Introduction: The incidence of early-onset colorectal cancer (EOCRC) has been increasing worldwide over the last decade. Genomic profiling of EOCRC compared with late-onset cancer (LOCRC) has not revealed significant mutational differences, suggesting that etiologic drivers of EOCRC largely influence the earliest stages of tumor initiation and progression, including colonic stem cell (cSC) development. In our Apc^min/+ mouse models, we have demonstrated that young adult mice (2-3 months), have more regenerative/tumorigenic intestinal Lgr5+ colonic stem cells (cSCs) than aged mice (20-22 months), whose cSCs have reverted to a fetal-like quiescent state. However, aged mice spontaneously develop more tumors after Cre-recombinase-induced loss of Apc. These data suggest that EOCRC may develop from transcriptionally distinct cell populations compared to LOCRC. To assess if such differences are found in humans, we examined age-related transcriptional changes that occur within human cSCs derived from intestinal biopsies.

Methods: Using 10X Genomics Chromium single-cell RNA sequencing, we sequenced EPCAM+/CD45- live sorted epithelial cells from rectosigmoid biopsies of ten “Younger” (18-55 years old) and ten “Older” (65+ years old) study participants with recent history of colorectal adenoma(s). Seurat was used to perform cell clustering and typing. Mixed-linear regression (Nebula) was used to discover age-related differentially expressed genes (DEGs).

Results: We identified 41,222 cSCs from cell clusters enriched for known cSC markers (SMOC2, LGR5, OLFM4, and high ribosomal read fraction). We found 412 age-associated cSC DEGs (FDR < 0.10), of which 96 had an absolute Log₂(Fold Change) > 0.50 (Fig. 1A). Gene set enrichment analysis revealed that Younger cSCs are enriched for upregulated genes involved in oxidative phosphorylation, MYC signaling, DNA repair, and interferon-gamma signaling. In contrast, Older cSCs were enriched for upregulated genes involved in tumor necrosis factor-alpha, sonic hedgehog, p53/p63, WNT, and PI3K/AKT signaling (Fig. 1B), which are markers of inflammaging and senescence. We confirmed this finding by studying the SenMayo senescence gene score and found that Older cSCs had significantly higher expression of senescence markers than Younger cSCs (Wilcoxon p < 2.2e⁻¹⁶). Further interrogation of age-related changes in other cell types is ongoing.

Conclusions: Together, these results suggest that at younger ages, cSCs support competent stem cell homeostasis, but with advancing age, the cSC pool appears to be less regenerative. These age-related biological differences in the cSC niche may result in differential susceptibility to etiological risk factors for neoplasia, such as diet and environmental exposures, in young individuals compared with older individuals, potentially explaining the recent rapid rise in the incidence of EOCRC.

Figure 1. A) A volcano plot demonstrating the differential gene expression analysis results. B) Significant (FDR < 0.10) gene sets found from the gene set enrichment analysis.

Thaddeus Stappenbeck

Appearances