Shp2 deletion in hepatocytes suppresses hepatocarcinogenesis driven by oncogenic β-Catenin, PIK3CA and MET

Abstract

Background & aims: Shp2 is an SH2-tyrosine phosphatase acting downstream of receptor tyrosine kinases (RTKs). Most recent data demonstrated a liver tumor-suppressing role for Shp2, as ablating Shp2 in hepatocytes aggravated hepatocellular carcinoma (HCC) induced by chemical carcinogens or Pten loss. We further investigated the effect of Shp2 deficiency on liver tumorigenesis driven by classical oncoproteins c-Met (receptor for HGF), β-catenin and PIK3CA.

Methods: We performed hydrodynamic tail vein injection of two pairs of plasmids expressing c-Met and ΔN90-β-catenin (MET/CAT), or c-Met and PIK3CA^H1047R (MET/PIK), into WT and Shp2^hep-/- mice. We compared liver tumor loads and investigated the pathogenesis and molecular mechanisms involved using multidisciplinary approaches.

Results: Despite the induction of oxidative and metabolic stresses, Shp2 deletion in hepatocytes suppressed hepatocarcinogenesis driven by overexpression of oncoproteins MET/CAT or MET/PIK. Shp2 loss inhibited proliferative signaling from c-Met, Wnt/β-catenin, Ras/Erk and PI3K/Akt pathways, but triggered cell senescence following exogenous expression of the oncogenes.

Conclusions: Shp2, acting downstream of RTKs, is positively required for hepatocyte-intrinsic tumorigenic signaling from these oncoproteins, even if Shp2 deficiency induces a tumor-promoting hepatic microenvironment. These data suggest a new and more effective therapeutic strategy for HCCs driven by oncogenic RTKs and other upstream molecules, by inhibiting Shp2 and also suppressing any tumor-enhancing stromal factors produced because of Shp2 inhibition.

Lay summary: Primary liver cancer is a malignant disease with poor prognosis, largely because there are limited systemic therapies available. We show here that a cytoplasmic tyrosine phosphatase Shp2 is required for liver tumorigenesis. This tumorigenesis is driven by two oncoproteins that are implicated in human liver cancer. This, together with our previous studies, uncovers the complexity of liver tumorigenesis, by elucidating the pro- and anti-tumor effects of Shp2 in mouse models. This data can be used to guide new therapies.

Keywords: Liver tumorigenesis; Met; PIK3CA; Shp2; β-Catenin.

Fig. 1. Shp2 deletion suppresses MET/CAT-induced hepatocellular cancer

(A). Macroscopic images and H&E staining of mouse liver sections at day 7 and week 4, 6 and 8 post-hydrodynamic injection of hMET and Δβ-Catenin (MET/CAT) constructs. Scale bars: 1 cm (macroscopic); 100 μm (microscopic). (B). Liver versus body weight ratios were measured at various time points (n=3, day-7; week 4 and week 6; n=8, week 8). Tumor incidences and maximal tumor sizes (n=8) were measured for WT and Shp2^hep−/− mice at week 8 post-injection, and compared using student’s t test. ***P<0.001, ****P<0.0001.

Fig. 2. Shp2 loss inhibits MET/PIK3CA-induced liver tumorigenesis

(A). Representative macroscopic images and H&E staining of mouse liver sections at various time points post injection of hMET and PIK3CA (MET/PIK) constructs. Scale bars: 1 cm (macroscopic), 100 μm (microscopic). (B). Liver versus body weight ratios of MET/PIK-transfected mice were measure at various time points (n=3, D-7; n=4, Wk-6; n=4, Wk-9; n=7, Wk-12) (**** P<0.0001 by t test). (C). Oil-Red-O staining of liver sections at week 12 post-injection of MET/PIK. Scale bars: 100 μm.

Fig. 3. Comparative analysis of transcriptomes in WT and Shp2^hep−/− livers

(A). Heatmap visualizing the FPKM values for differentially expressed genes across all 6 groups (n=3): WT and Shp2^hep−/− livers before (Day-0) or after (Day-3) transfection of MET/CAT or MET/PIK, with red and blue indicating high and low gene expression, respectively. Highly expressed gene clusters are highlighted and numbered. (B). GSEA analysis showing the most enriched gene sets in each highly expressed gene cluster. (C). Comparison of baseline transcripts of HGF/c-Met targets between WT and Shp2^hep−/− livers at Day 0. For each target gene, x- and y-axis values are its FPKM values in WT liver and Shp2^hep−/− livers, respectively. Printed in red are the genes expressed at significantly lower levels in Shp2^hep−/− than WT livers. (D). Comparison of transcript levels of Wnt target genes between Shp2^hep−/− and WT livers at Day 0. Red, blue and grey colors indicate the genes down-regulated, up-regulated or not affected by the Wnt pathway. (E). Ingenuity Pathway Analysis showing the top-ranked positively regulated pathways in Shp2^hep−/− livers, relative to WT controls at Day 0. (F). Overlap of differentially expressed genes (DEGs) in both Day-0 WT vs. Shp2^hep−/− livers, and GFP Day-7 WT vs. Shp2^hep−/− livers. The numbers indicate DEGs identified in both or only one comparison. Pathway analysis was performed on the DEGs from indicated groups.

Fig. 4. Transcriptomic analysis of MET/CAT- or MET/PIK-transfected WT and Shp2^hep−/− livers

(A)–(C). IPA of Day-3 MET/CAT-transfected WT or Shp2^hep−/− livers versus Day-0 WT or Shp2^hep−/− livers (A); Top-ranked up-regulated gene sets in Day-3 MET/PIK-transfected WT or Shp2^hep−/− livers versus Day-0 WT or Shp2^hep−/− livers (B); IPA of Day-3 MET/PIK-transfected WT or Shp2^hep−/− livers versus Day-0 WT or Shp2^hep−/− liver (C). (D). Heatmap representation of TNFa-induced NF-kB target genes. Gene expression variations are presented by fold change of expression with red and blue indicating increase and decrease respectively, compared to genotype-matched Day-0 controls. (E)–(F). Enrichment plots of gene set: CACGTG_V$MYC_Q2 comparing Day-3 MET/CAT-transfected Shp2^hep−/− and WT livers (E); gene sets: Hallmark_Fatty_Acid_Metabolism and Go_Oxidation_Reduction_Process comparing Day-3 MET/PIK-transfected Shp2^hep−/− and WT livers (F). (G). GSEA analysis of DEGs between MET/CAT-transfected Shp2^hep−/− and WT livers at Day 7. (H). GSEA analysis of DEGs identified by comparing MET/PIK-WT livers versus GFP-WT liver at day 7, or MET/PIK-Shp2^hep−/− livers versus GFP-Shp2^hep−/− livers at day 7.

Fig. 5. Shp2 loss suppresses critical proliferative signaling events induced by the injected oncogenes

(A). Immunoblotting of liver lysates for the exogenous human c-Met and ΔN90-β-catenin (Δβ-cat), and p-Erk and Erk proteins at various time points after injection of MET/CAT, or a GFP vector control (* the endogenous β-catenin). (B). Immunoblotting was performed for liver lysates collected at different time points after MET/PIK injection, to evaluate c-Met, p110α, and other signaling proteins as indicated, with a GFP vector control included. (C). Immunostaining for exogenous c-Met and β-catenin on liver sections at day 7, week 5 and 8 post injection of MET/CAT. Scale bars: 100 μm. (D). Immunostaining of glutamine synthetase (Gl syn) and p-Erk on liver sections prepared at day 7, week 5 and 8 post MET/CAT injection. Scale bars: 100 μm. (E). Immunostaining of WT and Shp2^hep−/− liver sections for exogenous c-Met at day 7, week 6, 9 and 12 post MET/PIK injection. Scale bars: 100 μm.

Fig. 6. Shp2 deficiency affects HGF and Wnt3a signaling in vivo and in vitro

(A). WT and Shp2^hep−/− livers were stimulated through portal vein with HGF or Wnt3a plus R-Spondin1 in vivo. Immunoblotting of liver lysates was performed to examine phosphorylation of c-Met and β-Catenin and more downstream effectors, using antibodies as indicated. (B). WT and Shp2^−/− primary hepatocytes were isolated and then stimulated with HGF or Wnt3a plus R-Spondin1 for 2 or 5 minutes in vitro. Immunoblotting of cell lysates was performed to examine phosphorylated and total proteins of the components in HGF/c-Met and Wnt pathway, using indicated antibodies. (C). qRT-PCR was performed to evaluate transcript levels of Wnt pathway target genes in WT and Shp2^−/− primary hepatocytes stimulated with Wnt3a plus R-spondin1 (Wnt3a+R-S) or Wnt3a plus R-spondin1 and HGF (HGF+Wnt3a+R-S) for 1 hr (n=3, #P<0.05, +P<0.05 by t test)

Fig. 7. Shp2 deficiency promotes oncogene-induced cell senescence

(A). Representative SA-β-galactosidase staining of WT and Shp2^hep−/− liver sections at day12 after injection of MET/CAT or MET/PIK constructs. Percentage of SA-β-Gal-positive cells versus total cells is presented on the right as mean ± SE (n=3–4). Scale bars: 200 μm, **P<0.01, ****P<0.0001 by t test. (B). qRT-PCR was performed to determine relative expression of the genes as indicated (n=3–4). *P<0.05, **P<0.01 by t test. (C). Representative images of co-staining of HNF4α and Ki67 of liver sections on day 12 post injection. Percentage of Ki67⁺ cells in total HNF4α⁺ cells was quantified and presented as mean ± SE (n=3–4). Scale bars: 100 μm, *P<0.05, **P<0.01 by t test.