NEW YORK – An Australian research group has set out to identify whether each possible genetic variant in two genes — RUNX1 and GATA2 — is likely to increase the risk for blood cancers in those who inherit them or if they're harmless.
The study is part of a larger project underway at the University of Melbourne's Walter and Eliza Hall Institute (WEHI) to better understand the role of variants uncovered via genomic testing in the diagnosis and targeted treatment of blood, breast, and ovarian cancers. Other participating institutions include the QIMR Berghofer Medical Research Institute, the Peter MacCallum Cancer Center, and NW Health Pathology in Australia. The University of Washington in the US is also part of the project.
The three-year project, funded by a $2.5 million grant from the Medical Research Future Fund, will link the MaveDB data repository with Shariant, a platform for sharing clinical data among Australian clinical molecular pathology laboratories. MaveDB holds data from experiments, dubbed multiplexed assays of variant effect (MAVEs), which measure the effect variants of a specific gene have on its function. By linking the experimental functional data on variants in MaveDB with data in Shariant, the researchers hope to more quickly determine whether detected variants are associated with cancer or harmless.
The Center for Cancer Biology, a research institute formed through a partnership between the University of South Australia and SA Pathology, received $1 million out of the $2.5 million grant to explore the significance of genetic variants in blood cancers. "One of the big obstacles is we can identify genetic changes very well, [but] many times we don't know how to interpret those," said Christopher Hahn, an adjunct senior research fellow at the Center for Cancer Biology. "For over 50 percent of the variants, we don't know if they're causative of the cancers or if they're benign."
Hahn's group is conducting MAVE assays on two transcription factor genes, RUNX1 and GATA2, in which pathogenic variants are known to predispose carriers to blood cancers. About 17,000 Australians and 1.24 million people worldwide are diagnosed with blood cancers each year, and Hahn estimated that for around 20 percent, a genetically inherited predisposition plays a role in the development of the disease.
For the MAVE assays, investigators synthetically generate genes that have every possible variation at every amino acid position in the protein chain. For example, GATA2 has 480 amino acids. Researchers would need to generate and study the role of 9,120 different genetic variants in MAVEs to identify the changes that impact the ability of the gene to make proteins.
"Then, we're going to throw all of these together in a single assay and work out which ones change the function of the protein and which ones don't," Hahn said. The assay involves transfecting the variants into a cell line with a reporter gene, such as light-emitting luciferase, that shows which genes are functioning normally and which aren't.
Beyond just a "yes/no" answer as to whether a variant is impacting gene function, Hahn said his group will develop a score for distinguishing the extent of that impact, such as whether a variant causes a total loss of function, a partial loss of function, or a gain of function. Hahn's group plans to develop this scoring system into a tool that a diagnostic lab or a clinician can use to help determine whether a variant detected in a patient is pathogenic or not. This functional scoring is very complex, according to Hahn, and will be handled by bioinformaticians led by Alan Rubin, a computational biologist at WEHI.
To illustrate the scale of the project, Hahn said, "In my lifetime, I've generated about 25 GATA2 mutations and tested them." Through this project, Hahn's team will try to classify more than 9,000 variants. He estimated it will take three years to generate all of the variants. But when it comes time to perform the MAVEs, it will all be done in a single run over about a two-week period.
While the Adelaide-based group at the Center for Cancer Biology is working on RUNX1 and GATA2 for blood cancers, collaborators at partner institutions in Melbourne will be performing a similar assay to explore the role of PARP1 gene variants in breast and ovarian cancer. Except, instead of testing for loss or gain of function, they'll be looking to see whether the variants may cause cancers to become resistant to PARP inhibitors.
When complete, the data and pathogenicity scores generated through this project may bolster researchers and diagnostics labs' ability to more definitively determine whether patients' lymphomas and leukemias are caused by an inherited RUNX1 and GATA2 variant. Such a finding has implications for blood relatives of the patient, who should receive cascade genetic testing to determine if they too harbor the same pathogenic variant, and if so, receive frequent screening to potentially diagnose cancer at an earlier stage.
"Searching for a genetic diagnosis is only performed when there is some indication of a familial cause of blood cancer or blood cancer with other solid cancers," said Hahn. He and others have found, for example, that when patients develop certain types of blood cancer, such as myelodysplastic syndrome or acute myeloid leukemia, as young children or as young adults, that's sufficient justification to search for a genetic cause even if there isn't a family history of these diseases. Pathogenic GATA2 variants are common in these cases, for example, and may show up in patients as inherited or de novo mutations, Hahn said.
This type of knowledge could also help when it comes to finding a donor for an allogeneic bone marrow transplant. Those donors are often family members, but it would be a mistake to use a family member that has the same familial pathogenic variant. "For a number of these genes, if you give a bone marrow transplant with a [cancer-]predisposed bone marrow, sometimes that in itself would generate mutations in the new bone marrow" and can lead to a condition known as donor leukemia, said Hahn.
There are currently no therapies specifically marketed for targeting RUNX1 and GATA2 pathogenic variants in cancer cells. However, he is optimistic that his group's efforts to improve the classification of these variants will be useful when such drugs do become available. For now, Hahn said that correct classification of these genetic variants will allow clinicians to better manage their patients and relatives and offer genetic counseling when appropriate.