Free fatty acid receptor 2 (Ffar2), also known as GPR43, is activated by short-chain fatty acids (SCFA) and expressed in intestine, adipocytes, and immune cells, suggesting involvement in lipid and immune regulation. In the present study, Ffar2-deficient mice (Ffar2-KO) were given a high-fat diet (HFD) or chow diet and studied with respect to lipid and energy metabolism. On a HFD, Ffar2-KO mice had lower body fat mass and increased lean body mass. The changed body composition was accompanied by improved glucose control and lower HOMA index, indicating improved insulin sensitivity in Ffar2-KO mice. Moreover, the Ffar2-KO mice had higher energy expenditure accompanied by higher core body temperature and increased food intake. The liver weight and content of triglycerides as well as plasma levels of cholesterol were lower in the Ffar2-KO mice fed a HFD. A histological examination unveiled decreased lipid interspersed in brown adipose tissue of the Ffar2-KO mice. Interestingly, no significant differences in white adipose tissue (WAT) cell size were observed, but significantly lower macrophage content was detected in WAT from HFD-fed Ffar2-KO compared with wild-type mice. In conclusion, Ffar2 deficiency protects from HFD-induced obesity and dyslipidemia at least partly via increased energy expenditure.
- G protein-coupled receptor 43
- energy expenditure
- food intake
- feces energy
- body composition
- white adipose tissue inflammation
the g protein-coupled receptors (GPCRs/GPRs) belong to the largest receptor families known today. A great variety of molecules, including biogenic amines, amino acids, proteins, fatty acids, lipids, nucleotides, and ions, serve as activating ligands for different GPCRs (22). This wide variety of ligands, the importance of GPCRs in many different biological processes, and involvement in many disease conditions have made this class of receptors important and useful targets for the pharmaceutical industry, and GPCRs are presently the target of an estimated 45% of marketed drugs (7).
GPR40, -41, -42, and -43 were first identified in cosmids, isolated from human chromosomal locus 19q13.1 as tandem encoded genes, and were found to be related members of a homologous family of orphan GPCRs (18). In mice, the gene encoding Gpr43 is situated on chromosome 7, and Gpr43 was renamed to free fatty acid receptor 2 (Ffar2) (14). Other free fatty acid receptors were also renamed (Gpr40 to Ffar1 and Gpr41 to Ffar3).
Ffar2 and Ffar3 are activated by acetate and other short-chain carboxylic acid anions, e.g., formate, propionate, butyrate, and pentanoate, albeit with different specificity for carbon chain length (5). Short-chain fatty acids (SCFA) are the major anions in the large intestine and are produced by bacterial fermentation of dietary fibers (21). Interestingly, dietary starch results in increased plasma levels of acetate and propionate associated with reduced adipose free fatty acid (FFA) release and improved insulin sensitivity in humans (16). Ffar1, on the other hand, is activated by saturated and unsaturated carboxylic acids with carbon chains longer than six (4).
Ffar2 is highly expressed in peripheral leukocytes and to a lesser extent in spleen, bone marrow, and fetal liver (5). Expression of Ffar2 is induced by leukocyte differentiation to monocytes and neutrophils, suggesting a role of Ffar2 in the activation process of leukocytes (19). In addition, Ffar2 expression is observed in rat and human colon wall, in mucosa, and in enteroendocrine cells expressing peptide YY (Pyy) (11, 21). Interestingly, SCFA are known to stimulate the release of Pyy from ileum and colon (6). Moreover, Ffar2 mRNA expression is detected in rat intestinal mucosa and whole distal ileum wall in smooth muscle and submucosal layers (11). Expression of Ffar2 mRNA is also observed in different adipose tissue depots, including subcutaneous, perirenal, mesenteric, and epididymal white adipose tissue (WAT), and Ffar2 expression is upregulated in adipose tissue in mice fed a high-fat diet (HFD) (10). Undifferentiated 3T3-L1 adipocytes have no baseline Ffar2 expression, but expression levels increase during cell differentiation (10). Propionate increases the expression of both Ffar2 and Pparγ2 (10), and reduced Ffar2 expression inhibits adipose differentiation, clearly showing an involvement of Ffar2 in adipocyte differentiation (10, 23).
The role of Ffar2 in the regulation of energy balance is poorly understood. Adipocytes treated with Ffar2 ligands show a reduced lipolytic activity, and in vivo infusion of SCFA reduced plasma levels of FFA. These effects were abolished in Ffar2-deficient mice (8). It has been stated that no gross developmental phenotype or effects on body weight could be detected in the Ffar2-deficient mice when they were fed normal chow diet (8). However, less is known about the phenotype of Ffar2-deficient mice under environmental stress conditions, e.g., by giving a HFD.
To study the in vivo importance of Ffar2 in different aspects of the energy balance, a novel Ffar2-deficient (Ffar2-KO) mouse line was generated. In this article, the effects of HFD on energy, glucose, and lipid metabolism were studied in Ffar2-KO mice compared with their littermate wild-type (WT) controls.
MATERIALS AND METHODS
Animals and diets.
Heterozygous Ffar2-knockout mice (Ffar2+/−) were obtained from Deltagen (San Carlos, CA). A schematic diagram of the mouse Ffar2 gene, including the disrupted region, is outlined in Fig. 1. The Ffar2 gene was targeted by homologous recombination in embryonic stem cells derived from 129/OlaHsd mouse substrain and injected into blastocysts. Mice carrying a disrupted Ffar2 gene were generated by heterozygous (Ffar2+/−) breeding to C57BL/6JOlaHsd mice (Harlan, Horst, The Netherlands). Backcrossing to C57BL/6JOlaHsd was done to produce generation N7 offspring. Heterozygous N7 intercross was performed to produce Ffar2-KO mice and WT littermate control mice that were studied. Genotyping was done by a PCR strategy from genomic DNA preparations. Three primers were used: one located upstream of the deleted region (5′-GCGGAAGTTGGATGCTGCTTCCACG-3′), one located within the deleted region (5′-GCACAGTTCCTTGATCCTCACGGCC-3′), and one located in the targeting cassette (5′-GGGCCAGCTCATTCCTCCCACTCAT-3′).
Male Ffar2-KO and WT mice were allocated in groups balanced by body weight and housed individually in a temperature-controlled room (22°C) with a 12:12-h light-dark cycle. They had access to normal chow diet (R36; Lactamin, Stockholm, Sweden) and water ad libitum. The R36 chow diet contains 3.5% cellulose (%weight), 22.9% protein (%energy), 67.1% carbohydrate, and 9.6% fat. The main sources of proteins are from soy, grain, and potatoes. Carbohydrate source is mainly grains, and the main fat source is soy. The energy density of R36 is 12.6 kJ/g (according to supplier). During the HFD experiments, the HFD used was R638 (Lactamin) containing 3.9% cellulose (%weight), 39.9% fat (%energy), 42.3% carbohydrates, and 17% protein with a total energy content of 15.6 kJ/g and supplemented with 0.15% cholesterol. The main source of proteins is casein, the carbohydrate source is mainly corn starch, and the glucose and fat source is cacao butter. The experiments were performed according to the approved ethical application of the Gothenburg Ethics Committee for Experimental Animals.
Body weight, food intake, and indirect calorimetry.
The mice were given either standard chow or HFD from 4 wk of age and onward. The body weight was recorded weekly from 5 wk of age. Food intake was analyzed over 48 h in food-deprived mice (12 h), as described previously (3), with a minor modification; no initial incubation (80°C for 1 h) of the cages was done due to the heat sensitivity of the HFD. Total feces produced over the measurement period were collected, and energy content of the feces was determined with a bomb calorimeter (C 5000; IKA Werke). Indirect calorimetry was performed at room temperature (22°C), as described previously, (99).
Body temperature and body composition.
Rectal core body temperatures were recorded in conscious nonanesthetized mice at day time (10–11 AM) using a rectal probe (9). Body composition was analyzed by dual energy X-ray absorptiometry in isoflurane-anesthetized mice, as described previously (9).
Oral glucose tolerance test and termination.
Oral glucose tolerance test was performed in 30-wk-old Ffar2-KO and WT mice, as described previously (1). The mice were terminated at 40 wk of age. Before euthanization, they were starved 4 h before blood glucose levels were measured (Accu-Check device; Roche Diagnostics, Mannheim, Germany). The mice were then anesthetized by isoflurane inhalation, and blood was collected in EDTA-coated tubes by cardiac puncture. Blood plasma was separated by centrifugation (2,500 rpm, 10 min, 4°C) and snap-frozen in liquid nitrogen. Epididymal WAT, retroperitoneal WAT, brown adipose tissue (BAT), liver, and heart was blotted and weighed. Tissue samples of epididymal WAT and BAT were placed in formalin for histological analysis. A liver tissue sample was taken for hepatic triglyceride content analysis. In addition, tissue samples of epididymal WAT, liver, hypothalamus, and BAT were snap-frozen in liquid nitrogen for expression analysis.
Plasma insulin and adiponectin were measured using a radioimmunoassay (RIA; Linco Research, St. Charles, MO) and plasma leptin with an ELISA from Chrystal Chem (Downers Grove, IL). Total plasma thyroxine (T4) and triiodothyronine (T3) levels were determined by using RIA (Coat-A-Count; Diagnostic Products, Los Angeles, CA). Corticosterone was measured using a RIA kit (Amersham Biosciences, Uppsala, Sweden). Plasma levels of alanine aminotransferase (ALAT) and albumin were measured with enzymatic colorimetric assays (Randox Laboratories, Crumlin, UK). Plasma cholesterol and triglyceride levels were measured with enzymatic colorimetric assays (Roche Diagnostics). The cholesterol distribution profiles were measured using a size exclusion high-performance liquid chromatography system, SMART, with column Superose 6 PC 3.2/30 (Amersham Pharmacia Biotech, Uppsala, Sweden), as described before (12). The lipoproteins in a 10-μl sample were separated within 60 min, and the area under the curves (AUC) represents the cholesterol content. Hepatic triglyceride content was measured after homogenization of the liver tissue sample in isopropanol (1 ml/50 mg), incubation (4°C, 1 h), and centrifugation (2,500 rpm, 5 min).
Expression level analysis.
RNA extraction, cDNA synthesis, and quantification by Taqman real-time PCR was performed as described previously (2). Gene-specific primer and probe sequences are given in Table 1. Taqman assays for Adipoq, Ffar2, and Ffar3 were purchased as assay on demand (Applied Biosystems, Foster City, CA).
Fresh WAT and BAT depots were fixed in 4% paraformaldehyde, dehydrated, embedded in paraffin, sectioned at 6 μm, deparaffinized, and rehydrated on slides. Macrophage content was measured on randomly selected sections from epididymal and retroperitoneal WAT depots. Briefly, tissue sections subjected to microwave antigen retrieval in 10 mM citrate buffer (pH 6.0) were stained (Techmate 500 Plus; DakoCytomation, Glostrup, Denmark) with the primary rat anti-mouse antibody Mac2/galectin3 (1:20,000, cl. M3/38; Cedarlane Laboratories) using the ABC method and counterstained by Mayer's Hematoxylin. Slides were scanned using the Mirax slide scanner (Carl Zeiss) and Mac2/Galectin3 staining was quantified using the BIOPIX software (BioPix, Gothenburg, Sweden). BAT was stained with hematoxylin and eosin. Representative pictures from BAT specimens were taken at ×20 magnification, and BAT density was determined using the Picsara software package (Euromed Networks, Stockholm, Sweden). For cell size analysis, rehydrated sections from epididymal and retroperitoneal fat depots were stained using 0.5% Basic Fuchsin (FLUKA, Buchs, Switzerland). Mounted sections were scanned, and automated determination of adipose tissue cell cross-sectional area (CSA) was performed on whole sections (∼1 × 1 cm) using the BIOPIX software. Areas between 1,000 and 12,000 μm2 were considered relevant and included in the calculation of mean CSA. For each WAT depot, morphometric data were obtained from >3,000 adipocytes.
All values are given as group means ± SE. Comparison between two groups was done by Student's t-test, and comparisons of four groups were done by one-way ANOVA followed by Bonferroni post hoc test for pairwise comparison. Parameters over time, e.g., energy expenditure and respiratory exchange ratio (RER), were analyzed by a two-way ANOVA by using the SPSS as computer software. The body weight data were analyzed by a separate two-way ANOVA model and a repeated measures analysis with age and group as factor, using SAS Proc Mixed as computer software. Values of P < 0.05 were considered significant. The data were log-normalized when appropriate.
Ffar2-KO mice were generated by targeted deletion of a 55-bp segment of the coding region of exon 3 and insertion of a reporter cassette (Fig. 1). To confirm absence of Ffar2 transcript in the Ffar2-KO mice, mRNA expression levels of Ffar2 were measured in liver biopsies. As expected, no expression of Ffar2 was observed in the Ffar2-KO mice (data not shown). Intercross of heterozygous Ffar2 (Ffar2+/−) mice resulted in offspring of normal litter sizes and a ratio of 49% male and 51% female offspring. Of the male offspring, 19.6% were homozygous, 51.5% were heterozygous, and 28.9% were WT mice. The female offspring distribution was 21.2% homozygous, 52.5% heterozygous, and 26.3% WT.
Body weight and body composition.
One group of male Ffar2-KO (n = 8) and male littermate WT control mice (n = 8) was given normal chow diet, whereas another group of male Ffar2-KO (n = 9) and male littermate WT control mice (n = 7) was given HFD from 4 wk of age.
As expected, the body weight gain was higher in both groups of mice given HFD compared with their respective control group receiving chow diet. The body weight was significantly higher from 9 wk of age and throughout the study period for the WT mice given HFD compared with WT on chow diet (Fig. 2). Ffar2-KO mice fed HFD had higher body weight after 12 wk of age compared with Ffar2-KO given the chow diet, but after 20 wk of age (15 wk on HFD), no significant difference in body weight was observed between Ffar2-KO mice given HFD and those given chow diet (Fig. 2).
At 13 wk of age, body composition analysis, including measurements of bone mineral content (BMC), bone mineral density (BMD), body lean mass, and body fat mass, did not reveal any significant differences between Ffar2-KO and WT mice given chow diet (data not shown). However, Ffar2-KO mice given a HFD had significantly higher absolute and relative body lean mass (relative to body length) compared with WT mice given HFD (Fig. 3A), but no significant difference in body fat mass (Fig. 3A), BMC, or BMD was observed between Ffar2-KO and WT mice on HFD at this age. The mice were reanalyzed at 33 wk of age. Again, no significant differences were observed between Ffar2-KO and WT mice given a chow diet with respect to BMD, BMC, body lean mass, or body fat mass (data not shown). However, HFD-fed Ffar2-KO mice had significantly higher relative body lean mass (Fig. 3B) and significantly lower body fat mass expressed as both absolute amount of fat and percentage of body weight compared with WT mice (Fig. 3B). Moreover, absolute BMC was decreased (WT: 0.496 ± 0.018 g; Ffar2-KO: 0.449 ± 0.013 g, P < 0.05), and relative (to body length) BMC tended to decrease (9% decrease, P = 0.08) in Ffar2-KO mice compared with the WT mice on HFD.
Food intake and fecal energy analysis.
Food intake was measured following a 12-h diet depletion over 48 h at 14–15 wk of age and then again at 37–38 wk of age.
At 14 wk of age, no significant difference in absolute or relative (to body weight) food intake was observed between Ffar2-KO and WT mice given chow diet. At 37 wk of age, Ffar2-KO mice had increased absolute food intake compared with WT mice when given a chow diet (WT: 9.08 ± 0.49 g/48 h; Ffar2-KO: 11.43 ± 0.61 g/48 h, P < 0.05). However, relative food intake was not significantly different between Ffar2-KO and WT mice given a chow diet.
At 15 wk of age, Ffar2-KO mice on the HFD had significantly higher absolute food intake compared with WT mice on the HFD (WT: 7.23 ± 0.29 g/48 h; Ffar2-KO: 8.22 ± 0.20 g/48 h, P < 0.05), but relative food intake was not significantly different between Ffar2-KO and WT mice on HFD at that age. However, at 38 wk of age, food intake was significantly higher in Ffar2-KO mice on HFD both in terms of absolute and relative (related to body weight) food intake compared with WT mice (Fig. 4A).
Feces production over 48 h and energy content in feces was determined from the feces samples collected at the end of the food intake measurement presented above. The Ffar2-KO mice produced significantly more feces compared with the WT mice fed the chow diet, 20% more at 14 wk of age (WT: 1.88 ± 0.13 g/48 h; Ffar2-KO: 2.25 ± 0.10 g/48 h, P < 0.05) and 61% more at 37 wk of age (WT: 1.69 ± 0.09 g/48 h; Ffar2-KO: 2.72 ± 0.23 g/48h, P < 0.05). Although no significant difference in fecal energy density (J/g feces) was observed at either age, the total fecal energy output tended to be higher at 14 wk of age (17% increase, P = 0.08) and was significantly higher at 37 wk of age for the chow-fed Ffar2-KO mice compared with the WT mice (WT: 26.97 ± 1.26 kJ/48 h; Ffar2 KO: 42.48 ± 3.93 kJ/48 h, P < 0.01).
Already at 15 wk of age, Ffar2-KO mice on the HFD had significantly higher fecal production, fecal energy density, and total energy output compared with WT mice (Fig. 4B). These differences persisted and were even more pronounced at 38 wk of age mainly because of enhanced loss of fecal energy in the Ffar2-KO mice at a higher age (Fig. 4C). A calculation of absorbed energy (energy intake minus energy lost in feces) revealed that there was a slight trend toward increased absorbed energy in the Ffar2-KO mice compared with WT mice [WT: 108.5 ± 5.4 kJ; Ffar2-KO: 139.4 ± 16.3 kJ, P = 0.08 (data log-normalized)]. Thus, there was no sign of malabsorption in the Ffar2-KO mice. Rather, the increased fecal energy content was mainly a consequence of the increased food intake.
Indirect calorimetry and body temperature.
The mice were analyzed by indirect calorimetry, locomotor activity, and water intake at 11–12 wk of age and again at 34–35 wk of age.
No significant differences in RER, energy expenditure, locomotor activity, or water intake were observed in Ffar2-KO compared with WT control mice given the chow diet at any age (data not shown).
At 12 wk of age, HFD-fed Ffar2-KO mice had significantly higher RER (Fig. 5A), but no significant difference in energy expenditure between Ffar2-KO mice and WT control mice was observed (Fig. 5B). Water intake over 72 h was significantly increased in Ffar2-KO compared with WT mice (WT: 6.2 ± 0.3 ml/72 h; Ffar2-KO: 8.3 ± 0.7 ml/72 h, P < 0.05) and also in relation to individual body weight (WT: 0.18 ± 0.01 ml·g−1·72 h−1; Ffar2-KO: 0.24 ± 0.02 ml·g−1·72 h−1, P < 0.05). Locomotor activity was not significantly different between Ffar2-KO and WT mice (data not shown) either in the metabolic cages (day and night activity measurements) or in the open-field activity measurement (daytime activity measurements).
At 35 wk of age, RER tended to be higher in the HFD-fed Ffar2-KO mice group compared with the HFD-fed WT mice, being significant during the s 24-h period (Fig. 5C). The Ffar2-KO mice had significantly higher energy expenditure compared with WT mice when fed the HFD (Fig. 5D). Water intake over 72 h was again significantly increased in Ffar2-KO compared with WT mice (WT: 6.5 ± 0.4 ml/72 h; Ffar2-KO: 12.2 ± 2.1 ml/72 h, P < 0.05) and also in relation to individual body weight (WT: 0.13 ± 0.01 ml·g−1·72 h−1; Ffar2-KO: 0.30 ± 0.07 ml·g−1·72 h−1, P = 0.051). Locomotor activity was not significantly different between Ffar2-KO and WT mice at this age in the metabolic cages (day and night activity measurements).
Body temperature was not significantly different at 12 wk of age but was significantly increased at 34 wk of age for the Ffar2-KO mice fed the chow diet compared with chow-fed WT mice (WT: 36.9 ± 0.1°C; Ffar2-KO: 37.5 ± 0.2°C, P < 0.05). At 13 wk of age, no significant difference in body temperature was observed between Ffar2-KO mice and WT mice on the HFD (Fig. 5E). However, Ffar2-KO mice had significantly higher body temperature than WT mice at 35 wk of age on HFD (Fig. 5E). To investigate whether the Ffar2-KO mice were cold resistant, the mice were placed in 6°C ambient temperature, and core body temperature was recorded after 30 and 60 min. However, no significant difference in core body temperature was observed between Ffar2-KO and WT mice fed the HFD following cold exposure.
Oral glucose tolerance test.
At 30 wk of age, glucose tolerance was analyzed following 5-h diet deprivation. No significant difference in integrated blood glucose or insulin AUC after oral glucose challenge was observed between Ffar2-KO and WT mice fed the chow diet (data not shown). However, Ffar2-KO mice given the HFD had significantly lower integrated blood insulin levels over the sample period of 120 min compared with WT mice (WT AUC: 592.3 ± 77.6; Ffar2-KO AUC: 365.3 ± 51.0, P < 0.05, Fig. 6A), whereas blood glucose levels were not significantly different between Ffar2-KO and WT mice (WT AUC: 1,835.1 ± 80.0; Ffar2-KO AUC: 1,614.6 ± 113.1, P = 0.15; Fig. 6B).
All data from blood and plasma analyses are collected in Table 2. After a 4-h fasting period, the animals were euthanized and blood was collected. Blood glucose levels were similar in 40-wk-old Ffar2-KO and WT mice fed the chow diet. No significant differences in plasma chemistry were observed between the two genotypes on chow diet.
At 40 wk of age, Ffar2-KO mice fed the HFD tended to have lower blood glucose and insulin levels (13 and 47% lower, respectively, compared with WT mice) compared with the WT mice, but the difference didn't reach statistical significance. Calculated homeostasis model assessment (HOMA) index was significantly lower in Ffar2-KO mice compared with WT mice on the HFD (WT:0.19 ± 0.03; Ffar2-KO: 0.09 ± 0.02, P < 0.05). Moreover, Ffar2-KO mice on the HFD had significantly lower plasma levels of albumin, ALAT, and cholesterol than WT control animals. Lipoproteins were separated by using size exclusion chromatography. The lipoprotein profile indicated that cholesterol levels were increased in HDL-LDL fractions by HFD in WT controls, whereas no obvious differences were observed in lipoprotein profile between HFD-fed and chow-fed Ffar2-KO mice (Fig. 7). In addition, plasma adiponectin levels were significantly higher, whereas leptin levels tended to be lower in the Ffar2-KO compared with WT mice (Table 2). Plasma levels of T3 and the T3/T4 ratio were significantly lower in the Ffar2-KO compared with WT mice.
Organ weight and histology.
The animals were killed at 40 wk of age. No significant difference was observed in body weight or the absolute or relative weight of the liver, epididymal WAT, or retroperitoneal WAT between the Ffar2-KO mice and the WT controls on chow diet. However, the absolute heart weight, but not the relative heart weight, was significantly higher for the Ffar2-KO compared with WT mice given the chow diet (Table 3). Several significant differences were observed between the HFD-fed Ffar2-KO and WT mice in terms of organ weights (Table 3). The HFD-fed Ffar2-KO mice had significantly lower weight of the liver and retroperitoneal WAT and significantly higher weight of the heart compared with HFD-fed WT control mice. In addition, hepatic triglyceride levels were significantly lower in the HFD-fed Ffar2-KO compared with WT mice (Table 3). The BAT in the intrascapular region was diffuse without a clear demarcation in both genotypes. Therefore, it was not possible to quantitatively dissect BAT to get an exact tissue weight. Histological examination of BAT showed a decreased amount of lipid droplets in the Ffar2-KO compared with WT mice fed the HFD (Fig. 8A).
Significantly lower macrophage content was detected in epididymal WAT from HFD-fed Ffar2-KO compared with WT mice, but no significant differences in WAT cell size between the genotypes were observed (Fig. 8B).
Gene expression level analysis.
All data from expression level analysis are presented in Table 4.
To investigate expression levels of genes that may correlate to the observed changes in energy expenditure and core body temperature, BAT expression of uncoupling protein 1 (Ucp1) mRNA and deiodinase iodothyronine type II (Dio2) were assessed. However, no significant difference was observed for these genes between Ffar2-KO and WT on either diet. Interestingly, there was a significant increase in Dio2 mRNA expression in Ffar2-KO on HFD compared with Ffar2-KO on the chow diet, whereas the expression was not changed by HFD in WT mice. Epididymal WAT expression of adiponectin was significantly lower in WT mice on HFD compared with WT mice on the chow diet as expected. However, no such down regulation of adiponectin expression was observed in Ffar2-KO by HFD. Interestingly, WAT expression level of Ffar3 (a. k. a. Gpr41) was significantly increased in the Ffar2-KO mice on the chow diet compared with WT mice on the chow diet. This difference was not observed in Ffar2-KO and WT mice on HFD. There were also significantly lower Ffar3 levels in the Ffar2-KO mice on the HFD compared with Ffar2-KO mice on the chow diet in WAT. Hypothalamic expression levels of neuropeptide Y (Npy) and proopiomelanocortin (Pomc) were determined to find changes that could explain the increased food intake in the Ffar2-KO mice. However, no significant differences were observed between Ffar2-KO and WT mice with respect to expression level of these genes. The hepatic triglyceride levels and plasma lipid levels were lower in the Ffar2-KO mice than in the WT controls on HFD. To investigate the potential mechanisms behind this phenotype of the Ffar2-KO mice, expression of genes involved in the regulation of hepatic lipid metabolism was investigated. Hepatic expression level of acetyl-coenzyme A carboxylase-α (Acaca), interleukin-6 (Il-6), peroxisome proliferator activated receptor-γ (Pparg), fatty acid synthase (Fasn), diacylglycerol O-acyltransferase 1 (Dgat1), and stearoyl-coenzyme A desaturase 1 (Scd1) was determined. However, neither levels of these genes were significantly different between Ffar2-KO and WT mice fed the HFD.
In the present study, we have investigated the in vivo importance of Ffar2 by studying Ffar2-KO mice fed either a normal chow diet or a HFD over a 35-wk period. The Ffar2-KO mice on the chow diet showed very few phenotypic alterations compared with chow-fed WT littermate controls. These differences included increased gene expression of Ffar3 in WAT as well as increased body temperature in Ffar2-KO mice compared with WT littermate control mice. However, on HFD several phenotypic differences between Ffar2-KO and littermate controls were revealed, indicating the importance of Ffar2 for adaptation to increased dietary intake of fat. Most importantly, the Ffar2-KO mice were protected from HFD-induced obesity and its consequences on glucose and lipoprotein metabolism. Strikingly, the effect of Ffar2 deficiency was late and occurred after more than 10 wk of HFD. After 8–11 wk of HFD, Ffar2-KO mice showed increased lean body mass and RER, but no significant differences in energy intake, energy expenditure, or body weight gain compared with littermate controls were observed. Thus, the earliest detectable effects of Ffar2 deficiency were increased lean body mass and slightly increased use of carbohydrates as energy source. After 29 wk on the HFD, the differences between the genotypes were more prominent. Lean body mass was still higher, and body fat mass was lower in the Ffar2-KO animals. These differences were associated with increased energy intake, energy expenditure, core body temperature, and RER. In addition to increased lean body mass, heart weight was significantly increased, indicating together with the effect on lean body mass that loss of Ffar2 signaling results in anabolic effects in muscle tissues.
Interestingly, WAT showed unaltered fat cell size compared with the more obese littermate controls. The unaltered adipocyte size and lower body fat mass indicates that the total number of adipocytes was reduced in the Ffar2-KO mice. This finding is in line with previous observations that Ffar2 is involved in adipocyte differentiation (10). In models of reduced adipocyte differentiation, fat will accumulate in other tissues such as liver and muscle, creating insulin resistance and diabetes (15). However, the liver of Ffar2-KO mice showed reduced triglyceride content compared with littermate controls. It is very likely that the Ffar2-KO mice are protected from ectopic fat on the HFD because of the increased energy expenditure. The increased energy expenditure is explained mainly by increased body temperature since locomotor activity was unchanged. BAT showed features of increased activity in the Ffar2-KO mice, but without increased expression of key genes in energy consumption such as Ucp1 and Dio2. Therefore, the increased BAT activity is governed by other physiological alterations involving Ffar2. Although the Ffar2-KO mice showed increased energy expenditure, they showed lower plasma levels of T3 and T3/T4 ratio, indicating reduced thyroid receptor signaling. Therefore, the finding of reduced T3 could be interpreted as a compensatory change.
It has previously been shown that Ffar2 is responsible for the antilipolytic effect of SCFA in adipocytes (8). The antilipolytic effect of SCFA could be understood as negative feedback to reduce the lipolysis in adipose tissue as a consequence of increased bacterial production or increased ingestion of SCFA. In the present study, it was shown that adipose quality was improved as indicated by reduced macrophage content and increased mRNA expression and plasma levels of adiponectin. This finding shows that Ffar2 deficiency protects from HFD-induced adipose inflammation. This is especially interesting in light of recent observations that Ffar2-KO mice showed exacerbated or unresolving inflammation in models of colitis, arthritis, and asthma (13). Moreover, Ffar2-KO mice reacted more strongly after acute dextrane sodium sulfate (DSS)-induced inflammation in terms of body weight reduction and lethality. In another study, Ffar2-KO mice showed lower signs of inflammation, distal colon tissue damage, and inflammatory cell infiltration, although plasma levels of TNFα were elevated (20). Hence, Ffar2 has a dual role in DSS-induced colitis. The reason for the difference in response between HFD-induced adipose inflammation and induction of inflammation in colon, joints, and lungs is not clear from the present study but indicates another role for SCFA and Ffar2 in adipose inflammation. Therefore, the role of Ffar2 in metabolically induced inflammation merits further investigation.
There was a trend toward lower fasting levels of glucose and insulin as well as a significantly lower HOMA index in the Ffar2-KO mice compared with the WT littermate controls on HFD. Moreover, the oral glucose tolerance test showed unchanged glucose tolerance but markedly lower insulin levels, which may point toward improved insulin sensitivity. Likely, the reduced body fat, reduced adipose inflammation, and higher plasma adiponectin levels could help to explain the improved glucose control. Also, HFD-fed Ffar2-KO mice had high RER, indicating that they used a higher proportion of carbohydrates as energy source. This finding may also help to explain the improved glucose control. Although the effect on glucose control was small, additional direct effects of Ffar2 deficiency on insulin receptor signaling cannot be excluded.
An unexpected finding was the remarkably low liver weight of HFD-fed Ffar2-KO mice that could not be explained only by the reduced triglyceride content. Expression levels of hepatic genes involved in de novo lipogenesis were not affected in the Ffar2-KO mice, indicating that reduced de novo lipogenesis was not the explanation for the reduced hepatic triglyceride content. Moreover, reduced plasma levels of ALAT, albumin, and plasma lipids indicate that there are fundamental changes in liver function in Ffar2-KO. Together with the observation that Ffar2 is expressed in the fetal liver (5), these results indicate that Ffar2 signaling could be important for normal liver growth and development. In contrast to the WT animals, the Ffar2-KO animals were completely protected from changes in plasma lipoproteins induced by HFD. It is likely that unchanged liver fat content protects these mice from dyslipidemia.
Another interesting observation was that the upregulation of Ffar3 in WAT in the Ffar2-KO mice on the chow diet was absent in Ffar2-KO on HFD. This finding indicates that Ffar3 could partly compensate for the loss of Ffar2 since there were no major changes in body weight, body fat, or glucose or lipid homeostasis between Ffar2-KO and WT mice fed the chow diet. It could be speculated that the loss of Ffar3 upregulation in Ffar2-KO mice on HFD feeding may partly explain some of the observed phenotypic alterations in Ffar2-KO mice on the HFD.
The increased food intake in the Ffar2-KO mice was accompanied by increased fecal energy in the Ffar2-KO mice. Calculations of absorbed energy (energy intake minus feces energy) showed that there was no significant difference between the Ffar2-KO and WT mice on HFD, although there was a trend toward increased absorbed energy in the Ffar2-KO mice (28.5% increase, P = 0.082). Therefore, reduced body weight and body fat gain in the Ffar2-KO mice is explained by increased energy expenditure and does not involve malabsorption. Interestingly, Ffar3-deficient mice show higher fecal energy, which is explained by increased gastric transit rate (17).
In summary, this article expands the understanding of the importance of Ffar2 in energy and lipid metabolism. Ffar2-KO mice are similar to their WT littermates when fed a normal chow diet with a few exceptions, including upregulation of Ffar3 in adipose tissue. However, several phenotypic differences occur after 25–30 wk of HFD feeding, including lower body fat mass, improved glucose control, lower plasma lipids, and lower WAT inflammation. These changes were not due to reduced energy intake or absorption but rather increased energy expenditure in the Ffar2-KO mice. These data demonstrated the importance of the SCFA/Ffar2 signaling system for metabolic control. The results of this study indicate that further studies are needed to clarify mechanisms responsible for the importance of Ffar2 for body temperature control, metabolic inflammation, adiponectin production, the interplay between Ffar2 and -3 as well as the interaction between the immune system and microbiota-produced ligands for Ffar.
No conflicts of interest, financial or otherwise, are declared by the authors.
We acknowledge Kasra Afsarinejad for assistance with the statistical analysis and Anders Elmgren, Lena Amrot Fors, Anna Tuneld, Charlotte Lindgren, Anne-Cristine Carlsson, Martina Johansson, Gisela Häggblad, and Marie-Louise Berglund Zackrisson, who kindly helped with the plasma assays. In addition, we thank Marie Jönsson for assistance with the experiments. Alison Davis and Wendy Vernon are acknowledged for tissue gene expression analysis.
- Copyright © 2011 the American Physiological Society