Glycolysis in hepatic stellate cells coordinates fibrogenic extracellular vesicle release spatially to amplify liver fibrosis

Abstract

Liver fibrosis is characterized by the activation of perivascular hepatic stellate cells (HSCs), the release of fibrogenic nanosized extracellular vesicles (EVs), and increased HSC glycolysis. Nevertheless, how glycolysis in HSCs coordinates fibrosis amplification through tissue zone-specific pathways remains elusive. Here, we demonstrate that HSC-specific genetic inhibition of glycolysis reduced liver fibrosis. Moreover, spatial transcriptomics revealed a fibrosis-mediated up-regulation of EV-related pathways in the liver pericentral zone, which was abrogated by glycolysis genetic inhibition. Mechanistically, glycolysis in HSCs up-regulated the expression of EV-related genes such as Ras-related protein Rab-31 (RAB31) by enhancing histone 3 lysine 9 acetylation on the promoter region, which increased EV release. Functionally, these glycolysis-dependent EVs increased fibrotic gene expression in recipient HSC. Furthermore, EVs derived from glycolysis-deficient mice abrogated liver fibrosis amplification in contrast to glycolysis-competent mouse EVs. In summary, glycolysis in HSCs amplifies liver fibrosis by promoting fibrogenic EV release in the hepatic pericentral zone, which represents a potential therapeutic target.

Fig. 1.. Glycolytic Hk2 deletion selectively in HSCs attenuates liver fibrosis.

(A) Breeding schema for obtaining the HK2^ΔHSC mice and injury strategy for inducing liver fibrosis in these mice. Prepared with BioRender.com. (B) Primary mouse HSCs were isolated from HK2^fl/fl or HK2^ΔHSC mice and Hk2 mRNA levels were examined (n = 4, unpaired t test). (C to F) HK2^fl/fl or HK2^ΔHSC male and female mice were treated with either olive oil or CCl₄ for 6 weeks. Livers were analyzed by quantitative polymerase chain reaction (qPCR) (C), Western blot (D), Sirius red (E), and Col1 immunofluorescence (F) (n = 3 to 7 animals per group, one-way ANOVA with Bonferroni multiple comparisons). CCl₄, carbon tetrachloride; Col1, collagen 1; HK2, hexokinase 2. Graph bars represent SEM. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Fig. 2.. Spatial transcriptomics defines two main clusters in the liver.

Livers from olive oil–treated HK2^fl/fl, CCl₄-treated HK2^fl/fl, and CCl₄-treated HK2^ΔHSC mice were studied using 10x Genomics Spatial Transcriptomics technology. R package was used for statistics. (A) Hematoxylin and eosin of the tissues and spatial plots for Col1a1 and Col3a1. (B) The clustering of the spots was performed based on a pathologist annotation of central and portal veins followed by a machine learning–based prediction of the clustering for each spot. (C) Uniform Manifold Approximation and Projection (UMAP) plot of the spots after pathologist-based clustering and violin plots of the expression of the periportal Cyp2f2 and pericentral Cyp2e1 markers. (D) Heatmap of the conserved genes in the pericentral and periportal clusters showing a distinct gene signature for each of these clusters. (E) Violin plots for some of the most up-regulated genes in each cluster, including Cyp2f2, Sds, and Hsd17b6 in the periportal cluster and Cyp2e1, Serpina7, and Lect2 in the pericentral cluster. (F) Violin plots of HSC markers, including Dcn, Reln, Col1a1, and Col3a1, from all conditions combined. Col1a1, collagen 1 alpha 1; Col3a1, collagen 3 alpha 1; Cyp2e1, cytochrome P450 family 2 subfamily E member 1; Cyp2f2, cytochrome P450 family 2 subfamily F member 2; Dcn, decorin; Hsd17b6, hydroxysteroid 17-beta dehydrogenase 6; Lect2, leukocyte cell–derived chemotaxin 2; Reln, reelin; Sds, serine dehydratase; Serpina7, serpin family A member 7.

Fig. 3.. Glycolysis inhibition selectively in HSCs down-regulates vesicle trafficking pathways in the fibrogenic pericentral cluster.

Livers from olive oil–treated HK2^fl/fl, CCl₄-treated HK2^fl/fl, and CCl₄-treated HK2^ΔHSC mice were studied using 10x Genomics Spatial Transcriptomics technology. R package was used for statistics. (A) Identification of the genes in the fibrogenic pericentral cluster that were up-regulated with fibrosis (HK2^fl/fl CCl₄ versus HK2^fl/fl olive oil) and down-regulated when HK2 was deleted selectively in HSCs (HK2^ΔHSC CCl₄ versus HK2^fl/fl CCl₄). (B) Gene Ontology pathway analysis of the molecular functions (Panther 2023) of the genes up-regulated with fibrosis and down-regulated with HSC-specific Hk2 deletion. (C) Gene Ontology pathway analysis of the cellular components (Panther 2023) of the genes up-regulated with fibrosis and down-regulated with HSC-specific Hk2 deletion. (D) Heatmap of the vesicle trafficking-related genes up-regulated with fibrosis and down-regulated with HSC-specific Hk2 deletion. (E) The top genes from the vesicle trafficking and EV-related pathways that were up-regulated with fibrosis and down-regulated with HSC-specific Hk2 deletion. (F) Violin plots of the top three genes from the EV-related pathways that were up-regulated with fibrosis and down-regulated with HSC-specific Hk2 deletion.

Fig. 4.. Glycolysis in HSCs is stimulated by PDGF in vitro.

(A) Seahorse glycolysis stress test measuring ECAR and OCR. Primary human HSCs, seeded at 20,000 cells per well, received direct injections of PDGF (20 ng/ml; 1 hour), 10 mM glucose (30 min), 1 μM oligomycin (30 min), and 50 mM 2-deoxyglucose (2-DG; 30 min) (n = 3). (B) Primary human HSCs were treated with vehicle or PDGF (20 ng/ml) for 4 hours and analyzed by qPCR (n = 4). (C) Primary human HSCs were treated with vehicle or PDGF (20 ng/ml) for 12 hours and analyzed by Western blot (n = 4). (D) Primary human HSCs were treated with vehicle or PDGF (20 ng/ml) for 24 hours and analyzed by nuclear magnetic resonance (n = 4). (E) Primary human HSCs were treated with vehicle or PDGF (20 ng/ml) for 8, 24, 48, or 72 hours and glucose consumption was measured (n = 3). Paired t test was used to measure the P value. *P < 0.05 and **P < 0.01.

Fig. 5.. PDGF-mediated glycolysis promotes the vesicle trafficking pathway through promoter region H3K9ac.

(A and B) Primary human HSCs were treated in the absence (No Glc) or in the presence of glucose (Glc; 1 g/liter) for 8 hours and analyzed by bulk RNA-seq. The heatmap of the differentially expressed genes is shown and the top differentially regulated RAB genes are presented in bar graphs (A). Gene counts for the top up-regulated RAB gene, RAB31, are shown (B) (n = 3). (C) Primary human HSCs were treated in the absence (No Glc) or in the presence of Glc (1 g/liter) for 8 hours and RAB31 mRNA expression was measured by qPCR (n = 3). (D) Primary human HSCs were treated with vehicle or PDGF (20 ng/ml) for 8 hours or 12 hours to measure RAB31 mRNA (n = 4) and protein levels (n = 3), respectively. (E) Primary human HSCs were transfected with control (shCtrl) or HK2 shRNA (shHK2), treated with either vehicle or PDGF (20 ng/ml) for 12 hours, and analyzed by Western blot (n = 3). Total H3 and H3K9ac were blotted simultaneously from the same samples, at the same concentration but in different lanes of the electrophoresis gel. (F) Primary human HSCs were treated with either vehicle or PDGF (20 ng/ml) for 12 hours. Chromatin was immunoprecipitated using an H3K9ac antibody and the RAB31 promoter region was measured by qPCR (n = 4). Paired t test was used to measure the P value when comparing two groups. ANOVA with Bonferroni comparison was used to measure the P value when comparing more than two groups. *P < 0.05 and **P < 0.01.

Fig. 6.. PDGF-mediated glycolysis promotes EV release through RAB31.

(A) Primary human HSCs were treated in the presence of Glc (1 g/liter) with vehicle or 2-DG for 12 hours and analyzed by transmission electron microscopy (n = 3) or immunofluorescence for CD63 (n = 5). (B) Primary human HSCs were treated in no glucose, Glc (1 g/liter) or Glc (1 g/liter) + 2-DG (1 g/liter) for 12 hours. EVs were purified using differential ultracentrifugation and analyzed by Western blot (n = 3). (C) Primary human HSCs were transfected with control (shCtrl) or HK2 shRNA (shHK2) and analyzed by qPCR (n = 4). (D) Primary human HSCs were transfected with control (shCtrl) or HK2 shRNA (shHK2) and analyzed by Western blot (n = 3). (E) Primary human HSCs were transfected with control (shCtrl) or HK2 shRNA (shHK2) and treated with either vehicle or PDGF (20 ng/ml) for 12 hours. EVs were purified using differential ultracentrifugation and analyzed by Western blot (n = 4) and nanoparticle tracking analysis (NTA; n = 4). (F) Primary human HSCs were transfected with control (siControl) or RAB31 siRNA (siRAB31) and analyzed by qPCR (n = 3). (G and H) Primary human HSCs were transfected with control (siCtrl) or RAB31 siRNA (siRAB31) and treated with either vehicle or PDGF (20 ng/ml) for 12 hours. EVs were purified using differential ultracentrifugation and analyzed by NTA, nanoflow cytometry (G), and Western blot (H). (n = 3) Paired t test was used to measure the P value when comparing two groups. ANOVA with Bonferroni comparison was used to measure the P value when comparing more than two groups. *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 7.. Glycolysis-dependent EVs are fibrogenic in vitro and amplify liver fibrosis in vivo.

(A and B) Primary human HSCs were treated in no glucose, Glc (1 g/liter), or Glc (1 g/liter) + 2-DG (1 g/liter) for 12 hours. EVs were purified using differential ultracentrifugation and analyzed by TMT-based quantitative proteomics (n = 1). Differentially regulated proteins among the three conditions (A) were subjected to Gene Ontology analysis using The PANTHER Classification System (www.pantherdb.org/) (B). (C) Donor primary human HSCs were treated in no glucose, Glc (1 g/liter), or Glc (1 g/liter) + 2-DG (1 g/liter) for 12 hours, EVs were purified from donor HSCs using differential ultracentrifugation and used to treat recipient primary human HSCs for 24 hours. Recipient HSCs were analyzed by qPCR (n = 4). (D and E) C57Bl/6 male and female mice were treated for 4 weeks with phosphate-buffered saline or an equal number of EVs derived from donor mice as indicated, in conjunction with CCl₄ for 4 weeks. Livers were analyzed by Sirius red (D) and Western blot (E) (n = 9 to 10 animals per group) ANOVA with Bonferroni comparison was used to measure the P value when comparing more than two groups. *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 8.. Overall schema of the study.

Image prepared with BioRender.com.