Enhanced Microbial Bile Acid Deconjugation and Impaired Ileal Uptake in Pregnancy Repress Intestinal Regulation of Bile Acid Synthesis

Pregnancy is associated with progressive hypercholanemia, hypercholesterolemia, and hypertriglyceridemia, which can result in metabolic disease in susceptible women. Gut signals modify hepatic homeostatic pathways, linking intestinal content to metabolic activity. We sought to identify whether enteric endocrine signals contribute to raised serum bile acids observed in human and murine pregnancies, by measuring fibroblast growth factor (FGF) 19/15 protein and mRNA levels, and 7α‐hydroxy‐4‐cholesten‐3‐one. Terminal ileal farnesoid X receptor (FXR)‐mediated gene expression and apical sodium bile acid transporter (ASBT) protein concentration were measured by qPCR and western blotting. Shotgun whole‐genome sequencing and ultra‐performance liquid chromatography tandem mass spectrometry were used to determine the cecal microbiome and metabonome. Targeted and untargeted pathway analyses were performed to predict the systemic effects of the altered metagenome and metabolite profiles. Dietary CA supplementation was used to determine whether the observed alterations could be overcome by intestinal bile acids functioning as FXR agonists. Human and murine pregnancy were associated with reduced intestinal FXR signaling, with lower FGF19/15 and resultant increased hepatic bile acid synthesis. Terminal ileal ASBT protein was reduced in murine pregnancy. Cecal bile acid conjugation was reduced in pregnancy because of elevated bile salt hydrolase‐producing Bacteroidetes. CA supplementation induced intestinal FXR signaling, which was not abrogated by pregnancy, with strikingly similar changes to the microbiota and metabonome as identified in pregnancy. Conclusion: The altered intestinal microbiota of pregnancy enhance bile acid deconjugation, reducing ileal bile acid uptake and lowering FXR induction in enterocytes. This exacerbates the effects mediated by reduced bile acid uptake transporters in pregnancy. Thus, in pregnant women and mice, there is reduced FGF19/15‐mediated hepatic repression of hepatic bile acid synthesis, resulting in hypercholanemia.

A bespoke Python script (available on request) was employed for re-formatting the output and adapting it to different statistical analysis software. White's non-parametric t-test was performed using Statistical Analysis of Metagenomic Profiles (STAMP) (10) and all p-values were corrected using Benjamini-Hochberg method. PCoA was calculated using weighted Unifrac distance method. Taxonomic tree was created using an internal development for visualization of taxonomic differences between experimental groups. ShortBRED (https://huttenhower.sph.harvard.edu/shortbred, Kaminski et al., in progress,) software was used for the functional targeted analyses of arylsulfatase and 7-α-dehydroxylase genes using a specific selection of sequences in Uniprot database and for the BSH analysis using an internal database of BSH sequences.
The microbial pathway ranking was created using the cross-referenced PFAM -EC KEGG output of Interproscan software and the following ranking criteria: Large pathways (more than 20 enzymes involved) were considered to be changed i.e. depleted or over-represented if a minimum of 10% of the involved enzymes were changed, in either direction viz increased or reduced in abundance. Small pathways (less than 20 enzymes involved) were considered to be changed if a minimum of 2 enzymes were changed. A list of KEGG pathways considered changed using these criteria is presented in Supplementary Table 5.
The integration of host and microbial disturbed pathways was performed using QIAGEN's Ingenuity Pathway Analysis (IPA®, QIAGEN Redwood City, www.qiagen.com/ingenuity).
The organic extracts of cecum and cecal content samples were reconstituted in a mixture of isopropanol/ acetonitrile/ water (2:1:1, 250μL), vortexed for 30s, sonicated for 5min and vortexed for 30s, followed by centrifugation at 20,000xg for 30min at 4 °C. Supernatant was transferred into glass inserts in the LC-MS vials. A Quality Control (QC) sample was prepared with 50μL of each sample, to assess analytical reproducibility. The column was conditioned by injecting the QC pooled sample, several times, until data showed adequate stability. The QC sample was then injected every 6 samples to monitor the instrument's performance. The analytical run was completed by the analysis of extraction and solvent blank samples.
Lipid profiling was performed on an Acquity UPLC system (Waters Corp, USA) coupled to a XEVO G2 QTof Mass Spectrometry system (Waters MS Technologies, UK). Chromatography was performed using an Acquity UPLC CSH C18 2.1x100mm, 1.7um and column (Waters Corporation, USA) was held at 55 o C. Separation was achieved using gradient elution with 0.1% (v/v) formic acid in acetonitrile/water (60:40) (A) and 0.1% (v/v) in isopropanol/acetonitrile (B) (90:10) at a flow rate of 0.4 mL/min. In both mobile phases ammonium formate (LC-MS grade, Fluka, USA) was diluted to 10mM. Starting conditions were 60%A and 40%B for 2min, changing linearly to 43%B over 2min, to 50%B within 0.1min and to 54%B over 10min, when it was changed to 70%B within 0.1min and to 99%B over 6min. The solvent composition then returned to starting conditions over 0.1min, followed by re-equilibration for 2min prior to the next injection. Mass spectrometry was performed using electrospray in both positive and negative ESI ionization modes. The capillary voltage was 1.5kV, sampling and extraction cone voltages were 20V and 4V respectively, desolvation temperature was 600 °C, and source temperature was 120 °C. The cone gas flow rate was 50L/h, and desolvation gas flow rate was 1000L/h. The MS was operated in sensitivity mode with a scan time of 0.2s. For mass accuracy, a LockSpray interface was used with a 5ng/L leucine enkephalin (555.2645 amu) solution (50/50 ACN/H2O with 0.1% v/v formic acid) at 15μL/min was used as the lock mass. Data were collected in centroid mode with a scan range of 50−2000 m/z, with lockmass scans collected every 30s and averaged over 4 scans to perform mass correction. Injection volumes of 4μL and 15μL were used for positive and negative ionization modes respectively. The auto-sampler was set at 4°C.
The aqueous extracts of cecum, cecal content and liver samples were reconstituted in a mixture of acetonitrile/water (1:1, 170μL), vortexed for 30s, sonicated for 5min and vortexed for 30s, followed by centrifugation at 20,000xg for 30min at 4 °C. The supernatant (100μL) from each sample was transferred into glass inserts in the LC-MS vials. A QC sample was prepared by aliquoting 50μL of each sample and the same QC strategy was used as above. The analytical run was completed by the analysis of extraction and solvent blank samples.
HILIC-UPLC-MS/MS analysis was performed using an identical Acquity UPLC system (Waters Corp, USA) coupled to a XEVO G2 QTof Mass Spectrometry system (Waters MS Technologies, UK) as the one used for lipid profiling of organic extracts. Column temperature was set at 40°C. Mobile phase A consisted of ACN/water (95:5) and mobile phase B ACN/water (50:50). In both solutions ammonium acetate was diluted to 10mM and formic acid to 0.1%. Separation was achieved using gradient elution: starting conditions were 99%A for 2min with flow rate 0.4ml/min, changing linearly to 55% B over the next 8min, and then to 99%B within 1 min, at which it was kept for 2min. Subsequently, the solvent composition returned to starting conditions over 0.1min, followed by re-equilibration for 9 min with increasing flow rate (up to 0.9ml/min) prior to the next injection. Mass spectrometry was performed using electrospray in both positive and negative ESI ionization modes. The capillary voltage was 1.5kV, sampling and extraction cone voltages were 30V and 4V respectively, desolvation temperature was 600 °C, and source temperature was 120 °C. The cone gas flow rate was 50L/h, and desolvation gas flow rate was 1000L/h. The MS was operated in sensitivity mode with a scan time of 0.2s. For mass accuracy, a LockSpray interface was used with a 5ng/L leucine enkephalin (555.2645 amu) solution (50/50 ACN/H2O with 0.1% v/v formic acid) at 15μL/min was used as the lock mass. Data were collected in centroid mode with a scan range of 50−1200 m/z, with lockmass scans collected every 30s and averaged over 4 scans to perform mass correction. Injection volume of 2μL was used for both positive and negative ionization modes. The auto-sampler was set at 4°C.
The remaining second dried aliquots of aqueous and organic extracts of cecum and cecal content samples were combined prior to analysis. The aqueous extracts were reconstituted in a mixture of propanol/ water (1:1, 150μL), vortexed for 30s, sonicated for 5min and vortexed for 30s. The supernatant was transferred into the dried organic extracts, vortexed for 30s, sonicated for 5min and vortexed for 30s, followed by centrifugation at 20,000xg for 30 min at 4 °C. 100μL supernatant from each sample was transferred into glass inserts in the LC-MS vials. A QC sample was prepared by aliquoting 40μL of each sample and the same QC strategy was used as above. The analytical run was completed by the analysis of extraction and solvent blank samples. Bile acid UPLC-MS/MS analysis was performed using an identical Acquity UPLC system (Waters Corp, USA) coupled to a XEVO G2 QTof Mass Spectrometry system (Waters MS Technologies, UK) as the one used above. An ACQUITY BEH C8 column (1.7μm, 100mm × 2.1mm) was used at an operating temperature of 60 °C. The mobile phase solvent A consisted of a volumetric preparation of 100mL of acetonitrile added to 1L of ultrapure water, with a final additive concentration of 1mM ammonium acetate and pH adjusted to 4.15 with acetic acid. Mobile phase solvent B consisted of a volumetric preparation of acetonitrile and 2-propanol in a 1:1 mixture. Separation was achieved using gradient elution: starting conditions were 90%A changing linearly to 35%B over the next 9.25min at 0.6ml/min, to 85%B within 2.25min at 0.6 ml/min, and then to 100%B within 0.3min at 0.8ml/min, at which it was kept for <1 min. Afterwards the solvent composition returned to starting conditions, followed by re-equilibration for 2.5min with increasing flow rate prior to the next injection.
Mass spectrometry was performed using electrospray in negative ESI ionization modes. The capillary voltage was 1.5kV, sampling and extraction cone voltages were 60V and 4V respectively, desolvation temperature was 600°C, and source temperature was 120°C. The cone gas flow rate was 150L/h, and desolvation gas flow rate was 1000L/h. The MS was operated in sensitivity mode with a scan time of 0.1s. For mass accuracy, a LockSpray interface was used with a 5ng/L leucine enkephalin (555.2645 amu) solution (50/50 ACN/H2O with 0.1% v/v formic acid) at 15μL/min was used as the lock mass. Data were collected in centroid mode with a scan range of 50−1200m/z, with lockmass scans collected every 30s and averaged over 4 scans to perform mass correction. The injection volume was 5μL and the auto-sampler was set at 4°C.

Tissue mRNA expression
Total RNA from duodenum, distal ileum and livers of mice was extracted using Qiazol lysis reagent or RLT buffer (both Qiagen, UK), and total RNA from human terminal ileal explants and gavaged murine terminal ileum was extracted using the RNeasy Mini Kit (Qiagen, UK) following bead beating with Qiagen Tissuelyser II, as per manufacturer's instructions. Reverse transcription was performed with the High-Capacity cDNA Reverse Transcription Kit, Thermo Fisher Scientific, UK. Real-time quantitative PCR was performed on Viia7 system (Thermo Fisher Scientific, UK), in a 384-well assay format using SYBR Green Mastermix (Sigma-Aldrich, UK). Primer sequences used are listed in Supplementary Table 1.

Tissue protein expression
Distal ileal ASBT protein levels were measured using western blotting. Protein was extracted by bead beating in phosphate-buffered saline and Halt Protease and Phosphatase Inhibitor Cocktail (Thermo Fisher Scientific, MA, USA), and quantified using Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, MA, USA). 15µg protein per sample was separated by gel electrophoresis on NuPAGE 4-12% bis-tris protein gels (Thermo Fisher Scientific, MA, USA). Following transfer to membranes, protein was incubated with SLC10A2 antibody ab203205 (Abcam, Cambridge UK) and βactin

Supplementary Tables
Supplementary Table 1. Primer sequences used for qRT-PCR.