RÉSUMÉ
We propose a shape prior representation-constrained multi-scale features fusion segmentation network for medical image segmentation, including training and testing stages. The novelty of our training framework lies in two modules comprised of the shape prior constraint and the multi-scale features fusion. The shape prior learning model is embedded into a segmentation neural network to solve the problems of low contrast and neighboring organs with intensities similar to the target organ. The latter can provide both local and global contexts to address the issues of large variations in patient postures as well as organ's shape. In the testing stage, we propose a circular collaboration framework strategy which combines a shape generator auto-encoder network model with a segmentation network model, allowing the two models to collaborate with each other, resulting in a cooperative effect that leads to accurate segmentations. Our proposed method is evaluated and demonstrated on the ACDC MICCAI'17 Challenge Dataset, CT scans datasets, namely, in COVID-19 CT lung, and LiTS2017 liver from three different datasets, and its results are compared with the recent state of the art in these areas. Our method ranked 1st on the ACDC Dataset in terms of Dice score and achieved very competitive performance on COVID-19 CT lung and LiTS2017 liver segmentation.
RÉSUMÉ
BACKGROUND AND OBJECTIVE: Aberrant accumulation of glycosphingolipids (GSLs) in the lysosome leads to GSL storage diseases. Glucosylceramide synthase inhibitors (GCSi) have the potential to treat several GSL storage diseases by reducing the synthesis of the disease-causing GSLs. AL01211 is a potent oral GCSi under investigation for Type 1 Gaucher disease and Fabry disease. Here, we evaluate the pharmacokinetics, pharmacodynamics, safety, and tolerability of AL01211 in healthy Chinese volunteers. METHODS: AL01211 was tested in a Phase 1, single-center, randomized, double-blind, placebo-controlled study with single-dose (15 and 60 mg) and multiple-dose (30 mg) arms. RESULTS: Results of AL01211 demonstrated dose-dependent pharmacokinetics, rapid absorption (median time to maximum plasma concentration [tmax] 2.5-4 hours), relatively slow clearance rate (mean apparent total clearance from plasma [CL/F] 88.3-200 L/h) and the mean terminal half-life above 30 hours. Repeated once-daily oral administration of AL01211 for 14 days had an approximately 2-fold accumulation, reaching steady-state levels between 7 and 10 days, and led to a 73% reduction in plasma glucosylceramide (GL1) on Day 14. AL01211 was safe and well tolerated, with no identified serious adverse events. CONCLUSION: AL01211 showed a favorable pharmacokinetic, pharmacodynamics, safety, and tolerability profile in healthy Chinese volunteers. These data support the further clinical development of AL01211 as a therapy for GSL storage diseases. CLINICAL TRIAL REGISTRY: Clinical Trial Registry no. CTR20221202 ( http://www.chinadrugtrials.org.cn ) registered on 6 June 2022 and ChiCTR2200061431 ( http://www.chictr.org.cn ) registered on 24 June 2022.
Sujet(s)
Asiatiques , Glucosyltransferases , Volontaires sains , Humains , Méthode en double aveugle , Mâle , Adulte , Administration par voie orale , Jeune adulte , Femelle , Glucosyltransferases/antagonistes et inhibiteurs , Relation dose-effet des médicaments , Chine , Adulte d'âge moyen , Antienzymes/pharmacocinétique , Antienzymes/administration et posologie , Antienzymes/effets indésirables , Antienzymes/pharmacologie , Peuples d'Asie de l'EstRÉSUMÉ
Pyroptosis induced by lipopolysaccharide (LPS) has an obvious impact on intestinal inflammation and immune regulation. Enterotoxigenic Escherichia coli (ETEC) K88 has been proved to induce inflammatory responses in several models, but whether E. coli K88 participates in the same process of pyroptotic cell death as LPS remains to be identified. We conducted a pilot experiment to confirm that E. coli K88, instead of Escherichia coli O157 and Salmonella typhimurium, promotes the secretion of interleukin-1 beta (IL-1ß) and interleukin-18 (IL-18) in macrophages. Further experiments were carried out to dissect the molecular mechanism both in vitro and in vivo. The Enzyme-Linked Immunosorbent Assay (ELISA) results suggested that E. coli K88 treatment increased the expression of pro-inflammatory cytokines IL-18 and IL-1ß in both C57BL/6 mice and the supernatant of J774A.1 cells. Intestinal morphology observations revealed that E. coli K88 treatment mainly induced inflammation in the colon. Real-time PCR and Western blot analysis showed that the mRNA and protein expressions of pyroptosis-related factors, such as NLRP3, ASC, and Caspase1, were significantly upregulated by E. coli K88 treatment. The RNA-seq results confirmed that the effect was associated with the activation of NLRP3, ASC, Caspase1, GSDMD, IL-18, and IL-1ß, and might also be related to inflammatory bowel disease and the tumor necrosis factor pathway. The pyroptosis-activated effect of E. coli K88 was significantly blocked by NLRP3 siRNA. Our data suggested that E. coli K88 caused inflammation by triggering pyroptosis, which provides a theoretical basis for the prevention and treatment of ETEC in intestinal infection.
RÉSUMÉ
Prolyl 4-hydroxylase subunit beta (P4HB) can catalyze the formation, breakage and rearrangement of disulfide bonds through two thioredoxin domains, which is important for the maintenance of oxidizing environment in endoplasmic reticulum. Recently, P4HB has been demonstrated its oncogenic role of tumorigenesis and development in cancers. Therefore, we comprehensively deciphered P4HB in human cancer from various aspects, including pan-cancer analysis and narrative summary. We also provided some possible interacted molecules and the top 10 predicted drugs targeting P4HB to contribute to future research. We proposed that P4HB was a potential target and brought new therapeutic opportunities for cancer patients.
RÉSUMÉ
Accurate segmentation of 3D medical images is vital for computer-aided diagnosis. However, the complexity of target morphological variations and a scarcity of labeled data make segmentation more challenging. Furthermore, existing models make it difficult to fully and efficiently integrate global and local information, which hinders structured knowledge acquisition. To overcome these challenges, we introduce the TNT Masking Network (TNT-MNet), a groundbreaking transformer-based 3D model that utilizes a transformer-in-transformer (TNT) encoder. For the first time, we present masked image modeling (MIM) in supervised learning, utilizing target boundary regions as masked prediction targets to enhance structured knowledge acquisition. We execute multiscale random masking on inner and outer tokens in online branch to tackle the challenge of segmenting organs and lesion regions with varying structures at multiple scales and to enhance modeling capabilities. In contrast, the target branch utilizes all tokens to guide the online branch to reconstruct the masked tokens. Our experiments suggest that TNT-MNet's performance is comparable, or even better, than state-of-the-art models in three medical image datasets (BTCV, LiTS2017, and BraTS2020) and effectively reduces the dependence on labeled data. The code and models are publicly available at https://github.com/changliu-work/TNT_MNet.
RÉSUMÉ
The biosynthetic process of selenium nanoparticles (SeNPs) by specific bacterial strain, whose growth directly affects the synthesis efficiency, has attracted great attentions. We previously reported that Bacillus paralicheniformis SR14, a SeNPs-producing bacteria, could improve intestinal antioxidative function in vitro. To further analyze the biological characteristics of SR14, whole genome sequencing was used to reveal the genetic characteristics in selenite reduction and sugar utilization. The results reviewed that the genome size of SR14 was 4,448,062 bp, with a GC content of 45.95%. A total of 4300 genes into 49 biological pathways was annotated to the KEGG database. EC: 1.1.1.49 (glucose-6-phosphate 1-dehydrogenase) and EC: 5.3.1.9 (glucose-6-phosphate isomerase), were found to play a potential role in glucose degradation and EC:2.7.1.4 (fructokinase) might be involved in the fructose metabolism. Growth profile and selenite-reducing ability of SR14 under different sugar supplements were determined and the results reviewed that glucose had a better promoting effect on the reduction of selenite and growth of bacteria than fructose, sucrose, and maltose. Moreover, RT-qPCR experiment proved that glucose supplement remarkably promoted the expressions of thioredoxin, fumarate reductase, and the glutathione peroxidase in SR14. Analysis of mRNA expression showed levels of glucose-6-phosphate dehydrogenase and fructokinase significantly upregulated under the supplement of glucose. Overall, our data demonstrated the genomic characteristics of SR14 and preliminarily determined that glucose supplement was most beneficial for strain growth and SeNPs synthesis.
RÉSUMÉ
PURPOSE: Identification of a role for, and the mechanism of action of, the acid-sensing ion channel 1a (ASIC1a) in M1 macrophage polarization, which results in osteoarthritis (OA)-associated chondrocyte senescence. METHOD: ASIC1a expression in synovial M1 macrophages of OA patients was assessed by immunofluorescence. A role for ASIC1a in M1 macrophage and chondrocyte senescence was assessed in a mouse OA model. RESULTS: ASIC1a expression was found to be upregulated in synovial M1 macrophages of OA patients. Extracellular acidification (pH 6.0) promoted M1 polarization of bone marrow derived macrophages (BMDMs), which was reversed by PcTx-1 or ASIC1a-siRNA. RNA-seq transcriptome results demonstrated a downregulation of M1 macrophage-associated genes in BMDMs after PcTx-1 treatment. Mechanistically, a role for the ASIC1a-cytidine/uridine monophosphate kinase 2 (CMPK2) axis in M1 macrophage polarization was demonstrated. The concentration of IL-18 was elevated in synovial fluid and supernatants of acid-activated BMDMs. In vitro, IL-18 stimulation or co-culture with acid-activated macrophages promoted chondrocyte senescence. In vivo, intra-articular administration of PcTx-1 reduced articular cartilage destruction and chondrocytes senescence in OA mice, which related to reduced numbers of M1 macrophages and IL-18 in affected joints. CONCLUSION: These results demonstrate a novel pathogenic process that results in OA cartilage damage, in which M1 macrophage derived IL-18 induces articular chondrocytes senescence. Further, the ASIC1a-CMPK2 axis was shown to positively regulate M1 macrophage polarization. Hence, ASIC1a is a promising treatment target for M1 macrophage-mediated diseases, such as OA.
RÉSUMÉ
In vivo articular cartilage degeneration is an essential hallmark of osteoarthritis (OA), involving chondrocyte senescence, extracellular matrix degradation, chondrocyte death, cartilage loss, and bone erosion. Among them, chondrocyte death is one of the major factors leading to cartilage degeneration. Many studies have reported that various cell death modes, including apoptosis, ferroptosis, and autophagy, play a key role in OA chondrocyte death. Currently, there is insufficient understanding of OA pathogenesis, and there remains a lack of treatment methods to prevent OA and inhibit its progression. Studies suggest that OA prevention and treatment are mainly directed to arrest premature or excessive chondrocyte death. In this review, we a) discuss the forms of death of chondrocytes and the associations between them, b) summarize the critical factors in chondrocyte death, c) discuss the vital role of chondrocyte death in OA, d) and, explore new approaches for targeting the regulation of chondrocyte death in OA treatment.
Sujet(s)
Cartilage articulaire , Arthrose , Humains , Chondrocytes/métabolisme , Arthrose/métabolisme , Cartilage articulaire/métabolisme , ApoptoseRÉSUMÉ
Radiomics refers to the high-throughput extraction of quantitative features from medical images, and is widely used to construct machine learning models for the prediction of clinical outcomes, while feature engineering is the most important work in radiomics. However, current feature engineering methods fail to fully and effectively utilize the heterogeneity of features when dealing with different kinds of radiomics features. In this work, latent representation learning is first presented as a novel feature engineering approach to reconstruct a set of latent space features from original shape, intensity and texture features. This proposed method projects features into a subspace called latent space, in which the latent space features are obtained by minimizing a unique hybrid loss function including a clustering-like loss and a reconstruction loss. The former one ensures the separability among each class while the latter one narrows the gap between the original features and latent space features. Experiments were performed on a multi-center non-small cell lung cancer (NSCLC) subtype classification dataset from 8 international open databases. Results showed that compared with four traditional feature engineering methods (baseline, PCA, Lasso and L2,1-norm minimization), latent representation learning could significantly improve the classification performance of various machine learning classifiers on the independent test set (all p<0.001). Further on two additional test sets, latent representation learning also showed a significant improvement in generalization performance. Our research shows that latent representation learning is a more effective feature engineering method, which has the potential to be used as a general technology in a wide range of radiomics researches.
RÉSUMÉ
Deep neural network (DNN) approaches have shown remarkable progress in automatic Chest X-rays classification. However, existing methods use a training scheme that simultaneously trains all abnormalities without considering their learning priority. Inspired by the clinical practice of radiologists progressively recognizing more abnormalities and the observation that existing curriculum learning (CL) methods based on image difficulty may not be suitable for disease diagnosis, we propose a novel CL paradigm, named multi-label local to global (ML-LGL). This approach iteratively trains DNN models on gradually increasing abnormalities within the dataset, i,e, from fewer abnormalities (local) to more ones (global). At each iteration, we first build the local category by adding high-priority abnormalities for training, and the abnormality's priority is determined by our three proposed clinical knowledge-leveraged selection functions. Then, images containing abnormalities in the local category are gathered to form a new training set. The model is lastly trained on this set using a dynamic loss. Additionally, we demonstrate the superiority of ML-LGL from the perspective of the model's initial stability during training. Experimental results on three open-source datasets, PLCO, ChestX-ray14 and CheXpert show that our proposed learning paradigm outperforms baselines and achieves comparable results to state-of-the-art methods. The improved performance promises potential applications in multi-label Chest X-ray classification.
Sujet(s)
Apprentissage machine , , Radiographie thoracique , Humains , Rayons X , Modèles biologiques , Jeux de données comme sujet , Radiographie thoracique/méthodesRÉSUMÉ
In deep learning-based image registration, the deformable region with complex anatomical structures is an important factor affecting the accuracy of network registration. However, it is difficult for existing methods to pay attention to complex anatomical regions of images. At the same time, the receptive field of the convolutional neural network is limited by the size of its convolution kernel, and it is difficult to learn the relationship between the voxels with far spatial location, making it difficult to deal with the large region deformation problem. Aiming at the above two problems, this paper proposes a cascaded multi-level registration network model based on transformer, and equipped it with a difficult deformable region perceptron based on mean square error. The difficult deformation perceptron uses sliding window and floating window techniques to retrieve the registered images, obtain the difficult deformation coefficient of each voxel, and identify the regions with the worst registration effect. In this study, the cascaded multi-level registration network model adopts the difficult deformation perceptron for hierarchical connection, and the self-attention mechanism is used to extract global features in the basic registration network to optimize the registration results of different scales. The experimental results show that the method proposed in this paper can perform progressive registration of complex deformation regions, thereby optimizing the registration results of brain medical images, which has a good auxiliary effect on the clinical diagnosis of doctors.
Sujet(s)
Algorithmes , , Traitement d'image par ordinateur/méthodesRÉSUMÉ
To provide a safe and effective supplement of the essential trace element selenium, we focused on the biosynthesis of nanoselenium (SeNPs) via probiotics. A novel kind of exopolymer-functionalized nanoselenium (SeEPS), whose average size was 67.0 ± 0.6 nm, was produced by Bacillus subtilis SR41, whereas the control consisted of exopolymers without selenium (EPS). Chemical composition analysis, Fourier transform infrared (FTIR) spectroscopy and high-performance liquid chromatography (HPLC) confirmed that SeEPS and EPS shared similar polysaccharide characteristic groups, such as COO- and C=O, and contained not only 45.2-45.4% of sugars but also 23.5-24.7% of proteins and some lipids. Both SeEPS and EPS were primarily composed of mannose, amino glucose, ribose, glucose and galactose. Furthermore, to identify the biologically active component of SeEPS, three kinds of selenium particles with different stabilizers [Se(0), bovine serum albumin-Se and EPS-Se] were synthesized chemically, and their ability to scavenge free radicals in vitro was compared with that of SeEPS and EPS. The results revealed that EPS itself exhibited weak superoxide and hydroxyl radical scavenging abilities. Nevertheless, SeEPS had superior antioxidant properties compared to all other products, possibly due to the specific structure of SeNPs and exopolymers. Our results suggested that exopolymer-functionalized SeNPs with specific monosaccharide composition and structure could eventually find a potential application as an antioxidant.
RÉSUMÉ
This paper proposes a new network framework, which leverages EfficientNetB4, attention gate, and residual learning techniques to achieve automatic and accurate liver segmentation. First, we use EfficientNetB4 as the encoder to extract more feature information during the encoding stage. Then, an attention gate is introduced in the skip connection to eliminate irrelevant regions and highlight features of a specific segmentation task. Finally, to alleviate the problem of gradient vanishment, we replace the traditional convolution of the decoder with a residual block to improve the segmentation accuracy. We verified the proposed method on the LiTS17 and SLiver07 datasets and compared it with classical networks such as FCN, U-Net, attention U-Net, and attention Res-U-Net. In the Sliver07 evaluation, the proposed method achieved the best segmentation performance on all five standard metrics. Meanwhile, in the LiTS17 assessment, the best performance is obtained except for a slight inferior on RVD. The proposed method's qualitative and quantitative results demonstrated its applicability in liver segmentation and proved its good prospect in computer-assisted liver segmentation.
Sujet(s)
Apprentissage , Foie , Humains , Foie/imagerie diagnostique , Évolution de la maladie , Tomodensitométrie , Traitement d'image par ordinateurRÉSUMÉ
Selenium-enriched polysaccharide (SeEPS) was prepared by reducing Se(IV) to elemental selenium and organic selenium in polysaccharide medium by the obtained Enterobacter cloacae strain Z0206 under aerobic conditions. In the present study, we focused on investigating the role of short-term supplementation of SeEPS at supernutritional doses in the regulation of growth performance, liver damage, antioxidant capacity, and selenium (Se) accumulation in C57 mice. Thirty-two C57 mice were randomly divided into four groups: the control group was gavaged with equal volume of phosphate-buffered saline, while the sodium selenite (Na2SeO3), selenomethionine (SeMet), and SeEPS groups were gavaged with 0.5 mg Se/kg BW of Na2SeO3, SeMet, and selenium-enriched polysaccharide (n = 8), respectively. We examined liver injury indicators, antioxidant capacity in the serum and liver, selenium deposition at different sites, selenoprotein levels, and selenocysteine-synthesizing and degradation-associated gene expression in mouse livers. SeEPS supplementation dramatically increased average daily weight gain but reduced the feed-to-gain ratio (F/G) of mice (P < 0.05). Compared to Na2SeO3 and SeMet supplementation, SeEPS supplementation at supernutritional doses did not cause the liver damage. SeEPS supplementation also markedly enhanced total antioxidant capacity (T-AOC), catalase (CAT), glutathione peroxidase (GSH-PX), and total superoxide dismutase (T-SOD) activities but reduced malondialdehyde (MDA) levels in the liver and serum (P < 0.05), while significantly increasing selenocysteine-synthesizing and degradation-related gene (SEPHS2, SEPSECS, Secisbp, Scly) expression at the mRNA level (P < 0.05), thus upregulating the mRNA levels of selenoproteins (SELENOP, SELENOK) (P < 0.05). We suggest that SeEPS could be a potential replacement for inorganic selenium to improve animals' growth performance, promote antioxidant capacity, and regulate selenium deposition.
Sujet(s)
Sélénium , Animaux , Antioxydants/métabolisme , Antioxydants/pharmacologie , Compléments alimentaires , Souris , Polyosides/pharmacologie , ARN messager/génétique , Sélénium/pharmacologie , Sélénocystéine , Sélénométhionine , Sélénoprotéines/génétiqueRÉSUMÉ
We previously obtained and characterized a novel sulfated derivative of the exopolysaccharides from Enterobacter cloacae Z0206 (SEPS). This study aimed at investigating the effects and mechanism of SEPS against dextran sulfate sodium (DSS) induced intestinal injury. The results showed that SEPS increased the proliferation and survival of intestinal epithelial cells during DSS stimulation. Furthermore, SEPS maintained the barrier function and inflammatory response via JAK2 and MAPK signaling to protect against DSS-induced intestinal injury. Mechanistically, SEPS elevated the DNA methylation in the promoter region to negatively regulate the JAK2 and MAPKs expression. Thus, the current study shows the potential effects and mechanism of SEPS on DSS-induced intestinal epithelial cell injury.
Sujet(s)
Colite/traitement médicamenteux , Méthylation de l'ADN , Enterobacter cloacae/composition chimique , Kinase Janus-2/génétique , Polyosides bactériens/administration et posologie , Animaux , Prolifération cellulaire/effets des médicaments et des substances chimiques , Survie cellulaire/effets des médicaments et des substances chimiques , Colite/induit chimiquement , Colite/génétique , Sulfate dextran , Modèles animaux de maladie humaine , Épigenèse génétique/effets des médicaments et des substances chimiques , Cellules épithéliales/composition chimique , Cellules épithéliales/cytologie , Cellules épithéliales/effets des médicaments et des substances chimiques , Intestins/composition chimique , Intestins/cytologie , Intestins/effets des médicaments et des substances chimiques , Système de signalisation des MAP kinases/effets des médicaments et des substances chimiques , Souris , Souris de lignée C57BL , Polyosides bactériens/composition chimique , Polyosides bactériens/pharmacologie , Agents protecteursRÉSUMÉ
Bacterial infections are among the major factors that cause stress and intestinal diseases in piglets. Lipopolysaccharide (LPS), a major component of the Gram-negative bacteria outer membrane, is commonly employed for inducing an immune response in normal organisms for convenience. The association between LPS stimulation and gut immunity has been reported. However, the effects of gut immunity on microbial homeostasis and metabolism of host, especially bile acid and lipid metabolism in piglets, remain unclear. Hence, in the current study, we elucidated the effect of gut immunity on microbial balance and host metabolism. Twenty-one-day-old healthy piglets (male) were randomly assigned into the CON and LPS groups. After 4 hours of treatment, related tissues and cecal contents were obtained for further analysis. The obtained results showed that stimulated LPS considerably damaged the morphology of intestinal villi and enhanced the relative expression of proinflammatory cytokines. Besides, LPS partially changed the microbial structure as indicated by ß-diversity and increased operational taxonomic units (OTUs) related to Oxalobacter and Ileibacterium. Furthermore, bile acid, a large class of gut microbiota metabolites, was also assessed by many proteins related to the enterohepatic circulation of bile acids. It was also revealed that LPS markedly inhibited the mRNA and protein expression of TGR5 and FXR (bile acid receptors) in the ileum, which expressed negative feedback on bile acid de novo synthesis. Additionally, results indicated upregulated mRNA of genes associated with the production of bile acid in the liver tissues. Moreover, LPS reduced the expression of bile acid transporters in the ileum and liver tissues and further disturbed the normal enterohepatic circulation. Taken together, gut immunity and microbial dysbiosis are associated with altered bile acid metabolism in LPS-challenged piglets, which provided theoretical basis for revealing the potential mechanism of intestinal inflammation in swine and seeking nutrients to resist intestinal damage.
Sujet(s)
Acides et sels biliaires/métabolisme , Dysbiose/immunologie , Microbiome gastro-intestinal , Immunité innée , Maladies des porcs/immunologie , Animaux , Cytokines/génétique , Cytokines/métabolisme , Dysbiose/microbiologie , Dysbiose/médecine vétérinaire , Muqueuse intestinale/immunologie , Muqueuse intestinale/métabolisme , Lipopolysaccharides/toxicité , Foie/métabolisme , Mâle , Récepteurs couplés aux protéines G/génétique , Récepteurs couplés aux protéines G/métabolisme , Suidae , Maladies des porcs/microbiologieRÉSUMÉ
This study investigated the effects of dietary C. butyricum ZJU-F1 on the apparent digestibility of nutrients, intestinal barrier function, immune response, and microflora of weaned piglets, with the aim of providing a theoretical basis for the application of Clostridium butyricum as an alternative to antibiotics in weaned piglets. A total of 120 weanling piglets were randomly divided into four treatment groups, in which piglets were fed a basal diet supplemented with antibiotics (CON), Bacillus licheniformis (BL), Clostridium butyricum ZJU-F1 (CB), or Clostridium butyricum and Bacillus licheniformis (CB-BL), respectively. The results showed that CB and CB-BL treatment increased the intestinal digestibility of nutrients, decreased intestinal permeability, and increased intestinal tight junction protein and mucin expression, thus maintaining the integrity of the intestinal epithelial barrier. CB and CB-BL, as exogenous probiotics, were also found to stimulate the immune response of weaned piglets and improve the expression of antimicrobial peptides in the ileum. In addition, dietary CB and CB-BL increased the proportion of Lactobacillus. The levels of butyric acid, propionic acid, acetic acid, and total acid were significantly increased in the ceca of piglets fed CB and CB-BL. Furthermore, we validated the effects of C. butyricum ZJU-F1 on the intestinal barrier function and immune response in vitro and found C. butyricum ZJU-F1 improved intestinal function and enhanced the TLR-2-MyD88-NF-κB signaling.
Sujet(s)
Clostridium butyricum/composition chimique , Compléments alimentaires/analyse , Microbiome gastro-intestinal/immunologie , Immunité/immunologie , Intestins/physiopathologie , Animaux , SuidaeRÉSUMÉ
The mechanism of message passing in graph neural networks (GNNs) is still mysterious. Apart from convolutional neural networks, no theoretical origin for GNNs has been proposed. To our surprise, message passing can be best understood in terms of power iteration. By fully or partly removing activation functions and layer weights of GNNs, we propose subspace power iteration clustering (SPIC) models that iteratively learn with only one aggregator. Experiments show that our models extend GNNs and enhance their capability to process random featured networks. Moreover, we demonstrate the redundancy of some state-of-the-art GNNs in design and define a lower limit for model evaluation by a random aggregator of message passing. Our findings push the boundaries of the theoretical understanding of neural networks.
Sujet(s)
, Analyse de regroupements , InfographieRÉSUMÉ
Lactic acid bacteria (LAB) play a key role in promoting health and preventing diseases because of their beneficial effects, such as antimicrobial activities, modulating immune responses, maintaining the gut epithelial barrier and antioxidant capacity. However, the mechanisms with which LAB relieve oxidative stress and intestinal injury induced by diquat in vivo are poorly understood. In the present study, Pediococcus pentosaceus ZJUAF-4 (LAB, ZJUAF-4), a selected probiotics strain with strong antioxidant capacities, was appointed to evaluate the efficiency against oxidative stress in diquat-induced intestinal injury of mice. Alanine transaminase (ALT) and aspartate aminotransferase (AST) were analyzed to estimate the liver injury. The intestinal permeability was evaluated by 4 kDa fluorescein isothiocyanate (FITC)-dextran (FD4), D-lactate (DLA), and diamine oxidase (DAO) levels. Jejunum reactive oxygen species (ROS) production was examined by dihydroethidium (DHE) staining. Western blotting was used to detect the expression of nuclear factor (erythroid-derived-2)-like 2 (Nrf2) and its downstream genes in jejunum. The gut microbiota was analyzed by high-throughput sequencing method based on the 16S rRNA genes. The results showed that ZJUAF-4 pretreatment was found to protect the intestinal barrier function and maintain intestinal redox homeostasis under diquat stimulation. Moreover, oral administration of ZJUAF-4 increased the expression of Nrf2 and its downstream genes. High-throughput sequencing analysis indicated that ZJUAF-4 contributed to restoring the gut microbiota influenced by diquat. Our results suggested that ZJUAF-4 protected the intestinal barrier from oxidative stress-induced damage by modulating the Nrf2 pathway and gut microbiota, indicating that ZJUAF-4 may have potential applications in preventing and treating oxidative stress-related intestinal diseases. KEY POINTS: ⢠ZJUAF-4 exerted protective effects against diquat-induced intestinal injury. ⢠Activation of Nrf2 and its downstream targets towards oxidative stress. ⢠ZJUAF-4 administration restoring gut microbiota.