Science Spotlight

IDH1 mutant (or methylation), stops monocyte migration

From the Holland Lab, Human Biology Division

Immunotherapy is an exciting and rapidly expanding aspect of cancer treatment and research. In the case of diseases like cancer, clinicians have to move fast, especially for aggressive malignancies with poor prognoses. Thus in the search for cures doctors and researchers are eager to translate immunotherapy based successes from one cancer to other types. However, as researchers we must remind ourselves – each patient and each tumor is unique. In fact, new work from the Holland Lab (Human Biology Division) and others at Fred Hutch shows that the relationship between immune cells and glioma, a form of brain cancer, is complex and immunotherapy approaches will need to be nuanced. In a recent Genes and Development publication the Holland Lab demonstrated that a specific oncogenic mutation in the gene isocitrate dehydrogenase (IDH1) alters the population of immune cells associated with glioma tumors. This study not only poses exciting new questions, but also develops the tools to answer them.

Mouse glioma models created with the RCAS system results in brain specific expression of shRNA targeting p53, elevated levels of platelet derived growth factor (PDGF), and increased levels of wild type or mutant IDH1. Both the wild type and mutant IDH1 form tumors expressing the expected IDH1 molecules by histology. In multiple genetic backgrounds survival time was increased in mutant IDH1 tumors compared to wild type.

IDH1 usually functions in glucose metabolism, specifically within the TCA cycle to produce α-ketoglutarate. A mutation to the active site of IDH1 (muIHD1) results in the production of 2-hydroxygluratate (2-HG) instead. Not only is this bad for metabolism, but 2-HG also inhibits many demethylases within the cell. Thus tumors expressing muIHD1 are characterized by hyper-DNA methylation, often focused in CpG islands, that affects tumor biology. In fact, if gliomas are grouped by the CpG island methylation phenotype (CIMP) and non-CIMP there are dramatic differences in survival of patients. Non-CIMP gliomas are more aggressive giving worse outcomes, and while many factors contribute to aggressiveness the role of immune cells in this behavior was of particular interest. First, researchers measured the levels of immune cells (CD45+) in human glioma tumor samples removed by surgery. This revealed that tumors with muIDH1 contained fewer of nearly all type of immune cells measured. While understanding this in a human context is important, there is limited experimental control. Thus researchers developed the first mouse model of muIDH1 glioma. To generate these gliomas platelet derived growth factor, p53 shRNA, and either mutant or wild type IDH1 were expressed only in the brain tissue. Expressing either wild type or mutant IDH1 resulted in glioma formation; however, only muIDH1 tumors demonstrated features of CIMP, including elevated levels of 2-HG and DNA methylation. Similar to observations in patients, the muIDH1 tumors tended to be less aggressive, resulting in better survival. As was observed in the human tissues, the mouse tumors expressing muIDH1 contained fewer immune cells (CD45+). Different from the human results, T cells were recruited to muIDH1 and wtIDH1 equivalently, while macrophages depletion was far greater in the mouse muIDH1.

Graph showing fewer CD45+ immune cells in human tumors with IDH1 mutation, which is recreated in the mouse model
Human glioma tumors excised as part of therapy contained fewer CD45+ immune cells if the metabolism gene IDH1 was mutated. The IDH1-driven mouse model of glioma created at Fred Hutch demonstrated similar behavior.

To begin understanding the mechanism for the difference in immune recruitment scientists analyzed the migratory activity and signals within tumor cells. Migratory behavior of muIDH1 tumor cells isolated from mouse tumors or grown in culture was reduced compared to wtIDH1 tumor cells. Moreover, gene expression profiles demonstrated that chemotactic signals like CCL-2 and CXCL-2 were expressed at lower levels in muIDH1 cells. These findings raise many interesting questions. Is immune recruitment a side-effect of chemotactic signals that drive metastatic behavior? Or do chemotactic signals recruit immune cells that aid in angiogenesis and microenvironment remodeling? Do these immune cells contribute to an oncogenic inflammatory phenotype? This study both generated many unanswered questions and provided the tools to answer many of them. This work was a collaboration between the Hockenbery, Houghton, and Holland Labs so it will be exciting to see how each group contributes to our understanding of IDH1 and CIMP in tumor biology.


Amankulor NM, Kim Y, Arora S, Kargl J, Szulzewsky F, Hanke M, Margineantu DH, Rao A, Bolouri H, Delrow J, Hockenbery D, Houghton AM, Holland EC. 2017. Mutant IDH1 regulates the tumor-associated immune system in gliomas. Genes Dev. Epub.


Funding for this research was provided by the National Cancer Institute.