Brain cells called astrocytes form during fetal development — typically during the second or third trimester of pregnancy. The cells are vital for keeping neurons healthy, but they also have a darker side: They are abundant in the deadly cancer glioblastoma (GBM).
To form in childhood or adulthood, GBMs pull off a kind of time travel, restarting the prenatal process of growing new astrocytes and other brain cells that then cluster into tumors and spread across the brain. The tumors’ diffuse nature makes them difficult to study and almost impossible to treat: There have been no new treatments for the cancer in half a century, and most people die within 15 months of diagnosis.
Many of the astrocytes inside GBMs are trapped in an immature state, according to a study published in Nature Cell Biology (1). Identifying ways to coax the cells into healthy maturity could be a path to treatment, said study coauthor Steven Sloan, a geneticist at Emory University School of Medicine.
“If we understand the rules of how these cells move developmentally from one stage to another, to maturity to quiescence,” Sloan said, researchers may find “that those rules may apply in these cancers as well.”
Sloan and his colleagues generated brain organoids and sampled astrocytes from them every 50 days starting at day 80 — when astrocytes are just beginning to proliferate — through their maturity at 550 days. They used RNA sequencing and Assay for Transposase-Accessible Chromatin using sequencing (ATAC-seq) to measure gene expression and chromatin accessibility in the cells.
Astrocytes mature in three distinct stages, Sloan’s analysis found: early (80-150 days), middle (200-350 days), and late (400-550 days). The signatures of the organoid astrocytes matched those of astrocytes from fetal samples at 17-20, 20-28, and 28-39 gestational weeks, respectively.
Spatial transcriptomics shows how genes from the different astrocyte developmental stages are expressed inside tumors and in nearby tissue.
Credit: Steven Sloan
The researchers performed the same sequencing on cells from patient glioblastomas. Tumor astrocyte signatures looked like those from the organoids in the early and middle stages. Astrocytes from tissue on the tumor margins, on the other hand, matched cells in the later stage.
Spatial transcriptomics showed that genes expressed during the early and middle stages dominated in the center of the tumor cells, while the genes associated with the late stage were expressed in cells found largely in the boundaries between the tumor and surrounding tissue.
“The more we know about the cell types that are there, the more we can either look at how they respond to or provide resistance against existing or developing therapies,” said Aparna Bhaduri, a glioblastoma researcher at the University of California, Los Angeles, who was not involved in the work.
The team next split their patient glioblastoma samples into those with and without a mutation in the gene IDH1, which codes for an enzyme called isocitrate dehydrogenase (IDH). IDH1 mutations are common in cancer cells. People with IDH-mutant gliomas tend to live slightly longer after diagnosis than those with the wild type gene.
Astrocytes from IDH-mutant tumors have a genetic signature similar to margin astrocytes, Sloan and his colleagues found. There are also more immature astrocytes in wild type than mutant tumors. That may explain why IDH-mutant tumors tend to grow more slowly and have a slightly longer prognosis.
“All of these genes that make an astrocyte mature were specifically different between the two conditions,” Sloan said.
IDH-mutant gliomas produce a metabolite called D-2-hydroxyglutarate (D2HG), which is rarely found in healthy human tissue. To test whether D2HG is slowing the growth of IDH-mutant gliomas, Sloan and his team added D2HG to fetal human astrocytes in a dish.
After one week, there were fewer actively dividing astrocytes in the dishes containing D2HG than control dishes, and astrocytes exposed to D2HG showed decreased expression of early astrocyte genes.
The more we know about the cell types that are there, the more we can either look at how they respond to or provide resistance against existing or developing therapies.
- Aparna Bhaduri, University of California, Los Angeles
“We do feel strongly that the D2HG, for whatever reason, does help the astrocytes move along their trajectory,” Sloan said. “That explains at least in part why those patients [live longer].”
The findings showed that coercing GBM cells into maturity could be therapeutic, Sloan said.
In a recent clinical trial, patients with a subtype of GBM who received a drug called ONC201 lived on average 22 months, compared with a typical survival rate of 11 to 15 months (2). ONC201 also increased the levels of L-2-hydroxyglutarate in tumor cells, a metabolite that, similar to D2HG, regulates cell proliferation and differentiation.
“It’s not a one-to-one,” Sloan said, but “it’s evidence that understanding how these molecules, whether normal or not, impact development could lead to therapeutics that then are beneficial.”
Future studies should probe what role the immature astrocytes and other cells are playing in the creation and growth of the tumor, Bhaduri said.
“It was a really beautiful paper,” she said. “I hope that the Sloan group or someone else is able to really understand [the astrocytes’] functional role and what are they doing.”
References
- Sojka, C. et al. Divergence from the human astrocyte developmental trajectory in glioblastoma. Nat Cell Biol (2024).
- Venneti, S. et al. Clinical Efficacy of ONC201 in H3K27M-Mutant Diffuse Midline Gliomas Is Driven by Disruption of Integrated Metabolic and Epigenetic Pathways. Cancer Discov 13, 2370–2393 (2023).
