chl in 2016 exceeded its values in 2017 by the same value of 0.1 mg/m 3 (Fig. 3b). Thus, the decrease in Chl in the first half of the year in the upper layer of 2016 was compensated by its increase in the summer period due to the lack of self-shading effect and the development of a deep Chl maximum directly in the nitroclyne. Because of this, the yearly-275 averaged integral Chl in the euphotic layer in 2016 and 2017 became comparable (Fig. 3b, Fig.7). Discussion The reasons for the variability of the characteristics of DCM is one of the important and actively investigated oceanographic tasks (Cullen, 2015;Leach et al., 2018;Barbieux et al., 2019). Variability of its position and strength are related in different studies to the vertical distribution of nutrients (Hartman et al., 2014;Barbieux et al., 2019), optical characteristics of water and light availability (Morel, 1991;Mignot et al., 2014;Leach et al., 2018), density stratification (Navarro & Ruiz, 2013). Our 285 study approves that all these factors are important and describe a relation between their impact on the vertical distribution of Chl on the example of the Black Sea (see scheme in Fig. 8). Nutricline in the highly-stratified waters of the Black Sea is closely related to the isopycnals position (Konovalov et al., 2005). The increase of density of the upper mixed layer in winter leads to the convective mixing reaching the isopycnals with the same density. A deep isopycnal layer with a high amount of nutrients mixes with surface waters and defines the concentration of nutrients in the winter mixed layer (0-40 m in the Black Sea) 290 (Kubryakova et al., 2018). Further thermal stratification stabilizes the water column. Entrained in winter period nutrients and the rise of the irradiance causes the following spring growth of phytoplankton. Part of them regenerates and another part sinks out to the nitroclyne. The regenerated nutrients further are used by the summer population of phytoplankton. For example, in the Black Sea, the strong coccolithophore bloom emerges after strong spring bloom of diatoms and their magnitude depends on the winter sea surface temperature (Burenkov, 2011;Silkin et al., 2014;Mikaelyan et al., 2015Mikaelyan et al., , 2018. Thus, the 295 phytoplankton concentration in the upper layer through the whole year largely depends on the density of the preceding winter mixed layer, which is approved by the difference in Chl distribution in warm and cold year in Fig. 4a, b, 5a, 6e, f. with cold winter, the larger amount of nutrients (grey color) is convectively entrained in the upper layer, which increases the growth of phytoplankton in the upper layer and causes self-shading of the deeper layer. Therefore, DCM moves to the upper layer. (b) In years with warm winter convective nutrient fluxes are low, the amount of phytoplankton and light attenuation decreases. In the summer period with the increase of PAR, light penetrates the upper layer of nitroclyne and causes intense and deep summer subsurface bloom. Therefore, the total amount of nutrients used by the phytoplankton in both years is 305 comparable. In the warm period of a year PAR strongly increases, the euphotic layer deepens, and the surface layer becomes overilluminated, which is one of the possible reasons for the formation of DCM in summer (Platt et al., 1982;Mignot et al., 2014). The phytoplankton pigments, absorb light, and its variability largely defines the optical characteristics of the water through the 310 self-shading mechanism (Morel, 1991;Churilova et al., 2020). Thereby in the water with higher nutrient concentration (e.g. after cold winter), the light penetration to the deeper layer weakens (Fig. 5b, c). The water clarity decreases and DCM is situated closer to the illuminated layers (Fig. 8a) near the surface (Mignot et al., 2014;Leach et al., 2018), as we observed in 2017 in the Black Sea (Fig. 7). On the opposite, in the water with relatively low winter nutrient fluxes (as in 2016 in the Black Sea), Chl and biomass of phytoplankton decreases. The absence of self-shading promotes the light penetration to the deeper 315 layer, which leads to the deepening and widening of the DCM (Fig. 7, Fig. 8b). These arguments suggest that DCM displacement is driven by the amount of nutrients and related self-shading by formed phytoplankton. Light attenuation depends not only on the phytoplankton or Chl, but also on the attenuation by clear water and DOM (Morel, 1991), which concentration in the Black Sea is very high due to the intense river discharge (Mankovsky et al., 2010;Organelli et al., 2017). Another strong source of light attenuation, which is nutrient-dependent, is autochthonous DOM formed due to 320 the release of lipids during lysis of phytoplankton cells. Kubryakov, Mikaeyan et al. (2019) based on the Bio-Argo data on diffuse attenuation at 390 nm and backscattering measurement show that one of the most significant sources of DOM is related to the lysis of coccolithophore cells during the termination of their early-summer bloom possibly mediated by a viral attack. As it is shown by Burenkov et al. (2011), Mikaelyan et al. (2011, 2015 the intensity of coccolithophore blooms in the Black Sea is significantly related to winter temperature and the amount of entrained nutrients. Particularly, the extremely strong 325 coccolithophore bloom was observed in the Black Sea after cold 2017, which results in the observed maximum light attenuation in July-August in 2017 at 10-30 m depth (Kubryakov, Mikaeyan et al., 2019). Such release of DOM caused by strong summer coccolithophore blooms or blooms, as well as other types of the phytoplankton in cold years additionally reduces the water transparency. This effect will additionally decrease the light penetration in cold years with a larger amount of nutrients. Therefore, the release of light-absorbing autochthonous DOM plays an additional role in equalizing the total productivity in 330 warm and cold years. It is also important that in the year with weak winter nutrient fluxes during summer seasonal maximum of irradiance light penetrates to the most deeper layers, where nutrient concentration is high. Particularly, in the considered case in warm 2016 PAR reached 40-60 m depth, where the upper border of nitrocline is located in the Black Sea. At these depths, the concentration of nitrates sharply increases ( Supplementary Fig. S1). At the same time, the winter mixed layer in the Black Sea is only about 335 40-45 m (Titov, 2004;Kubryakov, Belokopytov, et al., 2019) (Fig. 3a). Underlying rich in the nutrients layers were unaffected https://doi.org/10.5194/bg-2020-366 Preprint. Discussion started: 15 October 2020 c Author(s) 2020. CC BY 4.0 License. by winter dilution with poor in nutrients surface waters. The presence of light and nutrients cause the growth of phytoplankton. The maximum of Chl in this case forms closer to the layers with a high concentration of nutrients. Such a position should increase the efficiency of the upward nutrient fluxes, which are proportional to the vertical gradients of nutrient concentrations. This process is most important in the warm period of a year, when phytoplankton is fueled mainly from below by the intense 340 diapycnal mixing caused by storms (Iverson et al., 1974;Zhang et al., 2014;Chacko, 2017) or, e.g. eddy-driven upwelling (McGillicuddy et al., 1998;Mikaelyan et al., 2020). Particularly, in the Black Sea, several storms in August 2015 led to the intense subsurface maximum of Chl at the base of the euphotic layer reaching (Kubryakov, Zatsepin, et al., 2019). The closeness of DCM and nutricline increases the nutrient fluxes in the summer of the years with warm winter, which compensates for their decrease in winter caused by weak convective entrainment. 345 The proposed mechanism can help to explain the observed changes in seasonal succession of phytoplankton after cold and usual winters, studied in detail by Mikaelyan et al. (2018). Particularly Mikaelyan et al. (2018) show that after cold winters in the Black Sea the biomass of diatoms is high in spring and low in summer. The opposite is observed after warm winters. Large nutrient fluxes in cold winter cause intense spring diatom bloom (Mashtakova, 1985;Sorokin, 2002;Silkin et al., 2014) which is followed by strong coccolithophore bloom (Mikaelyan et al., 2015). The lysis of coccolithophores causes the intense 350 rise of DOM in the upper layer (Kubryakov, Mikaelyan, et al., 2019). Light attenuation decreases the thickness of the euphotic zone and the efficiency of the upward nutrient fluxes in it from the nutricline. The release of DOM also can trigger the microbial loop, which transfers the trophic energy in smaller species. Both these factors can explain the high biomass of diatoms in spring and low in summer after cold winters. In contrast, after warm winter phytoplankton in spring grows in low-nutrient conditions. Thereby in spring diatom biomass is lower and dinoflagellates biomass reaches maximum, according to (Mikaelyan 355 et al., 2018). However, in summer phytoplankton develops in the deeper layer with a higher amount of nutrients, which may explain the higher diatom biomass and diversity in warmer years documented by Mikaelyan et al. (2018). It should be noted, the phytoplankton after cold and warm winter grows at different depths and, therefore, the environmental (light and nutrients) conditions change significantly. Deepening of the euphotic layer may also promote the growth of species adapted to low light with low biomass and high Chl content in cells (Falkowski & La Roshe, 1991;MacIntyre et al., 2002;360 Latasa et al., 2017). Particularly, in the Black Sea, the lower border of the euphotic zone is characterized by the domination of small flagellates and unicellular cyanobacteria (Churilova et al., 2019;Mikaelyan et al., 2020), which may have an advantage in the years with warm winters. Globally-observed warming and the rise of thermal stratification are expected to decrease the vertical nutrient fluxes associated with winter cooling and primary production in the ocean (Behrenfeld et al., 2016). Particularly, in the Black Sea the long-term 365 rise of sea surface temperature on about 1.5℃ per 10 years was documented by Ginzburg et al. (2004), and already lead to the disappearance of the cold intermediate layer in recent years (Belokopytov & Shokurova, 2005;Knysh et al., 2011;Piotuch et al., 2011;Capet et al., 2020). The result of the presented manuscript suggests that the continuous weakening of thermal convection and associated weakening of the nutrient fluxes will also cause a decrease of self-shading by Chl and an increase of the water clarity. Such changes in water optical properties may cause the long-term deepening of DCM 370 in the ocean to the rich in nutrients deeper layers. We may suggest that, at first, such displacement may enhance the production in a warm period of the year during the seasonal peak of irradiance and compensate for the decrease of productivity in winterspring. However, with the deepening of DCM, remineralization of the organic matter and settling of the particles will occur also in deeper layers. These changes will cause the consequent deepening of the nutricline. In such a way the ecosystem of the temperate Black Sea will transit to a tropical state, with very deep nitroclyne and DCM. This trend is especially dangerous in 375 the oxygen minimum zones, such as observed in the Black Sea, where nitrates are effectively removed at the suboxic interface due to the process of denitrification, the oxidation of nitrates by particulate manganese, and others (Konovalov et al., 2008). In such areas, the intense removal of nitrates from below may cause the thinning of the nutricline, the additional decrease of biological productivity, and oxygen production. Such a process also can promote the changes in the taxonomic composition of phytoplankton to low-light adapted species. 380 Low winter nutrient fluxes in warm winters decrease the productivity in upper layers, which promote the light penetration in 395 summer to the upper layers of nitroclyne and intensify summer nutrient fluxes from below. As a result, the obtained estimates of the integral layer of Chl were comparable in years with warm and cold winters. At the same time, the intensity of winter convection led to significant changes in the vertical distribution of Chl. In the cold years, DCM was relatively sharp and located in the upper layers, while in the warm years DCM widens significantly, but Chl in it was less and distributed more uniformly. Conclusions The observed inverse relation between winter and summer nutrient fluxes can be one of the important reasons for the observed 400 interannual changes of seasonal succession in the ocean. Data availability Bio-Argo data were made freely available from IFREMER data archive [ftp://ftp.ifremer.fr/]. Author contribution Elena Kubryakova and Arseny Kubryakov processed the data of Bio-Argo buoys, performed the analysis. Elena Kubryakova 405 drafted the manuscript and designed the figures. Arseny Kubryakov was involved in planning and supervised the work. Competing interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.	2020-10-28T19:21:04.928Z	2020-10-15T00:00:00.000	{ "year": 2020, "sha1": "289e1ea8169ee8093c10fdc8d4b17693e8ec1416", "oa_license": "CCBY", "oa_url": "https://bg.copernicus.org/preprints/bg-2020-366/bg-2020-366.pdf", "oa_status": "GREEN", "pdf_src": "Anansi", "pdf_hash": "3f52c38b2a428eb330a2e2f6cf38e902ad8fed18", "s2fieldsofstudy": [ "Environmental Science" ], "extfieldsofstudy": [ "Environmental Science" ] }
209575232	pes2o/s2orc	v3-fos-license	Nutritional Characteristics and Potential Applications of Flour Prepared from Indonesian Wild White Yam (Dioscorea esculenta L.) Gembili or wild white yam (Dioscorea esculenta L.) is one of underutilized tubers that can be found during dry season in Indonesia and other Southeast Asian countries. Although it has been consumed as staple food by people for centuries during famine, no further studies have been conducted to explore its potentials in food applications. This work aims to study the preparation, characterization and potential analysis of Gembili flour for use as raw material in novel foods development. Creamy white flour was obtained from white yam tuber from milling and sieving of dried tuber chips. The flour has water content about 12.08%, which is acceptable for storage. No lipid was detected. In addition, the protein, ash, and fiber content of the flour were respectively 3.00%, 1.27%, and 9.04%. The carbohydrate and amylose contents of the flour were 86.69% and 29.92% indicating its suitability for energy source. As expected, the cyanide content is very low of about 1.688 ppm suggesting that it is safe for consumption. Refer to those properties, Gembili flour can be a promising raw material for the preparation of bakery, cookies, noodle and confectioneries.Keywords: proximate composition, flour, white yam, food, preparation INTRODUCTION Daily life exhibits the fact that human and animal cannot be separated from food issues. As can be seen recently, the rapid development of health science has increasingly influenced people's lifestyle and food choices. Accordingly, this condition encourages the food industries to extensively find raw material alternatives and food manufacturing process, specifically for breads and cookies. The steady increase of wheat flour price in the world due to climate change and higher demand also triggers the search of wheat flour alternatives (Retnowati et al., 2018). A lot of efforts have been performed to replace or substitute wheat flour as the main ingredient of breads and cookies, also noodles as the example of favorite dishes in Indonesia. Most areas in the main islands of Indonesian archipelago are rich in tubers and root plants. Literature study suggested that those plants show high potential for the production of flour for use in food applications. Cassava, taro and sweet potato are reported to be the most utilized tubers by Indonesian people, while other tubers and roots are not so popular because mostly are being seasonal bulbs (Kumoro et al., 2012). Since they cannot be harvested throughout the year, they are not always available in traditional market every time. One of these tubers is Gembili which is the member of Dioscorea family. It is a seasonal food plant and generally planted in Indonesia in the end of September to October. This tuber can only be harvested after being grown for 7 -9 months, usually during dry season. The yellowing leaves and withered vines indicate of mature crop, which is usually ready to harvest around May to July (Senanayake et al., 2012). Physically, Gembili tuber has a thin skin with some plumes attached on the surface, while the flesh exhibits a creamy color and is traditionally consumed by people upon steaming, boiling, baking, frying or roasting (Retnowati et al., 2018). According to Okunlola and Odeku (2011), Dioscorea tubers have a large quantity of starch, which is about 54% to 74%. Therefore, it could be a potential source of basic energy. Gembili flour is safe for consumption by people who are sensitive to gluten as celiac disease and gluten ataxia (Niland et al., 2018). Recent nutrition science development proves that gluten free foods are increasingly popular as a part of healthy diets. Unfortunately, high moisture content has caused yam tubers to become highly perishable and bulky (Coursey and Ferber, 1979). Hence, the conversion of Gembili tuber into flour as the most acceptable forms will not only increase its economic value, but also makes it easy to store for further uses as raw material in the manufacture of cakes, bread, cookies or noodles Processing of yam tuber into flour offers a higher possibility to extend the availability of yam supply during the off-season, thereby reducing the loss during storage and lowering the cost for marketing and transportation (Retnowati et al., 2018). Despite its allergen issue, gluten in wheat flour has an important role in the bread making (Toufeili et al., 1999). Therefore, a number of researches have been conducted to improve the gluten free flour through physical, chemical and biological modification or addition of hydrocolloids in dough flour to obtain closely similar baking properties of wheat flour (Naqash et al., 2017). Unfortunately, as the member of Dioscorea family yam tubers may contain different antinutrients, which prone to have deleterious effects to both human and animals when they are consumed (Polycarp et al., 2012). Therefore, the safe level antinutrients content in the tuber should be assured before further uses. The most common antinutrients contained in the yam tuber are phytic acid, tannin, and cyanogens. Yang and Lin (2008) reported that the age of harvest, variety, geographical condition or the storage condition after harvest may considerably affect its antinutritional content. The objectives of this work were to determine the proximate nutrient composition of Gembili flour and to analyze the potential uses of Gembili flour for future development as a raw material for food or other applications. Sample of Gembili flour preparation Gembili tubers were purchased from the local market in Gunungpati sub-district, Semarang City, Central Java province, Indonesia. Chemicals and standard Analytical grade of chemicals used for browning prevention (potassium metabisulfite) and Gembili flour proximate analysis were the products of Merck-Indonesia and were purchased from an authorized chemical store in Semarang-Indonesia. Analytical grade (petroleum ether) and standard (amylose) with a purity >98 % were the product of Merck (Germany) and were purchased and directly used without further treatment. Preparation of Gembili flour The Gembili tubers were carefully washed with clean tap water to ged rid of the adhering dirt and other undesirable materials from the tubers, and to inhibit the microbial growth on the final product. They were peeled and trimmed to remove defective parts, washed, and shreded with an automatic machine to obtain thin slices (5 mm). Then, the slices were soaked in potassium metabisulfite solution (0.075%) for 1 hour to prevent browning as suggested by Jayakody et al. (2007). The tuber slices were rinsed again with flowing water and were further spread in a single layer on drying trays and dried under sunshine for 3 days and were subsequently crushed in a laboratory scale crusher, milled into flour with a hammer mill to obtain flour. The flour was then passed through -180 μm +250 μm sieves. Only the Gembili flour retained on the 250 μm was used in this Composition analysis The proximate analysis (moisture (M), ash (A), lipid (L), and protein (P) content (N  6.25) of Gembili flour was conducted strictly according to AOAC methods (Latimer Jr, 2016). Then, the carbohydrate content was calculated by differences. In addition, the amylose content was determined following the methode developed by Juliano (Juliano, 1971). The total cyanide content in mg HCN equivalents/kg fresh weight (ppm) was determined based on the method previously used by Bradbury et al. (1999). All analyses were performed in triplicates. The reported values were the average of them. RESULTS AND DISCUSSIONS Nutritional compositions The proximate composition of Gembili flour is presented in Table 1. The moisture content of food samples is an indicator of shelf life and quality of solid foods. Although the moisture content of the lesser yam flour obtained in this study (12.08%) is slightly higher than those reported in the literature, it is still below the recommended safe level (12-13%) for storage of flour set by Codex Alimentarius Commission through joint food standards program of Food and Agriculture Organization/World Health Organization (FAO/WHO) (Codex Alimentarius Commission, 1995). Foods with moisture content higher than 14% are vulnerable to bacterial attacks and mould growth, which produce undesirable changes (Ihekoronye and Ngoddy, 1985). However, there was no lipid content detected in the Gembili flour. A similar observation was reported by Saskiawan and Nafi'ah (2014) who found an extremely low-fat content (0.15 %) on Gembili flour planted in Cibinong-West Java, Indonesia. While two researchers (Senanayake et al., 2012;Ukpabi, 2010) reported that fat contents of D. esculenta flour were comparable to wheat flour. Interestingly, the absence of lipid in Gembili flour suggests that the flour is a healthier choice of food, as perceived by consumers, especially in relation to cardio-vascular diseases (Tortoe et al., 2017). The protein content of the Gembili four is closely similar to that observed by Saskiawan and Nafi'ah (2014), but far lower than those found by Ukpabi (2010) and Senanayake et al. (2012) for D. esculenta farmed in Nigeria and Sri Lanka. The Gembili flour has a lower protein content compared to that of American wheat flour (Ukpabi, 2010). For some food product applications, the protein content of wheat flour is too high and can be diluted with other starches of lower protein content. This is the case with biscuit making. The protein content of the mixture required is around 7.0-8.5 % for sweet biscuits or 8.4-10 % for biscuit sponge (Snow and O'Dea, 1981). Therefore, the Gembili flour has the potential to partially substitute wheat flour to obtain composite flours with acceptable protein contents for certain food applications. Ash content, which reflects the mineral content in food (Tortoe et al., 2017) even though contamination can indicate a high mineral concentration in a food sample (Baah et al, 2009), Although Gembili contains lower ash content than that of D. esculenta grown in Nigeria and Sri Lanka (Senanayake at al., 2012;Ukpabi, 2010), it still could serve as a fairly good source of dietary minerals. Fiber is a type of carbohydrate that the body cannot digest. Though most carbohydrates are broken down into sugar molecules, fiber cannot be broken down into sugar molecules, and instead it passes through the body undigested. Fiber helps the peristaltic movement of food substances during digestion and regulates the body's use of sugars, helping to keep hunger and blood sugar in check. The Gembili four contains 9.04% of fiber, which is much higher than those reported by other researchers (Senanayake at al., 2012;Ukpabi, 2010). According to the provisions of Codex Alimentarius (2009), food containing more than 6g/100g dietary fiber can be considered as having a high content of dietary fiber. Thus, Gembili flour is a good source of dietary fibers. The carbohydrate content in Gembili flour is only slightly lower than 90%, but higher than those reported in the literature. This affirms the fact that Gembili flour is a valuable source of dietary carbohydrates, which provides caloric energy (Coursey and Ferber, 1979). This finding is in good agreement with Osunde (2008) who explained that the chemical composition of yam is characterized by high moisture and dry matter mainly of carbohydrates. Ukpabi (2010); and c Senanayake et al. (2012) Gembili flour contains high amylose content (29.92 ± 5.09 %). However, this value is slightly lower than the amylose content of D. alata var Krimbang flour obtained from Tulung Agung, East Java-Indonesia (33.1 ± 1.20 %) reported by Aprianita et al. (2014a). The proximate composition of tuber flour largely depends on the specific tissues, geographic location, variety, age, environmental conditions and the preparation technique (Garcia and Dale, 1999). Shimelis et al. (2006) explained that high amylose content tends to restrict flour swelling and the hot paste viscosity stabilizes. In addition, high amylose contents are desired in starches and flours that are to be used for the manufacture of extrudates (Chevanan et al., 2010). Having high amount of amylose, Gembili flour and starch could partially replace wheat flour in snack food formulations to obtain products with a crunchy texture. Amylose within flour or starch could strengthen the dough, which in turn improves the forming and cutting properties of dough to produce snack foods with a crunchy texture (Huang et al., 2006). However, the flour also contains small amount of cyanogens in the form of total hydrocyanic acid. This low antinutrient content suggests that no extensive pretreatment is required to make the flour becomes edible. The safe limit for cyanogens in food material was set by the WHO at 10 ppm for cassava flour (Codex Alimentarius Commission, 1991), and the acceptable limit in Indonesia is 40 ppm (Djazuli and Bradbury, 1999). Color appearance Visually, the Gembili flour is light brown in color. However, it is still darker than the commercial wheat flour. Lightness of flours was affected by browning reactions, which occur during its processing (Van Hal, 2000), and this may have extensively affected Gembili flour and reduced its lightness. As a member of Dioscorea family, Gembili tuber contains mucilage, which makes the flour immediately colored right after the tubers are cut. Krishnan et al. (2010) proposed that oxidation of phenols by polyphenol oxidases and peroxidases contained in the mucilage of the tuber generate brown color and generally known as enzymatic browning. Apart from the inherent color pigments present in yams, yellowness in yam flours has also been linked to total phenol content and the activity of polyphenoloxidase (Akissoe et al., 2003). If the color of a food product is one of the important criteria, then the use of native Gembili flour for its manufacture should be less considered. This is because usually white flours are more preferred in various applications (as in white bread making). Theoretically, the substitution of wheat flours with Gembili flour reduces the whiteness but increases the redness and yellowness of the flour composites (Aprianita et al., 2014b). Modification of flour either by physical, chemical, biological or their combination may be one way to improve the appearance of Gembili flour (Kumoro and Hidayat, 2018). CONCLUSIONS A comprehensive characterization of physicochemical, thermal and functional properties of Ubi Gembili (Dioscorea esculenta L.) flour has been successfully carried out. The moisture content of Gembili flour is below the recommended safe level (12.08%) for storage of flour set by Codex Alimentarius Commission. Gembili flour exhibits high carbohydrate and amylose content indicating its good potential as a source of energy. The flour is also rich in protein and fiber, which is an indication of good food. It contains low ash and a safe level of cyanide. However, it has no lipid content. The Gembili flour granules are creamy white in color. If this flour will be used alone or as substitute to wheat flour in bread making, addition of polar lipid can be arranged as needs to improve the loaf volume of bread.	2019-09-27T11:03:11.226Z	2019-08-11T00:00:00.000	{ "year": 2019, "sha1": "5482a168cb854afe7a0a2abcb7d874850b379c22", "oa_license": "CCBYNC", "oa_url": "https://ejournal.undip.ac.id/index.php/reaktor/article/download/23182/16009", "oa_status": "GOLD", "pdf_src": "Adhoc", "pdf_hash": "038b39556921ddcbea50ede1f259b92fbf1f0f73", "s2fieldsofstudy": [ "Agricultural And Food Sciences" ], "extfieldsofstudy": [ "Biology" ] }
238585992	pes2o/s2orc	v3-fos-license	Improved Secure and Lightweight Authentication Scheme for Next-Generation IoT Infrastructure the Introduction In recent years, the Internet of ings (IoT) [1][2][3][4] has become popular in our everyday life. IoT refers to the real-time collection of any information that needs to be monitored, connected, and interacted with through the use of various devices and technologies such as sensors, radio frequency identification technology, global positioning system, and laser scanners. In the IoT environment, every object (virtual or physical) can be perceived, identified, accessed, and interconnected in a dynamic, ubiquitous network through the Internet. IoT brings great convenience to our lives. Vehicular ad hoc networks [5,6] are considered to be one of the most promising applications of IoT. ey allow people, vehicles, and roadside units to cooperate closely. IoT is also applied to medical healthcare, which is also closely related to our lives. rough the use of IoT, medical healthcare environments have taken on a new look. In an IoT-enabled healthcare system [7][8][9], wearable sensors can be used to collect information about patients and the surrounding environment. Another example of an IoT application is the smart home [10,11]. Smart homes improve people's lifestyles and make them more comfortable, safer, and more efficient. In addition, the cloud system based on IoT can help the national government manage some resources to a great extent. e management data through the cloud system greatly reduces human resources and greatly improves the utilization rate of resources. ese advantages are mainly based on the principle of the cloud-based Internet of ings. e application of such technology supports legitimate users to access normal data from hospitals, homes, borders, and other areas, which can better manage data to a certain extent. Because IoT has grown so seamlessly, many end users are ignorant of the existence of these devices. Due to their invisibility, IoT device security is crucial, yet challenging to manage. Several IoT networks have recently been taken over to carry out malicious attacks. For these reasons, addressing these IoT security challenges is critical to their successful development. However, there has been a significant expansion in the number of IoT devices. Designing security mechanisms for all of these devices is complicated due to the heterogeneity and complexity of IoT networks. For an IoT network to be secure, all the entities (servers, end users, and devices) must mutually authenticate their identities. In addition, all communication should be encrypted to maintain data confidentiality. is means that a common session key for both sides of the communication is required. erefore, designing a secure and efficient authenticated key agreement (AKA) scheme is crucial [12][13][14][15]. Various AKA schemes for IoT have been proposed. In 2004, Kumari et al. [16] found that Chang et al.'s scheme [17] is vulnerable to offline password-guessing attacks, internal attacks, and server masquerading attacks. ey also pointed out that the protocol [17] has security vulnerabilities during the password update phase. To overcome these security weaknesses, Kumari et al. designed an improved scheme. Kumari et al. claimed that their scheme is more secure, more efficient, and more suitable for real-life IoT network use. However, Kaul and Awasthi [18] discovered that Kumari et al.'s protocol [16] is still vulnerable to some attacks. In their scheme, attackers can easily capture some security parameters transmitted on a public channel and then calculate the session key. In response to this, Kaul and Awasthi [18] proposed a robust and secure user authentication protocol based on resource-friendly symmetric cryptographic primitives. Unfortunately, Rana et al. [19] proved that the protocol [18] cannot resist various types of attacks. erefore, they proposed a secure, lightweight AKA scheme for next-generation IoT infrastructure. In this study, however, we found that Rana et al.'s scheme [19] is still vulnerable to offline password-guessing attacks and privileged-insider attacks. In their scheme [19], an illegal insider or malicious attacker can calculate the session key or guess passwords if they can capture a user's smart card. erefore, we propose a new AKA scheme. In the proposed scheme, we utilize the biological information of the users because it is difficult for attackers to obtain this information. To demonstrate that the proposed scheme is indeed secure, we analyze it using Burrows-Abadi-Needham (BAN) logic [20] and also show that it is secure against various types of attacks. Compared with the previous scheme, the proposed scheme has better performance in terms of memory overhead. e remainder of this paper is organized as follows. In Section 2, we briefly review the scheme proposed by Rana et al. [19]. Section 3 demonstrates that Rana et al.'s scheme [19] is vulnerable to offline password-guessing attacks and privileged-insider attacks. Our proposed scheme is described in Section 4. Sections 5 and 6 provide security and performance analyses and comparisons. Finally, Section 7 concludes the paper. Review of Rana et al.'s Scheme In this section, we briefly review Rana et al.'s AKA scheme. eir scheme contains three phases: user registration, login, and authentication, and the steps of their scheme are described below. Notations used in this paper are listed in Table 1. User Registration Phase. (1) First, the user U c selects their own identification ID c , password PW c , and an arbitrary number b. en, the following is calculated: and ID c , RPW c is transmitted to the server through a secure channel. (2) After the server receives the information from the user, it selects an arbitrary number y c and calculates (3) en, the server stores the parameters β c , c c , χ c , DI D c , h(·) in the smart card memory and sends them to the user U c through a secure channel. (4) Finally, the user calculates and stores η c in the smart card. Now, the smart card contains parameters β c , c c , χ c , η c , DI D c , h(·) . Login Phase. When a registered user wants to log in to the system, they perform the following operations: (1) User U c enters their ID c ′ and PW c ′ and inserts the smart card (2) e smart card reader extracts parameters m � η c ⊕h(ID c ′ ⊕PW c ′ ) and RPW c ′ � h(m ‖ PW c ′ ) (3) Further, the smart card reader can extract parameters α c ′ � β c ⊕h(ID c ′ ⊕RPW c ′ ) and y c ′ � c c ⊕h (α c ′ ⊕RPW c ′ ) and calculate If χ c ′ � χ c , it means that the legitimate user is allowed to log in; otherwise, the login is refused (4) After verifying the legitimacy of the user, the reader calculates e reader then sends the login request DI D c , ω c , ] c , T 1 to the server through a secure channel. Authentication Phase. In this phase, the smart card reader and server authenticate each other by performing the following steps: (1) S first verifies the validity of the timestamp by calculating T 2 -T 1 . If the calculated value is less than the given threshold δT, the login request proceeds; otherwise, it is rejected. (2) After that, S extracts and calculates ID c ′ using (ID c ′ ‖ y c ) � Dec ds (DI D c ) and then calculates the values: en, S verifies the validity of the login by comparing the calculated ] c ′ with the stored ] c . If the two are equal, the verification passes; otherwise, the verification fails and the server refuses to accept the login request. (3) After verifying the correctness of ] c , the server continues to calculatez en, S sends the calculated μ c and timestamp T 2 to U. (4) When U receives the information from the server, it first verifies the validity of T 2 and then calculates U checks whether μ c ′ is equal to μ c . If so, S is successfully verified. (5) Finally, after mutual verification, the session key SK can be calculated: Cryptanalysis of Rana et al.'s Scheme In this section, we first describe the threat model. en, we show that Rana et al.'s scheme is insecure against offline password-guessing attacks and privileged-insider attacks. reat Model. is threat pattern shows the capabilities of an adversary, which are also considered and discussed in [21,22]. A ′ s capabilities are as follows: (1) A can perform complete access control on the transmission channel. It can block, change, remove, replay, and hinder the messages passed between participants through a public channel. (2) A can get the information stored in the smart card using power analysis [23,24]. (3) A can obtain the information in the smart card and the information transmitted by the user on the secure channel during the registration process [25]. (4) A can simultaneously obtain the information in the smart card and perform offline password guessing as stated in [26,27]. (5) A can know any two of the user's password, smart card, and biological information. (6) A can obtain the session key that the user communicated with the server before. (7) A can register as a legitimate user in a legitimate way. (4) Finally, A obtains the session key SK according to the value of α c and y c calculated above: Privileged-Insider Attack. (1) First, the attacker A steals the smart card and gets the information (2) en, privileged insiders can obtain the information ID C and RPW c of legitimate users during registration (3) A can calculate the following parameters by using the information β c obtained in the smart card and the information ID C and RPW c obtained during user registration: (4) Finally, the attacker can calculate the session key SK according to the above parameters: Proposed Scheme In this section, we describe the specific process of the protocol and the overall architecture diagram. e main body of the protocol includes users and servers. e agreement consists of four phases: user registration, login, authentication, and password change. Figure 1 illustrates the architecture of the proposed protocol. User represents the main participant in the communication, and server represents the entity that communicates with the user. Figure 2 illustrates the user registration phase. e detailed steps are as follows: User Registration Phase. (1) First, U c selects their ID c , password PW c , and bio information R i , as well as an arbitrary number m, to calculate en, ds is used to encrypt ID c , with the result: U c then transmits DI D c , BRPW c to S through a secure channel. (2) After receiving the information from U, S selects an arbitrary number y c to decrypt DI D c , obtains the value of ID c , and then calculates (3) Finally, the calculated parameters β c , c c , χ c , DI D c , h(·) are stored in the smart card, and S sends the smart card to U through a secure channel. U calculates η c after receiving the message: en, η c is saved in the smart card, and the registration process of the user is complete. Login Phase. (1) U enters their own ID c ′ , PW c ′ , and bio information R i . (2) After inputting the information, calculate en, verify whether χ c ′ and χ c are equal. If they are equal, the verification passes; otherwise, the login request sent by U to S is rejected. (3) If the verification passes, the reader will calculate en, the login request DI D c , ω c , υ c , T 1 is sent to the server. Authentication Phase. is section describes the process of mutual authentication between S and U. After the user sends the login request to the server, the server starts to verify whether U is legitimate by calculating a series of parameters, and U verifies the validity of S by calculating the values of some parameters. e authentication process is described in detail below. e login phase and authentication phase are shown in Figure 3. (1) After S receives the request from U, it first verifies whether the present timestamp is reasonable. It then decrypts DI D c to obtain ID c and calculates S verifies whether υ c ′ and υ c are equal. If not, S rejects the login request from U. If equal, S receives the login request from U and then calculates the session key of both sides: (2) After calculating the session key, S continues to calculate en, S passes μ c , T 2 to U (3) After receiving the message from S, the user first verifies the validity of the timestamp T 2 and then calculates U verifies whether μ c ′ is equal to μ c . If it is equal, U calculates the session key: Here, the authentication process for U and S is completed. Password Change Phase. If U wants to change their password PW c to PW N c , the following steps are performed: (1) U first inserts their own smart card and enters their ID c , current password PW c , bio information R i , and new password PW N c . User User Register Server Internet Figure 1: Network architecture. Smart card SC c Select an arbitraty number y c (2) According to some parameter values in the smart card and their own identity information, the following are calculated: Security and Communication Networks If the calculated value of χ c ′ is equal to the value of χ c stored in the smart card, the user is considered legitimate and allowed to change the password. (3) Some parameter values need to be updated in the process of password modification. e specific calculation process is as follows: (4) Finally, the values β c , c c , χ c , η c stored in the smart card are updated to the modified values β N c , c N c , χ N c , η N c , and the process of password modification is completed. Formal Security Analysis. Burrows-Abadi-Needham (BAN) logic [20] has been used in several studies to prove whether a protocol can be executed securely. is section uses BAN logic to prove the security and reliability of our proposed protocol. is proof verifies that our protocol can successfully establish and share a session key between the user and server. In the following proof, U represents the user and S represents the server. e specific proof rules and process are as follows: Detailed Steps. By considering the message M1 and using the seeing rule, we get S1: S⊲ 〈ID c 〉 ds , 〈α c , y c 〉 ds , T 1 . Using S1, we get S2: S⊲ 〈ID c 〉 ds . Under the assumption of A2, using S2, R1 can be used to obtain S3: S \| ≡ U \| ∼ (ID c ). ROR Model. is paper follows the ROR (Random Oracles) model under the proof of security, and two participants U and S are mentioned in the paper. First, let H x U and H y S as the xth user and yth server, respectively. en, let U � H x U , H y S and A can perform the following operations. Execute(U): by executing this query, A can get the messages transmitted by U and S through the common channel. Send(U, M): with the help of send query, A can send messages to U and S. In addition, A can also receive response messages from two participants. Corrupt(U): with the help of this query, A can obtain the parameters information stored in the smart card as well as some temporary parameters information and long-term key. Hash(String): by performing this operation, A can obtain the value in the hash. Test(U): this operation is mainly used to verify whether the session key between the user and the server is secure. By tossing a homogeneous coin C, the result of the coin is known only to A. If C � 1, A can know the correct session key. If C � 0, a null value is an output. Definition 1 (one-way anticollision hash function): this is a common mathematical function that inputs a variable length field and then produces a fixed length output. If Adv(m) � Pr[(m, n)ε R A; h(m) � h(n)] ≤ t for at most run time m, the hash function is considered hash collision proof. Definition 2 Symmetric encryption method is used in the proposed protocol. Suppose E K 1 , E k 2 , . . . , E K n are encryption methods based on different keys K. In the model, the probability that A can crack the correct session key Theorem 1. If A is a polynomial time η opponent executing our scheme under the ROR model and we choose to look at Zipf's law [28] for the user's password, the possibility of A damaging the session key is Adv P send , t send /2 l + where l represents the length of the password. Security Proof Proof. In the proof process, we define six games GM 0 to GM 5 and prove the theorem mentioned above according to the defined six game rules. Succ GMi A (η) represents the probability of A ′ s success in the game. e specific proof is as follows. According to the birthday paradox, it can be concluded that the maximum probability of hash collision is t 2 hash /2 l+1 . erefore, it can be concluded that the maximum probability of hash collision of text transmitted by both sides of the session is (t send + t exe ) 2 /2 u . Finally, we can draw a conclusion \|Pr [Succ GM2 hash /2 l+1 + (t send + t exe ) 2 /2 u . e symbol l appearing in the formula represents the length of the hash value and u represents the length of the transmitted text. GM 3 : on the basis of the above game rules, we added the provision that A can obtain the parameters information stored in the smart card in the new round of game, that is, A can obtain the parameters β c , y c , c c , DI D c by executing the Corrupt operation. On this basis, we perform an offline password guessing attack. First, A calculates α c � β c ⊕h(ID c ′ ⊕BRPW ′ ), U ′ s biological information R i are confidential to us, so they cannot be obtained. According to Zipf's law [28], we can draw a conclusion: \|Pr[Succ GM3 send , t send /2 l . GM 4 : in this game rule, we analyze the security of the communication session key between both sides. We mainly analyze it from the following three aspects. e first is to prove that the protocol has perfect forward security. e second is to prove that A can block the user impersonation attacks. e third is that A can block the known session-specific temporary information attacks. Perfect forward security: A obtains the value of the long-term key ds through Corrupt. Known session-specific temporary information attacks: A obtains the value of temporary information m or y c through Corrupt query. User impersonation attacks: A obtains the information DI D c , ω c .v c , T 1 transmitted by both communication parties through the public channel through Exe query, but U ′ s identity ID c is obtained by symmetric encryption with the long-term key ds. However, the value of the long-term key ds cannot be obtained. e session key SK � h(ID c ⊕α c ⊕y c ⊕T 1 ⊕T 2 ) of both communication parties: in the first case, A must obtain the values of α c and y c in order to obtain the session key, but the value of α c needs U ′ s biological information. In the second case, A obtains the value of temporary information, but U ′ s identity ID c is obtained through symmetric encryption. In the third case, because U ′ s identity ID c is obtained through symmetric encryption, A cannot obtain U ′ s real identity, so it is impossible to carry out simulated attacks. erefore, we can conclude that \|Pr[Succ GM4 To sum up, we can get it according to the above formula: So, we come to the final conclusion Adv P Informal Security Analysis. In this section, we further show that the proposed scheme is secure against the following attacks. Privileged-Insider Attack. In this protocol, even if the attacker obtains the information DI D c , BRPW c of the user in the registration process and the information β c , c c , χ c , η c in the smart card, they cannot successfully obtain the session key. Because SK � h(ID c ⊕α c ⊕y c ⊕T 1 ⊕T 2 ) and the user's ID c is encrypted by ds before being transmitted to the server, even if the attacker obtains the value of DI D c and BRPW c , the attack is futile. erefore, this protocol can resist privileged-internal attacks. Offline Password-Guessing Attacks. Suppose the attacker gets the message in the smart card; then, based on this message, they can guess the password offline. Even if the η c value in the smart card is obtained and the values of ID c and PW c are guessed, the offline password-guessing operation cannot be successful. is is because the calculation of m also involves the value of the user's biological information R i , and the value of R i is difficult to obtain. erefore, this protocol can effectively resist offline password-guessing attacks. Replay Attack. Suppose that the malicious attacker intercepts the login information DI D c , ω c , ] c , T 1 and authentication information μ c , T 2 and attempts to replay the login request. e request is invalid because we use the timestamp T 1 in the protocol to verify whether the time difference is within the set time threshold. Similarly, if the attacker intercepts the authentication message and attempts to make the authentication request, the user will also test the validity of the timestamp. erefore, the protocol can effectively resist replay attacks. Forward Secrecy. Assuming that the attacker obtains the value of the long-term password ds, they can only use this value to decrypt DI D c to obtain the value of the user's ID c . However, because SK � h(ID c ⊕α c ⊕y c ⊕T 1 ⊕T 2 ), it is not Security and Communication Networks sufficient to only know the value of the user's ID c . e values of the parameters α c and y c cannot be obtained. erefore, this protocol can provide perfect forward security. Known Session-Specific Temporary Information Attacks. Assuming that the attacker obtains the value of temporary session information m or y c , the session key cannot be obtained successfully. Because the session key calculation is composed of ID c , but ID c is encrypted by long-term key ds, the ID c cannot be obtained by the attacker. erefore, this protocol can successfully resist known session-specific temporary information attacks. User Impersonation Attacks. Suppose that the attacker wants to carry out a user impersonation attack. ey must first obtain the value of ID c , but ID c is encrypted by the longterm key ds, and so, it is difficult for the attacker to obtain its value. In addition, assuming that the attacker intercepts the message DI D c , ω c , ] c , T 1 from the public channel and sends it to the server for verification, the user needs a certain amount of time to decrypt DI D c . erefore, when the server receives the message from the attacker for verification of the timestamp, it will find that the timestamp exceeds the set time domain and reject the login request. In this way, our protocol successfully resists user impersonation attacks. Mutual Authentication. In this protocol, users and servers can successfully authenticate each other. First of all, the server authenticates the user through the value of υ c sent by the user. Similarly, the user can verify whether the server is legitimate through the value of μ c sent by the server. Only legitimate users and servers can pass the authentication. erefore, this protocol can effectively provide mutual authentication between users and servers. Security and Performance Comparisons is section discusses the security and performance analysis of the proposed protocol. Security analysis is mainly conducted through a comparison with other proposed protocols in the resistance of some common attacks, and performance analysis is mainly performed through a comparison with the time and communication costs of other protocols. Security Comparisons. In this section, the protocol proposed in this study is compared with recent related protocols. Owing to the development of different types of attack technology and methods, previous protocols are now incapable of resisting some common attacks. At present, the common network attacks include A1: privileged-internal attack, A2: offline password-guessing attack, A3: replay attack, A4: perfect forward secrecy, A5: known sessionspecific temporary information attacks, and A6: user impersonation attacks. e comparison results are presented in Table 2. A "Yes" means that the protocol can resist the attack, whereas a "No" means that it cannot. While the other related protocols each fail in some of the security attacks mentioned above, our proposed protocol can resist all the attacks, making our proposed protocol more secure and reliable. Performance Comparisons. To better analyze the performance of this protocol, we compared it with a previous protocol. To obtain more convincing results, we analyzed the protocol using the same tools and under the same conditions and used the data provided by Rana et al. [19] e results show that different protocols have different execution times in the same execution environment. e time required for the connection operation and the noncollision hash function was 0.00014 ms and 0.00089 ms, respectively. e time required for the exception and encryption and decryption operations was extremely small, and so, it was not calculated. In addition, the number of bits required for the user name, password, arbitrary number, and integer was 160; the number of bits required for the private key and public key of the server was 256; the number of bits required for encryption and decryption was 512; and, the number of bits required for the exclusive or operation and noncollision hash function was 160 and 256, respectively. e symbols for each encryption operation are as follows: T ‖ : time required for connection operation T ⊕ : time required for XOR operation T Enc/Dec : time required for encryption/decryption T h : time required for hash operation First, we compared the communication cost of our proposed protocol with that of previous protocols. In particular, our protocol was compared with those proposed by Rana et al. [19], Kaul and Awasthi [18], Khan et al. [31], Chang et al. [17], and Kumari et al. [16]. e communication overhead of our protocol is 3136 bits, whereas that of the protocols proposed by Rana et al. [19], Kaul and Awasthi [18], Khan et al. [31], Chang et al. [17], and Kumari et al. [16] are 3296, 2668, 3744, 2336, and 3296 bits, respectively. As shown in Figure 4, the communication cost of our protocol is lower than that of Rana et al. and Khan et al., but slightly higher than that of Kaul and Awasthi [18]. Although the communication cost of Chang et al. is small, the protocol proposed by them cannot effectively resist privilege internal attacks, offline password guessing attacks, and replay attacks. Next, we compare the running time cost of our proposed protocol with those of the three protocols mentioned above. e operating cost of our protocol is 0.01512 ms, whereas that of the protocols proposed by Rana et al. [19], Kaul and Awasthi [18], Khan et al. [31], Chang et al. [17], and Kumari et al. [16] are 0.0215 ms, 0.021 ms, 0.01965 ms, 0.01318 ms, and 0.02191 ms, respectively. As shown in Figure 5, the running time of our proposed protocol is shorter than that of the four protocols mentioned above. Although the time consumption of the protocol proposed by us is a little higher than that proposed by Chang et al., the protocol proposed by Chang et al. has the problem of security. It can be said that our protocol has better performance than the ones mentioned above. rough the analysis of Tables 2 and 3, our protocol is slightly higher than Kaul and Awasthi's [18] protocol in terms of communication cost, but Kaul and Awasthi's [18] protocol cannot resist user simulation attacks. Because our proposed protocol can more effectively resist various security attacks, our protocol is more applicable in future works. Conclusions In this study, we analyzed the next generation Internet of ings remote protocol proposed by Rana et al., and found that their protocol cannot resist all kinds of security attacks as they claim. Specifically, we found that their protocols are vulnerable to offline password-guessing attacks and Data Availability No data were used to support this study. Conflicts of Interest e authors declare that they have no conflicts of interest.	2021-10-12T13:14:48.666Z	2021-09-30T00:00:00.000	{ "year": 2021, "sha1": "e78b56f66e2d5d0853a6afa84916fc63ce32e11f", "oa_license": "CCBY", "oa_url": "https://downloads.hindawi.com/journals/scn/2021/6537678.pdf", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "3733c3ad2f07586c839a927e00b192e8ff6e0290", "s2fieldsofstudy": [ "Computer Science" ], "extfieldsofstudy": [ "Computer Science" ] }
269812731	pes2o/s2orc	v3-fos-license	Analysis of Environmental Art in the Interior Design Concepts Interior space in the concept of art, mainly refers to the art of generalized acts, works of art are in the aesthetic characteristics of the concept of factors. It deconstruction, the common structure composed of colorful indoor environment, indoor deconstruction of the main structure is defined in the indoor "re-re-group structure," a re-interpretation of the "structure" and the extension of criticism. This extension of critical activities with the reunification of East and West is to be re-transformation, the original interior space, structure to be broken, the reorganization in an attempt to view art on the basis of established new visual order and psychological space. Introduction In interior design, try to retain the structure of the building or ' deconstruction ', to decorate and process the structure, rather than cover ; this concept is influenced by ' deconstruction '.Some large spaces with few building structures should be modeled with building blocks to enhance the sense of volume and quantity of space.Architectural design should consider more factors in the later stage of interior design in terms of structure or ' deconstruction ', so as to form the integration of the two, rather than the separation. Deconstruct the environment interior design from the artistic concept characteristic If this situation is pointed to a more specific field of interior design, it is not difficult for us to find that in the conceptual characteristics of art, it will inject new vitality into interior design activities, and at the same time give interior design thinking a new vision and development space.Similarly, in the past simple and superficial discussion of the relationship between interior design activities and architectural design, this process was once commonplace, but it was not much practical operational significance.However, if we can explore the impact and interaction between physical architecture and interior space design from another perspective, there will be a new common research platform and guidelines as well as the rules of the game.Especially in the early design of the building, it has laid a foreshadowing for the interior design, reserved the space for the development of thinking and creation, and also made the interior design have more flexible design tension and scalability. To interpret the diversified interior space from the perspective of art design To re-interpret from Duchamp's modern art to the current diversified art concepts from the perspective of art and design, we are in a position to sort out the theoretical model of today's interior design.If we simply understand and treat the components of interior design as we did in the past, we will find that they are still old and rigid in terms of lines, modelling, space division and even the use of materials.At present, interior design is still in the centre of "design" itself, focusing on "its own rules" and "practicality" of design, which of course has a great influence on the strong commercial penetration of interior design.This is of course related to the strong commercial penetration.Conceptual art in the contemporary cultural context actually originates from a cultural paradox, which is not only the inevitable product of global post-modern and post-colonial cultural discourse, but also the reconstruction of Chinese contemporary art texts with the help of Western conceptual art language.Modern art design and modern art are equally influenced by the wave of thinking, market system and economic laws, and the individual design mode and group design mode of modern design are also constantly and increasingly updated. In front of the influence of conceptual art, contemporary designers' design concepts are also facing the same painful choice, which is the value judgement between the powerful economic and market drive and the persistent pursuit of artistic creation, how to find the critical point between the two poles has become a huge and unavoidable problem for contemporary designers.It is not difficult for us to find that the conceptual conversion of interior design and the development of conceptual art should have similar spiritual qualities.In conceptual art, the flexibility of the expression of painting, cultural representativeness, in the field of the ultimate spiritual pursuit of art is far beyond the practicality of interior design, utilitarian, thus more operable and more intense aesthetic infectious force, which can touch on the essence of the human art of the deepest cultural context.It can touch the deepest cultural context of human art essence.And this is exactly the interior design by its own marginal disciplinary restrictions, it is difficult to reach the field, at least for the current situation is so, it touches the art category is different from the conceptual painting and other forms of art. Interior design in the commodity economy, science and technology, computer production, the influence of new materials derived from the new face, with a strong programme and pretentious but only the lack of the most infectious personalised features and cultural factors connotation.Therefore, in order to change this situation, the focus of interior design activities should be gathered to the conceptual conversion to accept more artistic conceptual factors. Interior design should focus on style and form of multi-level breakthroughs, advocate the academic atmosphere of the design community has commented on, advocate to adapt to each era or stage of the different mainstream, follow the art of development of their own laws and regulations, to achieve a certain period of time the mainstream concept of the design of the convergence of the diversity of forms of expression of the ideal state of the reconstruction of the new interior design thinking and concepts, in sorting out, understanding of the history of modern interior design, rethinking and studying the international and local It will rethink and study the relationship between internationalisation and localisation, focusing on the management and construction of aesthetics and taste, environment and atmosphere, rebellion and reconstruction, psychological feeling and cultural connotation in interior design activities. Analysing the concept of contemporary environmental interior art from the deconstructive elements of art Frank Gehry, the most famous contemporary architect in the United States, not only pays close attention to art, but also maintains a very close relationship with contemporary artists, and he once said, "In a certain sense, I may be an artist, and I may have crossed the valley between the two."It is also true to say that he is constantly trying to apply certain concepts and techniques from art to his architectural designs.Dr Yang Zhijiang of Southeast University, in his doctoral thesis "Architecture in the Vision of Contemporary Art", made a brief table of the interaction between art and architecture, corresponding to the influence and interaction between artists and architects on the characteristics of art, and proposed that the influence of conceptual art on architecture makes it a way of life and thinking, an exercise in thought, and a link with a variety of humanities.The interior design process, then, also requires a new shift in rational thinking, but this does not mean that the design is complicated and purely artistic. Conclusion Humanity has entered the 21st century, interior design is an important activity to improve the material and cultural life of human beings and beautify the living environment, and the core of its design is to continuously innovate and create activity space and aesthetic space that is more suitable for human habitation.Under the influence of conceptual art and architectural art, it is the responsibility of contemporary designers engaged in interior design to raise the art of interior design to a higher level, to take diversification and multidimensionality as the value orientation, and to take excellent cultural traditions as the aesthetic orientation, and to create more pioneering and original interior design works.Deconstructionism can not completely replace today's interior design, but today's interior design trend of a borrowed, its real significance lies in the "conceptual" touch and influence on interior design.Because, deconstructionism on the function of the resistance to the structure of the confrontation, and for the form and sacrifice function is the fundamental principle of interior design -both "functional" some of the backlash, which is precisely the need for interior designers in the process of design practice should rationally This is exactly what interior designers should rationally absorb and identify in the process of design practice.	2024-05-18T15:55:10.678Z	2024-04-27T00:00:00.000	{ "year": 2024, "sha1": "d993d4ccb8398be53c2e42a8b7b63b5d577cb174", "oa_license": "CCBYNC", "oa_url": "https://bcpublication.org/index.php/FHSS/article/download/6260/6078", "oa_status": "HYBRID", "pdf_src": "Anansi", "pdf_hash": "50c21f5ab984f82e4d5f9247ffc451ee99c764cf", "s2fieldsofstudy": [ "Art", "Environmental Science" ], "extfieldsofstudy": [] }
235645088	pes2o/s2orc	v3-fos-license	Prevotella melaninogenica, a Sentinel Species of Antibiotic Resistance in Cystic Fibrosis Respiratory Niche? The importance and abundance of strict anaerobic bacteria in the respiratory microbiota of people with cystic fibrosis (PWCF) is now established through studies based on high-throughput sequencing or extended-culture methods. In CF respiratory niche, one of the most prevalent anaerobic genera is Prevotella, and particularly the species Prevotella melaninogenica. The objective of this study was to evaluate the antibiotic susceptibility of this anaerobic species. Fifty isolates of P. melaninogenica cultured from sputum of 50 PWCF have been included. Antibiotic susceptibility testing was performed using the agar diffusion method. All isolates were susceptible to the following antibiotics: amoxicillin/clavulanic acid, piperacillin/tazobactam, imipenem and metronidazole. A total of 96% of the isolates (48/50) were resistant to amoxicillin (indicating beta-lactamase production), 34% to clindamycin (17/50) and 24% to moxifloxacin (12/50). Moreover, 10% (5/50) were multidrug-resistant. A significant and positive correlation was found between clindamycin resistance and chronic azithromycin administration. This preliminary study on a predominant species of the lung “anaerobiome” shows high percentages of resistance, potentially exacerbated by the initiation of long-term antibiotic therapy in PWCF. The anaerobic resistome characterization, focusing on species rather than genera, is needed in the future to better prevent the emergence of resistance within lung microbiota. Introduction The importance and abundance of strict anaerobic bacteria in the respiratory microbiota of people with cystic fibrosis (PWCF) is now established. These recent data are based on high-throughput sequencing [1,2] and cultural techniques [3][4][5]. Anaerobes colony count has been evaluated between 1.10 4 to 9.10 7 colony forming units per milliliter in sputum culture [3]. Numerous anaerobic genera have been identified as part of the CF core lung microbiota such as Prevotella, Veillonella, Porphyromonas, Fusobacterium, or Peptostreptococcus [2,6,7]. However, these bacteria remain the unknowns of the lung, both in terms of diversity, resistance or impact on the pathophysiology of CF disease [8]. One of the most prevalent and abundant anaerobic genera described in CF lung is Prevotella, and particularly the species Prevotella melaninogenica [4,5,9]. Contradictory hypotheses on the potential role of Prevotella genus have been previously developed [8]. On the one hand, it has been associated with better lung function and less inflammation [2,[10][11][12]. On the other hand, Prevotella species are able to enhance other bacteria's pathogenicity in lung [13], to produce pro-inflammatory short chain fatty acids [14] and are described to be resistant to antibiotics [15]. Focusing on this last point, previous studies have reported resistance acquisition of Prevotella isolates, due to the production of enzyme (mainly betalactamase [13,16,17]) or the carrying of the resistance gene [18,19]. Moreover, in CF, an over-expression of antibiotic resistance may be observed due to several reasons: repeated or chronic use of antimicrobial treatment, administration of high doses of antibiotic therapy, or prescription of antibiotic molecules at sub-clinical doses for anti-inflammatory properties. As the issue of antibiotic resistance is crucial, even more in PWCF, antibiotic susceptibility monitoring of this main anaerobic genus is necessary. The objective of this study was to evaluate the antibiotic resistance rate of Prevotella melaninogenica, the most prevalent anaerobic species of the bronchopulmonary microbiota of PWCF. Patients Characteristics Fifty PWCF were included in the study. The following sociodemographic and clinical parameters were recorded: age, sex, CFTR (CF transmembrane conductance regulator) gene mutation, lung function, diabetes, chronic antibiotics administration (azithromycin, aztreonam, colistin, and tobramycin), and antibiotic administration one month before the sample. The patients' lung involvement was assessed by the measurement of the maximum expiratory volume per second (FEV 1 ) by spirometry tests, which was defined by three stages: early (FEV 1 > 70%), intermediate (40 < FEV 1 < 70%) and advanced (FEV 1 < 40%) [20]. Statistical Analysis Statistical analysis was performed using the Chi-square and Fisher tests. The alpha risk has been reported significant when it was less than 0.05. Antibiotic Susceptibility Testing Antibiotic susceptibilities of P. melaninogenica isolates are presented in Table 2. All isolates were susceptible to the following antibiotics: amoxicillin/clavulanic acid, piperacillin/ tazobactam, imipenem, and metronidazole. A total of 96% of P. melaninogenica isolates (48/50) were classified as beta-lactamase producers (MIC for amoxicillin > 0.5 mg/L), and 38% of these isolates (18/48) had a MIC for amoxicillin superior to 256 mg/L ( Figure 1). The MIC for amoxicillin for the two non-beta-lactamase producers isolates have been evaluated at 0.094 mg/L and inferior to 0.016 mg/L (Figure 1). A total of 34% (17/50) and 24% (12/50) of isolates were resistant to clindamycin and moxifloxacin, respectively. Ten percent of isolates (5/50) were categorized MDR, due to a combined resistance to amoxicillin, clindamycin and moxifloxacin. Associations between Antibiotic Resistance and Antibiotic Administration Amoxicillin, clindamycin and moxifloxacin resistance percentages were compared to antibiotic administration one month before the sample (oral or intravenous: yes or no) and chronic antibiotic administration (azithromycin, aztreonam, colistin or tobramycin: yes or no). A significantly positive correlation was found between clindamycin resistance and oral chronic azithromycin administration (p = 0.002, Chi-square test). Discussion The potential impact of strict anaerobic bacteria in the pathophysiology of CF disease and in the respiratory function of PWCF is still at the hypothesis stage. The important and now undeniable place of these bacteria within the CF respiratory microbiota led to the study of their antibiotic susceptibilities to better understand the issue of resistance. For the genus Prevotella, disparities in resistance rates have been observed among the different studies but all the results agree on an increase in antibiotic resistance from clinical isolates in CF, as well as in other diseases [15,23]. Concerning beta-lactams, resistance is mainly due to the production of beta-lactamases, which induce a loss of susceptibility to penicillin, and that of the first, second, and third (oral) generation cephalosporins [21]. Several beta-lactamase detection methods can be used, combining phenotypic (mostly performed in routine practice) and molecular (cfxA resistance genes detection approaches [18,19]). Indeed, the detection of cfxA genes in Prevotella isolates has been associated with high MICs of amoxicillin [19,24]. However it remains important to determine if the isolates are functionally resistant to amoxicillin [25]. In this study focusing on P. melaninogenica, a high percentage of isolates (96%) were categorized as beta-lactamase producers using phenotypic detection based on the MIC of amoxicillin determination. This result is slightly higher than previous results for this species (Table 3) which can be explained by three main reasons. Firstly, the other studies used the nitrocefin test for phenotypic detection of beta-lactamase. However the recommendation of EUCAST/CASFM 2019 [21] is to avoid nitrocefin, considered as a poor substrate which may underestimate beta-lactamase activity detection. Moreover interference with the isolates pigmentation (as P. melaninogenica) represents a limit to this test based on a colorimetric reaction [26]. As formerly demonstrated, the determination of the MIC of amoxicillin is a more reliable method of detection [27]. Secondly, this study focused only on P. melaninogenica species while others' studies are based on several Prevotella species, for which resistance rates may differ [26]. Thirdly, in this study, isolates have been obtained from sputum of PWCF, for which antibiotic selective pressure is higher due to chronic antibiotic therapy administration and may lead to diminution of antibiotic susceptibility [15]. Moreover, CF lung has been described as a potential reservoir of resistance genes (as oral sphere) because cfxA genes coding beta-lactamase could be transferred among bacteria by horizontal transfer gene mechanisms [8,28,29]. Fourthly, as demonstrated for other genera known to carry cfxA genes (Capnocytophaga, Bacteroides, . . . ), the expression of beta-lactamase may differ according to major substitutions in the cfxA gene sequence or to the influence of the genetic environment of cfxA [30]. This molecular part has not been evaluated in this study. For the other beta-lactams, all of the isolates of the present study were susceptible to piperacillin/tazobactam, amoxicillin/clavulanic acid, and imipenem (Table 4). No resistance have been described to date for amoxicillin/clavulanic acid and imipenem, however resistance to piperacillin/tazobactam was previously reported in the literature for Prevotella spp. [3,31]. As beta-lactams, emerging resistance to macrolides and fluoroquinolone has been described for Prevotella spp. [42,43]. In our study, 24% of P. melaninogenica isolates were resistant to moxifloxacin, which was concordant with the results of Ulger Toprak et al. [26] (Table 4). Thirty-four percent of P. melaninogenica isolates were resistant to clindamycin, which was higher than previously described in non-CF individuals, as no resistance data are available for PWCF [18,26,32]. Chronic azithromycin administration, prescribed at sub-clinical doses for anti-inflammatory properties, was significantly correlated with clindamycin resistance (p < 0.05, Chi-square test), as previously highlighted in PWCF [16]. This result is consistent with the fact that one of the main factors leading to antibiotic resistance identified for anaerobic bacteria is long-term antibiotic use [43,44]. Sherrard et al. [15] showed that the administration of betalactam or tetracycline in the year prior to pulmonary sample collection in PWCF was associated with higher MICs of amoxicillin, ceftazidime and tetracycline for Prevotella spp. In the future, the worsening resistance of the Prevotella genus is a threat and should be monitored. Firstly, different studies highlighted the increase of MDR anaerobic isolates. Ulger Toprak et al. [26] have demonstrated that nearly 10% of Prevotella isolates were categorized MDR (antibiotics involved: ampicillin, clindamycin, moxifloxacin and tetracycline). In our study, a similar rate of MDR isolates was determined (five isolates of P. melaninogenica resistant to amoxicillin, moxifloxacin and clindamycin). Secondly, in addition to beta-lactamase activity, Prevotella species are able to produce extended-spectrum beta-lactamase (ESBLs), associated with treatment failure with cephalosporin antibiotics. In a study focused on Prevotella isolates isolated from respiratory samples of PWCF, 76% were found to produce ESBLs [17]. This resistance is rarely investigated in routine practice, as in our study, and may be underestimated. Finally, resistance to metronidazole, an anti-anaerobic molecule, has been reported for P. melaninogenica [18] and Prevotella spp. [31,42,[45][46][47] with a resistance rate under 10% (Table 4). Tunney et al. [3] have observed a higher resistance rate evaluated at 46% in Prevotella spp. isolates (including four species: P. corporis, P. disiens, P. melaninogenica and P. salivae). No metronidazole resistant isolates of P. melaninogenica were described in our study. For all the antimicrobial categories, results obtained in our study compared with previous ones showed the importance of the species level when considering antibiotic resistance and the difficulty to compare antibiotic resistance from isolates of different species belonging to the same genus [26] (Table 4). The targeted study based on species is therefore the most informative. In routine practice, this approach requires equipment (e.g., anaerobic chamber, collecting device for anaerobic atmosphere preservation) for the detection and identification of strict anaerobic bacteria in CF respiratory samples as well as for the realization of antibiogram. Conclusions This preliminary study of 50 clinical isolates of a predominant species of the lung "anaerobiome" showed a high percentage of resistance and the presence of MDR isolates, potentially exacerbated by the initiation of long-term antibiotic therapy in PWCF. In the future, it will be important to continue to characterize the anaerobic resistome by focusing on species rather than genera in order to better prevent the emergence of resistance within pulmonary microbiota.	2021-06-27T05:24:23.203Z	2021-06-01T00:00:00.000	{ "year": 2021, "sha1": "cf41fedf48b8621898ee536bcb740d57d5ed82f9", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2076-2607/9/6/1275/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "cf41fedf48b8621898ee536bcb740d57d5ed82f9", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
232161745	pes2o/s2orc	v3-fos-license	Reoperative Mitral Surgery Versus Transcatheter Mitral Valve Replacement: A Systematic Review Abstract Bioprosthetic mitral structural valve degeneration and failed mitral valve repair (MVr) have traditionally been treated with reoperative mitral valve surgery. Transcatheter mitral valve‐in‐valve (MVIV) and valve‐in‐ring (MVIR) replacement are now feasible, but data comparing these approaches are lacking. We sought to compare the outcomes of (1) reoperative mitral valve replacement (redo‐MVR) and MVIV for structural valve degeneration, and (2) reoperative mitral valve repair (redo‐MVr) or MVR and MVIR for failed MVr. A literature search of PubMed, Embase, and the Cochrane Library was conducted up to July 31, 2020. Thirty‐two studies involving 25 832 patients were included. Redo‐MVR was required in ≈35% of patients after index surgery at 10 years, with 5% to 15% 30‐day mortality. MVIV resulted in >95% procedural success with 30‐day and 1‐year mortality of 0% to 8% and 11% to 16%, respectively. Recognized complications included left ventricular outflow tract obstruction (0%–6%), valve migration (0%–9%), and residual regurgitation (0%–6%). Comparisons of redo‐MVR and MVIV showed no statistically significant differences in mortality (11.3% versus 11.9% at 1 year, P=0.92), albeit higher rates of major bleeding and arrhythmias with redo‐MVR. MVIR resulted in 0% to 34% mortality at 1 year, whereas both redo‐MVr and MVR for failed repairs were performed with minimal mortality and durable long‐term results. MVIV is therefore a viable alternative to redo‐MVR for structural valve degeneration, whereas redo‐MVr or redo‐MVR is preferred for failed MVr given the suboptimal results with MVIR. However, not all patients will be candidates for MVIV/MVIR because anatomical restrictions may preclude transcatheter options from adequately addressing the underlying pathology. S tructural valve degeneration (SVD) of bioprosthetic mitral valves is common and frequently occurs after 5 or more years after initial valve replacement. [1][2][3][4][5] Reoperative surgical mitral valve replacement (redo-MVR) has been the gold standard treatment for bioprosthetic SVD for many years and is required in up to 35% of patients at 10 years after index surgery. 6,7 Redo-MVR is, however, associated with significant mortality and morbidities, particularly among elderly patients. 8 Similarly, mitral valve replacement (MVR) and reoperative mitral valve repair (redo-MVr) are options for patients with failed MVr, but with significant attendant risks in inexperienced centers. At the time of SVD or repair failure, patients tend to be older and hence are at a higher risk for reoperation. To meet the needs of these high-risk populations, transcatheter MVR (TMVR) using valve-in-valve (MVIV) and valve-inring (MVIR) techniques have emerged in the past few years. However, there have been no clinical trials comparing the aforementioned approaches, with very few recurring adverse events such as left ventricular outflow tract obstruction (LVOTO), valve thrombosis, valve migration or embolization, major vascular complication, life-threatening or major bleeding, and postprocedural mitral regurgitation. Note that a meta-analysis was not performed for a variety of reasons. First, most of the transcatheter series included patients with relatively small sample sizes. Although small samples can still be combined with caution, we felt that this would detract from the generalizability of our results. Second, among the surgical series compiled, short-term outcomes were consistently reported, but mid-and long-term outcomes were reported on a case-by-case basis given the wide heterogeneity among these studies. Moreover, we do not have long-term data following MVIV and MVIR. Finally, although the conduct of surgical MV replacement and repair is generally well established, the technical aspects of MVIV and MVIR are evolving. Thus, earlier transcatheter series differ considerably from later series, especially with regard to the delivery technique used. RESULTS Our initial search identified 2123 studies, 708 of which remained after de-duplication. 79 full-text articles were assessed for eligibility and 32 were included in our qualitative synthesis (Figure 1). The vast majority of these studies were retrospective cohort and case-control studies, with 1 prospective comparative cohort trial. Reoperative Surgical Mitral Valve Replacement Bioprosthetic mitral valves are subject to SVD because of prosthesis degeneration, thrombosis, and paravalvular leak and may lead to significant stenosis, regurgitation, or both. Rates of SVD vary widely in the literature but have been reported at ≈25% to 30% at 10 years and ≈50% to 70% at 15 years. 9 Similarly, freedom from reoperation is estimated at 96.6%, 86.6%, and 75.3% at 5, 10, and 15 years, respectively. 10 Short-and long-term outcomes The volume of redo-MVR has increased steadily over the past few decades, with a concomitant improvement in outcomes. 8,11,12 Mortality at 30 days following redo-MVR ranges from 5% to 15% ( Figure 2). 7,8,[13][14][15][16][17][18][19][20][21] Age, sex, preoperative New York Heart Association class, indication for reoperation, type of prosthetic valve, number of previous operations, hepatic and renal failure, and timing of reoperation have all been associated with mortality and postoperative complications after redo-MVR. 15 In-hospital mortality was 12%, and mortality rates at 1 and 5 years were 18.4% and 27.2%, respectively. 7 In another report of 347 reoperations on mitral prostheses, Bortolotti et al reported actuarial survival rates of 63±3% at 5 years, 38±4% at 10 years, and 24±5% at 15 years. 16 Note that the choice of prosthesis at the index operation has implications for potential reintervention because transcatheter approaches cannot be performed in patients with prior mechanical valves. Transcatheter Mitral Valve-in-Valve Replacement In the United States, the SAPIEN 3 valve (Edwards Lifesciences LLC, Irvine, CA) was approved in 2017 for MVIV by the U.S. Food and Drug Administration in high surgical-risk patients and is the only transcatheter valve that is approved for MVIV. 24,25 Surgical and Transcatheter Options in Failed Mitral Valve Repair Patients with failed surgical MVr have historically undergone rerepair or MVR. Given the risks associated with reoperation, MVIR, although not approved by the Food and Drug Administration, has emerged as an alternative option, but robust data comparing these approaches are lacking. 39 Mitral Valve Rerepair Following Failed Surgical Annuloplasty Short-and long-term outcomes following redo-MVr are generally favorable in experienced mitral centers. 40 There was 0% operative mortality with 96% long-term freedom from mortality at 5 years. 22 Surgical Mitral Valve Replacement Following Failed Annuloplasty MVR is also an option for patients with failed mitral bands or rings. Evidence suggests that, although inhospital mortality and rates of postoperative complications may favor MVR, long-term mortality is likely similar to that of surgical rerepair. 18 In the study by Kilic et al, 257 of 305 patients underwent MVR following a previous repair. 8% operative mortality was noted, along with an increase in blood transfusion rates and duration of mechanical ventilation compared with the redo-MVr cohort. However, long-term freedom from mortality was comparable to the 48 patients who underwent rerepair (P=0.29). 22 Multiple groups have reported on the outcomes of reoperative mitral valve surgery without making the distinction between rerepair and reoperative replacement. 8,23 In an early series, specific to reoperation in failed MVr, Cerfolio et al reported an operative mortality of 4.0%, with 87.0% of patients in New York Heart Association class I or II at 5-year follow-up. 42 A contemporary cohort of reoperation for failed MVr was recently analyzed, with a reported mortality of 6.5%. A propensity-matched cohort from that series indicated that the failed MVr itself did not affect survival compared with other reoperative procedures with normal mitral valves. 43 Other groups have reported mortality rates between 0% and 9.0%. 40,41,44,45 Transcatheter Mitral Valve-in-Ring Replacement In contrast to surgical reoperation, MVIR is much less invasive and avoids issues related to redo sternotomies and cardiopulmonary bypass. Several articles have reported preliminary outcomes. Patient characteristics are shown in Table 3. Thirty-day and 1-year mortality are estimated at 0% to 18% and 0% to 34%, respectively ( Figure 4). Studies of MVIR have also noted several recurrent complications (Table 4). 30,35,36,[46][47][48][49][50][51][52][53][54][55][56][57] Of note, only 3 of these studies reported the characteristics of the rings that had been implanted. 29,36,49 Studies with long-term outcomes following MVIR are limited. The largest series of TMVR cases published to date by Yoon et al included 141 MVIR cases from an international registry. Procedural success was achieved in 80.9% of these cases with a 30-day mortality of 9.9% and 1-year mortality of 30.6%. 35 DISCUSSION Two key conclusions are evident from this systematic review: (1) given its safety profile, MVIV is a viable alternative to redo-MVR in high surgical-risk patients with bioprosthetic mitral SVD, with certain caveats; [26][27][28][29][30][31][32][33][34][35][36] and (2) for those with failed surgical mitral annuloplasty, reoperative MVr or MVR is preferred to MVIR in suitable candidates. For certain high-or extremesurgical risk patients, however, MVIR may be appropriate after a thorough assessment of all procedural risks and evaluation of the type of ring that has been implanted. Any discussion of bioprosthetic mitral SVD and failed MVr begins with a basic understanding of the surgical principles of MVR and MVr and their relation to the technical aspects of MVIV and MVIR, respectively. In contemporary practice, when repair is not feasible, a valve replacement operation that leaves leaflets and chords intact is preferred for optimal preservation of cardiac function. During chordal-sparing valve replacement, the posterior leaflet is usually preserved, and the anterior leaflet may be divided centrally and then plicated to avoid prosthetic leaflet impingement and LVOTO. 58,59 In contrast, the anterior leaflet is almost always left intact during MVr. Thus, the anatomy after MVr is significantly different from that following MVR. As is discussed next, this has significant implications for subsequent transcatheter therapies, especially with regard to the risk of postprocedural LVOTO, and may partially explain why we observe better outcomes with MVIV than with MVIR. Reoperative Mitral Valve Replacement and Transcatheter Mitral Valve-in-Valve Replacement Prosthetic valve reoperations carry significant risks. Thirty-day mortality is estimated at 5% to 15% with significantly higher rates of adverse events, including major bleeding and atrial arrhythmias, when compared with MVIV. In contrast, short-term mortality for MVIV ranges from 0% to 9%, with comparable valve hemodynamics and improvements in functional status at 1 year. There are several common themes that emerge from the literature when assessing the merits and risks of redo-MVR and MVIV. First, patients who develop SVD or require redo-MVR are generally very sick and have a high competing risk of death. Second, those who present in a higher New York Heart Association class and/or with a lower ejection fraction for redo-MVR have worse outcomes, and earlier intervention may be warranted in select cases. Third, MVIV via a transseptal approach, given its safety profile, is a reasonable alternative to redo-MVR in select high surgical-risk patients. However, redo-MVR should be performed in patients who are not high risk for surgery, those with paravalvular leak not amenable to percutaneous closure, prosthetic valve endocarditis, prosthetic valve thrombosis, patients with a high risk of LVOTO, or when concomitant cardiac surgical procedures (eg, coronary artery bypass grafting, tricuspid valve repair, Maze procedure, etc) are needed. The risk of LVOTO, which can be assessed by preoperative computed tomography imaging, is a particularly important anatomical reason that precludes use of MVIV (discussed later). Finally, transapical MVIV is associated with higher mortality and morbidity as compared with transseptal MVIV, and the latter approach should be pursued whenever possible. Comparisons of transcatheter and surgical options for patients with bioprosthetic mitral SVD have to be evaluated carefully, however, for a number of reasons. First, patients assigned to these 2 treatment modalities are generally preselected based on surgical risk, and it is difficult to directly compare outcomes with wide applicability. Given the risk of selection bias, a randomized trial comparing the safety, effectiveness, and durability of redo-MVR and MVIV is eagerly awaited. Furthermore, when comparing redo-MVR and MVIV, it is important to note whether patients in the transcatheter group underwent transapical or transseptal MVIV. Studies have shown that transapical MVIV has inferior outcomes compared with its transseptal counterpart. These 2 delivery techniques differ in numerous aspects. Transapical implantation may enable better control over the implant position given that the cardiac apex is considerably closer to the mitral valve than the peripheral veins. 60 However, the transseptal approach eliminates the need for either thoracotomy or trauma to the left ventricle and is associated with a mortality benefit and greater improvements in postprocedural functional performance. 61 Thus, although transapical MVIV may not be a viable alternative to redo-MVR, transseptal MVIV should be considered for high-risk patients with suitable anatomy, with the caveat that the transseptal approach may be more technically challenging requiring operator expertise at high-volume centers. 62 Failed Surgical Repairs and Transcatheter Mitral Valve-in-Ring Replacement For patients with a prior ring or band annuloplasty, MVIR seems to be inferior to both redo-MVr and MVR. As with redo-MVR/MVIV, selection bias may be at play given that patients being considered for MVIR tend to be at high or extreme risk for reoperative surgery. Mortality with MVIR at 30 days is ≈0% to 18%, and ≈0% to 34% at 1 year. In contrast, reoperative MVr can be performed with <5% mortality at most large institutions with excellent long-term outcomes. Similarly, although short-term mortality for reoperative MVR may be comparable to that of MVIR, the long-term benefits with regard to valve durability and freedom from reoperation are undeniable with the former. Thus, redo-MVr and MVR should both be preferred to MVIR in appropriately selected patients. A crucial consideration in preoperative planning for MVIR is assessment of the annuloplasty ring currently in place. The 3 most important characteristics to consider are ring rigidity, shape, and radio-opacity. 63 Valves currently used for TMVR are designed to anchor and fully expand within a circular geometry. The mitral annulus, however, takes on a D-shape, and many annuloplasty rings have been designed to accommodate this. The rigidity of a ring determines its ability to conform to the circular shape upon implantation and thereby optimize valve function. Generally, rings are classified as rigid, semirigid, or flexible. Rigid rings provide the most anchoring capacity for a transcatheter valve but are the least able to conform and thus pose the greatest risk of para-and intravalvular regurgitation at the 2 corners. Semirigid rings provide an optimal balance between a solid anchor and the ability to adopt a circular shape in response to MVIR. However, smaller-sized semirigid rings tend to be more rigid and are thus less likely to fully circularize. 64 Rings may also be incomplete or complete in shape. Flexible-incomplete rings are generally implanted in an effort to preserve the 3-dimensional saddle-geometry of the annulus. Incomplete rings have an opening at the anterior leaflet and do not provide a solid anchoring surface during MVIR procedures. Furthermore, the discontinuous portions of incomplete rings may result in significant paravalvular leak. 65 Thus, the ideal ring for MVIR appears to be a semirigid, complete annuloplasty ring with adequate radio-opacity. The "Mitral Valve-in-Valve" mobile application, developed by Dr. Vinayak Bapat, serves as a valuable reference for preprocedural planning and optimal valve positioning during the procedure. 66 Left Ventricular Outflow Tract Obstruction Previously underappreciated, LVOTO is a serious complication of TMVR that results from displacement of the native anterior mitral leaflet into the LVOT upon expansion of the transcatheter valve. 67 This occurs in ≈0% to 6% of patients following MVIV and ≈0% to 20% following MVIR. Furthermore, procedural mortality with TMVR is significantly higher among patients with LVOTO. Anatomical factors that increase the risk of LVOTO include protrusion of the device into the left ventricle, device flaring, and a shallow aortomitral angle between the plane of the aortic and mitral annuli. Device-specific preoperative assessment of the "Neo-LVOT" is useful to identify those at risk of this complication, with a sensitivity of 96.2% and specificity of 92.3%. 35 Several risk-reduction strategies have been studied for patients at risk of LVOTO. 70,71 Laceration of the anterior mitral leaflet to prevent LVOTO is one such technique that involves splitting the anterior leaflet with an electrified wire. [72][73][74] In patients at risk of LVOTO owing to a prominent septum, preoperative alcohol septal ablation has also been evaluated with promising results. 75 Other Limitations Recognized other adverse events associated with MVIV and MVIR include (1) procedural and delayed device migration or embolization, (2) elevated postprocedural gradients, (3) residual mitral regurgitation, and (4) valve thrombosis. These complications appear to be more common in MVIR than MVIV, with small valve size (<25 mm) appearing as a consistent predictor of significantly elevated postprocedural gradients. Improvements in TMVR, including imaging and valve technology, will undoubtedly address these issues in the near future. Sources of Funding None. Table S1	2021-03-10T06:23:19.275Z	2021-03-09T00:00:00.000	{ "year": 2021, "sha1": "b0e120f20a45991921620fc2da9007693de1a771", "oa_license": "CCBYNCND", "oa_url": "https://www.ahajournals.org/doi/pdf/10.1161/JAHA.120.019854", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "b654118078d94a806f7730c618c3074e36578a8b", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
207376155	pes2o/s2orc	v3-fos-license	Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers Adipose tissue is highly involved in whole-body metabolism and is the main site for lipid synthesis, storage and mobilization in ruminants. Therefore, knowledge about adipose tissue responses to different diets is important, especially in growing heifers as the feeding regimes of replacement heifers affect their future success as dairy cows. However, at gene expression level such knowledge is limited. As part of a larger feed trial, adipose tissue biopsies from 24 Norwegian Red heifers were collected at 12 months of age (12MO) and at month seven of gestation (PREG) and analyzed by next-generation mRNA sequencing. Between these two sampling points, all heifers had gone through a successful conception and a feed change from four dietary treatments of high or low energy (HE/LE) and protein (HP/LP) content (treatments LPHE, HPHE, LPLE and HPLE) to a low-energy, low-protein pregnancy feed given to all animals. Gene expression differences between different feed treatments at 12MO are described in an earlier publication from our group. The main objectives of this study were to investigate the long-term effects of diets differing in protein and energy density level on gene expression in adipose tissue of growing replacement dairy heifers. To achieve this, we examined the post-treatment effects between the treatment groups at month seven of gestation; 6 months after the termination of experimental feeding, and the long-term gene expression changes occurring in the adipose tissue between 12MO and PREG. Post-treatment group comparisons showed evidence of long-term effects of dietary treatment on adipose gene expression. Differences between protein treatments were smaller than between energy treatments. Adipose gene expression changes from 12MO to PREG were much larger for the HE than the LE treatments and seemed to mostly be explained by the characteristics of the diet change. 97 genes displayed a unidirectional expression change for all groups from 12MO to PREG, and are considered to be treatment-independent, possibly caused by pregnancy or increased age. This study provides candidate genes and key regulators for further studies on pregnancy preservation (TGFB1, CFD) and metabolic regulation and efficiency (PI3K, RICTOR, MAP4K4,) in dairy cattle. However, if one acknowledges the fact that a cow is not just a function of its genome and its environment at a given moment, but rather a function of its genome and all its previous environments, it is clear that if we are to optimize dairy management, we also need information about metabolic regulatory mechanisms at earlier life stages. In this regard, adipose tissue plays a key role as it is not only the end point for feed excess energy and an energy reserve for periods of higher energy consumption than input [24], but is also a transcriptionally and endocrinologically active organ strongly involved in the regulation of whole-body metabolism [25,26]. There are several published gene expression studies on adipose tissue in growing cattle [23,[27][28][29][30][31], but so far no reports of post-treatment dietary effects or effects of pregnancy on the total adipose transcriptome. As part of a larger feed trial, adipose tissue biopsies were collected at 12 months of age (12MO) and at month seven of gestation (PREG, 20 (HE) and 24 (LE) months of age) from 24 Norwegian Red heifers and analyzed by next-generation RNA sequencing (RNA-seq). Between these two sampling points, all heifers had gone through a successful conception and a feed change from four dietary treatments differing in energy and protein content, to a low-energy, low-protein pregnancy feed given to all animals. During pregnancy, we observed different growth responses to the pregnancy feed between the dietary treatment groups. Such post-treatment differences have been described by others, and have been found to account for a larger fraction of milk yield variability than the experimental feeding per se [32]. It is therefore of great interest to investigate the metabolic regulation underlying these post-treatment effects. Thus, the main objective of this study was to examine posttreatment gene expression differences between the experimental feeding groups at PREG; six months after change to a consistent feeding regime. A secondary objective of this study was to describe the gene expression changes occurring in the heifers between 12MO and PREG. Finally, differentially expressed genes associated with pregnancy, diet change or post-treatment dietary effects identified in this study could potentially yield candidate genes for further studies on breeding and feeding for improved feed efficiency, production and reproduction in Norwegian Red cattle. Ethics statement All experimental procedures involving animals were approved by the Norwegian Animal Research Authority (FOTS id 2955, reference no. 2010/203231). Biopsies were harvested under local anesthesia, and all animals were given 3 mg/ kg BW of Ketoprofen intramuscularly (Comforion vet., Orion Pharma Animal Health) to prevent pain or inflammation at the biopsy site. 850-1000 or 600-750 g/day from three months of age to confirmed pregnancy, respectively. Each of the energy groups were split into two protein groups, low (LP) or high (HP), to give four dietary treatment groups with 20 animals in each group, viz Low-protein high-energy (LPHE), high-protein high-energy (HPHE), low-protein low-energy (LPLE) and high-protein low-energy (HPLE). Target body weight at conception was 400 kg for all heifers. To achieve this, first insemination was carried out at first heat after passing 370 kg for the HE heifers (fastgrowing) and after passing 380 kg for the LE heifers (slower-growing). No hormone treatments or other fertility enhancing treatments were applied. The aim of a common BW for all heifers at conception also meant heifers on HE treatments were inseminated at an earlier age than LE heifers, because of their faster growth. Age at successful insemination was 13.6 months for LPHE, 12.8 months for HPHE, 17.1 months for LPLE and 16.9 months for LPHE heifers, and PREG samples were taken 7 months after these age points. After confirmed pregnancy (at day 28-42 of gestation), all heifers were group-fed the same diet ad libitum. The pregnancy diet was designed for a moderate growth of about 550 g/ d until calving to reach a target BW of 560 kg and a maximum BCS of 3.75 at calving. Nutrient content in the experimental diets at 12 months of age and in the pregnancy diet at month seven of gestation is shown in Table 1. Adipose biopsy sampling and sample preparation Adipose tissue biopsies were harvested at 12 months of age (12MO) and at month seven of gestation (PREG). These time points were chosen to assess the long-term between-treatment differences and within-treatment changes when the heifers were in a steady state, without the interference of a recent feed change or an imminent calving. Biopsy sampling procedure, RNA extraction and library preparation were performed as described by Waerp et al. [34]. Briefly, 0.5-1.5 g of adipose tissue was harvested from the tail base of the heifers under epidural anesthesia. Biopsies were snap frozen in liquid nitrogen, and subsequently stored at -80˚C. 48 Adipose tissue samples from 24 heifers were selected for RNA sequencing (n = 6 for each group). RNA extraction was carried out using the Qiagen RNeasy lipid tissue mini kit (Qiagen). The quality and RIN values of the extracted samples were assessed using an Agilent bioanalyzer. RIN values varied between 7.8 and 9.2. To avoid DNA contamination, all RNA samples were subject to a DNAse treatment after extraction (Turbo DNA-free, Ambion, Life technologies) and a subsequent cleanup (RNA Clean & Concentrator -25, Zymo Research Corp). Library preparation was performed using the Illumina TruSeq RNA sample prep kit v2, according to the manufacturer's manual.Sequencing was performed in an Illumina HiSeq Statistics and Bioinformatics The fastq files containing the raw sequence reads were examined using FASTQC (v 0.11.2.) Adapters were trimmed from all sample fastq files using Cutadapt (v 1.3). Reads were simultaneously quality trimmed by removing reads shorter than 20 bp, or with Phred scores below 25. Following adapter and quality trimming, all files were checked again in FASTQC. Paired-end read files were aligned to the UMD3.1 bovine genome assembly [35], using Tophat (v 2.0.12) [36]. The number of reads per gene in each sample was counted using HTSeq-count (v 0.6.1) [37]. Statistical analysis of read counts were carried out in EdgeR (v 3.1.2), a Bioconductor software package run in the statistical software environment R (v 3.1.2) [38]. All differential gene expression analyses were performed in EdgeR, using tagwise dispersions. The statistical tests were corrected for multiple testing using the Benjamini-Hochberg method as implemented in EdgeR [39,40]. To compare gene expression differences between treatments, a generalized linear model which included dietary treatment and genetic group was fitted to the data using the glmFit function [39]. As EdgeR cannot estimate multiple levels of variation, the 12MO and PREG sample sets were divided into subsets and analyzed for treatment differences in separate runs, to avoid bias caused by unaccounted within-individual variation (due to the paired nature of the total data set). The comparisons were performed as pairwise comparisons between contrasting treatments, i.e. LPHE-LPLE, HPHE-HPLE, LPHE-HPHE and LPLE-HPLE. Significantly differentially expressed genes (DEG) were called at a FDR (false discovery rate adjusted p-value as given in EdgeR) of 0.05. A full description of the DEG representing treatment differences at 12 months of age (12MO samples), is presented in a separate paper by our group [34]. To correct for any effect of individual animals, the difference in gene expression between 12MO and PREG was analyzed as a pairwise comparison within each treatment group. Subsequently, the resulting DEG lists were compared to find the DEG common to all treatments. Gene expression changes from 12MO to PREG which were common to all treatments was assumed to have occurred independently of previous dietary treatment, and to rather be caused by pregnancy or increased age. The lists of differentially expressed genes from each comparison were subsequently analyzed through the use of IPA (QIAGEN Inc., https://www.qiagenbioinformatics. com/products/ingenuity-pathway-analysis/ [41], to find over-represented biological pathways and upstream regulators represented by the DEG. Significance level for the pathway analyses was set at p < 0.05. Finally, we run an IPA comparison analysis of the 12MO-PREG DEG lists within each treatment was to compare the gene expression changes between 12MO and PREG and the upstream regulators controlling them for each dietary treatment. Pathways or functions specifically pertaining to irrelevant diseases, species or tissues were omitted. The RNAseq raw data are deposited and accessible through NCBI's Gene Expression Omnibus with the GEO Series accession number GSE79347. Animal performance Body weight and BCS at 12 months of age, conception, 7 months of pregnancy and calving are shown in Table 2. Both HE and LE heifers followed their planned growth trajectories throughout the experimental feeding period, with an average daily gain (ADG) of 930 and 660 g/d, respectively. Body weight at conception was close to 400 kg for all treatment groups, as shown in Table 2. The study design, aiming for a similar body weight at conception for all heifers, led to the slower-growing LE heifers being about 4 months older than the HE heifers at time of conception ( Table 2). The response to the pregnancy diet differed between treatment groups: LE heifers displayed a faster growth than expected with an ADG of 600 g/d, while the HE heifers only achieved an ADG of 470 g/d during pregnancy. This is thought to be caused by a late compensatory growth effect occurring in the LE heifers, which were restrictively fed during the experimental feeding period, while all heifers were fed ad lib during pregnancy. This led to the HE heifers calving in at a BW of approximately 535 kg at 22 months of age, while the LE heifers calved in at 26 months at a BW of approximately 565 kg. The different pregnancy growth rates were not planned, and thus represents a complicating factor which must be taken into account in the interpretation of our results. Read alignment and overall results Sequencing yielded on average 57.2 million raw reads per sample. Alignment gave an average overall read mapping rate of 95.1% and a concordant pair alignment rate of 89.6%. Genes with a minimum CPM (counts per million) of two in at least six (n = 6) samples were considered to be expressed and were included in the differential expression analysis. Out of 26,740 gene transcripts in the UMD3.1 bovine genome assembly, 13,396 gene transcripts fulfilled this criterion. The DEG lists resulting from the examined contrasts varied greatly in size. The number of DEG found for each contrast is presented in Table 3. At PREG, the LPHE-LPLE and HPHE-HPLE comparisons resulted in 79 and 57 DEG, respectively. The main affected pathways for these comparisons are shown in Figs 1 and 2. The HPLE-LPLE comparison yielded only five DEG, while we found no DEG when comparing HPHE and LPHE. All post-treatment comparison DEG are listed in S1 Table, and complete lists of significantly affected pathways for the LPHE-LPLE and HPHE-HPLE comparisons are given in S2 Table. The HP-LP comparisons yielded no significantly affected pathways. Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers The HE treated heifers underwent a larger change in ration characteristics after conception than heifers on LE treatments and this seems to be reflected in their degree of alteration in gene expression between 12MO and PREG, as these groups displayed more than 7 times as many DEG as any of the LE groups. The 12MO-to-PREG change in gene expression was massive for both HE treatments; The LPHE heifers displayed 3987 DEG, and the HPHE heifers displayed 3997 DEG with an FDR < 0.05. The 12MO-to-PREG gene expression change for the two LE treatments was less pronounced but still considerable, with 534 and 334 DEG (FDR < 0.05) for the LPLE and HPLE groups, respectively. The within-treatment 12MO-to-PREG gene expression changes were larger than the differences found between the dietary treatment groups during 12MO [34], and much larger than any differences found between treatments at PREG. The full lists of DEG between 12MO and PREG are shown in S3-S6 Tables for the LPHE, HPHE, LPLE and HPLE treatments, respectively. As illustrated in Fig 3, the 12MO-to-PREG DEG for each group displayed a complex pattern of overlapping: The two HE groups had the greatest change in total gene expression. Out of the 5184 DEG found for these two groups in total, 66% of the DEG were common to both groups. The two LE groups displayed a much smaller overall gene expression change, but comparable fractions of shared and unique DEG. IPA pathway analysis yielded 506, 493, 314 and 215 significantly affected pathways for LPHE, HPHE, LPLE and HPLE heifers from 12MO to PREG, respectively. Complete lists of significantly affected pathways for each treatment group are presented in S7-S10 Tables. The main over-represented pathways for each treatment group from 12MO to PREG are shown in Figs 4-7. Post-treatment differences between dietary treatments Although offered the same feed ad libitum during pregnancy and entering pregnancy at a similar BW, the HE and LE heifers were not the same age (they were inseminated at about 13 and 17 months of age), they did not respond to the pregnancy diet in the same way (LE heifers grew faster than HE heifers) and consequently seemed to be in two different physiological states during pregnancy. The LE-treated heifers maintained the same growth in response to the new ad libitum pregnancy feeding compared to the restricted rations they were offered earlier. As BCS also remained constant for the LE heifers through pregnancy, their growth may be assumed to consist of proportional amounts of lean and fat tissue ( Table 2), but our study does Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers not allow for precise assessment of lean and fat tissue distribution. Compared to the experimental feeding period, the HE-treated heifers decreased their growth rate and BCS during pregnancy. The LPHE-LPLE and HPHE-HPLE comparisons resulted in 79 and 57 DEG, respectively. Cholesterol biosynthesis, thyroid receptor complex (TR/RXR) activation, liver X receptor (LXR/RXR) activation and sirtuin signaling pathways were among the pathways which were significantly affected in both comparisons, with the majority of constituent genes displaying decreased expression in the HE-compared to the LE-heifers. Sirtuins are a class of proteins involved in a wide range of cellular processes, including ageing, transcription, apoptosis, inflammation, energy efficiency and adaptation to low calorie intake [42][43][44]. The TR/RXR and LXR/RXR activation pathways are involved in the regulation of both lipid, carbohydrate Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers and steroid metabolism [45,46], and the latter (cholesterol metabolism) was significantly affected and downregulated in HE heifers in both comparisons. As the HE heifers decreased their BCS through pregnancy, a decreased need for cholesterol in adipose cell membranes and downregulation of its production is logical. The TR/RXR and LXR/RXR activation and sirtuin signaling pathways were represented mainly by genes which are also highly involved in energy metabolism pathways, such as ACACA, FASN, ACLY, ACSS2, G6PD and SCD, and no genes coding directly for sirtuins, thyroid or liver X receptors or RXR complexes were DE. Therefore, we cannot ascertain whether the downregulation of these genes in HE heifers is mainly associated with cell signaling or with metabolic pathways. Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers A gene coding for an uncoupling protein, UCP2, is also associated with TR/RXR, LXR/RXR and Sirtuin signaling pathways and was upregulated in HE heifers in both comparisons. Uncoupling proteins are embedded in the inner mitochondrial wall of many cell types, but has been most extensively described in brown adipose tissue, where UCP1 is a prerequisite for the essential heat production function of this tissue [47]. The uncoupling proteins function as The diagram shows the number of differentially expressed genes unique to or shared between experimental treatment groups from experimental feeding at 12 months of age, to month 7 of gestation (at which point all heifers had received the same diet for 6 months). LPHE = Low-protein, High-energy; HPHE = High-protein, High-energy; LPLE = Low-protein, Low-energy and HPLE = High-protein, Low-energy (HPLE). Numbers in parentheses are the total number of differentially expressed genes found for each treatment. (Venn diagram modified from VennDiagram and Venneuler R packages). https://doi.org/10.1371/journal.pone.0218723.g003 Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers Main affected pathways for LPHE heifers during development from experimental feeding at 12 months of age proton carriers mediating a regulated proton leakage through the inner mitochondrial membrane, free-coupled from the reactions of the respiratory chain [48]. Free-coupling proton flow from ATP production in the respiratory chain causes a futile cycle where energy spent on pumping protons against an electrical gradient is lost as the protons flow back out along the gradient without energy-rich molecules being produced. However, the roles of the other UCPs (UCP2-5) seem to be more complex than the well-known thermogenetic function of UCP1 [49]. UCP2 has been found to act as a free radical scavenger, and it also plays a role in regulating insulin [49]. The UCP2 gene is almost ubiquitously expressed throughout body tissues [49]. It is regulated by TR and PPARγ and was recently shown to be regulated by diet in subcutaneous white adipose tissue of cattle [50]. However, our heifers had been offered the same diet for 6 mo before PREG biopsises were taken. Thus, the observed upregulation of UCP2 in HE heifers is not a direct, contemporary effect of diet, and the cause of this post-treatment effect is unclear. It may represent a lasting change in the regulation of insulin secretion and glucose metabolism or an increased need for free radical scavenging because of a higher lipid turnover (BCS decrease) in the adipose tissue of pregnant HE heifers, but further studies are required to confirm or reject these hypotheses. The difference in gene expression between HPLE and LPLE heifers, although small, increased post-treatment; from zero at 12MO to five at PREG. This is surprising but is in agreement with the results of Van Amburgh et al, who found that post-treatment effects of prepubertal feeding regimes differing in energy and protein content had a larger impact on subsequent milk yield than the experimental feeding itself [32]. Three DEG maintained their direction of differential expression between treatment groups from 12MO to pregnancy, despite the feed change, the age difference and the different growth profiles obtained during pregnancy. Two of these genes were found to be DEG in the LPHE-L-PLE comparison. The first gene was RNS1, coding for pancreatic ribonuclease. This was upregulated in LPHE relative to LPLE heifers at 12MO as well as PREG with an FDR<0.02, and fold changes of 2.5 and 5.7 at 12MO and PREG, respectively. The other gene found for the LPHE-LPLE contrasts was ENSBTAG00000007075, coding for a major histocompatibility complex class I (A-like precursor). This gene was also upregulated in LPHE heifers relative to LPLE heifers at both time points with an FDR<0.02. Log 2 fold change was 3.4 at 12MO and 12.0 at PREG. The third gene displaying a persisting treatment effect at PREG was ENSBTAG00000048049, a novel gene with immunoglobulin-like sequences, with several orthologues in other species found to be an IGHM gene [51]. This was downregulated for HPHE with a log 2 fold change of -2.9 and -6.1 in the HPHE-HPLE contrast at 12MO and PREG, respectively, and with an FDR <0.02 at both time points. As bovine pancreatic ribonuclease represents a very basic and general cellular function, it is hard to deduce exactly what function a persisting expression difference represents in the adipose tissue of heifers in our study. Furthermore, this gene was also one of the genes commonly upregulated across all treatment groups from 12MO to PREG. This indicates that several mechanisms may have influenced the expression of this gene, and the apparently persistent treatment effect may in fact be the result of different underlying causes. The two immunologically related genes on the other hand displayed different, treatment-dependent changes from 12MO to PREG as well as the month 7 of gestation (at which point all heifers had received the same diet for 6 months). Red bars indicate percent upregulated, and green bars indicate percent downregulated genes pertaining to the pathway during PREG versus 12MO. Number to the right of bars display total number of genes pertaining to each pathway. Orange squares indicate the negative logarithm of p-value of observation (-log p-value = 1.3 equals p-value = 0.05). (Reprinted from IPA under a CC BY license, with permission from Qiagen Bioinformatics, original copyright 2018). https://doi.org/10.1371/journal.pone.0218723.g004 Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers persistent DE between treatment groups at these time points. Together with the large expression differences (log 2 fold change), this may indicate that the experimental diets have in fact elicited a lasting difference in the expression of ENSBTAG00000007075 and ENSBTAG00000048049. This may suggest an ongoing potential for some type of epigenetic programming acting through an interaction between nutrition and the immune system after three months of age, which is when the experimental treatments were initiated. To our knowledge, this has not been shown before, but further studies are required to either confirm or reject this hypothesis. Within-treatment gene expression changes from 12MO to PREG LPHE and HPHE treatments. For the LPHE and HPHE heifers, the gene expression changes occurring between 12MO and PREG were very similar. Apart from slight differences in the ranking of the significantly affected pathways and number of DEG, the main pathway results were nearly identical. Therefore, gene expression changes for these two treatments are presented and discussed commonly in this section, under the term HE treatments (Any specific changes pertaining only to one group are specified). Gene expression changes from 12MO to PREG for the HE treatments were dominated by a downregulation of pathways pertaining to energy metabolism and mitochondrial functions, namely oxidative phosphorylation, mitochondrial dysfunction, cholesterol biosynthesis and the TCA cycle. Fifteen genes representing all eight steps in the TCA cycle and 85 genes associated with either of the five complexes in the respiratory chain were significantly downregulated. This finding is in concordance with our observations from the LPHE-LPLE contrast at 12MO: At this point, oxidative phosphorylation, mitochondrial dysfunction and the TCA cycle was clearly upregulated in the high-energy fed LPHE heifers compared to the restrictively fed LPLE heifers [34]. The transition to the pregnancy diet implied a reduction in energy intake for the HE heifers. This possibly explains a downregulation of pathways connected to the production of energy-rich molecules available for lipid synthesis. A small proportion of the DEG associated with mitochondrial dysfunction were upregulated instead of downregulated. None of these genes were directly associated with energyyielding reactions, and therefore do not change the overall massive downregulation of genes associated with energy production in the adipose of HE heifers. However, among the few upregulated DEG pertaining to mitochondrial dysfunction were NCSTN, APHA1, BACE1, APP, MAOB and CASP9. These genes are all part of a mitochondrial pathway leading to production of amyloid-beta from the amyloid precursor protein (APP) and apoptosis [52]. Such changes are associated with oxidative stress and lipid peroxidation [53], but studies on this pathway has so far been performed mainly with a focus on human diseases and little is known about its function in the adipose tissue of ruminants. However, apoptosis induced by caspases have also been associated with programmed cell death in lipoatrophy [54]. These DEG may therefore be an indication of apoptosis associated with the decreasing fatness of the HE heifers from 12MO to PREG (Table 2). Cholesterol biosynthesis was a major affected pathway for the HE treatments. 17 out of 28 genes pertaining to this pathway were downregulated at PREG in HE treatments. Cholesterol to month 7 of gestation (at which point all heifers had received the same diet for 6 months). Red bars indicate percent upregulated, and green bars indicate percent downregulated genes pertaining to the pathway during PREG versus 12MO. Number to the right of bars display total number of genes pertaining to each pathway. Orange squares indicate the negative logarithm of p-value of observation (-log p-value = 1.3 equals p-value = 0.05). (Reprinted from IPA under a CC BY license, with permission from Qiagen Bioinformatics, original copyright 2018). https://doi.org/10.1371/journal.pone.0218723.g005 Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers is synthesized in adipose tissue, where it is stored as a free sterol in the phospholipid membrane of lipid droplets [55]. Thus, it is logical to find a downregulation of cholesterol synthesizing pathways in adipose tissue when dietary energy is reduced. Besides metabolic pathways, the list of highly affected pathways for the HE treatments was dominated by signaling pathways, and the one most significantly affected for the LPHE heifers was axonal guidance signaling (144 DEG). PI3K signaling in B lymphocytes (55 DEG) and leukocyte extravasation signaling (71 DEG) was highly ranked for both HE treatments. The axonal guidance pathway was represented by 121 upregulated and 23 downregulated DEG for the LPHE heifers. Axonal guidance refers to the directional guidance of axons towards their target tissues or cells during tissue development or restructuring. White adipose tissue is sympathetically innervated, and the sympathetic system has been shown to regulate lipolysis [56]. Axonal guidance is signalled via a large range of molecules, collectively named "guidance cues" which includes semaphorins, netrins, ephrins, SLIT ligands, morphogens (Wnt), growth factors, EPH receptors, ROBO receptors and cytokines [57,58]. Genes coding for all these types of molecules were upregulated in HE heifers at PREG. Even though there are no neuronal cell bodies present in adipose tissue, the mRNA reads for these gene products in our data may originate from neurons, as transport to and local translation of neuronal mRNA in axons, far away from the nuclei in which they were transcribed, constitutes an important regulation mechanism in axon guidance [58]. As the axonal guidance pathway was significantly upregulated in LPHE heifers, in contrast to all other treatment groups, we assume it to be affected by their dietary treatment. If so, it is reasonable to assume that axonal guidance occurred to a greater degree in this group because these heifers experienced the greatest alteration in diet and the largest decrease in BCS causing these heifers to have the greatest degree of adipose tissue restructuring. Phosphatidylinositol 3-kinases (PI3K) exert their effects on membrane signaling complexes within many pathways and systems [59,60]. In B lymphocytes, PI3K regulates both activation, proliferation and differentiation [59]. More recently, PI3K signaling has also been shown to play an important role in energy balance regulation: Experimentally induced reduction of PI3K signaling increases energy expenditure and protects from obesity and obesity-related disorders [60,61]. Thus, the upregulation of PI3K signaling in adipose tissue of HE heifers after transition to low-energy feeding may be part of an energy-saving strategy. If so, manipulation of PI3K signalling may provide a means for increased feed efficiency in cattle, the opposite goal of ongoing human research, which focuses on alleviating obesity by increasing energy expenditure through PI3K functions. Leukocyte extravasation and immune cell infiltration of adipose tissue has been the subject of intense research over the last decades, as obesity, metabolic syndrome and adipose tissue dysfunction in humans is associated with low-grade chronic inflammation and increased immune cell trafficking of adipose tissue [62]. In contrast, our results show an upregulation of leukocyte extravasation signals in animals subjected to a reduction in feed energy, whilst displaying no increase in adiposity. This is in line with recent research on both cattle and other species, stating that adipocyte inflammation is not only associated with adipocyte-related diseases, but is also an essential part of physiological tissue remodeling, especially during lipolysis [63][64][65][66][67]. The type, degree, and duration of leukocyte migration and infiltration is month 7 of gestation (at which point all heifers had received the same diet for 6 months). Red bars indicate percent upregulated, and green bars indicate percent downregulated genes pertaining to the pathway during PREG versus 12MO. Number to the right of bars display total number of genes pertaining to each pathway. Orange squares indicate the negative logarithm of p-value of observation (-log p-value = 1.3 equals p-value = 0.05). (Reprinted from IPA under a CC BY license, with permission from Qiagen Bioinformatics, original copyright 2018). https://doi.org/10.1371/journal.pone.0218723.g006 Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers Fig 7. Main affected pathways from experimental feeding period (12MO) to pregnancy (PREG) for High-protein, Low-energy fed heifers (HPLE). Main affected pathways for HPLE heifers during development from experimental feeding at 12 months of age, interdependent on the type of physiological or pathological change occurring in the adipose tissue [62,63]. Current knowledge indicate that in dairy cattle, differences in immune cell infiltration in adipose tissue seem to be affected by stage of lactation [65][66][67], body condition score [68,69] and the presence, degree and underlying cause (dietary, physiological or pathological) of a positive or negative energy balance [64,66,67,69,70]. LPLE. The LPLE treatment expression changes from 12MO to PREG clearly differed from the other three dietary treatments: The LPLE experimental diet was the one most similar to the pregnancy diet, and this seems to be reflected in the gene expression of the LPLE heifers through the lack of affected metabolic pathways. The top list of affected pathways due to/or during diet change for this treatment group almost exclusively consisted of pathways associated with immunological and inflammatory functions (Fig 6). It is also worth mentioning that out of the eleven most affected pathways for this group, six were specifically associated with T lymphocytes (CD28 signaling in T helper cells, NFAT regulation of the immune response, PKCθ signaling in T-lymphocytes, iCOS-iCOSL signaling in T helper cells, Th1 and Th2 activation and Th2 pathway). The type of immune cells involved in adipose tissue infiltration is dependent on the current stage and physiology / pathology of adipose tissue [62]. T-cells infiltrating the adipose tissue has been described as a primary event in the process of macrophage recruitment, adipose inflammation and insulin resistance [71,72]. However, the LPLE heifers were never obese nor clinically ill during our trial. The T-cell marker CD3 is also found in both omental and subcutaneous adipose tissue of dairy cows. However, the occurrence was higher in the omental adipose tissue of dairy cows suffering from abomasal displacement. Together with our findings, this may indicate that T-cell infiltration in bovine adipose tissue is both a part of normal, physiological processes such as adipose tissue remodeling, as well as pathological inflammatory conditions [64,72]. HPLE. For the HPLE heifers, the diet change from their experimental feed treatment to the pregnancy feed meant a transition to a ration of lower protein density. From 12MO to PREG, we observed a downregulation of several pathways associated with amino acid degradation in the HPLE heifers; As shown in Fig 7, 11 of the 20 most highly affected pathways were implicated in the degradation of valine (Val), leucine (Leu), methionine (Met), alanine (Ala), phenylalanine (Phe), glycine (Gly) and glutamate (Glu). Except for Met and Gly, showing a mixed pattern of up-and downregulated DEG, all DEG pertaining to these pathways were downregulated at PREG. No other treatments displayed a similar downregulation of amino acid degrading pathways at PREG, and the reason for HPLE heifers to do so remains unknown. However, we may propose several hypotheses: The capacity of adipose tissue to both synthesize and degrade AA in reaction to concentrations of insulin, glucose or specific AA as part of whole-body metabolism and regulation has been known for decades [73][74][75][76]. The pregnancy diet and the experimental diet given to the HPLE heifers differed both in protein content and protein sources (because of the inclusion of concentrate in the experimental diet). Thus, both the total and relative amounts of different AA available to the animals could be expected to differ between the two diets, and the allocation and metabolic fate of specific AA likewise. However, most of the AA represented by significantly affected degradation pathways also have important functions other than acting merely as protein precursors and energy substrates: Methionine also acts as a methyl donor for transmethylation reactions in lipid to month 7 of gestation (at which point all heifers had received the same diet for 6 months). Red bars indicate percent upregulated, and green bars indicate percent downregulated genes pertaining to the pathway during PREG versus 12MO. Number to the right of bars display total number of genes pertaining to each pathway. Orange squares indicate the negative logarithm of p-value of observation (-log p-value = 1.3 equals p-value = 0.05). (Reprinted from IPA under a CC BY license, with permission from Qiagen Bioinformatics, original copyright 2018). https://doi.org/10.1371/journal.pone.0218723.g007 Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers biosynthesis and this function may also be the cause of Met degradation in the adipose tissue of HPLE heifers [77]. Levels of circulating branched chain amino acids (BCAA; Val, Leu, Isoleu) are associated with metabolic regulation, obesity and insulin resistance [78,79], and are regulated by BCAA metabolism in adipose tissue [80]. The other AAs represented by downregulated degradation pathways (Phe, Ala, Gly and Glu) also possess vital functions in addition to being energy substrates and protein precursors [81], and the background for the downregulation of their degradation pathways at PREG is almost certainly more complex than a simple up-and downregulation because of their availability as protein precursors in peripheral tissues. However, as HPLE was the only group displaying this effect, it must be assumed to be dietary related. The HPLE heifers received a diet of similar energy content but more protein than the LPLE heifers at 12MO, and the net energy intake for HPLE was very similar to that of the LPLE heifers (Table 1). Therefore, one would expect the change in energy intake from 12MO to PREG for the two LE groups to be similar, and small. Interestingly, the HPLE heifers still displayed a pathway analysis result with a closer resemblance to the results for the HE treatments, although at a smaller scale, with a downregulation of mitochondrial dysfunction and oxidative phosphorylation indicating a decrease in energy flux towards adipose tissue. Together with the downregulation of amino acid degradation pathways this indicates that the effect of total ration characteristics during 12MO has a larger effect than dietary energy and protein in isolation. Changes from 12MO to PREG common to all groups. Among the genes found to be differentially expressed between 12MO and PREG samples for all four groups, 97 displayed a similar up-or down-regulation in all groups. These genes are listed in S11 Table. Despite different feed treatments until confirmed pregnancy, different growth rates both before and during pregnancy, and different age at PREG, all heifers also had some factors in common over this period; all heifers became older, they all grew (though at different rates), they all became pregnant and at PREG they had all been offered the same feed ad lib for about 6 months. We interpret the differential expression of these common genes to be treatment-independent, and that they were caused by pregnancy, general progression of age and/or growth or by characteristics of the pregnancy feed per se. Pathway analysis of the common DEG yielded 23 significantly over-represented pathways. The main pathways are shown in Fig 8, and among these were the complement system, leucotriene biosynthesis and eicosanoid signaling. The immune system is known to undergo several changes during pregnancy, in order to avoid rejection and expulsion of the fetus while maintaining an adequate immune competence to combat infections and other threats. [82][83][84][85]. A characteristic part of these changes seems to be an activation of the innate immune system, including the complement system and natural killer cells [83,84,86]. Adipsin (also known as Complement factor D, coded by CFD, upregulated in PREG heifers), has been subject to extra interest as it has been associated with preeclampsia and abortion [87][88][89]. Consequently it seems complement activation may be part of both physiological and pathological immune changes related to pregnancy. The upregulation of genes coding for arachidonate 5-lipoxygenase (ALOX5) and its activating protein (ALOX5AP) together with leukotriene C4 synthase (LTC4) indicates an increased leukotriene production, especially of LTC4, from arachidonic acid in adipose tissue during pregnancy. Expression changes of ALOX5 and ALOX5AP through gestation and during labour in human chorionic membranes have been shown, suggesting a role for these genes both at parturition as well as earlier in pregnancy [90]. Concentration of LTC4 has also been shown to increase in human amniotic fluid during labour [91]. An increased capacity for leukotriene synthesis during pregnancy has recently been detected in human blood [82], but so far, no studies have reported such effects in adipose tissue. As the heifers in our study underwent normal pregnancies we assume the observed changes in their gene expression to be physiological. However, further studies should be performed to establish whether any of these genes are potential biomarkers for gestational health or preservation. Upstream regulators. Despite the differences in DEG, growth rates, age and BCS development during the period from 12MO to PREG, the treatment groups displayed many similarities in their main upstream regulators as analyzed by IPA. A heatmap containing key regulators and their most highly affected DEG is shown in Fig 9. The comparison analysis demonstrated transforming growth factor-β 1 (TGFB1) to be the highest ranked upstream regulator of identified DEG overall. At PREG versus 12MO, the TGFB1 gene itself was significantly upregulated in the adipose tissue of the HE groups, but not in the LE groups. In the HE groups, TGFB1 was directly or indirectly associated with > 450 of the identified DEG between Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers 12MO and PREG. These results indicate a widespread involvement in several cell functions. In adipose tissue, TGFB1 is mainly secreted by cells other than adipocytes, and an increased secretion of TGFB1 has been associated with obesity in other species [92]. In our study TGFB1 was upregulated only in the HE heifers at PREG when comparing to 12MO. This may seem paradoxical as the HE heifers underwent a decrease in energy intake and BCS during the period, but still suggest an association between TGFB1 and adiposity. Gene polymorphisms and plasma levels of TGFB1 have also shown associaton with the risk of preeclampsia and recurrent abortions in humans [93]. As such, it would be interesting to investigate whether TGFB1 or any of its affected downstream molecules proves to be indicators of pregnancy preservation capabilities in cattle as it is in humans. Rapamycin-insensitive companion of mechanistic target of rapamycin (RICTOR), encoded by RICTOR, is part of the mTORC2 protein complex. mTORC2 have both functional and structural similarities to the more thoroughly studied mTORC1, which regulates cell growth based on nutritional status and growth factor-derived signals. mTORC2 has so far been linked more tightly to regulatory functions within the insulin / PI3K signaling pathway [94]. Within mTORC2, RICTOR is thought to have a role as a facilitator for substrate binding and correct subcellular location of mTORC2, and is therefore essential for the function of the mTORC2 complex [95] RICTOR seem to be highly involved in the changes occurring in the HE-treated heifers from 12MO to PREG, but much less so in the LE-treated heifers. An IPA core analysis run on the DEG associated with RICTOR in the HE heifers showed it to be mostly involved in downregulation of genes associated with mitochondrial energy production. These functions were downregulated in the HE heifers compared to the LE heifers probably due to the larger reduction in dietary energy for these groups. This may explain the difference in RICTOR involvement between the two energy treatments, and suggest RICTOR and mTORC2 to have an important role in the regulation of energy production in adipose tissue. Interferon Gamma (IFNG) is a cytokine regulating a wide range of immune and inflammatory functions. In our study, IFNG was most highly associated with function of T cells, the complement system, cell differentiation and maturation and interleukin signaling across treatments. It was least active in HPLE heifers and most in HPHE and LPLE. Compared to the other treatments, the overall impact of IFNG in the HPLE heifers was also slightly skewed with involvement in citrulline metabolism highly ranked together with a more narrow range of immune functions, such as the complement system and phagosome formation. The reason for this is not known. Mitogen-activated protein kinase 4 (MAP4K4) displayed increased activity from 12MO to PREG for all treatments except LPLE. The activation also seemed lower for HPLE heifers than for the HE treatments. As for RICTOR, this seem to be explained by the DEG associated with MAP4K4 and to the changes occurring within each treatment group from 12MO to PREG: MAP4K4 was mostly associated with DEG pertaining to energy production, and these were downregulated in HE heifers and to a lesser degree in HPLE heifers (Figs 4, 5 and 7). The LPLE heifers, which underwent the smallest dietary change, did not display such changes ( Fig 6). Thus, the regulating function of MAP4K4 in our experimental heifers may have been mainly to decrease the production of energy rich compounds in adipose tissue when subjected to a decrease in dietary energy flux. Tumor necrosis factor (TNF) was also ranked high among the upstream regulators for all groups. Tumor necrosis factors are a family of multifunctional cytokines with known effects on apoptosis, inflammation, lipid metabolism and cellular redox balance among others [96]. In the context of our study, it is of interest that TNF is an activator of leukotriene production pathways, which were upregulated in all groups at PREG (Figs 4-7 and 9). TNF also acts upon TP53, and mediates some of its effects through this transcription factor [97], and TP53 was also ranked very high among upstream regulators for all treatment groups. TNF itself was not DE for any group, but all groups displayed between three and 19 DEG associated with activity regulation of TNF such as TNFAIP3, TNFAIP6, C1QTNF6, TNFSF18, LITAF, TNIP2, TRAF5 and TNFRSF1B. This is a good example of the versatility and flexibility of gene regulation: It may act directly upon the gene coding for a main regulator (as in this case with TGF1B), or it may act upon several other genes coding for regulators of the regulator (as here with TNF). Summary Gene expression differences between heifers fed different levels of energy and protein are evident even 6 months after the termination of different feeding regimes and transition to the same diet. Only a few of the gene differences observed while being fed different diets are still present post-treatment. These may indicate a possible immunologic or epigenetic programming, while most post-treatment differences seem to be reflections of different physiological responses to the new diet, rather than the dietary content itself. Still, diet also has a major direct impact on adipose gene expression, which is clear from the treatment-dependent and varied gene expression changes occurring within treatment groups from experimental feeding at 12 months of age until similar feeding at 7 months pregnant. Pregnancy is known to cause immunologic changes, and in the adipose tissue of our heifers, these changes seem to consist mainly of complement activation and increased leukotriene synthesis.Through our study, we have identified several candidate genes and key gene regulators which should be studied further with regards to pregnancy preservation (TGFB1, CFD) and metabolic regulation and efficiency (UCP2, PI3K, RICTOR, MAP4K4,) in dairy cattle. As this is a purely transcriptomic study, it represents a first step towards new knowledge about adipose tissue responses to diet and pregnancy in cattle, and warrants further studies in terms of functional assays, metabolomics and/ or proteomics for phenotypic confirmation. Supporting information S1 Table. Post-treatment differences between dietary treatments. All differentially expressed genes (DEG) from energy treatment comparisons at month 7 of pregnancy. (XLSX) S2 Table. All pathways affected by differently expressed genes between treatment comparisons at month seven of pregnancy. (XLSX) S3 Table. All DEG (FDR<0.05) between experimental feeding period (12MO) and month 7 of pregnancy (PREG) for heifers fed a high-protein, high-energy diet (HPHE) during the experimental feeding period. (XLSX) S7 Table. Over-represented pathways representing differentially expressed genes between experimental feeding period (12MO) and month 7 of pregnancy (PREG, 20 (HE) and 24 (LE) mo of age) for heifers on a low-protein, high-energy diet (LPHE) during the experimental feeding period. (XLSX) S8 Table. Over-represented pathways affected by transition between experimental feeding period (12MO) and month 7 of pregnancy (PREG, 20 (HE) and 24 (LE) mo of age, low-protein, low-energy diet) for heifers on a high-protein, high-energy diet (HPHE) during the experimental feeding period. (XLSX) S9 Table. Over-represented pathways affected by transition between experimental feeding period (12MO) and month 7 of pregnancy (PREG, 20 (HE) and 24 (LE) mo of age, low-protein, low-energy diet) for heifers on a low-protein, low-energy diet (LPLE) during the experimental feeding period. (XLSX) S10 Table. Over-represented pathways affected by transition between experimental feeding period (12MO) and month 7 of pregnancy (PREG, 20 (HE) and 24 (LE) mo of age, lowprotein, low-energy diet) for heifers on a high-protein, low-energy diet (HPLE) during the experimental feeding period. (XLSX) S11 Long-term effects of prior diets, dietary transition and pregnancy on adipose gene expression in dairy heifers	2019-07-05T13:14:59.565Z	2019-07-03T00:00:00.000	{ "year": 2019, "sha1": "620800b090b40b7c03ddbc7ffd6294249e2d8eaa", "oa_license": "CCBY", "oa_url": "https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0218723&type=printable", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "21b96d8517618a2495a1009dc125157082a41129", "s2fieldsofstudy": [ "Agricultural And Food Sciences" ], "extfieldsofstudy": [ "Medicine", "Biology" ] }
119204482	pes2o/s2orc	v3-fos-license	Role of dipole-dipole interactions in multiple quantum transitions in magnetic nanoparticles In order to better understand the origin of multiple quantum transitions observed in superparamagnetic nanoparticles, electron magnetic resonance (EMR) studies have been performed on iron oxide nanoparticles assembled inside the anodic alumina membrane. The positions of both the main resonance and"forbidden"(double-quantum, 2Q) transitions observed at the half-field demonstrate the characteristic angular dependence with the line shifts proportional to 3cos2q-1, where q is the angle between the channel axis and external magnetic field B. This result can be attributed to the interparticle dipole-dipole interactions within elongated aggregates inside the channels. The angular dependence of the 2Q intensity is found to be proportional to sin2qcos2q, that is consistent with the predictions of quantum-mechanical calculations with the account for the mixing of states by non-secular inter-particle dipole-dipole interactions. Good agreement is demonstrated between different kinds of measurements (magnetization curves, line shifts and 2Q intensity), evidencing applicability of the quantum approach to the magnetization dynamics of superparamagnetic objects. Introduction Atomic scale objects are described with the quantum approach while macroscopic systems are the common subject of the classical physics. Nanoscale magnetic objects can be used as a playground to study the transition between the quantum type of behavior and classical picture [1][2][3][4]. As was shown earlier [2,3], the electron magnetic resonance (EMR) in magnetic nanoparticles (NP) in certain conditions can be satisfactory described with the "quantatization" approach, considering a nanoparticle of a small size (5-10 nm) as a large exchange-coupled cluster with the spin S ~ 10 2 -10 3 of the ground spin multiplet. The EMR spectrum was ascribed to the quantum transitions between the energy levels associated with the projections, m, of this "giant spin" onto the direction of the magnetic field (with account made for the magnetic anisotropy), in a way similar to the common electron paramagnetic resonance consideration. Moreover, in a strong resemblance to multiple quantum transitions known for paramagnetic spins, the EMR signal in NPs demonstrates additional low-field signals at the fields B = B 0 /k (where B 0 is the field of the main resonance and k = 2, 3, 4… ) [3,5]. As it was discussed in [3], these features can be ascribed to "forbidden" transitions m =  2, 3, 4… which may be allowed due to the mixing of states caused by non-diagonal terms in the spin Hamiltonian. Note that recently the similar mixing of states was considered as an essential factor affecting magnetic relaxation in NP's. [4]. A "giant spin" approach was also applied to the description of nuclear spin relaxation stimulated by nanoparticles used as contrast agents in MRI treatment [6]. However, in spite of a number of evidences for the quantal effects in NPs, many questions still remain open. The modeling of the EMR line shape proposed in Ref. [2], though qualitatively consistent with the experiment, includes a number of phenomenological fitting parameters. Temperature dependence of the low-field signal intensity demonstrates a sharp drop upon cooling which is too steep in comparison with the theoretical predictions [3]. The exact mechanism of these "forbidden" lines is not clear yet. According to the quantatization model [3], it can be related to the effects of magnetic anisotropy or interparticle dipole-dipole interactions. Thus, further study is necessary to clarify the origin of the effects and search for the border between classical and quantal phenomena in fine magnetic objects. The idea of this work is to employ specially textured samples where randomly oriented nanoparticles are arranged in parallel chains, which are well aligned in one direction. Depending on the predominating mechanism, the model [3] predicts a specific angular dependence of the intensity of the double-quantum (2Q) transitions as compared with the anisotropy in the EMR spectrum measured on the aligned particles. The results obtained on such systems might clarify the origin of the effect as well as provide more independent data to confirm or reject the quantatization approach. Well-arranged chains of the particles can be fabricated using a porous anodic alumina membrane (PAA). PAAs are commonly used in fabrication and exploring parallel magnetic nanowires. In particular, detailed data on magnetic anisotropy of Ni nanowires were presented in Refs. [7][8][9][10][11][12][13], including the interplay between the easy axis and easy plane behavior depending on the pore diameter. Thorough static magnetic measurements on iron oxide NPs placed into the channels were reported in Ref. [14]. However, the anisotropic EMR spectra in the later system, not to mention the multiple quantum effects, were not studied yet until now. Experimental In the experiment we used two different series (A and B) of Fe 3 O 4 nanoparticles of nominally the same size but different magnetization. The chloroform suspensions of the particles with the diameter of 102.5 nm were purchased from Ocean Nanotech Inc. According to thermal gravitometric analysis, the solid content consists of 40 percent of iron oxide and 60 % of surfactant (oleic acid). Magnetization measurements were performed by SQUID magnetometer at room temperature in the field range 0 -40 kOe. Magnetization curves for the both samples are consistent with the Langevin law; however the saturation value of the magnetization, M, was different for the different series, which can be related to the different quality of the samples. For the sample A, M s A = 250 20 Oe which is lower but relatively close to the bulk value of 450 Oe. Such decreased values are typical for NPs and can be related to surface effects [15][16][17]. Magnetization of the sample B, M s B = 60  6 Oe, was significantly lower, which can be ascribed to the poorer quality of the magnetic content, presence of multiple defects or different phases. To prepare the textured samples, a porous alumina membrane with the dimensions 1cm x 1cm x 51µm, pore diameter of 35 nm and porosity of 15% purchased from Synkera Technologies was used as a template. Magnetic particles were embedded into the porous space of membrane by the standard technique of nanofiltration. As illustrated in Fig.1 (see also the data presented in Ref. [14]), the nanoparticles entered channels, forming parallel stacks. The experiments were performed using a standard Bruker EMX spectrometer, the microwave frequency was about 10 GHz (X band). The spectra were recorded using the standard field modulation technique, with the modulation amplitude 1 G and frequency of 100 kHz, resulting in the derivative of the EMR absorption. Angular dependences of EMR spectra were studied at room temperature. The orientation of the external magnetic field B with respect to the sample was adjusted through rotation of the sample in the horizontal plane, keeping the microwave field B 1 perpendicular to the external field B. A commercial gas flow cryostat was used in variable temperature experiments. The EMR technique was also used for independent estimation of the NPs magnetization. The EMR intensity (the double integrated signal) was compared with that of paramagnetic reference (CuSO 4 5H 2 O). Within the experimental error (10%), the results were consistent with the static measurements, indicating that nearly all magnetic material contributes to the EMR spectrum. Results The EMR spectra in the samples with the random orientation of NPs are presented in Fig.2 (for the convenience of comparison, the signals are normalized to the same integral intensity). The main EMR signal with the g-factor, g  2, consists of a broad line superimposed with a narrow one. Such a shape is typical for NPs as well as superparamagnetic and exchange-coupled clusters embedded in a non-magnetic matrix (see, for example, Refs. [2,5,[18][19][20][21][22]). In Ref. [2], the spectrum was interpreted in the framework of the quantatization model. Additionally, a weak low-field signal is seen at g  4; this line was previously reported and assigned to double quantum (2Q) transitions in Refs. [3,5]. To distinguish this half-field resonance from a well-known EPR feature with g = 4.3 expected for paramagnetic Fe 3+ ions [23], the temperature dependence of the low-field spectrum is shown in Fig. 3. A rapid decrease of the low-field signal upon cooling unambiguously testifies against paramagnetic origin and in favor of 2Q identification (see Refs. [3,5] for more details). As is expected, the main EMR signal is broader in the sample with the higher concentration of NPs. In this sample, the sharp central feature is not well-resolved due to the broadening effect of inter-particle dipole-dipole interactions. The broadening of the 2Q signal is observed as well. Let us now discuss the results obtained in the textured samples. The Fig. 4 demonstrates the EMR signals observed in the Sample A at various angles  between the external magnetic field B and the direction of the channels. The main EMR spectrum is broad, similar to that observed in the concentrated nonaligned sample (Fig. 2), pointing to a relatively close packing of the NPs in the matrix. The position of the main resonance strongly depends on the sample orientation, varying from lower fields at  = 0 (when the membrane channels are parallel to the external magnetic field) to higher fields at the perpendicular orientation. An additional weak and broad signal is observed as well with an opposite angular behavior (seen in Fig. 4 should also be noted that the 2Q intensity is considerably higher in the Sample A as compared to B, and in the textured samples as compared to the non-textured solid suspensions. The detailed analysis is presented in the next Section. Note that the sample A with the relatively high magnetization demonstrates much greater anisotropy and 2Q effects than the sample B. Likely, the latter has a much broader distribution of the particle magnetic moments, that can lead to higher experimental errors and uncertainty in the interpretation. Therefore for the quantitative analysis we will use the data obtained in the sample A, whereas the sample B is used for a qualitative comparison. Discussion Let us first discuss the anisotropic shifts of the main EMR spectrum, see Figs. 4, 5 and Eq.(1a). These data unambiguously show that the particles inside the membrane channels are strongly aligned along the channel direction. Note that similar behavior was observed previously on liquid suspensions of NPs frozen in high magnetic fields [24,2]; the effect was attributed to aligned aggregates (chains) formed by joint action of the external field and inter-particle dipolar interaction [2]. Since a layer of surfactant prevent the exchange interaction between adjacent particles, only the inter-particle dipole-dipole interaction should be taken into account. Note that our consideration does not require a compact filling of the channels with the particles. Instead, separate elongated aggregates are expected, randomly distributed along the channels. The magnetic interaction between the aggregates may contribute to the observed broadening of the EMR spectrum. As the first approximation, we consider an aggregate as a strongly elongated ellipsoid filled by randomly distributed magnetic entities (spins) with an average magnetization M av . In this case, the shift (first moment) of the EMR line reads [25]: The factor 2/3 accounts for the static (z-z) part of the dipolar interaction only since the contributions from resonant (flip-flop) inter-particle interactions are suppressed by the inhomogeneity of the EMR line due to a random distribution of individual anisotropy axes. Considering the EMR line width, the single-particle anisotropy field (due to both the shape and crystalline sources) does not exceed few hundred Gauss, in agreement with the literature data [14]. It is worth noting that the factor 2/3 in Eq.(2) increases to unity for a ferromagnet with an easy axis directed along the channel; in such a case, in the terms of classical magnetism, Eq.(2) describes the effect of demagnetizing field. This approach was successfully used in studying magnetic nanowires in membranes [7][8][9][10][11][12][13]. Further, we suppose Here c is the concentration of pure magnetite in the aggregates and L( 0 ) is the Langevin function with  0 =M s VB 0 /k B T, where V is the particle volume, T is the temperature, and k B is the Boltzmann constant. In our experiment, M s A =250 Oe, B 0 =3.4 kG, and T=295 K, which yield L( 0 ) = 0.9. Then, substituting experimental b A = 530 G for B el in Eq. (2) one gets c = 0.56. This value looks quite reasonable and suggests that nearly spherical NPs are packed closely enough in the elongated stacks inside the channels. Let us analyze the data using another approximation as well. Consider a linear chain which is parallel to the channel axis and consists of spherical NPs with the diameter d and magnetic moment =M s V. Using the fact that the field outside of a uniform magnetic sphere is equivalent to the field of a point magnetic moment  placed at the center of the sphere, the total z-component of the local field seen by every particle in a long linear chain is equal to: B ch () = 4.8MV P 2 ()/d 3 = 0.8M s L( 0 )P 2 () (4) where the distance between adjacent particles is taken equal to d, V=d 3 /6, and the numerical factor is caused by summation over the chain. Comparing Eqs. Note that the columns formed by the NPs are likely not strictly parallel to each other simply because of considerable difference between the particle size and pore diameter. Instead, a distribution of the chain directions around the channel axis is expected resulting in some asymmetry in the observed line shape. This effect is indeed observable: as seen in Fig.4, the signals taken at the extremal orientations are asymmetric with their centers-of-gravity being shifted with respect to the points of zero derivative toward higher fields at ~0 and lower fields at ~/2. In this case the center-of-gravity corresponds to the position averaged over the distribution of the anisotropy directions, whereas the absorption peaks (zero derivatives) are situated closer to those in strict parallel or perpendicular orientations. Simple estimations (not presented for brevity) show that the observed asymmetry can be explained with the Gaussian distribution of the chain directions with the normal deviation of 15-20 deg. This estimate will be used further. Now, let us discuss the low-field (2Q) signals. In the frames of the quantatization model [3], the 2Q transitions are explained by the admixing of the quantum states with m, m1. The two major possible mechanisms considered in [3] are: a) the magnetic anisotropy of an individual particle, which allows quantum transitions with m=2, or b) interactions between nanoparticles, when two particles coupled together through the dipole-dipole interaction absorb one photon of the microwave radiation. For the both mechanisms, the position of 2Q signal is expected to demonstrate the angular behavior similar to the behavior of the main resonance and be described with same Eqs.(2)-(4) but corrected for the reduced value of the magnetic field B. The decreased magnitude of the anisotropic shift of the half-field signals (Fig. 5) can be attributed to the lower M  L() in the low field range. An additional apparent decrease in the effective anisotropy of the 2Q signal can also be related to the specific angular dependence of the 2Q-line intensity, which has minima in the both extremal orientations, =0 and /2 (see below). Let us consider the angular dependence of the relative intensity of the half-field resonance, Fig.7, and compare the data with the predictions of the quantum model. According to the theory, the intensity of the "forbidden" 2Q transitions to be observed in the field B=B 0 /2 reads [3]:     Here B and  correspond to the conditions of the half-field resonance. These formulas were derived in the first approximation of perturbation theory as applied to the spin Hamiltonian of the ground spin multiplet of a particle considered as a giant exchange coupled cluster. Eq.(5) corresponds to the case when the mixing of the states (which allows the 2Q transitions) is due to the single-particle magnetic anisotropy; as mentioned above, the individual anisotropy axes are directed randomly and cannot be responsible for the angular dependence. Eq.(6) accounts for the dipole-dipole interaction between particles i and j, where r ij and  ij are the polar coordinates of the radius-vector connecting the particles. In the closely packed columns elongated along the channel direction, one can expect  ij   and r ij  d. The observed angular dependence of the 2Q intensity (Fig. 7) well resembles the sin 2 cos 2  function which enters Eq. (6). Certainly, this finding qualitatively confirms the model and clearly indicates the major role of the interparticle mechanism. However, there are two "anomalous" features to be explained. First, the data in Fig.7 are lifted above the abscissa axis, pointing to a noticeable isotropic contribution. The second peculiarity is an asymmetry of the observed angular dependence: the minimum at =0 is not as deep as that at =90 deg. This anomaly can be explained by taking into account a spread of the directions of NP stacks in respect to the channel axis. Such a possibility was already discussed in connection with the asymmetry of line shape seen in Fig. 4. Taking into account that the channel diameter (35 nm) is considerably larger than the mean particle size (~10 nm), this suggestion looks quite reasonable. Taking into account the distribution for the angles  between the channel and chain axes, one gets the corrected angular dependence in the form: where the brackets denote averaging over the distribution and  stands for the isotropic contribution. Using the geometrical relation between  and  and supposing the Gaussian distribution g() with a normal deviation  0 , we fitted the experimental data with the solid curves shown in Fig. 7. The agreement can be considered as satisfactory. In particular, the difference between the minima at perpendicular and parallel orientations is well reproduced by the fitting curves. The parameters found from the fitting are as follows: The fitting errors are about 20% for the sample A and considerably worse (about 50%) for B. In the frames of this accuracy, the results look quite reasonable. A larger value of the factor  A as compared with  B can be readily understood as a consequence of the difference in corresponding magnetizations which determine the dipolar strength. Further, the isotropic terms  A,B are close to the intensities obtained experimentally on the non-aligned samples, see the dashed lines in Fig. 7. Finally, the spread angles  0 A,B practically coincide with that obtained above from the analysis of the main spectrum. Table I. One can see that within the experimental errors, the values of h d determined from independent measurements are consistent, confirming validity of the model. In conclusion, elongated stacks of the magnetite nanoparticles have been fabricated by means of filtering the NP suspension through the porous alumina membrane. The positions of both the main EMR and weak half-field lines caused by "forbidden" 2Q transitions reveal the anisotropy typical of z-z dipolar interaction within the aligned columns, whereas the relative intensity of the 2Q signal demonstrates the sin 2 cos 2  dependence predicted by the quantatization model [3].	2019-04-12T20:13:09.892Z	2009-11-09T00:00:00.000	{ "year": 2009, "sha1": "ef4a3b90b530e17878304c410c80d89caad5126a", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "dd197c031dc95fc3c663e622bd6e1dba7c25e7f5", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Physics" ] }
1786153	pes2o/s2orc	v3-fos-license	Polarization characteristics of a reflective erbium doped fiber amplifier with temperature changes at the Faraday rotator mirror The temperature dependence of a Faraday rotator mirror (FRM) used in a reflective erbium doped fiber amplifier(R-EDFA) is reported in this paper. The influence of this dependence on the polarization state (PS) of amplified optical signals is also discussed. ©2005 Optical Society of America OCIS codes: (060.2320) Fiber optics amplifiers and oscillators; (260.5430) Polarization; (160.3820) Magneto-optical materials. References and Links 1. Goving P. Agrawal, Fiber-Optic Communication Systems (Third Edition), Ed. (Wiley & Sons, New York, 2002). 2. J.M. López-Higuera, Handbook of Optical Fiber sensing Technology, Ed. (Wiley & Sons, Chinchester, 2002). 3. J.C. Dung, S. Chi, C.C Chen, “Characteristics of the erbium doped fiber amplifier with polarization mode dispersion compensation,” Opt. Commun., 222, 207-212 (2003). 4. S. Yamashita, K. Hotate, M. Ito, “Polarization properties of a reflective fiber amplifier employing a circulator a faraday rotator mirror,” J. Lightwave Technol., 13, 385-389 (1996). 5. M. Martinelli, “A universal compensator for polarization changes induced by birefringence on retracing beam,” Opt. Commun, 72, 341-344 (1989). 6. Y. Honda, T. Hibiya and K. Shiroki, “DyBi garnet films with improved temperature dependence of Faraday rotation,” IEEE Trans. J. Magnet. Jpn. TJMJ-2, 142-144 (1987). 7. R. C. Jones, “A new calculus for the treatment of optical systems,” J. Opt. Soc. Am. 31, 488-503, (1941). 8. M. A. Quintela, C. Jauregui, F. J. Madruga, J.M. López-Higuera, “Experimental characterization of light polarization in active erbium doped fiber,” Microwave Opt. Technol. Lett., 42, 395-397 (2004). 9. M. A. Quintela, C. Jauregui, O. M. Conde, A. M. Cubillas, J.M. López-Higuera, “Optical signal polarization state instability on erbium doped fibers,” in Second European Workshop on Optical Fibre Sensors, Proc. SPIE 5502, 544-547 (2004). 10. E. Desurvire, D. Bayart, B. Desthieux, S. Bigo, Erbium-doped fiber amplifiers (Device and system developments), Ed. (Wiley & Sons, New York, 2002). Introduction Erbium Doped Fiber (EDF) has attracted attention in both sensing and telecommunication fields.This is due to its excellent characteristics for optical amplification.The behavior of the polarization state (PS) of the light in EDFs must be kept in mind because its influence on the characteristics of the system is very important.In digital optical fiber communications, EDF polarization phenomena [1] are important limiting factors to high-speed long-distance systems.These so-called polarization phenomena include polarization mode dispersion (PMD), polarization hole burning (PHB), polarization-dependent gain (PDG) and polarization-dependent loss (PDL).Fluctuations in the PS of the optical signal [2] can degrade the quality of the measurements made with optical fiber sensors.This is especially true for those techniques based on interferometry and polarimetry, where an EDF can be used to enhance both sensitivity and accuracy.Different techniques have been developed to improve the polarization behavior of light in erbium doped fiber amplifiers (EDFA).The use of a Faraday rotator mirror (FRM) at the end of the EDF is the most popular technique to stabilize the PS [3,4].This amplifier configuration is known as reflective erbium doped fiber amplifier (R-EDFA).Faraday rotator mirrors consist in a 45º Faraday rotator (based on the Faraday effect) and a mirror [5].These FRMs have the peculiarity that the PS of the reflected (backward) optical signal is orthogonal to the PS of the forward optical signal.Furthermore, this PS is, in theory, independent of the birefringence of the optical system.As a consequence the output PS is preserved regardless of the external perturbations and, therefore, is not affected by polarization phenomena. In general commercial Faraday rotators are made of ferromagnetic materials: yttrium iron garnet (YIG) or bismuth-substituted iron garnet (BIG).The magneto-optical properties of these ferromagnetic materials show great temperature dependence.The reason for this is to be found in the temperature dependence of the saturation magnetization.In crystals with a ferromagnetic garnet structure the saturation magnetization is the vectorial sum of magnetization of three sub-lattices corresponding to three magnetic ion crystallization sites (tetrahedron, octahedron and dodecahedron).Therefore, changing each magnetization contribution of the sub-lattices to the sum of saturation magnetization, or changing the quantity and types of the substituted ions or rare-earth ions in the sub-lattices, the temperature dependence of the Faraday rotation can be modified.Thus, small variations of the rotation angle can take place in the 45º Faraday rotator when it suffers temperature changes [6].However, the temperature sensitivities are different depending on the ferromagnetic material of the rotator.This temperature dependence has a negative influence on the capabilities of compensation of polarization changes provoked by externally-induced birefringence on a R-EDFA.To our knowledge this is the first time that the FRM temperature-dependence influence on the polarization behavior of the optical signal in a R-EDFA has been specifically studied and experimentally demonstrated. This paper is organized as follows: first a theoretical study is carried out, later on several experimental results are presented and discussed and, finally, some conclusions are extracted. Theory The R-EDFA configuration can be extracted from the general measurement set-up used in this work, which is shown in Fig. 1.The optical signal is launched into the EDF through an optical circulator and a wavelength division multiplexing (WDM).The amplified signal, after being reflected at the FRM, is amplified again and is extracted from the port 3 of the optical circulator.That is, the R-EDFA input and output are the circulator ports number 1 and 3, respectively. Jones matrices have been used in this work to analyze polarization properties of the R-EDFA.Jones matrices for both forward (T F ) and backward (T B ) directions in the EDF have been calculated neglecting polarization-dependent losses [7].They are given by: where a and b are complex values depending on the birefringence of the EDF and g is the average gain (including losses). The Jones matrix of the FRM is defined as where ∆ϑ is the rotation angle maladjustment with respect to the 45º rotation angle of the Faraday rotator. The only component remaining to be modeled with a Jones matrix is the circulator.Since a polarization independent circulator has two optical paths there will be a phase delay (∆) between these paths.Thus the Jones matrices that describe the transmission from port 1 to port 2 (T 12 ) and from port 2 to port 3 (T 23 ) of a polarization independent circulator are expressed, respectively, as where β=e j∆/2 and γ c is the circulator loss.It is assumed that the losses are the same for the two paths. The Jones matrices of the EDF in the forward and backward direction (T F and T B ), the Jones matrices of the FRM (T FRM ) and the circulator (T 12 and T 23 ) can be then multiplied by to obtain the complete PS behavior of the light going out the amplifier.Thus, the Jones matrix of the complete amplifier (T R-EDFA ) is: where T U is a unit matrix and T M is a Jones matrix accounting for the rotation angle maladjustment due to the temperature dependence of the magneto-optical Faraday effect: Here p=-ab+a * b * , q=β 2 (a 2 +b 2 ) and \|p\| 2 +\|q\| 2 =1. Applying the T R-EDFA matrix to an arbitrary input PS in the case of using an ideal FRM (∆ϑ=0), it can be demonstrated that the output PS is preserved regardless of the external perturbations.As a consequence, in these ideal circumstances this optical amplifier configuration is immune to any polarization phenomena (PMD, PHB, PDG and PDL).However, it is rather difficult to obtain an accurate 45º rotation angle with the Faraday rotator.Moreover, temperature changes at the FRM play an important role in the rotation angle maladjustment. A figure of merit should be defined [4] to theoretically evaluate the influence of the rotation angle maladjustment on the polarization behavior of the amplified optical signal.This figure of merit is the extinction ratio (r).It relates the amount of output optical power of the desired PS (parallel to the input PS) to that of the unwanted PS (perpendicular to the input PS).The desired PS is parallel to the input PS because, in the ideal case (∆ϑ=0), this fiber amplifier preserves the input PS, as can be seen in ec.(5).Being E the Jones vector of the input signal, the Jones vector of the output signal can be decomposed in two terms: one parallel to E (polarization behavior with an ideal FRM) and another one that can be expressed as a linear combination of E and its orthogonal vector ⊥ E .This last component is a product of the influence of temperature on the rotation angle of the Faraday rotator.Therefore, the extinction ratio r, that relates E (desired) and ⊥ E (unwanted), is given by The complex values u and v are calculated as the inner products of E and T M, and ⊥ E and T M respectively. The minimum extinction ratio r min (worst case) is then given by: Figure 2 shows the minimum extinction ratio for two FRMs (whose Faraday rotators are made of YIG and BIG materials respectively) as a function of the temperature deviation from the temperature at which the Faraday rotator was adjusted to 45º (T adjusted ).The temperature dependence of ∆ϑ is approximately 0.04 deg/ºC for YIG and 0.11 deg/ºC for BIG [6].From Fig. 2, a reduction of the extinction ratio as temperature increases can be observed.To get a minimum extinction ratio of 40 dB (typical value for commercial FRMs), the allowable temperature increment is 7ºC in commercial FRM-YIGs and only 2.5ºC in commercial FRM-BIGs.In short, the mentioned temperature-dependencies of the extinction ratio coefficient represent PS changes of the output optical signal of the R-EDFA.In the following, the latter will be experimentally checked. Experimental results The extinction ratio coefficient was used to theoretically demonstrate the PS changes of the optical signal induced by temperature fluctuations at the FRM.Nevertheless, the actual E and ⊥ E fields at the input and output at the amplifier (port 1 and 3 of the circulator, respectively) and their time stability are very difficult to be experimentally measured.Thus, as the main purpose of this work is to demonstrate the existence of the above mentioned PS changes, another figure of merit was used: the PS variance.This parameter is useful for measuring the PS fluctuations of an optical signal, as has recently been demonstrated by the authors [8] [9].In the present work it is used to experimentally quantify the PS changes of the R-EDFA output optical signal. (9) Fig. 2. Minimum extinction ratio for two FRM's: YIG(solid blue line) and BIG (dotted red line) as a function of the temperature deviation from the one at which the Faraday rotator was adjusted to 45º.On a Poincaré sphere the PS variance (σ PS 2 ) is defined as the average of the azimuth variance (σ AZI 2 ) and the ellipticity variance (σ ELLIP 2 ) of the changing PS.These parameters are determined by [8] [9]: where N is the number of measured points, x i and y i are, respectively, the azimuth and ellipticity values of the PS at each measurement time (T i ).On the other hand, η AZI and η ELLIP are the azimuth and ellipticity averages of the N PS's recorded during the experiment.All the measurements of this work have been carried out with N = 3000 over periods of 30 seconds. In order to measure the PS variance the experimental set-up shown in Fig. 1 was used.The optical signal from the tunable laser (HP-8168F) went through a polarization control device before being launched into the R-EDFA input (circulator port 1).The R-EDFA output (circulator port 3) was connected to an optical polarization analyzer (OPA) (HP-8509B).A tunable filter is additionally used to decrease the optical noise of the amplifier. A 26 meter-long erbium doped fiber (EDF) with an absorption peak of 5dB/m (which corresponds to a Er+3 concentration 300ppm at the core) was employed.The FRM used was a commercial ISOWAVE FRM made of YIG material.The FRM was put in a climatic chamber and the temperature was varied from 30ºC to 65ºC.The input signal had a wavelength of 1550 nm and a power of -5dBm.The EDF was pumped by a 1480 nm pump laser with a power of 17.6 dBm. A sample of the results obtained for a FRM temperature of 30ºC and 65ºC are shown in Fig. 3.The PS random fluctuations of the output optical signal can be observed on the Poincaré sphere.The output optical signal had always a high degree of polarization, that is, it was above 90%.The PS variations were quantified with the aid of the PS variance (σ PS 2 =6⋅10 - 4 and σ PS 2 =7⋅10 -4 for a FRM temperature of 30ºC and 65ºC respectively).As can be observed, the PS random fluctuations of the output optical signal were similar for both FRM temperatures (the surface S on the Poincaré sphere had the same diameter).Consequently the PS variances were also similar for both FRM temperatures.(10) (11) T=30ºC T=65ºC S Fig. 3. Evolution of the PS of the amplified signal for a FRM temperature of 30ºC (blue surface) and for a FRM temperature of 65ºC (red surface).S is the area drawn on the Poincaré sphere by the changing PS in 30 seconds. These fluctuations can be explained by the fact that during the amplification process some photons are spontaneously emitted with random phase, random direction and random PS.During its propagation through the EDF these random photons are replicated through a stimulated emission process, in a similar fashion as the photons of the input light [10].As a consequence, it can be stated that this amplified spontaneous emission (ASE) is a big deal responsible for the instabilities of the PS of the amplified optical signal.This also explains that the PS variances at both temperatures were similar.The gain and ASE of the R-EDFA were almost identical for these two FRM temperatures because the EDF was at a fixed temperature of 22ºC.Although in this work the measurements have been carried out at 1550 nm (the FRM was designed to work at this wavelength), it should be pointed out that the PS variance is wavelength-dependent due to its direct relation with the ASE and gain of the R-EDFA.The PS variance, the gain and the ASE present maximums and minimums at the same wavelengths.Maximums around 1555 nm and minimums around 1570 nm and 1540 nm, respectively [8] [9]. Nevertheless, a movement of the S surface on the Poincaré sphere can be observed in Fig. 3.As was theoretically predicted by the extinction ratio coefficient, this movement of the S surface represents the change of the PS of the output optical signal due to the FRM temperature change.The movement of the S surface on the Poincaré sphere with respect to that corresponding to the initial temperature (T=30ºC) was also measured with the PS variance.In this case the azimuth and ellipticity variances of the changing PS were calculated at each temperature with respect to the azimuth and ellipticity averages of the PS at the initial temperature.Experimental measurements of this PS variance for temperatures ranging from 30ºC to 65ºC are shown in Fig. 4. In this graph the PS variance at the initial FRM temperature (T=30ºC) was subtracted from the PS variances of the other temperatures.From the figure, it can be deduced that the PS variance and, consequently, the change of the PS of the output optical signal increase with FRM temperature.These experimental results are in good agreement with those theoretically obtained from the extinction ratio coefficient.The most important characteristic of the R-EDFA is its capacity to stabilize the PS of the output optical signal when externally-induced birefringences are present.However, the influence of the FRM temperature-dependence on the polarization has a negative effect on this characteristic of the R-EDFA.To experimentally verify this assertion, external perturbations were applied to an EDF section in order to induce linear birefringence.These external perturbations were induced by attaching a 8-turn 6-cm-diameter EDF coil to a fix holder placed on top of a TIRAvib 5100 shaker. Shaker Measurements of the PS variations of the output optical signal on the Poincaré sphere with perturbations applied to the EDF-coil for a FRM temperature of 30ºC and 65ºC are shown in Fig. 5(a) and Fig. 5(b), respectively.From Fig. 5(a) it can be seen that the PS of the output optical signal didn't suffer a substantial change with respect to the results of the same experiment but without perturbations (Fig. 3(a)).It can be seen that there was only a slight increment of the surface S on the Poincaré sphere when the system was perturbed.This might be due to a non-exactly 45º rotation angle.The results suggest that the FRM effectively cancels the effects of the linear birefringence.They also suggest that the PS variations of the output optical signal are exclusively caused by the ASE of the R-EDFA at this temperature. On the other hand, the PS of the output optical signal changed sharply when the EDF was perturbed and the FRM temperature was 65ºC (Fig. 5(b)).It can be observed that this externally-induced linear birefringence provoked important changes in the ellipticty of the PS. Taking profit of the ability of the PS variance to quantify PS variations, this variable was used to measure the externally-induced PS shifts of the output optical signal for FRM temperatures ranging from 30ºC to 65ºC.The results obtained are shown in Fig. 6.In that figure, the PS variance is normalized with respect to that corresponding to a FRM temperature of 30ºC.From the figure, it can be observed that, when the EDF was perturbed, the PS variance increased with FRM temperature.Thus, the PS variance at 65ºC was 2.4 times bigger than the PS variance at 30 ºC.This means that the birefringence effects on the PS of the output optical signal were bigger with increasing FRM temperatures because there was an increment of the rotation angle maladjustment. Conclusion In this paper the temperature dependence of FRMs and its influence on the behavior of the polarization in a R-EDFA is theoretically predicted and experimentally demonstrated.The effectiveness of the FRM gets worse with increasing temperature.As a consequence the PS shifts increase with FRM temperature when a linear birefringence is induced in the EDF.Thus, these measurements demonstrate that the FRM doesn't completely suppress the influence of external perturbations on the PS of the output optical signal.Therefore, for a stable operation of the R-EDFA polarization a temperature control at the FRM is required. Fig. 1 . Fig. 1.General measurement set-up used in this research work. Fig. 4 . Fig. 4. PS variance as a function of FRM temperature.This PS variance quantify the change of the PS of the output optical signal. Fig. 5 . Fig. 5. Evolution of the PS of the amplified signal with perturbations applied to the EDF-coil: a) at a FRM temperature of 30ºC, b) at a FRM temperature of 65ºC.a) b) Fig. 6 . Fig. 6.(a) PS variance as a function of FRM temperature when the EDF was perturbed.(b) Illustration of the EDF coil deformation mechanism used to induce linear birefringence changes	2017-06-16T13:56:35.734Z	2005-03-07T00:00:00.000	{ "year": 2005, "sha1": "0cf071aafb76266b9b4dda16b98371a3f051ae9b", "oa_license": "CCBY", "oa_url": "https://doi.org/10.1364/opex.13.001368", "oa_status": "GOLD", "pdf_src": "ScienceParseMerged", "pdf_hash": "0cf071aafb76266b9b4dda16b98371a3f051ae9b", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Materials Science", "Medicine" ] }
19530998	pes2o/s2orc	v3-fos-license	Zooxanthellae Harvested by Ciliates Associated with Brown Band Syndrome of Corals Remain Photosynthetically Competent ABSTRACT Brown band syndrome is a new coral affliction characterized by a local accumulation of yet-unidentified ciliates migrating as a band along the branches of coral colonies. In the current study, morphologically intact zooxanthellae (= Symbiodinium) were observed in great numbers inside the ciliates (>50 dinoflagellates per ciliate). Microscale oxygen measurements and variable chlorophyll a fluorescence analysis along with microscopic observations demonstrated that zooxanthellae within the ciliates are photosynthetically competent and do not become compromised during the progression of the brown band zone. Zooxanthellae showed similar trends in light acclimation in a comparison of rapid light curve and steady-state light curve measures of variable chlorophyll a fluorescence. Extended light exposure of steady-state light curves resulted in higher quantum yields of photosystem II. The brown band tissue exhibited higher photosynthetically active radiation absorptivity, indicating more efficient light absorption due to a higher density of zooxanthellae in the ciliate-dominated zone. This caused relatively higher gross photosynthesis rates in the zone with zooxanthella-containing ciliates compared to healthy coral tissue. The observation of photosynthetically active intracellular zooxanthellae in the ciliates suggests that the latter can benefit from photosynthates produced by ingested zooxanthellae and from photosynthetic oxygen production that alleviates diffusion limitation of oxic respiration in the densely populated brown band tissue. It remains to be shown whether the zooxanthellae form a stable symbiotic association with the ciliate or are engulfed incidentally during grazing on coral tissue and then maintained as active inside the ciliate for a period before being digested and replaced by new zooxanthellae. A recently identified coral disease, brown band syndrome (BrB), was described for the first time in three coral families (Acroporidae, Pocilloporidae, and Faviidae) in the northern and southern sectors of the Great Barrier Reef (GBR), Australia (55). Macroscopic symptoms of the syndrome manifest as a brown zone on the coral preceded by healthy tissue and followed by exposed white skeleton (see Fig. 1A). Sometimes, a white zone is observed between the brown band and healthy tissue which may comprise bleached tissue and/or denuded skeleton (55). Microscopic investigations of the brown zone revealed the massive presence of an unknown protozoan ciliate which accumulates zooxanthellae intracellularly, resulting in the characteristic brown coloring of the observed syndrome. The ciliate band has been observed to migrate along the length of branching corals from base to tip at a rapid rate (Ͼ5 cm per day [B. Willis, personal communication]). The ciliates appear to ingest coral tissue at the lesion interface, accumulating the symbiotic dinoflagellates, known as zooxanthellae (ϭ Symbiodinium), from the coral endoderm. However, it is unknown whether the zooxanthellae remain photosynthetically active inside the ciliates. In this study, we used variable chlorophyll a fluorescence analysis with Mini-and Imaging-PAM (pulse-amplitude-modulation) instruments (22,23,37) and the saturation-pulse tech-nique (reviewed in reference 45) to obtain measures of maximum and effective quantum yields of photosystem II (PSII) and derived measures of photosynthetic electron transport (24,36). These measurements were combined with oxygen microelectrodes (25,53) to assess the photosynthetic competency of zooxanthellae of corals affected with BrB. The combination of their fast response times, small tip diameters, and extremely small oxygen consumption means that O 2 microsensors are able to measure rapid changes in oxygen depletion as well as microprofiling of the diffusion boundary layer (DBL) (required to measure net photosynthesis and dark-respiration rates) at a spatial resolution of better than 100 m and with response times of Ͻ0.2 to 0.5 s (26,40,53). Only a limited number of studies have examined coral-protozoan associations (1,2,7,29,30,47,56). In this study, we compare the photosynthetic performances of zooxanthellae in healthy tissue preceding the brown band zone and within ciliates accumulating in the brown band zone of affected Acropora muricata corals. Direct microscopic and histological observations combined with photophysiological responses were analyzed to determine the competency of internalized zooxanthellae. Our results show that the dynamics of photosynthesis regulation of visually healthy zooxanthellae accumulated in the brown band zone were greatly altered (as shown by O 2 -derived measures) from that of zooxanthellae in the healthy tissue. Nevertheless, the capacity for photosynthesis appeared unchanged (as shown by variable chlorophyll a fluorescencederived measures). MATERIALS AND METHODS Sampling. Specimens of the branching coral Acropora muricata exhibiting symptoms of BrB were collected in November 2005 from an exposed reef flat at Davies Reef (18°49.86ЈS, 147°38.2ЈE) on the GBR and brought to the facilities of the Australian Institute of Marine Science in Townsville for further analysis. Histological and microscopic examination. For histological examination, the brown band mass including ciliates and coral tissue was scraped from a coral branch and concentrated by centrifugation (1,000 ϫ g) with excess liquid removed. Samples were fixed in Davidson's fixative (21) for 24 h prior to processing. All samples were sectioned (5 m) and stained (Mayer's hematoxylin and eosin) using routine histological procedures and examined using light microscopy (3). Microscopic visualization was performed with a Zeiss Axioskop II microscope (Carl Zeiss, Jena, Germany) with images captured by an AxioCam MRC5 (Zeiss) high-resolution, cooled charge-coupled device camera. Fluorescence images were obtained using a halogen lamp for incident light and a Cy3 filter set, HQCy3 (excitation, BP535/50 nm; dichroic mirror, Q565 LP; emission, BP610/75 nm) (Carl Zeiss, Jena, Germany). PAR Abs and chlorophyll a determinations. The Imaging-PAM estimates photosynthetically active radiation (PAR) absorptivity (Abs), i.e., the fraction of incident PAR that is absorbed, from the ratio of reflectance of red light (650 nm) (R) to the reflectance of nonabsorbed near-infrared light (780 nm) (NIR) from the tissue surface: This parameter was measured in both the healthy and brown band tissue of each replicate coral colony. It is important to note that Abs estimates the light absorption via reflectance from a surface. Highly pigmented tissue surfaces will thus show higher absorptivity than surfaces with pigments distributed more homogeneously or deeper in the tissue although the overall chlorophyll concentration may be the same. Tissue derived from ϳ1 cm 2 of healthy and brown band tissue, respectively, was extracted in 1 ml of methanol for 2 h. Extracts were centrifuged (15,000 rpm) for 2 min, and the clarified supernatant was mixed 1:1 in 70:30 (vol/vol) methanol-28 mM aqueous tetrabutyl ammonium acetate, pH 6.5. A 5-l sample of the mixture was injected onto a Phenomenex C 8 column (3-m particle size, 50 ϫ 4.6 mm) at room temperature (21°C). Analysis was performed by reverse-phase high-performance liquid chromatography on a personal computer-interfaced high-performance liquid chromatography system (Waters Corporation). System control, data collection, and integration were performed using Empower Pro software (Waters Corporation, 2002). Pigments were separated using a binary mobile phase system, 70:30 (vol/vol) methanol-28 mM aqueous tetrabutyl ammonium acetate, pH 6.5 (75% to 100%), over 10 min. The flow rate was kept constant at 1 ml min Ϫ1 . Chlorophyll a was detected using a photodiode array detector, and pigment identity was determined through spectral comparison to a chlorophyll a standard (Sigma-Aldrich). The height (h) and radius (r) of the cylinder-shaped and fingertip-sized coral pieces, from which tissue was extracted, were measured to the nearest mm using calipers and applied to the formula for calculating the lateral surface area of a cylinder ([2 ϫ ϫ r] ϫ h). The surface area was calculated excluding the top and base, since these represented the area of attachment to the colony and therefore had no surface tissue. Combined measures of variable chlorophyll a fluorescence and O 2 productivity. In order to assess the photosynthetic activity of zooxanthellae, three or four Acropora muricata branches from separate coral colonies showing symptoms of BrB were measured. The following setup and measuring sequence were applied to each colony: once in the asymptomatic healthy tissue 4 to 5 cm away from the lesion and once in the center (most dense region) of the brown band zone. An initial rapid light curve (RLC) with 10-second intervals between irradiances (0, 1, 73, 144, 365, 515, 690, 911, and 1204 mol photons m Ϫ2 s Ϫ1 ) was performed on a dark-acclimated coral using the Imaging-PAM fluorometer (Walz, Effeltrich, Germany). For oxygen microelectrode measurements, the coral was then placed in a custom-built flow chamber (25 ϫ 10 ϫ 10 cm) through which seawater from a reservoir was circulated at a flow velocity of ϳ1 cm s Ϫ1 . The reservoir contained approximately 20 liters of seawater and was maintained at 27 Ϯ 0.5°C using a glass aquarium heater (100 W; Aqua One; Kong's International, Australia). An O 2 microelectrode (39), which was linearly calibrated by recording signals in air-saturated seawater and O 2 -free seawater, was positioned at an angle of 25°r elative to the light source (Schott 2500), while the coral sample was observed with a dissecting microscope (Leica, Germany). The microsensor was placed in direct contact with the coral tissue surface by using a manually operated micromanipulator (MM33; Märtzhäuser, Germany), and the measuring signal was recorded via a picoamperemeter (PA2000; Unisense A/S, Denmark) on a strip-chart recorder (Kipp & Zonen, The Netherlands). Corals were illuminated with a fiber-optic tungsten-halogen light source with a collimating lens (Schott KL-2500; Germany) that was calibrated against a light sensor (LI-192SB; LI-COR). Each irradiance (0, 33, 59, 91, 130, 247, 494, and 845 mol photons m Ϫ2 s Ϫ1 ) was applied for 10 min in order to obtain steady-state conditions prior to O 2 measurements (53). At steady state for a particular irradiance O 2 transport, consumption, and production are in equilibrium and no changes to the O 2 level are detected over time (26). Monitoring of the tissue surface oxygen level on the strip-chart recorder after adjustments of irradiance showed that 10 min was sufficient to reach steady state. Oxygen microprofiles in darkness were measured in vertical steps of 10 m from the tissue surface through the DBL and out into the bulk water, which had a constant oxygen content (tabulated value of 204.3 mol O 2 liter Ϫ1 at 27°C and salinity of 35 ppt; Unisense). The oxygen flux (dark-respiration rate) was calculated from the oxygen concentration profile as where D o is the diffusion coefficient of O 2 at 27°C and salinity of 35 ppt (tabulated value, 2.35 ϫ 10 Ϫ5 cm 2 s Ϫ1 ; Unisense) and dC/dz is the constant slope of the oxygen gradient in the DBL; this calculation assumes a one-dimensional diffusion geometry (25). At each following irradiance level (33, 59, 91, 130, 247, 494, and 845 mol photons m Ϫ2 s Ϫ1 ), the microelectrode tip was positioned at the tissue surface, and when steady-state oxygen conditions were verified, a sudden and brief (1-to 3-s) light-dark shift was performed to estimate gross photosynthesis rate (26,53). The light-dark shift technique, which requires steady-state O 2 distribution before the light-dark shift, enables the measurement of gross photosynthesis rate as the immediate rate of oxygen depletion after onset of darkness (more details in reference 40). After gross photosynthesis measurements, oxygen microprofiles were measured and used to estimate net photosynthesis rates by equation 2. Finally, the Mini-PAM fluorometer (Walz, Effeltrich, Germany) fitted with a 2-mm fiber-optic probe was used at each irradiance level for an effective quantum yield (⌬F/F m Ј) reading using a single saturating flash. The probe was gently hand held to the surface of the coral adjacent to the O 2 microsensor. These measurements were used to generate a steady-state light curve (SSLC) of relative electron transport rate (rETR) versus irradiance. The initial RLC using the Imaging-PAM was repeated at the highest irradiance level (845 mol photons m Ϫ2 s Ϫ1 ). In order to maintain the high-lightacclimated state of PSII, actinic light (845 mol photons m Ϫ2 s Ϫ1 ) was turned on for 1 min after the coral was positioned in the Imaging-PAM prior to initiation of the RLC. The first saturating flash in the dark-acclimated RLC provided a measure of maximum quantum yield (F v /F m ), which was compared to the first measure of the light-acclimated RLC, which provided a comparative ⌬F/F m Ј (845 mol photons m Ϫ2 s Ϫ1 ). The maximum excitation pressure over PSII was then calculated as (24). Q m is used as a proxy for the physiological performance of the zooxanthellae, where values close to 0 indicate that photosynthesis is light limited even at high irradiance, while values closer to 1 indicate closure of PSII at high irradiance, e.g., due to photoinhibition. The rETR derived from the RLCs as well as from the steady-state light curves was calculated as rETR The light curves (both fluorescence and O 2 derived) were subsequently fitted to the formula of Platt et al. (35), whereby quantitative parameters of rETR max or P g,max (maximum gross photosynthesis rate), ␣ (light utilization coefficient), and E k (minimum saturating irradiance) were calculated (details are in references 27 and 36). Statistical analysis. One-way analysis of variance tests (SPSS v11.0.0) were used to determine if significant differences in the following parameters were present between healthy and brown band tissue: F v /F m , ⌬F/F m Ј, Q m , quantitative parameters of RLCs and steady-state light curves (fluorescence and O 2 derived), PAR absorptivity, and chlorophyll a content. Where the assumptions of normality and equal variance failed (P Ͻ 0.05), data were transformed using natural log; this applied only to rETR max values for RLCs and SSLCs. Transformed data successfully met the assumptions of normality and equal variance. Post hoc comparisons were performed on quantitative parameters of RLCs and fluorescence derived steady-state light curves using the Tukey-Kramer honestly significant difference test. RESULTS Field and microscopic observations of BrB. BrB exhibits distinctive macroscopic symptoms in the field (Fig. 1A), which VOL. 73, 2007 PHOTOPHYSIOLOGY OF BROWN BAND SYNDROME IN CORALS 1969 are derived from ciliates gliding over the exterior of the coral samples and into the coelenteron and cavities of the coral polyps. Light microscopic analysis of the ciliates demonstrated an accumulation of intact zooxanthellae within the ciliates (Fig. 1B). The ingested zooxanthellae maintained a strong localized chlorophyll autofluorescence, indicating intact chloroplasts and a potential capacity for photosynthesis (Fig. 1C). In addition, histological cross sections of the sampled brown band mass demonstrated that the ciliates contained intact zooxanthellae that were not surrounded by food vacuoles (Fig. 1D). PAR absorptivity and chlorophyll a determinations. Variable chlorophyll a fluorescence imaging showed high fluorescence along with high PAR absorptivity in the brown band tissue ( Fig. 2A and C), and active photosynthesis in the brown band tissue was evident by a maximum quantum yield (F v /F m ) comparable to or slightly higher than that in the healthy tissue (Fig. 2B). The quantitative parameters of RLCs (rETR max , ␣, and E k ) measured initially in the dark and again after increasing light exposure for 80 min were statistically compared to rETR max , ␣, and E k calculated for fluorescence-derived SSLCs (Table 1). rETR max calculated for light-and dark-acclimated RLCs of healthy and brown band tissue showed similar values. However, the RLC performed on dark-acclimated brown band tissue also shared similar values with rETR max calculated for SSLCs of healthy tissue which were significantly different (P ϭ 0.003). The rETR max calculated for SSLCs of brown band tissue was similar to values collected for all other groups of tissue and light curve types ( Table 1). The slope of the rETR versus irradiance curves, ␣, showed similar values between healthy and brown band tissue within all three groups of light curves (Table 1). However, RLCs performed in the dark showed a significantly lower (P ϭ 0.011) ␣ than those calculated for SSLCs. ␣ values calculated for RLCs performed in the light were similar to those for both dark-acclimated RLCs and SSLCs. There was no significant difference among groups of light curves and tissue types in minimum saturating irradiance, E k (P ϭ 0.571) (Table 1; Fig. 5). Healthy and brown band tissues exhibited widely different O 2 concentration levels over the entire irradiance range (Fig. 6). At the highest irradiance level (845 mol photons m Ϫ2 s Ϫ1 ), the O 2 concentration of healthy tissue was 223 Ϯ 4 mol liter Ϫ1 whereas the O 2 concentration of the brown band tissue was only 63 Ϯ 22 mol liter Ϫ1 . This amounts to ϳ30% of saturation O 2 concentration. In the dark, the O 2 concentration of the healthy tissue was 60 Ϯ 28 mol liter Ϫ1 but virtually 0 (3 Ϯ 3 mol liter Ϫ1 ) for brown band tissue due to high respiration activity coupled with O 2 transfer limitation imposed by the DBL (Fig. 6). The brown band tissue was almost completely hypoxic at irradiances below 60 mol photons m Ϫ2 s Ϫ1 and never reached the compensation irradiance (E c ), i.e., the irradiance above which the tissue exhibits net oxygen production. The healthy tissue reached compensation irradiance at ϳ230 mol photons m Ϫ2 s Ϫ1 . Microsensor measurements generally showed a thin DBL ranging from 100 to 200 m and strong oxygen gradients across the DBL. Motility of the ciliates caused fluctuating oxygen signals, and this resulted in relative high variability in our flux (Fig. 7). The compensation irradiance, E c , i.e., the irradiance above which the tissue exhibits net oxygen production, was reached between 130 and 247 mol photons m Ϫ2 s Ϫ1 in healthy tissue. However, maximum net production recorded beyond the compensation irradiance was low and highly variable (0.15 Ϯ 0.09 nmol O 2 cm Ϫ2 s Ϫ1 ). The brown band tissue exhibited net oxygen consumption at all irradiances, but consumption rates decreased with increasing irradiance, stabilizing around the E c determined for the healthy tissue (Fig. 7). Although no net oxygen production was measured in the brown band tissue due to intense respiration, there was substantial gross photosynthesis (Fig. 8). There was no significant difference between quantitative parameters (P g,max , ␣, and E k ) calculated from fitted gross photosynthesis rate and those calculated from irradiance curves ( Table 2). Brown band tissue showed higher P g,max than healthy tissue (Fig. 8), although the difference was not significant (P ϭ 0.170) ( Table 2). There was no significant difference between healthy and brown band tissue in ␣ (P ϭ 0.138) or E k (P ϭ 0.517) ( Table 1). DISCUSSION Intact zooxanthellae in ciliates associated with BrB. Reports concerning diseases of coral reef organisms have increased substantially over the last 2 decades (19,28,54). Coral diseases appearing with progressively greater frequency and wider distributions have been shown to alter total coral abundance and species diversity (20,31,44). A wide range of microorganisms including fungi, bacteria, cyanobacteria, and protozoans have been demonstrated to associate with both healthy and diseased corals (6,15,29,30,34,41,42,43). BrB was recently described as a newly identified coral disease on the GBR. The distinctive macroscopic field symptom of BrB is the formation of a brown zone on the coral harboring high densities of motile yet unidentified protozoan ciliates (55). Although the prevalence of BrB on three reefs in the northern GBR was quite low, i.e., Ͻ1% of total number of coral colonies surveyed, corresponding to 12 to 24 cases per investigated reef (55), the high spreading rates and efficient transmission of BrB may cause significant damage to coral reef assemblages and can Although other organisms such as bacteria and diatoms are also associated with the brown band zone of affected corals, the ciliates and ingested zooxanthellae account for the largest biomass. Light microscopy of the ciliates in the brown band tissue showed intact internal zooxanthellae with strong chlorophyll autofluorescence ( Fig. 1B and C), lending support to the notion of photosynthetic competency of the zooxanthellae. The degeneration of symbiotic zooxanthellae in symbio is a routine process as cells divide and senesce (13,14). Morphological changes associated with degeneration are characterized by vacuolization, disorganization of thylakoids, and enlargement of the accumulation body (5,9,13). Increased vacuolization in corals has previously been observed to result in loss of symbiont autofluorescence under green excitation light (13), but this was not observed in the grazed zooxanthellae investigated in this study. In addition, histological sections of ciliates also showed intact zooxanthellae within the cell cytostome (Fig. 1D) and without surrounding food vacuoles, which are otherwise indicative of the degradation of ingested food (18,38). Our microscopic observations thus clearly showed that in BrB the ingested zooxanthellae are not degraded rapidly and remain intact within the ciliates. Photophysiology of ingested zooxanthellae. Based on a novel methodological approach we showed that the zooxanthellae within the ciliates were photosynthetically competent. While microsensors for oxygen and variable chlorophyll a fluorescence have previously been employed separately to investigate the heterogeneity of zooxanthellar photosynthesis in hospite (25,36), this study provides the first combined measurements of PSII quantum yields and derived rETRs together with measurements of steady-state O 2 concentration, rates of O 2 flux, and gross photosynthesis in diseased coral tissues. Our comparison between RLCs and SSLCs obtained from diseased corals is also novel. Our results show that SSLCs are different from RLCs, which delay only 10 to 20 s between each irradiance level. It is evident that if zooxanthellae are allowed to acclimatize for an extended period of time to each irradiance level, their photosynthetic performance yields higher rETR values. Thus, fluorescence-derived light curves obtained using PAM instruments are dependent on previous light exposure (see also reference 27). Pigmentation and optical properties. Although the chlorophyll a content was similar across the disease lesion, the PAR absorptivity was significantly higher in the brown band zone (Fig. 2B and 3), suggesting that the light climate available to the zooxanthellae in this region was different from that of zooxanthellae associated with the healthy tissue. The higher absorptivity in the brown band indicates a more densely pigmented surface layer due to the high density of zooxanthellacontaining ciliates in this zone. In healthy tissue, the pigments are distributed differently and primarily in the coral endoderm below the surface. The different tissue composition of the healthy and brown band zones invariably results in differing spectral properties and intensities of the light reaching the zooxanthellae in the two zones (10,25). The paler tissue coloration in the healthy tissue (Fig. 1A) thus indicates a higher contribution of diffuse scattered light from the skeleton to the ambient light field of the zooxanthellae compared to the brown band tissue, and this will influence light-dependent measures of photosynthesis activity such as the effective quantum yield (⌬F/F m Ј) of PSII. Variable chlorophyll a fluorescence analysis. There was no significant difference in F v /F m between healthy and brown band tissue, indicating that PSII activity was not inhibited as a consequence of ingestion by ciliates ( Fig. 2 and 4). Actually, the efficiency of light utilization, characterized by the effective PSII quantum yield, ⌬F/F m Ј, apparently increased in zooxanthellae associating with ciliates. Interestingly, the healthy tissue exhibited slightly lower rETR values throughout the irradiance range in dark-acclimated RLCs than did brown band tissue (Fig. 5), and this was mirrored in the gross photosynthesis irradiance curves (Fig. 8). Although the quantitative parameters of the RLCs are not significantly different (Tables 1 and 2), it is notable that the trend is reproduced across different techniques. No photoinhibition, i.e., a decline in quantum yields or O 2 production measures at high irradiances, was observed except in the fluorescence-derived SSLC measures of healthy tissue (Fig. 5). The lack of significant differences in other fluorescence-derived light curves suggests that the photoinhibition observed is a reflection of a decrease in PAR absorptivity observed in healthy tissue that may cause higher scattering and light exposure (Fig. 3) (10). Oxygen dynamics. The O 2 dynamics measured at the tissue surface showed a typical pattern of increasing O 2 levels, which saturated and approached a maximum value asymptotically with increasing irradiance for healthy tissue (Fig. 6). From such curves, the compensation irradiance, E c , can be estimated as the irradiance at which there is no oxygen gradient across the DBL between the tissue surface and the surrounding water. At this irradiance there is no net exchange as oxygen consumption and production balance; above E c the system thus switches from being net heterotrophic to becoming autotrophic. In healthy tissue this occurred at a higher irradiance (ϳ230 mol photons m Ϫ2 s Ϫ1 ) than that previously reported for corals (25,53), reflecting the fact that the corals were high light adapted (the Acropora muricata branches were from the reef flat) and thus exhibited a high respiration rate. In contrast, the brown band tissue demonstrated much lower O 2 concentration at the tissue surface, and oxygen levels at the highest irradiance reached only ϳ25% of the O 2 concentration in the surrounding water due to high respiration of the ciliates and other constituents of the brown band tissue. In darkness and at the lowest irradiances, the brown band tissue was almost anoxic at the surface, indicating diffusion limitation of the oxygen supply across the DBL (Fig. 6). This limitation was clearly alleviated by the photosynthetic activity of the zooxanthellae at higher irradiance. Oxygen flux calculations for brown band tissue showed a net oxygen uptake, indicating that respiration was greater than gross photosynthesis at all experimental irradiances (Fig. 7). Oxygen levels at the brown tissue surface were fluctuating due to the presence of highly motile and dense ciliate populations, and this caused high variability in our oxygen profiles. It is also possible that boundary layer compression due to the microsensor (see reference 16) affected our oxygen microprofiles and the derived oxygen fluxes. Finally, high densities of ciliates have previously been shown to affect the local transport VOL. 73, 2007 PHOTOPHYSIOLOGY OF BROWN BAND SYNDROME IN CORALS 1973 coefficient of oxygen (17), and if such enhancement was present in the brown band tissue it would affect our flux calculations. Advection induced by the ciliates could alleviate diffusion limitation significantly, as has been shown for large motile bacteria (12,52). Enhancement of transport was shown to depend on both size and cell density of ciliates. For large filter-feeding ciliates such as Euplotes sp., it was shown that at cell densities above 10 3 cm Ϫ2 oxygen transport was significantly enhanced over the oxygen supply by diffusion alone (17). A zooxanthella-containing Euplotes uncinatus was recently described by Lobban et al. (30). A quantification of the ciliate density along with more detailed microsensor measurements and imaging of ciliate behavior would provide more insights into the mechanisms that affect oxygen fluxes in BrB, and our oxygen turnover rates reported here should thus be regarded as rough estimates. Nevertheless, the data clearly support our other data showing that zooxanthellae in the ciliates exhibit active photosynthesis. Photosynthesis in healthy versus brown band tissue. The A. muricata branches sampled in this study were collected from a reef flat exposed to high irradiance, and as such the healthy tissue was visibly pale, indicative of a relatively low density of zooxanthellae in the coral tissue (Fig. 1A). The low net photosynthesis rates observed suggest that these high-light-exposed corals acquire necessary photosynthates from reduced zooxanthellar densities and thereby also reduce the risk of exposure to excess production of reactive oxygen species (32). In addition, the causative agent of BrB syndrome has not been determined, and it is unknown if the net photosynthesis production of healthy tissue preceding the brown band zone is compromised by a bacterial or viral pathogen. Pantos et al. (33) demonstrated that the bacterial community of the whole coral colony is affected even when just a small part of the colony shows signs of disease. Results of gross photosynthesis measures (Fig. 8) confirm that the brown band tissue is indeed highly dynamic. The high variation observed in measurements, indicated by the large standard error, reflects the motility of the layer, which causes signal fluctuations even at steady state. The observed variation may also be due to variable health of the zooxanthellae, though this is unlikely since the PAM data do not support the assertion of compromised zooxanthellar health. Gross photosynthesis rates were higher, although not statistically significant, in the brown band tissue than in healthy tissue. Although the chlorophyll a contents of healthy tissue and brown band tissue were similar, the local density of zooxanthellae in the ciliates was enhanced, as seen in the higher absorptivity in the brown band tissue. Gross photosynthesis rate measurements with oxygen microsensors are influenced by the local density of active zooxanthellae around the tiny measuring tip, and this, along with a more efficient light capture, can explain the increased productivity in the brown band zone compared to healthy tissue. However, measurements of oxygen levels and fluxes showed that most of this productivity is channeled into respiration in the highly motile ciliate band. Species of oligotrichs in the genera Tontonia, Laboea, and Strombidium have been shown to enslave chloroplasts from ingested flagellates, and such kleptochloroplasts remain functional for periods ranging from hours to days fixing carbon that is metabolized by the ciliate (48,49). However, kleptochloroplasts do not reproduce inside the ciliate and must therefore be constantly replaced through feeding. Such mixotrophic marine oligotrichs use photosynthesis primarily to cover respiratory demands and to increase their growth efficiency, but photosynthesis may also alleviate starvation when food is scarce (8,50). To confirm such a survival strategy for brown band ciliates, life history studies of the ciliates are required. Recently, the marine ciliates Maristentor dinoferus (Heterotrichea) and Euplotes uncinatus have been demonstrated to form associations with symbiotic zooxanthellae (29,30). In the case of Maristentor dinoferus, approximately 50 to 800 dinoflagellates in the genus Symbiodinium clade C lineage have been observed within single ciliates (29). It is speculated that such ciliate-zooxanthella associations allow nutrients to pass from the ciliate to the dinoflagellate and photosynthates from dinoflagellate to the heterotrophic ciliate (30). Sommaruga et al. (47) detected UV-absorbing mycosporine-like amino acids in Maristentor dinoferus and proposed an additional mutualistic benefit of intracellular zooxanthellae, which in some cases are known to produce mycosporine-like amino acids (46). Whether similar UV protection occurs for ciliates in the brown band tissue requires a more detailed analysis of their pigmentation. Conclusion. Our results show that zooxanthellae ingested by ciliates associated with BrB are photosynthetically competent and do not become compromised during the progression of the ciliate band. Ciliates feeding on coral tissue within the brown band tissue can thus gain additional energy from photosynthates and possibly alleviate oxygen limitation due to the photosynthesis of ingested zooxanthellae. Whether this association is a stable symbiosis or whether the zooxanthellae undergo a turnover and slow degradation in the ciliate, which requires constant uptake of new photosynthetically competent zooxanthellae, awaits further investigation.	2017-06-17T14:43:10.290Z	2007-01-26T00:00:00.000	{ "year": 2007, "sha1": "593c86b47cc8d01fa5a5859e60322b2a96502e8c", "oa_license": "CCBY", "oa_url": "https://aem.asm.org/content/73/6/1968.full.pdf", "oa_status": "HYBRID", "pdf_src": "Highwire", "pdf_hash": "dfc850ec11ecedf9c9ce8bb07b9c3cb78a86033d", "s2fieldsofstudy": [ "Environmental Science" ], "extfieldsofstudy": [ "Biology", "Medicine" ] }
145862636	pes2o/s2orc	v3-fos-license	Frequency of Urinary Tract Infection among Febrile Pediatric Thalassemic Patients in Babylon Hereditary Blood Disease Center in Babylon Governorate / Iraq Background: There is an increased risk of infections in patients with thalassemia compared to general population including urinary tract infections (UTI). The susceptibility to infections in thalassemia arises from iron overload, transfusion therapy, splenectomy, zinc deficiency &disease itself. all these changes inherent to the pathological process which can interfere with the immune systems Aims of study: The study was carried out to determine the frequency of urinary tract infection among thalassemic patients, its correlation to risk factors including vaccination, splenectomy, iron overload, Hb level and to determine microorganism causing urinary tract infection among thalassemic patients and control group . Methods: A case control study was carried out at the Babylon hereditary blood disease center in Babylon Maternity and Children Teaching Hospital in Hilla / Iraq from March to October 2018 among 100 febrile thalassemic patients as patients group , comparison to 100 febrile non thalassemic patients as control group with age ranging between 6 months -15 years. Urine exam & culture, serum ferritin, Hb, C-reactive protein, ESR were sent for all patients. Results: The current study has revealed that 14 % of thalassemic patients developed UTI, in comparison to 5% in non thalassaemic patients .The infection is more among those with low Hb, incomplete vaccination, splenectomized patients but has no relation to gender &type of thalassemia and E. coli was the commonest microorganism among both group. Conclusion: This study revealed that the frequency of UTI among febrile thalassemic patients is 14 %, its increased among splenectomized patients, incomplete vaccination and iron overload. Introduction Thalassemia is a heterogeneous group of heritable Hypochromic anemia of various degrees of severity, resulting from genetic defects include total or partial deletions of globin chain genes and nucleotide substitutions, deletions or insertions that result in decreasing or total suppression of hemoglobin polypeptide chain. (1) Thalassemia is characterized by abnormal hemoglobin synthesis, which results in decreased www.jmscr.igmpublication.org Index Copernicus Value: 79.54 ISSN (e)-2347-176x ISSN (p) 2455-0450 DOI: https://dx.doi.org/10.18535/jmscr/v7i4.114 oxygen delivery to the tissues and to improve the oxygen-carrying capacity of the blood . Patients who receive regular transfusions may result in iron overload in various organs including heart, liver and endocrine organs. (2,3,4,5) The susceptibility to infections in thalassemia arises from iron overload, transfusion therapy, splenectomy, zinc deficiency& disease itself . All these changes inherent to the pathological process which can interfere with the immune systems. (6,7,8) One of the first lines of defense against bacterial infection is the withholding of nutrients to prevent bacterial outgrowth in a process termed nutritional immunity (the most significant form of nutritional immunity is the sequestration of nutrient iron). (9) Bacteria can acquire iron from transferrin and lactoferrin. The excess iron increases the virulence of numerous pathogens and in iron overload state . The effect of excess iron on immunological functions include Suppression of monocytemacrophage system, changes in subpopulations of T-lymphocytes, increasing CD8 and suppression of CD4, increased secretion of immunoglobulins and inhibiting the function of the complement system, impaired neutrophil function, impaired monocyte/macrophage activity against pathogens and decreased natural killer cells activity. (10,11,12) The patients after splenectomy, the immune system is modified, Expressed in quantitative changes of lymphocytes without any functional disorders. (13) Splenectomized patients have a significant infection risk, because the spleen is the largest accumulation of lymphoid tissue in the body including splenic macrophages that attack encapsulated organisms (in the absence of spleen, the ability to fight off these pathogens is severely diminished ), spleen is also a major site of early immunoglobulin M production, which is important in the acute clearance of pathogens from the bloodstream. (14) Efficient phagocytosis depends on splenic macrophages and on the production of many substances (opsonins, properdin, tufsin) which are reduced in splenectomized patients , however encapsulated pathogens (Streptococcus pneumoniae, Haemophilus influenza type B, Escherichia coli, Neisseria menigitidis) are the most fearsome. (15) In zinc deficiency, the Zinc is an immunoregulator and low level of zinc in thalassemic patients associated with changes in lymphocyte subpopulations which is adjusted after zinc administration. (16) Repeated transfusions lead to a continuous alloantigen stimulation& disruption of the immune balance in thalassemic patients. (15) Aims of study The study was carried out to: 1. Determine the frequency of UTI among thalassemic patients and Its correlation to risk factors include vaccination, splenectomy, iron overload and Hb level. 2. Determine microorganism causing UTI among thalassemi patients and control group. Patients and Methods A case control study was carried out at the Babylon hereditary blood disease center at Babylon maternity and children hospital From March to October 2018 and the study included 100 febrile thalassemic patients as patients group with age ranged from 6 month to 15 years and mean age 6 ± 2.4 years, classified into (44) males and (56) females, compared to 100 non thalassemic patients as control group, classified in to (43) males and (57) females with mean age of 7±1.2 years, consulted pediatric clinic or attended the emergency unit in the hospital. Inclusion criteria includes Bthalassemic patients (B thalassemia major, intermedia, minor and sickle cell thalassemia) with fever of more than 37.4 0 C, their age ranged between 6 months and 15 years, no history of antibiotic use in the preceding 3 days, no malnourished children or any conditions that increased the risk of UTI as vesico urethral reflux (VUR), obstructive uropathy, diabetes mellitus (DM) and uncircumcised child. The physical examination was initially directed at measuring the axillary temperature for those of less than 6 years with added 0.5 0 C and measured orally to those above 6 years, then complete examination was done including chest, abdomen, tonsils, meningeal signs and skin examination. Clinical diagnosis of UTI was made to those with abdominal pain, dysuria, frequency, renal angle tenderness and supra pubic tenderness. Examination for any signs & symptoms of respiratory distress and Chest X ray (CXR) for those with suspected of pneumonia. A through history and examination with investigations were done to prove or exclude other diagnosis including hepatitis, measles, mumps, otitis media and others. All patients (patients and control group) sent for general urine exam (GUE) & urine culture collected either through midstream urine in sterile universal container in toilet trained patients or through urine bag in infant to those who are not toilet trained( methods of supra pubic aspiration or urinary catheter used are refused by parents). Samples collected after cleaning the genital area with soap and water or povidone iodine then rinsing the area with water. The samples examined within 15 minute of collection and pyuria was considered as positive if pus cell of more than 5 cell/m 3 and Pure colony count of ≥10 5 /ml is considered positive culture. Blood was aspirated and sent for ferritin, Hb, WBC (white blood cell), CRP, ESR. Ferritin was not repeated to those patients measured within last 3 month prior to presentation, while in other patients, it was sent after stabilization of condition (not during the febrile illness). A value of more than 1000 ng/ml is considered a state of iron overload. Hb was classified as low if less than 8.5gm/dl (normal if more than 8.5gm/dl) in low transfusion regimen (in our patients, we used low transfusion regime). Statistical analysis Data was collected and included in a data based system and analyzed by statistical package of social sciences (SPSS version24). discrete variables presented as number and percentage, were analyzed using chi square like comparison between patients and control in presence of urinary tracts infection , continuous variables presented as mean & significance was set at the P ≤ 0.05 level in all analyses. Ethical Issue: informed consent was obtained from patient or patients accompany. The results In this study, it includes 100 febrile thalassemic patients classified into 44 males and 56 females with mean age of 6±2.4 years compared to 100 non thalassemia, classified into 43 males and 57 females with mean age 7±1.2 years . There is statistically significant increased in the frequency of UTI in thalassemic patients (14%), compared to control group (5%) with P. value <0.05 . Pneumonia is the commonest disease occurring in both thalassemia and control group . No statistical significant of UTI among males and females in both groups No statistical significance in distribution of cases according to the type of thalassemia. E. coli is the commonest organism in both groups Table 7: Signs, symptoms and investigations of UTI in patients and control group. The abdominal pain is commonest symptoms among both groups and UTI commonly occurred to those patients with Hb of less than 8gm/dl. Discussion The result of the current study shows that 14 % of patients group have UTI, in comparison to only 5% in control group and is statistically significant ( P value < 0.05). The mechanism for increasing susceptibility to infection are related to anemia , iron overload, altered immune response & splenectomy. This result is higher than the result done by Galia Rahav (17) < 1000 ng/ml ≥ 1000-< 2500 ng/ml ≥ 2500 ng/ml Serum ferritin supra pubic aspiration with variable sensitivity and specificity. Urine bag and midstream urine are methods used in this study resulting in false positive result, in comparison to supra pubic aspiration which used in Galia Rahav (17) which associated with high sensitivity and specificity. Urinary tract infection is forming the 3 rd important cause in febrile thalassemic patients after pneumonia& tonsillitis which account 25% ,16 % respectively. A distinct increased in UTI following splenectomy & incomplete vaccination, in splenectomized patients, there was 37 % (10/27 patients) associated with infection, in comparison to 5.4 % (4/73 patients) in non splenectomized patients. This may be related to spleen importance for the immunological surveillance and reservoir of immune competent lymphocyte, (15) also the antibody response to antigen is impaired , in addition to inefficient phagocytosis (which depend on splenic macrophages & production of many substance like (opsonin , properdin and tufsin)are reduced & impaired chemotaxis. (18) Most of the vaccine that is used for thalassmic patients given against S. pneumonia, N. meningitides & H. influenza, but in the current study, there is increased risk for E. coli among both groups which may be explained by small sample size of the patients and high predilection of E. coli infection into urinary tract. Urinary tract infection increased in patients with iron overload (its increased with increased level of iron overload) and statistically is insignificant, but its percentage is 16.9% (11/65 patients), in comparison to 8.5% (3/35 patients) in non iron overload as iron overload may be related to alteration of chemotactic phagocyte properties of neutrophil resulting in reduction their ability to kill invading pathogen, also the saturated transferrin allow the labile iron to be available for bacteria. (19) The level of Hb and gender are not associated with risk of UTI, even those patients with low Hb (< 8.5 gm / dl) associated with 16.3 %, compared to 10.2 % in those patients with Hb level (>8.5 gm /dl ). The frequency of UTI increased in first 10 years of age, especially between 2-10 years which explained by incomplete vaccine& poor hygiene. Urinary tract infection occurs more in females patients in both groups, attributed to the short female urethra with its close proximity to anal canal allowing easier contamination and ascending of infection. Generally, there was high incidence of UTI among patients with B-thalassemia major, compared to B-thalassemia intermedia, minor and sickle cell thalassemia as risk factor for infection occur more in B-thalassemia major like iron overload, splenectomy & allogenic blood transfusion, while in the current study there was no significant difference among all patients with different types of thalassemia which may be explained by small sample size. Escherichia coli is the commonest microorganism in our study as it is able to attach the bladder wall and produced biofilm that resist the body immune response (20) . The abdominal pain & dysuria are the concurrent symptoms in all patients, while tenderness occur in 50 % and no cases of abdominal mass. This indicate that signs and symptoms of UTI were not specific. The investigation like WBC, ESR, CRP, are negative in patients with UTI, while pyuria occur in most cases of UTI. Conclusion 1) The current study revealed that the frequency of UTI among febrile thalassemic patients is 14 %, and its increased with splenectomy, incomplete vaccination and in cases of iron overload. 2) Abdominal pain and dysuria are the commonest symptoms, while the pyuria occur in 92 % of all cases Recommendations 1) Urinary tract infection is common in febrile thalassemic patients, therefore routine screening is recommended during febrile illness. 2) Encouragement of giving vaccine to all thalassemic patients according to the national program& additional vaccine including E coli vaccine has shown promised in preventing UTI. 3) Splenectomy shouldn`t be done unless strictly indicated. 4) Advice the family to maintain pre transfusion Hb above 8.5 gm /dl with adequate chelation therapy.	2019-05-07T13:29:01.638Z	2019-04-19T00:00:00.000	{ "year": 2019, "sha1": "251794ea87133f039cccd784629f4aab9ac56f4a", "oa_license": null, "oa_url": "https://doi.org/10.18535/jmscr/v7i4.114", "oa_status": "GOLD", "pdf_src": "MergedPDFExtraction", "pdf_hash": "186797ced7fd101b2abd406551d0e4f34dd73df2", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
213327797	pes2o/s2orc	v3-fos-license	Platform-maker for preparing stumps for lowering below the surface of the soil with a variable boom of the hydraulic manipulator The technology of continuous lowering of stumps below the soil surface throughout the area of cutting is proposed to reduce the loads on the working bodies and frames of tillage tools and ensures complete mechanization of the reforestation process. A technological process has been developed to prepare stumps for lowering below the ground surface, which consists in creating platform-deepenings around stumps and cleaning their lateral surface from the soil within the depth of the platform. An aggregate for its implementation consisting of a tractor, a hydraulic manipulator, a screw rotator and a platform-maker with flexible working bodies has been designed. A mathematical model of the aggregate operation has been created. It enables to evaluate the efficiency of the manipulator placement of platform-maker and to determine the influence of the parameters of the hydraulic manipulator on the performance of the aggregate. The influences of the stump distribution density, the boom of the manipulator and the maximum angle of its rotation on the performance of the platform-maker have been determined. It was revealed that the productivity of the aggregate increases with an increase in the density of stump distribution on the cutting and the main geometrical parameters of the hydraulic manipulator. Introduction Existing technologies in forestry, many of which are based on the continuous onward movement of aggregates, have significant drawbacks. They include large energy and material costs. In addition, there is a violation of the natural environment: deep rutting from the propulsion means is removed, the ground cover is removed, the structure of the soil changes, and the undergrowth of valuable wood species is destroyed. The aggregates with hydraulic manipulators transporting the working parts during various technological operations are becoming more common in order to eliminate these drawbacks in logging operations, as well as during reforestation and afforestation. Thereby the technology provides a discrete way to carry them out [1]. Hydraulic manipulators are used in the aggregate with logging machines [2,3], equipment for cutting, crushing and milling stumps [4,5], cutting off of secondary hardwood, forming crowns, collecting seeds, preparing seats and platforms [6,7], digging seedlings [8] and others. It is also proposed to prepare stumps for lowering below the soil surface It consists in creating platformdeepenings around stumps (removing soil around stumps and cleaning them from contamination within the depth of a platform) using a hydraulic manipulator. The continuous lowering of stumps below the soil surface makes it possible to turn the felling into an area that, in many ways, is close to agricultural areas, free from the aerial part of the stumps, with the preserved fertile soil layer. When using this technology, there are opportunities to increase the speed of aggregate movements, reduce the dynamic loads on the working bodies of the tillage machines and, as a result, reduce metal intensity of the structures and material costs for their manufacture, create strictly parallel rows of forest plantations with predetermined row spacing. These enable to mechanize the whole technological process of the cultivation [9]. In addition, this operation will lead to an increase in the volume of energy raw materials. It is obtained as a result of crushing and collecting chips, which, unlike other traditional fuels, is renewable and the safest material for the environment. Removing the soil around the stump, for example, to a depth of 15-20 cm and a diameter of about 1.5 m, depending on the type of wood, will increase the raw material by 40-70%. The purpose of the study is to determine the dependencies of the platform-maker performance to prepare stumps for lowering below the soil surface on the overhang and angular range of the hydraulic manipulator boom, as well as on the stump distribution density in the cutting area. Materials and methods 2.1. The design and technological process of the platform-maker to prepare stumps for lowering below the soil surface A special aggregate was developed to prepare stumps for lowering below the soil surface ( figure 1). It includes a tractor 1, a telescopic hydraulic manipulator 2, a screw rotator 3, and a platform-maker for creating platform-deepenings around the stumps 4. The platform-maker (figure 2) consists of frame 1, rotary working bodies (rotors) 2, and two-piece hydraulic cylinder 3 for changing the distance between them, hydraulic motors 4 for driving the working bodies, hydraulic system and controls. The rotor is a shaft on which four sections are installed, consisting of disks 5 with fragments of cable 6 attached to them, designed to create platformdeepenings around the stumps and clean their side surfaces from the soil and other contaminants. In order to eliminate the rapid wear of the cables, their ends are hardened and covered by collars 7. The technological process of preparing stumps for lowering below the soil surface is as follows. The platform-maker is mounted above the stump with the help of a hydraulic manipulator and lowered to contact with the soil, after which the rotors are set in rotation by means of hydraulic motors, and the cables contacting the stump clean it. The profile of the platform-deepening (figure 3) -a depth of 15-20 cm and a radius of 60-90 cm -is provided by reversing rotation of the rotator, carried out simultaneously with the rotation of the rotors. After completion of the operation, the device moves to another stump, and the process repeats. Depending on the diameter of the stumps, the distance between the rotors is adjusted using a two-piece hydraulic cylinder. After cleaning the stumps can be crushed below the soil surface by machines for lowering, crushing and milling the aerial part of the stumps [10]. The use of a hydraulic manipulator for positioning the area of the distributor significantly improves the performance of the aggregate compared to the option of mounting the machine directly on the mounted mechanism of the tractor. The hydraulic manipulator reduces the number of crossings of the tractor on the felling due to the fact that at each stop of the aggregate it is possible to create platformdeepenings for -10 stumps (depending on the density of their distribution). Simulation of the technological process of the aggregate for the preparation of stumps to lower below the soil surface with a variable boom of the hydromanipulator To evaluate the effect of the manipulator placement of the platform-maker and determine the dependence of its performance on the parameters of the hydraulic manipulator, a mathematical model of the operation of the aggregate for cutting is developed. The proposed model allows to calculate the number of tractor moves on cuttings with different stump density distribution (from 200 to 1200 stumps/ha). Modeling is performed in the twodimensional space XY, and the modeling plane is a "top view" of the cutting. Initially, a model cutting is generated with a given density of stump distribution σ st (figure 4). The main objective of the simulation is to determine the number of moves tractor N t when processing a specified number of stumps N st . To solve this problem, the model simulates the movement of the tractor along a straight line of the rut and periodically checks the stumps surrounding the aggregate into the working area of the hydraulic manipulator, which is a sector of the ring and is determined by the range of the boom angles φ 1 -φ 2 and the range of its departure R 1 -R 2 (figure 5). In accordance with the design scheme, the condition for entering the i-th stump into the working area of the hydromanipulator can be written as follows: where R 1 , R 2 -minimum and maximum outreach of the boom of the manipulator; φ 1 , φ 2 -limit angles of rotation of the boom of the manipulator; х t , у t -coordinates of the manipulator column; R i and φ i -polar coordinates of departure and angle of rotation of the boom of the hydraulic manipulator for the i-th stump in the coordinate system associated with the axis of rotation of the rotary column (x t , y t ). At the beginning of the computer experiment, a model tractor with a hydraulic manipulator and a platform-maker is placed on the model cutting so that the working area of the hydraulic manipulator does not extend beyond the cutting area L x × L y . After counting the number of stumps that have fallen into the working area, the aggregate is moved along the OX axis to such a distance that the stump does not pass past the working area in the treated strip. Thus, the model reproduces the move of the tractor from one processing position to another, resulting in a gradual shift of the working area. In this case, each time the number of stumps caught in the work area is calculated. When reaching the total number of treated stumps, equal to some given value N max (in most experiments N max = 1000 stumps, so that the random error is small enough), computer experiment is completed, and the total processing time where N t -number of tractor moves; t t -time of moving the tractor from one working position to another, including the time of movement, time of fixation and time of removing the fixation of the tractor, time of preparing the operator for controlling the hydraulic manipulator, time of assessing the location of the stumps surrounding the aggregate and choosing the optimal sequence of their processing; N st -specified number of stumps; T st -processing time of one stump, including the time of positioning, lowering and raising of the platform-maker. The most important indicator determined by the model is productivity P of the platform-maker, hat is, the number of stumps processed per unit of time: . (4) The developed model makes it possible to find out how the productivity of the platform-maker with manipulator positioning depends on the basic geometrical parameters of the hydraulic manipulator: the minimum and maximum boom R 1 and R 2 and limiting angles of its rotation φ 1 and φ 2 . To study the model, a special computer program was written in Object Pascal in the programming environment Borland Delphi 7. Figure 6 shows examples of the operation of the aggregate: the sequence of processing areas is shown with the aggregate gradually moving along the rut line. In the case of a small number of stumps per unit of cutting area (up to 200-400 stumps/ha) ( figure 6, a), the platform-maker at each stop of the tractor processes approximately 1-3 stumps and makes long journeys (about 10 m). With a larger number (600-800 stumps/ha) ( figure 6, b), the number of stumps processed during one stop increases to 3-5, and crossings between stops are reduced to 5 m. Since the productivity of the platform-maker increases with an increase in the density of stump distribution at the cutting site ( figure 7, a), it can be concluded that the hydraulic manipulator positioning of the area of the landlord is reasonable at high density (more than 600 stumps/ha): in this case, it is possible to process from 3 to 5 stumps at each stop of the unit. With a low density (less than 400) stumps are located far from each other, and from one tractor stop it is rarely possible to process more than one stump. Therefore, in this case, it is advisable to place the platform-maker directly on the tractor hitch mechanism. Results With an increase in the maximum boom reach R 2 the productivity (calculated for σ st = 600 stumps/ha) increases almost linearly ( figure 7, b). Therefore, it is advisable for a platform-maker to choose a hydraulic manipulator with the boom as large as possible, but with simultaneous consideration of the mass and cost parameters, as well as the power characteristics of the tractor. An increase in the angular range of boom rotation leads to an increase in the working area of the manipulator, and, accordingly, to an approximately linear increase in productivity (figure 7, c). Therefore, it is advisable to install the hydraulic manipulator on the tractor so that the angular range of boom rotation is as large as possible. Conclusions Thus, the dependencies of the performance of the platform-maker for preparing stumps for lowering below the soil surface on the departure and angular range of the boom of the hydraulic manipulator, as well as on the density of stump distribution at the cutting site, are determined. It has been established that the manipulator placement of the platform-maker is necessary with 600 or more stumps per 1 hectare of logging. Otherwise, it is advisable to place it on the rear hitch of the tractor. In turn, an increase in the departure and angle of rotation of the boom of the hydromanipulator causes an approximately linear increase in the productivity of the platform-maker.	2019-11-28T12:36:28.645Z	2019-11-21T00:00:00.000	{ "year": 2019, "sha1": "a28f2230ef793dc3fbeeeedb9e5454505083d0b7", "oa_license": null, "oa_url": "https://doi.org/10.1088/1755-1315/392/1/012059", "oa_status": "GOLD", "pdf_src": "IOP", "pdf_hash": "52ec0550d8f7b5ab079ca1083265227691956533", "s2fieldsofstudy": [ "Engineering" ], "extfieldsofstudy": [ "Geology", "Physics" ] }
216365648	pes2o/s2orc	v3-fos-license	Effect of Photoanode Modification on Charge Transport, Recombination and Efficiency of Dye Sensitized Solar Cells Using Synthetic Organic Dyes The DSSCs (Dye-sensitized solar cells) with photoanode using different sizes of particulate TiO2 (18 nm, 30 nm and 200 nm) have been fabricated. The synthetic organic dyes (LEG4) was used as photosensitizer. The present work mainly investigates the influence of TiO2 photoanode modification on light absorption, charge transport and carrier recombination which are then correlated to the device efficiency measured under AM1.5 solar irradiance. The DSSCs with photoanode using larger 200-nm-TiO2 as light scatterer outperform other devices without larger TiO2. It is attributed to an increase in harvesting photons (by UV-vis spectroscopic measurement) via light scattering, smaller ideality factor (by dark current analysis), thus lower recombination possibility, lower charge transfer resistance and longer electron lifetime (by electrochemical impedance spectroscopy) which results in longer electron diffusion and higher charge collection efficiency. The photoanode modification in DSSCs has a strong impact on optical and charge transport properties, and eventually on the photovoltaic performance. Introduction  The DSSC (dye-sensitized solar cell) is a photo electrochemical solar cell using redox mediator to transport electrons and become an attractive alternative to the widely used Si-based solar cell due to its low processing cost, ease of fabrication, moderate efficiency [1][2][3][4]. In order to improve power conversion efficiency of DSSCs, many attempt shave been devoted to modification of the DSSC components such as photoanode, electrolyte, photosensitizer dye molecules and counter electrode [5]. Among them, the photoanode modification is of crucial importance in enhancing device efficiency since it works as both photon harvestor (by photosensitizer dye) as well as charge transporter (by n-type semiconductor) [6][7][8]. Mesoporous nanoparticulate TiO 2 semiconductor is a commonly used photoanode material for its relatively high energy conversion efficiency, high specific surface area [9], non-toxicity and low cost [10]. However one major constraint of this material is its random electron transport which would lead to charge carrier recombination in the device [11]. In addition, due to its small size, it is transparent to visible light, weak in visible light scattering and high recombination of photo generated electron-hole pairs [12]. The charge recombination, the electron loss in the device, is one of the principal factors limiting power conversion efficiency [13]. To address the charge recombination issue and achieve good scattering performance, several strategies have been utilized over the years. For instance, integration of photonic crystals and larger sized metal oxide nanostructures at photoanode which increases the absorption path-length of photons via D DAVID PUBLISHING Effect of Photoanode Modification on Charge Transport, Recombination and Efficiency of Dye Sensitized Solar Cells Using Synthetic Organic Dyes 31 light scattering [14], integration of plasmonic metal nanostructures in TiO 2 which offer light concentration and better light scattering via a phenomenon of surface plasmonics, and doping of metal or non-metal into TiO 2 which would change the sub-band states, modulating the band gap [15,16]. Many attempts on photoanode modification in DSSCs and associated efficiency improvements are widely acknowledged [17,18]. The rationales of efficiency enhancement via modifying the photoanode using different sizes of particulate TiO 2 in terms of current leakage, charge transport and carrier recombination have not yet widely been explored. In the present work, the photoanodes of DSSCs were fabricated using different sizes of particulate TiO 2 so as to study the effect of modifying photoanode on light absorption, current leakage, charge transport and charge recombination resistance, eventually on the device efficiency. Experimental Section The FTO (fluorine doped tin oxide) coated glass substrates were ultrasonically cleaned with detergent, alcohol and deionized water followed by nitrogen blow and UV exposure for 15 min to ensure full dryness and remove organic contamination on the substrate. The mesoporous TiO 2 films were screen-printed on FTO substrates using TiO 2 paste (Dyesol DSL 30 NRD-T) with different nanoparticle sizes of 18 nm, 30 nm and 200 nm. We prepared six different multilayer TiO 2 samples, namely (i) 18 The absorption spectra of dye loaded TiO 2 films were recorded using UV-Vis spectrophotometer (Thermo Scientific, Genesys 10S) in the wavelength range of 300-900 nm. The dye sensitized solar cells (DSSCs) were assembled by introducing the redox electrolyte (0.05 M/I 2 and 0.5 M/LiI in acetonitrile) between the dye loaded TiO 2 electrode (photoanode) and platinum counter electrode. The current-voltage (I-V) characteristics of DSSCs were measured at room temperature by using a source meter (Keithley 2400) and solar simulator (Newport/Oriel). The AM 1.5 illumination of intensity 1000 W/m 2 was calibrated using a standard Si solar cell certified by NREL (National Renewable Energy Laboratory). The I-V characteristics were also recorded in dark for extraction of ideality factor. EIS (Electrochemical impedance spectroscopic) measurement was also performed in dark to analyze the charge transport resistance and recombination lifetime in the DSSCs using the potentiostat/Galvanostat electrochemical analyzer (Corrtest: CS350). The impedance spectra were analyzed using the Z view software. The frequency range is 0.01-100 kHz and the magnitude of AC signal is 10 mV. Results and Discussion First, we studied the optical absorption of LEG4 dye loaded on different TiO 2 scaffolds [18 nm (4 m), 18/18 nm (7 m), 18/200 nm (9 m), 30 nm (4 m), 30/30 nm (7 m) and 30/200 nm (9 m)]. The values in the parenthesis represent the thickness of TiO 2 films. Their optical absorption spectra are shown in Fig. 1. For all samples, the absorption peak is centered at 475 nm which is the characteristic of LEG4 [19]. It is observed that the optical absorption increases with increasing TiO 2 layer thickness which is attributed to increased dye surface area (higher dye loading) with thicker TiO 2 [20]. The absorption enhancement is more pronounced in the 18/200 nm and 30/200 nm samples which is ascribed to the back light scattering by larger TiO 2 nanoparticles (200 nm) thereby increasing the path length of photons [21]. Next, we used the LEG4 dye/TiO 2 as photoanode in DSSCs and studied the effect of photoanode modification on the power conversion efficiency of DSSCs. Fig. 2 shows the I-V characteristics of DSSCs with six different TiO 2 photoanodes under AM 1.5 solar irradiance. The photovoltaic device parameters extracted from the I-V curves are listed in Table 1 photoanode is contributed from higher values of all device parameters (V oc , J sc and FF) with major contribution from J sc enhancement. This photocurrent enhancement would be due to an effective back scattering of light by larger TiO 2 nanoparticles (200 nm), which effectively increased dye absorption [22] evidenced by the absorption spectra (Fig. 1). The light scattering at TiO 2 scaffolds is thus an important phenomenon influencing the performance of DSSCs. It is also found that further increase in thickness of 200-nm-TiO 2 negated the efficiency of devices (the data is not shown here). Following the determination of device efficiency, charge recombination and leakage current in DSSCs upon modifying the photoanode were also investigated using dark current analysis. The informative parameters such as ideality factor (n), reverse saturation current (J o ), series resistance (R s ) and shunt resistance (R sh ) were extracted from the dark J-V curve. The device can be simply described by a one-diode equivalent model and the current flow across the device is represented by the general diode equation taking into account the series and shunt resistances. where, J o is the reverse saturation current density, q is the electronic charge, V is the applied voltage, A is the effective diode area, n is the ideality factor, k is the Boltzmann constant, T is the absolute temperature, R s is the series resistance and R sh is the shunt resistance. In dark, the photocurrent density (J ph ) is zero. Fig. 3 reveals the ln(J)-V characteristics in dark of DSSCs with 18 nm TiO 2 . In the plot [ln(J) vs. V], the three regions can be identified. Region I is a linear region at reverse voltages and low forward voltages where current is limited by shunt resistance R sh (leakage current). Region II exhibits an exponential behavior at immediate forward voltages where current is controlled by a diode. Region III is a linear region at high forward voltages where the current is limited by series resistance R s . Obviously, the R s and R sh can be extracted from the region I and III respectively. The slope and the extrapolated-to-zero-volt point of region II gives the ideality factor (n) and J o respectively [23]. The extracted parameters are listed in Table 2. The ideality factor (n) reflects the quality of interfaces within device and recombination behaviors at that interfaces. Enhanced disorder or defects induced recombination loss typically results in higher n values, deviating from the ideal diode behavior [24] The n value is 3.0 and 2.6 in the DSSC with 18 nm and 30 nm TiO 2 photoanode respectively. Upon modifying the photoanode with larger TiO 2 (200 nm), the n decreased to 1.7 and 1.3 respectively in the DSSCs with 18/200 nm and 30/200 nm photoanode. It is indicative of reduced charge recombination and consequently the V oc increased in these devices [25]. The reverse saturation current density (J o ) is a leakage current flowing across the charge collecting electrodes in the device. As seen in Table 2, decreased J o with larger TiO 2 suggests less current leakage in these devices. Also observed are the lower R s and higher R sh in DSSCs with larger TiO 2 which would result in higher FF in these devices. Electrochemical impedance spectroscopy (EIS) is one of the most important tools to elucidate the charge transfer and transport processes in various electrochemical systems including DSSCs. Fig. 4 (a, c) There are two semicircles in EIS spectra; the small one at high frequency fitted to a charge transfer resistance (R ct1 ) and chemical capacitance (C 1 ) is assigned to electrochemical reaction at the redox electrolyte/platinum interface while the larger one at low frequency fitted to a charge transfer resistance (R ct2 ) and chemical capacitance (C 2 ) is assigned to charge transfer reactions at the TiO 2 /dye/electrolyte interface and the accumulation/transport of the injected electrons within TiO 2 film. The total impedance of the DSSC is the sum of the impedance at the Pt and TiO 2 electrodes, and the impedance due to diffusion of tri-iodide in the electrolyte. These data were analyzed with a non-linear least square fitting program using Z view software. The resulting impedance parameters listed in Table 3 show that the DSSCs with TiO 2 scattering layer (18 nm/200 nm and 30 nm/200 nm) exhibits the lower values of R ct1 and R ct2 which indicates the more efficient charge transfer process at the interfaces of electrolyte/Pt and TiO 2 /dye/electrolyte. The Bode phase plots of EIS spectra (Fig. 4b, d) display the characteristic frequency peaks of the charge transfer process at different interfaces which are listed in Table 3. The characteristic frequency (f) is related to the inverse of electron lifetime (τ e ) [ e = 1/(2f max ) where, f max is the maximum frequency of the mid-frequency peak]. The electron lifetimes are longer with TiO 2 scattering layer, 24.2 ms with 18 nm/200 nm photoanode and 39.7 ms with 30 nm/200 nm photoanode. Longer electron lifetime could favor the electron transport through a longer distance which lowers the recombination resistance (R rec ) in the device [26]. Conclusions The DSSCs with a multilayer structure of TiO 2 using different nanoparticle sizes (18 nm, 30 nm and 200 nm) as photoanode were fabricated and characterized. It was found that the DSSCs with larger TiO 2 (200 nm) as scattering layer rendered a relatively higher device efficiency which is correlated to the optical and charge transport properties and charge recombination in the device. Optical absorption enhancement with TiO 2 scattering layer is due to increased dye loading in the TiO 2 and back light scattering by larger TiO 2 particles. Dark current analysis revealed that ideality factor and reverse saturation current are lower with TiO 2 scattering layer which indicate the low current leakage and charge recombination probability. Electrochemical impedance spectroscopy measurement indicated the lower charge transfer resistance suggesting high recombination resistance and longer electron lifetime which would result in higher charge collection and photocurrent generation in the device. To sum up, relatively higher device efficiency with photoanode using layer TiO 2 as scattering layer is attributed to increased light absorption, favorable charge transport and lower charge recombination in the device. The photoanode modification is critical in modulating the optical and charge transport properties and photovoltaic performance of DSSCs.	2020-03-05T10:38:41.878Z	2020-02-28T00:00:00.000	{ "year": 2020, "sha1": "3f0dae5a29382f6da81181b86bf8ab2d45855f6e", "oa_license": null, "oa_url": "https://doi.org/10.17265/2161-6213/2020.1-2.004", "oa_status": "GOLD", "pdf_src": "Adhoc", "pdf_hash": "32664f08fde415379ea704c9940e11265dd4b749", "s2fieldsofstudy": [ "Chemistry" ], "extfieldsofstudy": [ "Materials Science" ] }
214113884	pes2o/s2orc	v3-fos-license	Efforts to Improve Rhythmic Gymnastics Learning Outcomes Through The Audio Visual Media The purpose of this study was to determine the increase in rhythmic gymnastics learning outcomes through audio-visual media. The method used is CAR (Classroom Action Research). CAR has two cycles, namely in the first cycle making action plans, carrying out actions, observations and reflections. The results of reflection are used as a basis for the planning phase of the second cycle if the first cycle is not successful. So in the second cycle gives more learning through Audio Visual Media and gives more direction to students where errors are often more guided while in class before going back to practice in the field. The research instrument uses a portfolio and observation sheets. Data analysis techniques use percentages. The conclusion that through audio-visual media can improve rhythmic gymnastics learning outcomes. Keywords— Learning Outcomes, Rhythmic Gymnastics, Audio Visual I. INTRODUCTION According to Supandi (1992: 1) "Physical education is a process of systematic interaction between students and the environment that is managed through effective and efficient physical development towards the formation of a complete human being". Physical education, sports, and health are media to encourage physical growth, psychological development, motor skills, knowledge and reasoning, appreciation of values (attitude-mental-emotionalsportsmanship-spiritual-social), and habituation of healthy lifestyles that lead to stimulate the growth and development of a balanced physical and psychological quality. Suwarjo, et al (2009). Rhythmic gymnastics is gymnastics accompanied by musical rhythms, songs, or beats and prioritizes harmony in motion. Rhythmic gymnastics should follow the rhythm that accompanies it. Rhythmic gymnastics is a blend of dance and sports movements. The results of the researchers' initial observation interview regarding student learning outcomes in gymnastics lessons, apparently there are still many low grades. It was proven that there were still many students who did not understand how to do the right rhythmic gymnastic movements. Researchers do the learning process through audio-visual media in the form of movie shows (videos) or in the form of music used. This can encourage solving problems in doing rhythmic gymnastic movements. Purwanto (2008: 46), learning outcomes represent the achievement of educational goals in students who take part in the teaching and learning process. Learning outcomes are actual, meaning that the realization of educational goals is achieved, so the learning outcomes measured are highly dependent on the educational objectives. Husdarta and Saputra (2000: 4) suggest that the main task of the teacher is to create a climate or atmosphere so that the learning process takes place in the classroom. In the field, the main characteristic of the learning process is that students can be actively involved in the learning process. Teachers must always strive so students are motivated to play more roles. Nevertheless the teacher still functions as a manager of the learning and learning process. To support the implementation of learning in school, the creativity of a teacher is needed, so that the learning process can provide a good learning experience in full to students. This phenomenon is a problem due to the lack of ability of an education teacher to take advantage of his role as a teacher who has potential in accordance with the demands of curriculum targets and absorption and as creative educators in activating physical education learning in schools. The formulation of the problem of this research is how to improve rhythmic gymnastics learning outcomes through audio-visual media and know the Improvement of Rhythmic Gymnastics Learning Outcomes through the Audio Visual Media. II. METHOD Researchers use the Classroom Action Research (CAR) method or Classroom Action Research. The action is given by the teacher or with direction from the teacher done by students. III. RESULTS AND DISCUSSION The results of research on efforts to improve rhythmic gymnastics learning outcomes through audio-visual media can be seen in the following picture: Based on preliminary data, that the ability of students in rhythmic gymnastics learning is still relatively low. Of the 27 students who were sampled in this study, it turns out that only 10 students (37.03%) have mastery learning, while the remaining 17 people (62.96%) students have not reached mastery learning. On the results of the first cycle test of 27 students there were 16 students (59.25%) who had gotten the level of completeness, while 11 students (40.74%) had not reached the level of mastery learning. The percentage value of completeness of the minimum average grade obtained only reached 72.11%. While the results of the second cycle test obtained 22 students (81.48%) have achieved mastery learning while 5 students (18.51%) who have not met mastery learning. The average grade obtained reached (81.85%). The implementation in this second cycle, in general, went well and was conducive, because the completeness of classical learning (81.85%) had been achieved so rhythmic gymnastics learning through audio-visual media stopped. The following is a discussion of the results above: 1) Results of Rhythmic Gymnastics Learning Activities carried out by the Teacher -In arranging the class the teacher is good -Guidance given by teachers to students is evenly distributed -The teacher evaluates movements that were previously difficult for students to make -Teacher's skills in using media are good 2) Results of Rhythmic Gymnastics Learning Activities carried out by Students -High curiosity of the continuation of the material -Have discussions according to the worksheet -Already understand about things that have not been understood before -When the attitude of implementing a movement that was difficult to do before, namely the stretching and core movements, was well done 3) Reflection Based on the results of the observation, the following is described the failure and success in implementing the actions in the second cycle, namely: 1. Most students are able to do rhythmic gymnastics 2. Students' knowledge in rhythmic gymnastics learning has expanded. 3. The teacher has been able to motivate students to be active in teaching. 4. Involvement of student activities in doing movements is good. 5. On the attitude of the implementation of the stretching and core movements are very well done. 6. During the learning progress the student activity has increased. IV. CONCLUSION An increase in rhythmic gymnastics learning outcomes through audio visual media.	2020-01-09T09:17:58.372Z	2019-01-01T00:00:00.000	{ "year": 2019, "sha1": "edabe4968c735936e3001cd11106948582b93681", "oa_license": "CCBYNC", "oa_url": "https://www.atlantis-press.com/article/125928438.pdf", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "251bc7e0d1ac915909402bff36529e2a9e54a7f0", "s2fieldsofstudy": [ "Education" ], "extfieldsofstudy": [ "Psychology" ] }
11568255	pes2o/s2orc	v3-fos-license	Psychological Effects of False-Positive Results in Expanded Newborn Screening in China Objectives As more families participate expanded newborn screening for metabolic disorders in China, the overall number of false positives increases. Our goal was to assess the potential impact on parental stress, perceptions of the child's health, and family relationships. Methods Parents of 49 infants with false-positive screening results for metabolic disorders in the expanded newborn screening panel were compared with parents of 42 children with normal screening results. Parents first completed structured interview using likert scales, closed and open questions. Parents also completed the parenting stress index. Results A total of 88 mothers and 41 fathers were interviewed. More mothers in the false-positive group reported that their children required extra parental care (21%), compared with 5% of mothers in the normal-screened group (P<0.001). 39% of mothers in the false-positive group reported that they worry about their child's future development, compared with 10% of mothers in the normal-screened group (P<0.001). Fathers in the false-positive group did not differ from fathers in the normal-screened group in reporting worry about their child's extra care requirements, and their child's future development. Children with false-positive results compared with children with normal results were triple as likely to experience hospitalization (27%vs 9%, respectively; P<0.001). Conclusions The results showing false-positive screening results may affect parental stress and the parent-child relationship. Parental stress and anxiety can be reduced with improved education and communication to parents about false-positive results. Introduction Expanded newborn screening (NBS) using tandem mass spectrometry (MS/MS) to identify more than 30 biochemical genetic disorders is an important advance in early disease detection. It has greater sensitivity than past screening methods and allows for presymptomatic detection and identification of metabolic disorders [1]. However, expanded newborn screening has led not only to an increase in positive identifications but also to a dramatic increase in the overall number of out-of-range results, of which the majority are confirmed to be false positives after further testing [2]. Generally these results are not laboratory mistakes but rather are transient findings or indications of variant or carrier status. The current overall risk of a false positive result for expanded NBS in the United States is estimated to range between 1/1,500 to 1/3,600 [3][4]. False-positive screening results have been associated with increased anxiety and stress in parents of infants who require follow-up testing, even after the infant's good health is confirmed [5]. Studies also reported long-term negative effects including alterations in perceptions of their infant's health, an increase in the number of emergency room visits, and hospitalizations for the infant [5][6][7][8]. This report firstly describes china parents' responses to falsepositive newborn screening results among a cohort of children born after January 1, 2008, when voluntary expanded newborn screening began in Beijing. The psychological effects of falsepositive have not been studied in Chinese population before. Enrollment and Study Procedures Mothers and fathers of children with false-positive newborn screen results, defined as the initial result being abnormal or inconclusive for any of 35 biochemical disorders, were invited to participate after a referral was made for additional confirmatory testing. The authors contracted with a screening center, Center for Clinical Laboratory Development, Chinese Academy of Medical Science, which conducts newborn screening for more than 80% of birthing hospitals in Beijing, to recruit and interview parents of infants with false-positive newborn screen results. This was a comparative cross sectional study. Participants were enrolled by this laboratory between 2008 and 2009. Parents of children with false-positive results were sent a recruitment letter with a reply paid envelope more than 6 months after the diagnosis of a metabolic disorder had been ruled out. This inviting letter included, a short questionnaire (2 copies), two written informed consent (study purpose, methodology of the protocol, risks, direct and indirect potential benefits, the right to withdraw, duration of participation, possibility of alternative treatments, voluntariness) [9], and a letter of explanation the study. Parents of twins and triplets were sent a single letter. Parents who did not ''opt out'' by returning a response card indicating a preference not to be contacted were called to participate in a telephone interview [8,10]. Although both parents were invited, it was acceptable if only 1 parent participated. The comparison group for the false-positive cohort consisted of parents of 6-to 12-month-old children with normal screening results, selected randomly from the screening center database.The storage NBS card contain the date of birth, birth weight, parents' names, birth hospital, and address. All recruitment occurred between January 1, 2008, and December 31, 2009. All participants completed the study questionnaire once. Additional follow-up interviews were not included in the study design. Exclusions included parents of children who had died and parents of newborns whose birth weight was 2500 g. or gestation less than 32 weeks [10][11]. The latter exclusion avoided recruitment of parents of premature newborns, who frequently experience transient initial newborn screening abnormalities. Approval for this study was obtained annually from the institutional review board of Center for Clinical Laboratory Development, Chinese Academy of Medical Science and Peking Union Medical College. Written informed consent was obtained from all participants involved in our study. By design, the number of false-positive participants exceeded the number of participants in the normal group. Statistical methods that did not require balanced sample sizes were selected. Data Collection Instrument Complete details of data collection have been reported previously [10][11][12][13]. Parents responded to a structured questionnaire (study instrument) using likert scales and closed and open questions. This sought to determine from the parents (a) whether they would have the test performed again for another child, (b) in the first 6 months of life, whether [they] had ever taken [their] child to inpatient, and the dates of such visits; how many times did your babies visit a primary care physician, and the dates of such visits, (c) whether they knew the screening could lead to falsepositive result when they took part in expanded newborn screening, (d) whether their child required extra parental care, (e) whether they fear that their child might be developmentally delayed, or experience a false-positive screening result as a significant threat to the child's well-being, (f) reasons for repeat screen. Parents provided short answers or ratings on a 5-point Likert scale. The questionnaire also assessed sociodemographic factors (age and parity of the mother, level of education, income, and their site of residence) and one open ended question, ''what change can be made in the expanded screening process''. The same interview was given to both groups of parents, but only parents in the false-positive group were asked questions about (f). Parents next completed the parenting stress index (PSI), short form [8,14]. This is a 36-item questionnaire that provides a total stress score and 3 subscale scores, namely, parental distress, parent-child dysfunctional interaction, and difficult child. The normal range for total stress scores is 55 to 85, with scores of 85 considered to be in the clinical range in which treatment may be necessary. The PSI also provides a defensive responding index, which is an internal index of validity based on the parent's responses. Scores of 10 for this index indicate that the validity of the total stress and subscale scores is questionable [8,14]. All of the items in the PSI used in our sample were translated into Chinese by the first author and back translated into English by a professional translator. The first author followed the strictest translation procedure: back translation, informal interviews, pretest, and item analysis, to ensure cultural equivalence. The Chinese version of the PSI had high reliability and predictive validity. The cronbach's a coefficient in this study was 0.92. Data Analyses Statistical Package for Social Science, SPSS version 15.0 (SPSS Inc., Chicago, IL, USA) was used for data entry and analysis. Descriptive statistics such as means and standard deviation (SD) for age of mothers and frequency and percentages for categorical variables (such as race, gender of neonate, number of child, family income) were determined. The characteristics of children, parents, and families in the false-positive group were compared with those in the normal-screened group by using the student's unpaired t-test for continuous and scale variables and fisher' exact test for dichotomous variables. Student's unpaired t-test was also used to compare the PSI scores between different groups. The number of hospitalizations occurring before 6 months of age was compared for the 2 groups by using Poisson regression. For the PSI, subjects who failed the defensive responding index (scores of 10) were dropped from the analyses. The result of open ended questionnaire was analysis by frequency and percentages. All P values were2-sided, and values of ,0.05 were considered significant [8]. Sample The sample included parents of 49 children with false-positive newborn screen and 42 children with normal newborn screen results. A total of 88 mothers (47 false-positive and 41 normalscreened) and 41 fathers (23 false-positive and 18 normal-screened) were interviewed. For 38 infants (21 false-positive and 17 normalscreened), both parents responded. The number of enrolled families divided by the number of families contacted determined the participation rates, which were 48% for the false-positive group and 42% for the normal-screened group. As noted in Table 1, the false-positive group was similar to the comparison group in terms of parent age, gender, birth order, ethnicity, marry. In the false-positive group, children were older at the time of evaluation (mean: 12.4 months, SD: 3.2 months) compared with the normal screened group(mean: 6.7 months; SD: 1.2 months; P,0.001). The false-positive group was of lower economic or education status, compared with the normal-screened group (P,0.001). In additional, according to parental report, the median age of the infant's diagnosis was confirmed was 21 days (range: 7-94 days). Parental Stress and Parent-Child Relationship Although parents in the 2 groups reported both worry about their child's health, more mothers in the false-positive group reported that their children required extra parental care (21%), compared with 5% of mothers in the normal-screened group (P,0.001). Thirty-nine percent of mothers in the false-positive group reported that they worry about their child's future development, compared with 10% of mothers in the normalscreened group (P,0.001). Fathers in the false-positive group did not differ from fathers in the normal-screened group in reporting worry about their child's extra care requirements, and their child's future development. In additional, thirty-seven percent of parents in the false-positive group reported that they child have visited a primary care physician in the first 6 months of life, while 15% parents in the normal-screened group (P,0.001). The child hospitalizations during the first 6 months of life among falsepositive group was 0.27, compared with 0.09 in the normalscreened group (P,0.001). As shown in Table 2, mothers in the false-positive group reported higher overall stress on the PSI than did mothers in the normal-screened group. 17% of mothers in the false-positive group (n = 8) but no mothers in the normal-screened group scored in the clinical range. The differences between groups were more pronounced on the total score, parent-child dysfunctional interaction subscales, and difficult child subscales than on the parental distress subscale. Fathers in the false-positive group also registered higher overall stress on the PSI than did fathers in the normal screened group, especially on the total score (P = 0.01), and difficult child subscales (P,0.001). Parental Knowledge to the Screening Process As shown in Table 3, 55% (n = 26) of mothers and 50% (n = 11) of fathers knew the correct reason for their child needing a repeat screen. Mothers in the false-positive group who knew the correct reason for the repeat screen reported lower stress levels on the PSI, including the total score, difficult child subscales, and parent-child dysfunctional interaction subscales (Table 4). Fathers who knew a correct reason for the repeat screen did not exhibit lower stress levels on the PSI (P = 0.09). Parents in the false-positive group reported a lower tolerance of false-positive results. Ten (14%) parents in the false-positive group reported that they would not have the test performed again for another child, while no parents in the normal-screened group. Discussion Expanded newborn screening programs have expanded dramatically in the past decade. The main risks are related to false positive results and results with ambiguous implications for treatment-risks. Our results indicate that a false-positive result from an expanded newborn screening test can induce some parents to experience stress and affect parents' perceptions of their child's health, and the parent-child relationship. This finding is expressed by parents' higher overall stress on the PSI, more than four times required extra parental care, 3 times the number of children hospitalized, longer hospital stays in the false-positive group compared with the normal screen group. False-positive screening results have been associated with increased anxiety and stress in parents of infants who require follow-up testing, even after the infant's good health is confirmed. The true impact of false-positive newborn screening tests is just beginning to be well described. Early screening programs for phenylketonuria (PKU) showed poor parental understanding of false-positive results and a tendency for parents of such children to perceive their children as medically vulnerable [15]. Studies suggest that some parents of these infants remain anxious about their child's health, perceive the child as unhealthy, and, as a consequence, treat the child differently even after a result is deemed a false-positive finding [4,[16][17]. More than one third parents still have concerns about the health of their infant, on average, mothers report more stress [4]. These findings are consistent with our study that has shown 39% mothers experience a false-positive screening result as a significant threat to the child's well-being. Other studies on the impact of acute illnesses among children identified the ''vulnerable child syndrome'' [18][19].Those studies applied the vulnerable child syndrome to include (1) a condition or even a ''non-disease'' (eg, false-positive result) in a child, (2) parents who misinterpret hat condition or its sequelae, and (3) parents who exhibit sustained unjustifiable anxiety about the child's vulnerability to future events [8]. More recent research has demonstrated associations between false positive results and mothers' perceptions that their children with false positive results require increased parental care, and a trend towards increased hospitalization [5].Studies have led to speculation that a false-positive result would be associated with increased outpatient and inpatient health care utilization in early childhood [20][21][22][23]. This might explain a trend toward an increase in infant hospitalizations during the first 6 months of life among false-positive children in our study (mean: 0.27 hospitalization VS 0.09 hospitalization; P,0.001). Earlier studies have found an association between false-positive newborn screen results and negative psychosocial effects [24]. This association was also documented in studies screening for hearing, cystic fibrosis, diabetes [25][26][27][28]. We hypothesized that such psychosocial effects could lead to parents perceiving children with false-positive test results as vulnerable which in turn may lead to increased health care utilization. ''Our results are consistent with this hypothesis, despite other research reporting the contrary [12][13]. Studies have demonstrated that education of parents about false-positive results is lacking [29]. It is clear that even after routine NBS testing, a significant proportion of parents are confused about the meaning and reasons for repeat testing after an initial NBS test [8,10]. Similarly, in our study, 48% parents of children with false-positive results did not know the correct reason for their child's follow-up testing. It may be that physicians do not communicate the false-positive result to families, all the knowledge about false positive come from parents themselves. This idea is supported by the fact that, as detailed elsewhere, in our study sample 28% parents reported that they didn't know false-positive when they began participate in screening. Researchers have consistently shown providers' ability to communicate about newborn screening is poor in both training and primary care settings [30][31][32][33]. Nonetheless, in a recent study of paediatricians in Massachusetts, 42% were less than comfortable talking about newborn screening test results with families [32]. The results of Hewlett's review suggest that parental stress and anxiety can be reduced with improved education and communication to parents, specifically at the time of follow-up screening [5]. Physicians seem to be able to reduce parents' stress if they provide information about the process (as well as the false-positive results) of newborn screening, estimate the risk to the infant as low, or refer parents for additional information. Similarly, in our study, mothers who knew the correct reason for their child's repeat screening test experienced less total stress than did mothers who did not know. In additional, parents in our study suggested ways in which the process could be improved to reduce the influence of false-positive results, including provision of more information about newborn screening and false-positive results. They also suggested that providers should provide clearer explanations of the reasons of repeat screening. Table 3. Parents response to reason for repeat screen in false-positive group. This study has a number of limitations. The disparity in the children's ages between the false-positive and normal-screened groups could have biased results. However, the PSI used to measure parental stress, is considered age independent for small increments of age. It is possible that our study design, in which participants were interviewed 6 months after the resolution of the false-positive screening result, may not have fully captured the stress and anxiety experienced during the waiting period. In additional, sample sizes were small for both respondent groups. The samples were also geographically limited, potentially limiting the generalizability of our results. In the process of our study, the response rate was 45%, lower than other reported [10,[33][34][35][36]. This rate means that the finding should not be overinterpreted. These experiences relate largely to a single survey in one center. It cannot be assumed that they will apply to other center or population. This study suggests that false-positive screening results may affect parental stress. This is especially true for parents who have not received adequate information about newborn screening. Therefore, Parental stress and anxiety can be reduced with improved education and communication to parents about falsepositive results. Higher scores indicate higher stress; only PSI scores for subjects whose defensive responding index was .10 were included in the analysis [16]. b Student's unpaired t-test. doi:10.1371/journal.pone.0036235.t004	2014-10-01T00:00:00.000Z	2012-04-27T00:00:00.000	{ "year": 2012, "sha1": "9b370eb76be9c9e7b7a9870cf54499c241a0de1a", "oa_license": "CCBY", "oa_url": "https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0036235&type=printable", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "9b370eb76be9c9e7b7a9870cf54499c241a0de1a", "s2fieldsofstudy": [ "Medicine", "Psychology" ], "extfieldsofstudy": [ "Medicine" ] }
267370468	pes2o/s2orc	v3-fos-license	Mandibular Fracture following Dental Implant Protocol: Clinical Report and One-Year Follow-Up : Bone fractures following mandibular dental implant protocols associated with diagnosed osteoporosis are rare in the literature. We present a case in which a 55-year-old male patient with no previous medical history presented to the emergency department with pain in the left mandibular parasymphysis and gingival bleeding. Clinical examination revealed crepitus, mandibular mobility, and clinical signs of localized fracture and infection. Further radiographs conﬁrmed a mandibular fracture in the region of the alveolus of tooth 34 and four implants placed in the mandible as part of the patient’s immediate implant protocol. The infection developed into osteomyelitis, which was treated with a combination of antibiotics. After the infectious process had been eradicated, new complementary tests were carried out, which revealed that the patient had osteoporosis. Oral rehabilitation treatment and calcium replacement were carried out under specialist medical supervision. The importance of proper planning and clinical assessment of the patient is discussed, and the proposed long-term management of the case has been carried out. Introduction The etiology of mandibular fractures has always been associated with facial trauma, polytrauma, physical aggression, falls, and tumours [1][2][3][4][5][6][7][8].With the evolution of dental treatments and new technologies, dental implant surgery has become one of the best treatment options for oral rehabilitation when one or more teeth are missing.Since 1985, when the first long-term clinical studies were published, the treatment of totally edentulous patients with implant-borne prostheses has been gaining ground in the academic and clinical community [9][10][11].The so-called overdenture has two or more implants, which are placed in the mandible and/or maxilla and connected to a structural component that is connected to a laboratory-manufactured or 3D-printed total prosthesis [12,13].With advances in technology and understanding of the osseointegration process, more and more immediate treatments are being proposed, where the prosthesis is fitted on the same day or within a few weeks after the surgical placement of the dental implants.Also known as the "immediate implant protocol", this procedure has become the standard for fully edentulous patients, especially in the mandible. Mandibular fractures immediately following dental implant treatment are rare in the literature [14][15][16], and it is important to discuss and better understand when they occur.The limited literature on the subject always points to the atrophic mandible associated with age-related bone loss, peri-implantitis, and iatrogenesis as the causes of these fractures [15].Mandibular fractures with chronic bone infection, known as osteomyelitis, are difficult to treat and carry a higher risk of local and systemic complications that disrupt and prolong any proposed treatment [17,18].Particularly in cases of oral rehabilitation where bone healing is desired, such as dental implant surgery, the presence of a localised infectious process with necrotic bone detachment makes such treatment impossible.Several stages of surgical treatment are necessary, depending on the organic response to these proposals, until the final treatment is completed.In this paper we describe a complex case of fracture of the left mandibular parasymphysis (Figure 1) with osteomyelitis and submandibular suppurative drainage.Treatment of the osteomyelitis was performed in two stages (surgical and clinical): removal of the bone fragments in the fracture, debridement of the soft tissues and curettage of the necrotic hard tissues in the area, rigid internal fixation of the mandible with 2.0 mini-plates, and antibiotic treatment for 2 weeks.The patient was evaluated every 3 days until successful treatment of the osteomyelitis was confirmed.Given the complications that may be associated with this type of fracture, a DEXA scan was requested to assess the patient's risk of osteoporosis, which was confirmed.The patient was referred for specialist medical treatment and calcium replacement.Two months later, with no signs of infection and mandibular continuity restored, oral rehabilitation was performed with a prosthesis on the 3 remaining implants and occlusal adjustment.The patient was followed up for one year with no signs of recurrence or complaints about the proposed treatment. Surgical planning for dental implants must be thorough and always consider the patient's general condition [19].Assessment of the cause of tooth loss is essential for the proposed rehabilitation to ensure longevity and a better chance of success.Following the diagnosis of osteoporosis, the patient's panoramic radiograph (Figure 2) showed diffuse radiolucent images, suggesting a high degree of bone involvement and resorption in this unusual case of a man under 60 years of age.The anterior mandibular bone image was normal for the patient's age, with no limiting atrophy, making it even more difficult to determine the local bone density in advance.As the patient reported that extractions and immediate implant placement were performed during the dental implant surgery, the diagnosis of bone quality via surgery is limited by the smaller number of perforations required during the surgical procedure. The proposed treatment to eradicate the osteomyelitis, fix the mandibular fracture, and rehabilitate the occlusion was uneventful and well accepted by the patient, who preferred not to have further implants at time of writing.One-year follow-up revealed a good buccal condition with normal function and acceptable aesthetics. Case Report A 55-year-old male patient was admitted to the Unimed Hospital in Sao Carlos, Sao Paulo, Brazil, as an emergency patient on 28 December 2022.He reported having undergone dental implant surgery (four implants in the mandible for immediate prosthesis placement) in November 2022.A full denture was fitted over the implants one week later.He reported severe pain in the fracture area on the day the prosthesis was fitted.In December of the same year, the pain increased.Physical examination revealed gingival bleeding in the region of implant related to tooth 34, a left submandibular abscess with fistula and suppuration.Crepitus and mobility suggestive of fracture were noted, as well as pain on palpation and mouth opening.The complete blood count showed an elevated white blood cell count (11.1 million red blood cells per microliter of blood), confirming osteomyelitis, which was also clinically diagnosed via signs and symptoms for more than 3 weeks, as well as X-rays showing bone damage. Treatment planning can be divided into two parts: clinical and surgical.Both are complementary and depend on intensive monitoring to assess the evolution of the infectious process (osteomyelitis).Urgent therapy was started immediately in the emergency department due to localized pain and a localized abscess (Figure 1a).Antibiotic therapy was started with a combination of amoxicillin and clavulanic acid 875 g 8/8 h continued for 15 days.Mouthwash every 4/4 h (malvona solution) for the entire duration of the proposed treatment.A CT scan of the mandible and craniofacial region was obtained (Figure 3).The abscess was drained extra-orally and Penrose drains were placed to ensure maximum removal of necrotic material to facilitate organ response.During the drainage of the abscess, material was collected for culture and antibiogram in case the proposed antibiotic treatment failed, thus speeding up the identification of the causative agents of the infection.Analgesics were administered (metamizole 1 g 8/8 h 5 days), and the patient was discharged for outpatient follow-up.The Penrose drain was removed after 5 days. Three weeks after abscess drainage and antibiotic therapy, the patient had no clinical signs of infection, and the wound was not communicating with the oral environment.Laser therapy [wavelength: 660 nm (red laser) and 808 nm (infrared laser), MMO Optics, Sao Carlos SP, Brazil] was applied twice a week for the duration of the remaining treatment.Routine preoperative tests were ordered for the management of the pathological mandibular fracture, with the addition of a diagnostic test for osteoporosis (dual energy X-ray absorptiometry, DEXA) given the patient's fracture.DEXA confirmed the diagnosis of severe osteoporosis (a T-score greater than −2.5), which had not been observed prior to the previous implant surgery.The operation to treat the mandibular fracture was planned under general anaesthesia and naso-tracheal intubation with monitoring of vital parameters.The patient was placed in a supine position under aseptic and antiseptic conditions.Surgical drapes were placed to isolate the surgical field from contamination, leaving the left side of the mandible exposed.A regional anaesthetic block was applied to the mandibular inferior alveolar nerve with lidocaine 2% adrenaline (1:200,000) and local infiltration of the surgical area.A submandibular incision was made, following the facial lines, and dissected in planes until the fractured mandible was fully exposed.The necrotic tissue was debrided and removed, leaving vivid bleeding at the margins of the fractured segments of the mandible (Figure 4).The fracture was manually aligned and repositioned using stabilizing clamps, and a 2.0 mini plate was adapted to the anatomy of the mandible for fixation.Titanium screws included in the 2.0/2.4 mm locking system kit (Toride Company Ltd., Mogi Mirim, Sao Paulo, Brazil) were used to fix and stabilize the mandibular fracture.Tissue was repositioned and sutured in planes.The skin was sutured with 3.0 nylon sutures as a strong wound closure was required to avoid recontamination of the deep bone planes.The operation was uneventful, and the patient was taken to the post-anaesthesia room for 4 h and then returned to his bed.He was discharged from hospital the following day and continued to be monitored on an outpatient basis. The patient was treated with prophylactic antibiotics (amoxicillin and clavulanic acid 875 mg 8/8 h for 7 days) and the sutures were removed after 7 days with no evidence of recurrent infection.Three months after surgery, oral rehabilitation began with a prosthesis on the three remaining implants.A transfer mould was used to fabricate the metal structure in the laboratory to support the overdenture and also to adjust the occlusion with the upper dental arch with permanent teeth.Careful occlusal adjustment is essential, especially considering the force distribution of the natural teeth on the weakened mandible with osteoporosis and the presence of dental implants in the region.A less rigid material (composite resin) was used to fabricate the prosthetic teeth in order to better distribute the mechanical chewing forces.After the prosthesis was fitted, weekly occlusal adjustments were made until a good distribution of contact points was achieved with maximum intercuspation and mandibular laterality (Figure 5b).The patient was followed up for one year without any abnormalities, complaints of pain, or parafunctions. Results The results obtained were satisfactory considering the complexity of osteomyelitis treatment and the potential complicating treatment factors, particularly in cases of oral microbiota contamination.The patient's fracture was adequately aligned and immobilized without compromising occlusal rehabilitation with the prosthesis on the implants.It was decided not to place new implants in the region, mainly due to the patient's lack of enthusiasm for this treatment option.The patient was informed of the risks and benefits of the proposed treatment and gave his informed consent for his case to be published.The case was followed for one year postoperatively (Figure 5b) with no complaints or complications.There were no patient-reported symptoms or signs of paraesthesia at any time during treatment. Discussion The clinical case presented in this article discusses basic concepts and principles that all surgeons should be familiar with.It also examines an atypical situation of a mandibular fracture associated with osteomyelitis and osteoporosis in a male patient under the age of 60, following an immediate dental implant protocol in a fully edentulous mandible.The fracture in the left parasymphyseal area of the mandible may be related to osteoporosis, bone perforation for implants, and the general location of the mental foramen, which is usually posteriorized in totally edentulous individuals [20].Considering the anatomical features discovered during the surgical procedure, it is likely that there was a vestibular misalignment of the implant at this site, as shown in Figures 3 and 4. The adaptive bone response at the base of the mandible can also be seen in these images, altering its typical anatomy.The presence of a drainage and extraoral fistula, despite the presence of intraoral communication, suggests that the fracture occurred in microcracks during the drilling and screwing of the dental implant for fixation [21][22][23]. It is known that osteoporosis associated with bone loss due to tooth loss or edentulism and the corresponding alveolar bone are factors that contribute to the weakening of the mandibular resistance and the consequent increased risk of fracture [24,25].A better understanding of the fracture risks in immediate dental implant protocols enriches the safety and efficacy of treatment and provides more information that allows better control of the short and long-term outcomes of osseointegrated implants.The occurrence of localized infection and its development into chronic infection in cases of implant treatment demonstrates the need for a shorter follow-up period between consultations in these cases, in addition to the patient's own clarification of the necessary personal care [26,27]. The authors were unable to find a study describing a case of a patient with both osteoporosis and osteomyelitis.It was found that by removing the source of infection (implant at the fracture site, necrotic soft and hard tissues), reviving medullary bleeding in the mandible, repositioning, and immobilisation with rigid internal fixation, it was possible to treat the osteomyelitis and fracture simultaneously with the proposed treatment without major complications and without the need for prolonged antibiotic treatment.It is thought that the lower bone density did not compromise blood flow, allowing for continuous and appropriate antibiotic dosing to the infected area-thus allowing for permeability [28,29].Treatment of mandibular fractures in totally edentulous patients can be performed without an intermaxillary block, as the fine occlusal adjustment is made during prosthesis fitting.The use of the Miniplate 2.0, used in cases with a greater need to withstand mechanical forces, was determined based on the absence of the intermaxillary block, which guarantees the immobility of the bone segments in the presence of a poor fit between them due to resorption caused by chronic infection and low bone density, avoiding as much as possible the need for a second surgical treatment [30]. Clinical evaluations throughout the treatment period revealed a tendency for plaque to form around the implant healers.The probable cause of the loss of the natural teeth was a trauma to the occlusion, aggravating the periodontal disease, and the process was repeated with the dental implants, where the jaw was weakened by the vestibular insertion of the implant in the region of the alveolus of tooth 34 and the low bone density favoured the occurrence of fractures after the immediate placement of the implant prosthesis [14][15][16].There is also the possibility that the fracture was pathological, with periimplantitis contaminating the deep tissues due to the absence of the vestibular (external) cortical bone wall (Figures 3 and 4), and that the fracture occurred at the most fragile site due to the low bone density and high masticatory force of the natural teeth in the maxillary arch.In this case, the diagnosis would be a pathological fracture due to iatrogeny, which cannot be ruled out or confirmed because the patient was not followed by the same team during the initial rehabilitation and immediately after the fracture.Implant-supported overdentures conserve bone because of their positive load-related effect on the jaw, so keeping the implants in place and actively loading the overdenture improved the long-term results.Making general practitioners and specialists aware of the possibility of diagnosing osteoporosis through routine dental X-rays, such as panoramic X-rays (Figures 6 and 7), will increase the chances of preventing these diseases from becoming more serious and allow more accurate treatment decisions to be made [31,32].To assess osteoporotic changes in the oral cavity, several investigators have proposed indices derived from panoramic radiographs, such as the mandibular cortical index, the mandibular panoramic index, and the Wical and Swoope technique (Figure 8).All of these techniques assess the basal region of the mandible, posterior to the mandibular foramen, because this area meets the requirement for less inter-and intra-individual variation in anatomy, shape, bone structure, and function.This would allow dentists to correctly refer patients for specific treatment and also to assess the patient's bone quality, which is of great importance when planning dental implants [33,34].Identifying asymptomatic people at risk of osteoporosis in general dental practice could help to reduce the incidence of first fractures, secondary fractures, medical costs, and mortality associated with osteoporotic fragility fractures.Assessment of the cause of the loss of natural teeth, the relationship of occlusal forces between the dental arches based on the presence or absence of natural teeth and/or dentures, and fine occlusal adjustment are fundamental to the short-and long-term success of oral rehabilitation.The patient's general condition contributes to and determines the progress of the proposed treatment.The importance of the clinician managing the complications of mandibular fractures and mastering the occlusion and overall oral pathophysiology has already been established in order to achieve excellence in treatment.The management of craniofacial trauma is a multidisciplinary field and should therefore be considered whenever necessary for the benefit of the patient.The need for continuing professional development is undeniable, as is the need for further research in this area.Although this is a case report, the combined concepts described here may assist in the planning of future cases and highlight the need for further research into the effects of severe infectious processes in patients with osteoporosis. Conclusions An accurate initial emergency diagnosis, confirmed and supported as definitive by complementary examinations, is the fundamental factor in selecting and planning the emergency treatment that will provide the best biological conditions for achieving the best results in the face of the proposed definitive treatment.The attention paid to the accurate initial diagnosis of osteomyelitis in the presence of a pathological fracture following the fixation of dental implants and prostheses in the mandible made it possible to choose a simple and effective emergency therapy, consisting of draining the localized abscess, cleaning the wound, antibiotic therapy, taking a culture, and an antibiogram.The proposed definitive treatment, once the infection had been eliminated and the wound properly closed extra-and intra-orally, allowed the necrotic tissue to be removed through an extra-oral submandibular surgical approach and the mandibular segments to be prepared to restore the continuity of the mandible bone after the fracture had been repositioned and fixed with a rigid mini-plate and screws.Oral rehabilitation with an implant-borne prosthesis provided the function and aesthetics desired by the patient, and the one-year follow-up demonstrated the success of the proposed treatment. Figure 1 . Figure 1.(a) Clinical view of the patient on the emergency room; (b) intrabuccal view of the three remaining implants and the bleeding area of tooth 34, where the fracture and osteomyelitis occurred. Figure 2 . Figure 2. Panoramic X-ray after treatment before prosthesis placement. Figure 3 . Figure 3. Craniofacial CT scan showing: (a) parasymphysis fracture and discontinuity of bone tissue due to osteomyelitis progression and absence of vestibular (external) cortical bone tissue; and (b) coronal view of the mandible, implants, and fracture with precise location and size. Figure 4 . Figure 4. Surgical procedure; (a) exposure of the pathological fracture and removal of necrotic tissue; and (b) alignment of the fractured segments and fixation of the fracture with 2.0 mini plate. Figure 6 . Figure 6.DEXA exam confirming diagnosis of osteoporosis.The bone density of the femoral neck is 0.681 g/cm 2 and the T-score is −3.0, which is 27% less bone density than a person of the same age and weight. Figure 7 . Figure 7. Radiographs: (a) panoramic radiograph showing diffuse radiolucent areas that may go unnoticed on routine examination, and (b) CT scan showing suggestive images of mandibular osteoporosis. Figure 8 . Figure 8.The width of the mandibular cortex (between the parallel lines) is measured on both sides of the panoramic radiograph at a point below the mental foramen indicated by the arrow.	2024-02-02T16:21:58.880Z	2024-01-26T00:00:00.000	{ "year": 2024, "sha1": "1db5b3aef4d96460939595bd9c77a5ceaeea5659", "oa_license": "CCBY", "oa_url": "https://www.mdpi.com/2673-4036/4/1/1/pdf?version=1706277982", "oa_status": "GOLD", "pdf_src": "ScienceParsePlus", "pdf_hash": "67cc2e4a14f4fd8158cf7a0273add8f6992aa215", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [] }
248323528	pes2o/s2orc	v3-fos-license	Thermoelectric Inks and Power Factor Tunability in Hybrid Films through All Solution Process Thermoelectric (TE) materials can have a strong benefit to harvest thermal energy if they can be applied to large areas without losing their performance over time. One way of achieving large-area films is through hybrid materials, where a blend of TE materials with polymers can be applied as coating. Here, we present the development of all solution-processed TE ink and hybrid films with varying contents of TE Sb2Te3 and Bi2Te3 nanomaterials, along with their characterization. Using (1-methoxy-2-propyl) acetate (MPA) as the solvent and poly (methyl methacrylate) as the durable polymer, large-area homogeneous hybrid TE films have been fabricated. The conductivity and TE power factor improve with nanoparticle volume fraction, peaking around 60–70% solid material fill factor. For larger fill factors, the conductivity drops, possibly because of an increase in the interface resistance through interface defects and reduced connectivity between the platelets in the medium. The use of dodecanethiol (DDT) as an additive in the ink formulation enabled an improvement in the electrical conductivity through modification of interfaces and the compactness of the resultant films, leading to a 4–5 times increase in the power factor for both p- and n-type hybrid TE films, respectively. The observed trends were captured by combining percolation theory with analytical resistive theory, with the above assumption of increasing interface resistance and connectivity with polymer volume reduction. The results obtained on these hybrid films open a new low-cost route to produce and implement TE coatings on a large scale, which can be ideal for driving flexible, large-area energy scavenging technologies such as personal medical devices and the IoT. S-3 dev. of 0.7 µm), and Bi2Te3 exhibited average particle size about 200 nm (with std. dev. of 95 nm). In both the cases the thickness of the platelets is around 50-70 nm. Crystallite size was estimated from the XRD data, using Williamson-Hall plot, in the range 160-600 nm, with a weighted average of about 200 nm for Sb2Te3 , while it is in the range of 50-100 nm with a weighted average of 70 nm for Bi2Te3. Due to the anisotropy in the particle morphology the crystallites are also expected to exhibit some anisotropy, the obtained size being viable in the lateral plane. Percolation Theory As the system shows significant value of a percolation threshold, we are inclined to consider that direction/model. The critical exponent in a percolation system is assumed to depend only on the dimensionality of the lattice and is independent of the details of the lattice structure. 1 Namely, for 3D systems the critical exponent acquires values between 1.6 and 2.0 and, for 2D systems, the critical exponent varies between 1.1 and 1.3. 2 However, in conducting systems the critical exponent does not follow the universal trend. 3 In systems based on insulating matrices with embedded conductive fillers and where the conduction process is controlled by interfiller tunneling, the inverted Swiss-cheese model can be applied, which predicts values of the critical exponent between 0.8 and 1. 2 An example is found in a work published by Wang et al. 4 in which S-5 carbon nanotubes were introduced into a transparent PVDF matrix. In his case, the critical exponent has a value of 0.85, similar to that obtained by us when we use Bi2Te3 nanoparticles. The system can be considered 2D since the ratio between the thickness and the width of the film is very small. After the new measurements of electrical conductivity using the 4-point technique, the percolation curve settings have been updated in Figure 4 and the values of the critical exponent are close to 1. When the percolation threshold of Sb2Te3 and Bi2Te3 hybrid films with the DDT linker are compared, the values obtained for Sb2Te3 are lower than that of Bi2Te3. As the percolation threshold depends on the particle size, the results suggest that Bi2Te3 nanoparticles are smaller than Sb2Te3. This was confirmed by the SEM micrographs of as-made Sb2Te3 and Bi2Te3 nanoparticles presented in Figure S3. In order to demonstrate the viability of this suggested outcome, we chose Sb2Te3 -PMMA system for control experiments. We synthesized smaller Sb2Te3 platelets, by using the same synthetic process and only lowering the concentration of the precursors to half. DLS measurements are performed on the suspensions of previous and new batch of Sb2Te3 nanoparticles in isopropanol. Results are presented in Figure S4a, where the peak average dispersed size (by approximation to the volume of a sphere) of smaller sample is about 200 nm lower than the large ones used for the percolation study. Thereafter, a hybrid film is developed by using this and adapting the best performing film composition with 60% Sb2Te3 in the PMMA matrix, without the addition of DDT. The hybrid film with larger platelets showed a resistance of 600 W, while the films with smaller platelets showed about 600 kW ( Figure S4b), which is about three orders of magnitude higher. This finding confirms the predictions S-6 by the percolation theory, which allows the design of platelets of various lateral size to systematically study this correlation as a further research work. Figure S5. Flexibility test as a function of a) bending cycles and b) bending radius. Flexibility Tests As a proof of concept, two flexibility tests have been carried out on Sb2Te3 -PMMA hybrid film with 60% Sb2Te3 content and DDT linker. For this, the deposition of the ink has been applied to a flexible PET substrate. Figure S5a shows the results of (2.5 cm long hybrid film on PET substrate) bending the film 3000 times on a 2 cm diameter cylinder and measuring the change in electrical conductivity after every 100 bendings. The conductivity of the film gradually decreases with the number of flexes until reaching a loss in electrical conductivity of 50% after 3000 flexes. On the other hand, Figure S5b shows the variation of electrical conductivity as a function of the bending radius. In this case, the electrical conductivity also decreases as the bending radius decreases and a loss of electrical conductivity by 70% is reached when the bending radius is as small as 1 cm. These results clearly indicate that formulated hybrid films based on Sb2Te3 (and Bi2Te3) nanoparticles with DDT and in the PMMA matrix are not particularly flexible. However, we must remember that a film composed solely of Sb2Te3 (or Bi2Te3) nanoparticles deposited on a flexible substrate such as PET would almost completely lose electrical conductivity after a few bending cycles, since in this case there would be no glue effect of the polymeric matrix. In addition, we must remember that the polymeric matrix, PMMA, used in this work is a matrix that is rigid at room temperature since its glass transition temperature, Tg, is around 110 ºC. 5 Therefore, the developed hybrid-films will only have a S-7 flexible behavior when the working temperature is higher than 110 ºC since it is at this temperature when the polymeric matrix begins to be more fluid and, therefore, can be handled without breaking. With the choice of other flexible polymers as the matrix a higher flexibility could be achieved, with much less degradation of the transport performance. Transport Property Evaluation The electronic transport properties of the hybrid films were determined by the measurement of electrical conductivity ( ) and the Seebeck coefficient (S). The was determined by the Van der Pauw equation (Eq. S1), inserting four equidistant contacts of conductive silver paint on the surface of the films ( Figure S6). Next, a current was applied between two points and the potential between the other two points was measured, obtaining R1. To obtain R2, we applied a current intensity between two other points and the potential between the remaining two was measured. Knowing the values of R1, R2 and the thickness (d), the electrical resistivity ( ) of the film was obtained, which was then converted to electrical conductivity. (b) Voltage as a function of the current to determine R1 and R2. The S was determined at room temperature using a homemade system consisting of two copper blocks (Figure S7). One of the copper blocks was heated by a Peltier module while the other was kept at room temperature. The temperature values at the hot and cold end of the S-8 samples were recorded using two K-type thermocouples connected to PicoLog software. The hot-side temperature was gradually increased to reach a gradient of 60 °C by applying a voltage to the Peltier module with a source supplier (Keithley 2280S). The S potential generated was recorded with a Keithley 2450 source meter. By plotting the S voltage generated as a function of the temperature gradient, a linear distribution was obtained, where slope is the S. Figure S7. (a) Scheme of the home-made set up to measure the Seebeck coefficient. (b) Voltage generated as a function of the temperature difference, where the sloop is the Seebeck coefficient. In order to optimize the thermoelectric films, the influence of the film thickness on the electrical conductivity was studied by varying the number of layers deposited on the substrate. Figure S8 represents the study carried out previously to the percolation curve for the hybrid films containing 60wt% Sb2Te3 or Bi2Te3 nanoparticles. It was observed that for both the Sb2Te3	2022-04-23T06:22:57.673Z	2022-04-22T00:00:00.000	{ "year": 2022, "sha1": "fdbb2ecddf9d82279bda07a30466b46ac99d5213", "oa_license": "CCBY", "oa_url": null, "oa_status": null, "pdf_src": "PubMedCentral", "pdf_hash": "bfc189e427cc7958132ae784ac0f03ad4eefdf4a", "s2fieldsofstudy": [ "Materials Science" ], "extfieldsofstudy": [ "Medicine" ] }
221979273	pes2o/s2orc	v3-fos-license	Chimeric Virus-Like Particles and Capsomeres Induce Similar CD8+ T Cell Responses but Differ in Capacity to Induce CD4+ T Cell Responses and Antibody Responses Despite extensive research, the development of an effective malaria vaccine remains elusive. The induction of robust and sustained T cell and antibody response by vaccination is an urgent unmet need. Chimeric virus-like particles (VLPs) are a promising vaccine platform. VLPs are composed of multiple subunit capsomeres which can be rapidly produced in a cost-effective manner, but the ability of capsomeres to induce antigen-specific cellular immune responses has not been thoroughly investigated. Accordingly, we have compared chimeric VLPs and their sub-unit capsomeres for capacity to induce CD8+ and CD4+ T cell and antibody responses. We produced chimeric murine polyomavirus VLPs and capsomeres each incorporating defined CD8+ T cell, CD4+ T cell or B cell repeat epitopes derived from Plasmodium yoelii CSP. VLPs and capsomeres were evaluated using both homologous or heterologous DNA prime/boost immunization regimens for T cell and antibody immunogenicity. Chimeric VLP and capsomere vaccine platforms induced robust CD8+ T cell responses at similar levels which was enhanced by a heterologous DNA prime. The capsomere platform was, however, more efficient at inducing CD4+ T cell responses and less efficient at inducing antigen-specific antibody responses. Our data suggest that capsomeres, which have significant manufacturing advantages over VLPs, should be considered for diseases where a T cell response is the desired outcome. Despite extensive research, the development of an effective malaria vaccine remains elusive. The induction of robust and sustained T cell and antibody response by vaccination is an urgent unmet need. Chimeric virus-like particles (VLPs) are a promising vaccine platform. VLPs are composed of multiple subunit capsomeres which can be rapidly produced in a cost-effective manner, but the ability of capsomeres to induce antigen-specific cellular immune responses has not been thoroughly investigated. Accordingly, we have compared chimeric VLPs and their sub-unit capsomeres for capacity to induce CD8 + and CD4 + T cell and antibody responses. We produced chimeric murine polyomavirus VLPs and capsomeres each incorporating defined CD8 + T cell, CD4 + T cell or B cell repeat epitopes derived from Plasmodium yoelii CSP. VLPs and capsomeres were evaluated using both homologous or heterologous DNA prime/boost immunization regimens for T cell and antibody immunogenicity. Chimeric VLP and capsomere vaccine platforms induced robust CD8 + T cell responses at similar levels which was enhanced by a heterologous DNA prime. The capsomere platform was, however, more efficient at inducing CD4 + T cell responses and less efficient at inducing antigen-specific antibody responses. Our data suggest that capsomeres, which have significant manufacturing advantages over VLPs, should be considered for diseases where a T cell response is the desired outcome. INTRODUCTION The annual mortality rate of malaria is currently estimated at 405,000 people of whom 67% are children under 5 years of age (1). Eliminating the causative Plasmodium spp. parasite will likely require an effective vaccine (2) but conventional sub-unit vaccines strategies have thus far proved to be suboptimal. The identification and development of vaccine delivery platforms which induce long-lasting robust cellular and antibody immune responses is a global health priority. A specific goal is a vaccine against the pre-erythrocytic (sporozoite/liver) stage of Plasmodium sporozoites which would prevent both the clinical symptoms which develop during the blood stage, and the transmission of the diseases which occurs during the sexual stage. The most advanced malaria vaccine candidate, RTS,S (also known as Mosquirix TM ), is a virus-like particle (VLP) comprising multiple copies of the P. falciparum circumsporozoite protein (Pf CSP) B cell repeats and some CD4 + and CD8 + T cell epitopes fused with recombinant hepatitis B surface antigen (RTS) and co-expressed with free hepatitis B surface antigen (S) (3), coadministered with AS01 adjuvant (4). However, although early clinical studies showed some protective efficacy in the first year after vaccination, it is now established that RTS,S induced protection is low and wanes quickly (5,6). Significant research efforts have been directed at either improving the vaccine platform, or incorporating additional antigens to broaden the protective immune response. Additionally, ease and cost of manufacturing is an important consideration. VLPs are formed when recombinant viral structural proteins assemble into a highly repetitive array which resembles the native form of the virus. These units contain no genomic material so are incapable of replication, but they are highly immunogenic and particularly good at inducing antibody responses without the need for additional adjuvants (7)(8)(9). VLP vaccines that have been licensed include hepatitis B virus and human papillomavirus where protection is mediated through neutralizing antibodies (10,11). In those VLPs, the entire structural protein comes from the pathogen itself and their structure resembles the cognate native virus targeted by the vaccine. However, instead, a generic VLP independent of the pathogen target can be used to produce a chimeric VLP with antigenic epitopes inserted into regions of a flexible carrier virus, thereby making a vaccine platform which can theoretically target any organism [reviewed in (12)(13)(14)]. Chimeric VLPs have been shown to be effective at inducing robust antibody responses (12)(13)(14), but their efficacy in generating cellular responses has not been comprehensively investigated. In vitro studies have shown that chimeric hamster polyomavirus and SV40 VLPs incorporating T cell epitopes from mucin 1 and influenza, respectively, can induce activation of epitope-specific CD8 + T cells (15,16). Further evidence of CD8 + T cell induction was reported using hamster polyomavirus in a mouse study which demonstrated in vivo clonal proliferation of transferred epitope-specific CD8 + T cells, tumor growth inhibition and protection from lymphocytic choriomeningitis virus (17). We recently reported that immunization of mice with murine polyomavirus incorporating defined T cell epitopes derived from the P. yoelii CSP antigen (PyCSP) induced robust CD8 + T cell responses as well as high antibody titres, but poor CD4 + T cell responses (18). Murine polyomavirus VLPs are generated when VP1 structural proteins form pentameric capsomeres (19) which can then be chemically induced in vitro to self-assemble into a highly repetitive VLP containing 72 capsomeres (7,9,20). The crystal structure of the VP1 protein has been previously described (21) as well as the predicted structures of chimeric capsomeres and VLPs with the Group A Streptococcus J8 peptide epitopes (7). The formation of VLPs from the subunit capsomere components adds time and cost to vaccine production (22). Also, the introduction of foreign epitopes could interfere with the structural formation of the VLPs. These shortcomings could be overcome if the chimeric capsomeres themselves were sufficiently immunogenic so that the subsequent VLP production steps were not required. To address this, our colleagues constructed a truncated MuPy-VP1 which prevented the formation of VLPs, and further modified the protein with the addition of multiple epitope insertion sites to increase the resultant antigen to vector ratio (8). These truncated capsomeres were, however, less effective at inducing of antibody responses and required coadministration with adjuvants (7,8,23) including nanoparticles such as silica, poly (D,L-lactic-co-glycolic acid) (PLGA) and poly caprolactone (PCL) (23,24). In another study, it was shown that antibody responses induced by capsomeres could be enhanced to levels similar to those of VLPs without adjuvants, but this required a 20-40 times increase in the capsomere dose (7,25). However, IFN-γ cellular responses induced by capsomeres were similar to those of VLPs, even at low doses in a HPV16 L1 model (25). Additionally, both HPV16 L1 VLPs and their component capsomere induced robust cytotoxic CD8 + T cell responses which were capable of tumor regression in the absence of adjuvants (26). Herein, we extended those studies to comprehensively compare chimeric murine polyomavirus VLP and capsomere constructs incorporating defined CD8 + (280−288) (27) and CD4 + (59−79) (28) T cell and B cell repeat epitopes (29) from PyCSP for capacity to induce robust CD8 + and CD4 + T cell responses and antibody responses. In addition, we evaluated the potential benefit of co-administration of poly(I:C) adjuvant with VLPs, and of including a heterologous DNA prime-boost regimen. For plasmid DNA immunizations, plasmid DNA encoding full-length PyCSP (pVR2516) or empty vector without insert (pVR1020; Vical Inc, CA, USA) were commercially purchased from PureSyn Inc. (Malvern, PA, USA). Plasmid Construction The plasmid pGEX-4T-1 (GE Healthcare Biosciences, UK) with the murine polyomavirus VP1 sequence (accession number M34958) was obtained from the Protein Expression Facility (PEF, University of Queensland). For the VLP platform, the VP1 sequence was modified by PEF by inserting an AfeI restriction enzyme site flanked with Glycine 4 -Serine linker sequences at position 293; this was designated pGEX-VP1-S4-G4S (7). For the capsomere platform, the VP1 sequence was truncated to remove the first 28 and last 63 amino acids, and restriction enzyme sites were inserted at positions 28, 85, 293, and 380; this construct was designated VP1 N C (8). Human codon-optimized PyCSP CD8 + (280−288) and CD4 + (59−79) T cell epitopes and the B cell repeat epitope sequences were individually inserted into the AfeI site in pGEX-VP1-S4-G4S and into positions 28, 293 and 380 of VP1 N C using standard molecular biology techniques and constructs were confirmed by Sanger sequencing. Expression and Purification of Chimeric Capsomeres and VLPs Wild-type pGEX-VP1-S4-G4S and VP1 N C, or the chimeric constructs detailed above, were separately transformed into chemically competent E. coli Rosetta DE3 pLysS bacteria (Novagen, CA, USA). The GST-tagged VP1 proteins were expressed by bacteria culture in Terrific Broth and expression induced using 0.2 mM IPTG and purified as previously described (7,20). Briefly, filtered supernatant from sonicated bacteria were purified using a 5 ml GSTrap HP affinity column (GE Healthcare, UK), then the GST tag cleaved using thrombin (GE Healthcare UK). Capsomeres were then isolated by sizeexclusion chromatography using a Superdex 200 10/300 GL column (GE Healthcare, UK). Endotoxin levels were reduced to below 5 EU/ml using Vivapure Q maxi H ion exchange columns (Sartorius Stedim, Gottingen, Germany) (7). VP1 capsomeres were assembled into VLPs by dialysis against an assembly buffer (7,20) and then against PBS (7); VP1 N C capsomeres were dialyzed only against PBS (7). The characterization of VLPs for this project has been previously reported (18). VLPs were analyzed using asymmetric flow field-flow fractionation coupled to multi-angle light scattering (AF4-MALS) and transmission electron microscopy to assess size distribution as previously described (30,31). Immunization of Mice Female BALB/c mice (n = 5/group) aged 6-7 weeks (Animal Resources Center, WA, Australia) were immunized three times at 3-week intervals by (i) subcutaneous injection (s.c.) of pooled VLPs or pooled capsomeres (pools comprising 10 µg of each CD8, CD4 and B cell chimeric construct) on the lower back near the base of the tail; (ii) intramuscular injection (i.m.) of plasmid DNA (100 µg) into the anterior tibialis; or (iii) s.c. injection of pooled peptides (pools comprising 30 µg of each CD8 + , CD4 + or B cell peptide epitopes) on the lower back near the base of the tail. Capsomeres and peptides were co-administered with 50 µg of high molecular weight poly(I:C) adjuvant (Invivogen, USA); and VLPs were administered with or without this adjuvant, in parallel groups. Mice were immunized in both homologous and heterologous prime/boost regimens involving two priming doses of PyCSP plasmid DNA followed by a booster dose of capsomere, VLP or peptide. Negative control groups included PBS, wild-type capsomere with poly(I:C), or wild-type VLP with poly(I:C). The positive control group received three doses of PyCSP plasmid DNA. All murine experiments were approved by the QIMR Berghofer MRI Animal Ethics Committee and were conducted in accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes (2004). Cytometric Bead Array Splenocyte/A20 cultures were incubated at in 96-well U-bottom plates in 200 µl of complete media at 37 • C and 5% CO 2 for 72 h. Culture supernatant was collected and stored at −80 • C prior to assay. Secreted IFN-γ, TNF, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12p70, and IL-13 cytokines were analyzed using the mouse cytometric bead array flex kit (BD Biosciences, USA) following the manufacturer's protocol. Samples were acquired using a FACSArray instrument (BD Biosciences, USA) and data analyzed using the CBA array software (BD Biosciences, USA). Indirect Fluorescence Antibody Test (IFAT) Sporozoite-specific antibodies were assayed by Indirect fluorescence antibody test (IFAT) using a protocol modified slightly from that previously described (35). Briefly, cryopreserved P. yoelii 17XNL sporozoites (Sanaria Inc., MD, USA) were centrifuged at 10,000 × g for 5 min then resuspended to 10 5 sporozoites per ml of Medium 199 (Life Technologies, USA). Then, 10 µl was added to wells drawn on a microscope slide using a Barrier Pap pen and air dried at RT before long-term storage at −80 • C. Prior to use, slides were thawed to RT in a desiccator cabinet. Pooled sera from each immunization group collected 5 days after the final immunization was diluted 1:400 in PBS with 2% BSA, then 10 µl was added to each well and incubated at 37 • C in a humid box. Wells were gently washed with PBS then stained with 10 µl of FITC conjugated anti-mouse IgG antibodies (BD Biosciences, USA), diluted 1:30 in filtered PBS containing 0.005% Evans blue. Slides were incubated for 30 min in a humid chamber and then washed gently with PBS. A cover slide was mounted over PBS with 10% glycerol and slides viewed on an EVOS fluorescence microscope (Advanced Microscopy Group, USA) at x400 magnification. Capsomere and VLP Construction The genomic insertion of PyCSP peptide epitopes into the murine polyomavirus VP1 protein for generation of VLPs (7) or within the truncated VP1 proteins for capsomeres (8) was confirmed by Sanger sequencing. Size exclusion chromatography fractions post-GST cleavage showed the formation of capsomeres as previously described (8), and these were subsequently used for capsomere immunizations. The VLP forming proteins were assembled in vitro and analyzed by AF4-MALS analysis, which showed the mean radius of the chimeric VLPs [wild-type, 21.00 ± 1.27 nm; CD8 VLPs, 20.51 ± 0.67 nm; CD4 VLPs, 21.07 ± 0.61 nm; B cell VLPs, 20.85 ± 0.67 nm (mean radius ± SD)] with minimal amounts of aggregation (18). Transmission electron microscopy showed similar morphology between all VLP groups (18). Capsomeres Induce Similar CD8 + T Cell IFN-γ Responses to VLPs Significant antigen-specific IFN-γ responses were induced by homologous immunization with both capsomeres and VLP, when compared to their respective negative control, as detected after in vitro stimulation with PyCSP DNA transfected A20s or with pooled PyCSP peptides ( Figure 1A). The amount of detected IFN-γ was not significantly different between capsomeres or VLP immunized groups (p > 0.05). The peptide immunizations also FIGURE 1 \| IFN-γ ELISpot responses induced by chimeric VLP or capsomere constructs. BALB/c mice (n = 5/group) received three immunizations with either (A) homologous s.c. injections of VLPs ± poly(I:C), capsomeres or peptides with poly (I:C), or (B) a heterologous DNA prime/boost regimen with two i.m. PyCSP plasmid DNA primes followed by a single s.c. boost with VLPs ± poly(I:C), capsomeres or peptides with poly (I:C). Seven days after the final immunization splenocytes were harvested and single-cell suspensions were stimulated in vitro for 40 h with irradiated A20 cells transfected with PyCSP plasmid DNA, or with irradiated A20 cells pulsed with either CD8 (280−288) , or CD4 (58−79) peptides or a pool of both peptides plus the B cell repeat peptide. IFN-γ spot forming cells (SFCs) were quantified with data displayed as mean SFCs per 10 6 splenocytes plus SEM. Statistical comparisons made to the PBS control group and between immunization groups with significance determined using one-way ANOVA followed by Bonferroni's post-test. p < 0.01, p < 0.001, and **p < 0.0001. induced a robust IFN-γ responses which was dominated by the CD4 peptide stimulation. Heterologous prime/boost immunization resulted in an ∼5fold increase in IFN-γ responses over the homologous regimens for both capsomeres and VLPs ( Figure 1B). When compared to their homologous respective counterparts, the numbers of IFN-γ spot forming cells in the heterologous DNA prime/boost immunized mice were increased significantly for capsomeres (p < 0.001), VLPs and VLPs with poly(I:C) (p < 0.0001) when stimulated with the whole PyCSP antigen transfected into A20 cells, and capsomeres (p < 0.01), VLPs and VLPs with poly(I:C) (p < 0.0001) when stimulated in vitro with pooled peptides. This increase in IFN-γ responses was driven by the DNA prime as evidenced by comparison of the responses to those of DNA only immunized mice. Consistent with results from the homologous immunization regimens, IFN-γ responses induced by capsomeres and VLPs in the heterologous DNA prime/boost regimen were not significantly different. To differentiate the responses observed with the pooled peptide stimulation in vitro, we also stimulated with either PyCSP CD8 + or CD4 + T cell peptides separately. That study showed that both the capsomere and VLP induced IFN-γ responses were predominantly associated with the CD8 + T cell peptide, with no significant difference between the capsomere and VLP immunized groups (Figures 1, 2). ICS was used to determine whether the CD8 + or CD4 + T cells were responsible for the production of IFN-γ (Figure 3). For both PyCSP capsomere and VLP immunized mice, IFN-γ was predominantly produced by CD8 + T cells. Similarly, in the heterologous prime/boost regimen, responses were directed against the CD8 + T cell epitope and apparently driven by the DNA component with responses similar to that induced by homologous DNA only immunization. Capsomeres Induce More Robust CD4 + T Cell IFN-γ Responses Than VLPs Although capsomere and VLP platforms induced a very similar profile of CD8 + T cell responses, unexpectedly, they differed in ability to induce antigen-specific CD4 + T cell responses. Specifically, homologous capsomeres induced significantly higher IFN-γ responses than VLPs with poly(I:C) (p < 0.0001) when stimulated with the CD4 + T cell peptide ( Figure 1A). This trend was also evident in the CBA analysis (Figure 2) and in the CD4 + T cell response detected by ICS (Figure 3), although the responses did not reach the level of significance. As observed for the CD8 + T cell response, the heterologous prime/boost was more effective than homologous immunization in inducing robust responses to the CD4 + T cell epitope (Figures 1B, 2). With this heterologous regimen, which included two priming doses of DNA, the vaccine-induced CD4 + T cell response for capsomeres and VLPs were comparable FIGURE 2 \| Cytometric bead array analysis of antigen-specific cytokine responses induced by immunization with chimeric VLPs or capsomeres. BALB/c mice (n = 5/group) were immunized with either a homologous (no DNA prime) or heterologous (DNA prime-boost) regimen as described in the legend to Figure 1. Seven days after the third immunization splenocytes were harvested and single cell suspensions stimulated with either the CD8 + (280−288) , or CD4 + (58−79) T cell peptides or with pooled peptides including both peptides plus the B cell repeat peptide and incubated for 72 h at 37 • C and 5% CO 2 . Culture supernatant was assayed using a cytometric bead array to quantify IFN-γ, TNF, IL-2, and IL-6 cytokine levels. Data are displayed as mean pg/ml + SEM for each cytokine with statistical comparisons made to the PBS control group with significance determined using one-way ANOVA followed by Bonferroni's post-test. p < 0.05, p < 0.01, p < 0.001, and **p < 0.0001. FIGURE 3 \| Flow cytometry frequency of stimulated splenocyte CD8 + and CD4 + T cells expressing IFN-γ and TNF post-immunization. BALB/c mice (n = 5/group) received three immunizations with either homologous (No DNA prime) or a heterologous DNA prime-boost regimen (DNA prime) as described in Figure 1. Seven days after the final immunization splenocytes were harvested and single-cell suspensions were stimulated in vitro by culturing them with irradiated A20 cells either transfected with PyCSP plasmid DNA or pulsed with PyCSP CD8 + and CD4 + T cell and B cell repeat peptides and incubated in for 1 h before adding Golgi Plug followed by a further 5 h incubation. Cells were stained for CD8 and CD4 receptors, then permeabilized and stained for IFN-γ and TNF cytokines then assessed by flow cytometry. The frequency of CD8 or CD4 positive cytokine-expressing cells are shown as the group mean + SEM (n = 5 mice/group) with CD8 + and CD4 + T cells represented by solid black bars or open bars, respectively. Statistical comparisons made to the PBS control group with significance determined using one-way ANOVA with Bonferroni's post-test. p < 0.05, p < 0.01, p < 0.001, and ***p < 0.0001. and did not differ significantly from that of three doses of PyCSP plasmid DNA alone. Nonetheless, the highest IFN-γ responses to the CD4 + T cell peptide were seen with homologous immunization with pooled peptides with poly(I:C) showing beneficial effects of the repeat peptide doses. Multiplexed Cytokine Responses A broader spectrum of cytokine responses, including TNF, IL-2, and IL-6 as well as IFN-γ, were quantified in culture supernatant from in vitro immune assays using cytometric bead array (CBA) (Figure 2). The IFN-γ response profile was consistent with that detected by ELISpot (Figure 1). IFN-γ responses induced by capsomeres and VLPs were comparable and directed predominantly against the CD8 + T cell peptide, whereas the CD4 + T cell peptide stimulated cytokine profile observed by CBA was higher for capsomeres than for VLPs although these differences were not significant. Capsomeres did not induce a TNF response for either the CD8 + or CD4 + peptide, but significant TNF responses were induced by VLP immunization and this was further increased by poly(I:C) adjuvant; these responses were preferentially directed against the CD8 + T cell peptide but were also significant for the CD4 + T cell peptide epitopes (Figure 2). Significant IL-6 responses were induced by both capsomere and VLP groups with or without DNA priming, with a similar profile for both CD8 + and CD4 + T cell peptide epitopes. In the peptide-immunized mice, responses for robust for IFNγ, TNF, and IL-6 cytokines were preferentially directed to the CD4 + T cell peptide. Capsomeres Were Less Efficient Than VLPs at Inducing Antibody Responses ELISA assays against the PST-PyCSP B cell repeat peptide and PyCSP recombinant protein were performed using individual and pooled mouse sera collected 14 days after doses 1 and 2; and 5 days after final immunization (Figure 4). Consistent with previous studies, VLPs induced high anti-PyCSP B cell repeat IgG titres which were increased with the coadministration of poly(I:C) adjuvant ( Figure 4A). Capsomeres induced moderate titres but, importantly, these were significantly less than VLPs with and without poly(I:C) (p < 0.01 and p < 0.05, respectively). There was a trend to increased antibody response with VLP administered with adjuvant compared to the non-adjuvanted VLPs but this was not statistically significant. However, the adjuvant effect was significant when VLPs were administered with a DNA prime (p < 0.05) indicating that with only a single VLP dose, the inclusion of adjuvant is beneficial and this benefit becomes less important after multiple doses ( Figure 4B). A priming dose of DNA appeared to adversely affect the antibody induction, since responses in the heterologous prime-boost regimen tended to have lower titres than their homologous counterparts especially in the absence of poly(I:C) (p < 0.01). We assessed the anti-PyCSP B cell peptide total IgG responses using pooled sera collected after each immunization to determine the effect of the number of immunizations within each regimen (Figure 4B). After each dose of VLPs, there was a trend to increased antibody titres when poly(I:C) adjuvant was co-administered. Antibody isotypes contributing to the total IgG response reported above were also delineated, using pooled sera collected after the final immunization ( Figure 4C). Immunizations with capsomeres, VLPs and VLPs with poly(I:C) each induced robust IgG 1 responses which were similar for all groups, and the IgG2 response in the VLP-immunized mice was increased when administered with poly (I:C). VLP immunizations induced an IgG 3 response which was absent in the capsomereimmunized mice. Compared to other immunization regimens, the homologous VLP immunization groups had higher levels of IgG 3 but similar levels of IgG 1 indicating that the second and third doses caused a shift in isotype responses toward an IgG 3 bias. IgG 2a responses were increased with the inclusion of poly(I:C) with the VLP but this was not present in the capsomere with poly(I:C) immunized mice, indicating the difference was platform specific. Moderate IgG 2a responses were also present in the DNA group and the heterologous DNA prime groups. To confirm that the vaccine-induced antibodies had affinity to the PyCSP protein, pooled sera collected after the final immunization was assayed using ELISA against recombinant protein (Figure 4D). The endpoint titres and profile were similar to that reported above for the B cell peptide (Figure 4A). To confirm that the IgG 3 response was consistent within groups, an ELISA against the PyCSP B cell repeat peptide was done using individual sera (1:400) collected after the final immunization ( Figure 4E). Each VLP immunization group tested had significant amounts of IgG 3 which was absent in chimeric capsomeres. The inclusion of poly(I:C) resulted in increased levels of IgG 3 responses, and the homologous threedose regimen significantly increased levels as compared to the heterologous prime/boost regimen (p < 0.001). Antibody Recognition of Native Antigens Importantly, antibodies induced by immunized with capsomeres or VLPs with or without poly(I:C) adjuvant were able to recognize P. yoelii 17XNL sporozoites, as evidenced by surface staining in the homologous and heterologous immunization regimens (Figure 5). DISCUSSION The Plasmodium sp. CSP is the leading antigen target for subunit vaccines against malaria, and is the antigenic component of the virus-like particle RTS,S which is the most advanced human vaccine candidate against Plasmodium falciparum malaria. It was, therefore, a logical choice for evaluation of novel vaccine delivery platforms. We know that liver-stage protection from sporozoite challenge can be achieved by either CD8 + T cells (27,36,37), CD4 + T cell (38,39) or antibody responses (29,40). Thus, herein, we constructed chimeric VLPs and capsomeres incorporating a CD8 + T cell epitope, CD4 + T cell epitope, or B cell epitope derived from PyCSP in order to evaluate and compare the ability of capsomeres and VLP vaccine platforms to induce epitope-specific cellular and antibody responses. Although the antibody-inducing capacity of chimeric murine polyomavirus VLPs and capsomeres has been well-established (7)(8)(9)23), the ability of this vaccine platform to induce epitopespecific T cell responses has not previously been comprehensively evaluated. We set out to accomplish this. Our study established that capsomeres and VLPs induced similar and significant levels of antigen-specific IFN-γ responses and these responses were primarily directed against the immunodominant PyCSP CD8 + T cell epitope and produced by CD8 + T cells. This CD8 + T cell IFN-γ immune response is considered essential for protection (39,41). Moreover, responses were increased when DNA priming was included as a heterologous regimen, as expected based on previous studies (18,(42)(43)(44). Interestingly, however, while both platforms were similar in capacity to induce an IFN-γ response, it was only the VLPs which evoked a TNF response, albeit at low levels, suggesting that VLPs may be better than capsomeres at inducing polyfunctional T cell responses. In contrast to the comparability of the CD8 + T cell responses, VLPs and their subunit capsomeres differed significantly in capacity to induce CD4 + T cell responses and antibody responses. Unexpectedly, capsomeres proved to be the best platform at inducing responses to the CD4 + T cell epitope. The target PyCSP CD4 + T cell epitope (59−79) was selected for study because of its the ability to induce functional CD4 + T cell populations of either T H 1 or T H 2 subsets associated with T cell proliferation responses or help for antibody responses (28). In our previous study, we could not detect any synergistic CD4 + T cell helper effect on antibody titres by co-administering a chimeric CD4 VLP (in a VLP pool) with B cell VLPs (18). In the current study, we were unable to show any CD4 + T cell helper responses as both capsomeres and VLPs were only administered in pools. We found that immunization with pooled peptides adjuvanted with poly(I:C) induced the most robust responses, however this is likely related to the dose of the target peptide epitope, as the positive control peptide pools incorporated 30 µg of the CD4 peptide whereas the amount of epitope presented in the chimeric capsomeres and VLPs was ∼1.77 and 0.52 µg, respectively, per dose. This may also explain why the capsomeres were more effective than VLPs at inducing responses against this peptide epitope as one of the benefits of capsomeres over VLPs as a vaccine platform is the ability to present a higher antigen load, and in our study capsomeres achieved a 3-fold higher antigen dose than VLPs. It should be noted that the antigen dose difference was the same for the CD8 + T cell peptide epitope and we did not see similar increased immune responses against that antigen target. The ability of capsomeres to induce CD4 + T cell responses is an important advantage for vaccinology since vaccine platforms that preferentially induce CD8 + T cell responses rather than antibody responses are often poor inducers of CD4 + T cell responses (45). The essential role of CD4 + T cells for vaccine-induced protection against malaria has been demonstrated in CD4 + T cell-depleted sporozoite-immunized mice where an absence of CD4 + T cells resulted in reduced antisporozoite antibodies, a reduced effector capacity of CD8 + T cells, loss of protective efficacy (46). It is well-established that VLPs are very effective at inducing antibody responses, perhaps due to their ability to cross-link B cell surface receptors (25) acting in a T cell independent manner (47,48). Indeed, for all licensed VLP-based vaccines, protection is thought to be mediated through neutralizing antibodies (10,11). Furthermore, studies using VLPs as well as their capsomere components with adjuvants showed them to be strong inducers of antibodies to inserted antigens (7-9, 23). Here, we have shown that VLPs with and without adjuvant were significantly better than capsomeres at inducing antigen-specific antibody responses, and that inclusion of poly(I:C) adjuvant increased antibody titres (49) with a skewed T H 1 isotype profile. Importantly, these vaccine-induced antibodies were capable of recognizing the B cell repeat peptide, as well as recombinant PyCSP protein, and the parasite and IFAT showed antibody affinity to the surface of sporozoites establishing that they could recognize the whole parasite as well as the protein. The durability of the induced FIGURE 5 \| IFAT detection of anti-sporozoite antibodies induced by immunization with chimeric VLPs or capsomeres. Mice (n = 10/group) were immunized using a three-dose regimen with either homologous VLPs, capsomeres or PyCSP plasmid DNA; or a heterologous regimen with two PyCSP plasmid DNA doses followed by a single VLP or capsomere boost as described in Figure 1. Sera collected from mice 5 days after final immunization was pooled and assayed against P. yoelii 17XNL sporozoite-coated slides (1:400 dilution). Slides were stained using FITC-conjugated anti-mouse total IgG and viewed on an EVOS fluorescence microscope at x1000 magnification. Scale bar represents 10 µm. antibodies and their protective capacity has yet to be established. The value of anti-CSP antibodies has been shown by their ability to sterilely protect mice against sporozoite challenge (29), and by observations that anti-CSP IgG concentration and avidity contribute to RTS,S/AS01E-mediated protection (50) with anti-CSP antibodies estimated to prevent 32% of infections following RTS,S immunizations (51). An interesting observation from our study is that the IgG 3 antibody isotype observed with VLP immunizations contrasts with results obtained using the same platform incorporating the Group A Streptococcus J8 peptide, where IgG 1 was the dominant IgG isotype and no IgG 3 was detected (7). It appears, therefore, that the development of IgG 3 is associated with the target epitope presented in the VLP structure rather than with the VLP platform itself. We had previously observed that an IgG 3 response was induced by immunization with three doses of chimeric B cell epitope VLPs alone (our unpublished data), establishing that the CD8 + or CD4 + T cell chimeric VLPs were not responsible for the subclass, as seen in other platforms (52). It is curious that the IgG 3 response was induced by immunization with VLPs but not capsomeres. IgG 3 antibodies have previously been identified as important for protection against various pathogens (53,54) including Plasmodium spp. parasites (35,40), and IgG 3 monoclonal antibodies raised against the PyCSP repeat (QGPGAP) protected BALB/c mice from a P. yoelii 17XNL sporozoite challenge (29). There are many examples of enhanced immunogenicity and protection with DNA prime-boost regimens using various antigen delivery systems encoding the same antigen or epitope (44,(55)(56)(57)(58) including VLPs (59) as a boost immunogen. Consistent with those reports, our data shows that the heterologous prime-boost regimen was better at inducing cellular responses than a homologous immunization regimen, for both capsomere and VLP platforms. The gain in cellular responses observed by including a plasmid DNA prime was, however, countered by a decrease in antibody titres. This is consistent with the known ability of plasmid DNA to preferentially prime a CD8 + T cell response and its poor ability to prime an antibody response (45). Chimeric capsomeres are a promising vaccine platform which build on the established vaccine potential of VLPs, but are easier and cheaper to produce than VLPs (22), and can include more antigenic insertion sites because there is no reliance on structural VLP formation. Here, we show that capsomere and VLP platforms induced similar levels of CD8 + T cell responses but the capsomeres were significantly better at inducing epitopespecific CD4 + T cell responses than VLPs. This enhanced CD4 + T cell response may be of particular importance in the control of chronic viral infections (60), the maintenance of CD8 + T cells during prolonged viral infections (61) where CD4 + T cell help may be required for optimal CD8 + T cell activity (62) or for CD8 + T cell memory (63), or indeed for optimal responses against any pathogens where a CD4 + T cell vaccine induced response is required. On the other hand, since capsomeres have a limited capacity to induce antibody responses, VLPs would be the preferred platform for antibody mediated immunity. Given that most licensed vaccines target the induction of antibody response and the increasing interest at identifying vaccine platforms capable of inducing robust T cell responses, our data have important implications for the development of vaccines against those pathogens that have thus far proved challenging. DATA AVAILABILITY STATEMENT The datasets generated for this study can be obtained from the corresponding author upon reasonable request. ETHICS STATEMENT The animal study was reviewed and approved by QIMR Berghofer Medical Research Institute Animal Ethics Committee. AUTHOR CONTRIBUTIONS DP, SA, DD, and AM contributed conception and design of the study. DP, NW, TR-H, and AM contributed to construction of virus-like particles and capsomeres. DP, SA, and PG conducted the mouse experiments. DP performed the statistical analysis. DP and DD wrote the manuscript. All authors contributed to manuscript revision, have read, and approved the submitted version.	2020-09-29T13:09:06.745Z	2020-09-29T00:00:00.000	{ "year": 2020, "sha1": "e12521a5319aefdb3d0b19706f46ad5336119317", "oa_license": "CCBY", "oa_url": "https://www.frontiersin.org/articles/10.3389/fimmu.2020.564627/pdf", "oa_status": "GOLD", "pdf_src": "PubMedCentral", "pdf_hash": "e12521a5319aefdb3d0b19706f46ad5336119317", "s2fieldsofstudy": [ "Biology" ], "extfieldsofstudy": [ "Chemistry", "Medicine" ] }
133605709	pes2o/s2orc	v3-fos-license	Impact of co-amoxicillin-resistant Escherichia coli and Pseudomonas aeruginosa on the rate of infectious complications in paediatric complicated appendicitis Andrey Véroniquea, Crisinel Pierre-Alexb, Prod’hom Guyc, Croxatto Antonyc, Joseph Jean-Marca a Unit of Paediatric Surgery, Woman-Mother-Child Department, Lausanne University Hospital (CHUV), Lausanne, Switzerland b Unit of Paediatric and Infectious Diseases and Vaccinology, Woman-Mother-Child Department, Lausanne University Hospital (CHUV), Lausanne, Switzerland c Institute of Microbiology, Department of Laboratory, University of Lausanne and Lausanne University Hospital (CHUV), Lausanne, Switzerland Introduction Appendicitis is the most common abdominal surgical emergency in children.Complicated appendicitis accounts for 25% of all appendicitis cases in paediatric populations, and the infectious postoperative complication rate is 3-25% [1,2].Case management, including the choice and duration of postoperative antibiotic therapy, varies among centres and surgeons [3][4][5].Several authors have tried to suggest standardised management, but no real consensus has emerged [1,[5][6][7]. The emergence of resistant pathogens like Pseudomonas aeruginosa, which have been found in 15-35% of complicated appendicitis cases [3,4,8,9,[17][18][19], should be considered when choosing the appropriate antibiotic regimen.The initial choice of antibiotics is important, as studies have shown that the risk of complications increases when bacteria appear resistant to the chosen antibiotics [4,9,10,16].However, evidence that empirical coverage of P. aeruginosa decreases the risk of post-surgical infectious complications is lacking [20]. In our institution, triple antibiotherapy (co-amoxicillin (amoxicillin + clavulanic acid) + aminoglycoside + metronidazole) used to be prescribed for complicated appendicitis.Since 2013, after an evaluation of regional resistances showed that fewer than 15% of Escherichia coli cases were resistant to co-amoxicillin, and with regard to the IDSA guidelines [21,22], we reduced the antibiotic spectrum to limit the emergence of antibiotic resistance.Co-amoxicillin was prescribed alone for pre-and post-surgical management. In the months following the implementation of our new antibiotic strategy, an increase in P. aeruginosa documented in peritoneal swabs was observed. Antibiotic prescriptions were adjusted according to microbiological findings when fever persisted 48 hours after surgery or when a bad clinical evolution showed signs of postoperative infection.Until 2015, the initial empirical therapy remained co-amoxicillin alone in the majority of cases.Since then, the choice of antibiotics has been more variable and dependent on the surgeon.This study's objective was to describe the bacteriology of the peritoneal swabs taken from cases of complicated appendicitis in our paediatric patients, and to correlate the risk of infectious complications (wound or intra-abdominal abscesses) with regards to the initial empirical antibiotics prescribed and the resistance profile of the bacteria. Design, setting and population This retrospective, single-centre cohort study was carried out in a tertiary care hospital (Lausanne University Hospital, Lausanne, Switzerland).Children and adolescents under 18 years old who underwent an appendectomy between 1 January 2010 and 31 December 2016 for a complicated appendicitis (perforated, peritonitis or abscess) were eligible for inclusion if they also had a peritoneal swab culture.Due to its retrospective, anonymous design, with no anticipated harm to the patients, informed consent was not requested.The study was approved by the Human Research Ethics Committee of the canton of Vaud and conducted in accordance with the principles of the Declaration of Helsinki, the standards of Good Clinical Practice and Swiss regulatory requirements. Intervention A paediatric surgeon established the indication for surgery based on his/her clinical diagnosis of appendicitis or, when in doubt, with the help of a radiological examination (echography or CT scan).Surgical intervention consisted of appendectomy by open or laparoscopic surgery.The first dose of antibiotics was given by the anaesthetist at induction or, when surgery was delayed, at diagnosis (less than 24 hours before surgery in every case).The diagnosis of complicated appendicitis was made by the surgeon during the procedure.The usual duration of antibiotics was four to seven days when post-surgical evolution was uneventful, but could be changed according to the microbiological results and extended by up to two weeks when infectious complications occurred.Management of complications was decided individually for each patient and when possible, a conservative treatment was preferred.If not, we preferred a percutaneous drainage by the interventional radiologist but when this was not technically feasible, we decided to perform a second surgery. Data collection All eligible patients were identified from our surgical database.Relevant information relating to demographic characteristics, clinical presentation, microbiological investigations, management and evolution were retrospectively retrieved from medical records.Children who had not had bacterial cultures were subsequently excluded. Statistical analysis Demographic information (e.g., gender, age), clinical characteristics (e.g., duration of symptoms), microbiological information (e.g., bacteria isolated, antibacterial resistance) and management (e.g., surgical procedure, antibiotic treatment) were compared between groups using either the Student's t-test or the Kruskal-Wallis test for continuous variables, and the chi-squared test or Fisher's exact test for categorical variables.All variables with a p-value <0.05 were then included in a multivariable logistic regression model for which adjusted odds ratios (OR) and 95% confidence intervals (CI) were calculated.All tests were twotailed, and a p-value <0.05 was considered statistically significant.Statistical analyses were computed using Stata software (Stata/IC 11.2 for Mac; StataCorp, Lakeway, TX). Population and initial management Between 1 January 2010 and 31 December 2016, 742 appendectomies were performed in our institution.Diagnosis was made by clinical examination, without ultrasound or CT-scan.One hundred and fifty-one patients (20%) had complicated appendicitis, confirmed surgically.After the exclusion of 18 cases due to the absence of peritoneal swab cultures, 133 patients were available for analyses (table 1).Median age was 9.5 years old (interquartile range [IQR] 5.7-12.4);there were 53 girls (40%) and 80 boys (60%).The median duration of symptoms before surgery was two days (IQR 2-5 days).There was an abscess at presentation which was confirmed by the primary surgery in 23 patients (17%).A laparoscopic appendectomy was performed in 75 cases (56%), whereas the other 58 appendectomies (44%) were carried out by laparotomy.A washout of the peritoneal cavity was performed in all cases.Fortyseven patients (35%) were given a triple antibiotherapy (co-amoxicillin, amikacin and metronidazole), 63 (48%) a monotherapy using co-amoxicillin, 15 (11%) were treated with piperacillin-tazobactam and 8 (6%) a dual therapy (ceftriaxone + metronidazole) because of an allergy to coamoxicillin. Microbiology E. coli was the most common bacterium, found in 94 patients (71%) (fig.1).Thirteen cases (14%) were resistant or had intermediate resistance to co-amoxicillin.Only three strains (3%) were resistant to cephalosporins owing to the production of broad-spectrum beta-lactamase (ESBL), and only one strain was resistant to amikacin.The resistance rate to co-amoxicillin varied from 0% (0/9) in 2013 to 33% (6/18) in 2015. The second most common bacterium was Streptococcus anginosus, found in 64 patients (48%).They were all sensitive to co-amoxicillin.P. aeruginosa was retrieved from the peritoneal fluids of 31 patients (23%) and Bacteroides fragilis from the peritoneal fluids of 40 patients (30%).B. fragilis appeared to have an intermediate or complete resistance to co-amoxicillin in only four patients (10%).These four cases were all sensitive to piperacillin-tazobactam and metronidazole.The rate of P. aeruginosa prevalence increased significantly between 2010 and 2016, from 9% to 44%.(fig.1). Evolution and complications Postoperative, intra-abdominal abscesses occurred in 21 patients (16%) (table 1).Fourteen patients required followup surgery (nine laparotomies and five laparoscopies).Four initial laparoscopies were extended to laparotomies.Three required drainage under radiological guidance, and four were treated conservatively with antibiotics alone.Seven other patients were treated for a wound abscess.We did not document new bacteriologic species in the abscesses.The rate of postoperative abscesses was not correlated to the presence of a preoperative abscess.These abscesses are drained surgically during the appendectomy and do not recur in most patients.A longer duration of symptoms before surgery and the presence of co-amoxicillin-resistant E. coli were positively correlated with the occurrence of infectious complications (table 1).In a multivariable analysis, only co-amoxicillinresistant E. coli remained significantly associated with infectious complications (OR 4.8; 95% CI 1.4-16.8;p = 0.013) (table 3). The initial presentation (abscess, peritonitis or simple perforation) and the surgical technique (laparotomy vs laparoscopy) had no significant impact on the complication rate. The presence of P. aeruginosa had no impact on the rate of infectious complications (table 1).However, in patients infected with P. aeruginosa, the absence of adequate antibiotic coverage may increase the risk of a wound or intra-abdominal abscess.The rate of infectious complications was 38% (8/21 patients) when the empiric antibiotic did not cover P. aeruginosa and 0% (0/10 patients) when it was covered adequately (p = 0.03). The median duration of hospitalisation for the entire population was six days (IQR 5-9), but it was significantly longer when infectious complications occurred.The median duration of hospitalisation in patients with and without infectious complications was 14 days (IQR 7.5-18.5)and 6 days (IQR 5-7) respectively (p <0.001).Nine patients (7%) were rehospitalised, and in eight cases this was for infectious complications.Eight out of 28 patients (29%) with infectious complications were rehospitalised, compared to one of 105 patients (1%) without infectious complications (p < 0.001). Discussion The choice of antibiotics for complicated appendicitis should address local patterns of E. coli, which is the most frequently identified bacterium [9,21].In our study, we confirm the increased risk of complications in the presence of co-amoxicillin-resistant E. coli. The indication for covering P. aeruginosa is not clear, even if there is some evidence of a correlation between the presence of P. aeruginosa and the risk of postoperative infectious complications [19].Skarda et al. compared the rates of infectious complications before and after the implementation of a new protocol [20].The new protocol implemented the use of a dual therapy using ceftriaxone and metronidazole instead of monotherapy with piperacillintazobactam.In their population of 306 children with ruptured appendicitis (152 under the new protocol), they found no increase in infectious complications associated with the new protocol.The complication rate was 1.3% after the implementation of the new protocol, compared to 5.2% before.The difference was not statistically significant.The overall rate of P. aeruginosa was not reported.However, in a previous report [10] the same authors reported a P. aeruginosa rate of 32%.In recent reviews or guidelines, the question of how to cover P. aeruginosa has not been specifically addressed, even when coverage for this bacterium uses a regimen including aminoglycoside (mostly in a triple therapy regimen) or piperacillin-tazobactam (monotherapy) [6,14]. Our results showed that, even though the absence of coverage for P. aeruginosa increased the risk of complications when this bacterium was identified, its presence alone did not increase the risk of infectious complications.The adequate coverage of co-amoxicillin-resistant E. coli (4/13, 31%) and P. aeruginosa (10/31, 32%) were comparable in our study.However, the proportion of complications was higher with co-amoxicillin-resistant E. coli (7/13, 54%) than with P. aeruginosa (8/31, 26%).It is difficult to draw definitive conclusions with so few cases, but we can assume that in ruptured appendicitis, the relative pathogenic role of E. coli is higher than that of P. aeruginosa.A low inoculum for P. aeruginosa in comparison to E. coli might also explain this difference.Unfortunately, the bacterial cultures were not quantitative and could not determine the inoculum.If we wanted to empirically cover P. aeruginosa and coamoxicillin-resistant E. coli, our best choice would be piperacillin-tazobactam or triple antibiotherapy (table 2).However, because the benefit of empiric coverage of P. aeruginosa is not definitively proven and in the absence of a stable, high identification rate of P. aeruginosa, as was the case at our institution in 2013-2014, there is currently insufficient evidence to widen the antibiotic spectrum to cover P. aeruginosa.However, with the rate of co-amoxicillin resistance of E. coli peaking at more than 30% in 2015, we decided to change our choice of antibiotic from co-amoxicillin to a dual therapy of ceftriaxone and metronidazole.In the case of bad clinical evolution or cultures positive for P. aeruginosa, the treatment was changed to piperacillin-tazobactam as soon as possible to reduce the risk of postoperative complications. Our study has some limitations.First, its design is retrospective, and a significant proportion of patients had not had peritoneal swab cultures done.Second, our surgical and antibiotic strategies were not uniform throughout the study period.Finally, the small number of patients may have limited our ability to prove a significant impact of P. aeruginosa or other variables (such as the presence of abscess at presentation) on the risk of infectious complications, and the local conditions may limit the generalisability of our findings to other settings. Conclusion The multivariable analysis of this small, retrospective study revealed a statistically significant increase in the risk of postoperative complications in the presence of resistant E. coli.Knowledge of local resistance patterns should guide antibiotic strategies.The impact of the presence of P. aeruginosa and the absence of its antibiotic coverage on the rate of infectious complications remains unclear.More data would be needed to justify the systematic coverage of P. aeruginosa in children with ruptured appendicitis. Figure 1 : Figure 1: Microbiological results by year of identification. Table 1 : Patient characteristics and risks factors for infectious complications. IQR = interquartile range Triple antibiotherapy: co-amoxicillin, aminoglycoside and metronidazole * Univariable analysis comparing patients with and without infectious complications: Student's t-test or the Kruskal-Wallis test for continuous variables and chi-squared test or Fisher's exact test for categorical variables Table 2 : Antimicrobial resistant rates of peritoneal samples over time. Swiss Medical Weekly • PDF of the online version • www.smw.chPublished under the copyright license "Attribution -Non-Commercial -No Derivatives 4.0".No commercial reuse without permission.See http://emh.ch/en/services/permissions.html.	2019-04-27T13:03:56.250Z	2019-04-26T00:00:00.000	{ "year": 2019, "sha1": "ae61eb32660ea8c0c843a79a3d218026c17de862", "oa_license": "CCBY", "oa_url": "https://smw.ch/journalfile/view/article/ezm_smw/en/smw.2019.20055/dccfbc7ccbe366c57132057cfa46dc35596be37d/smw_149_w20055.pdf/rsrc/jf", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "dccfbc7ccbe366c57132057cfa46dc35596be37d", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
248326656	pes2o/s2orc	v3-fos-license	Developing a theoretically informed education programme within the context of a complex implementation strategy in UK primary care: an exemplar from the IMP2ART trial Background IMPlementing IMProved Asthma self-management as RouTine (IMP2ART) is a programme of work developing and evaluating a strategy for implementing supported asthma self-management in UK primary care. The strategy encompasses patient-facing resources, professional education, and organisational approaches to embed supported self-management. This paper reports the development of a theoretically informed interprofessional education programme which aims to raise awareness of and enable healthcare professionals to deliver effective supported self-management. Methods Aligned with the Medical Research Council (MRC) Complex Intervention Framework, the multidisciplinary team developed educational content in three phases: (1) developmental phase, identifying educational and behaviour change theory to guide development, in consultation with a professional advisory group; (2) feasibility pilot phase, testing the education using a ‘think-aloud’ method; and (3) pre-pilot phase, delivering the education within the IMP2ART strategy. Results The developmental phase identified educational and behaviour change theory and the need to provide two education modules: (1) a team module to raise awareness of supported asthma self-management for the whole team and (2) an individual study module for those who conduct asthma reviews with patients. The feasibility pilot highlighted content and design features in need of refinement and the pre-pilot identified substantial changes to the delivery strategy for the education modules. Conclusions A multi-stage development process, aligned with the MRC Framework, contributed to the module design and delivery. Prior explorative work, multi-disciplinary team discussions, and professional advisory group consultation, informed the initial development, and in-practice testing and pre-pilot stages enabled refinement. In our experience, there were important benefits of working together as an educationalist/researcher team. The education programme, a core component of the implementation strategy, is now being tested in the IMP2ART UK-wide cluster randomised controlled trial. team-based professional education is a core component of the strategy that seeks to embed supported selfmanagement within practice routines. This paper reports the development, within the context of the IMP 2 ART research, of a theoretically informed interprofessional education programme which aims to raise awareness of and enable healthcare professionals to deliver effective supported self-management. Methods Our programme of work aligns with the developmental and feasibility piloting stages of the Medical Research Council (MRC) Framework for developing and evaluating complex interventions [15] and follows the guidance for reporting intervention development studies in health research (GUIDED) [16]. The education is set within the multi-theories model of adult learning proposed by Taylor and Hamdy [17], which provides a framework for utilising and combining a range of educational principles and approaches. Ethical approval was provided by West Midlands -Black Country Research Ethics Committee (REC ref: 18/WM/0300), and all participants provided written informed consent. The multidisciplinary team The IMP 2 ART education programme was developed by a multidisciplinary team who, in addition to e-mail correspondence, held a series of eight workshop/meetings from September 2018 to August 2019 both virtually and in-person. Asthma UK Centre for Applied Research: IMP 2 ART team The IMP 2 ART research team consisted of academics, general practitioners, and health psychologists based within the Asthma UK Centre for Applied Research (AUKCAR). Funded by Asthma UK, AUKCAR is a network of researchers from universities across the UK, people affected by asthma, healthcare professionals, NHS partners, and other organisations. Education for Health Education for Health (EfH) is a health education charity that aims to improve the lives of people living with longterm conditions. The EfH team included educationalists and clinical educators with expertise in curriculum design for adult learners, as well as online learning technologists with expertise in instructional design. Professional advisory group We established a Professional Advisory Group (n = 10) including doctors and nurses from the Primary Care Respiratory Society (PCRS) to advise on educational content. They met formally (by video-conference) on two occasions during the education development, with additional informal input from some members. Specific topics discussed included insights into the primary care context, practical barriers to implementing supported selfmanagement and strategies for overcoming challenges. MRC framework The phases of the education development mapped to the MRC Framework are illustrated in Fig. 1. Developmental phase Preliminary work included a systematic review [9], qualitative exploration of the primary care context [18], and an understanding of both the educational and behaviour change literature [19]. The developmental work was led by EfH (VM) in partnership with IMP 2 ART researchers (KM, LS, and HP) with advice from the Professional Advisory Group (led by SH). The learning aims and outcomes for the education were developed using existing theory, e.g. multi-theories model of adult learning [17], Bloom's taxonomy [20], Theoretical Domains Framework [21], and the existing evidence-base [9,18,19]. Feasibility pilot phases We recruited staff from four demographically diverse general practices from around the UK and delivered prototype versions of the education. We used a 'thinkaloud' method [22], in which general practice staff were encouraged to verbalise their thoughts as they worked through the programme. A brief follow-up interview or focus group used a topic guide developed in-line with guidance by Creswell and Creswell [23], to clarify participants' thinking about whether the aims and learning outcomes of the education programme were met, whether theoretical elements were addressed, and explored potential implementation. The think-aloud method was used to capture any changes that needed to be made to (for example) user friendliness or content of the education. Think-aloud sessions and follow-up interview/focus groups were carried out by an IMP 2 ART researcher (KM), were audio-recorded following consent, and transcribed verbatim. Time to complete the education was not recorded, as vocalising thoughts within the think-aloud methodology would have distorted findings. The sessions were analysed using thematic analysis [24], and all analyses were performed in NVivo 11. Pre-piloting the education programme within the IMP 2 ART implementation strategy Following necessary refinements, the educational components were integrated with the patient resources and organisational components of the IMP 2 ART implementation strategy for pre-piloting in an additional four general practices. Following the delivery of the pre-pilot, a sample of general practice staff from each of the four pre-pilot practices were interviewed by KM between October and November 2019. The semi-structured interviews followed a topic guide that aimed to explore staff experiences of the programme. Interviews were audiorecorded, transcribed verbatim, and analysed using thematic analysis [24] in NVivo 11. Developmental phase Identify the evidence base Our systematic review of education for healthcare professionals implementing supported asthma selfmanagement informed the IMP 2 ART strategy [19]. Although findings were mixed, there was evidence to suggest that professional education, especially team-based education, can increase asthma action plan provision, improve asthma control, and reduce unscheduled care. Effective interventions were explicitly guideline-based, often involved local opinion leaders, and included interprofessional education. This echoes wider literature that suggests that educational outreach can change professional behaviour and improve processes of care by modifying peer group norms and expectations [25]. Selecting appropriate theory Utilising well-established educational theory enabled us to understand our learners and to plan approaches to meet our goals and their learning needs. From the outset, we applied the overarching principles of andragogy (adult learning), which considers that motivation to learn is intrinsic and education should be learnercentred [26]. The curriculum design was grounded in cognitivist and constructivist theory aiming to build upon existing knowledge and develop deeper levels of understanding to enable application to practice. Bloom's original taxonomy [20] supported the development of learning outcomes appropriate to the learner context. Taylor and Hamdy's [17] multi-theories model provided a framework for exploring learning "phases" and drawing upon multiple theories to meet learning needs. Online module design was informed by learning style theory [27,28] and transactional distance theory [29]. For example, visual, auditory, read/write, kinaesthetic (VARK) [27] principles allowed the full range of learning styles to be addressed via a variety of tasks, activities, interactions and narratives, and transactional distance was mitigated by interactive module design and scaffolding provided by additional resources and facilitator support. Implementing supported self-management involves a change in professional behaviour [9]. Throughout the IMP 2 ART programme, the 'COM-B' framework of understanding behaviour was applied [30], with specific theoretical elements needed in an education programme identified and addressed through the Theoretical Domains Framework (TDF) [21]. The TDF describes and aids application of theory to the understanding and development of interventions to promote behaviour change, including that of healthcare professionals [31,32]. The IMP 2 ART systematic review identified that the TDF domains 'social influences', 'environmental context and resources', 'behavioural regulation', 'beliefs about consequences', and 'social/professional role and identity', were associated with effective asthma education for healthcare professionals [19], and were therefore priorities for inclusion in the IMP 2 ART education. Understanding the setting We used findings from our prior qualitative study to inform the educational development, which explored how asthma care was delivered in UK general practice, and the views of clinicians and support staff on the skills they needed to support asthma self-management [18]. For practices in England, an important context is the Quality and Outcome Framework which rewards provision of an action plan [33], though there are concerns that this represents a 'tick box' rather than meaningful engagement with supported self-management [34,35]. Supporting self-management was described as 'high priority' (particularly by nurses) but was viewed as a nurse-led role, with general practitioners (GPs) often feeling unfamiliar with the practicalities of providing asthma action plans. Improving practitioner skills and confidence was thus a priority, and this needed to include GPs, nurses, and administrative staff. Linking with the COM-B element of 'social opportunity' [30], we decided that the IMP 2 ART education programme should target the whole general practice team, and include two modules with the aim of: 1. Addressing the importance of supported selfmanagement in asthma-for the whole practice team (team-awareness module: module 1), 2. Providing an in-depth education module on supporting self-management for the healthcare professionals who conduct annual asthma reviews with patients (most often nurses in the UK) (individual study module: module 2). The Professional Advisory Group highlighted the need to reiterate evidence that supported self-management reduces asthma attacks and deaths and to discuss team roles within the practice (including the receptionist/administrative role in supporting self-management). There was concern that some healthcare professionals delivering asthma care may not have the recommended prerequisite asthma training [36] and would not have the basic knowledge on which module 2 builds. This should thus be made explicit in the introduction to module 2 and suitable basic training signposted. In addition, the IMP 2 ART facilitators will be able to discuss training needs if they detect a lack of general asthma knowledge in the healthcare professional delivering care. The Professional Advisory Group also identified barriers that they had encountered in practice e.g. nurses lacking confidence in their ability to offer self-management advice, and patients viewing an action plan as 'non-essential'. Their suggested solutions to overcoming these barriers included specific training on personalising action plans so the healthcare professionals felt confident to adapt/ 'cross-out' sections of the action plan if they were irrelevant for individual patients. Establishing learning outcomes and mapping educational content The aims and learning outcomes for each module were underpinned by the educational and behaviour theory and the evidence-base [9,18,19] and informed by discussions with the Professional Advisory Group (see Table 1 for details). Both modules 1 and 2 were developed with Nimble Author, an interactive eLearning development software [37]. As module 1 highlights a team-based approach to supporting asthma self-management, it was designed to be delivered by a facilitator in a whole-team setting (although it could also be completed individually online). The facilitator role development has been guided by the integrated Promoting Action on Research Implementation in Health Services (i-PARIHS) framework [38]. The module included 20 min of content to be delivered by the facilitator over an hour. Module 2, the education module for those who conduct asthma reviews with patients, was developed with 60 min of content. The modules employed Taylor and Hamdy's [17] five phases for learning. Learners existing knowledge is challenged in the Dissonance phase; solutions are found in the Refinement phase and solutions applied in the Aim To raise awareness of the benefits of supported self-management and increase engagement, motivation and commitment to supporting selfmanagement (SSM) so that it becomes a priority across the whole practice team To enable healthcare professionals to use behaviour change strategies in clinical practice to deliver effective supported self-management Learning outcomes ▪ Understand the principles of supported self-management ▪ Recognise the benefits of supported self-management for patients and the practice ▪ Identify roles and contributions in the team approach to supported selfmanagement ▪ Understand the concept of supported selfmanagement ▪ Identify individual and organisational barriers to effective supported self-management ▪ Use a range of strategies to support behaviour change ▪ Reflect on various approaches to supported selfmanagement ▪ Evaluate self-management resources appropriate for their patient population and context Core module content ▪ Introduction (e.g. The National Review of Asthma Deaths (NRAD); definition of supported self-management) ▪ Principles of supported self-management (importance of SSM and barriers to SSM) ▪ Benefits of supported self-management (benefits for patients, general practice, and society) ▪ Teamwork (the team approach to supporting self-management) ▪ Summary, planning and evaluation ▪ Introduction (learners' approach to and confidence in supporting self-management) ▪ What is supported self-management? (evidence for SSM) ▪ Barriers to effective supportive self-management (patient and healthcare professional perceptions, and overcoming barriers) ▪ Helping patients to change behaviour (behaviour change and communication skills) ▪ Approaches to supported self-management (working in partnership with patients and motivational interviewing) ▪ Self-management in practice (personalising asthma action plans) ▪ Summary, planning and evaluation ▪ Additional resources Organisation phase. Review and refinement occurs in the Feedback phase and learners reflect on action in the Consolidation phase. Dissonance was addressed in both modules by clarifying the task before learners engaged with it. Instructional design techniques such as reflective activities and team planning aligned to the Refinement and Organisation phases. Reflective activities have previously been identified as an important component for learning about self-management support [39]. Feedback was provided within the online modules, as responses to interactive learning activities such as quizzes and reflective activities. Problem-based learning within both modules provided opportunities for learners to consolidate, and this was enhanced by the facilitator who stimulated reflection on actions [40]. Module content was mapped to the TDF domains (including those associated with effective interventions in our systematic review [19]) to ensure all relevant behaviour change elements were included [41] (see Table 2 for details). The 'social/professional role and identity' domain was addressed by clearly defining potential roles of various staff/professionals in general practice. The 'social influences' domain was addressed by incorporating local opinion leaders' views of supporting selfmanagement (known to be associated with successful promotion of evidence-based practice [19,43]). The domain 'beliefs about consequences' was addressed by providing information in the modules about health/societal costs of asthma and benefits of supported selfmanagement. Other TDF domains, by nature relevant to professional healthcare education and training, such as 'knowledge' and 'skills', were also included in the education modules. Drawing on the VARK principals [27], both modules delivered content using a variety of methods to appeal to different learning styles e.g. traditional text, animated video, whiteboard animation (e.g. statistics on asthma deaths), and live-action video (e.g. patient stories recorded by the IMP 2 ART Patient and Public Involvement (PPI) team and key opinion leader views). Feasibility pilot phases Learnings from the feasibility testing of the modules A total of 19 general practice staff across four general practices participated in testing the modules. Seventeen practice staff (administrative staff (n = 8), nurses (n = 5), and GPs (n = 4)) participated in testing the wholepractice team-awareness module (module 1) facilitated in the practices (n = 2) by one of the authors (VM). Three nurses from three practices tested module 2, the individual study module (one of whom had also taken part in module 1 testing). We describe below the overarching themes derived from the modules 1 and 2 feasibility pilot data. An overview of the themes is displayed in Table 3 (module 1) and Table 4 (module 2). Module 1: Delivery in a group setting. General practice staff were positive about module 1 being delivered to the whole team which they felt would raise awareness of supported self-management, though some clinicians were concerned that administrative staff may be less comfortable to contribute. User experience. The module was found to be clear. The included animations and videos were positively received by staff, particularly the videos with patients describing their experiences. There were a few practical issues, for example, sound tended to be disabled on practice computers, and some text inputted by the facilitator appeared in subsequent boxes. Content. The module raised awareness of supported self-management, though including an example of an asthma action plan was suggested. Clinicians appreciated some of the practical messages (e.g. 'remember to look at the patients' asthma action plan), though some felt that the content did not expand their existing knowledge. Staff suggested a certificate of completion would count towards their Continuing Professional Development (CPD). Module 2 User experience. Participating nurses found the module to be clear and easy to read, though there were some practical issues. The use of both animated and live video was welcomed, with the patient experience videos being well received. There were mixed opinions about the duration of the module, with some feeling that it would take about an hour, and others feeling it may take longer. Content. Positive aspects of the content included the use of credible sources in videos (e.g. two expert nurses describing the benefits and personalisation of action plans). Participants felt that the module would also be particularly useful for newer members of staff/GPs. The inclusion of behaviour change techniques to use with patients (e.g. motivational interviewing) was considered useful and a reminder for those already trained. One nurse suggested expanding on symptom information (e.g. coughing and expectorating). Refining the modules after feasibility testing Following feasibility testing, a number of refinements were made. Module 1 was designed to be delivered by a facilitator, who needed to highlight the whole-team approach, and encourage all members of staff to participate, specifically addressing concerns that administrative staff in some practices may lack confidence to contribute to the session. Examples of asthma action plans will also be brought to facilitated module 1 sessions. In order to keep module 2 to time, instructions for exercises where the participant had to input information, e.g. reflection exercises, were revised to specify 'brief notes' or 'three to five bullet points'. Further, some of the exercises were added to a separate 'resources' section to ensure that the learner could complete the module within 1 h. For both modules 1 and 2, design issues were fixed and, to circumvent the lack of sound on some general practice "It's probably the best way to get the maximum number of people aware at one point…" "…in a big group where you've got a huge range of staff, if the GPs start speaking, the admin staff will step back and not contribute." User experience Clarity Animation and videos Design issues "I think it certainly gets the information across." "I think the more visual the better, in my experience, I like animation and videos." (Observation; no illustrative quote) Content Raising awareness of SSM Relevance for all staff Examples of asthma action plans certification "I think it's really beneficial from a clinical point of view because it's a reminder to look for the asthma action plan." "I can imagine it would be interesting to non-medical staff probably more than medical staff. I didn't hear anything that I didn't know. But it wasn't aimed at me solely so that's fine." "If it's about a self-management plan, let's have a look at one." "…if you have, even if it's one sheet, if you have some kind of certificate at the end…" computers, a transcription was added to all videos. Finally, as an incentive to encourage completion of the modules, we added completion certificates, which learners could use towards their CPD. Findings related to the education modules within the IMP 2 ART implementation strategy pre-pilot The pre-pilot provided practices with the refined education modules, along with other components of the IM-P 2 ART implementation strategy including patient-facing resources (e.g. patient invitation letters, an asthma information website), organisational components (e.g. audit and feedback, a patient-centred asthma review template). Module 1 was delivered by a facilitator (VM or RL) in four pre-pilot general practices. Access to Module 2 was provided for the clinician(s) who conducted asthma reviews in each of the four practices. It became clear that a proactive approach was needed to ensure whole team attendance at the facilitated module 1 session; one practice had only invited core staff. Participants indicated that they would welcome the module link being available to them prior to the session "…giving it to us beforehand just to look through, you know." It also became clear that the facilitator needed flexibility to adjust the session to the priorities of the individual practice e.g. to introduce a specific IMP 2 ART resource that the practice might wish to adopt. Finally, for module 1, the team task of developing a 'practice plan' to embed supported selfmanagement needed greater emphasis. Completion of module 2 was low, and reminders about availability of this module were needed. Refinements after pre-pilot We developed a sequence of emails to be sent to all practice staff in the four weeks leading up to the facilitated module 1 session promoting the session and highlighting its applicability to all staff. The emails also provided a link to module 1, so that staff could (as an option) explore the material prior to the session. The approach to facilitation was changed so that module 1 could be delivered flexibly, tailoring the session to the individual practice, in order to improve motivation and engagement rather than "teaching". After developing the 'practice plan' in the facilitated module 1 session, scheduled follow-up contact with the practice reminded staff about the plan to which they had agreed and encouraged them to work towards implementing supported asthma self-management in their practice. We developed reminder emails for clinicians who were eligible to complete module 2. Adaptation to COVID-19 context In mid-2020 and post-refinements, we held a virtual workshop with IMP 2 ART educationalists and researchers to finalise the implementation strategy. UK primary care was adapting to remote consultations due to the COVID-19 pandemic [44], and we iteratively added content to module 2 that covered effective remote consultation skills. No content changes were made to module 1, as the flexible facilitation strategy meant the module could be delivered face-to-face (as originally intended) or remotely via video-conference. The final content of the two modules is summarised in Table 1. Discussion In summary, we describe the development within a research programme of two asthma self-management online education modules. One facilitated module highlighted the importance of a team approach and one provided education for those who conduct asthma reviews with patients. Aligned to the MRC Framework [15], the process included a developmental phase drawing on and mapping to educational and behaviour change theories, a feasibility phase to test module content and design, and a pre-pilot of the education modules as components of the IMP 2 ART whole-systems implementation strategy. Throughout the iterative process, changes were made to both the educational content and design, as well increasing flexibility of the facilitation. The IMP 2 ART strategy (incorporating the "…easy to read. Not too much on the screen, the screen wasn't too busy, which is always appreciated. Yeah, I thought the layout was fine, easy to read, flowed easily." "The patients, yeah, seeing the videos of the patients. That always kind of helps to put it in perspective, I think. Yeah, I really liked that." "…I've not actually got a speaker on here." "…the module itself is going to take a lot longer for people to do and I think they do need time then to go… especially if you're going to do a little bit of reflective things in it…" Content Positive aspects Behaviour change techniques Possible changes "…the fact that you've got two very senior people, very experienced people, explaining it and talking it through should emphases to those nurses who don't use it why they should." "Actually, what I did get from it was it reminded me, just watching that flu vaccination interaction, it reminded me how theoretical motivational interviewing is and how in practice it's much quicker." "…the concentration was on the peak flow but the other bits I think could have been expanded on as well." education programme) is now being tested in an internal pilot and UK-wide cluster randomised controlled trial (ISRCTN registry, ref: ISRCTN15448074). If successful, this has potential to change the way that asthma care is delivered in primary care, by directing focus and highlighting the importance of supported selfmanagement. Strengths and limitations A major strength was the cross-discipline empirical and theoretical work underpinning development of the asthma self-management education. The modules were developed using a systematic and standardised approach with a multi-disciplinary team and used appropriate educational and behaviour change theory, existing evidence, and stakeholder advice. Feasibility testing, pre-piloting and refining allowed for necessary changes to be made to the education programme. In addition, the modules were developed and tested with current general practice staff, strengthening applicability to real-world practice. Our thorough description of the development and content of the modules avoids the criticism of a recent systematic review that the training for asthma educators should be better described [45] and may act as a model/ framework for other primary care staff course development. A limitation was the poor initial uptake of module 2. Refinements to the delivery and strategies to facilitate engagement were made, though until the pilot trial is conducted, the impact of this strategy will not be clear. IMP 2 ART is a UK programme of work and would need to be adapted to other healthcare systems in which the professionals delivering asthma education and supporting self-management may be different. The principles of taking a multi-disciplinary approach to developing and adapting an education intervention are however, transferable. Interpretation of findings in relation to previously published work Recent commentaries have suggested that medical education might benefit from working within a complex intervention framework [46,47]. We used this approach and embedded the development, piloting and testing of educational modules within the MRC complex intervention framework [15]. Our multidisciplinary team worked together so that researchers learnt from educational theory [17,20] and the educationalists gained by discussing behaviour change techniques with health psychologists [21] and implementation frameworks with researchers [48]. The educational modules benefitted by being conceived within a whole systems implementation strategy, which added an over-arching dimension and changed the delivery of one of the modules. A systematic overview that used Normalization Process Theory to analyse studies of complex interventions [25] concluded that educational meetings and outreach were more effective when 'bundled together' with strategies such as audit and feedback and practical organisational tools that modified the peer group norm and influenced structures of practice (a target of IMP 2 ART module 1). This comprehensive developmental process has potential implications for the development of future healthcare education seeking to change professional behaviour. The MRC Framework imposed a structured process of feasibility testing and we built on this by incorporating a real-world pre-pilot as part of the intervention design, an approach taken by other complex interventions in primary care [49]. This iterative approach to development enabled us to refine the educational modules and adapt them further when the COVID-19 pandemic altered the primary care context. Furthermore, practices will be encouraged to adapt the components of the implementation strategy to their practice profile, the skill mix of their practice staff, and the routines of their practice organisation. Evaluation will assess implementation (action plan ownership) and health outcomes (unscheduled care) in a UK-wide cluster randomised trial (ISRC TN registry, ref: ISRCTN15448074). This corresponds to assessing changed behaviour and benefit to society described in (for example) Kirkpatrick's hierarchy of levels of evaluation [50]. Conclusion We conclude that a multi-stage development process aligned with the MRC complex intervention framework contributed to the design and delivery of the education modules. Prior explorative work, multi-disciplinary team discussions, professional advisory group consultation, informed the initial development, and in-practice testing and pre-pilot stages enabled refinement. Not all education can/should be part of a whole systems implementation strategy, and evaluating outcomes at the level of changing practice and improving health outcomes is not always feasible. There are, however, lessons to be learnt from our experience of working together in an educationist/researcher team. We thank the IMP 2 ART Patient and Public Involvement Group, co-ordinated by Dr Tracy Jackson, for their advice: Anthony McGuiness, Bill Day, David Weatherill, Daniel Russell, Elisabeth Ehrlich, Eve Smyth, Irena Paterson, Kieron Blake, Olivia Fulton, and Noelle Morgan. We thank the primary care general practices that participated in this study. We acknowledge the helpful advice of the Independent Programme Steering Committee chaired by Professor Robbie Foy.	2022-04-23T13:37:04.328Z	2022-04-23T00:00:00.000	{ "year": 2022, "sha1": "f26f3007aafd274a4952f8d78048c54c832532c3", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Springer", "pdf_hash": "f26f3007aafd274a4952f8d78048c54c832532c3", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
261088929	pes2o/s2orc	v3-fos-license	A rare form of hyperthyroidism leading to the diagnosis of acromegaly: A case report Acromegaly is a rare disease, usually caused by a pituitary tumor. It typically exhibits slow evolution and can result in numerous complications. In the present case report, the patient presented with hyperthyroidism associated with ophthalmopathy and right nodular goiter. The laboratory tests revealed persistent high levels of phosphorus without an apparent cause. After ruling out common pathologies associated with this finding, a focus was placed on the clinical aspects associated with acromegaly, a rare cause of hyperphosphatemia. Laboratory tests and MRI confirmed the diagnosis. The patient underwent transsphenoidal surgery, but the disease remained active, thus medical treatment was initiated, to a poor initial response. Associated with acromegaly, two distinct thyroid pathologies were diagnosed: Toxic adenoma and Graves' disease. This case highlights the challenges in diagnosing and managing a rare endocrine pathology. Introduction Acromegaly is a rare disease, usually caused by a pituitary tumor secreting growth hormone (GH), that has an annual incidence rate between 0.2 and 1.1 cases per 100,000 individuals (1). Given the clinical features associated with this pathology develop insidiously, the delay between onset and diagnosis can be >5 years (2). Numerous comorbidities are associated with exposure to high levels of GH and its peripheral mediator, insulin-like growth factor 1 (IGF-1). Goiter is a frequent endocrine comorbidity of acromegaly. It is the result of overstimulation by GH and IGF-1 of the follicular epithelium (3,4). Diffuse and nodular goiter represent ~57.7% of acromegaly cases, while toxic nodular goiter is an unusual occurrence in acromegalic patients, although it is more common in women (3,5). Usually, patients with acromegaly are euthyroid or develop central hypothyroidism due to the compressive nature of the tumor. Hyperthyroidism is diagnosed in 3.5-26% of acromegalic patients (3). Active acromegaly can impair thyroid-stimulating hormone (TSH) response to thyrotropin-releasing hormone (TRH), leading to abnormal TSH values, and increased conversion of free thyroxine (T4) to free triiodothyronine (T3), which results in central suppression of TSH secretion (6). The association between acromegaly and thyroid autoimmunity was also shown previously; 25% of the patients, the majority of whom were women, had positive anti-thyroid peroxidase (ATPO) antibodies (7). Autoimmune thyroid diseases (AITD) affect 2-5% of the population, being more commonly diagnosed in women, and these include Graves' disease among other pathologies (8,9). Clinical characteristics of Graves' disease are represented by hyperthyroidism, diffuse goiter, and Graves' ophthalmopathy (GO) (10). The primary clinical presentation of GO includes eyelid retraction and periorbital edema (11). Both signs may mimic soft tissue proliferation and coarsening of the facial features, characteristic of acromegaly (12). Other signs, such as increased sweating, heat intolerance, visual disturbances, and fatigue, are common both in hyperthyroidism and acromegaly (12). To summarize, a diagnosis of acromegaly may A rare form of hyperthyroidism leading to the diagnosis of acromegaly: A case report be overlooked, due to certain shared symptoms with Graves' disease. Acromegaly and hyperthyroidism have distinct pathophysiological mechanisms and hormonal profiles; biological parameters, such as those related to phospho-calcium metabolism, exhibit similar changes between the two diseases (12). In acromegaly, higher serum phosphate (due to increased renal tubular resorption) and increased plasma calcium levels, together with hypercalciuria, most likely as a result of GH or IGF-I direct actions, are observed (12,13). Hyperthyroidism is also associated with a catabolic state that leads to bone resorption and hypercalcemia (14). High levels of serum calcium lower the levels of parathormone and, as a consequence, increased renal tubular reabsorption of phosphorus, leading to high levels of this electrolyte are observed (14). The clinical case presented here describes the difficulties in diagnosing acromegaly due to its unusual association with two distinct thyroid pathologies. Case report A 47-year-old female patient was admitted to the Department of Endocrinology of the Emergency County Hospital Timisoara in October 2018, with the following symptoms: Hyperhidrosis, fatigue, paresthesia in the upper left limb, dyspnea, dysphagia, and a gritty sensation in the eyes accompanied by conjunctival edema and erythema. The personal history of the patient revealed that she was diagnosed with subclinical hyperthyroidism in 2012 (etiology not mentioned) and was treated with antithyroid drugs for 4 years. At that time, she was also diagnosed with hypertension and received treatment with β-blockers and angiotensin-converting-enzyme inhibitors. In 2016, the patient underwent neck ultrasound and thyroid scintigraphy resulting in a diagnosis of toxic adenoma and was thus treated with radioactive iodine. Later that year she developed hypothyroidism, for which she was given levothyroxine until September 2018. Therapy was discontinued due to the lab results that revealed hyperthyroidism and the patient was transferred to the development of bilateral ophthalmopathy. During the first admission to the Department of Endocrinology (October 2018), the clinical examination of the anterior cervical region revealed a palpable painless nodule of ~5 cm in the right lobe, which had increased consistency and was mobile while swallowing. The following abnormalities were also noted: Bilateral symmetric exophthalmos and increased blood pressure (160/80 mmHg). The neck ultrasound detected a hypoechoic thyroid parenchyma with an increased volume of the thyroid (26.8 ml, upper normal limit <16 ml). The right lobe was completely replaced by a spongiform nodule, with increased vascularity (Fig. 1), which was classified as benign by ACR-TIRADS (score 2) (15). The left lobe exhibited increased vascularity and, close to the posterior capsule, there was a partially cystic nodule of 6/5/7 mm (Fig. 2), also classified as benign by ACR-TIRADS (score 2). Laboratory tests confirmed that the patient presented subclinical hyperthyroidism, with low levels of TSH (0.025 mU/l; normal range, 0.55-4.78 mU/l) and normal levels of FT4 (14.93 pmol/l; normal range, 11.50-22.70 pmol/l), and FT3 (5.98 pmol/l; normal range, 3.54-6.47 pmol/l). Given the medical history of the patient and the morphological aspect of the right lobe, a thyroid scintigraphy was performed. It showed increased Tc 99 m-pertechnetate uptake in the right lobe, relative to the extra-nodular thyroid tissue, an aspect that confirmed the diagnosis of toxic adenoma (Fig. 3). The presence of the bilateral ophthalmopathy and increased vascularity of the thyroid left lobe raised the suspicion of Graves' disease, thus the levels of anti-TSH receptor antibodies were measured; their high titer (>40 IU/l, normal values <1.75 IU/l) confirmed this diagnosis. The patient had normal values of ATPO antibodies (48 IU/ml; normal range, 0-60 IU/ml) and high levels of antithyroglobulin antibodies (>500 IU/ml; normal range, 0-60 IU/ml). Given the fact that the patient denied the consumption of drugs containing phosphate, the normal renal function, and normal values of serum calcium, 25(OH)vitamin D, and parathormone, the most common causes of increased serum phosphate levels were excluded. A more detailed history and targeted clinical exam were performed. They revealed the following: Macroglossia, increased interdental spaces (Fig. 4A), enlarged lips, and prominent nasolabial folds (Fig. 4B). The patient also mentioned she was not able to take off her wedding ring for ~5 years. The suspicion of acromegaly was raised and, consequently, IGF-1 (891 ng/ml, normal range for age and sex 83.3-220 ng/ml) and basal GH (20.2 ng/ml, normal value for age and sex <10 ng/ml) were determined. A suppression test with 75 g glucose administered orally was also performed, showing the failure to suppress GH below 1 ng/ ml and thus confirming the diagnosis of acromegaly. The evaluation of the global function of the pituitary gland did not identify other hormonal deficiencies (prolactin=4 ng/ml, normal range=2.8-29.2 ng/ml; FSH=75.98 mIU/ml, normal range for menopause=23.0-116.3 mIU/ml; LH=43.05 mIU/ ml, normal range for menopause=7.9-53.8 mIU/ml; estradiol <11.80 pg/ml, normal range for menopause <32.2 pg/ml; and cortisol=12.38 mcg/dl, normal range=4.3-22.4 mcg/dl). A pituitary MRI was performed to detect and characterize the adenoma. It revealed a macroadenoma (13/9 mm), hypointense compared to the normal pituitary parenchyma (in T2-weighted signal), which invaded the left sphenoid sinus, without imprinting the optic chiasm (Fig. 5). To treat the hyperthyroidism, the patient was administered thiamazole 10 mg daily. Ophthalmopathy was classified according to the European guidelines (16) as active (clinical activity score, 3) and mild (Fig. 6). Consequently, a watchful strategy was implemented, and the patient received local treatment. Therapy of the pituitary pathology was prioritized and thyroid function and morphology were reassessed thereafter. Surgery was recommended and the patient was referred to the Neurosurgery Department. The patient was readmitted to the Department of Endocrinology in March 2019, 12 weeks after surgery (transsphenoidal adenectomy). Lab tests revealed high levels of IGF-1 (474 ng/ml, normal range 81.8-219 ng/ml), high levels of basal GH (9.14 ng/ml, normal range 0.05-8 ng/ml), and lack of suppression of GH <1 ng/ml during an oral glucose test with 75 g glucose. MRI was performed and a residual tumor was identified in the sphenoid sinus. The disease was considered active and therapy with octreotide (a somatostatin analogue) 20 mg intramuscularly (IM) every 28 days was initiated. The level of serum phosphate was 4.5 mg/dl (normal range, 2.5-4.9 mg/dl). Blood glucose and glycated hemoglobin were normal, and the response to oral glucose (blood glucose at 120 min=87 mg/dl) were normal, as well. The patient was screened for complications associated with acromegaly and was diagnosed with severe sleep apnea. The other screening tests were unremarkable. The patient did not present any gastrointestinal complications. The re-evaluation of thyroid function revealed hypothyroidism (low TSH levels associated with low FT4 levels), thus thiamazole was discontinued (the last daily dose was 5 mg). The thyroid aspects in ultrasound and the status of the ophthalmopathy did not change. In June 2019, 3 months after therapy with somatostatin analogues had been started, the patient was re-evaluated. Even though the clinical status improved (Fig. 7), the response to therapy was poor. Lab tests revealed high levels of IGF-1 (658 ng/ml, normal range 81.8-219 ng/ml), high levels of GH (8.49 ng/ml, normal range 0.05-8 ng/ml), and a lack of suppression of GH below 1 ng/ml during an oral glucose tolerance test; MRI remained unchanged. Taking into consideration the blood tests and the presence of the residual tumor, the dose of octreotide was increased (30 mg IM every 28 days) and cabergoline was added (2 mg orally every week). The function and morphology of the thyroid were reassessed, showing mild subclinical hyperthyroidism, while the neck ultrasound was unchanged. The ophthalmological consultation classified the eye disease as inactive and moderate (clinical score activity=1/10, Fig. 8) (16). Due to the history the thyroid disease and the tracheal compression by the thyroid nodule (proven by radiological studies), the patient was referred to an endocrine surgeon for total thyroidectomy. After surgery, the evolution was favorable and therapy with levothyroxine was initiated. The pathological result confirmed the concurrence of two distinct thyroid diseases in addition to the degenerative changes induced by therapy, a hyperfunctioning goiter with pronounced degenerative changes, imprecisely contoured adenomatous nodules, with areas of hyalinization and interstitial calcification, cystic degeneration, and areas of angiomatosis primarily expressed in the right lobe of the thyroid (Fig. 9A and B). The patient was re-evaluated periodically and, 9 months later, she was unresponsive to medical treatment. At this point, pegvisomant (up to 20 mg daily), was initiated which was well tolerated. After 6 months of therapy, IGF-1 was normal, and the clinical status of the patient improved significantly. The patient did not exhibit any adverse reactions, and the pituitary tumoral rest did not increase. Discussion Here, the case of an acromegalic patient, who was initially admitted to our department for hyperthyroidism, is described. After routine lab tests, acromegaly was suspected and confirmed thereafter by a dedicated workup. The etiology of hyperthyroidism was challenging due to the coexistence of both toxic adenoma and complicated Graves' disease. Secondary hypothyroidism was excluded due to the medical history of the patient, normal FT4 levels, positive anti-TSH receptor antibodies, thyroid ultrasound (thyroid volume was increased), and scintigraphy showing increased uptake in the right lobe (17,18). As a result of the insidious evolution of acromegaly, its signs and symptoms may remain unnoticed for long periods of time. In our case, due to the concomitant thyroid pathology, certain aspects of the clinical exam that suggested a diagnosis of acromegaly in a patient who had no complaints related to tumoral compression (headache, blurred vision, or vomiting) were overlooked. During routine lab workouts, a high level of serum phosphorus, accompanied by normal calcium and parathormone values, leads to a more detailed anamnesis and physical exam. After asking questions that targeted the signs and symptoms of acromegaly, the diagnosis was considered more seriously. The association between acromegaly and Graves' disease is rare, but may be due to a series of clinical manifestations such as acral enlargement, soft tissue overgrowth, coarsening of facial features, and joint pain. If these manifestations are discrete or absent, as was described in the present case, the diagnosis can be challenging. In such situations, even a minimal biological change, such as hyperphosphatemia, should receive more attention. Hyperphosphatemia is not uncommon among these patients, due to the increased tubular reabsorption of this electrolyte, given the action of IGF-1 on the apical membrane sodium-phosphate IIa cotransporters (19). It was hypothesized that phosphate may reflect the status of acromegaly, but in patients with preoperative serum phosphate >4.5 mg/dl, no correlation was found (20). Nevertheless, its decrease after 3 months may predict remission in patients with discordant GH and IGF-1 values (20). Conversely, a study published in 2017 revealed that hyperphosphatemia at diagnosis was more common in patients who had an active disease after therapy (21). Acromegaly is a rare disease; however, its association with thyroid pathology is not uncommon (3). Nevertheless, the coexistence of Graves' disease and acromegaly has a very low prevalence (<1%) (22). IGF-1 has a synergic effect with TSH on thyroid cell proliferation and may mediate an immunological response by stimulating the proliferation of T cells and B cells, while GH might stimulate the production of TSH receptor antibodies (22). All these mechanisms play a role in thyroid growth and can influence the severity of Graves' disease. The presence of both toxic adenoma and Graves' disease (Marine-Lenhart syndrome) is a very rare occurrence, with a prevalence of 1-2.5% (23), and their association with acromegaly has not previously been reported in the literature, to the best of our knowledge. Given the fact that the patient did not undergo a full medical check-up prior to being referred to our department, it was suggested that the radioiodine therapy may have induced Graves' disease (24). A study that included 191 patients with nontoxic goiter treated with 131 I found that 5% of the patients developed hyperthyroidism without eye involvement after 3 months of therapy, and this was more frequent in those who had elevated ATPO antibody levels. The suggested mechanism that induced autoimmunity was the release of antigens from the follicular cells (25). A retrospective study in patients treated with 131 I for toxic nodular goiter reported a prevalence of hyperthyroidism associated with high levels of anti-TSH receptor antibodies in 4% of the cases (26). Several risk factors are associated with hyperthyroidism, such as genetic predisposition and high levels of ATPO antibodies before therapy (25). In our patient, titers of ATPO antibodies were not evaluated before 131 I therapy, but they were increased when they were measured during the first hospital admission. In addition, the patient had high levels of antithyroglobulin antibodies, thus this type of antibody as a risk factor for developing an autoimmune disease in this case could not be excluded. Orbitopathy is a known extrathyroidal complication of Graves' disease, with an estimated incidence of 16 cases in 100,000 women, usually being present at the onset of the disease (27). The patient in the present case report provided little to no information regarding the onset of her eye symptoms, thus we were not able to establish a timeline regarding its evolution. A study by Coutu et al (28) found that high levels of GH were associated with extraocular muscle enlargement, a situation that may have been present in our case. After pituitary surgery, the condition of our patient's eye disease improved, without specific medication, an aspect also mentioned in the aforementioned study. Hypertension is a common complication of acromegaly, with a prevalence of 50% in patients with active disease (29). Its severity is related to the excess of GH and IGF-1, which cause expansion of plasma volume and increased systemic vascular resistance, leading to elevated diastolic blood pressure (29). Conversely, hypertension caused by hyperthyroidism is mostly due to increased systolic blood pressure as a result of renin release and sodium reabsorption (30). The patient in the present report had high blood pressure, most likely due to intricate mechanisms related to acromegaly and hyperthyroidism (stimulation of smooth muscle cell growth, increased arterial stiffness, renin release and sodium reabsorption) (29,30). Her blood pressure lowered, and the dosage of the oral medication was reduced after treating the acromegaly and hyperthyroidism. Diagnosing and treating the patient, in this case, was a challenge. She did not respond to somatostatin analogues, even though she had predictors for a better evolution, such as being a female of childbearing potential and presenting a hypointense T2-weighted MRI signal of pituitary adenoma (31,32). There are several factors that can influence the response to somatostatin analogues, such as a lack of somatostatin receptors type 2 and 5, Ki-67 nuclear protein, or cytokeratin staining pattern (33). However, they could not be analyzed in the present case. It has been noted that 25% of the patients do not respond to somatostatin analogues after 12 months (31). According to the current guidelines, our patient was completely unresponsive to therapy (34). Given the fact that her glucose metabolism was impaired before surgery and that she did not present any criterion for remission after the therapy with somatostatin analogues, we chose to initiate pegvisomant. ACROSTUDY showed that a dose of 18 mg/day was effective in 65% of patients treated with a GH receptor antagonist (35). In addition, complications associated with the treatment were very rare: 2.2% of patients reported local discomfort, lipoatrophy, or reversible lipohypertrophy; 3-5% had tumoral growth; and 9% presented liver enzyme abnormalities (34,35). In the present case, the patient responded well to the therapy and did not show any side effects. Due to the presence of multiple endocrine pathologies associated with glandular hyperfunction, the presence of a genetic defect in G-protein-coupled receptors cannot be excluded. However, its more frequent association with congenital endocrine pathologies, as well as the absence of a particular phenotype of the patient, made this hypothesis unlikely (36). In conclusion, acromegaly remains difficult to diagnose and treat, because it has a slow evolution and some of the clinical findings might be assigned to the patient's comorbidities. In such cases, all the clinical and biochemical abnormalities that may suggest a diagnosis of acromegaly should be considered. Acknowledgements Not applicable. Funding No funding was received. Availability of data and materials The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request. Authors' contributions MV, ISP, FV, MB, IG, AV, DA and MC collected, analyzed and interpreted the data. ISP, MB, IG, DA, AV, and MV wrote and critically revised the manuscript. All authors have read and approved the final manuscript. All authors agreed to be accountable for all aspects of the work. ISP, MV, IG, and AV confirm the authenticity of all the raw data. Ethics approval and consent to participate Not applicable. Patient consent for publication The patient provided written consent for the publication of their data.	2023-08-24T15:02:32.736Z	2023-08-22T00:00:00.000	{ "year": 2023, "sha1": "4b03d03fefe2cb07c139adcce56394cce292c1da", "oa_license": "CCBYNCND", "oa_url": "https://www.spandidos-publications.com/10.3892/etm.2023.12176/download", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "adfa0cac8037eeb4f4ce5c013802f7e1e03466d9", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [] }
56056303	pes2o/s2orc	v3-fos-license	Pedagogical integration and regionalisation The purpose of this research is concerned with the intentions of pedagogical integration and the notion of an expanded concept of teaching as described in the National School Curriculum of Sweden. In the analysis a theoretical synthesis comprising critical discourse analysis and sociological theories of institutional identities and processes is used through a policy ethnography approach. Purpose has been focus on the relationship between the formal, written curriculum and specific, local, classroom practices. The analysis shows that the Pedagogical dimensions of integration were transformed toward economic interests and control at the local political level. Education work is distinctly regionalising into two so-called safety zones, and a hybrid zone. Four different pedagogical identities are also identified in the analysis. The analysis thus suggests that the creation of safety zones restricts education work to pre-existing patterns. Only the hybrid zone is presently carrying a potential to significantly renew education and encourage and social change. ernment Proposition 1997Proposition /1998, s. 94), s. 94).This was a strategy to concretely and organizationally support the development of integration between the school and pre-school activities by providing a common site.A short analysis show that different policy texts from 1940 to 2006 motivated and represented the notion of integration and how this representation has changed.The changed has gone since one emphasizing developmental psychology and a collected curriculum code to one emphasizing integration and life long learning.The analysis shows and describes four discursive periods against pedagogical integration.The first of these is from 1940 to 1969 and is dominated by developmental psychology.The second period is from 1970 to 1990.During this period developmental psychology is challenged by pedagogical discourse emphasizing the need for pedagogical coherency between pre-school and school.The third period stretches from 1990 to 1995 and can be described as a period of transition.A period emphasising common goals and pedagogical coherence on the one hand to the concrete integration of activities on the other.It was within this period that the idea of a new school based on the ideal of life long learning was developed.The fourth period runs from 1995 to 2006 1 .It marks the further extension of discourses of integration from the subsequent period. Purpose of the research In the National School Curriculum (Ministry of Education, 1998) the intention of pedagogical integration between the school and pre-school is expressed.The aim for this pedagogical integration reform is to creating conditions for life long learning.The purpose of the research is concerned with the intentions of pedagogical integration and the notion of an expanded concept of teaching as described in National School Curriculum.The main focus has been on how different actors in the local arena consume, produce, interpret, negotiate and implement the National School Curriculum in their everyday local practices.The research has thus had a main analytical interest for the pedagogical integration between the formal compulsory school and pre school class.My interest behind this integration reform can be expressed as being an aim to relate policy as text (National School Curriculum) and policy as social practice to one another.The aim is to highlight social reproduction and social change in relation to reform intentions re-garding pedagogical integration and an expanded concept of teaching. Policy ethnography and analysis The present research has been carried out as a multi-layered analysis to describe and analyzing policy processes.Ball and Bowe (1989) have termed this kind of analysis as a policy ethnography.They described its distinctive purpose as being to focus on the relationship between the formal, written curriculum and specific, local, classroom practices.The goal here is to develop a link between detailed ethnographic descriptions of the social and discursive practices taking place within institutional practice and analyses of the broader education system. Using policy ethnography has had several concrete implications for the research.It has meant that I have participated in the interactions and day to day practices at different levels of the education system, particularly at the local level.In this policy ethnography I have work with the help of discourse analysis, interviews, field conversations and ethnographic observation.This has meant that, as Ball (1994Ball ( , 2008) ) suggests, the investigation has at times been unruly, difficult to predict and analytically complex.However, in the case of complex processes of policy formation, I believe that to take one single and simple analytical point of departure would be doomed to analytical failure.This failure would induce a thorough mismatch between the complexity of what was being observed and the simplicity of its analytical representation. 2 A theoretical synthesis comprising critical discourse analysis (Chouliaraki & Fairclough, 1999;Fairclough, 2003Fairclough, , 2009;;Wodak, 2009) and a combination of sociological and socialpsychological theories of institutional identities and processes from Bernstein (2009), Giddens (1984) and Stones (2005) has been used in the research.This synthesis has been necessary to analyse both the discursive and the non-discursive moments within the social practices of the investigated institutions.Critical discourse analysis has been used to analyse political texts and written and spoken texts of other kinds as well as the discursive dimensions of social practices.The sociological and social-psychological concepts and theories have been used to analyse the non-discursive moments of the same social practices.The analysis of the local education can in this way be described as a synthesis.The local education system is described and analysed in relation to different levels and the different arenas.I have an interest of how local actors try to construe meaningful conceptualisations of the intentions of the National Curriculum and other political texts of their acquaintance and how concrete pedagogical practices are constituted. THREE LEVELS OF ANALYSIS My theoretical synthesis policy processes are conceptualized as communicative events that take place in three levels.With the help of Chouliaraki and Fairclough (1999), Fairclough, (2003, 2009), Wodak (2009) and Ball (1994Ball ( , 2008) ) policy is defined as text, discursive practice and social practice.These three concepts have been of central importance in the definition of policy, the production of data and the processes of analysis.However, it has to be noted that the difference between these concepts is quite significant, even though as Ball suggests policy is not one or the other of them but rather comprises all three.In his view the concepts of text, discursive practice and social practice are implicit in each other and the discursive practice can be discerned as a mediating level.This has also been my point of departure.The three concepts have been more concretely applied in the following way: Policy as text has been used to refer to the written national and state documents and local policy documents.The documents came from the municipal council, the city district education office, the school and the teaching team.Central concepts at this level of analysis are those of inter-textuality, inter-discursivity, the discursive order, ambivalence and hegemony. Policy as discursive practice is related to the production and consumption processes that are tied to policy as text.Concerning discursive practices the research has focussed on how policy texts are produced and consumed within the local education system by teachers, recreation and leisure educators and pre-school teachers.Subject position, identity and meaning making are important concepts at this analytic level. Policy as a social practice is premised on the notion that communicative events are a social practice and that the relationship between texts and social practices is mediated by discursive practice.Chouliarki & Fairclough (1999) and Harvey (1996) have suggested that social practices 3 comprise both discursive and non-discursive moments and that these are dialectically re-lated to each other.It is also important to note that each moment internalises the other moments without being reduced to them.Because of this my interest can be described as being for Ravenhill School's social relations and identities, rules, regulations and power relations.Another important aspect of the level of social practices is the social material aspect of time and space.Time and space provide both limits and possibilities for educators to work according to their interpretations of curriculum intentions. Data production and analysis In my data production local and central policy text has played an important role.Data has also been produced by ethnographic observations (Hammersley & Atkinson 2005) in classroom, corridors and teachers' staff room.I have also made interviews with the local municipal manager, the local personal manager, the local finance director, headmasters, teachers, preschool teacher and recreation and leisure educators.The data has been summarized and analyzed through meaning concentration (Kvale 1997).In this process the concepts of pedagogical integration and expanded teaching are the key concepts in the investigation.The concept of extended teaching has a central position in the analysis and is examined in relation to pedagogical integration.This is because the concept of pedagogical integration advanced within the curriculum depends on the development of a broadened concept of teaching.The broadened concept (of teaching) has an effort to change the emphasis in teaching.From a traditional transmission model to one characterised by a holistic view of learning that emphasises active participation in knowledge production rather than passive reception.However, the broadened concept of teaching developed and expressed in the curriculum also has a somewhat eclectic theoretical foundation comprising socio-culturalist, cultivationist (bildung), progressivist and social constructivist elements.It can thus be described as a hybrid concept that is open to different interpretations. In concrete terms, by being a policy ethnographic study, the investigation behind the research has been carried out as a case study of pedagogical integration and expanded teaching at one particular school.This school has been called Ravenhill School from the Moe Valley School District in Greater Göteborg.Moe Valley is one of 21 city districts in Göteborg communi-ty.Both Ravenhill and Moe Valley are pseudonyms.This is in keeping with ethnographic tradition.Ravenhill School comprised formal compulsory school classes, pre-school classes and a recreation and leisure centre.Activities within formal school classes and the pre-school class have been researched in detail in the present investigation whilst those of the recreation and leisure centre have not.The activities in one teaching team, Ash, have been focussed on in particular.These activities mainly involved three teachers, three recreation and leisure educators, one pre-school-teacher and 75 children (between 6 and 12 year) and indirectly their parents or guardians. My ethnographic findings are presented in four parts.In the first part I focus on local policy texts from Göteborg community and Mae Valley city district with an aim to describe the process of pedagogical and economical restructuring.In the second part: Educational discourses, intertexuality and retraditionalisation, I show through a discourse analysis how the discursive practices of the school are structured by two educational discourses.Within Child centred pedagogy, thematic work and discursive struggle I describe the discursive struggle in the teaching team Ash around child centred pedagogy and thematic work.In Pedagogical identities and regionalisation as will be the fourth part I analysis the social practice of the teaching team Ash.I complete the article with a discussion of my findings and with a special interest against social reproduction and social change. Pedagogical restructuring and economic interests In the beginning of the 1990's an extensive reform of the administrative and political organisation of Greater Göteborg was conducted in a city district reform.This reform was initiated by the Municipal Council to improve political control and economical expediency and to encourage integration and decentralisation in decisionmaking.Through this reform Göteborg was administratively and politically divided into 21 new city districts.A more horizontal organisation was aimed for, with a shorter decision making channels and enhanced flexibility.This was in tune with developments of New Public Management (Almqvist, 2006) and what were taken to be new demands.Through the reform Göteborg's Municipal Council came to regard the 21 City Districts as local, independent authorities with responsibility for development, amongst other things, their own School Development Plan.The central position of political power in the City of Districts is held by the chairperson of the District Council (Stadsdelschef).He has responsibility for the organisation and implementation of political and economic administration, including the school leadership, education financing and aims realisation. Within Moe Valley the organisation of the political administration of education took on a particular form and became very much fashioned after an economic model and a form of organisational integration between school and pre-school.These thoughts were expressed by the Local municipality manager and personal manager: -In the new organization we work with the different solution of organizational and economic integration.About the school and preschools there has been much of the different solution of integration (interview, Sven Holm, Local municipality manager). -We work hard on different economic savings.In that work integration become a key issue.All manager of a middle level has to work with different models of integration (interview, Lisa Strand, personal manager). At the same time Moe Valley tries to create an integrated and horizontal, decentralised organisation for education and pare-school with a significantly reduced number of management levels and middle managers.Characteristic for Moe Valley was also the development of independent teaching teams comprising recreation and leisure educators, pre-school teachers and school teachers.This organisation has some resemblance to the thoughts of knowledge base economy.These thoughts and ideas are described by Jessop, Fairclough and Wodak (2008) plus Peters, Marginson, and Murphy (2009) that are expressed as structured towards the development of decision making expediency and organisational flexibility. Moe Valley is administratively divided into a number of sub-districts.Each of these comprises integrated educare units that are formed by schools, pre-schools and other child day-care facilities.During the period of the investigation these integrated units were economically steered through a resource distribution model (RDM) that determined which resources were made available when to which units.This RDM was locally termed as a machine of raw pedagogy by educators, pre-school teachers and the recreation and leisure educators providers in the district.They experienced it as excessively determining the level and character of the services that were offered. This feeling was complemented in the investigation by data relating to the domination of economic steering in Moe Valley and the absence some pedagogical discourse within curriculum decision-making amongst political leaders there.In Moe Valley pedagogical deliberations were excluded from the political decision making.The integration discourse was some hybrid discourse.Dominated by an organisational and an economic discourse respectively an neo-liberal 4 , and post-Fordist 5 discourse the neo-liberal elements emphasised individual autonomy, selfdetermination and economic control.Post-Fordist discourse emphasised flexibility, a decentralised organisation, reductions in management and the development of self-monitoring working group.The goals determined central aims and the local realisation of aims was determined through locally negotiated forms of practice.In this combination of elements the discourse of restructuring is a mirror of central reforms directed toward changing forms of steering in schools and improving the efficiency of the education system.However, it has to be emphasised that the discourse did not take up issues of content and the aims of education.The pedagogical dimensions of integration and restructuring were transformed toward economic interests and control at the local political level. EDUCATIONAL DISCOURSES, INTERTEXUALITY AND RETRADITIONALISATION The discourse analysis show two educational discourses that structure the pedagogical activities at Ravenhill School are described in particular.I have termed these discourses as the collected and integration discourses (Bernstein, 2009) respectively.The collected discourse is characterised by references to the value of strong classification and framing in the curriculum and in education work, whilst the integrated discourse is characterised by references to the value and need of weak classification and framing (Bernstein, 2009).These two discourses reflect distinctive perspectives in education that are related to the ways in which the different actors view the curriculum, subject specialisation, team teaching, knowledge and forms of work.The discourses however exist relatively independently of the curriculum and can also be described as two distinctive historical constructions.See table 1.At Ravenhill School the schoolteachers were the main bearers of the collected discourse whilst the integrated discourse was more often borne up by the pre-school teachers and the recreation and leisure educators.One of the teaching teams at Ravenhill, Star, only comprised schoolteachers.In this teaching team there was a relative consensus regarding the value of the collected discourse, whilst within Ash, a temporary compromise existed that allowed the two discourses to coexist.Regarding the Head teacher at Ravenhill, he adhered to three discourses at different times: an integrated discourse, a collected discourse, and the hybrid discourse of restructuring. The discourses operate within the day-to-day life of the education community at Ravenhill and in Ash in particular, within the teaching teams concerning the development of local subject plans.These were shown to be formed, mainly with the help of older plans, course textbooks and the experience of the educators involved.In this way the local subject plans were thus restricted by experience and intertextuality rather than formed through reflection over curriculum aims.One further issue,which emerged in the analysis, was that teachers emphasise the aims of the various course syllabi whilst preschool teachers and recreation and leisure educators emphasise methods rather than aims.In short we could state that work with the local subject plans is characterised by conservation and retraditionalisation and that the subject plans replace the local school plan in the governing of day-to-day work. CHILD CENTRED PEDAGOGY, THEMATIC WORK AND DISCURSIVE STRUGGLE The teaching team Ash developed local aims for its education practices in relation to the concepts of theme and time.In more concrete terms the research describes how a consensus developed at Ash around local aims and a child centred pedagogy.However, it also shows that this consensus was somewhat shallow.The agreement between different parties did not penetrate beyond surface details because issues such as basic education aims and the value foundations of education were not discussed.In fact consensus is maintained within the team through an avoidance strategy where arguments about value loaded issues in relation to education aims and the formations of practice are avoided.Some of the educators from the team has expressed that they would value these deeper kinds of argument and debate. About the collected and integration discourses mentioned earlier, two clearly distinctive definitions of theme have been identified in the analysis.The schoolteachers tend to bear up a definition that maintains the value of individual subjects and that suggests that themes should correspond to the subject knowledge of areas like civics and social science.Pre-school teachers and recreation and leisure educators tend to define theme with integration between subjects.Two distinctive discursive strategies have also been identified in the analysis, corresponding to the struggle over definitional rights between the purveyors of the two discourses.The school teachers have used a strategy where they try to bypass the need for an agreement over the idea of theme by developing long-term plans in subject areas that pre-structure thematic work. -I want a pedagogical plan for year. Whit out a plan I do not now what we will do in the classroom. In ordinary cases we usually look back and se what we have do in the subject (Stina, teacher). Pre-school teachers and recreation and leisure educators have tried to encourage conversations to develop in which to discuss and define the concept of theme. -We must ask our self how we think about knowledge and learning and our definition of theme (observation, Maud, recreation and leisure educator). As concerning the concept of theme, a discursive struggle exists regarding ways of conceiving and relating to the issue of time.This conflict concerns time for planning and reflection (so called independent or development time: in Swedish förtroendetid), which is considered to be unfairly distributed by pre-school teachers and recreation and leisure educators. Discussions about time within the teaching team are orchestrated from three distinctive discursive positions within the discursive order of Ravenhill.Pre-school teachers and recreation and leisure educators argue from a foundation in an integration discourse.They want a new system for the regulation time that treats the teaching team as a unit and distributes time-resources equally between team members. -... we have asked (recreation and leisure educators and pre-school-teachers) for more time for pedagogical planning.We need more time for reflection, planning observation and evaluation.Today we do not have this time.Instead we must work overtime if we want to do all necessary things (Olle, recreation and leisure educator). Pre-school teachers and recreation and leisure educators feel that teachers have time for planning and reflection but that they do not.They also feel that their own work has been intensified through integration and that this has led to increased tiredness and stress.About this work situation the teachers refuse to help the preschool teachers and recreation and leisure educators and work inside the pre-school class and recreation and leisure centres.The teachers want to maintain the current regulation of time and the current distribution of resources about time.The solutions for the teachers are instead a union contract for regulation of working hours: -We cannot solve that problem without more time and money.Perhaps we can help them but we need a new union contract for working hours (Stina, teacher). With a union contract for regulation of working hours there is a possibility of employing one more pre-school teacher.This would help the teaching team Ash to solve the time problem.Teachers depart from a collected discourse (Bernstein, 2009) position.From that position they mean that the team members from schools, pre-schools and recreation and leisure centres should be administratively acted towards as separate and as belonging to separate organisations.This will include different needs concerning the allocation of planning time and time for reflection; as at present.The Head-teacher argued from a position rooted within a restructuring discourse stressing the need to minimise the regulation of time in favour of a more flexible use of time. -As a teacher you must find a solution.It's important what you do and how you spread the expenditure and recourses.The professions have changed a great deal.The question is how we manage that.Time for planning and reflection is not a contract issue.You must need it in practice (Rolf, Head-teacher). The Head-teachers perspective corresponds with a discourse of New public management (Almqvist, 2006) with a focus on decentralization and flexibility.In this position the deregulation of time is stressed and educators are given the possibility of regulating the use of time within teaching teams. PEDAGOGICAL IDENTITIES AND REGIONALISATION Four different pedagogical identities are identified in the analysis concerning the concrete education work in Ash.These are termed the transmitter (förmedlaren), the helper (hjälparen), the role model (förebilden) and the controller (övervakaren).The constitution of the positions of the controller and transmitter is resonant with the discursive position of the collected discourse and its elements of a strong classification and framing.The constitution of the positions of helper and role model is resonant with the discursive position of the integrated discourse and its elements of a weak classification and framing.These four pedagogical identities and their concomitant forms of practice thus present a picture of education work in Ash as distinctly hybridised or mixed.There is no new construction of identity or practice!Rather, education has a reproductive character and the extended and broadened concept of teaching is compromised. The education work at Ash is distinctly regionalized into two so-called safety zones 6 , an hybrid 7 zone and a marginalized 8 zone.The teacher's safety zone corresponds (epistemologically and physically) to the subjects of mathematics, Swedish, English and pupil research whilst the pre-school teacher and the recreation and leisure educator safety zones are the preschool class.The hybrid zone consist of thematic work involving both teachers and Pre-school teachers and recreation and leisure educators.The collected discourse contributes to the structuring of the teachers' safety zone whilst the integrated discourse works resonantly in the preschool teacher and recreation and leisure educator safety zone.The hybrid zone is structured with the help of a discursive mix where the integration and collected discourses operates in par-allel.In these social practice teachers, pre-school teachers, recreation and leisure educators work together against a jointly aim.This articulation will also open up for a potential space for change and a new pedagogical order.The educa-tion work in Ash is structured by different rationalities (Ve, 1999), patterns of teaching and pedagogical identities.These patterns create three pedagogical zones and are described in table 2. Regarding art, music, data and digital media the teachers at Ash have given up this part of the curriculum to the pre-school teachers and recreation and leisure educators.Only pre-school teachers and recreation and leisure educators work in this fourth and marginalized zone.The pre-school teachers expressed her thoughts about the work with data and digital media: -I do not want to work with data any more.I am tired about that.It's not amusing any more.Computers must be integrated in the classroom and be a natural part of pupils work.Not as today.I am feeling like a chaotic and traditional teacher.It's Stina who has planned the lesson for me (Petra, pre-school teacher). In this practice the pre-school teacher and recreation and leisure educators filed marginalized and they do not work in line with their professional identity. The analysis thus suggests that the creation of safety zones and a marginalized zone restricts education work to pre-existing patterns.It is only the hybrid zone whom presently carrying a potential to significantly renew education and encourage and social change. DISCUSSION: SOCIAL REPRODUCTION AND SOCIAL CHANGE The research has an expressed interest in the relationships between social reproduction and social change in education.Through it the intentions of the curriculum in relation to integration Pedagogical identity The transmitter The helper The role model The controller The transmitter The helper The role model The controller and expanded teaching have been shown to be difficult for educators to implement in actual classroom practices.The analysis an also show that social reproduction has mainly taken precedence over social change.This idea has been asserted principally because conventionally articulated discourses have dominated the discursive order and discursive practices of the local arena. Because of that the education work has been predominantly organised around an economic model, without recourse to the intentions of National School Curriculum (Ministry of Education, 1998).However, on closer examination a more varied picture emerges through the richness of the empirical details of the ethnography. In this picture local practices can be characterised with both reproduction and change and aspects of time, space and the embodiment of practice are shown to be significantly important to the outcomes of reform.The implementation of a curriculum is a complex process.Within it the curriculum itself is not only locally interpreted and applied but is also locally negotiated and transformed.However, despite the above pointers, in the present case the primacy of economic steering has still been clear to see.A neo-liberal rationality has dominated from the political level.This rationality speaks in a number of ways in the different dimensions of practice.The present case budget control and the appropriation of raw pedagogy through an RDM have still set profoundly limiting frames on everyday availabilities, activities, possibilities and predisposition's to act.These findings are interesting in relation to the new teacher education program (prop. 2009/10, s. 89).In this policy text the thoughts of neo-liberal rationality is strong.The message is as follows: to create a border between pre-school, recreation and leisure centre and formal school.It's a reform of social reproduction and retraditionalisation. How these new thoughts will be implemented in local schools and fit with the institutional practice will be interesting to follow in the future.Agents have thus been the makers of their own identities, meanings, understandings, activities and histories.As often is the case, these processes have developed under historical and material conditions that are not directly under the control of agents and that may not necessarily be shaped in their own interests.The active agency is in other words a situational condition and a contingent issue to be struggled over, not a uni-versal, historically and materially independent fact of life for all humankind. NOTES 1.An interesting notice about pedagogical integration will be expressed by the new teacher education (prop. 2009/10:89).In this policy new document the ideology of integration is weak.The new teacher education program talks instead of borders between preschool and school.This reform sounds like an echo from the 1980ies The document expressed an institutional separation between pre-school and school. 2. A complex processes of policy formation with negotiation, re-construction and transformation in local curriculum development work. 3. Six dimensions of social practice: They are a) discourse and discursive practice, b) social relations and social interaction, c) power, d) material activities, e) beliefs/values/desires, and f) institutional routines and rituals. The finance director Ulf Lindell in Moe Valley expressed his thoughts about the resource distribution model:	2018-12-11T08:55:28.948Z	2010-12-09T00:00:00.000	{ "year": 2010, "sha1": "0e937d0021fbed65710c8c2a3ff72661ef67dc07", "oa_license": "CCBY", "oa_url": "https://nordiskbarnehageforskning.no/index.php/nbf/article/download/196/198", "oa_status": "GOLD", "pdf_src": "ScienceParseMerged", "pdf_hash": "0e937d0021fbed65710c8c2a3ff72661ef67dc07", "s2fieldsofstudy": [ "Sociology" ], "extfieldsofstudy": [ "Political Science" ] }
218657251	pes2o/s2orc	v3-fos-license	A putative new SARS-CoV protein, 3a, encoded in an ORF overlapping ORF3a Identification of the full complement of genes in SARS-CoV-2 is a crucial step towards gaining a fuller understanding of its molecular biology. However, short and/or overlapping genes can be difficult to detect using conventional computational approaches, whereas high throughput experimental approaches – such as ribosome profiling – cannot distinguish translation of functional peptides from regulatory translation or translational noise. By studying regions showing enhanced conservation at synonymous sites in alignments of SARS-CoV and related viruses (subgenus Sarbecovirus), and correlating with the conserved presence of an open reading frame and plausible translation mechanism, we identified a putative new gene, ORF3a, overlapping ORF3a in an alternative reading frame. A recently published ribosome profiling study confirmed that ORF3a* is indeed translated during infection. ORF3a* is conserved across the subgenus Sarbecovirus, and encodes a 40–41 amino acid predicted transmembrane protein. INTRODUCTION The aetiological agent of COVID-19 is the virus SARS-CoV-2, a coronavirus in the genus Betacoronavirus, subgenus Sarbecovirus. Like other coronaviruses, SARS-CoV-2 has a positivesense RNA genome, around 30,000 nt in size. The 5′ two thirds of the genome contain two long open reading frames (ORFs), ORF1a and ORF1b, which are translated from the viral genomic RNA (gRNA). ORFs 1a and 1b encode the polyproteins pp1a and pp1ab, where translation of pp1ab depends on a proportion of ribosomes making a programmed −1 nt ribosomal frameshift near the end of ORF1a to enter ORF1b. Polyproteins pp1a and pp1ab are proteolytically processed to produce the viral replication proteins (reviewed in Snijder et al., 2016). The 3′ third of the genome contains a number of ORFs that encode the viral structural and accessory proteins. These ORFs are translated from a nested series of subgenomic mRNAs (sgmRNAs) that are produced during the infection cycle (reviewed in Sola et al., 2015). In SARS-CoV-2, these ORFs comprise S, 3a, E, M, 6, 7a, 7b, 8, N, 9b and possibly 10 (Fig. 1a). The S, E, M and N ORFs encode respectively the virus spike, envelope, membrane and nucleocapsid proteins -key components of the virus particle that are conserved among coronaviruses. ORFs 3a, 6, 7a, 7b, 8 and 9b encode accessory proteins, also present in SARS-CoV-1, but less widely conserved across coronaviruses as a whole (reviewed in Liu et al., 2014). These 3′ ORFs each have a corresponding dedicated sgmRNA except for ORFs 7b, 9b and 10. ORFs 7b and 9b appear to be translated from the ORF7a and N sgmRNAs, respectively, via a leaky scanning mechanism (Schaecher et al., 2007;Xu et al., 2009) whereas the translation mechanism for ORF10 is unknown. There is considerable variability between coronavirus genera and subgenera in the complement of 3′-encoded accessory genes (Liu et al, 2014). Even within the sarbecovirus genus, there are differences. For example, SARS-CoV-1 has an ORF3b that overlaps the 3′ region of ORF3a but is absent in SARS-CoV-2. Also, in many human-adapted SARS-CoV-1 isolates, ORF8 is split by a frame-disrupting deletion (He et al., 2004). ORF10 is apparently translated in SARS-CoV-2 (Finkel et al., 2020) but is truncated in SARS-CoV-1. Identification of the full complement of genes in SARS-CoV-2 is a crucial step towards gaining a fuller understanding of its molecular biology, and may also help guide vaccine or other antiviral strategies. This information also facilitates rational manipulation of the viral genome (e.g. for developing replicon systems or for mutagenesis studies). However, short and/or overlapping genes can be particularly difficult to identify using traditional computational approaches. On the other hand, high throughput experimental techniques such as ribosome profiling and high resolution mass spectrometry -while powerful -do not necessarily distinguish between functional proteins, regulatory translation (where it is the act of translation rather than the encoded product that is biologically relevant), and translational noise. Comparative genomics offers a way forward: analysis of patterns of substitutions across alignments of related sequences can be used to reveal the signatures of "hidden" protein-coding genes. Analysis of synonymous substitution rates provides a particularly sensitive technique for identifying overlapping functional elements embedded within protein-coding genes, because such elements constrain synonymous changes which otherwise are selectively more-or-less neutral (Firth, 2014). When combined with the conserved presence of an open reading frame and the conserved presence of a plausible translation mechanism, overlapping genes may be distinguishable from overlapping non-coding elements such as functionally important RNA structures (Jagger et al., 2012;Fang et al., 2012;Loughran et al., 2012;Lulla & Firth, 2020). RESULTS & DISCUSSION A full genome alignment of 54 representative sarbecoviruses (SARS-CoV-1, SARS-CoV-2 and 52 bat coronaviruses) was constructed, and synonymous site conservation was analysed with the program synplot2 (Firth, 2014) (Fig. 1b). The analysis revealed conserved features towards the 5′ end of ORF1a, in the middle of ORF1a, near the end of ORF1a and start of ORF1b, and in many parts of the 3′ region of the viral genome. Most of these conserved elements do not correspond to conserved open reading frames and likely represent functional RNA sequence elements -a case in point being the frameshift-stimulating RNA pseudoknot (Baranov et al., 2005) at the junction of ORFs 1a and 1b. However a peak in conservation within ORF3a stood out since, although short, it coincides with an overlapping alternative-frame AUG-initiated ORF -hereafter ORF3apositioned close to the 5′ end of ORF3a where it might be accessible via ribosomal leaky scanning (reviewed in Firth & Brierley, 2012). Closer inspection revealed that the presence and location of the 3a initiation and stop codons were conserved across sarbecoviruses. The 3a* AUG codon is present in all except one of the 54 sequences, where it is replaced with GUG (MG772933, a bat CoV). In all 54 sequences, there is an A at the −3 position giving a strong initiation context (Kozak, 1986). The GUG in MG772933 may also serve as an initiation site as GUG codons can be utilized for initiation -at a reduced efficiency compared to AUG -when in a strong initiation context (reviewed in Firth & Brierley, 2012). In 52 sequences there are two upstream AUGs -the ORF3a AUG (weak context; C or U at −3) and another AUG also in the ORF3a reading frame (weak context; U at −3); in the remaining two sequences the second of these AUGs is absent. The two weak-context AUGs might enable a proportion of pre-initiation scanning 43S complexes loaded on the ORF3a sgmRNA to leaky scan to the 3a* initiation codon. The 3a* ORF has length 40 codons in all 54 sequences except one, namely SARS-CoV-2, where it is 41 codons in length (Fig. 2). The SARS-CoV-2 putative 3a* protein has a molecular mass of 4.9 kDa and a pI of 10.9. The protein is also predicted to contain a transmembrane domain (Fig. 2). Curiously, the transmembrane amino acids are relative highly conserved, suggesting they may form interactions within the lipid bilayer for example membranedisrupting or membrane-associated signalling activities. Ribosome profiling is a high throughput sequencing technique that allows footprinting of initiating and/or elongating ribosomes at sub-codon resolution and hence the global identification of initiation sites and/or sequence regions and reading frames undergoing translation (Ingolia et al., 2009;Irigoyen et al., 2016). A recent ribosome profiling study of cells infected with SARS-COV-2 revealed 23 novel translated ORFs (Finkel et al., 2020). Ten are very short (≤ 15 codons). Seven of the remainder comprise 5′ extensions or 5′ truncations of previously known ORFs (M, 6, 7a, 7b, 9b and 10). Two are uORFs positioned on the gRNA upstream of ORF1a that may play a role in regulating ORF1a/1b expression as previously proposed for uORFs in other coronaviruses (Wu et al., 2014;Irigoyen et al., 2016). After excluding these ORFs, only four novel translated ORFs remain: ORF3a* (25457..25582; 41 codons), another ORF overlapping ORF3a (25596..25697; 33 codons), an ORF overlapping the S ORF (21744..21863; 39 codons) and a truncated version of the same (21768..21863; 31 codons) (where numbers indicate ORF coordinates in NC_045512.2). We investigated the four overlapping ORFs (yellow rectangles, Fig. 3) and found only ORF3a* to coincide with a synonymous site conservation signal. Moreover, the other novel ORFs were not conserved: in many sarbecovirus sequences they lacked the AUG codon and were interrupted by stop codons (Fig. 3). Although synonymous site conservation can result from overlapping non-coding or coding elements, the conserved presence and conserved positions of the ORF3a* start and stop codons suggests the latter interpretation. Moreover, the ribosome profiling study of Finkel et al. (2020) confirms that the ORF3a* is indeed translated during infection. The combination of comparative genomics showing purifying selection (which to a large extent is synonymous with functional importance) and ribosome profiling showing expession strongly suggests that 3a* is a functional protein, conserved throughout sarbecoviruses. While the known SARS-CoV-2 genes have already been studied in SARS-CoV-1 (reviewed in Liu et al., 2014), 3a* has never before been studied. Clearly additional work with SARS-CoV reverse genetics systems will be required to elucidate the 3a* protein function, but it may eventually provide a new target for vaccine or antiviral strategies. The synplot2 analysis (Fig. 1b) can also reveal other functional elements embedded within the viral protein-coding genes, which may also be worthy of experimental investigation. During preparation of this manuscript, 3a* was independently discovered by Cagliani et al., (2020) (where it is termed 3h) who performed a similar analysis with synplot2 but using far fewer sarbecovirus sequences, and hence achieving lower statistical significance for conserved elements. METHODS Sequences with 100% coverage of, and ≥70% amino acid identity to the SARS-CoV-2 pp1ab sequence were identified with NCBI tblastn (Altschul et al., 1990) on 12 January 2020 and downloaded. These cut-off thresholds corresponded precisely to genus Sarbecovirus. Sequences that did not cover the entire protein-coding region of the genome, and all sequences with >99% amino acid identity in pp1ab to SARS-CoV-1 except a single reference sequence NC_004718. Synonymous site conservation was analyzed with synplot2 (Firth, 2014). For the full-genome analyses, codon-respecting alignments were produced using a procedure described previously (Firth, 2014). In brief, each individual genome sequence was aligned to the SARS-CoV-2 reference sequence (GenBank accession number NC_045512.2) using code2aln version 1.2 (Stocsits et al., 2005). Genomes were then mapped to NC_045512.2 coordinates by removing alignment positions that contained a gap character in the reference sequence, and these pairwise alignments were combined to give the multiple sequence alignment. To assess conservation at synonymous sites, the known virus coding regions were extracted from the alignment (with codons selected from the longer ORF in each overlap region), concatenated in-frame, and the alignment analyzed with synplot2 using a 25-codon sliding window. Conservation statistics were then mapped back to NC_045512.2 coordinates for plotting (Fig. 1). For the ORF S and 3a analyses (Fig. 3), the ORF S and 3a regions were extracted from all 54 sarbecovirus sequences, translated to amino acids, aligned using MUSCLE (Edgar, 2004), and the amino acid alignments were used to guide codon-respecting nucleotide sequence alignments (EMBOSS tranalign; Rice et al., 2000). These alignments were analyzed with synplot2, again using a 25-codon window size. In contrast to the full-genome alignments, all alignment gaps were retained instead of mapping to NC_045512.2 coordinates. Molecular mass and isoelectric point were calculated with EMBOSS pepstats (Rice et al., 2000). Transmembrane domains were predicted with Phobius (Käll et al., 2007). Amino acid alignment of sarbecovirus 3a* sequences. Amino acids are colour-coded according to their physicochemical properties. Asterisks indicate completely conserved columns in the alignment. The predicted transmembrane region is indicated with a pink bar below the alignment. Numbers at right indicate the number of times the particular sequence occurs among the 54 sarbecovirus sequences. † SARS-CoV-1. ‡ SARS-CoV-2. For the sequence beginning with GUG instead of AUG, the genetic decoding (i.e. valine) is shown, even though non-AUG initiation codons are normally expected to be decoded as methionine by initiator Met-tRNA. Figure 3. Conservation analyses of the sarbecovirus S and 3a ORFs. In each plot, the upper three panels show the positions of alignment gaps (grey), stop codons (black) and AUG codons (green) in each reading frame in each of the 54 aligned sequences. Below, is shown the analysis of conservation at synonymous sites (see Fig. 1 caption for details). Novel alternative-frame translated ORFs identified in the SARS-CoV-2 ribosome profiling study of Finkel et al. (2020) are indicated with yellow rectangles; for S, the two in-frame alternative initiation sites are indicated with different shades of yellow. ORF3a* is labelled. Note that only ORF3a* has conserved start and stop codon positions across sarbecoviruses and only ORF3a* coincides with a region of enhanced synonymous site conservation.	2020-05-17T13:07:48.503Z	2020-05-12T00:00:00.000	{ "year": 2020, "sha1": "6601a2bd784dda7c93325c17ec89e38100aeabe2", "oa_license": "CCBY", "oa_url": "https://www.biorxiv.org/content/biorxiv/early/2020/05/12/2020.05.12.088088.full.pdf", "oa_status": "GREEN", "pdf_src": "BioRxiv", "pdf_hash": "6601a2bd784dda7c93325c17ec89e38100aeabe2", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [ "Biology" ] }
219419836	pes2o/s2orc	v3-fos-license	Multiscale Laboratory Study and Numerical Analysis of Water-Weakening Effect on Shale Water-weakening effect is one of the most important factors inducing large deformation and stability problems of shale strata in the Huangjiazhai Tunnel.)e influence of water on shale with various water contents and its mechanism still needs more study. In this paper, the X-ray diffraction and water absorption test were first conducted to determine the influence of mineral components on water-rock interaction. )en, a series of rock mechanical tests, including uniaxial compressive test, Brazilian disk test, and triaxial compressive test, were carried out on shale under dried condition, wetted condition with various water contents, and saturated condition. For the uniaxial compression, tension, and triaxial compression, the softening coefficient could reach 0.38, 0.63, and 0.52, respectively. Based on the macroscale experiments, a numerical case study of tunnel excavation was investigated to evaluate the water effect. Finally, with the aid of scanning electron microscope and failure morphology, the mechanism of waterweakening effect on shale was discussed from microscopic and macroscopic perspectives. )e thickness of the double layer increases with the increase of water immersion time. )is paper provides a set of useful data for reference in construction in shale strata, and a methodology to evaluate water-weakening effect through laboratory and numerical modelling approaches. Introduction In rock engineering, there are external factors influencing the construction process, such as stress state [1][2][3] and stress path [4], temperature [5][6][7], and water. In these factors, water is generally regarded as a disadvantage in construction. Water plays a significant role in the mechanical properties of underground surrounding rocks and the development of landslide hazard [8]. erefore, water poses a great threat to the stability in rock engineering applications. It is recognized that evaluation of the water-weakening effect is a key to engineering classification and project design [9]. e water-weakening effect has been well studied in different types of rocks on their strength and stiffness. Colback and Wild [10] found that the uniaxial strength of saturated sandstone is nearly the half of the dried samples. Vásárhelyi [11] reported a 74% reduction in uniaxial compressive strength of the Miocene limestone and 53% reduction in modulus due to saturation. Similarly, Rajabzadeh et al. [12] and Gajic et al. [13] reported a 70% and 10% reduction in uniaxial compressive strength of carbonate rocks and limestones due to saturation, respectively. Mann and Fatt [14] conducted a series of compression tests on sandstones and found that modulus could be reduced by 6∼19% for the saturated samples. To investigate the effect of water on tensile strength, Brazilian tests [15,16], uniaxial tension tests [17], and point loading tests [18] have been performed. Ojo and Brook [18] and Karakul and Ulusay [19] reported that the tensile strength of saturated sandstone could be reduced by around 60%. Parate [20] studied the effect of water on limestone and attributed the strength loss to a reduction in the cohesive strength. Bell [21] proposed that the strength reduction is related to the porosity change. Hua et al. [22] used Brazilian test to study the fracture mechanical properties under periodic water-rock interactions. Dyke and Dobereiner [23] believe that the bond cohesion in sandstones is influenced by water. Burshtein [24] also observed a structure change to reduce elastic modulus in the saturated sandstone. Li et al. [8] studied the water effect on chlorite schist and applied grouting test. In former research studies, sandstone and limestone have been studied extensively. e mechanical properties of shale are usually focused in the field of oil/gas industry. Over 75% of the drilled formations are made up of shale, and over 70% of the borehole problems are caused by shale instability [25]. e stability is related to the shale intrinsic properties and the drilling fluid. Shale is a kind of clay-rich sedimentary rock and expands during adsorbing liquids or gas [26]. So shale properties are extremely sensitive to water. Jumikis [27] found that the compressive strength of saturated shale was one order of magnitude lower than that of the dry samples. Gutierrez et al. [28], Silva et al. [29], and Lin et al. [30] revealed that a number of wetting cycles dissolved the clay minerals and decreased the strength of shale. Chen et al. [31][32][33] studied the fracture mechanical properties of shales under the effect of water, salinity, pH, and temperature. Considering the water-weakening on shale, a better understanding of strength reduction by water under uniaxial compressive, triaxial compressive, and tensile stress condition needs to be explored in depth. e excavation of the Huangjiazhai Tunnel of Ma-Zhu Expressway in northwest Hubei Province in Central China revealed a shale formation. e average buried depth is about 250 m. Site visit shows the underground strata is rich in water in the form of dripping or linear seepage. ere is a lot of rainfall in this area, which enriches the water content in the strata. Under water wet condition, the shale is classified as soft rock of green gray colour, medium weathering, muddy, and fine laminated structure with developed fissures. e joint surfaces are filled with clay and cuttings. e thickness of each rock layer is 0.1∼0.5 m with muddy cementations. e degree of bonding between the structural surfaces is poor, which results in the weak integrity of the surrounding rocks. In this case, deformation rate of the surrounding rock increases during tunnel excavation. Figure 1 shows the vault crown settlement of the right tube at the stake of YK61 + 810. e accumulated crown settlement increases with monitoring time. e settlement rate increases rapidly at the first 5 days and slows down in the next 25 days approximately. e settlement rate increases again thereafter, and there is no trend towards stability after monitoring for 40 days. Consequently, shale in Huangjiazhai Tunnel has outstanding large deformation due to the effect of water weakening. To service for engineering design of Huangjiazhai Tunnel, Zheng et al. [34] conducted uniaxial test under water immersion. More tests are still needed. Based on the study, in this paper, the X-ray diffraction (XRD) and water absorption test were first conducted to determine the influence of mineral components on water-rock interaction. en, a series of rock mechanical tests, including uniaxial compressive test, Brazilian disk test, and triaxial compressive tests, were carried out on shale under natural condition, wetted condition with various water contents, and water condition. Based on the macroscale experiments, a numerical case study of tunnel excavation was investigated to evaluate the water effect. Finally, with the aid of scanning electron microscope (SEM) and failure morphology, the mechanism of water-weakening effect on shale was discussed from microscopic and macroscopic perspectives. Test Materials and Methods 2.1. Materials. Rock cores were taken from Silurian Longmaxi Formation shale in Huangjiazhai Tunnel. To quantify mineral components with aid of X-ray diffraction (XRD), the shale samples were triturated to powder with a diameter of 10 μm approximately. XRD was conducted on the D8 Advance X-ray diffractometer. e scanning angle and velocity are 3°-50°and 5°/min. Test result of one specimen is listed in Figure 2. It suggests that chlorite and muscovite contribute to more than half of the total components. e two clay minerals are sensitive to water [35,36]. Like other clay minerals, their special crystal structure can produce weak expansion under water immersion. Inspired by Chen et al. [32], crystalline swelling and osmotic swelling were used to describe clay swelling mechanism. Crystalline swelling is a stepwise process with the formation of multiple integer layers of hydrates, which could occur in all types of clay minerals. Osmotic swelling is driven by cation concentration differences between interlayer space and surrounding water, causing water diffusion into interlayer space to restore balance. Osmotic swelling is limited to clay minerals with exchangeable interlayer cations such as smectite. So the water-weakening effect in shales could be contributed primarily to the crystalline swelling. For example of chlorite schist, its unit cell consists of two layers, namely, brucite layer and mica-like layer. Due to the weak hydrogen bond connection between the two layers, water molecules enter the unit cell and expand the interlayer distance. In this study, the samples used for uniaxial and triaxial test were made into cylinder with a diameter of 50 mm and a height of 100 mm. To minimize the anisotropy effect, the shale sample was made as standard cylinder sample with its axis parallel to the vertical bedding planes. ree kinds of samples were used: dried sample, saturated sample, and water immersed sample. Dried samples were prepared by drying process in the oven for 24 hours. Parts of the dried samples were chosen for preparing the saturated and immersed samples. e saturated samples were obtained by immersing the dried samples in a vacuum space and then filled with water. In order to study the behavior of shale affected by water immersion, shale samples were put in water for various times and immersed samples were prepared. Figure 3 shows the water absorption response of shale. e water content is defined as the ratio of water weight to solid weight. e water content increases rapidly with immersion time in the first few days, then slows down, and tends to keep constant. After 90 days of water immersion, the samples could be viewed as saturated samples. e data could be fitted well in a power function. e natural water content of shale samples is about 0.1%. e saturated water absorption of samples is about 1.1%. e result shows that shale has a strong water absorption ability due to mineral components of chlorite and muscovite. Methods. e experimental study consists of macroscale mechanical tests and SEM analysis. To obtain mechanical properties of shale and provide basic information for numerical modelling, macroscale mechanical tests (uniaxial test, Brazilian disk test, and triaxial test) were performed on shales with various water contents. e three types of mechanical tests could give information under different conditions: (1) Uniaxial test applies a vertical load on the sample, which gives the rock a maximum principal compressive stress. Uniaxial strength and stiffness of the material are easily obtained. e uniaxial test was conducted on immersed samples. e failure characters of the samples with various immersion times could be observed from macroscopic and microscopic perspectives. (2) Brazilian disk test gives a vertical load on a lying cylinder sample. e rock is split due to tensile failure. e tensile strength was calculated for the dried and saturated samples. e failure characters were compared. (3) Triaxial test applies a confining pressure followed by a deviatoric pressure. It could simulate the engineering stress state in an approximate way. Before tunnel is excavated, the rock is suppressed with a confining pressure. After excavation, a deviatoric stress is applied and the rock is destroyed. In the triaxial test, strength of dried and saturated samples under various confining pressure was obtained and compared. All the tests were done in controlled humidity room equipped with air condition. In each situation, 3 samples were tested and the most regular results were choice for analysis. Advances in Materials Science and Engineering 3 Academy of Sciences. e loading rate for the uniaxial test is 0.002 mm/s. Figure 4 shows the stress-strain curve of shale with different immersion times. In the initial stage of loading, shale is compressed with pore collapse. e slope of the curve is relatively low, but becomes steeper. With the continuous loading, the slope of the curve increases gradually followed by an elastoplastic deformation stage. When the shale reaches the peak strength, the stress decreases with a sudden drop. It means brittle failure occurs when the shale reaches its peak strength. e peak strength is obtained as the maximum of the axial stress, and elastic modulus is the slope of the most linear part of the stress-strain curve. Figure 5 shows the evolution of peak strength and elastic modulus in uniaxial compression tests. At the beginning of the water immersion, the peak strength and modulus of elasticity drop rapidly. With increasing immersion time, the decreasing rate of the strength and stiffness gets smaller. For the sample without water immersion, the uniaxial strength and elastic modulus are 56.85 MPa and 27.38 GPa, respectively. In contrast, the uniaxial strength and elastic modulus under the water immersion time of 90 days are 21.54 MPa and 11.40 GPa, nearly dropped by more than half. Softening coefficient is defined as the ratio of strength of water immersed rock to dried rock [37]: where K is the softening coefficient, σ w is the peak strength of water immersed samples, and σ c is the peak strength of dried samples. e softening coefficient reflects the water effect on mechanical properties. As seen from Figure 5, the softening coefficient is 0.38 at the immersion time of 90 days. Except the data scatter, in general, the shale gets smaller strength and larger deformation with a longer time of water immersion. Tensile Behaviour in Brazilian Disk Test. Brazilian disk test was conducted on RMT-150°C test machine. e compressive load in the axial direction results in tensile stress in the middle of the sample. e tensile strength is smaller than the compressive strength in geomaterials, so the sample fails by tensile stress in Brazilian disk test. e loading rate for the uniaxial test is 0.05 kN/s. e axial forceaxial deformation curve is shown in Figure 6. e tensile strength may be calculated by the following formula based on elastic theory: where σ t is tensile strength in MPa, P is the maximum axial force in N, D is the diameter of the sample in mm, and h is the height of the sample in mm. Compared with dried samples, there is a remarkable compressive section at the beginning of loading for saturated samples, which is caused by pore collapse. After the pore collapse, there is a linear elastic section followed by a nonlinear elastoplastic section. e elastic modulus increases with increasing confining pressure. e water weakens the elastic modulus in saturated samples compared with the dried samples. e average of the maximum principal stress at failure is shown in Figure 8. As seen, the triaxial compressive strength is affected by both water and confining pressure. Under a confining pressure of 10 MPa, the average strength of the dried sample and saturated sample is 118.4 MPa and 89.2 MPa, respectively. e softening coefficient is 0.75. However, under a confining pressure of 20 MPa, the strength of the dried and saturated samples is 260.6 MPa and 136.9 MPa, respectively. e softening coefficient is 0.52. In a whole, the peak strength of dried samples is larger than that of saturated samples under the same confining pressure. On the other hand, for the sample of the same water content, the peak strength increases with the confining pressure. It is also worth noting that based on the test data, the ratio of strength in dried samples to the saturated samples becomes larger as the confining pressure increases. e gap between the dried strength and saturated strength is 29.1 MPa under a confining pressure of 10 MPa, while the gap is 123.7 MPa under a confining pressure of 20 MPa. A Numerical Case Study of Water Effect. e mechanical properties of shale are sensitive to water immersion. As seen from indoor test, a tunnel excavation could be numerically simulated based on the indoor test results. e parameters used in numerical modelling are determined through the following steps: (1) e uniaxial strength and elastic parameters under various water immersion times are obtained. From equation (1), softening coefficient could be solved. where φ and c are the Mohr-Coulomb yield strength parameters, representing the internal friction angle and cohesion, respectively. e numerical study is conducted with the aid of FLAC3D software. e ideal elastoplastic model with Mohr-Coulomb criterion and tension cutoff is used. Unlike nonlinear models [38,39], it is simple for calculations. e parameters are achieved from the above works and listed in Table 1), and in the triaxial test, the friction angle and cohesion are 42.5 MPa and 7.17 MPa, respectively, for the saturated samples, which could be deduced from Figure 7. As seen in Table 2, the gaps in the data between the calculation and test are acceptable. Although the failure mechanism in these tests is different, the calculation method offers an alternative and Advances in Materials Science and Engineering reasonable way to acquire parameters under various water contents. To simulate the water effect on tunnel excavation, the surrounding rock is first adopted as a homogeneous material. e diameter of the tunnel is 6.5 m. e block state and contour of z-displacement in meters are recalculated after gravity balance. e results are shown in Figures 9(a)-9(g). For the surrounding rock with a small value of moisture content, the plastic elements occur in the location between the directions of intermediate principal stress and minimum principal stress. With increasing water immersion time, the surrounding surface of the tunnel is gradually surrounded by plastic zones. In the contour of z-displacement, there is a horizontal line in the middle of the tunnel representing zero z-displacement. e vault crown and tunnel floor both expand to the free surface, and the absolute values of z-displacements are equal approximately. Another case of water influence is simulated, supposing the upper section of the surrounding rock is fully immersed by water due to rainfall compared with the lower section. e result is shown in Figure 9(h). e rainfall changes the mechanical properties of strata, resulting in the composite ground. It is like an up-soft/low-hard rock stratum with inhomogeneous characters [40,41]. Different from the homogeneous surrounding rock, the plastic zones are mainly generated in the upper section, and the settlement of the vault crown is larger than z-displacement of the tunnel floor. e z-displacement of the upper surrounding rocks is no longer symmetrical with the surrounding rocks of lower section. From a quantitative perspective, the volumes of plastic zones are calculated and the settlement of vault crown is determined. Figure 10 shows the volumes and settlements for comparison. e ratio in the figure means the ratio of plastic zone volumes or settlement under water immersed condition to dried condition. For the surrounding rocks with more moisture content, the settlement and plastic zones increase. e ratio of plastic zone volume and settlement under 90 days of water immersion is 8.35 and 2.4, respectively. e great water effect on plastic zones and deformation of the surrounding rock would cause serious problems in support. It is worth noting that the role of numerical modelling is to study the rules of physical phenomena rather than accurate simulation. e aim of numerical investigation is to obtain acquaintance with the influence of water on tunnel excavation. Although the parameters of surrounding rocks are used from indoor test, the quantitative analysis of water effect could be obtained for consideration in surrounding rock support. Discussions Although several water-weakening mechanisms have been proposed, it is hard to find a universally acceptable explanation [42]. It has been recognized that the weakening of macroscopic mechanical properties of rocks is closely related to the changes in microstructures of rocks. To investigate the mechanism in the weakening effect of water on shale, SEM technique was conducted by Quanta 250 in the Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. SEM slices were taken from the immersed samples without the mechanical tests. Figure 11 shows the scanned photos before mechanical test under various immersion times. e mechanism of water effect could be analyzed as follows: further destruction of cementation between minerals, the pores and cracks increase. (4) After 50 days of water immersion, the shale microstructure is further destroyed. e cementation between minerals becomes weaker. Parts of clay particles form a flocculent structure. e formation of the flocculent structure further destroys the original structure and cementation. Geomaterials generally consist of particles and pores. When clay particles swell in water, pore volume increases as well [43]. ere is a charge in the water, which may be called solution [44]. e interface between a charged surface and an aqueous electrolyte solution is known as the electric double layer [45]. e negative charge forms one half of the electric double layer, and a net positive charge in the solution forms the other half of the electric double layer. e thickness of the electric double layer is formulated as follows [46]: where κ is a measure of the reciprocal thickness of the diffuse double layer, z is the absolute value of the valence of an ion, ϵ is the charge on a single univalent cation, n 0i is the number of ions of type X i in a unit volume of solution measured at a point remote from the double layer, D is dielectric constant, D 0 is diabattivity of free space, k is Boltzmann' s constant, and T is absolute temperature. e concept of the electric double layer is introduced and used to illustrate the effect of water on shale from a microscopic perspective. As the immersion time increases, water enters the volume of shale and n 0i decreases. From equation (4) κ would decrease, and the thickness of the double layer would increase. In such case, the distance between particles increases. e cohesive strength would decrease and would induce cracks. Corresponding to the microscale results, the macroscale failure morphology is shown in Figure 12. Under water effect, the cohesion of shale becomes weaker. e strength gets smaller with the immersion time. Meanwhile, the cracks and pores generate. In such case, the shale is easy to get larger deformation and decreasing stiffness. In the uniaxial compressive state, the shale specimens fail mainly along the bedding surface or at a small angle with the bedding surface. It could be obviously observed that with the increase of water immersion time, more failure planes penetrate. It is the evidence of the waterweakening effect from a macroscopic viewpoint. Meanwhile, in the triaxial test, the saturated shale fails mainly along an inclined plane. ere are not so many Advances in Materials Science and Engineering fractures as generated in the uniaxial test. It could be inferred that the confining pressure works. As the confining pressure acts at the surface of the sample, lateral deformation is restrained. e shale could bear more loading stress. erefore, timely support and confining pressure supply is an effective and essential approach for construction [47,48]. is study is developed based on the research by Zheng et al. [34], which was conducted at the geological survey stage before construction. e studies both focus on the shale in Huangjiazhai Tunnel in China, for the serious condition and important problem during excavation. In Zheng et al.'s study, the main contents are two respects: uniaxial test and microscale study. e microscale study was through SEM observation and description, while lacking theoretical analysis. Considering the insufficiency, in our study, kinds of mechanical tests were investigated, including uniaxial test, triaxial test, and Brazilian test. rough these tests, the water-weakening effect was fully studied. On microscale, SEM results were analyzed with the theory of the electric double layer. e failure mechanism of rock block in water was elaborated on both macroscopic and microscopic scales. A numerical case study of tunnel excavation under water condition was simulated. erefore, more technical investigations and analysis (including test, theoretical analysis, and numerical simulation) were integrated in our paper. e comprehensive results in our paper offered suggestions and benefited the construction in shale strata. Conclusions (1) Under uniaxial compressive stress, water has an obvious influence on the peak strength and elastic modulus with a rapid drop at the beginning of the water immersion. With immersion time increases, the reduction of the strength and stiffness gets moderate. e average uniaxial compressive strength and modulus of saturated shale could be reduced by 62% and 58%. (2) With the Brazilian disk test, the tensile strength of the saturated samples is lower than that of the dried samples. e tensile strength of the saturated sample decreases by 36% under the effect of water. (3) In the triaxial test, there is a remarkable pore collapse at the beginning of loading for saturated samples. e mechanical behavior is affected by both water and confining pressure. e peak strength increases with less water content and increasing confining pressure. e ratio of strength in dried samples to the saturated samples becomes larger as the confining pressure increases. e softening coefficient is 0.75 and 0.52 under confining pressures of 10 MPa and 20 MPa approximately. (4) Based on macroscale experiment, the parameters used in numerical modelling were determined. In the evaluation of water effect on tunnel excavation with homogeneous surrounding rock, plastic zones and z-displacement increase with increasing water immersion time. e vault crown and tunnel floor both expand to the free surface, and the absolute values of z-displacements are equal approximately. e simulation of rainfall impact shows that plastic zones are mainly generated in the upper section, and the settlement of vault crown is larger than z-displacement of the tunnel floor. (5) With the increase of immersion time, more mineral particles are dispersed and peeled off. e flake particles are dispersed into smaller slices, and the connection between mineral particles gradually changes into point-to-surface contact. e microstructure becomes looser, and pores and cracks generate. e electric double layer was introduced and used for illustrating the effect of water on shale. e thickness of the double layer increases with the increase of water immersion time. From macroscale morphology, the number of penetrating planes increases, while confining pressure could inhibit the generation of cracks. It is essential to perform timely support and supply confining pressure in construction. Data Availability e data used to support the findings of this study are available from the corresponding author upon request.	2020-05-21T00:02:58.237Z	2020-05-11T00:00:00.000	{ "year": 2020, "sha1": "e5266181716966ec3ddfb86330785b3423ba90dc", "oa_license": "CCBY", "oa_url": "https://downloads.hindawi.com/journals/amse/2020/5263431.pdf", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "c7ae15ae8aa12d0bb7044f1dc8f0f1f0f633f518", "s2fieldsofstudy": [ "Engineering", "Environmental Science", "Geology" ], "extfieldsofstudy": [ "Materials Science" ] }
238059818	pes2o/s2orc	v3-fos-license	Pulse-width modulation in the autonomous voltage inverter in identifying parameters of the three-phase asynchronous motor . This paper deals with control algorithm of a two-level autonomous voltage inverter with pulse-width modulation in forming test phase voltages for identification of parameters of a three-phase asynchronous motor. Necessary conversion is carried out, the laws of change of phase voltages of an autonomous inverter are obtained. The feasibility of using third order active Butterworth filters for determining first harmonics of phase voltages is shown. By means of computer simulation and experimental research, it is revealed that the required shapes of phase voltages can be implemented in a two-level autonomous voltage inverter with pulse-width modulation, and the presence of Butterworth filters does not introduce significant distortion and allows to obtain necessary phase voltage shapes in the generalized electric machine. Introduction For implementation of laws of frequency regulation of speed in the asynchronous motor (AC), it is necessary to be armed with data on current parameter values of phase-substitution and load circuit. This information allows to implement more accurate algorithms of frequency regulation of speed. And this provides energy efficiency and resource saving. There are various methods of parameters identification [1 -5]. In [6][7][8][9][10][11][12][13] efficiency of using continuous gradient method of minimum search of positive definite function for parametric identification of a three-phase asynchronous motor (AC) and its load is shown. In [7,8,11] total number of identified parameters is seven. In [7 -10], a computer simulation of the identification processes of the parameters of a threephase asynchronous motor (AC) and its load was carried out. The basis of the asynchronous motor model was the mathematical description of the generalized electrical machine (GEM). When simulating, the phase voltages of the stator winding were set in the form of test signals. The specifics of the operation of the frequency converter and the influence of pulse-width modulation (PWM) were not taken into account. PWM algorithms are given in [14]. In this case, when identifying the parameters of the asynchronous motor, the shape and frequency of the output voltages of the frequency converter (FC) will depend on the laws of change in the test signals The purpose of this article is to analyze the feasibility of technical implementation of the output voltages of the frequency converter with test signals and filtering the high-frequency components of these voltages. Methods In most cases, algorithms of frequency regulation of speed are based on mathematical description and parameters of the generalized electric machine [14,15,16]. Thus, for identification of parameters we use the equation system of the generalized electric machine (GEM) on the basis of three-phase asynchronous motor (AC) [7,8,11,17,18]. Transition from parameters and values of a threephase asynchronous motor (AC) to the ones of the generalized electric machine (GEM) and vica versa is discussed in [17,18]. Conversion of phase voltages of a GEM stator in voltages of a stator of the three-phase asynchronous motor (AC) takes the form [17, 18] where 1  is shaft rotation angle of axis d of coordinate system d, q about axis A of stator. Computer simulation of parameters identification processes of the asynchronous electric drive with motor AIR80A6U2 was carried out in [8]. Voltages ) ( 1 t u d , ) ( 1 t u q and angle velocity ) were changing in accordance with timing diagrams in Figure 1 and were characterized by the parameters presented in Table 1. Results and Discussion Results of computer simulation in the form of timing diagrams of angular velocity ) ( 1 t  , rotation angle 1  and voltages ) ( 1 t u q , ) ( 1 t u А are given in Figure 2. shows that phase voltages are amplitude-frequency modulated oscillations. Amplitude changes according to variations of voltage amplitude ) ( 1 t u q , and frequency changes according to angular velocity variations ) In a larger timescale, for the first and the second halves of period 1 Т , the curve ) Figure 3 a, d. During the first half of period 1 Т , voltage ) ( 1 t u А varies with time according to harmonic law with amplitude 75 V and frequency 71.6 Hz, and during the second half of period 1 Т with amplitude 25 V and frequency 23.8 Hz. In a similar way, with angle shift π/3 2 − and π/3 2 , voltages ) will change correspondingly. To implement phase voltages ) it is necessary to change simultaneously their rootmean-square values and frequency. In most modern asynchronous electric drives, indirect frequency converters (FC) are used. They comprise uncontrolled rectifier (UR) and two-level autonomous voltage inverter (AVI). To operate transistors of the autonomous voltage inverter, pulse-width modulation (PWM) is applied. Phase voltages of AVI with PWM represent pulse string, amplitude and duration of which change directly with control voltages. Simulation of AVI with PWM was conducted providing three-phase mains supply of indirect frequency converter (FC) has voltage 390 V, frequency 50 Hz. In this case rectified voltage in DC link [21] takes the form of In simulation, voltage drop in diodes of uncontrolled rectifier and in on-transisters of AVI was taken into consideration. Control voltages of AVI were taken as proportional to voltages ) Results of computer simulation are presented in In pulse-width modulation (PWM), output voltages of frequency converter (FC) represent pulse string. Duration of the pulses changes proportionally to control voltages. Impact of PWM leads to occurrence of highfrequency components in phase voltage spectrum of FC. Thus, for identification of parameters of three-phase asynchronous motor it is necessary to perform filtration Evidently, that in phase voltage spectrum of FC there are high-frequency components. For high-frequency components cutoff, low-pass filters (LPF) are used. Taking into considerations [24 -28], analysis of circuitry and LPF characteristics was conducted. It was found that for filtration of phase voltages of (FC), it is reasonable to use third order Butterworth filters. Transfer function of such filter is given [28] where f k , f T are transmission factor and filter time Table 1; 2) circular frequency of reference voltage ref  is determined by PWM frequency which was set equal to 16 kHz; 3) effect of LPF on voltage conversion should be minimal. In Figure 5 frequency response of third order Butterworth filter with cutoff frequency 5 kHz is shown. The response is determined on account of transfer function proved that f k equals 1. Analysis of amplitude-frequency characteristics (AFC) in Figure 5 shows that with circular frequency 1  lower than circular cutoff frequency с  , it is at the level of 1, while with circular frequency 1  higher than с  , it has the slope of 60 dB/dec. This is decisive advantage when choosing LPF type. Evidently, that amplitude distortion with circular frequencies of the desired signal min 1  and max 1  will be minimal. Analysis of circuit phase responce in Figure 5 shows, that in the field of circular frequency of the desired signal min 1  and max 1  , the filter will not cause significant lag. Experimental setup scheme is given in Figure 6. The sensor unit in Figure 6 comprises voltage sensing devices and current sensing devices, Hall effect devices, rotation velocity sensors and angular accelerometers. Hall devices are installed in stator bore in orthogonal axes. Filter bank comprises six third order Butterworth filters. Frequency converter (FC) is not shown in Figure 6. The aim of the experiment is analysis of output voltage curves of third order Butterworth filters and analysis of results of conversion of output voltage of the filters into voltages of coordinate system d, q. Results of the experiment are presented in Figure 4 c, d, g, h in the form of timing diagrams of phase voltages of frequency converter (FC) and output voltages of LPF for the first and second halves of period 1 Т . In Figure 7b there are timing diagrams of voltages ) Curves of output voltages of LPF in Figure 4 d, h are not ideally smooth harmonic functions. In spectrum of LPF output voltages there is high-frequency ripple. Most significantly they are revealed in the second half of period 1 Т (Figure 4 h). This is explained by relatively small amplitude value of the first harmonic of phase voltage ) ( which is 25V. On account of (5), (9), (10), curves ) are formed, they are given in Figure 7 b. Becaue of the PWM effect, the curves contain high-frequency ripple with amplitude no more than 3 V. For qualitative evaluation of effect of third order Butterworth filters, analysis of frequency response was conducted, it is presented in Figure 5. Results are shown in Table 2. Conclusion The conducted analysis allowed to find out that third order Butterworth filters do not significantly influence the desired signal and provide efficient cutoff of highfrequency interference. Experimental research allowed to establish, that the use of indirect frequency converter (FC) with PWM and Butterworth filters makes it possible to a high precision to perform transition in phase voltages of GEM stator winding. When this happens, identification processes stability is reserved, and in identification of the aforementioned parameters of a three-phase asynchronous motor and its load, with initial deflection from true values by 50%, relative error is not higher than 7% [20]	2021-08-27T16:41:20.158Z	2021-01-01T00:00:00.000	{ "year": 2021, "sha1": "4723a04c7e41b722bbe318a1ed2d3101b76e9336", "oa_license": "CCBY", "oa_url": "https://www.e3s-conferences.org/articles/e3sconf/pdf/2021/64/e3sconf_suse2021_01054.pdf", "oa_status": "GOLD", "pdf_src": "Adhoc", "pdf_hash": "21ea3c7a3450fbe34dde55a178855e1e27aa8597", "s2fieldsofstudy": [ "Engineering" ], "extfieldsofstudy": [ "Computer Science" ] }
111105650	pes2o/s2orc	v3-fos-license	Essential Competences to the Pharmacy Industry Competitivity: Case Study With Cooperators in the Town of Natal, Rio Grande do Norte, Brazil The pharmaceutical industry is one of the most promising markets in the world. It is highly competitive, complex and requires huge investments in research, development, production, marketing, and sale of pharmaceutical products. One of the main strategies of the pharmaceutical industry is the promotion of its products for the medical fraternity through a drug sales representative, essential for the acceptance and prescription of your brand. This study identifies the essential drug promoter skills to competitiveness and the success of these organizations. Data were collected through semi-structured individual interviews and resulting 46 competences mentioned as the most important by own sales representatives. The research points to the need for new methods of empowerment, essential to the competitiveness of sales representatives, considering that these competences contribute to the strategies of establish industry. population purchasing power, increasing the job opportunities in this category. Supporters of pharmaceutical promotions claim that marketing expenditures give innovative pharmaceutical manufacturers a fair chance to recover high R&D expenditures (Kremer, Bijmolt, Leeflang, & Wieringa, 2008). The pharmaceutical market is a specialized field that requires knowledge in drugs and sales skills by the sellers that sell the product to a customer (doctor) who has more scientific knowledge. It is known that the importance of competency management in this scenario is to align human resources managements with business strategy incorporating the concept of competence. In this market, consumers do not always choose the product because it is the doctor who prescribes the medication and decides to be more appropriate to the patient. To Rubin (2003), marketing may serve as a communication channel to educate physicians and expose consumers to information that may improve their health outcomes and medical options. This is the reason to take promotional and commercial actions with the physicians and competent distributors (drugstores and wholesalers) in order to reach their planned objectives. Such spending benefits both patients and physicians in a variety of ways. The pharmaceutical representatives are valued highly by physicians as a reliable source of information. Kobayashi and Leite (2010) asserted that these professionals must be increasingly qualified and updated, so the investments could revert to keep the competitiveness of the organization. The question refers to the importance of core competencies of the pharmaceutical representatives as competitiveness in the city of Natal, Rio Grande do Norte, Brazil, outlining the following specific objectives: to identify individual skills, classification of competencies (knowledge, skills, and attitudes), and characterization of the pharmaceutical representatives activities. Competences and Strategy Organizational competencies stem from the specificity of each company to identify its resources and turn them into competitive advantage. It is noteworthy that the process to transform resources into organizational competencies happens in the competitive scenario and moves the dynamic of the organizational. This way, the strategy emerges linked to the company's competences, while its separation into personal competences goes through hierarchic levels and work teams. The effectiveness of the organization results in the ability to articulate their strategic guidance and expertise of the company, group, and individual skills. Therefore, establishing the strategic direction, aligning the organizational competences, and separating into individual competences reflect the organizational culture and the management model (Moura & Bitencourt, 2006). Mintzberg, Ahlstrand, Lampel, Quinn, and Ghoshal (2008) pointed out that both organizations and strategies exist independently. However, one influences the other, since strategies are constructed from a previously organizational framework, and evaluated by the strategic management at some moment. This can be conditioned, either by need to change as well as for the survival of the organization. Therefore, the strong and weak points and the opportunities and menaces must be considered as a starting point to build a new management strategy. The strategic management plays an important role in the characterization of the diverse and complex organizational processes, contributing to creating a kind of management focused on the efficiency and the search for better results (J. V. Queiroz, F. C. B. P. Queiroz, & Hékis, 2011). The word "competence" associated with human resource management originally appeared in a paper published by David C. McClelland in 1973 Testing for Competence Rather Than Intelligence (Daólio, 2004;Dutra, 2004). The concept of competence is though as a set of knowledge, abilities, and attitudes justifying a high professional performance (A. Fleury & M. T. Fleury, 2004). To Le Boterf (2003), by embracing knowledge, skills, and attitudes, competence is something that requires a continuous motion, a constant mobilization of diverse knowledge for application to a specific situation. Hence, it can be said that to be competent, the professional needs knowledge, abilities, and attitudes which are specific to practice his/her activities. In other words, the professional must have attitudes and abilities enabling him/her to mobilize his/her knowledge, at a precise time, to reach a particular goal, in order to solve a specific problem. For this study, we consider competence as a set of knowledge, abilities, and attitudes providing the professional and a better performance of his/her functional attributes, with efficiency in developing the activities and success. Organizations situated in competitive environments need to define its strategy and the necessary competences in a continuous learning. Individual Competences To Zarifian (2001), individual competences are those regarding to each professionals of the organization. These are skills that the individual possesses by his/her educational background, family creation, and his/her biography. Business competences refer to the comprehension that a person must have about the "business" of the organization and the organizational goals, as well as a comprehension of the social and political environment surrounding the organization. In this sense, a good professional must incorporate competences, or improve those already present, according to his/her experience in the environment. A switchover between organization and persons is perceivable, both sides have effective participation in the process of professional growth. A professional unable to move his/her knowledge, information, and abilities to solve an immediately solvable conflict, becomes unproductive for reaching the organizational goals (Le Boterf, 2003). Therefore, competence refers to the professional capacity to associate all his/her personal qualities in order to get a better performance in his/her acting area. Prahalad and Hamel (2004) developed the concept of essential competences in organizations. Such competences are a set of abilities and technologies resulting from contributing a fundamental difference for the competitiveness of the company. The identification and skills development are imperative for the organization to define its action strategies in a complex and ever-changing world. The companies are able to detect just a limited number of basic competences giving them strategic and sustainable advantage in the market. Competence management gains importance and is perceived as a management model derived from the resource-based theory of management. Bitencourt (2005) pointed out that competence can be summarized and associated with the development of concepts, abilities, and attitudes. Besides, they constantly rise questioning and trigger a process of individual learning. The competences are transferred and consolidated by the relationship (interaction). Identifying Competence A. Fleury and M. T. Fleury (2004) highlighted the importance of competence-based management. Two models: the American Society for Training and Development (ASTD) and competency-based management were the first ones to stand out identifying the competences. The research in skills management involves different approaches, including the views of managers, employees, and customers. The methods and techniques adopted to identify competences have the following features: observation of high performance employees behavior in order to delineate their qualities; application of semi-structured focused interviews; brainstorming; brain-writing; surveys; and modeling by specialists. Medical Advertisement Strategy Medical advertisement is defined by the World Health Organization (WHO) as all informational and persuasive activities by manufacturers and distributors to influence the prescription, supply, sale, or use of drugs (Norris, Herxheimer, Lexchin, & Mansfield, 2007). The industries concentrate its investments and efforts to develop drugs to treat chronic diseases that require ongoing prescriptions for longer periods (Rey, 2010). On the other hand, Palácios (2008) brought out that the use of sophisticated marketing and advertisement techniques, and the relationships established between physicians and pharmaceutical representatives are potentially harmful to the users and therefore it makes evident the importance of regulating medical advertising. One of the main links between the pharmaceutical industry and the physicians is its sales force, the pharmaceutical representatives, or medicine advertisers, who are employed and earn their salaries working to attend the interests of their employers (Brody, 2005). In this sense, some difficulties rise from different beliefs and interest values existing among doctors, patients, and the third parties, including the media and the pharmaceutical industry (Neelam & Jo-Ann, 2012). The Brazilian resolution establishes a difference: For over-the-counter (OTC) medicines, advertisements can be addressed to the consumer, but for prescription drugs (either with or without retention of the prescription), the advertisement must be addressed just to professionals entitled to prescribe them, though communication channels are restricted to them. There are two different segments for pharmaceutical marketing: the market of over-the-counter medicines and the markets for strictly prescription medicines, acquisition is done through medical prescription. There is one resolution regulating the activities of pharmaceutical representatives, determining that these professionals "must be limited to scientific information and characteristics of the medicines registered by the National Health Surveillance Agency". The ethical matter about physicians and pharmaceutical representatives relationship has been widely discussed in the literature, evidencing a debate about conflicts between medicine users (patients) and the pharmaceutical industry, which through the pharmaceutical marketing plans a series of promotion and advertisement techniques directed to the doctors. The use of freebies, free samples, invitations to events, lunches and gifts are the central point of this debate (Fugh & Ahari, 2007;Gagnon & Lexchin, 2008;Palácios, 2008;Wazana, 2000). Methodology The study was performed based on qualitative search (semi-structured interview) with the intention to release the expression of ideas, thoughts, and opinions from the own pharmaceutical representatives. From the point of the objectives, this research is characterized as descriptive, as it intends to describe the features of a given phenomenon and to establish the relationship among variables. Sampling is non-probabilistic and the selection is made by typicality (Vergara, 2000), developing medical advertisement activities in the Natal market. This work adopted a model praised by Dutra (2004) and A. Fleury and M. T. Fleury (2004): Competences can be surveyed based on their components of knowledge, abilities, and attitudes. The universe of the research was the group of pharmaceutical representatives, city of Natal, Rio Grande do Norte, Brazil. Information about functional competences was collected from 35 pharmaceutical representatives from a whole universe of 250, representing 18.8% of the professionals at state of Rio Grande do Norte, Brazil. These professionals develop their work though health professionals visits, mostly physicians of different specialties. They promote and advertise pharmaceutical products manufactured by several pharmaceutical companies. Interviews As a first procedure for collecting data on individual competences, a semi-structured individual interview was applied to 35 pharmaceutical representatives. For the semi-structured interview, a script was used containing: a heading indicating the date of the educational institution; identification of the post-graduate program; the name of the searcher and the title of the work related to the research; definitions of organizational and individual competence; and the question to be answered: "based on the typical activities of the pharmaceutical representatives and the competences that you consider important for the pharmaceutical representatives". The 35 pharmaceutical representatives were interviewed individually, in different moments and places according to their convenience. Firstly, a brief explanation was made about the reasons: What was the research about and what was it for? It was emphasized that the data provided were strictly intended for academic research. Results and Discussions Competences Identified in the Interviews Forty six competences were identified in the interviews in the requirements: knowledge, abilities, and attitudes according to Table 1. Among knowledge requirements, four competences were considered relevant, cited frequency higher than 60% by the interviewed professionals. There are the knowledge requirements standing out as the most relevant: • Sales techniques; • Planning techniques; • Advertisement techniques; • Negotiation techniques. These competences are related to the application of the daily activities knowledge of the pharmaceutical representatives, such as supplying information in an efficient and effective way to the highest number of health workers. For this, they prepare activities and promote products through advertising techniques, promotional material, freebies, and gifts, in order to gain the medical prescription, a fundamental point to increase the sales and consequently reach the planned goals of the company. The pharmaceutical companies frequently work out these competences in their trainings, providing this technical knowledge to their professionals in order to transform these individual competences into useful company competences. Other two competences identified were foreign language and product composition, which interviewed professionals believe to beessential knowledge in order to inform, negotiate, and convince the doctors to prescribe. Among the requirements of abilities, the competences stand out with more than 60%: • Good human relationship; • Communication skills. The results point toward a concern of the professionals about how to communicate and relate with the clients. The importance attached by pharmaceutical representatives to personal relationship skills and communication skills related to their daily activities are evident. Based on these data, knowing how to deal with customers is very important for the performance; as well as how to communicate in a clear concise and transparent way. Among these requirements, it is worth considering the importance perceived to the following competences: work under pressure taken as relevant in spite of receiving lower scores (less than 60%). The ability to work under pressure is one of the new realities of the modern world, and it is very evident in the pharmaceutical market. The strong competition and high investments on research and development, beyond establish goals to be attained by their collaborators, as a way to get higher profitability. None attitude requirement reached 60% in its score, but empathy, professional ethics and initiative got 54% and integrity reached 51%, hence being the most cited competences, evidencing the preoccupation about professional behavior and the doctors, colleagues, and patients relationship. The ethics issue got much attention as a result of discussions, raised by researchers and experts, about the limits and acceptable conduct of pharmaceutical representatives trying to convince the physicians about a medical prescription. Another competence to stress is empathy, which is also considered as a quite relevant competence, as professionals relate empathy with a better performance of their functions. Final Considerations The present work detected the essential competences to pharmaceutical representatives acting in state of the Rio Grande do Norte, Brazil, by own perspective. As a whole, 46 competences were identified. These competences were classified into knowledge, abilities, and attitudes. Knowledge on sales techniques, negotiation techniques, and planning techniques is competences related with knowing how to do daily activities of medical advertisers to plan their work, to know how to negotiate, and be able to identify the needs of their clients and demonstrate the advantages and benefits of the promoted products. As evidenced by A. Fleury and M. T. Fleury (2004), knowledge competences are those related with a potential to do, and competences identified among the requirements of attitude are those showing a capability to do and denote personal habits. Pharmaceutical representatives are professionals that transfer information about diseases and medicines, and in this case it is fundamental to have the ability to create and keep good relationships, and communicate in a clear and objective way. These relationships between physicians and pharmaceutical representatives are built through respect and ethical behavior, based on empathy and integrity (Palácios, 2008). It is necessary for the ability to create and keep good relationships, abilities to communicate in a clear and objective way, and work under pressure, in response to the high competitiveness of this market. Some attitudes are essential: empathy, professional ethics, and integrity, which are competences related to conduct and behavior, in the daily activities visiting and promoting pharmaceutical products to health professionals. It was observed that these professionals, in daily activities, carry out complex activities demanding these competences in order to reach objectives and the goals imposed by the pharmaceutical industry. This work highlights the importance of knowledge, abilities, and attitudes for the pharmaceutical representatives in order to attain results for the pharmaceutical industry. In this sense, it is necessary to increase the investments to develop new manuals for teaching and training pharmaceutical representatives in order to turn them more capable and aligned to new realities of this market. This will make pharmaceutical representatives force more conscious about the necessary essential competences. The survey still pointed out the competences of sales techniques, good human relationships, to know how to communicate or communication skills, and work under pressure as relevant. The competence considered as the most relevant by survey was sales techniques. This result indicates that knowledge of sales techniques is essential for the good performance of this profession, as evidenced in previous studies (Brody, 2005;Medawar, 2002;Wazana, 2000;Maguire, 1999;Norris et al., 2007). This competence is related to sales activities including: knowledge about the client's needs, about benefits and advantages of the product and being capable of keeping commercial relationships with the doctors. Planning and negotiation techniques are competences related to knowing how to do and knowledge. Planning and negotiation are related to the daily work of the pharmaceutical representatives. Before the visits, they need to structure a plan called "pre-visits", daily and weekly schedules and the client profile. The negotiation technique, on the other hand, is necessary to go around questions and to convince doctors to prescribe their brand. These results show that, besides knowing how to do, the pharmaceutical representatives must have the skills to do. In order to be able to transfer knowledge, the professionals need to read the study and retain knowledge in a constant and continuous way, and therefore, a good memory is relevant.	2019-04-13T13:12:00.727Z	2013-01-01T00:00:00.000	{ "year": 2013, "sha1": "5c1814b84a4e9c2fa6ba50988f5a52f02dd866b1", "oa_license": "CCBYNC", "oa_url": "http://www.davidpublisher.com/Public/uploads/Contribute/551cfdfeb7ceb.pdf", "oa_status": "HYBRID", "pdf_src": "Anansi", "pdf_hash": "01afe7dba3b0d1667bea27bfcfd25efb79d871f3", "s2fieldsofstudy": [ "Business" ], "extfieldsofstudy": [ "Engineering" ] }
145321503	pes2o/s2orc	v3-fos-license	Decision-making, human cognition, and equanimity of mind Background: Decision-making is one of the most puzzling issues in modern neuro-cognitive dynamics. It depends on how the brain behaves at that particular instance and identifies and responds to a signal among myriads of noises that are present in the surroundings (called external noise) as well as in the neurons themselves (called internal noise). The ability to predict the outcome of future events is, arguably, the most universal and significant of all global brain functions. The ability to anticipate the outcome of a given action depends on sensory stimuli from the outside world and previously learned experience and/or inherited instincts. So, there is a need to formulate a theory of inference using prior knowledge for decision-making and judgment, as well as, new empirical evidences. Aims: The current paper aims to shed new light on decision-making and judgment with the help of states of mind like neutral mind and equanimity. The paper also aims at highlighting the scientific aspects of these states, which are conducive to proper decision making by an individual. Method: The current paper makes use of methods of mathematical modeling based on a generalized version of probability argument in the Bayesian framework, which includes prior knowledge for decision making and human judgment, as well as, quantum theory in order to model the cognitive domain. Result: The states of mind like neutral mind and equanimity may help an individual to take correct decision with unbiased judgment, and accomplish right cognition. Conclusion: The general Bayesian framework when coupled with quantum theory may help us to understand states of mind like neutral mind and equanimity, in which decision-making happens with unbiased judgment. The current paper also opens up a new dialog between modern science and Indian philosophy as the latter is likely to offer an explanation to the superposed state that has been studied by the scientific community in quantum theory. INTRODUCTION For more than 200 years, mathematicians and philosophers have been using probability theory to describe the ambiguity or uncertainty in decision-making and judgment. Recently, through several experiments with human subjects (Aerts, Aerts & Gabora, 2009;Busemeyer & Trueblood, 2011), violation of traditional probability theory is clearly revealed in several cases. Literature survey clearly suggests that classical probability theory fails to model human cognition in relation to decision-making. The major problem seems to be the presence of epistemic uncertainty and its eff ect on cognition at any time point. Moreover, the stochasticity in the model arises due to the unknown path or trajectory (definite state of mind at each time point) a person is following. A generalized version of probability theory borrowing the idea from quantum paradigm may be a plausible approach. Quantum theory allows a person to be in an indefi nite state (superposition state) of mind at each moment of time. A person may be in an indefi nite state that allows all these states to be potentially (probability ym_7_14R8 Background: Decision-making is one of the most puzzling issues in modern neuro-cognitive dynamics. It depends on how the brain behaves at that particular instance and identifies and responds to a signal among myriads of noises that are present in the surroundings (called external noise) as well as in the neurons themselves (called internal noise). The ability to predict the outcome of future events is, arguably, the most universal and significant of all global brain functions. The ability to anticipate the outcome of a given action depends on sensory stimuli from the outside world and previously learned experience and/or inherited instincts. So, there is a need to formulate a theory of inference using prior knowledge for decision-making and judgment, as well as, new empirical evidences . 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 Roy: Cognition and equanimity of mind 2 Yoga Mimamsa \| Jan-Jun 2014 \| Vol 46 \| Issue 1 and 2 amplitude) expressed at each moment (Heisenberg, 1958). Thus, a superposition state seems to provide a better representation of the confl ict, ambiguity, or uncertainty that a person experiences at each moment (Busemeyer & Trueblood, 2011). Conte et al. (2009) demonstrated that mental states follow quantum mechanics during perception and cognition of ambiguous fi gures. However, the framework of quantum probability considering the superposition of mental states is an abstract framework devoid of material content like the concept of elementary particle, various fundamental constants in nature like the Planck constant, the speed of light, and the gravitational constant in modern physics. So, this framework can be applied to any branch of science dealing with decision-making, such as in biology, social science, etc. The proposition of superposition of mental states needs to be addressed in the context of brain function. Very few a empts have been done so far in this direction. AIMS The current paper aims to shed new light on decision-making and judgment with the help of states of mind like neutral mind and equanimity. The paper also aims at highlighting the scientifi c aspects of these states, which are conducive to proper decision making by an individual. METHODS The current paper makes use of methods of mathematical modeling based on a generalized version of probability argument in the Bayesian framework, which includes prior knowledge for decision making and human judgment, as well as, quantum theory in order to model the cognitive domain. RESULTS & DISCUSSION The superposition of mental states is linked with the neutral mind as well as to the equanimity of mental state as discussed in Indian philosophy (both the Hindu and the Buddhist). However, in Buddhist philosophy, there is a subtle diff erence between neutral mind and equanimity. In the state of neutral mind, there is a scope of decision-making, whereas in the state of equanimity, there is no such scope. It is expected that the states of mind like neutral mind and equanimity may help an individual to take correct decision with unbiased judgment, and accomplish right cognition. In order to see how this is possible, let us fi rst discuss the new empirical evidences and their implications in cognitive science. Empirical evidences in cognitive science Various group of scientists (Busemeyer & Trueblood, 2011) made a empts to show the inadequacy of the classical paradigm of probability to explain the new evidences related to modeling the cognitive domain. This cognition spectrum of human mind is usually classifi ed into six categories as follows: (1) Disjunction eff ect (2) Categorization -decision interaction (3) Perception of ambiguous fi gures (4) Conjunction and disjunction fallacies (5) Overextension of category membership (6) Memory recognition over-distribution eff ect. Fallacies over-distribution eff ect The data collected from various experiments related to these six categories clearly indicate the inadequacy of classical probability theory (Savage, 1972). For example, let us consider the fi rst one, that is, the disjunction eff ect. Tversky and Shafi r (1992) discovered a phenomenon called the disjunction eff ect in the process of testing a rational axiom of decision theory called the sure thing principle that was proposed by Savage (1954Savage ( , 1972. According to the sure thing principle, if under the state of the world X you prefer action A over B, and if under the complementary state of the world XC you also prefer action A over B, then you should prefer action A over B even when you do not know the state of the world. Tversky and Shafi r (1992) experimentally tested this principle by presenting 98 students with a two-stage gamble, that is, a gamble which can be played twice. Classical probability theory says that always P(A) > P(B) where, P(A) indicates probability of the event A occurring, say winning, and P(B) is the probability of the complementary event B occurring, say losing the game. However, the data from this study revealed that in the event of uncertainty of outcome, there was a disjunction eff ect, which was contrary to the sure thing principle. This meant that during uncertainty, instead of defi nitely being in the win or loss state, students entered a superposition state that prevented them from fi nding a reason to play in the second stage of the gamble. Let us look at Tversky and Shafi r's (1992) study in detail. In the fi rst stage of the study, if the students won they would get Rs. 2.0, but if they lost they would lose Rs. 1.0. The key point in this experiment was the decision for the second stage of play after the fi rst stage of play. The three conditions that were set for the second stage of the experiment were: fi rst, the students were informed that they won the fi rst stage of the gamble; second, the students were informed that they lost the gamble; third, the rest of the students were not told the outcome of the fi rst stage. Results revealed that 69% of the students who were told they won the game, 59% of the students who were told they lost, and 36 % of the students who were not told whether they won or lost wanted to play again. The explanations given for the above fi ndings were: in the fi rst group, the students who were told they won had extra money and, therefore, wanted to play again; in the second group, the students who were told they lost were interested to play again in order to recover the losses; in the case of the third group of students who did not know the outcome, very few were interested to play again. It is important to ask why the two reasons given by the fi rst and second groups to play again did not emerge in the minds of the third group as there could not be any other condition other than gain or loss and the reasons to play thereof. In order to answer the question, the probabilities of various outcomes were calculated and it was found that the simple additivity fails for probability law, that is, P(A + B) is not equal to P(A) + P(B). However, they can be explained with the probability rule of quantum framework, that is, P(A + B) = P(A) + P(B) + interference term. This is a clear indication of violation of classical probability theory. The interference eff ect of quantum framework can be related to the decision-making associated with the superposed states of mind. It is similar to the interference term associated with the superposition of wave functions in the double-slit experiment in quantum theory, where if one wants to detect one aspect of micro-particles, say wave aspect, then the other complementary aspect, that is, particle aspect will not be detected. But the total probability of two complementary aspects occurring will not be simply summation of two individual probabilities. At this point there is a need to emphasize that in all the above six categories of the empirical evidences, similar interference eff ects have been detected (Savage, 1972). This clearly shows the inadequacy of classical probability theory in the cognitive domain. Now, let us not discuss any more technical details here, but instead will discuss the implications in the cognitive domain. Equanimity of mind The challenging issue in modern neuroscience is how the neurons in the brain operate, so as to interpret the above kind of intermediate state which gives rise to the superposition of two complementary aspects. For example, it is possible to understand the operation yes or no with the operations performed by neurons. But if there exists interference term due to superposed states of mind, then one can think of an intermediate term corresponding to a state like: a \|HAPPINESS> + b \|UNHAPPINESS>, where a and b are positive constants. Here \|HAPPINESS> and \|UNHAPPINESS> are two vectors representing the happy and unhappy states of mind, respectively. According to the superposition principle, one can think of a state as neither happy nor unhappy. How brain interprets this operation at cellular level (i.e. at the level of neuron) is hard to think of. In quantum theory, such superposed states have been discussed and have raised lot of interest in the scientifi c community. The famous physicist Schrӧdinger, (1983) formulated this problem as "Cat Paradox," that is, one can think of a superposed state of the living state and the dead state of a cat. Only if one observes, it will be either in living or dead state. The observation procedure is considered as a classical process, and hence the superposed state or the wave function collapses to either of the states. Shimony (1993) proposed a "potentiality interpretation," that is, the superposed state function has the potentiality to be in either of the complementary states. This is a state of mind where the mind is capable of decision-making even while remaining in a state superposed of two complementary aspects. This type of issue has been debated in Eastern philosophy, especially in Hindu and Buddhist philosophy, many centuries ago. In the Hindu text, Bhagavad Gita (Sivananda, 2000), the concept of equanimity is described as yoga in the following way: Perform action, O Arjuna, being steadfast in yoga, abandoning a achment and balanced in success or failure. Such equanimity is called yoga. Swami Sivananda (2000) analyzed the situation in detail as: In Buddhist tradition, the term "equanimity" (Upeksha in Sanskrit and Upekkha in Pali) has been defined in various ways -the core of this defi nition rests on "gazing upon" or "observing without interference." In the Theravadan Buddhist School, two main usages of the term "equanimity" are used. The fi rst one refers to "neutral feeling" -a mental experience that is neither happy nor unhappy. This fi rst usage of equanimity corresponds to the Western psychological notion of "neutral valence" and is commonly experienced throughout any ordinary day. In this state, mind is considered to be capable of taking decision and making judgment. The second usage of the term "equanimity" refers to a state of mind that cannot be "swayed by biases and preferences" (Anuruddha, 2000). It is to be noted that Buddhist as well as Hindu philosophies agree on the issue that one needs practice to achieve the state of mind called "equanimity." The future developments of cognitive neuroscience and modeling in the cognitive domain may help to understand decision-making and judgment without biases and/or preferences.	2019-05-06T14:04:03.140Z	2014-01-01T00:00:00.000	{ "year": 2014, "sha1": "2aa44c0e534cc7a40e9e9cdddf16820673ade164", "oa_license": "CCBYNCSA", "oa_url": "https://doi.org/10.4103/0044-0507.140196", "oa_status": "GOLD", "pdf_src": "Anansi", "pdf_hash": "c15022e45f8ee05009fbc35f922c9721a2af9787", "s2fieldsofstudy": [ "Philosophy", "Psychology" ], "extfieldsofstudy": [ "Psychology" ] }
198397061	pes2o/s2orc	v3-fos-license	Development of an aerogel-based thermal coating for the energy retrofit and the prevention of condensation risk in existing buildings The energy retrofit of existing buildings, particularly historic and/or listed buildings, presents several issues; that is, the compatibility between the identified solutions and the heritage value or the internal space reduction (if internal interventions have to be adopted). For this reason, an emerging method to address the target of the energy retrofit of historic buildings is the use of advanced materials characterized by high thermal insulation performance. In the framework of an European research project (Horizon 2020), a novel aerogel-based insulating coating, particularly suitable for the mitigation of thermal bridges and for the prevention of the condensation risk, is under development. In this paper, both the laboratory and the in-fields research activities are described. The first was aimed at optimising the thermo-hygrometric properties of the coating; the latter dealt with the results of a monitoring activity carried out on a full-scale application. Results highlight that the internal application of 12 mm of the developed material can lead to a significant increase of the indoor surface temperature (about 1.4°C) with a decrease of the wall U-value of about 37%. Moreover, a mitigation of the thermal bridges was also observed with an increase of the minimum surface temperature up to 1.6°C. Introduction The building sector is responsible for nearly 40% of the total energy consumption in Europe (Directive 2010/31/EU). About 50% of the European stock was built before the first thermal regulations in the 1970s (EU Buildings Factsheets). Italy is not an exception, in fact, more than 60% of residential buildings were built before 1976, the year of the first law on energy savings, and 30% of buildings (12.5 million) is dated before 1945 (ISTAT 2011). Therefore, a large part of the building stock is characterized by traditional non- In the last decade, the importance of energy efficiency and thermal comfort in historical buildings has widely increased as it is also testified by a number of recent researches. While energy retrofit was previously seen as a potential threat to the character and fabric of historic and traditional buildings, it is now seen largely as an opportunity to protect these buildings and respond to global environmental concerns (Webb A.L -2017). There is a growing research activity facing the challenge of matching energy efficiency measures and internal thermal comfort with the requirement of maintaining the cultural and historic significance of buildings (De Bouw M. et al. -2016). Among the various technical solutions to reduce thermal losses, technologies for indoor insulation are particularly suitable for this purpose (Walker R. and Pavía S. -2018). Innovative materials and products, like vacuum insulation panels and aerogel-based materials, shall be explored because of their high insulation potential that allows guaranteeing significant indoor space saving compared to traditional insulating materials (Fantucci et al. -2019). In the framework of the on-going European H2020 research project Wall-ACE a novel aerogel-based thermal coating (to be applied together with an additional abrasive resistant top layer) has been developed, with the main aim of mitigating thermal bridges in the existing buildings. In this paper results related to the thermal characterisation in the lab and the performance in-field are presented. The study was aimed at: A c c e p t e d M a n u s c r i p t  the optimisation of the thermal coating  demonstrating (through a full-scale application) the effectiveness of the technology to prevent the risk of surface condensation and the improvement of the whole wall thermal performance. State of the art on aerogel-based wall plasters The use of aerogel granules as lightweight aggregate in a plaster mixture allows the final density and the thermal conductivity to be drastically reduced, reaching values of 150-200 kg/m 3 and 0.025-0.027 W/mK, respectively (Stahl et al. 2012, Ibrahim et al. 2015, Berardi and Nosrati 2018. Further, Buratti et al. (2014) demonstrated that by pushing the aerogel content up to 96-99% (by volume) allows reaching a density of about 115-125 kg/m 3 with a thermal conductivity of 0.014-0.016 W/mK. However, in the same study, for the in-situ application a more mechanically performant formulation with a lower aerogel content (80%) and a thermal conductivity of 0.05 W/mK was used. The effective thermal insulation capability together with the plaster features (i.e easy application on irregular substrates, relatively high compressive strength) make the aerogel-based plaster a suitable candidate for the energy retrofit of historic constructions, in which the old and damaged plaster can be easily replaced with up to 4-6 cm aerogel plaster layers, if the existing layer has no heritage value (Ghazi Wakili et al. 2015, Schuss et al. 2017, Stahl et al. 2017, Lisitano et al. 2018 For these reasons, in the last years, several aerogel-based plasters were developed and are nowadays available at least in the EU market, that are FIXIT 222 (2018), FIXIT 244 (2018), Heck-Aero iP (2017) and Interbran Premium 028 (2019). They typically have a declared thermal conductivity in the range 0.028-0.048 W/mK. Despite the fact that several studies demonstrated the high compatibility with historic buildings and the high energy retrofit capability, this technology is not, so far, largely diffused, because of its high cost, which remains one of the major market limitations. In fact, it was demonstrated that for a thermal insulating plaster improved with 80% of aerogel in the mixture, the cost reached a value between 80-90 €/m 2 for 1 cm A c c e p t e d M a n u s c r i p t of thickness (Buratti et al. 2016, Ibrahim et al. 2017). However, more recent market research highlight that for larger amount of aerogel plaster/render the cost updated at 2019 is in the range 30 €/ m 2 •cm (material cost only) up to 60 €/m 2 per 1 cm thick layer if the cost of application is considered (Aerogel applications, 2016; FIXIT, 2019 and personal communication from Fixit). A possible solution to reduce the final cost is based on minimising the aerogel content by partially replacing it with other lightweight aggregates (de Fátima Júlio et al. 2016). Nevertheless, in most of the studies done so far, the final thermal conductivity of such products revealed to be quite comparable with that of traditional thermal insulating plasters that are available on the market since a long time. Therefore, this solution is not optimal, and no significant benefits can be achieved compared to the state of the art. Other possible solutions aimed at reducing the final cost are based on the minimisation of the thickness. In a previous study of the authors , the development of an aerogel coating with a thickness between 3 and 12 mm has been proposed for thermal bridges mitigation and for slightly improving the wall performance in terms of surface temperature of the non-insulated wall. Methods Five different aerogel-based thermal coatings (called R0 to R4 in the paper) were developed adopting a different ratio of mineral and organic binders. Kwark® granular aerogel, produced by Enersens (2018), perlite, glass and ceramic spheres were used, in various percentage, as lightweight aggregates (LWA). A first series of preliminary tests aimed at determining the thermal and mechanical properties were carried out in the laboratory. The aim was to check the compliance of the plaster with the market needs and its potentials in mitigating thermal bridges and avoiding mould growth risk. Moreover, in order to test the plaster behaviour under actual operating conditions, a monitoring campaign was set-up and carried out on an actual building (full-scale case study) provided by ATC (Regional Agency for the Central Piedmont House). Laboratory characterization A c c e p t e d M a n u s c r i p t The laboratory measurements were aimed at determining: the dry bulk density, the mechanical resistance and the thermal conductivity of the developed aerogel-based plaster ( Figure 1a). The dry-bulk density was measured by weighing a specimen with a known volume (0.4 x 0.4 x 0.05 m), previously dried in a climatic chamber (ACS DM 340, figure 1c) as long as any weight variation of more than 0.2% in 24 h occurred. Thermal conductivity measurement The thermal conductivity was measured on the dried specimen according to UNI EN 12667:2002, by means of a heat flux meter apparatus (Lasercomp FOX 600, figure 1d). Thermal conductivity measurements were performed at an average temperature of 10°C on the dry specimen. The samples were enveloped in a polyethylene sheet in order to avoid any water vapour adsorption during the test. Moreover, the specimens were sandwiched between two natural rubber mats (2 mm thick), having a well-known thermal resistance. This was done in order to reduce the influence of the surface resistance (between the sample and the hot/cold plates) on the measured equivalent thermal conductivity of the plaster. Specific heat capacity tests The specific heat capacity measurements were performed according to the methodology proposed by Tleoubaev and Brzezinski 2017, through the adoption of the Heat Flow Meter apparatus. The instrument can measure the heat required to increase the temperature of the samples for a set of temperature difference. A c c e p t e d M a n u s c r i p t It is necessary to fix the initial temperature set point (T1) and the final temperature (T2), the HFM thus measures the heat absorbed by the sample to increase the temperature from T1 to T2 (Figure 2). The total specific absorbed heat is determined as the integration of the net specific heat flux from the two plates over the time-period, τ , that is: (1) Figure 2: Specific heat required by the sample to increase the temperature from T1 to T2. Finally, the specific heat capacity c p [J/kgK] can be determined as: Where: ΔT is the difference between the first and the second set-point temperatures [°C]; s is the sample thickness [m]; ρ is the sample density [kg/m 3 ]. The specific heat capacity tests were performed on all the aerogel-based thermal coating formulations (that is R0 to R4). The ΔT was set at 10°C, with a temperature variation of the plates from 15 °C (T1) to 25 °C (T2). The standard deviation (SD) of the specific heat capacity measurement was also calculated and reported within the results. Case study building The in-field application was aimed at investigating practical aspects related to the innovative plaster, like, e.g. installation techniques, technical feasibility and appearance, as well as its energy effectiveness. A c c e p t e d M a n u s c r i p t The coating was prepared and applied on CW wall and the adjacent thermal bridge (vertical intersection between wall and window). In order to improve the adhesion of the thermal coating on the existent wall, a mono-component primer in thin layer was firstly applied (figure 5a). The thermal coating, mechanically mixed with ~1:1 water ratio, was manually applied to reach a thickness of ~12 mm (figure 5b, c, d). The measurements were directly performed on the aerogel-based coating layer without applying an abrasive resistant top-layer. In-field monitoring The monitoring campaign was carried out for about two months in the winter period. Temperature and specific heat flux sensors were placed both on CW and RW walls and on the thermal bridges. To avoid the sensors being placed in a non-homogeneous area of the wall, a preliminary set of thermography were performed, by analogy to thermal insulating plasters analyses carried out by Ghazi Wakili et al. (2014) and Fenoglio et al. (2018). As shown in figure 6, two heat flux sensors (HFP01) were located in the centre of walls. T-type thermocouples were adopted for measuring air and surface temperatures, both on the walls and in the thermal bridges zones. Thermocouples were also placed at the interface between the thermal coating and the pre-existing wall (CW) (figure 6). A pyranometer sensor was used to measure the incident solar radiation. The indoor temperature was kept at 23 ± 1°C for all the monitoring period, in order to achieve a temperature difference, between the indoor and the outdoor, high enough to measure with sufficient accuracy the specific heat fluxes. To avoid the A c c e p t e d M a n u s c r i p t influence of the solar radiation on the monitored surfaces, the window shutters were kept closed during the measuremnts. The continuous average method (ISO 9869-1:2014) was used to assess the centre of wall thermal transmittance U (eq.3) and the thermal conductance C (eq.4): where: is the measured specific heat flux; and are the internal and external surface temperatures, respectively; and are the internal and external air temperatures, respectively. Laboratory results Laboratory tests showed that the aerogel coatings are characterised, in general, by good thermophysical properties (table 1). The density of the dry specimens is ~50% lower than that of traditional thermal insulating coating without aerogel (R0 table 1). As it is possible to observe from the table 1, the so far optimized formulation with the Kwark aerogel (R4), has an equivalent thermal conductivity of 0.051 W/mK. This corresponds to a reduction of the equivalent thermal conductivity of about 62% compared to the reference R0 formulation. A c c e p t e d M a n u s c r i p t As far as mechanical properties are concerned, the flexural strength of R4 is equal to 0.8 MPa, with a negligible reduction of performance with respect to the R0 formulation. The intermediate optimisation steps that were implemented for switching from the reference formulation, R0, to the optimized one, R4, were represented by:  the replacement of part of the LWA with the aerogel (from R0 to R1). This action provided a slight improvement of the thermal performance (~13%) with a reduction of the dry bulk density from 617 to 577 kg/m 3 ;  the replacement of part of the mineral binders with organic ones. This action led to a reduction of the thermal conductivity of about a 44% (from R2 to R3) and allowed to reduce the density significantly. Moreover, the differences in binders also determined an increase of the specific heat capacity of about 30%. In-field monitoring results The centre of wall temperature and the specific heat fluxes were analysed considering the entire monitoring period from December 21 st to March 3 rd (Figure 7). Moreover, additional analyses were carried out focusing on the potential thermal bridge mitigation attainable with the thermal coating finish. A c c e p t e d M a n u s c r i p t  decrease of about 1 °C and of about 1.4 °C the median and the minimum outdoor surface temperature, respectively (this aspects mainly concern the durability of the outer plaster layer that might be exposed to freezing-thaw cycles in cold climate) ( Figure 8b);  reduce of about 30% the median value of the centre of wall heat losses during the winter time which can provide heating energy demand reduction (Figure 8c). c c e p t e d M a n u s c r i p t layer. As it has been observed during the field tests, about 2 months are needed for the water in the mixture to be removed. During this time the moisture content is higher than in normal operating conditions, and this fact affects the thermal performance of the layer negatively. All the data presented in this paper have been measured after two months that is near the final condition of the wall. Thermal bridges mitigation The intersection between the wall and the window frame has been identified as the area mostly affected by the presence of the thermal bridge. For this reason, the surface temperatures of the wall side were analysed in detail (Figure 10a). Firstly, the infra-red images of the coated (CW) and uncoated (RW) A c c e p t e d M a n u s c r i p t thermal bridges were compared. Results are shown in figures 10b and 10c; the temperature increases significantly thanks to the presence of the R4 thermal coating. Figure 11 shows the boxplot analysis of the surface temperatures of the CW and RW configurations. The monitored results confirm that the application of the aerogel-based thermal coating allows the temperature in the proximity of the thermal bridge to be appreciably increased. In particular, an effective increase of the minimum surface temperature at 5 cm distance from the construction node (sensor Ts_tb_c), from 15.3°C to 16.9°C, can be observed; while the difference is negligible at 15 cm (sensor Ts_tb_b). Conversely, at 25 cm from the construction node (sensor Ts_tb_a) the application of the thermal coating finish determines slightly lower surface temperature. An additional analysis was also done as far as the potential mitigation of the surface condensation risk is concerned. This study was performed for a point at 5 cm distance from the node, and the results are resumed in Figure 12. The cumulative frequency distribution of the difference between the monitored surface temperature and the dew-point temperature for different indoor relative humidity classes (from 60 to 70%) are plotted. The analysis highlighted that:  no surface condensation occurs at 60% of indoor relative humidity for both the wall configurations (RW and CW);  condensation occurs for about 30% of the time in the reference wall (RW) at 70% indoor relative humidity, while the surface condensation risk can be considered neutralised in the coated wall (CW). A c c e p t e d M a n u s c r i p t A c c e p t e d M a n u s c r i p t Figure 12: Cumulative frequency distributions of the difference between the node temperature (Ts_tb_c) and the dew-point temperature, for different indoor relative humidity conditions. Conclusions In the framework of the Horizon-2020 project 'Wall ACE', a novel aerogel-based thermal coating was developed. As a first research step, the main physical properties of a set of four mixtures were assessed through laboratory analyses. The results allowed to optimise the material properties and to finally select a mixture that presents a thermal conductivity of about 0.05 W/mK. Furthermore, the best performant material was applied for an external wall energy retrofit in a 1920' demonstration building in Turin. The thermal behaviour of the retrofitted wall was monitored during the winter season, and its performance was compared with an uncoated reference wall. The obtained results highlight that the application of a thin thermal coating finishing layer can lead to a significant increase of the median indoor surface temperature (about 1.4°C), with a decrease of the wall thermal transmittance of about 37%. Moreover, a mitigation of the effect of the thermal bridge was also observed with an increment of the minimum node surface A c c e p t e d M a n u s c r i p t temperature (wall-window frame) of up to 1.6°C. The obtained results highlighted that the developed thermal coating can provide a non-negligible reduction of heat losses through the wall. Moreover, the aerogel-based coating represents a promising solution to mitigate the adverse effects of the thermal bridges on the condensation risks, even when the indoor environment is characterised by high relative humidity. These results are a good argument for the suitability of the developed product for the application in all the buildings in which usual thick internal insulating solutions cannot be addressed for space, historical and technical constraints.	2019-07-26T11:17:23.708Z	2019-08-19T00:00:00.000	{ "year": 2019, "sha1": "acd9a8b59ba057c5fb8a47c32564fcb2b3ebf16a", "oa_license": "CCBY", "oa_url": "https://www.tandfonline.com/doi/pdf/10.1080/23744731.2019.1634931?needAccess=true", "oa_status": "HYBRID", "pdf_src": "ScienceParsePlus", "pdf_hash": "ca9239c735f1e03becae9798dee51376c2b6f13d", "s2fieldsofstudy": [ "Engineering", "Materials Science", "Environmental Science" ], "extfieldsofstudy": [ "Engineering" ] }
18903529	pes2o/s2orc	v3-fos-license	Chaos and Asymptotical Stability in Discrete-time Neural Networks This paper aims to theoretically prove by applying Marotto's Theorem that both transiently chaotic neural networks (TCNN) and discrete-time recurrent neural networks (DRNN) have chaotic structure. A significant property of TCNN and DRNN is that they have only one fixed point, when absolute values of the self-feedback connection weights in TCNN and the difference time in DRNN are sufficiently large. We show that this unique fixed point can actually evolve into a snap-back repeller which generates chaotic structure, if several conditions are satisfied. On the other hand, by using the Lyapunov functions, we also derive sufficient conditions on asymptotical stability for symmetrical versions of both TCNN and DRNN, under which TCNN and DRNN asymptotically converge to a fixed point. Furthermore, generic bifurcations are also considered in this paper. Since both of TCNN and DRNN are not special but simple and general, the obtained theoretical results hold for a wide class of discrete-time neural networks. To demonstrate the theoretical results of this paper better, several numerical simulations are provided as illustrating examples. List of Symbols x i : output of neuron i y i : internal state of neuron i ω ij : connection weight from neuron j to neuron i I i : input bias of neuron i k : damping factor of nerve membrane ω ii , ω ii (t) : self-feedback connection weight W : connection weight matrix ǫ : steepness parameter of output function(ǫ > 0) f : Lyapunov function ∆t : time step of the Euler method β : damping factor for ω ii λ p : eigenvalues of matrix λ min : the smallest eigenvalue of matrix W λ max : the largest eigenvalue of matrix W I 0 : self-recurrent bias In particular, the transiently chaotic neural networks (TCNN) with chaotic simulated annealing (CSA) have been shown to be a promising tool for combinatorial optimization [3,4,5,6]. Since TCNN has rich dynamics with various coexisting attractors, not only of fixed points but also periodic and even chaotic attractors, it can be expected to have high ability of escaping from local minima and searching for globally optimal or near-optimal solutions. However, although it has numerically been observed that the numerical chaos with positive Lyapunov exponents exists in TCNN when the temperature, or the self-feedback connection weight is sufficiently large, it has not yet theoretically been proved that TCNN does have chaotic structure. On the other hand, there are a variety of mathematical definitions of chaos, such as geometrically constructed chaos (or geometric chaos) including Horseshoe, Solenoid, and Plykin systems, formal chaos (or topological chaos) with positive topological entropy where shift mapping is included as a subsystem, and numerical chaos in which the maximal Lyapunov exponent is numerically observed to be positive, and so on [16]. The definition of chaos widely used in discrete-time dynamical systems is the Li-York chaos which was proposed by Li and York and which actually corresponds to topological chaos [18]. In 1975, Li and York derived a surprisingly simple condition ensuring the existence of chaos in one-dimension systems, that is, "Period three implies chaos" [18]. In 1978, Marotto extended the Li-York's work to multi-dimensional systems by defining snap-back repellers, that is, "Snap-back repellers imply chaos in R n " [20]. In 1979, Shiraiwa and Kurata further generalized the theorem of Marotto to include not only repellers but also saddle points [26]. Shiraiwa and Kurata also showed that their theorem was a generalization of Smale's results [28] on transversal homoclinic points in some sense. On chaos in discrete-time systems obtained by temporally discretizing continuous-time systems, in 1979 Yamaguti and Matano proved that there exists chaotic structure in some class of one-dimensional difference equations with two fixed points when the difference time is sufficiently large, by using Li-York's Theorem [33]. In 1982, Hata extended the works of Yamaguti and Matano to multi-dimensional difference equations with two fixed points, by using Marotto's Theorem [10]. In 1983, Ushio and Hirai further showed that there exists chaotic behavior in nonlinear sampled-data control systems with two fixed points when the sampling period is sufficiently large [30,31], by using Marotto's Theorem and Shiraiwa-Kurata's Theorem. Motivated by the previous works, this paper aims to theoretically prove that both TCNN and more general discrete-time recurrent neural networks (DRNN) have chaotic structure by applying Marotto's Theorem. A significant property of TCNN and DRNN is that they have only one fixed point, when absolute values of the self-feedback connection weights in TCNN and the difference time in DRNN are sufficiently large. We show that this unique fixed point can actually evolve into a snap-back repeller which generates chaotic structure, if several conditions are satisfied. For the aspects of asymptotical stability, on the other hand, a number of progresses have been made for continuous-time neural networks. For instance, it has been proven that continuoustime neural networks with symmetrical connections always converge to a fixed point [13,14]. However, theoretical results of continuous-time neural networks do not necessarily hold for discrete-time neural networks with continuous-states because analysis of discrete-time neural networks, which also depends on the updating schemes such as synchronous updating and asynchronous updating, is quite different from that of continuous-time neural networks. For some special forms of discrete-time neural networks with continuous-states which are widely used in the analysis of mean field equations [22], Marcus and Westervelt indicated in 1989 that there exist only stable fixed points and stable period-two cycles [19]. Although it has numerically been observed that TCNN with CSA converges to a fixed point attractor if the absolute value of self-connections is sufficient small, it remains theoretically unclear what conditions are sufficient for the convergence and which fixed points are asymptotically stable in TCNN. In this paper, we also aim to analyse discrete-time neural networks of both TCNN and DRNN, and to investigate the asymptotical stability as well as the relations between fixed points and local minima of the Lyapunov functions. In other words, by using Lyapunov functions, we derive sufficient conditions on asymptotical stability for several symmetrical versions of both TCNN and DRNN, under which these networks asymptotically converge to a fixed point. We show that the convergent properties of discrete-time recurrent neural networks are quite different from those of continuous-time recurrent neural networks although many of them are similar in terms of asymptotical stability and locations of fixed points. Furthermore, the relationships between the asymptotically stable points and the minima of the Lyapunov functions are also investigated for both asynchronously and synchronously updating schemes, by which TCNN and DRNN can be applied to combinatorial optimization problems. In addition, we also give conditions of generic bifurcations, i.e., the fold and flip bifurcations, and explain the bifurcation structure of chaotic simulated annealing. Notation and Marotto's Theorem We use the following notation, similar to that of Marotto [20] throughout the paper. Let us define a discrete-time system as, where, Y (t) ∈ R n , (t = 1, 2, ...) and F ∈ C 1 : R n → R n . Let F t denote the composition of F with itself t times, then a point Y is a p-periodic x ∝ y means that x is directly proportional to y. 0 is the n × n zero matrix. A small letter stands for a scalar, and a capital letter represents a vector, a matrix or a set in this paper. Definition 2.1 (Marotto, 1978) A fixed point Y * is said to be a snap-back repeller of eqn.(1), if there exists a real number r (> 0) and Asymptotical Stability and Local Bifurcations of Transiently Chaotic Neural Networks In this section, first the transiently chaotic neural network (TCNN) is briefly summarized. Then, several convergence theorems for symmetrical versions of TCNN are derived. Besides, the relations between asymptotically stable points of TCNN and local minimal points of the Lyapunov function are also investigated. A model of the transiently chaotic neural network The model of the transiently chaotic neural network (TCNN) can be described in terms of scalar variables as follows [3,4,5,6]: where, ω ii = ω ii (t) is a variable depending on eqn.(4) and all variables and parameters are real numbers, and x i : output of neuron i y i : internal state of neuron i ω ij : connection weight from neuron j to neuron i a i : input bias of neuron i k : damping factor of nerve membrane ǫ : steepness parameter of the output function (ǫ > 0) : damping factor of the time-dependent ω ii (t) (0 ≤ β ≤ 1) a 0i : self-recurrent bias of neuron i. Since we investigate sufficient conditions of chaos and asymptotical stability for TCNN with temporally constant ω ii , ω ii is fixed by dropping eqn.(4) in the following analysis.Therefore, TCNN can be rewritten in terms of vector variables as follows. where, diag(ω 11 , ..., ω nn ) is a diagonal matrix and Obviously, F in eqn.(5) is a C 1 class function for any bounded Y . Besides, F can be viewed as an iterated map or a difference equation or a neural network with discrete-time and continuous-state. If I 0 = E 0 , eqn.(5) coincides with chaotic neural networks (CNN) [2,6]. Asymptotical stability for TCNN Next, we first show several convergent theorems for eqn. (6), which ensure that the system of eqn.(6) asymptotically converges to a stable fixed point, and then derive locally stable conditions by analysing the eigenvalues of the Jacobian matrix. Let us define two updating schemes, i.e., synchronously updating and asynchronously updating. In synchronously updating, all neurons are updated in parallel, using only old values as input. In asynchronously updating, neurons are updated one by one, using fresh values of previously updated neurons [22]. Generally, asynchronously updating can further be divided into random-order asynchronously updating where neurons are chosen in random order, and fixed-order asynchronously updating where neurons are chosen in fixed-order, e.g., cyclic updating. In this paper, asynchronously updating means both random-order and fixed-order asynchronously updating. Let λ min , λ max be the smallest eigenvalue and the largest eigenvalue of the matrix W , respectively. Define . Obviously, f (X) : R n → R is bounded for all 0 ≤ x i ≤ 1 (i = 1, ..., n). Theorem 3.1 Assume W = W T , and Then X(t) of the discrete-time system eqn. ≤ 0 which is used to derive eqn. (9). Furthermore, we get the following equations by expanding where, ϕ i is between x i (t) and x i (t + 1), and min 1 ϕ i (1−ϕ i ) = 4 for 0 ≤ ϕ i ≤ 1 is used in eqn. (11). Hence the following inequality holds. From the assumption of this theorem, the matrix W + 4(1 − k)ǫ 1 is positive definite. Hence, f (t + 1) − f (t) ≤ 0, where equality holds only for ∆X(t) = 0, which means that X(t) has reached a fixed point of the network dynamics. Note that since W is a symmetry matrix, all the eigenvalues of W are real numbers, i.e., λ min ∈ R. For condition 2 : We use a different approximation scheme to derive condition 2. We rearrange eqn. (8) as follows: where, The mean value theorem is applied in eqn. (15). Since ln x 1−x is a monotonously increasing function for 1 > x > 0 and ξ i is between x i (t) and 1−x i (t) take the same sign. Furthermore, Therefore for 1 ≥ k ≥ 0, because of eqn. (16) and the monotonously increasing function where η i is between x i (t) and x i (t + 1). In eqn. (17),eqn. (16) and condition 2 are used. The mean value theorem is applied in eqn. (18) again, and min 1 η i (1−η i ) = 4 for 0 < η i < 1 is substituted in eqn. (19). The equality holds only for ∆X(t) = 0, which means that X(t) has reached a fixed point of the network dynamics. Therefore, f (X) is a Lyapunov function for eqn. (6), which indicates that X(t) of the system asymptotically converges to a fixed point. ✷ [19] derived the condition 1 for k = 0. Therefore, Theorem 3.1 can be viewed as a generalization of the Marcus and Westervelt's result. These show that the system of X(t) is asymptotically stable. Theorem 3.2 Assume W = W T , and Then X(t) of the discrete-time system eqn.(6) asymptotically converges to a fixed point, as far as eqn. (6) is asynchronously updated. Proof of Theorem 3.2 . Without loss of generality, let neuron-l be the updated neuron at iteration-(t + 1), then the change in f defined by eqn. (7) due to this updating is, Then, as the same way as the proof of Theorem 3.1, we can prove this theorem by showing that f (X) is also a Lyapunov function of eqn. (6) with an asynchronously updating scheme. ✷ From the definition of x i (y i (t)) of eqn.(5) and the mean value theorem, where, ρ i is between x i (t) and x i (t + 1). In eqn.(23), eqns. (2) and (7) are used. Therefore, the proof of Lemma 3.1-1 is straightforward. For Lemma 3.1-2 : The stationary points satisfy If k = 1, while the second term and the third term on the right hand side in eqn. (24) are bounded, the first term becomes unbounded when x i → 0 or 1. This fact proves the Lemma Proof of Theorem 3.3. Assume that there exists a point X * which is an asymptotically stable point but not a local minimum of f (X). Then, according to the definition of asymptotical stability for discrete-time systems in the sense of Lyapunov [27], for an arbitrary neighbourhood U around X * , there exists V ⊂ U such that any trajectory with . Therefore, when the trajectory starts fromX at t = 0, which implies that the Lyapunov function increases with t. This contradicts Theorem 3.1 or Theorem 3.2. Furthermore, according to Lemma 3.1 any local minimum of f (X) is a fixed point of X(t) for eqn. (6). Therefore, any asymptotically stable point must be a local minimum On the other hand, since any local minimum of there exists a neighbourhood U around this local minimum X * such that f (X * ) −f = 0 and f (X)−f > 0 for X ∈ U and X = X * , wheref is a bounded constant. Moreover, from Theorem Therefore any local minimum is asymptotically stable. ✷ Since k = 1 and ω ii = 0, f (X) defined in eqn. (7) is a multilinear form which has no local minimum belonging to stationary points. In other words, f (X) has no local minimum for E 0 ≤ X ≤ E 1 except some vertices. Therefore, we get this Corollary from Theorem 3.3. ✷ Although Theorem 3.1 and Theorem 3.2 are simple and easy to apply, they are conservative estimators for the evaluation of asymptotical stability. Next, we give two more general theorems to evaluate the local stability in an exact manner. For synchronous updating, define the Jacobian matrix J X of eqn.(6) at X(t), If X = (x 1 , ..., x n ) is an internal fixed point, which means E 0 < X < E 1 , X(t) = X(t + 1) = X, then eqn.(26) simply becomes, or is expressed in terms of f XX , where On the other hand, if X is a fixed point on boundary or vertice, for any = lim Therefore, As the same way, for any Therefore, substituting these relations (32)-(33) into eqn. (26), combining with the Jacobian matrix J X eqn. (27) of internal point, we have the generalized Jacoban matrix of a fixed point ., x * n ) for internal, boundary or vertice points, where 1. For synchronous updating, a fixed point X * in the region S is asymptotically stable if the largest eigenvalue λ ′ max and smallest eigenvalue λ 2. For asynchronous updating, a fixed point X * in the region S is asymptotically stable if Proof of Theorem 3.4 . For synchronous updating: According to dynamical systems theory [11], the fixed point X * is asymptotically stable if absolute values of all eigenvalues for J X at X * are less than 1. On the other hand, if there exists an eigenvalue with the absolute value more than 1, X * is unstable. Note that all eigenvalues of J X are real numbers according to Theorem C.1 of Appendix C. For asynchronous updating: Without loss of generality, let neuron-l be the updated neuron at iteration-(t+1), then Analogously to the proof of synchronous updating, we obtain condition 2 for asynchronous updating, by using relations (32) 2. on the other hand, a fixed point X * in the region S is unstable if either \|λ cmax \| > 1 or \|λ cmin \| > 1 is satisfied; 3. when k < 1, no point on vertices or boundary points is asymptotically stable. Proof of Theorem 3.5 . Since the cyclic updating is proceeded neuron by neuron, the total differential is, where, m ij is the (i, j) element of the matrix J X defined in eqn.(34). Let us change the Jacobian matrix of cyclic updating into that of the effective synchronous updating. By dividing J X into an upper triangular U a , a diagonal D a and a lower triangular L a [22], eqn.(37) can be rewritten in terms of matrix form; Therefore, we have the effective synchronous updating Jacobian matrix, which means that f XX (X * ) is indefinite or negative definite (or semi-negative definite), then is negative according to Theorem C.1 of Appendix C. Hence from eqn. (28), there is at least one eigenvalue of J X whose absolute value is more than 1, which means that X * is unstable. Therefore, only local minimum of f (X) can possibly be asymptotically stable. ✷ According to eqn.(34), the closer x * i (i = 1, ..., n) are to 0 and 1, the smaller the absolute values of eigenvalues λ ′ min and λ ′ max are. Therefore, this implies that the system is likely to asymptotically converge to local minima near the vertices for −1 < k < 1. 1. When k < 1, no point on vertices or boundary points is asymptotically stable for both asynchronous and synchronous updating. 2. When k > 1, all vertices are asymptotically stable for both synchronous and asynchronous updating. 3. For both synchronous updating and cyclic updating, when k = 1, a vertex X * is asymp- 4. For random-order updating, when k = 1, a vertex is asymptotically stable if TCNN to practical problems, e.g., optimization or associative memeory, because they indicate which local minimum of the Lyapunov function is an attractor and which one is not. Local Bifurcations in TCNN Next, let us examine the bifurcations of the fixed points, in particular of the stationary points of f (X), by taking P as bifurcation parameters. P is a parametric column vector and may be defined by P = (ω 11 , ..., ω nn ) T or other parameters in eqn. (5). Since many applications take minima of f (X) as targets of information processing, we focus on analysing bifurcations for the stationary points of f (X), especially for internal fixed points. Instead of eqn.(6), we use eqn. (5) in the following analysis of the codimension one bifurcations of internal points, by applying the center manifold method [8,9]. Eqn.(5) is rewritten as, f (X; P ) is defined in eqn. (7) with parameters P expressed explicitly, and f X = ∂f ∂X . Then the Jacobian matrix J Y (X(Y (t)); P ) = ∂Y (t+1) .., n) be an internal fixed point, which is in fact a stationary point of f (X; P ) for W T = W according to Lemma 3.1. Let P c be the critical values for bifurcations. Then ∂f (X;P ) ∂X can be expanded in the vicinity of (Y * , P c ) in Taylor series with respect to both variables Y and parameters P , such that eqn.(40) becomes, Since X * or Y * is a fixed point and P c are the critical values of bifurcations, f X (X * ; P c ) = 0 and at least one of eigenvalues for the Jacobian J Y (X * ; P c ) at (X * , P c ) is λ = 1 or λ = −1. Let us first analyse the case of λ = 1 by using the center manifold method. Let U, U † be the normalized right and left eigenvectors of Note that U † is a row vector. Since f XY (X * ; .., n, eqn.(43) implies that f XX (X * ; P c ) has at least one zero eigenvalue. Then the center subspace at the vicinity of the bifurcation point can be defined by where a(t) is a scalar and a sufficiently small real number. Substituting eqn.(44) into eqn.(42) and then multiplying U † to eqn.(42), we have which is a one-dimensional difference equation for variable a(t) with parameters ∆P . Theorem 3.6 Let W T = W . Assume that X * is an internal fixed point of eqn.(40) or eqn. (5) and that f XY (X * ; P c ) has one zero eigenvalue with the right eigenvector U and the left eigenvector U † . Then for synchronous updating, a fold bifurcation (saddle-node bifurcation) occurs if Since the bifurcations in Theorem 3.6 occur when f XX (X * ; P c ) has zero eigenvalue and f X (X * ; P c ) = 0, they are actually stationary point bifurcations of f (X; P ). Next, let us analyse the case that the Jacobian matrix J Y has at least one eigenvalue λ = −1. Let X * be a fixed point and V, V † be the normalized right and left eigenvectors satisfying In other words, diag(x * 1 (1 − x * 1 ), ..., x * n (1 − x * n ))f XX (X * ; P c ) has at least one eigenvalue 2ǫ (> 0). By using Theorem C.1 of Appendix C, it is easy to show that f XX (X * ; P c ) has at least one positive eigenvalue at the critical point. Then the center subspace at the vicinity of the bifurcation point can be defined by where b(t) is a scalar and a sufficiently small real number. As the same way of derivation of eqn.(45), we have which is a one-dimensional difference equation for variable b(t) with parameters ∆P . Note that V † is a row vector. Then for synchronous updating, a flip bifurcation (a period-doubling bifurcation) occurs if and others [25]. In this subsection, we only examine the bifurcations of synchronous updating for the sake of the simplification of analysis. However, the bifurcation conditions of asynchronous updating can be derived analogously to synchronous updating. Obviously, all of analyses in this section are independent of the values of I and I 0 in eqn. (6). Chaos in Transiently Chaotic Neural Networks In the first of this section, one theorem for existence of a fixed point with sufficiently large absolute values of self feedback connection weights is derived. Then, we give sufficient conditions for chaos by applying Marotto's Theorem in the last of this section. Analysis in this section is mainly based on synchronous updating, but it is easy to apply it to asynchronous updating. In other words, all the theoretical results in this section hold for both synchronous and asynchronous updating. Fixed point in TCNN Firstly, we show that eqn.(5) possesses a fixed point for sufficiently large \|ω\|. such that the following three conditions hold for all \|ω\| > c 2 . Since X(Y ) is a monotonously increasing function for Y , and the value region of X(Y ) is From the condition 2 of Lemma 4.1, Then, according to the condition 3 of Lemma 4.1, which means Y =Ȳ (ω) is an approximate solution of Q(Y ; ω) = 0 for a sufficiently large number \|ω\|. Next, let us show that Y * is a unique fixed point for all bounded region if \|ω\| is sufficient large. Assume Y 0 to be another bounded fixed point different from Y * , in other words Y 0 ∈ U. Since both Y * and Y 0 are fixed points of eqn.(5) and W is invertible, the following relation holds. Therefore, X(Y * ) − X(Y 0 ) or Y * − Y 0 approaches zero when \|ω\| is sufficient large, because Y * and Y 0 are bounded. Since Y * ∈ U, there exists a sufficient large c 0 (> 0) such that Y 0 ∈ U for all \|ω\| > c 0 , which contradicts the uniqueness of the fixed point in U. Therefore, the theorem is proved. ✷ 1−a 01 , ..., ln a 0n 1−a 0n ] T , which is also a unique fixed point (bounded) for \|ω\| → ∞. Proof of Remark 4.1 . Since Y * ∞ is bounded, and Hence, the Remark is proven. ✷ The proof of this remark is straightforward from Remark 4.1. From eqns. (64) and (66), for sufficiently large \|ω\| which is negative because of the conditions of 1 and 2. Hence, is an approximate solution of Q(Y ; ω) = 0 for a sufficiently large number \|ω\|. Proof of condition 3: Let G = (g 1 , ..., g n ) T , where, Define Note that the following condition should be satisfied, 0 < g i < 1; (i = 1, .., n). Furthermore, Hence, for sufficiently large \|ω\|, which is positive because of the condition 3. Note that the condition 2 also satisfies eqns. Remark 4.4 Chaos in eqn. (5) is generated not from a saddle point, but from a repeller when \|ω\| is sufficiently large. For the case of k < 0 : As the same way, we can show thereby proving that any y i (t + 1) is bounded for any initial value y i (0). ✷ As a matter of fact, for steady states, i.e., when t is sufficiently large, k t+1 or (−k) t+1 −→ 0. We first use the condition 2 of Theorem 3.1 or Theorem 3.2 to estimate the stability of this example. According to the condition 2 of Theorem 3.2, the estimated value of ω is ω e = −8kǫ = −0.0288, below which the system can be ensured to be stable. If −ω < 0.0288, the system is stable and converges to a fixed point, which is verified by the numerical simulation in Figure 1. Next, we use Theorem 3.4 to calculate the first bifurcation point and to analyse the dynamics exactly. The fixed point satisfies According Theorem 3.4, the condition under which the fixed point is stable follows because we only analyse the dynamical behaviour for k > 0 and ω < 0. Therefore, by solving eqns. (86) and (87), we obtain the critical values of ω and the fixed point x * , Therefore, If −ω < 0.033104, the system is stable and converges to a fixed point; at −ω = 0.033104, the fixed point bifurcates into 2-periodic points which is confirmed by Theorem 3.7; when −ω > 0.033104, the system becomes unstable. These facts are verified by the numerical simulation in Figure 1. Finally, we use the condition 3 of Theorem 4.2 to analyze chaotic dynamics of this example. According to the condition 1 of Theorem 4.2, there exists chaotic behavior when −ω is sufficiently large. Actually, in this example, when −ω > 0.0596, the chaotic behavior appears, which is numerically verified by positive Lyapunov exponent in Figure 1, although there also exist many periodic windows. Asymptotical Stability in Discrete-time Recurrent Neural Networks In this section, we firstly give a general form of discrete-time recurrent neural networks. Then, several convergence theorems for symmetrical versions of DRNN are addressed. All the theorems in this section can be proved by similar arguments as in section 3. Difference scheme The continuous-time recurrent neural network can generally be written in forms of differential equations as follows (Amari, 1972;Hopfield,1984): where all variables and parameters are real numbers. Then, the discrete-time recurrent neural networks (DRNN) on the basis of the Euler method are given in eqns.(89). These show that the system of X(t) is asymptotically stable. Then the discrete Although Theorem 5.1 and Theorem 5.2 are simple and easy to apply, they are conservative estimators for the evaluation of asymptotical stability. Next, we give two more general theorems to evaluate the local stability in an exact manner in a similar way as Section 3. For synchronous updating, define the Jacobian matrix J X of eqn.(90) at X(t) as; Then for any x i = 0 or 1, Therefore, the Jacobian matrix for a fixed point X * = (x * 1 , ..., x * n ) is where Next, we give a theorem generalized to boundary points and vertices and internal points of X. For asynchronous updating, a fixed point in the region Proof of Theorem 5.4 . For synchronous updating. The Jacobian matrix is given in eqn.(96), and all eigenvalues of J X are real numbers according to Theorem C.1 of Appendix C. For asynchronous updating: Without loss of generality, let neuron-l be the updated neuron at iteration-(t+1), then According to eqn.(96), the closer x * i (i = 1, ..., n) are to 0 and 1, the smaller the absolute values of eigenvalues λ ′ min and λ ′ max are. Therefore, this Theorem implies that the system is likely to asymptotically converges to local minima near the vertices for −1 < 1 − µ∆t < 1. 1. When k < 1, no point on vertices or boundary points is asymptotically stable. 2. When k > 1, all vertices are asymptotically stable for synchronous and asynchronous updating. 3. For both synchronous updating and cyclic updating, when k = 1, a vertex X * is asymptot- 4. For random-order updating, when k = 1, a vertex is asymptotically stable if Chaos in Discrete-time Recurrent Neural Networks In this section, one theorem for a fixed point with sufficiently large difference times is firstly addressed, and then our main theorem for chaos is described in the last of this section. The function f is defined in eqn.(91). All the theorems in this section can be proved by similar arguments as in section 4. Fixed point in DRNN First, we show that eqn.(89) possesses a fixed point for sufficiently large ∆t. Theorem 6.1 Assume the following two conditions. Existence of chaos in DRNN Next, we establish a principal theorem of this section. Let G = (g 1 , ..., g n ) T , where, Then define Therefore, X(Y 0 ) = G by eqns. Note that lim ∆t→∞ g i = −H T i I(1/k + 1) which is between 0 and 1 ensured by the condition 1 of Theorem 6.1 and the conditions 1 and 2 of Theorem 6.2. Since, then for sufficiently large ∆t which is negative because of the conditions 1 and 2. Note that lim ∆t→∞ g i = −H T i I(1/k + 1) − 1/k which is between 0 and 1 ensured by the condition 1 of Theorem 6.1 and the conditions 3 and 4 of Theorem 6.2. Since then for sufficiently large ∆t which is positive because of the conditions 3 and 4 of Theorem 6.2. Then Theorem 6.2 can be proved as the same procedure as that of proof of Theorem 4.2. ✷ Obviously, there are two different type of snap-back repellers for Theorem 6.2. One is Figure 2 shows the output of neuron and the maximum Lyapunov exponent with increasing ∆t. Analogously to Example 4.1, we first use the condition 2 of Theorem 5.1 or Theorem 5.2 to estimate the stability of this example. According to the condition 2 of Theorem 5.2, the estimated value of ∆t is ∆t c = 8ǫ 8ǫµ−ω = 0.03101, below which the system can be ensured to be stable. If ∆t < 0.03101, the system is stable and converges to a fixed point, which is verified by the numerical simulation in Figure 2. Since the Theorem 5.1 and Theorem 5.2 are conservative although they are easy to use, next we apply Theorem 5.4 to exactly evaluate locally asymptotical stability of this example. The fixed point satisfies independent of ∆t. According Theorem 5.4, the condition under which the fixed point is stable is as follows because we only analyse the performances for k > 0. Therefore, by solving eqns. Therefore, If ∆t < 0.032744, the system is stable and converges to a fixed point; at ∆t = 0.032744, the fixed point bifurcates into 2-periodic points; when ∆t > 0.032744, the system becomes unstable. These facts are verified by the numerical simulation in Figure 2. Finally, we use the condition 1 of Theorem 6.2 to analyze chaotic dynamics of this example. According to the condition 1 of Theorem 6.2, there exists chaotic behavior when ∆t is sufficiently large. Actually, in this example, when ∆t > 0.09536, the chaotic behavior appears, which is numerically verified by positive maximum Lyapunov exponents in Figure 2, although there also exist many periodic windows. If ∆t < 0.04381, the system is stable and converges to a fixed point; at ∆t = 0.04381, the fixed point bifurcates into 2-periodic points; when ∆t > 0.04381, the system becomes unstable. These facts are verified by the numerical simulation in Figure 3. For the asynchronous updating: Here, we use cyclic updating for simulation. According to the condition 2 of Theorem 5.2, the estimated value of ∆t is ∆t e = 8ǫ 8ǫµ−ω ii = 0.01575, below which the system can be ensured to be stable. Actually, as shown in Figure 4, when ∆t < 0.01575, the system is stable and converges to a fixed point. Next, we use Theorem 5.4 to analyse stability of this example in more exact manner. Since the fixed point is independent of ∆t and updating schemes, the fixed point is X * = (x * 1 , x * 2 ) = (0.175392, 0.0615991) calculated in eqns.(112). According to the condition 2 of Theorem 5.4, the sufficient condition under which the fixed point is locally stable follows because of µ > 0 and ω ii < 0. Therefore, by solving eqns.(114), we obtain, ∆t s = 0.053825. If ∆t < 0.053825, the system is stable and converges to a fixed point. When ∆t > 0.0582, the system becomes unstable, as shown by the numerical simulation in Figure 4. On the other hand, according to the condition 1 of Theorem 6.2, there exists chaotic behavior when ∆t is sufficiently large. Actually, in this example, when ∆t > 0.06066 (for synchronous updating) or ∆t > 0.06915 (for asynchronous updating), the chaotic behavior appears, which is numerically verified by positive maximum Lyapunov exponents in Figures 3 and 4. Conclusion This paper theoretically proved that both TCNN and difference equations of discrete-time recurrent neural networks (DRNN) have chaotic structure by applying Marotto' Theorem, and gave sufficient conditions for existence of both a fixed point and chaotic behavior. A significant characteristic of TCNN and DRNN is that they have only one fixed point which generates chaotic behaviour of neural networks when the absolute value of self-feedback connection weights in TCNN and the difference time in DRNN are sufficiently large. Besides, this paper also investigated the relations between the asymptotically stable points and the minimal solutions of the Lyapunov function, and derived the conditions for which both TCNN and DRNN can asymptotically converge to a fixed point, which ensure the convergence of chaotic simulated annealing [5,6]. These convergent theorems are also useful for applications of TCNN and DRNN to practical problems, e.g., optimization and associative memory, because they indicate which local minimum of Lyapunov function is an attractor and which one is not. The local bifurcations were also investigated to show the bifurcation process for chaotic simulated annealing [6]. Since the models which we used in this paper are simple and general, the obtained theoretical results hold for a wide class of discrete-time neural networks. APPENDICES A Urabe's Proposition (1965) Let F : R n → R n be a continuously differentiable function in some region Ω ⊂ R n . Assume that equation has an approximate solution Y =Ȳ for which det(F Y (Ȳ )) is not zero, and there is a constant ǫ (> 0) and a constant µ (≥ 0) such that where r and d (> 0) are positive numbers such that F (Ȳ ) < r and F Y (Ȳ ) −1 < d. Lemma C.1 The matrix K + ZW has the same eigenvalues as those of the matrix K + √ ZW √ Z. Proof of Lemma C.1. Theorem C.1 Assume W is a symmetrical matrix, i.e., W = W T , and k ii , z ii are real numbers. Then the following results hold. 1. The matrix K + ZW has the same eigenvalues as those of the matrix K + √ ZW √ Z. All of these eigenvalues are real numbers. Proof of Theorem C.1. For condition 1. According to Lemma C. where √ Z i and A i are submatrices of √ Z and A defined as the same as W i , respectively. Note that different from an asymmetrical matrix ZW , √ ZW √ Z or A is a symmetrical matrix. Since z ii > 0 for i = 1, ..., n which means that √ Z is positive definite, det( √ Z i ) > 0 for i = 1, ..., n. Therefore, to ensure det( √ Z i )det(W i )det( √ Z i ) > 0 for i = 1, ..., n, the inequalities det(W i ) > 0 for i = 1, ..., n must hold, which means that W must be positive definite. On the other hand, if W > 0, then det( √ Z i )det(W i )det( √ Z i ) > 0 for i = 1, ..., n, which means that A = √ ZW √ Z is also positive definite. Therefore, the positive definiteness of √ ZW √ Z is identical to that of W . As the same way, we can prove the cases for semi-positive definite, indefinite, negative definite and semi-negative definite.	2014-10-01T00:00:00.000Z	1997-01-22T00:00:00.000	{ "year": 1997, "sha1": "6d553288a7db5cc647e3cd7a7925fb6fcba94cc6", "oa_license": null, "oa_url": "http://arxiv.org/pdf/chao-dyn/9701020v1.pdf", "oa_status": "GREEN", "pdf_src": "Arxiv", "pdf_hash": "2221e7e023c20cf85b8aa359491c634c6a454ca5", "s2fieldsofstudy": [ "Computer Science" ], "extfieldsofstudy": [ "Mathematics", "Physics" ] }
253627087	pes2o/s2orc	v3-fos-license	Rapid diagnosis of a complex oral mucosal infection using metagenomic next-generation sequencing: a case report The most commonly used methods for pathogen detection and identification in oral mucosal infectious diseases are DNA or RNA quantitative polymerase chain reaction detection, bacterial or fungal cultures, and immunohistochemical analysis. These traditional methods are time-consuming and can only detect one specific targeted pathogen at a time. An efficient and sensitive method with higher species richness is urgently needed. Metagenomic next-generation sequencing (mNGS) is a new method of pathogen detection with high efficiency and sensitivity. In this case report, mNGS was used to identify the pathogens in oral mucosal tissues of a patient with complex oral mucosal infections and oral leukoplakia. Candida albicans, human gamma herpesvirus 4, and many other pathogens were identified using this method. For complex oral mucosal infections, mNGS is a more efficient and sensitive approach that can replace conventional detection methods. Introduction The oral cavity has the second largest and most diverse microbiota, harboring over 700 species of bacteria. 1 The soft tissues of the oral mucosa are complex habitats where microbes colonize. Like systemic infections, oral mucosal infectious diseases are usually caused by bacteria, viruses, and fungi such as Porphyromonas gingivalis, herpes virus, and oral Candida, respectively. For viralinduced oral mucosal infectious diseases, the most common diagnostic method is quantitative polymerase chain reaction (qPCR) detection of viral DNA or RNA, which is highly efficient and sensitive. The limitation of this method is that only specifically targeted viruses can be detected, and the species richness is limited. For oral mucosal infectious diseases caused by bacteria and fungi, bacterial or fungal cultures are the most common diagnostic method. However, the time of culture can be relatively long, generally between 48 and 72 hours, and it can also be hard to distinguish different species from one another. 2 A method with high efficiency, sensitivity, and species richness is urgently needed to detect pathogens for oral mucosal infectious diseases. Metagenomic next-generation sequencing (mNGS) is a new detection method with high efficiency and sensitivity that involves high-throughput sequencing; additionally, the turnaround time of mNGS is shorter compared with other detection methods. Pathogen detection and identification by mNGS is emerging in clinical practice. Only a small amount of DNA needs to be extracted from a sample to simultaneously detect and identify pathogens using mNGS. 3 Moreover, mNGS is not dependent on culture so it is more sensitive than conventional culture methods. 3 mNGS has already been successfully applied in several clinical trials to diagnose infections, and the results have shown that mNGS provides more precise and faster pathogen detection and identification. 4 However, after a review of the literature, mNGS has rarely been used to diagnose oral mucosal infectious diseases. In this case report, we detail a patient who had undergone s/p allogeneic bone marrow transplantation and been taking immunosuppressants for a long time. Therefore, the patient was susceptible to oral mucosal infectious diseases. For safe and effective treatment, we used mNGS to identify the pathogens in oral mucosal tissues of the patient. We then discuss these results, which are of great significance to the treatment of patients with oral mucosal infectious diseases. Case report A 60-year-old male had been suffering from a rough feeling in the gingiva of his left maxillary posterior teeth and left buccal mucosa for 1 year. No tobacco-chewing or smoking habits were reported by the patient. The basic condition of this patient was also complex. In his medical history, the patient reported leukemia 13 years ago and had undergone s/p allogeneic bone marrow transplant in the same year, which led him to regularly take immunosuppressants (tab methylprednisolone, 4 mg orally per day) for several years. Suffering from insomnia and anxiety, the patient also takes zolpidem tartrate orally (10 mg at bedtime). Clinical examination of the mouth on the initial visit was as follows: the mouth opening was only 27 mm, there were painless white verrucous plaque lesions with multiple peaks on the gingiva and vestibulum oris mucosal located in the left maxillary first molar, and the mucosa of the lesion area was fragile and prone to bleeding ( Figure 1a). There was a white lesion with flat or slightly wrinkled surface located in the left buccal mucosa with an ulcer of approximately 3 mm in diameter (Figure 1b, c). According to the clinical symptoms, the patient was clinically diagnosed with oral leukoplakia (OLK) with infection. Incisional biopsy of the diseased mucosa was performed for a definitive diagnosis. The biopsy tissue was divided into two parts. One part was fixed in 10% formalin solution, and the other was stored at À80 C until DNA and RNA extraction. After embedding, the sample fixed in 10% formalin solution was used for hematoxylin-eosin staining for histopathology. The other tissue sample was used for mNGS. Briefly, 0.5 mL of BioSpec beads (0.5-mm diameter, ZIRCONIA/SILICA Cat. No. 11079105z, Thermo Fisher Scientific, Waltham, MA, USA) were added into the tissue sample. After vigorous agitation for 30 minutes, the sample was ready for DNA extraction. RNA was reverse transcribed and synthesized to double-stranded complementary DNA using the SuperScript II Reverse Transcription Kit (18064-014, Invitrogen, Waltham, MA, USA). After second-strand DNA synthesis, DNA libraries were constructed. Agilent 2100 (Santa Clara, CA, USA) was used for quality control, and DNA libraries were then sequenced using the BGISEQ-100 platform. Next, clean reads were mapped to a human reference database using Burrows-Wheeler Alignment (version 0.7.10). The mapped data were then processed for advanced data analysis. Histopathological examination revealed squamous epithelial hyperplasia, hyperkeratosis, and inflammatory cell infiltrated into the connective tissue ( Figure 2). Subsequently, a histopathological diagnosis of OLK was made. Further pathogenic microorganism DNA/RNA high-throughput genetic sequencing (PMseq)-DNA detection took approximately 36 hours, and the results showed overlapping infections with Candida albicans and Human gamma herpesvirus 4 (EBV). Other bacteria such as Rothia dentocariosa, Prevotella denticola, and Veillonella parvula, were also detected. The PMseq-DNA detection results are shown in Figure 3. On the basis of the PMseq-DNA detection results, 200 mg oral fluconazole was prescribed on the first day, and then 100 mg on the second day, for a total of 7 days. Acyclovir (200 mg orally) was prescribed 5 times per day for 7 days. Additionally, 2% sodium bicarbonate solution was used as mouthwash after eating. The follow-up visit was 2 months later, at which time the ulcer in the left buccal mucosa showed complete response after treatment and the surface of the white lesions had become flatter (Figure 4). The reporting of this study conforms to the CARE guidelines. 5 Discussion In this case, we used mNGS to detect the pathogens in the oral mucosal tissues of a patient with complex oral mucosal infection and OLK who was susceptible to oral mucosal infectious diseases due to s/p allogeneic bone marrow transplantation and use of immunosuppressants. The turnaround time of mNGS was 36 hours, which is shorter than other detection methods. mNGS detects pathogenic microorganisms in infected samples, not the tissue itself and can simultaneously detect almost all bacteria, viruses, and fungi. Compared with traditional detection methods, mNGS has higher efficiency, sensitivity, and species richness. For complex oral mucosal infections, mNGS can replace conventional detection methods. OLK is a common and potentially malignant oral disorder that has a high malignant transformation rate of between 0.13% and 34% into oral squamous cell carcinoma. 6 The etiology of OLK is complex, and infection is considered to be one of the major pathogenic factors of OLK. Candida infection has been found in 34% of OLK, and Candida invasion is a significant risk factor for malignant transformation in OLK. EBV infection 7 is also strongly associated with the malignant transformation of OLK. The oral microbial environment is complex. Early detection and inhibiting the growth of pathogens in the oral cavity is of great significance for the treatment and inhibition of malignant transformation in OLK. Candida infection may cause damage to the oral mucosa, such as atrophy, hyperplasia, and dysplasia. 8 Candida can be divided into many different species, such as Candida albicans, Candida tropical, and Candida glabrata. 9 Candida culture is the most commonly used method in the clinical diagnosis of oral Candida. The most frequently used primary isolation medium for Candida culture is Sabouraud Dextrose Agar (SDA). 10 However, the time of culture is relatively long, generally between 48 and 72 hours, and it is hard to distinguish different species from one another. 2 Pathogen detection and identification using mNGS can only take approximately 36 hours or less, and different species of Candida can be accurately distinguished. For patients with HIV infection, on corticosteroids or antibiotic therapies, and in post-transplant immunosuppression have higher infection rates for oral candidiasis. 11 In this case, the patient had been regularly taking immunosuppressants for 5 years, which is a predisposing systemic factor for pathogenic Candida. Polyene and azole are effective treatments for Candida albicans infection. Studies have found that fluconazole is superior to other drugs, showing a statistically significant improvement in its therapeutic effect for oral candidiasis. 12,13 Candida albicans prefers an acidic environment. Using 2% sodium bicarbonate solution to maintain an alkaline environment within the oral cavity can inhibit the growth of Candida albicans. Inhibiting the growth of Candida albicans is of great significance for the treatment and inhibition of malignant transformation in OLK. EBV is a human herpes virus and an opportunistic pathogen in immunocompromised patients. 14 EBV is associated with a large number of infectious diseases such as Hodgkin's lymphoma, Burkitt's lymphoma, and nasopharyngeal carcinoma. 15 Moreover, EBV infection is an important risk factor for malignant transformation in OLK. 16 Accordingly, anti-viral therapy is of great importance. Acyclovir is a commonly used drug to inhibit EBV, which is used as an oral treatment in this case. Besides, EBV infection requires regular follow-up. Rothia dentocariosa, Prevotella denticola, and Actinomyces gerencseriae were also identified at the same time. Rothia dentocariosa is a part of the normal flora of the human oral cavity and pharynx. 17 Prevotella denticola and Actinomyces gerencseriae are often regarded as principal causes of periodontitis. [18][19][20] In conclusion, we applied mNGS to pathogen detection and identification for a patient with complex oral mucosal infection. The results revealed that mNGS had higher efficiency, sensitivity and species richness compared with traditional detection methods. Thus, mNGS could significantly improve the treatment of oral mucosal infectious diseases. analysis and manuscript preparation. ZF Song and SF Liu contributed to the conception of the study and helped perform data analysis with constructive discussions. YH Song and FJ Song contributed equally to this work. Data availability All data are available from the corresponding author by request. Declaration of conflicting interests The authors declare no competing interests. Ethics statement and informed consent The patient provided informed consent for treatment at the initial visit. The patient was well informed of all purposes that might be used for research, and consent for publish was signed by the patient.	2022-11-19T06:16:19.198Z	2022-11-01T00:00:00.000	{ "year": 2022, "sha1": "a2991624f7a4deac3aa9d80b2ecf4f7976777310", "oa_license": "CCBYNC", "oa_url": null, "oa_status": null, "pdf_src": "PubMedCentral", "pdf_hash": "02f8a6d93eb8e8b44b8f1f57ff345bf4ebce6ce9", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
250998668	pes2o/s2orc	v3-fos-license	Clinical and Virological Response to Convalescent Plasma in a Chronic Lymphocytic Leukemia Patient with COVID-19 Pneumonia The burden of COVID-19 remains unchanged for immunocompromised patients who do not respond to vaccines. Unfortunately, Omicron sublineages are resistant to monoclonal antibodies authorized in Europe so far, and small chemical antivirals have contraindications and toxicities that have not been studied in these patients. We report here the successful treatment of COVID-19 pneumonia lasting for 4 months after the transfusion of COVID-19 convalescent plasma (CCP) in a patient with severe immunosuppression due to both chronic lymphocytic leukemia and venetoclax treatment. The patient achieved a complete clinical, radiological and virological response after six transfusions (600 mL each) of high-titer CCP collected from triple-vaccinated and convalescent donors. This dramatic case adds to the mounting evidence of CCP efficacy in immunocompromised patients, provided that high-titer and large volumes are infused. The COVID-19 pandemic has caused 554 million cases and 6.35 million deaths worldwide, as of 8 July 2022. While most immunocompetent subjects who have been triplevaccinated do not progress to severe disease after SARS-CoV-2 infection, immunocompromised patients, and especially the ones with B-cell depletion, do not respond to vaccines, and remain hence at increased risk of complications [1]. Such frail patients often have contraindications to or cannot tolerate the toxicity of small chemical antivirals, for which safety and efficacy issues still persist. Additionally, the advent of the Omicron sublineages of SARS-CoV-2 has caused a loss of efficacy for all therapeutic anti-SARS-CoV-2 Spike monoclonal antibodies (mAbs) authorized so far in Europe [2,3]. In this scenario, there is an urgent need for novel therapeutic agents. COVID-19 convalescent plasma (CCP) has been initially investigated as a treatment for COVID-19 inpatients [4], but findings from randomized controlled trials (RCT) have consistently shown definite efficacy only in outpatients treated within 5 days from onset of symptoms [5][6][7]. This is in line with indications for the usage of any other antiviral, including small-chemical and anti-Spike mAbs. On the basis of encouraging case reports and large case series [8,9], the FDA re-authorized CCP for immunosuppressed patients in January 2022 [10]. We report here the case of a 62-year-old male patient with chronic lymphocytic leukemia. He had been treated with rituximab plus bendamustine in 2015 and in 2019, with partial responses. In January 2021, the patient experienced a second relapse (showing de novo mutation of TP53 and 13q14.2 microdeletion) and received chlorambucil debulking Life 2022, 12, 1098 2 of 4 before starting rituximab plus venetoclax (400 mg) in March 2021. The last (third) rituximab infusion was in September 2021. In October 2021 the patient received his third dose of BNT162b2, but anti-Spike antibodies remained undetectable a few months later. On December 31, he tested positive for SARS-CoV-2 RNA via a nasopharyngeal swab (NPS). He was not admitted at hospital, was left at home under low-molecular-weight heparin, and received no small-chemical antiviral or anti-Spike mAb; while at home, despite azithromycin and ciprofloxacin courses, he became febrile and required low-flow oxygen support with nasal cannulas every day for 3 months. On 20 April 2022, a chest CT scan performed at the emergency room in Portoferraio showed diffuse interstitial pneumonia, and type 1 respiratory failure was seen in his arterial blood gases (ABG: pO2 64 mmHg, pCO2 33 mmHg, pH 7.6). Venetoclax was discontinued on April 24. The viral load on the NPS was very high (Ct = 19) and SARS-CoV-2 was also found in the peripheral blood (Ct = 38). Whole-genome sequencing of the virus showed that the strain belonged to BA.1. While such lineage was still sensitive to sotrovimab, the drug was not available at local hospitals at that time, and there was mounting evidence of treatment-emergent immune escape to sotrovimab in immunocompromised patients [11,12]. Small-chemical antivirals such as molnupiravir and nirmatrelvir/ritonavir were not offered because of poor efficacy [13] and contraindications, respectively; none of the pills had been investigated in immunosuppressed patients or for exposures longer than 5 days, which was likely needed in our patient because of the high initial viral load. The ethical committee of North-Western Tuscany granted permission to use CCP for this patient on April 24. The patient received six CCP units: the first (600 mL; >5500 BAU WHO/mL) on April 25, the second on May 1 (two 300 mL aliquots from the same donation; anti-authentic live Omicron-neutralizing antibody titer 1:640), the third on May 8 (600 mL; 3296 BAU WHO/mL), the fourth on May 23 (600 mL; >5500 BAU/mL), the fifth on May 30 (1279 BAU WHO/mL), and the sixth on June 4 (600 mL; >5500 BAU/mL). The first and fourth units were from the same donor. All donors were triple-vaccinated and convalescent from different SARS-CoV-2 VOCs. A few hours after the first CCP transfusion, the patient was apyretic (despite the suspension of corticosteroids and paracetamol), and oxygen supplementation was discontinued from April 29. A follow-up chest CT scan on May 9 showed the resolution of pneumonia, despite the persistence of focal fibrosis (Figure 1). Table 1 shows the progressive decrease in viral load and inflammatory serum markers with CCP treatment. Immunophenotyping by flow cytometry on May 22 confirmed a 0% presence of CD19+ B-lymphocytes. The sudden changes could not be explained by immune reconstitution after the discontinuation of venetoclax, and especially the sudden drop in viral load can only be attributed to neutralizing antibodies within the transfused CCP units. In a recent RCT on 133 immunosuppressed patients, two 300 mL plasma units (collected from either convalescent or double-vaccinated donors) with nAb titers > 1:80 led to a clinically significant benefit in patients with hematological malignancies, other cancers, or immunosuppression (group 1, group 2, n = 71), with a median time to improvement of 10 days for vaccinated plasma, 13 days for CCP, and 32 days for the control [14]. Mounting evidence has shown the superiority of hybrid plasma (i.e., CCP collected from donors who are both convalescent and at least double-vaccinated [15]), and the relevance of dose in CCP treatment [16]. Our patients received six units of plasma collected from donors who had been both infected and triple-vaccinated, and accordingly showed very high anti-Spike antibody levels. This case primed the startup of a dedicated path within the hospital network of the ASL Toscana Nord-Ovest, starting from CCP collection and resulting in stocking for transfusion within dedicated ambulatories [17]. In conclusion, we have shown here that CCP, a treatment with previously proven benefit in immunocompetent patients when given in the first 5 days from the onset of symptoms (when the patients are more likely to be seronegative), can be a life-saving treatment in immunocompromised patients, especially ones with B-cell depletion. In immunocompromised patients, benefit from CCP is expected at far later timepoints than in immunocompetent patients, likely until they become seronegative, in line with the principles of replacement therapy. While there are no standardized doses and schedules for In a recent RCT on 133 immunosuppressed patients, two 300 mL plasma units (collected from either convalescent or double-vaccinated donors) with nAb titers > 1:80 led to a clinically significant benefit in patients with hematological malignancies, other cancers, or immunosuppression (group 1, group 2, n = 71), with a median time to improvement of 10 days for vaccinated plasma, 13 days for CCP, and 32 days for the control [14]. Mounting evidence has shown the superiority of hybrid plasma (i.e., CCP collected from donors who are both convalescent and at least double-vaccinated [15]), and the relevance of dose in CCP treatment [16]. Our patients received six units of plasma collected from donors who had been both infected and triple-vaccinated, and accordingly showed very high anti-Spike antibody levels. This case primed the startup of a dedicated path within the hospital network of the ASL Toscana Nord-Ovest, starting from CCP collection and resulting in stocking for transfusion within dedicated ambulatories [17]. In conclusion, we have shown here that CCP, a treatment with previously proven benefit in immunocompetent patients when given in the first 5 days from the onset of symptoms (when the patients are more likely to be seronegative), can be a life-saving treatment in immunocompromised patients, especially ones with B-cell depletion. In immunocompromised patients, benefit from CCP is expected at far later timepoints than in immunocompetent patients, likely until they become seronegative, in line with the principles of replacement therapy. While there are no standardized doses and schedules for CCP yet [16], our case suggests that the duration of treatment should be driven by virological response (Ct in PCR). Institutional Review Board Statement: The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of Area Vasta Nord-Ovest (protocol). Informed Consent Statement: Written informed consent has been obtained from the patient(s) to publish this paper. Data Availability Statement: Data are available via corresponding author.	2022-07-24T15:14:12.903Z	2022-07-01T00:00:00.000	{ "year": 2022, "sha1": "fc9b5783e7da5b3a6aba69652f0e5ab95a98fa85", "oa_license": "CCBY", "oa_url": null, "oa_status": null, "pdf_src": "PubMedCentral", "pdf_hash": "70f4051ce72d5f85364c8df9eb028ed4952d2517", "s2fieldsofstudy": [ "Medicine", "Biology" ], "extfieldsofstudy": [ "Medicine" ] }
229421245	pes2o/s2orc	v3-fos-license	Physico-mechanical Properties of Low-density GeopolymerMortarSynthesized Using Inexpensive Foam Agent Low-density geopolymer(LDG) mortars have been synthesized through the use of an inexpensive car-wash foaming agent. LDG mortarswere made by mixing coal fly ash and sand with activator solution of NaOH and Na2SiO3. The slurry mixture was then mixed with a thick foam and stirred using a hand mixer. During the stirring process, the ordinary Portland cement (OPC)was added to the mixture as additive. Once homogeneous, the slurry mixture was then poured into mortar molds and left overnight at room temperature before heat treatment at 60 °C for 24 h. The geopolymer mortars formed were then characterized. The density measurement showed that the density of geopolymerdecreased to about half from the original density at the OPC compositions of 5-15% and the activator solution compositions of 20-26%. In this composition range, the highest compressive strength (2.15 MPa) was shown by the LDG synthesized using a 23% activator solution and 15% OPC. Further addition of activator solution decreased the compressive strength due to the increase of porosity. The thermal durability test indicated that the LDG structure was relatively stable up to a temperature of 300 C. A further increase in temperature would cause structural degradation. Introduction Research on low-density geopolymers(LDG) or lightweight geopolymers has attractedthe attention of researchers [1,2]. It is driven by various needs, including the desire to expand its applications, such as acoustic panels, thermal insulation, building wall materials, etc.Alvarezayusoetal. and Chengetal. in [2] reported that the low-density geopolymers had several characteristics including excellent mechanical properties, good thermal resistance, good adsorption ability of toxic metals, as well as high acid resistance and fire resistance. Apart from these advantages, the raw materials are easily obtained and cheap from an electric power plant fly ash waste or other agro-industrial fly ash. 1 amun.amri@eng.unri.ac.id The choice of manufacturing method will affect the quality of low-density geopolymer produced. There are two techniques in preparing of foam mortar, namely pre-foaming and mix-foaming techniques. In the pre-foaming technique, the foam is made separately from the mortar paste. The next step is mixing the paste and foam; meanwhile, the mix-foaming technique, from the beginning, the foam and pasta are mixed. Ducman and Korat[1] conducted a study on the manufacture of fly ashbased low-density geopolymers by comparing the addition of aluminum powder and hydrogen peroxide (H 2 O 2 ) foaming agents. The results of this study indicated that the hydrogen peroxide (H 2 O 2 ) had a better effect than aluminum powder, and the porosity test showed that the porosity of aluminum powder washigher than the hydrogen peroxide (H 2 O 2 ) by 59% and 48%, respectively. Wu et al., [2] have created fly ash -metakaolin based ultra-lightweight foamed geopolymer (UFG) with the addition of H 2 O 2 foaming agent. The results of this study indicated that the lightweight geopolymer showed better mechanical properties than the portland cement-based foam concrete at the same density. However, materials such as aluminum powder and hydrogen peroxide are quite expensive. Besides, hydrogen peroxide (H 2 O 2 ) is quite dangerous for health if inhaled and exposed to the body and environment. This work tries to create a low-density geopolymervia a pre-foaming technique using a car-wash foam agent that is cheap and safe for the environment. The effects of an increasing amount of activator solution and OPC additiveused were studied. The results showed that the density of the geopolymer decreased to about half of its original density at 5-15% of OPCand 20-26% activator solution compositions. The compositionsare acrucial factor in obtaining optimal density and compressive strength values Geopolymers preparation The activator solution was made by mixing 10 M NaOHand Na 2 SiO 3 solutions with a ratio of Na 2 SiO 3 : NaOHof 2.5.The car-wash foam was prepared through the pre-foaming technique by mixing the car-wash foaming agent with water in a ratio of 1:30 and then fed into foam generator equipment. Solid fly ash and sand were mixed with activator solution with a solid to liquid ratio of around 4:1. This mixture was then poured into a container that already contained thick car-wash foam. Stirring was done quickly using a hand mixer while adding the OPC little by little to the dough until homogeneous. This dough was then poured into molds (10x10x10 cm 3 ) and left overnight. The geopolymer mortar formed was then heated in an oven at 60 C for 24 hours and then cooled. The amount of activator solution and the amount of OPC used were varied and repeated three times (by triplo) Characterization The density of the formed geopolymer wasdetermined by dividing the dry mass by volume. The compressive strength of geopolymers was tested using a compressive testing machine following ASTM C579-01 (2012). The thermal durability test was performed by monitoring of compressive strength of geopolymer after heating to high temperatures at 300 ºC, 500 ºC, and 700 ºC in a furnace for 1 hour 30 minutes. Table 1and Figure 1 show the densities of geopolymers produced by various percentages of activator solution used and OPC added. It can be seen thatthe geopolymers have a reasonably low density compared to the control sample by about half. The decrease of density is due to the high porosity in the geopolymers matrix created by the foam integrated into the matrix at the molding process. The formation of pores initiated by the foam is then maintained by OPC, which has a fast setting time of hardening. This pores formation is difficult to be reached without the addition of OPC due to the slow geopolymer setting time. Density of Geopolymer From Figure 1, it can also be seen that the density of geopolymers increases as the increase of OPC content in all variations of activator solution addition. It is understandable because OPC is the adhesive between the components in the matrix and, at the same time, solidifies them. Likewise, the densities of geopolymers also increase as the increase of activator solution up to 23%. It is due to the rise of NaOH and Na 2 SiO 3 activator solution can disturb the stability of the foam bubbles, so they are easily broken into the liquid, which leads to a decrease in the number of pores and ends up with a higher density. However, when the activator solution increases to 26%, a strange phenomenon is observed, in which the density again decreases. So far, this is not fully understood. However, this is likely due to the high water content when the percentage of activator solution increases. In a specific composition, this water will counteract or offset the negative influence of NaOH and Na 2 SiO 3 on the foam, so that it will lead to foam stability. Therefore, the amount of foam, activator solution and OPC used must be precise to obtain optimal strength of lightweight geopolymers. understood that the more cement is added, the more compressive strength of the material will be[3].The highest compressive strength (2.15 MPa) is shown by LDG synthesized by OPC addition of 15% and activator solution of 23%. Linear to the densities data as discussed in the previous section, the compressive strength of LDG synthesized using activator solution of 26 % is lower compared to the LDG synthesized using activator solution of 20% and 23%. It is, of course, because their densities are the lowest than the others. According to its nature and without being associated with the addition of foam or formation of pores, the addition of too much activator solution that passes through a certain composition can decrease the compressive strength of LDG [4]. stated that the addition of excess activator solution could cause a decrease in the mechanical properties of geopolymers.It is similar to an increase in the ratio of water to cement. Special for LDG formation, the results of this study are also consistent with the work by Wu et al. [2]. They said that the addition of excess activator solution to the matrix will make the LDG tend to flow easily and thin, causing foam bubbles to move freely to form large cavities. These cavities can cause a decrease in mechanical properties of geopolymers, and when the curing process is applied, there is a shrinkage of LDG, resulting in cracks in the matrix, which will reduce the compressive strength of LDG. Based on the results, it can be concluded that the addition of OPC and activator solution at a certain maximum amount can increase the compressive strength of LDG. Thermal durability To obtain a low-density geopolymer (LDG) that can be applied for various practical applications, besides testing the compressive strength at room temperature, it is also necessary to know the effect of heat treatment on the geopolymers compressive strength (thermal durability). Figure 3 shows the compressive strength of LDG,which is heat-treated for 1.5 hours at different temperatures.Geopolymers tested are geopolymers with a composition of 15% OPC and 23% activator solution, which gave the best compressive strength values, as discussed in Section 3.2. From the figure, it can be seen that the structure of the low-density geopolymeris relatively stable up to a temperature of 300 ºC, which is indicated by a slight degradation in compressive strength of 11.62%. When heating exceeds a temperature of 300 ºC, the low-density geopolymer begins to experience a significant degradation of strength, with a decrease in compressive strength of 35% at 500 ºC and 72% at 700 ºC. This decrease in compressive strength is due to the dehydroxylation reaction of the geopolymer. Where when applying high heat, the bonds to the geopolymer will be broken. A visual inspection of the heat treatment results shows that the low-density geopolymer cracked only after heating above 500 °C. It is consistent with what was reported by Syamsidar et al. in [6], which said that the geopolymer has excellent heat resistance up to a temperature of 600 C. They also noted that the use of fly ash with high SiO 2 content and silica in the activator fluid (Na 2 SiO 3 ) could provide excellent stability in maintaining the geopolymer structure. These results indicate that the lowdensity geopolymer has excellent resistance to high temperatures. Besides, the addition of OPC has a positive effect on durability [3]. Overall, based on the results in this work, it can be concluded that the produced low-density geopolymer can be used for conventional applications such as building wall materials or other low-temperature applications. Conclusions Low-density geopolymer (LDG) mortars have been synthesized by adding car-wash foam agent and adjusting the composition of ordinary Portland cement (OPC) as well as activator solutions of NaOH and Na2SiO3. The density of the LDG decreased to half of the initial density at the OPC composition between 5-15% and 20-26% activator solution. In this composition range, the highest compressive strength (2.15 MPa) was shown by the geopolymer synthesized using a 23% activator solution and 15% OPC additions. Further addition of activator solution decreased the compressive strength due to increased porosity. The thermal durability test indicated that the geopolymer structure was relatively stable up to a temperature of 300 C. A further increase in temperature would cause the structural degradation of the geopolymer. Acknowledgment Thanks to DIKTI for funding this research via KRUPT research grant 2019. Compressive strength (MPa) Temperaure (C)	2020-12-03T09:06:21.182Z	2020-12-02T00:00:00.000	{ "year": 2020, "sha1": "e0ea7e5c6025853193ac6ccf13f843663c19210f", "oa_license": null, "oa_url": "https://doi.org/10.1088/1757-899x/990/1/012018", "oa_status": "GOLD", "pdf_src": "IOP", "pdf_hash": "82148023a4b9f2cb4e73a1966a713437bc4f6932", "s2fieldsofstudy": [ "Materials Science" ], "extfieldsofstudy": [ "Materials Science" ] }
119322264	pes2o/s2orc	v3-fos-license	Rigged configuration bijection and proof of the $X=M$ conjecture for nonexceptional affine types We establish a bijection between rigged configurations and highest weight elements of a tensor product of Kirillov-Reshetikhin crystals for all nonexceptional types. A key idea for the proof is to embed both objects into bigger sets for simply-laced types $A_n^{(1)}$ or $D_n^{(1)}$, whose bijections have already been established. As a consequence we settle the $X=M$ conjecture in full generality for nonexceptional types. Furthermore, the bijection extends to a classical crystal isomorphism and sends the combinatorial $R$-matrix to the identity map on rigged configurations. Introduction Kerov, Kirillov and Reshetikhin [KKR86] introduced rigged configurations as combinatorial objects to parameterize the Bethe vectors for Heisenberg spin chains. Moreover, they constructed a bijection from rigged configurations to highest weight elements of a tensor product of the vector representation of sl 2 . This was generalized to tensor products of the symmetric tensor representations of sl n in [KR86], where the generating function of rigged configurations with the cocharge statistic was shown to be the Kostka polynomials. This was generalized further to tensor products of multiples of fundamental weights (which can be interpreted as rectangles) in [KSS02], where these rigged configurations were connected with Littlewood-Richardson tableaux and generalized Kostka polynomials [KS02,Shi01a,Shi01b,Shi02,SW99,SW00]. In order to generalize this beyond type A, the notion of Kirillov-Reshetihkin (KR) crystals is needed. Let g be an affine Kac-Moody Lie algebra and U ′ q (g) be the quantum group of g ′ := [g, g]. A KR crystal is the crystal basis of a Kirillov-Reshetikhin module: a certain finite-dimensional U ′ q (g)-module that is the minimal affinization of a multiple of a fundamental weight [Cha95,CP95a,CP95b,CP96a,CP96b,CP98]. KR crystals were shown to exist for nonexceptional types in [OS08] and their U ′ q (g)-crystal structure was given in [FOS09]. A path is a classically highest weight element in the tensor product of KR crystals, where for type A (1) n , this agrees with the notion above. Paths also arise from calculations of 2D integrable lattice models using Baxter's corner transfer matrix method [Bax89]. This method leads to the quantity X, which is the sum over the intrinsic energy statistic of paths of B, a tensor product of KR crystals. Let M denote the sum over the cocharge statistic of rigged configurations of a fixed multiplicity matrix L, which is called the fermionic formula. The X = M conjecture of [HKO + 99, HKO + 02b] states that X is equal to M when L counts the factors in B, which suggests the existence of a bijection between paths and rigged configurations that sends the intrinsic energy to cocharge. As previously mentioned, the general case of the desired bijection was proven to be a bijection in type A (1) n [KSS02], building upon [KKR86,KR86]. Furthermore, for type A (1) n , the bijection was extended as a classical crystal isomorphism in [DS06] using the crystal structure of [Sch06a] and to a full U ′ q (g)-crystal isomorphism in [SW10]. For type D (1) n , an analogous bijection was proven in the general case in [OSSS17], building upon the special cases of [OSS13,Sch05]; the bijection was shown to intertwine with the classical crystal structure in [Sak14]. For other types, the bijection is known in a number of special cases [OSS03a, OSS03b, OSS03c, OS12, OSS13, SS06, SS15b, Scr16,Scr17]. As far as the X = M conjecture is concerned, there is an alternative proof by Naoi [Nao12] for type A (1) n and D (1) n using representation theory of Kirillov-Reshetikhin modules for a current algebra. KR crystals are also known to be a reservoir of perfect crystals [KKM + 92a, KKM + 92b]. Indeed, the condition for a KR crystal to be perfect of a fixed level ℓ is proven for all nonexceptional types in [FOS10]. In particular, this allows KR crystals to be used in the Kyoto path model of [KKM + 92b], an iterative method to construct highest weight U q (g)-crystals from the KR crystal. If all components of the tensor product of KR crystals B are perfect of level ℓ, then it is known that X, in a suitable limit, turns our to be a branching function for the coset g/g 0 , where g 0 is its underlying finite-dimensional simple Lie algebra of g. Hence, the proof of X = M conjecture implies a fermionic formula of such branching functions (see [HKO + 99,Theorem 5.4] and [HKO + 02b, Proposition 4.5]). The goal of this paper is to give a combinatorial proof of the X = M conjecture for all nonexceptional types. Our main result is the construction of an explicit bijection Φ from paths to rigged configurations for all nonexceptional affine types. Furthermore, we show that Φ sends the combinatorial R-matrix on paths to the identity on rigged configurations and can be extended to a classical crystal isomorphism. The bijection also sends the intrinsic energy to cocharge up to a simple involution θ on rigged configurations which interchanges each rigging with its corigging. The map θ is related to the Lusztig involution. These facts were shown in type A (1) n (resp. D (1) n ) in [KSS02] (resp. [OSSS17]). Our techniques for constructing Φ use virtual crystals [Kas96,OSS03b,OSS03c,SS15b], which are crystals constructed using diagram foldings of a simply-laced type. Explicitly, we construct Φ for nonexceptional type g by lifting to ambient type A or D, doing the type A or D bijection and then retracting back to type g. Then, using the algorithm for the bijection in simply-laced types, we give an explicit algorithm for Φ in all nonexceptional types, where the basic operation δ is given in [OSS03a]. In [FOS09], KR crystals were constructed using Kashiwara-Nakashima (KN) tableaux [KN94]. However, this does not explicitly describe the tableaux in the image of Φ −1 except in type A (1) n . By taking the image of the bijection Φ −1 on a single KR crystal, a new tableau model is obtained, coined Kirillov-Reshetikhin (KR) tableau. KR tableaux have been explicitly described on classically highest weight elements for all nonexceptional types [OSS13,SS15b], type G [Scr16] and some additional special cases [Scr17]. From [OSSS17], the image of Φ −1 in type D (1) n for arbitrary factors is precisely described by a tensor product of KR tableaux, which gives an explicit algorithm for computing Φ. Furthermore, KR tableaux distinguish each classical component. As a consequence of our construction, we have that the image of Φ −1 are tensor products of KR tableaux for all nonexceptional types. We expect our techniques to apply to the exceptional cases as well. To achieve this, the existence of KR crystals for exceptional type g needs to be established first, which is sufficient for extensions to type G and E (2) 6 . In [Scr17], a more conceptual approach was given to describe Φ as relating to the crystal s tructure of the factors added. In [SS16], the map θ was shown to be the star involution on the rigged configuration model for B(∞) given in [SS15a,SS17]. Our results are more evidence that Φ has a natural crystal-theoretic description since Φ respects the virtualization of the KR crystals. Furthermore, our results might have applications to construct extremal level-zero crystals [Kas02] using rigged configurations, parallel to [HN06,LL15,NS03,NS05,NS06,NS08a,NS08b,PS17]. This would allow a direct description of a U q (g)-crystal structure on rigged configuration. This paper is organized as follows. In Section 2, we give the necessary background on crystals and rigged configurations. In Section 3, we construct the bijection Φ and prove our main results. In Section 4, we show some properties of Φ and prove the X = M conjecture for nonexceptional types. was partially supported by NSF grant DMS-1500050. TS was partially supported by NSF RTG grant DMS-1148634 and JSPS grant K15K13429. TS and MO would like to thank the University of California Davis for hospitality during their stay in March, 2017, where the majority of this work took place. TS would like to thank Osaka City University for hospitality during his visit in July, 2017. This work benefitted from computations and experimentations in Sage [SCc08, Sag17]. 2. Background 2.1. Crystals. Let g be an affine Kac-Moody Lie algebra with index set I, Cartan matrix (A ij ) i,j∈I , simple roots (α i ) i∈I , simple coroots (α ∨ i ) i∈I , fundamental weights (Λ i ) i∈I , weight lattice P and canonical pairing ·, · such that α ∨ i , α j = A ij . We follow the labeling of I given in [Kac90]. For our purpose we also need the opposite labeling of I for A (2) 2n denoted by A (2) † 2n . The Dynkin diagrams of all nonexceptional types including A (2) † 2n are given in Table 1. Let U q (g) denote the corresponding quantum group, and let U ′ q (g) = U q ([g, g]) be the quantum group corresponding to the derived subalgebra of g. Let g 0 denote the canonical simple Lie algebra given by the index set I 0 := I \ {0}. Let ̟ i (i ∈ I 0 ) denote the funamental weights of the weight lattice P 0 of type g 0 . Let c i and c ∨ i denote the Kac and dual Kac labels [Kac90, Table Aff1-3]. The null root is given by δ = i∈I c i α i . The canonical central element is given by , for i ∈ I, and weight function wt : B → P . Let ε i , ϕ i : B → Z 0 be statistics given by The following conditions should be satisfied: We say an element b ∈ B is J-highest weight if e i b = 0 for all i ∈ J ⊂ I. Table 1. Dynkin diagrams for all nonexceptional affine types. The labeling of the nodes by elements of I is specified under or to the right of the nodes. For further details regarding crystals, see [BS17,HK02]. In particular note that in this paper we use the convention for tensor products of crystals as in [BS17], which is opposite to the convention used by Kashiwara [Kas91]. Let B 1 and B 2 be two U q (g)-crystals. A crystal morphism ψ : B 1 → B 2 is a map B 1 ⊔ {0} → B 2 ⊔ {0} with ψ(0) = 0 such that the following properties hold for all b ∈ B 1 : An embedding (resp. isomorphism) is a crystal morphism such that the induced map B 1 ⊔ {0} → B 2 ⊔ {0} is an embedding (resp. bijection). A crystal morphism is strict if it commutes with all crystal operators. (where r ∈ I 0 and s ∈ Z >0 ) have a multiplicity free decomposition as U q (g 0 )crystals: for certain (distinct) λ ∈ P + 0 . Here B(λ) is the highest weight U q (g 0 )-crystal of highest weight λ ∈ P + 0 . An explicit combinatorial construction of B r,s for all nonexceptional types was given in [FOS09] 2n by relabeling the nodes i ↔ n − i, but we need to be careful about the weight. In particular, we have g 0 is type B n for type A (2) † 2n , and so we need twice ̟ n of type B n . Let κ r = 1 unless g = A (2) † 2n and r = n, in which case κ n = 2. We note that there is a unique element u sκr̟r ∈ B(sκ r ̟ r ) ⊂ B r,s of weight sκ r ̟ r , called the maximal element . Furthermore, it is known that tensor products of KR crystals N i=1 B ri,si are connected [FSS07,Oka13] with a unique maximal element u s1κr 1 ̟r 1 ⊗ · · · ⊗ u sN κr 1 ̟r N . Therefore, there exists a unique U ′ q (g)-crystal isomorphism R : B ⊗ B ′ → B ′ ⊗ B called the combinatorial R-matrix defined by R(u⊗u ′ ) = u ′ ⊗u, where u and u ′ are the maximal elements of B and B ′ respectively. 2.3. Dualities. Denote by τ : where w 0 is the longest element of the Weyl group of g 0 . Explicitly, we have (g 0 = D n for n odd), τ (i) = i (g 0 = D n for n even, B n , C n ). Define the Lusztig involution ⋆ : B(λ) → B(λ) as the unique involution satisfying We note that the Lusztig involution sends highest weight elements to a lowest weight element. It can be extended to ⋆ : B r,s → B r,s by defining τ (0) = 0 and requiring that ⋆ satisfies (2.1). Let B ∨ denote the contragredient dual crystal of B. As a set B ∨ = {b ∨ \| b ∈ B} with the crystal structure given by For the highest weight U q (g 0 )-crystal B(λ), we note that B(λ) ∨ is naturally isomorphic to B(−w 0 λ). We can also extend the Lusztig involution and the contragredient dual to tensor products by a natural isomorphism n , we consider the diagram automorphism σ given by This induces a twisted crystal isomorphism from B r,s to B σ(r),s given by By abuse of notation, we denote this twisted crystal isomorphism by σ. We can extend σ to tensor products by a natural isomorphism n for nonspin columns, we have σ(b) by interchanging the letters n ↔ n, which follows from considering the map on the highest weight U q (g 0 )-crystal B(Λ 1 ). For the spin columns in type D 2.4. Virtual crystals. Virtual crystals were introduced in [OSS03b, OSS03c] as a way to realize crystals for nonsimply-laced types as embeddings into simplylaced types. At the time these papers were written, the existence of some of the KR crystals had not yet been established, explaining the choice of name "virtual crystals." In the meantime, the existence of all KR crystals of nonexceptional types was established in [Oka07,OS08] and explicit combinatorial realizations were constructed in [FOS09]. Even though this means that virtual crystals are now true realizations of crystals, we will stick with the terminology virtual crystals for historical reasons. Let I X denote the index set of ambient type X (i.e., g is of type A ). We denote the corresponding map on the index sets by φ : I X ց I as in Figure 1 (resp. Figure 2 ). For ease of notation, if Z is an object for type g, we denote the corresponding object for type g (type A Note that if \|φ −1 (a)\| = 1, then γ a = 1. Furthermore, we have a natural embedding Ψ : P → P X given by where the map on simple roots is induced from the embedding of the root lattice into the weight lattice. Note that this implies that δ X = c 0 γ 0 δ. Let φ and (γ a ) a∈I be the folding and the scaling factors given above. The virtual crystal operators (of type g) are defined as Consider a set B ⊂ B X with a fixed φ and (γ a ) a∈I . If for all v ∈ B and a ∈ I, we have , then we say that B X is aligned . Note that in particular Definition 2.1 requires that virtual crystals are aligned. When there is no danger of confusion, we simply denote the virtual crystal by B. We say that a type g crystals B is realized as a virtual crystal B if there exists a U ′ q (g)-crystal isomorphism χ : B → B. We denote the composition of χ with the natural inclusion B ⊂ B X by emb : B → B X , which we call the virtualization map. Furthermore, we will also denote λ = Ψ(λ). It is straightforward to see that virtual crystals are closed under direct sums. Moreover, they are closed under tensor products. Next, we provide explicit virtual crystal realizations for all nonexceptional types. We distinguish the cases that embed into type A (1) 2n−1 (denoted ambient type A) and those that embed into type D We are using the embeddings of types C It was shown in [Oka13, Thm. 5.1] that these crystals are aligned, proving [OSS03b, Conj. 6.6]. Hence they give realizations for the corresponding KR crystals. For types B (1) n and A (2) 2n−1 , we use the realizations n . The realization of the first two lines are given in [SS15b, Thm. 5.14]. The last realization is given in A rigged configuration (ν, J) ∈ RC(L) is a sequence of partitions (ν (a) ) a∈I0 , where to each row of ν (a) we associate an integer 1 x, called rigging. The pair (i, x), where i is the length of a row and x is the associated rigging, is called a string. Let 2n , A (2) † 2n respectively. If g is of simply-laced type, we set γ a = 1 for all a ∈ I. Here m Remark 2.3. The rigged configurations that we give here differ slightly from those given in [OSS03a]. In particular, for type B (1) n (resp. type C (1) n ), our rigged configurations use full-width boxes instead of half-width boxes (resp. double width boxes) for ν (n) . To go to the rigged configurations of [OSS03a], simply half (resp. double) the partition ν (n) for type B Note that for i ≫ 1, we have p where k 0 is such that c, wt(ν, J) = 0. There exists an extension of rigged configurations to a U q (g 0 )-crystal that was given in [Sch06a,SS15b]. The complement rigging involution θ : RC(L) → RC(L) is defined by replacing every rigging x ∈ J Moreover, this completely characterizes emb RC(L) . Also note that n , and let B be a tensor product of KR crystals. The set of paths corresponding to B consists of all I 0 -highest weight elements in B: When also restricting to a given weight space, we write where wt 0 is the g 0 -weight. For type A n , a bijection Φ : P(B) → RC(L) was established in [KSS02], generalizing ideas in [KKR86,KR86]. Moreover, the paths P(B) are described using the usual semistandard tableaux whose entries are in {1, 2, . . . , n + 1}. For type D (1) n , a bijection Φ : P(B) → RC(L) was established in [Sch05,OSS13,OSSS17]. In this case, the KR crystals are described in terms of Kirillov-Reshetikhin (KR) tableaux [Sch05,OSS13] rather than Kashiwara-Nakashima tableaux [KN94] that were previously used to describe B r,s [FOS09]. In the KR tableau formulation, the elements in B r,s are represented by rectangular tableaux of width s and height r with entries in B 1,1 . For both types A n and D (1) n , the bijection Φ : P(B) → RC(L) is defined recursively. Let B • be a tensor product of KR crystals. On the path side, the composition of the following maps is used: until the empty path is reached. The analogous maps on the rigged configuration side are denoted by: i < j. Consider step j. Let ℓ j denote a minimal i a ℓ j−1 for all a ∈ I 0 such that f a b j−1 = 0 and (ν, J) (a) has a singular string of length i a that has not been previously selected. If no such string exists, terminate, set all ℓ j ′ = ∞ for j ′ j and return b j−1 . Otherwise set b j = f a b j−1 and ℓ j = i a and repeat. We form a new rigged configuration by removing a box from each string selected by the algorithm above, making the resulting rows singular and keeping all other strings the same. We note that δ was first described for type A n in [KR86,KKR86] and for type D (1) n in [OSS03a]. The map δ for the other nonexceptional types was described in [OSS03a]. The map δ sp as given here was first described in [OS12,Scr17], where it was shown to be equivalent to the definition given in [Sch05]. The map β is given for all nonexceptional types by adding a length 1 singular string to (ν, J) (a) for all a < r. The map γ is the identity map, but we note that some vacancy numbers will change. In the sequel, we will need to indicate the left factor when applying γ, and so we denote this by γ r,s : We also have right analogs of the maps given above. Let ↑(b) denote the map which takes an element b ∈ B(λ) ⊂ B to the corresponding I 0 -highest weight element u λ ∈ B(λ). We define We need an additional operation lb (s) on P(B) and β (s) on RC(L) for type D n . For 1 < r n − 2, the map lb (s) : For other elements b, the map lb (s) is determined by assuming that lb (s) commutes with f i for i ∈ I 0 . For a rigged configuration (ν, J) ∈ RC L(B r,s ⊗ B • ) , the operation β (s) adds a singular string of length s to ν (a) for all 1 a < r. Since p (a) i 0, this implies all quantities in (2.11) must be zero. For b > 1, we s−1 = 0, and so any row of length s − 1 must be cosingular since all riggings are nonnegative. Therefore, we must have m from (2.10) at i = s − 1. Hence, we must have k (b−1) = ℓ (b−1) s − 2. Therefore, by iterating the previous argument, we can show that m Let emb 2× be the doubling map: the virtualization (or similarity) map given by g = g and γ a = 2 for all a ∈ I [Kas96]. Recall from [Sch05,Scr17] where β [r] (resp. β) adds a singular row of length 1 to ν (a) for all a ∈ I 0 \ {r} (resp. a n − 2). Define β : Furthermore, the application of γ makes all strings of length i < s nonsingular in ν (1) , but keeps all other strings singular. Hence δ starts by selecting the singular string of length s added by β (s) and continues selecting them until it reaches ν (r) , where we have two main cases. We first consider the case N < s. In this case δ terminates and returns r, and similarly for each subsequent application of δ until the leftmost factor is B 1,N . In this case, we continue selecting rows until we return again to the r-th partition, at which point there is a second singular string of length N − 1 we can select since k N > 0. We continue selecting singular strings until we reach the (r − k N )th partition, at which point there are no other singular strings and we return r − k N + 1. The next application of δ is similar to the previous application except it terminates at the (r − k N −1 + 1)-th partition and returns r − k N −1 + 1. By iterating this argument, the claim follows. Since the remaining factor is multiplicity free and determined by the weight, the claim follows. Proposition 2.6. For types A (1) n and D (1) n , the diagram commutes. Proof. For type A (1) n , this was shown in [OSS03b, Thm. 5.7]. Thus, let g be of type D (1) n . Recall that σ interchanges n ↔ n. It is straightforward to see that σ intertwines with lh, lb and ls. Because β and γ affect neither ν (n−1) nor ν (n) (since γ is the identity map and β only applies to ν (a) for a < n − 1), the map ς intertwines with β and γ. It remains to show that ς intertwines with δ and δ sp . We first consider the case when δ does not terminate with n or n. In this case, the choices of ν (n−1) and ν (n) are done independently, and so the same rows are selected in ς(ν, J) := (ν ς , J ς ). If δ terminates at n or n, then the algorithm selected a row in ν (n−1) or ν (n) , respectively, but not both. Therefore, we select the same row in ν (n) ς or ν (n−1) ς , respectively, and terminate with n or n, respectively. Hence, the map ς intertwines with δ. For δ sp , we note that σ : B n−1,1 ↔ B n,1 . From the description of δ sp , every f a arrow that would normally be applied when doing δ sp , is replaced by an f σ(a) arrow. However, we have (ν ς , J ς ) (a) = (ν, J) (σ(a)) and the claim follows. The bijection In this section, we show that there exists a bijection Φ from paths to rigged configurations for any nonexceptional type g. We construct Φ as a generalization of the map Φ : P(B) → RC(L) of types A (1) n and D (1) n of Section 2.6 to nonexceptional types using generalizations of the maps (2.9) (the maps for paths remain the same). In order to do this, we will define Φ by using the map Φ X in ambient type X and lifted versions of the maps on paths and rigged configurations. Afterwards, we will show that Φ can be computed by using generalizations of (2.9). We begin by defining the generalizations of the rigged configuration operations (2.9). The map δ in (2.9) is defined in [OSS03a] and β and γ are the same maps as in (2.9). For type D n , we define δ sp and lh sp by (3.1) except emb 2× is the virtualization map with g of type A (2) 2n−1 and γ a = 1 + δ an for all a ∈ I. The KR tableaux used for P(B) are defined in [SS15b] and precisely characterize the image of Φ under this bijection. We also need the fact where B is a tensor product of KR crystals, which is a consequence of Proposition 2.2 and emb P(B r,s ) = P emb(B r,s ) . In this section, we continue with the notation of Section 2.4, where f X denotes a map f in ambient type X. 3.1. Lifted operations on paths. We define the operations lh, lh sp , lb and ls on paths for the ambient types. 3.1.1. For ambient type A. Let B • be a tensor product of KR crystals of type A (1) 2n−1 . We define lh as a composition of the following maps (3.2) For D (2) n+1 , the operation lh sp is defined by Next, we define lb for 1 < r < n in types C (1) n and D (2) n+1 and 1 < r n in types A (2) 2n and A (2) † 2n by as (3.4a) Recall that in type D (2) n+1 , we use lh sp in the case r = n, so lb is not needed in this case. For r = n in type C (1) n , we define lb by (3.4b) Finally for s > 1, we define ls for 1 r < n in types C (1) n and D (2) n+1 and 1 r n in types A (2) 2n and A (2) † 2n by (3.5a) For r = n in type D (2) n+1 , we define (3.5b) ls := ls A . For r = n in type C (1) n , we define ls by (3.5c) 3. For r = n in type B (1) n , we use lh sp , so we do not need this case. For r = n in type A n tableaux rules of [FOS09]; the usual lb D map is then applied to the height n tableaux. Finally, for type B n , we define ls for 1 r < n (resp. for r = n) by the same composition of maps as (3.5c) (resp. (3.5a)) with ls A and R A replaced by ls D and R D , respectively. For type A (2) 2n−1 , define ls := ls D for 1 r < n. For r = n in type A (2) 2n−1 , we define ls as the same sequence of maps as (3.5a) with ls A and R A replaced by ls D and R D , respectively. 3.2. Lifted operations on rigged configurations. We define the operations δ, δ sp , β and γ on rigged configurations for the ambient types. 3.2.1. For ambient type A. Here we restrict ourselves to ambient type A. First, for a rigged configuration with multiplicity array L( B 1,1 ⊗ B • ), we define δ by On a rigged configuration with multiplicity array L( B n,1 ⊗ B • ) for type D (2) n+1 , we define δ sp by Next we define β on a rigged configuration with multiplicity array L( B r,1 ⊗ B • ) for 1 < r < n in type D (2) n+1 and 1 < r n for all other types as Finally, we define γ on a rigged configuration with multiplicity array L( B r,s ⊗B • ) with s > 1 as follows: if r = n for type C n , γ A 2n−r,s • γ A r,s otherwise. 3.2.2. For ambient type D. Now we restrict ourselves to ambient type D. On a rigged configuration with multiplicity array L( B 1,1 ⊗ B • ), we define δ by 2n−1 . For type B (1) n , we define δ sp := δ D sp • δ D sp , where the first δ D sp is for B n,1 and the second is for B n+1,1 . Note that for r = n in type A if r = n. Preparatory statements. In order to prove our main theorem, we begin by proving some facts about the maps given in Sections 3.1 and 3.2. Proof. We show that δ is well-defined and that δ • emb = emb •δ. We explain the most fundamental case g = D (2) n+1 as the rest of the cases are similar. The involution σ is given by Here a ∨ stands for the element of B 2n−1,1 A , which is naturally isomorphic to (B 1,1 A ) ∨ , without letter a. Hence, the embedding emb : By [SS06], the composition ς •δ A •ς is similar to δ A except we start at the rightmost rigged partition (ν, J) (2n−1) and proceed to the left. Therefore, doing δ means that we apply the usual box-removing procedure δ A from the leftmost rigged partition moving right first and then perform the same procedure from the rightmost one moving left after. Let us consider 2n⊗(2n) ∨ ∈ B 1,1 A ⊗B 2n−1,1 A as an example and see what happens on the rigged configuration side. The fact that the left factor is 2n means that in the first lap, a box is removed from ν (a) for all a = 1, . . . , 2n− 1. If the length of the string ℓ (2n−1) from which a box is removed is larger than 1, then a box in ν (2n−1) can be removed in the second lap, which contradicts that the second component is (2n) ∨ . Hence, we have ℓ (2n−1) = 1, which forces that ℓ (a) = 1 for all a in the first lap. This case corresponds to (P) in [OSS03a,§4.6]. Let us take n ⊗ n ∨ for next example. In this case, a box is removed from ν (a) for a = 1, . . . , n−1 and stop in the first lap, and then a box is removed from ν (2n−a) for a = 1, . . . , n and stop in the second lap. Since the box in ν (n) removed in the second lap was not removed in the first lap, it should have been quasi-singular before the procedure. This case corresponds to (Q) in [OSS03a,§4.6]. Finally let us take n+1⊗n ∨ (which corresponds to n in type D (2) n+1 ). In this case, a box is removed from ν (a) for a = 1, . . . , n and stop in the first lap, and then a box is removed from ν (2n−a) for a = 1, . . . , n and stop in the second lap. At ν (n) there are two possibilities. The first one is that the string from which a box is removed in the second lap coincides with the one in the first lap. This corresponds to case (S) in [OSS03a,§4.6]. The second one is that the string from which a box is removed in the second lap is strictly smaller than the one in the first lap. Since the former string should have been quasi-singular before the procedure, this corresponds to case (QS) in [OSS03a,§4.6]. Proof. The proof is essentially done in [OSS03a,§4]. We check for each g with ambient type D that the operation δ is consistent with the one given there. For g of type B (1) n , we have B 1,1 = B 1,2 D . As an illustration let us look at the case when nn ∈ B 1,2 D corresponding to 0 ∈ B 1,1 is removed and compare it with the algorithm of [OSS03a, §4.2] as the other cases are similar. Note that all the configurations ν (a) in the ambient rigged configurations are doubled from the ones there for all a < n. The first case of (Q) occurs exactly when a box is removed from a string of length 2ℓ (a) in ν (a) for a = 1, . . . , n−1 and stopped in the first application of δ D and a box is removed from a string of length 2ℓ (a) − 1 for a = 1, . . . , n − 1 and also from a string of length 2ℓ (n−1) − 1 and stopped in the second application. The second case of (Q) is the same as the previous one except that in the second application of δ D the box is removed from a string of length not less than 2ℓ (n−1) . This box was not removed in the first application of δ D , since it was quasi-singular. Proposition 3.4. Let g be of type B (1) n . Consider an element (s 1 , . . . , s n ) ∈ B n,1 given by the ±-vector from [KN94]. Denote the elements in B r,1 , for r = n, n + 1, of type D , where s ± n+1 is such that s 1 · · · s n s ± n+1 = ±1. Then emb is the virtualization map and the image is characterized by Proof. This is straightforward from the definition of the (virtual) crystal operators. The uniqueness comes from the fact that tensor products of KR crystals are generated by the (unique) maximal vector. The characterization of the images comes from the fact that the maximal vector must map to the maximal vector, which has the desired property. Proof. The proof of (1) is given by Lemma 3.1 (resp. Lemma 3.3) for ambient type A (resp. ambient type D). For (1') in type D (2) n+1 , this follows from [Scr17, Lemma 3.9] and the fact that there exists a virtualization map (see also [Scr17,Prop. 8.3]). For (1') in type B (1) n , this is given by [Scr17,Thm. 6.1]. For (2), this is a straightforward computation. For (3), this is the identity map on rigged configurations and a straightforward computation of the changes in vacancy numbers shows that γ is well-defined. Proposition 3.7. The operations lh, lb and ls for ambient types A and D as well as lh sp for types D (2) n+1 and B (1) n have the following properties: Proof. The proofs of (1) and (1') follow immediately from the fact emb is defined component-wise. Next, we show (2) by induction on r, where the base case r = 1 is trivial. Recall that Φ X intertwines lh X , lb X , ls X and R X with δ X , β X , γ X and id X respectively for ambient type A [KSS02] and ambient type D [OSSS17], as well as lb (s)D with β (s)D by Proposition 2.5. Moreover, recall that Φ X intertwines σ and ς by Proposition 2.6. Hence, the bijection Φ X intertwines lb with β. From Proposition 3.5, the map β is well-defined and, thus, so is the map lb. Since lb only affects the leftmost factor(s) and is a composition of strict U q (g 0 )-crystal embeddings, it is sufficient to consider p ∈ emb P(B r,1 ) . We note that emb(B r,1 ) is multiplicity free as a U q (g 0 )-crystal, and therefore (ν X , J X ) = Φ( p) ∈ emb RC(L(B r,1 )) . Next, we note that β(ν X , J X ) ∈ emb RC(L(B 1,1 ⊗ B r−1,1 )) by Proposition 3.5. Therefore, we have lb( p) ∈ emb P(B 1,1 ⊗ B r−1,1 ) by induction. The proof of (3) is similar to (2) by using induction on s. Additionally note that (ls A ) −1 in (3.5c) can indeed be applied because of the previous two applications of ls A and the corresponding rigged configuration is never changed. Proof. Since emb is applied component-wise and B r,s is multiplicity free as a U q (g 0 )crystal, we have lx • emb = emb • lx. Proof. For ξ = δ, the claim was shown in Lemma 3.1 (resp. Lemma 3.3) for ambient type A (resp. ambient type D). For ξ = β, this is a straightforward computation on the change in vacancy numbers and that β and β do not change the vacancy numbers. For ξ = γ, this is a straightforward computation on the vacancy numbers. For ξ = δ sp , the statement for type D 3.4. Main theorem. We now give our main results. We will use diagrams of the following kind: We regard this diagram as a cube with front face given by the large square. Suppose that the square diagrams given by the faces of the cube except for the front face commute and ι is the injective map. Then the front face also commutes since we have Consider the following diagram: (3.15) where lx = lh, lh sp , lb, ls and ξ = δ, δ sp , β, γ respectively, B ′ = lx(B) and L ′ = lx(L). Thanks to Proposition 3.7, the left face is well-defined, and it commutes by Lemma 3.8. The right face is also well-defined due to Proposition 3.5 and commutes by Lemma 3.9. The back face commutes since lx and ξ are compositions of maps, as defined in Section 3.1, intertwining with Φ X by [KSS02,OSSS17]; Proposition 2.5; and Proposition 2.6. We first show that there exists an injective map Φ from P(B) to RC(L). We use an induction on B such that the application of any lx on B decreases its order. Suppose in the bottom of (3.15) that there exists an injective map from P(B ′ ) to RC(L ′ ) and take p ∈ P(B) such that lx(p) ∈ P(B ′ ). Let (ν ′ , J ′ ) = Φ lx(p) . Then from the commutativity of the left, right and back faces of diagram (3.15) and the bijectivity of Φ X , there must exist ( ν, J) such that ξ( ν, J) = emb(ν ′ , J ′ ). However, from the explicit algorithm of ξ, the rigged configuration ( ν, J) belongs to emb RC(L) . Hence, to p ∈ P(B) one can associate (ν, J) ∈ RC(L) such that emb(ν, J) = ( ν, J). One can thus define Φ(p) = (ν, J). To show that Φ is a bijection, we make (3.15) left and right reversed and replace Φ X with (Φ X ) −1 . Moreover, the top and bottom faces commute by the definition of Φ. Hence, we have Theorem 3.10. Proof. We first show (1) for = ⋆. Let u λ denote the lowest weight vector in B(λ). We have emb(u λ ) ⋆ = u ⋆ λ = u λ = emb(u λ ) = emb(u ⋆ λ ). We also have since the orbits of τ are contained in the orbits of φ, and similarly interchanging e i and f i . Recall that u λ generates B(λ), thus the claim follows from (2.2) and Proposition 2.2. The proof for = ∨ is similar. The proof for = ⋄ follows from [Oka13, Thm. 5.1] and the = ⋆ case. and that the combinatorial R-matrix (resp. R X -matrix) is defined by u → u ′ (resp. u → u ′ ), the left face commutes. Therefore, the front face commutes as desired. Let cc : RC(L) → Z denote the cocharge statistic on rigged configurations following [Sch06b,(3.3)], except for type A (2) † 2n , in which we set t ∨ a = 1 for all a ∈ I. Let D : P(B) → Z be the intrinsic energy statistic on paths [OSS03b, (2.13)] (alternatively [HKO + 02b, (3.8)], but the tensor factors are reversed from our tensor product convention). Proof. We first note that the claim was shown for type A (1) n in [KSS02] and D (1) n in [OSSS17]. For the remaining nonexceptional types g, it is sufficient to show that cc( ν, J) = γ 0 cc(ν, J), Proof. It is known that the claim holds in type A	2018-09-07T00:49:53.000Z	2017-07-16T00:00:00.000	{ "year": 2017, "sha1": "9694a7bad3a0b5f4922efa141a8a9cb844b3952e", "oa_license": null, "oa_url": "https://www.sciencedirect.com/science/article/am/pii/S0021869318305209", "oa_status": "BRONZE", "pdf_src": "Arxiv", "pdf_hash": "9694a7bad3a0b5f4922efa141a8a9cb844b3952e", "s2fieldsofstudy": [ "Mathematics" ], "extfieldsofstudy": [ "Mathematics" ] }
3044563	pes2o/s2orc	v3-fos-license	Tuberculosis contact investigation in an intermediate burden setting: implications from a large tuberculosis contact cohort in Taiwan There are still several unanswered research questions regarding tuberculosis (TB) contact investigations [1–4]. For example, up-to-date epidemiological studies regarding TB contacts in the area of intermediate incidence are not readily available [5]. The long-term risk of TB, especially beyond 5 years, remains undetermined [6]. Risk factors for TB development among contacts are also not well defined [2]. To expand current understanding and knowledge of TB contact investigation, we recruited a population-based TB contacts cohort utilising the National Health Insurance Research Database of Taiwan, which included medical care claims data of 99% Taiwan residents [7]. Tuberculosis contact investigation in an intermediate burden setting: implications from a large tuberculosis contact cohort in Taiwan To the Editor: There are still several unanswered research questions regarding tuberculosis (TB) contact investigations [1][2][3][4]. For example, up-to-date epidemiological studies regarding TB contacts in the area of intermediate incidence are not readily available [5]. The long-term risk of TB, especially beyond 5 years, remains undetermined [6]. Risk factors for TB development among contacts are also not well defined [2]. To expand current understanding and knowledge of TB contact investigation, we recruited a population-based TB contacts cohort utilising the National Health Insurance Research Database of Taiwan, which included medical care claims data of 99% Taiwan residents [7]. Taiwan is a country with an intermediate TB burden. The incidence rate of TB was 72.5 per 100 000 population in 2005 and decreased gradually to 45.7 in 2015 [8]. Supported by the national budget and a single, government-sponsored insurance medical service system, the completion rate of TB contact investigation has reached more than 95% [9]. Participants were included if they had an outpatient diagnosis of International Classification of Diseases, ninth revision, clinical modification (ICD-9-CM) code of V01.1 or 795.5. At the same time, the participants should also fulfil the criteria of co-payment exemption. Under the National TB Program of Taiwan, TB contacts were exempt from a certain fraction of the actual cost of using the medical services, the so-called co-payment exemption, since July 2007. The date of this outpatient visit was set to be the cohort entry date. Contacts were followed until December 31, 2013. The primary outcome was development of active TB. Early adherent isoniazid preventive therapy (IPT) was defined as starting isoniazid within 180 days after the entry date and lasting for at least 180 days. Non-adherent IPT was defined as starting isoniazid within 180 days of the entry date but lasting for fewer than 180 days. Contacts with ICD-9 code 795.5 or a prescription for IPT within 6 months after the entry date were defined as having latent tuberculosis infection (LTBI). The Cox proportional hazard regression model was used to analyse factors associated with development of TB. Adjusted covariates included age, gender, medical visits within 3 months prior to the cohort entry date, contact area TB incidence, urban contact area, economic status, underlying diseases and steroid usage. A total of 312 322 participants were recruited in the cohort (figure 1). The median follow-up time was 955 days (IQR 468-1476 days). Among the 312 322 TB close contacts, 0.7% developed active TB during follow-up (n=2322). Among all TB close contacts, 140 849 (45.1%) were male and the median age was 36 years (IQR 21-52 years). The most common comorbidities were diabetes mellitus (5.8%), chronic obstructive pulmonary disease (2.1%), malignancy (1.6%) and end-stage renal disease (1.2%). Of the overall contacts, 88.6% had no previous underlying diseases and were considered to be healthy. Among overall TB close contacts, the incidence of active TB was 3236, 889 and 415 ( per 100 000 person-years) in the first, second and third months, respectively. At 1 year, the incidence decreased to 202 ( per 100 000 person-years), and at the third year the rate decreased to 119 ( per 100 000 person-years). The incidence rate remained stable within the third to fifth year (143 per 100 000 person-years) (figure 2), but always higher than the baseline population in Taiwan (the incidence rate of TB was 33.6 per 100 000 person-years for those aged between 45 and 55 in 2015 in Taiwan) [8]. The highest TB incidence group was among those aged more than 65 years (634 at first year and 403 at the fifth year per 100 000 Among contacts with LTBI, early adherent IPT was associated with a lower risk of active TB development (HR 0.02, 95% CI 0.02-0.03) while non-adherent IPT remained associated with a lower risk of active TB development (HR 0.11, 95% CI 0.09-0.14). The long-term risk of TB among close contacts remained undetermined since most studies conducted previously were limited by the case number and follow-up period (usually less than 5 years) [3,6]. Our study revealed that the TB risk was still higher than the general population 5 years after TB contact. While it can be impractical in public health policy to regularly follow such a large number of contacts for years, integrating contact history (even beyond years) as a risk factor for tuberculosis into routine clinical practice may be considered. In a meta-analysis pooling five randomised control trials, those who received IPT had a relative risk of 0.31 of developing active TB [10]. Our study revealed a stronger association between IPT and TB risk reduction. This interpretation should be taken with caution because of the observational nature of our study. For example, after contacts were coded as LTBI because of positive tuberculin skin tests and referred, IPT would not be given for those suspected to have coprevalent TB. In addition to what was already included in the guidelines, we found that older age, especially those above 65 years old, was associated with a significant risk for TB (HR 26.1, 95% CI 17.37-39.23). The fact that older age was associated with significantly higher risk does not equal the necessity of treatment of LTBI among elderly people. The risk of TB drug-related hepatotoxicity and systemic drug reaction remains a concern, especially in the elderly population [11,12]. If elderly contacts had some underlying comorbidities such as diabetes mellitus, however, the risk may be further elevated [13,14]. Careful selection of patients is mandatory before initiating LTBI treatment among elderly people. Our study also has limitations. We did not have data on an index case, especially sputum mycobacterial study results. Second, not until 2016 was universal screening for LTBI performed in Taiwan TB contact investigations. Also, the identification of LTBI is through a diagnosis code or use of IPT and has not yet been validated. In conclusion, we found that under a setting of intermediate TB incidence, TB contact investigation resulted in 0.5% yield with a 1-year follow-up. TB risk among contacts remained higher than the general population after a 5-year interval. Specific subgroups of contacts (e.g. age⩾65, AIDS) were at higher risk for developing TB and warrant special attention.	2018-04-03T03:23:36.995Z	2017-08-01T00:00:00.000	{ "year": 2017, "sha1": "0f20625b16f6783a5d249c6477646b0115475629", "oa_license": "CCBYNC", "oa_url": "https://erj.ersjournals.com/content/erj/50/2/1700851.full.pdf", "oa_status": "HYBRID", "pdf_src": "PubMedCentral", "pdf_hash": "0f20625b16f6783a5d249c6477646b0115475629", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
249994024	pes2o/s2orc	v3-fos-license	Deep Proteomics Network and Machine Learning Analysis of Human Cerebrospinal Fluid in Japanese Encephalitis Virus Infection Japanese encephalitis virus is a leading cause of neurological infection in the Asia-Pacific region with no means of detection in more remote areas. We aimed to test the hypothesis of a Japanese encephalitis (JE) protein signature in human cerebrospinal fluid (CSF) that could be harnessed in a rapid diagnostic test (RDT), contribute to understanding the host response and predict outcome during infection. Liquid chromatography and tandem mass spectrometry (LC–MS/MS), using extensive offline fractionation and tandem mass tag labeling (TMT), enabled comparison of the deep CSF proteome in JE vs other confirmed neurological infections (non-JE). Verification was performed using data-independent acquisition (DIA) LC–MS/MS. 5,070 proteins were identified, including 4,805 human proteins and 265 pathogen proteins. Feature selection and predictive modeling using TMT analysis of 147 patient samples enabled the development of a nine-protein JE diagnostic signature. This was tested using DIA analysis of an independent group of 16 patient samples, demonstrating 82% accuracy. Ultimately, validation in a larger group of patients and different locations could help refine the list to 2–3 proteins for an RDT. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD034789 and 10.6019/PXD034789. ■ INTRODUCTION Japanese encephalitis virus (JEV) is a mosquito-borne flavivirus and a leading cause of neurological infection as Japanese encephalitis (JE) in Asia and the Pacific. It is of considerable public health importance, with recent estimates based on sparse data suggesting 1.5 billion people at risk with 42,000 cases per year. 1,2 It is an emerging disease, with recent evidence of JEV in multiple territories in Australia. 3 Patients may experience devastating socioeconomic consequences; JE predominantly affects children in poor rural areas with a 20− 30% case fatality rate and 30−50% of survivors suffer long-term disability. 4 Although no specific treatment is available, several vaccines are available and recommended by the World Health Organization (WHO). 5,6 Although recent efforts have strengthened JEV vaccination programs, still only 15 of 24 endemic countries include JEV vaccine in routine immunization policies, and even then, it is not uniformly nationwide, with vaccine coverage in targeted areas reported to be as low as 39%. 7 JEV is a zoonosis, and sustained vaccine coverage is essential to control disease. A fundamental limitation in the control of JE is the poor accuracy of existing diagnostic tests, requirement for lumbar puncture and laboratory capacity for diagnosis. 8 Surveillance data suggest that only 11 of 24 countries meet the minimum surveillance standards, equivalent to diagnostic testing in a sentinel site. 7 This is a threat to vaccine implementation, as accessible and accurate diagnostics are essential to understand epidemiology and effectiveness of vaccination, identify associated research knowledge gaps, and facilitate public engagement. This also has implications for appropriate risk-assessment for travelers. Aside from JEV control, diagnosis is crucial for patients, families, and health-workers, to be able to institute appropriate supportive and rehabilitation care, stop unnecessary antibiotics, or if the test is negative to prompt further investigation. The gold-standard JEV test is a neutralization assay. However, this requires paired acute and convalescent sera, is laborious, time-consuming, requires specialist skills, high-level isolation facilities for viral cell culture, and may not define the infecting virus in secondary flavivirus infections. 8 The WHO recommended diagnostic test is anti-JEV IgM antibody capture ELISA (MAC-ELISA) of cerebrospinal fluid (CSF). There are limited data from field studies comparing CSF MAC-ELISA with neutralization assays. The manufacturer of the only available commercial kit for clinical diagnosis (InBios) quotes a sensitivity of >90% for well-characterized CSF samples, but sensitivity in the field is as low as 53%. 9 There are also increasingly recognized problems with specificity related to prior vaccination and cross-reactivity with other flaviviruses. 10,11 Reported specificity is >90%; however, a study by our group demonstrated that 13% of patients with anti-JEV IgM detected in CSF by MAC-ELISA had another pathogen detected that may have explained the presentation. 10 Detection of JEV RNA would be highly specific, but the period of viraemia is brief and hard to capture clinically, often occurring before the onset of neurological symptoms and signs. RT-qPCR remains insensitive irrespective of the analytical sensitivity or gene targets. 12 For this reason, the application of metagenomics is not likely to significantly improve JEV RNA detection. Non-targeted, discovery-based liquid chromatography− tandem mass spectrometry (LC−MS/MS) proteomics represents an underused technology for improving diagnosis of JE from the analysis of clinical samples. 13,14 Such an approach is based on the hypothesis that there is a protein signature in the CSF specific for JE and that diagnostic protein biomarkers could be harnessed in an antibody-based point-of-care test. Furthermore, deep proteomics exploration provides insights into disease processes and potential therapeutic targets. Network science and machine learning are two complementary disciplines enabling insights into complex high-dimensional data. 15,16 Networks, composed of nodes and links, are naturally attuned to problems where features have a relational structure 17 and have a track record of success in understanding networks of biological interactions. 18 Weighted correlation network analysis (WGCNA) was developed for the analysis of transcriptomics datasets but is increasingly used in proteomics research, enabling assignment of lengthy lists of proteins into modules with biological insights. 19−21 On the other hand, machine learning can uncover signals in data related to outcome variables and identify predictive markers of disease, a vital exploratory process for constructing diagnostics. 22 Used in conjunction, network science and machine learning provide novel characterization of disease states and can identify robust predictive markers of disease. 23 Herein, we aimed to test the hypothesis that there is a diagnostic protein signature of JE by performing LC−MS/MS in patient samples recruited as part of the Laos CNS study, incorporating differential expression, network, and machine learning analysis. A subsidiary aim was to utilize the data in the same workflow to evaluate proteins associated with outcome of JE. We first performed a pilot feasibility tandem mass tag labeling (TMT) LC−MS/MS study (n = 15) and then a larger verification TMT LC−MS/MS study (n = 148) including a sample size based on a power calculation. These data were combined in the final analysis (n = 163). The results were verified by data-independent acquisition (DIA) LC−MS/MS in 16 (10%) of the samples. Weighted correlation network analysis (WGCNA) was used to explore the data. For the purposes of feature selection and training a machine learning model for classifying JE vs non-JE patients, the TMT LC−MS/ MS data was used excluding the patients analyzed by DIA LC−MS/MS (n = 147).. The model was tested using the DIA LC−MS/MS data (n = 16), providing an independent group of patient samples and alternative methodological analysis. ■ EXPERIMENTAL SECTION Patient Samples A prospective study of central nervous system (CNS) infection has been conducted at Mahosot Hospital, Vientiane, Laos, since 2003. Methods and results from 2003 to 2011 have been described. 24 Patients from 2014 to 2017 were included in the Southeast Asia Encephalitis Project. 25 Inpatients of all ages were recruited for whom diagnostic lumbar puncture was indicated for suspicion of CNS infection because of altered consciousness or neurologic findings and for whom lumbar puncture was not contraindicated. There was no formal definition for CNS infection; patient recruitment was at the discretion of the responsible physician, reflecting local clinical practice. The laboratory also received samples from patients from other hospitals in Vientiane: Friendship, Children's, and Setthathirat Hospitals. Written informed consent was obtained from patients or responsible guardians. Ethical clearance was granted by the Ethical Review Committee of the former Faculty of Medical Sciences, National University of Laos and the Oxford University Tropical Ethics Research Committee. The confirmed etiology was determined by the results of a panel of diagnostic tests which included tests for the direct detection of pathogens in CSF or blood, specific IgM in CSF, seroconversion, or a 4-fold rise in antibody titer between admission and follow-up serum samples. 24 Pathogen detection was confirmed after critical analysis of test results to rule out possible contamination. JEV infection was confirmed, as recommended by the World Health Organization, by detection of anti-JEV IgM by ELISA in CSF or seroconversion in paired serum samples. All anti-JEV IgM positive samples were subsequently confirmed by the gold-standard virus neutralization assay see cited reference 26. Power analysis was performed to estimate the sample size that would be required using different values. A schematic representation of the study design and methods is illustrated in Figure 1. LC−MS Sample Preparation CSF samples were diluted 1:5 in 9 M urea and vortexed intermittently at room temperature for 30 min, to solubilize and denature proteins, inactivating any pathogens and rendering the sample acellular. Protein concentration was assessed with a Nanodrop assay ND-1000 spectrophotometer (Thermo Scientific) by measuring the absorbance at 280 nm and normalized by aliquoting different volumes of each sample dependent on the protein concentration, and then the total volume equalized with 7.5 M urea. An equal volume of 100 mM dithiothreitol (DTT) in 50 mM ammonium bicarbonate (AmBic) was added as a reducing agent, and the samples were vortexed and incubated at 56°C for 45 min. An equal volume of 100 mM iodoacetamide (IAA) in 50 mM AmBic was added as an alkylating agent, vortexed, and incubated at room temperature for 1 h in the dark. 50 mM AmBic was added to each sample to reduce the urea concentration to below 1 M. Digestion was performed with trypsin in a ratio of 1:20 m:m protein:trypsin (Promega, P/N V5072 for the pilot study; V5117 for the larger study); first 75% of the total amount of trypsin added and incubated at 37°C for 18 h overnight and then the remaining 25% added and incubated at 37.5°C for 6 h. The samples were frozen at −20°C to quench the trypsin digestion reaction. A pooled aliquot of each sample was analyzed by label-free LC−MS to verify protein digestion. Reverse-phase (RP) C18 solid-phase extraction (SPE) was used to desalt the digested proteins, as per the manufacturer's instructions (Waters P/N WAT023590 for the pilot study; Thermo Scientific P/N 60109-001 for the larger study). The total eluate was dried completely using a vacuum concentrator (Savant SpeedVac or Eppendorf concentrator) and for the samples to be labeled by tandem mass tag (TMT), resuspended in 100 mM triethylammonium bicarbonate (TEAB). The samples were vortexed, centrifuged, and sonicated for 3 min, and then this was repeated. The Pierce Quantitative Colorimetric Peptide Assay (Thermo Scientific, UK) was performed as per the manufacturer's instructions. The samples were normalized for peptide concentration with TEAB to make up a final volume of 100 μL required for TMT labeling. TMT labeling was performed as per the manufacturer's instructions, in two batches of TMT 11-plex (Thermo Scientific, P/N A37724) for the pilot study and 10 batches of 16-plex (Thermo Scientific, P/N A44520) for the larger study. For the larger study, in order to examine technical variability and adjust for batch effects, each batch contained one reference pool and the batches 9 and 10 had two replicate samples. A pooled sample was analyzed by LC−MS to verify labeling efficiency. Offline High-pH Reverse-Phase Fractionation For the pilot study, offline high-pH reverse-phase fractionation was performed using a Hypersil Gold column (Thermo Scientific, P/N 25002-202130). The mobile phase A was water adjusted with ammonium hydroxide to pH 10 and B was 10 mM ammonium bicarbonate in 80% acetonitrile (ACN) adjusted with ammonium hydroxide to pH 10 and a flow rate of 300 μL/min. The samples were separated into 91 fractions with each fraction collected every 60 s from the start of the run and using the gradient shown in Supplementary Data (S1 Data). For the larger study, offline high-pH reversephase fractionation was performed using an Xbridge BEH C18 column (Waters P/N 186006710). The mobile phase A was water adjusted to pH 10 with ammonium hydroxide and B was 90% ACN adjusted to pH 10 with ammonium hydroxide, at a flow rate of 200 μL/min. Fractions were collected every 60 s from the start of the run (100 fractions) and then concatenated into 44 fractions using the gradient shown in Supplementary Data (S1 Data). The samples analyzed by DIA LC−MS/MS were not processed by offline fractionation. Liquid Chromatography-Mass Spectrometry Online peptide desalting was performed with a Dionex Ultimate 3000 nano UHPLC (Thermo Scientific) using 100% of loading mobile phase A = 0.05% TFA in water at a flow rate of 10 μL/min for 4.6 min. The online desalting column (trap column) used was a C18 column (Thermo Scientific P/N 160454). At 4.6 min, the flow from the nanopump was diverted to the trap column in a backward flush direction. For online low-pH reverse-phase fractionation, the trapped peptides were eluted from the column over the gradient time specified in Supplementary Data (S1 Data). For the pilot study, Accucore C18 columns (Thermo Scientific P/ N 16126-507569) were used with a nanosource, at a flow rate of 250 nL/min. For the larger study, EASY-Spray PepMap C18 columns (Thermo Scientific P/N ES903) were used with an EASY-Spray source and a flow rate of 300 nL/min. Mobile phase A was 0.1% FA and B was 0.1% FA in 80% ACN. MS was performed with a Q Exactive benchtop hybrid quadrupole-Orbitrap MS (Thermo Scientific); the settings are described in detail in Supplementary Data (S1 Data). For the CSF samples processed by DIA LC−MS/MS, samples were analyzed using a Dionex Ultimate 3000 nano UPLC (Thermo Scientific) coupled to an Orbitrap Fusion Lumos mass spectrometer (Thermo Scientific). Briefly, peptides were trapped on a PepMap C18 trap columns (Thermo) and separated on an EasySpray column (50 cm, P/N ES803, Thermo) over a 60min linear gradient from 2% buffer B to 35% buffer B (A: 5% DMSO, 0.1% formic acid in water. B: 5% DMSO, 0.1% formic acid in acetonitrile) at a flow rate of 250 nL/min. The instrument was operated in data-independent mode as previously described. 27 Data Processing and Statistical Analysis The sample size was estimated using a power calculation based on a t test and multiple testing correction, with data from the pilot study and the R package "FDRsampsize". 28 TMT LC−MS/MS analysis protein identification, quantification, missing value imputation and batch correction: Thermo raw files were imported into Proteome Discoverer v2.5 (Thermo Scientific, UK) for peptide identification using the SEQUEST algorithm 29 searching against the SwissProt Homo sapiens and pathogen databases according to the included samples with precursor mass tolerance 10 ppm and fragment mass tolerance 0.02 Da. Carbamidomethylation of cysteine, TMT at N-termini and lysine were set as fixed modifications, and oxidation of methionine was set as a variable modification. False discovery rate (FDR) estimation was performed using the Percolator algorithm. 30 The criteria for protein identification included FDR < 1%, ≥2 peptides per protein, ≥1 unique peptide per protein, ≤2 missed cleavages and ≥6 and ≤144 peptide length (amino acids), coisolation threshold <50%, average S/N threshold >10, and at least two channels with quantification data. Protein quantification was performed in R v 4.1.2 with the package MSstatsTMT. 31 Proteins with >50% missing data were removed, and the data were imputed with the package DreamAI. 32 To incorporate peptide count per protein, jitter was added proportional to 1/median peptide count for each protein. The pilot and larger study data were merged and normalized with the package RobNorm, 33 and then batch correction was performed with the function ComBat 34 in the package sva without modifiers as covariates. 35 The protein list was filtered to remove potential contaminant proteins from the skin or red blood cells, see Supplementary Data S5_contaminants for the list of proteins removed. The effectiveness of batch correction was performed by visualizing the processed data using principal component analysis and hierarchical clustering in MetaboAnalyst v5.0. 36 TMT LC−MS/MS Data Differential Protein Expression. Differential expression between the protein abundance in the JE vs non-JE patient samples was performed using a t test and Benjamini−Hochberg correction for multiple testing. TMT LC−MS/MS Data Protein Set Enrichment Analysis. Functional analysis of human proteins identified in JE vs non-JE patient samples was performed using the WebGestalt online tool 37 using gene set enrichment analysis (GSEA) and gene ontology. TMT LC−MS/MS Data Network Analysis. WGCNA was performed using the package WGCNA: constructing a signed weighted coexpression network with a soft power threshold of 12 to produce a power distribution, that is, scale-free topology; applying hierarchical clustering to detect modules of highly interconnected proteins with a minimum module size of five, deepSplit 4 and merge threshold 0.3; classifying intramodular hub proteins as the five proteins with the highest module membership for each module; and then correlating the modules with patient sample data. 38 Data-Independent Acquisition (DIA) Data Processing. For robustness, final verification was performed on 10% of the samples independently processed via a separate mass spectrometry pipeline using label-free DIA LC−MS/MS. DIA data were analyzed using DIA-NN software (v0.8) with the library-free approach as previously described, 39 using the default settings as recommended. Briefly, for the library-free processing, a library was created from human UniProt SwissProt database (downloaded 24/2/21 containing 20,381 sequences) using deep learning. Trypsin was selected as the enzyme (1 missed cleavage), with carboamidomethylation of C as a fixed modification, oxidation of methionine as a variable modification, and N-term M excision. Identification and quantification of raw data were performed against the in silico library applying 1% FDR at precursor level and match between runs (MBR). The DIA-NN "report.proteingroup" matrix output was further analyzed. Missing values were imputed with half the minimum value for each protein. Feature selection and predictive modeling: This was performed using the TMT LC−MS/MS data without the samples processed by DIA (n = 147) with the Boruta algorithm (using the random forest classifier) using the package Boruta 40 and with Lasso (least absolute shrinkage and selection operator) regression using the package glmnet. 16,41 A final list of proteins based on the intersect between Boruta and Lasso was selected. 42 Classification of JE vs non-JE was performed with selected proteins using several different machine learning models (random forest, support vector machine, logistic regression, and naıve bayes with the package caret and caretEnsemble). 43 Models were trained using tenfold cross-validation repeated 10 times evaluated on AUC-ROC. The ensemble model was tested with the DIA LC−MS/MS data (n = 16). An analysis of feature importance was performed to identify proteins that best predicted the outcome (alive/ died) in JE patients, however due to the small sample size this was considered an exploratory analysis. Feature selection was performed with Boruta and Lasso, and then fivefold crossvalidation was performed on the entire TMT LC−MS/MS dataset using different machine learning models. Protein involvement in biological, molecular, and cellular processes was explored using gene ontology using the webserver STRING, 44 Power analysis was performed to estimate the sample size that would be required to compare differential expression of proteins in JE vs non-JE using different values: with 1,000− 3,000 biomarkers to be tested, 50−150 finally verified, effect size 0.8, power 90%, FDR < 5%, the total sample size with an equal number of JE cases and non-JE controls of 122. Overall, including the pilot and larger study, 163 patients were included: 68 JE and 95 Non-JE; see Table 1 Figure 4; suggesting that 15 modules were associated with JE (p value <0.05), 9 upregulated (red) and 6 downregulated (green). 10 of the modules included proteins in the top five intramodular proteins, that is, proteins with the highest modular membership, with significant differences in abundance between the JE and non-JE group. JEV has a predilection for the thalamus and substantia nigra of the basal ganglia. 26 One of the proteins were "group enriched" in the thalamus, MMP9, from the HPA database. Four proteins were associated with the GO term "substantia nigra development", associated with BASP1, Glucose-6phosphate dehydrogenase (G6PD), YWHAH, and 14-3-3 protein epsilon (14-3-3epsilon). The HPA database includes mRNA expression data from 13 brain regions, including the basal ganglia and thalamus; substantia nigra expression on its own is not reported (https://www.proteinatlas.org/ humanproteome/brain). Feature selection identified a final set of nine proteins which together exhibited high predictive performance ( Figure 6). When examined using the ensemble model, using 10-fold cross-validation, JE classification demonstrated an AUC-ROC of 98.7 (98.0−99.4), in addition to high sensitivity and specificity�metrics in Table 2 and ROC in Supplementary Data S13 and S14. Data acquired by DIA LC−MS/MS of 16 samples were used to verify the nine-protein JE diagnostic predictive model. The test metrics are reported in Table 2. Establishing CSF Molecular Signatures as Predictors of the JE Outcome Feature Selection. Subgroup analysis was performed using 42 JE samples for which outcome data at hospital discharge (died vs alive) were available. Seven proteins were identified as important in predicting outcome using the Boruta algorithm and two proteins using Lasso, such that two proteins were identified by both Boruta and Lasso, see Supplementary Data S15. In view of the small sample size, the data were not split into a training and test set. These proteins were used to train different models with fivefold CV repeated ten times evaluated on ROC and then combined in an ensemble model with crossvalidation scores reported in Table 2, see the list of proteins in Supplementary Data S15 and ROC in Data S16. There were five JE patients in the DIA LC−MS analysis of which 3 had outcome data, and this was considered too small to report test metrics. ■ CONCLUSIONS We performed deep untargeted analysis of well-characterized patient CSF samples from a large number of different confirmed neurological infections. To our knowledge, the highest number of proteins in CSF identified to date has been 3,174; 48 thus, this research represents a notable improvement in terms of the numbers of proteins identified, and this serves as a marker of the depth of analysis and prospects for biomarker identification. 49 Offline fractionation into 90 fractions in the pilot study, and 100 fractions concatenated into 44 in the larger study, with two-hour online LC gradients and multiplexing with TMT-16plex contributed to the depth of Table 1. continued WGCNA identified 20 clusters of highly correlated proteins and provided insight into the proteins and how they associate with disease mechanisms. The modules were allocated a descriptor, according to gene ontology analysis, as well as the clinical and biological significance of the proteins. For example, one module was associated with IgM (proteins in the module included immunoglobulin heavy constant mu and immunoglobulin J chain) and correlated with JE and Orientia tsutsugamushi (OT), as well as the duration of illness. Other important modules associated with upregulation in JE included neuronal damage, antiapoptosis, heat shock response, unfolded protein response, cell adhesion, and macrophage and dendritic cell activation. In contrast, in comparison to other non-JE neurological infections, there was an association with downregulated acute inflammatory response, hepatotoxicity, activation of coagulation, extracellular matrix, and actin regulation. Predictive modeling using the nine-protein ensemble model enabled classification of JE and non-JE samples with a CV accuracy of 97.0 (95% CI 95.7−98.0) using TMT labeled DDA data, and 81.3% (95% CI 54.4−96.0) in verification with 16 (10%) of the samples by DIA. DIA is a label-free method of analysis, with ongoing improvements in depth and throughput; in this case providing a complementary method to verify the TMT data rather than performing traditional targeted LC− MS/MS proteomics such as parallel reaction monitoring. Three proteins selected as the best disease classifiers were not "significant", i.e., p value <0.05 with t-test and adjustment for multiple testing, highlighting the limitations of univariate analysis in biomarker identification. 50 Biomarker discovery is a lengthy process, akin to the pharmaceutical pipeline. 13 The work demonstrates important CSF proteins in classifying JE vs non-JE. However, there is no doubt that the protein signature needs to be validated with orthogonal antibody-based methods in additional patient groups. It will also be useful to compare this with protein profiling in other body fluids. This will inform the use of a smaller subset of proteins in an ELISA or rapid diagnostic test to be tested alongside the existing anti-JEV IgM assay. To date, to our knowledge, two studies have utilized unbiased techniques to examine the CSF proteome in human patients with confirmed JEV infection; while they demonstrate the feasibility of the methods, the patients were not confirmed by seroneutralization and included relatively small numbers of patients (10 and 26 JE patients). 51,52 There have been a handful of studies utilizing ELISA methods to target specific proteins; however, these rarely used power calculations in their experimental design, nor did they include adequate controls. 53 Furthermore, mRNA expression does not directly correlate with that of the corresponding protein. 70 As expected, while we included JEV proteins in the search database, we did not identify any JEV proteins. This is compatible with previous publications; non-structural protein 1 is the major secreted protein during flavivirus infections, harnessed widely as a diagnostic biomarker for dengue virus infection, but not a useful diagnostic biomarker for JE. 71 The data provide useful interrogation of the host response to JEV infection. The identified proteins fit well into the existing literature on the host response in JEV and other closely associated flavivirus infections, most importantly West Nile virus infection. 72,73 MAPT and MAP2 are both closely associated microtubule stabilizing proteins specific to neuronal cells. 74 Both proteins were identified in this study as being biomarkers of JE in CSF, and the high levels in comparison to other neurological infections are striking. The association of the former has previously been demonstrated by ELISA, in one of the only studies of this type. 75 The role of actin, microtubule, and intermediate filament cytoskeletal reorganization in flavivirus infection has been described 76 and upregulation of MAPT and MAP2 may represent neuronal damage following transneural spread of JEV. Other proteins that were associated with JE in this study, all within the red WGCNA module, that may reflect neuronal damage include Paralemmin, Calbindin 1, MAP2, Parvalbumin, Secernin 1, and cell cycle exit and neuronal differentiation. The upregulation of ISG15 and ISG20 fit in with the known upregulation of a host of ISGs as part of the innate immune response to a viral infection. 77,78 Additional functional enrichments reflecting different WGCNA modules have previously been described anti-apoptosis, 79 heat shock response, 80,81 unfolded protein response, 82 translation, 83 IgM, 84 cell adhesion and pathogen attachment, 85 endothelial activation, 86 and macrophage activation. 87,88 In comparison to other neurological infections, there was a downregulation in acute phase response proteins and neutrophil enriched proteins, as has been seen by other studies. 89−91 In these, however, the sample size for the analysis of proteins predictive of outcome was less substantial and not supported by an a priori power calculation. Incomplete coverage and missing data between LC−MS runs is an ongoing issue in the field. 32 It is notable that comparing with other similar studies in the literature, the important proteins may not be exactly the same but are closely related. These issues are now being improved by DIA methods. Further limitations are that the demographics of the cases and Additional data using antibody-based methods will allow the nine-protein signature to be refined. This could be performed by purchasing or developing ELISA assays and comparing the specific protein abundance in JE and non-JE patients. These data will need to be validated in a larger group of patients, in Results of MSstatsTMT analysis for the pilot (tab 1) and larger (tab 2) study; S5_Contaminants: List of potential protein contaminant proteins that were excluded from downstream analysis; S6_Combined data processed: Normalized and batch corrected data from the pilot and larger study; S7: Assessment of data processing and effectiveness of batch correction; S8_ JE vs non-JE protein set enrichment analysis: Results of protein set enrichment analysis performed using the WebGestalt online tool; S9_WGCNA sample clustering to detect outliers: WGCNA hierarchical clustering of the samples to identify outliers; S10_WGCNA output: Results of WGCNA analysis; S11 WGCNA summary of modules: WGCNA eigengene dendrogram to illustrate groups of correlated modules; S12 Boruta Lasso proteins: List of proteins identified as important in the classification of JE vs non-JE by either Boruta or Lasso algorithms; S13_Predictive modeling to identify a JE diagnostic protein signature; S14_ROC: Receiver operator characteristic curve of the nine-protein model built to classify JE and non-JE patients; S15_Boruta proteins outcome: List of proteins identified as important in predicting the outcome (dead vs alive) of JE patients; and S16	2023-05-24T06:17:49.829Z	2023-05-23T00:00:00.000	{ "year": 2023, "sha1": "3819e5cb8108f343e78e2858e2a0ca8c8902ab53", "oa_license": "CCBY", "oa_url": "https://doi.org/10.1021/acs.jproteome.2c00563", "oa_status": "HYBRID", "pdf_src": "PubMedCentral", "pdf_hash": "f4bc1a41612cd91cdb37baef9a7c24c9bef33626", "s2fieldsofstudy": [ "Biology", "Medicine" ], "extfieldsofstudy": [ "Biology" ] }
237433655	pes2o/s2orc	v3-fos-license	Text-to-Table: A New Way of Information Extraction We study a new problem setting of information extraction (IE), referred to as text-to-table. In text-to-table, given a text, one creates a table or several tables expressing the main content of the text, while the model is learned from text-table pair data. The problem setting differs from those of the existing methods for IE. First, the extraction can be carried out from long texts to large tables with complex structures. Second, the extraction is entirely data-driven, and there is no need to explicitly define the schemas. As far as we know, there has been no previous work that studies the problem. In this work, we formalize text-to-table as a sequence-to-sequence (seq2seq) problem. We first employ a seq2seq model fine-tuned from a pre-trained language model to perform the task. We also develop a new method within the seq2seq approach, exploiting two additional techniques in table generation: table constraint and table relation embeddings. We consider text-to-table as an inverse problem of the well-studied table-to-text, and make use of four existing table-to-text datasets in our experiments on text-to-table. Experimental results show that the vanilla seq2seq model can outperform the baseline methods of using relation extraction and named entity extraction. The results also show that our method can further boost the performances of the vanilla seq2seq model. We further discuss the main challenges of the proposed task. The code and data are available at https://github.com/shirley-wu/text_to_table. Introduction Information extraction (IE) is a task that aims to extract information of interest from text data and represent the extracted information in a structured form. Traditional IE tasks include named entity recognition which recognizes entities and their types (Huang, Xu, and Yu 2015;Ma and Hovy 2016;Lample et al. 2016;Devlin et al. 2019), and relation extraction which identifies the relationships between entities (Zheng et al. 2017;Zeng et al. 2018;Luan et al. 2019;Zhong and Chen 2020). Since the results of IE are structured, they can be easily used by computer systems in different applications such as text mining. In this work, we study IE in a new setting, referred to as text-to-table. First, the system receives a training dataset containing text-table pairs. Each text-table pair contains a text and a table (or tables) representing information extracted * Work done in Bytedance AI Lab Figure 1: A example of text-to-table from the Rotowire dataset. The text is a report of a basketball game, and the tables are the scores of the teams and players. from the text. The system learns a model for information extraction. Next, the system employs the learned model to conduct information extraction from a new text and outputs the result in a table (or tables). Figure 1 gives an example of text-to-table, where the input (above) is a report of a basketball game, and the output (below) is two tables summarizing the scores of the teams and players from the input. Our work is inspired by research on the so-called table-totext (or data-to-text) problem, which is the task of generating a description for a given table. Table-to-text is useful in applications where the content of a table needs to be described in natural language. Thus, text-to-table can be regarded as an inverse problem of table-to-text. However, there are also differences. Most notably, their applications are different. Text-to-table can be applied to document summarization, text mining, etc. Text-to-table is unique compared to the traditional IE approaches. First, it is mainly designed to extract information on complex relations between items from a long text (e.g., the whole document). Second, the schemas for extraction are implicitly included in the training data, and there is no need to explicitly define the schemas. As a result, one can easily build a system to perform the task. In this work, we formalize text-to-table as a sequenceto-sequence (seq2seq) task. More specifically, we translate the text into a sequence representation of a table (or tables), where the schema of the table is implicitly contained in the representation. We also assume that the seq2seq model is built on top of a pre-trained language model, such as BART (Lewis et al. 2019) and T5 (Raffel et al. 2020). Although the approach is a natural application of existing technologies, as far as we know, there has been no previous study to investigate to what extent the approach works. We also develop a new method for text-to -table within the seq2seq approach with two additional techniques, table constraint and table relation embeddings. Table constraint controls the creation of rows in a table and table relation embeddings affect the alignments between cells and their row headers and column headers. Both are to make the generated table well-formulated. The approach to IE based on seq2seq has already been proposed. Methods for conducting individual tasks of relation extraction (Zeng et al. 2018;Nayak and Ng 2020), named entity recognition (Chen and Moschitti 2018;Yan et al. 2021), and event extraction (Lu et al. 2021) have been developed. Methods for jointly performing multiple tasks of named entity recognition, relation extraction, and event extraction have also been devised (Paolini et al. 2021). Most of the methods exploit suitable pre-trained models such as BERT. However, all the existing methods rely on pre-defined schemas for extraction. Moreover, their models are designed to extract information from short texts, rather than long texts, and extract information with simple structures (such as an entity and its type), rather than information with complicated structures (such as a table). We conduct extensive experiments on four existing tableto-text datasets. Experimental results show that the vanilla seq2seq model fine-tuned from BART (Lewis et al. 2019) can outperform the state-of-the-art IE models fine-tuned from BERT (Devlin et al. 2019;Zhong and Chen 2020). Furthermore, results show that our proposed approach to text-to-table with the two techniques can further improve the extraction accuracies. We also summarize the challenging issues with the seq2seq approach to text-to-table for future research. We make the following contributions in this work: 1. We propose the new task of text-to-table for IE. We derive four new datasets for the task from existing datasets. 2. We formalize the task as a seq2seq problem and propose a new method within the seq2seq approach using the techniques of table constraint and table relation embeddings. 3. We conduct extensive experiments to verify the effectiveness of the proposed approach. Related Work Information Extraction (IE) is a task of extracting information (structured data) from a text (unstructured data). For example, named entity recognition (NER) recognizes entities appearing in a text. Relation extraction (RE) identifies the relationships between entities. Another example is event extraction (EE), which discovers events occurring in a text. Traditionally, researchers formalize the task as a language understanding problem. The state-of-the-art methods for NER perform the task on the basis of the pre-trained language model BERT (Devlin et al. 2019). The pipeline approach to RE divides the problem into NER and relation classification, and conducts the two sub-tasks in a sequential manner (Zhong and Chen 2020), while the end-to-end approach jointly carries out the two sub-tasks (Zheng et al. 2017;Zeng et al. 2018;Luan et al. 2019). The state-of-the-art methods for EE also employ BERT and usually jointly train the models with other tasks such as NER and RE Zhang, Ji, and Sil 2019;Lin et al. 2020). All the methods assume the use of pre-defined schemas (e.g., entity types for NER, entity and relation types for RE, and event templates for EE). Besides, most methods are designed for extraction from short texts. Therefore, existing methods for IE cannot be directly applied to text-to-table. IE is also conducted at document level, referred to as doclevel IE. For example, some NER methods directly perform NER on a long document (Strubell et al. 2017;Luo et al. 2018), and others encode each sentence in a document, use attention to fuse document-level information, and perform NER on each sentence (Hu et al. 2020;Xu, Wang, and He 2018). There are also RE methods that predict the relationships between entities in a document (Yao et al. 2019;Nan et al. 2020a). However, existing doc-level IE approaches usually do not consider extraction of complex relations between many items. Sequence-to-sequence (seq2seq) is the general problem of transforming one text into another text (Sutskever, Vinyals, and Le 2014;Bahdanau, Cho, and Bengio 2014), which includes machine translation, text summarization, etc as special cases. The use of the pre-trained language models of BART (Lewis et al. 2019) and T5 (Raffel et al. 2020) can significantly boost the performances of seq2seq, such as machine translation (Lewis et al. 2019;Raffel et al. 2020;Liu et al. 2020) and text summarization (Lewis et al. 2019;Raffel et al. 2020;Huang et al. 2020). Recently, some researchers also formalize the IE problems as seq2seq, that is, transforming the input text into an internal representation. One advantage is that one can employ a single model to extract multiple types of information. Experimental results show that this approach works better than or equally well as the traditional approach of language understanding, in RE (Zeng et al. 2018;Nayak and Ng 2020), NER (Chen and Moschitti 2018;Yan et al. 2021) and EE (Lu et al. 2021). Data-to-text aims to generate natural language descriptions from the input structured data such as sport commentaries (Wiseman, Shieber, and Rush 2017). The structured data is usually represented as tables (Wiseman, Shieber, and Rush 2017;Thomson, Reiter, and Sripada 2020;Chen et al. 2020), sets of table cells (Parikh et al. 2020;Bao et al. 2018), semantic representations (Novikova, Dušek, and Rieser 2017), or sets of relation triples (Gardent et al. 2017;Nan et al. 2020b). The task requires the model to select the salient information from the data, organize it in a logical order, and generate an accurate and fluent natural language description (Wiseman, Shieber, and Rush 2017). Data-totext models usually adopt the encoder-decoder architecture. The encoders are specifically designed to model the input data, such as multi-layer perceptron (Puduppully, Dong, and Lapata 2019a,b), recurrent neural network (Juraska et al. 2018;Liu et al. 2018;Shen et al. 2020 Problem Formulation Text-to-table takes a text as input and produces one or several tables to summarize the content of the text, as shown in Figure 1. This can be considered as an inverse problem of data-to-text, and each has its applications. The input is a text denoted as x = x 1 , x 2 , · · · , x \|x\| . The output is one table or multiple tables. For simplicity suppose that there is only one table denoted as T . Further suppose that T has n r rows and n c columns. There are n r × n c cells in T , where the cell of row i and column j is a sequence of There are three types of table: one that has both column headers and row headers, one that has only column headers, and one that only has row headers. For example, the player table in Figure 1 has both column headers ("Assists", "Points", etc) and row headers ("Al Horford", "Isaish Thomas", etc). We let t 1,j , j = 2, 3, · · · , n c to denote the column headers, let t i,1 , i = 2, 3, · · · , n r to denote the row headers, and let t 1,j , i = 2, 3, · · · , n r , j = 2, 3, · · · , n c to denote the nonheader cells of the table. For example, in the player table in Figure 1, t 1,2 = Assists, t 2,1 = Al Horford, and t 2,2 = 5. We consider using machine learning to perform text-totable. In learning, a number of text-table pairs are given as training data, and a model is trained from the data. In inference, the learned model is utilized to generate a table or tables given a new text. Once the information of a text is extracted into tables via text-to-table, it can be leveraged in many different applications such as document summarization and text mining. For example, in Figure 1, one can quickly obtain the key information of the text by simply looking at the tables summarized from the text. There are differences between text-to-table and traditional settings of information extraction. As can be seen from the example in Figure 1, extraction of information is performed from the entire document. The extracted information (structured data) is in a complex form, specifically multiple types of scores of teams and players in a basketball game. Furthermore, the data-driven approach is taken, and the schemas of the tables do not need to be explicitly defined. Our Method We develop a method for text-to-table using the seq2seq approach and the two techniques of table constraint and table relation embeddings. Seq2Seq Framework We formalize text-to-table as a sequence-to-sequence (seq2seq) problem (Sutskever, Vinyals, and Le 2014;Bahdanau, Cho, and Bengio 2014). Specifically, given an input text, we generate a sequence representing the output table (or tables). We introduce two special tokens, a separation token denoted as " s " and a new-line token denoted as " n ". For a table t, we represent each row t i with a sequence of cells delimited by separation tokens: We represent the entire table with a sequence of rows delimited by new-line tokens: s , t 2,1 , s , · · · , s , t 2,nc , s , n , · · · · · · s , t nr,1 , s , · · · , s , t nr,nc , s Figure 2 shows the sequence of the player table in Figure 1. When there are multiple tables, we create a sequence of tables using the captions of the tables as delimiters. Let x = x 1 , · · · , x \|x\| and y = y 1 , · · · , y \|y\| denote the input and output sequences respectively. In inference, the model generates the output sequence based on the input sequence. We conduct decoding in an auto-regressive way, which generates one token at each step based on the tokens it has generated so far. Formally, the model calculates the conditional probability: We use decoding algorithms such as beam search or greedy search to find an output sequence that approximately maximizes the conditional probability. In training, we learn the model based on the text-table pairs {(x 1 , y 1 ), (x 2 , y 2 ), · · · , (x n , y n )}. The objective of learning is to minimize the cross-entropy loss arg min where θ denotes the parameter of the model. In our work, we adopt Transformer as the model (Vaswani et al. 2017), which is the state-of-the-art method for seq2seq. We build the model on top of the pre-trained language model BART (Lewis et al. 2019) by fine-tuning. We refer to the method described above as "vanilla seq2seq". There is no guarantee, however, that the output sequence of vanilla seq2seq represents a well-formulated table. We add a post-processing method to ensure that the output sequence is a table. The post-processing method takes the first row generated as well-defined, deletes extra cells at the end of the other rows and inserts empty cells at the end of the other rows. Techniques We develop two techniques to improve table generation, called table constraint and table relation embeddings. We use "our method" to denote the seq2seq approach with these two techniques. Table Constraint Our method exploits a constraint in the decoding process to ensure that the output sequence represents a well-formulated table. Specifically, our method calculates the number of cells in the first row it generates, and then forces the following rows to contain the same number of cells. The algorithm of the decoder is shown in Algorithm 1. Table Relation Embeddings Our method also incorporates table relation embeddings including row relation embeddings and column relation embeddings into the self-attention of the Transformer decoder. Given a token in a non-header cell, the row relation embeddings τ K r and τ V r indicate which row header the token is aligned to, and the column relation embeddings τ K c and τ V c indicate which column header the token is aligned to. Let us consider the self-attention function in one block of Transformer decoder: at each position, self-attention only attends to the previous positions. For simplicity, let us only consider one head in the self-attention. At the t-th position, the input of self-attention is the sequence of representations z = (z 1 , · · · , z t ) and the output is the sequence of representations h = (h 1 , · · · , h t ), where z i ∈ R d and h i ∈ R d are the representations at the i-th position (i = 1, · · · , t). In a conventional Transformer decoder, self-attention is defined as follows, i = 1, · · · , t, j = 1, · · · , i where W Q , W K , W V ∈ R d×d k are the query, key, and value weight matrices respectively, and W O ∈ R d k ×d is the output weight matrix. Algorithm 1: Decoding using table constraint. eos , s , and n denote the end of sentence, separation token, and new-line token respectively. Seq2seq denotes the seq2seq model. Decode denotes the decoding algorithm such as beam search and greedy search. The relation vectors r K ij and r V ij are defined as follows. For the token at the i-th position, if the token at the j-th position is a part of its row header, then r K ij and r V ij are set to the row relation embeddings τ K r and τ V r . Similarly, for the token at the i-th position, if the token at the j-th position is a part of its column header, then r K ij and r V ij are set to the column relation embeddings τ K c and τ V c . Otherwise, r K ij and r V ij are set to 0. To identify the row header or the column header of a token, we parse the sequence generated so far to create a partial table using the new-line tokens and separation tokens in the sequence. Figure 3 illustrates how relation vectors are constructed. Datasets We make use of four existing datasets which are traditionally utilized for data-to-text: Rotowire (Wiseman, Shieber, and Rush 2017), E2E (Novikova, Dušek, and Rieser 2017), Wik-iTableText (Bao et al. 2018), and WikiBio (Lebret, Grangier, and Auli 2016). In each dataset, we filter out the content in the tables that does not appear in the texts. We plan to make the processed datasets publicly available for future research. Table 2 gives the statistics of the Rotowire dataset and Table 1 gives the statistics of the other three datasets. Rotowire is from the sports domain. Each instance is composed of a text and two tables, where the text is a report of a basketball game and the two tables represent the scores of teams and players respectively (cf., Figure 1). Each table has column headers describing the types of scores, and row headers describing the names of teams or players. The texts are long and may contain irrelevant information such as the performance of players in other games. Thus this is a challenging dataset for text-to- Procedure Methods: We conduct experiments with vanilla seq2seq and our method, as well as baselines. We know of no existing method that can be directly employed in text-to-table. For each dataset, we first define the schemas based on the training data, then use an existing method of relation extraction (RE) or named entity extraction (NER) to extract information, and finally create tables based on the schemas and extracted information. We take it as the baseline for the dataset. No baseline can be applied to all four datasets. For RE, we use PURE, a state-of-the-art method (Zhong and Chen 2020). For NER, we use BERT model (Devlin et al. 2019). Training: For vanilla seq2seq and our method, we finetune the seq2seq models from BART-base, which has 12 layers, 768 hidden dimensions, 16 heads, and 139M parameters. For RE and NER, we fine-tune the models from BERT-baseuncased, which has 12 layers, 768 hidden dimensions, 12 heads, and 110M parameters. All models are trained with the Adam optimizer until convergence, and the hyper-parameters are tuned on the development sets. Appendix A shows the hyper-parameters. For the small datasets of Rotowire and WikiTableText, we run experiments five times with different random seeds and take average of results to reduce variance. Table 4: Experimental results of our method, vanilla seq2seq, and the baseline of NER, on WikiTableText and WikiBio. Evaluation: We evaluate the performance of a method based on the number of correct non-empty cells in the tables (i.e., we ignore empty cells). To judge whether a cell is correctly generated in the table, we use not only its content but also its row header and column header to ensure that the cell is on the right row and right column. Exact match is used to compare the content of the generated cell and the ground truth. We adopt precision, recall, and F1 score as evaluation measures. We calculate the measures on each generated table and then take the average on all tables. We also evaluate the percentage of output sequences that cannot represent wellformulated tables, referred to as error rate. Table 3 shows the results on the Rotowire dataset. One can see that in terms of F1 score, our method performs the best followed by vanilla seq2seq, and both outperform the baselines of doc-level RE and sent-level RE. The RE baselines perform quite well, but they heavily rely on rules and cannot beat the seq2seq approach. Among them the doc-level RE performs better than sent-level RE, because some information in Rotowire can only be extracted when cross-sentence context is provided. Results on Rotowire We implement two baselines of RE, namely doc-level RE and sent-level RE. We take team names, player names, and numbers of scores as entities and take types of scores as relations. Sent-level RE predicts the relations between entities within each sentence. Doc-level RE predicts the relations between entities within a window (the window size is 12 entities) and uses the approximation model proposed by Zhong and Chen (2020) to speed up inference. Table 4 shows the results of our method, vanilla seq2seq, and the baseline of NER on E2E, WikiTableText, and WikiBio. Again, the seq2seq approach outperforms the baseline. The NER baseline has slightly higher precision, but the seq2seq approach has significantly higher recall and F1. Our method and vanilla seq2seq are comparable, because the table structures in the three datasets are very simple (there are only two columns in the tables), and the use of the two techniques does not further improve the performances. The NER baseline has high precision but low recall, mainly because NER can only make the right decision when it is clear. Results on E2E, WikiTableText and WikiBio We implement the baseline of NER in the following way. We view the non-head cells in the tables as entities and their row headers as entity types. In training, we match the nonhead cells into the texts and take them as "entities" in the texts. Only a proportion of the non-header cells can be matched into the texts (85% for E2E, 74% for WikiTableText, and 69% for WikiBio). Additional Study We carry out ablation study on our method, specifically we exclude pre-trained language model, table constraint (TC) and table relation embeddings (TRE) from our method. Note that our method without TC and TRE is equivalent to vanilla seq2seq. Table 5 gives the results on the four datasets. It can be seen that the use of both TC and TRE can significantly improve the performance on Rotowire, which indicates that our method is particularly effective when the tables are large with many rows and columns. There are not significant improvements on E2E, WikiTableText, and WikiTableText, apparently because formulation of tables is easy for the three datasets. Therefore, we conclude that the two techniques of TC and TRE are helpful when the task is difficult. The use of pre-trained language model can boost the performances on all datasets, especially on Rotowire and Wik-iTableText. This indicates that pre-trained language model is particularly helpful when the task is difficult and the size of training data is small. We observe that vanilla seq2seq makes more formatting errors than our method, especially on player tables in Rotowire that have a large number of columns. It indicates that for vanilla seq2seq, it is difficult to keep track of the columns in each row and make alignments with the column headers. In contrast, the two techniques of our method can help effectively cope with the problem. Figure 4 shows a bad case of vanilla seq2seq, where the model correctly infers the column of "assists" but fails to infer the columns of "personal fouls", "points", and "total rebounds" for the row of "Rajon Rondo". In contrast, our method can successfully handle the case, be- cause TC can eliminate the incorrectly formatted output, and TRE can make correct alignments with the column headers. We also investigate the effect of the scale of pre-trained language model BART. We use both BART-base and BARTlarge and conduct fine-tuning on top of them for vanilla seq2seq and our method. Table 6 gives the results on the four datasets. The results show that the use of BART-large can further boost the performances on all four datasets, indicating that it is better to use larger pre-trained models when computation cost is not an issue. Discussions We analyze the experimental results on the four datasets and identify five challenging issues. (1) Text Diversity: Extraction of the same content from different expressions is one challenge. For example, the use of synonyms is very common in Rotowire. The team of "Knicks" is often referred to as "New York", its home city. Identification of the same entities from different expressions is needed in the task. (2) Text Redundancy: There are cases such as those in Wik-iBio, in which the texts contain much redundant information. This poses a challenge to the text-to-table model to have a strong ability in summarization. It seems that the seq2seq approach works well to some extent but further improvement is undoubtedly necessary. (3) Large Table: The tables in Rotowire have large numbers of columns, and the extraction from them is challenging even for our method of using TC and TRE. (4) Background Knowledge: WikiTableText and WikiBio are from open domain. Thus, performing text-to-table on such kind of datasets require the use of much background knowledge. The use of more powerful pre-trained language models can be further explored in the future. (5) Reasoning: Sometimes the information is not explicitly presented in the text, and reasoning is required to conduct correct extraction. For example, an article in Rotowire reports a game between the two teams "Nets" and "Wizards". From the sentence: "The Nets seized control of this game from the very start, opening up a 31 -14 lead after the first quarter", humans can infer that the point of "Wizards" is 14, which is still difficult for machines. Conclusion We propose employing text-to-table as a new way of information extraction (IE), which extracts information of interest from the input text and summarizes the extracted information in tables. The advantage of the approach is that one can easily conduct information extraction from either short texts or long texts to create simple tables or complex tables without explicitly defining the schemas. Text-to-table can be viewed as an inverse problem of table-to-text. We formalize text-to-table as a sequence-to-sequence problem on top of a pre-trained model. We further propose an improved method using a seq2seq model and table constraint and table relation embeddings techniques. We conduct experiments on four datasets derived from existing table-to-text datasets. The experimental results demonstrate that our proposed approach outperforms existing methods using conventional IE techniques. We further analyze the challenges of text-to-table for future study. The issues include diversity of text, redundancy of text, large-table, background knowledge, and reasoning.	2021-09-08T01:15:54.739Z	2021-09-06T00:00:00.000	{ "year": 2021, "sha1": "0a06ec55462aa5c438a776fd8ca7d6f3bd1c9088", "oa_license": "CCBY", "oa_url": "https://aclanthology.org/2022.acl-long.180.pdf", "oa_status": "HYBRID", "pdf_src": "MergedPDFExtraction", "pdf_hash": "c200950cf6ae80219342123e2d8b4dcddca7c40e", "s2fieldsofstudy": [ "Computer Science" ], "extfieldsofstudy": [ "Computer Science" ] }
215182497	pes2o/s2orc	v3-fos-license	Chalazion A 21 year old girl presents with a painless lump in her right upper lid. She says that this has been present for a few weeks and seems to be enlarging slowly. Her pharmacist suggested an antibiotic ointment, which she has been using for a few days with no benefit. A 21 year old girl presents with a painless lump in her right upper lid. She says that this has been present for a few weeks and seems to be enlarging slowly. Her pharmacist suggested an antibiotic ointment, which she has been using for a few days with no benefit. What you should cover A chalazion, or meibomian cyst, is a focus of granulomatous inflammation within the eyelid. It arises from retained meibomian secretions. It is benign and often self limiting. It can occur in all age groups and is common in primary care. • Patients report a slowly enlarging lump with some variability in size on a day to day basis. • Ask about skin conditions which predispose to meibomian gland dysfunction-acne rosacea and seborrhoeic dermatitis. • Larger chalazions may be associated with visual symptoms. Ask about blurry vision from induced astigmatism or an awareness of visual field obstruction from mechanical ptosis. • Ask about pain, as this allows the chalazion to be differentiated from a hordeolum (a small abscess); chalazion is painless. • Ask if the eyes have been unusually gritty, uncomfortable, or "tired"-these symptoms point to blepharitis, which predisposes to chalazion formation. Patients with severe blepharitis also report sticky eyes, especially on waking. What you should do During examination, look and feel: • Observe the facial skin for midface telangiectasia (rosacea) or an unusually "oily" or "dry" appearance (dermatitis). The patient will usually point to an obvious lump in the eyelid. Note the appearance, which should be smooth. The overlying skin is usually normal but may occasionally be indurated. Ptosis may be obvious. Swelling and redness indicate early or established infection. Observe the lid margins for crusting and a red rimmed appearance (blepharitis). Ulceration and destructive changes of the lid margin are red flag signs indicating possible malignancy and warrant referral. • Run the tip of your finger along the eyelid margin. The lump is non-tender and can be either firm (longstanding) or soft and slightly fluctuant (early). There may be more than one lump. Tenderness and erythema are signs of infection (hordeolum). Management consists of dealing with the presenting lump and preventing recurrence: • Most chalazions will respond to conservative treatment of applying heat and massage at least twice a day (figure⇓). Explain that the lump is caused by an obstructed eyelid gland that contains fatty material; heat softens this material, and massage disperses it. Early treatment may lead to faster resolution. To improve compliance, explain that resolution often takes time and that several weeks of regular hot bathing may be required. Topical antibiotic preparations are of no benefit and should be avoided. Surgical management, involving incision and curettage or intralesional triamcinolone, should be regarded as second line treatment. Infected chalazions (hordeolums), with or without cellulitis, require oral antibiotics. • Prevention relies on management of blepharitis and its causes. Advise twice daily lid hygiene (heating and cleaning; figure⇓), on a long term basis, to those with obvious blepharitis. Explain that blepharitis is a chronic condition and that flare-ups can be managed with regular lid hygiene. Lid hygiene also reduces the likelihood of a gland becoming obstructed. Consider a course of tetracycline (doxycycline 50-100 mg once daily or lymecycline 408 mg once daily for at least three months) for blepharitis that is severe or associated with rosacea. Erythromycin is a good alternative in children and pregnancy. Have a low threshold for referring young children with large chalazions on account of the risk of amblyopia. In elderly people, consider the possibility of sebaceous cell carcinoma if a chalazion fails to settle or recurs. Contributors: ZIC and EMA had the idea for the article; all authors contributed equally to preparing and editing the manuscript. ZIC is guarantor. Competing interests: All authors have completed the unified competing interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare no support from any organisation for the submitted work; no financial relationships with any organisation that might have an interest in the submitted work in the previous three years; and no other relationships or activities that could appear to have influenced the submitted work. Provenance and peer review: Not commissioned; externally peer reviewed.	2018-04-03T02:22:21.446Z	2010-01-01T00:00:00.000	{ "year": 2010, "sha1": "be3d4eed0758c7e09bee272c9013bd47f386b505", "oa_license": "implied-oa", "oa_url": "https://europepmc.org/articles/pmc2974575", "oa_status": "GREEN", "pdf_src": "BMJ", "pdf_hash": "0a49bf8ba447a000444086f41f17fd9763acc00d", "s2fieldsofstudy": [ "Medicine" ], "extfieldsofstudy": [ "Medicine" ] }
237755550	pes2o/s2orc	v3-fos-license	An Ultrahigh Narrowband Absorber Close to the Information Communication Window In this paper, we demonstrate a plasmonic ultrahigh narrowband perfect absorber, which realizes an absorption intensity of up to 99.99% in the near-infrared electromagnetic spectrum regime. Different dimensional effects on absorption properties are studied using computer simulation technology (CST) with finite element method (FEM) solver. For both transverse electric (TE) and magnetic (TM) polarization, the absorber shows high stability over a wide range of incident angles. This ultrahigh absorption is attributed to synergy effect of magnetic resonance and surface plasmon resonance. Furthermore, the sensitivity goes to 300 nm/RIU for various refractive indices of different analytes. Our proposed absorber working wavelength is very close to the communication window of information; therefore, except for remarkable sensing abilities, it can also be well utilized in optoelectronic applications like optical switching, amplifiers, and all-optical plasmonic modulators. Introduction Metamaterials and metasurfaces are artificially patterned materials consisting of periodically arranged subwavelengths "metaatoms." The meta-atoms design and geometry flexibility have branded metamaterial as designable permittivity and permeability materials [1][2][3][4] and make them widely used in electromagnetic (EM) induced transparency (EIT) [5][6][7], imaging [8][9][10], EM wave detections [11], ultrahigh sensitive sensors [12][13][14], EM invisibility cloaking [15], and solar cells [16][17][18][19]. Because of the huge investment in the solar energy harvesting systems, the plasmonic and metamaterial perfect absorbers have aroused great attention in various fields of technologies. The perfect absorbers are categorized into two, the narrow-and broad-band absorbers [20], where broadband absorbers have prominent applications in ultrashort pulse generation [21], photovoltaics [22], and thermal emitters [23], and narrowband absorbers have numerous applications in nonlinear photonics like optical modulators [24], and switches [25]. Moreover, the ultra-narrow band perfect absorbers in both visible and nearinfrared regions are also of great importance in refractive index sensors. Although some narrow-band absorbers have been studied [26][27][28][29], most of them have polarization or angle sensitivity or lower absorption intensity, which limit their potential applications. In this work, we propose an ultrahigh narrowband absorber with absorbance up to 99.99% close to the information telecommunication window. The absorber also reveals an outstanding performance in polarization and angle independence. The sensitivity can achieve 300 nm/RIU for various refractive indices of different analytes. Result and Discussion The schematic representation of the tunable hybrid metasurface is shown in Fig. 1 as a three-dimensional structure diagram. The device's topmost layer is made of goldalumina nanoring with a common air hole of radius Rin. The thicknesses of dielectric alumina Al 2 O 3 and plasmonic gold is h 1 = 30 nm and h 2 = 35 nm, with inner and outer radii Rin = 65 nm and Rout = 130 nm, respectively. The bottom gold film with thickness h 3 = 150 nm is used as a ground layer to guarantee total reflection and eradicate any incident EM wave transmission. The periodicity of the unit cell of metamaterial is P = 500 nm. Alumina's dielectric function is taken from Harris et al. [30], while noble gold is obtained from Johnson and Christy [31]. The whole structure is illuminated with plane electromagnetic waves of propagation constant K directed in negative z-axis. Under the FEM solver for the Maxwell equation, here CST microwave studio is employed to simulate and study the response of planner EM waves for both TE and TM polarization at various incident angles. For the meta-atom simulation purposes, the unit cell boundary conditions are employed periodically in the x and y direction and open (add space) in the z-direction. Evaluating the proposed hybrid metasurface's ability in both absorption A and reflection R, Fig. 1b gives the optical responses at normal incident of plane EM waves. The absorption intensity can reach 99.99% at 1490 nm wavelength, which is also very close to the communication window (1550 nm). In this device, the suppression of the reflectance R at normal incidence is because of the electric and magnetic dipole matching or, in other words, the matching of the impedance of the device with free space at the frequency of operation. The meta-device ensures near-perfect absorption of the EM waves in a specific regime from zero reflectance Fig. 1 a The schematic diagram of the proposed metasurface device consisting of goldalumina ring resonators on gold mirror backplane. b The absorption A and reflection R at normal incidence of the electromagnetic waves. Fig. 2 Electric and magnetic field distribution in both front and side view. a, b The absolute amplitude of electric field distribution at resonance peak 1490 nm while c, d the absolute value of H-field at resonance peak value 1490 nm of the absorption spectrum and transmittance. Figure 2 displays the absolute values of electric and magnetic field distribution of the proposed device at normal incidence to the plane of the device. As the device has a metal-insulator-metal (MIM) configuration, therefore, the Fabry Perot cavity in the alumina region has a high tendency of the field confinement (at appropriate thickness), as shown in Fig. 2a, b. The strongly enhanced magnetic field shown in Fig. 2c, d gives a clear view of magnetic resonance inside the alumina spacer and at the edges of the ring hole. To ensure the best matching of the impedance, we simulated various dielectric alumina thickness "h 1 ," and the results are plotted in Fig. 3a. The results show that 30 nm thick Al 2 O 3 can provide the optimal absorption results at the normal incidence of electromagnetic waves. For the spacer of a constant thickness (i.e., h 1 = 30 nm), the varied thickness of top gold nanodisk is studied from h 2 = 25 to 45 nm, and we can see Fig. 4 The absorption spectra at a different periodicity of the unit cell of the metasurface 35 nm is the optimum choice for 30 nm alumina, which makes almost completely matching of the impedance and improves the absorption intensity nearly to 100%. In both cases, we can observe that as the thickness decrease of either nanostructure, there is a red shift in the absorption peaks. The thinner dielectric enhances image charge distribution on the bottom layer due to the top metallic nanostructure's plasmonic resonance. Similarly, a thinner gold layer provides more EM transmission to the cavity region, and hence, the development of standing waves enhances the total absorption ability of the device. As our device consists of the gold-alumina disk with common inner and outer radii, therefore, we investigate in Fig. 3c, d the effect of both radii separately. From Fig. 3c, we observe that the absorption intensity for a solid disk or ring with R in = 0 nm reaches 90% at resonance peak 1300 nm of the EM wave spectrum, not as good as the optimum ring resonator. This amount of absorption is high enough to be utilized in different types of applications, like sensing and imaging. To further enhance the absorption intensity, we introduce an air hole in the nanodisk to make more chances of field confinement as elucidated for different inner radius values in Fig. 3c. As the inner radius increases, the plasmon resonance position shifts to a higher wavelength, and the absorption intensity increases up to 99.99% at R in = 50 nm for R out = 130 nm. After 50 nm, a further increase in inner radius reduces the absorption intensity, and at 100 nm, the intensity reduces to 70%, as shown in Fig. 3c. As for as the outer radius is concern, for constant R in , the lower R out gives lower absorption, but bandwidth is sharper, as shown in Fig. 3d. The optimum value of R out is 130 nm, and with the further increase, the absorption intensity decreases. The proposed device is further investigated for the unit cell's periodicity in Fig. 4 for five different values starting from 300 nm up to 700 nm. Among this 500 nm, center-tocenter separation of the periodic layer gives the optimum result of absorption. The plasmon resonance position at all periods is nearly the same with slightly different intensity of the absorption. In various applications, whether the absorber in a broad or narrow band range, the most important property is the less sensitivity to incidence angle and the polarization of the impinging EM waves. Here, we investigate the angular stability of the proposed absorbing device for a very broad range of the incidence angles (in Fig. 5) for both TE and TM polarization. Figure 5a represents the absorption at 1490 nm for different incidence angles when the incident waves have TE polarization, and the device is insensitive to the whole incident angles from 0° to 70°. Furthermore, there is no critical shifting of the resonance position in this wide range at all, and such property of the absorber is very suitable for biological sensing applications. Nearly the same response is observed for TM polarization, as shown in Fig. 5b. Here, the average absorption intensity is about 99.89% until 65 0 of the incidence angle, and after that, there is a minor decline to 97% at the same resonance wavelength. Thus, the absorber overall response in both TE and TM polarization at a broad range of the incidence angle is nearly independent. Because of such high absorption and tunability, this proposed configuration is highly recommended for many optical applications, especially for refractive index ultrahigh sensitivity. Finally, we are going to discuss the refractive index sensing ability of our ultrahigh plasmonic metamaterial absorber. The analytes here used with different refractive indices represent different concentrations of the glucose solution. Figure 5c, d displays the proposed metamaterial structure's sensing ability for various refractive indices from 1.302 to 1.352, and water is considered as a reference medium. As the refractive index changes a little bit, there is a very clear red shift in the absorption spectra's resonance peaks with no critical change in its intensity, as shown in Fig. 5c Fig. 6a gives a straight line, which indicates that there is no critical change in the absorption intensity and the plasmon resonance of the nanostructure is highly sensitive to the change in the surrounding medium. Sensitivity S n of refractive index based sensor can be calculated by using the following relations between wavelength resonance shift ∆λ and refractive index ∆n The sensitivity of our proposed structure reached 300 nm/RIU, as shown in Fig. 6b. We observe that the proposed device has a high sensitivity for very low refractive index values and high linearity in sensing. It is not far from practical realization and implementation due to the advanced electron beam lithographic technologies. Furthermore, our proposed absorber is insensitive to the angle S n = △ ∕ △ n Fig. 6 The resonance wavelength as a function of refractive index a wavelength shift versus the refractive index and b sensitivity of the device for the various analyst of refractive index having a range of 1.302-1.352 of incidence and polarization. Therefore, it has advantages over a prism-based sensor whose work is good only at oblique incidence. Conclusion In summary, we proposed and theoretically designed a perfect metamaterial absorber whose absorption intensity reaches 99.99% in the near-infrared range. The ultrahigh absorption of the device is attributed to the plasmon resonance of the plasmonic nanostructure and magnetic resonance at the spacer alumina region. The device geometrical effects have been investigated for various dimensions and found the optimum thicknesses, periodicity, and inner and outer radii at which ultrahigh absorption can achieved. Furthermore, the absorber ability at both TE and TM polarization was also evaluated at a very long range of the incidence angles and found no critical change in absorption intensity as well as the plasmon resonance position. Finally, the device is tested for the sensing property for analytes with different refractive indices and to get the sensitivity of the system which reached 300 nm/RIU. The proposed plasmonic device can also be utilized in the hot electron-based plasmonic modulators and switches because of the narrow perfect absorption close to the information's communication window.	2021-09-28T01:08:52.291Z	2021-07-15T00:00:00.000	{ "year": 2021, "sha1": "14a0970ab640e60e1b306ffebe2406121e9523ba", "oa_license": "CCBY", "oa_url": "https://www.researchsquare.com/article/rs-680454/latest.pdf", "oa_status": "GREEN", "pdf_src": "Springer", "pdf_hash": "a599cdb9ef5a931d55696a828f820e03334fab6c", "s2fieldsofstudy": [ "Physics" ], "extfieldsofstudy": [ "Computer Science" ] }
113199024	pes2o/s2orc	v3-fos-license	Empirical analysis of three wind simulation tools to support urban planning in early stages of design Este artigo apresenta a experiencia de utilização de três ferramentas de simulação de vento: Flow Design, ODS-Studio e um mini túnel de vento. O objetivo foi avaliar as vantagens e desvantagens de incorpora-las nas fases iniciais de projeto. Como tal, foram simulados três cenários urbanos e os resultados obtidos foram posteriormente comparados com as simulações efetuadas por Sousa (2014) com o programa CFD Phoenics. Como resultado o Flow Design é o mais intuitivo, mas menos preciso nos resultados. O ODS-studio é mais complexo para manipular e o mini túnel de vento o mais apropriado para analise de pontos especificos. Introduction In the last few years it has been well documented that urbanization modifies the natural wind circulation generating areas of turbulent airflow (Gandemer, 1978;Oke,1987).For Boris (2005) is important to analyse urban aerodynamics phenomenon to determine environmental air quality, wind pressure on buildings, urban heat islands, pedestrian comfort, and ambient noise level in the surrounding.Bittencourt and Cândido (2010) complement this statement reminding that understanding the wind dispersion through the city is essential to reduce the consumption of energy in indoor spaces with artificial mechanisms of comfort like the air conditioning. Despite natural ventilation be an important environmental parameter to be considered for the sustainability of the city it is still neglected by most architects and urban planners in Brazil.According to Melo and Barbirato (2011).the development of Brazilian cities is determined by its land use policy that defines different urban parameters, such as, land use rates, natural soil percentage, distance between constructions and building height.Some of this rules arose to ensure human comfort and healthier urban spaces with natural ventilation and light.However, the deficiency of studies involving urban prescriptions with the climatic data turns urban planning, paradoxically, responsible for the detriment of the welfare of the population.Assis (2006), reported that there was an increase in the number of studies that relate wind flow and urban development in Brazil, but in general the investigations are limited to a specific case study and there is lack of integration between the academic knowledge and the professionals that regulate and design the future of the city. Several tools can be applied to simulate wind phenomenon in digital or physical models, providing visual and numeric information that can be used by urban planners and architects' to support their decisions.Some of these tools are based on computational fluid dynamics (CFD) system and can elaborate very sophisticated calculations, such as, ANSYS CFX, Fluent, Phoenics and Flow 3D.(Cfd online, n.d.).These computer applications are frequently used by engineers to analyze a variety of fluids like water, gas, wind, smoke.Other simulation tools such as wind tunnels can also be considered a reliable technology for airflow analysis; however it is not very common to find laboratories with this type of equipment because is extremely expensive fully incorporated into the design process because the equipment is extremely expensive (Williams et al., 2015). These technologies are very advanced, but they demand some degree of knowledge to manipulate and to be able to interpret the results.In addition, a high initial investment is necessary, challenging their use in early stages of design process by architects (Kirkegaard, Hougaard, Staerdahl, 2008).This aspect along with a very time consuming simulation has been pointed out as the main reasons for the lack of integration of wind analysis in architectural design exploration (Salim and Moya, 2012;Moya, 2015).As a result, usually, when wind simulations are applied it is hired a team of experts to analyse the final design project missing the opportunity to develop more sophisticated strategies in the concept design stage (Kirkegaard et al., 2008). Considering these difficulties of workflow, some CFD simulation tools have been developed to be used by architects for rapid visualization of airflow, interoperability with other architectural design applications and easier set up of wind parameters, looking to promote the use of these tools in the initial stages of design (Williams et al., 2015).Naboni (2013) realized in his study that some of the world's leading architecture firms have been incorporated progressively this environmental simulation tools, into their design process, analyzing energy consumption, radiation, natural light effect and airflow movement."These simulation tools do not only have a significant impact on the way buildings are designed, but they are also changing the internal structures of architectural offices" (Naboni, 2013). With this in mind, some investigations have been done to develop methodologies and tools to improve the integration of CFD and wind tunnel in architectural practice.For instance, Prohasky, Moya, Watkins, Burry and Burry (2014), used wind-sensing platform inside a wind tunnel to evaluate the airflow around one building concept.Similarly, Salim and Moya (2012) analyzed wind in outdoor spaces to evaluate the performance of two digital simulation software's (CFD Vasari and CFD ANYS CFX) and two physical simulation devices ( a wind tunnel and a mini wind tunnel).The study of Williams et al. (2015) tested different designs solutions of porous screen for building façade using CFD Vasari and the Mini airflow tunnel.The investigation of Moya (2015) compared the use of Vasari, ODS-Studio and ANSYS CFX for the analysis of a windbreak shelter.These studies were all focused on the building design scale. The purpose of the current paper is to evaluate the use of two CFD programs and a mini airflow tunnel, for a brother scale of analysis, visualizing airflow in urban scenarios to support urban planner decision.The study was designed to evaluate the following aspects: 1.The tools currently developed for wind simulation in early design stages can generate a similar airflow of a sophisticated tool like CFD Phoenics at the pedestrian level? 2. What are the advantages and disadvantages in the use of these simulations tools to analyze urban scenarios?3. How is the workflow experience during the simulation? Methodology The research was developed in three different phases.In the first phase was selected three wind analysis tools that were specifically developed to be used during the design process and for educational purposes.The tools chosen were two computer fluid dynamic (CFD) programs (Autodesk Flow Design, and ODS-Studio) and one mini airflow tunnel developed by Moya and Prohasky (n.d.). In the second step, it was designated a case of study to be used as mean of comparison between all simulation tools.The area selected was simulated before by Sousa (2014) with a sophisticated and reliable CFD software (Phoenics), largely used by engineers and architects for professional and academic research.The aim was to use the results obtained in the previous investigation as analytic standard (see figure 1).The last step was to analyse the performance and the feedback of each wind simulation tool considering time and complexity of manipulation, identifying the advantages and disadvantages of their use.Moreover, the wind velocity results generated by the simulation were compared among all tools seeking for similar pattern of airflow.For this purpose, it was established sixteen points of comparison between all images obtained at two meters above surface, representing the pedestrian level of comfort.Considering that two of the tools studied show the simulation results in dynamic images of air movement (Flow Design and mini airflow tunnel) it was necessary to record one minute video to be then translated in an excel file and transformed into a value of wind speed average.All the information was transformed into a graphic chart for further evaluation. Digital Simulation tools for airflow Visualization of airflow inside and around buildings usually is accomplished through digital simulation using CFD software's or with scaled models inside a wind tunnel.In Brazil this two techniques have been widely applied in academic research.For instance, the work developed by Brandão (2009) and Leite (2010) used CFD Ansys CFX, Marques (2003) and Costa (2001), made use of CFD Phoenics, while Prata (2005) and Peregrino (2005) used a wind tunnel.All of them analysed the development of the city based on urban parameters established by a masterplan and a land use policy.Sousa (2014), during her master research also used the software Phoenics (Parabolic Hyperbolicor Elliptic Numerical Integration Code Series), created by CHAM (Concentration Heat and Momentum), a company founded in 1974.This software is used by engineers (civil, naval, nuclear, and chemical) architects, and designers because it allows the investigation of various processes involving fluid flow, heat and mass transfer, chemical reactions and / or combustion equipment (Cham, n.d.). Phoenics runs three sequential modules: the pre-processor, the processor and post-processor.The preprocessor is composed of VR-Editor and Satellite, which are virtual reality editors.The processor contains a system of subroutines, called EARTH, which generates the simulation model through complex numerical calculations (Costa, 2001).Finally, the visualization of the results is in the post-processor, which is composed of the VR Viewer, PHOTON and AUTOPLOT.Sousa (2014) used the VR Viewer, which allows observing the results in the form of vectors of velocity fields or isolines. This CFD tool has more than thirty years of development.The simulation used as analytic standard in the research was runned in the 3.6.1 version, from 2005, available in the laboratory of the Federal University of Alagoas (UFAL).It provides a very reliable demonstration of aerodynamic flow around buildings; however, this is an expensive software, difficult to be found in Universities or architectural firms.In addition, the version used did not have interoperability with other software.As a result, every step of the simulation had to be implemented in Phoenics, including creating the geometric model of the scenario.Therefore, it was very time consuming and difficult to integrate into usual architect workflows.Based on this reason, this research proposes to analyse other CFD tools available for free in the internet, trying to make a correlation between precision and usability.The following paragraphs describes each tool main characteristics: Flow Design is a virtual wind tunnel developed by Autodesk for rapid visualization of airflow around buildings and other objects.This tool was design to be easy and intuitive to use.The interoperability with other architectural softwares like, Autocad and Revit gives the chance to easily test different design concepts.Moreover, the visualization system was projected for an interactive 2D and 3D simulation of airflow movement (Autodesk, n.d.); ODS-studio is a platform developed to bring together many existing open-source tools.The aim was to provide rapid design and advanced technical analysis of building concepts in the early stages of design.The platform connects a modeling software called Blender (were users can create the 3D models or import from other architectural softwares, and also they can manipulate the digital mesh), with three environmental simulation programs: OpenFOAM (for CFD analysis of airflow), Radiance (for evaluation of daylight routine) and Energyplus (for calculation of energy consumption).To visualize Open FOAM simulation is also necessary the program Paraview.Furthermore, ODSstudio was developed to work in Linux, Mac and Windows operational systems.(ODS-Engineering, n.d.). Simulation tools of airflow -the mini airflow tunnel Since the beginning of 1890, architects have been using the wind tunnel for analysis wind flow inside and around buildings (Aynsley, 1977).Some of the techniques used for visualization of air movement are fog/smoke emission, floating particles or gas bubbles.Williams et al., (2015).Claims that "these techniques provide a qualitative approach to visualizing the pattern of wind vortices, separation flows, turbulence and changes of wind direction.Quantitative techniques typically involve the analysis of results after the experiment has taken place.Nevertheless, because of the expensive facilities required, this technology has not been fully incorporated into the design process or the education of architecture students" Bearing in mind this limitation, two researchers from SIAL (Spatial Information Architecture Laboratory) in RMIT University developed a portable and affordable version of a regular wind tunnel they call it the mini airflow tunnel project (Moya, Prohasky, n.d.).Is a modular wind tunnel made of MDF and acrylic panels that can be cut and assembled by anyone.The template is available for free download on the web site of the project, expanding the opportunities for its use.The test chamber can have three to nine modules depending on the user need and space.Each model has 0.9m wide, 0.9m high and 0.6m long, that can be taken apart and transported to other laboratories, when is no longer needed (see figure 2) (Moya, Prohasky, n.d.). Four fans are installed in the inlet zone to produce a continuous airflow of approximately 4m/s.In general, the wind tunnel is not designed to reproduce full scale wind flow conditions such as atmospheric boundary layer or turbulence intensity profiles.However, it presents a controlled and stable wind flow environment for reasonable observations of the dynamics in environmental parameters that, in situ, exist in a similar manner (Williams et al., 2015). This mini airflow tunnel was developed along with a microelectronic wind-sensing platform for immediate feedback of the simulation.Williams et al., (2015).explainsthat this method combines physical simulations with digital visualization where micro anemometer sensors are placed inside the test chamber to collect wind speed data in meters per second (m/s) temperature and humidity.This information is transmitted to an arduino board that transforms this data into an interactive visualization of airflow using a digital interface of Grasshopper script inside Rhinoceros software and with the Firefly plugin (Moya, Prohasky, n.d.).Author. Case study The case study chosen to assist the research is located in the city of Recife, northeast region of Brazil, where the climate is hot and humid and natural ventilation is considered the most important bioclimatic strategy to promote its citizens thermal comfort (Lamberts, Dutra e Pereira, 2004).Bittencourt and Cândido (2010) analyzed wind speed, frequency and direction based on data collected in the past 40 years in the weather station installed in Varzea neighborhood at 10 meters above the ground level.Sousa (2014) realized that the average wind frequency and velocity throughout the day should be most of the time satisfying for its citizens. However, due to the development of the city without deep analysis and consideration of airflow pattern around buildings is easy to find urban areas of discomfort for wind speed (sometimes too high other times too low).In order to understand airflow process in Recife Sousa (2014) analyzed a central area of the city that is speculated to be redevelop with the construction of a building complex comprising a set of skyscrapers.The project, named Novo Recife, will include new housing, office spaces and hotels.The study goal was to understand the impact of the new building complex in the natural wind ventilation and provide guidelines to designers and urban planners, according to different patterns of urban settlement. In the research, Sousa (2014) simulated the airflow in six scenarios of possible future development of the city.Nevertheless, for the purpose of this paper it was selected a part of the total area as case study since the focus during this investigation was not the area itself but the simulation tools used.In addition, part of the studied scenarios were excluded and the model was scaled to fit inside the mini airflow tunnel.Therefore, it was selected a part of three distinguished urban arrangements that varies on land use rates; distance between buildings and maximum height (see figure 3). The first scenario was created based on the Novo Recife proposal.This model contains two buildings of 140 meters (46 floors), one of 130 meters (43 floors) and another of 110 meters (36 floors), in the building basement was also added an 11 meters high garage building (equivalent to 3 floors) connecting all skyscrapers.In the second scenario the distance between the buildings was increased in the ground level, the garage was reduced to the projection of the skyscrapers base and the area that was taken from the garage was transferred to the buildings, increasing its height.Therefore, the second model has two buildings of 152 metros ( 50 only sixteen points were considered in the following step (comparing the simulation tools). Visualization step All tools have different visualization systems, but apart from the mini airflow tunnel all of them represent wind velocity (m/s) in a graphic color scale.Flow Design gives an immediate feed-back so, users can change the configurations of wind set up while they are observing the air movement.In addition, the program offers four types of graphic representation of airflow (plans, lines, tubes and particles), similar to ODS-studio visualization options, but ODS-studio results are generated only after a few hours of processing an advanced numerical calculation.Therefore, ODS-studio simulation tool is more time consuming and more complex, and in this sense less attractive to the user.The data collected by the wind sensors inside the mini airflow tunnel was transmitted in real time to the computer where it was possible to visualized wind speed by bubbles that increase or decrease their size according to air velocity. Experimentation set up With this investigation, it was possible to compare the use of three different tools of wind simulation according to three different steps: modeling, wind set up and visualization. Modelling step The models used in the CFD simulations were all drawn in AutoCAD 2D and manipulated on Sketchup for 3D volumetric definition.This was the most efficient way to create all models at the same time and it was possible because all software's have the option to import a vast variety of file formats making easier for professionals to analyze different design solutions developed using other programs.Additionally, Blender (the modular software that works with ODS-Studio) and Rhinoceros (the software necessary for the mini-airflow tunnel visualization) also offer options to import and develop tridimensional models.The mini airflow tunnel simulations required virtual scenarios and also physical models.The physical models were made of cardboard cut and assemble by hand at 1/500 scale of original to fit inside the wind tunnel and leave some area free of barrier around the model, providing enough space to the wind naturally circulate (see figure 2). Wind setup step Every tool has different levels of complexity to set up wind parameters.The manipulations of its definitions determine the quality of the results obtained in the simulations.In Flow Design user can control the domain size, wind direction, wind velocity and level of resolution.The graphic display is intuitive, easily understood by anyone without deep knowledge of airflow.On the other hand, ODS-Studio requires a period of training with tutorials, since it offers a wider range of options to control airflow configurations and different possibilities to refine the digital meshes for more detailed analysis. For the simulation using Flow Design and ODS-studio, the wind speed inlet was define at 2,8m/s with orientation from southeast, the same configuration used by Sousa (2014), in the previous research.While, the wind speed inside the mini airflow tunnel was around 4m/s, the only possible option, since the fans used in the study did not allow any power regulation. Inside the mini airflow tunnel were installed nine wind sensors aligned in a row.This sensors were positioned at two different distances from the models: at 3 cm and 14 cm (representing 15 meters and 70 meters of distance) to evaluate the wind flow turbulence at leeward of the projected buildings.So for each scenario was necessary to run two simulations in order to have the 18 different points of analysis proposed in the methodology.The sensors went through a calibration process, which consisted in verifying if all sensors were providing the same wind speed data before placing any model inside the test chamber.Unfortunately, after calibration the number of sensors were reduced, as one of them was not collecting data properly.As a result, Results Analyzing the outcome of the simulations the three scenarios exposed very distinct aerodynamic phenomena.The first scenario revealed a continuous area of low wind speed at leeward, due to the row effect produced by the garage building, and the acceleration of the wind speed on the streets aside of the buildings because of the corner effect and also because they were exposed to the main airflow. In the second scenario the reduction of the garage building footprint provided greater spacing between the building at ground level this caused a reduction of the area of wake effect at leeward.On the other hand, the increase on buildings height intensified the corner effect and triggered the acceleration of wind speed (above 6m/s) between the buildings, displaying that this is also not a good solution for this urban area.Finally, the third scenario with the decrease of building height and distance between buildings was the model that showed the highest turbulence wake and the most significant area of reduction in the air velocity at leeward of the buildings proving to be the worst solution for the ventilation in outdoor spaces affecting more houses from the surrounding area.regions far behind the buildings.In the first column the leeward side is shown as a region of two islands of low wind velocity by Phoenics software, while the other simulations presents a single low wind speed region.In the same way, the third scenario simulated with Flow Design is less evident the deflection of the wind at the corners of the second line of buildings.This deflection is clearer in the simulations conducted with Phoenics and ODS-Studio.The reasons for these differences are probably related with several parameters from each tool used (flow domain dimensions, quality of the digital mesh, etc.).This makes evident that a simplified tool like Flow Design is limited to the observation of areas closed to buildings, while a more complete tool like Phoenics is efficient to visualise regions far from the buildings.In figure 5, the maps of four analyses, it is possible to observe how the patterns of wind speed areas are similar in the vicinity of the buildings but with differences at the In figure 6 it is clear that the results between simulation techniques are more similar at regions closer to the buildings.The variation of wind intensity detected in each point by these tools shows patterns of higher wind velocities where the wind is 'funnelled' in-between the buildings.From the results presented it can be seen that these regions will have the greatest impact on pedestrian comfort.The magnitude of these accelerations of wind could be the evaluation criteria for the building designs. Taken as reference the simulation results from Phoenics the mini airflow tunnel shows the best agreement with the relative patterns in wind speed results in close range to the buildings (from point one to eight, located at 15 meters of distance from the buildings).This cannot be said for the values measured further down-wind from the buildings (from point nine to sixteen, located at 70 meters of distance from the buildings).where effects from atmospheric boundary layer (ABL) conditions may be more evident, thus emphasising the importance to simulate urban wind conditions with ABL velocity profile conditions. Moreover, this method of analysis was not efficient with Flow Design and ODS-Studio, because the plan view generated by the program is with perspective, making impossible to identify the exact same spot.This condition, is easily identified in figure 5 and 6, where the graphical chart demonstrates the wind speed data, making clear that this too softwares have a more resembling pattern of airflow because of this problem. Conclusion All tools revealed advantages and disadvantages in the interface systems they provide.Reinforcing the importance of studying this wind simulation tools for early design stages.Autodesk Flow Design had very limited options to set up wind parameters.For instance, the grid resolution is only defined by high or low standard and the user can not define the exact coordinates in the (x,y,z) axes for better evaluation of the model.The ODS-Studio displayed more options to regulate airflow however; it has a steep learning curve than the other tools because it requires basic knowledge in three different computer applications.The mini wind tunnel has some limitations in reproducing the boundary layer effect as well some restrictions in relation to the size of the model.Finally, the results generated by all simulation tools were not as precise as the ones shown on CFD program Phoenics however the overall air flow performance was similar. Considering the visualization system provided, two of the selected tools ran a rapid and interactive simulation of airflow (Flow design, and the mini airflow tunnel) and the other two (ODS-studio and Phoenics) generate a sequence of images of wind speed data after a few hour of processing.This means that Flow design and mini airflow tunnel show air movement and wind speed variations (m/s) that could be more interesting for educational enlightenment of fluid dynamic behavior than ODS-studio or Phoenics that demonstrate a final result of wind speed average.On the other hand, the two CFD tools (ODS-studio and Phoenics) show a more accurate final answer of the overall scenario and its details, while Flow Design lacks precision, the opposite of the mini airflow tunnel that provides the opportunity to visualize wind speed in the exact point were the sensor is located omitting the rest of the overall research site.In summary, all simulation tools demonstrate constraints so future work will seek to find others. Figure 1 : Figure 1: Methodology applied in the research.Author. Figure 2 : Figure 2: The mini airflow tunnel chamber, the connection of the sensors to Arduino board and the visualization on the computer. floors), one of 142 meters (47 floors) another of 122 meters (40 floors) and the garage buildings has 12 meters (4 floors).The third scenario has the double amount of buildings with approximately half the height of the second model.This means that this model has four buildings of 76 meters (25 floors) two buildings of 71 meters (23 floors) and another two buildings of 61 meters (20 floors). Figure 3 : Figure 3: The three scenarios used in the testes.Author. Figure 4 : Figure 4: Graphic representation of time and complexity of each tool during the three different phases of simulation.Author. Figure 5 : Figure 5: Representation of three scenarios simulated in Phoenics, Flow Design, ODS-Studio and mini wind tunnel.Author. Figure 6 : Figure 6: Graphic charts of wind speed in 16 different points.Author. Table 1 : Summary of experience using the simulation tools.Author.	2018-12-20T18:24:08.554Z	2015-11-01T00:00:00.000	{ "year": 2015, "sha1": "cd13dc81c86fd6bf6d77ea19e855e98741a6f4c8", "oa_license": "CCBY", "oa_url": "http://pdf.blucher.com.br/designproceedings/sigradi2015/80189.pdf", "oa_status": "HYBRID", "pdf_src": "MergedPDFExtraction", "pdf_hash": "cd13dc81c86fd6bf6d77ea19e855e98741a6f4c8", "s2fieldsofstudy": [], "extfieldsofstudy": [ "Engineering" ] }
233346820	pes2o/s2orc	v3-fos-license	Time series analysis with dynamic law exploration In this paper we examine, how the dynamic laws governing the time evolution of a time series can be identified. We give a finite difference equation as well as a differential equation representation for that. We also study, how the required symmetries, like time reversal can be imposed on the laws. We study the compression performance of linear laws on sound data. Introduction Intelligence is hard to define, and a lot of effort is made to bring up a sensible interpretation [1,2]. The significance of the definition itself is that it provides a framework of thinking, and as such it initiates the possible questions and thus the possible answers, too. As an example, in the field of artificial intelligence (AI) the study of the visual cortex of the brain motivated the creation of feed forward (deep) neural networks, in particular the convolutional networks [3]. Complemented by the mathematical notions like Bayesian analysis lead to numerous brilliant achievements [4]. In all types of tasks, like classification, regression, data compression and generation the present day neural networks have achieved considerable success. Nevertheless, AI systems face serious challenges. They are still vulnerable to unexpected errors, adversarial attacks [5], catastrophic forgetting [6]. Usually long range correlations are hard to maintain with current AI methods (see for example [7]). All these difficulties make it necessary to rethink the basic concepts of AI over and over again. As classification is concerned, the basic view of the intelligent actor is that as output it provides the (conditional) probability p(y\|x) that a certain class y can be observed in the input x [8]. One can give arguments why this approach is correct [8]. On the other hand human perception recognize classes not based on a single probability, but recognizing numerous features in the input. If all, or at least most of the features characteristic to a given class can be found in the input, then we confidentally can single out that class as a result of classification. If characteristic features are missing, or, even worse, we find features that contradict to the given class, then we exclude the input from that class. It is not a linear process, but a hard veto. In [9] we tried to lay the mathematical foundation of a theory that supports the intelligent perception through features. The goal of the present paper is to apply these ideas for the analysis of time series with linear features. Linear mappings have a long history, from support vector machines [10] to Linear Predictive Coding [11]. The presented method of identifying laws are close to these techniques, but uses a different line of thought. This paper is organized as follows. First we overview the relevant and irrelevant features of [9] in Section 2. Then we recap the treatment of time series with embedding 3. In the next Section (Section 4) we examine the properties of linear features. Then we apply our technique for musical data compression example in Section 5. Finally we close the discussion with a Conclusions section (Section 6). Relevant and irrelevant features Intelligence can be thought to be an ability to rephrase a complex input in terms of new notions that fit best to the input set. For example the notion of quadrupedality applies for most mammals, the carnassial dentition is characteristic for carnivora, while the property to be rided applies mostly for horses (or, eventually donkeys, ostriches, water buffalo etc.). But a useful propery is also that a given person has a name Betty; although there are a lot of Betties in the world, but in our closer environment probably there are just few. Thus a property is useful if either it characterizes a whole set, taking a constant value on each of its elements (for example the property of having caranssial dentition is "true" for carnivores and "false" for other animals); or if it can be used to make distinction between elements within the set (she is the only person among our friends called Betty). These useful properties can be called features, and, following [9], they are called relevant, if they are constant on a set, and irrelevant if they can tell elements of a set apart. For classification the relevant features are the most useful, since each relevant feature singles out a set, the common appearance of several relevant features single out the intersection of the given sets, thus restricting the interpretation of the input. On the other hand for compression (when we know the set) we need to remember only the irrelevant features; for example we can tell our friends to call Betty without needing to specify other properties of the person. The mathematical basis to deal with relevant and irrelevant features are published in Ref. [9], here we only recall the most important findings. In this approach the probability space of the intelligent actor is (X, F X , P), where X is the set of all possible inputs (considered to be a finite set), F X is its power set, and P is the uniform discrete probability measure (i.e. P(A ⊂ X) = \|A\|/\|X\|). The intelligent actor (e.g. a neural network) is a set of random variables {ξ i \|i = 1, . . . , N } over this probability space 1 , with joint probability distribution over p (A) (ξ 1 , . . . , ξ N ) over A. To represent a subset A ⊂ X we choose random variables that are independent over A, i.e. the joint probability distribution factorizes p (A) (ξ 1 , . . . , ξ N ) = p . Moreover these random variables should be either constant ξ i (x ∈ A) = σ i , or they have uniform distribution (i.e. the distribution is constant p (A) (ξ i = σ) = 1/\|A\|). The former are called relevant feature, the latter irrelevant feature. One can prove that any set can be completely described by giving appropriate number of relevant and irrelevant features [9]. In practice instead of determining the exact relevant features, we are usually satisfied with an approximation. This means that instead ξ i (x ∈ A) = σ i constant we can only ensure \|ξ i (x ∈ A) − σ i \| < ε for some small ε. A relevant feature restricts the possible input set to be in . Several relevant features restrict more and more the possible elements of the input set, and so it is the basic building block of classification. Irrelevant features, on the other hand, are useful to tell apart the elements of the studied subset of input. Indeed, if we denote the range ξ i (A) by R, then ξ −1 i (r), r ∈ R yields a pairwise disjoint subdivision of our subset. Using independent irrelevant features, these subdivisions will differ, and so with enough number of irrelevant features all elements of A can be uniquely described. Thus the task of compression is equivalent to find all the irrelevant features. Following this line of thoughts we can think about an intelligent actor as one that is capable to find the best features of the studied input set. Depending on the task one should seek the relevant or the irrelevant ones. In the following we discuss the application of the above ideas for the treatment of time series. We will mostly concentrate on relevant features, or laws, because it leads to a recursive description of the series, from where one can restore the complete series. Time series, laws and symmetries Time series are mathematically y : R → V functions, where V is a finite dimensional Hilbert space. To be able to handle this function, we have to prepare a finite, representative set from it. We will choose a finite set {t k \| k ∈ {1, . . . , K}} ⊂ R as initial points, and then we will sample the time series at finite past points starting from t k . In this way we generate the database We will need to find relevant features (laws), which satisfy This construct can be thought of as a neural network (c.f. Fig. 1), if we choose t k = k∆t. In this case we can build a network, where we have n memory slots for the past i = 1, . . . n elements, and we always put the new element into the first (0th) place, shifting all the others down. The content of the ith slot after k ≥ n steps is Y (k) i , and the single output is simply We may have restrictions on the actual form of F from requirements (prior knowledge) about the time series. For example we may want that the law is time reversal symmetric: this may come from the knowledge that basic physical laws are time reversal invariant, and if we want to describe only a steady state then the state also does not single out an arrow of time. Time reversal symmetry requires that if we feed the function in a reverse direction, the same law applies: time reversal symmetry: In some cases, like acoustic signals, or market prices we also expect that the law is independent of the amplitude. Therefore if x satisfies the law, then αx should also satisfy it: scaling symmetry: This requirement can be fulfilled if i.e. F is a scaling function with (an arbitrary) scaling dimension z. This requires that all activation functions are power functions. Linear mapping The simplest form of a law is a linear map, then the network is similar to the linear support vector machine, albeit we apply a different logics here. If the maps of the feature function are linear, then it is worth to apply the linear algebra notations. In particular we denote the input set as a matrix A linear map is equivalent to giving a vector w ∈ V n+1 for which where the first dot denotes the V scalar product, while in the second form we use the usual matrix notation. Applying it to our sample set we get K results by denoting it as a matrix multiplication. We remark here that we could use a more general Ansatz here. If we define a coefficient matrix R ∈ V n+1 ⊗ V n+1 , then the form of the law could be If R is invertible, then this simply redefines the w weights. Practical use of this form comes when R is diagonal and have zero and one elements. In this way we can leave out certain past points from the consideration, which can be a useful technique to reduce the number of effective weights. Technically this can be taken into account with Y ki → r i Y ki , and in the resulting weights we put w i → r i w i . Thus this general case can be completely embedded in the present discussion. This is equivalent to \|Y w\| = 0 with some norm. With the usual quadratic norm we may define where is the correlation matrix of the given problem (the same matrix appears in the Principal Component Analysis, PCA). This can be simply satisfied with w = 0. To avoid this trivial solution we require \|w\| = 1. To implement this constraint we realize that the minimum of χ 2 is zero, so instead of an equation, we can think the problem as minimization. Using the Lagrange multiplicator method we should minimize The solution of this requirement is eigenvalue equation. Since Y T Y is symmetric and positive definite, all eigenvalues are real and positive. Rewriting it to χ 2 we find This means that an exact law would be equivalent to find a zero eigenvalue in C P CA . In practice it is usually not possible to satisfy, we always have a (small) breaking of the law. We can describe it as where ξ k can be thought of as a random process. It is remarkable that This means that the variance of the accuracy of the law is given by the eigenvalue. We remark that although in the PCA analysis we use the same matrix, but there we seek the largest eigenvalue. The PCA logic is to find the direction that aligns with the data the most, i.e. the direction where the data vary the most, which corresonds to the largest eigenvalue of the PCA matrix. We, however, look for a direction that characterizes all the data, so we need the direction where the data differ the least. This corresponds to the smallest eigenvalue of the PCA matrix. In the language of [9], the eigenvectors belonging to the largest eigenvalues are the irrelevant, the ones belonging to the smallest eigenvalues are the relevant features. In realistic situations when we analyze the eigensystem of the PCA matrix we find a spectrum, similar to what can be seen in Fig. 4. As we see, the PCA matrix has a few large(ish) eigenvalues and a lot of small ones. If we went on to determine longer laws, we would have determined more and more small, i.e. relevant eigenvalues. This is a very general observation for natural processes: we have much more relevant than irrelevant features [9]. Any of the relevant features can be used as a law. A law over real valued time series can be organized as a recursion. First, using the definition of Y (7) we can rewrite the law (12) as Function class compatible with linear law Now choosing t k = k∆t we can define a series y k = y(k∆t). It satisfies then Being a linear recursion, sum of solutions is also a solution. For a particular solution we try an Ansatz Rewriting this Ansatz into (22) we find That means that the solutions are roots of the nth order polynomial. This equation yields n solutions, that are either real, or complex, appearing in conjugated pairs (so altogether n real data). Let us denote the roots as {q a \| a = 1 . . . n}. With all these roots we can fully fit the n initial conditions of the recursion, thus it yields the complete solution. As a side remark we can establish a relation between ws and qs from the condition that Thus the time series with generic initial conditions that satisfies the linear law can be written as where the c a coefficients are complex numbers. With the notation we can write Although originally it is valid for discrete times, but it is easy to verify that the above form is a solution for any starting time. This form gives the definition of that function class that is compatible with a linear law. To give still another view, we observe that Therefore (30) satisfies n a=1 (∂ t + α a )y(t) = 0. Thus y(t) satisfies an nth order linear differential equations with constant coefficients. A corollary is that in the linear case a law can be given in multiple ways: • give an w ∈ R n+1 weight function, then the element satisfy the law lie in the orthogonal subspace: y T w = 0 • give a linear nth order recursion: y 0 = − n i=1w i y i , wherew i = w i /w 0 • give the q a roots of the polynomial P(x) = n i=1 w i x i : the weights can be reconstructed from • give the corresponding differential equation. The recursive form of the law (22) can be generalized to nonlinear laws, then the law should be formulated as a (nonlinear) recursion. Stability of a recursion Let us consider the asymptotic behavior of the recursion, i.e. for large t. If the real part of α > 0, then for large t it yields an exponentially growing solution, that is to be excluded. If α < 0, then for long times it is vanishing, and we will not observe it. Therefore for long times only the imaginary α survive. Since y(t) must be real for all t, the roots e iωt and e −iωt both appear in the series (with complex conjugate coefficients). This means that in the differential equation form we find Therefore the differential equation is time reversal invariant. What does it imply for the weights in the discretized law? We should consider the relation (27) which connects the roots with the coefficients. Now the roots appear in pairs: Here the coefficient of the x 2 and 1 terms are the same. This will be inherited to all of the weights. To see it, we observe that the polynomial coming from the above pairs satisfies (using Z a = −2 cos(ω∆t)) where is the number of frequencies (i.e. n = 2 ). If we consider the same polynomial in weight representation, we can write (note that 2 = n) This is equal to P(x), if This also means that the recursion forward and backward happens with the same law: this is the manifestation of time reversal symmetry. This is the property that can be easily generalized to nonlinear recursions: we should require that the forward and backward iteration happens with the same law. Example: a single sin To see an example, consider a simple function: This satisfies (∂ 2 t + ω 2 )y(t) = 0, so we know that it can be represented a law consisting of 3 terms (it can be longer, too, see later). We could use the relation between the root and weight representation of a polynomial (27) to find the same result. Example: sum of two sin functions Let us consider the function This satisfies The corresponding polynomial reads We see that the composite law maintains time reversal symmetry, as it should. Data compression using linear laws Laws, as we have seen, represent a function class, and so they can be used to represent data with appropriate coefficients. Linear laws include polynomials, exponentials and harmonic functions (and also their mixture or product of them), and thus they can replace existing methods with broader possibilities. If the data support our function class, then the law represents it for a long time, in extreme cases all points. We remark that these include also ax n e −bx like transient or saturation processes besides other examples. For the description an actual data series which (approximately) satisfy a law, we need also the initial conditions for the recursion. Given the law, i.e. the function class, then each initial condition represents a different element of this class. These are useful to tell apart the elements of the class, and so they correspond to the irrelevant quantities in the language of [9]. Coefficient fitting We can determine the initial conditions in several ways. The first is to determine the q a roots of the corresponding polynomial. Then we know the the data can be described by the function (30), we just have to fit the coefficients. In order to do that we introduce the matrix and use least square method This means that the coefficient vector comes from the solution of Once we have the coefficients, we can propose a continuous function that approximates data the best. Another method is that we use the recursion directly to describe the data. The problem is that if our law is not exact, i.e. the data are noisy, then we can not just take simply the first n terms to start the recursion. We should start from an optimalȳ k starting values in order to achieve the best overall fit to the data. The best way to find these optimal values is to define a series basis starting from initial condition y k = δ k with given , for = 0, . . . , n − 1. Using the recursion (22) we will have a series s ( ) . Then we require Introducing a matrix again the initial condition vectorȳ comes from the solution of In case of nonlinear laws we have to use the recursive form of the law to determine the best description of the data. There the sum of solutions is not a solution and so the initial conditions should be fitted. Results Usually the accuracy of a single law will decrease with increasing data length. An n+1 long linear law means n roots and n amplitudes, together 2n parameters: it can therefore describe a N data = 2n long data vector. For longer vectors we unavoidably loose precision, the question is only, what is the rate of this loss. To numerically characterize the loss, we can plot the accuracy A versus the compression rate R, defined as where X is the data which is N data long, X sym is the simulated data using the n + 1 long law. We always get A = 1 for R = 1. An extreme situation is when the data exactly support our function class: there A = 1 for any R. The other extreme is the random data series, where there is no law at all, see Fig. 5.2. There we can observe that A = R −3/2 is the average trend, independently on the number of laws we use. In all intermediate situation we lose precision, the rate depends on the actual data. As an example we study the sound of a flute, and determine the accuracy of the representation. It may remain close to a perfect desciption for a large compression rate, as we can see from Fig. 5.2. As it can be seen, even for very large compression rates of 40, the best laws may provide reasonable 90% accuracy. It is not the level of HiFi acoustic experience, but for recognizing a sound it is perfectly usable. Usually a the best accuracy for a given compression rate comes from an appropriate choice of the number of terms in the law and ∆t. In case of clear acoustic data, like the sound of the flute, one can reach a large compression, while for more noisy situation, like in percussion instruments, it is hard to increase the compression rate too much, c.f. Fig. 5.2. The problem with percussion instruments (even if the are imitated by human voice, like in Bobby McFerrin's "Don't worry" song) is that they consist mainly of "switch-on" effects, when the device starts to emit sound. These are sudden and short burst, which means that the laws change frequently. The linear law method we can only give a homogeneous law, so it will have difficulty to represent these musical instruments. Nevertheless, the nearly 97% accuracy in the randomly chosen part of the song, at compression rate about 4, is again not a bad description. It can also be noted that for small compression rate (up to about 1.5), the compression is parctically lossless. This property can be used to lossless compression techniques. Comparison with Linear Predictive Coding technique Linear Predictive Coding (LPC), applied in many areas of sound file compression (e.g. FLAC [12]), uses a similar formalism. There are, however some important differences. LPC uses a recursive form with w 0 = −1 prescription for the weight, which also means that the w n>0 coefficients must be determined by matrix inversion. In our case the weights are eigenvectors belonging to the smallest eigenvalue, which is more stable procedure. A further difference is that we only determine the dominant frequencies (relevant features), the actual coefficients of the recursion comes with an extra coefficient fitting step. Another difference is that LPC wants to predict the next data point, while we give a continuous model that describes it. As a consequence we are not forced to use ∆t = 1/sampling rate, which extends our possibilities. Usually the best compression comes from choosing not the smallest ∆t. In addition, a law can be formed to satisfy symmetry requirements, like time translational symmetry. Conclusions The goal of this paper was to demonstrate how the idea of finding laws in input data sets is working in the linear case. We used time series to analyze, in particular sound tracks, embedded in finite n + 1 dimensions. Linear laws can be represented by an n + 1 dimensional vector that is perpendicular to all embedded data. These vectors can be found as the eigenvectors belonging to the zero eigenvalue of the corresponding PCA problem. In realistic case the data do not support the law exactly. In this case we choose the smallest eigenvalue, then we know that the variance of the law breaking term is equal to the value of this eigenvalue. As it was shown, the function class that satisfies a linear law consists of the solutions of nth order differential equations with constant coefficients. We determined the relation between the law and the coefficients, which allows us to give a continuous function that describes the data. Laws can be constrained to satisfy certain requirements. Long time stability, for example, requires time reversal symmetry, which is manifested in the weights as a symmetry under reversing it. These ideas are checked to describe realistic data, in this paper we considered music instruments. We plotted the accuracy of the description of the data vs. the compression ratio. Depending on the data the efficiency of linear laws are different. In the lack of laws, like in random data set, the accuracy decreases very fast. The more the data support a linear law, the longer a high precision maintains. Among our examples the best compression rate could be achieved in case of piano music. A lot of finding of the linear laws can be taken over for the nonlinear case. In particular in neural networks the last layer, which is usually a fully connected one, can be substituted by a linear law finder. Then in the result we will find several optimized features, corresponding to the corresponding laws. With more relevant features hopefully the classification tasks can be performed more reliably, since it is more and more difficult to find such adversarial directions that satisfy all the laws, but still takes us out of the given class. A Relation between root finding of a polynomial and diagonalization Consider a recursion This recursion can be written two ways. First we may assume that the solution of the recursion is The M matrix is diagonalizable M = U −1 DU , then with Z = U Y we find Thus The original y N is a linear combination of these expressions, any of this is a solution of the recursion. Therefore the roots are equal to the eigenvalues	2021-04-23T01:16:05.577Z	2021-04-22T00:00:00.000	{ "year": 2021, "sha1": "945be56e14f4940225d1b3f05e99b8f36a7c2d57", "oa_license": null, "oa_url": null, "oa_status": null, "pdf_src": "Arxiv", "pdf_hash": "945be56e14f4940225d1b3f05e99b8f36a7c2d57", "s2fieldsofstudy": [ "Mathematics" ], "extfieldsofstudy": [ "Computer Science" ] }

Subsets and Splits

No community queries yet

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