- Email: [email protected]
The gut-liver axis: impact of a mouse model of small-bowel bacterial overgrowth
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
The gut-liver axis: impact of a mouse model of small-bowel bacterial overgrowth Qingqing Wang, MD,a,1 Bin Wang, MD,a,1 Vijay Saxena, PhD,a Lili Miles, MD,b Josh Tiao,a Joel E. Mortensen, PhD,b and Jaimie D. Nathan, MDa,* a b
Division of Pediatric General and Thoracic Surgery, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Division of Pathology and Laboratory Medicine, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio
article info
abstract
Article history:
Background: The mechanisms by which intestinal bacteria impact liver diseases remain
Received 29 March 2017
poorly understood. The aim of this study was to develop a mouse model of small-bowel
Received in revised form
bacterial overgrowth and to determine its impact on hepatobiliary injury.
7 August 2017
Materials and methods: A jejunal self-filling blind loop (SFBL) was created in C57BL/6 mice.
Accepted 30 August 2017
Three weeks after surgery, the mice were euthanized, and bacterial cultures of luminal
Available online xxx
content of the loop and extraintestinal tissues were performed. Liver and jejunum were collected for histological grading of inflammation and injury. Serum liver biochemistry
Keywords:
assays were performed. Hepatobiliary transporter mRNA expression in liver was measured
Gut-derived factors
by quantitative real-time polymerase chain reaction. Bile and blood were collected for
Small-bowel bacterial overgrowth
measurement of total bile acids, phospholipid, and cholesterol. Mice undergoing jejunal
Bacterial translocation
transection and reanastomosis and laparotomy only served as control groups.
Self-filling blind loop
Results: SFBL induced a dramatic increase in intraluminal bacterial counts, mesenteric
Liver injury
lymph node bacterial translocation, and evidence of jejunal and hepatobiliary injury. Sig-
Bile composition
nificant reductions in hepatic expression of hepatobiliary transporters involved in biliary canalicular export and basolateral uptake were observed in SFBL mice. SFBL resulted in a significant increase in biliary total bile acid concentration, decreases in bile phospholipid and cholesterol output, and an increase in the bile acid/phospholipid ratio. Conclusions: We have developed a reproducible mouse model of small-bowel bacterial overgrowth with evidence of liver inflammation, altered hepatobiliary transporter expression, and alterations in bile composition. This model may help to elucidate the mechanisms by which gut-derived bacterial factors impact the liver and contribute to the exacerbation of liver diseases and biliary injury. ª 2017 Elsevier Inc. All rights reserved.
Introduction Emerging evidence suggests that intestinal microbiota or bacteria-derived factors play a critical role in the modulation
and exacerbation of a number of liver diseases,1-5 such as primary biliary cirrhosis (PBC), primary sclerosing cholangitis (PSC), nonalcoholic fatty liver disease (NAFLD), parenteral nutrition-associated liver disease (PNALD), alcoholic liver
* Corresponding author. Division of Pediatric General and Thoracic Surgery, Cincinnati Children’s Hospital Medical Center, MLC 2023, 3333 Burnet Avenue, Cincinnati, OH 45229-3039. Tel.: þ513 636-4371; fax: þ513 636-7657. E-mail address: [email protected] (J.D. Nathan). 1 These authors contributed equally to this work. 0022-4804/$ e see front matter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2017.08.049
wang et al liver impact of bacterial overgrowth
diseases, and cystic fibrosis (CF)eassociated liver disease.6-12 However, the pathogenic link between intestinal microbiota and chronic liver disease requires further elucidation. Furthermore, several studies support the notion that small-bowel bacterial overgrowth (SBBO) and bacterial translocation (BT) may play a crucial role in the progression of liver diseases.13-16 The normal human microflora maintains a delicate balance between its constituent parts, numbering 1011 bacteria per gram content of the small intestine with over 400 different species.17 SBBO is defined as an increased number and/or abnormal type of bacteria in the small bowel, and BT is defined as the passage of viable endogenous bacteria or their products from the intestinal tract through the epithelial barrier to the mesenteric lymph nodes (MLNs), systemic circulation, or extraintestinal organs.18,19 Although studies from both animal models8,20 and patients with liver disease13,21 have shown that SBBO and BT are critical in disease progression, the mechanisms by which they modulate intestinal barrier dysfunction and contribute to liver disease are still unclear. One mechanism by which microbial factors play a role in hepatic function is in the metabolism of bile acids because they participate in the generation of secondary bile acids, which are subsequently taken up by the ileum to enter the enterohepatic circulation.22 Through modulation of bile, microbiota can interact with the host and impact not only the liver diseases but also other organs and metabolic pathways.23 This represents one of the crucial manners by which intestinal microbiota communicate with the liver. One such microbe-derived factor, lipopolysaccharide (LPS), present in the outer membrane of gramnegative bacteria, has been demonstrated to be a key mediator in inflammation-induced cholestasis, a process resulting from either a defect in bile formation at the level of hepatocyte or from an impairment of bile secretion and flow at the level of bile ducts.24,25 LPS and/or LPS-induced proinflammatory cytokines, such as tumor necrosis factor-alpha and various interleukins, mediate the cholestatic effect via inhibition of the expression and function of the hepatobiliary transporter system.26 Despite the recognition of the importance of microbial factors, including SBBO and BT in liver disease, there remains a lack of relevant animal models for the investigation of disease pathogenesis. Approximately two decades ago, Lichtman et al. surgically created jejunal self-filling blind loops (SFBLs) in rat, and observed bacterial overgrowth in the jejunal loop and mesenteric translocation. Although histologic and biochemical changes were found in rats with SFBL, and antibiotic treatment prevented hepatic injury in SFBL rats, further light could not be shed on the pathogenesis of liver injury associated with SBBO/BT.27-31 Studying such interactions in a rat model of SBBO is hampering, and genetically-modified mice represent powerful tools to study such interactions. A recent study has described a surgical model of SFBL in mice, using a blind loop created in the distal ileum, to study the role of microbiota in the pathogenesis of ulcerative colitis.32 In the current study, we describe a mouse model of SBBO/BT using a jejunal blind loop, and we investigate its impact on hepatic injury and function. We hypothesize that SBBO/BT induced by the creation of a jejunal SFBL in mice would result in hepatic injury and alterations in hepatobiliary transporter expression and bile composition.
247
Materials and methods Mice Approximately, 5- to 6-week-old C57BL/6 male mice (weight range, 20-25 gm) were purchased from Charles River (Wilmington, MA). Mice were fed standard chow and had access to tap water ad libitum. The mice were maintained in a 12:12-hr day-night rhythm at a constant temperature of 23 C and a relative humidity of 40%-60%. The mice were given liquid diet (microstabilized rodent liquid diet LD101; TestDiet, St. Louis, MO) 24 h before surgery. After surgery, only water was provided ad libitum for the first 24 h. Thereafter, all mice were fed liquid diet until euthanasia. The mice’s body weights were measured daily. A protocol for this study was approved by the Cincinnati Children’s Hospital Medical Center Institutional Animal Care and Use Committee (protocol no. 9D12102).
Experimental design and surgical procedures The mice were anesthetized with isoflurane. To create SFBL mice, a 2-cm midline laparotomy incision was made, the jejunum was divided 8 cm distal to the ligament of Treitz, and a 3-cm jejunal SFBL was created by an end-to-side anastomosis of the distal jejunal stump to the side of the proximal jejunal limb 5 cm distal to the ligament of Treitz using interrupted 9-0 monofilament suture (Ethicon Endo-Surgery, Cincinnati, OH) (Fig. 1A). For jejunal transection and reanastomosis (JTR) control mice, the jejunum was partially transected at 50% of diameter 5 cm distal to the ligament of Treitz and resutured in interrupted fashion using 9-0 monofilament suture. Shamoperated (Sham) control mice underwent laparotomy without any intestinal transection. We use two cohorts of mice for the study. The first cohort of mice was used for most of the experiments including bacterial cultures, LPS measurement, biochemical study, liver and small-intestine histology, and hepatobiliary transporter expression. For this cohort, we have 11 mice in the Sham group, 13 mice in the JTR group, and 10 mice in the SFBL group, with the unbalanced totals secondary to early postoperative mortality. The second cohort of mice was used for measuring bile flow and composition. We have two groups of mice in this cohort, including eight in the Sham group and eight in the SFBL group.
Bacterial cultures of organs All mice were euthanized at week 3 after surgery. Mice were anesthetized by triple-mix (12 mL triple-mix solution contained 8.0 mL ketamine at 100 mg/mL, 0.4 mL xylazine at 100 mg/mL, 2 mL acepromazine at 10 mg/mL, and 1.6 mL normal saline), given intramuscularly. Using sterile techniques to expose the abdominal cavity, the skin was opened and then the muscle and peritoneal cavity were opened with a second set of sterile instruments. Peritoneal cultures were performed by swabbing the viscera with sterile cotton-tipped applicator sticks. Portal vein blood of 50 mL was diluted in 500-ml Baltimore Biological Laboratoriesenriched thioglycollate medium (BD Biosciences, Sparks,
248
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
Fig. 1 e Creation of SFBL in C57BL/6 mice. (A) A schematic diagram of SFBL surgical model showing a 3-cm jejunal blind loop, created 5 cm distal to the ligament of Treitz; (B) A representative mouse intraabdominal photograph showing the dilated SFBL (indicated by white arrow) at the time of euthanasia; (C) Changes in mouse body weight at indicated time points after surgery (Sham, n [ 11; JTR, n [ 13; SFBL, n [ 10). (Color version of figure is available online.)
MD). Each diluted blood sample of 100 mL was plated on trypticase soy agar with 5% sheep blood agar plates (BD Biosciences, Sparks, MD) for aerobic culture and on Brucella agar with 5% sheep blood agar plates (PML microbiologicals Inc, Wilsonville, OR) for anaerobic cultures. Liver (50-100 mg) and MLNs (5-20 mg) were collected and homogenized in a 500-mL culture medium. Each sample of 100 mL was cultured aerobically and anaerobically. The aerobic cultures were incubated at 37 C for 24 h, and anaerobic cultures were incubated in BD GasPak EZ Gas Generating System (BD Biosciences, Sparks, MD) for 40 h.
Bacterial cultures of luminal contents Sterile saline of 2 mL was flushed through the loop or corresponding 3-cm segment of jejunum of control mice. The fluid was collected and diluted serially in medium from 101 to 108 and cultured aerobically and anaerobically.
Biochemical assays Cardiac blood was collected for liver biochemical profile at the time of euthanasia and stored at 80 C until analysis. The
249
wang et al liver impact of bacterial overgrowth
liver biochemical profile was measured by the clinical laboratory at TriHealth laboratories, Cincinnati, OH.
Lipopolysaccharide measurement Portal vein serum was collected at the time of euthanasia for LPS measurement using Limulus Amebocyte Lyisate assay kit (Lonza, Walkersville, MD). Briefly, serum samples were preheated at 70 C for 30 min to deactivate the inhibitory factors for the reaction. Standard LPS from the kit was used to create a standard curve. Bacterial LPS from mouse small intestinal lumen was used as positive control. Pyrogen-free water was used as a negative control. Absorbance of all samples and standards was read at 405 nm. The LPS concentration (EU/ml) was calculated using the standard curve.
Real-time polymerase chain reaction for hepatobiliary transporter expression Total liver RNA was isolated by TRIzol and subjected to cDNA synthesis using oligo-dT primers dT12-18 (Invitrogen, Carlsbad, CA) and SuperScript II Reverse Transcriptase enzyme (Invitrogen, Carlsbad, CA). Using a specific primer set (Table 1), polymerase chain reaction was performed using SYBR green QPCR mix (Agilent technologies, Santa Clara, CA). All samples were loaded in duplicate wells, and data analysis was performed using DDCt method. b-Actin was used as internal control. Polymerase chain reaction conditions were as follows: 95 C for 10 min to activate the DNA enzyme, followed by 40 cycles of 95 C for 30 s (denaturation), 55 C-60 C for 1 min (annealing), and 72 C for 30 s (extension).
Histological scoring Liver and jejunum were collected, immediately placed into 10% neutral-buffered formalin for 24 h and transferred to 70% alcohol thereafter. Tissues were embedded in paraffin, and 5-mm thick sections were mounted and stained with hematoxylin and eosin. To quantify the differences between Sham, JTR, and SFBL liver and jejunum, histology scoring was performed on hematoxylin and eosinestained sections by a pathologist blinded to the experimental design and groups.
The liver scoring system was adopted from a scoring system developed by Kleiner et al.,6 and included portal inflammation, lobular inflammation, hepatocellular injury, and cholangitis. Each category was graded based on severity and extent of involvement. Jejunal sections were scored for cryptitis, crypt architecture, lamina propria inflammation, and bowel-wall inflammation.
Bile collection and composition measurement In additional cohorts of mice, bile was collected as previously described by Jahnel et al.,26 with modifications. Briefly, mice were anesthetized with 1.5% isoflurane with 1% oxygen after 4 h of fasting. The common bile duct was ligated just below the junction of the cystic duct and hepatic duct. Bile from the gallbladder was aspired and discarded. The gallbladder was then cannulated with 2-French silicone catheter tubing (inner diameter 0.5 mm, outer diameter 0.9 mm, HelixMark standard silicone tubing; Helix Medical, Carpinteria, CA) for bile collection. Bile was collected for 30 min in preweighed tubes and stored at 80 C until bile composition measurements. To examine the bile flow rate in mice, bile specimens were weighed at the end of collection, and the volume of bile was measured gravimetrically by assuming a specific gravity of 1.0. Bile flow rate was normalized to total liver weight. Biliary concentrations of total bile acids, phospholipids, and cholesterol were measured by commercially available kits (total bile acids: BIOQUANT Image Analysis Corporation, Nashville, TN; phospholipids: Wako Chemicals USA, Inc, Richmond, VA; cholesterol: Point Scientific Inc, Canton, MI), and biliary output (nmol/g/min) of each component was calculated. Bile acid/ phospholipid ratio was calculated by dividing the bile acid output by phospholipid output. Portal vein serum was collected after bile collection and stored at 80 C until measurements of serum total bile acids, phospholipids, and cholesterol were performed.
Statistical analysis Data are presented as mean standard error of the mean. Statistical differences between group mean values were assessed by unpaired t-test, analysis of variance, Fisher’s
Table 1 e Primer sequences for mouse hepatobiliary transporters. Target
Primer sense
Primer antisense
GGC TTC TTC AAT AAC TGT CG
GTT TCT ATT TCC CGC TCT CT
Abcg8
CTC ACA CAG GAC ACT GAC TG
GGA AGT CAT TGG AAA TCT GA
Bsep1
CAT TAA CAG CGA AGT CAT CA
CAG TGA TTA CCC ACA ACC TT
Mdr1a
AAG AAA CCA GCA GTC AGT GT
TGG ATA ATA GCA GCC AGA GT
Mdr2
GTT CAG CCT GTT TGA AAG AC
CAC GCT TCC TTC AAA CTT AT
Mrp2
ATG TGG CAT ATT CCA GTT TC
CCA TAG GAG ACG AAG AAC AG
Ntcp
CTC TGC TCT CTT CCG ACT AA
GGT GAC ATT GAG GAT GGT AG
Oatp1
CGC AGT CTT CAT TCT AAT CC
CCT TCT CTG TGA GCT TCA TC
Oatp2
CAT CTG TCT TTG GCT ACC TC
TGG TAG GAG ACA GTC AGT CC
Oatp4
CAC GTT GAC CTG TTT AGT CA
AAT CCA GGT GTA TGA GTT GG
Abcg5
250
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
exact test and ManneWhitney U-test, and P < 0.05 was considered statistically significant.
Table 2 e Extraintestinal organ culture of aerobic and anaerobic bacteria. Surgery (n)
Results
Sham (11) JTR (13)
Creation of SFBL in mice and its impact on general health of mice All groups of surviving mice were generally healthy at the time of euthanasia. At 3 weeks after surgery, the jejunal SFBL appeared considerably dilated in all SFBL mice (Fig. 1B), and MLN enlargement was also evident in all SFBL mice. Gross examination of visceral organs was otherwise normal in all groups of mice, and none of the mice had abscesses or other evidence of peritonitis at the time of necropsy. After surgery, mice in the SFBL group grew slowly, and weight gain plateaued at postoperative day 10, whereas control group mice continued to gain weight until euthanasia at postoperative day 21 (Fig. 1C).
Bacterial overgrowth and translocation in SFBL mice The hallmark of successful SFBL surgery is the overgrowth of bacteria in the blind loop. The creation of an SFBL induced dramatic increases in both aerobic and anaerobic intraluminal bacterial counts (8.48 0.11 log10 CFU/ml and 8.87 0.10 log10 CFU/ml, respectively), compared to JTR mice (3.0 0.34 log10 CFU/ml and 3.78 0.30 log10 CFU/ml, respectively) and Sham mice (2.63 0.16 log10 CFU/ml and 3.13 0.45 log10 CFU/ ml, respectively; P < 0.001) (Fig. 2). In 10 of 10 SFBL mice (100%), MLN bacterial cultures were positive, compared with 1 of 13 JTR mice (7.7%) and 1 of 11 Sham mice (9%) (Table 2). Liver bacterial cultures were positive in 10% of SFBL mice and were negative in all JTR and Sham mice. The bacterial cultures of peritoneal swabs and portal vein blood were negative in all mice. We measured the LPS concentration in portal vein blood because LPS is a major bacteria-derived factor and has been
SFBL (10)
Liver
MLN
Peritoneum
PV blood
0%
9%
0%
0%
0% 10%
7.7% 100%*
0%
0%
0%
0%
PV ¼ portal vein. * P < 0.0001 compared to Sham and JTR mice.
shown to play an important role in hepatic inflammation and cholestasis. The LPS concentration in portal vein serum in SFBL mice, as well as in JTR and Sham mice, was below the detection level of the assay but well detected in the positive control.
Biochemical and histological evidence of liver and jejunal injury in SFBL mice To understand whether SBBO in SFBL mice is associated with evidence of liver injury, plasma aspartate aminotransferase, alanine aminotransferase, and alkaline phosphatase were measured, and liver histology was examined. Plasma alkaline phosphatase was significantly elevated in SFBL mice, compared with JTR and Sham mice (P < 0.05), whereas alanine aminotransferase and aspartate aminotransferase were not significantly different among all three group of mice (Fig. 3). Histologically, SFBL livers were characterized by liver injury with evidence of hepatocyte necrosis, as well as infiltration of neutrophils into the biliary epithelium consistent with cholangitis (Fig. 4A), whereas JTR and Sham livers did not demonstrate such findings. Histologic scoring of liver sections revealed that SFBL mice had significantly greater mean scores for hepatocellular injury and cholangitis, compared with JTR and Sham mice (Fig. 4B). MLN of SFBL mice had focal acute lymphadenitis, characterized by neutrophilic infiltration, whereas MLN of JTR and Sham mice did not demonstrate such findings (data not shown). Histologically, the blind jejunal loop in SFBL mice was characterized by shortened villi, crypt proliferation, areas of focal severe cryptitis and crypt distortion, inflammation in the lamina propria, and transmural inflammation (Fig. 5A). These findings are consistent with disruption of the jejunal barrier in SFBL mice. The jejunum in JTR and Sham mice had normal crypt architecture and no inflammation. SFBL mice exhibited significantly higher mean scores for cryptitis, crypt distortion, lamina propria inflammation, and bowel wall inflammation in the blind jejunal loop, compared with jejunal sections from JTR and Sham mice (Fig. 5B).
Alterations in hepatobiliary transporters induced by creation of SFBL Fig. 2 e Bacterial overgrowth in jejunal blind loop in SFBL mice. At week 3 postoperatively, blind loop content from SFBL mice was isolated and cultured for the presence of aerobic and anaerobic bacteria. Jejunal luminal content from Sham and JTR groups were used as control. Data are represented as mean bacterial concentration (log10 CFU/ ml) ± standard error of the mean. ***P < 0.001.
Because SBBO in SFBL mice resulted in disruption of the jejunal barrier and BT to MLN, we hypothesized that chronic exposure of the liver to microbe-derived factors from the jejunal blind loop may cause changes in hepatobiliary transporter expression. Using real-time polymerase chain reaction to measure expression of hepatobiliary transporters in the
wang et al liver impact of bacterial overgrowth
Fig. 3 e Liver biochemical abnormalities in SFBL mice. Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), and alkaline phosphatase (Alk phos) from Sham, JTR and SFBL mice were measured at the time of euthanasia. *P < 0.05. liver, our data demonstrates that SBBO induces alterations in mRNA expression of multiple hepatobiliary transporters. Expression of 8 out of 10 transporters examined was significantly lower in SFBL mice, compared with Sham controls (Fig. 6). Significant reductions were observed in mRNA expression of four transporter genes associated with biliary canalicular export: Abcg8 (P < 0.0001), Bsep1 (P < 0.0001), Mrp2 (P < 0.01), and Mdr2 (P < 0.05). In addition, we found significant decreases in hepatic expression of all four transporter genes associated with basolateral uptake: Ntcp (P < 0.01), Oatp1 (P < 0.0001), Oatp2 (P < 0.01), and Oatp4 (P < 0.05).
Alterations in bile composition and bile acid/phospholipid ratio in SFBL mice To test whether SBBO/BT-associated reductions in hepatobiliary transporter expression result in altered bile flow and/or composition, we measured these parameters in SFBL mice and compared them with Sham mice. We found that the creation of SFBL caused significant perturbation of the bile composition in the hepatobiliary system, compared to Sham mice, but no differences in the portal vein serum concentrations of total bile acid, phospholipid, or cholesterol were observed between the two groups (Table 3). Although the mean bile flow rate in SFBL mice was 34% lower than the rate in Sham mice, this difference did not reach statistical significance (P ¼ 0.09). Biliary mean total bile acid concentration was significantly higher in SFBL mice, compared with Sham (39.90 2.57 mmol/L versus 23.78 3.85 mmol/L, P < 0.01), but total bile acid output was not different between the two groups. Significant reductions in bile phospholipid and cholesterol output were observed in SFBL mice, which resulted in a significant increase in the bile acid/phospholipid ratio in SFBL mice, suggestive of the formation of toxic bile.
Discussion In the present study, we have recapitulated SBBO in mice by surgically creating a jejunal SFBL, providing a reproducible model that physiologically resembles SBBO in the clinical
251
setting. We have shown inflammation and barrier disruption in the blind loop with MLN translocation, and we have demonstrated histologic evidence of liver injury, as well as a reduction in hepatobiliary transporter expression and alterations in bile composition. Our data clearly demonstrate the effect of SBBO and BT on the liver, thereby providing critical insight into the gut-liver axis. Importantly, our mouse model has the distinct advantage of application to genetically modified animals for further dissection of mechanisms underlying the impact of SBBO and translocation on liver disease pathogenesis. There is increasing evidence that intestinal microbiota or bacteria-derived factors play an important role in liver injury in many chronic liver diseases, including NAFLD,31,33,34 PSC,10,35,36 PBC,7 PNALD,11,12 and CF-associated liver disease.37-40 Studies suggest that PBC is associated with an antibody response to particular alphaproteobacteria (Novosphingobium aromaticivorans)41; however, the pathogenic mechanisms that result in the manifestations of PBC remain obscure. Approximately, 80% of cases of PSC are associated with inflammatory bowel disease, supporting the “leaky gut hypothesis” whereby bacteria and/or bacterial components enter the portal venous system by penetrating the inflamed intestinal mucosa and induce an inflammatory response in the liver leading to fibro-obliterative sclerosis of intrahepatic and extrahepatic bile ducts.42 An increased prevalence of SBBO associated with increased intestinal permeability has been reported in NAFLD patients,15 and in animal studies, dextran sulfate sodium treatment induces colitis and promotes hepatic inflammation and fibrosis in mice with nonalcoholic steatohepatitis.34 The small bowel in infants with short-bowel syndrome is inherently affected by intestinal stasis, and interval dilation of small-bowel loops in PNALD patients can promote bacterial overgrowth, translocation, and recurrent episodes of sepsis, further contributing to liver dysfunction and intrahepatic cholestasis.43 Approximately, 30%-40% of CF patients have been reported to have SBBO, and CF transmembrane receptor knock-out mice demonstrate bacterial overgrowth in the small intestine.38,39 Despite these studies supporting an important role for bacteria in various liver diseases, the pathogenic mechanisms of such an interaction remain poorly understood. A rat model of SBBO was reported over 20 y ago, yet the mechanisms of SBBO-associated liver injury remain unclear.27,29,30 This is largely due to the lack of availability of genetically modified rats. Although a recently reported SFBL mouse model, using a distal ileal blind loop, has shed light on the role of microbiota in ulcerative colitis, there has been no report of the hepatic effects of SBBO/BT in mice.32 Clinically, SBBO is defined as an increase in number and/or alteration in the type of bacteria in the upper gastrointestinal tract.19 In the present study, SFBL mice had 105- to 106-fold higher bacterial colonies in the jejunal blind loop compared with control groups, whereas the number of bacterial colonies in Sham and JTR groups was similar to human studies of the small intestine.17 Overgrowth of bacteria in the blind loop can disrupt the barrier function of the small intestine, leading to translocation of endogenous bacteria or their products through the injured epithelial barrier to MLNs, the systemic circulation, or extraintestinal organs. Indeed, we demonstrated 100% MLN translocation in SFBL mice. The absence of BT to liver or portal venous blood in SFBL mice allows speculation that the hepatic
252
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
Fig. 4 e Histological evidence of hepatobiliary injury in SFBL mice. (A) Hematoxylin and eosin staining of liver sections from Sham, JTR, and SFBL mice at 3 weeks after surgery. SFBL mice showed evidence of hepatocyte necrosis (represented by arrows), neutrophil infiltration (arrow with one asterisk), and cholangitis (arrow with two asterisks). Sham and JTR controls exhibited no evidence of inflammation or injury. (B) Mean liver histology score among Sham, JTR, and SFBL mice. *P < 0.05 compared with Sham and JTR mice. (Color version of figure is available online.)
wang et al liver impact of bacterial overgrowth
253
Fig. 5 e Histological evidence of jejunal wall injury in SFBL mice. (A) Hematoxylin and eosin (H&E) staining of jejunal sections from Sham and SFBL mice. SFBL mice had evidence of shortened villi (represented by dashed lines), crypt proliferation (solid lines), areas of focal, severe cryptitis and crypt distortion (arrow), inflammation in the lamina propria, and transmural inflammation (arterisks). Sham mice had normal crypt architecture and no inflammation. (B) Mean jejunal histology score among Sham, JTR, and SFBL mice. *P < 0.05 compared with Sham and JTR mice. (Color version of figure is available online.)
254
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
Fig. 6 e Alterations in hepatobiliary transporters in SFBL mice. Hepatic mRNA was subjected to real-time polymerase chain reaction analysis for evaluation of hepatobiliary transporter expression. Relative gene expression was calculated according to DDCt method after normalization to b-actin internal control. (A) Transporter genes associated with biliary canalicular export. (B) Transporter genes associated with basolateral uptake. *P < 0.05, **P < 0.01, ***P < 0.001.
Table 3 e Effect of SFBL on bile secretion and composition. Sham (n ¼ 8)
SFBL (n ¼ 8)
1.64 0.19
1.09 0.23
Total bile acid
23.78 3.85
39.90 2.57y
Phospholipids
14.32 1.18
12.44 2.02
2.92 0.51
1.89 0.33
Total bile acid
39.90 7.72
40.45 7.39
Phospholipids
24.47 4.01
12.20 2.44*
Cholesterol
5.16 1.38
1.84 0.53*
Bile acid/phospholipid ratio
1.72 0.27
3.91 0.85*
Total bile acid
19.80 5.87
27.20 3.70
Phospholipids
2.76 0.34
2.09 0.28
Cholesterol
2.99 0.24
2.54 0.25
Bile analysis parameter Bile flow rate (mL/g/min) Biliary concentrations (mmol/L)
Cholesterol Biliary output (nmol/g/min)
Serum concentrations (mmol/L)
*
P < 0.05. y P < 0.01 compared to Sham mice.
effects related to SBBO and BT may not be due to bacteria themselves but rather to other factors triggered by SBBO and BT, such as immune response to LPS or other bacteria-derived signals, such as bacterial DNA, single-stranded DNA, or double-stranded RNA.24,44 Alternatively, efficient hepatic clearance of bacteria in early stages of SBBO and intermittent portal venous bacteremia may be explanations for these findings. Our mouse model provides evidence that SBBO in wild-type C57BL/6 mice causes injury and inflammation in the blind loop, supporting the notion that SBBO results in intestinal barrier dysfunction with potential for hepatic consequences. We hypothesize that the local injury and inflammation of the blind loop increases local intestinal permeability in our model, thereby exposing the liver to bacteria or bacteria-derived factors, such as LPS. This is supported by prior studies that have demonstrated that derangement of the homeostasis between bacteria and the host, as occurs in SBBO, may cause disruption of the intercellular tight junctions and subsequent increase in intestinal permeability leading to translocation of bacteria and bacterial products.45,46 Creation of SFBL in mice induced mild liver injury evidenced by significantly higher alkaline phosphatase levels in the blood and elevated histological scores of hepatocellular injury and cholangitis. Portal and lobular inflammation and fibrosis were not observed in SFBL mice, and serum bilirubin and g-GGT levels were not different between SFBL and Sham mice (data not shown). These data support the notion that gut-derived factors are more likely to act as the “second hit” or environmental trigger in chronic liver diseases, as suggested by Jahnel et al.26 and others.47 The two-hit model for liver diseases such as PSC has been proposed, suggesting that specific mutations (first hit) in an individual can predispose the individual to an inappropriate response to an environmental trigger (second hit). On this basis, we speculate that our mouse SFBL model would be particularly useful in investigating the role of gut-derived factors in predisposed mouse strains, such as in models of PSC, NAFLD, and PNALD. Furthermore, since specific bacterial species are implicated in the pathogenesis of autoimmune liver diseases such as PBC,7 this mouse model may provide the opportunity to dissect its pathogenesis via inoculation of specific bacteria into the blind loop. Evidence suggests that intestinal microbiota affect bile acid metabolism.22 Microbial products, especially LPS, have been shown to cause changes in hepatic transporter expression and bile formation, as well as result in cholestasis.48-50 In the present study, we hypothesized that SBBO and translocation induces changes in bile composition through reduction in hepatobiliary transporter expression. A significant reduction in expression of hepatobiliary transporters involved in biliary canalicular export (Abcg8, Bsep, Mrp2, and Mdr2), as well as basolateral uptake (Ntcp, Oatp1, Oatp2, and Oatp4), was observed in SFBL mice. Our findings are consistent with those of Jahnel et al.,26 in which LPS administration to mice resulted in decreased expression of hepatobiliary transporters, further suggestive of the role of LPS as a major gut-derived factor. Although we did not find a significant increase in LPS levels in portal vein serum at euthanasia, it is quite possible that LPS levels increased intermittently during the postoperative course after SFBL creation and that our failure to detect an increase was related to the fact that we only attempted portal
wang et al liver impact of bacterial overgrowth
vein LPS measurement at a single time point. Dynamic monitoring of LPS levels during the experiment would be needed to further confirm its alteration. In contrast to LPS administration by Jahnel et al.,26 which resulted in increased bile acid output, no change in phospholipids or cholesterol output, and decreased bile acid/phospholipid ratio, we observed no change in bile acid output, decreased phospholipid and cholesterol output and increased bile acid/ phospholipid ratio. These results indicate that hepatic pathophysiology associated with SBBO is complex and likely multifactorial, and it may not be caused by one single bacterial factor such as LPS. Other microbe-associated molecular patterns or pathogen-associated molecular patterns, such as flagellin, peptidoglycan, and microbial metabolites such as short-chain fatty acids, could have significant effects on the alteration of bile composition and hepatobiliary injury.51 From this point of view, our SFBL model would be considered a more physiologic model to study the link between intestinal microbial homeostasis and liver injury. Bile containing a relatively high bile acid/phospholipid ratio is potentially deleterious to the bile duct system due to the inadequate neutralization of the detergent and cytotoxic effects of bile salts by phospholipids.52 Geuken et al.52 reported that cytotoxic bile formation caused by more rapid recovery of bile salt secretion than phospholipid secretion after human liver transplantation correlated with bile duct injury. In the current study, we observed increased bile acid/phospholipid ratio of bile, as well as histologic evidence of cholangitis, in SFBL mice, providing evidence that SBBO alters bile composition with formation of toxic bile and may play a pathogenic role in biliary and hepatocellular injury. In parallel with the observed decreases in cholesterol and phospholipid output, we observed decreases in Abcg8 and Mdr2 mRNA expression in the liver. These results are consistent with the bile alterations found in Mdr2 knockout mice,53 which demonstrated a primary role of Mdr2 for phospholipid excretion in the liver. Surprisingly, we observed an increase in bile acid concentration, but a decrease in Bsep mRNA expression in livers from SFBL mice. Wang et al.54 have previously demonstrated that Bsep knockout mice exhibit a 70% reduction in total bile salt output, but an unexpectedly large amount of tetrahydroxylated bile acid secretion, compared with wild-type mice. Moreover, Smit et al.53 have demonstrated that total bile acid output tended to be higher both in (/þ) and (/) Mdr2 mice than in wild-type control animals. All of the above data suggest that there are likely alternative transport pathways for bile acid excretion, which are independent of Bsep. In addition, studies using mass spectrometry to analyze the primary and secondary bile salts are needed for further assess the bile salt composition in our model.
Conclusions Intestinal microbiota or bacteria-derived factors are likely to play a critical role in the modulation and exacerbation of a number of chronic liver diseases. We have demonstrated that SBBO can initiate histologic evidence of liver injury, thereby providing an important perspective of the gut-liver axis in terms of the effect of bacteria or bacteria-derived products.
255
Using this surgical mouse model, we demonstrated that SBBO induced mild liver injury with decreased hepatobiliary transporter expression and altered bile composition. This model can be used in various genetically modified mice and/or applied to in vivo treatments to further study the interactions between gut-derived factors and liver disease. The applicability of genome- and transcriptome-level approaches to the SFBL mouse model has the potential to greatly advance our understanding of the mechanisms by which SBBO and translocation may impact liver disease pathogenesis.
Acknowledgment The authors thank Dr Michael Trauner and Dr Jo¨rg Jahnel for their advice and assistance regarding the techniques of bile collection and bile composition measurements. Author contributions: Q.W., B.W., and V.S. performed the experiments, collected the data, and drafted the manuscript. L.M. provided substantial technical support and data analysis and conceptual advice regarding data analysis, and critically revised the manuscript. J.T. and J.E.M. helped carry out the experiments and collect the data, and critically revised the manuscript. J.D.N. contributed to the conception, design and coordination of the study and revision of the manuscript. All authors gave the final approval of the submitted version of the manuscript. Funding: This research was supported by internal institutional funding provided by Cincinnati Children’s Hospital Medical Center. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Disclosure The authors report no proprietary or commercial interest in any product mentioned or concept discussed in the article.
references
1. Son G, Kremer M, Hines IN. Contribution of gut bacteria to liver pathobiology. Gastroenterol Res Pract. 2010;2010:1e13. 2. Schnabl B. Linking intestinal homeostasis and liver disease. Curr Opin Gastroenterol. 2013;29:264e270. 3. Seki E, Schnabl B. Role of innate immunity and the microbiota in liver fibrosis: crosstalk between the liver and gut. J Physiol. 2012;590(Pt 3):447e458. 4. Machado MV, Cortez-Pinto H. Gut microbiota and nonalcoholic fatty liver disease. Ann Hepatol. 2012;11:440e449. 5. Chassaing B, Etienne-Mesmin L, Gewirtz AT. Microbiota-liver axis in hepatic disease. Hepatology. 2014;59:328e339. 6. Kleiner DE, Brunt EM, Van Natta M, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology. 2005;41:1313e1321. 7. Mattner J, Savage PB, Leung P, et al. Liver autoimmunity triggered by microbial activation of natural killer T cells. Cell Host Microbe. 2008;3:304e315. 8. Fouts DE, Torralba M, Nelson KE, Brenner DA, Schnabl B. Bacterial translocation and changes in the intestinal microbiome in mouse models of liver disease. J Hepatol. 2012;56:1283e1292.
256
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
9. Abu-Shanab A, Quigley EM. The role of the gut microbiota in nonalcoholic fatty liver disease. Nat Rev Gastroenterol Hepatol. 2010;7:691e701. 10. Riordan SM, Williams R. The intestinal flora and bacterial infection in cirrhosis. J Hepatol. 2006;45:744e757. 11. Pollheimer MJ, Halilbasic E, Fickert P, Trauner M. Pathogenesis of primary sclerosing cholangitis. Best Pract Res Clin Gastroenterol. 2011;25:727e739. 12. Kelly DA. Intestinal failure-associated liver disease: what do we know today? Gastroenterology. 2006;130(2 Suppl 1):S70eS77. 13. Kalambokis GN, Tsianos EV. Rifaximin reduces endotoxemia and improves liver function and disease severity in patients with decompensated cirrhosis. Hepatology. 2012;55:655e656. 14. Schnabl B, Brenner DA. Interactions between the intestinal microbiome and liver diseases. Gastroenterology. 2014;146:1513e1524. 15. Compare D, Coccoli P, Rocco A, et al. Guteliver axis: the impact of gut microbiota on non alcoholic fatty liver disease. Nutr Metab Cardiovasc Dis. 2012;22:471e476. 16. Cesaro C, Tiso A, Del Prete A, et al. Gut microbiota and probiotics in chronic liver diseases. Dig Liver Dis. 2011;43:431e438. 17. Gorbach SL, Bengt E. Gustafsson memorial lecture. Function of the normal human microflora. Scand J Infect Dis Suppl. 1986;49:17e30. 18. Berg RD, Garlington AW. Translocation of certain indigenous bacteria from the gastrointestinal tract to the mesenteric lymph nodes and other organs in a gnotobiotic mouse model. Infect Immun. 1979;23:403e411. 19. Bures J, Cyrany J, Kohoutova D, et al. Small intestinal bacterial overgrowth syndrome. World J Gastroenterol. 2010;16:2978e2990. 20. Yan AW, Fouts DE, Brandl J, et al. Enteric dysbiosis associated with a mouse model of alcoholic liver disease. Hepatology. 2011;53:96e105. 21. Pinzone MR, Celesia BM, Di Rosa M, Cacopardo B, Nunnari G. Microbial translocation in chronic liver diseases. Int J Microbiol. 2012;2012:694629. 22. de Aguiar Vallim TQ, Tarling EJ, Edwards PA. Pleiotropic roles of bile acids in metabolism. Cell Metab. 2013;17:657e669. 23. Swann JR, Want EJ, Geier FM, et al. Systemic gut microbial modulation of bile acid metabolism in host tissue compartments. Proc Natl Acad Sci U S A. 2011;108(Suppl 1):4523e4530. 24. Szabo G, Bala S, Petrasek J, Gattu A. Gut-liver axis and sensing microbes. Dig Dis. 2010;28:737e744. 25. Trauner M, Fickert P, Stauber RE. Inflammation-induced cholestasis. J Gastroenterol Hepatol. 1999;14:946e959. 26. Jahnel J, Fickert P, Langner C, et al. Impact of experimental colitis on hepatobiliary transporter expression and bile duct injury in mice. Liver Int. 2009;29:1316e1325. 27. Lichtman SN, Sartor RB. Hepatobiliary injury associated with experimental small-bowel bacterial overgrowth in rats. Immunol Res. 1991;10:528e531. 28. Lichtman SN, Keku J, Schwab JH, Sartor RB. Evidence for peptidoglycan absorption in rats with experimental small bowel bacterial overgrowth. Infect Immun. 1991;59:555e562. 29. Lichtman SN, Keku J, Schwab JH, Sartor RB. Hepatic injury associated with small bowel bacterial overgrowth in rats is prevented by metronidazole and tetracycline. Gastroenterology. 1991;100:513e519. 30. Lichtman SN, Keku J, Clark RL, Schwab JH, Sartor RB. Biliary tract disease in rats with experimental small bowel bacterial overgrowth. Hepatology. 1991;13:766e772. 31. Shanab AA, Scully P, Crosbie O, et al. Small intestinal bacterial overgrowth in nonalcoholic steatohepatitis: association with toll-like receptor 4 expression and plasma levels of interleukin 8. Dig Dis Sci. 2011;56:1524e1534.
32. Ward MA, Pierre JF, Leal RF, et al. Insights into the pathogenesis of ulcerative colitis from a murine model of stasis-induced dysbiosis, colonic metaplasia, and genetic susceptibility. Am J Physiol Gastrointest Liver Physiol. 2016;310:G973eG988. 33. Miele L, Valenza V, La Torre G, et al. Increased intestinal permeability and tight junction alterations in nonalcoholic fatty liver disease. Hepatology. 2009;49:1877e1887. 34. Gabele E, Dostert K, Hofmann C, et al. DSS induced colitis increases portal LPS levels and enhances hepatic inflammation and fibrogenesis in experimental NASH. J Hepatol. 2011;55:1391e1399. 35. Schrumpf E, Boberg KM. Epidemiology of primary sclerosing cholangitis. Best Pract Res Clin Gastroenterol. 2001;15:553e562. 36. O’Mahony CA, Vierling JM. Etiopathogenesis of primary sclerosing cholangitis. Semin Liver Dis. 2006;26:3e21. 37. Herrmann U, Dockter G, Lammert F. Cystic fibrosis-associated liver disease. Best Pract Res Clin Gastroenterol. 2010;24:585e592. 38. Lewindon PJ, Robb TA, Moore DJ, Davidson GP, Martin AJ. Bowel dysfunction in cystic fibrosis: importance of breath testing. J Paediatr Child Health. 1998;34:79e82. 39. O’Brien S, Mulcahy H, Fenlon H, et al. Intestinal bile acid malabsorption in cystic fibrosis. Gut. 1993;34:1137e1141. 40. Norkina O, Kaur S, Ziemer D, De Lisle RC. Inflammation of the cystic fibrosis mouse small intestine. Am J Physiol Gastrointest Liver Physiol. 2004;286:G1032eG1041. 41. Mattner J. Genetic susceptibility to autoimmune liver disease. World J Hepatol. 2011;3:1e7. 42. Llorente C, Schnabl B. The gut microbiota and liver disease. Cell Mol Gastroenterol Hepatol. 2015;1:275e284. 43. Beath SV, Davies P, Papadopoulou A, et al. Parenteral nutrition-related cholestasis in postsurgical neonates: multivariate analysis of risk factors. J Pediatr Surg. 1996;31:604e606. 44. Kesar V, Odin JA. Toll-like receptors and liver disease. Liver Int. 2014;34:184e196. 45. Ferolla SM, Armiliato GN, Couto CA, Ferrari TC. The role of intestinal bacteria overgrowth in obesity-related nonalcoholic fatty liver disease. Nutrients. 2014;6:5583e5599. 46. Vajro P, Paolella G, Fasano A. Microbiota and gut-liver axis: their influences on obesity and obesity-related liver disease. J Pediatr Gastroenterol Nutr. 2013;56:461e468. 47. Mayer L. Redefining autoimmunity. Gastroenterology. 2003;125:1574. 48. Kosters A, Karpen SJ. The role of inflammation in cholestasis: clinical and basic aspects. Semin Liver Dis. 2010;30:186e194. 49. Bolder U, Ton-Nu HT, Schteingart CD, Frick E, Hofmann AF. Hepatocyte transport of bile acids and organic anions in endotoxemic rats: impaired uptake and secretion. Gastroenterology. 1997;112:214e225. 50. Moseley RH, Wang W, Takeda H, et al. Effect of endotoxin on bile acid transport in rat liver: a potential model for sepsisassociated cholestasis. Am J Physiol. 1996;271(1 Pt 1):G137eG146. 51. Ha CW, Lam YY, Holmes AJ. Mechanistic links between gut microbial community dynamics, microbial functions and metabolic health. World J Gastroenterol. 2014;20:16498e16517. 52. Geuken E, Visser D, Kuipers F, et al. Rapid increase of bile salt secretion is associated with bile duct injury after human liver transplantation. J Hepatol. 2004;41:1017e1025. 53. Smit JJ, Schinkel AH, Oude Elferink RP, et al. Homozygous disruption of the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease. Cell. 1993;75:451e462. 54. Wang R, Salem M, Yousef IM, et al. Targeted inactivation of sister of P-glycoprotein gene (spgp) in mice results in nonprogressive but persistent intrahepatic cholestasis. Proc Natl Acad Sci U S A. 2001;98:2011e2016.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
The gut-liver axis: impact of a mouse model of small-bowel bacterial overgrowth Qingqing Wang, MD,a,1 Bin Wang, MD,a,1 Vijay Saxena, PhD,a Lili Miles, MD,b Josh Tiao,a Joel E. Mortensen, PhD,b and Jaimie D. Nathan, MDa,* a b
Division of Pediatric General and Thoracic Surgery, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Division of Pathology and Laboratory Medicine, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio
article info
abstract
Article history:
Background: The mechanisms by which intestinal bacteria impact liver diseases remain
Received 29 March 2017
poorly understood. The aim of this study was to develop a mouse model of small-bowel
Received in revised form
bacterial overgrowth and to determine its impact on hepatobiliary injury.
7 August 2017
Materials and methods: A jejunal self-filling blind loop (SFBL) was created in C57BL/6 mice.
Accepted 30 August 2017
Three weeks after surgery, the mice were euthanized, and bacterial cultures of luminal
Available online xxx
content of the loop and extraintestinal tissues were performed. Liver and jejunum were collected for histological grading of inflammation and injury. Serum liver biochemistry
Keywords:
assays were performed. Hepatobiliary transporter mRNA expression in liver was measured
Gut-derived factors
by quantitative real-time polymerase chain reaction. Bile and blood were collected for
Small-bowel bacterial overgrowth
measurement of total bile acids, phospholipid, and cholesterol. Mice undergoing jejunal
Bacterial translocation
transection and reanastomosis and laparotomy only served as control groups.
Self-filling blind loop
Results: SFBL induced a dramatic increase in intraluminal bacterial counts, mesenteric
Liver injury
lymph node bacterial translocation, and evidence of jejunal and hepatobiliary injury. Sig-
Bile composition
nificant reductions in hepatic expression of hepatobiliary transporters involved in biliary canalicular export and basolateral uptake were observed in SFBL mice. SFBL resulted in a significant increase in biliary total bile acid concentration, decreases in bile phospholipid and cholesterol output, and an increase in the bile acid/phospholipid ratio. Conclusions: We have developed a reproducible mouse model of small-bowel bacterial overgrowth with evidence of liver inflammation, altered hepatobiliary transporter expression, and alterations in bile composition. This model may help to elucidate the mechanisms by which gut-derived bacterial factors impact the liver and contribute to the exacerbation of liver diseases and biliary injury. ª 2017 Elsevier Inc. All rights reserved.
Introduction Emerging evidence suggests that intestinal microbiota or bacteria-derived factors play a critical role in the modulation
and exacerbation of a number of liver diseases,1-5 such as primary biliary cirrhosis (PBC), primary sclerosing cholangitis (PSC), nonalcoholic fatty liver disease (NAFLD), parenteral nutrition-associated liver disease (PNALD), alcoholic liver
* Corresponding author. Division of Pediatric General and Thoracic Surgery, Cincinnati Children’s Hospital Medical Center, MLC 2023, 3333 Burnet Avenue, Cincinnati, OH 45229-3039. Tel.: þ513 636-4371; fax: þ513 636-7657. E-mail address: [email protected] (J.D. Nathan). 1 These authors contributed equally to this work. 0022-4804/$ e see front matter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2017.08.049
wang et al liver impact of bacterial overgrowth
diseases, and cystic fibrosis (CF)eassociated liver disease.6-12 However, the pathogenic link between intestinal microbiota and chronic liver disease requires further elucidation. Furthermore, several studies support the notion that small-bowel bacterial overgrowth (SBBO) and bacterial translocation (BT) may play a crucial role in the progression of liver diseases.13-16 The normal human microflora maintains a delicate balance between its constituent parts, numbering 1011 bacteria per gram content of the small intestine with over 400 different species.17 SBBO is defined as an increased number and/or abnormal type of bacteria in the small bowel, and BT is defined as the passage of viable endogenous bacteria or their products from the intestinal tract through the epithelial barrier to the mesenteric lymph nodes (MLNs), systemic circulation, or extraintestinal organs.18,19 Although studies from both animal models8,20 and patients with liver disease13,21 have shown that SBBO and BT are critical in disease progression, the mechanisms by which they modulate intestinal barrier dysfunction and contribute to liver disease are still unclear. One mechanism by which microbial factors play a role in hepatic function is in the metabolism of bile acids because they participate in the generation of secondary bile acids, which are subsequently taken up by the ileum to enter the enterohepatic circulation.22 Through modulation of bile, microbiota can interact with the host and impact not only the liver diseases but also other organs and metabolic pathways.23 This represents one of the crucial manners by which intestinal microbiota communicate with the liver. One such microbe-derived factor, lipopolysaccharide (LPS), present in the outer membrane of gramnegative bacteria, has been demonstrated to be a key mediator in inflammation-induced cholestasis, a process resulting from either a defect in bile formation at the level of hepatocyte or from an impairment of bile secretion and flow at the level of bile ducts.24,25 LPS and/or LPS-induced proinflammatory cytokines, such as tumor necrosis factor-alpha and various interleukins, mediate the cholestatic effect via inhibition of the expression and function of the hepatobiliary transporter system.26 Despite the recognition of the importance of microbial factors, including SBBO and BT in liver disease, there remains a lack of relevant animal models for the investigation of disease pathogenesis. Approximately two decades ago, Lichtman et al. surgically created jejunal self-filling blind loops (SFBLs) in rat, and observed bacterial overgrowth in the jejunal loop and mesenteric translocation. Although histologic and biochemical changes were found in rats with SFBL, and antibiotic treatment prevented hepatic injury in SFBL rats, further light could not be shed on the pathogenesis of liver injury associated with SBBO/BT.27-31 Studying such interactions in a rat model of SBBO is hampering, and genetically-modified mice represent powerful tools to study such interactions. A recent study has described a surgical model of SFBL in mice, using a blind loop created in the distal ileum, to study the role of microbiota in the pathogenesis of ulcerative colitis.32 In the current study, we describe a mouse model of SBBO/BT using a jejunal blind loop, and we investigate its impact on hepatic injury and function. We hypothesize that SBBO/BT induced by the creation of a jejunal SFBL in mice would result in hepatic injury and alterations in hepatobiliary transporter expression and bile composition.
247
Materials and methods Mice Approximately, 5- to 6-week-old C57BL/6 male mice (weight range, 20-25 gm) were purchased from Charles River (Wilmington, MA). Mice were fed standard chow and had access to tap water ad libitum. The mice were maintained in a 12:12-hr day-night rhythm at a constant temperature of 23 C and a relative humidity of 40%-60%. The mice were given liquid diet (microstabilized rodent liquid diet LD101; TestDiet, St. Louis, MO) 24 h before surgery. After surgery, only water was provided ad libitum for the first 24 h. Thereafter, all mice were fed liquid diet until euthanasia. The mice’s body weights were measured daily. A protocol for this study was approved by the Cincinnati Children’s Hospital Medical Center Institutional Animal Care and Use Committee (protocol no. 9D12102).
Experimental design and surgical procedures The mice were anesthetized with isoflurane. To create SFBL mice, a 2-cm midline laparotomy incision was made, the jejunum was divided 8 cm distal to the ligament of Treitz, and a 3-cm jejunal SFBL was created by an end-to-side anastomosis of the distal jejunal stump to the side of the proximal jejunal limb 5 cm distal to the ligament of Treitz using interrupted 9-0 monofilament suture (Ethicon Endo-Surgery, Cincinnati, OH) (Fig. 1A). For jejunal transection and reanastomosis (JTR) control mice, the jejunum was partially transected at 50% of diameter 5 cm distal to the ligament of Treitz and resutured in interrupted fashion using 9-0 monofilament suture. Shamoperated (Sham) control mice underwent laparotomy without any intestinal transection. We use two cohorts of mice for the study. The first cohort of mice was used for most of the experiments including bacterial cultures, LPS measurement, biochemical study, liver and small-intestine histology, and hepatobiliary transporter expression. For this cohort, we have 11 mice in the Sham group, 13 mice in the JTR group, and 10 mice in the SFBL group, with the unbalanced totals secondary to early postoperative mortality. The second cohort of mice was used for measuring bile flow and composition. We have two groups of mice in this cohort, including eight in the Sham group and eight in the SFBL group.
Bacterial cultures of organs All mice were euthanized at week 3 after surgery. Mice were anesthetized by triple-mix (12 mL triple-mix solution contained 8.0 mL ketamine at 100 mg/mL, 0.4 mL xylazine at 100 mg/mL, 2 mL acepromazine at 10 mg/mL, and 1.6 mL normal saline), given intramuscularly. Using sterile techniques to expose the abdominal cavity, the skin was opened and then the muscle and peritoneal cavity were opened with a second set of sterile instruments. Peritoneal cultures were performed by swabbing the viscera with sterile cotton-tipped applicator sticks. Portal vein blood of 50 mL was diluted in 500-ml Baltimore Biological Laboratoriesenriched thioglycollate medium (BD Biosciences, Sparks,
248
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
Fig. 1 e Creation of SFBL in C57BL/6 mice. (A) A schematic diagram of SFBL surgical model showing a 3-cm jejunal blind loop, created 5 cm distal to the ligament of Treitz; (B) A representative mouse intraabdominal photograph showing the dilated SFBL (indicated by white arrow) at the time of euthanasia; (C) Changes in mouse body weight at indicated time points after surgery (Sham, n [ 11; JTR, n [ 13; SFBL, n [ 10). (Color version of figure is available online.)
MD). Each diluted blood sample of 100 mL was plated on trypticase soy agar with 5% sheep blood agar plates (BD Biosciences, Sparks, MD) for aerobic culture and on Brucella agar with 5% sheep blood agar plates (PML microbiologicals Inc, Wilsonville, OR) for anaerobic cultures. Liver (50-100 mg) and MLNs (5-20 mg) were collected and homogenized in a 500-mL culture medium. Each sample of 100 mL was cultured aerobically and anaerobically. The aerobic cultures were incubated at 37 C for 24 h, and anaerobic cultures were incubated in BD GasPak EZ Gas Generating System (BD Biosciences, Sparks, MD) for 40 h.
Bacterial cultures of luminal contents Sterile saline of 2 mL was flushed through the loop or corresponding 3-cm segment of jejunum of control mice. The fluid was collected and diluted serially in medium from 101 to 108 and cultured aerobically and anaerobically.
Biochemical assays Cardiac blood was collected for liver biochemical profile at the time of euthanasia and stored at 80 C until analysis. The
249
wang et al liver impact of bacterial overgrowth
liver biochemical profile was measured by the clinical laboratory at TriHealth laboratories, Cincinnati, OH.
Lipopolysaccharide measurement Portal vein serum was collected at the time of euthanasia for LPS measurement using Limulus Amebocyte Lyisate assay kit (Lonza, Walkersville, MD). Briefly, serum samples were preheated at 70 C for 30 min to deactivate the inhibitory factors for the reaction. Standard LPS from the kit was used to create a standard curve. Bacterial LPS from mouse small intestinal lumen was used as positive control. Pyrogen-free water was used as a negative control. Absorbance of all samples and standards was read at 405 nm. The LPS concentration (EU/ml) was calculated using the standard curve.
Real-time polymerase chain reaction for hepatobiliary transporter expression Total liver RNA was isolated by TRIzol and subjected to cDNA synthesis using oligo-dT primers dT12-18 (Invitrogen, Carlsbad, CA) and SuperScript II Reverse Transcriptase enzyme (Invitrogen, Carlsbad, CA). Using a specific primer set (Table 1), polymerase chain reaction was performed using SYBR green QPCR mix (Agilent technologies, Santa Clara, CA). All samples were loaded in duplicate wells, and data analysis was performed using DDCt method. b-Actin was used as internal control. Polymerase chain reaction conditions were as follows: 95 C for 10 min to activate the DNA enzyme, followed by 40 cycles of 95 C for 30 s (denaturation), 55 C-60 C for 1 min (annealing), and 72 C for 30 s (extension).
Histological scoring Liver and jejunum were collected, immediately placed into 10% neutral-buffered formalin for 24 h and transferred to 70% alcohol thereafter. Tissues were embedded in paraffin, and 5-mm thick sections were mounted and stained with hematoxylin and eosin. To quantify the differences between Sham, JTR, and SFBL liver and jejunum, histology scoring was performed on hematoxylin and eosinestained sections by a pathologist blinded to the experimental design and groups.
The liver scoring system was adopted from a scoring system developed by Kleiner et al.,6 and included portal inflammation, lobular inflammation, hepatocellular injury, and cholangitis. Each category was graded based on severity and extent of involvement. Jejunal sections were scored for cryptitis, crypt architecture, lamina propria inflammation, and bowel-wall inflammation.
Bile collection and composition measurement In additional cohorts of mice, bile was collected as previously described by Jahnel et al.,26 with modifications. Briefly, mice were anesthetized with 1.5% isoflurane with 1% oxygen after 4 h of fasting. The common bile duct was ligated just below the junction of the cystic duct and hepatic duct. Bile from the gallbladder was aspired and discarded. The gallbladder was then cannulated with 2-French silicone catheter tubing (inner diameter 0.5 mm, outer diameter 0.9 mm, HelixMark standard silicone tubing; Helix Medical, Carpinteria, CA) for bile collection. Bile was collected for 30 min in preweighed tubes and stored at 80 C until bile composition measurements. To examine the bile flow rate in mice, bile specimens were weighed at the end of collection, and the volume of bile was measured gravimetrically by assuming a specific gravity of 1.0. Bile flow rate was normalized to total liver weight. Biliary concentrations of total bile acids, phospholipids, and cholesterol were measured by commercially available kits (total bile acids: BIOQUANT Image Analysis Corporation, Nashville, TN; phospholipids: Wako Chemicals USA, Inc, Richmond, VA; cholesterol: Point Scientific Inc, Canton, MI), and biliary output (nmol/g/min) of each component was calculated. Bile acid/ phospholipid ratio was calculated by dividing the bile acid output by phospholipid output. Portal vein serum was collected after bile collection and stored at 80 C until measurements of serum total bile acids, phospholipids, and cholesterol were performed.
Statistical analysis Data are presented as mean standard error of the mean. Statistical differences between group mean values were assessed by unpaired t-test, analysis of variance, Fisher’s
Table 1 e Primer sequences for mouse hepatobiliary transporters. Target
Primer sense
Primer antisense
GGC TTC TTC AAT AAC TGT CG
GTT TCT ATT TCC CGC TCT CT
Abcg8
CTC ACA CAG GAC ACT GAC TG
GGA AGT CAT TGG AAA TCT GA
Bsep1
CAT TAA CAG CGA AGT CAT CA
CAG TGA TTA CCC ACA ACC TT
Mdr1a
AAG AAA CCA GCA GTC AGT GT
TGG ATA ATA GCA GCC AGA GT
Mdr2
GTT CAG CCT GTT TGA AAG AC
CAC GCT TCC TTC AAA CTT AT
Mrp2
ATG TGG CAT ATT CCA GTT TC
CCA TAG GAG ACG AAG AAC AG
Ntcp
CTC TGC TCT CTT CCG ACT AA
GGT GAC ATT GAG GAT GGT AG
Oatp1
CGC AGT CTT CAT TCT AAT CC
CCT TCT CTG TGA GCT TCA TC
Oatp2
CAT CTG TCT TTG GCT ACC TC
TGG TAG GAG ACA GTC AGT CC
Oatp4
CAC GTT GAC CTG TTT AGT CA
AAT CCA GGT GTA TGA GTT GG
Abcg5
250
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
exact test and ManneWhitney U-test, and P < 0.05 was considered statistically significant.
Table 2 e Extraintestinal organ culture of aerobic and anaerobic bacteria. Surgery (n)
Results
Sham (11) JTR (13)
Creation of SFBL in mice and its impact on general health of mice All groups of surviving mice were generally healthy at the time of euthanasia. At 3 weeks after surgery, the jejunal SFBL appeared considerably dilated in all SFBL mice (Fig. 1B), and MLN enlargement was also evident in all SFBL mice. Gross examination of visceral organs was otherwise normal in all groups of mice, and none of the mice had abscesses or other evidence of peritonitis at the time of necropsy. After surgery, mice in the SFBL group grew slowly, and weight gain plateaued at postoperative day 10, whereas control group mice continued to gain weight until euthanasia at postoperative day 21 (Fig. 1C).
Bacterial overgrowth and translocation in SFBL mice The hallmark of successful SFBL surgery is the overgrowth of bacteria in the blind loop. The creation of an SFBL induced dramatic increases in both aerobic and anaerobic intraluminal bacterial counts (8.48 0.11 log10 CFU/ml and 8.87 0.10 log10 CFU/ml, respectively), compared to JTR mice (3.0 0.34 log10 CFU/ml and 3.78 0.30 log10 CFU/ml, respectively) and Sham mice (2.63 0.16 log10 CFU/ml and 3.13 0.45 log10 CFU/ ml, respectively; P < 0.001) (Fig. 2). In 10 of 10 SFBL mice (100%), MLN bacterial cultures were positive, compared with 1 of 13 JTR mice (7.7%) and 1 of 11 Sham mice (9%) (Table 2). Liver bacterial cultures were positive in 10% of SFBL mice and were negative in all JTR and Sham mice. The bacterial cultures of peritoneal swabs and portal vein blood were negative in all mice. We measured the LPS concentration in portal vein blood because LPS is a major bacteria-derived factor and has been
SFBL (10)
Liver
MLN
Peritoneum
PV blood
0%
9%
0%
0%
0% 10%
7.7% 100%*
0%
0%
0%
0%
PV ¼ portal vein. * P < 0.0001 compared to Sham and JTR mice.
shown to play an important role in hepatic inflammation and cholestasis. The LPS concentration in portal vein serum in SFBL mice, as well as in JTR and Sham mice, was below the detection level of the assay but well detected in the positive control.
Biochemical and histological evidence of liver and jejunal injury in SFBL mice To understand whether SBBO in SFBL mice is associated with evidence of liver injury, plasma aspartate aminotransferase, alanine aminotransferase, and alkaline phosphatase were measured, and liver histology was examined. Plasma alkaline phosphatase was significantly elevated in SFBL mice, compared with JTR and Sham mice (P < 0.05), whereas alanine aminotransferase and aspartate aminotransferase were not significantly different among all three group of mice (Fig. 3). Histologically, SFBL livers were characterized by liver injury with evidence of hepatocyte necrosis, as well as infiltration of neutrophils into the biliary epithelium consistent with cholangitis (Fig. 4A), whereas JTR and Sham livers did not demonstrate such findings. Histologic scoring of liver sections revealed that SFBL mice had significantly greater mean scores for hepatocellular injury and cholangitis, compared with JTR and Sham mice (Fig. 4B). MLN of SFBL mice had focal acute lymphadenitis, characterized by neutrophilic infiltration, whereas MLN of JTR and Sham mice did not demonstrate such findings (data not shown). Histologically, the blind jejunal loop in SFBL mice was characterized by shortened villi, crypt proliferation, areas of focal severe cryptitis and crypt distortion, inflammation in the lamina propria, and transmural inflammation (Fig. 5A). These findings are consistent with disruption of the jejunal barrier in SFBL mice. The jejunum in JTR and Sham mice had normal crypt architecture and no inflammation. SFBL mice exhibited significantly higher mean scores for cryptitis, crypt distortion, lamina propria inflammation, and bowel wall inflammation in the blind jejunal loop, compared with jejunal sections from JTR and Sham mice (Fig. 5B).
Alterations in hepatobiliary transporters induced by creation of SFBL Fig. 2 e Bacterial overgrowth in jejunal blind loop in SFBL mice. At week 3 postoperatively, blind loop content from SFBL mice was isolated and cultured for the presence of aerobic and anaerobic bacteria. Jejunal luminal content from Sham and JTR groups were used as control. Data are represented as mean bacterial concentration (log10 CFU/ ml) ± standard error of the mean. ***P < 0.001.
Because SBBO in SFBL mice resulted in disruption of the jejunal barrier and BT to MLN, we hypothesized that chronic exposure of the liver to microbe-derived factors from the jejunal blind loop may cause changes in hepatobiliary transporter expression. Using real-time polymerase chain reaction to measure expression of hepatobiliary transporters in the
wang et al liver impact of bacterial overgrowth
Fig. 3 e Liver biochemical abnormalities in SFBL mice. Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), and alkaline phosphatase (Alk phos) from Sham, JTR and SFBL mice were measured at the time of euthanasia. *P < 0.05. liver, our data demonstrates that SBBO induces alterations in mRNA expression of multiple hepatobiliary transporters. Expression of 8 out of 10 transporters examined was significantly lower in SFBL mice, compared with Sham controls (Fig. 6). Significant reductions were observed in mRNA expression of four transporter genes associated with biliary canalicular export: Abcg8 (P < 0.0001), Bsep1 (P < 0.0001), Mrp2 (P < 0.01), and Mdr2 (P < 0.05). In addition, we found significant decreases in hepatic expression of all four transporter genes associated with basolateral uptake: Ntcp (P < 0.01), Oatp1 (P < 0.0001), Oatp2 (P < 0.01), and Oatp4 (P < 0.05).
Alterations in bile composition and bile acid/phospholipid ratio in SFBL mice To test whether SBBO/BT-associated reductions in hepatobiliary transporter expression result in altered bile flow and/or composition, we measured these parameters in SFBL mice and compared them with Sham mice. We found that the creation of SFBL caused significant perturbation of the bile composition in the hepatobiliary system, compared to Sham mice, but no differences in the portal vein serum concentrations of total bile acid, phospholipid, or cholesterol were observed between the two groups (Table 3). Although the mean bile flow rate in SFBL mice was 34% lower than the rate in Sham mice, this difference did not reach statistical significance (P ¼ 0.09). Biliary mean total bile acid concentration was significantly higher in SFBL mice, compared with Sham (39.90 2.57 mmol/L versus 23.78 3.85 mmol/L, P < 0.01), but total bile acid output was not different between the two groups. Significant reductions in bile phospholipid and cholesterol output were observed in SFBL mice, which resulted in a significant increase in the bile acid/phospholipid ratio in SFBL mice, suggestive of the formation of toxic bile.
Discussion In the present study, we have recapitulated SBBO in mice by surgically creating a jejunal SFBL, providing a reproducible model that physiologically resembles SBBO in the clinical
251
setting. We have shown inflammation and barrier disruption in the blind loop with MLN translocation, and we have demonstrated histologic evidence of liver injury, as well as a reduction in hepatobiliary transporter expression and alterations in bile composition. Our data clearly demonstrate the effect of SBBO and BT on the liver, thereby providing critical insight into the gut-liver axis. Importantly, our mouse model has the distinct advantage of application to genetically modified animals for further dissection of mechanisms underlying the impact of SBBO and translocation on liver disease pathogenesis. There is increasing evidence that intestinal microbiota or bacteria-derived factors play an important role in liver injury in many chronic liver diseases, including NAFLD,31,33,34 PSC,10,35,36 PBC,7 PNALD,11,12 and CF-associated liver disease.37-40 Studies suggest that PBC is associated with an antibody response to particular alphaproteobacteria (Novosphingobium aromaticivorans)41; however, the pathogenic mechanisms that result in the manifestations of PBC remain obscure. Approximately, 80% of cases of PSC are associated with inflammatory bowel disease, supporting the “leaky gut hypothesis” whereby bacteria and/or bacterial components enter the portal venous system by penetrating the inflamed intestinal mucosa and induce an inflammatory response in the liver leading to fibro-obliterative sclerosis of intrahepatic and extrahepatic bile ducts.42 An increased prevalence of SBBO associated with increased intestinal permeability has been reported in NAFLD patients,15 and in animal studies, dextran sulfate sodium treatment induces colitis and promotes hepatic inflammation and fibrosis in mice with nonalcoholic steatohepatitis.34 The small bowel in infants with short-bowel syndrome is inherently affected by intestinal stasis, and interval dilation of small-bowel loops in PNALD patients can promote bacterial overgrowth, translocation, and recurrent episodes of sepsis, further contributing to liver dysfunction and intrahepatic cholestasis.43 Approximately, 30%-40% of CF patients have been reported to have SBBO, and CF transmembrane receptor knock-out mice demonstrate bacterial overgrowth in the small intestine.38,39 Despite these studies supporting an important role for bacteria in various liver diseases, the pathogenic mechanisms of such an interaction remain poorly understood. A rat model of SBBO was reported over 20 y ago, yet the mechanisms of SBBO-associated liver injury remain unclear.27,29,30 This is largely due to the lack of availability of genetically modified rats. Although a recently reported SFBL mouse model, using a distal ileal blind loop, has shed light on the role of microbiota in ulcerative colitis, there has been no report of the hepatic effects of SBBO/BT in mice.32 Clinically, SBBO is defined as an increase in number and/or alteration in the type of bacteria in the upper gastrointestinal tract.19 In the present study, SFBL mice had 105- to 106-fold higher bacterial colonies in the jejunal blind loop compared with control groups, whereas the number of bacterial colonies in Sham and JTR groups was similar to human studies of the small intestine.17 Overgrowth of bacteria in the blind loop can disrupt the barrier function of the small intestine, leading to translocation of endogenous bacteria or their products through the injured epithelial barrier to MLNs, the systemic circulation, or extraintestinal organs. Indeed, we demonstrated 100% MLN translocation in SFBL mice. The absence of BT to liver or portal venous blood in SFBL mice allows speculation that the hepatic
252
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
Fig. 4 e Histological evidence of hepatobiliary injury in SFBL mice. (A) Hematoxylin and eosin staining of liver sections from Sham, JTR, and SFBL mice at 3 weeks after surgery. SFBL mice showed evidence of hepatocyte necrosis (represented by arrows), neutrophil infiltration (arrow with one asterisk), and cholangitis (arrow with two asterisks). Sham and JTR controls exhibited no evidence of inflammation or injury. (B) Mean liver histology score among Sham, JTR, and SFBL mice. *P < 0.05 compared with Sham and JTR mice. (Color version of figure is available online.)
wang et al liver impact of bacterial overgrowth
253
Fig. 5 e Histological evidence of jejunal wall injury in SFBL mice. (A) Hematoxylin and eosin (H&E) staining of jejunal sections from Sham and SFBL mice. SFBL mice had evidence of shortened villi (represented by dashed lines), crypt proliferation (solid lines), areas of focal, severe cryptitis and crypt distortion (arrow), inflammation in the lamina propria, and transmural inflammation (arterisks). Sham mice had normal crypt architecture and no inflammation. (B) Mean jejunal histology score among Sham, JTR, and SFBL mice. *P < 0.05 compared with Sham and JTR mice. (Color version of figure is available online.)
254
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
Fig. 6 e Alterations in hepatobiliary transporters in SFBL mice. Hepatic mRNA was subjected to real-time polymerase chain reaction analysis for evaluation of hepatobiliary transporter expression. Relative gene expression was calculated according to DDCt method after normalization to b-actin internal control. (A) Transporter genes associated with biliary canalicular export. (B) Transporter genes associated with basolateral uptake. *P < 0.05, **P < 0.01, ***P < 0.001.
Table 3 e Effect of SFBL on bile secretion and composition. Sham (n ¼ 8)
SFBL (n ¼ 8)
1.64 0.19
1.09 0.23
Total bile acid
23.78 3.85
39.90 2.57y
Phospholipids
14.32 1.18
12.44 2.02
2.92 0.51
1.89 0.33
Total bile acid
39.90 7.72
40.45 7.39
Phospholipids
24.47 4.01
12.20 2.44*
Cholesterol
5.16 1.38
1.84 0.53*
Bile acid/phospholipid ratio
1.72 0.27
3.91 0.85*
Total bile acid
19.80 5.87
27.20 3.70
Phospholipids
2.76 0.34
2.09 0.28
Cholesterol
2.99 0.24
2.54 0.25
Bile analysis parameter Bile flow rate (mL/g/min) Biliary concentrations (mmol/L)
Cholesterol Biliary output (nmol/g/min)
Serum concentrations (mmol/L)
*
P < 0.05. y P < 0.01 compared to Sham mice.
effects related to SBBO and BT may not be due to bacteria themselves but rather to other factors triggered by SBBO and BT, such as immune response to LPS or other bacteria-derived signals, such as bacterial DNA, single-stranded DNA, or double-stranded RNA.24,44 Alternatively, efficient hepatic clearance of bacteria in early stages of SBBO and intermittent portal venous bacteremia may be explanations for these findings. Our mouse model provides evidence that SBBO in wild-type C57BL/6 mice causes injury and inflammation in the blind loop, supporting the notion that SBBO results in intestinal barrier dysfunction with potential for hepatic consequences. We hypothesize that the local injury and inflammation of the blind loop increases local intestinal permeability in our model, thereby exposing the liver to bacteria or bacteria-derived factors, such as LPS. This is supported by prior studies that have demonstrated that derangement of the homeostasis between bacteria and the host, as occurs in SBBO, may cause disruption of the intercellular tight junctions and subsequent increase in intestinal permeability leading to translocation of bacteria and bacterial products.45,46 Creation of SFBL in mice induced mild liver injury evidenced by significantly higher alkaline phosphatase levels in the blood and elevated histological scores of hepatocellular injury and cholangitis. Portal and lobular inflammation and fibrosis were not observed in SFBL mice, and serum bilirubin and g-GGT levels were not different between SFBL and Sham mice (data not shown). These data support the notion that gut-derived factors are more likely to act as the “second hit” or environmental trigger in chronic liver diseases, as suggested by Jahnel et al.26 and others.47 The two-hit model for liver diseases such as PSC has been proposed, suggesting that specific mutations (first hit) in an individual can predispose the individual to an inappropriate response to an environmental trigger (second hit). On this basis, we speculate that our mouse SFBL model would be particularly useful in investigating the role of gut-derived factors in predisposed mouse strains, such as in models of PSC, NAFLD, and PNALD. Furthermore, since specific bacterial species are implicated in the pathogenesis of autoimmune liver diseases such as PBC,7 this mouse model may provide the opportunity to dissect its pathogenesis via inoculation of specific bacteria into the blind loop. Evidence suggests that intestinal microbiota affect bile acid metabolism.22 Microbial products, especially LPS, have been shown to cause changes in hepatic transporter expression and bile formation, as well as result in cholestasis.48-50 In the present study, we hypothesized that SBBO and translocation induces changes in bile composition through reduction in hepatobiliary transporter expression. A significant reduction in expression of hepatobiliary transporters involved in biliary canalicular export (Abcg8, Bsep, Mrp2, and Mdr2), as well as basolateral uptake (Ntcp, Oatp1, Oatp2, and Oatp4), was observed in SFBL mice. Our findings are consistent with those of Jahnel et al.,26 in which LPS administration to mice resulted in decreased expression of hepatobiliary transporters, further suggestive of the role of LPS as a major gut-derived factor. Although we did not find a significant increase in LPS levels in portal vein serum at euthanasia, it is quite possible that LPS levels increased intermittently during the postoperative course after SFBL creation and that our failure to detect an increase was related to the fact that we only attempted portal
wang et al liver impact of bacterial overgrowth
vein LPS measurement at a single time point. Dynamic monitoring of LPS levels during the experiment would be needed to further confirm its alteration. In contrast to LPS administration by Jahnel et al.,26 which resulted in increased bile acid output, no change in phospholipids or cholesterol output, and decreased bile acid/phospholipid ratio, we observed no change in bile acid output, decreased phospholipid and cholesterol output and increased bile acid/ phospholipid ratio. These results indicate that hepatic pathophysiology associated with SBBO is complex and likely multifactorial, and it may not be caused by one single bacterial factor such as LPS. Other microbe-associated molecular patterns or pathogen-associated molecular patterns, such as flagellin, peptidoglycan, and microbial metabolites such as short-chain fatty acids, could have significant effects on the alteration of bile composition and hepatobiliary injury.51 From this point of view, our SFBL model would be considered a more physiologic model to study the link between intestinal microbial homeostasis and liver injury. Bile containing a relatively high bile acid/phospholipid ratio is potentially deleterious to the bile duct system due to the inadequate neutralization of the detergent and cytotoxic effects of bile salts by phospholipids.52 Geuken et al.52 reported that cytotoxic bile formation caused by more rapid recovery of bile salt secretion than phospholipid secretion after human liver transplantation correlated with bile duct injury. In the current study, we observed increased bile acid/phospholipid ratio of bile, as well as histologic evidence of cholangitis, in SFBL mice, providing evidence that SBBO alters bile composition with formation of toxic bile and may play a pathogenic role in biliary and hepatocellular injury. In parallel with the observed decreases in cholesterol and phospholipid output, we observed decreases in Abcg8 and Mdr2 mRNA expression in the liver. These results are consistent with the bile alterations found in Mdr2 knockout mice,53 which demonstrated a primary role of Mdr2 for phospholipid excretion in the liver. Surprisingly, we observed an increase in bile acid concentration, but a decrease in Bsep mRNA expression in livers from SFBL mice. Wang et al.54 have previously demonstrated that Bsep knockout mice exhibit a 70% reduction in total bile salt output, but an unexpectedly large amount of tetrahydroxylated bile acid secretion, compared with wild-type mice. Moreover, Smit et al.53 have demonstrated that total bile acid output tended to be higher both in (/þ) and (/) Mdr2 mice than in wild-type control animals. All of the above data suggest that there are likely alternative transport pathways for bile acid excretion, which are independent of Bsep. In addition, studies using mass spectrometry to analyze the primary and secondary bile salts are needed for further assess the bile salt composition in our model.
Conclusions Intestinal microbiota or bacteria-derived factors are likely to play a critical role in the modulation and exacerbation of a number of chronic liver diseases. We have demonstrated that SBBO can initiate histologic evidence of liver injury, thereby providing an important perspective of the gut-liver axis in terms of the effect of bacteria or bacteria-derived products.
255
Using this surgical mouse model, we demonstrated that SBBO induced mild liver injury with decreased hepatobiliary transporter expression and altered bile composition. This model can be used in various genetically modified mice and/or applied to in vivo treatments to further study the interactions between gut-derived factors and liver disease. The applicability of genome- and transcriptome-level approaches to the SFBL mouse model has the potential to greatly advance our understanding of the mechanisms by which SBBO and translocation may impact liver disease pathogenesis.
Acknowledgment The authors thank Dr Michael Trauner and Dr Jo¨rg Jahnel for their advice and assistance regarding the techniques of bile collection and bile composition measurements. Author contributions: Q.W., B.W., and V.S. performed the experiments, collected the data, and drafted the manuscript. L.M. provided substantial technical support and data analysis and conceptual advice regarding data analysis, and critically revised the manuscript. J.T. and J.E.M. helped carry out the experiments and collect the data, and critically revised the manuscript. J.D.N. contributed to the conception, design and coordination of the study and revision of the manuscript. All authors gave the final approval of the submitted version of the manuscript. Funding: This research was supported by internal institutional funding provided by Cincinnati Children’s Hospital Medical Center. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Disclosure The authors report no proprietary or commercial interest in any product mentioned or concept discussed in the article.
references
1. Son G, Kremer M, Hines IN. Contribution of gut bacteria to liver pathobiology. Gastroenterol Res Pract. 2010;2010:1e13. 2. Schnabl B. Linking intestinal homeostasis and liver disease. Curr Opin Gastroenterol. 2013;29:264e270. 3. Seki E, Schnabl B. Role of innate immunity and the microbiota in liver fibrosis: crosstalk between the liver and gut. J Physiol. 2012;590(Pt 3):447e458. 4. Machado MV, Cortez-Pinto H. Gut microbiota and nonalcoholic fatty liver disease. Ann Hepatol. 2012;11:440e449. 5. Chassaing B, Etienne-Mesmin L, Gewirtz AT. Microbiota-liver axis in hepatic disease. Hepatology. 2014;59:328e339. 6. Kleiner DE, Brunt EM, Van Natta M, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology. 2005;41:1313e1321. 7. Mattner J, Savage PB, Leung P, et al. Liver autoimmunity triggered by microbial activation of natural killer T cells. Cell Host Microbe. 2008;3:304e315. 8. Fouts DE, Torralba M, Nelson KE, Brenner DA, Schnabl B. Bacterial translocation and changes in the intestinal microbiome in mouse models of liver disease. J Hepatol. 2012;56:1283e1292.
256
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 2 4 6 e2 5 6
9. Abu-Shanab A, Quigley EM. The role of the gut microbiota in nonalcoholic fatty liver disease. Nat Rev Gastroenterol Hepatol. 2010;7:691e701. 10. Riordan SM, Williams R. The intestinal flora and bacterial infection in cirrhosis. J Hepatol. 2006;45:744e757. 11. Pollheimer MJ, Halilbasic E, Fickert P, Trauner M. Pathogenesis of primary sclerosing cholangitis. Best Pract Res Clin Gastroenterol. 2011;25:727e739. 12. Kelly DA. Intestinal failure-associated liver disease: what do we know today? Gastroenterology. 2006;130(2 Suppl 1):S70eS77. 13. Kalambokis GN, Tsianos EV. Rifaximin reduces endotoxemia and improves liver function and disease severity in patients with decompensated cirrhosis. Hepatology. 2012;55:655e656. 14. Schnabl B, Brenner DA. Interactions between the intestinal microbiome and liver diseases. Gastroenterology. 2014;146:1513e1524. 15. Compare D, Coccoli P, Rocco A, et al. Guteliver axis: the impact of gut microbiota on non alcoholic fatty liver disease. Nutr Metab Cardiovasc Dis. 2012;22:471e476. 16. Cesaro C, Tiso A, Del Prete A, et al. Gut microbiota and probiotics in chronic liver diseases. Dig Liver Dis. 2011;43:431e438. 17. Gorbach SL, Bengt E. Gustafsson memorial lecture. Function of the normal human microflora. Scand J Infect Dis Suppl. 1986;49:17e30. 18. Berg RD, Garlington AW. Translocation of certain indigenous bacteria from the gastrointestinal tract to the mesenteric lymph nodes and other organs in a gnotobiotic mouse model. Infect Immun. 1979;23:403e411. 19. Bures J, Cyrany J, Kohoutova D, et al. Small intestinal bacterial overgrowth syndrome. World J Gastroenterol. 2010;16:2978e2990. 20. Yan AW, Fouts DE, Brandl J, et al. Enteric dysbiosis associated with a mouse model of alcoholic liver disease. Hepatology. 2011;53:96e105. 21. Pinzone MR, Celesia BM, Di Rosa M, Cacopardo B, Nunnari G. Microbial translocation in chronic liver diseases. Int J Microbiol. 2012;2012:694629. 22. de Aguiar Vallim TQ, Tarling EJ, Edwards PA. Pleiotropic roles of bile acids in metabolism. Cell Metab. 2013;17:657e669. 23. Swann JR, Want EJ, Geier FM, et al. Systemic gut microbial modulation of bile acid metabolism in host tissue compartments. Proc Natl Acad Sci U S A. 2011;108(Suppl 1):4523e4530. 24. Szabo G, Bala S, Petrasek J, Gattu A. Gut-liver axis and sensing microbes. Dig Dis. 2010;28:737e744. 25. Trauner M, Fickert P, Stauber RE. Inflammation-induced cholestasis. J Gastroenterol Hepatol. 1999;14:946e959. 26. Jahnel J, Fickert P, Langner C, et al. Impact of experimental colitis on hepatobiliary transporter expression and bile duct injury in mice. Liver Int. 2009;29:1316e1325. 27. Lichtman SN, Sartor RB. Hepatobiliary injury associated with experimental small-bowel bacterial overgrowth in rats. Immunol Res. 1991;10:528e531. 28. Lichtman SN, Keku J, Schwab JH, Sartor RB. Evidence for peptidoglycan absorption in rats with experimental small bowel bacterial overgrowth. Infect Immun. 1991;59:555e562. 29. Lichtman SN, Keku J, Schwab JH, Sartor RB. Hepatic injury associated with small bowel bacterial overgrowth in rats is prevented by metronidazole and tetracycline. Gastroenterology. 1991;100:513e519. 30. Lichtman SN, Keku J, Clark RL, Schwab JH, Sartor RB. Biliary tract disease in rats with experimental small bowel bacterial overgrowth. Hepatology. 1991;13:766e772. 31. Shanab AA, Scully P, Crosbie O, et al. Small intestinal bacterial overgrowth in nonalcoholic steatohepatitis: association with toll-like receptor 4 expression and plasma levels of interleukin 8. Dig Dis Sci. 2011;56:1524e1534.
32. Ward MA, Pierre JF, Leal RF, et al. Insights into the pathogenesis of ulcerative colitis from a murine model of stasis-induced dysbiosis, colonic metaplasia, and genetic susceptibility. Am J Physiol Gastrointest Liver Physiol. 2016;310:G973eG988. 33. Miele L, Valenza V, La Torre G, et al. Increased intestinal permeability and tight junction alterations in nonalcoholic fatty liver disease. Hepatology. 2009;49:1877e1887. 34. Gabele E, Dostert K, Hofmann C, et al. DSS induced colitis increases portal LPS levels and enhances hepatic inflammation and fibrogenesis in experimental NASH. J Hepatol. 2011;55:1391e1399. 35. Schrumpf E, Boberg KM. Epidemiology of primary sclerosing cholangitis. Best Pract Res Clin Gastroenterol. 2001;15:553e562. 36. O’Mahony CA, Vierling JM. Etiopathogenesis of primary sclerosing cholangitis. Semin Liver Dis. 2006;26:3e21. 37. Herrmann U, Dockter G, Lammert F. Cystic fibrosis-associated liver disease. Best Pract Res Clin Gastroenterol. 2010;24:585e592. 38. Lewindon PJ, Robb TA, Moore DJ, Davidson GP, Martin AJ. Bowel dysfunction in cystic fibrosis: importance of breath testing. J Paediatr Child Health. 1998;34:79e82. 39. O’Brien S, Mulcahy H, Fenlon H, et al. Intestinal bile acid malabsorption in cystic fibrosis. Gut. 1993;34:1137e1141. 40. Norkina O, Kaur S, Ziemer D, De Lisle RC. Inflammation of the cystic fibrosis mouse small intestine. Am J Physiol Gastrointest Liver Physiol. 2004;286:G1032eG1041. 41. Mattner J. Genetic susceptibility to autoimmune liver disease. World J Hepatol. 2011;3:1e7. 42. Llorente C, Schnabl B. The gut microbiota and liver disease. Cell Mol Gastroenterol Hepatol. 2015;1:275e284. 43. Beath SV, Davies P, Papadopoulou A, et al. Parenteral nutrition-related cholestasis in postsurgical neonates: multivariate analysis of risk factors. J Pediatr Surg. 1996;31:604e606. 44. Kesar V, Odin JA. Toll-like receptors and liver disease. Liver Int. 2014;34:184e196. 45. Ferolla SM, Armiliato GN, Couto CA, Ferrari TC. The role of intestinal bacteria overgrowth in obesity-related nonalcoholic fatty liver disease. Nutrients. 2014;6:5583e5599. 46. Vajro P, Paolella G, Fasano A. Microbiota and gut-liver axis: their influences on obesity and obesity-related liver disease. J Pediatr Gastroenterol Nutr. 2013;56:461e468. 47. Mayer L. Redefining autoimmunity. Gastroenterology. 2003;125:1574. 48. Kosters A, Karpen SJ. The role of inflammation in cholestasis: clinical and basic aspects. Semin Liver Dis. 2010;30:186e194. 49. Bolder U, Ton-Nu HT, Schteingart CD, Frick E, Hofmann AF. Hepatocyte transport of bile acids and organic anions in endotoxemic rats: impaired uptake and secretion. Gastroenterology. 1997;112:214e225. 50. Moseley RH, Wang W, Takeda H, et al. Effect of endotoxin on bile acid transport in rat liver: a potential model for sepsisassociated cholestasis. Am J Physiol. 1996;271(1 Pt 1):G137eG146. 51. Ha CW, Lam YY, Holmes AJ. Mechanistic links between gut microbial community dynamics, microbial functions and metabolic health. World J Gastroenterol. 2014;20:16498e16517. 52. Geuken E, Visser D, Kuipers F, et al. Rapid increase of bile salt secretion is associated with bile duct injury after human liver transplantation. J Hepatol. 2004;41:1017e1025. 53. Smit JJ, Schinkel AH, Oude Elferink RP, et al. Homozygous disruption of the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease. Cell. 1993;75:451e462. 54. Wang R, Salem M, Yousef IM, et al. Targeted inactivation of sister of P-glycoprotein gene (spgp) in mice results in nonprogressive but persistent intrahepatic cholestasis. Proc Natl Acad Sci U S A. 2001;98:2011e2016.