Colonocyte keratins stabilize mitochondria and contribute to mitochondrial energy metabolism.

Keratin intermediate filaments form dynamic filamentous networks, which provide mechanical stability, scaffolding and protection against stress to epithelial cells. Keratins and other intermediate filaments have been increasingly linked to the regulation of mitochondrial function and homeostasis in different tissues and cell types. While deletion of keratin 8 (K8‒/‒) in mouse colon elicits a colitis-like phenotype, epithelial hyperproliferation and blunted mitochondrial ketogenesis, the role for K8 in colonocyte mitochondrial function and energy metabolism is unknown. We used two K8 knockout mouse models and CRISPR/Cas9 K8‒/‒ colorectal adenocarcinoma Caco-2 cells to answer this question. The results show that K8‒/‒ colonocyte mitochondria in vivo are smaller and rounder, and that mitochondrial motility is increased in K8‒/‒ Caco-2 cells. Furthermore, K8-/- Caco-2 cells displayed diminished mitochondrial respiration and decreased mitochondrial membrane potential compared to controls, whereas glycolysis was not affected. The levels of mitochondrial respiratory chain complex proteins and mitochondrial regulatory proteins mitofusin-2 and prohibitin were decreased both in vitro in K8‒/‒ Caco-2 cells and in vivo in K8‒/‒ mouse colonocytes, and re-expression of K8 into K8‒/‒ Caco-2 cells normalizes the mitofusin-2 levels. Mitochondrial Ca2+ is an important regulator of mitochondrial energy metabolism and homeostasis, and Caco-2 cells lacking K8 displayed decreased levels and altered dynamics of mitochondrial matrix and cytoplasmic Ca2+. In summary, these novel findings attribute an important role for colonocyte K8 in stabilizing mitochondrial shape and movement and maintaining mitochondrial respiration and Ca2+ signaling. Further, how these metabolically compromised colonocytes are capable of hyperproliferating presents an intriguing question for future studies.

Co-factor independent oxidases ncnN and actVA-3 are involved in the dimerization of benzoisochromanequinone antibiotics in naphthocyclinone and actinorhodin biosynthesis.

Streptomyces produce complex bioactive secondary metabolites with remarkable chemical diversity. Benzoisochromanequinone polyketides actinorhodin and naphthocyclinone are formed through dimerization of half-molecules via single or double carbon-carbon bonds, respectively. Here we sequenced the genome of S. arenae DSM40737 to identify the naphthocyclinone gene cluster and established heterologous production in S. albus J1074 by utilizing direct cluster capture techniques. Comparative sequence analysis uncovered ncnN and ncnM gene products as putative enzymes responsible for dimerization. Inactivation of ncnN that is homologous to atypical co-factor independent oxidases resulted in the accumulation of fogacin, which is likely a reduced shunt product of the true substrate for naphthocyclinone dimerization. In agreement, inactivation of the homologous actVA-3 in S. coelicolor M145 also led to significantly reduced production of actinorhodin. Previous work has identified the NAD(P)H-dependent reductase ActVA-4 as the key enzyme in actinorhodin dimerization, but surprisingly inactivation of the homologous ncnM did not abolish naphthocyclinone formation and the mutation may have been complemented by an endogenous gene product. Our data suggests that dimerization of benzoisochromanequinone polyketides require two-component reductase-oxidase systems.

Melanocortin 1 receptor regulates cholesterol and bile acid metabolism in the liver.

Melanocortin 1 receptor (MC1-R) is widely expressed in melanocytes and leukocytes and is thus strongly implicated in the regulation of skin pigmentation and inflammation. MC1-R has also been found in the rat and human liver, but its functional role has remained elusive. We hypothesized that MC1-R is functionally active in the liver and involved in the regulation of cholesterol and bile acid metabolism. We generated hepatocyte-specific MC1-R knock-out (Mc1r LKO) mice and phenotyped the mouse model for lipid profiles, liver histology, and bile acid levels. Mc1r LKO mice had significantly increased liver weight, which was accompanied by elevated levels of total cholesterol and triglycerides in the liver as well as in the plasma. These mice demonstrated also enhanced liver fibrosis and a disturbance in bile acid metabolism as evidenced by markedly reduced bile acid levels in the plasma and feces. Mechanistically, using HepG2 cells as an in vitro model, we found that selective activation of MC1-R in HepG2 cells reduced cellular cholesterol content and enhanced uptake of low- and high-density lipoprotein particles via a cAMP-independent mechanism. In conclusion, the present results demonstrate that MC1-R signaling in hepatocytes regulates cholesterol and bile acid metabolism and its deficiency leads to hypercholesterolemia and enhanced lipid accumulation and fibrosis in the liver.

Semi-Supervised Adversarial Auto-Encoder to Expedite Human Activity Recognition.

The study of human activity recognition concentrates on classifying human activities and the inference of human behavior using modern sensing technology. However, the issue of domain adaptation for inertial sensing-based human activity recognition (HAR) is still burdensome. The existing requirement of labeled training data for adapting such classifiers to every new person, device, or on-body location is a significant barrier to the widespread adoption of HAR-based applications, making this a challenge of high practical importance. We propose the semi-supervised HAR method to improve reconstruction and generation. It executes proper adaptation with unlabeled data without changes to a pre-trained HAR classifier. Our approach decouples VAE with adversarial learning to ensure robust classifier operation, without newly labeled training data, under changes to the individual activity and the on-body sensor position. Our proposed framework shows the empirical results using the publicly available benchmark dataset compared to state-of-art baselines, achieving competitive improvement for handling new and unlabeled activity. The result demonstrates SAA has achieved a 5% improvement in classification score compared to the existing HAR platform.

Detection of Image Level Forgery with Various Constraints Using DFDC Full and Sample Datasets.

The emergence of advanced machine learning or deep learning techniques such as autoencoders and generative adversarial networks, can generate images known as deepfakes, which astonishingly resemble the realistic images. These deepfake images are hard to distinguish from the real images and are being used unethically against famous personalities such as politicians, celebrities, and social workers. Hence, we propose a method to detect these deepfake images using a light weighted convolutional neural network (CNN). Our research is conducted with Deep Fake Detection Challenge (DFDC) full and sample datasets, where we compare the performance of our proposed model with various state-of-the-art pretrained models such as VGG-19, Xception and Inception-ResNet-v2. Furthermore, we perform the experiments with various resolutions maintaining 1:1 and 9:16 aspect ratios, which have not been explored for DFDC datasets by any other groups to date. Thus, the proposed model can flexibly accommodate various resolutions and aspect ratios, without being constrained to a specific resolution or aspect ratio for any type of image classification problem. While most of the reported research is limited to sample or preview DFDC datasets only, we have also attempted the testing on full DFDC datasets and presented the results. Contemplating the fact that the detailed results and resource analysis for various scenarios are provided in this research, the proposed deepfake detection method is anticipated to pave new avenues for deepfake detection research, that engages with DFDC datasets.

Semi-Supervised Adversarial Learning Using LSTM for Human Activity Recognition.

The training of Human Activity Recognition (HAR) models requires a substantial amount of labeled data. Unfortunately, despite being trained on enormous datasets, most current models have poor performance rates when evaluated against anonymous data from new users. Furthermore, due to the limits and problems of working with human users, capturing adequate data for each new user is not feasible. This paper presents semi-supervised adversarial learning using the LSTM (Long-short term memory) approach for human activity recognition. This proposed method trains annotated and unannotated data (anonymous data) by adapting the semi-supervised learning paradigms on which adversarial learning capitalizes to improve the learning capabilities in dealing with errors that appear in the process. Moreover, it adapts to the change in human activity routine and new activities, i.e., it does not require prior understanding and historical information. Simultaneously, this method is designed as a temporal interactive model instantiation and shows the capacity to estimate heteroscedastic uncertainty owing to inherent data ambiguity. Our methodology also benefits from multiple parallel input sequential data predicting an output exploiting the synchronized LSTM. The proposed method proved to be the best state-of-the-art method with more than 98% accuracy in implementation utilizing the publicly available datasets collected from the smart home environment facilitated with heterogeneous sensors. This technique is a novel approach for high-level human activity recognition and is likely to be a broad application prospect for HAR.

Epileptic-Net: An Improved Epileptic Seizure Detection System Using Dense Convolutional Block with Attention Network from EEG.

Epilepsy is a complex neurological condition that affects a large number of people worldwide. Electroencephalography (EEG) measures the electrical activity of the brain and is widely used in epilepsy diagnosis, but it usually requires manual inspection, which can be hours long, by a neurologist. Several automatic systems have been proposed to detect epilepsy but still have some unsolved issues. In this study, we proposed a dynamic method using a deep learning model (Epileptic-Net) to detect an epileptic seizure. The proposed method is largely heterogeneous and comprised of the dense convolutional blocks (DCB), feature attention modules (FAM), residual blocks (RB), and hypercolumn technique (HT). Firstly, DCB is used to get the discriminative features from the EEG samples. Then, FAM extracts the essential features from the samples. After that, RB learns more vital parts as it entirely uses information in the convolutional layer. Finally, HT retains the efficient local features extracted from the layers situated at the different levels of the model. Its performance has been evaluated on the University of Bonn EEG dataset, divided into five distinct classes. The proposed Epileptic-Net achieves the average accuracy of 99.95% in the two-class classification, 99.98% in the three-class classification, 99.96% in the four-class classification, and 99.96% in classifying the complicated five-class problem. Thus the proposed approach shows more competitive results than the existing model to detect epileptic seizures. We also hope that this method can support experts in achieving objective and reliable results, lowering the misdiagnosis rate, and assisting in decision-making.

Melanocortin 1 Receptor Deficiency in Hematopoietic Cells Promotes the Expansion of Inflammatory Leukocytes in Atherosclerotic Mice.

Melanocortin receptor 1 (MC1-R) is expressed in leukocytes, where it mediates anti-inflammatory actions. We have previously observed that global deficiency of MC1-R signaling perturbs cholesterol homeostasis, increases arterial leukocyte accumulation and accelerates atherosclerosis in apolipoprotein E knockout (Apoe-/-) mice. Since various cell types besides leukocytes express MC1-R, we aimed at investigating the specific contribution of leukocyte MC1-R to the development of atherosclerosis. For this purpose, male Apoe-/- mice were irradiated, received bone marrow from either female Apoe-/- mice or MC1-R deficient Apoe-/- mice (Apoe-/- Mc1re/e) and were analyzed for tissue leukocyte profiles and atherosclerotic plaque phenotype. Hematopoietic MC1-R deficiency significantly elevated total leukocyte counts in the blood, bone marrow and spleen, an effect that was amplified by feeding mice a cholesterol-rich diet. The increased leukocyte counts were largely attributable to expanded lymphocyte populations, particularly CD4+ T cells. Furthermore, the number of monocytes was elevated in Apoe-/- Mc1re/e chimeric mice and it paralleled an increase in hematopoietic stem cell count in the bone marrow. Despite robust leukocytosis, atherosclerotic plaque size and composition as well as arterial leukocyte counts were unaffected by MC1-R deficiency. To address this discrepancy, we performed an in vivo homing assay and found that MC1-R deficient CD4+ T cells and monocytes were preferentially entering the spleen rather than homing in peri-aortic lymph nodes. This was mechanistically associated with compromised chemokine receptor 5 (CCR5)-dependent migration of CD4+ T cells and a defect in the recycling capacity of CCR5. Finally, our data demonstrate for the first time that CD4+ T cells also express MC1-R. In conclusion, MC1-R regulates hematopoietic stem cell proliferation and tissue leukocyte counts but its deficiency in leukocytes impairs cell migration via a CCR5-dependent mechanism.

Improving R Peak Detection in ECG Signal Using Dynamic Mode Selected Energy and Adaptive Window Sizing Algorithm with Decision Tree Algorithm.

R peak detection is crucial in electrocardiogram (ECG) signal analysis to detect and diagnose cardiovascular diseases (CVDs). Herein, the dynamic mode selected energy (DMSE) and adaptive window sizing (AWS) algorithm are proposed for detecting R peaks with better efficiency. The DMSE algorithm adaptively separates the QRS components and all non-objective components from the ECG signal. Based on local peaks in QRS components, the AWS algorithm adaptively determines the Region of Interest (ROI). The Feature Extraction process computes the statistical properties of energy, frequency, and noise from each ROI. The Sequential Forward Selection (SFS) procedure is used to find the best subsets of features. Based on these characteristics, an ensemble of decision tree algorithms detects the R peaks. Finally, the R peak position on the initial ECG signal is adjusted using the R location correction (RLC) algorithm. The proposed method has an experimental accuracy of 99.94%, a sensitivity of 99.98%, positive predictability of 99.96%, and a detection error rate of 0.06%. Given the high efficiency in detection and fast processing speed, the proposed approach is ideal for intelligent medical and wearable devices in the diagnosis of CVDs.

A Deep Machine Learning Method for Concurrent and Interleaved Human Activity Recognition.

Human activity recognition has become an important research topic within the field of pervasive computing, ambient assistive living (AAL), robotics, health-care monitoring, and many more. Techniques for recognizing simple and single activities are typical for now, but recognizing complex activities such as concurrent and interleaving activity is still a major challenging issue. In this paper, we propose a two-phase hybrid deep machine learning approach using bi-directional Long-Short Term Memory (BiLSTM) and Skip-Chain Conditional random field (SCCRF) to recognize the complex activity. BiLSTM is a sequential generative deep learning inherited from Recurrent Neural Network (RNN). SCCRFs is a distinctive feature of conditional random field (CRF) that can represent long term dependencies. In the first phase of the proposed approach, we recognized the concurrent activities using the BiLSTM technique, and in the second phase, SCCRF identifies the interleaved activity. Accuracy of the proposed framework against the counterpart state-of-art methods using the publicly available datasets in a smart home environment is analyzed. Our experiment's result surpasses the previously proposed approaches with an average accuracy of more than 93%.

Melanocortin 3 receptor activation with [D-Trp8]-γ-MSH suppresses inflammation in apolipoprotein E deficient mice.

The melanocortin MC1 and MC3 receptors elicit anti-inflammatory actions in leukocytes and activation of these receptors has been shown to alleviate arterial inflammation in experimental atherosclerosis. Thus, we aimed to investigate whether selective targeting of melanocortin MC3 receptor protects against atherosclerosis. Apolipoprotein E deficient (ApoE-/-) mice were fed high-fat diet for 12 weeks and randomly assigned to receive either vehicle (n = 11) or the selective melanocortin MC3 receptor agonist [D-Trp(8)]-gamma-melanocyte-stimulating hormone ([D-Trp8]-γ-MSH; 15 µg/day, n = 10) for the last 4 weeks. Lesion size as well as macrophage and collagen content in the aortic root plaques were determined. Furthermore, leukocyte counts in the blood and aorta and cytokine mRNA expression levels in the spleen, liver and aorta were quantified. No effect was observed in the body weight development or plasma cholesterol level between the two treatment groups. However, [D-Trp8]-γ-MSH treatment significantly reduced plasma levels of chemokine (C-C motif) ligands 2, 4 and 5. Likewise, cytokine and adhesion molecule expression levels were reduced in the spleen and liver of γ-MSH-treated mice, but not substantially in the aorta. In line with these findings, [D-Trp8]-γ-MSH treatment reduced leukocyte counts in the blood and aorta. Despite reduced inflammation, [D-Trp8]-γ-MSH did not change lesion size, macrophage content or collagen deposition of aortic root plaques. In conclusion, the findings indicate that selective activation of melanocortin MC3 receptor by [D-Trp8]-γ-MSH suppresses systemic and local inflammation and thereby also limits leukocyte accumulation in the aorta. However, the treatment was ineffective in reducing atherosclerotic plaque size.

Discovery of the Showdomycin Gene Cluster from Streptomyces showdoensis ATCC 15227 Yields Insight into the Biosynthetic Logic of C-Nucleoside Antibiotics.

Nucleoside antibiotics are a large class of pharmaceutically relevant chemical entities, which exhibit a broad spectrum of biological activities. Most nucleosides belong to the canonical N-nucleoside family, where the heterocyclic unit is connected to the carbohydrate through a carbon-nitrogen bond. However, atypical C-nucleosides were isolated from Streptomyces bacteria over 50 years ago, but the molecular basis for formation of these metabolites has been unknown. Here, we have sequenced the genome of S. showdoensis ATCC 15227 and identified the gene cluster responsible for showdomycin production. Key to the detection was the presence of sdmA, encoding an enzyme of the pseudouridine monophosphate glycosidase family, which could catalyze formation of the C-glycosidic bond. Sequence analysis revealed an unusual combination of biosynthetic genes, while inactivation and subsequent complementation of sdmA confirmed the involvement of the locus in showdomycin formation. The study provides the first steps toward generation of novel C-nucleosides by pathway engineering.

Structure-based engineering of angucyclinone 6-ketoreductases.

Angucyclines are tetracyclic polyketides produced by Streptomyces bacteria that exhibit notable biological activities. The great diversity of angucyclinones is generated in tailoring reactions, which modify the common benz[a]anthraquinone carbon skeleton. In particular, the opposite stereochemistry of landomycins and urdamycins/gaudimycins at C-6 is generated by the short-chain alcohol dehydrogenases/reductases LanV and UrdMred/CabV, respectively. Here we present crystal structures of LanV and UrdMred in complex with NADP(+) and the product analog rabelomycin, which enabled us to identify four regions associated with the functional differentiation. The structural analysis was confirmed in chimeragenesis experiments focusing on these regions adjacent to the active site cavity, which led to reversal of the activities of LanV and CabV. The results surprisingly indicated that the conformation of the substrate and the stereochemical outcome of 6-ketoreduction appear to be intimately linked.

Molecular evolution of the bacterial pseudouridine-5'-phosphate glycosidase protein family.

Pseudouridine is a noncanonical C-nucleoside commonly present in RNA, which is not metabolized in mammals, but can be recycled by the unique enzyme family of bacterial pseudouridine glycosidases such as YeiN from Escherichia coli. Here, we present rigorous bioinformatic and biochemical analyses of the protein family in order to find sequences that might code for nonpseudouridine glycosidase activities. To date, the only other function reported for the enzyme family occurs during the biosynthesis of the antibiotic alnumycin A in Streptomyces species, where AlnA functions as an unusual C-glycosynthase. Bioinformatics analysis of 755 protein sequences identified one group of sequences that were unlikely to harbour pseudouridine glycosidase activities. This observation was confirmed in vitro with one representative protein, IdgA from Streptomyces albus, which was unable to synthesize pseudouridine monophosphate, but was able to attach d-ribose-5-phosphate to juglone. Furthermore, our analyses provide evidence for horizontal gene transfer of pseudouridine glycosidases that may have occurred in Streptomyces and Doria species. Inspection of the genomic loci in the vicinity of pseudouridine glycosidases revealed that in 77% of the strains a kinase gene putatively involved in the phosphorylation of pseudouridine was found nearby, whereas the sequences encoding nonpseudouridine glycosidases coexisted with a phosphatase of the haloacid dehalogenase enzyme family. The investigation suggested that these unknown sequences might be involved in the biosynthesis of soluble blue pigments because of the presence of genes homologous to nonribosomal peptide synthetases.