NEW YORK – By comparing the transcriptional features found in individual cells from developing mouse or human brain samples to those in pediatric brain tumors, investigators from Canada, the US, and France have identified points of development when such cancers may arise due to "stalled differentiation."
"Our findings identify impaired differentiation of specific neural progenitors as a common mechanism underlying these pediatric cancers and provide a rational framework for future modeling and therapeutic interventions," first author Selin Jessa, a doctoral student affiliated with McGill University and Jewish General Hospital, and her co-authors wrote.
The researchers identified 191 cell populations with single-cell RNA sequencing on more than 65,000 cells from two tumor-prone parts in samples from developing human or mouse brains: the embryonal pons and the forebrain. They compared gene expression profiles with RNA sequences for hundreds of bulk samples from cancer patients or from normal adult or fetal brain samples.
The team's findings, appearing online today in Nature Genetics, pointed to transcriptomic ties between specific brain cell populations and pediatric tumor types previously profiled by bulk sequencing.
"Childhood brain tumors have a spatiotemporal distribution that mirrors cellular waves of brain development, and several of their known drivers have developmental roles," the authors explained, noting that "[a] major challenge in understanding, modeling, and treating these tumors has been the absence of a comprehensive blueprint for normal brain development and lack of knowledge regarding their cell of origin."
To better spell out those origins, the researchers relied on single-cell RNA-seq to assess transcriptome profiles in 61,595 cells from developing mouse embryonal pons and forebrain samples and in more than 3,900 cells from two human samples representing brain development at 17 to 19 weeks post-conception.
With these profiles, they tracked down 191 cell populations that were set alongside bulk RNA-seq profiles for 186 patient samples, as well as 43 normal adult and 11 fetal brain samples.
"In all cases, [single-sample gene-set enrichment analysis] scores for human populations were extremely close to their mouse counterparts," the authors noted, "indicating no major cross-species differences at this level of analysis."
When it came to medulloblastoma tumors from the WNT subtype, gene expression markers in the bulk tumor profiles lined up with those found with the new transcriptome profiles for cells from a so-called "rhombic lip-derived mossy fiber neuronal lineage" in the pontine pre-cerebellar region, the authors reported. Together with prior findings in WNT medulloblastoma, the results point to "a strong differentiation block in this medulloblastoma subgroup," they noted.
On the other hand, the researchers noted that the neuronal lineage cells had transcriptomic features closely resembling those in embryonal tumors with multilayered rosettes (ETMRs) — a type of central nervous system tumor with both neuroblastoma and ependymoblastoma features that is typically diagnosed in very young children — consistent with prenatal origins for ETMRs in early neural progenitor cells.
The analysis revealed still other similarities between developing brain cell populations and pediatric tumors, they reported, including expression features in group 2a/b atypical teratoid/rhabdoid tumors (ATRTs) and non-neuroectoderm cell types such as the microglia, macrophages, and endothelial cells.
The team was also able to trace pediatric high-grade glioma (pHGG) tumors back to a population of neural progenitor cells on their way to becoming glial cells — a differentiation trajectory that could be nudged along by lopping out a pHGG-related mutation known as H3K27M in tumor-derived cell lines.
"[O]ur data reveal a common theme across subtypes of pediatric brain tumors where genetic alterations impact restricted developmental windows during the differentiation of neural lineages, retaining cells in a self-renewing, progenitor-like phenotype," the authors wrote, noting that a "deep understanding of the biology, timing, and transitional states of the developmental hierarchies at the root of childhood brain tumors may allow the rational design of preclinical models, an essential step toward improved tumor diagnostics and new therapeutics."