A growing number of long noncoding RNAs (lncRNAs) have emerged as vital metabolic regulators. However, most human lncRNAs are nonconserved and highly tissue specific, vastly limiting our ability to identify human lncRNA metabolic regulators (hLMRs). In this study, we established a pipeline to identify putative hLMRs that are metabolically sensitive, disease relevant, and population applicable. We first progressively processed multilevel human transcriptome data to select liver lncRNAs that exhibit highly dynamic expression in the general population, show differential expression in a nonalcoholic fatty liver disease (NAFLD) population, and respond to dietary intervention in a small NAFLD cohort. We then experimentally demonstrated the responsiveness of selected hepatic lncRNAs to defined metabolic milieus in a liver-specific humanized mouse model. Furthermore, by extracting a concise list of protein-coding genes that are persistently correlated with lncRNAs in general and NAFLD populations, we predicted the specific function for each hLMR. Using gain- and loss-of-function approaches in humanized mice as well as ectopic expression in conventional mice, we validated the regulatory role of one nonconserved hLMR in cholesterol metabolism by coordinating with an RNA-binding protein, PTBP1, to modulate the transcription of cholesterol synthesis genes. Our work overcame the heterogeneity intrinsic to human data to enable the efficient identification and functional definition of disease-relevant human lncRNAs in metabolic homeostasis.

Mouse is the most widely used animal model in biomedical research, but it remains unknown what causes the large number of differentially regulated genes between human and mouse livers identified in recent years. In this report, we aim to determine whether these divergent gene regulations are primarily caused by environmental factors or some of them are the result of cell-autonomous differences in gene regulation in human and mouse liver cells. The latter scenario would suggest that many human genes are subject to human-specific regulation and can only be adequately studied in a human or humanized system. To understand the similarity and divergence of gene regulation between human and mouse livers, we performed stepwise comparative analyses in human, mouse, and humanized livers with increased stringency to gradually remove the impact of factors external to liver cells, and used bioinformatics approaches to retrieve gene networks to ascertain the regulated biological processes. We first compared liver gene regulation by fatty liver disease in human and mouse under the condition where the impact of genetic and gender biases was minimized, and identified over 50% of all commonly regulated genes, that exhibit opposite regulation by fatty liver disease in human and mouse. We subsequently performed more stringent comparisons when a single specific transcriptional or post-transcriptional event was modulated in vitro or vivo or in liver-specific humanized mice in which human and mouse hepatocytes colocalize and share a common circulation. Intriguingly and strikingly, the pattern of a high percentage of oppositely regulated genes persists under well-matched conditions, even in the liver of the humanized mouse model, which represents the most closely matched in vivo condition for human and mouse liver cells that is experimentally achievable. Gene network analyses further corroborated the results of oppositely regulated genes and revealed substantial differences in regulated biological processes in human and mouse cells. We also identified a list of regulated lncRNAs that exhibit very limited conservation and could contribute to these differential gene regulations. Our data support that cell-autonomous differences in gene regulation might contribute substantially to the divergent gene regulation between human and mouse livers and there are a significant number of biological processes that are subject to human-specific regulation and need to be carefully considered in the process of mouse to human translation.

Dietary supplementation is a widely adapted strategy to maintain nutritional balance for improving health and preventing chronic diseases. Conflicting results in studies of similar design, however, suggest that there is substantial heterogenicity in individuals’ responses to nutrients, and personalized nutrition is required to achieve the maximum benefit of dietary supplementation. In recent years, nutrigenomics studies have been increasingly utilized to characterize the detailed genomic response to a specific nutrient, but it remains a daunting task to define the signatures responsible for interindividual variations to dietary supplements for tissues with limited accessibility. In this work, we used the hepatic response to omega-3 fatty acids as an example to probe such signatures. Through comprehensive analysis of nutrigenomic response to eicosapentaneoid acid (EPA) and/or docosahexaenoic acid (DHA) including both protein coding and long noncoding RNA (lncRNA) genes in human hepatocytes, we defined the EPA- and/or DHA-specific signature genes in hepatocytes. By analyzing gene expression variations in livers of healthy and relevant disease populations, we identified a set of protein coding and lncRNA signature genes whose responses to omega-3 fatty acid exhibit very high interindividual variabilities. The large variabilities of individual responses to omega-3 fatty acids were further validated in human hepatocytes from ten different donors. Finally, we profiled RNAs in exosomes isolated from the circulation of a liver-specific humanized mouse model, in which the humanized liver is the sole source of human RNAs, and confirmed the in vivo detectability of some signature genes, supporting their potential as biomarkers for nutrient response. Taken together, we have developed an efficient and practical procedure to identify nutrient-responsive gene signatures as well as accessible biomarkers for interindividual variations.

Fibrotic liver features excessive deposition of extracellular matrix (ECM), primarily produced from “activated” hepatic stellate cells (HSCs). While targeting human HSCs (hHSCs) in fibrosis therapeutics shows promise, the overall understanding of hHSC activation remains limited, in part because it is very challenging to define the role of human long non-coding RNAs (lncRNAs) in hHSC activation. To address this challenge, we identified another cell type that acts via a diverse gene network to promote fibrogenesis. Then, we identified the lncRNAs that were differentially regulated in activated hHSCs and the other profibrotic cell. Next, we conducted concurrent analysis to identify those lncRNAs that were specifically involved in fibrogenesis. We tested and confirmed that transdifferentiation of vascular smooth muscle cells (VSMCs) represents such a process. By overlapping TGFβ-regulated lncRNAs in multiple sets of hHSCs and VSMCs, we identified a highly selected list of lncRNA candidates that could specifically play a role in hHSC activation. We experimentally characterized one human lncRNA, named CARMN, which was significantly regulated by TGFβ in all conditions above. CARMN knockdown significantly reduced the expression levels of a panel of marker genes for hHSC activation, as well as the levels of ECM deposition and hHSC migration. Conversely, gain of function of CARMN using CRISPR activation (CRISPR-a) yielded the completely opposite effects. Taken together, our work addresses a bottleneck in identifying human lncRNAs that specifically play a role in hHSC activation and provides a framework to effectively select human lncRNAs with significant pathophysiological role.

Bile acids, regarded as the body's detergent for digesting lipids, also function as critical signaling molecules that regulate cholesterol and triglyceride levels in the body. Bile acids are the natural ligands of the nuclear receptor, FXR, which controls an intricate network of cellular pathways to maintain metabolic homeostasis. In recent years, growing evidence supports that many cellular actions of the bile acid/FXR pathway are mediated by long non-coding RNAs (lncRNAs), and lncRNAs are in turn powerful regulators of bile acid levels and FXR activities. In this review, we highlight the substantial progress made in the understanding of the functional and mechanistic role of lncRNAs in bile acid metabolism and how lncRNAs connect bile acid activity to additional metabolic processes. We also discuss the potential of lncRNA studies in elucidating novel molecular mechanisms of the bile acid/FXR pathway and the promise of lncRNAs as potential diagnostic markers and therapeutic targets for diseases associated with altered bile acid metabolism.

Growing evidence links human long noncoding RNAs (lncRNAs) to metabolic disease pathogenesis, yet no FDA-approved drugs target human lncRNAs. Most human lncRNAs lack conservation in other mammals, complicating efforts to define their roles and identify therapeutic targets. Here, we leveraged the concept of functionally conserved lncRNAs (FCLs) — lncRNAs that share function despite no sequence similarity — to develop a framework for identifying human lncRNAs as therapeutic targets for metabolic disorders. We used expression quantitative trait loci mapping and functional conservation analyses to pinpoint human lncRNAs influenced by disease-associated SNPs and with potential functionally conserved mouse equivalents. We identified human and mouse GULLs (glucose and lipid lowering), which regulate glucose and lipid metabolism by binding CRTC2, thereby modulating gluconeogenic genes via CREB and lipogenic genes via SREBP1. Despite their lack of sequence similarity, both lncRNAs demonstrated similar metabolic effects in obese mice, with more pronounced benefits from long-term activation. To identify druggable sites, we mapped GULLs’ binding motifs to CRTC2 (termed GULFs). Standalone human GULF, an RNA oligomer resembling FDA-approved siRNAs, significantly improved glucose and lipid levels in obese mice. This framework highlights functionally conserved human lncRNAs as promising therapeutic targets, exemplified by GULLs’ potential as a glucose- and lipid-lowering therapeutic.