Intestinal deletion of leptin signaling alters activity of nutrient transporters and delayed the onset of obesity in mice

Abstract

The importance of B-isoform of leptin receptor (LEPR-B) signaling in the hypothalamus, pancreas, or liver has been well characterized, but in the intestine, a unique site of entry for dietary nutrition into the body, it has been relatively ignored. To address this question, we characterized a mouse model deficient for LEPR-B specifically in intestinal epithelial cells (IECs). ^IECLEPR-B-knockout (KO) and wild-type (WT) mice were generated by Cre-Lox strategy and fed a normal or high-fat diet (HFD). The analyses of the animals involved histology and immunohistochemistry of intestinal mucosa, indirect calorimetric measurements, whole-body composition, and expression and activities of nutrient transporters. ^IECLEPR-B-KO mice exhibited a 2-fold increase in length of jejunal villi and have normal growth on a normal diet but were less susceptible (P<0.01) to HFD-induced obesity. No differences occurred in energy intake and expenditure between ^IECLEPR-B-WT and -KO mice, but ^IECLEPR-B-KO mice fed an HFD showed increased excreted fats (P<0.05). Activities of the Na⁺/glucose cotransporter SGLT-1 and GLUT2 were unaffected in LEPR-B-KO jejunum, while GLUT5-mediated fructose transport and PepT1-mediated peptide transport were substantially reduced (P<0.01). These data demonstrate that intestinal LEPR-B signaling is important for the onset of diet-induced obesity. They suggest that intestinal LEPR-B could be a potential per os target for prevention against obesity.—Tavernier, A., Cavin, J.-B., Le Gall, M., Ducroc, R., Denis, R. G. P., Cluzeaud, F., Guilmeau, S., Sakar, Y., Barbot, L., Kapel, N., Le Beyec, J., Joly, F., Chua, S., Luquet, S., Bado, A. Intestinal deletion of leptin signaling alters activity of nutrient transporters and delayed the onset of obesity in mice.

Leptin is a 16 kDa predominant adipocyte circulating hormone (1) with a large range of biological functions, within the regulation of body weight, neuroendocrine and immune functions, fertility, and angiogenesis (2). The biological actions of leptin on target tissues are mediated by leptin receptors (LEPRs), which belong to the gp130 family of cytokine receptors (3, 4). These LEPRs occur in several variants (LEPR-A to LEPR-F) generated by alternative splicing and proteolytic cleavage. Both the A- and B-isoforms of LEPR (LEPR-A and LEPR-B) can transduce signals through insulin receptor substrate and mitogen-activated protein kinase pathways, whereas only LEPR-B can also activate the JAK/STAT signaling pathway (5). The importance of LEPR-B signaling in the hypothalamus (6–8) and in peripheral tissues (9) like pancreas (10, 11) or liver (12) has been relatively well characterized. However, the importance of LEPR-B in the intestine requires further investigation. One study demonstrated that deletion of almost all peripheral LEPR, including the intestine, did not affect whole-body energy metabolism but increased leptin production (9), consistent with a negative-feedback regulation of leptin production in adipose tissue (13, 14). It has also been reported that deletion of LEPRs in either proopiomelanocortin (POMC)- or agouti-related protein (AgRP)-expressing intestinal cells reduces microsomal triglyceride transfer protein (MTP) and controls lipid absorption across enterocytes (15). Moreover, leptin has been reported to exert a protective role in mucosal resistance to amebic infection and does so primarily via LEPRs expressed on intestinal epithelial cells (IECs; refs. 16, 17).

The gastrointestinal (GI) tract is recognized as a key player in the development of obesity (18, 19) and GI surgery is currently the most long-term effective treatment for morbid obesity-associated pathologies (reviewed in ref. 20). Over the past years, much data have been provided for the implication of leptin acting through its receptors in the control of GI functions and food intake (21, 22). Recently, it was reported that intestinal leptin signaling lowers glucose production in high-fat-diet (HFD)-fed rodents, suggesting a role of LEPRs in the early antidiabetic effect of GI weight-loss surgeries (23). However, whether intestinal LEPRs contribute to peripheral mechanisms during diet-induced obesity remains unknown. To directly assess the importance of leptin signaling in the intestine, we generated a mouse model that is deficient for LEPR-B signaling in IECs. Here, we present evidence for the importance of IEC LEPR-B in peripheral mechanisms of diet-induced obesity and for the possibility that intestinal LEPR-B accessible per os could be a potential target for obesity treatment.

MATERIALS AND METHODS

Generation of IEC LEPR-B-knockout (^IECLEPR-B-KO) mice

All animal use conformed to the U.S. National Institutes of Health Guide for the Care and Use of Laboratory Animals and was approved by the Ministry of Higher Education and Research and the local ethics committee (Comité d'Ethique Paris-Nord). We used previously characterized Lepr^flox/floxmice (N7 on FVB/NJ strain), which have loxP sites flanking exon 17 of the LEPR gene (Lepr) (24). These mice were bred into C57BL/6J background for 8+ generations before subsequent breeding with C57BL/6J Villin-creTg⁺ mice (from laboratory of Dr. Sylvie Robine, Institut Curie, Paris, France; ref. 25). The Lepr^flox/flox mice were crossed with Villin-cre Tg⁺ mice to produce Lepr^flox/+ heterozygous offspring that were mated to each other to generate Lepr^flox/flox Villin-creTg⁺ (^IECLEPR-B-KO) and Lepr^flox/flox Villin-creTg⁻ [^IECLEPR-B–wild-type (WT) control] mice, which were used for all the experiments. For comparison purposes, Lepr^+/+ Villin-creTg⁺ mice, LEPR-deficient db/db (C57BL/KS/J Lepr^db/db) mice, and their control (C57BL/KS/J Lepr^db/+) mice were also used. Only male mice were used in the study.

Genotyping

The mouse genotype was assessed by PCR on genomic DNA extracted from tail tips with Kappa Mouse Genotyping kits (Clinisciences, Nanterre, France) using specific primers (Table 1). To assess the effectiveness of the Villin-Cre-mediated recombination of Lepr-B, PCR was performed on genomic DNA extracts from mucosa scrapings of stomach, intestine, isolated intestinal cells, pancreas, liver, spleen, and brain of 6- to 8-wk-old ^IECLEPR-B-KO or -WT mice.

Table 1.

Sequences used for genotyping and for RT-PCR

Gene	Accession number	Primer names	Primer sequences
Leptin receptor transcript variant 2 (Lepr)	NM_010704.2	106	GTCTGATTTGATAGATGGTCTT
		65	AGAATGAAAAAGTTGTTTTGG
		105	ACAGGCTTGAGAACATGAACAC
Villin Cre transgenic construct		Villin-Cre13-Left Villin-Cre13-Right	CAAGCCTGGCTCGACGGCC CGCGAACATCTTCAGGTTCT
Hypocretin (HPRT)	NM_010410.2	Orexin-Left	TTGGACCACTGCACTGAAGA
		Orexin-Right	CCCAGGGAACCTTTGTAGAAG
Proopiomelanocortin-α (POMC)	NM_001278581.1	POMC-Left	AGTGCCAGGACCTCACCA
		POMC-Right	CAGCGAGAGGTCGAGTTTG
CART prepropeptide (CARTPT)	NM_013732.6	CART-Left	CGAGAAGAAGTACGGCCAAG
		CART-Right	CTGGCCCCTTTCCTCACT
Neuropeptide Y (NPY)	NM_023456.2	NPY-Left	AGAAAACGCCCCCAGAAC
		NPY-Right	AAGTCGGGAGAACAAGTTTCAT
Thyrotropin-releasing hormone (TRH)	NM_009426.3	TRH-Left	AAGACCTCCAGCGTGTGC
		TRH-Right	CCTCCTTCTCCTCCCTTTTG
Apolipoprotein A-IV (Apo-AIV)	NM_007468.2	ApoA4-Left	TGTCCTGGAAGAGGGTACTGA
		ApoA4-Right	ACCCAGCTAAGCAACAATGC
Microsomal triglyceride transfer protein (MTP)	NM_001163457.1	MTP-55-Left	TGCTGGCCAACACGTCTA
		MTP-55-Right	GCCCAACGTACTTCTAATTTATGG
Solute carrier family 5 member 1 Slc5a1 (SGLT-1)	NM_019810.4	SGLT1-Left	GACATCCCAGAGGACTCCAA
		SGLT1-Right	ACCACTGTCCTCCACAAAGG
Solute carrier family 2 member 2 Slc2a2 (GLUT2)	NM_031197.2	GLUT2-Left	CTGCTCTTCTGTCCAGAAAGC
		GLUT2-Right	TGGTGACATCCTCAGTTCCTC
Hypoxanthine guanine phosphoribosyl transferase	NM_013556.2	HRPT1-Left	TGGATACAGGCCAGACTTTGTT
(Hprt)		HRPT1-Right	CAGATTCAACTTGCGCTCATC
Glyceraldehyde-3-phosphate dehydrogenase (GAPDH)	NM_008084.2	GAPDH-Left	GTGTCCGTCGTGGATCTG
		GAPDH-Right	CCTGCTTCACCACCTTCTTG
TATA box binding protein (TBP)	NM_013684.3	TBP-Left	CAAACCCAGAATTGTTCTCCT
		TBP-Right	ATGTGGTCTTCCTGAATCCCT

Morphometric and immunohistochemical studies

Intestinal segments were fixed overnight in 10% neutral buffered formalin, paraffin embedded, and sectioned at 5 μm. After dewaxing and hydration, the sections were stained with hematoxylin and eosin, and mucus staining was performed with periodic acid-Schiff and Alcian blue. An observer blinded to origin of the sections assessed the morphology. For each sample slide, microscopic measurements of villus height were recorded from well-oriented villus/crypt units using an optical microscopic (Leica DMD108; Leica Microsystems, Wetzlar, Germany). For immunohistochemistry, the sections were pretreated with 3% H₂O₂ for 10 min at room temperature and incubated with rat monoclonal anti-Ki67 (Dako, Les Ulis, France). Horseradish peroxidase (HRP)-conjugated secondary antibodies were detected with the diaminobenzidine peroxidase substrate kit (Vector Laboratories, Burlingame, CA, USA).

Diet studies

Mice housed individually in cages (12-h light-dark, 21°C) with free access to water were fed ad libitum a normal diet (ND) or HFD (Altromin 1320 or C1090-45; Genestil, Royaucourt, France) for 13 wk. The ND provides 2820 kcal/kg of food and contains 3% fat, 48% complex carbohydrates (primarily starch), 16% protein, and 6% fiber. The HFD provides 4630 kcal/kg and contains 22% fat (primarily lard), 43% carbohydrates, 21% protein, and 3% fiber. Food consumption and body weight were measured. Whole-body composition was measured in unanesthetized mice using an EchoMRI 100 whole-body composition analyzer (EchoMRI, Houston, TX, USA).

Indirect calorimetric measurements

^IECLEPR-B-WT and -KO mice were fed an ND or HFD for 3 wk, and indirect calorimetric measurements of whole energy expenditure, oxygen consumption and carbon dioxide production, respiratory exchange rate, food intake, and locomotor activity were performed using calorimetric cages with food and water (Labmaster; TSE Systems, Bad Homburg, Germany), as described previously (26). The stools were collected for determination of excreted fats and total energy by the method of Van de Kamer and bomb calorimetry (PARR 1351 bomb calorimeter; Parr Instrument Co., Moline, IL, USA) as described previously (27, 28).

Analyses were performed using O₂ consumed, CO₂ production (ml/h), and energy expenditure (kcal/h). Subsequently, each value was expressed either by total body weight or whole lean tissue mass extracted from the EchoMRI analysis. Displayed values are expressed as means ± sem. The test of homogeneity of variance (F test) was performed prior to analysis. Comparisons between groups were carried out using the nonparametric Mann-Whitney test. When appropriate, results were analyzed by the general linear model followed by Tukey's test, with regimen, genotype, and their interactions as factors.

Plasma analysis

Collected blood was used for the determination of glucose, triglyceride, cholesterol, alanine aminotransferase (ALAT), and aspartate aminotransferase (ASAT) activities using an automatic analyzer (AU400; Olympus Diagnostics, Rungis, France). Leptin and insulin levels were determined with the radioimmunoassay kits from Linco Research (Labodia, France).

Western blot analysis

Intestinal mucosa extracts were prepared and used for Western blot analysis as described previously (29) using rabbit polyclonal antibodies anti-LEPR (OBR 12-A; Biotrend Chemikalien, Cologne, Germany), anti-sodium-glucose cotransporter 1 (anti-SGLT-1; Life Technologies, Carlsbad, CA, USA), anti-GLUT2, anti-GLUT5 (Santa Cruz Biotechnologies, Santa Cruz, CA, USA), and anti-PepT-1 (provided by N.K.). Equal protein loading was verified by reprobing membranes for β-actin. Peroxidase-conjugated secondary antibodies were used, and immune complexes were detected by enhanced chemiluminescence (ECL; Perkin Elmer Life Sciences, Boston, MA, USA).

Studies in Ussing chamber

The proximal jejunum (5 cm distal to the ligament of Treitz) was dissected out and rinsed to removed luminal content in cold Krebs-Ringer bicarbonate (KRB) solution as described previously (30). The mesenteric border was carefully stripped off, and the intestine was opened along this border and placed between the halves of an Ussing chamber with circular aperture of 0.2 cm² (EM-CYS-2; Physiological Instruments, San Diego, CA, USA). Electrogenic ion transport was monitored continuously as the short-circuit current (I_sc) by using an automated voltage clamp apparatus (DVC 1000; WPI, Aston, UK). At steady state, 10 nM leptin was introduced or not in mucosal bath, and after 3 min, tissues were challenged mucosally by 10 mM glucose as described previously (30). Glucose-induced I_sc, which reflects SGLT-1 activity, was recorded for the following 15 min. Results were expressed as the difference (ΔI_sc μA/cm²) between the peak I_sc (measured after 3 min) and the basal I_sc (measured before addition of glucose).

Measurement of GLUT2, GLUT5, and PepT1 activities

Jejunum segments from ^IECLEPR-B-WT and -KO mice were studied. We used d-[1-¹⁴C]glucose (specific activity 49.5 mCi/mmol), [¹⁴C]fructose (specific activity 267 mCi/mmol; Amersham, Little Chalfont, UK), and [³H]Gly-Sar (specific activity 57 mCi/mmol; Isobio, Fleurus, Belgium) as substrates for GLUT2, GLUT5, and PepT1, respectively. Briefly, a segment of jejunum was filled with a solution (100 μl/cm) of 30 mM glucose or 30 mM fructose or 20 μM Gly-Sar (Sigma-Aldrich, Saint Quentin, France), containing respective radiolabeled substrates and 500 μg/ml phenol red, to assess paracellular permeability. The segment was placed in a 37°C bath of KRB buffer and bubbled with 95% O₂-5% CO_2. Mucosal to serosal (M/S) apparent permeability to glucose, fructose, or Gly-Sar was estimated as follows: P_app = (dQ/dt)·(V/Q₀·A), where V is the volume of the bath, A is the area of the loop, Q₀ is the total amount of radiolabeled substrate introduced into the loop, and dQ/dt is the flux across the intestinal loop.

RNA isolation and real-time PCR

Total RNA was extracted from frozen intestinal mucosa scrapings and the hypothalamus with Trizol reagent, and first-strand cDNA was synthesized by reverse transcription with SuperScript II reverse transcriptase (Invitrogen, Carlsbad, CA, USA). cDNA was quantified as previously reported (29) with the Light Cycler System (Roche Diagnostics, Indianapolis, IN, USA) according to the manufacturer's instructions and using specific primers (Table 1) synthesized by Eurogentec (Liege, Belgium).

Statistical analysis

The results are expressed as means ± sem. One-way ANOVA followed by Tukey-Kramer multiple comparison posttests was performed using Prism (GraphPad Software, Inc., San Diego, CA, USA). Values of P < 0.05 were considered statistically significant.

RESULTS

Genotype characterization of ^IECLEPR-B-KO mice

The intestine-specific nature of Cre-mediated recombination in LEPR-B-KO mice was characterized by performing PCR on genomic DNA using primers flanking the loxP sites of the Lepr^flox allele (Fig. 1A). With the use of primers (106 forward and 105 reverse), the deleted allele Lepr^Δ17 was detected only in duodenum, jejunum, ileum, and colon mucosa but not in liver, pancreas, stomach, or brain from ^IECLEPR-B-KO mice (Fig. 1B). The deleted Lepr^Δ17 allele was also detected in isolated intestinal cells from LEPR-B-KO but not -WT mice (Fig. 1C). The inactivation was very efficient, since an undetectable Lepr-B mRNA or a drastic reduced amplicon was seen in the jejunum mucosa of ^IECLEPR-B-KO (Fig. 1D). Accordingly, LEPR-B protein was reduced or undetectable in mucosa extracts of jejunum LEPR-B-KO mice (Fig. 1E). These data indicate that ^IECLEPR-B-KO mice have an intestine-specific deletion of LEPR-B.

Figure 1.

Genotype characterization of ^IECLEPR-B-KO mice. A) Genomic DNA was extracted from tissues of ^IECLEPR-B-KO and -WT mice and was PCR amplified using primers flanking the loxP sites to ascertain the Cre-mediated recombination in intestinal and nonintestinal tissues from ^IECLEPR-B-KO mice. B) With the use of primers (106 forward and 105 reverse), a 238 bp product corresponding to the deleted Lepr^flox allele, Lepr^Δ17, was detected in intestinal tissues but not in liver, pancreas, stomach, and brain. C) Lepr^Δ17 was also detected in intestinal isolated cells from ^IECLEPR-B-KO but not from WT mice. D) RT-PCR analysis to detect Lepr-B mRNA in the jejunum of ^IECLEPR-B-WT and –KO mice. E) Western blot analysis of LEPR-B protein in jejunum mucosa of ^IECLEPR-B-KO and -WT mice.

At 8–10 wk of age, ^IECLEPR-B-KO and -WT control mice had similar body weight and fat mass by contrast to Lepr^db/db mice (complete loss of LEPRs), which exhibited a 2-fold increase (P<0.05) in fat mass (Table 2). In addition, ^IECLEPR-B-KO mice had similar levels of blood glucose, HDL, and total cholesterol. Moreover, they had similar circulating levels of ALAT and ASAT, suggesting normal hepatic function. Remarkably, they exhibited a significant reduction of circulating levels of triglycerides (30%, P<0.05 vs. WT; Table 2).

Table 2.

Biochemical parameters of male 8- to 10-wk-old leptin-receptor-deficient db/db mice, db/+, and ^IECLEPR-B-WT control mice and ^IECLEPR-B-KO mice

Parameter	Leprdb/+	Lepr db/db	IECLEPR-B WT	IECLEPR-B KO
Body weight (g)	21.6 ± 0.15	42.9 ± 0.86*	23.5 ± 1.43	22.8 ± 1.3
Fat mass (%)	13.0 ± 0.18	29.4 ± 0.48*	11.2 ± 0.60	11.9 ± 0.40
Small intestine (cm)	39.4 ± 1.60	50.8 ± 0.55**	39.3 ± 1.2	40.3 ± 1.35
Large intestine (cm)	8.22 ± 0.50	10.2 ± 0.43***	6.60 ± 0.20	7.04 ± 0.62
Fasting glycemia (mM)	6.51 ± 0.41	24.3 ± 1.52*	7.54 ± 0.59	7.25 ± 0.88
Triglyceridemia (mM)	0.75 ± 0.06	1.45 ± 0.27***	1.44 ± 0.47	0.94 ± 0.19#
HDL-cholesterol (mM)	1.02 ± 0.28	1.05 ± 0.27	1.18 ± 0.20	0.99 ± 0.34
Total cholesterol (mM)	ND	ND	1.23 ± 0.21	1.13 ± 0.22
ALAT (UI)	62.5 ± 20.3	230.4 ± 26.2**	96 ± 37.1	83.2 ± 27.8
ASAT (UI)	364 ± 107.3	304 ± 93.6	260 ± 40.5	315 ± 19.9

Results are means ± se of n ≥ 6/group. Data were analyzed by 1-way ANOVA followed by a multiple comparison Tukey-Kramer test. ND, not determined.

^*P < 0.05,

^**P < 0.01,

^***P < 0.001 vs. db/+;

^#P < 0.05 vs. WT.

Deletion of intestinal LEPR-B affects morphology of mucosa

^IECLEPR-B-WT and KO mice had similar lengths of small and large intestines, while Lepr^db/db mice exhibited a greater (+35%, P<0.05 vs. Lepr^db/+) length of small intestine (Table 2). Histological analysis of the intestine (Fig. 2A) showed no modification in the architecture of the epithelium. Morphometric analysis (Fig. 2A, B) revealed an increased height of the mucosa throughout the small and large intestine. In the ^IECLEPR-B-KO jejunum, the length of villi was 2-fold (P<0.001) longer than that of ^IECLEPR-B-WT jejunum. This increased mucosal height in the small intestine occurred to a lesser extent in the large intestine, whereas no significant change could be observed in fundus or antrum of the stomach of ^IECLEPR-B-KO mice (Fig. 2A, B).

Figure 2.

Morphology of GI mucosa of ^IECLEPR-B-KO mice. A) Representative photomicrographs of fundic (a), antral (b), and colon mucosa (d) stained for mucus with periodic acid-Schiff and Alcian blue and jejunum mucosa (c) stained with hematoxylin and eosin. B) Morphometric analysis of the height of the mucosa in fundus and antrum, of the length of villi in the small intestine (duodenum, jejunum, and ileum) of ^IECLEPR-B-WT and -KO mice (note the marked increased height of the jejunal villi of KO mice). C) Immunostaining of Ki-67 proliferating cells in formalin-fixed jejunum sections from ^IECLEPR-B-WT and -KO and Lepr^+/+ Villin-CreTg⁺ mice. Scale bars = 100 μm.

In the jejunum, further examination of Ki67-positive cells (Fig. 2C) showed no significant changes in cell proliferation between ^IECLEPR-B-KO and -WT (248.2 ± 13.1 vs. 274.6 ± 27.9 Ki67-positive cells/10 crypts/mouse). Thus, the transient amplifying compartment that normally occupies the entire crypt in the small intestine appears not significantly different between ^IECLEPR-B-KO and -WT mice. Taken together, these data strongly suggest a major role of intestinal LEPR-B signaling in the control of intestinal homeostasis.

^IECLEPR-B-KO mice have normal weight on an ND but are less susceptible to diet-induced obesity

^IECLEPR-B-KO and -WT mice fed for 13 wk on an ND have no significant difference in weight gain (Fig. 3A) and fat mass (not shown), and there were no significant differences in daily calorie intake between the 2 groups (Fig. 3B). We next analyzed the temporal changes in body weights of ^IECLEPR-B-KO and -WT mice fed an HFD for 13 wk.

Figure 3.

Loss of intestinal LEPR-B signaling delays the onset of HFD-induced obesity. A, B) Body-weight curves (A) and mean daily food intake (B) of male ^IECLEPR-B-KO and -WT mice fed ND or HFD. C–F) Circulating levels of glucose (C), triglycerides (D), HDL cholesterol (E), and total-cholesterol (F) in ^IECLEPR-B-KO and -WT mice fed ND or HFD for 13 wk. Results are presented as box and whiskers (min to max) of n = 6–8 mice/group. G, H) Variations of plasma leptin (G) and insulin (H) levels in ^IECLEPR-B-KO and -WT mice after 13 wk of ND or HFD. Results are means ± sem for 6–8 mice/group. Statistical analysis used 1-way ANOVA followed by Tukey-Kramer multiple comparison posttests.

As shown, ^IECLEPR-B-KO mice fed an HFD gained less (P<0.01 vs. WT) weight in comparison to WT mice (Fig. 3A) despite similar consumption of dietary calories (Fig. 3B). In addition, they exhibited a reduced fat mass that remained similar to that of WT mice fed an ND (data not shown). The lesser susceptibility to weight gain of ^IECLEPR-B-KO mice fed an HFD was associated with a reduction in HFD-induced hyperglycemia (Fig. 3C), hypertriglyceridemia (Fig. 3D) and increased HDL and total cholesterol (Fig. 3E, F), with levels remaining similar to those of ^IECLEPR-B-KO mice fed an ND. They have also reduced circulating leptin levels, reflective of adipocyte production (Fig. 3G) accordingly with the reduced fat mass (P<0.05 vs. WT) of ^IECLEPR-B-KO mice and with the histological analysis of mesenteric adipose tissue showing smaller adipocytes (data not shown) in these mice.

Plasma fasting insulin levels (Fig. 3H) rose by 2-fold (P<0.01) in both ^IECLEPR-B-KO and -WT mice fed an HFD in comparison to corresponding mice fed an ND. Note that ^IECLEPR-B-KO mice fed an ND have a trend toward increase in fasting insulin levels vs. WT mice, but this was not significant. Collectively, these results indicate that deletion of IEC leptin signaling delays the onset of HFD-induced obesity.

To determine whether the loss of intestinal LEPR-B signaling would modify the crosstalk between the intestine and brain, we analyzed the expression of some hypothalamic neuropeptides controlling food intake (Supplemental Fig. S1). We found no significant change in the levels of neuropeptide Y (NPY), orexin, POMC, and thyrotropin-releasing hormone (TRH) mRNA in the hypothalamus of ^IECLEPR-B-KO vs. -WT mice fed an HFD (Supplemental Fig. S1).

Because the lesser susceptibility of ^IECLEPR-B-KO mice to obesity could reflect increased energy expenditure, we undertook extensive analysis of energy intake and expenditure using indirect calorimetry in mice fed an ND and HFD (Supplemental Fig. S2 and Fig. 4). After 3 wk, the body weights were similar in KO and WT mice fed an ND or HFD (Fig. 4B). On an HFD, ^IECLEPR-B-KO mice displayed a significantly reduced (−30%, P<0.05 vs. WT) fat mass (Fig. 4C) without any change in lean mass (Fig. 4D). Both groups fed an HFD had similar cumulative calorie intakes (Fig. 4E) and displayed no significant difference in their energy expenditure (Fig. 4F) or spontaneous locomotor activity (Fig. 4G). The respiratory exchange rate (RER; Vco₂/Vo₂) did not change in ^IECLEPR-B-KO and -WT in mice (Fig. 4H). Finally, bomb calorimetry analysis of the collected stools revealed that ^IECLEPR-B-KO mice had increased fecal calorie output (28%, P<0.05 vs. WT), mainly fats (Fig. 4I). There was no significant difference in jejunum APO-A4 (Fig. 4J) and MTP mRNA levels (Supplemental Fig. S3) in ^IECLEPR-B-KO mice as compared with WT mice.

Figure 4.

Body composition, food intake, and energy expenditure in mice fed an HFD for 3 wk. A) Schematic representation of the calorimetric analyses. B–D) Means of body weight (B), relative fat content (g/body weight; C), relative lean body mass (g/body weight; D) of male ^IECLEPR-B-KO and -WT mice after 3 wk exposure to ND or HFD. E–H) Cumulative food intake (kcal; E), whole energy expenditure (F), spontaneous locomotor activity (beam break; G), and respiratory exchange ratio (Vco₂/Vo₂; H) of ^IECLEPR-B-KO and -WT mice during 4 d. Values are expressed as means ± sem. I) Bomb calorimetry analyses of fecal excreted fats in ^IECLEPR-B-KO vs. -WT mice. J) Apolipoprotein-A4 mRNA levels in ^IECLEPR-B-WT and -KO mice fed ND or HFD. Each point represents 1 mouse; bar represents median of the points of each group.

Overall, our data support the contention that at least a reduction in intestinal fat absorption in ^IECLEPR-B-KO mice rather than a change in food intake or energy expenditure was instrumental in the lesser susceptibility to the obesogenic action of an HFD observed in ^IECLEPR-B-KO mice.

Intestinal LEPR-B signaling regulates activities of nutrient transporters

The increase in absorptive jejunal area in ^IECLEPR-B-KO mice contrasts with their normal body weight on an ND and reduced body weight on an HFD. Because ^IECLEPR-B-KO mice displayed no change in caloric intake and energy expenditure, we hypothesized that their lesser susceptibility to the obesogenic action of an HFD was related to alterations of some IEC nutrient transporter expressions and functions. Thus, we studied the expression and activity of intestinal sugar and peptide transporters.

In ^IECLEPR-B-KO jejunum, the abundance of SGLT-1 protein was significantly reduced (−45%, P<0.05 vs. WT; Fig. 5A) whereas mRNA levels were not affected (Supplemental Fig. S3), and this was accompanied by a slight but nonsignificant decrease of glucose-induced I_sc (Fig. 5A). Addition of leptin at the mucosal side of WT jejunum segment in the Ussing chamber reduced by 30% (P<0.05 vs. saline) glucose-induced I_sc, consistent with our earlier data (30). As expected, in ^IECLEPR-B-KO jejunum, leptin's inhibitory effect on SGLT-1 no longer occurred. On the other hand, the amounts of GLUT2 mRNA and protein were unchanged (Supplemental Fig. S3 and Fig. 5B), and the passive component of glucose transport mediated by GLUT2 was not affected in LEPR-B-KO jejunum (Fig. 5B).

Figure 5.

Intestinal LEPR-B leptin signaling is essential for the activity of glucose transporters. A) Transport activity and expression of SGLT-1. Jejunal segments from ^IECLEPR-B-WT and -KO mice were mounted in Ussing chambers and challenged with glucose (10 mM). A) Left panel: increase in glucose-induced I_sc is an index of electrogenic Na⁺-coupled glucose SGLT-1 transport activity. Columns represent means ± sem of 6–8 different experiments from different mice. A) Right panel: representative immunoblot of SGLT-1 from WT and KO jejunum extracts and densitometric analyses normalized to β-actin. Note that SGLT-1 abundance was significantly reduced, and leptin inhibitory no longer occurred. B) Transport activity and expression of GLUT2. Left panel: time course of mucosal to serosal glucose apparent diffusion coefficient across the jejunum of ^IECLEPR-B-WT control and ^IECLEPR-B-KO mice. Columns represent means ± sem of n = 6/group. Right panel: representative immunoblot of GLUT2 from WT and KO jejunum extracts and densitometric analysis normalized to β-actin. Note that GLUT2 transport activity and abundance in the jejunum did not change.

As shown in Fig. 6A, the abundance of fructose transporter GLUT5 proteins was reduced by 2-fold (P<0.05 vs. WT) in LEPR-B-KO jejunum, and, accordingly, GLUT5-mediated fructose transport was significantly decreased (Fig. 6A). We also found a significant reduction of PepT1 transport activity in the LEPR-B-KO jejunum in comparison to WT (Fig. 6B), and changes in PepT-1 protein levels trend toward reduction, but without reaching statistical significance.

Figure 6.

Loss of intestinal LEPR-B leptin signaling reduces GLUT5 and Pept-1 transport activities A) Left panel: time course of mucosal to serosal fructose apparent diffusion coefficient across the jejunum of ^IECLEPR-B-WT and -KO mice. Points are means ± sem of n = 6/group. A 2-way ANOVA was used for statistical comparison. Right panel: representative immunoblot of GLUT5 from WT and KO jejunum extracts and densitometric analysis normalized to β-actin. Note that GLUT5 transport activity and abundance are reduced 2-fold in the KO jejunum in comparison to WT jejunum. B) Left panel: time course of mucosal to serosal Gly-Sar apparent diffusion coefficient across the jejunum of ^IECLEPR-B-WT and -KO mice. Points are means ± sem of n = 6/group. A 2-way ANOVA was used for statistical comparison. Right panel: PepT-1 immunoblot densitometric analysis normalized to β-actin Note the significant reduced PepT-1 transport activity in the KO jejunum compared with WT jejunum.

DISCUSSION

Here, by analyzing the phenotype of ^IECLEPR-B-KO mice, we provide evidence for a crucial role of intestinal LEPR-B signaling in intestinal homeostasis, nutrient absorption, and the onset of obesity. Indeed, ^IECLEPR-B-KO mice exhibited an increased height of the small and large intestinal mucosa with a strongly enhanced absorptive area in the jejunum mainly due to increased length of the villi. Surprisingly, and despite this enhanced absorptive area, ^IECLEPR-B-KO mice have normal body weight on an ND and are less susceptible to HFD-induced obesity, consistent with their reduced fat mass. Overall, these data reveal new insights into the contribution of intestinal leptin signaling in peripheral mechanisms of diet-induced obesity.

The phenotype of ^IECLEPR-B-KO mice is quite distinct from other animal models of leptin signaling deficiency. Most of the studies have shown that deletion of LEPRs in the central nervous system (CNS) recapitulates the effects of complete loss of functional leptin (ob/ob) or leptin-receptor (db/db) leading to hyperphagia and expansion of fat mass, i.e., obesity (6–8, 31, 32). On the other hand, specific disruption of leptin signaling at the periphery (in liver, adipose tissue, and intestine but not in the CNS) has no effect on body weight but results in hyperleptinemia, secondary to inactivation of negative-feedback regulation in the adipose tissue (9). The phenotype of ^IECLEPR-B-KO mice also differs from the phenotype of mice with deletion of LEPR-B signaling in β-pancreatic cells that become obese and glucose intolerant (10). However, the phenotype of ^IECLEPR-B-KO mice is close to that of mice with specific invalidation of LEPR-B signaling in the liver. Indeed, unlike a complete deficiency of leptin action, the loss of hepatic leptin signaling did not alter body weight and body composition but increases hepatic insulin sensitivity and protects against age- and diet-related glucose intolerance (12).

The inability of ^IECLEPR-B-KO mice to increase their body weight on an HFD over that of their WT littermates could be related to a possible increase of energy expenditure and/or impairment of intestinal absorption of nutrients. Whole-animal indirect calorimetry analyses of energy intake and expenditure under both ND and HFD did not reveal any differences between ^IECLEPR-B-WT and –KO mice. To assess whether loss of intestinal LEPR-B signaling alters nutrients absorption, total energy intake and concurrent energy content in the stool was measured by bomb calorimetry. We found that ^IECLEPR-B-KO and -WT mice fed an HFD had similar daily calorie intakes and daily fecal output. Remarkably, however, ^IECLEPR-B-KO mice displayed an enhanced excretion of fats, arguing for a reduction of intestinal lipid absorption. This finding apparently contrasts with the previous demonstration that leptin can reduce APO-A4 mRNA, leading to reduced plasma triglyceride levels (33). However, it is consistent with data showing that LEPR-B deficiency in intestinal POMC-expressing cells leads to decreased expression of intestinal MTP, a key enzyme for the biosynthesis of lipoproteins (15). This reduced intestinal MTP was further shown to be associated with reduced absorption of dietary triglycerides. Therefore, the increased excretion of fats (i.e., reduction of lipid absorption) could represent physiological mechanisms required to attenuate hyperlipidemia and obesity in ^IECLEPR-B-KO mice, mechanisms in line with reduced circulating levels of triglycerides. These observations support the hypothesis that in ^IECLEPR-B-KO mice, a reduction in intestinal fat absorption rather than a change in food intake or energy expenditure was instrumental in their lesser susceptibility to the obesogenic action of an HFD. This prompts us to question whether other nutrient absorption known to be regulated by leptin is affected in the LEPR-B signaling-deficient intestine. In this line, we focus on the SGLT-1 and the GLUT2 transporters, which mediate active and passive components of intestinal glucose transport, respectively (34). In fasted state, glucose and galactose transport from the intestinal lumen to the enterocytes is an active process mediated by SGLT-1 located at the apical membrane of IECs (35). During feeding, when higher concentrations of glucose or galactose are found in the intestinal lumen, activity of SGLT-1 decreases, and GLUT2 becomes active at the apical membrane, providing the small intestine with an absorptive capacity to match dietary intake during the meal. We previously demonstrated that in the intestine, leptin inhibits SGLT-1 activity (30) and enhances GLUT2 activity (29), inducing a reduction in SGLT-1:GLUT2 activity ratio (favoring GLUT2 over SGLT-1), similarly to what occurs under physiological conditions.

We show here that loss of LEPR-B signaling in intestinal mucosa leads to a modest down-regulation of SGLT-1 without affecting the expression of GLUT2 transporter. In addition, LEPR-B-KO jejunum SGLT-1 transport activity, as well as GLUT2 activity, did not change at the basal state (i.e., in absence of leptin). However, and as expected, the previously reported inhibitory effect of leptin on SGLT-1-mediated active glucose transport (30) was no longer observed in LEPR-B-KO jejunum. The reduced SGLT-1:GLUT2 activity ratio was no longer observed, arguing against an increase in glucose absorption in ^IECLEPR-B-KO mice. Thus, it appears from these data that LEPR-B-specific signaling is required for leptin regulation of expression/activity of SGLT-1 and GLUT2.

In addition, in LEPR-B-KO jejunum, the fructose transporter GLUT5 is down-regulated, and its transport activity is reduced, indicating differences in GLUT5 vs. SGLT-1:GLUT2 regulation by intestinal LEPR-B signaling. At this stage, the overall data suggest that the lesser susceptibility of ^IECLEPR-B-KO mice to the obesogenic action of an HFD may be linked to the reduced intestinal fat and sugar (especially fructose) absorption.

The ND and HFD regimen offered to ^IECLEPR-B-KO and -WT mice in this study contains proteins that are degraded by enzymes originating from the stomach, pancreas, and small intestine, resulting in a mixture of free amino acids and small peptides, mainly di- and tripeptides that are efficiently absorbed by enterocytes through apical membrane PepT-1 protein that cotransports di- and tripeptides with protons. Thus, the next question we addressed was how this PepT-1 transporter is affected in the LEPR-B-deficient jejunum. We found that deletion of intestinal LEPR-B signaling is associated with a reduced absorption of peptides mediated by PepT-1 transporter, a finding in agreement with data showing that leptin increases PepT-1 activity, leading to enhanced peptide transport across the intestinal epithelial barrier (36).

Altogether, our data indicate that LEPR-B-KO mice exhibit a lesser susceptibility to diet-induced obesity, which could be explained by reduced intestinal nutrient absorption (fats, sugar, and proteins). These alterations of nutrient absorption are consistent with the reported key role of leptin in intestinal absorption of nutrients (29, 30, 33, 36, 37). Thus, the longer villi observed in LEPR-B-KO jejunum, which should have constituted adaptive mechanisms to make efficient intestinal absorption, are counterbalanced by reduced function of several nutrient transporters. This intestinal adaptation after loss of LEPR-B signaling appears sufficient in response to normal chow but not in HFD conditions.

In summary, leptin secreted mainly by adipose tissue (1), but also by the stomach (21, 38), acts in the brain and on peripheral tissues to regulate several metabolic actions including the intestinal absorption of nutrients delivered from ingested food. In this report, we provide evidence that loss of LEPR-B signaling in IEC leads to defects in the regulation of nutrient transporters, which could contribute to a resistance to diet-induced obesity.