Quantifying equol-producing and beta-glucuronidating gut bacteria in breast cancer using 16S rDNA sequencing methods.

10601 Background: Gut bacteria with hormone-modulating potential may influence breast cancer (BC) risk: Equol-producing bacteria convert dietary isoflavones into equol, a phytoestrogen that binds estrogen receptors, while beta-glucosidating bacteria affect glycosides of dietary phytoestrogens, and beta-glucuronidating bacteria regulate the deconjugation and reabsorption of endogenous estrogens. We aimed to investigate whether gut bacterial species that have the potential to affect phytoestrogens or endogenous estrogens can be detected using 16S rDNA sequencing methods, and to what extent this detection may be influenced by the choice of 16S hypervariable region targeted for sequencing. Methods: DNA extracted from stool samples of women with BC (n = 112) and healthy controls (n = 171) was subjected to PCR amplification and high-throughput sequencing of the V1-V3 and V3-V4 16S rRNA hypervariable regions, commonly used regions in microbiome research. Data was processed using QIIME, and taxonomy assignment was performed with the SILVA database to the species level. Relative abundance and detection of taxa in cases vs controls were evaluated using Wilcoxon and McNemar’s tests in R. Taxa were classified as equol-producing based on Setchell and Clerici (2010) or beta-glucuronidating and beta-glucosidating based on Dabek et al (2008). Results: The relative abundance of equol-producing bacteria was significantly higher in the BC group vs controls in the V1–V3 (p = 0.0007); V3–V4 (p = 0.0089) regions, and when both regions were summed (p = 0.0011). Adlercreutzia equolifaciens was detected 29.6% more often in a presence/absence analysis of the V1-V3 region (V1-V3 presence: 93.0%, V3-V4 presence: 63.4%, p < 0.001), while Finegoldia magna was detected 21.1% more often in the V3-V4 region (V1-V3 presence: 18.0%, V3-V4 presence: 39.1%, p < 0.001). The relative abundance of beta-glucuronidating and beta-glucosidating bacteria was significantly higher in the BC group vs controls in the V1-V3 region (p = 0.0094) and when both regions were summed (p = 0.0254), and approached significance in the V3-V4 region (p = 0.0534) . Presence of Bifidobacterium adolescentis was detected 21.8% more often in the V1-V3 region (V1-V3 presence: 21.8%, V3-V4 presence: 0.00%, p < 0.001), and Roseburia inulinivorans was detected 21.1% more often in the V3-V4 region (V1-V3 presence: 63.0%, V3-V4 presence: 83.1%, p < 0.001). Conclusions: Equol-producing and beta-glucuronidating and beta-glucosidating bacteria appear enriched at the species-level analysis in BC subjects, but detection of specific organisms is sensitive to the 16S region sequenced. Accurate identification of hormone-modulating taxa requires both the V1-V3 and V3-V4 regions, as either alone is insufficient to capture the full range of taxa. Additional analyses with absolute quantification such as qPCR are needed to validate the results.