The 16α-hydroxylated Bile Acid, Pythocholic Acid Decreases Food Intake and Increases Oleoylethanolamide in Male Mice

Abstract

Modulation of bile acid (BA) structure is a potential strategy for obesity and metabolic disease treatment. BAs act not only as signaling molecules involved in energy expenditure and glucose homeostasis, but also as regulators of food intake. The structure of BAs, particularly the position of the hydroxyl groups of BAs, impacts food intake partly by intestinal effects: (1) modulating the activity of N-acyl phosphatidylethanolamine phospholipase D, which produces the anorexigenic bioactive lipid oleoylethanolamide (OEA) or (2) regulating lipid absorption and the gastric emptying-satiation pathway. We hypothesized that 16α-hydroxylated BAs uniquely regulate food intake because of the long intermeal intervals in snake species in which these BAs are abundant. However, the effects of 16α-hydroxylated BAs in mammals are completely unknown because they are not naturally found in mammals. To test the effect of 16α-hydroxylated BAs on food intake, we isolated the 16α-hydroxylated BA pythocholic acid from ball pythons (Python regius). Pythocholic acid or deoxycholic acid (DCA) was given by oral gavage in mice. DCA is known to increase N-acyl phosphatidylethanolamine phospholipase D activity better than other mammalian BAs. We evaluated food intake, OEA levels, and gastric emptying in mice. We successfully isolated pythocholic acid from ball pythons for experimental use. Pythocholic acid treatment significantly decreased food intake in comparison to DCA treatment, and this was associated with increased jejunal OEA, but resulted in no change in gastric emptying or lipid absorption. The exogenous BA pythocholic acid is a novel regulator of food intake and the satiety signal for OEA in the mouse intestine.

bile acids, pythocholic acid, OEA, NAPE-PLD, gastric emptying

Highlights

• Pythocholic acid decreases food intake.

• Pythocholic acid increases intestinal oleoylethanolamide.

• The effects of pythocholic acid on oleoylethanolamide and hypophagia are greater than the effects of deoxycholic acid.

• Pythocholic acid does not affect lipid absorption or gastric emptying.

Bile acids (BAs) are synthesized from cholesterol in the liver and promote lipid absorption in the intestine. They act as signaling molecules by activating BA receptors, Takeda G protein-coupled receptor 5 (TGR5) (1-3), and transcription factor farnesoid X receptor (FXR) (4-6). Activation of TGR5 increases energy expenditure and prevents diet-induced obesity (7, 8). TGR5 signaling is also involved in insulin secretion and GLP-1 secretion (9-14). FXR activation mitigates insulin resistance in diabetic rodents by improving insulin signaling (15) and GLP-1 secretion (16) and inhibiting gluconeogenic gene expression (17). Recently, BAs were shown to act in the hypothalamus to stimulate satiety (8, 18). Thus, BA receptors may be a potential therapeutic target for obesity and metabolic diseases because of their traits as signaling molecules involved in energy balance and glucose homeostasis.

In addition to the involvement of BA receptors in energy balance and glucose homeostasis, BAs can also regulate food intake via BA receptor-independent mechanisms. One is an intestinal lipid-sensing mechanism. Our previous study indicated that BAs regulate gastric emptying and satiation by determining lipid access into the distal intestine and the distally abundant fat-sensing receptor, GPR119 (19). A second mechanism is via BA-induced allosteric modulation of bioactive lipid synthesis. Magotti et al showed that BAs hydroxylated at carbon 3α and 12α bind and activate the enzyme N-acyl phosphatidylethanolamine phospholipase D (NAPE-PLD), an enzyme that produces fatty acid ethanolamides (FAEs) (20). In line with this, we showed that oleoyl ethanolamide (OEA) concentrations were reduced in the jejunal epithelium of mice lacking the 12α-hydroxylated BAs, cholic acid and deoxycholic acid (DCA) (19). Jejunal OEA is a bioactive signaling lipid that induces satiety, defined as an increase in the time between meals, via PPARα activation and the vagally mediated gut-brain axis (21-24). Thus, the structure of BAs, particularly the position of the hydroxyl groups, determines the production of the satiety molecule OEA.

Enhancing intestinal OEA production may be a therapeutic strategy to treat obesity and metabolic diseases based on clinical and animal studies. A human study suggests that (1) impaired OEA synthesis is a cause of obesity and (2) OEA treatment is effective for weight loss. A Norwegian population-based cohort study shows that a single nucleotide polymorphism of NAPE-PLD is associated with severe obesity (25). OEA treatment decreases body weight and fat mass by reducing the desire to eat and appetite for sweet foods in obese subjects (26, 27). Additionally, preclinical studies have established the physiological role of OEA. Intestinal OEA induces dopamine release in the brain to reduce appetite for fatty food in mice (28). OEA treatment reduces total food intake and meal size in diet-induced obese animal models as well as a mouse model of Prader-Willi syndrome, which is a genetic disorder that causes childhood-onset hyperphagia and obesity (29, 30). The mechanism of OEA-induced satiety and hypophagia is suggested to be mediated by PPARα activation in the intestine (22, 23). However, some adverse effects of systemic PPARα activation have been reported such as hepatocellular carcinoma or chemotaxis (31-33). Overexpression of PPARα induces lipid accumulation in cardiomyocytes (34, 35) and glucose intolerance in skeletal muscle (36). Thus, locally increasing intestinal OEA presents an attractive alternative against obesity because it may mitigate the side effects of systemic PPARα activation.

Increasing intestinal OEA locally is challenging because the chemical compounds that promote OEA in the intestine are not known. To address this, we focused on an atypical research animal, pythons, which have a unique system to promote and sustain high levels of OEA in the intestine. In pythons, jejunal OEA is increased 300-fold and sustained for at least 2 days after a meal (37). For comparison, jejunal OEA levels in rodents are increased 2- to 10-fold after a meal and are reduced to premeal levels within a few hours (19, 23, 38, 39). Pythons consume large meals at infrequent intervals (∼2-12 months). The dramatic increase in postprandial OEA in pythons is considered a potential mechanism contributing to pythons’ long intermeal intervals (40). However, the regulation of intestinal OEA production in pythons is incompletely defined.

In this work, we examined the role of the unique BA in pythons, pythocholic acid, in the production of OEA. Pythocholic acid was discovered in 1950 by Haslewood and Wootton as a major BA in pythons (41, 42). Pythocholic acid is a unique BA because of its structure, which has hydroxyl groups at positions 3α, 12α, and 16α (42). Hydroxylation at 16α is common in specific snakes (Cylindrophiidae, Uropeltidae, Boidae, and Pythonidae) and birds (Shoebill), but is not seen in mammals (43). Pythocholic acid has been found in specific snakes such as boas and pythons that possess long meal intervals (43). We hypothesized that the unique structure of pythocholic acid promotes NAPE-PLD activity and increases intestinal OEA production (19, 20, 44, 45). Here, we tested the effects of pythocholic acid on OEA production and satiety in mice.

Materials and Methods

Animals

We used 8-week-old male C57BL/6J mice (The Jackson Laboratory #000664, RRID:IMSR_JAX:000664). The mice were fed a normal chow diet (3.4 kcal/g, Purina 5053; 24.7% kcal from protein, 62.1% carbohydrate, and 13.2% fat). The mice were provided with the diet and water ad libitum and maintained on a 12-hour light/dark cycle, with lights on at 7 Am. All experiments were approved and conducted according to the guidance of Columbia University and St. John's University Institutional Animal Care and Use Committee.

Food Intake Measurements

For food intake measurements, 8-week-old mice were individually housed, and the food dispenser was located inside the cage. The mice were trained to take food from the food dispenser for 3 days before the food intake measurement.

Gastric Emptying

Solid gastric emptying was measured following ingestion of a normal chow diet as we previously reported (19). We measured individual food intake and stomach contents to determine the percentage of food eaten that had emptied from the stomach (19, 46-48). To estimate the accurate stomach contents, the mice were fasted for 16 hours with free access to water, then allowed access to a chow diet for 1 hour (19, 46-48). The mice were food-deprived again for 2 hours before euthanasia. Food intake during the 1-hour feeding period was measured by using a food dispenser. Food content in the stomach was measured at euthanasia. Solid gastric emptying was calculated by the following formula: solid gastric emptying (%) = {1 – (food content in stomach/food intake ×100. For python bile or BA treatment, mice were orally gavaged with 20 mg/kg of BAs in 1.5% NaHCO₃ or an equivalent dose of pythocholic acid-containing python bile for 2 consecutive days at 6 Pm.

Extraction of Pythocholic Acid From Python Bile

Euthanized ball pythons (Python regius) were donated by Mr. Dustin Leahy (Piedthonidae Exotics). Bile was collected from the gall bladder. Pythocholic acid (PubChem CID 5283890, https://pubchem.ncbi.nlm.nih.gov/compound/5283890) and tauro-conjugated pythocholic acid were purified using HyperSep C8 solid-phase extraction (SPE) cartridge (Thermo Scientific, Waltham, MA, Cat # 60108-309) and silica gel chromatography. Python bile (0.5 mL) was suspended in 2 mL of aqueous HCl (0.01 M). The acidified suspension was extracted with ethyl acetate (2 mL × 5). The ethyl acetate extract was subjected to liquid chromatography/mass spectrometry (LC/MS) on Agilent iFunnel 6550 Q-ToF LC/MS System. The total ion chromatogram in the positive ion mode showed a major peak with m/z 431.2769, which was consistent with the sodium adduct of pythocholic acid [M + Na]⁺ (calculated m/z 431.2765). The ethyl acetate extract was then subjected to further purification with SPE cartridge, from which the major constituent was eluted with 20% methanol in water. NMR of the purified material (4 mg, clear solid) confirmed its identity as pythocholic acid (49). The aqueous layer from the ethyl acetate extraction was also subjected to LC/MS to give a predominant peak with m/z 516.2999, which corresponded to the protonated form of tauro-conjugated pythocholic acid [M + H]⁺ (calculated m/z 516.2989). The major compound was purified with silica gel chromatography to give a clear solid (45.5 mg, R_f 0.2, 30% methanol in dichloromethane). The steroidal moiety of the purified material was confirmed to be pythocholic acid after deconjugation in the base (50).

Blood and Plasma Analysis

Blood glucose was measured in mice by tail vein bleeding using a OneTouch glucose monitor and strips (LifeScan). One hundred microliters of blood samples were collected and mixed with 1% DPP-4 inhibitor (Millipore) and 5% aprotinin (Fisher) and kept on ice. Plasma was obtained after centrifuging blood collected in EDTA-coated tubes for 15 minutes at 2500g. Plasma GLP-1 was measured by using an immunoassay (total GLP-1 assay kit; Meso Scale Discovery Cat# K150JVC, RRID:AB_2801383). Plasma insulin was measured using an ELISA kit (Mercodia Cat# 10-1247-01, RRID:AB_2783837) according to the manufacturer's protocol. Plasma alanine aminotransferase (ALT) (MAK052, Sigma) and aspartate aminotransferase (AST) (MAK055, Sigma) were measured using dedicated kits according to the manufacturer's protocol.

Fecal Lipid Extraction and Measurement

To evaluate lipid absorption, feces were collected for 24 hours from each cage after 2 consecutive days of treatment of BA or python bile. Collected feces were dried in an incubator for 16 hours at 42 °C and 100 mg of feces were homogenized with 1 mL of 1 M NaCl. One milliliter of homogenized solution was added into 6 mL of chloroform: methanol solution (2:1). Chloroform layers were collected after centrifugation, then evaporated under N₂ flow until dry. One milliliter of 2% Triton X-100 in chloroform was added and evaporated again under N₂ flow. One milliliter of ddH₂O was added and vortexed until the sample dissolved. Fecal free fatty acid (FFA, Wako; HR Series NEFA-HA(2), ), triglycerides (Infinity; Thermo Scientific) and cholesterol (Cholesterol E, Wako Diagnostics) content were assessed using a colorimetric assay.

Bile Acid Profile

The liver samples were homogenized in 50% methanol using a rotor-stator homogenizer. Deuterated BA standard (20 mL of 25 mM d4-cholic acid) was added to 200 mL of each sample, and calibrator curves were generated for each BA in charcoal-stripped tissue. To each sample/calibrator, 2 mL of ice-cold acetonitrile was added, then samples were vortexed for 1 hour at 2000 rpm and centrifuged for 10 minutes at 11 000g. Supernatants were transferred to a clean glass tube and dried down at 45 °C under N₂. Each sample/calibrator was extracted a second time in 1 mL of ice-cold acetonitrile, vortexed for 1 hour at 2000 rpm and centrifuged for 10 minutes at 11 000g. The supernatant of the second extraction was combined with the first and dried down at 45 °C under N₂. Each sample was resuspended in 200 mL of 55:45 (vol/vol) methanol:water, both with 5 mM ammonium formate. Samples were centrifuged in UltraFree MC 0.2-mm centrifugal filters (Millipore) and transferred to LC-MS vials, and 10 mL were injected into ultraperformance LC-tandem MS vials (Waters Corporation, Milford, MA, USA).

Intestinal Lipid Extraction and Analysis of FAEs and Monoacylglycerols

Lipid extraction and analysis were performed as previously described (51-53). Frozen tissue of the mucosal intestine was blade-homogenized in 1 mL of a methanol solution containing the internal standards, [²H₅] 2-arachidonylglycerol (2-AG), [²H₄]-anandamide (AEA), [²H₄]-OEA (Cayman Chemical, Ann Arbor, MI, USA). Lipids were extracted with 2 mL of chloroform and washed with 1 mL of water. Organic (lower) phases were collected and separated by open-bed silica gel column chromatography as previously described (54). Eluate was gently dried under N₂ stream (99.998% pure) and resuspended in 0.2 mL of methanol:chloroform solution (9:1), with 1 μL injection for ultra-performance LC/tandem MS analysis.

Data acquisition was performed using an Acquity I Class UPLC with in-line connection to a Xevo TQ-S Micro Triple Quadrapole Mass Spectrometer (Waters Corporation) and accompanying electrospray ionization (ESI) sample delivery. Lipids were separated using an Acquity UPLC BEH C₁₈ column (2.1 × 50 mm inner diameter, 1.7 µm, Waters) and inline guard column UPLC BEH C₁₈ VanGuard PreColumn (2.1 × 5 mm inner diameter; 1.7 µm, Waters). Lipids were eluted by a gradient of water and methanol (containing 0.25% acetic acid, 5 mM ammonium acetate) at a flow rate of 0.4 mL/min and gradient: 80% methanol 0.5 minutes, 80% to 100% methanol 0.5 to 2.5 minutes, 100% methanol 2.5 to 3.0 minutes, 100% to 80% methanol 3.0 to 3.1 minutes, and 80% methanol 3.1 to 4.5 minutes. The column was maintained at 40 °C and samples were kept at 10 °C in sample manager. MS detection was in positive ion mode with capillary voltage maintained at 1.10 kV and Argon (99.998%) was used as collision gas. Cone voltages and collision energies for respective analytes: AEA = 30v, 14v; OEA = 28v, 16v; docosahexaenoyl ethanolamide (DHEA) = 30v, 16v; 2-AG (20:4) = 30v, 12v; 2-oleoylglycerol (2-OG; 18:1) = 42v, 10v; 2-decanoyl-rac-glycerol (2-DG; 22:6) = 34v, 14v; 2-linoleoylglycerol (2-LG; 18:2) = 30v, 10v; [²H₄]-AEA = 26v, 16v; [²H₄]-OEA = 48v, 14v; and [²H₅]-2-AG = 25v, 44v. Lipids were quantified using a stable isotope dilution method detecting proton or sodium adducts of the molecular ions [M + H/Na]⁺ in multiple reaction monitoring mode. For many monoacylglycerols, acyl migration from sn-2 to sn-1 glycerol positions is known to occur; for these analytes, the sum of these isoforms is presented. Tissue processing and LC/MS analyses for experiments occurred independently of other experiments. Extracted ion chromatograms for multiple reaction monitoring transitions were used to quantitate analytes: AEA (m/z = 348.3 > 62.0), OEA (m/z = 326.4 > 62.1), DHEA, (m/z = 372.3 > 62.0), 2-AG (m/z = 379.3 > 287.3), 2-OG (m/z = 357.4 > 265.2), 2-DG (m/z = 403.3 > 311.2), and 2-LG (m/z = 355.3 > 263.3), with [²H₄]-AEA (m/z = 352.4 > 66.1) as the internal standard for AEA and DHEA, [²H₄]-OEA (m/z = 330.4 > 66.0) as the internal standard for OEA, and [²H₅]-2-AG (m/z = 384.3 > 93.4) as internal standard for 2-AG, 2-OG, 2-DG, and 2-LG. One “blank” sample was processed and analyzed in the same manner as all samples, except no tissue was included. This control revealed no detectable endocannabinoids and related lipids included in our analysis.

Statistics

Results are presented as mean ± SEM. Data were analyzed by 1-way ANOVA with Tukey multiple comparisons test or Student t-test.

Results

Python Bile Treatment Reduced Food Intake Independent of Slow Gastric Emptying and Lipid Absorption

Pythocholic acid is not commercially available. Thus, we first used python bile to test the effect of pythocholic acid on food intake because the major bile acid (>70%) in pythons is reported to be pythocholic acid (42). We measured pythocholic acid in the bile of ball pythons by LC-tandem MS and found 6 mg/mL of pythocholic acid in the bile of ball pythons. We treated wild-type mice with a single dose of vehicle or python bile (100 μL, which is estimated to yield a dose of approximately 20 mg/kg of pythocholic acid) by oral gavage for food intake measurement. As a positive control, we treated a parallel set of mice with DCA because kinetic experiments demonstrate that DCA increases NAPE-PLD activity better than other BAs (45). Mice were fed ad libitum before the gavage, which was delivered at 6 Pm. We measured food intake every 3 or 6 hours. DCA decreased food intake in the first 3 hours after treatment (Fig. 1A), but subsequent food intake returned to normal (Fig. 1B-F). On the other hand, python bile decreased food intake more profoundly (Fig. 1A and 1B), and this was sustained over 6 hours before returning to normal (Fig. 1C-F). Python bile strongly decreased cumulative food intake for 24 hours relative to DCA treatment (Fig. 1G). We then examined putative mechanisms underlying the hypophagic effect of python bile in mice.

Figure 1.

Ad libitum food intake after single dose administration of deoxycholic acid (DCA) or python bile at 6 Pm. Food intake after treatment (A) 0 to 3 hours, (B) 3 to 6 hours, (C) 6 to 9 hours, (D) 9 to 12 hours, (E) 12 to 18 hours, (F) 18 to 24 hours, and (G) total food intake for 24 hours (n = 6-7 for each group). *P < .05, **P < .01 (1-way ANOVA).

Lipids in the distal small intestine suppress food intake by slowing gastric emptying and inducing satiation (55-58). Previously, we reported that modulation of bile acid composition by Cyp8b1 deficiency caused slow gastric emptying because of impaired lipid absorption (19). To evaluate the effect of python bile on gastric emptying and lipid absorption, we gave DCA or python bile once per day for 2 days before measuring solid phase gastric emptying. After 2 days of treatment with DCA or python bile, we fasted mice for 16 hours, then allowed them to access chow for 1 hour, then removed food for 2 hours (Fig. 2A). We measured individual food intake and stomach contents to determine the percentage of food eaten that had emptied from the stomach. There was no significant difference in gastric emptying after DCA or python bile treatment (Fig. 2B). We determined that “refed” food intake was not different during the 1-hour feeding period after 16 hours of fasting (Fig. 2C). DCA and python bile did not affect body weight or glucose levels during the experiment (Fig. 2E and 2F). Unexpectedly, mice treated with python bile had lower plasma insulin levels after 1-hour refeeding compared with the vehicle treatment group (Fig. 2G). However, there were no differences in plasma insulin levels between groups at the 16-hour fasting or 2-hour refasting time points (Fig. 2G). We observed an interesting increase in the plasma GLP-1 levels in the python bile-treated group at the 2-hour refasting time points (Fig. 2D). However, with no increase in insulin and no suppression of gastric emptying, the significance of the elevated GLP-1 is not yet known. Snake bile from cobras or vipers is potentially toxic, but pythocholic acid has not been identified in their bile (59, 60). To rule out the possibility that python bile is toxic, we measured plasma AST and ALT. At the end of the experiment, we observed no differences in plasma AST and ALT levels resulting from DCA or python bile treatment, suggesting that oral gavage of python bile at hypophagic doses is not toxic for mice (Fig. 2H and 2I).

Figure 2.

Effects of DCA and python bile on gastric emptying, blood glucose, and insulin levels after 2 days of once-daily treatment. (A) Experimental schedule, (B) gastric emptying, (C) 1-hour refed food intake after fasting, (D) plasma GLP-1 levels, (E) body weight, (F) blood glucose levels, (G) plasma insulin levels, (H) plasma AST levels, and (I) plasma ALT levels. (n = 6-7 for each group). *P < .05 (1-way ANOVA).

It is well-known that BAs promote lipid absorption and the effects of BA composition on lipid absorption can contribute to systemic energy homeostasis (19, 61-64). To evaluate the lipid absorption after 1 day of DCA and python bile treatment, we measured fecal excretion of FFAs, acylglycerols, and cholesterol. We did not find any significant differences in these parameters between treatment groups (Fig. 3 A-C). These data imply that python bile induced a hypophagic effect that was independent of gastric emptying and lipid absorption.

Figure 3.

Influence of DCA and python bile on lipid absorption. (A) Fecal free fatty acid, (B) fecal acylglycerol, and (C) fecal cholesterol levels. Feces were collected for 24 hours before the gastric emptying experiment (n = 6-7 for each group).

Extraction and Purification of Pythocholic Acid From Python Bile

The bile in pythons contains not only bile acids but also other lipids and biliary contents. The bile from ball python that we used contained 957 mM of FFA, 5.7 mg/dL of triglycerides, and 70.8 mg/dL of cholesterol. To specifically test the effects of pythocholic acid on food intake, we extracted and purified the pythocholic acid from ball python bile. Python bile was first acidified to fully protonate pythocholic acid, which made it easier to extract with ethyl acetate (Fig. 4A). The tauro-conjugate remained in the aqueous phase during the extraction because of its polar taurine moiety. LC/MS analyses revealed molecular ions that are consistent with pythocholic acid and tauro-conjugate in the ethyl acetate extract and the aqueous phase, respectively (Fig. 4B-E). The ethyl acetate extract was further purified by reversed-phase C₈ SPE to give pythocholic acid (4 mg), whereas the aqueous phase was subjected to silica gel chromatography to obtain tauro-conjugated pythocholic acid (45.5 mg).

Figure 4.

Purification of pythocholic acid and its taurine conjugate from python bile. (A) Python bile was first fractionated by partition between organic (ethyl acetate) and aqueous phases. The ethyl acetate phase was further purified by C8 solid phase extraction to give pythocholic acid, whereas the aqueous phase was subjected to silica gel chromatography to give tauro-conjugate. (B, C) Liquid chromatography/mass spectrometry (LC/MS) analysis of the ethyl acetate extract from python bile. (B) Total ion chromatogram (TIC) in the positive mode. The major peak was observed around 4.2 to 4.3 minutes. (C) High-resolution electrospray mass spectrometry (HR ESI MS) of the major peak (4.2-4.3 minutes). The MS signal at m/z 431.2769 was consistent with the sodium adduct of pythocholic acid [M + Na]⁺ (calculated m/z 431.2765). (D-E) LC/MS analysis of the aqueous phase from python bile (after ethyl acetate extraction). (D) TIC in the positive mode. The major peak was observed around 3.5 minutes. (E) HR ESI MS of the major peak (∼3.5 minutes). The MS signal at m/z 516.2999 was consistent with the protonated form of tauro-conjugated pythocholic acid [M + H]⁺ (calculated m/z 516.2989).

Pythocholic Acid Treatment Is Sufficient to Reduce Food Intake and Does Not Influence Gastric Emptying or Lipid Absorption

To test the effect of pythocholic acid on food intake, we gave a single dose of pythocholic acid (20 mg/kg) by oral gavage to ad libitum-fed mice. DCA (20 mg/kg) was given to mice as a positive control. Pythocholic acid and DCA were delivered at 6 Pm of the test day, and food intake was measured every 3 or 6 hours. Pythocholic acid decreased food intake in the first 6 hours and between 9 and 12 hours after treatment (Fig. 5A, 5B, and 5D). Pythocholic acid treatment decreased food intake compared with DCA between hours 3 through 6 and 9 through 18 (Fig. 5B, 5D, and 5E). There were no significant differences between groups during the 6- to 9-hour and 18- to 24-hour time points (Fig. 5C and 5F). Cumulatively, pythocholic acid significantly decreased food intake over 24 hours compared with the vehicle or DCA groups (Fig. 5G).

Figure 5.

Ad libitum food intake after single-dose administration of DCA or pythocholic acid (PA) at 6 Pm. Food intake after treatment (A) 0 to 3 hours, (B) 3 to 6 hours, (C) 6 to 9 hours, (D) 9 to 12 hours, (E) 12 to 18 hours, (F) 18 to 24 hours, and (G) total food intake for 24 hours (vehicle; n = 12, DCA; n = 11, PA; n = 10). *P < .05, **P < .01 (1-way ANOVA).

To evaluate the effect of pythocholic acid on gastric emptying, DCA or pythocholic acid was given by oral gavage once per day for 2 days (Fig. 6A). There were no significant differences in gastric emptying after DCA or pythocholic acid treatment in mice (Fig. 6B). Food intake after overnight fasting was not different during the 1-hour refeeding period between any groups (Fig. 6C), and there were no differences in plasma GLP-1 levels, body weight glucose levels, or insulin levels during the experiment (Fig. 6D-G). To evaluate the BA composition after pythocholic acid treatment, we measured BA levels in the liver. Pythocholic acid and tauro-conjugated pythocholic acid were detected in liver samples of the pythocholic acid-treated mice. However, there were no differences between all groups in the total amount of conjugated or unconjugated BAs and no difference in the total amounts of BAs (Fig. 6H and 6I). At the end of the experiment there were no differences in plasma AST and ALT levels between groups (Fig. 6J and 6K). We also did not find any differences in fecal FFA, acylglycerol, or cholesterol levels between DCA and pythocholic acid-treated groups (Fig. 7A-C). Thus, we concluded that pythocholic acid is sufficient to induce hypophagia independent of gastric emptying and lipid absorption.

Figure 6.

Effects of DCA and PA on gastric emptying, blood glucose, and insulin levels after 2 days of once-daily treatment. (A) Experimental schedule, (B) gastric emptying, (C) 1-hour refed food intake after fasting, (D) plasma GLP-1 levels, (E) body weight, (F) blood glucose levels, (G) plasma insulin levels, (H) BA species measured by UPLC-MS/MS from liver samples, (I) the absolute quantities of BAs, (J) plasma AST levels, and (K) plasma ALT levels. (vehicle; n = 12, DCA; n = 11, PA; n = 10). *P < .05, vehicle vs PA, **P < .01, Vehicle vs PA, ^#P < .05, DCA vs PA, ^##P < .01, DCA vs PA (1-way ANOVA). α-MCA, α-muricholic acid; β-MCA, β-muricholic acid; PA, pythocholic acid; CA, cholic aid; CDCA, chenodeoxycholic acid; DCA, deoxycholic acid; GCA, glycocholic acid; GCDCA, glycochenodeoxycholic acid; GDCA, glycodeoxycholic acid; GUDCA, glycoursodeoxycholic acid; HDCA, hyodeoxycholic acid; LCA, lithocholic acid; TabMCA, tauro-α- and tauro-β-muricholic acid; TCA, tauro-conjugated cholic acid; TCDCA, taurochenodeoxycholic acid; TDCA, taurodeoxycholic acid; TLCA, taurolithocholic acid; TPA, tauro-conjugated pythocholic; TUDCA, tauroursodeoxycholic acid; UDCA, ursodeoxycholic acid.

Figure 7.

Influence of DCA and PA on lipid absorption. (A) Fecal free fatty acid, (B) fecal acylglycerol, and (C) fecal cholesterol levels. Feces were collected for 24 hours before the gastric emptying experiment; (vehicle; n = 12, DCA; n = 11, PA; n = 10).

Pythocholic Acid Promoted Jejunal NAPE-PLD-Dependent Fatty Acid Ethanolamide Synthesis

Another mechanism by which intestinal lipids suppress food intake is by promoting synthesis of FAEs. In particular, intestinal OEA synthesized from dietary lipid after a meal is known to induce satiety (22, 23, 38, 65). Recent studies indicate that BAs stabilize and activate the FAE-producing enzyme NAPE-PLD (20, 45). To investigate the involvement of pythocholic acid in NAPE-PLD and FAE production, we measured 3 FAE species (OEA, AEA, and DHEA) in jejunal epithelia. Mice were given pythocholic acid or DCA once per day for 2 days. We collected jejunal epithelia after food intake and gastric emptying experiment (Fig. 8A). Jejunal OEA levels after refeeding were robustly increased in pythocholic acid-treated mice relative to the vehicle or DCA groups (Fig. 8B). We did not find differences in other FAE such as AEA and DHEA (Fig. 8C and 8D).

Figure 8.

Effects of DCA and PA on fatty acid ethanolamides and cannabinoids in jejunal intestine after refeeding. (A) Experimental schedule, effect of DCA and PA after refeeding on the levels of (B) OEA, (C) AEA, (D) DHEA, (E) 2-AG, (F) 2-OG, (G) 2-DG, and (H) 2-LG. (vehicle; n = 12, DCA; n = 11, PA; n = 10). *P < .05 (1-way ANOVA).

We measured several other bioactive lipids, 2-AG, 2-OG, 2-DG, and 2-LG, which are synthesized from phospholipids by diacylglycerol lipase. We found no differences in jejunal 2-AG, 2-OG, 2-DG, or 2-LG between treatment groups (Fig. 8E-H). These data indicate that pythocholic acid specifically promotes production of OEA but not other FAE or diacylglycerol lipase-dependent bioactive lipids.

Discussion

In this study, we successfully isolated pythocholic acid from pythons for experimental use and tested the effect of pythocholic acid on food intake in mice. Our data showed that pythocholic acid treatment decreased ad libitum food intake and increased OEA in the jejunum. In addition, the anorexigenic effect of pythocholic acid was independent of lipid absorption and gastric emptying. Our findings indicate that pythocholic acid is a novel tool to promote intestinal OEA and suppress food intake.

The biological pathway of OEA in production and deactivation has been studied over the past 20 years. Feeding promotes jejunal OEA formation in chow diet-fed lean rodents (19, 23, 38, 39). There are 3 essential steps to produce and maintain OEA: (1) production of oleic acid-containing NAPEs, (2) NAPE-PLD activation to produce OEA from oleic acid-containing NAPEs, and (3) inhibition of fatty acid amide hydrolase that hydrolyzes OEA (24, 38). Among these biological events that maintain OEA, BAs are known to play a role in NAPE-PLD activation. Hydroxyl groups at carbon 3α and 12α in BAs bind the substrate-binding pocket of the NAPE-PLD dimer, which reinforces the enzyme's stability and activity (20). Kinetic experiments demonstrated that DCA (3α, 12α) increases NAPE-PLD activity better than chenodeoxycholic acid, another dihydroxy BA (3α, 7α) (45). Our data showed that the tri-hydroxylated (3α, 12α, 16α) BA, pythocholic acid, increased jejunal OEA levels much higher than DCA in mice. But we found no effect of pythocholic acid on other bioactive lipids levels such as AEA, DHEA, 2-AG, 2-OG, 2-DG, and 2-LG. These findings suggest that the specific structure of pythocholic acid activates OEA synthesis through NAPE-PLD activity in the intestine. However, further studies would be needed to show the direct binding between NAPE-PLD and pythocholic acid.

Though the mechanism is not known, OEA mobilization is disrupted in diet-induced obese rodents (66). On the other hand, OEA treatment decreases food intake and body weight in diet-induced obese rodents (29). This evidence suggests that impaired OEA synthesis contributes to obesity. Examining the effect of pythocholic acid on food intake, weight gain, and glucose homeostasis in diet-induced obese rodents is a future direction of this work.

Another possible pathway to regulate food intake by BAs is TGR5-mediated signaling in the brain. Feeding or oral gavage of BAs allows access of BAs into the hypothalamus, which decreases food intake (18). Intracerebroventricular injection of TGR5 specific agonist, 3-(2-chlorophenyl)-N-(4-chlorophenyl)-N,5-dimethyl-4-isoxazolecarboxamid, decreases food intake (8). Deletion of TGR5 in AgRP neurons cancels the anorexigenic effect of TGR5 agonist INT-777 (18). Thus, TGR5 signaling in the central nervous system is an alternative pathway for food intake regulation by BAs. Further studies to evaluate the transport of pythocholic acid into the brain after oral gavage and its effects on TGR5 will be of interest to fully understand the anorexigenic mechanism(s) of pythocholic acid.

In conclusion, our findings reveal an anorexigenic effect of the python-derived 16α hydroxylated bile acid, pythocholic acid. Results in this study indicate that modification of BA, particularly 16α hydroxylation or pythocholic acid treatment, may be a novel strategy for obesity and metabolic disease treatment.

Acknowledgments

The authors thank members of the Haeusler laboratory, Dr. Domenico Accili (Columbia University), Dr. Kimberly Lackey, Dr. Stephen Secor (University of Alabama), Mr. Dustin Leahy (Piedthonidae Exotics), Dr. Reiko Toyama (NICHD) and Dr. Jurgen Wess (NIDDK), and Arclev Academia Strategists Network members for helpful discussions, suggestions, and/or providing materials. The authors are grateful to Columbia University Biomarkers Core Laboratory and the Core Laboratory director Dr. Renu Nandakumar for assistance with the BA profiling. We thank Ms. Tasfia I. Anushka and Ms. Elma Radoncic (St. John's University) for proofreading our manuscript. This research idea originated from the pet snake (Albino ball python) of the Higuchi household, so we thank Reiri and Akira Higuchi for inspiring research ideas and caring for their pet snake.

Funding

Funding was from the National Institutes of Health R01DK115825 to R.A.H., R01DK119498 to N.V.D., the American Diabetes Association 7-20-IBS-130 to R.A.H., NIH-funded center grants P30DK026687, P30DK132710, and UL1TR001873 (Biomarkers Core), and by faculty startup funds to S.H. from the College of Pharmacy and Health Sciences, St. John’s University.

Author Contributions

S.H. and R.A.H conceived and designed research; S.H., C.W., L.J.G., R.H.N., V.M., R.D., and A.K. performed experiments; S.H., N.V.D., and A.K. analyzed data; S.H. and R.A.H drafted the manuscript. All authors edited the manuscript.

Disclosures

The authors have nothing to disclose.

Data Availability

Original data generated and analyzed during this study are included in this published article.

Abbreviations

 
• 2-AG

  2-arachidonylglycerol

 
• 2-OG

  2-oleoylglycerol

 
• 2-DG

  2-decanoyl-rac-glycerol

 
• 2-LG

  2-linoleoylglycerol

 
• AEA

  arachidonoylethanolamide (also known as anandamide)

 
• ALT

  alanine transaminase

 
• AST

  aspartate transaminase

 
• BA

  bile acid

 
• DCA

  deoxycholic acid

 
• DHEA

  docosahexaenoyl ethanolamide

 
• ESI

  electrospray ionization

 
• FAE

  fatty acid ethanolamide

 
• FFA

  free fatty acid

 
• FXR

  farnesoid X receptor

 
• LC/MS

  liquid chromatography/mass spectrometry

 
• NAPE-PLD

  N-acyl phosphatidylethanolamine phospholipase D

 
• OEA

  oleoylethanolamide

 
• SPE

  solid-phase extraction

 
• TGR5

  Takeda G protein-coupled receptor 5