Cytoplasmic tyrosine phosphatase Shp2 coordinates hepatic regulation of bile acid and FGF15/19 signaling to repress bile acid synthesis

Abstract

Bile acid (BA) biosynthesis is tightly controlled by intrahepatic negative feedback signaling elicited by BA binding to farnesoid X receptor (FXR) and also by enterohepatic communication involving ileal BA reabsorption and FGF15/19 secretion. However, how these pathways are coordinated is poorly understood. We show here that nonreceptor tyrosine phosphatase Shp2 is a critical player that couples and regulates the intrahepatic and enterohepatic signals for repression of BA synthesis. Ablating Shp2 in hepatocytes suppressed signal relay from FGFR4, receptor for FGF15/19, and attenuated BA activation of FXR signaling, resulting in elevation of systemic BA levels and chronic hepatobiliary disorders in mice. Acting immediately downstream of FGFR4, Shp2 associates with FRS2α and promotes the receptor activation and signal relay to several pathways. These results elucidate a molecular mechanism for the control of BA homeostasis by Shp2 through the orchestration of multiple signals in hepatocytes.

Figure 1. Hepatobilliary defects in Shp2^hep−/− mice

(A) Macroscopic view of the whole livers from two-month-old WT (Alb-cre⁻:Shp2^fl/fl) and Shp2^hep−/− (Alb-cre⁺:Shp2^fl/fl) mice. (B) Gallbladder volumes were adjusted by liver weight from WT and Shp2^hep−/− mice (n = 6–7). Data is shown as mean ± s.e.m. *** p < 0.001. (C–F) Liver sections were stained with H&E (C), Manson’s Trichrome (D), and Reticulin (E) and CK-19 (F). Scale bars in (D, E and F) were same as in (C).

Figure 2. Severe hepatobiliary damages in Shp2^hep−/− mice following bile duct ligation

(A) Kaplan-Meier survival analysis of WT and Shp2^hep−/− mice after BDL. ** p = 0.0014, as determined by Log-rank (Mantel-Cox) Test. (B) Macroscopic views of WT and Shp2^hep−/− livers were taken 24 and 48 hrs after BDL. (C) Gallbladder volumes were adjusted to liver weight after BDL (n = 4–10). (D) Macroscopic view of palms was shown 24 and 48 hrs after BDL. Serum bilirubin levels were measured (n = 6–12). (E) Serum BA levels were measured after BDL (n = 6–12) (F) Liver sections were stained with H&E (left), and statistical analysis (n = 5–7) of necrotic areas (right). Data in (E, D, E and F) are shown as the mean ± s.e.m. ** p < 0.01 and *** p < 0.001, as determined by Student’s t test.

Figure 3. Elevation of systemic BA levels in Shp2^hep−/− mice

(A) BA pool (liver, gallbladder and intestine) sizes were measured in both genders of the two genotypes (n = 6–10). (B–E) BA levels in liver (B), gallbladder (C), feces (D), serum (E) were measured (n = 5–11). All data were collected in males, hepatic BA concentration was adjusted to every gram liver weight, and fecal BA excretion was adjusted to 100 g body weight/day. (F) Bile flow rate was adjusted to 100 g body weight/min (n = 3). (G–I) BA composition in BA pool (n=6–9), liver (n=6–7) and feces (n=6–9) was analyzed by liquid chromatography/mass spectrometry. The fold changes of BA species in Shp2^hep−/− mice were calibrated to WT (the average value was designated as 1, left panels in G–I). The percentile representations of each conjugated and unconjugated BA species are shown in two panels separately in the right. Data in (A–I) are shown as the mean ± s.e.m. * p<0.05, ** p < 0.01 and *** p < 0.001, as determined by Student’s t test.

Figure 4. Lowering BA levels in Shp2^hep−/− mice alleviates hepatobilliary defects

(A) Macroscopic views of WT and Shp2^hep−/− livers fed with chow without or with 2% cholestyramine from age of 3 weeks to 2 months. (B) Liver sections were stained with Manson’s Trichrome. (C) Liver sections were stained with reticulin. (D) The ratios of liver/body weight were determined for each group (n = 5–8). (E) Collagen areas (blue) were measured from images in (B) (n = 4–7).

Figure 5. BA synthesis-related genes are significantly up-regulated in Shp2^hep−/− liver

(A) The expression of genes as indicated was determined by qRT-PCR in 2-month-old WT or Shp2^hep−/− livers (n = 4–5). (B) Cyp7a1, Shp2 and β-actin protein levels were determined by immunoblotting of liver lysates from WT and Shp2^hep−/− mice. Each lane represents each mouse. (C) Cyp7a1, Shp2 and β-actin protein levels were determined by immunoblotting of liver lysates from Control and Shp2^(H+K)−/− mice. Each lane represents each mouse. (D) Relative expression of Shp2 and Cyp7a1 was measured by qRT-PCR in liver extracts of WT and Shp2^(H+K)−/− mice following poly-I:C injection (n = 5). (E) Hepatic expression of Cyp7a1, Cyp8b1,SHP and FXR mRNAs was determined by qRT-PCR in mice fed with chow without or with 2% cholestyramine from 3 weeks to 2 months. (n = 4–8). (F) Cyp7a1, Shp2 and β-actin protein levels were determined by immunoblotting of liver lysates as in (F). Each lane represents each mouse. (G) Cytoplasmic (C) and nuclear (N) fractions were prepared from freshly-isolated liver samples. FXR, HNF4α, LRH-1, Lamin B (nuclear marker), Shp2 and Hsp90 (cytoplasmic marker) protein levels were determined by immunoblot analysis. Each pair of (C) and (N) samples was prepared from the same mouse. (H) Chromatin Immunoprecipitation (ChIP) was performed with liver samples (n=3) using FXR antibody. qPCR was performed with FXR binding region on SHP promoter (SHP) and coding region (con). Data are shown as fold enrichment. (I–J) ChIP assay was performed with HNF4α or LRH-1 antibodies, and different DNA sequences in Cyp7a1 promoter and proximal regions (n=4). Data are shown as fold enrichment. (M–O) Cholesterol (chol) levels of serum (h), liver (i) and gallbladder (j) were measured. Hepatic cholesterol was adjusted to mg/liver weight (g). (P) Hepatic expression of HMGCR and ACAT2 mRNA was determined by qRT-PCR (n = 4–5). All PCR data was normalized against β-actin, and fold change was calibrated to WT group. Data (A, D, F, H, I, J and K) are shown as the means ± s.e.m. * p < 0.05, ** p < 0.01 and *** p < 0.001, as determined by Student’s t test.

Figure 6. Shp2^hep−/− mice are refractory to FGF15/19 repression of BA synthesis

(A) Relative expression of SHP and FGF15 mRNA was determined by qRT-PCR in ileum samples (n = 5–7). (B) Relative expression of Cyp7a1, Cyp8b1 and SHP mRNAs in liver samples was determined by qRT-PCR. The animals (n = 5–10) were injected with PBS or hFGF19 (1 mg/kg body weight) and fasted for 6 hrs before sample collection. (C) Cyp7a1 and Shp2 protein levels were determined by immunoblot analysis of liver lysates from mice as in (B). Each lane represents one mouse. (D) Immunoblotting of liver lysates was performed with antibodies against pFRS2α(Y196), pErk, Erk1, p-p90RSK, p-PKC(pan) (βII Ser660), pJNK, JNK, p-p38, p38, β-Klotho and GAPDH. Two-month-old WT or Shp2^hep−/− mice were fasted for 5.5 hours before IP injection of PBS or hFGF19 (1 mg/kg body weight). The animals were sacrificed 30 min after injection. (E) Relative expression of SHP, Cyp7a1 and Cyp8b1 mRNA was determined by qRT-PCR in liver samples. The mice (n = 4–5) were injected with 2×10⁹ virions of VP16, VP16-FXR or SHP adenoviruses through tail vein and liver samples were collected 5 days later. (F) Cyp7a1, V5, Shp2 and GAPDH protein levels were determined by immunoblotting of liver samples collected as in (E). Each lane represents one mouse. Relative gene expression was normalized to β-actin, and fold change was calibrated to WT group. Data are shown as the means ± s.e.m. * p < 0.05, ** p < 0.01 and *** p < 0.001, as determined by Student’s t test.

Figure 7. Shp2 is required for FGF15/19-stimulated FGFR4 activation

(A–F) Serum-starved Hep3B cells were stimulated with 100 ng/ml hFGF19 as indicated. Immunoblotting was performed with indicated antibodies for total cell lysate (TCL) or immunoprecipitates. In E–F, the cells were treated with lenti-viruses expressing either scrambled (sh-scr) or Shp2-specific (sh-Shp2) shRNAs for 72 hrs before starvation. (G) The same samples as in Figure 6D were blotted with antibodies to FGFR4 and β-actin. (H) The same samples as in Figure 6C were immunoblotted for FGFR4 and Shp2. (I) A model shows how Shp2 orchestrates BA and FGF15/19 signaling in control of BA biosynthesis.