Hepatic uptake of conjugated bile acids is mediated by both sodium taurocholate cotransporting polypeptide and organic anion transporting polypeptides and modulated by intestinal sensing of plasma bile acid levels in mice

Abstract

The Na⁺ -taurocholate cotransporting polypeptide (NTCP/SLC10A1) is believed to be pivotal for hepatic uptake of conjugated bile acids. However, plasma bile acid levels are normal in a subset of NTCP knockout mice and in mice treated with myrcludex B, a specific NTCP inhibitor. Here, we elucidated which transport proteins mediate the hepatic uptake of conjugated bile acids and demonstrated intestinal sensing of elevated bile acid levels in plasma in mice. Mice or healthy volunteers were treated with myrcludex B. Hepatic bile acid uptake kinetics were determined in wild-type (WT), organic anion transporting polypeptide (OATP) knockout mice (lacking Slco1a/1b isoforms), and human OATP1B1-transgenic mice. Effects of fibroblast growth factor 19 (FGF19) on hepatic transporter mRNA levels were assessed in rat hepatoma cells and in mice by peptide injection or adeno-associated virus-mediated overexpression. NTCP inhibition using myrcludex B had only moderate effects on bile acid kinetics in WT mice, but completely inhibited active transport of conjugated bile acid species in OATP knockout mice. Cholesterol 7α-hydroxylase Cyp7a1 expression was strongly down-regulated upon prolonged inhibition of hepatic uptake of conjugated bile acids. Fgf15 (mouse counterpart of FGF19) expression was induced in hypercholanemic OATP and NTCP knockout mice, as well as in myrcludex B-treated cholestatic mice, whereas plasma FGF19 was not induced in humans treated with myrcludex B. Fgf15/FGF19 expression was induced in polarized human enterocyte-models and mouse organoids by basolateral incubation with a high concentration (1 mM) of conjugated bile acids.

Conclusion: NTCP and OATPs contribute to hepatic uptake of conjugated bile acids in mice, whereas the predominant uptake in humans is NTCP mediated. Enterocytes sense highly elevated levels of (conjugated) bile acids in the systemic circulation to induce FGF15/19, which modulates hepatic bile acid synthesis and uptake. (Hepatology 2017;66:1631-1643).

Figure 1

TCA kinetics in WT and Slco1a/1b (OATP1A/1B) knockout mice. (A,B) Immunofluorescent visualization of mouse Oatp1a1 (A) or –Oatp1a4 (B) in male WT and Slc10a1 knockout (NTCP‐KO) mice with high plasma bile acid levels. Pericentral hepatocytes were visualized by costaining of glutamine synthetase (green, as insert). Scale bar is 40 μm. (C) Plasma TCA clearance in WT and OATP1A/1B knockout mice after myrcludex B or vehicle administration. TCA with tracer amounts of [³H]TCA was injected into the tail vein of mice and tritium activity (dpm) was measured in plasma (C) or bile (D) at indicated time points. (E) Hepatic bile flow pre‐TCA and post‐TCA injection in WT and OATP1A/1B knockout mice injected with myrcludex B or vehicle. (C‐E) N = 5/6 mice per group. Asterisk indicates significant difference between vehicle and myrcludex B in WT mice, and hash indicates significant difference between vehicle and myrcludex B in OATP1A/1B knockout mice (P < 0.05; Mann‐Whitney U test).

Figure 2

Myrcludex B increases bile acid levels in plasma and urine, but not in feces, of (OATP1B1‐humanized) Slco1a/1b knockout mice. Myrcludex B or vehicle was injected for 5 days. Plasma composition of (A) conjugated and (B) unconjugated bile acid species as quantified by HPLC. Concentrations (μM) are given on a ¹⁰log scale for 6 mice/group. (C) Urinary bile acid concentrations corrected for creatinine concentrations in each mouse (μmol/mmol). (D) Total fecal bile acid excretion (μmol/24 hours/100 g BW). White and light‐gray bars indicate OATP‐KO mice, injected with vehicle or myrcludex B, respectively. Dark‐gray or black bars indicate OATP1B1‐humanized OATP‐KO mice, injected with vehicle or mycludex B, respectively. Asterisk indicates significant change for OATP‐KO mice, and hash indicates significant change for OATP1B1‐humanized OATP‐KO mice. Dagger indicates significant change between genotypes upon myrcludex B treatment (P < 0.05; Mann‐Whitney U test). Abbreviations: BA, bile acid; CDCA, chenodeoxycholic acid; DCA, deoxycholic acid; KO, knockout; MCA, muricholic acid; Myr B, myrcludex B; TDCA, taurodeoxycholic acid; TUDCA, tauroursodeoxycholic acid.

Figure 3

OATP isoforms mediate uptake of conjugated bile acids in vitro. (A) [³H]TCA uptake in U2OS transiently overexpressing mouse Ntcp, Oatp1b2, Oatp1a1, and rat Oatp1a4. N = 6 wells/group from two independent experiments. NTCP‐ and OATP‐mediated transport was compared to parental cells (mock), and asterisk indicates significant difference (P < 0.05). (B) FRET‐based sensing of NTCP‐ and OATP‐mediated TCDCA influx. The nuclear FRET sensor for bile acids (NucleoBAS) was cotransfected with empty plasmid (mock) or each individual bile acid uptake transporter in U2OS cells. The emission ratio of citrin (525 nm)/cerulean (475 nm) is plotted in time. Compound addition is indicated by arrows. To fully activate NucleoBAS, each experiment ended with addition of 5 μM of GW4064. N = 6 cells/group, from two independent experiments. (C) Evident plasma membrane expression of mouse OATP1B2, OATP1A1, and rat OATP1A4 was demonstrated by cell surface biotinylation. (+) indicates with biotin. (–) indicates without biotin.

Figure 4

Cyp7a1 expression after acute or prolonged myrcludex B treatment. (A) Hepatic mRNA expression of the rate‐limiting bile acid synthesis enzyme, Cyp7a1, shortly (1.5 hours) after vehicle or myrcludex B treatment in WT or Slco1a/1b knockout (KO) mice. Values are normalized to vehicle‐treated WT mice. Asterisk indicates significant increase (P < 0.05) between vehicle and myrcludex B injection in Slco1a/1b knockout mice. (B) Hepatic Cyp7a1 mRNA expression after 5 daily injections with vehicle or myrcludex B in (OATP1B1‐humanized) Slco1a/1b knockout mice. Asterisk or hash symbols indicate significant decrease (P < 0.05; 5‐6 fasted male mice/group) compared to vehicle controls. Data are calculated using the geometric mean of reference genes Rplp0 and Tbp and normalized to vehicle‐treated Slco1a/1b knockout mice. (C) C4 levels, a marker for bile acid synthesis, were reduced in plasma of Slc10a1 knockout compared to WT mice. Data from 5‐6 fasted female mice/group. Abbreviations: KO, knockout; Myr B, myrcludex B.

Figure 5

Millimolar concentrations of conjugated bile acids, present in blood after myrcludex B treatment, induce intestinal FXR‐FGF15/19 signaling. (A) Relative mRNA expression of FXR target genes Shp and Fgf15 in the distal ileum in (OATP1B1‐humanized) Slco1a/1b knockout mice after 5‐day injections with vehicle or myrcludex B. (B) Relative mRNA expression of Shp and Fgf15 in the distal ileum for WT and Slc10a1 knockout mice with high plasma bile acid levels. Data are calculated using the geometric mean of reference genes Hprt and Ppib and normalized to vehicle‐treated Slco1a/1b KO mice or Slc10a1 KO mice. Asterisk and hash symbols indicate significant changes (P < 0.05; 6 fasted male mice/group). (C) Relative Fgf15 mRNA expression in the distal ileum after 5‐day BDL (5‐6 fasted male mice/group), normalized to vehicle‐treated WT mice. (D) Effects of myrcludex B on human plasma FGF19 levels. Blood samples were drawn at 9:45 am (fasting), immediately before myrcludex B administration. After 3 hours, a consecutive blood sample was taken. FGF19 levels were measured by enzyme‐linked immunosorbent assay in 12 individuals. (E) Mouse and human conjugated bile acids (200 μM and 1 mM) were added to the basolateral compartment of filter‐grown T84 cells. mRNA expression of FXR target genes FGF19, OSTα, and IBABP was determined after 24‐hour incubation. GW4064 (5 μM) was used as a positive control. Data are calculated using the geometric mean of reference genes Ppib and Hprt and normalized to the vehicle‐treated condition. Asterisk indicates significant change (P < 0.05; 5‐6 replicates/group). Abbreviations: DMSO, dimethyl sulfoxide; KO, knockout; Myr B, myrcludex B.

Figure 6

FGF19 modulates the hepatic bile acid uptake machinery in rat hepatocytes (HPCT‐1e3) and mice upon peptide injection or AAV‐mediated overexpression. (A) Effects of TCA (100 μM) or hFGF19 (40 ng/mL) on Ntcp, Oatp1b2, Oatp1a1, and Oatp1a4 expression in rat hepatoma cells (HPCT‐1e3) cells was determined by qRT‐PCR using the geometric mean of Rplp0 and Tbp as a reference. Data are normalized to vehicle (6 replicates/group). Asterisk indicates significant change (P < 0.05). (B) Relative hepatic mRNA expression of Ntcp and Oatp isoforms was determined 6 hours (acutely) after a single injection of hFGF19 (1 mg/kg intraperitoneally) or vehicle in male WT mice, using the housekeeping gene Rplp0 as a reference. Data are normalized to vehicle‐treated WT mice. Asterisk symbols indicate significant (P < 0.05; 5 fasted male mice/group). (C) Relative hepatic mRNA expression of Ntcp and Oatp isoforms after 24 weeks of AAV‐induced (3 × 10¹¹ v.g.) FGF19 overexpression in db/db male mice, using the housekeeping gene, Rplp0, as a reference. Data are normalized to AAV‐GFP treatment. Asterisk indicates significant change (P < 0.05; 8‐10 fasted male mice/group).