Background: Therapeutic approaches to impair intestinal lipid absorption are widely employed in the management of obesity and metabolic dysfunction. Intestine-specific microsomal triglyceride transfer protein deletion (MttpIKO) blocks chylomicron secretion, impairs lipid absorption and enterohepatic bile acid (BA) metabolism and reduces hepatic steatosis. Here we asked if those adaptations alter hepatic tumorigenesis. Methods: Chow-fed WT and MttpIKO mice received either diethylnitosamine (DEN) or underwent hydrodynamic tail vein injection (HDI) of CRISPR-Cas9 and transposon vectors to disrupt Trp53 and overexpress C-Myc (Trp53^KO/C-Myc^OE). Results: DEN treated MttpIKO mice exhibit increased tumor burden at 6 and 12-month with increased proliferation and CCND1 expression (Fig1A-D). We observed increased BA pool size (WT 60.9±14.9 vs MttpIKO 111.1±19.1µmol/100g BW), with a shift to small intestine BA accumulation (WT 105.3±14.1 vs MttpIKO 184.4 ± 17.7 nmol/mg protein) along with liver Fxr inactivation (increased Cyp7a1) and intestinal Fxr activation (increased Ilbp, Asbt, Osta) (Fig1G,H). We found increased hepatic oxysterol 7KC (0.06 vs 0.12 peak area/ mg protein) along with hepatic Lxr activation (increased Abcg5, Hmgcr) and decreased Cyp7b1 (Fig 1G). Those findings together suggest altered LXR/FXR signaling in the liver of MttpIKO mice. We also observed increased serum FD4 and decreased mRNA expression of Zo-1 and Lamb1 suggesting impaired intestinal barrier in MttpIKO mice (Fig 1E, F and H). Those MttpIKO mice also exhibited increased portal vein LPS and hepatic TLR4 and NFkb proteins, suggesting pathogen associated liver inflammation (Fig 1F, G and I). We observed decreased hepatic PTEN signaling and increased STAT3 protein in MttpIKO (Fig 1 I), which are known signaling pathways in HCC development. Next, we explored liver tumor formation by suppressing p53 and overexpressing c-myc through HDI. Again, we observed increased mortality and hepatic tumor burden in MttpIKO mice (Fig 2 A to C). Because we observed increased hepatic CCND1 expression (Fig 1D, Fig. 2D and E) in both tumor models, we asked if CCND1 protein expression could be altered by manipulation of intestinal Fxr signaling. We found CCND1 expression is increased at baseline in MttpIKO mice and following Fxr agonist administration (GW4064) but decreased in mice following Fxr inactivation (Tempol) (Fig. 2F). Conclusion: Intestinal Mttp deletion induced alterations in enterohepatic bile acid and oxysterol metabolism, which in turn signal through altered Fxr/Lxr activation. These changes, together with impaired intestinal barrier function and bacterial translocation, promote hepatic proliferation, inflammation and tumorigenesis. Therapeutic strategies to impair intestinal lipid absorption may have long term adverse consequences for hepatic tumorigenesis.

Figure1. Increased proliferation and tumor formation in MttpIKO after DEN treatment, along with altering sterol metabolism, impairing intestinal barrier and related gene expression changes. A. Visual tumor counts. B. Tumor area of HE staining tissue quantitated by Image J. The bar graph data associated with different letters are significant different. p<0.05. C. BrdU staining cells%. D. Cyclin D1 (CCND1) protein measured by western blot, corrected to albumin (ALB), normalized to WT. E. Serum FITC-Dextran 4Kd (FD4) 4hr after gavage. F. Portal vein Lipopolysaccharides (LPS) content. G. and H. mRNA expression of genes related to Fxr, Lxr pathway, inflammation (liver), and barrier (Ileum) were measured by qPCR, corrected to Gapdh, normalized to WT. I. Western blot images of PTEN, NFκB (left), STAT3 (middle) and their quantitation by Kodak system (right), corrected to β-actin, normalized to WT. All the Bar graph are mean ± SE, *p<0.05.

Figure2. <b>A.</b> to <b>E.</b> Increased death, liver tumor formation, liver proliferation in MttpIKO after hydrodynamic tail vein injections (HDI) of CRISPR-Cas9 and transposon vectors to disrupt Trp53 and overexpress C-Myc (Trp53<sup>KO</sup>/C-Myc<sup>OE</sup><b>. A. </b>Survival curve. n=9-10 per group. <b>B.</b> Liver weight corrected to body weight %. <b>C. </b>Visible tumor counts per liver. <b>D. </b>mRNA expression of genes related to sterol metabolism and proliferation measured by qPCR, corrected to <i>Gapdh</i>, normalized to WT. <b>E.</b> Representative of Cyclin D1 (CCND1) protein expression by western blot. <b>F.</b> Representative images of western blot of CCND1 and quantitation by Image Lab software in WT and MttpIKO at baseline (left), FXR agonist GW4064 treated (middle) and Tempol (Right) treated MttpIKOs. All the Bar graphs are mean ± SE, *p<0.05.

Figure2. A. to E. Increased death, liver tumor formation, liver proliferation in MttpIKO after hydrodynamic tail vein injections (HDI) of CRISPR-Cas9 and transposon vectors to disrupt Trp53 and overexpress C-Myc (Trp53^KO/C-Myc^OE. A. Survival curve. n=9-10 per group. B. Liver weight corrected to body weight %. C. Visible tumor counts per liver. D. mRNA expression of genes related to sterol metabolism and proliferation measured by qPCR, corrected to Gapdh, normalized to WT. E. Representative of Cyclin D1 (CCND1) protein expression by western blot. F. Representative images of western blot of CCND1 and quantitation by Image Lab software in WT and MttpIKO at baseline (left), FXR agonist GW4064 treated (middle) and Tempol (Right) treated MttpIKOs. All the Bar graphs are mean ± SE, *p<0.05.

Elizabeth Brunt

Appearances