Search

For poisoned patients, ambulance services may be the first point of contact for medical attention. With limited training in toxicology, ambulance services are encouraged to contact the Poisons Information Centre (PIC) for advice. This study aims to characterise referrals to a PIC from a state ambulance service with the purpose of improving information delivery and efficient use of these services.This was a retrospective observational series of referrals to an Australian state PIC from ambulance staff from 1 January 2020 to 31 December 2020. Referrals were identified through the PIC Pharmhos database where the call originated from either a paramedic or emergency dispatch officer. Call reports were reviewed to extract data on patient demographics, exposure details and advice provided by the PIC.There were 1537 calls regarding 1420 poisoning exposures over the 12-month period, with 117 (7.6%) follow-up calls, representing 4.1% (1537/37835) of total calls to the PIC. Initial calls originated from paramedics in 999/1420 (70.4%) referrals, with dispatch officers referring 421/1420 (29.6%). Paediatric patients aged15 years were involved in 492/1420 (34.6%) exposures with the commonest age range being 1-4 years. Most referrals involved pharmaceuticals exposures (756/1420 [53.2%]) followed by chemicals (557/1420 [39.2%]) and drugs of abuse (69/1420 [4.9%]). The commonest agents involved were paracetamol followed by quetiapine and sertraline. The PIC advised no treatment following benign exposures in 617/1420 (43.5%) calls, first aid measures in 333/1420 (23.5%) calls, supportive measures in 339/1420 (23.9%) calls and specific treatment in 32/1420 (2.3%) calls. Referral to the hospital was advised in 761/1420 (53.6%) calls, the majority of these were following deliberate self-poisonings (428/1420 [30.1%]).Ambulance staff commonly contact the PIC following benign exposures where no treatment is required. Ambulance referral to a PIC following suspected poisonings may have a role in preventing unnecessary transfer to hospital in poisoned patients.

Mitochondria are required for energy production and even give brown adipose tissue (BAT) its characteristic color due to their high iron content and abundance The physiological function and bioenergetic capacity of mitochondria are connected to the structure, folding, and organization of its inner-membrane cristae. During the aging process, mitochondrial dysfunction is observed, and the regulatory balance of mitochondrial dynamics is often disrupted, leading to increased mitochondrial fragmentation in aging cells. Therefore, we hypothesized that the morphological changes in BAT mitochondria and cristae observed across aging would reflect alterations to energy dynamics. We developed a quantitative three-dimensional (3D) electron microscopy approach to map cristae network organization in mouse BAT to test this hypothesis. Using this methodology, we investigated the 3D morphology of mitochondrial cristae in adult (3-month) and aged (2-year) murine BAT tissue via serial block face-scanning electron microscopy (SBF-SEM) and3D reconstruction software for manual segmentation, analysis, and quantification. Upon investigation, we found increases in mitochondrial volume, surface area, and complexity and decreased sphericity in aged BAT. Cristae examination revealed decreased cristae volume, perimeter, and complexity, with cristate appearing more heterogeneous. Overall, these data define the nature of the mitochondrial structure in BAT, showing loss of cristae definition across aging, suggesting alterations in functionality and bioactivity.Graphical AbstractWorkflow overview of serial block facing-scanning electron microscopy (SBF-SEM), data segmentation, and 3D analysis of mitochondria using Amira software in murine BAT.

Glucagon has emerged as a key regulator of extracellular amino acid (AA) homeostasis. Insufficient glucagon signaling results in hyperaminoacidemia, which drives adaptive proliferation of glucagon-producing α cells. Aside from mammalian target of rapamycin complex 1 (mTORC1), the role of other AA sensors in α cell proliferation has not been described. Here, using both genders of mouse islets and glucagon receptor (gcgr)-deficient zebrafish (Danio rerio), we show α cell proliferation requires activation of the extracellular signal-regulated protein kinase (ERK1/2) by the AA-sensitive calcium sensing receptor (CaSR). Inactivation of CaSR dampened α cell proliferation, which was rescued by re-expression of CaSR or activation of Gq, but not Gi, signaling in α cells. CaSR was also unexpectedly necessary for mTORC1 activation in α cells. Furthermore, coactivation of Gq and mTORC1 induced α cell proliferation independent of hyperaminoacidemia. These results reveal another AA-sensitive mediator and identify pathways necessary and sufficient for hyperaminoacidemia-induced α cell proliferation.

Context: There is little research examining clinician adherence to specialist toxicological phone advice. Efforts to improve adherence should be sought to optimise the management of poisoned patien...

Glucagon stimulates glycogenolysis and gluconeogenesis in liver resulting in increased hepatic glucose output. Thus, interrupting glucagon signaling in liver reduces blood glucose levels in animal models and individuals with diabetes. However, interrupting glucagon signaling also leads to hyperaminoacidemia and proliferation of glucagon secreting alpha cells, evidence of an endocrine liver-alpha cell axis. Among the high amino acids, we previously identified high glutamine levels are required for alpha cell proliferation. We recently found that glutaminase (GLS) is expressed in alpha cells at much higher levels than any other pancreatic endocrine cells. Additionally, mouse pancreatic islets cultured in high amino acid containing media and treated with the small molecule inhibitor of GLS (CB839) have reduced alpha cell proliferation (Vehicle: 7.6 ± 2.2%, 0.1µM CB839: 0.11 ± 0.05%**, 10µM: 0.02 ± 0.01%**; n=3-4, **p < 0.01). To further characterize the role of glutamine metabolism in glutamine-induced alpha cell proliferation, GLS activity was specifically knocked out in mouse alpha cells by breeding Gcg-CRE ERT2 with Glsflox mice (aGlsKO). Targeted deletion of Gls expression in most alpha cells was confirmed by staining for GLS and glucagon (aGlsWT: 95% GLS+, aGlsKO: 18% GLS+). Mice were treated with a glucagon receptor monoclonal antibody (GCGR mAb/mAb-4) for 10 days or with isotype control IgG. aGlsWT mice treated with GCGR mAb have robust alpha cell proliferation while aGlsKO mice treated with GCGR mAb for 10 days showed a 4.5-fold decrease in alpha cell proliferation and were similar to mice treated with the control IgG (aGlsWT Ab: 15.7 ± 3.6% alpha cell proliferation, aGlsWT IgG: 1.0 ± 0.4%***, aGlsKO Ab: 3.4 ± 0.6%**; n=4-6, ***p < 0.001). In addition, aGlsWT mice treated with GCGR mAb showed a robust activation of mTOR in alpha cells as measured by phospho-S6 expression while this was lost in aGlsKO mice treated with GCGR mAb. Induced expression of the mTOR-dependent glutamine transporter SLC38A5 was also lost (aGlsWT Ab: 25.0 ± 4.8% alpha cell SLC38A5 expression, aGlsWT IgG: 0.40 ± 0.27%***, aGlsKO Ab: 3.7 ± 3.5%***, n= 2-4, ***p < 0.001). Similar results on proliferation and mTOR activation in alpha cells were observed after 4 weeks of GCGR mAb treatment. Together these data suggest a critical role for glutamine metabolism via glutaminase in alpha cell proliferation. Presentation: Saturday, June 11, 2022 1:36 p.m. - 1:41 p.m., Sunday, June 12, 2022 12:30 p.m. - 2:30 p.m.

Glucagon has emerged as the main regulator of extracellular amino acid homeostasis. Insufficient glucagon signaling results in hyperaminoacidemia, which drives adaptive proliferation of glucagon-producing α cells. Aside from mammalian target of rapamycin complex 1 (mTORC1), the role of other amino acid sensors in the α cell proliferation has not been described. Here, using gcgr-deficient zebrafish and cultured mouse islets, we show that α cell proliferation requires the calcium sensing receptor (CaSR) and downstream extracellular signalregulated protein kinase (ERK1/2). Inactivation of casr dampened α cell proliferation, which can be rescued by re-expression of CaSR or activation of the downstream Gq, but not Gi, signaling in α cells. CaSR was also unexpectedly necessary for mTORC1 activation in α cells. Furthermore, co-activation of Gq and mTORC1 induced α cell proliferation independent of hyperaminoacidemia. These results reveal another amino acid sensitive mediator, and identify major pathways necessary and sufficient for hyperaminoacidemia-induced α cell proliferation.

Glucagon and its partner insulin are dually linked in both their secretion from islet cells and their action in the liver. Glucagon signaling increases hepatic glucose output, and hyperglucagonemia is partly responsible for the hyperglycemia in diabetes, making glucagon an attractive target for therapeutic intervention. Interrupting glucagon signaling lowers blood glucose but also results in hyperglucagonemia and α-cell hyperplasia. Investigation of the mechanism for α-cell proliferation led to the description of a conserved liver–α-cell axis where glucagon is a critical regulator of amino acid homeostasis. In return, amino acids regulate α-cell function and proliferation. New evidence suggests that dysfunction of the axis in humans may result in the hyperglucagonemia observed in diabetes. This discussion outlines important but often overlooked roles for glucagon that extend beyond glycemia and supports a new role for α-cells as amino acid sensors.

Interrupted glucagon signaling reduces hyperglycemia. However, interrupted glucagon signaling increases serum amino acid and glucagon levels and stimulates amino acid-dependent alpha cell proliferation as part of an endocrine feedback loop, the liver-alpha cell axis. Our previous studies showed that glutamine and SLC38A5, a transporter of glutamine and other amino acids, are required for amino acid-stimulated alpha cell proliferation. To understand the mechanism of glutamine-stimulated proliferation, we analyzed published transcriptomics data for enzymes involved in the glutamine metabolism pathway and found 3-fold greater gene expression of glutaminase (Gls/GLS) in mouse and human alpha cells than in other pancreatic endocrine cells, while Gls2/GLS2 expression was extremely low in islet cells. Immunohistochemical analyses of mouse and human pancreata showed greater protein expression of GLS in islets versus the surrounding exocrine tissue. Within islets, both human and mouse alpha cells had higher expression than beta cells, suggesting that alpha cells may have increased capacity for glutamine metabolism. This expression pattern was also observed in individuals with either type 1 or type 2 diabetes. To examine the role of GLS activity in alpha cell proliferation, we treated the αTC1-6 mouse glucagonoma cell line with a competitive GLS inhibitor CB-839. GLS inhibition resulted in a 1.6-fold decrease in cell count over an 8-day cell growth assay (DMSO: (9.7 ± 1.3)x105, 10 μM CB-839: (6.2 ± 1.4)x105; n=2-4). In isolated mouse islets cultured in high glutamine and amino acid containing media, GLS inhibition inhibited alpha cell proliferation (DMSO: 7.6 ± 2.2%, 0.1 μM CB-839: 0.11 ± 0.05%**, 1.0 μM: 0.04 ± 0.02**, 10 μM: 0.02 ± 0.01%**, n=3-4, p Disclosure W. Siv: None. K. Sellick: None. A. Reuter: None. M. Shou: None. J. Stanley: None. E. Spears: None. W. Chen: None. D. Dean: None. Funding National Institutes of Health (DK117969, DK117147, DK020593); JDRF (2-SRA-2016-149-Q-R)

Significance Both type 1 and type 2 diabetes are associated with reduced β-cell mass or function, resulting from decreased proliferation and increased apoptosis. Understanding the signals governing β-cell survival and regeneration is critical for developing strategies to maintain healthy populations of these cells in individuals. Both forms of diabetes are associated with hyperglucagonemia and an increased plasma glucagon:insulin ratio. Glucagon excess contributes to metabolic dysregulation of the diabetic state and glucagon receptor antagonism is a potential target area for the treatment and prevention of diabetes. Our studies presented here suggest that blockade of glucagon signaling lowers glycemia in mouse models of type 1 diabetes while enhancing formation of functional β-cell mass and production of insulin-positive cells from α-cell precursors., We evaluated the potential for a monoclonal antibody antagonist of the glucagon receptor (Ab-4) to maintain glucose homeostasis in type 1 diabetic rodents. We noted durable and sustained improvements in glycemia which persist long after treatment withdrawal. Ab-4 promoted β-cell survival and enhanced the recovery of insulin+ islet mass with concomitant increases in circulating insulin and C peptide. In PANIC-ATTAC mice, an inducible model of β-cell apoptosis which allows for robust assessment of β-cell regeneration following caspase-8–induced diabetes, Ab-4 drove a 6.7-fold increase in β-cell mass. Lineage tracing suggests that this restoration of functional insulin-producing cells was at least partially driven by α-cell-to-β-cell conversion. Following hyperglycemic onset in nonobese diabetic (NOD) mice, Ab-4 treatment promoted improvements in C-peptide levels and insulin+ islet mass was dramatically increased. Lastly, diabetic mice receiving human islet xenografts showed stable improvements in glycemic control and increased human insulin secretion.

Interruption of glucagon signaling by glucagon receptor gene inactivation or treatment with a glucagon receptor blocking monoclonal antibody (GCGR-Ab) stimulates alpha cell proliferation and expansion. Surprisingly, pancreas weight is also increased in mice with GCG gene inactivation. To investigate mechanisms linking glucagon signaling and pancreas size, we treated C57BL/6J mice with GCGR-Ab or a control IgG. Eight weeks of treatment (N=4-15 mice/treatment) increased absolute pancreas weight (0.42±0.013 vs. 0.32±0.008; p Disclosure C. Dai: None. A. Bradley: None. K.C. Coate: None. G. Poffenberger: None. E. Spears: None. J.J. Wright: None. S. Shrestha: None. D. Dean: None. M. Brissova: None. A.C. Powers: None. Funding National Institutes of Health (DK117147)

Interrupting glucagon signaling (IGS) elicits robust alpha cell proliferation in an amino acid-dependent manner in mouse and human pancreatic islets. To determine whether IGS triggers beta cell proliferation in mouse and human islets in vivo, we quantified the percentage of Ki67-positive beta cells in adult mouse and human islets after IGS. Wild type (WT) mouse islets transplanted into liver-specific glucagon receptor (GCGR) knockout (LKO) mice exhibited a 4.1-fold increase in beta cell proliferation (WT to LKO: 0.98±0.34%, WT to WT: 0.24±0.19%, P=0.008, n=8). Likewise, treatment of C57BL6 mice with an anti-GCGR monoclonal antibody (GCGR-Ab) elicited a 3.6-fold increase in beta cell proliferation (GCGR-Ab: 1.09±0.21%, control (CTR): 0.30±0.14%, P=0.004, n=10). Given that the cationic amino acid transporter, Slc7a2, is one of the most highly expressed amino acid transporters in pancreatic islets and required for IGS-induced alpha cell proliferation, we assessed GCGR-Ab-induced beta cell proliferation in adult mice with inactivation of Slc7a2. There was a 3.2-fold increase in beta cell proliferation in WT mice treated with GCGR-Ab (GCGR-Ab: 1.45±0.46%, CTR: 0.46±0.15%, P=0.001, n=10). Notably, this effect was absent in Slc7a2 knockout mice (GCGR-Ab: 0.35±0.35%, CTR: 0.41±0.28%, n=10), suggesting that GCGR antagonism induces beta cell proliferation in an amino acid-dependent manner in mice. Furthermore, transplanted human islets from normal donors (ages 10-55, n=8) into immunodeficient mice exhibited a 2.8-fold increase in beta cell proliferation (GCGR-Ab: 0.61±0.60%, CTR: 0.22±0.21%, P=0.05, n=8) following GCGR-Ab treatment. These data indicate that IGS stimulates beta cell proliferation in mouse and human islets and that in mouse islets, this is Slc7a2-dependent. Disclosure K.C. Coate: None. E. Spears: None. C. Dai: None. S. Wisniewski: None. G. Poffenberger: None. L.D. Shultz: None. D.L. Greiner: None. D.J. Drucker: Advisory Panel; Self; Merck Sharp & Dohme Corp. Consultant; Self; Eli Lilly and Company, Intarcia Therapeutics. Research Support; Self; Novo Nordisk Inc. H. Yan: Stock/Shareholder; Self; REMD Biotherapeutics Inc. K. Sloop: None. A.C. Powers: None. D. Dean: None. Funding National Institutes of Health (DK117147)

Selective inhibitors of sodium glucose cotransporter-2 (SGLT2) are widely used for the treatment of type 2 diabetes and act primarily to lower blood glucose by preventing glucose reabsorption in the kidney. However, it is controversial whether these agents also act on the pancreatic islet, specifically the α cell, to increase glucagon secretion. To determine the effects of SGLT2 on human islets, we analyzed SGLT2 expression and hormone secretion by human islets treated with the SGLT2 inhibitor dapagliflozin (DAPA) in vitro and in vivo. Compared to the human kidney, SLC5A2 transcript expression was 1600-fold lower in human islets and SGLT2 protein was not detected. In vitro, DAPA treatment had no effect on glucagon or insulin secretion by human islets at either high or low glucose concentrations. In mice bearing transplanted human islets, 1 and 4 weeks of DAPA treatment did not alter fasting blood glucose, human insulin, and total glucagon levels. Upon glucose stimulation, DAPA treatment led to lower blood glucose levels and proportionally lower human insulin levels, irrespective of treatment duration. In contrast, after glucose stimulation, total glucagon was increased after 1 week of DAPA treatment but normalized after 4 weeks of treatment. Furthermore, the human islet grafts showed no effects of DAPA treatment on hormone content, endocrine cell proliferation or apoptosis, or amyloid deposition. These data indicate that DAPA does not directly affect the human pancreatic islet, but rather suggest an indirect effect where lower blood glucose leads to reduced insulin secretion and a transient increase in glucagon secretion.

Posttransplantation diabetes mellitus (PTDM) is a common and significant complication related to immunosuppressive agents required to prevent organ or cell transplant rejection. To elucidate the effects of 2 commonly used agents, the calcineurin inhibitor tacrolimus (TAC) and the mTOR inhibitor sirolimus (SIR), on islet function and test whether these effects could be reversed or prevented, we investigated human islets transplanted into immunodeficient mice treated with TAC or SIR at clinically relevant levels. Both TAC and SIR impaired insulin secretion in fasted and/or stimulated conditions. Treatment with TAC or SIR increased amyloid deposition and islet macrophages, disrupted insulin granule formation, and induced broad transcriptional dysregulation related to peptide processing, ion/calcium flux, and the extracellular matrix; however, it did not affect regulation of β cell mass. Interestingly, these β cell abnormalities reversed after withdrawal of drug treatment. Furthermore, cotreatment with a GLP-1 receptor agonist completely prevented TAC-induced β cell dysfunction and partially prevented SIR-induced β cell dysfunction. These results highlight the importance of both calcineurin and mTOR signaling in normal human β cell function in vivo and suggest that modulation of these pathways may prevent or ameliorate PTDM.

Interrupting glucagon receptor signaling results in hyperaminoacidemia, which stimulates alpha cell hyperplasia. Arginine, a potent secretagogue for both insulin and glucagon, is required for hyperaminoacidemia-stimulated alpha cell proliferation. To investigate the role of arginine transport, we compared the expression of amino acid transporters in published datasets of sorted mouse and human islet cells and found that Slc7a2, cationic amino acid transporter, is the most highly expressed amino acid transporter in both mouse and human alpha cells and that its expression is three-fold higher in alpha cells than in beta cells. Following glucose/arginine injection, global Slc7a2 knockout mice (Slc7a2-/-) had 20% higher blood glucose (+/+ vs. -/-: 412±23 vs. 517±25mg/dL, n=11-12, p=0.004), 80% lower serum insulin (3.7±0.6 vs. 0.7±0.4nM, p=0.0008) and 59% lower serum glucagon (64.3±23.1pM vs. 26.3±11.4pM, p=0.009) than Slc7a2+/+ animals. Slc7a2+/- mice had 55% decreased insulin levels (+/+ vs. +/-: 3.7±0.6 vs. 1.7±0.3nM, n=8, p=0.012), but showed no difference in serum glucose or glucagon levels, indicating that Slc7a2 expression may be rate limiting for beta cell, but not alpha cell, function. Isolated islets from Slc7a2-/- mice had normal morphology, but reduced high amino acid-stimulated alpha cell proliferation by 90% and 58% in Slc7a2-/- and Slc7a2+/- islets, respectively (+/+ vs.-/-: 3.7±1.7 vs. 0.4±0.2%, p=0.007, n=4; +/+ vs. +/-: 3.7±1.7 vs. 1.5±0.6%, p=0.038, n=4). Knockdown with Slc7a2 shRNA in the aTC1-6 mouse alpha cell line also decreased proliferation. Taken together, these data indicate a central role for Slc7a2 in amino acid-stimulated alpha cell proliferation and glucagon and insulin secretion. Disclosure E. Spears: None. M. Shou: None. W.A. Siv: None. C. Dai: None. W. Chen: None. A.C. Powers: None. D. Dean: None.

Dapagliflozin (DAPA), a selective Sodium Glucose Cotransporter-2 (SGLT2) inhibitor, is widely used for the treatment of type 2 diabetes (T2D). In T2D individuals, DAPA increases plasma glucagon, decreases insulin secretion, and improves muscle insulin sensitivity. DAPA has been reported to increase glucagon secretion in cultured normal human islets possibly because SGLT2 was more highly expressed in α cells. To investigate DAPA effect on human islets in vivo, we transplanted normal human islets (n=3 donors) into immunodeficient mice and treated them with DAPA (2mg/kg/day) or NaCl by oral gavage for 4 weeks. Basal blood glucose (NaCl vs. DAPA: 101±8 vs. 85±5 mg/dL, p=0.0858), glucagon (199.5±24.4 vs. 169.9±15.9 pg/mL, p=0.3107), and human insulin (1.15±0.17 vs. 0.99±0.14 ng/mL, p=0.4626) levels were similar in both groups. However, 15 minutes after glucose challenge, DAPA-treated mice had lower blood glucose (375±21 vs. 287±16 mg/dL, p=0.003), unchanged glucagon (86±14 vs. 75±8 pg/mL, p=0.4883), but lower human insulin levels (2.05±0.24 vs. 0.96±0.ng/mL, p Disclosure C. Dai: None. A. Shostak: None. Y.H. Bouchi: None. N. Hart: None. G. Poffenberger: None. D.C. Saunders: None. R. Haliyur: None. D. Dean: None. R. Aramandla: None. M. Brissova: None. M. Shiota: None. D.L. Greiner: Consultant; Self; The Jackson Laboratory, Allakos, Inc. L.D. Shultz: None. R. Bottino: None. A.C. Powers: None.

The previous chaptercovered what constitutes an AI model, the different types of models we can create, and how to train and build these models. An AI model does not become useful until it is deployed somewhere and consumed by the end user. This chapter describes the various options available on Azure to deploy your models. We provide general guidelines on what to use and when, but this is by no means an exhaustive guide to the Azure platform. In the following sections we discuss the metrics over which we compare the various deployment platforms. Then we discuss the platforms we have found to be suitable for deploying ML models and highlight their pros and cons. We also present simple use cases and architectures for each of them so that you get an idea of how they would fit into a larger solution. We also provide a step-by-step tutorial for deployment of a CNN to Azure Kubernetes Services (AKS) with GPU nodes as a hands-on guide for one recommended option for building a real-time request–response AI system.

The previous chapter showed how a deep learning model—specifically CNNs—could be applied to images. The process could be decoupled into a feature extractor that figures out the optimal hidden-state representation of the input (in this case a vector of feature maps) and a classifier (typically a fully connected layer). This chapter focuses on the hidden-state representation of other forms of data and explores RNNs. RNNs are especially useful for analyzing sequences, which is particularly helpful for natural language processing and time series analysis.

In Chapter 1, we gave an overview of AI and the basic idea behind deep learning. We discussed how deep learning—applying artificial neural network models with a large number of layers—has yielded state-of-the art results for several research areas, such as image classification, object detection, speech recognition, and natural language processing.

Intelligence can be defined in many ways, from the ability to learn to deal with new situations to the ability to make the right decisions according to some criterion, for example (Bengio, 2010). Standard computers and even basic calculators can be thought to be intelligent in some ways, as they can compute an outcome based on human-programed rules. Computers are extremely useful for mundane operations such as arithmetic calculations, and the speed and scale at which they can tackle these problems has greatly increased over time.

This chapter introduces the Microsoft AI Platform, which is a set of services, infrastructure, and tools for building intelligent applications powered by AI. The Microsoft AI Platform runs on the Microsoft Azure cloud computing environment, which provides computing as a utility where you pay for what you use rather than what you own. For more details on the broader Azure Platform, please see the e-book Developer’s Guide to Microsoft Azure (Crump & Luijbregts, 2017). The Microsoft AI Platform enables data scientists and developers to create AI solutions in an efficient and cost-effective manner.

This chapter discusses some of the trends in deep learning and related fields. We cover specifically which trends might be useful for what tasks as well as discuss some of the methods and ideas that could have far-reaching implications but have yet to be applied to many real-world problems. We finish by covering briefly some of the current limitations of deep learning as well as some other areas of AI that seem to hold promise for future AI applications, and discuss briefly some of the ethical and legal implications of deep learning applications.

For many AI projects, deep learning techniques are increasingly being used as the building blocks for innovative solutions ranging from image classification to object detection, image segmentation, image similarity, and text analytics (e.g., sentiment analysis, key phrase extraction). GANs, first introduced by Goodfellow et al. (2014), are emerging as a powerful new approach toward teaching computers how to do complex tasks through a generative process. As noted by Yann LeCun (at http://bit.ly/LeCunGANs ), GANs are truly the “coolest idea in machine learning in the last 20 years.”

CNNs are a prime example of neuroscience influencing deep learning (LeCun, Bottou, Bengio, & Haffner, 1998). These neural networks are based on the seminal work done by Hubel and Wiesel (1962). They discovered that individual neuronal cells in the visual cortex responded only to the presence of visual features such as edges of certain orientations. From their experiments they deduced that the visual cortex contains a hierarchical arrangement of neuronal cells. These neurons are sensitive to specific subregions in the visual field, with these subregions being tiled to cover the entire visual field. They in fact act as localized filters over the input space, making them well suited to exploiting the strong spatial correlation found in natural images. CNNs have been immensely successful in many computer vision tasks not just because of the inspiration drawn from neuroscience, but also due to the clever engineering principles employed. Although they have traditionally been used for applications in the field of computer vision such as face recognition and image classification, CNNs have also been used in other areas such as speech recognition and natural language processing for certain tasks.

Chapter 4 introduced the tools, infrastructure, and services that are available to build the next generation of intelligent applications. These together form a platform that empowers data scientists and developers to build, train, and deploy ML and deep learning models on the intelligent cloud and intelligent edge.

Training AI models is usually more demanding than training standard ML models because they are processing intensive and often the data sets involved are larger. That is why if you are serious about deep learning you have to have access to GPUs. In Azure there are a number of ways you can make use of GPUs, on single VMs or in orchestrated clusters of them. In this chapter, we summarize several of the most common methods available as well as the pros and cons of each. Then we expand on the code we wrote in Chapter 6, which used a VGG-like CNN to tackle the CIFAR10 data set using the DLVM as the computing environment. In this chapter, we extend to other training options such as Batch AI and Batch Shipyard, which can both be useful for scaling up or scaling out training. We finish by highlighting briefly some of the other methods of training AI models on Azure that are not as common but might be useful depending on the problem at hand.

Glucagon plays a major role in the regulation of glucose homeostasis during fed and fasting states. However, the mechanisms responsible for the regulation of pancreatic α cell mass and function are not completely understood. In the current study, we identified mTOR complex 1 (mTORC1) as a major regulator of α cell mass and glucagon secretion. Using mice with tissue-specific deletion of the mTORC1 regulator Raptor in α cells (αRaptorKO), we showed that mTORC1 signaling is dispensable for α cell development, but essential for α cell maturation during the transition from a milk-based diet to a chow-based diet after weaning. Moreover, inhibition of mTORC1 signaling in αRaptorKO mice and in WT animals exposed to chronic rapamycin administration decreased glucagon content and glucagon secretion. In αRaptorKO mice, impaired glucagon secretion occurred in response to different secretagogues and was mediated by alterations in KATP channel subunit expression and activity. Additionally, our data identify the mTORC1/FoxA2 axis as a link between mTORC1 and transcriptional regulation of key genes responsible for α cell function. Thus, our results reveal a potential function of mTORC1 in nutrient-dependent regulation of glucagon secretion and identify a role for mTORC1 in controlling α cell-mass maintenance.

Summary Identification of cell-surface markers specific to human pancreatic β cells would allow in vivo analysis and imaging. Here we introduce a biomarker, ectonucleoside triphosphate diphosphohydrolase-3 (NTPDase3), that is expressed on the cell surface of essentially all adult human β cells, including those from individuals with type 1 or type 2 diabetes. NTPDase3 is expressed dynamically during postnatal human pancreas development, appearing first in acinar cells at birth, but several months later its expression declines in acinar cells while concurrently emerging in islet β cells. Given its specificity and membrane localization, we utilized an NTPDase3 antibody for purification of live human β cells as confirmed by transcriptional profiling, and, in addition, for in vivo imaging of transplanted human β cells. Thus, NTPDase3 is a cell-surface biomarker of adult human β cells, and the antibody directed to this protein should be a useful new reagent for β cell sorting, in vivo imaging, and targeting.

Decreasing glucagon action lowers the blood glucose and may be useful therapeutically for diabetes. However, interrupted glucagon signaling leads to α-cell proliferation. To identify postulated hepatic-derived, circulating factor(s) responsible for α-cell proliferation, we used transcriptomics/proteomics/metabolomics in three models of interrupted glucagon signaling and found that proliferation of mouse, zebrafish, and human α-cells was mTOR- and FoxP transcription factor-dependent. Changes in hepatic amino acid (AA) catabolism gene expression predicted the observed increase in circulating AA. Mimicking these AA levels stimulated α-cell proliferation in a newly developed in vitro assay with L-glutamine being a critical AA. α-cell expression of the AA transporter Slc38a5 was markedly increased in mice with interrupted glucagon signaling and played a role in α-cell proliferation. These results indicate a hepatic-α-islet cell axis where glucagon regulates serum AA availability and AA, especially L-glutamine, regulates α-cell proliferation and mass via mTOR-dependent nutrient sensing.

Glucagon antagonism is a potential treatment for diabetes. One potential side effect is α-cell hyperplasia, which has been noted in several approaches to antagonize glucagon action. To investigate the molecular mechanism of the α-cell hyperplasia and to identify the responsible factor, we created a zebrafish model in which glucagon receptor (gcgr) signaling has been interrupted. The genetically and chemically tractable zebrafish, which provides a robust discovery platform, has two gcgr genes (gcgra and gcgrb) in its genome. Sequence, phylogenetic, and synteny analyses suggest that these are co-orthologs of the human GCGR. Similar to its mammalian counterparts, gcgra and gcgrb are mainly expressed in the liver. We inactivated the zebrafish gcgra and gcgrb using transcription activator-like effector nuclease (TALEN) first individually and then both genes, and assessed the number of α-cells using an α-cell reporter line, Tg(gcga:GFP). Compared to WT fish at 7 days postfertilization, there were more α-cells in gcgra−/−, gcgrb−/−, and gcgra−/−;gcgrb−/− fish and there was an increased rate of α-cell proliferation in the gcgra−/−;gcgrb−/− fish. Glucagon levels were higher but free glucose levels were lower in gcgra−/−, gcgrb−/−, and gcgra−/−;gcgrb−/− fish, similar to Gcgr−/− mice. These results indicate that the compensatory α-cell hyperplasia in response to interruption of glucagon signaling is conserved in zebrafish. The robust α-cell hyperplasia in gcgra−/−;gcgrb−/− larvae provides a platform to screen for chemical and genetic suppressors, and ultimately to identify the stimulus of α-cell hyperplasia and its signaling mechanism.

To examine whether caffeine, the most widely used xenobiotic compound, would induce insect cytochrome P450 or CYP gene expression, upstream DNA fragments of Cyp6a2 (0.12, 0.26, 0.52 and 0.98-kb) and Cyp6a8 (0.06, 0.1, 0.2, 0.5 and 0.8-kb) genes of Drosophila melanogaster were individually fused to the firefly luciferase (luc) reporter gene. Promoter activities of these constructs were examined in Drosophila SL-2 cells using luciferase assays. Activity of 0.2- and 0.8-kb upstream DNA of Cyp6a8 was also measured in transgenic female flies. When these flies were treated with 2 mM pure caffeine or Vivarin caffeine, both DNA fragments showed a 4–5-fold induction of promoter activity. Endogenous Cyp6a8 and Cyp6a2 genes in these flies also showed caffeine-induced expression. In addition, both 0.2- and 0.8-kb DNAs showed differential basal and caffeine-induced activity in head, ovaries, gut, cuticle plus fat body and malpighian tubules. However, in all tissues 0.8-kb DNA always showed higher basal and caffeine-induced activities compared to the 0.2-kb DNA, suggesting that the additional DNA present in the 0.8-kb fragment has sequences that enhance both activities. In SL-2 cells, all reporter constructs of each Cyp6 gene showed significantly higher basal activity than the empty vector. Sequences that boost basal activity are located in − 265/− 129 and − 983/− 522 DNA of Cyp6a2, and − 199/− 109 and − 491/− 199 DNA of Cyp6a8 genes. While the 0.12- and 0.1-kb upstream DNAs of Cyp6a2 and Cyp6a8 genes respectively did not show caffeine-inducibility in SL-2 cells, the longest upstream DNA of each gene gave the highest level of induction. Caffeine-responsive sequences are not clustered at one place; they appear to be dispersed in − 983/− 126 and − 761/− 109 regions of Cyp6a2 and Cyp6a8 genes which also contain many binding sites for activator protein 1 (AP1) and cyclic AMP response element binding protein (CRE-BP). Significance of these binding sites in caffeine-inducibility has been discussed.

Twenty-two patients with plantar fasciitis, 23 patients with other types of foot pain, and 30 patients in a control group were evaluated with the Windlass test performed in a weightbearing and non-weightbearing position. In the non-weightbearing test, the first MP joint was maximally dorsiflexed with the ankle stabilized. The weightbearing test was performed with the toes hanging off the edge of a stool and dorsiflexion of the first MP was performed. Seven of the 22 patients in the plantar fasciitis group had a positive weightbearing Windlass test (31.8%), while only three had a positive test result in a non-weightbearing position (13.6%). None of the patients in the other foot pain group or control group had pain in the weightbearing and non-weightbearing positions. Despite its high rate of specificity, the low rate of sensitivity of the Windlass test may limit its usefulness in the clinical evaluation in patients with plantar fasciitis.

In recent years, the occupational therapy profession has been under increased pressure to document the outcomes of interventions in order 10 improve the quality of client care and to provide evidence of the effectiveness and efficiency of services to recipients and sponsors. The literature to date has focused primarily on establishing therapy outcomes from a therapist perspective. However, attention must also be paid to the views of our consumers regarding the outcomes of therapy. The purpose of this study was to determine whether clients and therapists agree regarding client scores on three outcome measures. The study included 4 occupational therapists and 20 clients with stroke who rated clients' status in terms of personal activities of daily living, instrumental activities of daily living and quality of life, at admission of clients as well as discharge from rehabilitation. The results indicated that there was generally only moderate agreement between therapists' and clients' ratings at both admission...

Glucagon is a critical regulator of glucose homeostasis; however, mechanisms regulating glucagon action and α-cell function and number are incompletely understood. To elucidate the role of the hepatic glucagon receptor (Gcgr) in glucagon action, we generated mice with hepatocyte-specific deletion of the glucagon receptor. Gcgr(Hep)(-/-) mice exhibited reductions in fasting blood glucose and improvements in insulin sensitivity and glucose tolerance compared with wild-type controls, similar in magnitude to changes observed in Gcgr(-/-) mice. Despite preservation of islet Gcgr signaling, Gcgr(Hep)(-/-) mice developed hyperglucagonemia and α-cell hyperplasia. To investigate mechanisms by which signaling through the Gcgr regulates α-cell mass, wild-type islets were transplanted into Gcgr(-/-) or Gcgr(Hep)(-/-) mice. Wild-type islets beneath the renal capsule of Gcgr(-/-) or Gcgr(Hep)(-/-) mice exhibited an increased rate of α-cell proliferation and expansion of α-cell area, consistent with changes exhibited by endogenous α-cells in Gcgr(-/-) and Gcgr(Hep)(-/-) pancreata. These results suggest that a circulating factor generated after disruption of hepatic Gcgr signaling can increase α-cell proliferation independent of direct pancreatic input. Identification of novel factors regulating α-cell proliferation and mass may facilitate the generation and expansion of α-cells for transdifferentiation into β-cells and the treatment of diabetes.

Parkinson's disease (PD) is primarily thought of as a disease of aging. However, recent evidence points to the potential for exposure to xenobiotics during development to increase risk of PD. Here, we report that developmental exposure to the organochlorine pesticide heptachlor alters the dopamine system and increases neurotoxicity in an animal model of PD. Exposure of pregnant mice to heptachlor led to increased levels of the dopamine transporter (DAT) and vesicular monoamine transporter 2 (VMAT2) levels at both the protein and mRNA level in their offspring. Increased DAT and VMAT2 levels were accompanied by alterations of mRNA levels of nuclear transcription factors that control dopamine neuron development and regulate DAT and VMAT2 levels in adulthood. At 12 weeks of age, control and heptachlor-exposed offspring were administered a moderate dose (2× 10 mg/kg) of the parkinsonism-inducing agent MPTP. Greater neurotoxicity as evidenced by a greater loss of striatal dopamine and potentiation of increased levels of glial fibrillary acidic protein and α-synuclein was observed in heptachlor-exposed offspring. The neurotoxicity observed was greater in the male offspring than the female offspring, suggesting that males are more susceptible to the long-term effects of developmental heptachlor exposure. These data suggest that developmental heptachlor exposure causes long-term alterations of the dopamine system thereby rendering it more susceptible to dopaminergic damage in adulthood.

Exposure to pesticides has been suggested to increase the risk of Parkinson's disease (PD), but the mechanisms responsible for this association are not clear. Here, we report that perinatal exposure of mice during gestation and lactation to low levels of dieldrin (0.3, 1, or 3 mg/kg every 3 days) alters dopaminergic neurochemistry in their offspring and exacerbates MPTP toxicity. At 12 wk of age, protein and mRNA levels of the dopamine transporter (DAT) and vesicular monoamine transporter 2 (VMAT2) were increased by perinatal dieldrin exposure in a dose-related manner. We then administered MPTP (2 x 10 mg/kg s.c) at 12 wk of age and observed a greater reduction of striatal dopamine in dieldrin-exposed offspring, which was associated with a greater DAT:VMAT2 ratio. Additionally, dieldrin exposure during development potentiated the increase in GFAP and alpha-synuclein levels induced by MPTP, indicating increased neurotoxicity. In all cases there were greater effects observed in the male offspring than the female, similar to that observed in human cases of PD. These data suggest that developmental exposure to dieldrin leads to persistent alterations of the developing dopaminergic system and that these alterations induce a "silent" state of dopamine dysfunction, thereby rendering dopamine neurons more vulnerable later in life.