Mice lacking the intestinal peptide transporter display reduced energy ...

Am J Physiol Gastrointest Liver Physiol 304: G897–G907, 2013. First published March 14, 2013; doi:10.1152/ajpgi.00160.2012.

Mice lacking the intestinal peptide transporter display reduced energy intake and a subtle maldigestion/malabsorption that protects them from diet-induced obesity Dominika Kolodziejczak,* Britta Spanier,* Ramona Pais, Judith Kraiczy, Tamara Stelzl, Kurt Gedrich, Christian Scherling, Tamara Zietek, and Hannelore Daniel ZIEL Research Center of Nutrition and Food Sciences, Abteilung Biochemie, Technische Universität München, Freising, Germany Submitted 27 April 2012; accepted in final form 14 February 2013

Kolodziejczak D, Spanier B, Pais R, Kraiczy J, Stelzl T, Gedrich K, Scherling C, Zietek T, Daniel H. Mice lacking the intestinal peptide transporter display reduced energy intake and a subtle maldigestion/ malabsorption that protects them from diet-induced obesity. Am J Physiol Gastrointest Liver Physiol 304: G897–G907, 2013. First published March 14, 2013; doi:10.1152/ajpgi.00160.2012.—The intestinal transporter PEPT1 mediates the absorption of di- and tripeptides originating from breakdown of dietary proteins. Whereas mice lacking PEPT1 did not display any obvious changes in phenotype on a high-carbohydrate control diet (HCD), Pept1⫺/⫺ mice fed a high-fat diet (HFD) showed a markedly reduced weight gain and reduced body fat stores. They were additionally protected from hyperglycemia and hyperinsulinemia. Energy balance studies revealed that Pept1⫺/⫺ mice on HFD have a reduced caloric intake, no changes in energy expenditure, but increased energy content in feces. Cecal biomass in Pept1⫺/⫺ mice was as well increased twofold on both diets, suggesting a limited capacity in digesting and/or absorbing the dietary constituents in the small intestine. GC-MS-based metabolite profiling of cecal contents revealed high levels and a broad spectrum of sugars in PEPT1deficient mice on HCD, whereas animals fed HFD were characterized by high levels of free fatty acids and absence of sugars. In search of the origin of the impaired digestion/absorption, we observed that Pept1⫺/⫺ mice lack the adaptation of the upper small intestinal mucosa to the trophic effects of the diet. Whereas wild-type mice on HFD adapt to diet with increased villus length and surface area, Pept1⫺/⫺ mice failed to show this response. In search for the origin of this, we recorded markedly reduced systemic IL-6 levels in all Pept1⫺/⫺ mice, suggesting that IL-6 could contribute to the lack of adaptation of the mucosal architecture to the diets. PEPT1; interleukin-6; morphological adaptation; fecal energy; cecal biomass; sodium/hydrogen exchanger-3 THE MAJOR ROLE OF THE SMALL intestine is the absorption of dietary constituents from the ingested food. Digestion of dietary proteins by proteases and peptidases results in the release of short-chain peptides and free amino acids. While various transporters mediate the uptake of free amino acids across the brush-border membrane, di- and tripeptides are selectively transported into enterocytes via the SLC15 member A1 peptide transporter, PEPT1 (8). Peptide transporters are found in all living organisms (29). They act as electrogenic symporters that utilize the proton-motive force for uptake of di- and tripeptides and peptidomimetics (10) with an acidic microclimate on the

* D. Kolodziejczak and B. Spanier contributed equally to the manuscript. Address for reprint requests and other correspondence: H. Daniel, Biochemistry, Technische Universität München, ZIEL Research Center of Nutrition and Food Sciences, Gregor-Mendel-Straße 2, D-85350 Freising, Germany (e-mail: [email protected]). http://www.ajpgi.org

luminal surface providing the protons for influx (7). The proton gradient is maintained by the Na⫹/H⫹ antiporter NHE3 in the apical membrane and by Na⫹-K⫹-ATPase in the basolateral membrane for sodium efflux (11). PEPT1 has unique features and broad substrate specificity. It transports, with a few exceptions, almost all of the 400 di- and 8,000 tripeptides resulting from dietary breakdown of proteins. A variety of drugs that mimic peptides, e.g., amino ␤-lactam antibiotics, selected angiotensin-converting enzyme inhibitors, peptidase inhibitors, and antiviral compounds, were shown to serve as substrates as well (4, 25). Despite extensive analysis of peptide transporter function in various expression systems (18, 2), it is not yet defined what the protein contributes to gastrointestinal physiology and protein nutrition. PEPT1 is mainly expressed in the small intestine with higher levels in the proximal parts than in distal regions. Studies employing Pept1⫺/⫺ mice have demonstrated a lack of intestinal transport of model dipeptides such as glycylsarcosine (5, 15), but animals otherwise do not show any obvious phenotypic alterations (21, 22). We here provide on the analysis of changes in the gastrointestinal tract and of other phenotypic measures in mice deficient of PEPT1 when the animals are fed a diet containing 48 energy% from fat (HFD) compared with a high-carbohydrate/low-fat diet (13 energy% fat). Such high-fat diets are commonly used for inducing obesity in rodent models. Body weight development in Pept1⫺/⫺ animals on HFD but not HCD diet revealed impaired weight gain, and this appeared to derive mainly from a reduced energy intake and a subtle maldigestion and/or malabsorption that, in the case of the HFD, seems not to be fully compensated by bacterial fermentation thus leading to a loss of energy in feces. METHODS

Animals. Pept1⫺/⫺ mice were obtained from Deltagen (San Mateo, CA) (15) and backcrossed for 10 generations to a C57BL/6 background. They were bred and kept like wild-type C57BL/6 mice (WT) in a special pathogen-free animal facility at 22 ⫾ 2°C and a 12:12-h light-dark cycle. All procedures were performed according to the German guidelines for animal care and approved by the state ethics committee under reference number 55.2-1-54-2531-39-10. Body weight and food/water consumption. From 6 wk of age on, Pept1⫺/⫺ and WT mice (n ⫽ 15, respectively) were housed individually. The animals had access to tap water and semisynthetic purified high-fat pellet diet [HFD, 48 energy% fat from palm oil and 28% (wt/wt) starch; Ssniff S5745-E722, Soest, Germany] ad libitum for 12 wk. The control groups received a chemically defined carbohydraterich diet comprised of 48% starch, 5% maltodextrin, and 5% sucrose (all wt/wt) and providing 13 energy% of fat (HCD; Ssniff S5745-

0193-1857/13 Copyright © 2013 the American Physiological Society

G897

G898

PEPT1-DEFICIENT MICE ON HIGH-FAT DIET

E720). Body weight changes as well as food/water intake were recorded one time a week. Rectal body temperature. The rectal body temperature was assessed with a digital precision thermometer (Almemo 2490; Ahlborn, Holzkirchen, Germany). Temperature was determined from each mouse at five time points during the feeding trial. Indirect calorimetry. After 4 wk on the experimental diet, five mice of each group were individually housed for three consecutive days in metabolic cuvettes (volume 2.5 l) connected to an open flow respiratory system (Phenomaster; TSE-Systems, Bad Homburg, Germany). The temperature was adjusted to 22°C, and the mice received food and water ad libitum. Oxygen consumption, CO2 production, and energy expenditure were analyzed and calculated as previously described (3). The median energy expenditure was calculated per hour and was adjusted to gram body weight of each individual animal. Physical activity measurements. After 4 wk on the experimental diet, five mice of each group were housed individually for 3 days in special feeding-drinking-activity cages (TSE-Systems, Bad Homburg, Germany) to monitor activity in the x-, y-, and z-direction. It consists of sensor frames with 32 light barriers on the longitudinal and 16 light barriers on the transversal side. Furthermore, there are sensors in the holding device for food and water containers that allow the detection of the frequency and amount of food and water consumption. All cages are connected to a computer with LabMaster version 2.8.3 software (TSE-Systems), which records and visualizes all measurements for 24 h/day. The mice received food and water ad libitum. Body composition analysis via NMR. Relative lean and fat mass of mice was determined throughout the experiment after 4, 8, and 12 wk using a Minispec mq 7.5 whole body NMR analyzer with Bruker Minispec plus and OPUS v5.5. software. Movement of animals was reduced by a polycarbonate restrainer placed horizontally in the NMR device with measurements performed at 37°C for 3 min. The NMR analyzer was calibrated with dissected adipose tissue and lean muscle. Small intestinal morphology and organ weights. Small intestinal length was determined, and tissues were dissected. Morphology changes were assessed in 4=,6-diamidino-2-phenylindole- and hematoxylin and eosin (H&E)-stained paraffin-embedded tissue sections. Mean villus length was determined in proximal and distal small intestine from 5 animals in each group and from 10 individual villi/sample. Representative epithelial cell numbers were determined by counting cells along a straight 100-␮m section of at least 40 villi/group in H&E-stained slides using a brightfield microscopy (Leica Microsystems, Wetzlar, Germany). For interaction plots, the dataset consisted of five variables and a total of 400 observations: 2 types of diet [HFD and control diet (HCD)], 2 genotypes of mice (WT and Pept1⫺/⫺ mice), 5 individuals per mouse genotype, 2 different regions of the intestine (proximal and distal), and 10 measurements of the villus length per each sample of the intestine. Interaction of the factor variables diet, genotype, and region of the intestine was analyzed using the software R version 2.15.1 using the function “interaction.plot.” Ki67 immunoreactive cells were determined in paraffinembedded sections of the small intestine with a rat anti-mouse Ki67 antibody (Monoclonal Rat Anti-Mouse Ki-67 Antigen Clone TEC-3; Dako, Hamburg, Germany) in 1:25 dilution and with a donkey anti-rat-Cy3-coupled secondary antibody. Organ weights of ceca with and without contents of liver and of epididymal, inguinal, and perirenal fat depots were determined as well. Gastrointestinal transit time. Gastrointestinal transit time was determined according to Koopman et al. (17). Briefly, 10 steel balls of 0.5 mm in diameter were given by gavage at 7:00 P.M. During the following 10 h, feces of individual mice were collected every 30 min and examined for occurrence of the steel balls. The mean recovery time of the 10 steel balls was taken as the apparent gastrointestinal transit time. Because the experiment was performed on two subsequent days, the final transit time was calculated as mean from the values of both days.

RNA isolation and real-time RT-PCR. RNA was isolated from the mucosa of the proximal small intestine using a combination of phenol-chloroform extraction with Isol-RNA Lysis reagent (5 Prime, Hamburg, Germany) and an RNeasy Mini Kit (Qiagen, Hilden, Germany). Ethanol-precipitated RNA was transferred onto minispin columns of the RNeasy kit and further proceeded according to the manufacturer’s instructions. RNA was transcribed into cDNA using the Moloney murine leukemia virus reverse transcriptase, RNase H minus, and point mutant enzyme (Promega, Mannheim, Germany). Expression analysis was performed with a Roche Light-Cycler 2.0 capillary system applying the fluorescent dye SYBR Green (Roche, Mannheim, Germany). Isolated RNA was diluted to 20 ng/␮l, and primers were used at a concentration of 20 ␮M. The primer sequences were as follows: mGAPDH: forward atcccagagctgaacg, reverse gaagtcgcaggagaca; mGLUT-2: forward gcctgtgtatgcaacc, reverse gctcacgtaactcatcca; mGLUT-5: forward acacctactacgacagaaa, reverse gctcaaacgattgggc; mSglt1: forward taccgttggaggcttc, reverse agatactccggcatcg; mFatp4: forward gtagtgtggcaacttcct, reverse gcatggatctcacagatagc; mFabp2: forward gaagcttggagctcatgaca, reverse gcttggcctcaactccttc; mCD36: forward gccaagctattgcgacatga, reverse caatggttgtctggattctgg; mPat1: forward gctaccatgtccacacagagg, reverse tgcagtgcacggccacgatac. The data of the expression analysis were analyzed by the 2-(⌬⌬Ct) method of Livak and Schmittgen (19a) and normalized against the housekeeper gene mGAPDH. Fecal analysis. Feces of five mice of each group were sampled every 4 wk and analyzed for the excreted energy using a bomb calorimeter. For analysis, 0.9 –1.1 g of feces were ground and pressed. The pellets were burned in an isoperibol oxygen bomb calorimeter 6300 (Parr Instrument, Frankfurt, Germany). Furthermore, fecal composition was analyzed via Fourier-transform infrared spectroscopy (FT-IR). For analysis, the FT-IR spectrometer Tensor 27 (Bruker Optik, Ettlingen, Germany) was used. The spectrometer was precooled with liquid nitrogen. Each sample was measured in triplicates, and data were analyzed using the OPUS software (Bruker Optik). GC-MS analysis of cecal contents. Metabolites were extracted from 10 –30 mg (measured samples were corrected for weighted dry mass) of lyophilized cecal contents with 500 ␮l of ice-cold methanol. After samples were incubated for 10 min at 4°C on a shaker, they were centrifuged for 5 min at 4°C and 14,000 g. The supernatant was divided equally into two tubes and dried using a vacuum concentrator (SPD 111V SpeedVac; Thermo Savant). One sample was used for analysis of free fatty acids. Processing of cecal samples was performed as described before (20). Briefly, for free fatty acid analysis, 20 ␮l of pyridine containing 40 mg/ml methoxyamine hydrochloride were added to the dried pellets and incubated for 90 min at 30°C. Derivatization was performed by treatment of samples for 30 min at 37°C with 32 ␮l N-methyl-N-trimethylsilyl-trifluoroacetamide. One microliter of the derivatized sample was finally injected for analysis. GC was operated on a VF 5-ms capillary column, 30 m long (0.25 mm inner diameter, 25 ␮m film thickness; VARIAN, Palo Alto, CA) at a constant flow of 1 ml/min helium. The temperature program started isocratic with 1 min at 70°C followed by temperature ramping at 10°C/min to a final temperature of 330°C, which was held for 8 min. Metabolite profiling was performed via GC coupled to a Quadropole (Agilent 5975) mass analyzer with scan rates of 2.5/s and mass ranges of 70 to 600 Da. Significant differences among the sample groups resulting from cluster analysis were obtained by Kruskal-Wallis tests (selected false discovery rate limit ⫽ 0.05; Bonferroni adjustment), and clustering was based on Euclidean distance and average linkage. Plasma analysis. Plasma samples were collected by puncturing the retro-orbital sinus with heparin-coated capillaries (Neolab, Heidelberg, Germany). General blood parameters and plasma lipids were quantified using the Piccolo General Chemistry 13 Reagent Disc or the Piccolo Lipid Panel Reagent Disc (HITADO, Möhnesee, Germany) according to the manufacturer’s instructions. Briefly, 100 ␮l of plasma were added to a well in the middle of the disc and measured in a Piccolo xpress Chemistry Analyzer (HITADO). Plasma nonest-

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00160.2012 • www.ajpgi.org

G899

PEPT1-DEFICIENT MICE ON HIGH-FAT DIET

erified fatty acids (NEFAs) were determined using the NEFA-HR(2) kit (Wako Chemicals, Neuss, Germany). Phospholipid concentrations in plasma were assayed with the LabAssay Phospholipid kit (Wako Chemicals). Preparation of reagents and analysis of all parameters were carried out according to the manufacturer’s instructions. Total glucagonlike peptide 2 (GLP-2) was measured in the plasma of mice with a commercially available GLP-2 EIA kit (Yanaihara Institute), which shows a high specificity to mouse GLP-2 and no cross-reactivity with mouse glucagon and mouse GLP-1. Plasma interleukin-6 (IL-6) was measured using the multiplex technology (MILLIPLEX MAP kit; Millipore) following the manufacturer’s instructions. Lipid content in liver. Twenty milligrams of liver tissue were homogenized in liquid nitrogen and mixed with 0.9% NaCl. Two hundred microliters of the suspension were transferred to a fresh tube and mixed with 500 ␮l of 0.5 M ethanolic KOH. Samples were shaken in a thermo block for 30 min at 71°C. After 1 ml of 0.15 M MgSO4 was added, samples were centrifuged for 10 min at room temperature and 14,000 g. The supernatant was used for triglyceride analysis. Analysis was performed using the Triglycerides liquicolormono kit (Human, Wiesbaden, Germany) according to the manufacturer’s instructions. NEFAs were determined using the NEFA-HR(2) kit (Wako Chemicals). The phospholipid concentrations were assayed with the LabAssay Phospholipid kit (Wako Chemicals). Preparation of reagents and analysis of all parameters were carried out according to the manufacturer’s instructions. Statistical analysis. All data are expressed as means ⫾ SE. Statistical analysis was performed using GraphPad Prism 4.01 (San Diego, CA). One-way, two-way, or three-way ANOVA and Bonferroni test or unpaired Student’s t-test were used to test for statistical significance. RESULTS

Body weight and food/water intake and organ weights. Body weight was not significantly different between Pept1⫺/⫺ mice and WT animals at 6 wk of age when the feeding trial started. After 12 wk on the HFD, Pept1⫺/⫺ mice showed a significantly lower body weight (35.9 ⫾ 1.4 g, P ⬍ 0.001) compared with WT mice (42.8 ⫾ 1.2 g). Animals fed the HCD for 12 wk showed no significant differences in body weight (Pept1⫺/⫺ 26.5 ⫾ 0.5 g, WT mice 28.7 ⫾ 0.5 g). The HCD was used instead of a standard chow, since it is chemically defined with a well-known stable carbohydrate composition and its energy content per gram is comparable to the HFD. The total weight gain during the feeding period (Fig. 1A) for Pept1⫺/⫺ mice on HFD was 15.0 ⫾ 0.9 g, whereas the WT mice on HFD gained 19.6 ⫾ 1.0 g. On HCD, Pept1⫺/⫺ mice gained 5.1 ⫾ 0.3 g and WT mice 4.6 ⫾ 0.5 g. The cumulative food intake of the animals on the different diets is shown in Fig. 1B. Whereas WT mice on HFD had a mean energy intake of 81.9 ⫾ 2.1 kJ/day,

Pept1⫺/⫺ mice displayed a slightly lower energy intake of 75.7 ⫾ 2.0 kJ/day. WT mice on HCD consumed 59.7 ⫾ 4.00 kJ compared with Pept1⫺/⫺ mice with 56.8 ⫾ 0.8 kJ/day. Water consumption revealed no significant differences, neither by diet nor by genotype. When organ weights relative to body weight were determined, the liver weights did not differ significantly. However, epididymal fat depots in Pept1⫺/⫺ mice on a HFD were significantly (P ⬍ 0.001) lower than in WT mice on HFD, accounting to 4.18 ⫾ 0.51% of body weight compared with 5.75 ⫾ 0.18%. For the perirenal fat 1.99 ⫾ 0.51% of body weight in Pept1⫺/⫺ mice on a HFD and 2.68 ⫾ 0.18% for WT animals on HFD (P ⬍ 0.01) were determined. Body composition analysis via NMR. Body composition differences mirrored differences in body weight over time, whereas WT and PEPT1-deficient animals on HCD had similar lean mass accounting to 70.75 ⫾ 1.51% (WT) and 74.45 ⫾ 0.40% (Pept1⫺/⫺ mice) of total body mass. On HFD, WT animals displayed a lean mass of 49.69 ⫾ 1.91% and a fat mass of 38.20 ⫾ 2.23% of total body mass, whereas mice lacking PEPT1 had 59.65 ⫾ 2.59% lean mass and only 28.24 ⫾ 4.32% of fat mass. Energy expenditure and body temperature. One possible cause for a reduced weight gain could be increased energy expenditure. We found that neither the locomotor activity (Fig. 2A) nor the energy expenditure (Fig. 2B) was changed in PEPT1-deficient mice compared with WT mice on a HFD. The respiratory quotient (RQ) as calculated from the ratio of CO2 production and oxygen consumption showed a clear diet-specific effect. The RQ in mice on the HFD was as expected lower than in animals on a HCD, indicating the major use of fatty acids instead of glucose. However, no genotype-specific differences in the RQ were detected (Fig. 2C). Another factor that might lead to increased energy expenditure is the investment into thermoregulation. However, there were no significant differences detectable in the body temperature between the four study groups (WT-HFD: 37.6 ⫾ 0.1°C, Pept1⫺/⫺-HFD: 37.0 ⫾ 0.2°C, WT-HCD: 37.9 ⫾ 0.3°C, and Pept1⫺/⫺-HCD: 36.8 ⫾ 0.1°C). These results suggest that the reduced weight gain and body fat mass of Pept1⫺/⫺ mice is not a consequence of increased energy expenditure or heat production. Fecal energy excretion. To assess further energy balance, fecal energy excretion was measured at 4, 8, and 12 wk of the feeding trial. Pept1⫺/⫺ mice on HFD showed significantly higher energy contents in feces (15.5 ⫾ 0.3 kJ/g, P ⬍ 0.001) compared with WT animals on HFD that excreted 13.4 ⫾ 0.2 kJ/g (Fig. 3A). Pept1⫺/⫺ and WT animals on HCD had a mean

Fig. 1. Mean body weight and energy intake of Pept1⫺/⫺ and wild-type (WT) mice fed a high-fat diet (HFD) or high-carbohydrate diet (HCD) for 12 wk. A: mean body weight over the cause of the 12 wk of feeding the different diets. B: cumulative energy intake as calculated from mean food intake rates. Data are expressed as means ⫾ SE of n ⫽ 15 mice with significant differences denoted as WT animals on HFD different from all other groups (P ⬍ 0.001) (a) and Pept1⫺/⫺ animals on HFD different from all other groups (P ⬍ 0.001) (b).

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00160.2012 • www.ajpgi.org

G900

PEPT1-DEFICIENT MICE ON HIGH-FAT DIET

Fig. 2. Locomotor activity and energy expenditure in Pept1⫺/⫺ and WT mice. After 4 wk on HFD and HCD, selected mice (n ⫽ 5, respectively) were placed individually in special feeding-drinking-activity cages of the TSE system for three consecutive days. The resting (photophase, day) and activity (scotophase, night) phases are indicated by white and black bars. A: cumulative activity over the total period of 72 h. B: median (left) and total (right) energy expenditure during day and night cycles. C: median (left) and total (right) metabolic rate (respiratory quotient, RQ) during day and night cycles. Data are expressed as means ⫾ SE. The SE was calculated based on n ⫽ 5 mean values/animal. Every 10 min, a data point was measured, which were summarized as hourly mean values, and these 12 hourly mean values per animal were again summarized to the animal-specific mean values. Statistical differences are denoted as P ⬍ 0.05 (*).

energy excretion of 13.2 ⫾ 0.3 and 12.7 ⫾ 0.2 kJ/g, respectively, without significant differences. FT-IR analysis of fecal samples revealed distinct differences in composition that allowed all four study groups of animals to be clearly separated by a cluster analysis (Fig. 3B). In Pept1⫺/⫺ mice on HFD, the signal derived from lipids at wave numbers around 2,800/cm was higher than in WT animals. Intestinal morphology and nutrient transporter mRNA expression levels. Inspection of the intestinal anatomy revealed Pept1⫺/⫺ mice to have a significantly increased length of the

small intestine with 36.5 ⫾ 0.6 and 39.7 ⫾ 0.7 cm on HFD and HCD, respectively, compared with WT animals with 32.5 ⫾ 0.8 (HFD) and 33.6 ⫾ 0.6 (HCD) cm in length. Although Pept1⫺/⫺ mice possessed this increased length of the small intestine, gastrointestinal transit time of orally administered tracers remained unchanged with a mean of 6.0 ⫾ 0.1 h in WT animals and 6.4 ⫾ 0.3 h in Pept1⫺/⫺ mice. The mean villus length was determined in 5-␮m tissue sections by microscopy. This revealed in duodenum and upper jejunum but not in the distal part of the small intestine marked differences between

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00160.2012 • www.ajpgi.org

PEPT1-DEFICIENT MICE ON HIGH-FAT DIET

A

WT - HCD

Faeces Energy [kJ/g]

17

WT - HFD

-/-

-/-

Pept1 - HCD

Pept1 - HFD

16

a

15 14 13 12 11 0 1 2 3 4 5 6 7 8 9 10 11 12 13 Feeding weeks

B

Pept1-/- - HCD

WT- HCD

WT- HFD

Pept1-/- - HFD

5

1.0

2.0

Absorbance Units

Heterogenity

0 6

3

4

2

1

7

8 Pept1-/- - HCD Pept1-/- - HFD WT – HCD WT - HFD

Wavenumber cm-1

Fig. 3. Fecal energy excretion and composition of feces of Pept1⫺/⫺ and WT mice on HFD and HCD. A: fecal energy content of fecal samples collected during weeks 4, 8, and 12 determined by bomb calorimetry. B: cluster analysis of fecal samples analyzed by Fourier-transform infrared spectroscopy (FT-IR) as described in METHODS. Data in A are expressed as means ⫾ SE of n ⫽ 15 mice with significance provided as Pept1⫺/⫺ mice on HFD different from all other groups (P ⬍ 0.001) (a).

Pept1⫺/⫺ and WT animals with a pronounced diet effect (Fig. 4, A and B). In WT mice on HCD, mean villus length was 434 ⫾ 10 ␮m, whereas in Pept1⫺/⫺ mice it was reduced by around 20% (353 ⫾ 9 ␮m, P ⬍ 0.01). In animals on HFD, this difference increased drastically, resulting in WT mice in a mean villus length of 548 ⫾ 5 ␮m, whereas it remained low in Pept1⫺/⫺ mice with 318 ⫾ 10 ␮m, resulting in a 40% difference between the groups. In contrast to the villus length, the mean villus width did not differ among diets or genotypes (data not shown), and inspection of tissue sections also did not reveal any other obvious morphological differences (Fig. 4C). By using interaction plots, diet and genotype effects on proximal but not distal regions of the intestine became obvious (Fig. 4D). Immunofluorescence staining of tissue samples with Ki67 as a marker for cell proliferation revealed no differences in the number of proliferating cells neither in the crypts nor the villus region in WT and Pept1⫺/⫺ mice (data not shown). Furthermore, the number of epithelial cells per 100 ␮m of villus length counted based on nuclei detected in H&E-stained sections did also not reveal any differences (data not shown). In assessing whether adaptive changes in mRNA expression of selected genes involved in intestinal nutrient absorption take place, we determined apparent mRNA levels of transporters involved in carbohydrate uptake such as GLUT-2, GLUT-5,

G901

and Sglt1, for lipids CD36, Fabp2, and Fatp4, and for amino acids the representative proton-dependent Pat1 transporter. After data normalization, only two transporter genes were found to change in mRNA levels depending on diet or genotype. The fructose transporter GLUT-5 was significantly lower in WT mice on HFD compared with HCD, whereas no dietspecific regulation was observed in Pept1⫺/⫺ mice. mRNA expression of the fatty acid transporter Fatp4 in WT mice was reduced on HFD compared with HCD. In Pept1⫺/⫺ mice, the Fatp4 expression level was generally reduced compared with WT on HCD, and no diet-specific changes were observed (Fig. 5). Taken together, the findings suggest that, despite the dietdependent differences in the apparent surface area of the upper small intestine, only minor changes in nutrient transporter expression occur in either adaptation to diet or with respect to genotype. This suggests that a surface-dependent reduction in overall digestion and absorption capacity causes the effects in PEPT1-deficient mice on HFD. Analysis of cecal contents. As a maldigestion/malabsorption should affect the cecal nutrient/metabolite composition, we analyzed cecal contents. All mice fed the HFD displayed a reduced cecal weight compared with mice on HCD. In WT mice on HCD, cecal weight accounted for 1.18 ⫾ 0.08% of the body weight but only to 0.54 ⫾ 0.02% when fed the HFD. Pept1⫺/⫺ mice on HCD had the highest relative cecal weight with 2.30 ⫾ 0.15% of the body weight, which was reduced to 1.08 ⫾ 0.11% on HFD. When cecal contents were submitted to GC-MS analysis, marked differences in metabolite concentrations were obtained (Fig. 6A) although WT animals on HCD or HFD revealed only subtle differences. Pept1⫺/⫺ mice, on the other hand, displayed pronounced changes in cecal metabolites when HCD and HFD were compared. The principal component analysis shown in Fig. 6B, with normalized metabolite data [x ⫽ (value ⫺ average)/SD], revealed a clear separation caused by different metabolite abundances in all animals based on either diet or genotype. Cecal contents of Pept1⫺/⫺ mice fed HFD were classified by high concentrations of glycerol, medium- and long-chain fatty acids with chain length of C-15 to C-22, but low levels of carbohydrates (including glucose, fructose, and mannose) and amino acids (including alanine and glycine). In contrast, WT animals on a HFD displayed higher levels of a variety of amino acids such as alanine, isoleucine, threonine, tyrosine, and lysine but also elevated levels of glucose, fructose, galactose, N-acetylglucosamine, mannose, or xylose. The metabolite profile in the PEPT1-deficient animals on a HCD revealed high levels of carbohydrates and amino acids and low concentrations of fatty acids, tricarboxylic acid cycle intermediates (malic acid, oxalic acid), and metabolites included in RNA processing such as uridine, guanosine, and inosine. These findings suggest that, in Pept1⫺/⫺ mice on both diets, larger quantities of nonabsorbed food constituents reach the cecum and are submitted to bacterial degradation, which results in differences in biomass and cecal compositions. Plasma clinical chemistry parameters. After 12 wk of feeding the specific diets, plasma of Pept1⫺/⫺ and WT animals on HFD and HCD had comparable concentrations of albumin, calcium, creatinine, ␥-glutamyltransferase, total bilirubin, total protein, blood urea nitrogen, and uric acid (data not shown). Although PEPT1-deficient animals on HFD had a reduced body weight, the plasma lipid concentrations of cholesterol, high- and low-density lipoprotein, very low density lipoprotein, and triglyc-

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00160.2012 • www.ajpgi.org

G902

PEPT1-DEFICIENT MICE ON HIGH-FAT DIET

A

B

C

D a

b

Villi length [µm]

Fig. 4. Morphological analysis of small intestinal mucosa of Pept1⫺/⫺ and WT mice fed HFD or HCD. A: mean villus length as determined in duodenum and upper jejunum (proximal). B: mean villus length of lower jejunum and ileum (distal). Data are expressed as means ⫾ SE of 5 mice in each group with n ⫽ 10 villi/animal and significances provided as WT animals on HFD different from all other groups (P ⬍ 0.001) (a). C: representative tissue section of duodenum samples obtained from a WT and a Pept1⫺/⫺ mouse on HFD. Arrows indicate the representative villus length and width used for analysis. The bar represents 50 ␮m. D: interaction plots of the villus length in the proximal and distal small intestine of WT and Pept1⫺/⫺ mice on a HFD and control diet (HCD). Whereas the diet has no significantly altered impact on the villus length in the various regions of the intestine (P ⫽ 0.4876) (a), the effect of the diet on the villus length depends significantly on the mouse genotype (highest effect in WT, P ⫽ 0.00155) (b), and the effect of the genotype on the villus length depends significantly on the various intestinal regions (highest effect in proximal intestine, P ⬍ 0.001) (c). Each data point shown represents the mean ⫾ SE of 5 animals with 10 villi measurements/intestinal region (N ⫽ 10, n ⫽ 100).

Pept1-/- mouse, HFD

WT mouse, HFD

Villi length [µm]

HCD

Diet

Intestinal region

Genotype

--1-- proximal 2 distal

--1-- WT 2 Pept1-/-

HFD

WT

Genotype

Pept1-/-

Intestinal region

c

--1-- proximal 2 distal

HCD

erides were comparable to those in WT mice (Table 1). In contrast, Pept1⫺/⫺ mice had significantly lower plasma concentrations of glucose, alanine aminotransferase, and aspartate aminotransferase and increased concentration of alkaline phosphatase. Although the glucose concentration and plasma insulin was also lower in Pept1⫺/⫺ mice on HFD, neither reached significance. Concentrations of GLP-2 in plasma were determined since GLP-2 was reported to alter intestinal PEPT1 expression (14), but

Diet

HFD

no differences were found between any of the four groups. Because IL-6 was recently reported to affect intestinal growth (16), we also analyzed IL-6 plasma concentrations, and this revealed that in WT mice on a HFD IL-6 concentrations in plasma increased twofold, whereas systemic IL-6 levels in all Pept1⫺/⫺ mice were significantly lower than in WT, indicating that a diet-induced compensatory mechanism is missing in PEPT1-deficient animals (Table 1).

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00160.2012 • www.ajpgi.org

mGlut2

1.5

mRNA expression ratio

1.0

0.5

1.0

0.5

mRNA expression ratio

*

1.0

0.5

0.0

mFatp4 * *

1.5

PEPT1-/--HCD

*

1.0

0.5

mRNA expression ratio

mSglt1

1.5

1.0

0.5

PEPT1-/--HCD

WT - HCD

mFabp2

2.0 1.5

Fig. 5. mRNA expression of selected carbohydrate (mGLUT-2, mGLUT-5, mSglt1), fatty acid (mCD36, mFatp4, mFabp2), and amino acid (mPat1) transporters in the proximal small intestine of WT and Pept1⫺/⫺ mice on HFD and control diet (HCD). Normalized values are indicated as means of 5 individuals with appropriate averaged SE. RNA expression ratios (⌬Ct) are shown with reference to WT mice on control diet, with a fixed mean ⌬Ct value of 100%. Statistical significance is shown as P ⬍ 0.05 (*).

1.0 0.5

PEPT1-/--HCD

WT - HCD

PEPT1-/--HCD PEPT1-/--HCD

WT - HFD

WT - HCD WT - HCD

0.0 PEPT1-/--HFD

0.0

PEPT1-/--HFD

WT - HFD

PEPT1-/--HCD

WT - HCD

PEPT1-/--HFD

WT - HFD

0.0

WT - HFD

mRNA expression ration

WT - HCD

WT - HFD

PEPT1-/--HCD

WT - HCD

PEPT1-/--HFD

WT - HFD

mRNA expression ratio

mGlut5 1.5

PEPT1-/--HFD

0.0

0.0

mRNA expression ratio

G903

mCD36

1.5

PEPT1-/--HFD

mRNA expression ratio

PEPT1-DEFICIENT MICE ON HIGH-FAT DIET

mPat1

2.0 1.5 1.0 0.5

PEPT1-/--HFD

WT - HFD

0.0

Liver lipids. WT and Pept1⫺/⫺ animals on HFD showed increased liver weight (2.19 ⫾ 0.72 and 1.69 ⫾ 0.47 g, respectively) compared with animals on HCD with 1.22 ⫾ 0.21 g (WT) and 1.29 ⫾ 0.12 g (Pept1⫺/⫺). Only WT mice on HFD developed a fatty liver with high levels in triglycerides (32.7 ⫾ 1.9 mg/g liver tissue), phospholipids (80.8 ⫾ 5.7 mg/g liver tissue), and NEFAs (99.4 ⫾ 17.6 mmol/g liver tissue). In contrast, Pept1⫺/⫺ mice on HFD displayed HCD-comparable concentrations of triglycerides (36.3 ⫾ 5.4 mg/g liver tissue) and phospholipids (71.4 ⫾ 3.1 mg/g liver tissue) but showed

significantly (P ⬍ 0.05) lower NEFA levels (37.2 ⫾ 2.3 mmol/g liver tissue). DISCUSSION

Although the structure and function of PEPT1 has been studied extensively in heterologous expression systems or cell models, its role in gut physiology, energy assimilation, and protein nutrition has not been well defined. With the availability of mice lacking the peptide transporter, this can now be

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00160.2012 • www.ajpgi.org

G904

PEPT1-DEFICIENT MICE ON HIGH-FAT DIET

A -2.0 PEPT1-/- HFD

WT HCD

WT HFD

0.9693894

0.3774506

PEPT1-/- HCD -0.21448821

2.0

0.0

Pentadecanoic acid Heptadecanoic acid n-heneicosan-1-ol Octadecanoic acid peptide 2321.64 Docosanoic acid Tetracosanoic acid 9,12-(Z,Z)-octadecadienoic acid 9-(Z)-octadecanoic acid Glycerol Hexadecanoic acid Tetradecanoic acid Fumaric acid Ornithine Threose Glycerol (conjugated) 2852.93 n-Eicosan-1-ol Glycerophosphoglycerol n-octadecan-1-ol fatty acid 2261.39 gamma-tocopherol 13-(Z)-docosenoic acid 1-O-octadecylglycerol Cholesterol Lanosterol 3-beta-cholest-5-en-3-ol alpha-tocopherol n-hexadecan-1-ol 2,2-dimethyl-succinic acid Ethanolaminephosphate fatty acid 1913.3 Glucose-6-phosphate Glycerol-3-phosphate Nicotinic acid Pyroglutamic acid Arabinose Hypoxanthine Uracil Thymine Xanthine alpha-tocopherolacetate 4-amino-bencoic acid 1-monooleoylglycerol fatty acid 2589.18 myo-inositol Inosine Guanosine Uridine Sitosterol Oxalic acid Ursodeoxycholic acid Glyceric acid Malic acid 2-amino-butanoic acid 2-oxo-isovaleric acid 3-methyl-2-oxopentanoic acid 2 2-oxo-isocaproic acid Alanine Glycine D-Glucopyranose Glucose Hexose 2009.31 Methionine N-acetyl-glucosamine Galactose N-acetyl-mannosamine Rhamnose Pyruvic acid Fructose Mannose N-acetyl-glucosamine Xylose/Lyxose

Fig. 6. GC-MS analysis of cecal contents of Pept1⫺/⫺ and WT mice fed HFD or HCD. A: heat map showing hierarchical clustering of metabolites as determined by semiquantitative GC-MS analysis of cecal contents of 11–15 animals in each group. Color coding displays significant changes in the relative metabolite levels for standardized data, with blue indicating lower concentrations and yellow indicating higher levels with metabolite identities provided on the right. Characteristic clusters of metabolites separating Pept1⫺/⫺ mice from WT controls fed HFD are indicated by red boxes. B: principal component analysis blot with normalized metabolite data [x ⫽ (value ⫺ average)/SD] obtained from Pept1⫺/⫺ and WT control mice fed either HFD or HCD with clear separation by genotype and diet effects.

B

addressed at the organismic level. Surprisingly, PEPT1-deficient mice with a C57BL/6 genetic background show no obvious phenotypical changes when fed a standard high-carbohydrate diet. The animals grow to WT-like size with normal body and organ weights and no genotype-related abnormalities

in organ histology (15, 21). We recently demonstrated that Pept1⫺/⫺ mice challenged with a high-protein diet providing 45% of energy from protein showed impairments in development and a markedly reduced food intake and severe weight loss during the first days on the diet (22). Although food intake

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00160.2012 • www.ajpgi.org

G905

PEPT1-DEFICIENT MICE ON HIGH-FAT DIET

Table 1. Plasma parameters obtained from Pept1⫺/⫺ and WT mice on HFD and HCD Pept1⫺/⫺-HFD

WT-HCD

Pept1⫺/⫺-HCD

162.67 ⫾ 5.16 95.67 ⫾ 4.94 51.33 ⫾ 6.59 16.00 ⫾ 1.26 80.33 ⫾ 6.99 263.33 ⫾ 16.29 1.86 ⫾ 0.39 87.00 ⫾ 13.64 982.33 ⫾ 23.73 31.67 ⫾ 6.64 65.67 ⫾ 4.57 1.01 ⫾ 0.09 0.63 ⫾ 0.08

89.00 ⫾ 11.32 65.50 ⫾ 8.52 12.00 ⫾ 2.68 11.67 ⫾ 1.02 58.33 ⫾ 4.79 244.83 ⫾ 44.71 0.74 ⫾ 0.09 57.33 ⫾ 4.94c 757.17 ⫾ 29.52c 36.67 ⫾ 9.53 100.50 ⫾ 16.89 1.06 ⫾ 0.20 1.42 ⫾ 0.37c

88.00 ⫾ 3.86 61.50 ⫾ 2.86 15.83 ⫾ 1.82 10.50 ⫾ 0.96 52.00 ⫾ 5.27 177.50 ⫾ 6.35 0.60 ⫾ 0.16 76.33 ⫾ 4.06 694.50 ⫾ 44.71 19.83 ⫾ 2.12 57.67 ⫾ 2.64 1.02 ⫾ 0.06 0.64 ⫾ 0.07

WT-HFD

Cholesterol, mg/dl HDL, mg/dl LDL, mg/dl VLDL, mg/dl Triglycerides, mg/dl Glucose, mg/dl Insulin, ng/ml Alkaline phosphatase, U/l ␣-Amylase, U/l ALT, U/l AST, U/l GLP-2, ng/ml IL-6, pg/ml

187.00 ⫾ 6.17 109.00 ⫾ 4.12 63.00 ⫾ 3.92 15.67 ⫾ 1.20 78.67 ⫾ 5.77 366.00 ⫾ 28.45b 3.49 ⫾ 0.67b 71.67 ⫾ 7.01c 1,102 ⫾ 86.59c 105.33 ⫾ 25.54 129.67 ⫾ 21.28 0.96 ⫾ 0.05 2.25 ⫾ 0.53c a

a

Data are expressed as means ⫾ SE; n ⫽ 6 mice in each group. WT, wild type; HFD, high-fat diet; HCD, high-carbohydrate diet; HDL, high-density lipoprotein; LDL, low-density lipoprotein; VLDL, very low density lipoprotein; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GLP-2, glucagon-like peptide-2; IL-6, interleukin-6. Clinical chemistry data and plasma GLP-2 and IL-6 levels of mice after 12 wk of feeding the HFD or HCD. Significance (in superscript) is denoted by only diet effect (a), diet and genotype effect (b), and only genotype effect (c).

returned to normal levels after 5 days, Pept1⫺/⫺ mice did not regain weight and remained significantly leaner than WT animals. These findings were the first indication that different diets can cause different phenotypic changes and that such an approach may provide additional insights into the role of PEPT1 in gastrointestinal physiology. Here we provide data demonstrating that Pept1⫺/⫺ mice fed a 48 energy% HFD compared with a HCD (13 energy% fat) do show significant changes in energy balance, gut morphology, and physiology and selected parameters of clinical chemistry. HFD are frequently used in rodent models, including C57BL/6 mice (30, 32) or rats (27), to cause obesity associated with metabolic disturbances such as hyperglycemia, hyperinsulinemia, and nonalcoholic fatty liver disease. These phenotypic impairments grossly mimic abdominal obesity phenotypes in humans and are linked to the development of type 2 diabetes (6). Such HFD challenge the intestine by the need to maximize fat digestion and absorption and are associated with major changes in gene transcription in the intestine of genes related to fat handling but also of cell cycle control, inflammation, and immune function (9). Surprisingly, PEPT1-deficient mice fed such a HFD displayed a significantly reduced weight gain compared with WT animals. Significant differences in body weight development over 12 wk can arise from small differences in energy intake and/or energy excretion. We observed that Pept1⫺/⫺ mice had a mean daily energy intake that was around 6% lower than in WT mice. Quantifying small differences in energy intake, including food spillage and energy disposal, is challenging. Within the boundaries of technical limitations, no significant differences in energy expenditure, as assessed by indirect calorimetry, locomotion analysis, and body temperature recordings, were found that could account for the reduced weight gain in PEPT1-deficient animals on HFD. However, NMRbased determination of body composition after 12 wk on HFD clearly established that PEPT1-deficient mice have more lean mass and around 10% less body fat than WT mice. This is presumably the result of both slightly lower caloric intake and a significantly increased energy excretion in Pept1⫺/⫺ mice, specifically on the HFD. In search of the origin of the impairments in energy assimilation in the intestine of Pept1⫺/⫺ mice, morphometric anal-

ysis of the intestinal mucosa indicated that the HFD caused a 26% increase in mean villus length in the proximal but not in distal small intestine of WT animals. In contrast, in PEPT1deficient animals that already had a reduced villus length (around 20% shorter than controls) on HCD, the HFD failed to cause the morphological adaptation seen in control animals on HFD. This resulted in an around 40% reduced villus length in the proximal small intestine. However, the length of the small intestine increased significantly by around 10% in all Pept1⫺/⫺ mice independent of the diet when measured under standardized conditions. We hypothesize that an increase in gut length might be a sign for a partial compensation of the reduced mucosal surface area. Assessment of mRNA expression levels of selected glucose, amino acid, and fatty acid transporters in the mucosa revealed no major differences between WT and Pept1⫺/⫺ mice on HFD, indicating a lack of compensatory mechanisms in knockout mice to overcome the morphology changes. The missing adaptation to dietary conditions in PEPT1⫺/⫺ mice is not only visible on the transcript level but was also found in the systemic IL-6 concentration and in the villus length of the proximal small intestine. The finding that feeding a HFD causes an adaptive change in the mucosal architecture is not new but is rarely reported in mouse studies, yet an increase in mucosal mass and average villus length is a consistent finding in both rats and mice subjected to HFD trials (1, 12, 23, 26). An increase in villus length in C57/Bl6 mice fed a HFD was reported to be associated with a slight, yet not significant, increase in proliferation rate as revealed by Ki67 staining (9). In the current study, Ki67 immunostaining also did not reveal significant differences in the number of stained cells along the crypt-villus axis. Because the number of cells per 100 ␮m of villus remained unchanged, the total number of cells per villus increased in accordance with previous findings reported by de Wit et al. (9). That these alterations in the intestinal architecture that reduce the overall surface area can reduce fat absorption and cause a partial resistance to dietinduced obesity was also shown in another mouse model (31). Major increases in total cecal weight and cecal contents in animals lacking PEPT1 indicated that the capacity for digestion and absorption of dietary constituents in the intestine in transporter-deficient animals was impaired. This may be attributed to the reduced mucosal surface in the proximal small intestine.

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00160.2012 • www.ajpgi.org

G906

PEPT1-DEFICIENT MICE ON HIGH-FAT DIET

Whereas the HFD compared with the HCD caused a reduced cecal weight in WT animals, all PEPT1-deficient animals had ceca two times as large as controls with increases in luminal biomass of the cecum. A qualitative analysis of fecal contents by FT-IR and GC-MS indicated an enrichment of free fatty acids in cecum that together with the twofold increase in total cecum mass in Pept1⫺/⫺ mice on HFD strongly suggest an insufficient small intestinal fat digestion and/or absorption. That the microbiota cannot efficiently extract the fat contained in cecum becomes obvious by the increased energy content in feces found in PEPT1-deficient animals on HFD. Changes in metabolite profiles in cecal contents in Pept1⫺/⫺ mice suggested that, not only in the HFD but also the HCD, a higher nutrient influx into cecum may occur. However, in this case, the higher cecal carbohydrate load appears to be compensated by increased fermentation that increases both bacterial biomass and production of short-chain fatty acids. With an increased production and absorption of short-chain fatty acids, the diet energy is made available to the host, and this could explain the absence of differences in fecal energy excretion between Pept1⫺/⫺ mice and controls on HCD and a lack of difference in weight gain. Because a prominent difference between WT and Pept1⫺/⫺ mice when fed a HFD was found in a markedly reduced surface area in upper small intestine, we asked which systemic factors might induce this difference between the two genotypes. Intestinal adaptation to dietary factors is considered to be mainly controlled by hormones secreted from the gastrointestinal tract, and here GLP-2 seems to play the most important role (19, 24). Both, in the context of short bowel syndrome or intestinal atrophy during total parenteral nutrition, GLP-2 proved to promote mucosal growth and functional adaptation of the intestine (13, 32). However, when we analyzed nonfasting plasma levels of GLP-2 in the mice, no differences between genotypes or diets were observed. In contrast, IL-6 in plasma revealed significantly lower concentrations in PEPT1-deficient animals on both diets. Recently, it was demonstrated that systemically overexpressed IL-6 has pronounced effects on intestinal morphology in mice with an increase in villus length and mucosal mass caused by an inhibition of apoptosis in enterocytes (16). IL-6 may therefore be considered as a natural gut growth-promoting chemokine. The observation that all Pept1⫺/⫺ mice displayed reduced systemic IL-6 levels suggests that, in these animals, gut growth and the adaptation to diet may be impaired because of low IL-6 levels. What, however, causes the reduced systemic IL-6 level in the transporter-deficient animals awaits further studies. One possible mechanistic explanation for our finding that PEPT1⫺/⫺ mice are protected from diet-induced obesity may relate to the coupling of epithelial proton flux via the sodiumproton exchanger NHE3 to PEPT1 activity. NHE3 is essential for PEPT1 activity by exporting protons back to the lumen, leading to the recovery of pHin from the acid load. As reported previously, PEPT1-deficient mice have impaired intestinal fluid absorption that is mediated via NHE3 in concert with a chloride-bicarbonate exchange system (5). This impairment in water absorption could also contribute to the observed dietindiced obesity (DIO) resistance phenotype. Most interestingly, mice lacking NHE3 also display reduced body weight, increased gut length, increased mass of cecum and colon (28), and thus very similar phenotypic changes as in PEPT1⫺/⫺ mice

on HFD. However, whether the DIO resistance phenotype in PEPT1⫺/⫺ mice on HFD is partially driven by an impaired function of NHE3 is still under analysis. In summary, Pept1⫺/⫺ mice submitted to HFD to induce obesity revealed a resistance with impaired weight gain that derives from a slightly reduced food intake but mainly originates from a limited capacity to digest and absorb the fat in the small intestine with an increased fecal energy loss. A mild malabsorption was similarly observed when animals were fed a HCD, but this appears to be compensated by increased bacterial fermentation and cecal biomass with extraction and delivery of energy to the host in the form of short-chain fatty acids. The main origin for all alterations observed seems the inability of Pept1⫺/⫺ mice to increase the surface area in the upper small intestine that occurs on a HFD. This was associated with markedly reduced systemic levels of IL-6, which was recently shown to be a prominent growth-promoting cytokine in the intestine. Although the cause of these differences in systemic IL-6 levels in Pept1⫺/⫺ mice remains to be determined, PEPT1-deficient animals appear generally as less flexible in adapting to diets leading to quite different phenotypes on diets rich in carbohydrates, protein, or high fat. ACKNOWLEDGMENT We thank Ronny Scheundel, Christine Schulze, Pia Röder, Pieter Giesbertz, Johanna Welzhofer, and Margot Siebler for excellent technical assistance. GRANTS This work was supported by the Deutsche Forschungsgemeinschaft as part of the Graduiertenkolleg 1482. DISCLOSURES No conflicts of interest exist for any of the authors listed above. AUTHOR CONTRIBUTIONS Author contributions: D.K., B.S., T.Z., and H.D. conception and design of research; D.K., R.P., J.K., T.S., C.S., and T.Z. performed experiments; D.K., B.S., R.P., J.K., T.S., K.G., C.S., T.Z., and H.D. analyzed data; D.K., B.S., R.P., J.K., K.G., T.Z., and H.D. interpreted results of experiments; D.K., B.S., R.P., J.K., T.S., K.G., C.S., T.Z., and H.D. prepared figures; D.K. drafted manuscript; B.S., R.P., J.K., T.Z., and H.D. edited and revised manuscript; B.S., T.Z., and H.D. approved final version of manuscript. REFERENCES 1. Balint JA, Fried MB, Imai C. Ileal uptake of oleic acid: evidence for adaptive response to high fat feeding. Am J Clin Nutr 33: 2276 –2280, 1980. 2. Biegel A, Gebauer S, Hartrodt B, Brandsch M, Neubert K, Thondorf I. Three-dimensional quantitative structure-activity relationship analyses of beta-lactam antibiotics and tripeptides as substrates of the mammalian H⫹/peptide cotransporter PEPT1. J Med Chem 48: 4410 –4419, 2005. 3. Bolze F, Rink N, Brumm H, Kühn R, Mocek S, Schwarz AE, Kless C, Biebermann H, Wurst W, Rozman J, Klingenspor M. Characterization of the melanocortin-4-receptor nonsense mutation W16X in vitro and in vivo. Pharmacogenomics J. doi:10.1038/tpj.2011.43. 4. Carl SM, HRD, Bhardwaj RK, Gudmundsson OS, Knipp G. Mammalian oligopetide transporters. In: Drug Transporters: Molecular Characterization and Role in Drug Disposition, edited by You G, Morris ME. Hoboken, NJ: Wiley, 2007. 5. Chen M, Singh A, Xiao F, Dringenberg U, Wang J, Engelhardt R, Yeruva S, Rubio-Aliaga I, Nässl AM, Kottra G, Daniel H, Seidler U. Gene ablation for PEPT1 in mice abolishes the effects of dipeptides on small intestinal fluid absorption, short-circuit current, and intracellular pH. Am J Physiol Gastrointest Liver Physiol 299: G265–G274, 2010.

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00160.2012 • www.ajpgi.org

PEPT1-DEFICIENT MICE ON HIGH-FAT DIET 6. Collins S, Martin TL, Surwit RS, Robidoux J. Genetic vulnerability to diet-induced obesity in the C57BL/6J mouse: physiological and molecular characteristics. Physiol Behav 8: 243–248, 2004. 7. Daniel H, Fett C, Kratz A. Demonstration and modification of intervillous pH profiles in rat small intestine in vitro. Am J Physiol Gastrointest Liver Physiol 257: G489 –G495, 1989. 8. Daniel H. Molecular, and integrative physiology of intestinal peptide transport. Annu Rev Physiol 66: 361–384, 2004. 9. de Wit NJ, Bosch-Vermeulen H, de Groot PJ, Hooiveld GJ, Bromhaar MM, Jansen J, Müller M, van der Meer R. The role of the small intestine in the development of dietary fat-induced obesity and insulin resistance in C57BL/6J mice (Abstract). BMC Med Genomics 1: 14, 2008. 10. Ganapathy V, Leibach FH. Role of pH gradient and membrane potential in dipeptide transport in intestinal and renal brush-border membrane vesicles from the rabbit. Studies with L-carnosine and glycyl-L-proline. J Biol Chem 258: 4189 –4192, 1983. 11. Ganapathy V, Brandsch ME, Leibach FH. Intestinal transport of amino acids and peptides. In: Physiology of the Gastrointestinal Tract, edited by Johnson LR. New York, NY: Raven, 1994. 12. Goda T, Takase S. Effect of dietary fat content on microvillus in rat jejunum. J Nutr Sci Vitaminol 40: 127–136, 1994. 13. Hornby PJ, Moore BA. The therapeutic potential of targeting the glucagon-like peptide-2 receptor in gastrointestinal disease. Expert Opin Ther Targets 15: 637–646, 2011. 14. Howard A, Goodlad RA, Walters JR, Ford D, Hirst BH. Increased expression of specific intestinal amino acid and peptide transporter mRNA in rats fed by TPN is reversed by GLP-2. J Nutr 134: 2957–2964, 2004. 15. Hu Y, Smith DE, Ma K, Jappar D, Thomas W, Hillgren KM. Targeted disruption of peptide transporter Pept1 gene in mice significantly reduces dipeptide absorption in intestine. Mol Pharm 5: 1122–1130, 2008. 16. Jin X, Zimmers TA, Zhang Z, Pierce RH, Koniaris LG. Interleukin-6 is an important in vivo inhibitor of intestinal epithelial cell death in mice. Gut 59: 186 –196, 2010. 17. Koopman JP, Kennis HM, van Druten JA. Colonization resistance of the digestive tract and gastrointestinal transit time in SPF mice. Lab Anim 12: 223–226, 1978. 18. Kottra G, Stamfort A, Daniel H. PEPT1 as a paradigm for membrane carriers that mediate electrogenic bidirectional transport of anionic, cationic, and neutral substrates. J Biol Chem 277: 32683–32691, 2002. 19. Litvak DA, Hellmich MR, Evers BM, Banker NA, Townsend CM Jr. Glucagon-like peptide 2 is a potent growth factor for small intestine and colon. J Gastrointest Surg 2: 146 –150, 1998. 19a.Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2-⌬⌬CT method. Methods 25: 402–408, 2001.

G907

20. Morgenthal K, Wienkoop S, Scholz M, Selbig J, Weckwerth W. Correlative GC-TOF-MS based metabolite profiling and LC-MS based protein profiling reveal time-related systemic regulation of metaboliteprotein networks and improve pattern recognition for multiple biomarker selection. Metabolomics 1: 109 –121, 2005. 21. Nässl AM, Rubio-Aliaga I, Fenselau H, Marth MK, Kottra G, Daniel H. Amino acid absorption and homeostasis in mice lacking the intestinal peptide transporter PEPT1. Am J Physiol Gastrointest Liver Physiol 301: G128 –G137, 2011. 22. Nässl AM, Rubio-Aliaga I, Sailer M, Daniel H. The intestinal peptide transporter PEPT1 is involved in food intake regulation in mice fed a high-protein diet. PLoS One 6: e26407, 2011. 23. Petit V, Arnould L, Martin P, Monnot MC, Pineau T, Besnard P, Niot I. Chronic high-fat diet affects intestinal fat absorption and postprandial triglyceride levels in the mouse. J Lipid Res 48: 278 –287, 2007. 24. Rowland KJ, Brubaker PL. The “cryptic” mechanism of action of glucagon-like peptide-2. Am J Physiol Gastrointest Liver Physiol 301: G1–G8, 2011. 25. Rubio-Aliaga I, Daniel H. Mammalian peptide transporters as targets for drug delivery. Trends Pharmacol Sci 23: 434 –440, 2002. 26. Sagher FA, Dodge JA, Johnston CF, Shaw C, Buchanan KD, Carr KE. Rat small intestinal morphology and tissue regulatory peptides: effects of high dietary fat. Br J Nutr 65: 21–28, 1991. 27. Schemmel R, Mickelsen O, Gill JL. Dietary obesity in rats: body weight and body fat accretion in seven strains of rats. J Nutr 100: 1041–1048, 1970. 28. Schultheis PJ, Clarke LL, Meneton P, Miller ML, Soleimani M, Gawenis LR, Riddle TM, Duffy JJ, Doetschman T, Wang T, Giebisch G, Aronson PS, Lorenz JN, Shull GE. Renal and intestinal absorptive defects in mice lacking the NHE3 Na⫹/H⫹ exchanger. Nat Genet 19: 282–285, 1998. 29. Steiner HY, Naider F, Becker JM. The PTR family: a new group of peptide transporters. Mol Microbiol 16: 825–834, 1995. 30. Surwit RS, Kuhn CM, Cochrane C, McCubbin JA, Feinglos MN. Diet-induced type II diabetes in C57BL/6J mice. Diabetes 37: 1163–1167, 1988. 31. Tocchetti A, Soppo CB, Zani F, Bianchi F, Gagliani MC, Pozzi B, Rozman J, Elvert R, Ehrhardt N, Rathkolb B, Moerth C, Horsch M, Fuchs H, Gailus-Durner V, Beckers J, Klingenspor M, Wolf E, Hrabé de Angelis M, Scanziani E, Tacchetti C, Scita G, Di Fiore PP, Offenhäuser N. Loss of the actin remodeler Eps8 causes intestinal defects and improved metabolic status in mice. PLoS One 5: e9468, 2010. 32. Wallis K, Walters JR, Gabe S. Short bowel syndrome: the role of GLP-2 on improving outcome. Curr Opin Clin Nutr Metab Care 12: 526 –532, 2009.

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00160.2012 • www.ajpgi.org