Altered systemic bile acid homeostasis contributes to liver disease in pediatric patients with intestinal failure

Intestinal failure (IF)-associated liver disease (IFALD), as a major complication, contributes to significant morbidity in pediatric IF patients. However, the pathogenesis of IFALD is still uncertain. We here investigate the roles of bile acid (BA) dysmetabolism in the unclear pathogenesis of IFALD. It found that the histological evidence of pediatric IF patients exhibited liver injury, which was characterized by liver bile duct proliferation, inflammatory infiltration, hepatocyte apoptosis and different stages of fibrosis. The BA compositions were altered in serum and liver of pediatric IF patients, as reflected by a primary BA dominant composition. In IF patients, the serum FGF19 levels decreased significantly, and were conversely correlated with ileal inflammation grades (r = −0.50, p < 0.05). In ileum, the inflammation grades were inversely associated with farnesoid X receptor (FXR) expression (r = −0.55, p < 0.05). In liver, the expression of induction of the rate-limiting enzyme in bile salt synthesis, cytochrome P450 7a1 (CYP7A1) increased evidently. In conclusion, ileum inflammation decreases FXR expression corresponding to reduce serum FGF19 concentration, along with increased hepatic bile acid synthesis, leading to liver damages in IF patients.

The IF patients exhibit histological liver injury. As seen in Fig. 1A, the control liver tissues exhibited normal liver histology. In contrast, the liver sections from pediatric IF patients exhibited liver damages characterized by bile duct proliferation, lymphocytes infiltration, and hepatocyte ballooning (Fig. 1A). Apoptotic hepatocytes were detected by TUNEL staining of liver sections. As expected, few TUNEL-positive cells were observed in the control liver specimens. Conversely, TUNEL-positive hepatocytes increased significantly in liver sections from patients ( Fig. 1A and B). When compared with the controls, extensive portal fibrosis was indicated in liver slides of pediatric IF patients ( Fig. 1A and C).
As seen in Fig. 4A, the total BA contents in IF patients' liver were higher (442.49 ± 312.93 nmol/mg) compared to controls (199.15 ± 134.59 nmol/mg) (Fig. 4A). The BA composition of the liver tissues of patients was also altered when compared with controls. Both unconjugated and conjugated primary BA, including cholic acid (CA), glycocholic acid (GCA), taurocholic acid (TCA), chenodeoxycholic acid (CDCA) and glycodeoxycholic acid (GDCA), increased in livers of patients related to the controls ( Fig. 4B and C). As seen in Fig. 4D-F, contents of CA (r = − 0.70, p < 0.05), CDCA (r = − 0.73, p < 0.05) and GCA (r = − 0.71, p < 0.05) inversely correlated with serum FGF19 (Fig. 4D-F). FXR is an important BA receptor and essential to the BA homeostasis [14][15][16] . We here showed that the hepatic FXR expression was lower in IF patients compared to controls ( Fig. 4G-J). In liver, the classic bile acid synthesis is controlled by cholesterol 7a-hydroxylase (CYP7A1) 17 . We here found that the hepatic CYP7A1 protein expression increased significantly in patients when compared to the controls ( Fig. 4G-J).

Discussion
In this study, we firstly showed that pediatric IF patients exhibited liver injury that characterized by cholestasis, portal inflammation, as well as hepatic apoptosis and fibrosis. Secondly, It demonstrated that the altered FXR/ FGF19 signaling was contributed to the cholestasis and liver injury in pediatric IF patients. As alteration in BA composition can cause hepatotoxicity 18 , thus analysis of the BA composition is essential to assess the impact of altered BA composition on the development of IFALD. In this study, the IF patients exhibited increased levels of primary BA in blood and liver, including CA and CDCA, concurrent with decreased levels of secondary or tertiary BA, such as DCA. In human, the primary BA, including CA and CDCA, can converted into secondary BA, DCA, LCA, through gut microbial 7-dehydroxylation 19 . The primary BA increased in IF patients might be attributed to differences in gut microbiota composition that shifts the microbial modification of the BA. Similar changes in BA composition levels have been reported in various liver diseases, including cirrhosis 20 , alcoholic liver disease 21 , cholestasis 22 , in which it was observed increased CDCA levels. In addition, alterations in BA metabolism are likely to induce the liver damage through affecting the solubilization of phospholipids, cholesterol and other lipids 23 . It reported that pro-inflammatory cytokines, including IL-6 and TNF-α , had been shown to be important mediators of cholestatic liver injury 24,25 . We here found that serum IL-6 and TNF-α concentrations Conversely, serum levels of IL-6 and TNF-α were significantly higher in patients than in controls. Scale bar = 25 or 50 μ m. **p < 0.01; ***p < 0.001.
were higher in IF patients compared to controls, which suggest alterations of bile acids metabolism potentiate hepatotoxicity may partly through pro-inflammatory mechanisms.
During bile acids enterohepatic circulation, FGF19 mediates bile acids hemostasis through a negative feedback way 9,26 . In the enterocyte, bile acids reclaimed by terminal ileum can upregulate FGF19 gene expression via activation of FXR 27 . After releasing into circulation, FGF19 reaches the liver and inhibits hepatic bile acids synthesis through suppression of CYP7A1 gene that encodes the rate limiting enzyme in synthesis of bile acids 28 . In this study, serum FGF19 concentrations were markedly decreased in IF patients compared to healthy controls. Annika and colleagues recently showed that loss of ileum led to reduced FGF19 production in patients with intestinal failure 12 . However, the mechanisms underlying the FGF19 associated with liver injury is still not fully established. Although patients with ileum preserved in this study, the inflammatory infiltrating presented in ileum, resulted in intestinal mucosa injury and further reduced FXR expression. In addition, we also indicated serum FGF19 concentrations were inversely associated with pro-inflammatory cytokine IL-6. IL-6 has been previously reported to involved in cholestasis among infants and children with intrahepatic and extrahepatic cholestasis 29 . The intestinal injury in IF patients could cause terminal ileum epithelial cells damaged and reduced the terminal ileum FXR expression. In contrast, the intestinal injury may allow translocation of bacterial antigens in the circulation and cause increased expression of pro-inflammatory including IL-6, leading to disturbed hepatocyte bile acid homeostasis. These findings here support the hypothesis that ileal inflammation leads to impaired intestinal FGF19 expression and altered bile acid homeostasis may through inhibiting ileal FXR activation in patients with IF. Experimental study indicated that administration of FGF19 protected the liver from cholestasis by reducing hepatic synthesis and primary hydrophobic BA 30 . In our study, we showed that serum FGF19 levels inversely correlated with hepatic primary bile acids including CA, CDCA and GCA. In addition, the expression of CYP7A1 was markedly increased in liver tissues from patients compared controls, which reinforce the concept that FGF19 can protect liver from cholestatic liver injury via repression of CYP7A1 expression, which suggests that FGF19 may be an important mediator in the pathogenesis of IFALD. Indeed, we showed that FGF19 were conversely related to the markers of liver injury, such as alkaline phosphatase, (ALP). Interestingly, we here did not found that FGF19 was tightly correlated to the bilirubin or lipids in the serum.
As pediatric IF is a rare disease, a relatively limited sample has been one of several limitations in this study. Although these results of our study show association rather than causality, this study adds novel data on FXR/ FGF19 pathway in the pathogenesis of IF in pediatric patients. Further studies conducted in more patients with biopsy-proven steatosis, portal inflammation, and histological cholestasis or with different fibrosis stages are warranted to analysis relationship between them and FGF19. Collectively, these results bring new insight to possible approaches for prevention and treatment of IF.

Materials and Methods
Patients. This study was approved by the Faculty of Medicine's Ethics Committee of Xin Hua hospital (XHEC-C-2016-063), School of Medicine, Shanghai Jiao Tong University, Shanghai, China. A total of 23 serum samples and 7 liver specimens were obtained from pediatric IF patients who underwent surgery. A total of 21 serum specimens were obtained from healthy controls with matched age. 6 normal adjacent non-tumour tissues, taken from hepatoblastoma patients, were used as liver controls. All patients' guardians provided written informed consent. The patients' characteristics are presented in Table 1. All methods in this study were carried out in accordance with the relevant guidelines.  Histological analyses and fibrosis determination. Histological examination was stained with hematoxylin and eosin (H&E). Fibrosis was determined by mason's trichrome stain according to the method described in a previous study 31 . Masson's trichrome staining was performed according to the manufacturer's protocol (Genmed Scientifics, Wilmington, DE). The collagen fiber was stained blue, the nuclei were stained black, and the background was stained red. Liver tissues were analyzed for apoptosis using the TUNEL test (TdT-mediated dUTP nick end labeling) was performed using the "In Situ Cell Death Detection Kit" from Roche Diagnostics according to the manufacturers' instructions. Sections were analyzed with a fluorescence microscope. Liver biopsies were analyzed by two researchers and a pathologist, blinded to clinical data, until consensus was reached.

Western-blot and Immunohistochemistry (IHC). Western-blot and Immunohistochemistry (IHC)
assays were performed as previously described 32 . For western-blot, the protein was extracted from liver tissues of IF patients using RIPA buffer with protease inhibitor cocktail (Pierce). The soluble fraction of the cell lysates was isolated by centrifugation at 12, 000 g for 10 minutes in a microfuge. BCA reagent (Pierce, Rockford, IL, USA) was applied to determine protein concentration. The equal amounts of proteins (150 ug/well) were separated by 4-12% SDS-PAGE, and transferred to nitrocellulose membranes. The membranes were incubated overnight at 4 °C with primary antibodies. Antibodies for GAPDH (Cell signaling technology, Danvers, MA, USA), CYP7A1 (Millipore, Darmstadt, Germany) and FXR (Invitrogen, Camarillo, CA) were used here. After incubation, the membranes were washed with PBS (containing 0.1% Tween) and incubated with horseradish-peroxidase conjugated detected the antigen-antibody complexes using an ECL Plus chemiluminescence reagent kit (Pierce, Rockford, IL, USA). For immunohistochemistry (IHC) analysis, paraffin-embedded tissues were incubated with xylol and descending concentrations of ethanol. Antigen retrieval was performed using citrate buffer, pH 6.0 or PH 8.0. Endogenous peroxidases were removed by incubation with 0.3% H 2 O 2 for 15 minutes at room temperature (RT) and blocking was performed using 10% bovine serum albumin (BSA) for 1 hour at RT. Primary   supernatant was transferred to a clean tube. The supernatant was dried under vacuum and reconstituted with 40 μ L of acetonitrile (with 0.1% formic acid) and 40 μ L of water (with 0.1% formic acid). After centrifugation, 5 μL of supernatant was injected for measurement. Liver tissue was weighed and homogenized in 50 μ L of icecold 40% methanol, and then centrifuged at 20 000 g for 10 min. The supernatant was transferred to a clean tube. Then, 80 μL of ice-cold methanol/chloroform (3:1, v/v) was added to the remaining pellet and rehomogenized. After centrifugation, the two supernatants were combined, spiked with 10 μ L of IS (1200 nmol/L of CA-D4 and LCA-D4), and dried under vacuum. The extracts were reconstituted with 40 μ L of acetonitrile (with 0.1% formic acid) and 40 μ L of water (with 0.1% formic acid). After centrifugation, 5 μ L of supernatant was injected for measurement. The analysis was performed with a Waters ACQUITY ultra performance liquid chromatography (BEHC-18, 1.7 μ m 2.1 × 100 mm column) coupled with Waters Xevo TQ-S triple quadrupole mass spectrometry (Waters MS Technologies, Ltd.). Data acquisition and bile acids quantification were performed using the MassLynx 4.1 software (Waters, Ltd.).
Statistical analysis. Data statistics are presented as medians IQR or mean ± SD. For comparisons of different groups, statistical significance was determined based on the Student's t test. Correlations were tested with Spearman rank correlation. P values < 0.05 were considered statistically significant.