Interplay between WNT/PI3K-mTOR axis and the microbiota in APC-driven colorectal carcinogenesis: data from a pilot study and possible implications for CRC prevention

Abstract

Background: Wnt/β-catenin signalling impairment accounts for 85% of colorectal cancers (CRCs), including sporadic and familial adenomatous polyposis (FAP) settings. An altered PI3K/mTOR pathway and gut microbiota also contribute to CRC carcinogenesis. We studied the interplay between the two pathways and the microbiota composition within each step of CRC carcinogenesis.

Methods: Proteins and target genes of both pathways were analysed by RT-qPCR and IHC in tissues from healthy faecal immunochemical test positive (FIT+, n = 17), FAP (n = 17) and CRC (n = 15) subjects. CRC-related mutations were analysed through NGS and Sanger. Oral, faecal and mucosal microbiota was profiled by 16 S rRNA-sequencing.

Results: We found simultaneous hyperactivation of Wnt/β-catenin and PI3K/mTOR pathways in FAP-lesions compared to CRCs. Wnt/β-catenin molecular markers positively correlated with Clostridium_sensu_stricto_1 and negatively with Bacteroides in FAP faecal microbiota. Alistipes, Lachnospiraceae, and Ruminococcaceae were enriched in FAP stools and adenomas, the latter also showing an overabundance of Lachnoclostridium, which positively correlated with cMYC. In impaired-mTOR-mutated CRC tissues, p-S6R correlated with Fusobacterium and Dialister, the latter also confirmed in the faecal-ecosystem.

Conclusions: Our study reveals an interplay between Wnt/β-catenin and PI3K/mTOR, whose derangement correlates with specific microbiota signatures in FAP and CRC patients, and identifies new potential biomarkers and targets to improve CRC prevention, early adenoma detection and treatment.

Keywords: Colorectal cancer; Familial adenomatous polyposis; Microbiota; PI3K/mTOR; Wnt/β-catenin.

Fig. 1

Transcriptional and translational signatures of Wnt/β-catenin and PI3K/mTOR in FIT+, FAP and CRC tissues. A1-D2) Representative images and relative quantifications of cytosolic and nuclear β-catenin (A1-A2), nuclear Ki67 (B1-B2), cytosolic p-p70S6K (C1-C2) and p-S6R (D1-D2); representative fields with a staining value close to the median of the corresponding group were shown for each marker. Values are shown as median with 95% CI, scale bars: 100 μm, magnification: 200x, magnification in small panels (A1) was three times higher that shown in the corresponding figures. Ten independent fields were quantified for CRC tissues while six to eight for FIT + and FAP tissues; n = 14 FIT+, 13 FAP NM, 13 FAP P, 15 CRC. E) RNA expression levels of AXIN2, cMYC, CCND1, LGR5 and RPS6 genes. Values are shown as mean ± SD; n = 17 FIT+, 16 FAP NM, 13 FAP P, 15 CRC. Ordinary one-way ANOVA with Tukey’s multiple comparison or Kruskal-Wallis with Dunn’s multiple comparison was performed accordingly to the data distribution checked with the Shapiro-Wilk test; * P < .05, ** P < .01, *** P < .001, **** P < .0001

Fig. 2

Oral microbiota in FAP and CRC patients compared to FIT + subjects. (A) Distribution of alpha diversity, calculated using Faith’s phylogenetic diversity, in the oral microbial profiles of FAP, CRC and FIT + subjects. No significant differences were found (Wilcoxon test, P > .05). (B) PCoA of Bray-Curtis dissimilarity between microbial profiles. Ellipses include a 95% confidence area based on the standard error of the weighted average of sample coordinates. A significant separation was found between the FAP and FIT + groups (Adonis, P = .0084). (C) Ternary plot showing the genus-level composition of the three groups, with point size representing the mean relative abundance in the cohort. The position of each point indicates which group is more represented by that taxon. (D) Relative abundance distribution of bacterial genera differentially represented between groups. Wilcoxon test, ° P < .1; * P < .05; ** P < .01. n = 16 FIT+, 17 FAP and 15 CRC.

Fig. 3

Faecal microbiota in FAP and CRC patients compared to FIT + subjects. (A) Distribution of alpha diversity, calculated using Faith’s phylogenetic diversity, in the faecal microbial profiles of FAP, CRC and FIT + subjects. Wilcoxon test, * P < .05. (B) PCoA of Bray-Curtis dissimilarity between microbial profiles. Ellipses include a 95% confidence area based on the standard error of the weighted average of sample coordinates. A significant separation was found between FAP and the other groups (Adonis, P = .0005). (C) Ternary plot showing the genus-level composition of the three groups, with point size representing the mean relative abundance in the cohort. The position of each point indicates which group is more represented by that taxon. (D) Relative abundance distribution of bacterial genera differentially represented between groups. Wilcoxon test, ° P < .1; * P < .05; ** P < .01. n = 17 FIT+, 16 FAP and 15 CRC.

Fig. 4

Mucosa-associated microbiota in FAP and CRC patients compared to FIT + subjects. (A) Distribution of alpha diversity, calculated using Faith’s phylogenetic diversity, in the microbial profiles associated with adenomatous polyps from FAP patients (FAP P), cancerous mucosa from CRC patients (CRC) and normal mucosa from FAP (FAP NM), CRC (CRC NM) and FIT + subjects (FIT+). Wilcoxon test, ° P < .1; * P < .05. (B) PCoA of Bray-Curtis dissimilarity between microbial profiles. Ellipses include a 95% confidence area based on the standard error of the weighted average of sample coordinates. A significant separation was found between all groups (Adonis, P < .05). (C) Ternary plot showing the genus-level composition of the three groups, with point size representing the mean relative abundance in the cohort. The position of each point indicates which group is more represented by that taxon. Wilcoxon test, ° P < .1; * P < .05; ** P < .01; *** P < .001. n = 17 FIT+, FAP NM, 14 FAP P, 13 CRC NM and 15 CRC.

Fig. 5

Associations between Wnt/β-catenin and mTOR downstream effectors and relative abundances of oral, faecal and mucosa-associated bacterial genera in FAP patients. Correlation plots between RNA (i.e. MYC, AXIN2 and CCND1) or protein (i.e. nuclear β-Catenin and p-S6R) expression levels and relative taxon abundances in the oral (A), faecal (B) and mucosal (C) microbiota of FAP patients. Only significant Spearman’s correlations (P < .05) at family and genus level with |rho| > 0.3 are shown

Fig. 6

Associations between mTOR downstream effectors and relative abundances of faecal and mucosa-associated bacterial genera in CRC patients. Correlation plots between RNA (i.e. RPS6) or protein (i.e. p-S6R) expression levels and relative taxon abundances in the faecal (A) and mucosal (B) microbiota of CRC patients. Only significant Spearman’s correlations (P < .05) at family and genus level with |rho| > 0.3 are shown