A study led by researchers at Washington University School of Medicine in St. Louis, Pacific Northwest National Laboratory, Case Western Reserve University, and the National Cancer Institute (NCI), has revealed a detailed map of the genes, proteins, infiltrating cells and signaling pathways that play key roles in driving glioblastoma (GBM).
The researchers say the study, through which they analyzed 99 tumors from patients, offers the largest and most detailed schematic of this deadly brain tumor, and could aid in the design of clinical trials, and potentially point to new therapeutic strategies. The research forms part of the NCI’s Clinical Proteomic Tumor Analysis Consortium (CPTAC).
“To improve therapies for this deadly cancer, understanding the tumor cells themselves is important but not enough,” said Li Ding, PhD, a professor of medicine and of genetics and director of computational biology in the Division of Oncology at Washington University. “We also must understand the tumor cells’ interactions with the surrounding environment, including immune cells and the connective tissues and blood vessels. In our study, we performed high-resolution and high-depth analyses on 99 glioblastoma tumors. Harnessing new technologies, including proteomics, metabolomics and single cell sequencing, this study is an extremely deep dive into glioblastoma tumor biology, revealing new possibilities for therapy.”
Ding is co-senior author of the team’s paper, which is published in Cancer Cell, and titled, “Proteogenomic and metabolomics characterization of human glioblastoma.”
Glioblastoma is among the most aggressive and devastating of cancers. While it is rare compared with other cancers—the incidence is about 12,000 new cases in the U.S. every year—glioblastoma is the most common type of brain cancer. Current standard of care includes surgery, chemotherapy and radiotherapy. “Promising immunotherapies have been proposed, including immune checkpoint inhibitors, vaccines, chimeric antigen receptor T cell (CAR-T) therapy, and viral therapy, though none have cleared Phase III trials,” the authors pointed out.
Yet even with intensive therapy, relatively few patients survive longer than two years after diagnosis, and fewer than 10% of patients survive beyond five years. And despite extensive studies focused on genomic features of glioblastoma, relatively little progress has been made in improving treatment for patients with this deadly disease. “GBM was one of the earliest subjects of deep genomic and transcriptomic analysis and targeted MS studies, the investigators continued. “However, most patients are still treated with a standard of care developed almost two decades ago, underscoring the need for deeper insights.
For their newly reported study, the researchers carried out an integrated analysis of genomic, proteomic, post-translational modification and metabolomic data for 99 treatment-naive GBMs. “Here, we extended classical sequencing approaches with comprehensive integration of MS-based proteome, phosphoproteome, acetylome, metabolome, and lipidome analyses and single-cell transcriptomics,” they wrote.
Among their reported results, the study findings identified new activated proteins— particularly PTPN11 and PLCG1—that serve as signaling hubs driving tumor growth in some patients. “Phosphoproteomic data indicate that PLCG1 and PTPN11 act as a common signaling hub for multiple RTKs [receptor tyrosine kinases],” the authors noted. The results also revealed gene expression patterns involved in epithelial-to-mesenchymal transition (EMT) that is common in tumor formation.
Interestingly the analysis identified four different categories by which to classify glioblastoma, based on the number and types of immune cells present in the tumors. “RNA and protein expression data from bulk tumors indicate that GBM subtypes differ in infiltrating macrophages and the distribution of specific immune cell types,” the team noted. “What’s especially appealing about this study is the clustering of glioblastoma into four groups based on immune subtypes that emerged by combining proteomic and genomic comprehensive analysis,” added Henry Rodriguez, director of the Office of Cancer Clinical Proteomics Research at the NCI. “This may open the door for effective responses to immune therapies.”
With the finding that the immune landscape of these tumors varied widely, and fit into four separate categories, the new data could indicate how well individual tumors are likely to respond differently to targeted therapies. For example, Type 1 tumors contain high numbers of immune cells called macrophages and a few T cells. Type 2 tumors have a moderate number of macrophages. Type 3 tumors include high numbers of T cells and a few macrophages. And type 4 tumors are what Ding calls an immune desert, with few or no immune cells of any type. So, an immunotherapy that targets macrophages, for example, might work well in patients with type 1 tumors but not at all in patients with type 4. Also, its possible that a clinical trial in which all patients are lumped together may not show that a drug works at all, when averaged across all patients.
Added co-author Albert H. Kim, MD, PhD, a professor of neurological surgery at Washington University and director of the Brain Tumor Center at Siteman: “Immunotherapy clinical trials in glioblastoma have been negative so far. And the fact that there are four different immune subgroups may be one of the reasons behind that. We can’t treat all glioblastoma tumors as one disease.”
The study also determined how an understudied protein modification, acetylation, may explain some functional differences between glioblastoma subtypes. “Acetylation changes a protein’s shape and often results in opening up DNA-protein complexes to facilitate gene expression,” said co-senior author Karin Rodland, PhD, chief scientist for biomedical research at Pacific Northwest National Laboratory. “By adding protein acetylation to our study, we were able to complete the loop from proteins to genes and gene expression, shedding light on important regulatory changes in glioblastoma.”
A group led by co-senior author Tao Liu, PhD, of Pacific Northwest National Laboratory, measured all of the proteins in the tumor samples as well as phosphorylation and acetylation, which affect biological functions such as cell signaling. Adding these data into the genomic analysis of the tumors revealed a small subset of glioblastomas that did not fit neatly into any of the typical genomic subtypes. “Multi-omics analysis identified a subset of patients with mixed subtypes compared with traditional sequencing-based subtypes, who exhibit shortened overall survival.
These mixed-subtype tumors were associated with a poor clinical outcome, providing the researchers with clues as to factors affecting the aggressiveness of a tumor, which were not evident from genetic information alone. “These patterns provide additional information for researchers to understand how the glioblastoma subtypes they identified may vary in biological function,” Liu said. “This multifaceted analysis provides an unprecedented level of detail, which is starting to connect the missing dots in glioblastoma.”
Summarising their findings, the authors noted, “The multidimensional analysis of patient specimens described in this investigation adds context to prior genomic and transcription-based investigations of GBM and suggests avenues for further mechanistic studies.”
Co-author Milan G. Chheda, MD, an assistant professor of medicine who treats patients at Siteman Cancer Center at Barnes-Jewish Hospital and Washington University School of Medicine, further commented, “The most immediate implications for these findings are better design of clinical trials. For most clinical trials, we take all comers and give them the same treatment. We are not designing trials in the most precise way because we have not fully understood the molecular differences between each patient’s tumor. This leads us to call a treatment a failure when in fact it may be helping specific people.” The researchers are conducting further studies to identify the best drugs to investigate in glioblastoma patients, depending on where their diseases fall on the new tumor map.
The authors concluded, “Rapid advancement of single-cell genomics and proteomics technologies will facilitate deeper analyses of GBM heterogeneity and TME interactions. We hope these advances will improve patient stratification for clinical trials and lead, ultimately, to personalized treatments.” Added Chheda, “This paper is an example of the advances that can be made when there is deep collaboration between many experts across the country who the National Cancer Institute has the ability to bring together.”
