In recent years, with the development of metagenomic studies, there has been a growing awareness of the important role of the gastrointestinal microbiome in public health and diseases, including digestive system diseases such as inflammatory bowel disease [28], colorectal cancer [29], irritable bowel symptoms [30], and parenteral diseases such as diabetes [31, 32], coronary heart disease [33, 34], and Alzheimer's disease [35, 36]. An imbalance of the gastrointestinal microbiome occurs early in the development of SAP. In addition to the gastrointestinal microbiome, metabolites, such as short-chain fatty acids (SCFAs), also influence the progression of AP [37]. In this study, we performed two-sample MR analyses to assess the relationship between the gastrointestinal microbiome and AP, using aggregated statistics of the gastrointestinal microbiome from the largest GWAS meta-analysis conducted by the MiBioGen Consortium and aggregated statistics of AP published by the UK Biobank and FinnGen consortium. We found that the abundance of Eubacterium rectale group was positively correlated with the severity of AP. Holdemania and Terrisporobacter had a protective effect on AP patients. The occurrence and development of AP were positively correlated with the accumulation of Allisonella, Anaerotruncus and Paraprevotella in the gastrointestinal tract. On the contrary, the occurrence of AP had a negative impact on the colonization of Erysipelotrichaceae UCG003, Family XIII AD3011 group, Haemophilus, Intestinibacter, Turicibacter and Tyzzerella 3.
It is well known that cholelithiasis and cholecystitis secondary to cholelithiasis are common pathogenesis factors of biliary pancreatitis. Another MR analysis that explored the relationship between the gastrointestinal microbiome and cholelithiasis showed that Holdemania had a protective effect against cholelithiasis [38]. Our research found that Holdemania was not only protective against cholelithiasis, it also plays a protective role in the occurrence and development of AP disease. This connection suggested that Holdemania may be a direction for the treatment of biliary pancreatitis. Holdemania also plays a role in other diseases, for example, people who drink excessively have an increased abundance of Holdemania in the gut [39]. Alcohol consumption is one of the predisposing factors of AP, and the increased abundance of Holdemania in patients with excessive alcohol consumption also indicates that Holdemania has a protective effect in the development of AP. Holdemania was associated with clinical indicators of impaired lipid and glucose metabolism [40], suggesting that Holdemania may be associated with AP prognosis. In addition to Holdemania, Terrisporobacter is also a bacterial genus that we found to have a protective effect on AP patients through SNPs. Hyperlipidaemia is one of the factors inducing AP, and the incidence of hyperlipidemic AP is characterized by young age. One study confirmed that the abundance of Terrisporobacter in the gut was positively correlated with low-density lipoprotein cholesterol (LDL-C) and total cholesterol (TC) [41]. We speculated that there may be a genetic susceptibility factor for the association between AP and hyperlipidaemia, and the association of Terrisporobacter with both of these factors confirms our conjecture in another way. The abundance of Eubacterium rectale group differed among patients with chronic pancreatitis. Our results suggest that the variation of SNPs in Eubacterium rectale group may cause it to increase the severity of AP, but there are no in-depth reports on the mechanism of action of Eubacterium rectale group and AP related studies.
A study of chronic radiation proctitis (CRP) confirmed that the abundance of Allisonella was elevated in patients with CRP combined with haematochezia, adversely affecting intestinal function [42]. Although no direct evidence has been found between AP and Allisonella in the existing research, our study suggested that the occurrence of AP was positively correlated with the abundance of Allisonella, so we had reason to suspect that AP could affect gastrointestinal function through changing the colonization of Allisonella at the genetic susceptibility level. There was no direct evidence that Anaerotruncus is associated with the course of AP for the moment, but several previous studies had demonstrated a correlation between Anaerotruncus and the development and prognosis of digestive diseases [43], and the impact of AP on intestinal function and the intestinal barrier cannot be ignored. Anaerotruncus correlated with cholesterol diet-induced nonalcoholic fatty liver disease (NAFLD)-associated hepatocellular carcinoma (HCC), and a study demonstrated a progressive increase in the abundance of Anaerotruncus during the NAFLD-HCC disease process [44]. The accumulation of Anaerotruncus in the intestine increased the risk of anal fistula, and Anaerotruncus was increased in unresectable pancreatic ductal adenocarcinoma (PDAC) [45]. The overexpression of granulocyte-macrophage colony-stimulating factor (GM-CSF) in hepatocellular carcinoma could reduce the abundance of Anaerotruncus in the intestine [46]. Moreover, the gastrointestinal microbiome and fecal metabolites induced by GM-CSF can participate in pathways related to reducing the inflammatory response, biotin metabolism and intestinal barrier dysfunction, which provided us with ideas for treating intestinal dysfunction caused by AP. Researchers isolated the Paraprevotella strain from the fecal microbiome of healthy humans and confirmed that it is an effective trypsin-degrading symbiotic [47]. Paraprevotella is a recently discovered genus of Prevotellaceae that contains only two species, P. clara and Paraprevotella xylaniphila. Mechanistically, Paraprevotella plays a role in promoting trypsin autolysis by the mechanism of action of recruiting trypsin to the bacterial surface by a polysaccharide anchor protein dependent on the type IX secretion system. This study also suggested that symbiotic colonization via trypsin degradation may help maintain intestinal homeostasis and prevent pathogen infection. As we all know, abnormal activation of trypsin is one of the pathogenic factors of AP, beyond that, the occurrence of infection necrotizing pancreatitis (INP) after AP is closely related to damage to the intestinal barrier. Taken together, these findings indicate that the occurrence and development of AP are negatively correlated with the abundance of Paraprevotella, which may suggest that Paraprevotella has a positive effect on the prevention of INP in patients with later-stage of AP.
We know that diet and reduced gut microbiome diversity are associated with AP [48]. One study showed that high-chicken-protein (HFHCH) or high-pork-protein (HFHP) diets increased the abundance of the Family XIII AD3011 group in the cecum and colon, which may be related to the production of skatole and indole [49]. In a mouse model of DSS-induced colitis, cepharanthine (CEP) reduced macrophage infiltration and ACOD1 expression in colon tissue, and correlation analysis also revealed that the Family-XIII-AD3011 group was positively correlated with ACOD1 expression levels [50]. A Previous study provided ideas for regulating the effects of the gastrointestinal microbiome on the course of AP disease. Erysipelotrichaceae UCG003 was enriched in postcholecystectomy diarrhea (PCD) and may be associated with bile acid metabolism [51]. Moreover, there was a study confirmed that the gastrointestinal microbiome was enriched to different degrees in colorectal cancer (CRC) with different degrees of differentiation, and Erysipelotrichaceae UCG003 was enriched in moderately differentiated CRC [52]. However, in addition to changes in gut microbes, oral microbes were also affected. Changes in microbial composition occurred with increasing severity of AP, for example, the abundance of beneficial bacteria such as Neisseria, Haemophilus, and Gemella is reduced in MSAP and SAP compared to MAP [53]. The abundance of Turicibacter in a mouse model of chronic pancreatitis was positively correlated with the degree of pancreatic fibrosis [54]. Turicibacter can also be used to distinguish alcoholic fatty liver disease from nonalcoholic fatty liver disease [55]. At present, although there was no direct evidence showing the role of Family-XIII-AD3011, Erysipelotrichaceae UCG003 and Turicibacter in the pathogenesis of AP, they have been confirmed to be correlated with inflammation and immunity in a variety of digestive diseases. This suggests that we can include them in the future research direction. There has been relatively little research on Tyzzerella 3, but a decreased abundance of Tyzzerella 3 has been reported to be associated with acute myocardial infarction [56].
This study has several advantages. First, we determined the causal relationship between the gastrointestinal microbiome and AP through MR analysis to exclude the interference of confounding factors and reverse causal inference. Second, the genetic variation in the gastrointestinal microbiome was obtained from the largest GWAS meta-analysis available, ensuring the strength of the instruments used for MR analysis. Horizontal pleiotropy was detected and excluded by MR-PRESSO and MR-Egger regression intercept tests. In addition, we used two-sample MR analysis to avoid bias due to overlapping exposure and outcome data.
However, aggregated statistics of AP patients were used in the MR analysis, which has some limitations due to the lack of raw data and the inability to perform subgroup analysis. Secondly, in order to perform sensitivity analysis and horizontal pleiotropy testing, more genetic variants need to be included as instrumental variables, so the SNPs used in the analysis did not reach the most significant threshold for traditional GWAS (P < 5×10-8). In addition, the data included in this study are mainly for the European population, so they are affected by ethnic genetic differences, and the research results cannot be fully applied to populations in Asia and the Americas; therefore, a more comprehensive analysis is needed after improving the GWAS data.