Abstract
Humans harbor a very complex and diverse collection of microbes, mostly bacteria, in their intestines that are collectively called the intestinal microbiota. It is known that we cannot live without the microbiota but on the other hand, the microbiota is associated to a number of intestinal and systemic diseases, including colorectal cancer (CRC). Nevertheless, the mechanisms underlying the key roles of the microbiota in health and disease are largely unknown. Compared to human populations in the Western countries, native Africans have a very low incidence and mortality of sporadic colorectal cancer (CRC). In contrast, African Americans, who are genetically similar to Africans, have a very high rate of CRC compared to other ethnic groups in the USA. Therefore, we hypothesize that differences in CRC risk between these groups is that the differential diets consumed by both groups (high fat and protein diets by African Americans; high starch and fibers diets by native Africans) impact the microbiota which as a result contributes to the differential CRC risk development.
To test this hypothesis, we compared and contrasted the microbiota between African Americans, urbanized Africans (who live in a more Western semi-urban environment) and rural Africans at baseline, followed by a longitudinal study focusing on healthy middle aged African and African Americans volunteers who were subjected to reciprocal diet exchange under strict controlled in-house conditions over two weeks. Fresh fecal samples were collected and frozen immediately at baseline, and before and after diet exchange. DNA was isolated from these samples and subsequently used for microbiota profiling using the Human Intestinal Tract Chip (HITChip), which is a 16S ribosomal RNA-based phylogenetic microarray covering over 1,000 of the currently known bacterial species from the human intestine.
HITChip profiling demonstrated that the microbiota composition is significantly different between the three groups at baseline with Prevotella (average abundance 22%), Oscillospira (average abundance 16%) and Bacteroides (28%) as key genera in urbanized Africans, rural Africans and African Americans, respectively. Co-occurrence analysis of microbial groups demonstrated that despite the differences in composition, the microbiota connectivity was similar between urban and rural Africans with potential butyrate producing and fiber consuming groups co-occurring at high correlation (p>0.8) in both groups. Remarkably, these groups did not show strong co-occurrences within African Americans. We speculate that these co-occurrences represent interacting bacterial groups that form the structural core
Compared to differences at baseline, the microbiota compositional changes were subtle after the short dietary exchange with only ten and two bacterial groups changing significantly (p<0.01) in Africans and African Americans, respectively. Remarkably, the co-occurrence between bacterial groups changed reciprocally upon the diet exchange with more highly correlated (p>0.8) co-occurring groups after consumption of the traditional African diet. We speculate that the decrease of co-occurring groups in the Western diet is associated to a weaker microbiota structure, which is more vulnerable for pertubations.
In conclusion, our study demonstrated that long-term dietary patterns are likely responsible for the microbiota compositional differences observed between Africans and Americans, but that a short term dietary exchange can modify microbiota composition and structure. The fact that the microbiota structure changes reciprocally upon diet exchange holds promise for the prevention of diet-microbiota induced CRC risk.
Citation Format: Erwin G. Zoetendal, Leo Lahti, Sebastian Tims, Philippe G.B. Puylaert, Junhai Ou, Kishore Vipperla, Willem M. de Vos, Stephen J. O'Keefe. Can variations in the intestinal microbiota explain differences in colon cancer risk?. [abstract]. In: Proceedings of the Sixth AACR Conference: The Science of Cancer Health Disparities; Dec 6–9, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2014;23(11 Suppl):Abstract nr SS01-03. doi:10.1158/1538-7755.DISP13-SS01-03