Molecular patterns of the NO-sGC-cGMP pathway in progressive and regressive liver fibrosis models

Abstract

The nitric oxide (NO)-soluble guanylate cyclase (sGC)-cyclic guanosine-monophosphate (cGMP) pathway is impaired in liver fibrosis. We investigated expression patterns of NO-sGC-cGMP components via RT-qPCR in various rat models of liver fibrosis and murine models of liver fibrosis regression. Hepatic cGMP-levels were measured chromatographically. All models demonstrated portal-hypertension and liver fibrosis, which significantly regressed in murine models. The rat models showed etiology-specific differences in NO-sGC-cGMP pathway regulation. We observed strong upregulation of sGCa1 and sGCb1 subunits in a rat choline-deficient high-fat diet model (1.75-fold, p = 0.004 and 2.04-fold, p = 0.004, respectively). The sGCa2 subunit was markedly downregulated in a rat thioacetamide model (0.66-fold, p = 0.026). The rat bile-duct-ligation model was characterized by strong upregulation of inducible nitric oxide synthetase (28.10-fold, p = 0.029). The rat thioacetamide and bile-duct-ligation models displayed downregulation of sGCb2 (0.15-fold, p = 0.002, and 0.19-fold, p = 0.029, respectively). Regardless, hepatic cGMP-levels in rat models remained unchanged. Both mouse models demonstrated upregulation of NO-sGC-cGMP pathway nodes during regression, further accompanied by increased hepatic cGMP-levels in murine carbon tetrachloride (peak-fibrosis: 3.86 nM vs. 1-week regression: 6.28 nM, p = 0.006; vs. 2-week regression: 5.49 nM, p = 0.091) and thioacetamide (peak-fibrosis: 2.87 nM vs. 1-week regression: 5.22 nM, p = 0.007; vs. 2-week regression: 6.68 nM, p < 0.001) models. The NO-sGC-cGMP pathway exhibits etiology-specific and temporal regulation patterns during liver fibrogenesis and fibrosis regression. We further highlight the functional contribution of the pathway via increases in hepatic cGMP during fibrosis regression.

Keywords: Liver cirrhosis; Liver fibrosis; Liver fibrosis regression; Portal hypertension; Soluble guanylate cyclase.

Fig. 1

Model characterization. A Fibrosis was induced over 12 weeks (W) in all models except for rat bile-duct ligation, which was assessed 4 W post-surgery. Mouse models included groups with a 1 W (R1) or 2 W (R2) fibrosis regression period (no toxic agent administration), respectively. B (top) Representative images from rat models highlight unique fibrosis patterns associated with each model. B (bottom) Representative images from the mouse models show gradual thinning and resolution of fibrosis. C–F All animal models exhibited pronounced fibrosis (measured by collagen-proportionate area) and elevated portal-pressure. In mice, both fibrosis and portal-pressure decreased during regression, reaching statistical significance at the R2 timepoint. p values were calculated against the diseased (positive) controls using Welch’s t test (rat models) or Dunn’s test (mouse models). BDL bile-duct-ligation, CCl4 carbon tetrachloride-induced, CDHFD choline-deficient high-fat diet-fed, CHOW Standard balanced chow-fed, m mouse, NaCl saline-induced, OO olive oil-induced, r rat, R1 1-week regression, R2 2-week regression, SHAM sham-operated, TAA thioacetamide-induced.

Fig. 2

Model-specific regulation of NO-sGC-cGMP pathway nodes in rats. A Schematic representation of the NO-sGC-cGMP pathway. B iNOS (Nos2) showed the strongest expression in the rBDL model, whereas eNOS (Nos3) was consistently upregulated in all rat models. C The sGCa1 subunit (Gucy1a3) showed robust upregulation in the rCDHFD model and an upwards trend in the rCCl4 and rTAA models. The sGCb1 subunit (Gucy1b3) showed significant upregulation in the rCDHFD, rCCl4 and rTAA models. D Significant upregulation of PKG1 (Prkg1) was observed in the rCDHFD and rTAA models. PDE5 (Pde5a) showed consistent upregulation in all rat models. E Heatmap depicting the observed changes across all genes and rat models. GAPDH-normalized fold-change values were MinMax-scaled gene-wise. The control (CTRL) column represents averaged values of all control groups as a reference point. B–E Changes are expressed as fold-change compared to healthy controls. Statistical comparisons between groups were performed using Mann–Whitney U test. BDL bile-duct-ligation, CCl4 carbon tetrachloride-induced, CDHFD choline-deficient high-fat diet-fed, CHOW standard balanced chow-fed, eNOS endothelial nitric oxide synthetase, GAPDH glyceraldehyde-3-phosphate dehydrogenase, iNOS inducible nitric oxide synthetase, NaCl saline-induced, OO olive oil-induced, PDE phosphodiesterase, PKG proteinkinase G, r rat, sGC soluble guanylate cyclase, SHAM sham-operated, TAA thioacetamide-induced.

Fig. 3

Protein-level changes in the NO-sGC-cGMP pathway nodes in rats. A sGCa1 and B sGCb1 were significantly increased in the rCDHFD model, whereas a trend was observed in the rTAA and rBDL models. C sGCa2 was strongly upregulated in the rCDHFD model. D PKG1 showed a trend for upregulation in all models, except for rBDL. E Strong upregulation of PDE5 was found in rCDHFD, rCCl4 and rTAA models. F Representative western blots (cropped to protein of interest) from the rCDHFD model. Full western blots are provided within Supplementary Fig. 3. Statistical comparisons between groups were performed using Mann–Whitney U test. BDL bile-duct-ligation, CCl4 carbon tetrachloride-induced, CDHFD choline-deficient high-fat diet-fed, CHOW standard balanced chow-fed, NaCl saline-induced, OO olive oil-induced, PDE phosphodiesterase, PKG proteinkinase G, r rat, sGC soluble guanylate cyclase, SHAM sham-operated, TAA thioacetamide-induced.

Fig. 4

Temporal regulation of the NO-sGC-cGMP pathway during fibrosis regression. A iNOS (Nos2) was significantly upregulated at R2 of the mCCl4 model and remained high during mTAA-regression. B All sGC subunits (Gucy1a3, Gucy1a2, Gucy1b3) were markedly increased during regression in both models, albeit at different timepoints. At peak fibrogenesis, sGCa2 (Gucy1a2) was downregulated in both models, but expression-levels recovered during R2 and R1 of mCCl4 and mTAA, respectively. C PKG1 (Prkg1) was upregulated at R2 of the mCCl4 model and at R1 of the mTAA model. PKG2 (Prkg2) showed strong downregulation in fibrosis and was slowly reconstituted during regression. D PDE5 (Pde5a) was strongly upregulated at R2 of the mCCl4 model. In the mTAA model, PDE5 was upregulated in peak fibrosis and during the R1 and R2 regression period. PDE9 (Pde9a) was strongly downregulated during fibrosis of both mouse models, but expression was subsequently upregulated during regression. E The heatmap highlights changes at R2 of the mCCl4 model, while changes in the mTAA model mostly occur already at R1. ActinB-normalized fold-change values were MinMax-scaled gene-wise. Changes are expressed as fold-change compared to the healthy control group. Statistical comparisons between groups were performed using Dunn’s test. CCl4 carbon tetrachloride-induced, iNOS inducible nitric oxide synthetase, m mouse, NaCl saline-induced, OO olive oil-induced, PDE phosphodiesterase, PKG proteinkinase G, R1 1-week regression, R2 2-week regression, sGC soluble guanylate cyclase, TAA thioacetamide-induced.

Fig. 5

Hepatic cGMP content during peak fibrogenesis and fibrosis regression. A Despite significant differences in NO-sGC-cGMP pathway component expression, cGMP-levels in the rat models remained similar when comparing healthy controls and diseased animals. B In the murine models, the hepatic cGMP content decreased at peak fibrosis in the mTAA model. Importantly, during liver fibrosis regression, there were significant increases in hepatic cGMP-levels in both models, with peak-levels achieved at R1 and R2 of mCCl4 and mTAA, respectively. The p values were calculated using Welch’s t test (A) or Dunn’s test (B). BDL bile-duct-ligation, CCl4 carbon tetrachloride-induced, CDHFD choline-deficient high-fat diet-fed, cGMP cyclic guanosine monophosphate, CHOW standard balanced chow-fed, m mouse, NaCl saline-induced, OO olive oil-induced, r rat, R1 1-week regression, R2 2-week regression, SHAM sham-operated, TAA thioacetamide-induced.