Deletion of intestinal epithelial insulin receptor attenuates high-fat diet-induced elevations in cholesterol and stem, enteroendocrine, and Paneth cell mRNAs

Abstract

The insulin receptor (IR) regulates nutrient uptake and utilization in multiple organs, but its role in the intestinal epithelium is not defined. This study developed a mouse model with villin-Cre (VC) recombinase-mediated intestinal epithelial cell (IEC)-specific IR deletion (VC-IR^Δ/Δ) and littermate controls with floxed, but intact, IR (IR^fl/fl) to define in vivo roles of IEC-IR in mice fed chow or high-fat diet (HFD). We hypothesized that loss of IEC-IR would alter intestinal growth, biomarkers of intestinal epithelial stem cells (IESC) or other lineages, body weight, adiposity, and glucose or lipid handling. In lean, chow-fed mice, IEC-IR deletion did not affect body or fat mass, plasma glucose, or IEC proliferation. In chow-fed VC-IR^Δ/Δ mice, mRNA levels of the Paneth cell marker lysozyme (Lyz) were decreased, but markers of other differentiated lineages were unchanged. During HFD-induced obesity, IR^fl/fl and VC-IR^Δ/Δ mice exhibited similar increases in body and fat mass, plasma insulin, mRNAs encoding several lipid-handling proteins, a decrease in Paneth cell number, and impaired glucose tolerance. In IR^fl/fl mice, HFD-induced obesity increased circulating cholesterol; numbers of chromogranin A (CHGA)-positive enteroendocrine cells (EEC); and mRNAs encoding Chga, glucose-dependent insulinotrophic peptide (Gip), glucagon (Gcg), Lyz, IESC biomarkers, and the enterocyte cholesterol transporter Scarb1. All these effects were attenuated or lost in VC-IR^Δ/Δ mice. These results demonstrate that IEC-IR is not required for normal growth of the intestinal epithelium in lean adult mice. However, our findings provide novel evidence that, during HFD-induced obesity, IEC-IR contributes to increases in EEC, plasma cholesterol, and increased expression of Scarb1 or IESC-, EEC-, and Paneth cell-derived mRNAs.

the insulin receptor (IR) mediates insulin action on glucose and lipid metabolism in a number of organs, including the liver, skeletal muscle, and adipose tissue. The IR is a heterotetrameric receptor tyrosine kinase that shares significant structural homology with insulin-like growth factor (IGF) 1 receptor (IGF1R) (11). The IR exists as two isoforms produced by alternative splicing. IR isoform A (IR-A) binds both insulin and IGFs and is implicated in growth during fetal development and in cancer (11, 36, 55). IR isoform B (IR-B) is considered the metabolic IR and is highly expressed in differentiated tissues and those sensitive to the metabolic actions of insulin (4, 6, 14, 33, 66, 72). IR-A, IR-B, and IGF1R and their ligands, insulin, IGF1, and IGF2, comprise the insulin-IGF system, which controls metabolism and growth in numerous tissues (37). The IR and IGF1R display the highest affinity for insulin and IGFs, respectively, but also exhibit ligand cross-reactivity (11, 37). High levels of insulin can bind and activate IGF1R. IGF1 and IGF2 can also bind and activate IR, although IR-A binds IGFs, particularly IGF2, with higher affinity than IR-B. In vitro evidence suggests that, in some situations, signaling through the remaining receptor can compensate for loss of the IR or IGF1R (19, 20, 23, 27, 78).

The role of the IR has been most thoroughly characterized in the liver, adipose tissue, and skeletal muscle, even though the IR is expressed in most, if not all, other tissues (37). The development of tissue-specific IR knockout mice has enabled studies of the role of the IR in individual tissues, including skeletal muscle, liver, white and brown adipose tissue, the central nervous system, and the myeloid lineage (17, 21, 22, 42, 52, 60). Recent work from our group demonstrates that the IR is expressed in the intestinal epithelium, but its role in growth or function of this highly proliferative tissue is not defined (4). A number of studies in hamsters (1, 34, 43), rats (84), desert gerbils (88), and humans (32, 48, 83) link hyperinsulinemia, insulin resistance, and/or type 2 diabetes mellitus with biomarkers of intestinal insulin resistance and increased intestinal chylomicron production or aberrant lipid handling by enterocytes (2). All of these studies were performed in situations of whole body insulin resistance. The functional roles of the IR in the intestinal epithelium, specifically its potential contributions to normal growth or renewal of the epithelium and/or to the growth or metabolic responses to high-fat diet (HFD)-induced obesity and insulin resistance, are not defined.

The intestinal epithelium is a single layer of cells lining the luminal surface of the intestine that is responsible for nutrient digestion, absorption, and barrier functions. This tissue is also the most proliferative in the body; it renews itself every 3–10 days from a small population of intestinal epithelial stem cells (IESC) located at the crypt base (9, 24, 39, 86, 87). Actively cycling IESC are marked by mRNA expression of leucine-rich repeat-containing G protein-coupled receptor 5 (Lgr5), achaete-scute complex homolog 2 (Ascl2), and olfactomedin 4 (Olfm4) and low levels of sex-determining region Y-box 9 (Sox9). Other studies support the existence of a more quiescent, activatable IESC population identified by expression of Bmi1 polycomb ring finger oncogene (Bmi1), HOP homeobox (Hopx), leucine-rich repeats and immunoglobulin-like domains 1 (Lrig1), and high levels of Sox9 (9, 35, 41, 63, 70, 75, 80, 82). The IESC self-renew and generate rapidly proliferating, transit-amplifying progenitor cells (9, 74, 79). As the progenitor cells migrate out of the crypts, they undergo cell cycle arrest and differentiate into postmitotic specialized cells, including enterocytes, enteroendocrine cells (EEC), goblet cells, and Paneth cells (9, 16, 24, 74, 79). Terminally differentiated, absorptive enterocytes are marked by expression of the brush border enzyme sucrase isomaltase (Sis), goblet cells by mucin-2 (Muc2), EEC by chromogranin A (Chga), and Paneth cells by lysozyme (Lyz). Enterocytes also express a number of receptors responsible for mediating nutrient absorption (44, 49, 51); EEC secrete hormones involved in metabolic regulation, including the incretins glucose-dependent insulinotrophic peptide (GIP) and glucagon-like peptide (GLP) 1 (GLP1) (62); and Paneth cells produce a variety of antimicrobial peptides, including the defensins (7).

IGF1 is known to play a prominent role in regulating intestinal epithelial renewal, adaptation, and recovery from injury (18, 29–31, 65, 82). However, the role of insulin in intestinal epithelial growth is not well defined. One study in rats demonstrated that insulin increased proliferation in colonic epithelium (77), while another study demonstrated that oral insulin increased cell proliferation in vitro and in a rat model of short bowel syndrome (12, 13). In humans, elevated levels of circulating insulin have been linked to decreased apoptosis in the normal colonic mucosa and an increased risk of colorectal adenoma (50, 71).

IR-A, IR-B, and IGF1R are each expressed within the intestinal epithelium (4, 65). We recently demonstrated that IR-A expression predominates in IESC and proliferative progenitor cells of the crypt, while IR-B expression is enriched in differentiated intestinal epithelial cells (IEC) (4). In Caco-2 cells, increases in IR-B accompanied differentiation and forced IR-B expression accelerated increases in differentiation markers (4). Other studies provide evidence that IGF1R is enriched in IESC (57) and that IGF1R is required for GLP2-mediated increases in intestinal growth (65). Recent work in cancer cell lines and other cancer models suggests that, in a tumor environment, the IR, particularly IR-A, may mediate proliferation and compensate for loss or inhibition of signaling through IGF1R (4, 23, 78). The specific roles of IGF1R or the two IR isoforms in mediating proliferation, differentiation, or metabolic functions of insulin in the normal intestinal epithelium, in vivo or during obesity and hyperinsulinemia induced by long-term HFD, are not well defined.

In the present study we generated mice with specific disruption of the IR gene in all IEC. We cross-bred mice with loxP-modified IR alleles (IR^fl/fl) and mice expressing villin-Cre (VC) recombinase to generate VC-IR^Δ/Δ mice with IEC-IR gene disruption. The VC transgene is expressed in all small intestinal IEC in both crypts and villi (56). We hypothesized that loss of IEC-IR would 1) alter intestinal epithelial growth via loss of IR-A or, potentially, by favoring signaling through the remaining IGF1R, 2) affect numbers or biomarkers of differentiated IEC lineages due to loss of IR-B, and 3) alter the effect of HFD-induced obesity and hyperinsulinemia on intestinal growth, differentiated lineages, or nutrient handling. We first examined the effects of IEC-IR loss on basal intestinal epithelial structure. We then performed a HFD challenge, which induces obesity, hyperinsulinemia, and elevated plasma IGF1 in wild-type mice (3, 28, 38, 64, 67, 85). By comparing IR^fl/fl mice with intact IR and VC-IR^Δ/Δ mice with IEC-IR deletion fed chow or HFD, we aimed to assess whether loss of IEC-IR affects metabolic responses to HFD or HFD-induced changes in intestinal epithelial growth or gene expression. Our findings indicate that loss of IEC-IR does not affect normal intestinal morphology, measures of intestinal epithelial growth, or glucose tolerance in lean, chow-fed animals. After HFD consumption that leads to obesity and hyperinsulinemia, loss of IEC-IR has little impact on intestinal morphology or measures of intestinal epithelial growth. In mice with HFD-induced obesity and hyperinsulinemia, loss of IEC-IR attenuated HFD-associated increases in plasma cholesterol, mRNAs encoding the cholesterol transporter Scarb1, and biomarkers of IESC. Loss of IEC-IR also attenuated HFD-associated increases in numbers of CHGA-positive EEC, Chga mRNA, and mRNAs encoding GIP (Gip) and glucagon (Gcg), as well as Lyz and Def1a mRNAs, produced in Paneth cells.

MATERIALS AND METHODS

IR^fl/fl and VC-IR^Δ/Δ mice.

Mice were maintained in a specific-pathogen-free facility at the University of North Carolina at Chapel Hill; food (PMI Prolab RMH 3000, LabDiet, St. Louis, MO) and water were provided ad libitum. The IR^fl/fl mice were originally characterized and generously provided by C. Ronald Kahn (22). The VC mice were purchased from Jackson Laboratory (Bar Harbor, ME). To generate mice with IEC-specific IR disruption, IR^fl/fl mice were cross-bred with VC-IR^+/+ mice to generate mice heterozygous for the floxed IR allele (VC-IR^Δ/+). These animals were bred with mice homozygous for the floxed IR allele (IR^fl/fl) to generate mice with homozygous IR disruption (VC-IR^Δ/Δ). Study animals were generated by crossing VC-IR^Δ/Δ and IR^fl/fl mice. Genotyping was performed as described elsewhere (22, 61). All data were collected from co-housed, sex-matched littermate pairs of ≥4-mo-old male or female mice. All animal studies were approved by the Institutional Animal Care and Use Committee at the University of North Carolina at Chapel Hill.

Diet studies.

Four-week-old IR^fl/fl and VC-IR^Δ/Δ sex-matched littermate pairs were fed a diet with 60% of kilocalories from fat (HFD; D12492, Research Diets, New Brunswick, NJ) for 22–26 wk. Control sex-matched littermate pairs were fed standard rodent chow, with 14% of kilocalories from fat (PMI Prolab RMH 3000). Body weight was monitored weekly.

Glucose tolerance tests.

After an overnight (∼16-h) fast, mice were given an oral gavage of glucose (1.5 g/kg body wt; Gibco, Grand Island, NY) in PBS. Glucose in blood taken from the tail was measured using a OneTouch Ultra glucometer (LifeScan, Milpitas, CA) prior to glucose administration and at 15, 30, 60, and 120 min after glucose gavage.

Body fat mass measurements and tissue collection.

Fat and lean body mass were measured by MRI (EchoMRI, Houston, TX). At 90 min prior to euthanasia, mice were given an intraperitoneal injection of 5-ethyl-2′-deoxyuridine (EdU, 100 μg/25 g body wt; Sigma) to mark cells in the S phase. Animals were euthanized with a lethal dose of pentobarbital sodium (Nembutal; 150 μg/g body wt). The small intestine was removed and flushed with ice-cold PBS (0.137 M NaCl, 3 mM KCl, 8 mM Na₂HPO₄, and 2 mM KH₂PO₄, pH 7.4). Mesentery and mesenteric fat were removed. Mass of gonadal fat pads and mesenteric fat surrounding the small intestine were measured. Mass and length of the small intestine were measured. Length was measured using a 3-g clip attached to the end of the tissue to avoid any impact of differences in peristalsis. The most-proximal quarter of the small intestine was designated the duodenum, the middle two quarters the jejunum, and the most-distal quarter the ileum (Fig. 1). The most-proximal 2 cm of each small intestinal segment were fixed as intact tubes overnight in 10% zinc-buffered formalin (Thermo Fisher Scientific, Pittsburgh, PA) at 4°C before paraffin embedding for histology and morphometric measures of growth. The next most-proximal segment from the duodenum, jejunum, and ileum was fixed overnight in fresh 4% paraformaldehyde in 1× PBS at 4°C (Fig. 1) for immunofluorescence. Paraformaldehyde-fixed segments were taken through a gradient of 10% and 30% sucrose sequentially overnight at 4°C and then cryoembedded in optimum cutting temperature medium. Embedded tissues were sectioned (∼5 μm), and sections were placed on positively charged microscope slides.

Fig. 1.

Schematic for intestinal tissue harvest and fixation. Small intestine was divided into 3 segments: duodenum, jejunum, and ileum. Designated regions were isolated and fixed or the epithelium was isolated for RNA, DNA, or protein assays. H&E, hematoxylin and eosin; IEC, intestinal epithelial cell(s).

Morphological measurements, submucosal circumference, and crypt and villus density.

All measurements were taken on paraffin-embedded, hematoxylin-eosin-stained cross sections. Tissue was visualized using Axio Imager.A2 (Zeiss, Thornwood, NY) and a ProgRes CF Scan camera (Jenoptik, Jena, Thuringia, Germany). Measurements were made with ProgRes Capture Pro 2.7 software. Crypt depth and villus height were measured in ≥15 well-oriented crypts and 10 well-oriented villi for each intestinal region and animal. For measurement of crypt density, the number of visible crypts within a jejunal cross section was counted and then divided by the submucosal circumference. For measurement of villus density, the number of villi attached to the surrounding epithelium within a cross section was counted and then divided by the submucosal circumference. All scoring was performed by an observer blinded to animal genotype and diet.

Immunofluorescent staining and quantification of EdU-, Lyz-, and CHGA-positive cells.

Cryopreserved sections were brought to room temperature and washed with 1× PBS to remove the optimal cutting temperature compound. Heat-induced epitope retrieval was performed in a decloaking chamber at 120°C for 30 s followed by 90°C for 10 s. Samples were then incubated with 10% normal goat serum in Tris (Thermo Fisher Scientific, Rockford, IL) for 30 min at room temperature. EdU staining was performed using the Click-iT EdU Alexa Fluor 594 kit according to the manufacturer's instructions (Invitrogen, Carlsbad, CA). For Lyz and CHGA staining, tissue sections were incubated in a primary antibody [rabbit α-lysozyme (1:400 dilution; Leica, Buffalo Grove, IL) or rabbit α-CHGA (1:400 dilution; Abcam, Cambridge, MA)] overnight at 4°C and then with the secondary antibody goat anti-rabbit Cy3 (1:500 dilution; Jackson ImmunoResearch, West Grove, PA). Slides were rinsed in 1× PBS and mounted with medium containing 4′,6-diamidino-2-phenylindole (DAPI; Fluoro-Gel II, Electron Microscopy Services, Hatfield, PA).

EdU-, Lyz-, or CHGA-positive cells were visualized and quantified in the jejunum using an IX83 fluorescence microscope (Olympus, Tokyo, Japan) and an ORCA-Flash4.0 C11440 camera (Hamamatsu, Hamamatsu City, Japan) with Cellsens (Olympus) accompanying software. Individual cells were identified and quantified using DAPI staining. Cells containing clear, nuclear EdU staining, cytoplasmic Lyz granules/staining, or cytoplasmic CHGA staining were considered positive; all others were considered negative. The position of EdU-positive cells in the crypt was recorded to assess if loss of IEC-IR preferentially altered proliferation of cells within the IESC or progenitor zones of the crypt. For the analyses, the two cells at the crypt base were designated cells at position 1. Each adjacent cell between position 1 and the mouth of the crypt was assigned a positional value. At least 15 well-oriented crypts were scored per animal by an observer blinded to animal genotype and diet.

ELISAs.

Blood was collected by cardiac puncture. Plasma was isolated by centrifugation (1,200 g, 10 min, 4°C), snap-frozen in liquid nitrogen, and stored at −80°C. Plasma insulin levels were assessed by ELISA (Mercodia, Uppsala, Sweden) following the manufacturer's instructions. Prior to IGF1 ELISA, IGF-binding proteins were removed from plasma samples by acid-ethanol extraction. Plasma samples were incubated for 30 min at room temperature in acid-ethanol (87.5% ethanol-12.5% 2 M HCl). Samples were centrifuged at 10,000 g for 13 min at 4°C, and supernatant was neutralized with 0.855 mM Tris base. Samples were incubated at −80°C for 30 min. After cryoprecipitation, the samples were spun down at 10,000 g for 15 min at 4°C. Unbound IGF1, present in the supernatant, was quantified using a mouse IGF1 ELISA kit (R & D Systems, Minneapolis, MN) according to the manufacturer's instructions.

Cholesterol assay.

Plasma cholesterol levels were assessed using the Cholesterol E assay kit (Wako Diagnostics, Mountain View, CA) according to the manufacturer's instructions.

IEC isolation.

IEC were isolated from the distal half of the jejunum (Fig. 1) as described elsewhere (82). Isolated cells were used for subsequent DNA, RNA, or protein isolation.

DNA isolation and PCR.

For extraction of DNA, isolated IEC were boiled at 95°C in 50 mM NaOH for 10 min and then neutralized with 1 M Tris·HCl, pH 6.8. Contents were pelleted by centrifugation (13,000 rpm, 6 min, room temperature). Supernatant was used for subsequent PCR. IEC-specific recombination of the IR allele was confirmed at the DNA level by PCR as described elsewhere (78).

Protein isolation and Western blotting.

IEC and liver samples were lysed in Laemmli lysis buffer (200 mM Tris, pH 6.8, 20% glycerol, 5% β-mercaptoethanol, 4% sodium dodecyl sulfate, and 0.03% bromophenol blue). Lysates were boiled, sonicated, and resolved on NuPAGE 4–12% Bis-Tris 1.5-mm gels (Invitrogen, Grand Island, NY) and transferred to a PVDF membrane (0.45-μm pore; Millipore, Billerica, MA). The membrane was blocked in Blocker Casein in PBS (Thermo Fisher Scientific, Pittsburgh, PA). Primary antibodies [mouse anti-α-tubulin monoclonal (1:500 dilution, Sigma) or rabbit anti-β-actin polyclonal (1:2,000 dilution; Sigma) as invariant controls; rabbit anti-IR-β polyclonal (1:500 dilution; sc-711, Santa Cruz Biotechnology, Santa Cruz, CA) or rabbit anti-IGF1R-β polyclonal (1:1,000 dilution; Cell Signaling, Beverly, MA)] were incubated overnight at 4°C.

RNA isolation and quantitative RT-PCR.

RNA was isolated from IEC using the RNeasy Mini Kit with on-column DNase digestion (Qiagen, Valencia, CA) after homogenization with Lysing Matrix D tubes (MP Biomedicals, Santa Ana, CA) containing RLT lysis buffer and 1% β-mercaptoethanol (Gibco) and Precellys24 (Bertin Technologies, Rockville, MD). cDNA synthesis was performed using 0.5 μg of RNA and the High Capacity cDNA Reverse Transcription kit, including RNase inhibitor (Applied Biosystems, Carlsbad, CA). Quantitative RT-PCR (qRT-PCR) was performed using Platinum Quantitative PCR 2× Supermix-UDG (Invitrogen, Grand Island, NY) and TaqMan primer/probe sets (Applied Biosystems). Reactions were run using the Applied Biosystems Step One Plus Real Time PCR system (Applied Biosystems). Gene expression values were calculated using a standard curve generated from a pool of samples, and expression values reflect an average of duplicates. All expression values were normalized to the invariant control gene β-actin (ActB). The primer/probe sets were as follows: ActB (Mm00607939_s1), apolipoprotein A-4 (Apoa4, Mm00431814_m1), apolipoprotein B-48 (Apob-48, Mm01545156_m1), Ascl2 (Mm01268891_g1), Bmi1 (Mm03053308_g1), Chga (Mm00514341_m1), Def1a (Mm 02524428_g1), Gcg (Mm01269055_m1), Gip (Mm00433601_m1), Hopx (Mm00558630_m1), Igf1r (Mm00802831_m1), Lgr5 (Mm00438890_m1), Lrig1 (Mm00456116_m1), Lyz (Mm00727183_s1), Muc2 (Mm00458299_m1), Olfm4 (Mm01320260_m1), peroxisome proliferator-activated receptor-γ (Pparγ, Mm01184322_m1), Scarb1 (Mm00450234_m1), Sis (Mm01210305_m1), Sox9 (Mm00448840_m1), and sterol regulatory element-binding transcription factor 1 (Srebf1, Mm00550338_m1).

Statistical analyses.

Values are means ± SE. Data were analyzed using a two-way ANOVA for main effect of diet, main effect of genotype, or significant diet × genotype interaction. Post hoc comparisons were conducted using Sidak's test, as appropriate. P < 0.05 was considered statistically significant. All analyses were performed using GraphPad Prism 6.

RESULTS

VC-mediated IR disruption is specific to the intestinal epithelium at the DNA and the protein level.

Mice homozygous for loxP sites (floxed) flanking exon 4 of the IR were crossed with animals carrying the transgene for Cre recombinase, located downstream of genomic regulatory elements for villin (VC; Fig. 2A). VC is expressed along the entire crypt-villus axis and throughout the entire length of the intestine and colon (56). Expression begins 14–15 days postcoitum (56). IEC were isolated from the small intestine of adult mice homozygous for the intact, but floxed, allele of the IR (IR^fl/fl) and mice homozygous for the recombined allele of IR due to VC-mediated recombination (VC-IR^Δ/Δ). DNA was isolated from the IEC, and PCR was performed using three primers that amplify the intact, but floxed, or the recombined allele (Fig. 2A). In animals with intact, but floxed, IR, these primers generate a 293-bp amplicon. In animals with recombined IR (lacking exon 4 and genomic DNA lying between the loxP sites), primers will amplify the remaining DNA between introns 3 and 4, yielding a 220-bp amplicon. As shown in Fig. 2B, recombination occurred only in the IEC from VC-IR^Δ/Δ animals. Recombination did not occur in the IR^fl/fl animals or in the underlying serosa in animals of either genotype (data not shown). Evaluation of IR protein by Western blotting on lysates of epithelial cells from the small intestine and colon, as well as the liver, confirmed that recombination resulted in undetectable IR only in the IEC from the small intestine and colon of VC-IR^Δ/Δ mice (Fig. 2C). Loss of IR expression in small intestinal epithelial cells did not result in statistically significant increases in Igf1r mRNA (Fig. 2D) or protein (Fig. 2E).

Fig. 2.

Villin-Cre (VC)-mediated insulin receptor (IR) disruption is specific to the epithelium at DNA and protein levels. A: IR^fl/fl mice are homozygous for a floxed allele of the IR gene, where exon 4 is flanked by loxP sites, as originally described by Bruning et al. (22). Mice homozygous for the recombined allele (VC-IR^Δ/Δ) were generated by crossing IR^fl/fl animals to mice carrying VC recombinase (expressing Cre recombinase only in intestinal and colonic epithelial cells). Insr, insulin receptor gene. B: level of VC-mediated recombination in IR DNA in IEC from IR^fl/fl and VC-IR^Δ/Δ animals assessed by PCR using primers in introns 3 and 4 and exon 4. SI, small intestine. C: Western blot assessment of IR protein levels in small intestinal and colonic IEC and liver from IR^fl/fl and VC-IR^Δ/Δ mice. α-Tubulin was used as loading control. D: quantitative RT-PCR measurement of insulin-like growth factor 1 receptor (Igf1r) mRNA expression in small intestinal IEC from IR^fl/fl and VC-IR^Δ/Δ mice. E: Western blot assessment of IGF1R protein levels in small intestinal IEC from IR^fl/fl and VC-IR^Δ/Δ mice. mRNA and protein expression levels are normalized to β-actin invariant control. Values are means ± SE; n ≥ 3.

Loss of IEC-IR does not affect body weight or adiposity in chow- or HFD-fed animals.

We challenged the IR^fl/fl and VC-IR^Δ/Δ animals with a HFD (60% of kilocalories from fat) to assess if loss of intestinal epithelial IR influenced 1) metabolic effects of HFD, 2) impact of obesity/hyperinsulinemia on intestinal epithelial growth, or 3) IESC or differentiated IEC lineages. It is well established that 14–20 wk of HFD feeding in C57BL/6 mice or other rodent models leads to increased fat mass, obesity, and increases in circulating plasma insulin (15, 28, 57), IGF1 (38, 57, 67, 85), and cholesterol (15) levels.

After chow or HFD feeding, body weight and percent lean and fat mass were assessed by MRI. Both HFD-fed IR^fl/fl and VC-IR^Δ/Δ mice lost significant lean mass and gained significant fat mass relative to chow-fed controls and showed significant increases in mass of gonadal and mesenteric fat (P < 0.05, by 2-way ANOVA with Sidak's post hoc pair-wise comparisons for an effect of diet; Table 1). Body weight, fat and lean body mass, gonadal fat pad mass, and small intestinal mesenteric fat did not differ between IR^fl/fl and VC-IR^Δ/Δ mice (Table 1).

Table 1.

Intestinal epithelial IR disruption does not alter body mass or adiposity in mice fed chow or HFD

	Chow (n ≥ 6)		HFD (n ≥ 11)
	IRfl/fl	VC-IRΔ/Δ	IRfl/fl	VC-IRΔ/Δ
Body wt, g	30.8 ± 1.5	30.7 ± 1.6	41.5 ± 2.4*	42.3 ± 1.8*
Lean mass, %	75 ± 2	75 ± 3	57 ± 3*	56 ± 2*
Fat mass, %	16 ± 2	17 ± 3	39 ± 4*	40 ± 3*
Gonadal fat pad wt, g	0.57 ± 0.1	0.65 ± 0.2	1.9 ± 0.4*	2.0 ± 0.2*
SI mesenteric fat wt, g	0.15 ± 0.02	0.19 ± 0.04	0.63 ± 0.12*	0.58 ± 0.08*

Values are means ± SE.

IR, insulin receptor; HFD, high-fat diet (60% of kilocalories from fat); SI, small intestine; VC, villin-Cre.

^*P < 0.05 (by 2-way ANOVA for main effect of diet).

VC-IR^Δ/Δ and IR^fl/fl mice show identical responses to oral glucose and impact of HFD on plasma insulin or IGF1 levels, but the HFD-induced increase in cholesterol is attenuated in VC-IR^Δ/Δ mice.

Given known roles of IR in glucose uptake and metabolism in other tissues, we examined oral glucose tolerance in IR^fl/fl and VC-IR^Δ/Δ mice in the basal state and after HFD feeding. Glucose tolerance was assessed after ≥23 wk of HFD. HFD feeding significantly increased the area under the curve in both genotypes (P < 0.05, by 2-way ANOVA for main effect of diet), but we found no difference in glucose tolerance between genotypes (Fig. 3, A and B).

Fig. 3.

Loss of intestinal epithelial IR does not impact glucose tolerance, blood insulin, or insulin-like growth factor 1 (IGF1) levels after chow or high-fat diet (HFD) feeding but attenuates HFD-induced increases in circulating cholesterol. A and B: blood glucose levels following intestinal absorption and glucose secretion into the circulation (A) and concentration of blood glucose over time [area under the curve (AUC), B] in IR^fl/fl and VC-IR^Δ/Δ mice measured by oral glucose tolerance test. C–E: fed plasma insulin and IGF1 levels measured by ELISA (C and D) and plasma cholesterol levels assessed by colorimetric cholesterol assay (E) in IR^fl/fl and VC-IR^Δ/Δ mice fed chow or HFD (60% of kilocalories from fat). Values are means ± SE; n ≥ 5. *P ≤ 0.05, HFD vs. chow (by 2-way ANOVA with Sidak's post hoc pair-wise comparisons for main effect of diet).

After 22–26 wk on the HFD, both IR^fl/fl and VC-IR^Δ/Δ mice exhibited significant increases in plasma insulin levels compared with their chow-fed counterparts of the same genotype (Fig. 3C). Plasma IGF1 levels increased significantly following HFD feeding, independent of genotype (Fig. 4D; P < 0. 05, by 2-way ANOVA for main effect of diet). Circulating cholesterol levels increased significantly in the IR^fl/fl, but not VC-IR^Δ/Δ, mice after HFD feeding (Fig. 3E).

Fig. 4.

Intestinal epithelial-specific disruption of the IR does not alter crypt depth or villus height. Representative images of hemotoxylin-eosin-stained sections show crypt-villus (A) and crypt (B) from duodenum, jejunum, and ileum of chow-fed IR^fl/fl and VC-IR^Δ/Δ mice. Magnification ×20 (a) and ×40 (b); scale bars = 50 μm (A) and 20 μm (B).

Loss of IEC-IR does not dramatically alter small intestinal morphology in chow-fed animals.

Hematoxylin-eosin-stained sections of duodenum, jejunum, and ileum from chow-fed IR^fl/fl and VC-IR^Δ/Δ mice show no obvious differences in crypt or villus size or morphology (Fig. 4). Subsequent quantitative comparisons verified that there was no significant effect of IEC-IR loss in chow-fed mice.

Intestinal epithelial-specific IR disruption does not alter morphometric measures of intestinal growth or IEC proliferation in the basal state or with HFD feeding.

Table 2 summarizes data for measures of intestinal growth and morphometry in IR^fl/fl and VC-IR^Δ/Δ littermates fed chow or HFD. HFD feeding significantly decreased small intestine length and increased crypt depth in IR^fl/fl and VC-IR^Δ/Δ mice. There was no difference in intestinal length or mass, jejunal crypt depth, or villus height between genotypes fed the same diet. There was also no difference in duodenal or ileal crypt depth or villus height between chow-fed IR^fl/fl and VC-IR^Δ/Δ mice (data not shown). Jejunal submucosal circumference and crypt and villus density were also similar in IR^fl/fl and VC-IR^Δ/Δ mice fed chow or HFD (Table 2). Although loss of IEC-IR did not significantly affect the majority of the parameters measured, one exception was villus height in the jejunum, which was modestly increased in VC-IR^Δ/Δ animals after HFD but decreased in IR^fl/fl littermates, resulting in a significant interaction between diet and genotype (P < 0.05, by 2-way ANOVA). As shown in Fig. 5A, there was no difference in the total number of cells per crypt across genotypes or diet groups. The total number of EdU-positive cells per crypt was quantified as a measure of proliferation. There was no difference in the average number of EdU-positive cells per crypt between genotypes fed chow or HFD (Fig. 5B). Assessment of the position of EdU-labeled cells revealed no significant differences in the position of proliferating cells in the crypts of IR^fl/fl vs. VC-IR^Δ/Δ mice when diet-matched littermates were compared (Fig. 5, C–E).

Table 2.

Intestinal epithelial IR disruption does not alter intestinal structure in mice fed chow or HFD

	Chow (n ≥ 5)		HFD (n ≥ 4)
	IRfl/fl	VC-IRΔ/Δ	IRfl/fl	VC-IRΔ/Δ
SI length, cm	39.7 ± 1.0	41.0 ± 0.9	37.4 ± 0.6*	36.1 ± 0.6*
SI mass, g	1.5 ± 0.09	1.5 ± 0.08	1.35 ± 0.05	1.36 ± 0.06
Jejunal crypt depth, μm	66.7 ± 1.4	62.1 ± 2.4	75.8 ± 2.1*	75.1 ± 2.5*
Jejunal villus height, μm†	350.7 ± 15.5	321.3 ± 8.3	312.7 ± 12.2	343.8 ± 13.0
Jejunal circumference, mm	7.91 ± 0.35	7.66 ± 0.51	7.83 ± 0.33	7.72 ± 0.32
Crypt density, mm−1	14.8 ± 1.1	16.3 ± 0.6	16.5 ± 1.1	16.8 ± 0.9
Villus density, mm−1	4.4 ± 0.42	5.4 ± 0.36	4.7 ± 0.33	5.2 ± 0.27

Values are means ± SE.

^*P < 0.05 (by 2-way ANOVA for main effect of diet).

^†P < 0.05 (by 2-way ANOVA for diet × genotype interaction).

Fig. 5.

Intestinal epithelial-specific disruption of the IR does not alter cell proliferation in the basal state or during HFD feeding. A and B: cells per crypt assessed by quantification of 4′,6-diamidino-2-phenylindole (DAPI)-stained nuclei (A) and cells in the S phase assessed by quantification of 5-ethyl-2′-deoxyuridine (EdU) staining (B) in IR^fl/fl and VC-IR^Δ/Δ mice fed chow or HFD (60% of kilocalories from fat). C: schematic showing positions of EdU-positive cells within the crypt. D and E: location of proliferating cells in IR^fl/fl and VC-IR^Δ/Δ mice fed chow (D) or HFD (60% of kilocalories from fat; E) assessed by quantification of number of EdU-positive cells at each position along the crypt. Values are means ± SE; n ≥ 5.

Our previous study demonstrated that HFD-induced obesity expanded IESC (57). We therefore assessed mRNAs encoding biomarkers of crypt-based columnar (CBC), actively cycling IESC and quiescent/reserve IESC. HFD resulted in significant increases in Olfm4 and Ascl2, which are enriched in CBC IESC, a similar trend for Lgr5, and increased Sox9 mRNA, which is expressed in both types of IESC. HFD and obesity were also associated with increased expression of Hopx, which marks the quiescent/reserve IESC, but not Bmi1 or Lrig1 (Table 3). Interestingly, the impact of HFD on these biomarkers was attenuated or abolished in mice with IEC-IR deletion, such that HFD induced no significant increase in any IESC biomarker in VC-IR^Δ/Δ mice (Table 3).

Table 3.

IESC biomarker gene expression in HFD-fed mice

		HFD-Fed Mice
Gene	CBC/Quiescent	IRfl/fl	VC-IRΔ/Δ
Lgr5	CBC	1.42 ± 0.15	1.15 ± 0.10
Olfm4	CBC	1.67 ± 0.17*	1.08 ± 0.12◇
Ascl2	CBC	2.92 ± 0.40*	1.77 ± 0.19
Sox9†	CBC/quiescent	1.69 ± 0.11*	1.10 ± 0.08
Bmi1	Quiescent	0.93 ± 0.10	0.76 ± 0.09*
Hopx†	Quiescent	1.38 ± 0.15	0.75 ± 0.06◇
Lrig1	Quiescent	1.02 ± 0.12	0.70 ± 0.06

Values (means ± SE; n ≥ 9) are expressed as fold change vs. chow-fed mice.

CBC, crypt-based columnar intestinal epithelial stem cell (IESC).

^*P < 0.05 (by 2-way ANOVA with Sidak's multiple comparisons for effect of diet).

^†P < 0.05 (by 2-way ANOVA for diet × genotype interaction).

^◇P < 0.05 (by 2-way ANOVA with Sidak's multiple comparisons for effect of genotype).

IEC-IR deletion does not affect Sis or Muc2 expression.

We next examined mRNA expression of markers of differentiated IEC lineages. The mRNA expression of Sis, which marks enterocytes, was similar between diet and genotype groups (Fig. 6A). Expression of Muc2, a goblet cell marker, decreased with HFD feeding but was not affected by IEC-IR loss (Fig. 6B).

Fig. 6.

Loss of IEC-IR has no effect on mRNA expression of enterocyte and goblet cell markers. Sucrase isomaltase (Sis, A) and mucin-2 (Muc2, B) mRNA levels in jejunal epithelial cells isolated from IR^fl/fl and VC-IR^Δ/Δ mice fed chow or HFD (60% kilocalories from fat) were measured using quantitative RT-PCR. Expression levels were normalized to β-actin invariant control. Values are means ± SE; n ≥ 6. *P ≤ 0.05, HFD vs. chow (by 2-way ANOVA with Sidak's post hoc pair-wise comparisons for main effect of diet).

HFD-induced obesity is associated with increased numbers and biomarkers of EEC, and loss of IEC-IR attenuates these effects.

In IR^fl/fl mice, HFD-induced obesity was associated with significant increases in the number of cells positive for CHGA, which marks a majority of EEC (Fig. 7, A and B), and also increased the levels of Chga mRNA (Fig. 7C). Gip and Gcg mRNAs were examined, because levels of these EEC products are altered in obesity (62). Gip mRNA was increased almost threefold (Fig. 7D), while Gcg mRNA was increased almost twofold, in obese IR^fl/fl mice (Fig. 7E). Effects of obesity on CHGA-positive cells and Chga, Gip, and Gcg mRNAs were lost in VC-IR^Δ/Δ mice lacking IEC-IR.

Fig. 7.

IEC-IR loss prevents HFD-induced increases in chromogranin A (CHGA)-positive cells and Chga, glucose-dependent insulinotrophic peptide (Gip), and glucagon (Gcg) mRNAs. A: representative images of CHGA immunostaining (red) and DAPI (blue). Magnification ×20; scale bar = 50 μm. B: quantification of number of CHGA-positive enteroendocrine cells in chow- or HFD-fed IR^fl/fl and VC-IR^Δ/Δ mice from images in A. C–E: quantitative RT-PCR assessment of Chga (C), Gip (D), and Gcg (E) mRNA levels in jejunal epithelial cells isolated from chow- or HFD-fed IR^fl/fl and VC-IR^Δ/Δ mice. Data are normalized to β-actin invariant control. Values are means ± SE; n ≥ 6. *P < 0.05 (by 2-way ANOVA with Sidak's post hoc pair-wise comparisons for main effect of diet). ◇P < 0.05 (by 2-way ANOVA with Sidak's post hoc pair-wise comparisons for main effect of genotype).

HFD-induced increase in Paneth cell antimicrobial peptides is abolished in mice lacking IEC-IR.

Quantification of Lyz-positive Paneth cells revealed similar Paneth cell numbers per crypt in chow-fed IR^fl/fl and VC-IR^Δ/Δ mice (Fig. 8, A and B). HFD reduced the number of Lyz-positive Paneth cells in both genotypes. HFD and obesity were associated with increases in Lyz mRNA in IR^fl/fl, but not VC-IR^Δ/Δ, mice (Fig. 8C). Another Paneth cell-specific mRNA (Def1a) was also increased after HFD, but only in IR^fl/fl mice (Fig. 8D).

Fig. 8.

Loss of intestinal epithelial IR attenuates production of mRNAs encoding Paneth cell antimicrobial peptides. A: representative images of lysozyme (LYZ) immunostaining (red) and DAPI (blue). Magnification ×40; scale bar = 20 μm. B: quantification of number of Paneth (Lyz⁺) cells in chow- or HFD-fed IR^fl/fl and VC-IR^Δ/Δ mice from images in A. Magnification ×40; scale bar = 20 μm. C and D: quantitative RT-PCR assessment of Lyz (C) and Def1a (D) mRNA expression in jejunal epithelial cells isolated from chow- or HFD-fed IR^fl/fl and VC-IR^Δ/Δ mice. Data are normalized to β-actin invariant control. Values are means ± SE; n ≥ 6. *P < 0.05 (by 2-way ANOVA with Sidak's post hoc pair-wise comparisons for main effect of diet). ◇P < 0.05 (by 2-way ANOVA with Sidak's post hoc pair-wise comparisons for main effect of genotype).

IEC-specific IR loss attenuates HFD-induced increase in cholesterol transporter Scarb1 mRNA expression.

The levels of mRNAs encoding a number of proteins involved in lipid handling or lipid metabolism were evaluated in IEC prepared from the jejunum of IR^fl/fl and VC-IR^Δ/Δ mice. These mRNAs were chosen on the basis of previous studies in the fructose-fed hamster model of insulin resistance or in short-term HFD studies that demonstrated that insulin resistance or HFD was associated with increased expression of a number of genes involved in intestinal lipid handling (5, 25, 45, 46). Consistent with these reports, HFD feeding significantly increased levels of mRNAs encoding the transcription factors Pparγ and Srebf1, as well as Apoa4, in both IR^fl/fl and VC-IR^Δ/Δ mice (Fig. 9, A–C). Neither diet nor genotype resulted in significant differences in the levels of mRNAs encoding Apob-48, which is involved in chylomicron synthesis (Fig. 9D). In IR^fl/fl and VC-IR^Δ/Δ mice fed HFD, levels of mRNAs encoding the intestinal cholesterol transporter Scarb1 were significantly increased; however, the increase was significantly attenuated in VC-IR^Δ/Δ mice (Fig. 9E). The same effect was also seen in animals fed HFD over a short-term [0.49 ± 0.2 and 1.93 ± 0.5 in chow- and HFD-fed IR^fl/fl mice, respectively (P = 0.0063, by 2-way ANOVA for main effect of diet; P = 0.02, by 2-way ANOVA for main effect of genotype) and 0.32 ± 0.03 and 0.64 ± 0.1 in chow- and HFD-fed VC-IR^Δ/Δ mice, respectively]. Overall results suggest that effects of IEC-IR loss may be specific to cholesterol handling.

Fig. 9.

IEC-specific IR loss attenuates HFD-induced increase in mRNA encoding the cholesterol transporter Scarb1. Peroxisome proliferator-activated receptor-γ (Pparγ, A), sterol regulatory element-binding transcription factor 1 (Srebf1, B), apolipoprotein A-4 (Apoa4, C), apolipoprotein B-48 (Apob-48, D), and Scarb1 (E) mRNA expression in jejunal epithelial cells isolated from chow- or HFD-fed IR^fl/fl and VC-IR^Δ/Δ mice was determined by quantitative RT-PCR. All gene expression data are normalized to β-actin invariant control. Values are means ± SE; n ≥ 5. *P ≤ 0.05, HFD vs. chow (by 2-way ANOVA with Sidak's multiple-comparisons for effect of diet). ◇P < 0.05 (by 2-way ANOVA with Sidak's post hoc pair-wise comparisons for main effect of genotype).

DISCUSSION

This work provides new evidence that genetic deletion of the IR, specifically in IEC, has a minimal effect on intestinal epithelial growth and morphology, as well as plasma markers of metabolism, in lean, chow-fed animals. In contrast, in obesity due to HFD, loss of IEC-IR attenuates or prevents increases in EEC, mRNAs encoding IESC biomarkers, Paneth cell products, and an intestinal cholesterol transporter and also attenuates elevations in plasma cholesterol.

We initially hypothesized that adult VC-IR^Δ/Δ mice would exhibit altered intestinal growth, resulting from loss of the growth-promoting effects of IR-A (4, 55) or, potentially, enhanced proliferative signaling, via the remaining IGF1R. Our findings of similar intestinal weight, length, morphometry, morphology, and proliferation levels in adult, chow-fed VC-IR^Δ/Δ mice and IR^fl/fl littermates indicate that the IR is dispensable for normal maintenance of the intestinal epithelium in adult animals. The lack of a basal morphological phenotype could be explained if the IGF1R can compensate for the growth-promoting roles of the IR (or IR-A) in the basal state. Although there was no significant change in Igf1r mRNA or protein expression in the basal state, this does not exclude the possibility that activation of the IGF1R is increased to compensate for IR loss.

One report in a colon cancer cell line with predominantly IR-A expression indicated that siRNA-mediated loss of IR enhanced cell viability and IGF1 signaling via the IGF1R (20). This led us to hypothesize that IR deletion might enhance intestinal growth by favoring IGF1R-mediated growth effects in a setting of diet-induced obesity and hyperinsulinemia, where circulating IGF1 is also elevated. However, our studies demonstrate that in vivo loss of IR does not result in enhanced intestinal growth, even after HFD feeding.

Our group previously demonstrated that IESC express high levels of both IR-A and Igf1r mRNA (4, 57) and that HFD expands Sox9-EGFP^Low IESC, which rely on insulin/IGF1 signaling for intrinsic function following HFD feeding (57). Consistent with this study, HFD-fed IR^fl/fl mice exhibited an increase in the IESC biomarkers Olfm4, Ascl2, Sox9, and Hopx that was not seen in VC-IR^Δ/Δ mice after HFD feeding. These data suggest that the IR impairs HFD-induced IESC expansion, at least on the basis of abundance of mRNAs enriched in these cells, and that IGF1R cannot fully compensate for IR loss, in a setting of obesity and elevated circulating insulin levels.

These findings indicate that it will be of interest to assess the effect of IEC-IR loss in a model carrying a stem cell reporter to permit direct analyses of the role of IR in IESC. Consistent with prior findings, HFD-induced obesity did not affect the total number of EdU-labeled proliferating cells (57), and loss of IEC-IR had no effect on the number of EdU-labeled cells.

We recently reported evidence for predominance of IR-B in differentiated IEC lineages vs. proliferating IESC or progenitors (4). On the basis of this finding and the findings of our group and others that HFD impacts Paneth cells, goblet cells, and EECs (40, 57, 69, 73), we examined differentiated IEC lineages in chow- and HFD-fed IR^fl/fl and VC-IR^Δ/Δ mice. In chow-fed animals, there was little evidence that IR-B loss affected differentiated lineages on the basis of similar villus height, similar relative expression of mRNA biomarkers of enterocytes, goblet cells, or EEC, and direct evaluation of Paneth cell and EEC numbers.

Elevations in the number of CHGA-positive EEC and levels of Chga mRNA in obese mice after prolonged HFD feeding are consistent with prior reports of EEC hyperplasia in obesity (8, 40). The significant increases in CHGA-positive cells and Chga mRNA levels in HFD-fed mice with intact IR were not observed in VC-IR^Δ/Δ mice, indicating roles for IR in these effects. EEC produce and secrete a number of hormones involved in regulating metabolism, including the incretins GIP and GLP1, which stimulate insulin synthesis and secretion from the pancreas (76). Several studies demonstrate that GIP levels are elevated during obesity or HFD feeding (8, 40, 62). GLP1 is produced from the intestinal precursor proglucagon and, as well as stimulating insulin synthesis and secretion in islets, is responsible for numerous metabolic functions, including increased glucose uptake by the liver, muscle, and adipocytes, decreased gastric emptying, and increased satiety (58). Insulin-resistant mice display elevated basal levels of GLP1 (54). Increases in Gip and Gcg mRNA levels in IEC from the HFD-fed IR^fl/fl animals in the present study are consistent with these previous observations and suggest that HFD increases synthesis of these hormones. Findings in VC-IR^Δ/Δ mice suggest that IEC-IR contributes to the increase in Gip and Gcg expression during HFD-induced obesity. Investigation of the effects of IEC-IR loss on additional EEC products and hormone secretion is of future interest.

Our group recently demonstrated that HFD-induced obesity and hyperinsulinemia reduce Paneth and goblet cell numbers in wild-type mice (57). Consistent with these observations, HFD was associated with decreases in Paneth cell numbers and Muc2 mRNA expression in the present study. This occurred in both IR^fl/fl and VC-IR^Δ/Δ mice, indicating no role of IR in these effects of HFD. A recent human study demonstrated elevated LYZ and DEFA5 mRNA levels in obese human patients (47), indicating an effect of diet-induced obesity on Paneth cell expression of antimicrobial products. Consistent with these findings, we observed increases in Lyz and Def1a mRNA levels in the HFD-fed IR^fl/fl mice. Our findings that loss of IEC-IR led to reduced Lyz mRNA in chow-fed animals, prevented a HFD-induced increase in Lyz mRNA, and attenuated HFD-induced increases in Def1a mRNA suggest that loss of IR affects Paneth cell expression of antimicrobial peptides. Our current data suggest that IR may directly or indirectly impact effects of obesity on Paneth cell antimicrobial mRNAs. Whether the reduction in antimicrobial mRNA production in the HFD-fed IEC-IR-deficient mice is sufficient to alter intestinal barrier function, the microbial community, or immune mechanisms requires further investigation. Further studies are also needed to elucidate the physiological significance and role of IR in Paneth cells.

HFD-fed IR^fl/fl and VC-IR^Δ/Δ animals accumulated significant amounts of fat relative to chow-fed controls. Despite these gross similarities in fat accumulation, we examined mRNA expression of a number of lipid-handling proteins in IEC, as increasing evidence suggests a potential link between intestinal insulin sensitivity and lipid production by the intestine (26, 32, 34, 43, 45, 53, 83, 84, 88). Consistent with previously published work, we found that levels of mRNAs encoding the transcription factors Pparγ and Srebf1, which control expression of enterocyte lipid-handling genes, as well as mRNA for the chylomicron-associated protein Apoa4, were significantly increased in small intestinal IEC following HFD feeding (46), and this was unaffected by IR loss. An exception was Scarb1 mRNA encoding a cholesterol receptor (44). Consistent with previous work (81), Scarb1 was elevated in IEC of HFD-fed IR^fl/fl controls. Loss of IEC-IR attenuated HFD-induced increases in Scarb1 mRNA, providing evidence for a role of IR in regulating this transporter. It is noteworthy that a prior pilot study revealed elevated Scarb1 mRNA in IR^fl/fl mice fed a HFD with 45% of kilocalories from fat, and this effect was not observed in VC-IR^Δ/Δ animals. Thus the impact of loss of IEC-IR on Scarb1 mRNA is consistent across two independent HFD challenges.

Although we have only examined Scarb1 at the mRNA level, available evidence suggests that decreases in SCARB1 protein could impact lipid sensing and handling within the enterocytes of VC-IR^Δ/Δ mice. In vitro studies have demonstrated that SCARB1 is present at the apical membrane, where it is required for sensing postprandial lipid, as well as lipid transport, and for mobilizing APOB-48 from the plasma membrane to the endoplasmic reticulum for chylomicron generation (10, 68). This is further supported by Hayashi et al. (45), who showed that knockdown of Scarb1 resulted in reduced APOB-48 chylomicron secretion in vitro. Although VC-IR^Δ/Δ animals are not protected from HFD-induced adiposity, they are protected from increases in circulating cholesterol present in HFD-fed IR^fl/fl littermates, suggesting that IEC-IR impacts both Scarb1 expression and enterocyte cholesterol handling.

There is growing evidence that obesity and insulin resistance lead to alterations in intestinal nutrient handling, which have been attributed to intestinal epithelial insulin resistance. This includes increased APOB-48 chylomicron production and secretion, leading to dyslipidemia (34, 43, 45, 48, 83, 84, 88). Direct evidence that intestinal epithelial insulin resistance contributes to these problems is, however, lacking. Our studies provide new evidence that complete loss of IEC-IR expression does not lead to systemic insulin resistance or increased weight gain in animals fed a chow diet, as glucose tolerance tests, body weight, and mRNA expression levels of lipid-handling proteins in chow-fed VC-IR^Δ/Δ mice are identical to those in IR^fl/fl littermates. These studies in mice provide important insights that loss of intestinal epithelial IR signaling is not sufficient to promote or exacerbate weight gain or systemic hyperinsulinemia. We cannot exclude the possibility that IGF1R or other metabolic receptors are able to fulfill some metabolic roles of IR.

Although we originally hypothesized that there would be a more robust intestinal growth or metabolic phenotype in these IEC-specific IR knockout mice, our findings add to a body of evidence for relatively mild phenotypic effects of deletion of IR in nonhepatic tissues. For example, loss of IR in skeletal muscle did not alter circulating glucose or insulin levels but did increase fat mass (22). Ablation of IR in adipose tissue resulted in increased longevity and reduced fat mass but did not dramatically alter glucose homeostasis and actually improved age-associated glucose intolerance (17). Knocking out IR in the central nervous system had broad effects on energy homeostasis and reproductive maturation (21). Thus it is not surprising that a pathway as critical to survival as the insulin system has some redundancy or compensatory mechanisms to guard against phenotypic effects of IR loss on IR ablation in adult animals. Enhanced IGF1R signaling may compensate for IR loss, although this has not been reported for other tissues following IR ablation (17, 21, 22, 42, 52, 59, 60).

Overall, our data demonstrate that loss of IR signaling in IEC does not impact gross or morphometric measures of small intestinal growth, systemic insulin sensitivity, or adiposity in lean animals or animals fed HFD to induce obesity. However, our findings do point to functional roles of IEC-IR in plasma cholesterol, EEC expansion, biomarkers of IESC, Paneth cell antimicrobial peptides, and intestinal cholesterol transporter mRNA expression during HFD-induced obesity.

GRANTS

This work was supported by National Institutes of Health Grants F31-AG-040943 (S. F. A), 5-R01-DK-040247-19 and 5-R01-AG-041198-01A1 (P. K. Lund), and P30-DK-034987 (CGIBD).

AUTHOR CONTRIBUTIONS

S.F.A. designed, conducted, and analyzed the majority of the experiments or oversaw those performing them and wrote the manuscript. M.A.S. conducted the PCR and Western blot experiments to validate the knockout specificity and took some images of immunofluorescence staining. A.T.M. assisted with experiments and performed some quantitative RT-PCRs. J.A.K. analyzed tissue sections for CHGA-positive cells and morphological measures and performed some quantitative RT-PCRs. A.E.B. analyzed tissue sections for proliferation. R.E.B. assisted with all tissue harvests and performed some quantitative RT-PCRs. P.K.L. oversaw all studies and data analysis, contributed to the writing of the manuscript, and provided guidance throughout the study. All authors reviewed and approved the manuscript.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.