Saravanan, C. | bioRxiv Science

bioRxiv2021

Loss of hepatic Lgr4 and Lgr5 promotes nonalcoholic fatty liver disease

Background & AimsThe Rspo-Lgr4/5-Znrf3/Rnf43 module is a master regulator of hepatic Wnt/{beta}-catenin signaling and metabolic zonation, but its impact on nonalcoholic fatty liver disease (NAFLD) remains unclear. We studied whether liver-specific loss of the Wnt/{beta}-catenin modulators Leucine-Rich Repeat-Containing G Protein-Coupled Receptor 4/5 (Lgr4/5) promotes nonalcoholic fatty liver disease (NAFLD). MethodsMice with liver-specific deletion of both receptors Lgr4/5 (Lgr4/5dLKO) were fed with normal diet (ND) or high fat diet (HFD). Livers of these mice were analyzed for lipid and fibrotic content by tissue staining and immunohistochemistry (IHC), and lipoproteins, inflammation and liver enzyme markers were measured in blood. Mechanistic insights into hepatic lipid accumulation were obtained by using ex vivo primary hepatocyte cultures derived from the Lgr4/5dLKO mice. Lipid analysis of mouse livers was performed by mass spectrometry (MS)-based untargeted lipidomic analysis. ResultsWe demonstrated that liver-specific ablation of Lgr4/5-mediated Wnt signaling resulted in hepatic steatosis, impaired bile acid (BA) secretion and predisposition to liver fibrosis. Under HFD conditions, we observed progressive intrahepatic fat accumulation, developing into macro-vesicular steatosis. Serum lipoprotein levels in HFD-fed Lgr4/5dLKO mice were decreased, rather than increased, suggesting that accumulation of fat in the liver was due to impaired lipid secretion by hepatocytes. Our lipidome analysis revealed a severe alteration of several lipid species in livers of Lgr4/5dLKO mice, including triacylglycerol estolides (TG-EST), a storage form of bioactive free fatty acid (FA) esters of hydroxy FAs (FAHFAs). ConclusionsLoss of hepatic Wnt/{beta}-catenin activity by Lgr4/5 deletion led to deregulation of lipoprotein pathways, loss of BA secretion, intrinsic alterations of lipid homeostasis and the onset of NAFLD. Lay summaryThe Wnt/{beta}-catenin pathway plays an important role during development and tissue homeostasis. Loss of Wnt/{beta}-catenin activity in mouse liver leads to loss of liver zonation, but the impact on nonalcoholic fatty liver disease (NAFLD) remains unclear. We show that livers of mice developed steatosis upon deletion of the positive pathway regulators Lgr4/5. Livers of knock-out (KO) mice exhibited altered lipid composition due to impaired lipid secretion. Furthermore, livers of these mice developed a nonalcoholic steatohepatitis (NASH)-like phenotype and fibrotic features derived from activated hepatic stellate cells. Our data demonstrate a protective role of Wnt/{beta}-catenin pathway activity towards the development of NAFLD. HighlightsO_LIAbrogation of hepatic Wnt/{beta}-catenin activity and liver zonation upon Lgr4/5 deletion in mice led to hepatic steatosis. C_LIO_LILiver fat accumulation was caused by impaired lipid secretion from hepatocytes. C_LIO_LISteatotic livers contained increased levels of diverse lipid species, including polyunsaturated fatty acids and triglycerol-estolides. C_LIO_LIThese data confirmed that a decrease in Wnt/{beta}-catenin signaling led to the development of nonalcoholic fatty liver disease (NAFLD) in mice. C_LI

physiology↗

bioRxiv2019

A deep learning-based model of normal histology

Deep learning models have been applied on various tissues in order to recognize malignancies. However, these models focus on relatively narrow tissue context or well-defined pathologies. Here, instead of focusing on pathologies, we introduce models characterizing the diversity of normal tissues. We obtained 1,690 slides with rat tissue samples from the control groups of six preclinical toxicology studies, on which tissue regions were outlined and annotated by pathologists into 46 different tissue classes. From these annotated regions, we sampled small patches of 224 x 224 pixels at six different levels of magnification. Using four studies as training set and two studies as test set, we trained VGG-16, ResNet-50, and Inception-v3 networks separately at each of these magnification levels. Among these models, Inception-v3 consistently outperformed the other networks and attained accuracies up to 83.4% (top-3 accuracy: 96.3%). Further analysis showed that most tissue confusions occurred within clusters of histologically similar tissues. Investigation of the embedding layer using the UMAP method revealed not only pronounced clusters corresponding to the individual tissues, but also subclusters corresponding to histologically meaningful structures that had neither been annotated nor trained for. This suggests that the histological representation learned by the normal histology network could also be used to flag abnormal tissue as outliers in the embedding space without a need to explicitly train for specific types of abnormalities. Finally, we found that models trained on rat tissues can be used on non-human primate and minipig tissues with minimal retraining. Author contributionT.S. and H.H. contributed equally to this work. Significance statementLike many other scientific disciplines, histopathology has been profoundly impacted by recent advances in machine learning with deep neural networks. In this field, most deep learning models reported in the literature are trained on pathologies in specific tissues/contexts. Here, we aim to establish a model of normal tissues as a foundation for future models of histopathology. We build models that are specific to histopathology images and we show that their embeddings are better feature vectors for describing the underlying images than those of off-the shelf CNN models. Therefore, our models could be useful for transfer learning to improve the accuracy of other histopathology models.

pathology↗