Tian-Yu Lin ,Yi-Fan Zhang ,Yang Wang ,Yun Liu ,Jun Xu ,Yu-Lan Liu ,?

a Department of Gastroenterology, Peking University People’s Hospital, Beijing 10 0 044, China

b Clinical Center of Immune-Mediated Digestive Diseases, Peking University People’s Hospital, Beijing 10 0 044, China

c Central Laboratory & Institute of Clinical Molecular Biology, Peking University People’s Hospital, Beijing 10 0 044, China

Keywords: Nonalcoholic fatty liver disease Acute pancreatitis Bacterial translocation Cholesterol metabolism

ABSTRACT Background: Nonalcoholic fatty liver disease (NAFLD) is an independent risk factor for severe acute pancreatitis (AP). The underlying mechanism remains unclear. We sought to determine how bacterial translocation and cholesterol metabolism in the liver and pancreas affect the severity of AP in NAFLD mice.Methods: C57BL/6N mice were fed on a high-fat diet (HFD) to generate the NAFLD model,and mice in the control group were provided with a normal diet (ND). After being anesthetized with ketamine/xylazine,mice got a retrograde infusion of taurocholic acid sodium into the pancreatic duct to induce AP,and sham operation (SO) was used as control. Serum amylase and Schmidt’s pathological score system were used to evaluate AP severity. Bacterial loads,total cholesterol level,and cholesterol metabolic-associated molecules [low-density lipoprotein receptor (LDLR) and ATP-binding cassette transporter A1 (ABCA1)]were analyzed in the liver and pancreas.Results: Compared with the ND-AP group,mice in the HFD-AP group had severer pancreatitis,manifested with higher serum amylase levels and higher AP pathologic scores,especially the inflammation and hemorrhage scores. Compared with the HFD-SO group and ND-AP group,bacterial loads in the liver and pancreas were significantly higher in the HFD-AP group. Mice in the HFD-AP group showed a decreased LDLR expression and an increased ABCA1 expression in the pancreas,although there was no significant difference in pancreas total cholesterol between the HFD-AP group and the ND-AP group.Conclusions: NAFLD aggravates AP via increasing bacterial translocation in the liver and pancreas and affecting pancreas cholesterol metabolism in mice.

Introduction

Acute pancreatitis (AP) is a common clinical disease. Severe acute pancreatitis (SAP) comprises about 20%-30% of the patients,with consistent organ dysfunction,high complication rate,and morbidity [ 1,2 ]. Early recognition of SAP,intravenous fluid administration,intensive care unit (ICU) management and nutritional support are important for a better outcome [3] . A previous study showed that fatty liver disease is an independent risk factor for moderate and severe AP. A meta-analysis showed that fatty liver disease could worsen the outcomes of AP [4] . Local complications,organ failure,and morbidity increase with the severity of fatty liver in AP patients [ 5,6 ]. However,the underlying mechanism remains unclear. Wang et al. showed that hepatic steatosis could cause depression of serum alpha-1-antitrypsin levels,down-regulation of liver PPARαsignaling pathway and fatty acid degradation pathway,which may be associated with SAP [ 7,8 ]. However,few studies focused on the changes of the liver and pancreas at the same time when studying how fatty liver disease aggravated AP.

Recently,an international panel of experts proposed the nomenclature of metabolic dysfunction associated fatty liver disease (MAFLD) instead of nonalcoholic fatty liver disease (NAFLD),which means fatty liver diseases are associated with metabolic dysfunction [9] . We usually use a high-fat diet (HFD) model to study fatty liver disease,and a high-fat diet could have an influence on multiple systems besides the liver. An HFD can damage the intestinal mucosal barrier. A previous study showed that,after a high-sucrose and HFD for 48 weeks,rats got a metabolic syndrome,fatty liver,and a reduction of small intestine occluding protein expression and intestinal endotoxemia [10] . And it has been previously reported that bacterial translocation is associated with the pathology and progression of AP [11] . Although the presence of bacteria in the livers of mice has been proven through 16S ribosomal RNA gene detection [12],we know little about the alteration of bacteria load in both pancreases and livers in SAP of NAFLD animals. NAFLD is highly associated with pancreatic fat accumulation,and about 50%-80% of nonalcoholic steatohepatitis (NASH) patients have fatty pancreas [13] . Matsuda et al.showed that triglyceride accumulated in pancreas acinar cells in obese rats [14] . Recently,some population-based studies found that serum total cholesterol level was associated with SAP [15] .However,alteration of pancreas cholesterol metabolism in SAP of NAFLD animals has not been investigated.

Although mice and rats are both used to establish the NAFLD [16] and AP models [17],we found that mice established a stable NAFLD model in our feeding conditions and showed typical pancreatitis 24 h after retrograde infusion of taurocholic acid sodium into the pancreatic duct. In this study,we induced the combined AP and NAFLD model in mice to investigate the effect of fatty liver disease on AP. We are interested in how bacterial translocation and cholesterol metabolism are changed in the NAFLD mice model with AP and investigated bacterial loads and cholesterol metabolism in both the liver and pancreas. We hypothesized that the changes of bacteria and cholesterol in the liver and pancreas could explain how fatty liver disease aggravated AP in mice.

Methods

Animal model

Male C57BL/6N mice of 6 weeks old (weighing 20.97-23.18 g)were purchased from Beijing Vital River Laboratory Animal Technology Co. Ltd. (Beijing,China),raised in a specific pathogen-free(SPF) facility with a temperature of 24 °C,humidity of 50% and a 12-h light-dark cycle,and were fedadlibitum. After one week of adaptive feeding,12 mice were fed with an HFD (45% kcal fat;Mediscience Ltd.,Yangzhou,China),while 12 mice were fed with a normal diet (ND). The mice were weighed every two weeks until the body weight achieved 40 g. Six mice of each diet group were randomly selected to establish an experimental AP,and 6 mice received sham operation.

AP was surgically induced in the SPF laboratory as described previously [18] . Briefly,first,a stock solution of ketamine/xylazine(Animal Laboratory of Peking University People’s Hospital,Beijing,China) was prepared by adding 0.2 mL of xylazine (50 mg/mL)to 2 mL of ketamine (50 mg/mL) and diluted with saline to 20 mL. Intraperitoneal injection of 20μL/g of the ketamine (100 mg/kg)/xylazine (10 mg/kg) mix was used for effective anesthetization for about 30 minutes which was enough for the AP surgery.The mouse was placed on its back,the abdomen was cleaned and disinfected,and a middle incision was made for about 2-3 cm. The stomach was pulled out and the white papillary was found in descending duodenum. Then the duodenum wall was punctured opposite the papillary with a thin needle of 0.3 mm in diameter. Taurocholic acid sodium salt hydrate (Sigma-Aldrich,Shanghai,China)was dissolved in saline to the concentration of 5% and one part of methyl blue solution (Jichuan Pharmaceutical Group Co.,Ltd.,Taixing,China) was added to 99 parts of bile acid solution on the day of surgery. TwoμL of the bile acid solution per gram body weight of the mouse was injected into the biliopancreatic duct at a constant speed and a microvascular clamp was placed on the bile duct closed to the hilum of the liver at the same time. When the infusion finished,the microvascular clamp and needle from the duct were gently removed and the abdominal wall was closed with a suture. The mouse was placed on a heated pad (37 °C) until recovered. With practice,the surgery could be finished within 15 minutes with a>95% success rate. For the sham operation (SO) group,the pancreas was touched several times at the descending duodenum as described previously [19] . The mice were sacrificed 24 h after induction of AP. Serum,pancreas,liver,mesenteric adipose tissue (MAT),and epididymal adipose tissue (EAT) were collected for analysis. All protocols were approved by the Ethics Committee of Peking University People’s Hospital (2020PHE074).

Serum biochemical measurement

The serum concentrations of amylase (AMY),alanine aminotransferase (ALT),aspartate aminotransferase (AST),total cholesterol (TC),and triglyceride (TG) were determined using an automatic biochemical detector (Labospect 008; Hitachi Ltd.,Ibarakiken,Japan).

Histopathological scoring system

Samples from the liver and pancreas head were fixed in 4%paraformaldehyde and embedded in paraffin. The slices were stained with hematoxylin-eosin (H&E) and observed by light microscopy to analyze pathology. Fatty liver disease activity was evaluated by two independent pathologists using a general NAFLD scoring system for rodents which comprises micro- and macro-vesicular steatosis,inflammation,and hypertrophy. Three fields were randomly selected,and the slices were evaluated at 100 × magnification [20] . Two independent pathologists evaluated AP severity using Schmidt’s pathological score system,including edema,acinar necrosis,hemorrhage,and fat necrosis,inflammation,and perivascular infiltration. Three fields were randomly selected. Pancreas edema and hemorrhage were evaluated at 40 × magnification,while acinar necrosis and inflammation were evaluated at 100 × magnification [21] .

Bacterial culture and PCR

As Balmer et al. reported,fresh pancreas and liver tissues (30-50 mg each) were removed aseptically and homogenized in 0.5 mL germ-free phosphate-buffer saline [22] . The homogenates were then planted on Luria broth plates (Solarbio Science & Technology Co.,Ltd.,Beijing,China) and cultured at 37 °C for 24 h. The colonyforming units (CFUs) were then counted.

Pancreas and liver tissues (50-100 mg each) were removed aseptically. Total bacterial DNA in the pancreas and liver was extracted using the TIANamp Genomic DNA Kit (TIANGEN BIOTECH Co.,Ltd.,Beijing,China). Real-time PCR was performed using the StepOne Plus Real-Time PCR System (Applied Biosystems,Waltham,MA,USA) and the SYBR Premix (Toyobo,Kita-Ku,Osaka,Japan).Primers for total bacteria detection,forward (5’ to 3’) GTG STG CAY GGY TGT CGT CA; reverse (5’ to 3’) ACG TCR TCC MCA CCT TCC TC [23] . All reactions were run in duplicate.

Cholesterol metabolism analysis

Total cholesterol in liver and pancreas tissues was detected with a tissue total cholesterol assay kit (E1015,Applygen Technologies Inc.,Beijing,China). The liver and pancreas were stored at -80?C for RNA extraction. Total RNA was extracted using TRIzol (Ambion,Carlsbad,CA,USA),and cDNAs were synthesized using the Revert Aid First Strand cDNA Synthesis Kit (Thermo,Vilnius,Lithuania).Real-time PCR was performed using the StepOne Plus Real-TimePCR System (Applied Biosystems) and the SYBR Premix (Toyobo).Primers list:β-actin,forward (5’ to 3’): TGT CCA CCT TCC AGC AGA TGT; reverse (5’ to 3’): AGC TCA GTA ACA GTC CGC CTA G. Lowdensity lipoprotein receptor (LDLR),forward (5’ to 3’): GGG CAA TCG GAA AAC CAT; reverse (5’ to 3’): CAT CTG TCC AAT ACA CTT TGT CCT. ATP-binding cassette trsnaporter A1 (ABCA1),forward (5’to 3’): CGT TTC CGG GAA GTG TCC TA; reverse (5’ to 3’): CTA GAG ATG ACA AGG AGG ATG GA.

Statistical analysis

Data were presented as mean ± standard error of the mean(SEM). The two-tailed Student’st-test or one-way analysis of variance (ANOVA) was performed to compare values between the groups. Nonparametric test was used when two groups had unequal variances. Spearman’s correlation analysis was performed between the NAFLD score and the AP pathological score. SPSS version 20.0 (IBM,Amonk,NY,USA) was used to conduct statistical analyses. GraphPad Prism 8.0 (GraphPad Software Inc.,La Jolla,CA,USA) was used to plot the graphs. APvalue<0.05 was considered statistically significant.

Results

HFD-induced fatty liver and metabolic dysfunction in mice

The mouse’s body weight increased with time. Since two weeks after the HFD,the HFD group was heavier than the ND group( Fig. 1 ). Compared with the ND-SO group,MAT weight and EAT weight were significantly higher in the HFD-SO group ( Table 1,P<0.05). Serum TC was higher in the HFD-SO group compared with the ND-SO group,while serum TG showed no significant difference between the two groups ( Table 1 ). Compared with the NDAP group,MAT weight and EAT weight were significantly higher in the HFD-AP group ( Table 1,P<0.05). Serum TC was more elevated in the HFD-AP group compared with the ND-AP group,while serum TG showed no significant difference between the two groups ( Table 1 ).

Fig. 1. Body weight between the ND and HFD groups. ND: normal diet; HFD: highfat diet.

Table 1Detailed data about basic information of mice in each group.

The liver in the HFD group was yellower than that in the ND group ( Fig. 2 A). The primary pathological manifestation of the liver was steatosis ( Fig. 2 B). The NAFLD score was higher in the HFD-AP group compared with the ND-AP group ( Table 1,P<0.05). Serum ALT and AST showed no significant difference between the HFD and ND groups ( Table 1 ).

Fig. 2. NAFLD aggravates the severity of acute pancreatitis in mice. A: Representative gross appearance of liver and pancreas in each group; B: representative hematoxylineosin (H & E) staining of mouse liver (original magnification ×200) and pancreas (original magnification ×100) in each group; C: the relationship between NAFLD score and AP pathological score was assessed by Spearman correlation analysis ( r = 0.71,P = 0.014). Data were presented as mean ± standard error of the mean. ND: normal diet;HFD: high-fat diet; AP: acute pancreatitis; SO: sham operation; NAFLD: nonalcoholic fatty liver disease.

NAFLD aggravated AP severity

In the ND-SO group,the pancreas was light yellow lying between descending duodenum and spleen. After AP induction,the pancreas became swollen and white; some may have a hemorrhage ( Fig. 2 A). The pathological manifestations of AP were edema,hemorrhage,necrosis,and inflammation ( Fig. 2 B,Table 2 ). Serum amylase was elevated after AP induction ( Table 2 ).

Table 2The severity of AP in each group.

The fatty liver was thought to be a risk factor for SAP. Compared with the ND-AP group,mice in the HFD-AP group had more severe pancreatitis,manifested with elevated serum amylase level,AP pathological score,and the inflammation and hemorrhage scores were higher as well ( Table 2,P<0.05). NAFLD score was correlated with AP histopathologic score (r= 0.71,P<0.05,Fig. 2 C).

Bacterial was translocated to the liver and pancreas of HFD-AP mice

To assess the role of bacterial translocation in SAP of NAFLD mice,we evaluated bacterial load in the liver and pancreas. Liver bacterial load was not significantly different between the ND-AP group and ND-SO group. The liver bacterial load of the HFD-SO group showed no significant difference from that of the ND-SO group. While compared with the HFD-SO group,the liver bacterial load was higher in the HFD-AP group,manifested with elevated CFUs of tissue bacterial culture (P<0.05) and total bacterial DNA(P<0.05) in the liver ( Fig. 3 A-C).

Fig. 3. Bacterial translocation increased in AP mice combined with NAFLD. ( A ) Representative appearance of tissue bacterial culture of the liver in each group. CFUs of tissue bacterial culture ( B ) and total bacterial DNA ( C ) in the liver. ( D ) The representative appearance of tissue bacterial culture of the pancreas in each group. CFUs of tissue bacterial culture ( E ) and total bacterial DNA ( F ) in the pancreas. Data were presented as mean ± standard error of the mean. ?: P < 0.05. ND: normal diet; HFD: high-fat diet; AP: acute pancreatitis; SO: sham operation; CFU: colony-forming units.

Compared with the ND-SO group,the pancreas bacterial load was higher in the ND-AP group,manifested with elevated CFUs of tissue bacterial culture (P<0.05) and total bacterial DNA (P<0.05) in the pancreas. The pancreas bacterial load of the HFDSO group showed no significant difference from the ND-SO group.Compared with the ND-AP group,the pancreas bacterial load was higher in the HFD-AP group,manifested with elevated CFUs of tissue bacterial culture (P<0.05) and total bacterial DNA (P<0.05)in the pancreas ( Fig. 3 D-F).

Cholesterol metabolic dysregulation in the pancreas of HFD-AP mice

We found that serum TC level was higher in the HFD-AP group.To assess the role of cholesterol metabolism in SAP of NAFLD mice,TC level and cholesterol metabolic-associated molecules levels in the liver and pancreas were determined. Compared with the NDSO group,TC in the liver was higher in the HFD-SO group,with decreased LDLR expression and increased ABCA1 expression (P<0.05,Fig. 4 A-C).

Fig. 4. Cholesterol metabolic dysregulation in AP mice combined with NAFLD. (A) Total cholesterol of liver in each group. LDLR expression ( B ) and ABCA1 expression ( C ) of the liver. ( D ) Total cholesterol in the pancreas in each group. LDLR expression ( E ) and ABCA1 expression ( F ) of the pancreas. Data are presented as mean ± standard error of the mean. ?: P < 0.05. LDLR: low-density lipoprotein receptor; ABCA1: ATP-binding cassette transporter A1; AP: acute pancreatitis; NAFLD: nonalcoholic fatty liver disease.

Compared with the ND-SO group,pancreas TC in the ND-AP group was higher (P<0.05),while pancreas LDLR and ABCA1 expression showed no difference between the two groups. Compared with the ND-SO group,total pancreas cholesterol in the HFD-SO group was higher (P<0.05),while pancreas LDLR and ABCA1 expression showed no difference between the two groups. TC in the pancreas was not significantly different between the HFD-AP group and ND-AP group. In contrast,pancreas LDLR expression was decreased (P<0.05),and ABCA1 expression was increased (P<0.05) in the HFD-AP group compared to those in the ND-AP group( Fig. 4 D-F).

Discussion

A previous study has shown that rats and mice with fatty liver had more severe AP than those with a healthy liver [24] . In this study,we used the mice model to investigate how fatty liver diseases aggravate AP,especially on the organ level of the liver and pancreas. Our results showed that the HFD-AP group had severer pancreatitis,consistent with the previous study of different AP mice models [24] .

We found that pancreas bacterial load was increased after AP induction,and the pancreas bacterial load was higher when combined with fatty liver. While AP or liver steatosis did not increase the bacterial burden in the liver of mice,fatty liver in AP mice showed increased bacterial load. We concluded that the HFD-AP group had increased bacterial loads in the liver and pancreas,while mice with normal liver had lower bacterial loads in both liver and pancreas. Emerging evidence showed that the pancreas is not sterile [25] . Previous studies via tissue bacterial culture have shown that AP could increase the pancreas bacteria load in different rodent AP models [ 26,27 ]. The liver serves as a gateway between the abdominal organ and the systemic circulatory system. The liver receives pathogenic factors from pancreatitis [28] . At the same time,the liver participates in modulating pro-inflammatory cytokine distribution to distant organs in AP [29] . The liver acts as a firewall to sequestrate bacteria from the portal vein and hepatic artery. In addition,several multiple studies have demonstrated that neutrophils or monocytes support Kupffer cells in the liver on bacterial clearance [ 22,30 ]. However,the defense against microorganisms of the liver becomes impaired when liver cirrhosis is present [ 22,31 ]. We proposed that healthy liver takes part in the clearance of bacteria while fatty liver impaired the clearance of bacteria in AP mice,and NAFLD aggravates AP via increasing bacterial translocation in the liver and pancreas. On the one hand,there are more bacteria translocated to the pancreas and liver,and on the other hand,the fatty liver had a declined ability on bacterial clearance. All of these changes lead to the aggravation of AP. There might be a vicious circle between bacterial translocation and AP,in which pancreas inflammation may induce bacterial translocation from the intestine,and bacteria translocated to the pancreas could aggravate pancreatitis.

In this study,we used tissue bacterial culture and total bacterial DNA detection to evaluate the bacterial load in the liver and pancreas. Although culture and sequencing display their own bias [32],both of them are widely used in tissue pathogen detection [33] . Therefore,it is helpful to apply multiple techniques for bacterial detection in tissue. We mainly focus on the bacterial load in the liver and pancreas. The previous study showed that fatty liver disease and SAP are systemic inflammatory diseases in which intestinal mucosal permeability increases and bacterial translocation occurs [34] . Therefore,the role of the intestinal mucosal barrier in HFD-AP rodents deserves further study.

In our study,serum TC was significantly higher in the HFD group. We further investigated the role of cholesterol metabolism on SAP in NAFLD mice. Cholesterol is an important component in cells. Cell cholesterol homeostasis is regulated by cholesterol influx and efflux,controlled by feedback mechanisms. LDLR is an important molecule that mediates cholesterol influx. In comparison,ABCA1 is an important molecule that mediates cholesterol efflux [35] . Our results showed that the HFD group had an increased liver TC level,with a change of cholesterol metabolicassociated molecules in the liver. It is consistent with the previous studies that liver cholesterol accumulation could down-regulate LDLR-mediated cholesterol influx and up-regulate ABCA1-mediated cholesterol efflux [ 36,37 ].

Our study found that HFD group had a higher pancreas total cholesterol level. AP may increase the pancreas cholesterol level due to the lipolysis of peripancreatic adipocytes [38],and NAFLD mice also accumulate cholesterol in the pancreas. However,our results showed that neither AP nor fatty pancreas could affect LDLR or ABCA1 expression; while HFD-AP mice showed a significantly changed cholesterol metabolism of the pancreas with a downregulation of LDLR expression and an up-regulation of ABCA1 expression. We therefore concluded that cholesterol metabolism might be involved in the aggravation of AP in NAFLD.

Cholesterol may be involved in the inflammatory response.A previous study showed that hypercholesterolemia promoted the transformation of macrophages into foam cells,leading to macrophages apoptosis and a reduction of inflammatory macrophage responses [39] . Increased plasma membrane cholesterol encourages the recognition of lipopolysaccharide by toll-like receptors and enhances inflammatory response [40] . Hypercholesterolemia can cause oxidative stress,mitochondrial dysfunction,and the activation of pro-apoptotic factors,thus leading toβ-cell apoptosis [41] . The present study preliminarily studied the cholesterol metabolism in the HFD-AP mice model. The relationship between pancreatic cholesterol metabolism and AP needs further investigation.

In conclusion,NAFLD aggravates AP via increasing bacterial translocation in the liver and pancreas and affecting pancreas cholesterol metabolism in mice.

Acknowledgments

None.

CRediT authorship contribution statement

Tian-YuLin:Conceptualization,Data curation,Formal analysis,Investigation,Writing – original draft.Yi-FanZhang:Methodology,Validation,Visualization.YangWang:Methodology,Validation.YunLiu:Validation.JunXu:Data curation,Formal analysis.Yu-LanLiu:Conceptualization,Funding acquisition,Supervision,Writing – review & editing.

Funding

This study was supported by grants from the National Natural Science Foundation of China (82070539 and 81873549) and the Youth Innovation Foundation of Aerospace Center Hospital(2021QN02).

Ethical approval

This study was approved by the Ethics Committee of Peking University People’s Hospital (2020PHE074).

Competing interest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.