APC mutations in human colon lead to decreased neuroendocrine maturation of ALDH+ stem cells that alters GLP-2 and SST feedback signaling: Clue to a link between WNT and retinoic acid signalling in colon cancer development.

APC mutations drive human colorectal cancer (CRC) development. A major contributing factor is colonic stem cell (SC) overpopulation. But, the mechanism has not been fully identified. A possible mechanism is the dysregulation of neuroendocrine cell (NEC) maturation by APC mutations because SCs and NECs both reside together in the colonic crypt SC niche where SCs mature into NECs. So, we hypothesized that sequential inactivation of APC alleles in human colonic crypts leads to progressively delayed maturation of SCs into NECs and overpopulation of SCs. Accordingly, we used quantitative immunohistochemical mapping to measure indices and proportions of SCs and NECs in human colon tissues (normal, adenomatous, malignant), which have different APC-zygosity states. In normal crypts, many cells staining for the colonic SC marker ALDH1 co-stained for chromogranin-A (CGA) and other NEC markers. In contrast, in APC-mutant tissues from familial adenomatous polyposis (FAP) patients, the proportion of ALDH+ SCs progressively increased while NECs markedly decreased. To explain how these cell populations change in FAP tissues, we used mathematical modelling to identify kinetic mechanisms. Computational analyses indicated that APC mutations lead to: 1) decreased maturation of ALDH+ SCs into progenitor NECs (not progenitor NECs into mature NECs); 2) diminished feedback signaling by mature NECs. Biological experiments using human CRC cell lines to test model predictions showed that mature GLP-2R+ and SSTR1+ NECs produce, via their signaling peptides, opposing effects on rates of NEC maturation via feedback regulation of progenitor NECs. However, decrease in this feedback signaling wouldn't explain the delayed maturation because both progenitor and mature NECs are depleted in CRCs. So the mechanism for delayed maturation must explain how APC mutation causes the ALDH+ SCs to remain immature. Given that ALDH is a key component of the retinoic acid (RA) signaling pathway, that other components of the RA pathway are selectively expressed in ALDH+ SCs, and that exogenous RA ligands can induce ALDH+ cancer SCs to mature into NECs, RA signaling must be attenuated in ALDH+ SCs in CRC. Thus, attenuation of RA signaling explains why ALDH+ SCs remain immature in APC mutant tissues. Since APC mutation causes increased WNT signaling in FAP and we found that sequential inactivation of APC in FAP patient tissues leads to progressively delayed maturation of colonic ALDH+ SCs, the hypothesis is developed that human CRC evolves due to an imbalance between WNT and RA signaling.