An Unprecedented Covalent Binding Mechanism of ML210 to GPX4

GPX4 is a selenoprotein that plays a critical role in protecting cells against oxidative stress and ferroptosis. The nucleophilic selenocysteine residue (Sec) located in the catalytic center of GPX4 can be exploited to design covalent inhibitors. Previously reported covalent GPX4 inhibitors always contain reactive alkyl chloride groups, conferring poor selectivities. Recently, the Schreiber group reported that ML210, a direct covalent inhibitor targeting GPX4 in cells, exhibits much more potency and remarkable selectivity compared with chloroacetamide inhibitors. More interestingly, an unprecedented covalent binding mechanism is involved in the ML210 targeting GPX4 in cells.

ML210 was previously reported as a ferroptosis inducer. Due to lack of an apparent covalent reactive group and inability to react with small molecule thiols directly, the covalent binding action of ML210 with its target proteins has never been disclosed. To identify that ML210 acts as a direct covalent inhibitor targeting GPX4, pulldown and competitive pulldown experiments were performed, suggesting that ML210 interacts with GPX4 in cells. The cellular thermal shift assay (CETSA) also revealed that ML210 can stabilize GPX4^U46C instead of GPX4^U46A, which further confirmed that the covalent interaction may be dependent on the selenocysteine residue of GPX4.

Intriguingly, ML210 can only bind GPX4 in cells instead of the recombinant counterpart. This is distinctly different from other reported chloroacetamide GPX4 inhibitors, which can covalently bind both GPX4 in intact cells and in a purified form. Intact protein mass spectrometry-based assay showed that ML210-GPX4 complex (+434 Da) can be detected after cells treated with ML210. The binding feature and the loss of 41 Da mass suggested that ML210 maybe acts as a prodrug, which can be metabolized in cells to form new intermediates targeting GPX4 via a covalent bond.

This unusual covalent binding action prompted researchers to further investigate the detailed mechanism. Previous studies have demonstrated the nitroisoxazole group is likely hydrolyzed to release a carboxylic acid and α-nitroketoxime under basic conditions. The SAR studies of the isoxazole group reveled that both isoxazole and nitro groups are essential for the activity of ML210. The 5-position substitution has little effect on the cellular potency of ML210. Similar to ML210, isopropyl-ML210 also induced the same ML210-GPX4 (+434 Da) complex formation in cells. This is consistent with the hypothesis of nitroisoxazole group hydrolysis.

Therefore, JKE-1674 was proposed as a key intermediate after the hydrolysis of ML210 in cells. Indeed, after incubating ML210 in cells, the formation of JKE-1674 and the consumption of ML210 were correlated in a time-dependent manner. JKE-1674 showed a similar activity to ML210 in terms of the cellular potency and proteome-wide selectivity. However, JKE-1674 still cannot react with the purified GPX4, suggesting that an additional metabolite acts as the key intermediate to react with GPX4 directly. Subsequently, the hypothesized nitrile-oxide electrophile JKE-1777 was synthesized. The nitrile-oxide JKE-1777 showed considerable cellular activity and can react with purified GPX4 directly. Additionally, the molecular weight of JKE-1777 is consistent with the +434 Da mass of ML210-GPX4 complex. In contrast to JKE-1777, no activity was observed in cells for its precursor 39, suggesting that JKE-1777 was generated through other precursor in cells, a possible precursor is the furoxan. Collectively, JKE-1777 may be the final electrophile that forms the covalent bond with the selenocysteine residue of GPX4 in cells.

Based on the findings, the researchers further expanded the diversity of the masked nitrile-oxide electrophile. Compound 43, JKE-1708, JKE-1716 and diacylfuroxan showed pretty good potency in ferroptosis-inducing and fer-1 rescue experiments. It would be interesting to see whether the potency and selectivity will be improved if replacing the chloride of RSL3 and ML162 with the masked nitrile-oxide electrophiles.

Although the in vivo activities of these masked nitrile-oxide compounds remain to be evaluated, this novel covalent binding mechanism in cells indicates that masked nitrile-oxide electrophiles may be potentially designed as covalent inhibitors for targeting other (seleno)cysteine residues in proteins.

Reference:

1. Eaton J. K., et.al. Selective covalent targeting of GPX4 using masked nitrile-oxide electrophiles. Nat. Chem. Biol.  2020, 16, 497. https://doi.org/10.1038/s41589-020-0501-5

2.  Eaton J. K., et.al. Diacylfuroxans Are Masked Nitrile Oxides That Inhibit GPX4 Covalently. J. Am. Chem. Soc. 2019, 141, 20407.