Spatial molecular and cellular determinants of STAT3 activation in liver fibrosis progression in non-alcoholic fatty liver disease

Abstract

Background & aims: The prevalence of non-alcoholic fatty liver disease (NAFLD) and its severe form, non-alcoholic steatohepatitis (NASH), is increasing. Individuals with NASH often develop liver fibrosis and advanced liver fibrosis is the main determinant of mortality in individuals with NASH. We and others have reported that STAT3 contributes to liver fibrosis and hepatocellular carcinoma in mice.

Methods: Here, we explored whether STAT3 activation in hepatocyte and non-hepatocyte areas, measured by phospho-STAT3 (pSTAT3), is associated with liver fibrosis progression in 133 patients with NAFLD. We further characterized the molecular and cellular determinants of STAT3 activation by integrating spatial distribution and transcriptomic changes in fibrotic NAFLD livers.Results: pSTAT3 scores in non-hepatocyte areas progressively increased with fibrosis severity (r = 0.53, p <0.001). Correlation analyses between pSTAT3 scores and expression of 1,540 immune- and cancer-associated genes revealed a large effect of STAT3 activation on gene expression changes in non-hepatocyte areas and confirmed a major role for STAT3 activation in fibrogenesis. Digital spatial transcriptomic profiling was also performed on 13 regions selected in hepatocyte and non-hepatocyte areas from four NAFLD liver biopsies with advanced fibrosis, using a customized panel of markers including pSTAT3, PanCK+CK8/18, and CD45. The regions were further segmented based on positive or negative pSTAT3 staining. Cell deconvolution analysis revealed that activated STAT3 was enriched in hepatic progenitor cells (HPCs) and sinusoidal endothelial cells. Regression of liver fibrosis upon STAT3 inhibition in mice with NASH resulted in a reduction of HPCs, demonstrating a direct role for STAT3 in HPC expansion.

Conclusion: Increased understanding of the spatial dependence of STAT3 signaling in NASH and liver fibrosis progression could lead to novel targeted treatment approaches.

Impact and implications: Advanced liver fibrosis is the main determinant of mortality in patients with NASH. This study showed using liver biopsies from 133 patients with NAFLD, that STAT3 activation in non-hepatocyte areas is strongly associated with fibrosis severity, inflammation, and progression to NASH. STAT3 activation was enriched in hepatic progenitor cells (HPCs) and sinusoidal endothelial cells (SECs), as determined by innovative technologies interrogating the spatial distribution of pSTAT3. Finally, STAT3 inhibition in mice resulted in reduced liver fibrosis and depletion of HPCs, suggesting that STAT3 activation in HPCs contributes to their expansion and fibrogenesis in NAFLD.

Keywords: DSP, digital spatial profiler; FC, fold change; HCC, hepatocellular carcinoma; HFD, high-fat diet; HPCs, hepatic progenitor cells; HSCs, hepatic stellate cells; IPA, Ingenuity® Pathway Analysis; LSECs, liver sinusoidal endothelial cells; NAFLD; NAFLD, non-alcoholic fatty liver disease; NAS, NAFLD activity score; NASH; NASH, non-alcoholic steatohepatitis; SECs, sinusoidal endothelial cells; STAT, signal transducer and activator of transcription; STAT3; cirrhosis; fibrosis; liver cancer; pSTAT3, phospho-STAT3.

Graphical abstract

Fig. 1

Nuclear staining of pSTAT3. (A) In hepatocytes, and (B) in non-hepatocyte areas, of representative liver biopsies from patients with NAFLD. Scale bar: 200 Μm. NAFLD, non-alcoholic fatty liver disease; pSTAT3, phospho-STAT3.

Fig. 2

Quantification of pSTAT3 nuclear staining in hepatocytes (top) and non-hepatocyte areas (bottom) in liver biopsies. (A) With F1 to F4 fibrosis stages. (B) With or without NASH. (C) With different degrees of steatosis. (D) With different degrees of inflammation. (E) With different degrees of ballooning. Log2-transformed pSTAT3 scores were used. Error bar: interquartile range. Statistical significance was calculated using Kruskal–Wallis test (more than two groups) or Mann-Whitney test (two groups). NASH, non-alcoholic steatohepatitis; pSTAT3, phospho-STAT3.

Fig. 3

Correlation between expression levels of selected genes and pSTAT3 scores. (A) With hepatocytes pSTAT3 scores. (B) With non-hepatocytes pSTAT3 scores. r, Spearman’s correlation coefficient. Log10-transformed pSTAT3 scores and log2-transformed gene expression data were used. pSTAT3, phospho-STAT3.

Fig. 4

Digital spatial transcriptome profiling of ROIs from NASH liver biopsies with advanced fibrosis. (A) Representative ROIs using the following color fluorescence: panCK+CK8/18 (cyan), CD45 (yellow), pSTAT3 (red), and DNA (blue) in hepatocytes (top) and non-hepatocyte/hepatocyte areas (bottom). Scale bar, 100 μm. (B) PCoA of gene expression data from whole transcriptome atlas based on hepatocyte and non-hepatocyte areas and pSTAT3 staining. Ellipses were drawn using the standard deviation of point scores. Log₁₀-transformed gene expression data were used. NASH, non-alcoholic steatohepatitis; PC, principle component; PCoA, principle component analysis; pSTAT3, phospho-STAT3; ROI, region of interest.

Fig. 5

Cell deconvolution plots using adult liver matrix and ROI transcriptome profiles show changes of cell proportions. In (A) hepatocyte and (B) non-hepatocyte areas. The mean of each cell proportion was calculated for hepatocyte and non-hepatocyte areas. Statistical significance was calculated using an unpaired t test. ROI, region of interest.

Fig. 6

Cell deconvolution analysis using mouse liver matrix and transcriptomic profiles show changes of cell proportions in NASH livers between HepPten^- mice treated with C188-9 or placebo. The mean of each cell proportion was calculated for treated mice and placebo. Statistical significance was calculated using an unpaired t test. NASH, non-alcoholic steatohepatitis.