Targeting YAP-mediated HSC death susceptibility and senescence for treatment of liver fibrosis

Abstract

Background and aims: Liver fibrosis results from the accumulation of myofibroblasts (MFs) derived from quiescent HSCs, and yes-associated protein (YAP) controls this state transition. Although fibrosis is also influenced by HSC death and senescence, whether YAP regulates these processes and whether this could be leveraged to treat liver fibrosis are unknown.

Approach and results: YAP activity was manipulated in MF-HSCs to determine how YAP impacts susceptibility to pro-apoptotic senolytic agents or ferroptosis. Effects of senescence on YAP activity and susceptibility to apoptosis versus ferroptosis were also examined. CCl 4 -treated mice were treated with a ferroptosis inducer or pro-apoptotic senolytic to determine the effects on liver fibrosis. YAP was conditionally disrupted in MFs to determine how YAP activity in MF-HSC affects liver fibrosis in mouse models. Silencing YAP in cultured MF-HSCs induced HSC senescence and vulnerability to senolytics, and promoted ferroptosis resistance. Conversely, inducing HSC senescence suppressed YAP activity, increased sensitivity to senolytics, and decreased sensitivity to ferroptosis. Single-cell analysis of HSCs from fibrotic livers revealed heterogeneous sensitivity to ferroptosis, apoptosis, and senescence. In mice with chronic liver injury, neither the ferroptosis inducer nor senolytic improved fibrosis. However, selectively depleting YAP in MF-HSCs induced senescence and decreased liver injury and fibrosis.

Conclusion: YAP determines whether MF-HSCs remain activated or become senescent. By regulating this state transition, Yap controls both HSC fibrogenic activity and susceptibility to distinct mechanisms for cell death. MF-HSC-specific YAP depletion induces senescence and protects injured livers from fibrosis. Clarifying determinants of HSC YAP activity may facilitate the development of novel anti-fibrotic therapies.

Fig. 1. YAP mediates HSC ferroptosis susceptibility.

(A-D) LX2 cells were grown at low density (< 30% confluence) or high density (100% confluence) and some were treated with erastin or RSL3 for 48h. (A) YAP localization. (B) YAP and p-YAP in total cell lysates, cytosolic and nucleus fractions. (C) Expression of Hippo-YAP pathway target genes. (D) Cell viability in low or high cell density cultures treated with indicated doses of erastin or RSL3. *p<0.05 vs low cell density. (E-F) LX2 cells were treated with nontargeting RNA (siNT) or small interfering RNA targeting either YAP (siYAP) or TAZ (siTAZ). (E) Protein expression of YAP, TAZ and myofibroblast or proliferation markers. (F) Viability of cells treated with indicated doses of erastin or RSL3. *p<0.05 vs siNT. (G-K) LX2 cells were treated with TGFβ or vehicle for 48h; erastin or vehicle were added for another 48h. (G) Expression of MF-HSC markers. (H) Ferroptosis susceptibility. (I) GSEA enrichment plot of YAP1-driven-ECM-axis pathway in TGFβ-treated HSCs in RNA-seq dataset GSE11954. (J) Increased expression of YAP target genes in TGFβ-treated HSCs. (K) Differential effects of YAP vs TAZ knockdown on TGFβ-induced ferroptosis sensitivity. *p<0.05 vs vehicle ctrl. ^# p<0.05 vs TGFβ + siNT. (L) HSC ferroptosis sensitivity tracks with mesenchymal status. Data are graphed as mean ± sem of n=3–5 assays/group.

Fig. 2. YAP controls ferroptosis sensitivity of HSCs by regulating P21-GPX4 axis.

(A) Knock-down YAP but not TAZ by siRNA increased P21 expression. (B, C) Inhibiting P21 by siRNA increased HSC ferroptosis sensitivity and partially abolished YAP-depletion induced ferroptosis resistance. (D, E, F) Inhibiting P21 by pharmacological inhibitor UC2288 increased HSC ferroptosis sensitivity and partially abolished ferroptosis resistance induced by YAP-depletion. (G, H) P21 overexpression decreased HSC sensitivity to erastin or RSL3 induced ferroptosis. (I, J) Knock-down YAP but not TAZ maintained P21 and GPX4 expression during erastin treatment. (K) High cell density increased P21 and GPX4 expression. (L) TGFβ treatment decreased P21 and GPX4 expression. (M) Hippo-YAP pathway activity determines ferroptosis sensitivity in HSCs by regulating P21-GPX4 axis. Data are graphed as mean ± sem of n=3–5 assays/group. *p<0.05 vs siNT (B, I), or empty vector ctrl (G, H), or low cell density (K), or vehicle ctrl (D, L). ^# p<0.05 vs siYAP (C, E).

Fig. 3. YAP deficiency in MF-HSCs induces senescence and differential sensitivity to ferroptosis and pro-apoptotic senolytics.

LX2 cells or primary human HSCs were treated with nontargeting RNA (siNT) or small interfering RNA targeting either YAP (siYAP) or TAZ (siTAZ) for 7 days. LX2 cells: (A, B) Representative SA-β-Gal staining and quantification of positive stained cells. (C, D, E) YAP, but not TAZ, deficiency upregulated SASP factors, Bcl2 family proteins and sensitivity to ABT-263. Primary human HSCs: (F) Effect of YAP and/or TAZ deficiency on the expression of indicated proteins. (G) SA-β-Gal staining and corresponding quantification. (H) YAP and TAZ deficiency induced expression of SASP factors. (I) Cells were treated with indicated doses of erastin or ABT263 for 2 days, and cell viability was measured by CCK8 assay. Data are graphed as mean ± sem of n=3–5 assays/group. *p<0.05 vs siNT; ^#p<0.05 vs siYAP and siTAZ.

Fig 4.. Senescent HSCs exhibit impaired YAP/TAZ activity, decreased sensitivity to ferroptosis but increased sensitivity to Bcl2 inhibitor-based senolytics.

LX2 cells were treated with 250nM etoposide or P21 plasmid for 7 days to induce HSC senescence. (A) SA-β-Gal staining and quantification of positively stained cells. (B) Expression of SASP factors. (C) Expression of P21, YAP and TAZ. (D) Expression of YAP/TAZ target genes. (E, F) Comparison of cell viability of etoposide-treated HSCs treated with indicated doses of erastin or ABT-263. *p<0.05 vs ctrl. (G-I) P21 overexpression (OE) by P21 plasmid vector increased SA-β-Gal staining, upregulated SASP factors and increased susceptibility to senolytics ABT263 and A-1331852. *p<0.05 vs ctrl empty vector. Data are graphed as mean ± sem of n=3–5 samples/group.

Fig. 5. Cysteine depletion increases liver injury and exacerbates liver fibrosis.

C57BL/6J male mice were treated with corn oil vehicle or CCl₄ for 4 weeks, and cyst(e)inase or its vehicle was administered i.p. during the final 2 weeks. (A, B) Serum levels of AST and bilirubin. (C, D) Representative images of liver sections stained for H&E, TUNEL and corresponding quantification of TUNEL positive hepatocytes. (E) mRNA expression of liver fibrosis markers. (F, G) Representative images of liver sections stained for Sirius red, αSMA and desmin and corresponding quantification of the positive IHC stained area. Data are graphed as mean ± sem (n=3 mice/group in corn oil groups, and n=7 mice/group in CCl₄ groups). *p<0.05 vs corn oil; ^#p<0.05 vs CCl₄ + veh.

Fig. 6. Systemic administration of senolytic A-1331852 exacerbates liver injury and fibrosis.

C57BL/6J male mice were treated with corn oil vehicle or CCl₄ for 6 weeks, and senolytic A-1331852 was orally administered during the final 2 weeks. (A, B) Serum levels of ALT and AST. (C, D) Representative images of liver sections stained for H&E, αSMA and Sirius red and corresponding quantification of positively stained areas. (E, F, G) Expression of MF-HSC and liver fibrosis markers and p21. (H, I) Representative staining for HSC senescence (SA-β-Gal⁺ & Desmin⁺) and quantification of SA-β-Gal⁺ areas. (J) Expression of p16, p21 and SASP factors. Data are graphed as mean ± sem (n=4–5 mice/group). *p<0.05 vs CCl₄ + veh.

Fig 7.. Selective YAP depletion in MF-HSCs induces senescence and protects against BDL -induced liver injury, fibrosis and ductular reaction.

αSMA-CreER^T2 x Yap1^flox/flox mice were subjected to sham or BDL surgery for 14 days to promote MF-HSC accumulation and liver fibrosis. Tamoxifen or its vehicle was intraperitoneally administered every other day from day 4 until liver tissue was harvested. (A, B) Expression of Yap1 and its target genes. (C, D) Yap1 and aSMA co-immunofluorescence staining and quantification of the double positive cells. (E) Quantification of SA-β-Gal positively-stained areas. (F) mRNA expression of SASP factors. (G) Representative staining for HSC senescence (SA-β-Gal⁺ & Desmin⁺_, as indicated by arrows), liver histology (H&E), MF-HSC accumulation and liver fibrosis (Sirius Red, aSMA, desmin), and ductular reaction (CK7 and CK19). (H) Serum levels of ALT, AST and bilirubin to assess liver injury. (I, J) Quantification of the IHC positive area in staining of Sirius Red, aSMA, desmin, CK7 or CK19. Data are graphed as mean ± sem of n= 4 in sham mice and n=6–7 mice /group in BDL mice. *p<0.05 vs sham, ^#p<0.05 vs Yap1^flox/flox mice.

Fig 8.. Selective YAP depletion in MF-HSCs induces senescence and protects against CCl₄-induced liver injury and fibrosis.

αSMA-CreER^T2 x Yap1^flox/flox mice were treated with CCl₄ for 6 weeks. Tamoxifen or its vehicle was intraperitoneally administered every other day in the last two weeks. (A, B) Yap1 and aSMA co-immunofluorescence staining and quantification of the double positive cells. (C, D) Expression of Yap1 target genes and SASP factors. (E) Serum levels of ALT and AST. (F, G) Representative staining for HSC senescence (SA-β-Gal⁺Desmin⁺_, as indicated by arrows) and quantification of SA-β-Gal positively-stained areas and markers of liver fibrosis (Sirius Red, aSMA, desmin). (H) Representative staining for liver fibrosis markers (Sirius Red, aSMA, desmin). Data are graphed as mean ± sem of n=6–7 mice /group. *p<0.05 vs Yap1^flox/flox mice.