A team of researchers at the University of Michigan has developed a new mouse model that replicates the aggressive features of tubo-ovarian high-grade serous carcinoma (HGSC), the most common and lethal form of ovarian cancer. Their findings, published in the Proceedings of the National Academy of Sciences, demonstrate that the gene CDK12 acts as a tumor suppressor in this cancer type—and that its loss creates a vulnerability that could be therapeutically exploited by targeting a closely related gene, CDK13.
The study offers both critical insight into the biology of ovarian cancer and a roadmap for developing more effective treatments, particularly for patients whose tumors carry CDK12 mutations.
“CDK12 is frequently mutated in human ovarian cancer, but its exact role wasn’t fully understood,” said first author Jean Ching-Yi Tian. “We’ve now shown that losing CDK12 function accelerates tumor progression in mice and creates a specific vulnerability that can be targeted with precision therapies.”
A critical need for new models and therapies
Ovarian cancer is the sixth leading cause of cancer death among women in the U.S., with most fatalities attributed to HGSC. These tumors are notorious for developing resistance to chemotherapy, making the development of new treatment strategies an urgent priority. Genetically engineered mouse models (GEMMs) have emerged as powerful tools to study cancer in a system that closely mimics human disease, but until now, few robust models of CDK12-mutant HGSC existed.
To address this, Tian and colleagues engineered a new mouse model to create aggressive tumors with hallmark features of HGSC, including genomic instability, rapid progression, and reduced survival.
“This model faithfully captures what we see in patients with CDK12-mutant ovarian cancer,” Tian explained. “It provides a critical platform for testing therapies and studying disease mechanisms.”
Tumor suppressor with a twist
CDK12 mutations have also been observed in metastatic castration-resistant prostate cancer (mCRPC), and the team’s interest in the gene originally stemmed from their work in sequencing prostate tumors.
“We first identified CDK12 mutations in mCRPC,” she said. “When we knocked out CDK12 in the prostate mouse model, we found that it caused precancerous lesions and accelerated tumor growth, particularly when combined with Trp53 loss. That made us wonder—what’s CDK12 doing in ovarian cancer?” CDK12 is known to play a role in about 3% of tubo-ovarian high-grade serous cancers.
The team’s new data confirm that CDK12 serves as a tumor suppressor in ovarian cancer as well. Mice with intact CDK12 had slower-growing tumors and better survival outcomes compared to their CDK12-deficient counterparts.
Yet paradoxically, the researchers also found that CDK12-mutant tumors could be treated by targeting a different molecule: CDK13.
Synthetic lethality and the power of targeting CDK13
To identify therapeutic vulnerabilities in CDK12-deficient tumors, the team conducted a CRISPR/Cas9-based synthetic lethality screen. The standout result: CDK13, a close paralog of CDK12, emerged as an essential gene in CDK12-mutant cells. When CDK13 was inhibited, the cancer cells died—a classic case of paralog synthetic lethality.
“This was a major insight,” Tian said. “CDK12 and CDK13 share some overlapping functions. So, when a tumor loses CDK12, it becomes dependent on CDK13 to survive. That creates a therapeutic window.”
To exploit this vulnerability, the researchers tested a compound called YJ1206, a degrader that simultaneously targets both CDK12 and CDK13. In cell lines derived from their mouse models, YJ1206 caused robust tumor cell death, especially in the CDK12-deficient setting. In mouse models, tumors shrank and immune cell infiltration increased—suggesting the potential for combining this treatment with immune checkpoint inhibitors.
Context matters: tumor suppressor and therapeutic target
One aspect of the study that may seem counterintuitive is the idea of targeting CDK12 with a degrader when it also acts as a tumor suppressor. Tian acknowledged the complexity.
“CDK12 is context-dependent,” she said. “In some cancers, like breast cancer and Ewing’s sarcoma, CDK12 is amplified and might drive tumor growth. But in ovarian and prostate cancers, it’s frequently inactivated, and its loss leads to more aggressive disease.”
Despite this, CDK12 and CDK13 are also essential genes—meaning tumors still rely on their residual functions for survival. This makes them viable drug targets, particularly when one is lost and the other can be selectively targeted.
“We found that even some ovarian cancer cells without CDK12 mutations are sensitive to CDK12/13 degraders,” Tian added. “That opens the door to broader applications.”
Implications for precision oncology
Perhaps the most exciting implication of the study is its potential for precision medicine. By sequencing tumors to identify CDK12 mutations, clinicians may be able to select patients who are most likely to benefit from CDK13-targeted therapies.
“CDK12 can serve as a biomarker,” Tian said. “If we can identify patients with CDK12 loss-of-function mutations, we can treat them with this dual degrader and potentially achieve better outcomes.”
This work represents a significant step toward that goal. It not only uncovers a new tumor suppressor function for CDK12 in ovarian cancer but also lays the foundation for a targeted therapeutic strategy that could benefit a group of patients currently underserved by existing treatments.
“We’re at an exciting moment,” Tian said. “We’ve developed the models, identified the vulnerabilities, and demonstrated that the strategy works in mice. Now the next step is to move this into clinical development and help patients.”
