- Email: [email protected]
Altered diversity and composition of gut microbiota in Chinese patients with chronic pancreatitis
Journal Pre-proof Altered diversity and composition of gut microbiota in Chinese patients with chronic pancreatitis Chun-Hua Zhou, Yu-Ting Meng, Jia-Jia Xu, Xue Fang, Jiu-Long Zhao, Wei Zhou, Jianhua Zhao, Ji-Chen Han, Ling Zhang, Kai-Xuan Wang, Liang-Hao Hu, Zhuan Liao, Wen-Bin Zou, Zhao-Shen Li, Duo-Wu Zou PII:
S1424-3903(19)30794-X
DOI:
https://doi.org/10.1016/j.pan.2019.11.013
Reference:
PAN 1123
To appear in:
Pancreatology
Received Date: 9 June 2019 Revised Date:
14 November 2019
Accepted Date: 23 November 2019
Please cite this article as: Zhou C-H, Meng Y-T, Xu J-J, Fang X, Zhao J-L, Zhou W, Zhao J, Han JC, Zhang L, Wang K-X, Hu L-H, Liao Z, Zou W-B, Li Z-S, Zou D-W, Altered diversity and composition of gut microbiota in Chinese patients with chronic pancreatitis, Pancreatology (2019), doi: https:// doi.org/10.1016/j.pan.2019.11.013. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier B.V. on behalf of IAP and EPC.
1
Altered Diversity and Composition of Gut Microbiota in Chinese Patients with Chronic Pancreatitis
Chun-Hua Zhoua,b,d,#, Yu-Ting Menga,b,#, Jia-Jia Xua,b,#, Xue Fanga,b, Jiu-Long Zhaoa,b, Wei Zhoua,b, Jianhua Zhaoc, Ji-Chen Hanc, Ling Zhangd, Kai-Xuan Wanga,b, Liang-Hao Hua,b, Zhuan-Liaoa,b, Wen-Bin Zoua,b,*, Zhao-Shen Lia,b,*, Duo-Wu Zou a,b,d*
a
Department of Gastroenterology, Changhai Hospital, The Second Military Medical University,
Shanghai, China b
c
Shanghai Institute of Pancreatic Diseases, Shanghai, China
Shanghai Majorbio Bio-pharm Technology Co., Ltd.
d
Department of Gastroenterology, Ruijin Hospital Affiliated to Shanghai Jiao Tong University School
of Medicine, Shanghai, China
#
These authors contributed equally to this work.
Short title: Gut Microbiota Alterations in Chronic pancreatitis
#
Corresponding authors:
Duo-Wu Zou Department of Gastroenterology, Changhai Hospital, The Second Military Medical University, Shanghai, China No.168, Changhai Road,Yangpu District, Shanghai,China, 200433 Department of Gastroenterology, Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China
2
No.197,Rui Jin Er Road,Shanghai,China,200025 Tel.: +86 21 32162770 Fax: +86 21 55621735 Email address: [email protected]
Zhao-Shen Li Department of Gastroenterology, Changhai Hospital, The Second Military Medical University, Shanghai, China No.168, Changhai Road,Yangpu District, Shanghai,China, 200433 Tel.: +86 21 32162770 Fax: +86 21 55621735 Email address: [email protected] Wen-Bin Zou Department of Gastroenterology, Changhai Hospital, The Second Military Medical University, Shanghai, China No.168, Changhai Road,Yangpu District, Shanghai,China,200433 Tel.: +86 21 32162770 Fax: +86 21 55621735 Email address: [email protected]
3
Abstract Background/Objectives: Gut microbiota alterations in chronic pancreatitis (CP) are seldomly described systematically. It is unknown whether pancreatic exocrine insufficiency (PEI) and different etiologies in patients with CP are associated with gut microbiota dysbiosis. Methods: The fecal microbiota of 69 healthy controls (HCs) and 71 patients with CP were compared to investigate gut microbiome alterations in CP and the relationship among gut microbiome dysbiosis, PEI and different etiologies. Fecal microbiomes were analyzed through 16S ribosomal RNA gene profiling, based on next-generation sequencing. Pancreatic exocrine function was evaluated by determining fecal elastase 1 activity. Results: Patients with CP showed gut microbiota dysbiosis with decreased diversity and richness, and taxa-composition changes. On the phylum level, the gut microbiome of the CP group showed lower Firmicutes and Actinobacteria abundances than the HC group and higher Proteobacteria abundances. The abundances of Escherichia-Shigella and other genera were high in gut microbiomes in the CP group, whereas that of Faecalibacterium was low. Kyoto Encyclopedia of Genes and Genomes pathways (lipopolysaccharide biosynthesis and bacterial invasion of epithelial cells) were predicted to be enriched in the CP group. Among the top 5 phyla and 8 genera (in terms of abundance), only Fusobacteria and Eubacterium rectale group showed significant differences between CP patients, with or without PEI. Correlation analysis showed that Bifidobacterium and Lachnoclostridium correlated positively with fecal elastase 1 (r = 0.2616 and 0.2486, respectively, P < 0.05). Conclusions: The current findings indicate that patients with CP have gut microbiota dysbiosis that is partly affected by pancreatic exocrine function.
4
Key words: 16S rRNA gene, fecal elastase 1, pancreatic exocrine insufficiency, pathway, profiling
5
Introduction Chronic pancreatitis (CP) is a complex disease that manifests with a highly variable clinical symptomatology and disease course. CP is characterized by repeated episodes of mild acute pancreatic inflammation, resulting in inflammatory cell infiltration and fibrosis. CP can affect the exocrine and endocrine functions of the pancreas. However, the pathogenesis of CP remains poorly understood [1-3]. In humans, the gut microbiota is a key contributor to host metabolism and plays an important role in immune homeostasis and immune disorders [4, 5]. Microbial dysbiosis-related diseases affect many organs, including the pancreas. The roles of microbiota in the development of pancreatic disorders, such as diabetes [6, 7] and acute pancreatitis [8, 9], are becoming increasingly appreciated. However, the role of intestinal microbiota in CP pathogenesis has yet to be understood [10, 11]. Previous studies showed that 36% of patients with CP have bacterial overgrowth in the small intestine, which could be associated with an imbalance of colonic microbiota, thereby further exacerbating the symptoms of CP [12, 13]. A recent study demonstrated that the intestinal microbiota of patients with CP and diabetes exhibited altered intestinal microbiota, although a relatively small number of patients were included in the study [14]. In another study, the intestinal microbiomes of patients with alcoholic chronic pancreatitis (ACP) were compared with those of alcoholic controls, but patients with CP of different etiologies were not included [15]. Frost et al. [16] recently showed that exocrine pancreatic function was the most important host factor involved in shaping the human intestinal microbiome in a population-based study. In this study, we investigated gut microbiome changes in patients with CP without diabetes, as well as the effects of pancreatic exocrine insufficiency (PEI) and different etiologies on the diversity and taxa composition of gut microbiomes. We present a comprehensive overview of the microbiota in
6
Chinese patients with CP, which provides novel insight into the pathogenic role of gut dysbiosis in CP pathogenesis. Methods
Patient information
All recruited patients with CP were inpatients of Changhai Hospital Affiliated to The Second Military Medical University, China. Written informed consent was obtained from each participant, and the study protocol was approved by the Research Ethics Committee of Changhai Hospital. The diagnosis of CP was established when one of the following conditions was met: (1) presence of pancreatic calcification, (2) pancreatic ductal changes (moderate or marked changes according to the Cambridge classification system), (3) abnormal results of pancreatic function tests, (4) endoscopic ultrasound abnormalities indicating CP, and (5) histological proof of CP as described by the Asia-Pacific consensus [17]. Alcoholic CP (ACP) was considered when the alcohol intake exceeded 80 g/day for men or 60 g/day for women for at least 2 years in the absence of other causes. Patients with CP were considered idiopathic (ICP) when none of definite etiologies, such as alcohol, abnormal anatomy of pancreatic duct, hyperlipidemia, or post-traumatic and hereditary factors were found. There was no use of opioids by any of the CP patients. Family members without medical problems who resided with the patients were enrolled as healthy controls (HCs). Participants who received antibiotics or probiotics within 1 month before sample collection were excluded from this study. Participants with gastrointestinal illnesses, such as inflammatory and irritable bowel disease or severe liver, neurological, cardiac, psychiatric, or metabolic diseases (including hypertension and diabetes mellitus) were also
7
excluded. The study protocol was based on the inclusion and exclusion criteria shown in Figure 1.
Sample collection and DNA extraction
Fecal samples from all participants were collected in sterile plastic tubes, transported to our laboratory in ice-filled coolers within 2 h of collection, and then stored at -80°C until analysis. Clinical data were collected through the inpatient medical record system. Total bacterial genomic deoxyribonucleic acid (DNA) samples were extracted using the Fast DNA SPIN Extraction Kit (MP Biomedicals, Santa Ana, CA, USA) in accordance with the manufacturer’s instructions and stored in H2O at -80°C prior to further analysis. The concentration of DNA extracted from each fecal sample was quantified using a NanoDrop 2000 UV-vis spectrophotometer. DNA integrity and size were assessed by agarose gel electrophoresis.
Fecal elastase-1 (FE1) determinations by enzyme-linked immunosorbent assay (ELISA) analysis
Fecal samples were diluted at 1:90, and FE1 concentrations (µg/g stool) were calculated spectraphotometrically (optical density, 405 mm) in comparison with a standard solution, according to manufacturer’ instructions of FE1 ELISA kit [18].
Polymerase chain reaction (PCR)-based amplification and 16S rRNA gene sequencing
The V3–V4 variable regions of the 16S rRNA gene extracted from each fecal sample were
8
amplified
using
primers
338F
(5′-ACTCCTACGGGAGGCAGCA-3′)
and
806R
(5′-GGACTACHVGGGTWTCTAAT-3′) on a GeneAmp 9700 thermal cycler PCR system (Applied Biosystems, USA). Sample-specific 7-base barcodes were incorporated into the primers for multiplex sequencing. PCR components contained 4 µL of 5 × FastPfu Buffer, 2 µL of 2.5 mM dNTPs, 0.8 µL of each primer (5 µM), 0.4 µL of FastPfu Polymerase, and 10 ng of template DNA. PCR analyses were conducted using the following thermocycling program: 3 min of denaturation at 95°C; 27 cycles of 30 s of denaturation at 95°C, 30s of annealing at 55°C, and 45s of elongation at 72°C; and a final extension step at 72°C for 10 min. PCR amplicons were purified with the AxyPrep DNA Gel Extraction Kit (Axygen, USA) and quantified using a QuantiFluor™-ST fluorometer (Promega, USA). After the individual quantification step, amplicons were pooled in equal amounts, and pair-end 2× 300-base pair (bp) sequencing was performed using the Illumina MiSeq platform. Raw reads were deposited into the NCBI Sequence Read Archive (SRA) database (accession numbers SRP157953 and SRP176493).
Processing of sequencing data
Raw sequencing reads with exact matches to the barcodes were assigned to the corresponding samples and identified as valid sequences. Raw fastq files were quality-filtered using Trimmomatic software and merged using FLASH software, based on the following criteria: (i) Reads truncated at any site received an average quality score of < 20 over a 50-bp sliding window. (ii) Sequences with overlaps of > 10 bp were merged in accordance with their overlaps with mismatches of ≤ 20%. (iii) The sequences of each sample were separated based on the barcodes (exact matches) and primers (allowing two nucleotide mismatches), and reads containing ambiguous bases were removed. Operational
9
taxonomic units (OTUs) were clustered with a cutoff of 97% similarity by using UPARSE (version 7.1, http://drive5.com/uparse/). The taxonomy of each 16S rRNA gene sequence was analyzed using the RDP Classifier Algorithm (http://rdp.cme.msu.edu/) against the Silva (SSU128) 16S rRNA database, with a confidence threshold of 70% [19].
Bioinformatics and statistical analysis
Sequence data were mainly analyzed using the QIIME and R packages (v3.2.0). In addition, sequencing data were analyzed using the free online Majorbio I-Sanger Cloud Platform (www.i-sanger.com). Principal coordinates analysis (PCoA) was conducted based on the OTU level using the R package. Analysis of similarities (ANOSIM) analysis was conducted on the OTU level using R-vegan. Statistically significant differences in the relative abundances of taxa were calculated by using the linear discriminant analysis (LDA) effect size method. Taxa with LDA results of >3 were considered statistically significantly enriched. Metabolic profiles of bacterial communities were predicted based on the 16S rRNA sequences using Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt) analysis. Before performing PICRUSt analysis, the OTU table was rarefied and then normalized by dividing the abundance of each organism by its predicted 16S rRNA gene-copy number. Predicted metagenomic contents were collapsed into Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway level 3 (http://www.genome.jp/kegg/pathway.html), and the pathway abundances were compared between groups using statistical analysis of taxonomic and functional profiles (STAMP). Correlations between FE1 levels and the abundances of fecal microbiota genera in
10
patients with CP were calculated using Spearman’s analysis. The data presented were expressed as the means ± SD and medians (25th–75th percentiles), and a P-value (P) < 0.05 was considered to reflect a statistically significant difference.
Results
Cohort description and sequencing data
Seventy-one patients with CP and 69 HC subjects were recruited, and the study protocol and participant distribution based on the inclusion and exclusion criteria are shown in Fig. 1. The age, gender ratio, and body–mass index (BMI) of the CP and HC groups were comparable (P > 0.05, Table 1). After quality filtering and trimming, 2,353,680 high-quality sequences were obtained, with an average of 16,812 reads per sample. The rarefaction curves showed clear asymptotes and indicated a near-complete sampling of the communities (Fig. 2A). A total of 1,589 OTUs and 2,528 OTUs were identified for the CP and HC groups, respectively. Both groups shared 1,370 OTUs. The demographic and clinical characteristics of HCs and patients with CP, CP groups with or without PEI, and CP groups with different etiologies are summarized in Table 1, Supplemental Table 1, and Supplemental Table 2.
CP-associated gut microbiota changes
The observed OTUs and Shannon, ACE, Chao1, phylogenetic diversity (PD), and coverage indexes of the fecal microbiota of patients with CP were significantly lower (P < 0.01) than those of the
11
HCs. In addition, the Simpson index of the fecal microbiota of patients with CP was significantly higher (P < 0.01) than that of the HCs (Table 2). Significant differences were also found in β -diversity based on the unweighted (ANOSIM r = 0.0336, P = 0.002) and the weighted (ANOSIM r = 0.0587, P = 0.001) unifrac distances between the CP and HC groups (Fig. 2B, C). Our data showed that the fecal microbial richness, diversity, and community coverage were significantly lower in the CP group than in the HC group. In addition, the fecal microbiome structure and composition in the CP group was significantly different from that of HC group. At the phylum level, the abundances of Firmicutes and Actinobacteria in the CP group were significantly lower than those in the HC group. The abundance of Proteobacteria in the CP group was significantly higher (P < 0.05) than that in the HC group (Fig. 2D). The top 15 bacterial genera with statistically significant different abundances were identified. The abundances of Escherichia-Shigella, Dialister, Parabacteroides, and Prevotella_7 in the fecal microbiomes of the CP group were significantly higher than those in the HC group. In contrast, the
abundances
of
Faecalibacterium,
unclassified_f_Lachnospiraceae,
Subdoligranulum,
Prevotella_9,
unclassified_f_Peptostreptococcaceae,
Megamonas, Collinsella,
Erysipelotrichaceae_UCG-003, Butyricicoccus, and Dorea were significantly lower in the fecal microbiomes of the CP group than in the HC group (Fig. 2E). The Spearman correlation test was performed to evaluate the relationship among the abundances of these 15 different genera. Significant negative correlations were found between Escherichia-Shigella and Faecalibacterium (r = -0.410, P < 0.001), Subdoligranulum (r = -0.303, P < 0.001), and Prevotella_9 (r = -0.226, P < 0.05). The results suggested that antagonistic effects exist between harmful and beneficial bacteria (Fig. 2F). Cladogram and LDA analyses identified taxa specifically associated with CP (Fig. 2G). A logarithmic LDA score cutoff of 3.0 was used to identify important taxonomic differences between the CP and HC groups.
12
Escherichia-Shigella, Prevotella_7, Parabacteroides, Eubacterium_hallii_group, and Sutterella were identified as dominant genera in the fecal microbiomes of patients with CP, but not in those of HCs, whereas Faecalibacterium, Megamonas, Subdoligranulum, and Prevotella_9 were found as dominant genera in the fecal microbiomes of HCs (Fig. 2H). Change in the gut microbiomes of CP patients, with or without PEI
FE1 determination by ELISA is the non-invasive gold standard for testing exocrine pancreatic function. It has high specificity and sensitivity, based on monoclonal antibodies, and the results are not influenced by digestive enzyme substitution therapy. According to the FE1 ELISA kit specification and literature by C Löser [18], the criteria for PEI is 200 µg/g. FE1 levels were determined in 71 patients with CP and the data from 70 patients were valid, with FE1 levels of > 200 µg/g found in 28 patients and FE1 values of < 200 µg/g found in 42 patients. Patients with CP were further classified into two subgroups (CP with or without PEI). The two subgroups showed no significant difference in the numbers of observed OTUs, or in the Shannon, Simpson, ACE, Chao1, PD, and coverage indexes of the fecal microbiota (Supplemental Table 3). Significant differences were found in the β-diversity, based on the unweighted unifrac test (ANOSIM r = 0.1064, P = 0.004), whereas no significant difference was found based on the weighted unifrac test (ANOSIM r = 0.0177, P = 0.197; Fig. 3A, B). Among the top 5 most abundant phyla, the abundance of only one phylum (Fusobacteria) was statistically different between the two subgroups. Among the top 8 most abundant genera, the abundance of only one genus (Eubacterium_rectale_group) was significantly different between the two subgroups (Fig. 3C, D). Cladogram and LDA analyses identified taxa specifically associated with both
13
subgroups (Fig. 3E). A logarithmic LDA score cutoff of 3.0 was used to identify important taxonomic differences between both subgroups. Eubacterium_rectale_group, Coprococcus, Sutterella, and Eubacterium_ruminantium_group were identified as dominant genera in the fecal microbiomes of CP patients with PEI, whereas Pseudomonas, Fusobacterium, and Ruminococcus_gnavus_group were found as dominant genera in the fecal microbiomes of CP patients without PEI (Fig. 3F). Change in the gut microbiomes of CP patients with different etiologies
According to the etiologies, patients with CP were further classified into CP_alcoholic group (n = 15) and CP_idiopathic group (n = 56). The two subgroups showed no significant difference in the numbers of observed OTUs, or in the Shannon,Simpson, ACE, Chao1, PD, and Coverage indexes of the fecal microbiota (Supplemental Table 4). There was no significant difference in β-diversity based on the unweighted unifrac test (ANOSIM r = -0.0799, P = 0.816) and weighted unifrac test (ANOSIM r = 0.0083, P = 0.428; S-Fig. 1A and 1B.) Among the top 5 most abundant phyla in fecal microbiomes, there was no statistical difference between the two subgroups. Among the top 10 most abundant genera in fecal microbiomes, there was no significant difference between the two subgroups (S-Fig. 1C and 1D). Cladogram and LDA analyses identified taxa specifically associated with both subgroups. A logarithmic LDA score cutoff of 3.0 was used to identify important taxonomic differences between both subgroups. g_Ruminococcus_torques_group were identified as dominant genera in the fecal microbiomes of ACP patients, whereas g_Dialisterwere found as dominant genera in the fecal microbiomes of ICP patients (S-Fig. 1E and 1F).
14
Microbial functions in CP and associations between FE1 and gut microbiota
We predicted functional-composition profiles from the 16S rRNA data of the CP and HC groups using PICRUSt analysis to characterize functional alterations in the gut microbiomes of patients with CP. Multiple level-3 KEGG categories were disturbed in patients with CP. The pathways of lipopolysaccharide (LPS) biosynthesis (Fig. 4A) and bacterial invasion of epithelial cells (Fig. 4B) were enriched in the CP group. Arginine and proline metabolism and the glycolysis/gluconeogenesis pathways were highly enriched in the HC group. Correlation analysis was also performed between taxa and predicted KEGG pathways at the OTU level. OTU2627 from Escherichia-Shigella correlated positively with the pathways of LPS biosynthesis and the bacterial invasion of epithelial cells (r = 0.454 and 0.694, respectively; P < 0.001), as shown in Fig. 4C, D. Correlation analysis was used to evaluate the relationships between the level of FE1 and the top 15 most abundant genera in patients with CP. We found that Bifidobacterium and Lachnoclostridium correlated positively with FE1 (r = 0.2616 and 0.2486, respectively; P < 0.05), as shown in Fig. 4E, F.
Discussion
We applied 16S rRNA sequencing to demonstrate that intestinal microbiotas associated with CP have different structures and compositions compared with HCs. CP patients with PEI showed no significant differences in α-diversity and β-diversity, as determined using weighted UniFrac testing (ANOSIM r = 0.0177, P = 0.197), when compared with those without PEI. In addition, both subgroups
15
had little differences in phylum and genus taxa composition abundances. ACP patients showed no significant differences in α-diversity and β-diversity when compared with ICP patients. The two subgroups also had no differences in the top 5 phylum and top 10 genus taxa composition abundances. To our knowledge, this is the largest study to date where the intestinal microbiotas of patients with CP were profiled, not including CP patients with diabetes. Our results indicated that the abundances of potential intestinal pathogens and commensal bacteria were imbalanced in patients with CP. These results partially agreed with those of Ciocan et al. [15] who found that the intestinal microbiota of patients with ACP had a lower bacterial diversity and different composition than that of alcoholic control patients. However, we did not observe increased abundances of Klebsiella, Enterococcus, and Sphingomonas in patients with ACP in the present study, possibly due to the relatively small number of ACP patients (n = 15, 21.1%) and different gender ratio between the two subgroups. A decreased abundance of Faecalibacterium found by Jandhyala et al. was also found in this study. However, the decreased abundance of the phylum Bacteroidetes and increased ratio of Firmicutes:Bacteroidetes in both the non-diabetic CP and diabetic CP groups they observed [14] was not found in this study. We also did not observe a reduced abundance of Ruminococcus bromii. Chronic Escherichia infections can induce cell cycle disorders, DNA damage, and inflammation. Members of Escherichia-Shigella are involved in the pathogenesis of inflammatory bowel disease and are associated with mucosal inflammation [20]. Moreover, the Escherichia-Shigella genus is a primary contributor to hepatic disease progression [21, 22]. LPS produced by members of the Escherichia-Shigella genus activates Toll-like receptor 4 to elicit inflammatory signaling [23]. High LPS levels activated the NF-κB pathway, induced the activation of proinflammatory cytokines (TNF-α, IL-6, and IL-1), and elicited inflammatory and oxidative damage [24, 25]. LPS challenge activated
16
pancreatic stellate cells and drastically increased collagen production [26]. Faecalibacterium is among the most abundant commensals in the human intestines. It is a key producer of butyrate and has been associated with anti-inflammatory activities. In addition, members of this genus improved gut barrier function by stimulating the synthesis of mucin and tight-junction proteins [27, 28]. The anti-inflammatory effects of Faecalibacterium in colitis mouse models were partially associated with metabolites that block NF-κB and IL-8 production [29]. In this research, a markedly decreased abundance of Subdoligranulum in the gut microbiomes of the CP group was also observed. Subdoligranulum can produce butyrate and enhance the intestinal health by providing energy for host cells and maintaining gut barrier integrity. Low Subdoligranulum abundance was previously correlated with increased incidences of inflammatory bowel disease [30, 31]. Diet, as well as geographic and ethnic origins can considerably affect the gut microbiota composition [32, 33]. Recruiting healthy family members as controls is an important approach that may have mitigated the influence of dietary and geographic differences on our results [34]. We only enrolled participants with Han Chinese ethnicity to prevent the effects of ethnic origins on our results. Patients with type-2 diabetes are characterized by a moderate degree of gut microbial dysbiosis [35]. Diabetes and glucose intolerance are common complications of CP. Given that islet cell loss distinguishes pancreatogenic diabetes from type-2 diabetes, diabetes mellitus resulting from CP is classified as pancreatogenic diabetes which is also known as type-3c diabetes [36]. In this study, we excluded CP patients with type-3c diabetes to prevent the potential effects of metabolic imbalances on intestinal microbiota. In addition to being involved in digestive proteases, pancreatic acinar cells also secrete antimicrobial peptides. A recent study showed that the secretion of antimicrobial peptides by pancreatic
17
acinar cells regulated the intestinal microbiota composition[37]. Levels of the antibacterial compound cathelicidin-related peptide (CRAMP) in pancreatic extracts of Orai1-/- mice were markedly reduced; supplementation of digestive enzymes failed to prevent intestinal dysbiosis, whereas CRAMP substitution substantially increased survival rates [37]. In dogs with exocrine pancreatic insufficiency, digestive enzyme supplementation did not prevent dysbiosis, further evidencing that digestive enzymes do not regulate the microbiome composition [38, 39]. However, Frost et al. [16] found that cathelicidin antimicrobial peptide levels were not significantly different in patients with reduced exocrine pancreatic function and control subjects. In this study, we found that gut microbiota dysbiosis was partly affected by pancreatic exocrine function in patients with CP. We speculate that the dysbiosis of intestinal microbiomes in patients with CP could be associated with the decreased levels of antimicrobial peptides secreted by pancreatic acinar cells. Isaiah et al. [38] found that the abundances of Lachnospiraceae and Ruminococcaceae were significantly lower in dogs with PEI. Members of the Lachnospiraceae family are butyrate producers. Butyrate production in the human gut could promotes Treg cell differentiation, which can ultimately suppress proinflammatory responses [40]. The current research showed that the Lachnoclostridium and Bifidobacterium genera correlated positively with FE1 levels, meaning that the abundance of beneficial bacteria decreased with the development of PEI. Our data indicated that for CP patients with PEI, supplementation with Lachnoclostridium and Bifidobacterium could be a viable option for regulating the dysbiosis. In summary, the strength of our study lies in its large cohort and optimized study design. These characteristics ensure that our study captured microbial diversity in patients with CP. Several limitations of this study must be addressed. First, stool samples provide an unselected representation of microbial ecology and microbiomes on the mucosal surface show marked heterogeneity. Second, in
18
contrast to quantitative metagenomics methods based on deep shotgun sequencing, the 16S rRNA gene sequencing method does not provide direct data related to functionally important changes in microbiomes. Third, the cross-sectional nature of the study did not enable us to determine the mechanisms and temporal associations of gut microbiota dysbiosis. Fourth, our study did not include animal experiments to examine probable mechanisms linking gut microbiota dysbiosis to CP development. The effects of changes in gut microbiota occurring in CP warrant further investigation. A logical extension of this study is to compare bacteria-derived metabolites of CP patients with that from HCs and identify candidate metabolites that could exacerbate pancreatic disease. Elucidating the regulatory mechanisms and functions of gut microbiomes may help improve our understanding of CP pathogenesis and could support potentially novel therapeutic options that could modify the gut microbiota of patients with CP.
Acknowledgements This study was supported by grants from National Natural Science Foundation of China (81670485 and 81800570), the Doctor Research Program of Naval Medical University (D-W Zou), and the Young Medical Talents Program and Natural Science Research Project of Universities in Jiangsu Province (C-H Zhou, QNRC2016863 and 18KJB320021). The authors declare that they have no conflicts of interest.
Author contributions
19
C-H Zhou, Y-T Meng, and J-J Xu designed the study. X Fang, J-L Zhao, and W-Zhou recruited the participants and collected the samples. J-H Zhao, J-C Han, W-B Zou, and K-X Wang performed the microbiota and medical analyses. L-H Hu and Z Liao coordinated the research. C-H Zhou, Y-T Meng, and L-Zhang wrote the manuscript. All authors critically revised the manuscript and approved the final version. References [1] Majumder S, Chari ST. Chronic pancreatitis. Lancet 2016; 387: 1957-66. [2] Yadav D, Slivka A. Managing chronic pancreatitis: the view from medical pancreatology. Am J Gastroenterol 2018; 113: 1108-10. [3] Hao L, Zeng XP, Xin L, Wang D, Pan J, Bi YW, et al. Incidence of and risk factors for pancreatic cancer in chronic pancreatitis: a cohort of 1656 patients. Dig Liver Dis 2017; 49: 1249-56. [4] Rooks MG, Garrett WS. Gut microbiota, metabolites and host immunity. Nat Rev Immunol 2016; 16: 341-52. [5] Marchesi JR, Adams DH, Fava F, Hermes GD, Hirschfield GM, Hold G, et al. The gut microbiota and host health: a new clinical frontier. Gut 2016; 65: 330-9. [6] Leal-Lopes C, Velloso FJ, Campopiano JC, Sogayar MC, Correa RG. Roles of commensal microbiota in pancreas homeostasis and pancreatic pathologies. J Diabetes Res 2015; 2015: 284680. [7] Sohail MU, Althani A, Anwar H, Rizzi R, Marei HE. Role of the gastrointestinal tract microbiome in the pathophysiology of diabetes mellitus. J Diabetes Res 2017; 2017: 9631435. [8] Tan C, Ling Z, Huang Y, Cao Y, Liu Q, Cai T, et al. Dysbiosis of intestinal microbiota associated with inflammation involved in the progression of acute pancreatitis. Pancreas 2015; 44: 868-75.
20
[9] Cen ME, Wang F, Su Y, Zhang WJ, Sun B, Wang G. Gastrointestinal microecology: a crucial and potential target in acute pancreatitis. Apoptosis 2018; 23: 377-87. [10] Memba R, Duggan SN, Ni Chonchubhair HM, Griffin OM, Bashir Y, O'Connor DB, et al. The potential role of gut microbiota in pancreatic disease: a systematic review. Pancreatology 2017; 17: 867-74. [11] Akshintala VS, Talukdar R, Singh VK, Goggins M: The gut microbiome in pancreatic disease. Clin Gastroenterol Hepatol 2019; 17: 290-5. [12] Capurso G, Signoretti M, Archibugi L, Stigliano S, Delle Fave G. Systematic review and meta-analysis: Small intestinal bacterial overgrowth in chronic pancreatitis. United European Gastroenterol J 2016; 4: 697-705. [13] Signoretti M, Roggiolani R, Stornello C, Delle Fave G, Capurso G. Gut microbiota and pancreatic diseases. Minerva Gastroenterol Dietol 2017; 63: 399-410. [14] Jandhyala SM, Madhulika A, Deepika G, Rao GV, Reddy DN, Subramanyam C, et al. Altered intestinal microbiota in patients with chronic pancreatitis: Implications in diabetes and metabolic abnormalities. Sci Rep 2017; 7: 43640. [15] Ciocan D, Rebours V, Voican CS, Wrzosek L, Puchois V, Cassard AM, et al. Characterization of intestinal microbiota in alcoholic patients with and without alcoholic hepatitis or chronic alcoholic pancreatitis. Sci Rep 2018; 8: 4822. [16] Frost F, Kacprowski T, Ruhlemann M, Bulow R, Kuhn JP, Franke A, et al. Impaired exocrine pancreatic function associates with changes in intestinal microbiota composition and diversity. Gastroenterology 2019; 156: 1010-15. [17] Tandon RK, Sato N, Garg PK, Consensus Study Group. Chronic pancreatitis: Asia-Pacific
21
consensus report. J Gastroenterol Hepatol 2002; 17: 508-18. [18] Loser C, Mollgaard A, Folsch UR: Faecal elastase 1: a novel, highly sensitive, and specific tubeless pancreatic function test. Gut 1996; 39: 580-6. [19] Cole JR, Wang Q, Cardenas E, Fish J, Chai B, Farris RJ, et al. The ribosomal database project: improved alignments and new tools for rrna analysis. Nucleic Acids Res 2009; 37: D141-5. [20] Xu J, Chen N, Wu Z, Song Y, Zhang Y, Wu N, et al. 5-aminosalicylic acid alters the gut bacterial microbiota in patients with ulcerative colitis. Front Microbiol 2018; 9: 1274. [21] Shen F, Zheng RD, Sun XQ, Ding WJ, Wang XY, Fan JG. Gut microbiota dysbiosis in patients with non-alcoholic fatty liver disease. Hepatobiliary Pancreat Dis Int 2017; 16: 375-81. [22] Woodhouse CA, Patel VC, Singanayagam A, Shawcross DL. Review article: the gut microbiome as a therapeutic target in the pathogenesis and treatment of chronic liver disease. Aliment Pharmacol Ther 2018; 47: 192-202. [23] Beutler B. Tlr4 as the mammalian endotoxin sensor. Curr Top Microbiol Immunol 2002; 270: 109-20. [24] Vonlaufen A, Spahr L, Apte MV, Frossard JL. Alcoholic pancreatitis: a tale of spirits and bacteria. World J Gastrointest Pathophysiol 2014; 5: 82-90. [25] Lamping N, Dettmer R, Schroder NW, Pfeil D, Hallatschek W, Burger R, et al. Lps-binding protein protects mice from septic shock caused by lps or gram-negative bacteria. J Clin Invest 1998; 101: 2065-71. [26] Sun L, Xiu M, Wang S, Brigstock DR, Li H, Qu L, et al. Lipopolysaccharide enhances tgf-beta1 signalling pathway and rat pancreatic fibrosis. J Cell Mol Med 2018; 22: 2346-56. [27] Arumugam M, Raes J, Pelletier E, Le Paslier D, Yamada T, Mende DR, et al. Enterotypes of the
22
human gut microbiome. Nature 2011; 473: 174-80. [28] Wrzosek L, Miquel S, Noordine ML, Bouet S, Joncquel Chevalier-Curt M, Robert V, et al. Bacteroides thetaiotaomicron and faecalibacterium prausnitzii influence the production of mucus glycans and the development of goblet cells in the colonic epithelium of a gnotobiotic model rodent. BMC Biol 2013; 11: 61. [29] Sokol H, Pigneur B, Watterlot L, Lakhdari O, Bermudez-Humaran LG, Gratadoux JJ, et al. Faecalibacterium prausnitzii is an anti-inflammatory commensal bacterium identified by gut microbiota analysis of crohn disease patients. Proc Natl Acad Sci U S A 2008; 105: 16731-6. [30] Deaver JA, Eum SY, Toborek M. Circadian disruption changes gut microbiome taxa and functional gene composition. Front Microbiol 2018; 9: 737. [31] Thomas LV, Ockhuizen T, Suzuki K. Exploring the influence of the gut microbiota and probiotics on health: a symposium report. Br J Nutr 2014; 112: S1-18. [32] Escobar JS, Klotz B, Valdes BE, Agudelo GM. The gut microbiota of colombians differs from that of americans, europeans and asians. BMC Microbiol 2014; 14: 311. [33] Graf D, Di Cagno R, Fak F, Flint HJ, Nyman M, Saarela M, et al. Contribution of diet to the composition of the human gut microbiota. Microb Ecol Health Dis 2015; 26: 26164. [34] Song SJ, Lauber C, Costello EK, Lozupone CA, Humphrey G, Berg-Lyons D, et al. Cohabiting family members share microbiota with one another and with their dogs. Elife 2013; 2: e00458. [35] Qin J, Li Y, Cai Z, Li S, Zhu J, Zhang F, et al. A metagenome-wide association study of gut microbiota in type 2 diabetes. Nature 2012; 490: 55-60. [36] Rickels MR, Bellin M, Toledo FG, Robertson RP, Andersen DK, Chari ST, et al. Detection, evaluation and treatment of diabetes mellitus in chronic pancreatitis: recommendations from
23
PancreasFest 2012. Pancreatology 2013; 13: 336-42. [37] Ahuja M, Schwartz DM, Tandon M, Son A, Zeng M, Swaim W, et al. Orai1-mediated antimicrobial secretion from pancreatic acini shapes the gut microbiome and regulates gut innate immunity. Cell Metab 2017; 25: 635-46. [38] Isaiah A, Parambeth JC, Steiner JM, Lidbury JA, Suchodolski JS. The fecal microbiome of dogs with exocrine pancreatic insufficiency. Anaerobe 2017; 45: 50-58. [39] Rubinstein E, Mark Z, Haspel J, Ben-Ari G, Dreznik Z, Mirelman D, et al. Antibacterial activity of the pancreatic fluid. Gastroenterology 1985; 88: 927-32. [40] Singh N, Gurav A, Sivaprakasam S, Brady E, Padia R, Shi H, et al. Activation of gpr109a, receptor for niacin and the commensal metabolite butyrate, suppresses colonic inflammation and carcinogenesis. Immunity 2014; 40: 128-39.
24
Figure legends Fig. 1. Study protocol and allocation of participants to different groups, based on inclusion and exclusion criteria
Fig. 2. Structure and composition of the gut microbiomes of patients with CP. Rarefaction curves tended to approach the saturation plateau at the operational taxonomic unit (OTU) level (A). Principal coordinates analysis (PCoA) with weighted and unweighted unifrac testing were used to compare the microbiota compositions of CP with HC groups. ANOSIM revealed significant differences in the structures of both groups (r = 0.0336, P = 0.002, unweighted unifrac; r = 0.0587，P = 0.001, weighted unifrac). The axes represent the three dimensions explaining the greatest proportion of variance in the communities. Each symbol represents a sample (B, C). The top 15 fecal microbiota with significantly different abundances between the CP and HC groups at the phylum (D) and genus (E) levels are shown. Correlation analysis of the 15 significantly different genera identified in the fecal microbiomes of HCs and patients with CP (F). Cladogram showing taxa with the largest differences in relative abundances between the CP and HC groups. The circle sizes in the cladogram plot are proportional to the bacterial abundances. Going from the inside to the outside, the circles represent the phylum, class, order, family, and genus. Only taxa with an LDA score of >3 and P < 0.05 in the Wilcoxon signed-rank test are shown. (G). A logarithmic LDA-score cutoff of 3.0 was used to identify important taxonomic differences between the fecal microbiomes of HCs and patients with CP (H).
Fig. 3. Structures and compositions of the gut microbiomes of CP patients with or without PEI. Principal coordinates analysis (PCoA) with weighted and unweighted unifrac testing were used to
25
compare the microbiota compositions of CP patients with or without PEI. ANOSIM revealed significant differences in the structures of both subgroups using unweighted unifrac test (r = 0.1064, P = 0.004), whereas no significant differences were observed using weighted unifrac test (r = 0.0177, P = 0.197) (A, B). Top 15 fecal microbiota in abundance of the CP patients with PEI and without PEI on the phylum (C) and genus (D) levels are shown. Cladogram showing taxa with the largest differences in relative abundances between CP patients with or without PEI. Circle sizes in the cladogram plot are proportional to bacterial abundances. Going from the inside to the outside, the circles represent the phylum, class, order, family, and genus. Only taxa with an LDA score >3 and P < 0.05 in the Wilcoxon signed-rank test are shown (E). A logarithmic LDA-score cutoff of 3.0 was used to identify important taxonomic differences between the fecal microbiomes of CP patients with or without PEI (F).
Fig. 4. Predicted Kyoto Encyclopedia of Genes and Genomes (KEGG) level 3 pathways, i.e., lipopolysaccharide (LPS) biosynthesis and bacterial invasion of epithelial cells, which were enriched in patients with CP (A, B). Pathways associated with LPS biosynthesis and bacterial invasion of epithelial cells positively correlated with Escherichia-Shigella OTU2627 (r = 0.454 and 0.694, respectively; P < 0.001), which was enriched in the gut microbiomes of patients with CP (C, D). Correlation analysis was used to evaluate the relationship between FE1 levels and the top 15 most abundant genera in patients with CP. Bifidobacterium and Lachnoclostridium correlated positively with FE 1 (r = 0.2616 and 0.2486, respectively; P < 0.05) (E, F) .
26
Supplemental Fig.1. Structures and compositions of the gut microbiomes of CP_alcholic group and CP_idiopathic group. Principal coordinates analysis (PCoA) with weighted and unweighted unifrac test were used to compare the microbiota compositions of the two subgroups. ANOSIM revealed no significant differences in the structures of fecal microbiota in both subgroups using unweighted and weighted unifrac test (r = -0.0799, P = 0.816 and r = 0.0083, P = 0.428) (A, B). Top 15 fecal microbiota in CP_alcholic group and CP_idiopathic group on the phylum (C) and genus (D) levels are shown. Cladogram showing taxa with the largest differences in relative abundances between CP_alcholic group and CP_idiopathic group. Circle sizes in the cladogram plot are proportional to bacterial abundances. Going from the inside to the outside, the circles represent the phylum, class, order, family, and genus. Only taxa with an LDA score >3 and P < 0.05 in the Wilcoxon signed-rank test are shown (E). A logarithmic LDA-score cutoff of 3.0 was used to identify important taxonomic differences between the fecal microbiomes of the two subgroups (F).
Tables
Table 1 Demographic and Clinical data of healthy controls (HCs) and patients with chronic pancreatitis (CP) CP (n = 71)
HC (n = 69)
P value
Male/female
41/30
29/40
0.06
Age, median (mean ± SD)
44 ± 11
47 ± 10
0.07
BMI, median (min–max)
21.0 (18.3–28.0)
22.3 (18.7–29.8)
0.26
Disease duration, median
24 (0.5–276)
-
-
Dilated pancreatic duct
68
-
-
Pancreatic duct stone
68
-
-
Pancreatic atrophy
40
-
-
Pancreatic pseudocyst
11
-
-
Surgery history
6
-
-
ERCP history
39
-
-
ESWL history
27
-
-
Hemoglobin level (g/L), (mean ± SD)
141 ± 10
142 ± 15
0.87
FBG, median (min–max)
5.3 (4.2–6.8)
4.7 (4.1–6.1)
<0.0001
(min–max)
Data are expressed as the mean ± SD or median (25th–75th percentiles) in accordance with the normality of distribution. BMI, body–mass index; ERCP, endoscopic retrograde cholangiopancreatography; ESWL, extracorporeal shock wave lithotripsy
Table 2 Richness and diversity indexes of the gut microbiotas of HCs and patients with CP Parameter
CP
HCs
P value
Observed OTUs
218.76 ± 63.36
283.09 ± 146.79
0.000934*
Shannon
3.09 ± 0.60
3.46 ± 0.70
0.001277*
Simpson
0.13 ± 0.083
0.096 ± 0.071
0.02527*
ACE
297.47 ± 96.40
376.24 ± 200.73
0.00348*
Chao1
285.56 ± 83.70
369.13 ± 194.04
0.00113*
PD
19.53 ± 4.92
24.63 ± 13.24
0.00284*
Coverage
0.997 ± 0.001
0.996 ± 0.003
0.00744*
The diversities of fecal microbiomes in patients with CP decreased relative to those in HCs. α-diversity was estimated based on the observed OTUs and the Shannon, ACE, Chao 1, and phylogenetic diversity (PD) indexes. The α-diversities of the fecal microbiomes of patients with CP showed statistically significant reductions relative to those in HCs. In contrast, the Simpson index of the gut microbiotas of patients with CP was significant higher than that of HCs. Data are expressed as the mean ± SD. Student’s t-test was used to identify statistically significant differences in each variable between the HC and CP groups. P < 0.05 was considered to reflect statistical significance.
S1424-3903(19)30794-X
DOI:
https://doi.org/10.1016/j.pan.2019.11.013
Reference:
PAN 1123
To appear in:
Pancreatology
Received Date: 9 June 2019 Revised Date:
14 November 2019
Accepted Date: 23 November 2019
Please cite this article as: Zhou C-H, Meng Y-T, Xu J-J, Fang X, Zhao J-L, Zhou W, Zhao J, Han JC, Zhang L, Wang K-X, Hu L-H, Liao Z, Zou W-B, Li Z-S, Zou D-W, Altered diversity and composition of gut microbiota in Chinese patients with chronic pancreatitis, Pancreatology (2019), doi: https:// doi.org/10.1016/j.pan.2019.11.013. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier B.V. on behalf of IAP and EPC.
1
Altered Diversity and Composition of Gut Microbiota in Chinese Patients with Chronic Pancreatitis
Chun-Hua Zhoua,b,d,#, Yu-Ting Menga,b,#, Jia-Jia Xua,b,#, Xue Fanga,b, Jiu-Long Zhaoa,b, Wei Zhoua,b, Jianhua Zhaoc, Ji-Chen Hanc, Ling Zhangd, Kai-Xuan Wanga,b, Liang-Hao Hua,b, Zhuan-Liaoa,b, Wen-Bin Zoua,b,*, Zhao-Shen Lia,b,*, Duo-Wu Zou a,b,d*
a
Department of Gastroenterology, Changhai Hospital, The Second Military Medical University,
Shanghai, China b
c
Shanghai Institute of Pancreatic Diseases, Shanghai, China
Shanghai Majorbio Bio-pharm Technology Co., Ltd.
d
Department of Gastroenterology, Ruijin Hospital Affiliated to Shanghai Jiao Tong University School
of Medicine, Shanghai, China
#
These authors contributed equally to this work.
Short title: Gut Microbiota Alterations in Chronic pancreatitis
#
Corresponding authors:
Duo-Wu Zou Department of Gastroenterology, Changhai Hospital, The Second Military Medical University, Shanghai, China No.168, Changhai Road,Yangpu District, Shanghai,China, 200433 Department of Gastroenterology, Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China
2
No.197,Rui Jin Er Road,Shanghai,China,200025 Tel.: +86 21 32162770 Fax: +86 21 55621735 Email address: [email protected]
Zhao-Shen Li Department of Gastroenterology, Changhai Hospital, The Second Military Medical University, Shanghai, China No.168, Changhai Road,Yangpu District, Shanghai,China, 200433 Tel.: +86 21 32162770 Fax: +86 21 55621735 Email address: [email protected] Wen-Bin Zou Department of Gastroenterology, Changhai Hospital, The Second Military Medical University, Shanghai, China No.168, Changhai Road,Yangpu District, Shanghai,China,200433 Tel.: +86 21 32162770 Fax: +86 21 55621735 Email address: [email protected]
3
Abstract Background/Objectives: Gut microbiota alterations in chronic pancreatitis (CP) are seldomly described systematically. It is unknown whether pancreatic exocrine insufficiency (PEI) and different etiologies in patients with CP are associated with gut microbiota dysbiosis. Methods: The fecal microbiota of 69 healthy controls (HCs) and 71 patients with CP were compared to investigate gut microbiome alterations in CP and the relationship among gut microbiome dysbiosis, PEI and different etiologies. Fecal microbiomes were analyzed through 16S ribosomal RNA gene profiling, based on next-generation sequencing. Pancreatic exocrine function was evaluated by determining fecal elastase 1 activity. Results: Patients with CP showed gut microbiota dysbiosis with decreased diversity and richness, and taxa-composition changes. On the phylum level, the gut microbiome of the CP group showed lower Firmicutes and Actinobacteria abundances than the HC group and higher Proteobacteria abundances. The abundances of Escherichia-Shigella and other genera were high in gut microbiomes in the CP group, whereas that of Faecalibacterium was low. Kyoto Encyclopedia of Genes and Genomes pathways (lipopolysaccharide biosynthesis and bacterial invasion of epithelial cells) were predicted to be enriched in the CP group. Among the top 5 phyla and 8 genera (in terms of abundance), only Fusobacteria and Eubacterium rectale group showed significant differences between CP patients, with or without PEI. Correlation analysis showed that Bifidobacterium and Lachnoclostridium correlated positively with fecal elastase 1 (r = 0.2616 and 0.2486, respectively, P < 0.05). Conclusions: The current findings indicate that patients with CP have gut microbiota dysbiosis that is partly affected by pancreatic exocrine function.
4
Key words: 16S rRNA gene, fecal elastase 1, pancreatic exocrine insufficiency, pathway, profiling
5
Introduction Chronic pancreatitis (CP) is a complex disease that manifests with a highly variable clinical symptomatology and disease course. CP is characterized by repeated episodes of mild acute pancreatic inflammation, resulting in inflammatory cell infiltration and fibrosis. CP can affect the exocrine and endocrine functions of the pancreas. However, the pathogenesis of CP remains poorly understood [1-3]. In humans, the gut microbiota is a key contributor to host metabolism and plays an important role in immune homeostasis and immune disorders [4, 5]. Microbial dysbiosis-related diseases affect many organs, including the pancreas. The roles of microbiota in the development of pancreatic disorders, such as diabetes [6, 7] and acute pancreatitis [8, 9], are becoming increasingly appreciated. However, the role of intestinal microbiota in CP pathogenesis has yet to be understood [10, 11]. Previous studies showed that 36% of patients with CP have bacterial overgrowth in the small intestine, which could be associated with an imbalance of colonic microbiota, thereby further exacerbating the symptoms of CP [12, 13]. A recent study demonstrated that the intestinal microbiota of patients with CP and diabetes exhibited altered intestinal microbiota, although a relatively small number of patients were included in the study [14]. In another study, the intestinal microbiomes of patients with alcoholic chronic pancreatitis (ACP) were compared with those of alcoholic controls, but patients with CP of different etiologies were not included [15]. Frost et al. [16] recently showed that exocrine pancreatic function was the most important host factor involved in shaping the human intestinal microbiome in a population-based study. In this study, we investigated gut microbiome changes in patients with CP without diabetes, as well as the effects of pancreatic exocrine insufficiency (PEI) and different etiologies on the diversity and taxa composition of gut microbiomes. We present a comprehensive overview of the microbiota in
6
Chinese patients with CP, which provides novel insight into the pathogenic role of gut dysbiosis in CP pathogenesis. Methods
Patient information
All recruited patients with CP were inpatients of Changhai Hospital Affiliated to The Second Military Medical University, China. Written informed consent was obtained from each participant, and the study protocol was approved by the Research Ethics Committee of Changhai Hospital. The diagnosis of CP was established when one of the following conditions was met: (1) presence of pancreatic calcification, (2) pancreatic ductal changes (moderate or marked changes according to the Cambridge classification system), (3) abnormal results of pancreatic function tests, (4) endoscopic ultrasound abnormalities indicating CP, and (5) histological proof of CP as described by the Asia-Pacific consensus [17]. Alcoholic CP (ACP) was considered when the alcohol intake exceeded 80 g/day for men or 60 g/day for women for at least 2 years in the absence of other causes. Patients with CP were considered idiopathic (ICP) when none of definite etiologies, such as alcohol, abnormal anatomy of pancreatic duct, hyperlipidemia, or post-traumatic and hereditary factors were found. There was no use of opioids by any of the CP patients. Family members without medical problems who resided with the patients were enrolled as healthy controls (HCs). Participants who received antibiotics or probiotics within 1 month before sample collection were excluded from this study. Participants with gastrointestinal illnesses, such as inflammatory and irritable bowel disease or severe liver, neurological, cardiac, psychiatric, or metabolic diseases (including hypertension and diabetes mellitus) were also
7
excluded. The study protocol was based on the inclusion and exclusion criteria shown in Figure 1.
Sample collection and DNA extraction
Fecal samples from all participants were collected in sterile plastic tubes, transported to our laboratory in ice-filled coolers within 2 h of collection, and then stored at -80°C until analysis. Clinical data were collected through the inpatient medical record system. Total bacterial genomic deoxyribonucleic acid (DNA) samples were extracted using the Fast DNA SPIN Extraction Kit (MP Biomedicals, Santa Ana, CA, USA) in accordance with the manufacturer’s instructions and stored in H2O at -80°C prior to further analysis. The concentration of DNA extracted from each fecal sample was quantified using a NanoDrop 2000 UV-vis spectrophotometer. DNA integrity and size were assessed by agarose gel electrophoresis.
Fecal elastase-1 (FE1) determinations by enzyme-linked immunosorbent assay (ELISA) analysis
Fecal samples were diluted at 1:90, and FE1 concentrations (µg/g stool) were calculated spectraphotometrically (optical density, 405 mm) in comparison with a standard solution, according to manufacturer’ instructions of FE1 ELISA kit [18].
Polymerase chain reaction (PCR)-based amplification and 16S rRNA gene sequencing
The V3–V4 variable regions of the 16S rRNA gene extracted from each fecal sample were
8
amplified
using
primers
338F
(5′-ACTCCTACGGGAGGCAGCA-3′)
and
806R
(5′-GGACTACHVGGGTWTCTAAT-3′) on a GeneAmp 9700 thermal cycler PCR system (Applied Biosystems, USA). Sample-specific 7-base barcodes were incorporated into the primers for multiplex sequencing. PCR components contained 4 µL of 5 × FastPfu Buffer, 2 µL of 2.5 mM dNTPs, 0.8 µL of each primer (5 µM), 0.4 µL of FastPfu Polymerase, and 10 ng of template DNA. PCR analyses were conducted using the following thermocycling program: 3 min of denaturation at 95°C; 27 cycles of 30 s of denaturation at 95°C, 30s of annealing at 55°C, and 45s of elongation at 72°C; and a final extension step at 72°C for 10 min. PCR amplicons were purified with the AxyPrep DNA Gel Extraction Kit (Axygen, USA) and quantified using a QuantiFluor™-ST fluorometer (Promega, USA). After the individual quantification step, amplicons were pooled in equal amounts, and pair-end 2× 300-base pair (bp) sequencing was performed using the Illumina MiSeq platform. Raw reads were deposited into the NCBI Sequence Read Archive (SRA) database (accession numbers SRP157953 and SRP176493).
Processing of sequencing data
Raw sequencing reads with exact matches to the barcodes were assigned to the corresponding samples and identified as valid sequences. Raw fastq files were quality-filtered using Trimmomatic software and merged using FLASH software, based on the following criteria: (i) Reads truncated at any site received an average quality score of < 20 over a 50-bp sliding window. (ii) Sequences with overlaps of > 10 bp were merged in accordance with their overlaps with mismatches of ≤ 20%. (iii) The sequences of each sample were separated based on the barcodes (exact matches) and primers (allowing two nucleotide mismatches), and reads containing ambiguous bases were removed. Operational
9
taxonomic units (OTUs) were clustered with a cutoff of 97% similarity by using UPARSE (version 7.1, http://drive5.com/uparse/). The taxonomy of each 16S rRNA gene sequence was analyzed using the RDP Classifier Algorithm (http://rdp.cme.msu.edu/) against the Silva (SSU128) 16S rRNA database, with a confidence threshold of 70% [19].
Bioinformatics and statistical analysis
Sequence data were mainly analyzed using the QIIME and R packages (v3.2.0). In addition, sequencing data were analyzed using the free online Majorbio I-Sanger Cloud Platform (www.i-sanger.com). Principal coordinates analysis (PCoA) was conducted based on the OTU level using the R package. Analysis of similarities (ANOSIM) analysis was conducted on the OTU level using R-vegan. Statistically significant differences in the relative abundances of taxa were calculated by using the linear discriminant analysis (LDA) effect size method. Taxa with LDA results of >3 were considered statistically significantly enriched. Metabolic profiles of bacterial communities were predicted based on the 16S rRNA sequences using Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt) analysis. Before performing PICRUSt analysis, the OTU table was rarefied and then normalized by dividing the abundance of each organism by its predicted 16S rRNA gene-copy number. Predicted metagenomic contents were collapsed into Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway level 3 (http://www.genome.jp/kegg/pathway.html), and the pathway abundances were compared between groups using statistical analysis of taxonomic and functional profiles (STAMP). Correlations between FE1 levels and the abundances of fecal microbiota genera in
10
patients with CP were calculated using Spearman’s analysis. The data presented were expressed as the means ± SD and medians (25th–75th percentiles), and a P-value (P) < 0.05 was considered to reflect a statistically significant difference.
Results
Cohort description and sequencing data
Seventy-one patients with CP and 69 HC subjects were recruited, and the study protocol and participant distribution based on the inclusion and exclusion criteria are shown in Fig. 1. The age, gender ratio, and body–mass index (BMI) of the CP and HC groups were comparable (P > 0.05, Table 1). After quality filtering and trimming, 2,353,680 high-quality sequences were obtained, with an average of 16,812 reads per sample. The rarefaction curves showed clear asymptotes and indicated a near-complete sampling of the communities (Fig. 2A). A total of 1,589 OTUs and 2,528 OTUs were identified for the CP and HC groups, respectively. Both groups shared 1,370 OTUs. The demographic and clinical characteristics of HCs and patients with CP, CP groups with or without PEI, and CP groups with different etiologies are summarized in Table 1, Supplemental Table 1, and Supplemental Table 2.
CP-associated gut microbiota changes
The observed OTUs and Shannon, ACE, Chao1, phylogenetic diversity (PD), and coverage indexes of the fecal microbiota of patients with CP were significantly lower (P < 0.01) than those of the
11
HCs. In addition, the Simpson index of the fecal microbiota of patients with CP was significantly higher (P < 0.01) than that of the HCs (Table 2). Significant differences were also found in β -diversity based on the unweighted (ANOSIM r = 0.0336, P = 0.002) and the weighted (ANOSIM r = 0.0587, P = 0.001) unifrac distances between the CP and HC groups (Fig. 2B, C). Our data showed that the fecal microbial richness, diversity, and community coverage were significantly lower in the CP group than in the HC group. In addition, the fecal microbiome structure and composition in the CP group was significantly different from that of HC group. At the phylum level, the abundances of Firmicutes and Actinobacteria in the CP group were significantly lower than those in the HC group. The abundance of Proteobacteria in the CP group was significantly higher (P < 0.05) than that in the HC group (Fig. 2D). The top 15 bacterial genera with statistically significant different abundances were identified. The abundances of Escherichia-Shigella, Dialister, Parabacteroides, and Prevotella_7 in the fecal microbiomes of the CP group were significantly higher than those in the HC group. In contrast, the
abundances
of
Faecalibacterium,
unclassified_f_Lachnospiraceae,
Subdoligranulum,
Prevotella_9,
unclassified_f_Peptostreptococcaceae,
Megamonas, Collinsella,
Erysipelotrichaceae_UCG-003, Butyricicoccus, and Dorea were significantly lower in the fecal microbiomes of the CP group than in the HC group (Fig. 2E). The Spearman correlation test was performed to evaluate the relationship among the abundances of these 15 different genera. Significant negative correlations were found between Escherichia-Shigella and Faecalibacterium (r = -0.410, P < 0.001), Subdoligranulum (r = -0.303, P < 0.001), and Prevotella_9 (r = -0.226, P < 0.05). The results suggested that antagonistic effects exist between harmful and beneficial bacteria (Fig. 2F). Cladogram and LDA analyses identified taxa specifically associated with CP (Fig. 2G). A logarithmic LDA score cutoff of 3.0 was used to identify important taxonomic differences between the CP and HC groups.
12
Escherichia-Shigella, Prevotella_7, Parabacteroides, Eubacterium_hallii_group, and Sutterella were identified as dominant genera in the fecal microbiomes of patients with CP, but not in those of HCs, whereas Faecalibacterium, Megamonas, Subdoligranulum, and Prevotella_9 were found as dominant genera in the fecal microbiomes of HCs (Fig. 2H). Change in the gut microbiomes of CP patients, with or without PEI
FE1 determination by ELISA is the non-invasive gold standard for testing exocrine pancreatic function. It has high specificity and sensitivity, based on monoclonal antibodies, and the results are not influenced by digestive enzyme substitution therapy. According to the FE1 ELISA kit specification and literature by C Löser [18], the criteria for PEI is 200 µg/g. FE1 levels were determined in 71 patients with CP and the data from 70 patients were valid, with FE1 levels of > 200 µg/g found in 28 patients and FE1 values of < 200 µg/g found in 42 patients. Patients with CP were further classified into two subgroups (CP with or without PEI). The two subgroups showed no significant difference in the numbers of observed OTUs, or in the Shannon, Simpson, ACE, Chao1, PD, and coverage indexes of the fecal microbiota (Supplemental Table 3). Significant differences were found in the β-diversity, based on the unweighted unifrac test (ANOSIM r = 0.1064, P = 0.004), whereas no significant difference was found based on the weighted unifrac test (ANOSIM r = 0.0177, P = 0.197; Fig. 3A, B). Among the top 5 most abundant phyla, the abundance of only one phylum (Fusobacteria) was statistically different between the two subgroups. Among the top 8 most abundant genera, the abundance of only one genus (Eubacterium_rectale_group) was significantly different between the two subgroups (Fig. 3C, D). Cladogram and LDA analyses identified taxa specifically associated with both
13
subgroups (Fig. 3E). A logarithmic LDA score cutoff of 3.0 was used to identify important taxonomic differences between both subgroups. Eubacterium_rectale_group, Coprococcus, Sutterella, and Eubacterium_ruminantium_group were identified as dominant genera in the fecal microbiomes of CP patients with PEI, whereas Pseudomonas, Fusobacterium, and Ruminococcus_gnavus_group were found as dominant genera in the fecal microbiomes of CP patients without PEI (Fig. 3F). Change in the gut microbiomes of CP patients with different etiologies
According to the etiologies, patients with CP were further classified into CP_alcoholic group (n = 15) and CP_idiopathic group (n = 56). The two subgroups showed no significant difference in the numbers of observed OTUs, or in the Shannon,Simpson, ACE, Chao1, PD, and Coverage indexes of the fecal microbiota (Supplemental Table 4). There was no significant difference in β-diversity based on the unweighted unifrac test (ANOSIM r = -0.0799, P = 0.816) and weighted unifrac test (ANOSIM r = 0.0083, P = 0.428; S-Fig. 1A and 1B.) Among the top 5 most abundant phyla in fecal microbiomes, there was no statistical difference between the two subgroups. Among the top 10 most abundant genera in fecal microbiomes, there was no significant difference between the two subgroups (S-Fig. 1C and 1D). Cladogram and LDA analyses identified taxa specifically associated with both subgroups. A logarithmic LDA score cutoff of 3.0 was used to identify important taxonomic differences between both subgroups. g_Ruminococcus_torques_group were identified as dominant genera in the fecal microbiomes of ACP patients, whereas g_Dialisterwere found as dominant genera in the fecal microbiomes of ICP patients (S-Fig. 1E and 1F).
14
Microbial functions in CP and associations between FE1 and gut microbiota
We predicted functional-composition profiles from the 16S rRNA data of the CP and HC groups using PICRUSt analysis to characterize functional alterations in the gut microbiomes of patients with CP. Multiple level-3 KEGG categories were disturbed in patients with CP. The pathways of lipopolysaccharide (LPS) biosynthesis (Fig. 4A) and bacterial invasion of epithelial cells (Fig. 4B) were enriched in the CP group. Arginine and proline metabolism and the glycolysis/gluconeogenesis pathways were highly enriched in the HC group. Correlation analysis was also performed between taxa and predicted KEGG pathways at the OTU level. OTU2627 from Escherichia-Shigella correlated positively with the pathways of LPS biosynthesis and the bacterial invasion of epithelial cells (r = 0.454 and 0.694, respectively; P < 0.001), as shown in Fig. 4C, D. Correlation analysis was used to evaluate the relationships between the level of FE1 and the top 15 most abundant genera in patients with CP. We found that Bifidobacterium and Lachnoclostridium correlated positively with FE1 (r = 0.2616 and 0.2486, respectively; P < 0.05), as shown in Fig. 4E, F.
Discussion
We applied 16S rRNA sequencing to demonstrate that intestinal microbiotas associated with CP have different structures and compositions compared with HCs. CP patients with PEI showed no significant differences in α-diversity and β-diversity, as determined using weighted UniFrac testing (ANOSIM r = 0.0177, P = 0.197), when compared with those without PEI. In addition, both subgroups
15
had little differences in phylum and genus taxa composition abundances. ACP patients showed no significant differences in α-diversity and β-diversity when compared with ICP patients. The two subgroups also had no differences in the top 5 phylum and top 10 genus taxa composition abundances. To our knowledge, this is the largest study to date where the intestinal microbiotas of patients with CP were profiled, not including CP patients with diabetes. Our results indicated that the abundances of potential intestinal pathogens and commensal bacteria were imbalanced in patients with CP. These results partially agreed with those of Ciocan et al. [15] who found that the intestinal microbiota of patients with ACP had a lower bacterial diversity and different composition than that of alcoholic control patients. However, we did not observe increased abundances of Klebsiella, Enterococcus, and Sphingomonas in patients with ACP in the present study, possibly due to the relatively small number of ACP patients (n = 15, 21.1%) and different gender ratio between the two subgroups. A decreased abundance of Faecalibacterium found by Jandhyala et al. was also found in this study. However, the decreased abundance of the phylum Bacteroidetes and increased ratio of Firmicutes:Bacteroidetes in both the non-diabetic CP and diabetic CP groups they observed [14] was not found in this study. We also did not observe a reduced abundance of Ruminococcus bromii. Chronic Escherichia infections can induce cell cycle disorders, DNA damage, and inflammation. Members of Escherichia-Shigella are involved in the pathogenesis of inflammatory bowel disease and are associated with mucosal inflammation [20]. Moreover, the Escherichia-Shigella genus is a primary contributor to hepatic disease progression [21, 22]. LPS produced by members of the Escherichia-Shigella genus activates Toll-like receptor 4 to elicit inflammatory signaling [23]. High LPS levels activated the NF-κB pathway, induced the activation of proinflammatory cytokines (TNF-α, IL-6, and IL-1), and elicited inflammatory and oxidative damage [24, 25]. LPS challenge activated
16
pancreatic stellate cells and drastically increased collagen production [26]. Faecalibacterium is among the most abundant commensals in the human intestines. It is a key producer of butyrate and has been associated with anti-inflammatory activities. In addition, members of this genus improved gut barrier function by stimulating the synthesis of mucin and tight-junction proteins [27, 28]. The anti-inflammatory effects of Faecalibacterium in colitis mouse models were partially associated with metabolites that block NF-κB and IL-8 production [29]. In this research, a markedly decreased abundance of Subdoligranulum in the gut microbiomes of the CP group was also observed. Subdoligranulum can produce butyrate and enhance the intestinal health by providing energy for host cells and maintaining gut barrier integrity. Low Subdoligranulum abundance was previously correlated with increased incidences of inflammatory bowel disease [30, 31]. Diet, as well as geographic and ethnic origins can considerably affect the gut microbiota composition [32, 33]. Recruiting healthy family members as controls is an important approach that may have mitigated the influence of dietary and geographic differences on our results [34]. We only enrolled participants with Han Chinese ethnicity to prevent the effects of ethnic origins on our results. Patients with type-2 diabetes are characterized by a moderate degree of gut microbial dysbiosis [35]. Diabetes and glucose intolerance are common complications of CP. Given that islet cell loss distinguishes pancreatogenic diabetes from type-2 diabetes, diabetes mellitus resulting from CP is classified as pancreatogenic diabetes which is also known as type-3c diabetes [36]. In this study, we excluded CP patients with type-3c diabetes to prevent the potential effects of metabolic imbalances on intestinal microbiota. In addition to being involved in digestive proteases, pancreatic acinar cells also secrete antimicrobial peptides. A recent study showed that the secretion of antimicrobial peptides by pancreatic
17
acinar cells regulated the intestinal microbiota composition[37]. Levels of the antibacterial compound cathelicidin-related peptide (CRAMP) in pancreatic extracts of Orai1-/- mice were markedly reduced; supplementation of digestive enzymes failed to prevent intestinal dysbiosis, whereas CRAMP substitution substantially increased survival rates [37]. In dogs with exocrine pancreatic insufficiency, digestive enzyme supplementation did not prevent dysbiosis, further evidencing that digestive enzymes do not regulate the microbiome composition [38, 39]. However, Frost et al. [16] found that cathelicidin antimicrobial peptide levels were not significantly different in patients with reduced exocrine pancreatic function and control subjects. In this study, we found that gut microbiota dysbiosis was partly affected by pancreatic exocrine function in patients with CP. We speculate that the dysbiosis of intestinal microbiomes in patients with CP could be associated with the decreased levels of antimicrobial peptides secreted by pancreatic acinar cells. Isaiah et al. [38] found that the abundances of Lachnospiraceae and Ruminococcaceae were significantly lower in dogs with PEI. Members of the Lachnospiraceae family are butyrate producers. Butyrate production in the human gut could promotes Treg cell differentiation, which can ultimately suppress proinflammatory responses [40]. The current research showed that the Lachnoclostridium and Bifidobacterium genera correlated positively with FE1 levels, meaning that the abundance of beneficial bacteria decreased with the development of PEI. Our data indicated that for CP patients with PEI, supplementation with Lachnoclostridium and Bifidobacterium could be a viable option for regulating the dysbiosis. In summary, the strength of our study lies in its large cohort and optimized study design. These characteristics ensure that our study captured microbial diversity in patients with CP. Several limitations of this study must be addressed. First, stool samples provide an unselected representation of microbial ecology and microbiomes on the mucosal surface show marked heterogeneity. Second, in
18
contrast to quantitative metagenomics methods based on deep shotgun sequencing, the 16S rRNA gene sequencing method does not provide direct data related to functionally important changes in microbiomes. Third, the cross-sectional nature of the study did not enable us to determine the mechanisms and temporal associations of gut microbiota dysbiosis. Fourth, our study did not include animal experiments to examine probable mechanisms linking gut microbiota dysbiosis to CP development. The effects of changes in gut microbiota occurring in CP warrant further investigation. A logical extension of this study is to compare bacteria-derived metabolites of CP patients with that from HCs and identify candidate metabolites that could exacerbate pancreatic disease. Elucidating the regulatory mechanisms and functions of gut microbiomes may help improve our understanding of CP pathogenesis and could support potentially novel therapeutic options that could modify the gut microbiota of patients with CP.
Acknowledgements This study was supported by grants from National Natural Science Foundation of China (81670485 and 81800570), the Doctor Research Program of Naval Medical University (D-W Zou), and the Young Medical Talents Program and Natural Science Research Project of Universities in Jiangsu Province (C-H Zhou, QNRC2016863 and 18KJB320021). The authors declare that they have no conflicts of interest.
Author contributions
19
C-H Zhou, Y-T Meng, and J-J Xu designed the study. X Fang, J-L Zhao, and W-Zhou recruited the participants and collected the samples. J-H Zhao, J-C Han, W-B Zou, and K-X Wang performed the microbiota and medical analyses. L-H Hu and Z Liao coordinated the research. C-H Zhou, Y-T Meng, and L-Zhang wrote the manuscript. All authors critically revised the manuscript and approved the final version. References [1] Majumder S, Chari ST. Chronic pancreatitis. Lancet 2016; 387: 1957-66. [2] Yadav D, Slivka A. Managing chronic pancreatitis: the view from medical pancreatology. Am J Gastroenterol 2018; 113: 1108-10. [3] Hao L, Zeng XP, Xin L, Wang D, Pan J, Bi YW, et al. Incidence of and risk factors for pancreatic cancer in chronic pancreatitis: a cohort of 1656 patients. Dig Liver Dis 2017; 49: 1249-56. [4] Rooks MG, Garrett WS. Gut microbiota, metabolites and host immunity. Nat Rev Immunol 2016; 16: 341-52. [5] Marchesi JR, Adams DH, Fava F, Hermes GD, Hirschfield GM, Hold G, et al. The gut microbiota and host health: a new clinical frontier. Gut 2016; 65: 330-9. [6] Leal-Lopes C, Velloso FJ, Campopiano JC, Sogayar MC, Correa RG. Roles of commensal microbiota in pancreas homeostasis and pancreatic pathologies. J Diabetes Res 2015; 2015: 284680. [7] Sohail MU, Althani A, Anwar H, Rizzi R, Marei HE. Role of the gastrointestinal tract microbiome in the pathophysiology of diabetes mellitus. J Diabetes Res 2017; 2017: 9631435. [8] Tan C, Ling Z, Huang Y, Cao Y, Liu Q, Cai T, et al. Dysbiosis of intestinal microbiota associated with inflammation involved in the progression of acute pancreatitis. Pancreas 2015; 44: 868-75.
20
[9] Cen ME, Wang F, Su Y, Zhang WJ, Sun B, Wang G. Gastrointestinal microecology: a crucial and potential target in acute pancreatitis. Apoptosis 2018; 23: 377-87. [10] Memba R, Duggan SN, Ni Chonchubhair HM, Griffin OM, Bashir Y, O'Connor DB, et al. The potential role of gut microbiota in pancreatic disease: a systematic review. Pancreatology 2017; 17: 867-74. [11] Akshintala VS, Talukdar R, Singh VK, Goggins M: The gut microbiome in pancreatic disease. Clin Gastroenterol Hepatol 2019; 17: 290-5. [12] Capurso G, Signoretti M, Archibugi L, Stigliano S, Delle Fave G. Systematic review and meta-analysis: Small intestinal bacterial overgrowth in chronic pancreatitis. United European Gastroenterol J 2016; 4: 697-705. [13] Signoretti M, Roggiolani R, Stornello C, Delle Fave G, Capurso G. Gut microbiota and pancreatic diseases. Minerva Gastroenterol Dietol 2017; 63: 399-410. [14] Jandhyala SM, Madhulika A, Deepika G, Rao GV, Reddy DN, Subramanyam C, et al. Altered intestinal microbiota in patients with chronic pancreatitis: Implications in diabetes and metabolic abnormalities. Sci Rep 2017; 7: 43640. [15] Ciocan D, Rebours V, Voican CS, Wrzosek L, Puchois V, Cassard AM, et al. Characterization of intestinal microbiota in alcoholic patients with and without alcoholic hepatitis or chronic alcoholic pancreatitis. Sci Rep 2018; 8: 4822. [16] Frost F, Kacprowski T, Ruhlemann M, Bulow R, Kuhn JP, Franke A, et al. Impaired exocrine pancreatic function associates with changes in intestinal microbiota composition and diversity. Gastroenterology 2019; 156: 1010-15. [17] Tandon RK, Sato N, Garg PK, Consensus Study Group. Chronic pancreatitis: Asia-Pacific
21
consensus report. J Gastroenterol Hepatol 2002; 17: 508-18. [18] Loser C, Mollgaard A, Folsch UR: Faecal elastase 1: a novel, highly sensitive, and specific tubeless pancreatic function test. Gut 1996; 39: 580-6. [19] Cole JR, Wang Q, Cardenas E, Fish J, Chai B, Farris RJ, et al. The ribosomal database project: improved alignments and new tools for rrna analysis. Nucleic Acids Res 2009; 37: D141-5. [20] Xu J, Chen N, Wu Z, Song Y, Zhang Y, Wu N, et al. 5-aminosalicylic acid alters the gut bacterial microbiota in patients with ulcerative colitis. Front Microbiol 2018; 9: 1274. [21] Shen F, Zheng RD, Sun XQ, Ding WJ, Wang XY, Fan JG. Gut microbiota dysbiosis in patients with non-alcoholic fatty liver disease. Hepatobiliary Pancreat Dis Int 2017; 16: 375-81. [22] Woodhouse CA, Patel VC, Singanayagam A, Shawcross DL. Review article: the gut microbiome as a therapeutic target in the pathogenesis and treatment of chronic liver disease. Aliment Pharmacol Ther 2018; 47: 192-202. [23] Beutler B. Tlr4 as the mammalian endotoxin sensor. Curr Top Microbiol Immunol 2002; 270: 109-20. [24] Vonlaufen A, Spahr L, Apte MV, Frossard JL. Alcoholic pancreatitis: a tale of spirits and bacteria. World J Gastrointest Pathophysiol 2014; 5: 82-90. [25] Lamping N, Dettmer R, Schroder NW, Pfeil D, Hallatschek W, Burger R, et al. Lps-binding protein protects mice from septic shock caused by lps or gram-negative bacteria. J Clin Invest 1998; 101: 2065-71. [26] Sun L, Xiu M, Wang S, Brigstock DR, Li H, Qu L, et al. Lipopolysaccharide enhances tgf-beta1 signalling pathway and rat pancreatic fibrosis. J Cell Mol Med 2018; 22: 2346-56. [27] Arumugam M, Raes J, Pelletier E, Le Paslier D, Yamada T, Mende DR, et al. Enterotypes of the
22
human gut microbiome. Nature 2011; 473: 174-80. [28] Wrzosek L, Miquel S, Noordine ML, Bouet S, Joncquel Chevalier-Curt M, Robert V, et al. Bacteroides thetaiotaomicron and faecalibacterium prausnitzii influence the production of mucus glycans and the development of goblet cells in the colonic epithelium of a gnotobiotic model rodent. BMC Biol 2013; 11: 61. [29] Sokol H, Pigneur B, Watterlot L, Lakhdari O, Bermudez-Humaran LG, Gratadoux JJ, et al. Faecalibacterium prausnitzii is an anti-inflammatory commensal bacterium identified by gut microbiota analysis of crohn disease patients. Proc Natl Acad Sci U S A 2008; 105: 16731-6. [30] Deaver JA, Eum SY, Toborek M. Circadian disruption changes gut microbiome taxa and functional gene composition. Front Microbiol 2018; 9: 737. [31] Thomas LV, Ockhuizen T, Suzuki K. Exploring the influence of the gut microbiota and probiotics on health: a symposium report. Br J Nutr 2014; 112: S1-18. [32] Escobar JS, Klotz B, Valdes BE, Agudelo GM. The gut microbiota of colombians differs from that of americans, europeans and asians. BMC Microbiol 2014; 14: 311. [33] Graf D, Di Cagno R, Fak F, Flint HJ, Nyman M, Saarela M, et al. Contribution of diet to the composition of the human gut microbiota. Microb Ecol Health Dis 2015; 26: 26164. [34] Song SJ, Lauber C, Costello EK, Lozupone CA, Humphrey G, Berg-Lyons D, et al. Cohabiting family members share microbiota with one another and with their dogs. Elife 2013; 2: e00458. [35] Qin J, Li Y, Cai Z, Li S, Zhu J, Zhang F, et al. A metagenome-wide association study of gut microbiota in type 2 diabetes. Nature 2012; 490: 55-60. [36] Rickels MR, Bellin M, Toledo FG, Robertson RP, Andersen DK, Chari ST, et al. Detection, evaluation and treatment of diabetes mellitus in chronic pancreatitis: recommendations from
23
PancreasFest 2012. Pancreatology 2013; 13: 336-42. [37] Ahuja M, Schwartz DM, Tandon M, Son A, Zeng M, Swaim W, et al. Orai1-mediated antimicrobial secretion from pancreatic acini shapes the gut microbiome and regulates gut innate immunity. Cell Metab 2017; 25: 635-46. [38] Isaiah A, Parambeth JC, Steiner JM, Lidbury JA, Suchodolski JS. The fecal microbiome of dogs with exocrine pancreatic insufficiency. Anaerobe 2017; 45: 50-58. [39] Rubinstein E, Mark Z, Haspel J, Ben-Ari G, Dreznik Z, Mirelman D, et al. Antibacterial activity of the pancreatic fluid. Gastroenterology 1985; 88: 927-32. [40] Singh N, Gurav A, Sivaprakasam S, Brady E, Padia R, Shi H, et al. Activation of gpr109a, receptor for niacin and the commensal metabolite butyrate, suppresses colonic inflammation and carcinogenesis. Immunity 2014; 40: 128-39.
24
Figure legends Fig. 1. Study protocol and allocation of participants to different groups, based on inclusion and exclusion criteria
Fig. 2. Structure and composition of the gut microbiomes of patients with CP. Rarefaction curves tended to approach the saturation plateau at the operational taxonomic unit (OTU) level (A). Principal coordinates analysis (PCoA) with weighted and unweighted unifrac testing were used to compare the microbiota compositions of CP with HC groups. ANOSIM revealed significant differences in the structures of both groups (r = 0.0336, P = 0.002, unweighted unifrac; r = 0.0587，P = 0.001, weighted unifrac). The axes represent the three dimensions explaining the greatest proportion of variance in the communities. Each symbol represents a sample (B, C). The top 15 fecal microbiota with significantly different abundances between the CP and HC groups at the phylum (D) and genus (E) levels are shown. Correlation analysis of the 15 significantly different genera identified in the fecal microbiomes of HCs and patients with CP (F). Cladogram showing taxa with the largest differences in relative abundances between the CP and HC groups. The circle sizes in the cladogram plot are proportional to the bacterial abundances. Going from the inside to the outside, the circles represent the phylum, class, order, family, and genus. Only taxa with an LDA score of >3 and P < 0.05 in the Wilcoxon signed-rank test are shown. (G). A logarithmic LDA-score cutoff of 3.0 was used to identify important taxonomic differences between the fecal microbiomes of HCs and patients with CP (H).
Fig. 3. Structures and compositions of the gut microbiomes of CP patients with or without PEI. Principal coordinates analysis (PCoA) with weighted and unweighted unifrac testing were used to
25
compare the microbiota compositions of CP patients with or without PEI. ANOSIM revealed significant differences in the structures of both subgroups using unweighted unifrac test (r = 0.1064, P = 0.004), whereas no significant differences were observed using weighted unifrac test (r = 0.0177, P = 0.197) (A, B). Top 15 fecal microbiota in abundance of the CP patients with PEI and without PEI on the phylum (C) and genus (D) levels are shown. Cladogram showing taxa with the largest differences in relative abundances between CP patients with or without PEI. Circle sizes in the cladogram plot are proportional to bacterial abundances. Going from the inside to the outside, the circles represent the phylum, class, order, family, and genus. Only taxa with an LDA score >3 and P < 0.05 in the Wilcoxon signed-rank test are shown (E). A logarithmic LDA-score cutoff of 3.0 was used to identify important taxonomic differences between the fecal microbiomes of CP patients with or without PEI (F).
Fig. 4. Predicted Kyoto Encyclopedia of Genes and Genomes (KEGG) level 3 pathways, i.e., lipopolysaccharide (LPS) biosynthesis and bacterial invasion of epithelial cells, which were enriched in patients with CP (A, B). Pathways associated with LPS biosynthesis and bacterial invasion of epithelial cells positively correlated with Escherichia-Shigella OTU2627 (r = 0.454 and 0.694, respectively; P < 0.001), which was enriched in the gut microbiomes of patients with CP (C, D). Correlation analysis was used to evaluate the relationship between FE1 levels and the top 15 most abundant genera in patients with CP. Bifidobacterium and Lachnoclostridium correlated positively with FE 1 (r = 0.2616 and 0.2486, respectively; P < 0.05) (E, F) .
26
Supplemental Fig.1. Structures and compositions of the gut microbiomes of CP_alcholic group and CP_idiopathic group. Principal coordinates analysis (PCoA) with weighted and unweighted unifrac test were used to compare the microbiota compositions of the two subgroups. ANOSIM revealed no significant differences in the structures of fecal microbiota in both subgroups using unweighted and weighted unifrac test (r = -0.0799, P = 0.816 and r = 0.0083, P = 0.428) (A, B). Top 15 fecal microbiota in CP_alcholic group and CP_idiopathic group on the phylum (C) and genus (D) levels are shown. Cladogram showing taxa with the largest differences in relative abundances between CP_alcholic group and CP_idiopathic group. Circle sizes in the cladogram plot are proportional to bacterial abundances. Going from the inside to the outside, the circles represent the phylum, class, order, family, and genus. Only taxa with an LDA score >3 and P < 0.05 in the Wilcoxon signed-rank test are shown (E). A logarithmic LDA-score cutoff of 3.0 was used to identify important taxonomic differences between the fecal microbiomes of the two subgroups (F).
Tables
Table 1 Demographic and Clinical data of healthy controls (HCs) and patients with chronic pancreatitis (CP) CP (n = 71)
HC (n = 69)
P value
Male/female
41/30
29/40
0.06
Age, median (mean ± SD)
44 ± 11
47 ± 10
0.07
BMI, median (min–max)
21.0 (18.3–28.0)
22.3 (18.7–29.8)
0.26
Disease duration, median
24 (0.5–276)
-
-
Dilated pancreatic duct
68
-
-
Pancreatic duct stone
68
-
-
Pancreatic atrophy
40
-
-
Pancreatic pseudocyst
11
-
-
Surgery history
6
-
-
ERCP history
39
-
-
ESWL history
27
-
-
Hemoglobin level (g/L), (mean ± SD)
141 ± 10
142 ± 15
0.87
FBG, median (min–max)
5.3 (4.2–6.8)
4.7 (4.1–6.1)
<0.0001
(min–max)
Data are expressed as the mean ± SD or median (25th–75th percentiles) in accordance with the normality of distribution. BMI, body–mass index; ERCP, endoscopic retrograde cholangiopancreatography; ESWL, extracorporeal shock wave lithotripsy
Table 2 Richness and diversity indexes of the gut microbiotas of HCs and patients with CP Parameter
CP
HCs
P value
Observed OTUs
218.76 ± 63.36
283.09 ± 146.79
0.000934*
Shannon
3.09 ± 0.60
3.46 ± 0.70
0.001277*
Simpson
0.13 ± 0.083
0.096 ± 0.071
0.02527*
ACE
297.47 ± 96.40
376.24 ± 200.73
0.00348*
Chao1
285.56 ± 83.70
369.13 ± 194.04
0.00113*
PD
19.53 ± 4.92
24.63 ± 13.24
0.00284*
Coverage
0.997 ± 0.001
0.996 ± 0.003
0.00744*
The diversities of fecal microbiomes in patients with CP decreased relative to those in HCs. α-diversity was estimated based on the observed OTUs and the Shannon, ACE, Chao 1, and phylogenetic diversity (PD) indexes. The α-diversities of the fecal microbiomes of patients with CP showed statistically significant reductions relative to those in HCs. In contrast, the Simpson index of the gut microbiotas of patients with CP was significant higher than that of HCs. Data are expressed as the mean ± SD. Student’s t-test was used to identify statistically significant differences in each variable between the HC and CP groups. P < 0.05 was considered to reflect statistical significance.