Early-onset colorectal tumors may show greater stromal cell enrichment relative to epithelial cells compared with average-onset tumors, with additional differences in gene expression and cell signaling.
Researchers conducted single-nucleus RNA sequencing of 189 primary colorectal tumors, including 54 from patients younger than 50 years and 135 from patients aged 50 years or older. The analysis included 850,222 high-quality nuclei from tumors spanning 8 decades of life. The researchers stratified tumors by microsatellite status and evaluated age-associated differences while accounting for sex, tumor location and stage, and neoadjuvant treatment. Independent cohorts were used to validate selected findings.
The study examined tumor cell composition, gene expression, copy number alterations, and inferred cell-to-cell communication. Particular attention was given to the stromal-to-epithelial cell ratio, long noncoding RNA expression, and pathways associated with epithelial-to-mesenchymal transition. The primary cellular composition analyses included 134 microsatellite-stable tumors, comprising 45 early-onset and 89 average-onset cases.
In the primary cohort, younger age was associated with a greater proportion of stromal cells relative to epithelial cells. The proportion of epithelial cells increased with age, whereas stromal and other nonepithelial populations were more abundant in younger patients' tumors.
The association was supported by an independent immunofluorescence cohort of 712 microsatellite-stable tumors. After adjustment, patients younger than 50 years had 2.70 times the likelihood of having a higher stromal-to-epithelial cell ratio compared with patients aged 70 years or older. The association followed a statistically significant trend across age groups.
An additional bulk RNA sequencing cohort of 564 microsatellite-stable tumors supported the association.
Gene expression analyses identified 215 genes with higher expression at younger ages across epithelial cell subsets vs. 44 genes with higher expression at older ages. Long noncoding RNAs accounted for about 30% to 45% of genes with increased expression at younger ages in stem-like, transit-amplifying-like, and colonocyte-like epithelial populations. Enrichment was particularly pronounced in stem-like cells.
The researchers identified higher expression of the FN1 gene in tumors from younger patients. Several other age-associated transcripts were poorly characterized, indicating areas for further investigation. Comparison with an independent bulk RNA sequencing data set identified 19 overlapping differentially expressed genes, including 14 with higher expression at younger ages and 5 with higher expression at older ages.
Cell-to-cell communication analyses identified additional differences in left-sided microsatellite-stable tumors. Early-onset tumors showed greater interaction strength across 29 signaling pathways vs. 7 pathways in average-onset tumors.
Several pathways enriched in early-onset tumors involved fibronectin, growth differentiation factors, activin, and osteopontin signaling. These pathways were associated with epithelial-to-mesenchymal transition, stromal remodeling, and angiogenesis. However, the interactions were computationally inferred rather than directly demonstrated through functional experiments.
Several molecular features were conserved across age groups. Broad epithelial cell states and core gene expression programs were similar in early-onset and average-onset tumors. The researchers found no distinct early-onset copy number alteration profile or statistically significant association between age and overall copy number alteration burden in microsatellite-stable tumors. A trend toward greater intratumoral copy number alteration heterogeneity in younger patients did not reach statistical significance.
The study had several limitations. Its observational design prevented the researchers from establishing whether the identified differences contributed directly to tumor progression or clinical outcomes. Younger patients were more likely to have left-sided tumors, complicating efforts to separate age-associated molecular differences from those related to tumor location. Some subgroup analyses included relatively few tumors and could not fully account for clinical confounders. The absence of matched normal tissue also required comparisons with external datasets. The findings require further functional validation.
“Our observations of increased stromal composition, [long noncoding] RNA activity and [epithelial-to-mesenchymal transition] in [early-onset colorectal cancer] provide hypotheses for possible mechanistic drivers of the advanced stage and poorer survival that characterize patients with [early-onset disease],” wrote lead study author Caroline B. M. Porter, of the Klarman Cell Observatory at the Broad Institute of the Massachusetts Institute of Technology and Harvard, and colleagues.
Full disclosures of the study authors can be found in the study.
Source: Nature Medicine
