The septate junction component bark beetle is required for Drosophila intestinal barrier function and homeostasis.

Age-related loss of intestinal barrier function has been documented across species, but the causes remain unknown. The intestinal barrier is maintained by tight junctions (TJs) in mammals and septate junctions (SJs) in insects. Specialized TJs/SJs, called tricellular junctions (TCJs), are located at the nexus of three adjacent cells, and we have shown that aging results in changes to TCJs in intestines of adult Drosophila melanogaster. We now demonstrate that localization of the TCJ protein bark beetle (Bark) decreases in aged flies. Depletion of bark from enterocytes in young flies led to hallmarks of intestinal aging and shortened lifespan, whereas depletion of bark in progenitor cells reduced Notch activity, biasing differentiation toward the secretory lineage. Our data implicate Bark in EC maturation and maintenance of intestinal barrier integrity. Understanding the assembly and maintenance of TCJs to ensure barrier integrity may lead to strategies to improve tissue integrity when function is compromised.

SelfPose: 3D Egocentric Pose Estimation From a Headset Mounted Camera.

We present a new solution to egocentric 3D body pose estimation from monocular images captured from a downward looking fish-eye camera installed on the rim of a head mounted virtual reality device. This unusual viewpoint leads to images with unique visual appearance, characterized by severe self-occlusions and strong perspective distortions that result in a drastic difference in resolution between lower and upper body. We propose a new encoder-decoder architecture with a novel multi-branch decoder designed specifically to account for the varying uncertainty in 2D joint locations. Our quantitative evaluation, both on synthetic and real-world datasets, shows that our strategy leads to substantial improvements in accuracy over state of the art egocentric pose estimation approaches. To tackle the severe lack of labelled training data for egocentric 3D pose estimation we also introduced a large-scale photo-realistic synthetic dataset. xR-EgoPose offers 383K frames of high quality renderings of people with diverse skin tones, body shapes and clothing, in a variety of backgrounds and lighting conditions, performing a range of actions. Our experiments show that the high variability in our new synthetic training corpus leads to good generalization to real world footage and to state of the art results on real world datasets with ground truth. Moreover, an evaluation on the Human3.6M benchmark shows that the performance of our method is on par with top performing approaches on the more classic problem of 3D human pose from a third person viewpoint.

Connections among Land Use, Water Quality, Biodiversity of Aquatic Invertebrates, and Fish Behavior in Amazon Rivers.

Rivers in the Amazon have among the greatest biodiversity in the world. The Xingu River, one of the tributaries of the Amazon River, has a length of 1640 km, draining 510,000 km2 in one of the most protected regions on the planet. The Middle Xingu region in Brazil has been highly impacted by mining and livestock farming, leading to habitat fragmentation due to altered water quality. Therefore, comparing two rivers (the preserved Xingu River and the impacted Fresco River) and their confluence, the aims of the present study were to (1) assess the land uses in the hydrographic basin; (2) determine the water quality by measurements of turbidity, total solids, and metals (Cd, Cu, Fe, Mn, Pb, Zn, and Hg); (3) compare the zooplankton biodiversity; and (4) to evaluate the avoidance behavior of fish (Astyanax bimaculatus) when exposed to waters from the Xingu and Fresco Rivers. Zooplankton were grouped and counted down to the family level. For the analysis of fish avoidance, a multi-compartment system was used. The forest class predominated at the study locations, accounting for 57.6%, 60.8%, and 63.9% of the total area at P1XR, P2FR, and P3XFR, respectively, although since 1985, at the same points, the forest had been reduced by 31.3%, 25.7%, and 27.9%. The Xingu River presented almost 300% more invertebrate families than the Fresco River, and the fish population preferred its waters (>50%). The inputs from the Fresco River impacted the water quality of the Xingu River, leading to reductions in local invertebrate biodiversity and potential habitats for fish in a typical case of habitat fragmentation due to anthropic factors.

Deregulation of phenylalanine biosynthesis evolved with the emergence of vascular plants.

Phenylalanine (Phe) is the precursor of essential secondary products in plants. Here we show that a key, rate-limiting step in Phe biosynthesis, which is catalyzed by arogenate dehydratase, experienced feedback de-regulation during evolution. Enzymes from microorganisms and type-I ADTs from plants are strongly feedback-inhibited by Phe, while type-II isoforms remain active at high levels of Phe. We have found that type-II ADTs are widespread across seed plants and their overproduction resulted in a dramatic accumulation of Phe in planta, reaching levels up to 40 times higher than those observed following the expression of type-I enzymes. Punctual changes in the allosteric binding site of Phe and adjacent region are responsible for the observed relaxed regulation. The phylogeny of plant ADTs evidences that the emergence of type-II isoforms with relaxed regulation occurred at some point in the transition between nonvascular plants and tracheophytes, enabling the massive production of Phe-derived compounds, primarily lignin, a hallmark of vascular plants.

3D Human Pose, Shape and Texture From Low-Resolution Images and Videos.

3D human pose and shape estimation from monocular images has been an active research area in computer vision. Existing deep learning methods for this task rely on high-resolution input, which however, is not always available in many scenarios such as video surveillance and sports broadcasting. Two common approaches to deal with low-resolution images are applying super-resolution techniques to the input, which may result in unpleasant artifacts, or simply training one model for each resolution, which is impractical in many realistic applications. To address the above issues, this paper proposes a novel algorithm called RSC-Net, which consists of a Resolution-aware network, a Self-supervision loss, and a Contrastive learning scheme. The proposed method is able to learn 3D body pose and shape across different resolutions with one single model. The self-supervision loss enforces scale-consistency of the output, and the contrastive learning scheme enforces scale-consistency of the deep features. We show that both these new losses provide robustness when learning in a weakly-supervised manner. Moreover, we extend the RSC-Net to handle low-resolution videos and apply it to reconstruct textured 3D pedestrians from low-resolution input. Extensive experiments demonstrate that the RSC-Net can achieve consistently better results than the state-of-the-art methods for challenging low-resolution images.

Oxidative stress responses after exposure to triclosan sublethal concentrations: an integrated biomarker approach with a native (Corydoras paleatus) and a model fish species (Danio rerio).

Triclosan (TCS) is a synthetic broad-spectrum antimicrobial agent commonly used world-wide in a range of personal care and sanitizing products detected frequently in aquatic ecosystems. The aim of this study was to examine biochemical markers responses triggered by TCS in Danio rerio and in a native South American fish species (Corydoras paleatus). Further, an integrated approach comparing both test fish species was undertaken. These fish organisms were exposed to 100 or 189 µg TCS/L for 48 h. The activities of catalase (CAT), glutathione-s-transferase (GST), superoxide dismutase (SOD), and lipid peroxidation levels (LPO) and total antioxidant capacity against peroxyl radicals (ACAP) were determined in liver, gills, and brain. Acetylcholinesterase activity (AChE) was measured in the brain. Multivariate analysis showed that the most sensitive hepatic parameters were activities of GST and SOD for C. paleatus while LPO levels were for D. rerio. In gills the same parameters were responsive for C. paleatus but CAT in D. rerio. ACAP and GST activity were responsive parameters in brain of both species. Integrated biomarker responses (IBR) index demonstrated similar trends in both species suggesting this parameter might serve as a useful tool for quantification of integrated responses induced by TCS.

Ivermectin: A multilevel approach to evaluate effects in Prochilodus lineatus (Valenciennes, 1836) (Characiformes, Prochilodontidae), an inland fishery species.

Ivermectin (IVM) is one of the most widely used antiparasitics worldwide. It is a potent and effective drug for treatment and prevention of internal and external parasitic infections of livestock and humans. IVM is excreted unchanged in manure of treated animals. Thus, residues of IVM may reach aquatic systems, affecting non-target organisms such as fish. Although the presence of IVM in aquatic environments has been reported, a multilevel approach (from cellular to behavioral responses) is necessary to determine the health of exposed organisms and the environmental risks associated. The aim of the present study was to investigate the response of the Neotropical fish Prochilodus lineatus, one of the main target species of South American freshwater fisheries, exposed to environmental concentrations of IVM: low (0.5 µg L-1) and high (1.5 µg L-1). Behavioral responses were assessed in juvenile fish and included water column use, routine swimming, total distance travelled, total activity time and Maximum swimming speed achieved during the escape response. Biochemical/oxidative stress responses assessed included brain acetylcholinesterase (AChE), catalase (CAT) and glutathione S-transferase (GST) activities; total antioxidant competence against peroxyl radicals (ACAP) and lipid oxidative damage (TBARs). Hematological biomarker responses included blood glucose levels, hematocrit, hemoglobin concentration, mean corpuscular hemoglobin, mean corpuscular hemoglobin concentration, and mean corpuscular volume. Condition factor and hepatosomatic index were also calculated. The lowest IVM concentration caused a significant decrease in GST activity and maximum swimming speed during the escape response. Multivariate analysis with biochemical/stress and behavioral data revealed overall effects of IVM treatments. This multilevel analysis shows detrimental effects related to swimming behavior and predator avoidance which could affect population size and size-structure of P. lineatus. To our knowledge this is the first attempt to assess the effects of IVM on Neotropical fishes using an integrative approach based on biomarkers from different levels of biological organization.

Multibiomarker responses and bioaccumulation of fipronil in Prochilodus lineatus exposed to spiked sediments: Oxidative stress and antioxidant defenses.

Fipronil is a current use pesticide, widely used in many crops, commonly adsorbed to sediments of aquatic environments. The purpose of this study was to evaluate the biomarker responses and fipronil distribution pattern in different matrixes (fish, sediment and water) after juveniles P. lineatus exposure at two environmental concentrations (5.5 and 82 µg kg--1) of fipronil-spiked sediments. The levels of oxidized proteins (PO), lipid peroxidation (LPO), and enzymatic activity of superoxide dismutase (SOD), reduced glutathione content (GSH), antioxidant capacity against peroxyls (ACAP) and acetylcholinesterase (AChE) were evaluated in liver, gills and brain. Concentrations of fipronil and its metabolites (f. desulfinyl, f sulphpHide and f. sulfone) were quantified by GC-ECD. F. desulfinyl was the major metabolite found in all matrixes, followed by f. sulphide in sediments, while f. sulfone was mainly accumulated in fish. Fipronil promoted oxidative stress in P. lineatus, as evidenced by the increases in LPO and PO levels and the decrease brain AChE activity. Fish exposed at both concentrations showed significant decrease in antioxidant capacity. Alterations in the antioxidant defenses system was evidenced in all organs. These results suggest that the occurrence of fipronil in aquatic environments can generate oxidative stress at different levels in P. lineatus, showing that this species is highly sensitive to the deleterious effects of fipronil and metabolites.

Nanosilver and Silver Nitrate Toxicity in Ex Vivo-Exposed Gills of Fish and Mitigation by Humic Acids.

Silver nanoparticles (AgNP) are unique because of their biocide properties. Once released to environment, AgNP interact with the natural organic matter which impact on their fate, dispersion, and ultimate toxicity. We carried out an ex vivo exposure of gill of Corydoras paleatus fish to 100 µg L-1 of AgNP or AgNO3, alone and in combination with 10 mg L-1 of humic acids (HA), with the aim to evaluate the potential mitigation of HA on AgNP toxic effects. We analyzed Ag accumulation and oxidative stress biomarkers. The results showed high bioaccumulation after the AgNO3+HA exposure. An inhibition of glutathione-S-transferase enzymatic activity and depletion of reduced glutathione levels were registered after the AgNO3 exposure, and increased lipid peroxidation levels in the case of AgNP one. Oxidative responses were mitigated when the HA were present in the media. Overall, the knowledge about the fate of this emergent pollutant was deepened through this study.

Exposure to environmental concentrations of fipronil induces biochemical changes on a neotropical freshwater fish.

Fipronil is a broad-use insecticide with severe toxicity to fish. Biomarkers responses and bioaccumulation were evaluated on Prochilodus lineatus after exposure to environmentally relevant concentrations of fipronil (0.5 µg L-1, 9 µg L-1, and 100 µg L-1) in a prolonged flow-through assay and ex vivo gills short-term exposition. Lipid peroxidation (LPO), oxidatively modified proteins (PO), the activity of superoxide dismutase (SOD), the content of reduced glutathione (GSH), antioxidant capacity against peroxyles (ACAP), and acetylcholinesterase (AChE) were evaluated. Besides, levels of fipronil and metabolites were analyzed by GC-ECD. At the end of the flow-through assay, fipronil, Fp. sulfone and Fp. desulfinyl were detected in fish, being liver the target organ. Fipronil prolonged exposition promoted oxidative damage in lipids and proteins, alterations in the defense system and low-antioxidant capacity in organs of P. lineatus. The brain AChE activity was affected after prolonged exposition. Ex vivo gills exposition to fipronil promoted changes in antioxidant capacity and damage to lipids, providing a fast and suitable test to assess the pesticide exposure in fish. The results revealed that fipronil at environmental concentrations would be an inducer of oxidative stress in this fish, becoming a vulnerable species to the effects of fipronil in aquatic environments.

Mitigation of silver nanoparticle toxicity by humic acids in gills of Piaractus mesopotamicus fish.

Silver nanoparticles (AgNPs) are one of the most produced nanoproducts due to their unique biocide properties. The natural organic matter has an important impact on nanoparticle's dispersion as it may alter their fate and transport, as well as their bioavailability and toxicity. Therefore, this study aimed to evaluate the mitigatory effect of humic acids (HAs) on AgNP toxicity. For this purpose, we carried out an ex vivo exposure of gill of Piaractus mesopotamicus fish to 100 µg L-1 of AgNPs or AgNO3, alone and in combination with 10 mg L-1 of HAs. In parallel, a complete AgNP characterization in the media, including the presence of HAs, was provided, and the Ag+ release was measured. We analyzed Ag bioaccumulation, antioxidant enzymes activities, lipid peroxidation, antioxidant capacity against peroxyl radicals, and reduced glutathione levels in fish tissue. Our results indicated the Ag+ release from AgNPs decreased 28% when the HAs were present in the media. The Ag accumulation in gill tissue exposed to AgNPs alone was higher than the AgNO3 exposure, and sixfold higher than the treatment with the HA addition. Moreover, after both Ag forms, the catalase enzyme augmented its activity. However, those responses were mitigated when the HAs were present in the media. Then, our results suggested the mitigation by HAs under the exposure to both Ag forms, providing valuable information about the fate and behavior of this emergent pollutant.

Weakly-Supervised Learning of Category-Specific 3D Object Shapes.

Category-specific 3D object shape models have greatly boosted the recent advances in object detection, recognition and segmentation. However, even the most advanced approach for learning 3D object shapes still requires heavy manual annotations on large-scale 2D images. Such annotations include object categories, object keypoints, and figure-ground segmentation for the instances in each image. In particular, annotating figure-ground segmentation is unbearably labor-intensive and time-consuming. To address this problem, this paper devotes to learn category-specific 3D shape models under weak supervision, where only object categories and keypoints are required to be manually annotated on the training 2D images. By exploring the underlying relationship between two tasks: object segmentation and category-specific 3D shape reconstruction, we propose a novel weakly-supervised learning framework to jointly address these two tasks and combine them to boost the final performance of the learned 3D shape models. Moreover, learning without using figure-ground segmentation leads to ambiguous solutions. To this end, we develop the confidence weighting schemes in the viewpoint estimation and 3D shape learning procedure. These schemes effectively reduce the confusion caused by the noisy data and thus increase the chances for recovering more reliable 3D object shapes. Comprehensive experiments on the challenging PASCAL VOC benchmark show that our framework achieves comparable performance with the state-of-the-art methods that use expensive manual segmentation-level annotations. In addition, our experiments also demonstrate that our 3D shape models improve object segmentation performance.

Parkinson's Disease Tremor Detection in the Wild Using Wearable Accelerometers.

Continuous in-home monitoring of Parkinson's Disease (PD) symptoms might allow improvements in assessment of disease progression and treatment effects. As a first step towards this goal, we evaluate the feasibility of a wrist-worn wearable accelerometer system to detect PD tremor in the wild (uncontrolled scenarios). We evaluate the performance of several feature sets and classification algorithms for robust PD tremor detection in laboratory and wild settings. We report results for both laboratory data with accurate labels and wild data with weak labels. The best performance was obtained using a combination of a pre-processing module to extract information from the tremor spectrum (based on non-negative factorization) and a deep neural network for learning relevant features and detecting tremor segments. We show how the proposed method is able to predict patient self-report measures, and we propose a new metric for monitoring PD tremor (i.e., percentage of tremor over long periods of time), which may be easier to estimate the start and end time points of each tremor event while still providing clinically useful information.

Comparing laboratory and in-the-wild data for continuous Parkinson's Disease tremor detection.

Passive, continuous monitoring of Parkinson's Disease (PD) symptoms in the wild (i.e., in home environments) could improve disease management, thereby improving a patient's quality of life. We envision a system that uses machine learning to automatically detect PD symptoms from accelerometer data collected in the wild. Building such systems, however, is challenging because it is difficult to obtain labels of symptom occurrences in the wild. Many researchers therefore train machine learning algorithms on laboratory data with the assumption that findings will translate to the wild. This paper assesses how well laboratory data represents wild data by comparing PD symptom (tremor) detection performance of three models on both lab and wild data. Findings indicate that, for this application, laboratory data is not a good representation of wild data. Results also show that training on wild data, even though labels are less precise, leads to better performance on wild data than training on accurate labels from laboratory data.

Enzymes Involved in the Biosynthesis of Arginine from Ornithine in Maritime Pine (Pinus pinaster Ait.).

The amino acids arginine and ornithine are the precursors of a wide range of nitrogenous compounds in all living organisms. The metabolic conversion of ornithine into arginine is catalyzed by the sequential activities of the enzymes ornithine transcarbamylase (OTC), argininosuccinate synthetase (ASSY) and argininosuccinate lyase (ASL). Because of their roles in the urea cycle, these enzymes have been purified and extensively studied in a variety of animal models. However, the available information about their molecular characteristics, kinetic and regulatory properties is relatively limited in plants. In conifers, arginine plays a crucial role as a main constituent of N-rich storage proteins in seeds and serves as the main source of nitrogen for the germinating embryo. In this work, recombinant PpOTC, PpASSY and PpASL enzymes from maritime pine (Pinus pinaster Ait.) were produced in Escherichia coli to enable study of their molecular and kinetics properties. The results reported here provide a molecular basis for the regulation of arginine and ornithine metabolism at the enzymatic level, suggesting that the reaction catalyzed by OTC is a regulatory target in the homeostasis of ornithine pools that can be either used for the biosynthesis of arginine in plastids or other nitrogenous compounds in the cytosol.

Structural and Functional Characteristics of Two Molecular Variants of the Nitrogen Sensor PII in Maritime Pine.

High levels of nitrogen are stored as arginine during the last stages of seed formation in maritime pine (Pinus pinaster Aiton). The protein sensor PII regulates the feedback inhibition of arginine biosynthesis through interaction with the key enzyme N-acetylglutamate kinase (NAGK). In this study, the structural and functional characteristics of PII have been investigated in maritime pine to get insights into the regulation of arginine metabolism. Two different forms of PII have been identified, PpPIIa and PpPIIb, which differ in their amino acid sequence and most likely correspond to splicing variants of a single gene in the pine genome. Two PII variants are also present in other pine species but not in other conifers such as spruces. PpPIIa and PpPIIb are trimeric proteins for which structural modeling predicts similar tridimensional protein core structures. Both are located in the chloroplast, where the PII-target enzyme PpNAGK is also found. PpPIIa, PpPIIb, and PpNAGK have been recombinantly produced to investigate the formation of NAGK-PII complexes. The interaction of PpPIIa/PpPIIb and PpNAGK may be enhanced by glutamine and contribute to relieve the feedback inhibition of PpNAGK by arginine. Expression analysis of PpPII genes revealed that PpIIa transcripts were predominant during embryogenesis and germination. The potential roles of PpPIIa and PpPIIb in the regulation of arginine metabolism of maritime pine are discussed.

Transcriptional analysis of arogenate dehydratase genes identifies a link between phenylalanine biosynthesis and lignin biosynthesis.

Biogenesis of the secondary cell wall in trees involves the massive biosynthesis of the phenylalanine-derived polymer lignin. Arogenate dehydratase (ADT) catalyzes the last, and rate-limiting, step of the main pathway for phenylalanine biosynthesis. In this study, we found that transcript levels for several members of the large ADT gene family, including ADT-A and ADT-D, were enhanced in compression wood of maritime pine, a xylem tissue enriched in lignin. Transcriptomic analysis of maritime pine silenced for PpMYB8 revealed that this gene plays a critical role in coordinating the deposition of lignin with the biosynthesis of phenylalanine. Specifically, it was found that ADT-A and ADT-D were strongly down-regulated in PpMYB8-silenced plants and that they were transcriptionally regulated through direct interaction of this transcription factor with regulatory elements present in their promoters. Another transcription factor, PpHY5, exhibited an expression profile opposite to that of PpMYB8 and also interacted with specific regulatory elements of ADT-A and ADT-D genes, suggesting that it is involved in transcriptional regulation of phenylalanine biosynthesis. Taken together, our results reveal that PpMYB8 and PpHY5 are involved in the control of phenylalanine formation and its metabolic channeling for lignin biosynthesis and deposition during wood formation in maritime pine.

Characteristics, quality of life and control of respiratory allergic diseases caused by house dust mites (HDMs) in Spain: a cross-sectional study.

This multicentre, cross-sectional study conducted in Spain assessed the clinical characteristics and quality of life of patients who were aged 14-55 years and had allergic rhinitis and/or asthma, which was due to house dust mite sensitisation, for at least 2 years. Overall, 296 patients were included; 60% had allergic rhinitis (mostly persistent moderate-to-severe) and 40% had rhinitis and asthma (mostly intermittent or mild-to-persistent). Patients with rhinitis had moderately reduced quality of life, which was significantly worse in adults than adolescents. The impact of asthma on quality of life was less pronounced than that of rhinitis. Our findings show that allergic asthma and/or rhinitis due to house dust mites have a moderately negative impact on quality of life, particularly in adults, and that quality of life is significantly influenced by rhinitis.

Discriminative Optimization: Theory and Applications to Computer Vision.

Many computer vision problems are formulated as the optimization of a cost function. This approach faces two main challenges: designing a cost function with a local optimum at an acceptable solution, and developing an efficient numerical method to search for this optimum. While designing such functions is feasible in the noiseless case, the stability and location of local optima are mostly unknown under noise, occlusion, or missing data. In practice, this can result in undesirable local optima or not having a local optimum in the expected place. On the other hand, numerical optimization algorithms in high-dimensional spaces are typically local and often rely on expensive first or second order information to guide the search. To overcome these limitations, we propose Discriminative Optimization (DO), a method that learns search directions from data without the need of a cost function. DO explicitly learns a sequence of updates in the search space that leads to stationary points that correspond to the desired solutions. We provide a formal analysis of DO and illustrate its benefits in the problem of 3D registration, camera pose estimation, and image denoising. We show that DO outperformed or matched state-of-the-art algorithms in terms of accuracy, robustness, and computational efficiency.

Learning Facial Action Units with Spatiotemporal Cues and Multi-label Sampling.

Facial action units (AUs) may be represented spatially, temporally, and in terms of their correlation. Previous research focuses on one or another of these aspects or addresses them disjointly. We propose a hybrid network architecture that jointly models spatial and temporal representations and their correlation. In particular, we use a Convolutional Neural Network (CNN) to learn spatial representations, and a Long Short-Term Memory (LSTM) to model temporal dependencies among them. The outputs of CNNs and LSTMs are aggregated into a fusion network to produce per-frame prediction of multiple AUs. The hybrid network was compared to previous state-of-the-art approaches in two large FACS-coded video databases, GFT and BP4D, with over 400,000 AU-coded frames of spontaneous facial behavior in varied social contexts. Relative to standard multi-label CNN and feature-based state-of-the-art approaches, the hybrid system reduced person-specific biases and obtained increased accuracy for AU detection. To address class imbalance within and between batches during training the network, we introduce multi-labeling sampling strategies that further increase accuracy when AUs are relatively sparse. Finally, we provide visualization of the learned AU models, which, to the best of our best knowledge, reveal for the first time how machines see AUs.