Enhanced gastrointestinal expression of cytosolic malic enzyme (ME1) induces intestinal and liver lipogenic gene expression and intestinal cell proliferation in mice

Abstract

The small intestine participates in lipid digestion, metabolism and transport. Cytosolic malic enzyme 1 (ME1) is an enzyme that generates NADPH used in fatty acid and cholesterol biosynthesis. Previous work has correlated liver and adipose ME1 expression with susceptibility to obesity and diabetes; however, the contributions of intestine-expressed ME1 to these conditions are unknown. We generated transgenic (Tg) mice expressing rat ME1 in the gastrointestinal epithelium under the control of the murine villin1 promoter/enhancer. Levels of intestinal ME1 protein (endogenous plus transgene) were greater in Tg than wildtype (WT) littermates. Effects of elevated intestinal ME1 on body weight, circulating insulin, select adipocytokines, blood glucose, and metabolism-related genes were examined. Male Tg mice fed a high-fat (HF) diet gained significantly more body weight than WT male littermates and had heavier livers. ME1-Tg mice had deeper intestinal and colon crypts, a greater intestinal 5-bromodeoxyuridine labeling index, and increased expression of intestinal lipogenic (Fasn, Srebf1) and cholesterol biosynthetic (Hmgcsr, Hmgcs1), genes. The livers from HF diet-fed Tg mice also exhibited an induction of cholesterol and lipogenic pathway genes and altered measures (Irs1, Irs2, Prkce) of insulin sensitivity. Results indicate that gastrointestinal ME1 via its influence on intestinal epithelial proliferation, and lipogenic and cholesterologenic genes may concomitantly impact signaling in liver to modify this tissue's metabolic state. Our work highlights a new mouse model to address the role of intestine-expressed ME1 in whole body metabolism, hepatomegaly, and crypt cell proliferation. Intestinal ME1 may thus constitute a therapeutic target to reduce obesity-associated pathologies.

Figure 1. ME1-Tg mice on chow diet exhibit increased ME1 protein and mRNA abundance in small intestine.

A) Schematic representation of the mouse villin1-ME1 transgene construct in which a complete open reading frame for rat ME1 was placed downstream of the murine villin1 gene promoter-enhancer (12.4 kb fragment). The SV40 polyA signal-containing region was located downstream of the Me1 cDNA sequence. Red arrows indicate the location of genotyping primers, while black arrows indicate the location of primers used to detect transgene expression by RT-PCR. B) ME1-Tg mRNA expression in the jejunum and colon of WT and ME1-Tg mice detected by RT-PCR (n = 3 mice/group; Exp.1). C) Representative Western blots of ME1 protein in the jejunum, ileum and colon of WT and ME1-Tg mice (n = 2−3/group; Exp. 1). The ME1 antibody detected endogenous murine and Tg-derived rat ME1 proteins. D) Densitometric analysis of relative ME1 protein levels in panel C. E) Representative images of immunohistochemical staining of ME1 in the Ileum and colon of WT and ME1-Tg Tg mice. Scale bars = 100 µM. Arrows indicate villous epithelial and luminal epithelial staining of ME1 in the ileum and colon, respectively. F) Weight gain calculated as percentage increase of final body weight from initial body weight of WT and ME1-Tg male mice (n = 8−10 mice/group) from Exp. 1. Bar graphs represent mean ± SEM; *Significant difference at P<0.05 between genotypes. **Significant difference at P<0.01 between genotypes. P values are indicated for those tending to have significant differences between groups (0.05<P<0.10).

Figure 2. Enhanced intestinal ME1 expression promotes weight gain during consumption of HF-diet.

(A) Body and (B) liver, gonadal (GF) and retroperitoneal (RPF) fat depot weights of WT and ME1-Tg mice (n = 10 mice/group) from Exp. 2. Weight gain (A) was calculated as percentage increase of final body weight from initial body weight. C) Fasting (3–4 h) serum glucose levels at 12 wk and 18 wk of age in WT and ME1-Tg mice (Exp. 2). D) Serum levels of insulin, (E) HOMA-IR index mice (at 18 wk), and F) serum leptin and adiponectin levels of WT and ME1-Tg (at 18 wk); n = 7−8 mice/group for C–F. Bar graphs present mean ± SEM; * Significant differences at P<0.05 between genotypes. P value was indicated for those tending to have significant differences between groups (0.05<P<0.10).

Figure 3. Increased crypt cell proliferation in jejunums of ME1-Tg mice fed HF diet.

A) Western blot of ME1 protein and corresponding band densitometry analysis of WT and ME1-Tg mice fed HF diet (Exp. 2). B) ME1 enzyme activity in the jejunum of WT and ME1-Tg mice. Each data point represents an individual mouse. C) Tissue ratio of NADPH/NADPt in jejunum samples of WT and ME1-Tg mice. D–E) Representative images of jejunum villi, crypts, and BrdU staining of WT and ME1-Tg mice; scale bars = 100 µM (D and E). F) Quantification of jejunum villus height and crypt depth, and BrdU staining of jejunum of WT and Me1-Tg mice (Exp. 2). G–H) Representative images of colon crypts, and BrdU labeling of WT and ME1-Tg mice; scale bars = 100 µM (G) and 50 µM (H). I) Quantification of colon crypt depth, and BrdU staining of colons of WT and Me1-Tg mice (Exp. 2). Significant differences at *P<0.05 and ** P<0.01 between genotypes. P values are indicated for those values tending to exhibit significant differences between groups (0.05<P<0.10). For panels B, C, F and I, data were from 5–6 mice/group.

Figure 4. Enhanced intestinal ME1 expression with HF diet induces jejunum lipogenic- and proliferation-associated gene expression.

A) Relative expression of jejunum genes of WT and ME1-Tg mice [Exp. 2 (n = 7−9/group)]. B) Western blot and corresponding band densitometry analysis of FASN in the jejunum of WT and ME1-Tg mice (Exp. 2). C–E) Effects of ME1 over-expression on intestinal epithelial cell proliferation and gene expression in vitro. Rat intestinal epithelial cells (IEC-6) were transfected with control or hME1 expression vectors; overexpression of human ME1 mRNA, but not beta-actin mRNA, was observed by RT-PCR (C). At 48 h, cells were evaluated for proliferation by MTT assay (D), and at 96 h (E) evaluated for expression of Fasn, Scd1, Rxrg and Lpl genes (n = 3 replicates/group). F) Gene expression of Fasn, Angptl4, Lpl, Fgr, Irs1 and Irs2 in jejunums of WT and MOD-1 mice fed HF diet (n = 5/group). RT-PCR was repeated twice in all experiments, and MTT proliferation assay was repeated thrice. Bar graphs represent mean ± SEM. Significant differences at *P<0.05 and ** P<0.01 between genotypes.

Figure 5. Expression of lipogenic pathway and cholesterol synthesis pathway genes in livers of WT and ME1-Tg mice fed HF diet.

A) qRT-PCR analysis of major lipogenic pathway genes in livers of WT and ME1-Tg mice fed HF diet (Exp. 2; n = 8−9/group). B) Western blot of FASN protein in livers of WT and ME1-Tg mice fed HF diet. C) Densitometric analysis of immunoreactive bands in (B) relative to α-Tubulin protein. D) Western blot of IRS1, pSer307-IRS1, and IRS2 in livers of WT and ME1-Tg mice fed HF diet (n = 5/group). E–G) Densitometric analysis of immunoreactive bands of total liver IRS1 (E) and IRS2 (G) and the relative ratio of immunoreactive pSer307-IRS1/total IRS1 band densities (F). H) mRNA expression of select cholesterol synthesis- and cholesterol uptake-related genes in the livers of WT and ME1-Tg mice (Exp. 2; n = 8−9/group). qRT-PCR reactions were repeated twice in all experiments (A, H). Bar graphs represent mean ± SEM. Significant differences at *P<0.05 and ** P<0.01 between genotypes.

Figure 6. Liver lipid content correlates with Pparg gene expression in mice fed HF diet.

A) Quantification of Oil Red O staining intensity of liver samples using Aperio Image Scope software. B) Pearson’s correlation analysis for the level of liver mRNA expression of Pparg with Oil Red O score of WT and ME1-Tg mice. C) Comparisons of liver mRNA levels of Pparg in WT and ME1-Tg mice that exhibited Oil Red O score of >20; n = 4−5/group. Bar graphs represent mean ± SEM. Significant differences at *P<0.05 between genotypes. P value is indicated for that tending to have a significant difference between groups (0.05<P<0.10).

Figure 7. Proposed model summarizing the major findings and implications of this study.

(?) Represents current unknowns.