Ostα^−/− mice exhibit altered expression of intestinal lipid absorption genes, resistance to age-related weight gain, and modestly improved insulin sensitivity

Abstract

The organic solute transporter OSTα-OSTβ is a key transporter for the efflux of bile acids across the basolateral membrane of ileocytes and the subsequent return of bile acids to the liver. Ostα^−/− mice exhibit reduced bile acid pools and impaired lipid absorption. In this study, wild-type and Ostα^−/− mice were characterized at 5 and 12 mo of age. Ostα^−/− mice were resistant to age-related weight gain, body fat accumulation, and liver and muscle lipid accumulation, and male Ostα^−/− mice lived slightly longer than wild-type mice. Caloric intake and activity levels were similar for Ostα^−/− and wild-type male mice. Fecal lipid excretion was increased in Ostα^−/− mice, indicating that a defect in lipid absorption contributes to decreased fat accumulation. Analysis of genes involved in intestinal lipid absorption revealed changes consistent with decreased dietary lipid absorption in Ostα^−/− animals. Hepatic expression of cholesterol synthetic genes was upregulated in Ostα^−/− mice, showing that increased cholesterol synthesis partially compensated for reduced dietary cholesterol absorption. Glucose tolerance was improved in male Ostα^−/− mice, and insulin sensitivity was improved in male and female Ostα^−/− mice. Akt phosphorylation was measured in liver and muscle tissue from mice after acute administration of insulin. Insulin responses were significantly larger in male and female Ostα^−/− than wild-type mice. These findings indicate that loss of OSTα-OSTβ protects against age-related weight gain and insulin resistance.

bile acids are important mediators of dietary lipid absorption. Synthesized in the liver from cholesterol, they are secreted into bile and delivered into the small intestine, where they aid in the emulsification and absorption of dietary lipids, cholesterol, fat-soluble vitamins, lipophilic drugs, and electrolytes (7, 36). Recently, attention has focused on the role of bile acids as signaling molecules acting through nuclear receptors [e.g., the farnesoid X receptor (FXR)], G protein-coupled receptors (e.g., TGR5), and other cell-signaling pathways to modulate their own levels, as well as triglyceride, cholesterol, glucose, and energy metabolism (8, 9, 14, 16, 34).

Bile acids activate the FXR and then decrease their own synthesis through two signaling pathways that regulate cholesterol 7α-hydroxylase (CYP7A1). FXR activation increases expression of the short heterodimer partner (SHP), which interacts with the liver receptor homolog-1 and prevents liver receptor homolog-1 from associating with and activating CYP7A1 (14). In the ileum, activation of the FXR by bile acids increases release of FGF15 and FGF19 (mouse and human forms, respectively), which activates FGF receptor 4 in the liver to repress Cyp7a1 expression and, thus, decrease bile acid synthesis (16). In addition, the FXR decreases expression of the apical sodium-dependent bile acid transporter (ASBT) and increases expression of the organic solute transporter α- and β-subunits (OSTα and OSTβ) (2, 21). These changes are expected to reduce the concentration of bile acids in the enterocyte and decrease bile acid movement through the enterohepatic circulation. FXR activation also reduces de novo lipogenesis by decreasing expression of sterol regulatory element-binding protein 1c (SREBP1c) and its target genes and improves glucose tolerance and insulin sensitivity (14). In particular, the FXR target FGF19 increases metabolic rate, decreases adiposity, increases glycogen synthesis, and decreases gluconeogenesis (27).

Bile acids are cycled between the intestine and the liver, creating an enterohepatic cycle that participates in regulating whole body lipid and, in turn, energy homeostasis (7, 36). After secretion into the intestinal lumen, bile acids are efficiently taken up into ileocytes via the ASBT; their subsequent efflux across the basolateral membrane, a requirement for return to the liver via the portal circulation, is mediated largely by OSTα-OSTβ. OST consists of two subunits that heterodimerize for membrane delivery and generation of transport activity (2, 4, 15, 31, 33, 37). OSTα-OSTβ was originally implicated in the basolateral efflux of bile acids in the ileum on the basis of its substrate specificity, tissue distribution, subcellular localization, and transcriptional regulation (2, 5). The in vivo functions of the transporter have subsequently been characterized in two Ostα-deficient (Ostα^−/−) mouse models (1, 25). Neonatal Ostα^−/− mice are ∼25% smaller than wild-type mice, but this difference disappears shortly after weaning (1, 25). The mechanism for the growth delay in Ostα^−/− mice is unknown. The small intestine is longer and heavier in Ostα^−/− than wild-type mice, with apparent ileal enterocyte dysplasia (1, 13, 25, 32). Ostα^−/− mice exhibit impaired transileal bile acid absorption coupled with an increase in ileal FGF15 production and a decrease in hepatic Cyp7a1 expression resulting from activation of the FXR in the ileocyte (1, 12, 13, 25). Ostα^−/− mice have a markedly reduced bile acid pool size and decreased cholesterol and lipid absorption (1, 13, 25).

Because bile acids serve important roles in regulating lipid levels and glucose handling and initial studies identified potentially beneficial changes in lipid metabolism in Ostα^−/− mice, we set out to establish the role of OSTα-OSTβ in the regulation of lipid and glucose homeostasis in adult (5-mo-old) and aged (12-mo-old) mice. The rationale for studying adult and aged mice was to determine whether the absence of the OST alters the tendency of wild-type C57BL/6 mice to accumulate body fat and become insulin-resistant with age. We show that Ostα^−/− mice are resistant to the age-related accumulation of body fat and decline in insulin sensitivity.

MATERIALS AND METHODS

Animal maintenance.

Ostα^−/− mice generated and maintained on a C57BL/6 background (1, 15) were housed in a full-barrier facility with a 12:12-h light-dark cycle and maintained on standard chow (Laboratory Autoclavable Rodent Diet 5010, Purina Mills, St. Louis, MO) containing 12.7 kcal% fat and 0.027% cholesterol. For the cholic acid study, 12-mo-old male mice were fed a modified LabDiet 5001 enriched with 0.2% cholic acid and containing 13.4 kcal% fat and 0.025% cholesterol for 3 wk. Genotyping was performed by PCR analysis of DNA isolated from tail biopsies (1). Mice had free access to food and water, unless otherwise noted. The age of death (in months) for mice that died spontaneously was recorded for the longevity study. Mice were monitored for general health and well-being, and mice that exhibited an inability to ambulate, an inability to maintain food or water intake, or clinical signs of pain or illness were removed from the study and euthanized. All procedures were approved by the local Animal Care and Use Committees, according to criteria outlined in the Guide for the Care and Use of Laboratory Animals issued by the Institute of Laboratory Animal Research of the National Academy of Sciences.

Body composition and indirect calorimetry.

Body composition and indirect calorimetry studies were conducted at the National Institutes of Health-Yale Mouse Metabolic Phenotyping Center and approved by the Yale University Animal Care and Use Committee. Mice were individually housed under controlled temperature (23°C) and lighting (12:12-h light-dark cycle) with free access to water and fed ad libitum a standard chow (2018S, Harlan Teklad, Indianapolis, IN). Body composition was assessed by ¹H-magnetic resonance spectroscopy (BioSpin, Bruker, Billerica, MA). Whole-body energy metabolism, including O₂ uptake, CO₂ production, energy expenditure, food and water intake, and locomotor activity, was measured for 72 h by indirect calorimetry (CLAMS, Columbus Instrument, Columbus, OH).

Analysis of organ and fat pad weights.

Mice at 5 ± 1 and 12 ± 1 mo of age were fasted for 4 h and then anesthetized with pentobarbital sodium (50 mg/kg ip). Whole-body weights were recorded, and liver, kidney, small intestine, and gonadal and perirenal fat pads were removed and weighed. Where noted, weights are expressed relative to body weight. The small intestine was emptied before measurement.

Analysis of fecal and liver lipids.

Feces were collected from mice fed ad libitum and dried overnight. Liver samples were obtained from mice fasted for 4 h. Fecal and liver lipid content was determined gravimetrically (22). Briefly, 0.5 g of dried feces or 100- to 300-mg pieces of liver were homogenized in water or phosphate-buffered saline and extracted using chloroform-methanol (2:1). Sulfuric acid (0.1%) was added to liver extractions. A portion of the organic phase was dried in a vacuum evaporator to determine lipid mass. Cholesterol was analyzed in liver lipids resuspended in 2% Triton X-100 using an enzymatic assay (Wako Chemicals USA, Richmond, VA).

Bile acid analysis.

Dried feces were collected from mice fed ad libitum, and liver samples were obtained from mice restricted from food for 4 h. The bile acid pool samples include liver, gallbladder, and small intestine of mice after an overnight fast, and serum was also collected from mice after an overnight fast. Bile acids were extracted from feces, liver, and bile acid pool samples using methanol as described by Setchell et al. (30), and total bile acids in serum and tissue samples were analyzed using an enzymatic assay (19). Where noted, values are expressed relative to tissue wet weight.

Serum lipid analysis.

Mice were fasted overnight and anesthetized with pentobarbital sodium (50 mg/kg ip). Whole blood was collected from the vena cava; serum was separated, and total cholesterol, HDL, LDL, and triglycerides were analyzed using enzymatic kits from Wako Chemicals or Thermo Fisher and nonesterified free fatty acids were analyzed using an enzymatic kit from Zenbio (Research Triangle Park, NC).

Glucose and insulin tolerance tests.

Overnight-fasted mice were used for glucose tolerance tests (GTTs). At 7 AM, mice were anesthetized using isoflurane, their tails were snipped, and the mice were allowed to recover for 2 h. Glucose was administered (1 mg/g ip), and tail blood glucose was measured at 0, 15, 30, 60, 90, and 120 min using a OneTouch Ultra glucometer. Additional blood was collected at 0, 30, 60, and 90 min for an insulin ELISA (Crystal Chem, Chicago, IL). For insulin tolerance tests (ITTs), mice were fasted for 6 h (morning to afternoon) and anesthetized, and insulin was administered (1.5 U/kg ip); glucose levels were measured as described above.

Analysis of Akt phosphorylation.

Male and female 12-mo-old wild-type and Ostα^−/− mice were fasted overnight. After they were anesthetized with pentobarbital sodium (50 mg/g ip), the mice were injected with saline or insulin (1.5 U/kg, the dose used in the ITTs), and liver and quadriceps were removed 10 min postinjection and immediately snap-frozen in liquid nitrogen. Frozen tissues were homogenized using an Ultra Turrax homogenizer in lysis buffer (150 mM NaCl, 1.0% Triton X-100, 50 mM Tris, pH 8.0, mammalian protease and phosphatase inhibitor cocktails, and 5 mM activated sodium orthovanadate) and centrifuged at 20,000 g for 20 min. Protein concentration was measured using the Lowry assay (17). Western blots were performed as follows: 30 μg of protein were run on a 10% polyacrylamide gel and transferred to a polyvinylidene difluoride membrane. Milk (5%) was used as a blocking agent, and primary antibodies to Akt1/2 (catalog no. sc-1619-r, Santa Cruz Biotechnology, Santa Cruz, CA) or phosphorylated (Ser⁴⁷³) Akt (catalog no. 4060, Cell Signaling Technology, Danvers, MA) were used at 1:200 and 1:2,500 dilutions, respectively. Bands were visualized using secondary horseradish peroxidase-conjugated antibody to rabbit IgG at 1:5,000 dilution and the LumiGlo system (Kirkegaard & Perry Laboratories, Gaithersburg, MD).

Total mRNA isolation from mouse tissues.

Liver and intestine were removed from anesthetized mice fasted for 4 h. The intestine was sectioned into three equal-length pieces, and the contents were removed. Total RNA was isolated from these tissues using guanidinium thiocyanate-cesium chloride and treated with DNase I (Fermentas/Thermo Scientific).

Real-time quantitative reverse-transcriptase PCR analyses.

Gene-specific oligonucleotide primers were designed using National Center for Biotechnology Information Primer-BLAST and purchased from Integrated DNA Technologies (Coralville, IA) and sequences are available on request. Each primer pair was tested using liver and/or small intestine mRNA and a Superscript One-Step RT-PCR kit (Life Technologies, Carlsbad, CA) on a ThermoHybaid Px2 RT-PCR unit. The resulting PCR product was separated by gel electrophoresis, and the cDNA product was extracted with a kit from Qiagen (Valencia, CA) and diluted to produce gene-specific cDNA standards containing 10–10⁷ copies diluted in yeast total RNA. Relative gene expression was determined on a real-time cycler (Rotor-Gene 3000, Corbett, San Francisco, CA). Samples (n = 6–7) were analyzed in duplicate using an iScript one-step reverse-transcriptase PCR kit with SYBR Green (Bio-Rad, Hercules, CA) or Stratagene Brilliant III Ultra-Fast SYBR Green QRT-PCR Master Mix (Agilent Technologies, Santa Clara, CA). A 50-ng sample of total RNA was analyzed per reaction.

Statistical analyses.

Values are means ± SE. Mean values were considered to be significantly different when P < 0.05 by two-way ANOVA followed by Tukey's or Bonferroni's multiple comparisons test or Student's t-test where applicable.

RESULTS

Decreased weight gain and fat accumulation in Ostα^−/− mice.

Interestingly, body weights were lower and perirenal and gonadal fat pads were smaller in male Ostα^−/− mice as they matured through adulthood (Fig. 1, A and B). Differences in body weight and fat pad weight were also observed in female mice (Fig. 2, A and B), but the differences between genotypes were not as dramatic, because female wild-type mice did not exhibit the age-related weight gain to the extent of male wild-type mice. The small intestines of Ostα^−/− mice were heavier than those of wild-type animals at 5 mo (males: 1.07 ± 0.06 vs. 0.80 ± 0.05 g; females: 1.24 ± 0.10 vs. 0.87 ± 0.08 g) and at 12 mo (males: 1.24 ± 0.07 vs. 1.00 ± 0.04 g; females: 1.55 ± 0.30 vs. 1.02 ± 0.10 g), extending previous findings (1, 13, 25). Ongoing studies of intestinal morphology indicate that Ostα^−/− mice display ileocyte enterocyte dysplasia at an early age that persists but does not appear to worsen as the animals age (Soroka and Boyer, unpublished data). In male and female mice, liver and kidney weights were similar for both genotypes at both ages (data not shown). Male Ostα^−/− mice lived significantly longer than wild-type mice (Fig. 1C), whereas female Ostα^−/− and wild-type mice had similar lifespans (Fig. 2C).

Fig. 1.

Age-related weight gain is less rapid in male organic solute transporter (OST)-deficient (Ostα^−/−) mice. A: body weight of male mice from 2 to 18 mo of age (n = 4–44). B: combined weight of gonadal and perirenal fat pads from 5-mo-old (5m) and 12-mo-old (12m) mice (n = 10–16). Values are means ± SE. *P < 0.05 vs. wild-type (Ostα^+/+). #P < 0.05 vs. 5m. C: age of mice at the time of death (n = 16–49). Median ages of wild-type and Ostα^−/− mice at time of death were 23.60 and 28.65 mo, respectively (P < 0.05).

Fig. 2.

Age-related weight gain is less rapid in female Ostα^−/− mice. A: body weight of female mice from 2 to 15 mo of age (n = 6–52). B: combined weight of gonadal and perirenal fat pads from 5- and 12-mo-old mice (n = 14–23). bw, Body weight. Values are means ± SE. *P < 0.05 vs. wild-type. C: age of mice at the time of death (n = 6–12). Median ages of wild-type and Ostα^−/− mice at time of death were 26.00 and 27.10 mo, respectively [P = not significant (NS)].

Altered energy utilization in Ostα^−/− mice.

At 5 mo of age, male Ostα^−/− mice had ∼5% lower total body fat and 4% higher lean body mass than male wild-type mice (Table 1). These changes were not due to a decrease in food consumption or an increase in activity of Ostα^−/− mice, although Ostα^−/− mice did consume more water than their wild-type counterparts. The respiratory quotient was also higher for Ostα^−/− mice, indicating that more carbohydrates were utilized for energy. The increase in the respiratory quotient occurred during the light cycle while the mice were at rest.

Table 1.

Body composition and indirect calorimetry data for 5-mo-old male wild-type and Ostα^−/− mice

	Wild-Type	Ostα−/−
Body weight, g	28.2 ± 0.5	27.3 ± 0.5
Body fat, %	13.4 ± 1.0	8.9 ± 0.7*
Lean mass, %	69.8 ± 0.9	73.6 ± 0.5*
	24-h cycle
V̇o2, ml·kg−1·h−1	4,738 ± 132	5,120 ± 186
V̇co2, ml·kg−1·h−1	4,367 ± 115	4,804 ± 170
Respiratory quotient (V̇co2/V̇o2)	0.915 ± 0.003	0.933 ± 0.005*
Energy expenditure, kcal·kg−1·h−1	23.5 ± 0.7	25.4 ± 0.9
Calories consumed, kcal·kg−1·h−1	18.8 ± 0.9	21.1 ± 0.9
Water consumed, ml·kg−1·h−1	0.93 ± 0.05	1.20 ± 0.07*
Activity, counts/h	173 ± 31	130 ± 16
	12-h light cycle
V̇o2, ml·kg−1·h−1	4,365 ± 149	4,781 ± 193
V̇co2, ml·kg−1·h−1	3,838 ± 118	4,293 ± 160*
Respiratory quotient (V̇co2/V̇o2)	0.876 ± 0.005	0.896 ± 0.005*
Energy expenditure, kcal·kg−1·h−1	21.4 ± 0.7	23.5 ± 1.0
Calories consumed, kcal·kg−1·h−1	11.8 ± 1.0	13.4 ± 0.6
Water consumed, ml·kg−1·h−1	0.50 ± 0.04	0.72 ± 0.08*
Activity, counts/h	86 ± 19	67 ± 10
	12-h dark cycle
V̇o2, ml·kg−1·h−1	5,252 ± 135	5,587 ± 180
V̇co2, ml·kg−1·h−1	5,095 ± 142	5,507 ± 188
Respiratory quotient (V̇co2/V̇o2)	0.969 ± 0.004	0.984 ± 0.008
Energy expenditure, kcal·kg−1·h−1	26.3 ± 0.7	28.0 ± 0.9
Calories consumed, kcal·kg−1·h−1	28.6 ± 1.5	31.6 ± 1.9
Water consumed, ml·kg−1·h−1	1.53 ± 0.09	1.85 ± 0.08*
Activity, counts/h	291 ± 51	217 ± 31

Values are means ± SE (n = 6–7). V̇o₂, O₂ uptake; V̇co2, CO₂ output.

^*P < 0.05 vs. wild-type.

Lipid and bile acid levels in Ostα^−/− mice.

Male Ostα^−/− mice excreted more lipids in feces than wild-type mice at 5 and 12 mo of age (Fig. 3A). Fecal bile acid excretion was similar in both genotypes and remained unchanged between 5 and 12 mo (Fig. 3B). Liver total lipid and cholesterol content increased with age in male wild-type, but not Ostα^−/−, mice (Fig. 3, C and E). Liver bile acids were also lower in male Ostα^−/− mice (Fig. 3D), likely due to lower Cyp7a1 expression and bile acid synthesis (1, 25). The bile acid pool decreased with age in the wild-type, but not Ostα^−/−, mice and was much smaller in male Ostα^−/− mice at both ages (Fig. 3F). Skeletal muscle total lipid (Fig. 3G) and triglyceride (Fig. 3H) content were lower in 5- and 12-mo-old male Ostα^−/− than wild-type mice and increased with age in both genotypes.

Fig. 3.

Male Ostα^−/− mice excrete more lipid and have lower liver lipids and bile acids. A and B: total lipids and bile acids in feces collected from 5- and 12-mo-old mice over 48 h. In A and B, Ostα^−/− mice ate more than wild-type mice (4.2 ± 0.2 vs. 3.7 ± 1.2 g/day at 5 mo and 4.4 ± 0.3 vs. 3.1 ± 0.4 g/day at 12 mo). C and D: total lipids and bile acids in liver samples from 5- and 12-mo-old animals expressed relative to wet weight (ww). E and F: liver cholesterol and bile acid levels in liver, gallbladder, and small intestine from 5- and 12-mo-old mice. G and H: total lipids and triglycerides in skeletal muscle from 5- and 12-mo-old animals. Values are means ± SE (n = 6–12). *P < 0.05 vs. wild-type. #P < 0.05 vs. 5m.

Female Ostα^−/− mice also excreted more lipids than wild-type mice at 12 mo of age (Fig. 4A). Female knockout mice tended to have lower fecal and liver bile acids by 12 mo of age (Fig. 4, B and D). In female mice, the effects of age and genotype on total lipid and cholesterol in liver were small relative to the differences in male mice (Fig. 4, C and E). Total lipid and triglyceride contents in skeletal muscle of female mice were not affected by age or genotype (Fig. 4, G and H). The bile acid pool of female mice was unchanged with age but, as with males, was dramatically lower in Ostα^−/− animals (Fig. 4F).

Fig. 4.

Female Ostα^−/− mice excrete more lipid and have lower liver lipids. A and B: total lipids and bile acids in feces collected from 5- and 12-mo-old mice over 48 h. C and D: total lipids and bile acids in liver samples from 5- and 12-mo-old animals. E and F: liver cholesterol and bile acid levels in liver, gallbladder, and small intestine from 5- and 12-mo-old mice. G and H: total lipids and triglycerides in skeletal muscle from 5- and 12-mo-old animals. Values are means ± SE (n = 6–12). *P < 0.05 vs. wild-type. #P < 0.05 vs. 5m.

Serum chemistry in Ostα^−/− mice.

To determine whether the increase in fecal lipid excretion led to decreased serum lipids and whether the decrease in bile acid pool size was reflected in serum bile acids, total cholesterol, HDL, LDL, triglycerides, free fatty acids, and bile acids were measured in serum from 5- and 12-mo-old mice (Table 2). Total serum cholesterol increased modestly with age in males, but there were no statistically significant differences between wild-type and Ostα^−/− mice in any of these parameters, except for higher serum bile acids in 5-mo-old female Ostα^−/− mice. It is noteworthy that the results from mature female mice differ from previously reported findings from younger female Ostα^−/− mice, which showed lower serum cholesterol and triglyceride levels in Ostα^−/− than wild-type mice (1, 25).

Table 2.

Serum chemistry in wild-type and Ostα^−/− mice

	5 Months		12 Months
	Wild-type	Ostα−/−	Wild-type	Ostα−/−
	Male
Total cholesterol, mg/dl	55 ± 5	54 ± 3	66 ± 3#	63 ± 3
HDL, mg/dl	57 ± 4	50 ± 5	65 ± 2	62 ± 4
LDL, mg/dl	6.9 ± 0.4	5.6 ± 0.8	7.1 ± 0.5	6.9 ± 0.5
Triglycerides, mg/dl	41 ± 6	40 ± 7	42 ± 4	42 ± 7
Free fatty acids, μM	951 ± 91	785 ± 94	865 ± 61	821 ± 28
Bile acids, μM	35 ± 8	29 ± 6	30 ± 2	38 ± 3
	Female
Total cholesterol, mg/dl	62 ± 5	59 ± 3	61 ± 2	62 ± 2
HDL, mg/dl	44 ± 2	43 ± 1	41 ± 1	41 ± 2
LDL, mg/dl	13 ± 0.7	12 ± 1	11.1 ± 0.4	10.1 ± 0.9
Triglycerides, mg/dl	47 ± 6	58 ± 5	40 ± 2	45 ± 3
Free fatty acids, μM	830 ± 74	999 ± 98	725 ± 70	822 ± 72
Bile acids, μM	5.1 ± 0.4	9.5 ± 0.7*	6.0 ± 0.8	5.5 ± 0.6#

Values are means ± SE (n = 4–14). Serum from overnight-fasted animals was analyzed.

^*P < 0.05 vs. wild-type.

^#P < 0.05 vs. 5 mo.

Expression of genes involved in dietary lipid absorption.

Gene expression was analyzed by real-time PCR using RNA obtained from liver and the proximal, middle, and lower third of the small intestine from 5- and 12-mo-old male wild-type and Ostα^−/− mice. The amount of each mRNA relative to actin mRNA and the expression of each mRNA in Ostα^−/− relative to wild-type mice is given in Tables 3 and 4.

Table 3.

Real-time RT-PCR analysis of intestinal gene expression

	5 mo			12 mo
Gene	Wild-type	Ostα−/−	Ostα−/−/wild-type	Wild-type	Ostα−/−	Ostα−/−/wild-type
	Intestinal mediators of bile acid synthesis, transport, and signaling
Proximal small intestine
Asbt	BD	BD	BD	BD	BD	BD
Fgf15	BD	BD	BD	BD	BD	BD
Fxr	0.31 ± 0.03	0.33 ± 0.04	1.05 ± 0.12	0.25 ± 0.02	0.33 ± 0.02*	1.29 ± 0.08*
Ibabp	BD	BD	BD	BD	BD	BD
Mrp3	0.84 ± 0.2	0.84 ± 0.04	0.99 ± 0.05	1.3 ± 0.2	0.72 ± 0.08*	0.57 ± 0.07*
Shp	0.021 ± 0.002	0.020 ± 0.003	0.93 ± 0.16	0.010 ± 0.002#	0.029 ± 0.002*#	2.78 ± 0.18*
Middle small intestine
Asbt	0.12 ± 0.02	0.25 ± 0.02*	2.11 ± 0.18*	0.11 ± 0.02	0.12 ± 0.04#	1.14 ± 0.40
Fgf15	0.22 ± 0.05	0.091 ± 0.02*	0.41 ± 0.07*	0.12 ± 0.03	0.098 ± 0.02	0.85 ± 0.15
Fxr	0.34 ± 0.03	0.36 ± 0.02	1.07 ± 0.06	0.35 ± 0.02	0.48 ± 0.05*	1.38 ± 0.14*
Ibabp	0.42 ± 0.10	26 ± 5*	61.7 ± 13.1*	0.25 ± 0.05	29 ± 3*	116 ± 13.8*
Mrp3	0.32 ± 0.05	0.51 ± 0.03	1.58 ± 0.12*	0.51 ± 0.03#	0.34 ± 0.06*#	0.66 ± 0.12*
Shp	0.0072 ± 0.002	0.022 ± 0.002*	3.13 ± 0.25*	0.0096 ± 0.001	0.020 ± 0.002*	2.06 ± 0.20*
Distal small intestine
Asbt	0.82 ± 0.04	0.82 ± 0.05	1.00 ± 0.06	0.64 ± 0.07	0.94 ± 0.07*	1.46 ± 0.11*
Fgf15	3.88 ± 0.4	6.45 ± 1.3	1.66 ± 0.34	5.02 ± 0.37	6.96 ± 1.7	1.39 ± 0.33
Fxr	0.55 ± 0.04	0.68 ± 0.05	1.22 ± 0.09	0.53 ± 0.03	0.71 ± 0.09	1.34 ± 0.18
Ibabp	41 ± 2	91 ± 6*	2.24 ± 0.14*	42 ± 3	100 ± 9*	2.37 ± 0.22*
Mrp3	0.42 ± 0.04	0.91 ± 0.07*	2.15 ± 0.16*	0.40 ± 0.04	0.65 ± 0.1*	1.63 ± 0.28*
Shp	0.017 ± 0.001	0.067 ± 0.02*	3.94 ± 1.04*	0.015 ± 0.002	0.074 ± 0.02	4.89 ± 1.51
	Intestinal lipid transporters
Proximal small intestine
Abca1	0.73 ± 0.09	0.28 ± 0.08*	0.39 ± 0.10*	0.66 ± 0.03	0.21 ± 0.04*	0.32 ± 0.06*
Abcg5	1.3 ± 0.1	1.1 ± 0.1	0.84 ± 0.09	1.7 ± 0.1#	1.6 ± 0.1#	0.95 ± 0.08
Abcg8	0.84 ± 0.1	0.85 ± 0.07	1.01 ± 0.08	1.1 ± 0.1	1.2 ± 0.1	1.02 ± 0.10
Fat	0.92 ± 0.2	1.2 ± 0.1	1.27 ± 0.16	0.36 ± 0.07	0.65 ± 0.1#	1.78 ± 0.27
Npc1L1	0.94 ± 0.1	0.81 ± 0.2	0.86 ± 0.17	1.2 ± 0.2	1.8 ± 0.2#	1.54 ± 0.19
Srb1	0.084 ± 0.005	0.091 ± 0.005	1.09 ± 0.06	0.16 ± 0.01#	0.27 ± 0.01*#	1.72 ± 0.08*
Middle small intestine
Abca1	0.68 ± 0.1	0.16 ± 0.01*	0.24 ± 0.01*	0.79 ± 0.05	0.20 ± 0.02*	0.25 ± 0.03*
Abcg5	1.3 ± 0.1	0.86 ± 0.04	0.66 ± 0.03*	1.1 ± 0.09	0.94 ± 0.09	0.82 ± 0.08
Abcg8	0.87 ± 0.1	0.65 ± 0.03	0.76 ± 0.04*	0.64 ± 0.06	0.71 ± 0.04	1.10 ± 0.07
Fat	0.017 ± 0.001	0.019 ± 0.002	1.12 ± 0.10	0.024 ± 0.002#	0.027 ± 0.002#	1.11 ± 0.08
Npc1L1	1.2 ± 0.2	1.3 ± 0.1	1.07 ± 0.08	1.4 ± 0.2	1.3 ± 0.2	0.98 ± 0.16
Srb1	0.013 ± 0.001	0.010 ± 0.0006	0.78 ± 0.05	0.016 ± 0.001	0.018 ± 0.002#	1.16 ± 0.11
Distal small intestine
Abca1	0.42 ± 0.08	0.28 ± 0.03*	0.68 ± 0.06*	0.62 ± 0.02	0.20 ± 0.04*	0.32 ± 0.07*
Abcg5	1.14 ± 0.07	0.76 ± 0.08*	0.66 ± 0.07*	0.96 ± 0.03	0.67 ± 0.08	0.70 ± 0.08
Abcg8	1.1 ± 0.1	0.73 ± 0.05*	0.69 ± 0.05*	0.84 ± 0.01	0.75 ± 0.1	0.88 ± 0.13
Fat	0.66 ± 0.05	0.76 ± 0.07	1.16 ± 0.10	0.61 ± 0.06	0.83 ± 0.07	1.37 ± 0.11*
Npc1L1	2.0 ± 0.2	0.91 ± 0.09*	0.46 ± 0.05*	1.6 ± 0.1	0.96 ± 0.08*	0.61 ± 0.05*
Srb1	0.019 ± 0.001	0.018 ± 0.001	0.95 ± 0.06	0.020 ± 0.0007	0.018 ± 0.002	0.92 ± 0.10

Values are means ± SE (n = 6). Small intestine from 5- and 12-mo-old mice was sectioned into 3 equal pieces, with proximal section attached to the stomach and distal section attached to the cecum. Columns labeled wild-type and Ostα^−/− show gene expression normalized to β-actin expression; columns labeled Ostα^−/−/wild-type show ratio of Ostα^−/− to wild-type expression.

BD, below detection.

^*P < 0.05 vs. wild-type.

^#P < 0.05 vs. 5 months.

Table 4.

Real-time RT-PCR analysis of hepatic gene expression

	5 Months			12 Months
Gene	Wild-type	Ostα−/−	Ostα−/−/wild-type	Wild-type	Ostα−/−	Ostα−/−/wild-type
	Hepatic mediators of bile acid synthesis, transport, and signaling
Bsep	0.28 ± 0.01	0.21 ± 0.02	0.73 ± 0.09	0.20 ± 0.01#	0.19 ± 0.02	0.94 ± 0.08
Cyp27a1	1.7 ± 0.1	1.63 ± 0.07	0.96 ± 0.04	1.1 ± 0.1#	0.85 ± 0.04#	0.79 ± 0.03
Cyp7a1	0.14 ± 0.01	0.035 ± 0.01*	0.26 ± 0.08*	0.085 ± 0.01#	0.044 ± 0.01*	0.52 ± 0.13*
Cyp8b1	0.090 ± 0.01	0.071 ± 0.007	0.78 ± 0.08	0.039 ± 0.002#	0.032 ± 0.004#	0.82 ± 0.12
Fxr	0.036 ± 0.003	0.046 ± 0.01	1.28 ± 0.28	0.025 ± 0.002#	0.027 ± 0.003	1.10 ± 0.12
Shp	0.029 ± 0.04	0.022 ± 0.04	0.77 ± 0.15	0.017 ± 0.001#	0.013 ± 0.001	0.79 ± 0.09
	Hepatic mediators of lipid synthesis
Acaca	0.035 ± 0.001	0.036 ± 0.003	1.05 ± 0.07	0.016 ± 0.001#	0.022 ± 0.001#	1.40 ± 0.08
Acacb	0.023 ± 0.003	0.025 ± 0.002	1.12 ± 0.07	0.0095 ± 0.0006#	0.016 ± 0.002*#	1.68 ± 0.16*
Dgat2	0.65 ± 0.08	0.54 ± 0.03	0.83 ± 0.05	0.60 ± 0.08	0.60 ± 0.06	0.99 ± 0.10
Fas	0.039 ± 0.004	0.053 ± 0.004	1.37 ± 0.11	0.018 ± 0.002#	0.045 ± 0.004*	2.44 ± 0.23*
Fdft1	0.11 ± 0.02	0.28 ± 0.02*	2.47 ± 0.19*	0.033 ± 0.004#	0.13 ± 0.01*#	3.90 ± 0.34*
Gpat	0.090 ± 0.008	0.086 ± 0.003	0.96 ± 0.03	0.075 ± 0.003	0.090 ± 0.009	1.20 ± 0.12
Hmgcr	0.084 ± 0.008	0.21 ± 0.03*	2.45 ± 0.36*	0.046 ± 0.006#	0.11 ± 0.01*#	2.47 ± 0.32*
Insig1	0.28 ± 0.02	0.50 ± 0.01*	1.78 ± 0.05*	0.10 ± 0.01#	0.16 ± 0.01*#	1.66 ± 0.14*
Insig2	0.14 ± 0.02	0.18 ± 0.04	1.22 ± 0.30	0.12 ± 0.01	0.12 ± 0.02	1.04 ± 0.14
Lpin1	0.019 ± 0.002	0.064 ± 0.01*	3.37 ± 0.73*	0.013 ± 0.002	0.018 ± 0.002#	1.45 ± 0.15
Lpin2	0.21 ± 0.03	0.30 ± 0.03	1.43 ± 0.16	0.13 ± 0.01#	0.13 ± 0.01#	0.97 ± 0.08
Lxrα	0.061 ± 0.005	0.052 ± 0.003	0.85 ± 0.05*	0.042 ± 0.006	0.042 ± 0.003	1.00 ± 0.07
Mvd	0.0053 ± 0.002	0.015 ± 0.002*	2.83 ± 0.30*	0.0019 ± 0.0003	0.0091 ± 0.001*#	4.88 ± 0.62*
Pparγ	0.20 ± 0.04	0.21 ± 0.05	1.03 ± 0.25	0.15 ± 0.01	0.12 ± 0.01	0.79 ± 0.08
Scap	0.072 ± 0.004	0.076 ± 0.003	1.05 ± 0.05	0.043 ± 0.001#	0.058 ± 0.007	1.32 ± 0.15
Srebp1a	0.0127 ± 0.0007	0.011 ± 0.001	0.84 ± 0.07	0.0074 ± 0.001#	0.0068 ± 0.0002#	0.93 ± 0.03
Srebp1c	0.061 ± 0.01	0.034 ± 0.003*	0.55 ± 0.05*	0.037 ± 0.005	0.023 ± 0.003*#	0.63 ± 0.07*
Srebp2	0.043 ± 0.002	0.055 ± 0.007	1.28 ± 0.16	0.029 ± 0.003#	0.042 ± 0.003*	1.43 ± 0.12*
	Hepatic mediators of cholesterol transport
Abca1	0.22 ± 0.02	0.17 ± 0.01	0.76 ± 0.07	0.10 ± 0.02#	0.089 ± 0.005#	0.90 ± 0.05
Abcg5	0.053 ± 0.005	0.063 ± 0.006	1.19 ± 0.11	0.034 ± 0.005#	0.039 ± 0.003#	1.14 ± 0.09
Abcg8	0.070 ± 0.007	0.085 ± 0.008	1.20 ± 0.11	0.042 ± 0.007#	0.048 ± 0.002#	1.13 ± 0.05
Ldlr	0.117 ± 0.007	0.14 ± 0.02	1.20 ± 0.15	0.061 ± 0.007#	0.083 ± 0.002#	1.37 ± 0.04

Values are means ± SE (n = 6). Liver from 5- and 12-mo-old mice was used to analyze gene expression. Columns labeled wild-type and Ostα^−/− show gene expression normalized to β-actin expression; columns labeled Ostα^−/−/wild-type show ratio of Ostα^−/− to wild-type expression.

BD, below detection.

^*P < 0.05 vs. wild-type.

^#P < 0.05 vs. 5 months.

Analysis of genes involved in intestinal lipid absorption revealed changes consistent with decreased dietary lipid absorption in Ostα^−/− mice (Table 3). In 5- and 12-mo-old mice, mRNAs encoding the apical cholesterol uptake transporter Niemann-Pick C1-like 1 (Npc1L1) in the distal intestinal segment and the basolateral cholesterol efflux transporter ATP-binding cassette transporter 1 (Abca1) in each intestinal section were lower in Ostα^−/− mice, in agreement with a previous study of ileal Npc1L1 and Abca1 in 3-mo-old male mice (13).

Hepatic expression of the cholesterol synthetic genes hydroxymethylglutaryl-CoA reductase (Hmgcr), mevalonate (diphospho) decarboxylase (Mvd), and farnesyldiphosphate farnesyltransferase (Fdft) was significantly upregulated in Ostα^−/− mice (Table 4), suggesting that reduced dietary cholesterol absorption was partially compensated by increased cholesterol synthesis. The expression of Srebp1a and Srebp2, which activate lipogenic and cholesterogenic enzymes, was unchanged in the livers of Ostα^−/− mice, whereas the expression of Srebp1c, the major transcriptional activator for lipogenic enzymes, was lower. However, several genes downstream of SREBP1c [Fas, acetyl-CoA carboxylase-α and -β (Acaca and Acacb), diglyceride acyltransferase (Dgat2), and glycerol-3-phosphate acyltransferase (Gpat)] were either unaffected or upregulated. This may be partly due to the actions of LIPIN 1, a transcriptional coactivator of lipid homeostasis genes (3, 23). mRNA encoding LIPIN 1 was significantly increased in the livers of 5-mo-old Ostα^−/− mice. Overall, enhanced hepatic Lpin1 expression in Ostα^−/− mice is consistent with increased hepatic triglyceride synthesis necessary to restore triglyceride and fatty acid homeostasis in the face of impaired dietary lipid absorption.

Expression of FXR-related genes.

To examine the possibility of altered FXR activation in male Ostα^−/− mice, expression of Fxr and Shp1 was measured in liver and small intestine sections and expression of the FXR target gene Fgf15 and the ileal bile acid-binding protein (IBABP) gene Ibabp and the bile acid transporters Asbt and multidrug resistance-associated protein 3 (Mrp3) in small intestine sections from 5- and 12-mo-old mice. In Ostα^−/− mice, Fxr was not significantly altered in the liver or small intestine, whereas Shp1 was unchanged in the liver but strongly increased in the middle and distal segments of the small intestine at 5 and 12 mo and in the proximal section of the small intestine at 12 mo.

Patterns of expression of FXR target genes were complex. Fgf15 expression was highest in the distal section of the small intestine but was not significantly affected by age or the absence of OSTα-OSTβ. FGF15 protein, however, is dramatically increased in the ileum of young adult Ostα^−/− mice (13). Remarkably, Ibabp expression was increased by 60- to 100-fold in the middle section and ∼2-fold in the distal section of the small intestine in Ostα^−/− mice of both ages. However, Ibabp expression was by far the highest in the distal small intestine (Table 3). Because the small intestine was sectioned into equal thirds and Ostα^−/− mice have elongated ileums and relatively normal-length jejunums and duodenums (13), there was probably more ileal tissue in the middle sections of small intestine from Ostα^−/− mice, accounting for the apparent increase in Ibabp expression in those sections.

Expression of the bile acid transporter Asbt was unchanged or mildly upregulated in 5- and 12-mo-old Ostα^−/− mice, with undetectable expression in the proximal sections. This finding was unexpected, because ASBT protein is markedly lower in the ileum of Ostα^−/− mice (13, 32), possibly because protein levels were measured in younger animals. Mrp3 expression was higher in the middle and distal sections of small intestine from Ostα^−/− mice at 5 mo and increased in the distal section at 12 mo. This upregulation of Mrp3 presumably partially compensates for the loss of the OST and is similar to findings in previous studies of male and female Ostα^−/− mice (25, 32). Mrp3 expression increased with age in the proximal and middle sections of small intestine from wild-type, but not Ostα^−/−, mice (Table 3).

Improved glucose tolerance and insulin sensitivity in Ostα^−/− mice.

Intracellular lipid metabolites and bile acids have been shown to affect insulin action and glucose tolerance (14, 26). GTTs and ITTs were performed on 5- and 12-mo-old male (Fig. 5) and female (Fig. 6) mice. Interestingly, subtle improvements in glucose tolerance, as evidenced by lower blood glucose levels 15, 30, 60, and 90 min after glucose injection at 5 and 12 mo of age, were observed in male Ostα^−/− mice (Fig. 5, A–D). Plasma insulin concentrations during the GTTs were similar in 5-mo-old mice (Fig. 5E), as were the declines in plasma glucose during the ITTs (Fig. 5G).

Fig. 5.

Male Ostα^−/− mice have improved glucose tolerance and insulin sensitivity. Glucose tolerance tests were performed in 5-mo-old (A, C, and E) and 12-mo-old (B, D, and F) animals following an overnight fast. Glucose (1 mg/g ip) was administered, and tail blood was analyzed for glucose at 15- to 30-min intervals. In E and F, additional blood was collected from tail snips at 0, 30, 60, and 90 min for insulin analysis. G and H: insulin tolerance tests in 5- and 12-mo-old animals following a 6-h fast. Insulin (1.5 U/kg ip) was administered, and tail blood was analyzed for glucose at 15- to 30-min intervals. Body weights of 5-mo-old wild-type and Ostα^−/− mice were 28 ± 4 and 29 ± 3 g, respectively, whereas 12-mo-old wild-type and Ostα^−/− mice weighed 39 ± 5 and 31 ± 3 g, respectively. C and D: total area under the curve relative to wild-type mice. Values are means ± SE (n = 9–12). *P < 0.05 vs. wild-type.

Fig. 6.

Glucose tolerance and insulin sensitivity in female mice. Glucose tolerance tests were performed in 5-mo-old (A, C, and E) and 12-mo-old (B, D, and F) mice following an overnight fast. Glucose (1 mg/g ip) was administered, and tail blood was analyzed for glucose at 15- to 30-min intervals. In E and F, additional blood was collected from tail snips at 0, 30, 60, and 90 min for insulin analysis. G and H: insulin tolerance tests in 5- and 12-mo-old mice following a 6-h fast. Insulin (1.5 U/kg ip) was administered, and tail blood was analyzed for glucose at 15- to 30-min intervals. Body weights of 5 mo-old wild-type and Ostα^−/− mice were 28 ± 4 and 23 ± 2 g, respectively, whereas 12-mo-old wild-type and Ostα^−/− mice weighed 30 ± 3 and 25 ± 2 g, respectively. C and D: total area under the curve relative to wild-type mice. Values are means ± SE (n = 9–10). *P < 0.05 vs. wild-type.

In 12-mo-old mice, although the differences in glucose tolerance were diminished, there were marked differences in the insulin responses. Fasting plasma insulin concentrations were higher in wild-type than Ostα^−/− mice (Fig. 5F). In addition, the glucose challenge resulted in higher plasma insulin concentrations in wild-type mice at 12 than 5 mo of age, whereas Ostα^−/− mice produced similar low levels of insulin at 5 and 12 mo of age. There was also a greater reduction in plasma glucose during the ITT in 12-mo-old Ostα^−/− than wild-type mice (Fig. 5H). These results suggest that aged Ostα^−/− mice are protected from age-associated decreases in insulin sensitivity.

Glucose tolerance was not significantly different between female wild-type and Ostα^−/− mice at either age (Fig. 6, A–D). Insulin levels were lower in female Ostα^−/− than wild-type mice at 5 and 12 mo (Fig. 6, E and F), and ITTs in female mice suggested improved insulin sensitivity at 5 mo (Fig. 6G). The major difference between male and female animals, however, was that wild-type females did not undergo the same age-related increase in insulin levels as males (cf. Fig. 5, E and F, and Fig. 6, E and F). These results may reflect the fact that female mice did not undergo the same age-related fat accumulation and increased liver and muscle lipids as their male counterparts.

Intestinal bile acid levels determine lipid excretion in Ostα^−/− mice.

Lipid excretion was increased and glucose tolerance and insulin responsiveness were improved in male Ostα^−/− compared with wild-type mice. Bile acid levels were also decreased in liver, gallbladder, and small intestine in Ostα^−/− mice, and bile acid levels and composition are known to play a role in fat absorption and glucose homeostasis. To determine the contribution of bile acid levels to lipid absorption and insulin sensitivity, male mice were fed a diet enriched in 0.2% cholic acid for 3 wk. The cholic acid diet did not affect food consumption, fecal production, or the differences in body and fat pad weight between wild-type and Ostα^−/− mice (data not shown). The cholic acid diet did introduce enough bile acid into the lumen of the small intestine to reduce fecal lipid excretion in Ostα^−/− mice to levels measured in wild-type mice (Fig. 7A). Glucose responses were the same for both genotypes with or without the addition of cholic acid (Fig. 7, B and C). However, the amount of insulin produced during the GTT did differ between wild-type and Ostα^−/− mice, with Ostα^−/− mice displaying substantially lower insulin levels before and after glucose administration (Fig. 7, D and E). Insulin levels were higher after glucose injection in Ostα^−/− mice fed the cholic acid diet than in Ostα^−/− mice fed standard chow but were still below levels measured in wild-type mice (Fig. 7E).

Fig. 7.

Cholic acid-enriched diet restores fecal lipid absorption in male Ostα^−/− mice. Male 12-mo-old wild-type and Ostα^−/− mice were fed a diet enriched with 0.2% cholic acid. A–C: fecal lipid excretion was measured (n = 5; A) and glucose tolerance tests (n = 9–11) were performed before (B) and 1 wk after (C) the animals were switched to the cholic acid-enriched diet (n = 5). D and E: insulin levels during the glucose tolerance tests. Values are means ± SE. *P < 0.05 vs. wild-type.

Improved insulin responses in liver and muscle in Ostα^−/− mice.

Akt is phosphorylated downstream of insulin receptor activation; therefore, Akt phosphorylation provides a direct measure of insulin responsiveness. Wild-type and Ostα^−/− mice at 12 mo of age were injected with saline or insulin, and liver and skeletal muscle tissue were collected 10 min later for immunoblot analysis of total and phosphorylated Akt.

In male liver tissue, insulin-injected wild-type mice showed a 2.2-fold increase in phosphorylated Akt [ratio of phosphorylated to total Akt (p-Akt/Akt)] compared with their saline-injected counterparts, whereas Ostα^−/− mice showed a 3.3-fold increase (Fig. 8). Liver tissue from female mice showed similar responses to insulin, with 2.9-fold (wild-type) and 3.5-fold (Ostα^−/−) increases in Akt phosphorylation (Fig. 9). Interestingly, p-Akt/Akt levels were lower in skeletal muscle tissue from saline-injected Ostα^−/− than saline-injected wild-type mice, indicating lower basal Akt phosphorylation in male and female Ostα^−/− mice. Whereas male and female wild-type mice showed 2.0- and 2.3-fold increases in skeletal muscle p-Akt/Akt following insulin injection, muscle tissue from male and female Ostα^−/− mice showed significantly stronger responses (5.3- and 9.9-fold for males and females, respectively).

Fig. 8.

Insulin-induced phosphorylated Akt (p-Akt) and total Akt expression in liver and muscle tissue from male mice. Male 12-mo-old wild-type and Ostα^−/− mice were fasted overnight. At 10 min after intraperitoneal injection of insulin (1.5 U/kg) or saline, liver (A) and muscle (B) tissue were removed, and Akt phosphorylation was analyzed by Western blotting. Graphs represent densitometric values of p-Akt-to-Akt expression ratio relative to average insulin-stimulated wild-type p-Akt-to-Akt ratio. Values are means ± SE (n = 6, each analyzed 3 separate times). *P < 0.05.

Fig. 9.

Insulin-induced p-Akt and Akt expression in liver and muscle tissue of female mice. Female 12-mo-old wild-type and Ostα^−/− mice were fasted overnight. At 10 min after intraperitoneal injection of insulin (1.5 U/kg) or saline, liver (A) and muscle (B) tissue were removed, and Akt phosphorylation was analyzed by Western blotting. Graphs represent densitometric values of p-Akt-to-Akt expression ratio relative to average insulin-stimulated wild-type p-Akt-to-Akt ratio. Values are means ± SE (n = 4–5, analyzed 3 separate times). *P < 0.05.

DISCUSSION

The present study indicates that the loss of OSTα-OSTβ has beneficial effects on lipid and glucose homeostasis. Ostα^−/− mice were protected from age-related weight gain, and male mice, which gain more weight than female mice during aging, demonstrated improved glucose tolerance and insulin sensitivity, as well as increased lifespan. Interestingly, caloric intake and activity levels in the Ostα^−/− and wild-type mice were similar and do not account for the lower body fat and higher lean body mass of the Ostα^−/− mice. In contrast, increased fecal lipid excretion in the Ostα^−/− mouse, which indicates a defect in lipid absorption, may be an important mechanism contributing to decreased fat accumulation.

Ostα^−/− mice had lower bile acid levels in the enterohepatic circulation at 5 and 12 mo of age. Ostα^−/− mice also have an altered bile acid pool composition, with a lower percentage of cholic acid and a smaller hydrophobic bile acid pool (13). Bile acids facilitate the formation of micelles with dietary cholesterol and other lipid constituents, allowing their absorption, and micelle formation is favored by more hydrophobic bile acids. The changes in bile acid pool size and composition explain why cholesterol and lipid absorption from the small intestine are reduced in Ostα^−/− mice and other mouse models with similar phenotypes (the Cyp7a1^−/− mouse and mice with activated FXR or increased FGF15) (10, 29, 37, 38). Increased fecal lipid excretion in Ostα^−/− mice depended on reduced levels or altered composition of bile acids in the intestinal lumen, because supplementation with 0.2% cholic acid returned lipid excretion to levels measured in wild-type mice. These results confirm a previous report by Lan et al. (13). The decrease in fecal lipid excretion caused by cholic acid feeding was accompanied by, at most, a partial reversal of the improved glucose tolerance and insulin sensitivity in Ostα^−/− mice fed normal chow. This result, obtained when overall body weight and composition were unchanged by the diet, shows that increased lipid excretion in Ostα^−/− mice does not fully account for their improved glucose tolerance and insulin sensitivity.

Despite decreased lipid absorption in Ostα^−/− mice, circulating levels of cholesterol, triglycerides, and fatty acids were not reduced. Synthesis of cholesterol, triglycerides, and fatty acids may have increased to maintain levels measured in wild-type mice. Significant increases in hepatic expression of the cholesterol synthesis genes Hmgcr, Mvd, and Fdft were observed in Ostα^−/− mice, as were small increases in Srebp2. However, expression of the repressor of cholesterol synthesis insulin-induced gene 1 (Insig1) was also increased. The altered expression of genes involved in cholesterol synthesis did not raise hepatic cholesterol levels; in fact, liver cholesterol was lower in Ostα^−/− than wild-type mice at 12 mo of age. Patterns of lipogenic gene expression in the Ostα^−/− mouse were also difficult to interpret. Srebp1c was decreased, implying that lipogenesis is lower in the Ostα^−/− mice, but several downstream targets of SREBP1c (Fas, Acaca, Acacb, Dgat2, and GpatI) were not. Interestingly, the Cyp7a1^−/− mouse, like the Ostα^−/− mouse, has a markedly reduced bile acid pool due to deficient synthesis and a resulting defect in cholesterol absorption, but cholesterol levels are not reduced because of a compensatory increase in cholesterol synthesis (28, 29).

The decrease in bile acid synthesis in Ostα^−/− models has been attributed to hyperplasia of the ileum, providing increased opportunities for bile acid-mediated signaling events. Increased ileal FXR signaling results in increased FGF15, which enters the enterohepatic circulation and acts via FGF receptor 4 in the liver to decrease Cyp7a1 mRNA expression and bile acid synthesis. The central role of FXR signaling was confirmed by Lan et al. (13), who showed lower FGF15, higher Cyp7a1 mRNA, and larger bile acid pools in double-knockout mice lacking Ostα and Fxr than mice lacking Ostα alone. Increased bile acid signaling through FXR may also be potentiated through the increase in IBABP in the ileum of the Ostα^−/− mouse. Bile acids bound to IBABP activate FXR within the ileum (6, 20), and signaling by less hydrophobic bile acids, such as those present in Ostα^−/− mice, is facilitated by IBABP (6).

Important findings in this study are improved glucose tolerance and insulin sensitivity in male Ostα^−/− mice, particularly as they age. Several factors may explain how the altered bile acid and lipid homeostasis in Ostα^−/− mice protect from age-related insulin resistance. Lipid (specifically diacylglycerol) accumulation in insulin-responsive tissues can activate nPKCs, with subsequent impairment of insulin signaling (27). Male Ostα^−/− mice had reduced body fat and hepatic and skeletal muscle lipid content combined with normal or increased lean body mass. This was associated with improved insulin sensitivity in older mice, as reflected by a reduction in insulin secretion during the GTT and a greater decrease in plasma glucose during the ITT. These improvements could be attributed to improvements in insulin signaling, as insulin-stimulated Akt phosphorylation was enhanced in the Ostα^−/− mice. In contrast, female mice accumulated less body fat than male mice as they aged. Female wild-type mice did not display an age-dependent increase in skeletal muscle lipid content. Female Ostα^−/− mice exhibited only minor improvements in glucose handling compared with their wild-type counterparts; however, 12-mo-old female Ostα^−/− mice showed a stronger phosphorylated Akt response to insulin than female wild-type mice of the same age. Taken together, the data support the conclusion that reduced ectopic lipid accumulation in the Ostα^−/− mice enhances insulin signaling and insulin sensitivity.

Bile acids have recently been shown to regulate glucose homeostasis by signaling through FXR and other pathways. FXR^−/− mice have impaired glucose tolerance and insulin sensitivity and increased fasting blood glucose and serum triglyceride levels (18). FXR activation and adenoviral-mediated overexpression of constitutively active FXR improve these parameters in genetic diabetic obese mouse models (14, 18). It is thought that these effects of bile acids are, in part, mediated through FGF15/19 produced in response to intestinal FXR activation. Transgenic mice expressing FGF19 are resistant to the diabetogenic effects of a high-fat diet and display improved glucose tolerance and insulin sensitivity (35), and Fgf15^−/− mice have impaired glucose tolerance that can be restored by FGF19 treatment (11). FGF15/19 inhibits gluconeogenesis by inhibiting the cAMP response element-binding protein-peroxisome proliferator-activated receptor-γ coactivator-1α pathway and increases glycogen synthesis by activating glycogen synthase (11, 24). Because Ostα^−/− mice produce more FGF15 than their wild-type counterparts (13), it is likely that FGF15 also contributes to the improvement in glucose tolerance.

The main findings of this study are summarized in Fig. 10. Regardless of age, Ostα^−/− mice exhibit a phenotype consistent with impaired dietary lipid absorption: reduced bile acid pools and increased fecal lipids. Gene expression of the cholesterol uptake transporter Npc1L1 and the efflux transporter Abca1 are decreased in the small intestine of Ostα^−/− mice. Expression of the bile acid transporter Mrp3 is increased, partially compensating for the loss of OSTα-OSTβ. This study confirms previous work showing that the loss of OSTα-OSTβ and the subsequent reduction in bile acid pool size impair dietary lipid absorption and alter bile acid signaling pathways and demonstrates that these effects persist throughout life. A novel and important discovery is that mice lacking OSTα-OSTβ gain less weight, accumulate less lipid in liver and skeletal muscle, and remain more insulin-sensitive as they age, providing compelling evidence that OSTα-OSTβ is a physiological regulator of carbohydrate, as well as lipid, metabolism.

Fig. 10.

Age-independent phenotype of Ostα^−/− mice (left) and age-related changes in wild-type and Ostα^−/− mice (right). In Ostα^−/− mice, lower concentrations of bile acids (BA) impair absorption of dietary lipids (○) in the small intestine. Lipid transporters in enterocytes are depicted in the cell at left and bile acid transporters in the cell at right. Differences in expression of genes encoding transporters are shown schematically by font size. ABCA1, ATP-binding cassette transporter 1; SRB1, scavenger receptor class B type 1; NPC1L1, Niemann-Pick C1-like 1; MRP3, multidrug resistance-associated protein 3.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Research Grants DK-067214 (N. Ballatori and P. M. Hinkle) and DK-19974 (P. M. Hinkle), National Institute of Environmental Health Sciences Training Grant ES-070236 (S. G. Wheeler) and Center Grant ES-01247 (N. Ballatori), NIDDK Grants DK-40936 (G. I. Shulman) and U24 DK-076169 (V. T. Samuel and G. I. Shulman), Veterans Health Administration Biomedical Laboratory Research and Development Grant I01 BX-000901 (V. T. Samuel), and NIDDK Grants DK-25636 and DK-34989 (J. L. Boyer).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

S.G.W., C.L.H., C.J.S., J.L.B., P.M.H., and N.B. are responsible for conception and design of the research; S.G.W., C.L.H., and F.R.J. performed the experiments; S.G.W., C.L.H., F.R.J., V.T.S., and G.I.S. analyzed the data; S.G.W., C.L.H., F.R.J., V.T.S., G.I.S., and P.M.H. interpreted the results of the experiments; S.G.W. and C.L.H. prepared the figures; S.G.W., C.L.H., and P.M.H. drafted the manuscript; S.G.W., C.L.H., F.R.J., V.T.S., G.I.S., C.J.S., J.L.B., and P.M.H. edited and revised the manuscript; S.G.W., C.L.H., F.R.J., V.T.S., G.I.S., C.J.S., J.L.B., and P.M.H. approved the final version of the manuscript.