Selectively targeting inosine monophosphate dehydrogenase-2 impairs brain metastatic potential while preserving immune cell function.

Brain metastases (BM) occur in 26% of cancer patients and have a 90% mortality rate within 1 y of diagnosis, yet the current standard of care remains palliative. We have previously shown that de novo GTP synthesis is a druggable metabolic vulnerability in BM cells, through its rate-limiting enzyme, inosine monophosphate dehydrogenase (IMPDH). IMPDH inhibitors have progressed to phase-II oncology trials in the past, failing largely due to dose-limiting toxicities associated with off-target inhibition of IMPDH1, the constitutively expressed isoenzyme in normal human lymphocytes. Here, we determined that a single subtype (isoenzyme) of IMPDH, IMPDH2, is specifically upregulated in brain metastasis-initiating cells (BMICs), absent in normal brain tissue, and is sufficient to drive the formation of BM. Moreover, we show that genetic knockout of IMPDH2 stops the proliferation of BM cells in vitro and the onset of BM in vivo. We synthesized IMPDH2-selective compounds and showed that they maintain a potent antiproliferation effect on BMICs, but spare immune cell function compared to previously developed pan-IMPDH inhibitors. Furthermore, we introduce a positive correlation between compound selectivity for IMPDH2 and the ability to synergize with Osimertinib: the standard of care for EGFR-mutant non-small cell lung cancer. Overall, our results suggest that specifically blocking IMPDH2 is an effective therapeutic strategy for BM by overcoming the immune suppressive effects that have hindered the clinical development of pan-IMPDH inhibitors in the past. An IMPDH2 specific therapy could be coadministered with primary tumor standard of care treatments to provide a safe and interceptional approach for BM.