Neurotensin differentially regulates bile acid metabolism and intestinal FXR-bile acid transporter axis in response to nutrient abundance

Abstract

Studies demonstrate a role for neurotensin (NT) in obesity and related comorbidities. Bile acid (BA) homeostasis alterations are associated with obesity. We determined the effect of NT on BA metabolism in obese and non-obese conditions. Plasma and fecal BA profiles were analyzed by LC-MS/MS in male and female NT^+/+ and NT^−/− mice fed low-fat (LFD) or high-fat diet (HFD) for 6 weeks (early stage of obesity) or greater than 20 weeks (late stage of obesity). The nuclear farnesoid X receptor (FXR) and BA transporter mRNA expression was assessed in ileum, mouse enteroids and human cell lines. HFD decreased plasma primary and secondary BAs in NT^+/+ mice; HFD-induced decrease of plasma BAs was improved in NT-deficient mice. In NT^+/+ mice, HFD inhibited ileal FXR and BA transporter expression; HFD-decreased expression of FXR and BA transporters was prevented in NT^−/− mice. Compared with LFD-fed NT^+/+ mice, LFD-fed NT^−/− mice had relatively lower levels of ileal FXR and BA transporter expression. Moreover, NT stimulates expression of FXR and BA transporters in Caco-2 cells; however, stimulated expression of BA transporters was attenuated in NT^−/− enteroids. Therefore, we demonstrate that HFD disrupts the BA metabolism and ileal FXR and BA transporter axis which are improved in the absence of NT, suggesting that NT contributes to HFD-induced disruption of BA metabolism and plays an inhibitory role in the regulation of ileal FXR and BA transporter signaling under obese conditions. Conversely, NT positively regulates the expression of ileal FXR and BA transporters under non-obese conditions. Therefore, NT plays a dual role in obese and non-obese conditions, suggesting possible therapeutic strategies for obesity control.

Neurotensin (NT), a tridecapeptide gut hormone released from intestinal enteroendocrine cells (1), facilitates fat ingestion and absorption (2), which in turn stimulates NT release in animals and humans (3, 4). NT is involved in other gastroenteropancreatic functions, including inhibition of gastric motor activity (5), stimulation of the exocrine pancreatic secretion (6) and pancreaticobiliary secretion (7). NT stimulates insulin secretion at low glucose concentration but inhibits glucose-induced insulin release (8). Recent studies demonstrate a critical role for NT and its receptors in human diseases, including obesity and its associated comorbid conditions. For example, Melander, et al. (9) reported that increased fasting plasma levels of pro-NT (a stable 117-amino acid fragment from the neurotensin precursor) are significantly associated with increased risk of diabetes, cardiovascular disease and mortality. Our laboratory has shown that NT deficiency attenuates the weight gain, hepatic steatosis and insulin resistance associated with feeding mice a high-fat diet (HFD) (10). Moreover, in human longitudinal studies among non-obese subjects, high levels of pro-NT denote a doubling of the risk of developing obesity later in life (10). These studies suggest a functional role for NT in whole body metabolism under physiologic and obese conditions.

Bile acids (BAs) have recently been considered as signaling molecules regulating glucose, lipid and energy homeostasis (11) and can be affected by gut hormones, including NT. For example, intravenous infusion of NT increases taurocholate but not cholate absorption (12). Moreover, pediatric patients with liver disease associated with intestinal failure have lower serum NT concentrations (13). The levels of NT mRNA in the ileum of patients positively correlates with apical sodium dependent bile acid transporter (ASBT, also called SLC10A2) mRNA levels (13). In mice and in cultured intestinal cells, NT treatment stimulates the expression of ASBT and leads to increased BA uptake via NT receptor 1 (NTR1) and NTR3 (also called sortilin 1) (13). Cholesterol 7α-hydroxylase (CYP7A1) catalyzes the initial and rate-limiting step in the classical pathway and produces cholic acid (CA) and chenodeoxycholic acid (CDCA) (14). Sterol-12α-hydroxylase (CYP8B1) catalyzes the biosynthesis of CA. In mice, CDCA is converted to α-muricholic acid (α-MCA); ursodeoxycholic acid (UDCA), a primary BA in mice (15), is converted to β-muricholic acid (β-MCA) (11). Sterol-27-hydroxylase (CYP27A1) mediates the alternative BA biosynthesis pathway and synthesizes CDCA in humans or muricholic acid in mice through oxysterol 7α-hydroxylase (16). In humans, about 95% of primary BAs are reabsorbed in the ileum and the remainder reach the colon, where they are de-conjugated to secondary BAs by intestinal bacteria, and CA is converted to deoxycholic acid (DCA) and CDCA to lithocholic acid (LCA), 2 major secondary BAs in humans. In mice, CDCA is converted to ɷ-MCA. Secondary BAs are reabsorbed and in humans, about 0.5 g of BAs is secreted into the feces per day (17).

Farnesoid X receptor (FXR) is a major regulator of BA homeostasis and enterohepatic circulation (18). Sinal, et al. demonstrated that the expression of CYP7A1 and CYP8B1 genes is negatively regulated in an FXR-dependent manner. Elevated BAs in hepatocytes activate FXR, resulting in suppression of BA biosynthesis, enhanced BA transport to the small intestine and reduced BA uptake from the blood (19). Fxr null mice fed 1% CA diet had elevated serum BAs but reduced BA pools and fecal BA excretion due to decreased expression of the major hepatic canalicular BA transport protein (20). In a recent study, HFD feeding in wild-type and whole body Fxr-deficient germ-free and conventionally raised mice revealed that microbiota-induced obesity requires functional FXR signaling (21). Conventionally raised wild-type mice gained significantly more weight than germ-free wild-type mice after 10 weeks on a HFD. In the absence of FXR, the gut microbiota failed to affect weight gain, adipose tissue inflammation and hepatic steatosis.

Here, we used our NT deficient mouse model to test the hypothesis that HFD feeding disrupts BA homeostasis, which is prevented or significantly attenuated by the absence of NT. In addition, transcriptional profiles of ileal FXR and BA transporters were investigated.

Experimental Procedures

Materials and Methods

Reagents.

Fetal bovine serum (FBS), DMEM, Minimum Essential Medium Eagle (MEM), CDCA, n-acetylcysteine (NAC), [Leu]15-Gastrin 1, NT peptide and β-actin antibody were from Sigma-Aldrich (St. Louis, MO). GW4064 was from Tocris (Minneapolis, MN). The phospho ERK1/2 and ERK1/2 antibodies were from Cell signaling (Danvers, MA). FXR and NTR1 antibodies were from Santa Cruz (Dallas, TX). Noggin-conditioned medium was purchased from U-Protein Express BV (Netherlands). Advanced DMEM/F12 medium, OptiMEM Reduced Serum Medium, growth factor–reduced Matrigel, B-27 Supplement, N-2 Supplement, HEPES, GlutaMAX and Zeocin were from ThermoFisher (Grand Island, NY). Mouse EGF was from PeproTech (Rocky Hill, NJ).

Mice

All procedures were carried out according to protocols approved by the IACUC at the University of Kentucky. NT^−/− mice and their wild type littermates (NT^+/+) were bred from NT^+/− mice (22) and randomly grouped for the experiment. For induction of obesity, male mice were fed a 60% HFD or as a control, 10% low-fat diet (LFD) (catalogue no. D12492 and D12450B, respectively; Research Diets, New Brunswick, NJ) at weaning for 5–6 weeks or 26–28 weeks, as described in our previous studies (10, 23). Mice were housed in standard cages in temperature-controlled environments under a 12/12 h light-dark cycle with ad libitum access to food and water. Prior to sacrifice, feces were collected from individually caged mice and stored at −80°C. Mice were fasted overnight and anesthetized with isoflurane inhalation. Blood was collected from the inferior vena cava in K₂EDTA blood collection tubes from Fisher Scientific (Lenexa KS). Plasma was obtained by centrifuging the blood at 8,000 × g for 10 min at 4°C and aliquots stored in −80°C.

BA analysis.

Tissue extraction was performed as described previously (24). Fecal samples (15 mg) were placed on dry ice and extracted twice with 500 μl cold methanol. The 2 extracts were combined, dried and reconstituted in 100 μl methanol:acetonitrile (50:50, v/v) with internal standards (Sigma-Aldrich, St. Louis, MO). A 50 μl plasma sample was processed and quantified by LC-MS/MS.

Mouse intestinal crypt isolation and three dimensional culture.

Small intestinal crypts were isolated from either NTR1^−/− or NT^−/− mice as well as their wild-type littermates (8-week-old) remained on normal chow as previously described with modifications (25). Briefly, the small intestine from the pylorus to the ileocecal valve was dissected. After 5 cm from the pylorus was removed, the rest of the small intestine was divided into 2 equal sections; after 3 cm from the fold point in each fragments was further removed, the rest proximal segment was designated as jejunum and the distal ileum. The crypts were isolated from either jejunal or ileal fragments with 2 mM EDTA/PBS (pH 7.4) and embedded in complete Matrigel overlaid by ENR medium [Advanced DMEM/ F12 supplemented with B27, N2, NAC, HEPES, GlutaMAX (basal medium) containing EGF (50 ng/ml), noggin-conditioned medium (1:100) and R-spondin 1 conditioned medium (1:20) (ENR medium)]. Mature enteroids were subseeded every 4–7 days. Enteroids were generated from 3 separate animals for each condition, generating 3 biological replicates. R-spondin conditioned medium was made from Cultrex HA-R-Spondin1-Fc 293T cells, purchased from Trevigen (Gaithersburg, MD) following the instructions provided by the manufacturer.

Caco-2 cells.

Human Caco-2 cell line was purchased from ATCC. Caco-2 cells were cultured in MEM containing 0.1 mM nonessential amino acids, 1.5 g/L sodium bicarbonate and 0.75 g/L sodium pyruvate supplemented with 10% FBS. To induce the differentiation of Caco-2 cells, cells were seeded at a density of 5×10⁴/cm² and the medium changed every 2 days after the cells reached confluence.

RNA extraction and quantitative reverse transcription PCR analysis.

The small intestine, after 5 cm from the pylorus was removed, was divided into 2 equal sections (the proximal half was designated as jejunum and the distal half ileum). Ileal mucosa was scraped with a glass slide and used for RNA analyses. For mouse ileal mucosal scrapings and Caco-2 cells, total RNA was extracted using RNeasy kit (Qiagen, Valencia, CA). Each cDNA was synthesized using High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA). For enteroids, total RNA isolation and cDNA synthesis were performed using TaqMan^™ Gene Expression Cells-to-CT^™ Kit (Thermo Fisher, Grand Island, NY). Quantitative PCR (qPCR) was performed using a TaqMan Gene Expression Master Mix and TaqMan probes (Thermo Fisher, Waltham, MA). Mouse β-actin and human GAPDH probes were used as the internal controls for mouse samples and Caco-2 cells, respectively. Expression levels were assessed by evaluating threshold cycle (Ct) values. The relative amount of mRNA expression was calculated by the comparative ΔΔCt method. Each sample was loaded into wells in triplicate in the qPCR assay.

Protein preparation and western blotting.

Enteroids were lysed with lysis buffer (Cell Signaling Technology, Danvers, MA) and equal amounts of protein were loaded into 4–20% Criterion TGX gels (Bio-Rad, Hercules, CA) and electrophoretically transferred to polyvinylidene difluoride (PVDF) membranes; the membranes were incubated with primary antibodies overnight at 4°C followed by secondary antibodies conjugated with horseradish peroxidase (ThermoFisher, Waltham, MA). Membranes were developed using Amersham ECL Western Blotting Detection Reagent from GE Healthcare Life Science (Piscataway, NJ) or Immobilon Western Chemiluminescent HRP substrate (ThermoFisher, Waltham, MA).

Statistical analysis.

All data were summarized using bar graphs with means and standard deviations. BAs and relative levels of mRNA expression were measured in triplicate samples per mouse. Biological replicates were used for assessment of protein levels based on western blots. Comparisons between genotypes and diet were performed using repeated measures linear mixed model with 2 factors (genotype and diet) and an interaction term which accounts for the differential effect of genotype and diet. Two-way analysis of variance (ANOVA) for genotype and diet and interaction between factors was utilized for western blot studies. One-way ANOVA was utilized for analysis of NT and GW4064 dose levels. For statistical analysis, data were log-transformed as necessary and normal distribution were utilized for optimal fit of the model. Adjustment for multiple pairwise comparisons was performed using the Holm’s method. Statistical analyses were performed using the SAS system 9.4 (Cary, NC).

Results

NT deficiency attenuates HFD-induced decrease of plasma primary unconjugated BAs.

Observations from a human study demonstrate that obesity is associated with increased BA biosynthesis and impaired serum BA fluctuations (26). We previously demonstrated that NT deficiency prevented HFD-induced obesity in both male and female mice, indicating the contribution of endogenous NT to obese development (10). To determine whether NT affects BA metabolism, we quantified plasma BA composition at an early stage (i.e., 6 weeks) and at a later stage (i.e., 28 weeks) of obesity. We have shown that HFD consumption significantly increased body weight and fat mass in NT^+/+ but not NT^−/− mice at both early and later stages of obesity in our previous reports (10, 23). The plasma BA profiles of mice fed HFD for 6 weeks are summarized in Table 1 (glycine-conjugated BAs were detected at noise levels and thus excluded). Taurine-conjugated forms of primary BAs including T-CA, T-UDCA and T-β-MCA were found the dominant species in LFD-fed NT^+/+ mice (Table 1). Primary BAs were reduced by HFD in NT^+/+ mice (Table 1); however, only plasma CA, CDCA and UDCA were significantly decreased (Fig. 1A). There were no significant differences in primary BAs between LFD-fed NT^+/+ and NT^−/− mice (Table 1); however, the levels of primary BAs were significantly increased in HFD-fed NT^−/− mice compared with HFD-fed NT^+/+ mice (Fig. 1A–B). The secondary BAs were also decreased by HFD in NT^+/+ mice (Table 1), although only DCA, ω-MCA and T-LCA were significantly decreased (Fig. 1C).

Table 1.

Plasma BAs of male NT^+/+ and NT^−/− mice fed LFD or HFD for 6 weeks.

Primary	NT+/+ LFD	NT+/+ HFD	NT−/− LFD	NT−/− HFD
CA	159.5±218.3	20.3±37.1 *	148.1±85.6	95.1±23.6 #
CDCA	2.4±1.9	0.6±0.6 *	4.2±3	1.4±1.2 † #
UDCA	3.2±5	0.5±0.5 *	2.9±2.1	1.7±1 #
α-MCA	0.8±1.3	0.2±0.2	0.7±0.3	0.8±0.5 #
β-MCA	14.6±28.7	3.8±6.3	10.5±8.6	17.2±14.5 #
T-CA	314.2±446.3	186.8±110.9	172.8±153	198.4±142.6
T-CDCA	10.6±24.9	4.4±7.4	5.5±9.7	2±1.3
T-UDCA	335.7±271.7	231.6±84.9	194.8±95.2	209.5±45.6
T-α-MCA	58.1±108.7	5±3.3	29.9±61.7	26.6±40.2
T-β-MCA	285±528.4	85.9±72.3	95.4±132.3	110.1±106
Secondary	NT+/+ LFD	NT+/+ HFD	NT−/− LFD	NT−/− HFD
DCA	76.7±83.4	18.7±10.7 *	214±194.4 ‡	39.1±15.2 † #
LCA	1.2±0.8	1±0.7	2±1	1.8±0.5
ω-MCA	11.7±31.9	0.7±0.3 *	2.1±1.3	2.1±0.9
T-DCA	26.4±26	25±21.1	42.2±28.1	22.9±22.5
T-LCA	1.5±1	0.6±0.3 *	1.9±1.6	0.9±0.5
T-ω-MCA	46.4±93.9	10.6±14.6	15.3±25	20.1±21.9

Primary and secondary BAs (nM) in the plasma of male NT^+/+ and NT^−/− mice fed LFD or HFD for 6 weeks. Results are given as mean ± SD, n=10 mice/group.

^*p<0.05 vs. NT^+/+ LFD;

^†p<0.05 vs. NT^−/− LFD;

^‡p<0.05 vs. NT^+/+ LFD;

^#p<0.05 vs. NT^+/+ HFD.

Figure 1. Plasma BA profile in male NT^+/+ and NT^−/− mice fed LFD or HFD for 6 weeks.

Male NT^+/+ and NT^−/− mice were fed LFD or HFD at weaning for 6 weeks (n=10 mice/group) and plasma BA profile analysis was performed as described in the Materials and Methods section. A. Plasma CA, CDCA and UDCA. B. Plasma α-MCA and β-MCA. C. Plasma DCA, ɷ-MCA and T-LCA. *p<0.05 vs. NT^+/+ LFD; ^†p<0.05 vs. NT^−/− LFD; ^‡p<0.05 vs. NT^+/+ LFD; ^#p<0.05 vs. NT^+/+ HFD. ANOVA was utilized to analyze the data.

Consistently, the plasma profile from mice fed HFD for 28 weeks also showed that the plasma uncojugated primary BAs were significantly decreased by HFD in NT^+/+ mice, which were increased in HFD-fed NT^−/− mice (Suppl. Table 1). Unconjugated secondary BAs, DCA and ω-MCA were also reduced in NT^+/+ mice in which ω-MCA was significantly increased in HFD-fed NT^−/− mice (Suppl. Table 1). In addition, LFD-fed NT^−/− mice showed relatively higher levels of certain primary and secondary BAs compared with LFD-fed NT^+/+ mice (Suppl. Table 1). The unconjugated primary BAs and secondary BAs, except LCA, were not as reduced by HFD feeding in NT^−/− mice as compared with NT^+/+ mice (Suppl. Table 1). Therefore, we detected the HFD-induced decrease on plasma BA levels at both early and later stages of obesity. These decreases were significantly attenuated in NT-deficient mice, suggesting that endogenous NT contributes to the disruption of BA homeostasis associated with obesity.

NT contributes to the increased fecal BA excretion at the later stage of obesity.

It was reported that the plasma BA profiles did not reflect hepatic BA synthetic pathways, but rather transport and metabolism within the enterohepatic circulation (27). We hypothesized that NT deficiency improved HFD-decreased plasma BA profiles through controlling the fecal BA excretion. We next profiled fecal BAs in mice fed HFD for 6 or 26 weeks (Table 2 and Suppl. Table 2, respectively). Similar to the plasma BA profile as indicated above that plasma unconjugated primary BAs were decreased by HFD in NT^+/+ mice, the fecal primary BAs tended to be increased in NT^+/+ mice fed HFD for 6 weeks (Table 2); however, only β-MCA (Fig. 2B) and T-CA (Fig. 2C, left panel) were found significantly increased. Surprisingly, we noticed that the levels of certain fecal BAs were relatively elevated in LFD-fed NT^−/− mice (Table 2). Furthermore, fecal CDCA, T-CA and T-CDCA were significantly increased in NT^−/− mice fed HFD (Fig. 2A and C); only DCA was significant higher in HFD-fed NT^−/− mice vs. HFD-fed NT^+/+ mice (Fig. 2D). In addition, HFD significatly increased T-DCA and T-β-MCA (Fig. 2E), which were not correlated with their decreased levels in plasma.

Table 2.

Fecal BAs of male NT^+/+ and NT^−/− mice fed LFD or HFD for 6 weeks

Primary	NT+/+LFD	NT+/+ HFD	NT−/− LFD	NT−/− HFD
CA	60.7±55.3	187.9±423.1	399±698.3	1267.3±1870.6
CDCA	0.3±0.0	0.3±0.0	0.3±0.0	0.6±0.7 †
UDCA	4150±2370.6	6149.5±3406.6	6929.2±4180.2	7160.6±2967.1
α-MCA	1099.7±646.5	1535.4±1280.4	1761.9±1307.7	1900.3±1018.5
β-MCA	3.1±3	8.1±2.5 *	7.6±7	20.1±38.7
T-CA	46.3±72.4	210.2±92.6 *	48.8±62.8	291.8±451.7 †
T-CDCA	846±300.5	1102.2±530.4	708.8±448	1191.1±446.5 †
T-α-MCA	29.2±37.2	39.6±10.5	34.6±40.9	264.7±732.8
T-β-MCA	56.1±62.1	100.5±27.9	64.6±78.8	537.6±1444.5
Secondary	NT+/+ LFD	NT+/+ HFD	NT−/− LFD	NT−/− HFD
DCA	5.1±3.7	7.1±6.9	9.4±8.3	89.3±209.9 #
LCA	1033.4±567.4	1261.5±683.7	1957.2±1231.8	1658.6±876
ω-MCA	1.8±0.2	1.6±0.1	1.8±0.2	1.8±0.3
T-DCA	5.5±4.2	19.5±7.3 *	13.2±19.6	36.6±61.1 †
T-LCA	4648.5±3133.1	3032.2±2417.8	5577.4±4691.6	3379.2±2630.8
T-ω-MCA	6±8.4	5.1±2.4	2.5±3	36.9±97 †
T-UDCA	33.3±28.3	50.2±13	33.4±43.6	57.2±73.6
GHDCA	527.1±289.5	526.4±453	1101±765.8	1137.7±789.4

Primary and secondary BAs in the feces (pg/mg) of male NT^+/+ and NT^−/− mice fed LFD or HFD for 6 weeks. Results are given as mean ± SD. n=10 mice/group.

^*p< 0·05 vs. NT^+/+ LFD;

^†p<0·05 vs. NT^−/− LFD;

^‡p< 0·05 vs. NT^+/+ LFD;

^#p< 0·05 vs. NT^+/+ HFD.

Figure 2. Fecal BA profile in male NT^+/+ and NT^−/− mice fed LFD or HFD for 6 weeks.

Mice and fecal BA profile analysis are the same as in Figure 1. A. Fecal CDCA. B. Fecal β-MCA. C. Fecal T-CA and T-CDCA. D. Fecal DCA. E. Fecal T-DCA and T-ω-MCA. *p<0.05 vs. NT^+/+ LFD; ^†p<0.05 vs. NT^−/− LFD; ^‡p<0.05 vs. NT^+/+ LFD; ^#p<0.05 vs. NT^+/+ HFD. ANOVA was utilized to analyze the data.

However, fecal unconjugated primary and secondary BAs (except CDCA) were dramatically increased in NT^+/+ mice fed HFD for 26 weeks (Suppl. Table 2). Unconjugated primary and secondary BAs as well as T-DCA were upregulated in LFD-fed NT^−/− mice (Suppl. Table 2). HFD feeding significantly increased CA in NT^−/− mice in a pattern similar to NT^+/+ mice (Suppl. Table 2) but did not increase α-MCA and β-MCA in NT^−/− mice as effectively as in NT^+/+ mice (Suppl. Table 2). Compared with HFD-fed NT^+/+ mice, the levels of CDCA, T-DCA, T-LCA and T-ω-MCA were significantly higher in NT^−/− mice fed HFD. Together, fecal unconjugated primary and secondary BAs were increased in NT^+/+ mice with long-term HFD feeding, which are correlated with the decreased plasma BAs. NT deficiency increases basal fecal BA levels of unconjugated primary and secondary BAs as noted in LFD-fed NT^−/− mice; feeding a HFD results in less of an increase in the levels of fecal unconjugated primary and secondary BAs in NT^−/− mice. Therefore, the beneficial effects of NT deficiency on HFD-decreased plasma BAs at the early stage of obesity appear not to be the result of decreased fecal BA excretion; other mechanisms are likely involved. It is possible that NT contributes to the increased fecal BA excretion at the later stage of obesity, thus resulting in decreased levels of plasma BAs.

NT deficiency decreases Fxr expression, which is upregulated in HFD-fed NT^−/− mice.

We next evaluated the genes involved in intestinal BA transport. We found that ileal Fxr mRNA expression was reduced in NT^+/+ mice fed HFD for 5 weeks (Fig. 3A); relative to LFD-fed NT^+/+ mice, Fxr mRNA expression was significantly decreased in NT^−/− mice which was significantly increased by HFD feeding. Consistently, ileal FXR protein expression was also reduced in NT^+/+ mice fed HFD for 28 weeks (Suppl. Fig. 1A); relative to LFD-fed NT^+/+ mice, FXR protein expression was significantly decreased in NT^−/− mice, which was not further influenced by HFD feeding. Therefore, the ileal expression of FXR is decreased by HFD feeding at both early and later stages of obesity; importantly, NT deficiency decreases the basal FXR expression, which is either upregulated at the early stage or not influenced at the later stage of obesity by HFD feeding.

Figure 3. HFD inhibits ileal Fxr and BA transporter expression which is upregulated in NT-deficient mice.

Total RNA was isolated from ileal mucosal scrapings of male NT+/+ and NT−/− mice fed LFD or HFD for 5 weeks. A-D. qPCR analysis on ileal Fxr, Asbt, Ilbp, Ostα and Ostβ. n=4 mice/group. *p<0.05 vs. NT^+/+ LFD; ^†p<0.05 vs. NT^−/− LFD; ^‡p<0.05 vs. NT^+/+ LFD; #p<0.05 vs. NT^+/+ HFD. ANOVA was utilized to analyze the data.

NT deficiency downregulates BA transporter mRNA expression in non-obese conditions, which is upregulated under obese conditions.

Active BA absorption in the ileum is mediated by BA transporters, including ASBT, ileal lipid binding protein (ILBP) and organic solute transporter α and β (OSTα and OSTβ) (28). We detected a trend as noted by a decrease of ileal Asbt, Ilbp, Ostα and Ostβ mRNA expression in NT^+/+ mice fed HFD for 5 weeks (Fig. 3B–D); mRNA levels were significantly decreased in LFD-fed NT^−/− mice, which was increased by HFD except Ostβ.

Similarly, the expression of ileal Asbt, Ilbp, Ostα and Ostβ was also decreased in NT^+/+ mice fed HFD for 28 weeks (Suppl. Fig. 1B–E); these expression levels were relatively lower in NT^−/− mice fed LFD, which were not further altered by HFD. Together, the expression levels of the BA transporters are influenced at the early stage as well as the later stage of obesity. The expression levels of BA transporters are downregulated by HFD in wild-type mice, which are upregulated in the absence of NT, suggesting a negative regulation of NT on BA transporter signaling under obese conditions. The expression levels of BA transporters are decreased in LFD-fed NT-deficient mice, indicating a positive regulation of NT on BA transporter signaling under non-obese conditions.

NT stimulates FXR and BA transporter expression in differentiated Caco-2 cells.

Caco-2 cells, a human colon cancer cell line (29), have been widely used as a normal enterocyte model of differentiation by numerous investigators. NTR1 expression is increased on the apical surface of differentiated Caco-2 cells (30). We next utilized differentiated Caco-2 cells to confirm the regulation of NT on ileal FXR and BA transporters in vitro. Consistently, we detected lower levels of NTR1 protein expression in undifferentiated Caco-2 cells, which were increased on day 6 post-confluency and reached a peak on day 12 (Fig. 4A, left panel). FXR protein and mRNA expression increased in a similar fashion as NTR1 (Fig. 4A). Treatment with NT alone failed to induce ILBP but increased OSTα mRNA expression (Fig. 4B). Both ILBP and OSTα mRNA expression were increased by treatment of GW4064, an agonist of FXR (31), and the increases were significantly enhanced in the presence of NT (Fig. 4B). These data demonstrate that NT positively regulates activated FXR transcriptional activity thus BA transporter expressions under normal conditions.

Figure 4. NT increases the mRNA expression of FXR and BA transporters in Caco-2 cells.

A. Caco-2 cells were collected prior (d0) and post confluency (d3, d6 and d12) and western blotting analysis (left panel) and qPCR (right panel) performed. *p<0.05 vs. d0; ^†p<0.05 vs. d3; ^‡p<0.05 vs. d6. B. Caco-2 cells at d12 as in A were pretreated with or without GW4064 in serum-free medium for 30 min followed by addition of NT for 24 h; cells were collected and qPCR performed. *p<0.05 vs. DMSO; ^†p<0.05 vs. NT alone; ^‡p<0.05 vs. GW4064 alone. ANOVA was utilized to analyze the data.

NT positively regulates Fxr and Ilbp mRNA expression in mouse ileal enteroids.

We next confirmed the positive regulation of NT on FXR-BA transporter signaling in mouse ileal enteroids ex vivo. The enteroids were established from jejunum or ileum of NTR1^+/+ and NTR1^−/− mice remained on normal chow; there were no obvious morphorlogical difference in both jejunal and ileal enteroids between NTR1^+/+ (Fig. 5A) and NTR1^−/− (data not shown) mice. NT treatment increased phosphorylation of ERK1/2 (p-ERK1/2), an activation marker of NT/NTR1 signaling (32), in NTR1^+/+ jejunal and ileal enteroids, which was decreased in NTR1^−/− enteroids (Fig. 5B). Ileal Fxr mRNA expression was increased by NT treatment in a dose-dependent fashion in wild-type mouse enteroids (Fig. 5C). CDCA is the highest affinity natural ligand for FXR (33), while GW4064 a synthetic selective, non-steroidal agonist of FXR (EC₅₀=15 nM) (31). Both CDCA and GW4064 treatments increased Fxr mRNA expression in NT^+/+ ileal enteroids, whereas expression was attenuated in NT^−/− ileal enteroids (Fig. 5D) thus indicating the specific effects of NT on Fxr expression. Consistently, ileal Ilbp mRNA expression was increased by NT treatment in wild-type mouse enteroids (Fig. 5E). GW4064 treatment dramatically elevated ileal Ilbp mRNA levels in NT^+/+ enteroids, which was attenuated in NT^−/− enteroids (Fig. 5F). These results provide strong evidence confirming the positive role for NT in the regulation of FXR and its downstream BA transporters under normal conditions.

Figure 5. NT increases Fxr and Ilbp mRNA expression in mouse enteroids.

A. Jejunal and ileal enteroids were established from NTR1^+/+ mice and cultured for d7 after subseeding. Images were taken by inverted microscopy. Scale bars, 100 μm. B. Jejunal and ileal enteroids, established from NTR1^+/+ and NTR1^−/− mice, were cultured for d7 and treated with or without NT (10 nM) in serum-free DMEM/F12 medium for 15 min; proteins were isolated and western blotting analysis performed. C. Ileal enteroids from wild-type mice were treated at d7 after subseeding with various dosages of NT in serum-free DMEM/F12 medium for 24 h and qPCR performed. *p<0.05 vs. control (−); ^†p<0.05 vs. 1 nM NT. D. NT^+/+ or NT^−/− ileal enteroids were treated with or without CDCA (left panel) or GW4064 (right panel) in serum-free DMEM/F12 medium for 24 h and qPCR performed. *p<0.05 vs. vehicle control (–) (DMSO) in NT^+/+ or NT^−/− enteroids; ^†p<0.05 vs. 10 nM CDCA in NT^+/+ or NT^−/− enteroids; ^‡p<0.05 vs. 10 nM CDCA or 0.1 μM GW4064 in NT^+/+ enteroids; ^#p<0.05 vs. 100 nM CDCA or 0.5 μM GW4064 in NT^+/+ enteroids. E. Ileal enteroids from wild-type mice were treated at d7 with different concentrations of NT in serum-free DMEM/F12 medium for 24 h and qPCR performed. *p<0.05 vs. control (−). F. NT^+/+ or NT^−/− ileal enteroids were treated at d7 with or without GW4064 in serum-free DMEM/F12 medium for 24 h and qPCR performed. *p<0.05 vs. vehicle control (−) (DMSO) in NT^+/+ or NT^−/− ileal enteroids; ^†p<0.05 vs. GW4064 in NT^+/+. ANOVA was utilized to analyze the data.

Discussion

To investigate the association of NT with BA metabolism in diet-induced obesity, we analyzed BA concentrations in plasma and feces in HFD-induced obesity at an early stage (6 weeks) and a later stage (greater than 20 weeks) in NT^+/+ and NT^−/− mice. We further evaluated the regulation of NT on intestinal FXR and BA transporter signaling. Our current study include 3 main findings: i) HFD decreases plasma primary and secondary BAs in NT^+/+ mice; NT deficiency improves plasma profile. ii) HFD inhibits the ileal FXR-BA transporter axis in NT^+/+ mice; NT deficiency prevents HFD-induced decrease of ileal FXR and BA transporter expression. iii) HFD feeding increased fecal BA excretion which is correlated with the decrease of plasma BAs at the later stage of obesity; these effects are attenuated by NT deficiency.

We observed that primary and secondary plasma BAs were downregulated by HFD feeding and NT deficiency improved this HFD-induced defects. Consistent with these beneficial effects from NT deficiency, Ippagunta, et al. (34) demonstrated that CA administration improved metabolic disorders. UDCA has been shown to improve glucose metabolism (35). Observations from human studies showed an association of increased circulating BAs with obesity and other metabolic disorders (36). For example, plasma CA was significantly increased in obese patients (37). Despite an apparent discrepancy between our results from mice and findings from humans, the beneficial observations from bariatric surgery, specifically Roux-en-Y gastric bypass (RYGB) and vertical sleeve gastrectomy (VSG), demonstrated an increase in circulating BAs, which correlated with improvement in metabolic parameters (38).

We found that HFD decreased FXR expression in ileum in NT^+/+ mice; NT deficiency maintained ileal FXR expression at relatively higher levels and resisted to the decrease by HFD feeding, suggesting that NT played a negative role in the regulation of FXR under obese conditions. However, ileal FXR expression was lower in NT^−/− mice compared with LFD-fed NT^+/+ mice, indicating that NT played a positive role in the regulation of FXR expression under non-obese conditions. Consistent with our findings, FXR is required to prevent obesity and metabolic disorders. For example, CA prevents hepatic triglyceride accumulation, very low density lipoprotein secretion, and elevated serum triglycerides in mouse models of hypertriglyceridemia by activating FXR (39). Activation of FXR by the synthetic agonist GW4064 or hepatic overexpression of constitutively-active FXR significantly lowered blood glucose levels in both diabetic db/db and wild-type mice, and in contrast, FXR null mice exhibited glucose intolerance and insulin insensitivity (40). One study showed that up-regulating FXR with a gut-restricted FXR agonist, fexaramine, reduced diet-induced weight gain and hepatic glucose production while enhancing thermogenesis and browning of white adipose tissue (41). The effects of RYGB and VSG on body weight and glucose tolerance depend on FXR, and in the absence of FXR, the beneficial effects of bariatric surgery were substantially reduced (38).

In the intestine, FXR controls the absorption of BAs by regulating the expression of BA transporters, ASBT, OSTα and OSTβ and ILBP (42). Sinal, et al. (20) provide convincing evidence showing undetectable ILBT mRNA in Fxr null mice under either control or CA diet. Human ileal ASBT and OSTα expression was significantly increased in obesity compared to non-obese conditions (43). In our current study, however, we showed a clear trend for a decrease in the expression of ileal Asbt, Ilbp, Ostα and Ostβ in HFD-fed NT^+/+ mice at both the early stage and the later stage of obesity; consistent with the expression of Fxr, the expression of these genes also remained at relatively high levels in NT^−/− mice fed HFD, indicating that NT contributed to HFD-induced disruption of BA transporter signaling under obese conditions. Moreover, the basal levels of these BA transporters were downregulated in NT^−/− mice compared with LFD-fed NT^+/+ mice, demonstrating that NT plays a positive role in the regulation of those genes under non-obese conditions. The alterations in the FXR and BA transporter axis correlated with the fecal BA profile and the plasma BA profile at the later stage of obesity. We assume that the FXR-BA transporter signaling pathway takes place more rapidly in response to the HFD consumption than the actual outcomes of fecal BAs.

In summary, the main findings in the current study are that NT, released from N cells in response to fat ingestion, contributes to HFD-induced disruption of BA metabolism. NT stimulates FXR and BA transporter signaling under non-obese conditions, however, inhibits FXR and BA transporter signaling under obese conditions. Our study demonstrates, for the first time, that NT plays different roles in the regulation of BA metabolism and ileal FXR and BA transporters in response to various nutrient abundancy, thus providing valuable insights into the development of obesity and possible therapeutic strategies for its control.

Nonstandard Abbreviations:

α-MCA

α-muricholic acid

β-MCA

β-muricholic acid

ASBT

apical sodium dependent bile acid transporter

BA

Bile acid

BSEP

bile-salt export pump

CA

cholic acid

CDCA

chenodeoxycholic acid

CYP7A1

Cholesterol 7α-hydroxylase

CYP8B1

Sterol-12α-hydroxylase

DCA

deoxycholic acid

FBS

Fetal bovine serum

FXR

Farnesoid X receptor

HFD

high-fat diet

L-WRN CM

L-WRN conditioned medium

LCA

lithocholic acid

LFD

low-fat diet

LM1

lipid mixture 1

LXRα

liver X receptor-α

LXRβ

liver X receptor-β

MEM

Minimum Essential Medium Eagle

MRP

multidrug resistance protein

NAC

n-acetylcysteine

NT

Neurotensin

NTR

NT receptor

OATP2

organic anion transporting polypeptide 2

p-ERK1/2

phosphorylation of ERK1/2

PA

palmitic acid

PVDF

polyvinylidene difluoride

RYGB

Roux-en-Y gastric bypass

UDCA

ursodeoxycholic acid

VSG

vertical sleeve gastrectomy