Association Between Sulfur-Metabolizing Bacterial Communities in Stool and Risk of Distal Colorectal Cancer in Men

Abstract

Background & aims: Sulfur-metabolizing microbes, which convert dietary sources of sulfur into genotoxic hydrogen sulfide (H₂S), have been associated with development of colorectal cancer (CRC). We identified a dietary pattern associated with sulfur-metabolizing bacteria in stool and then investigated its association with risk of incident CRC using data from a large prospective study of men.

Methods: We collected data from 51,529 men enrolled in the Health Professionals Follow-up Study since 1986 to determine the association between sulfur-metabolizing bacteria in stool and risk of CRC over 26 years of follow-up. First, in a subcohort of 307 healthy men, we profiled serial stool metagenomes and metatranscriptomes and assessed diet using semiquantitative food frequency questionnaires to identify food groups associated with 43 bacterial species involved in sulfur metabolism. We used these data to develop a sulfur microbial dietary score. We then used Cox proportional hazards modeling to evaluate adherence to this pattern among eligible individuals (n = 48,246) from 1986 through 2012 with risk for incident CRC.

Results: Foods associated with higher sulfur microbial diet scores included increased consumption of processed meats and low-calorie drinks and lower consumption of vegetables and legumes. Increased sulfur microbial diet scores were associated with risk of distal colon and rectal cancers, after adjusting for other risk factors (multivariable relative risk, highest vs lowest quartile, 1.43; 95% confidence interval 1.14-1.81; P-trend = .002). In contrast, sulfur microbial diet scores were not associated with risk of proximal colon cancer (multivariable relative risk 0.86; 95% CI 0.65-1.14; P-trend = .31).

Conclusions: In an analysis of participants in the Health Professionals Follow-up Study, we found that long-term adherence to a dietary pattern associated with sulfur-metabolizing bacteria in stool was associated with an increased risk of distal CRC. Further studies are needed to determine how sulfur-metabolizing bacteria might contribute to CRC pathogenesis.

Keywords: Cancer Biogeography; Colorectal Carcinogenesis; FFQ; Fecal Microbes.

Figure 1:. Experimental Design.

(A) Study population and sampling details: Participants in the MLVS provided up to four stool samples over a six-month study period with measurement of dietary intake via FFQ, identical to the FFQ given to participants in the HPFS. Stool underwent metagenomic and metatranscriptomic sequencing. (B) Creation of the sulfur microbial diet: In the first stage, among MLVS participants with longitudinal stool metagenomes and FFQs, we used supervised clustering and regression techniques to determine the foods most commonly associated with increased abundance of sulfur-metabolizing bacteria to generate the sulfur microbial diet score. (C) Predicted microbial carriage & risk of colorectal cancer: In the second stage, leveraging access to the much larger HPFS cohort with diet assessed every 4 years since inception, as well as other factors that may confound the relationship between diet and CRC, we calculated sulfur microbial diet scores in all 48,246 eligible participants of the HPFS, with higher scores reflecting closer adherence to a diet predicted to enrich for sulfur-metabolizing bacteria. Abbreviations: MLVS - Men’s Lifestyle Validation Study, SAA - sulfur-containing amino acids, H₂S - hydrogen sulfide, FFQ - food frequency questionnaire, CRC - colorectal cancer, ASA - aspirin or acetylsalicylic acid, BMI - body mass index

Figure 2:. Barplot of correlation between sulfur-related enzyme functional potential (DNA) and functional activity (RNA) level.

Modest concordance was observed between observed metagenomes and their downstream metatranscriptomes among enzymes intimately involved in sulfur modification and sulfur metabolism. This suggests that for reactions catalyzed by our subset of enzymes, the sulfur microbial diet trained on the species that encode for these gene products is reflective of the underlying sulfur-metabolizing activity of the human gut in stool. Only enzymes represented in both DNA and RNA with p-value <0.05 shown.

Figure 3:. The association between the sulfur microbial diet and two representative sulfur-metabolizing bacteria.

Higher sulfur microbial diet scores were associated with a relative enrichment of two sulfur-metabolizing microbes previously implicated in dysbiosis-associated CRC, Bilophila wadsworthia and Erysipelotrichaeceae bacterium. Trend line fit to non-zero data.

Figure 4:. Contributional genetic diversity among sulfur-metabolizing enzymes by sulfur-metabolizing bacteria.

Enzymes are arranged along the x-axis from least to greatest number of attributable species encoding for it at the DNA level (EC 1.8.1.8=1 species, EC 2.8.1.7=36 species). In general, the greater number of species encoding for a given enzyme (DNA), the greater the number of species that will contribute to that enzyme’s functional activity or transcript level. Only ECs with one or more assignable taxon for both DNA and RNA shown.

Figure 5:. Sulfur microbial diet and risk of colorectal cancer.

Multivariable modeling demonstrating an association between increased adherence to the sulfur microbial diet and risk of distal colon and rectal cancer. Models adjusted for age, family history of CRC, BMI, physical activity, smoking, aspirin use, total caloric intake, prior endoscopy, and recent physical exam. Tests for trend were conducted using the median value of each quartile category as a continuous variable. Abbreviations: RR, relative risk; CI, confidence interval.

Figure 6:. Sulfur microbial diet and risk of overall colorectal cancer and distal colon & rectal cancer by various subgroups.

Multivariable relative risks comparing extreme quartiles of sulfur microbial diet scores were adjusted for age, family history of CRC, BMI, physical activity, smoking, aspirin use, total caloric intake, prior endoscopy, and recent physical exam with the exception of a given stratification variable.