1. Introduction

2. Material and Method

2.1. Study Design

This study focused on the causal relationships between intestinal microbes and CRC and its subtypes (such as colon cancer, rectal cancer, and rectal anastomosis cancer). Utilizing publicly available GWAS data and a European CRC cohort 16S rRNA dataset, the relationship between intestinal flora traits and CRC across diverse pathological sites and stages was explored through two-sample univariate MR and MR-Bayesian Model Averaging (MR-BMA) methods [12]. The MR-BMA method is particularly suitable for identifying causal factors in high-correlation data, such as intestinal microbes associated with CRC [13]. Moreover, the NetMoss method was employed to construct a multi-population cohort and validate the causal relationships between intestinal microbes and CRC. The overall study design aimed to deepen the understanding of the potential role of intestinal microbes in the development of CRC.

Figure 1. The MR Model 1.

Figure 1. The MR Model 1.

2.2. Data Source

2.2.1. Intestinal Microbe Dataset

GWAS data from the Dutch LifeLines cohort were collected, involving 207 intestinal microbial taxa and 205 functional pathways. LifeLines is a large-scale prospective study with complete fecal and phenotypic information from 7,738 participants, covering 5,584,686 SNPs. Based on further metagenomic sequencing data, an in-depth analysis of microbial composition was conducted and subjected to rigorous quality control.

2.2.2. CRC Dataset

GWAS summary data from the Finnish Biobank (FinnGen) were adopted, encompassing CRC across various pathological locations and stages. The specific data include CRC (CRC, IV=221,814), colon malignancy (CC, IV=220,595), rectal malignancy (CR, IV=219,870), benign colorectal tumors (BCR, IV=228,104), benign colon tumors (BC, IV=not mentioned), and benign rectal tumors (BR, IV=220,900). The FinnGen project aims to investigate the genomes of 500,000 Finns and data from the National Health Registry, currently covering 224,737 genotypic and phenotypic participants. The study identified genome-wide significant associations at 2,491 independent loci based on 1,932 clinical endpoints defined by the National Health Registry. Researchers applied mixed-effects logistic regression to adjust for age, sex, ten principal components, and genotype batch to ensure the accuracy of results.

Table 1. Detailed Information of GWAS Data.

Table 1. Detailed Information of GWAS Data.

Type	Phenotype	Sample Size	GWAS ID	Institution	Race
CRC	colorectal malignant tumor	221,814	fmn-b-c3	FinnGen	European
	colon malignancy	220,595	fmn-b-c3	FinnGen	European
	rectal malignancy	219,870	fmn-b-c3	FinnGen	European
	benign colorectal tumors	228,104	finn-b-CD2	FinnGen	European
	benign colon tumors	218,792	finn-b-CD2	FinnGen	European
	benign rectal tumors	220,900	finn-b-CD2	FinnGen	European
Intestinal microbes	Microbial taxa	7,738	NA	NA	European
	Microbial functional pathways	7,738	NA	NA	European
Other diseases	Crohn's disease	211,107, 211,268	finn-b-CHRONLARGE, finn-b-CHRONSMALL	FinnGen	European
	Ulcerative colitis	214,620	fmn-b-KLL_ULCER	FinnGen	European
	Irritable bowel syndrome	187,028	fmn-b-KLL_IBS	FinnGen	European
	Non-alcoholic fatty liver disease	218,792	fmn-b-NAFLD	FinnGen	European
	Type 2 diabetes	215,654	fmn-b-E4_DM2	FinnGen	European
Dietary habits	Alcohol consumption status	360,726	ukb-d-2	NA	European
	Bread intake	452,236	ukb-b-11348	MRC-IEU	European
	Grain type: Biscuits	299,898	ukb-d-14682	NA	European
	Lamb/Lamb Intake	460,006	ukb-b-14179	MRC-IEU	European
	Ferritin	23,986	ieu-a-1050	GIS	European
	Liver intake	64,944	ukb-b-6373	MRC-IEU	European
	The type of milk used	360,806	ukb-d-14186	NA	European
	Minerals and other dietary supplements	336,314	ukb-a-495	Neale Lab	European
	Single layer pastry intake	64,949	ukb-b-2024, ukb-b-11189	MRC-IEU	European
	Vitamin and trace elements	335,591, 460,351	ukb-a-464, ukb-b-15175	Neale Lab, MRC-IEU	European

2.5. Two-Sample MR Analysis

MR analysis software (TwoSampleMR) was used to verify the specificity of microbial biomarkers closely related to the development of CRC. First, GWAS summary data on six non-CRC diseases (including Crohn's disease of the colon, Crohn's disease of the small intestine, ulcerative colitis, irritable bowel syndrome, non-alcoholic fatty liver disease, and type 2 diabetes) were obtained from the FinnGen database. Subsequently, detailed MR analysis was conducted using these data along with previously identified microbial traits associated with CRC. Parameters were first set, including but not limited to the type of MR estimates (e.g., IVW), methods for adjusting effect sizes, and criteria for determining significance levels. Corrected P-values were then utilized to assess statistical significance, with a threshold set at P < 0.2, aiming to clearly identify microbial biomarkers specifically related to the development of CRC [17].

Figure 2. The MR Analysis Framework.

Figure 2. The MR Analysis Framework.

3. Result

3.2. Univariate MR Analysis

After selecting appropriate IVs, various univariate MR methods were applied to explore the potential causal relationships between intestinal microbial markers and CRC across different pathological locations and stages. Through IVW (M), Wald ratio, and MR Egger, multiple groups of intestinal flora diversity were identified with significant associations with CRC and its subtypes (CRC, colon cancer, rectal cancer, rectosigmoid colon cancer [BCR], unspecified rectal cancers [BC], and unspecified rectal cancers [BR]) (P < 0.05). Notably, 12 and 20 microbial traits having significant causal relationships with CRC and its subtypes were discovered (IVW(M) < 0.2), respectively, revealing differences in microbial markers between the colon and rectum.

A total of 12 key microbial traits were identified in CRC, including but not limited to Clostridium, Bacteroides, and Bifidobacterium. These floras significantly influenced the occurrence and progression of CRC through distinct metabolic activities and interactions with the host. For colon cancer, major associated floras were concentrated in Ruminococcus and Lactobacillus, which may play a key regulatory role in the local environment of the colon. Regarding rectal cancer and its subtypes, more complex distribution patterns were observed, such as significant associations between benign colorectal tumors and Akkermansia and potential links between unspecified rectal cancers and benign rectal tumors with Prevotella and Fusobacterium. These findings provide valuable insights into the mechanisms underlying CRC.

The univariate MR analysis identified 19 to 29 intestinal microbial taxa that had potential causal relationships with CRC, colon cancer, rectal cancer, rectosigmoid colon cancer, unspecified rectal cancers, and BR. Among them, Clostridium, whose metabolites influence intestinal immunity and inflammation, was related to the overall risk of CRC; Bacteroides, promoting intestinal health through polysaccharide degradation, was significantly associated with the subtypes of colon cancer; Bifidobacterium, a probiotic that regulates intestinal flora balance, showed potential benefits for various CRC subtypes. Additionally, Ruminococcus exhibited prominent characteristics in rectal cancer and rectosigmoid colon cancer, affecting the local environment through cellulose degradation; Lactobacillus was commonly linked to the subtypes of CRC, enhancing intestinal barrier function.

In order to probe into the causal relationships between these microbes and disease outcomes, a potential biomarker ranking method based on MR-BMA was utilized, with a prior probability of 0.1 as a benchmark for causal importance ranking. Upon stringent screening and pruning, 14 to 23 microbial taxa were analyzed and measured by 87 to 157 genetic variants as IVs. Preliminary MR-BMA analysis revealed significant effects of specific SNPs (e.g., rs12736307 and rs76321722) on colon cancer and rectal cancer outcomes. Through further adjustment for potential confounders (such as age, sex, and lifestyle), more robust and reliable results were obtained. The MR-BMA analysis indicated that the previously identified key SNPs maintained significant associations with CRC subtypes, with their effect sizes and directions remaining substantially unchanged.

Table 2. MR-BMA Ranking of MIP and MACE for the Microbial Biomarkers of CRC.

Table 2. MR-BMA Ranking of MIP and MACE for the Microbial Biomarkers of CRC.

Microbial Biomarker	MIP	MACE	Pval
Pseudoflavonifractor	0.906	0.257	0.020
Streptococcus thermophilus	0.562	-0.106	0.020
Coprococcus catus	0.433	-0.114	0.010
Parabacteroides distasonis	0.076	0.011	0.347
Erysipelotrichaceae	0.065	-0.010	0.535
Sutterella wadsworthensis	0.056	-0.009	0.802
Streptococcus	0.055	-0.003	0.455
Roseburia inulinivorans	0.029	0.002	1.000
Flavonifractor	0.028	0.002	0.960
Gammaproteobacteria	0.026	-0.001	1.000

Table 3. MR-BMA Ranking of MIP and MACE for Colonic Malignant Tumors.

Table 3. MR-BMA Ranking of MIP and MACE for Colonic Malignant Tumors.

Microbial Biomarker	MIP	MACE	Pval
Gammaproteobacteria	0.204	0.049	0.089
Eubacterium ventriosum	0.169	-0.025	0.030
Holdemania unclassified	0.148	-0.020	0.059
Sutterella wadsworthensis	0.136	-0.024	0.158
Prevotella copri	0.088	0.012	0.188
Roseburia inulinivorans	0.082	0.012	0.198
Betaproteobacteria	0.078	-0.010	0.366
Paraprevotella xylaniphila	0.071	-0.007	0.188
Sutterellaceae	0.071	-0.008	0.495
Parabacteroides goldsteinii	0.064	0.005	0.079

3.3. Two-Sample MR Analysis

Through the results of two-sample univariate MR analysis and potential biomarker ranking, several key microorganisms were identified to have potential relationships with CRC and its subtypes. For colon cancer, emopMw, as a unique microbial strain, exhibited significant protective effects by modulating the intestinal microenvironment, promoting beneficial bacterial growth, and inhibiting the proliferation of harmful bacteria [18]. Its protective effects primarily stem from reducing inflammatory responses and oxidative stress in the intestine, thereby maintaining the health of colon cells [19].

In the subgroup of rectal cancer, members of the Erysipelotrichaceae family are particularly notable. By enhancing intestinal barrier function and reducing local inflammatory responses, they protect rectal health [20]. This protective effect is not only evident in the prevention of CRC but also significantly lowers the risk of inflammatory bowel disease (IBS), further validating their extensive benefits for intestinal health.

In contrast, for CRC subtypes, including rectosigmoid colon cancer (BCR) and basal cell carcinoma (BC), microbial species such as Orato were identified as potential risk factors. These microbes may increase the risk of specific CRC subtypes by promoting inflammation, disrupting the intestinal barrier, or producing harmful metabolites [21].

Figure 3. The scatter plot of five MR models of intestinal microbes causally related to CRC.

Figure 3. The scatter plot of five MR models of intestinal microbes causally related to CRC.

The two-sample univariate MR analysis was conducted to further explore whether these microbes could serve as specific biomarkers for CRC. The results showed that Erysipelotrichaceae significantly reduced the risk of IBS (OR = 0.897, 95% CI = 0.844-0.954, P = 0.0006), indicating its potentially positive role in preventing IBS. Additionally, Streptococcus thermophilus, Eubacterium ventriosum, Streptococcus, Coprococcus sp. strain ART55-1, Eubacterium siraeum, Bacteroides ovatus, and Bifidobacterium ado/cew were identified as specific biomarkers associated with CRC. These findings provide novel perspectives and potential targets for the diagnosis and treatment of CRC.

4. Discussion

References

1. Zhao, W.; Yin, Z.; Wang, Y.; Li, W. Interpretation of 2024 American Cancer Statistics Report and Comparison of Cancer Epidemiology between China and the U.S. Cancer Research on Prevention and Treatment, 1–12.
2. Tang, M.; Wang, X.; Li, X. . Research Progress on Statins in the Prevention and Treatment of Colorectal Cancer. The Journal of Practical Medicine 2024, 40, 2041–2046. [Google Scholar]
3. Wang, G.; Zhang, G.; Tang, J.; Zhang, Y.; Yang, Y.; Wang, J. The Impact of SOX4 on Immune Evasion and Cell Migration in Colorectal Cancer Cells by Targeting miR-17 Expression Levels. Laboratory Medicine and Clinic 2024, 21, 1825–1830. [Google Scholar]
4. Yi, D.; Tian, X. Screening Results of Colorectal Cancer in Fecal DNA Positive Cases in the Wuchang Area of Wuhan from 2021 to 2022. Journal of Public Health and Preventive Medicine 2024, 35, 103–106. [Google Scholar]
5. Liu, Z.; Zhang, B.; Li, S. Application Effects of Bifidobacterium Tri-Active Capsules in Postoperative Chemotherapy for Colorectal Cancer. Chinese Journal of Clinical Rational Drug Use 2024, 17, 8–11. [Google Scholar]
6. He, R.; Zhang, Y.; Li, P.; Song, G.; Li, A.; Wang, Y. Progress in Imaging Assessment of Perineural Invasion in Rectal Cancer. Chinese Journal of Digestion and Medical Imageology (Electronic Edition) 2024, 14, 378–381. [Google Scholar]
7. Chen, L.; Lu, N.; Chen, H. Cardiovascular Disease Risk Factors Associated with Prognosis in Esophageal and Colorectal Cancer Patients after Surgery. Chinese Journal of Hypertension (Bilingual Edition), 1–7.
8. Zhao, J. Study on the Effects of Different Enteral Nutritional Preparations on Nutritional Status and Gastrointestinal Function in Colorectal Cancer Patients after Radical Surgery. The Journal of Medical Theory and Practice 2024, 37, 2228–2230. [Google Scholar]
9. Hou, W.; Li, Y.; Bi, Y.; Han, L.; Ren, X. Research Progress on Lipid Metabolism Reprogramming in Colorectal Cancer. Chinese Journal of Gerontology 2024, 44, 3323–3327. [Google Scholar]
10. Wang, W.; Xu, S. Research Progress on Correlation between intestinal flora and Colorectal Cancer. Journal of Anhui Medical University 2024, 59, 1285–1289. [Google Scholar] [CrossRef]
11. Liu, Y.; Xu, H.; Ji, G. Bile Acids Driving Gastrointestinal Inflammation-Cancer Transformation and Traditional Chinese Medicine Prevention Strategies. Bulletin of National Natural Science Foundation of China 2024, 38, 461–468. [Google Scholar]
12. Han, W.; Fan, Z.; Li, Y.; Miao, L. Changes in Intestinal Flora in Colorectal Cancer Patients and Their Relationship with Serum Inflammatory Factors and Pathological Staging. Clinical Journal of Medical Officer 2024, 52, 641–643. [Google Scholar]
13. Xu, J.; Han, L.; Guo, K. Mendelian Randomization Analysis of the Causal Relationship between intestinal flora and Systemic Sclerosis. Progress in Microbiology and Immunology, 1–8.
14. Guo, Y.; Cui, Y. Exploring the Causal Relationship between intestinal flora and Insomnia Based on Mendelian Randomization. Chinese General Practice, 1–8.
15. Liang, J.; Yang, S.; Wang, H. Exploring the Causal Relationship between Intestinal Flora and Childhood Asthma Based on Mendelian Randomization. West China Medical Journal 2024, 39, 553–560. [Google Scholar]
16. Guo, Y.; Cui, Y.; Kong, Y. Causal Relationship between Intestinal Flora and Cluster Headache: A Two-Sample Mendelian Randomization Analysis. Practical Journal of Cardiac Cerebral Pneumal and Vascular Disease 2024, 32, 43–48. [Google Scholar]
17. Ma, J.; Zhou, Y.; Chen, W. Two-Sample Bidirectional Mendelian Randomization Analysis of Causal Relationship between Gastroesophageal Reflux Disease and Migraine. Journal of Xi'an Jiaotong University (Medical Edition) 2024, 45, 262–270. [Google Scholar]
18. Chai, J.; Li, S.; Li, W. intestinal flora and Drug-Related Avascular Necrosis: A Two-Sample Bidirectional Mendelian Randomization Analysis. Chinese Journal of Tissue Engineering Research 2024, 28, 4325–4331. [Google Scholar]
19. Wu, M.; An, Z.; He, Y. Exploring the Causal Relationship between intestinal Microbes and Non-Alcoholic Fatty Liver Disease Using Mendelian Randomization. Acta Universitatis Medicinalis Anhui 2024, 59, 236–242. [Google Scholar]
20. Wang, X.; Chai, R. The Causal Relationship between Intestinal Flora and Allergic Diseases: Insights from Bidirectional Two-Sample Mendelian Randomization Analysis. Chinese Journal of Allergy & Clinical Immunology 2023, 17, 579. [Google Scholar]
21. Yu, L.; Zhang, M. Progress in the Application of Mendelian Randomization Analysis in Diabetic Retinopathy. Chronic Pathematology Journal 2023, 24, 1790–1793. [Google Scholar]
22. Wei, Y.; Lan, Y.; Huang, X. A Study on the Association between Intestinal Flora and Frailty Based on Two-Sample Mendelian Randomization Analysis. Guangxi Medical Journal 2023, 45, 2880–2885. [Google Scholar]
23. Dong, X.; Xue, J.; Liang, X. The Application of Mendelian Randomization in the Study of Risk Factors for Primary Carcinoma of Liver. Infectious Disease Information 2023, 36, 473–476. [Google Scholar]
24. Ren, Q.; He, W. Two-Sample Mendelian Randomization Analysis of Causal Relationship between Intestinal Flora and Lung Cancer. Chinese Journal of Clinical Thoracic and Cardiovascular Surgery 2023, 30, 1618–1627. [Google Scholar]
25. Wang, C.; Sun, X.; Chen, W. The Relationship between Type 2 Diabetes and Sepsis: A Two-Sample Mendelian Randomization Study Based on GWAS Database. Chinese Journal of Critical Care Medicine 2023, 43, 794–799. [Google Scholar]
26. Wang, D. Mendelian Randomization Study of intestinal flora and Metabolic Cardiovascular Disease. Central South University, 2023. [Google Scholar]
27. Lü, B. Systematic Genetic Studies of Intestinal Flora and Associated Diseases. Huazhong Agricultural University, 2020. [Google Scholar]