Search

Interactive semantic segmentation pursues high-quality segmentation results at the cost of a small number of user clicks. It is attracting more and more research attention for its convenience in labeling semantic pixel-level data. Existing interactive segmentation methods often pursue higher interaction efficiency by mining the latent information of user clicks or exploring efficient interaction manners. However, these works neglect to explicitly exploit the semantic correlations between user corrections and model mispredictions, thus suffering from two flaws. First, similar prediction errors frequently occur in actual use, causing users to repeatedly correct them. Second, the interaction difficulty of different semantic classes varies across images, but existing models use monotonic parameters for all images which lack semantic pertinence. Therefore, in this article, we explore the semantic correlations existing in corrections and mispredictions by proposing a simple yet effective online learning solution to the above problems, named correction-misprediction correlation mining ( CM 2 ). Specifically, we leverage the correction-misprediction similarities to design a confusion memory module (CMM) for automatic correction when similar prediction errors reappear. Furthermore, we measure the semantic interaction difficulty by counting the correction-misprediction pairs and design a challenge adaptive convolutional layer (CACL), which can adaptively switch different parameters according to interaction difficulties to better segment the challenging classes. Our method requires no extra training besides the online learning process and can effectively improve interaction efficiency. Our proposed CM 2 achieves state-of-the-art results on three public semantic segmentation benchmarks.

Knowledge distillation (KD) is a learning paradigm for boosting resource-efficient graph neural networks (GNNs) using more expressive yet cumbersome teacher models. Past work on distillation for GNNs proposed the local structure preserving (LSP) loss, which matches local structural relationships defined over edges across the student and teacher's node embeddings. This article studies whether preserving the global topology of how the teacher embeds graph data can be a more effective distillation objective for GNNs, as real-world graphs often contain latent interactions and noisy edges. We propose graph contrastive representation distillation (G-CRD), which uses contrastive learning to implicitly preserve global topology by aligning the student node embeddings to those of the teacher in a shared representation space. Additionally, we introduce an expanded set of benchmarks on large-scale real-world datasets where the performance gap between teacher and student GNNs is non-negligible. Experiments across four datasets and 14 heterogeneous GNN architectures show that G-CRD consistently boosts the performance and robustness of lightweight GNNs, outperforming LSP (and a global structure preserving (GSP) variant of LSP) as well as baselines from 2-D computer vision. An analysis of the representational similarity among teacher and student embedding spaces reveals that G-CRD balances preserving local and global relationships, while structure preserving approaches are best at preserving one or the other.

Background: SARS-CoV-2-neutralizing antibodies (nABs) showed great promise in the early phases of the COVID-19 pandemic. The emergence of resistant strains, however, quickly rendered the majority of clinically approved nABs ineffective. This underscored the imperative to develop nAB cocktails targeting non-overlapping epitopes., Methods: Undertaking a nAB discovery program, we employed a classical workflow, while integrating artificial intelligence (AI)-based prediction to select non-competing nABs very early in the pipeline. We identified and in vivo validated (in female Syrian hamsters) two highly potent nABs., Findings: Despite the promising results, in depth cryo-EM structural analysis demonstrated that the AI-based prediction employed with the intention to ensure non-overlapping epitopes was inaccurate. The two nABs in fact bound to the same receptor-binding epitope in a remarkably similar manner., Interpretation: Our findings indicate that, even in the Alphafold era, AI-based predictions of paratope-epitope interactions are rough and experimental validation of epitopes remains an essential cornerstone of a successful nAB lead selection., Funding: Full list of funders is provided at the end of the manuscript., Competing Interests: Declaration of interests Ghent University has filed for patent protection on the antibody sequences described herein, and D.D.A., M.W., R.W., W.W., S.G. and L.V. are named as co-inventors on this patent (European Patent Application: 21186206.5). A.P. is employee of the MAbSilico, H.R. holds a patent regarding neutralizing VHH antibodies binding the Spike RBD (PCT/EP2021/052885) and has filed a priority application for neutralizing VHH antibodies binding Spike S2 (EP 23160838.1). X.S. is a recipient of FWO research project COVID-19 (G0G4920N) and FWO-FNRS project VIREOS (EOS ID: 30981113) grants., (Copyright © 2023. Published by Elsevier B.V.)

Existing approaches towards anomaly detection (AD) often rely on a substantial amount of anomaly-free data to train representation and density models. However, large anomaly-free datasets may not always be available before the inference stage; in which case an anomaly detection model must be trained with only a handful of normal samples, a.k.a. few-shot anomaly detection (FSAD). In this paper, we propose a novel methodology to address the challenge of FSAD which incorporates two important techniques. Firstly, we employ a model pre-trained on a large source dataset to initialize model weights. Secondly, to ameliorate the covariate shift between source and target domains, we adopt contrastive training to fine-tune on the few-shot target domain data. To learn suitable representations for the downstream AD task, we additionally incorporate cross-instance positive pairs to encourage a tight cluster of the normal samples, and negative pairs for better separation between normal and synthesized negative samples. We evaluate few-shot anomaly detection on 3 controlled AD tasks and 4 real-world AD tasks to demonstrate the effectiveness of the proposed method.

Current COVID-19 vaccines are based on prototypic spike sequences from ancestral 2019 SARS-CoV-2 strains. However, the ongoing pandemic is fueled by variants of concern (VOC) escaping vaccine-mediated protection. Here we demonstrate how immunization in hamsters using prototypic spike expressed from yellow fever 17D (YF17D) as vector blocks ancestral virus (B lineage) and VOC Alpha (B.1.1.7) yet fails to fully protect from Beta (B.1.351). However, the same YF17D vectored vaccine candidate with an evolved antigen induced considerably improved neutralizing antibody responses against VOCs Beta, Gamma (P.1) and the recently predominant Omicron (B.1.1.529), while maintaining immunogenicity against ancestral virus and VOC Delta (B.1.617.2). Thus vaccinated animals resisted challenge by all VOCs, including vigorous high titre exposure to the most difficult to cover Beta, Delta and Omicron variants, eliminating detectable virus and markedly improving lung pathology. Finally, vaccinated hamsters did not transmit Delta variant to non-vaccinated cage mates. Overall, our data illustrate how current first-generation COVID-19 vaccines may need to be updated to maintain efficacy against emerging VOCs and their spread at community level., (© 2022. The Author(s).)

Self-amplifying RNA vaccines may induce equivalent or more potent immune responses at lower doses compared to non-replicating mRNA vaccines via amplified antigen expression. In this paper, we demonstrate that 1 μg of an LNP-formulated dual-antigen self-amplifying RNA vaccine (ZIP1642), encoding both the S-RBD and N antigen, elicits considerably higher neutralizing antibody titers against Wuhan-like Beta B.1.351 and Delta B.1.617.2 SARS-CoV-2 variants compared to those of convalescent patients. In addition, ZIP1642 vaccination in mice expanded both S- and N-specific CD3 + CD4 + and CD3 + CD8 + T cells and caused a Th1 shifted cytokine response. We demonstrate that the induction of such dual antigen-targeted cell-mediated immune response may provide better protection against variants displaying highly mutated Spike proteins, as infectious viral loads of both Wuhan-like and Beta variants were decreased after challenge of ZIP1642 vaccinated hamsters. Supported by these results, we encourage redirecting focus toward the induction of multiple antigen-targeted cell-mediated immunity in addition to neutralizing antibody responses to bypass waning antibody responses and attenuate infectious breakthrough and disease severity of future SARS-CoV-2 variants., Competing Interests: Declaration of interest S.M., A.K.M.A.H., A.V., B.W., M.P., M.S., S.V., and I.S. are current employees of Ziphius Vaccines NV and may own Ziphius shares/warrants. S.M., A.K.M.A.H., S.V., I.S., and N.N.S. are inventors on a patent application claiming subject matter related to the SARS-CoV-2 saRNA vaccine candidates described herein. The Virology Unit of the Institute of Tropical Medicine, supervised by K.K.A., and the Department of Microbiology, Immunology, and Transplantation of Rega Institute for Medical Research (KUL), supervised by J.N., received compensation from Ziphius Vaccines NV to perform the neutralization assays and hamster studies, respectively. The authors declare no other competing interests., (Copyright © 2022 The Author(s). Published by Elsevier Inc. All rights reserved.)

The SARS-CoV-2 Omicron variant has very high levels of transmission, is resistant to neutralization by authorized therapeutic human monoclonal antibodies (mAb) and is less sensitive to vaccine-mediated immunity. To provide additional therapies against Omicron, we isolated a mAb named P2G3 from a previously infected vaccinated donor and showed that it has picomolar-range neutralizing activity against Omicron BA.1, BA.1.1, BA.2 and all other variants tested. We solved the structure of P2G3 Fab in complex with the Omicron spike using cryo-electron microscopy at 3.04 Å resolution to identify the P2G3 epitope as a Class 3 mAb that is different from mAb-binding spike epitopes reported previously. Using a SARS-CoV-2 Omicron monkey challenge model, we show that P2G3 alone, or in combination with P5C3 (a broadly active Class 1 mAb previously identified), confers complete prophylactic or therapeutic protection. Although we could select for SARS-CoV-2 mutants escaping neutralization by P2G3 or by P5C3 in vitro, they had low infectivity and 'escape' mutations are extremely rare in public sequence databases. We conclude that this combination of mAbs has potential as an anti-Omicron drug., (© 2022. The Author(s).)

Ancestral severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) lacks the intrinsic ability to bind to the mouse ACE2 receptor, and therefore establishment of SARS-CoV-2 mouse models has been limited to the use of mouse-adapted viruses or genetically modified mice. Interestingly, some of the variants of concern, such as the Beta B.1.351 variant, show an improved binding to the mouse receptor and hence better replication in different wild-type (WT) mouse species. Here, we describe the establishment of a SARS-CoV-2 Beta B.1.351 variant infection model in male SCID mice as a tool to assess the antiviral efficacy of potential SARS-CoV-2 small-molecule inhibitors. Intranasal infection of male SCID mice with 10 5 50% tissue culture infective doses (TCID 50 ) of the Beta B.1.351 variant resulted in high viral loads in the lungs and moderate signs of lung pathology on day 3 postinfection. Treatment of infected mice with the antiviral drugs molnupiravir (200 mg/kg, twice a day [BID]) or nirmatrelvir (300 mg/kg, BID) for 3 consecutive days significantly reduced the infectious virus titers in the lungs by 2 and 3.9 log 10 TCID 50 /mg of tissue, respectively, and significantly improved lung pathology. Together, these data demonstrate the validity of this SCID mouse Beta B.1.351 variant infection model as a convenient preclinical model for assessment of potential activity of antivirals against SARS-CoV-2. IMPORTANCE Unlike the ancestral SARS-CoV-2 strain, the Beta (B.1.351) variant of concern has been reported to replicate to some extent in WT mice (C57BL/6 and BALB/c). We demonstrate here that infection of SCID mice with the Beta variant resulted in high viral loads in the lungs on day 3 postinfection. Treatment of infected mice with molnupiravir or nirmatrelvir for 3 consecutive days markedly reduced the infectious virus titers in the lungs and improved lung pathology. The SARS-CoV2 SCID mouse infection model, which is ideally suited for antiviral studies, offers an advantage in comparison to other SARS-CoV2 mouse models, in that there is no need for the use of mouse-adapted virus strains or genetically modified mice. Mouse models also have advantages over hamster models because (i) lower amounts of test drugs are needed, (ii) more animals can be housed in a cage, and (iii) reagents to analyze mouse samples are more readily available than those for hamsters.

In the absence of drugs to treat or prevent COVID-19, drug repurposing can be a valuable strategy. Despite a substantial number of clinical trials, drug repurposing did not deliver on its promise. While success was observed with some repurposed drugs (e.g., remdesivir, dexamethasone, tocilizumab, baricitinib), others failed to show clinical efficacy. One reason is the lack of clear translational processes based on adequate preclinical profiling before clinical evaluation. Combined with limitations of existing in vitro and in vivo models, there is a need for a systematic approach to urgent antiviral drug development in the context of a global pandemic. We implemented a methodology to test repurposed and experimental drugs to generate robust preclinical evidence for further clinical development. This translational drug development platform comprises in vitro, ex vivo, and in vivo models of SARS-CoV-2, along with pharmacokinetic modeling and simulation approaches to evaluate exposure levels in plasma and target organs. Here, we provide examples of identified repurposed antiviral drugs tested within our multidisciplinary collaboration to highlight lessons learned in urgent antiviral drug development during the COVID-19 pandemic. Our data confirm the importance of assessing in vitro and in vivo potency in multiple assays to boost the translatability of pre-clinical data. The value of pharmacokinetic modeling and simulations for compound prioritization is also discussed. We advocate the need for a standardized translational drug development platform for mild-to-moderate COVID-19 to generate preclinical evidence in support of clinical trials. We propose clear prerequisites for progression of drug candidates for repurposing into clinical trials. Further research is needed to gain a deeper understanding of the scope and limitations of the presented translational drug development platform.

Background: Consumer-grade wearable devices enable detailed recordings of heart rate and step counts in free-living conditions. Recent studies have shown that summary statistics from these wearable recordings have potential uses for longitudinal monitoring of health and disease states. However, the relationship between higher resolution physiological dynamics from wearables and known markers of health and disease remains largely uncharacterized., Objective: We aimed to derive high-resolution digital phenotypes from observational wearable recordings and to examine their associations with modifiable and inherent markers of cardiometabolic disease risk., Methods: We introduced a principled framework to extract interpretable high-resolution phenotypes from wearable data recorded in free-living conditions. The proposed framework standardizes the handling of data irregularities; encodes contextual information regarding the underlying physiological state at any given time; and generates a set of 66 minimally redundant features across active, sedentary, and sleep states. We applied our approach to a multimodal data set, from the SingHEART study (NCT02791152), which comprises heart rate and step count time series from wearables, clinical screening profiles, and whole genome sequences from 692 healthy volunteers. We used machine learning to model nonlinear relationships between the high-resolution phenotypes on the one hand and clinical or genomic risk markers for blood pressure, lipid, weight and sugar abnormalities on the other. For each risk type, we performed model comparisons based on Brier scores to assess the predictive value of high-resolution features over and beyond typical baselines. We also qualitatively characterized the wearable phenotypes for participants who had actualized clinical events., Results: We found that the high-resolution features have higher predictive value than typical baselines for clinical markers of cardiometabolic disease risk: the best models based on high-resolution features had 17.9% and 7.36% improvement in Brier score over baselines based on age and gender and resting heart rate, respectively (P<.001 in each case). Furthermore, heart rate dynamics from different activity states contain distinct information (maximum absolute correlation coefficient of 0.15). Heart rate dynamics in sedentary states are most predictive of lipid abnormalities and obesity, whereas patterns in active states are most predictive of blood pressure abnormalities (P<.001). Moreover, in comparison with standard measures, higher resolution patterns in wearable heart rate recordings are better able to represent subtle physiological dynamics related to genomic risk for cardiometabolic disease (improvement of 11.9%-22.0% in Brier scores; P<.001). Finally, illustrative case studies reveal connections between these high-resolution phenotypes and actualized clinical events, even for borderline profiles lacking apparent cardiometabolic risk markers., Conclusions: High-resolution digital phenotypes recorded by consumer wearables in free-living states have the potential to enhance the prediction of cardiometabolic disease risk and could enable more proactive and personalized health management., (©Weizhuang Zhou, Yu En Chan, Chuan Sheng Foo, Jingxian Zhang, Jing Xian Teo, Sonia Davila, Weiting Huang, Jonathan Yap, Stuart Cook, Patrick Tan, Calvin Woon-Loong Chin, Khung Keong Yeo, Weng Khong Lim, Pavitra Krishnaswamy. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 29.07.2022.)

Several techniques for multivariate time series anomaly detection have been proposed recently, but a systematic comparison on a common set of datasets and metrics is lacking. This article presents a systematic and comprehensive evaluation of unsupervised and semisupervised deep-learning-based methods for anomaly detection and diagnosis on multivariate time series data from cyberphysical systems. Unlike previous works, we vary the model and post-processing of model errors, i.e., the scoring functions independently of each other, through a grid of ten models and four scoring functions, comparing these variants to state-of-the-art methods. In time-series anomaly detection, detecting anomalous events is more important than detecting individual anomalous time points. Through experiments, we find that the existing evaluation metrics either do not take events into account or cannot distinguish between a good detector and trivial detectors, such as a random or an all-positive detector. We propose a new metric to overcome these drawbacks, namely, the composite F-score (Fc 1 ), for evaluating time-series anomaly detection. Our study highlights that dynamic scoring functions work much better than static ones for multivariate time series anomaly detection, and the choice of scoring functions often matters more than the choice of the underlying model. We also find that a simple, channel-wise model-the univariate fully connected auto-encoder, with the dynamic Gaussian scoring function emerges as a winning candidate for both anomaly detection and diagnosis, beating state-of-the-art algorithms.

Nelfinavir is an HIV protease inhibitor that has been widely prescribed as a component of highly active antiretroviral therapy, and has been reported to exert in vitro antiviral activity against SARS-CoV-2. We here assessed the effect of Nelfinavir in a SARS-CoV-2 infection model in hamsters. Despite the fact that Nelfinavir, [50 mg/kg twice daily (BID) for four consecutive days], did not reduce viral RNA load and infectious virus titres in the lung of infected animals, treatment resulted in a substantial improvement of SARS-CoV-2-induced lung pathology. This was accompanied by a dense infiltration of neutrophils in the lung interstitium which was similarly observed in non-infected hamsters. Nelfinavir resulted also in a marked increase in activated neutrophils in the blood, as observed in non-infected animals. Although Nelfinavir treatment did not alter the expression of chemoattractant receptors or adhesion molecules on human neutrophils, in vitro migration of human neutrophils to the major human neutrophil attractant CXCL8 was augmented by this protease inhibitor. Nelfinavir appears to induce an immunomodulatory effect associated with increasing neutrophil number and functionality, which may be linked to the marked improvement in SARS-CoV-2 lung pathology independent of its lack of antiviral activity. Since Nelfinavir is no longer used for the treatment of HIV, we studied the effect of two other HIV protease inhibitors, namely the combination Lopinavir/Ritonavir (Kaletra™) in this model. This combination resulted in a similar protective effect as Nelfinavir against SARS-CoV2 induced lung pathology in hamsters., (Copyright © 2022 The Authors. Published by Elsevier B.V. All rights reserved.)

Background: Many transcription factors (TFs), such as multi zinc-finger (ZF) TFs, have multiple DNA binding domains (DBDs), and deciphering the DNA binding motifs of individual DBDs is a major challenge. One example of such a TF is CCCTC-binding factor (CTCF), a TF with eleven ZFs that plays a variety of roles in transcriptional regulation, most notably anchoring DNA loops. Previous studies found that CTCF ZFs 3-7 bind CTCF's core motif and ZFs 9-11 bind a specific upstream motif, but the motifs of ZFs 1-2 have yet to be identified., Results: We developed a new approach to identifying the binding motifs of individual DBDs of a TF through analyzing chromatin immunoprecipitation sequencing (ChIP-seq) experiments in which a single DBD is mutated: we train a deep convolutional neural network to predict whether wild-type TF binding sites are preserved in the mutant TF dataset and interpret the model. We applied this approach to mouse CTCF ChIP-seq data and identified the known binding preferences of CTCF ZFs 3-11 as well as a putative GAG binding motif for ZF 1. We analyzed other CTCF datasets to provide additional evidence that ZF 1 is associated with binding at the motif we identified, and we found that the presence of the motif for ZF 1 is associated with CTCF ChIP-seq peak strength., Conclusions: Our approach can be applied to any TF for which in vivo binding data from both the wild-type and mutated versions of the TF are available, and our findings provide new potential insights binding preferences of CTCF's DBDs., (© 2022. The Author(s).)

Ivermectin, an FDA-approved antiparasitic drug, has been reported to have in vitro activity against SARS-CoV-2. Increased off-label use of ivermectin for COVID-19 has been reported. We here assessed the effect of ivermectin in Syrian hamsters infected with the SARS-CoV-2 Beta (B.1.351) variant. Infected animals received a clinically relevant dose of ivermectin (0.4 mg/kg subcutaneously dosed) once daily for four consecutive days after which the effect was quantified. Ivermectin monotherapy did not reduce lung viral load and even significantly worsened SARS-CoV-2-induced lung pathology. Additionally, it did not potentiate the activity of molnupiravir (Lagevrio TM ) when combined with this drug. This study contributes to the growing body of evidence that ivermectin does not result in a beneficial effect in the treatment of COVID-19. These findings are important given the increasing, dangerous off-label use of ivermectin for the treatment of COVID-19.

There is an urgent need for potent and selective antivirals against SARS-CoV-2. Pfizer developed PF-07321332 (PF-332), a potent inhibitor of the viral main protease (Mpro, 3CLpro) that can be dosed orally and that is in clinical development. We here report that PF-332 exerts equipotent in vitro activity against the four SARS-CoV-2 variants of concerns (VoC) and that it can completely arrest replication of the alpha variant in primary human airway epithelial cells grown at the air-liquid interface. Treatment of Syrian Golden hamsters with PF-332 (250 mg/kg, twice daily) completely protected the animals against intranasal infection with the beta (B.1.351) and delta (B.1.617.2) SARS-CoV-2 variants. Moreover, treatment of SARS-CoV-2 (B.1.617.2) infected animals with PF-332 completely prevented transmission to untreated co-housed sentinels., (© 2022. The Author(s).)

The emergence of SARS-CoV-2 variants of concern (VoCs) has exacerbated the COVID-19 pandemic. End of November 2021, a new SARS-CoV-2 variant namely the omicron (B.1.1.529) emerged. Since this omicron variant is heavily mutated in the spike protein, WHO classified this variant as the 5th variant of concern (VoC). We previously demonstrated that the ancestral strain and the other SARS-CoV-2 VoCs replicate efficiently in and cause a COVID19-like pathology in Syrian hamsters. We here wanted to explore the infectivity of the omicron variant in comparison to the ancestral D614G strain in the hamster model. Strikingly, in hamsters that had been infected with the omicron variant, a 3 log 10 lower viral RNA load was detected in the lungs as compared to animals infected with D614G and no infectious virus was detectable in this organ. Moreover, histopathological examination of the lungs from omicron-infected hamsters revealed no signs of peri-bronchial inflammation or bronchopneumonia., (Copyright © 2022 The Authors. Published by Elsevier B.V. All rights reserved.)

Large-scale distributed training of deep neural networks results in models with worse generalization performance as a result of the increase in the effective mini-batch size. Previous approaches attempt to address this problem by varying the learning rate and batch size over epochs and layers, or ad hoc modifications of batch normalization. We propose scalable and practical natural gradient descent (SP-NGD), a principled approach for training models that allows them to attain similar generalization performance to models trained with first-order optimization methods, but with accelerated convergence. Furthermore, SP-NGD scales to large mini-batch sizes with a negligible computational overhead as compared to first-order methods. We evaluated SP-NGD on a benchmark task where highly optimized first-order methods are available as references: training a ResNet-50 model for image classification on ImageNet. We demonstrate convergence to a top-1 validation accuracy of 75.4 percent in 5.5 minutes using a mini-batch size of 32,768 with 1,024 GPUs, as well as an accuracy of 74.9 percent with an extremely large mini-batch size of 131,072 in 873 steps of SP-NGD.

Recent work on curvilinear structure segmentation has mostly focused on backbone network design and loss engineering. The challenge of collecting labelled data, an expensive and labor intensive process, has been overlooked. While labelled data is expensive to obtain, unlabelled data is often readily available. In this work, we propose SemiCurv, a semi-supervised learning (SSL) framework for curvilinear structure segmentation that is able to utilize such unlabelled data to reduce the labelling burden. Our framework addresses two key challenges in formulating curvilinear segmentation in a semi-supervised manner. First, to fully exploit the power of consistency based SSL, we introduce a geometric transformation as strong data augmentation and then align segmentation predictions via a differentiable inverse transformation to enable the computation of pixel-wise consistency. Second, the traditional mean square error (MSE) on unlabelled data is prone to collapsed predictions and this issue exacerbates with severe class imbalance (significantly more background pixels). We propose a N-pair consistency loss to avoid trivial predictions on unlabelled data. We evaluate SemiCurv on six curvilinear segmentation datasets, and find that with no more than 5% of the labelled data, it achieves close to 95% of the performance relative to its fully supervised counterpart.

Background clutters pose challenges to defocus blur detection. Existing approaches often produce artifact predictions in background areas with clutter and relatively low confident predictions in boundary areas. In this work, we tackle the above issues from two perspectives. Firstly, inspired by the recent success of self-attention mechanism, we introduce channel-wise and spatial-wise attention modules to attentively aggregate features at different channels and spatial locations to obtain more discriminative features. Secondly, we propose a generative adversarial training strategy to suppress spurious and low reliable predictions. This is achieved by utilizing a discriminator to identify predicted defocus map from ground-truth ones. As such, the defocus network (generator) needs to produce 'realistic' defocus map to minimize discriminator loss. We further demonstrate that the generative adversarial training allows exploiting additional unlabeled data to improve performance, a.k.a. semi-supervised learning, and we provide the first benchmark on semi-supervised defocus detection. Finally, we demonstrate that the existing evaluation metrics for defocus detection generally fail to quantify the robustness with respect to thresholding. For a fair and practical evaluation, we introduce an effective yet efficient AUF β metric. Extensive experiments on three public datasets verify the superiority of the proposed methods compared against state-of-the-art approaches.

Broadly neutralizing antibodies are an important treatment for individuals with coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Antibody-based therapeutics are also essential for pandemic preparedness against future Sarbecovirus outbreaks. Camelid-derived single domain antibodies (VHHs) exhibit potent antimicrobial activity and are being developed as SARS-CoV-2–neutralizing antibody-like therapeutics. Here, we identified VHHs that neutralize both SARS-CoV-1 and SARS-CoV-2, including now circulating variants. We observed that the VHHs bound to a highly conserved epitope in the receptor binding domain of the viral spike protein that is difficult to access for human antibodies. Structure-guided molecular modeling, combined with rapid yeast-based prototyping, resulted in an affinity enhanced VHH-human immunoglobulin G1 Fc fusion molecule with subnanomolar neutralizing activity. This VHH-Fc fusion protein, produced in and purified from cultured Chinese hamster ovary cells, controlled SARS-CoV-2 replication in prophylactic and therapeutic settings in mice expressing human angiotensin converting enzyme 2 and in hamsters infected with SARS-CoV-2. These data led to affinity-enhanced selection of the VHH, XVR011, a stable anti–COVID-19 biologic that is now being evaluated in the clinic.

Control of the ongoing SARS-CoV-2 pandemic is endangered by the emergence of viral variants with increased transmission efficiency, resistance to marketed therapeutic antibodies, and reduced sensitivity to vaccine-induced immunity. Here, we screen B cells from COVID-19 donors and identify P5C3, a highly potent and broadly neutralizing monoclonal antibody with picomolar neutralizing activity against all SARS-CoV-2 variants of concern (VOCs) identified to date. Structural characterization of P5C3 Fab in complex with the spike demonstrates a neutralizing activity defined by a large buried surface area, highly overlapping with the receptor-binding domain (RBD) surface necessary for ACE2 interaction. We further demonstrate that P5C3 shows complete prophylactic protection in the SARS-CoV-2-infected hamster challenge model. These results indicate that P5C3 opens exciting perspectives either as a prophylactic agent in immunocompromised individuals with poor response to vaccination or as combination therapy in SARS-CoV-2-infected individuals., Competing Interests: Declaration of interests C.F., P.T., G.P., and D.T declare that they are co-inventors on a patent application titled “Neutralizing antibodies and use thereof in the treatment of SARS-CoV-2 infection” with filing number PCT/IB2021/050621 that covers newly identified antibodies described in this manuscript. The remaining authors declare no competing interests., (Copyright © 2021 The Author(s). Published by Elsevier Inc. All rights reserved.)

Background: Favipiravir and Molnupiravir, orally available antivirals, have been reported to exert antiviral activity against SARS-CoV-2. First efficacy data have been recently reported in COVID-19 patients., Methods: We here report on the combined antiviral effect of both drugs in a SARS-CoV-2 Syrian hamster infection model. The infected hamsters were treated twice daily with the vehicle (the control group) or a suboptimal dose of each compound or a combination of both compounds., Findings: When animals were treated with a combination of suboptimal doses of Molnupiravir and Favipiravir at the time of infection, a marked combined potency at endpoint is observed. Infectious virus titers in the lungs of animals treated with the combination are reduced by ∼5 log10 and infectious virus are no longer detected in the lungs of >60% of treated animals. When start of treatment was delayed with one day a reduction of titers in the lungs of 2.4 log10 was achieved. Moreover, treatment of infected animals nearly completely prevented transmission to co-housed untreated sentinels. Both drugs result in an increased mutation frequency of the remaining viral RNA recovered from the lungs of treated animals. In the combo-treated hamsters, an increased frequency of C-to-T mutations in the viral RNA is observed as compared to the single treatment groups which may explain the pronounced antiviral potency of the combination., Interpretation: Our findings may lay the basis for the design of clinical studies to test the efficacy of the combination of Molnupiravir/Favipiravir in the treatment of COVID-19., Funding: stated in the acknowledgment., Competing Interests: Declaration of Competing Interest None to declare., (Copyright © 2021 The Author(s). Published by Elsevier B.V. All rights reserved.)

Deep learning models achieve strong performance for radiology image classification, but their practical application is bottlenecked by the need for large labeled training datasets. Semi-supervised learning (SSL) approaches leverage small labeled datasets alongside larger unlabeled datasets and offer potential for reducing labeling cost. In this work, we introduce NoTeacher, a novel consistency-based SSL framework which incorporates probabilistic graphical models. Unlike Mean Teacher which maintains a teacher network updated via a temporal ensemble, NoTeacher employs two independent networks, thereby eliminating the need for a teacher network. We demonstrate how NoTeacher can be customized to handle a range of challenges in radiology image classification. Specifically, we describe adaptations for scenarios with 2D and 3D inputs, with uni and multi-label classification, and with class distribution mismatch between labeled and unlabeled portions of the training data. In realistic empirical evaluations on three public benchmark datasets spanning the workhorse modalities of radiology (X-Ray, CT, MRI), we show that NoTeacher achieves over 90-95% of the fully supervised AUROC with less than 5-15% labeling budget. Further, NoTeacher outperforms established SSL methods with minimal hyperparameter tuning, and has implications as a principled and practical option for semi-supervised learning in radiology applications., Competing Interests: Declaration of Competing Interest 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., (Copyright © 2021. Published by Elsevier B.V.)

The emergence of SARS-CoV-2 variants of concern (VoCs) has exacerbated the COVID-19 pandemic. Currently available monoclonal antibodies and vaccines appear to have reduced efficacy against some of these VoCs. Antivirals targeting conserved proteins of SARS-CoV-2 are unlikely to be affected by mutations arising in VoCs and should therefore be effective against emerging variants. We here investigate the efficacy of molnupiravir, currently in phase 2 clinical trials, in hamsters infected with Wuhan strain or B.1.1.7 and B.1.351 variants. Molnupiravir proved to be effective against infections with each of the variants and therefore may have potential combating current and future emerging VoCs., (© The Author(s) 2021. Published by Oxford University Press for the Infectious Diseases Society of America.)

Background: Within one year after its emergence, more than 108 million people acquired SARS-CoV-2 and almost 2·4 million succumbed to COVID-19. New SARS-CoV-2 variants of concern (VoC) are emerging all over the world, with the threat of being more readily transmitted, being more virulent, or escaping naturally acquired and vaccine-induced immunity. At least three major prototypic VoC have been identified, i.e. the United Kingdom, UK (B.1.1.7), South African (B.1.351) and Brazilian (B.1.1.28.1) variants. These are replacing formerly dominant strains and sparking new COVID-19 epidemics., Methods: We studied the effect of infection with prototypic VoC from both B.1.1.7 and B.1.351 variants in female Syrian golden hamsters to assess their relative infectivity and virulence in direct comparison to two basal SARS-CoV-2 strains isolated in early 2020., Findings: A very efficient infection of the lower respiratory tract of hamsters by these VoC is observed. In line with clinical evidence from patients infected with these VoC, no major differences in disease outcome were observed as compared to the original strains as was quantified by (i) histological scoring, (ii) micro-computed tomography, and (iii) analysis of the expression profiles of selected antiviral and pro-inflammatory cytokine genes. Noteworthy however, in hamsters infected with VoC B.1.1.7, a particularly strong elevation of proinflammatory cytokines was detected., Interpretation: We established relevant preclinical infection models that will be pivotal to assess the efficacy of current and future vaccine(s) (candidates) as well as therapeutics (small molecules and antibodies) against two important SARS-CoV-2 VoC., Funding: Stated in the acknowledgment., Competing Interests: Declaration of Competing Interest Authors declare no conflict of interests., (Copyright © 2021 The Authors. Published by Elsevier B.V. All rights reserved.)

There is an urgent need for antivirals targeting the SARS-CoV-2 virus to fight the current COVID-19 pandemic. The SARS-CoV-2 main protease (3CLpro) represents a promising target for antiviral therapy. The lack of selectivity for some of the reported 3CLpro inhibitors, specifically versus cathepsin L, raises potential safety and efficacy concerns. ALG-097111 potently inhibited SARS-CoV-2 3CLpro (IC 50  = 7 nM) without affecting the activity of human cathepsin L (IC 50  > 10 μM). When ALG-097111 was dosed in hamsters challenged with SARS-CoV-2, a robust and significant 3.5 log 10 (RNA copies/mg) reduction of the viral RNA copies and 3.7 log 10 (TCID 50 /mg) reduction in the infectious virus titers in the lungs was observed. These results provide the first in vivo validation for the SARS-CoV-2 3CLpro as a promising therapeutic target for selective small molecule inhibitors., (Copyright © 2021 Elsevier Inc. All rights reserved.)

State-of-the-art methods for semantic segmentation are based on deep neural networks trained on large-scale labeled datasets. Acquiring such datasets would incur large annotation costs, especially for dense pixel-level prediction tasks like semantic segmentation. We consider region-based active learning as a strategy to reduce annotation costs while maintaining high performance. In this setting, batches of informative image regions instead of entire images are selected for labeling. Importantly, we propose that enforcing local spatial diversity is beneficial for active learning in this case, and to incorporate spatial diversity along with the traditional active selection criterion, e.g., data sample uncertainty, in a unified optimization framework for region-based active learning. We apply this framework to the Cityscapes and PASCAL VOC datasets and demonstrate that the inclusion of spatial diversity effectively improves the performance of uncertainty-based and feature diversity-based active learning methods. Our framework achieves 95% performance of fully supervised methods with only 5 - 9% of the labeled pixels, outperforming all state-of-the-art region-based active learning methods for semantic segmentation.

Members of the PR/SET domain-containing (PRDM) family of zinc finger transcriptional regulators play diverse developmental roles. PRDM10 is a yet uncharacterized family member, and its function in vivo is unknown. Here, we report an essential requirement for PRDM10 in pre-implantation embryos and embryonic stem cells (mESCs), where loss of PRDM10 results in severe cell growth inhibition. Detailed genomic and biochemical analyses reveal that PRDM10 functions as a sequence-specific transcription factor. We identify Eif3b, which encodes a core component of the eukaryotic translation initiation factor 3 (eIF3) complex, as a key downstream target, and demonstrate that growth inhibition in PRDM10-deficient mESCs is in part mediated through EIF3B-dependent effects on global translation. Our work elucidates the molecular function of PRDM10 in maintaining global translation, establishes its essential role in early embryonic development and mESC homeostasis, and offers insights into the functional repertoire of PRDMs as well as the transcriptional mechanisms regulating translation.

Background and Purpose: Animal toxins have contributed significantly to our understanding of the neurobiology of receptors and ion channels. We studied the venom of the coral snake Micrurus fulvius fulvius and identified and characterized the structure and pharmacology of a new homodimeric neurotoxin, fulditoxin, that exhibited novel pharmacology at nicotinic ACh receptors (nAChRs)., Experimental Approach: Fulditoxin was isolated by chromatography, chemically synthesized, its structure determined by X-ray crystallography, and its pharmacological actions on nAChRs characterized by organ bath assays and two-electrode voltage clamp electrophysiology., Key Results: Fulditoxin's distinct 1.95-Å quaternary structure revealed two short-chain three-finger α-neurotoxins (α-3FNTxs) non-covalently bound by hydrophobic interactions and an ability to bind metal and form tetrameric complexes, not reported previously for three-finger proteins. Although fulditoxin lacked all conserved amino acids canonically important for inhibiting nAChRs, it produced postsynaptic neuromuscular blockade of chick muscle at nanomolar concentrations, comparable to the prototypical α-bungarotoxin. This neuromuscular blockade was completely reversible, which is unusual for snake α-3FNTxs. Fulditoxin, therefore, interacts with nAChRs by utilizing a different pharmacophore. Unlike short-chain α-3FNTxs that bind only to muscle nAChRs, fulditoxin utilizes dimerization to expand its pharmacological targets to include human neuronal α4β2, α7, and α3β2 nAChRs which it blocked with IC 50 values of 1.8, 7, and 12 μM respectively., Conclusions and Implications: Based on its distinct quaternary structure and unusual pharmacology, we named this new class of dimeric Micrurus neurotoxins represented by fulditoxin as Σ-neurotoxins, which offers greater insight into understanding the interactions between nAChRs and peptide antagonists., (© 2019 The British Pharmacological Society.)

The emergence of multi-drug (MDR-TB) and extensively-drug resistant tuberculosis (XDR-TB) is a major threat to the global management of tuberculosis (TB) worldwide. New chemical entities are of need to treat drug-resistant TB. In this study, the mode of action of new, potent quinazoline derivatives was investigated against Mycobacterium tuberculosis (M. tb). Four derivatives 11626141, 11626142, 11626252 and 11726148 showed good activity (MIC ranging from 0.02-0.09 μg/mL) and low toxicity (TD50 ≥ 5μg/mL) in vitro against M. tb strain H37Rv and HepG2 cells, respectively. 11626252 was the most selective compound from this series. Quinazoline derivatives were found to target cytochrome bc1 by whole-genome sequencing of mutants selected with 11626142. Two resistant mutants harboured the transversion T943G (Trp312Gly) and the transition G523A (Gly175Ser) in the cytochrome bc1 complex cytochrome b subunit (QcrB). Interestingly, a third mutant QuinR-M1 contained a mutation in the Rieske iron-sulphur protein (QcrA) leading to resistance to quinazoline and other QcrB inhibitors, the first report of cross-resistance involving QcrA. Modelling of both QcrA and QcrB revealed that all three resistance mutations are located in the stigmatellin pocket, as previously observed for other QcrB inhibitors such as Q203, AX-35, and lansoprazole sulfide (LPZs). Further analysis of the mode of action in vitro revealed that 11626252 exposure leads to ATP depletion, a decrease in the oxygen consumption rate and also overexpression of the cytochrome bd oxidase in M. tb. Our findings suggest that quinazoline-derived compounds are a new and attractive chemical entity for M. tb drug development targeting two separate subunits of the cytochrome bc1 complex., Competing Interests: The authors have declared that no competing interests exist.

Answering questions using multi-modal context is a challenging problem as it requires a deep integration of diverse data sources. Existing approaches only consider a subset of all possible interactions among data sources during one attention hop. In this paper, we present a Holistic Multi-modal Memory Network (HMMN) framework that fully considers interactions between different input sources (multi-modal context, question) at each hop. In addition, to hone in on relevant information, our framework takes answer choices into consideration during the context retrieval stage. Our HMMN framework effectively integrates information from the multi-modal context, question, and answer choices, enabling more informative context to be retrieved for question answering. Experimental results on the MovieQA and TVQA datasets validate the effectiveness of our HMMN framework. Extensive ablation studies show the importance of holistic reasoning and reveal the contributions of different attention strategies to model performance.

The efficacy of the standardized four-drug regimen (comprising isoniazid, rifampin, pyrazinamide, and ethambutol) for the treatment of tuberculosis (TB) is menaced by the emergence of multidrug-resistant (MDR) and extensively drug-resistant (XDR) strains of Mycobacterium tuberculosis Intensive efforts have been made to develop new antibiotics or to repurpose old drugs, and several of these are currently being evaluated in clinical trials for their antitubercular activity. Among the new candidate drugs is macozinone (MCZ), the piperazine-containing benzothiazinone PBTZ169, which is currently being evaluated in phase I/II clinical trials. Here, we determined the in vitro and in vivo activity of MCZ in combination with a range of anti-TB drugs in order to design a new regimen against active TB. Two-drug combinations with MCZ were tested against M. tuberculosis using checkerboard and CFU enumeration after drug exposure assays. MCZ was observed to have no interactions with all first- and second-line anti-TB drugs. At the MIC of each drug, MCZ with either bedaquiline (BDQ), clofazimine (CLO), delamanid (DMD), or sutezolid (STZ) reduced the bacterial burden by 2 logs compared to that achieved with the drugs alone, indicating synergism. MCZ also displayed synergism with clomiphene (CLM), a potential inhibitor of the undecaprenyl pyrophosphate synthase (UppS) in mycobacteria. For all the other drugs tested in combination with MCZ, no synergistic activity was observed. Neither antagonism nor increased cytotoxicity was found for most combinations, suggesting that MCZ could be added to different TB treatment regimens without any significant adverse effects., (Copyright © 2018 American Society for Microbiology.)

New drugs are needed to control the current tuberculosis (TB) pandemic caused by infection with Mycobacterium tuberculosis We report here on our work with AX-35, an arylvinylpiperazine amide, and four related analogs, which are potent antitubercular agents in vitro All five compounds showed good activity against M. tuberculosis in vitro and in infected THP-1 macrophages, while displaying only mild cytotoxicity. Isolation and characterization of M. tuberculosis -resistant mutants to the arylvinylpiperazine amide derivative AX-35 revealed mutations in the qcrB gene encoding a subunit of cytochrome bc 1 oxidase, one of two terminal oxidases of the electron transport chain. Cross-resistance studies, allelic exchange, transcriptomic analyses, and bioenergetic flux assays provided conclusive evidence that the cytochrome bc 1 -aa 3 is the target of AX-35, although the compound appears to interact differently with the quinol binding pocket compared to previous QcrB inhibitors. The transcriptomic and bioenergetic profiles of M. tuberculosis treated with AX-35 were similar to those generated by other cytochrome bc 1 oxidase inhibitors, including the compensatory role of the alternate terminal oxidase cytochrome bd in respiratory adaptation. In the absence of cytochrome bd oxidase, AX-35 was bactericidal against M. tuberculosis Finally, AX-35 and its analogs were active in an acute mouse model of TB infection, with two analogs displaying improved activity over the parent compound. Our findings will guide future lead optimization to produce a drug candidate for the treatment of TB and other mycobacterial diseases, including Buruli ulcer and leprosy. IMPORTANCE New drugs against Mycobacterium tuberculosis are urgently needed to deal with the current global TB pandemic. We report here on the discovery of a series of arylvinylpiperazine amides (AX-35 to AX-39) that represent a promising new family of compounds with potent in vitro and in vivo activities against M. tuberculosis AX compounds target the QcrB subunit of the cytochrome bc 1 terminal oxidase with a different mode of interaction compared to those of known QcrB inhibitors. This study provides the first multifaceted validation of QcrB inhibition by recombineering-mediated allelic exchange, gene expression profiling, and bioenergetic flux studies. It also provides further evidence for the compensatory role of cytochrome bd oxidase upon QcrB inhibition. In the absence of cytochrome bd oxidase, AX compounds are bactericidal, an encouraging property for future antimycobacterial drug development., (Copyright © 2018 Foo et al.)

Macozinone (MCZ) is a tuberculosis (TB) drug candidate that specifically targets the essential flavoenzyme DprE1, thereby blocking synthesis of the cell wall precursor decaprenyl phosphoarabinose (DPA) and provoking lysis of Mycobacterium tuberculosis As part of the MCZ backup program, we exploited structure-guided drug design to produce a new series of sulfone-containing derivatives, 2-sulfonylpiperazin 8-nitro 6-trifluoromethyl 1,3-benzothiazin-4-one, or sPBTZ. These compounds are less active than MCZ but have a better solubility profile, and some derivatives display enhanced stability in microsomal assays. DprE1 was efficiently inhibited by sPBTZ, and covalent adducts with the active-site cysteine residue (C387) were formed. However, despite the H-bonding potential of the sulfone group, no additional bonds were seen in the crystal structure of the sPBTZ-DprE1 complex with compound 11326127 compared to MCZ. Compound 11626091, the most advanced sPBTZ, displayed good antitubercular activity in the murine model of chronic TB but was less effective than MCZ. Nonetheless, further testing of this MCZ backup compound is warranted as part of combination treatment with other TB drugs., (Copyright © 2018 Piton et al.)

The flavoenzyme DprE1 catalyses a crucial step in arabinan production for cell wall biosynthesis in Mycobacterium tuberculosis and is a highly vulnerable drug target. It was first discovered using benzothiazinones (BTZ): exquisitely potent bactericidal agents that are being developed as drugs to treat tuberculosis. Subsequently, many compounds with diverse scaffolds were found to act as either covalent or noncovalent DprE1 inhibitors. Covalent inhibitors, like the BTZ, are all nitroaromatic compounds that serve as suicide substrates after DprE1-mediated nitroreduction. Here, we describe how high-resolution structures of DprE1, alone and in complex with various ligands, explain enzyme activity and inhibition., (Copyright © 2016 Elsevier Ltd. All rights reserved.)

Benzothiazinones (BTZs) are a class of compounds found to be extremely potent against both drug-susceptible and drug-resistant Mycobacterium tuberculosis strains. The potency of BTZs is explained by their specificity for their target decaprenylphosphoryl-d-ribose oxidase (DprE1), in particular by covalent binding of the activated form of the compound to the critical cysteine 387 residue of the enzyme. To probe the role of C387, we used promiscuous site-directed mutagenesis to introduce other codons at this position into dprE1 of M. tuberculosis The resultant viable BTZ-resistant mutants were characterized in vitro, ex vivo, and biochemically to gain insight into the effects of these mutations on DprE1 function and on M. tuberculosis Five different mutations (C387G, C387A, C387S, C387N, and C387T) conferred various levels of resistance to BTZ and exhibited different phenotypes. The C387G and C387N mutations resulted in a lower growth rate of the mycobacterium on solid medium, which could be attributed to the significant decrease in the catalytic efficiency of the DprE1 enzyme. All five mutations rendered the mycobacterium less cytotoxic to macrophages. Finally, differences in the potencies of covalent and noncovalent DprE1 inhibitors in the presence of C387 mutations were revealed by enzymatic assays. As expected from the mechanism of action, the covalent inhibitor PBTZ169 only partially inhibited the mutant DprE1 enzymes compared to the near-complete inhibition with a noncovalent DprE1 inhibitor, Ty38c. This study emphasizes the importance of the C387 residue for DprE1 activity and for the killing action of covalent inhibitors such as BTZs and other recently identified nitroaromatic inhibitors., (Copyright © 2016, American Society for Microbiology. All Rights Reserved.)

The complexities of DNA recognition by transcription factors (TFs) with multiple Cys2-His2 zinc fingers (C2H2-ZFs) remain poorly studied. We previously reported a mutation (R1092W) in the C2H2-ZF TF Zfp335 that led to selective loss of binding at a subset of targets, although the basis for this effect was unclear. We show that Zfp335 binds DNA and drives transcription via recognition of two distinct consensus motifs by separate ZF clusters and identify the specific motif interaction disrupted by R1092W. Our work presents Zfp335 as a model for understanding how C2H2-ZF TFs may use multiple recognition motifs to control gene expression., (© 2016 Han et al.; Published by Cold Spring Harbor Laboratory Press.)

Anticoagulant therapy is used for the prevention and treatment of thromboembolic disorders. Blood coagulation is initiated by the interaction of factor VIIa (FVIIa) with membrane-bound tissue factor (TF) to form the extrinsic tenase complex which activates FX to FXa. Thus, it is an important target for the development of novel anticoagulants. Here, we report the isolation and characterization of a novel anticoagulant ringhalexin from the venom of Hemachatus haemachatus (African Ringhals Cobra). Amino acid sequence of the protein indicates that it belongs to the three-finger toxin family and exhibits 94% identity to an uncharacterized Neurotoxin-like protein NTL2 from Naja atra. Ringhalexin inhibited FX activation by extrinsic tenase complex with an IC50 of 123.8 ± 9.54 nM. It is a mixed-type inhibitor with the kinetic constants, Ki and Ki' of 84.25 ± 3.53 nM and 152.5 ± 11.32 nM, respectively. Ringhalexin also exhibits a weak, irreversible neurotoxicity on chick biventer cervicis muscle preparations. Subsequently, the three-dimensional structure of ringhalexin was determined at 2.95 Å resolution. This study for the first time reports the structure of an anticoagulant three-finger toxin. Thus, ringhalexin is a potent inhibitor of the FX activation by extrinsic tenase complex and a weak, irreversible neurotoxin.

Mutations in the leucine-rich repeat kinase 2 (LRRK2) gene cause late-onset, autosomal dominant Parkinson's disease (PD). The clinical and neurochemical features of LRRK2-linked PD are similar to idiopathic disease although neuropathology is somewhat heterogeneous. Dominant mutations in LRRK2 precipitate neurodegeneration through a toxic gain-of-function mechanism which can be modeled in transgenic mice overexpressing human LRRK2 variants. A number of LRRK2 transgenic mouse models have been developed that display abnormalities in dopaminergic neurotransmission and alterations in tau metabolism yet without consistently inducing dopaminergic neurodegeneration. To directly explore the impact of mutant LRRK2 on the nigrostriatal dopaminergic pathway, we developed conditional transgenic mice that selectively express human R1441C LRRK2 in dopaminergic neurons from the endogenous murine ROSA26 promoter. The expression of R1441C LRRK2 does not induce the degeneration of substantia nigra dopaminergic neurons or striatal dopamine deficits in mice up to 2years of age, and fails to precipitate abnormal protein inclusions containing alpha-synuclein, tau, ubiquitin or autophagy markers (LC3 and p62). Furthermore, mice expressing R1441C LRRK2 exhibit normal motor activity and olfactory function with increasing age. Intriguingly, the expression of R1441C LRRK2 induces age-dependent abnormalities of the nuclear envelope in nigral dopaminergic neurons including reduced nuclear circularity and increased invaginations of the nuclear envelope. In addition, R1441C LRRK2 mice display increased neurite complexity of cultured midbrain dopaminergic neurons. Collectively, these novel R1441C LRRK2 conditional transgenic mice reveal altered dopaminergic neuronal morphology with advancing age, and provide a useful tool for exploring the pathogenic mechanisms underlying the R1441C LRRK2 mutation in PD., (Copyright © 2014 Elsevier Inc. All rights reserved.)

The generation of naïve T lymphocytes is critical for immune function yet the mechanisms governing their maturation remain incompletely understood. We have identified a mouse mutant, bloto, that harbors a hypomorphic mutation in the zinc finger protein Zfp335. Zfp335(bloto/bloto) mice exhibit a naïve T cell deficiency due to an intrinsic developmental defect that begins to manifest in the thymus and continues into the periphery, affecting T cells that have recently undergone thymic egress. The effects of Zfp335(bloto) are multigenic and cannot be attributed to altered thymic selection, proliferation or Bcl2-dependent survival. Zfp335 binds to promoter regions via a consensus motif, and its target genes are enriched in categories related to protein metabolism, mitochondrial function, and transcriptional regulation. Restoring the expression of one target, Ankle2, partially rescues T cell maturation. These findings identify Zfp335 as a transcription factor and essential regulator of late-stage intrathymic and post-thymic T cell maturation.

Introduction: Obesity is a real and dangerous problem with rising incidence. This study aimed to examine public perceptions of obesity as a disease, as well as the community's impression of the various modalities in the management of obesity, particularly bariatric surgery, in Singapore., Methods: Volunteers from the public were approached in 2010 to complete a questionnaire that collected both their demographic data and perceptions on obesity, management of obesity and bariatric surgery. The perceptions of the respondents were analysed based on various demographic factors., Results: The incidence of obesity was 16.8%. Consistent with previous studies in developed countries, the highest mean body mass index was noted in the age range of 41-50 years. 95.6% of the population surveyed were aware that obesity was related to significant medical conditions. 60% of the population surveyed had attempted weight loss in the past, with 41.7% expressing lack of success, and 58.4% expressed that they had not heard of surgery as a modality., Conclusion: This study shows that the majority of the population understands the significance of obesity as a medical problem and would take steps to combat it. However, there is suboptimal knowledge of the various modalities of managing obesity as a disease, especially with regard to surgical options. More education on the modalities available and access to them would help in the combat of this obesity epidemic.

Background: The acute abdomen accounts for up to 40% of all emergency surgical hospital admissions and a large proportion are secondary to gastrointestinal perforation. Studies have shown the superiority of the abdominal CT over upright chest radiographs in demonstrating free intraperitoneal air. Spontaneous perforated pyometra is a rare cause of the surgical acute abdomen with free intraperitoneal air. Only 38 cases have been reported worldwide., Case Report: We report 2 cases of spontaneously perforated pyometra in our hospital's general surgery department. Both underwent exploratory laparotomy: one had a total hysterectomy and bilateral salpingo-oophorectomy, while the other had an evacuation of the uterine cavity, primary repair of uterine perforation and a peritoneal washout. A literature search was conducted and all reported cases reviewed in order to describe the clinical presentations and management of the condition. Of the 40 cases to date, including 2 of our cases, the most common presenting symptoms were abdominal pain (97.5%), fever (37.5%) and vomiting (25.0%). The main indication for exploratory laparotomy was pneumoperitoneum (97.5%)., Conclusions: Pyometra is an unusual but serious condition in elderly women presenting with an acute abdomen. A high index of suspicion is needed to make the appropriate diagnosis.

Snake venoms are a mixture of pharmacologically active proteins and polypeptides that have led to the development of molecular probes and therapeutic agents. Here, we describe the structural and functional characterization of a novel neurotoxin, haditoxin, from the venom of Ophiophagus hannah (King cobra). Haditoxin exhibited novel pharmacology with antagonism toward muscle (alphabetagammadelta) and neuronal (alpha(7), alpha(3)beta(2), and alpha(4)beta(2)) nicotinic acetylcholine receptors (nAChRs) with highest affinity for alpha(7)-nAChRs. The high resolution (1.5 A) crystal structure revealed haditoxin to be a homodimer, like kappa-neurotoxins, which target neuronal alpha(3)beta(2)- and alpha(4)beta(2)-nAChRs. Interestingly however, the monomeric subunits of haditoxin were composed of a three-finger protein fold typical of curaremimetic short-chain alpha-neurotoxins. Biochemical studies confirmed that it existed as a non-covalent dimer species in solution. Its structural similarity to short-chain alpha-neurotoxins and kappa-neurotoxins notwithstanding, haditoxin exhibited unique blockade of alpha(7)-nAChRs (IC(50) 180 nm), which is recognized by neither short-chain alpha-neurotoxins nor kappa-neurotoxins. This is the first report of a dimeric short-chain alpha-neurotoxin interacting with neuronal alpha(7)-nAChRs as well as the first homodimeric three-finger toxin to interact with muscle nAChRs.

Malignant gastrocolic fistula formation is a rare complication of gastric carcinoma. We report a cachectic 82-year-old woman who presented with upper abdominal pain, diarrhoea, loss of weight and loss of appetite. Further investigation of her symptoms revealed a gastrocolic fistula connecting the ulcerated tumour of the body of the stomach to the splenic flexure of the colon.

A novel heterodimeric three-finger neurotoxin, irditoxin, was isolated from venom of the brown treesnake Boiga irregularis (Colubridae). Irditoxin subunit amino acid sequences were determined by Edman degradation and cDNA sequencing. The crystal structure revealed two subunits with a three-finger protein fold, typical for "nonconventional" toxins such as denmotoxin, bucandin, and candoxin. This is the first colubrid three-finger toxin dimer, covalently connected via an interchain disulfide bond. Irditoxin showed taxon-specific lethality toward birds and lizards and was nontoxic toward mice. It produced a potent neuromuscular blockade at the avian neuromuscular junction (IC(50)=10 nM), comparable to alpha-bungarotoxin, but was three orders of magnitude less effective at the mammalian neuromuscular junction. Covalently linked heterodimeric three-finger toxins found in colubrid venoms constitute a new class of venom peptides, which may be a useful source of new neurobiology probes and therapeutic leads.

The present study aimed to investigate the role of hydrogen sulphide (H2S) in the cardioprotection induced by ischemic postconditioning and to examine the underlying mechanisms. Cardiodynamics and myocardial infarction were measured in isolated rat hearts. Postconditioning with six episodes of 10-s ischemia (IPostC) significantly improved cardiodynamic function, which was attenuated by the blockade of endogenous H2S production with d-l-propargylglycine. Moreover, IPostC significantly stimulated H2S synthesis enzyme activity during the early period of reperfusion. However, d-l-propargylglycine only attenuated the IPostC-induced activation of PKC-alpha and PKC-epsilon but not that of PKC-delta, Akt, and endothelial nitric oxide synthase (eNOS). These data suggest that endogenous H2S contributes partially to the cardioprotection of IPostC via stimulating PKC-alpha and PKC-epsilon. Postconditioning with six episodes of a 10-s infusion of NaHS (SPostC) or 2 min continuous NaHS infusion (SPostC2) stimulated activities of Akt and PKC, improved the cardiodynamic performances, and reduced myocardial infarct size. The blockade of Akt with LY-294002 (15 microM) or PKC with chelerythrine (10 microM) abolished the cardioprotection induced by H2S postconditioning. SPostC2, but not SPostC, also additionally stimulated eNOS. We conclude that endogenous H2S contributes to IPostC-induced cardioprotection. H2S postconditioning confers the protective effects against ischemia-reperfusion injury through the activation of Akt, PKC, and eNOS pathways.

Motivation: The need for accurate and efficient tools for computational RNA structure analysis has become increasingly apparent over the last several years: RNA folding algorithms underlie numerous applications in bioinformatics, ranging from microarray probe selection to de novo non-coding RNA gene prediction. In this work, we present RAF (RNA Alignment and Folding), an efficient algorithm for simultaneous alignment and consensus folding of unaligned RNA sequences. Algorithmically, RAF exploits sparsity in the set of likely pairing and alignment candidates for each nucleotide (as identified by the CONTRAfold or CONTRAlign programs) to achieve an effectively quadratic running time for simultaneous pairwise alignment and folding. RAF's fast sparse dynamic programming, in turn, serves as the inference engine within a discriminative machine learning algorithm for parameter estimation., Results: In cross-validated benchmark tests, RAF achieves accuracies equaling or surpassing the current best approaches for RNA multiple sequence secondary structure prediction. However, RAF requires nearly an order of magnitude less time than other simultaneous folding and alignment methods, thus making it especially appropriate for high-throughput studies., Availability: Source code for RAF is available at:http://contra.stanford.edu/contrafold/.

Multiprotein complexes play central roles in many cellular pathways. Although many high-throughput experimental techniques have already enabled systematic screening of pairwise protein-protein interactions en masse, the amount of experimentally determined protein complex data has remained relatively lacking. As such, researchers have begun to exploit the vast amount of pairwise interaction data to help discover new protein complexes. However, mining for protein complexes in interaction networks is not an easy task because there are many data artefacts in the underlying protein-protein interaction data due to the limitations in the current high-throughput screening methods. We propose a novel DECAFF (Dense-neighborhood Extraction using Connectivity and conFidence Features) algorithm to mine for dense and reliable subgraphs in protein interaction networks. Our method is devised to address two major limitations in current high throughout protein interaction data, namely, incompleteness and high data noise. Experimental results with yeast protein interaction data show that the interaction subgraphs discovered by DECAFF matched significantly better with actual protein complexes than other existing approaches. Our results demonstrate that pairwise protein interaction networks can be effectively mined to discover new protein complexes, provided that the data artefacts in the underlying interaction data are taken into account adequately.

Introduction: Laparoscopic adjustable gastric banding (LAGB) has been used for the treatment of obesity, being shown in western populations to be a safe and effective option. We present the results of our experience in the use of LAGB in the treatment of obesity., Methods: A retrospective review of all patients who underwent LAGB from February 1999 to June 2004 was made from a prospectively collected database. Pre-operative comorbidities, height, weight and body mass index (BMI) were recorded and compared with post-operative results. Operative times, length of stay and complications were also noted., Results: 38 consecutive patients underwent LAGB. The 16 male and 22 female patients had a mean age of 37.6 years (range 19 to 62 years) and a mean BMI of 42.7 kg per square metre (range 28.8 to 78.4 kg per square metre). Nine were done utilising the perigastric approach, and the other 29 with the pars flaccida approach. There were no open conversions. Overall peri-operative morbidity was encountered in two (5.3 percent) patients, with a minor liver laceration and a capsular splenic laceration in separate patients. There were no post-operative complications. One (2.7 percent) patient had the band removed at 29 months post-operatively on request. There were no mortalities. Mean follow-up period was 13 months (range one to 56 months). Mean BMI decreased to 40.7, 38.7, 37.4, 34.1 and 32.9 kg per square metre, respectively, at one week, one, six, 12 and 18 months post-operatively., Conclusion: LAGB is a feasible modality in the multifaceted approach to treatment of morbid obesity. The results from our follow-up are comparable to larger series in western populations, with a similar safety profile.