Loss of Cdkn1a protects against MASLD alone or with alcohol intake by preserving lipid homeostasis

Abstract

Background & aims: Expression of P21, encoded by the CDKN1A gene, has been associated with fibrosis progression in steatotic liver disease (SLD); however, the underlying mechanisms remain unknown. In the present study, we investigated the function of CDKN1A in SLD.

Methods: CDKN1A expression levels were evaluated in different patient cohorts with SLD, fibrosis, and advanced chronic liver disease (ACLD). Cdkn1a ^-/- and Cdkn1a ^+/+ mice were fed with either a Western diet (WD), a Lieber-DeCarli (LdC) diet plus multiple EtOH (ethanol) binges, or a DuAL diet (metabolic dysfunction-associated fatty liver disease and alcohol-related liver). Primary hepatocytes were isolated and functional assays performed.

Results: A significant increase in CDKN1A expression was observed in patients with steatohepatitis and fibrosis (with a positive correlation with both NAFLD Activity Score and fibrosis staging scores), cirrhosis and ACLD. Cdkn1a ^+/+ mice, fed a DuAL diet exhibited liver injury and cell death increased reactive oxygen species (ROS), and markers of senescence (γH2AX, β-GAL, Cdkn1a/p53) contributing to steatosis and inflammation. In contrast, Cdkn1a ^-/- mutant mice showed a significant decrease in senescence-associated markers as well as in markers of liver injury, hepatic steatosis and an increase in fatty acid oxidation and reduction in free fatty acid uptake as well as de novo lipogenesis. Mechanistically, activation of the AMPK-SIRT3 was observed in Cdkn1a-deleted animals.

Conclusions: Cdkn1a deletion protected against preclinical SLD by promoting fatty acid oxidation and preventing free fatty acid uptake and de novo lipogenesis via the AMPK-SIRT3 axis. CDKN1A expression was found to be directly correlated with increased severity of NAFLD Activity Score and fibrosis in patients with SLD. CDKN1A could be a potential theragnostic target for the treatment of metabolic dysregulation in patients with SLD, with and without alcohol consumption.

Impact and implications: Expression of p21, encoded by the CDKN1A gene, has been associated with fibrosis progression in steatotic liver disease (SLD), but the molecular mechanisms remain elusive. Interestingly, in this study we found that Cdkn1a deletion protected against preclinical SLD by promoting fatty acid oxidation and preventing free fatty acid uptake and de novo lipogenesis, via the AMPK-SIRT3 axis. Translationally, Cdkn1a expression was found to be directly correlated with increased severity of NAFLD Activity Score (NAS) and fibrosis in SLD patients, and therefore, CDKN1A might be used potential theragnostic target for the treatment of metabolically induced SLD, with and without alcohol consumption.

Keywords: CDKN1A; Hepatocyte; Metabolic dysregulation; Palbociclib; Senescence; Steatotic liver disease (SLD).

Graphical abstract

Fig. 1

CDKN1A expression is characteristic of patients with SLD and fibrosis. (A) Cohort #1. CDKN1A mRNA relative expression to GAPDH in livers of patients with MASLD and MASH, identified with a NAS score ≥5 (n = 187). (B) Cohort #1. CDKN1A mRNA relative expression to GAPDH in livers of patients with MASLD with a fibrosis score between F0–F2 and F3–F4 (n = 225). (C) Cohort #1. AUROC when assessing CDKN1A for fibrosis prediction. AUROC of 0.81 (95% CI 0.72–0.89) for fibrosis prediction. (D) Cohort #1. Correlation between CDKN1A mRNA relative expression to GAPDH in livers of patients with MASLD with a NAS score >4 (n = 187). (E) Cohort #1. Correlation between CDKN1A mRNA relative expression to GAPDH in livers of patients with MASLD with a fibrosis score F0–F2 and F3–F4 (n = 225). (F) Cohort #3. CDKN1A mRNA relative expression to GAPDH in livers of non-affected patients and patients with a fibrosis score between F2 and F4 and compensated cirrhosis (CC) and decompensated cirrhosis (DC) (n = 22). (G) Cohort #4. CDKN1A mRNA relative expression to GAPDH in livers of patients with ACLD and non-affected patients (n = 13). (H) Cohort #1. Correlation between IL-8 mRNA relative expression to GAPDH in livers of patients with MASLD. (I) Immunostaining for CDKN1A was tested in paraffin sections of patients with a diagnosis of MASH with the presence of steatosis and inflammation. Microphotographs were taken at 300 μm. (J) Quantification of positive cells was performed and graphed, 200 hepatocytes per field were counted. Student t test was performed (∗p <0.05; ∗∗∗∗p <0.0001). ACLD, advanced chronic liver disease; AUROC, area under the receiver operating characteristic curve; CDKN1A, cyclin-dependent kinase inhibitor 1A; GADPH, glyceraldehyde-3-phosphate dehydrogenase; MASLD, metabolic dysfunction-associated steatotic liver disease; MASH, metabolic-associated steatohepatitis; NAS, NAFLD Activity Score; SLD, steatotic liver disease.

Fig. 2

Liver injury is significantly reduced after a DuAL diet in Cdkn1a^-/- mice. Schematic description of a DuAL model performed in Cdkn1a^+/+ and Cdkn1a^-/- mice. (A) Liver weight to body weight ratio (%). (B) Serum ALT levels. (C) H&E staining in Cdkn1a^+/+ and Cdkn1a^-/- mice after DuAL diet. Scale bar = 100 μm. (D) α-Sma mRNA relative liver expression to Gapdh. (E) Sirius red staining quantification in Cdkn1a^+/+ and Cdkn1a^-/- mice after DuAL diet and (F) representative paraffin liver sections stained. Scale bar = 100 μm. (G) Total liver protein (two individuals’ samples from each group are shown as representative of the group) were isolated from DuAL diet-fed Cdkn1a^+/+ and Cdkn1a^-/- mice and analyzed for CC-3, CC-8, pRIPK1, pRIPK3, and pMLKL. Protein expression levels were normalized to the levels of total GAPDH and the ratio was calculated. (H) Quantification of TUNEL-positive cells (%) was done and graphed. (I) CcnA2, (J) CcnD1, and (K) CcnE1 measured in mice of each group relative to Gapdh. Scale bar = 100 μm. n = 6–10; one-way ANOVA with post hoc Tukey test was used for the calculations (intragroup: ∗p <0.05, ∗∗∗∗p <0.0001; intergroup: ^#p <0.05, ^####p <0.0001). ALT, alanine aminotransferase; BW, body weight; CC3, cleaved-caspase-3; CC8, cleaved-caspase-8; CcnA2, ciclyn A2; CcnD1, cyclin D1; CcnE1, ciclin E1; Cdkn1a, cyclin-dependent kinase inhibitor 1A; DuAL, metabolic dysfunction-associated fatty liver disease and alcohol-related liver; GADPH, glyceraldehyde-3-phosphate dehydrogenase; LW, liver weight; RIPK, receptor interacting protein kinase; TUNEL, terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling.

Fig. 3

Deletion of Cdkn1a in a preclinical model of SLD and in murine ALD. (Left panel) Schematic description of a WD model performed in of Cdkn1a^+/+ and Cdkn1a^-/- mice. (A) LW/BW ratio (%). (B) ALT measurement in serum (n = 7–11). (C) H&E representative images in liver of Cdkn1a^+/+ and Cdkn1a^-/- mice after WD. Scale bar = 100 μm. (D) Representative liver images stained with Sirius red (E) TUNEL staining was performed in liver cryosections of Cdkn1a^+/+ and Cdkn1a^-/- mice after WD and (F,G) quantification of positive Sirius Red-stained area and TUNEL-positive cells. Scale bar = 100 μm. (n = 8–12); one-way ANOVA with post hoc Tukey test was used for the calculations (intragroup: ∗∗∗p <0.001, ∗∗∗∗p <0.0001; intergroup ^###p <0.001, ^####p <0.0001). (Right panel) Schematic description of an LdC model plus multiple binges performed in Cdkn1a^+/+ and Cdkn1a^-/- mice. (H) LW/BW ratio (%). (I) ALT measurement in serum. (J) H&E staining in Cdkn1a^+/+ and Cdkn1a^-/- mice after LdC EtOH diet. Scale bar = 100 μm. (I) Sirius Red staining performed in liver of Cdkn1a^+/+ and Cdkn1a^-/- mice after LdC EtOH diet. Scale bar = 100 μm. (J) TUNEL staining was performed in liver cryosections of liver of Cdkn1a^+/+ and Cdkn1a^-/- mice after LdC EtOH diet. Scale bar = 100 μm. (M,N) Quantification of Sirius Red-positive area and TUNEL-positive cells was done and graphed, respectively. n = 7–8; one-way ANOVA with post hoc Tukey test was used for the calculations (intragroup: ∗p <0.05, ∗∗∗∗p <0.0001). ALD, alcohol-related liver disease; ALT, alanine aminotransferase; BW, body weight; Cdkn1a, cyclin-dependent kinase inhibitor 1A; EtOH, ethanol; LdC, Lieber-DeCarli; LW, liver weight; SLD, steatotic liver disease; WD, Western diet.

Fig. 4

Amelioration of liver senescence is characteristic of Cdkn1a^-/- mice fed a DuAL diet. Cdkn1a/p21 (A) and P53 (B) mRNA relative liver expression to Gapdh was analyzed in Cdkn1a^+/+ and Cdkn1a^-/- mice after DuAL diet. (C) Relative telomere length was calculated using PCR. (D) Total liver protein (two individuals’ samples from each group are shown as representative of the group) were isolated from DuAL diet-fed Cdkn1a^+/+ and Cdkn1a^-/- mice and analyzed for γH2AX, pRb, p16, and PCNA. Protein expression levels were normalized to the levels of total GAPDH and the ratio was calculated. (E–G) Representative liver sections of Cdkn1a^+/+ and Cdkn1a^-/- mice stained with 4-HNE (E), γH2AX (F) and β-GAL (G). Scale bar = 100 μm. (H–J) Quantification of 4-HNE, γH2AX and β-GAL-positive cells per field. One-way ANOVA with post hoc Tukey test was used for the calculations. n = 6–10 (intragroup: ∗∗∗p <0.001, ∗∗∗p <0.001; intergroup: ^#p <0.01, ^###p <0.001). β-Gal, beta galactosidase; γH2AX, gamma H2A histone family member X; 4-HNE, 4-hydroxynonenal; Cdkn1a, cyclin-dependent kinase inhibitor 1A; DuAL, metabolic dysfunction-associated fatty liver disease and alcohol-related liver; Gadph, glyceraldehyde-3-phosphate dehydrogenase; pRb, phospho retinoblastoma tumor suppressor; PCNA, proliferating cell nuclear antigen.

Fig. 5

Inflammation is significantly reduced in Cdkn1a^-/- mice fed a DuAL diet. (A,B) Representative immunofluorescence staining for CD45 (A) and CD11b (B) in DuAL diet-fed Cdkn1a^+/+ and Cdkn1a^-/- mice. Quantification of CD45- (C) and CD11b- (D) positive cells in the same samples. Arrows (→) indicate positive cells. (E) Circulating F4/80^hiLy6C^hi cells in DuAL diet-fed Cdkn1a^+/+ and Cdkn1a^-/- mice. Tnfα (F), Il6 (G), Il1α (H), Il8 (I), Igfbp4 (J), and Pge2 (K) mRNA expression determined by qPCR and normalized to the amount of Gapdh in the liver of DuAL-fed mice and controls. n = 4–10; one-way ANOVA with post hoc Tukey test was used for the calculations; intragroup: ∗∗p <0.01, ∗∗∗∗p <0.0001; intergroup: ^#p <0.05, ^###p <0.001. Cdkn1a, cyclin-dependent kinase inhibitor 1A; DuAL, metabolic dysfunction-associated fatty liver disease and alcohol-related liver; Igfbp4, Insulin like growth factor binding protein 4; Gadph, glyceraldehyde-3-phosphate dehydrogenase; Pge2, prostaglandin E2; qPCR, quantitative PCR; Tnfα, tumor necrosis factor alpha.

Fig. 6

Lipid metabolism in experimental models of SLD and MetALD. (A) ORO staining performed in liver cryosection of DuAL-fed mice and (B) each quantification. Scale bar = 100 μm. (C) Quantification of hepatic triglycerides after DuAL diet (mg/mg liver). (D) Illustrative ORO-stained liver sections from each group of mice after WD. Scale bar = 100 μm. (E) Quantification of ORO-stained area (n = 5–10) (F) Quantification of hepatic triglycerides after a WD in Cdkn1a^+/+ and Cdkn1a^-/- after WD and controls was done and graphed. (G) Serum triglycerides levels after a DuAL diet and controls (mg/dl). (H) Pparα, (I) Fxr1, (J) Pparγ, (K) Cd36, and (L) Fas mRNA expression determined by qPCR and normalized to the amount of Gapdh in liver of DuAL-fed mice and controls. (N) Total liver protein (two individuals’ samples from each group are shown as a representative of the group) were isolated from DuAL diet-fed Cdkn1a^+/+ and Cdkn1a^-/- mice and analyzed for pAMPK, pAKT, and SIRT3. GAPDH was used as a loading control. (O) FAO was measured in liver tissue of DuAL diet-fed Cdkn1a^+/+ and Cdkn1a^-/- mice and the concentration of ASM + CO₂ (mmol/g) was quantified and graphed. n = 6–10, one-way ANOVA with post hoc Tukey test was used for the calculations; intragroup: ∗p <0.05, ∗∗∗∗p <0.0001; intergroup: ^#p <0.05, ^####p <0.0001. ASM, acid-soluble metabolites; CO₂, carbon dioxide; Cdkn1a, cyclin-dependent kinase inhibitor 1A; DuAL, metabolic dysfunction-associated fatty liver disease and alcohol-related liver; FAO, fatty acid oxidation; Fxfr1, fragile X-related protein 1; Gadph, glyceraldehyde-3-phosphate dehydrogenase; MetALD, metabolic-associated alcoholic liver disease; ORO, Oil Red O; qPCR, quantitative PCR; Pprα, peroxisome proliferation-activated receptor alpha; Pparγ, peroxisome proliferation-activated receptor gamma; SLD, steatotic liver disease; WD, Western diet.

Fig. 7

Lipid balance is dysregulated in Cdkn1a-deficient hepatocytes. A schematic description of siRNA transfection is shown. (A) Cdkn1a mRNA relative expression to Gapdh in siLuc and siCdkn1a hepatocytes. (B) Comparative protein expression of CDKN1A in siLuc and siCdkn1a hepatocytes by immunoblotting analysis. β-ACTIN was used as a loading control. Ratio between CDKN1A and β-ACTIN was calculated. (C) Incorporation of glucose into lipid fractions of siLuc and siCdkn1a hepatocytes treated with 0.25 uCi of D-[¹⁴C(U)] glucose overnight. Data are presented as ratio of labeled glucose incorporation to lipids. (D) Srebp1c and (E) Chrebp1 mRNA relative expression to Gapdh in siLuc and siCdkn1a hepatocytes. Statistical significance between siCtrl and siCdkn1a primary hepatocytes in qPCR assays was determined by Student t test (∗p <0.05). (F) ORO staining performed in primary hepatocytes isolated from Cdkn1a^+/+ and Cdkn1a^-/-. (G) quantified in a graph and (H) protein expression for pAMPK, pAKT, and SIRT3 was studied in primary hepatocytes isolated from Cdkn1a^+/+ and Cdkn1a^-/- mice (two individuals’ samples from each group are shown as a representative of the group). GAPDH was used as a loading control. Scale bar = 50 μm. n = 3, one-way ANOVA with post hoc Tukey test was used for the calculations, intragroup: ∗p <0.05; ∗∗∗p <0.001; intergroup: ^#p <0.05, ^####p <0.0001). AMPK, AMP-activated protein kinase; AKT, serin/theonine kinase 1; Cdkn1a, cyclin-dependent kinase inhibitor 1A; Gadph, glyceraldehyde-3-phosphate dehydrogenase; Luc, luciferase; ORO, Oil red O; PIH, Primary hepatocytes isolated; qPCR, quantitative PCR; SIRT3, sirtuin 3.

Fig. 8

The inhibition of CDK4/6 protects mice from liver injury in advanced SLD. Schematic description of a DuAL model plus oral palbociclib administration performed in wild type BL/6 mice. (A) Cdkn1a, (B) p53, (C) Cyclin A2, and (D) Cyclin B1 mRNA expression determined by qPCR and normalized to the amount of Gapdh in liver of DuAL-fed mice and controls. (E) H&E staining in liver of DuAL-fed mice and controls was performed. Scale bar = 100 μm. (F) LW/BW ratio (%). (G) AST measurement in serum. (H) Sirius red staining was performed in livers of DuAL-fed mice and controls, and quantification was represented (I). Scale bar = 100 μm. (J) ORO staining was performed in liver cryosections of DuAL-fed mice and controls, and quantification was graphed (K). n = 5–6, Student t test was used for the calculations, ∗p <0.05. AST, aspartate amino transferase; CDK4/6, cyclin-dependent kinase 4/6; Cdkn1a, cyclin-dependent kinase inhibitor 1A; DuAL, metabolic dysfunction-associated fatty liver disease and alcohol-related liver; Gadph, glyceraldehyde-3-phosphate dehydrogenase; LW/BW, liver weight/body weight; ORO, Oil Red O; qPCR, quantitative PCR; SLD, steatotic liver disease.