The Effect Of Gut Microbiota And Synbindin Knockout In Colorectal Cancer

Purpose:Predicting colorectal cancer?CRC?based on fecal microbiota presents a promising method for non-invasive screening of CRC,but the optimization of classification models remains an unaddressed question.Meanwhile,our previous studies indicated an oncogenic role of synbindin in colon and gastric cancer by activating ERK on Golgi,mainly based on in vitro experiments.The role of synbindin in vivo is unclear.The purpose of this study was to systematically evaluate the effectiveness of different supervised machine-learning models in predicting CRC in two independent eastern and western populations and to explore the effect of synbindin on colon tumorigenesis in vivo.Methods:The structures of intestinal microflora in feces in Chinese population?N=141?were determined by 454 FLX pyrosequencing,and different supervised classifiers were employed to predict CRC based on fecal microbiota operational taxonomic unit?OTUs?.We validated the results in a French population.We constructed intestinal specific deletion of synbindin mice(Synbindin^?IEC).Then mice were applied to AOM/DSS model.Its littermates(Synbindin^flox)were used as controls.We examined the colon tumorigenesis and colon inflammation in both genotypes.We further examined the involved mechanisms by integrated analysis of gene expression levels of colon tissue via RNA sequence.We also analyzed the gut microbiota composition in both genotypes via16s RNA sequence.Result:Bayes Net displayed higher accuracy than other algorithms in both populations.When unclassified OTUs were used as input,models reached higher performance.Synbindin^?IEC mice displayed much less colon tumors compared with Synbindin^flox mice in AOM/DSS model.Reduced enterocyte proliferation rate was detected in Synbindin^?IECIEC mice.RNA-sequence analysis showed that Wnt signaling was greatly inhibited in colon tissue of Synbindin^?IEC mice.We confirmed the reduced activation of Wnt signaling both in vivo and in vitro.16s RNA sequence showed the gut microbiota in both genotypes were quite different.Conclusion:Bayes Net classification model combined with unclassified OTUs may present an accurate method for predicting CRC based on the compositions of gut microbiota.Enterocyte-specific deletion of synbindin inhibited colon tumorigenesis via attenuation of Wnt signaling and altering gut microbiota.