Oncogenic mutant KRAS inhibition through oxidation at cysteine 118

Abstract

Specific reactive oxygen species activate the GTPase Kirsten rat sarcoma virus (KRAS) by reacting with cysteine 118 (C118), leading to an electron transfer between C118 and nucleoside guanosine diphosphate (GDP), which causes the release of GDP. Here, we have mimicked permanent oxidation of human KRAS at C118 by replacing C118 with aspartic acid (C118D) in KRAS to show that oncogenic mutant KRAS is selectively inhibited via oxidation at C118, both in vitro and in vivo. Moreover, the combined treatment of hydrogen-peroxide-producing pro-oxidant paraquat and nitric-oxide-producing inhibitor N(ω)-nitro-l-arginine methyl ester selectively inhibits human mutant KRAS activity by inducing oxidization at C118. Our study shows for the first time the vulnerability of human mutant KRAS to oxidation, thereby paving the way to explore oxidation-based anti-KRAS treatments in humans.

Keywords: KRAS C118; NSCLC; ROS; cysteine modification; oncogene; redox regulation.

Fig. 1

The introduction of either the C118S or C118D substitution renders KRAS redox‐insensitive without compromising its biochemical function or altering the protein levels of KRAS. (A) Hras^−/−; Nras^−/−; Kras^lox/lox mouse embryonic fibroblasts (MEFs) stably transduced with human influenza hemagglutinin‐tagged (HA‐tagged) human KRAS^WT and KRAS^G12C, KRAS^G12D or KRAS^G12V mutant with or without the in cis C118S or C118D substitution (KRas^lox KRAS^MUT cells) were cultured in the presence or absence of 4‐hydroxytamoxifen (4OHT) and analyzed by western blot to confirm endogenous KRAS removal and to determine exogenous KRAS expression. Results are representative of one of three similar experiments. (B) Ras‐GTP levels in KRas^lox KRAS^MUT MEFs expressing KRAS^WT or KRAS^G12V with or without the in cis C118S or C118D substitution. Results are representative of one of three similar experiments. (C) Polyethylene glycol maleimide (PEG‐PC‐Mal) labeling of KRAS cysteine‐thiol residues in KRas^lox KRAS^MUT MEFs expressing KRAS^WT and KRAS^G12V, KRAS^G12C, or KRAS^G12D with or without the in cis C118S or C118D substitution. Results are representative of one of three similar experiments.

Fig. 2

The C118D substitution strongly inhibits KRAS mutant signaling activity, while the C118S substitution exerts a weaker inhibitory effect. (A) Growth rates of in KRas^lox KRAS^MUT MEFs expressing KRAS^WT, KRAS^G12C, KRAS^G12D or KRAS^G12V with or without the in cis C118S or C118D substitution grown in 10% fetal bovine serum (FBS) medium assessed by IncuCyte measurements. Unpaired Student's test was used to evaluate statistical significance between KRAS mutant cells with or without the C118S substitution and between KRAS mutant cells with or without the C118D substitution (ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001). Results are representative of one of three similar experiments (scale bar: 400 μm), error bars represent mean ± SEM. (B) Growth rates of in KRas^lox KRAS^MUT MEFs expressing KRAS^WT or KRAS^G12V with or without the in cis C118S or C118D substitution grown in 0.5% FBS medium assessed by IncuCyte measurements. Unpaired Student's test was used to evaluate statistical significance between KRAS G12V cells with or without the C118S substitution and between KRAS G12V cells with or without the C118D substitution (ns, not significant; ***P < 0.001). Results are representative of one of three similar experiments (scale bar: 400 μm), error bars represent mean ± SEM. (C) RAS downstream effector signaling in KRas^lox KRAS^MUT MEFs expressing KRAS^WT, KRAS^G12C, KRAS^G12D or KRAS^G12V with or without the in cis C118S or C118D substitution. Results are representative of one of three similar experiments. (D) Mean mitochondrial Reactive Oxygen Species (ROS) levels detected with MitoSOX in KRas^lox KRAS^MUT MEFs expressing KRAS^WT, KRAS^C118S, KRAS^C118D, KRAS^G12V, KRAS^G12V/C118S, or KRAS^G12V/C118D. Error bars represent mean ± SEM (P = 0.0005 and P = 0.0065; one‐way ANOVA test followed by the Bonferroni's multiple comparison test to correct for multiple comparisons).

Fig. 3

Through inducing oxidation at C118, Paraquat (PQ) inhibits the growth signaling of KRAS^G12V. (A) Growth rates of KRas^lox KRAS^MUT mouse embryonic fibroblasts (MEFs) expressing KRAS^G12V, KRAS^G12V/C118S, and KRAS^G12V/C118D treated or not treated with 35 μm PQ or 1 mm  N(ω)‐nitro‐l‐arginine methyl ester (L‐NAME) assessed by IncuCyte measurements. Unpaired Student's test was used to evaluate statistical significance between KRAS^G12V cells treated or not treated with PQ (**P < 0.01). Results are representative of one of three similar experiments (scale bar: 400 μm), error bars represent mean ± SEM. (B) Growth rates of KRas^lox KRAS^MUT MEFs expressing KRAS^WT, KRAS^C118S, and KRAS^C118D treated or not treated with 35 μm PQ or 1 mm L‐NAME assessed by IncuCyte measurements. Unpaired Student's test was used to evaluate statistical significance between KRAS^WT cells treated or not treated with PQ (*P < 0.05). Results are representative of one of three similar experiments (scale bar: 400 μm), error bars represent mean ± SEM. (C) Growth rates of KRas^lox KRAS^MUT MEFs expressing KRAS^G12V, KRAS^G12V/C118S, and KRAS^G12V/C118D treated or not treated with a combination of 35 μm PQ and 1 mm L‐NAME assessed by IncuCyte measurements. Unpaired Student's test was used to evaluate statistical significance between KRAS^G12V cells treated or not treated with the combination of 35 μm PQ and 1 mm L‐NAME (***P < 0.001). Results are representative of one of three similar experiments (scale bar: 400 μm), error bars represent mean ± SEM. (D) Growth rates of KRas^lox KRAS^MUT MEFs expressing KRAS^WT, KRAS^C118S, and KRAS^C118D treated or not treated with a combination of 35 μm PQ and 1 mm L‐NAME assessed by IncuCyte measurements. Unpaired Student's test was used to evaluate statistical significance between KRAS^WT cells treated or not treated with the combination of 35 μm PQ and 1 mm L‐NAME (ns, not significant). Results are representative of one of three similar experiments (scale bar: 400 μm), error bars represent mean ± SEM. (E) Polyethylene glycol maleimide (PEG‐PC‐Mal) labeling of KRAS cysteine‐thiol residues MEFs expressing KRAS^G12C, KRAS^G12D or KRAS^G12V treated or not treated with 500 μm PQ for 3 h. Results are representative of one of three similar experiments.

Fig. 4

In vivo, the C118D substitution robustly inhibits KRAS mutant signaling activity, whereas the C118S substitution exerts a weaker inhibitory effect. (A) KRas^lox KRAS^MUT mouse embryonic fibroblasts (MEFs) expressing KRAS^G12V, KRAS^G12V/C118S, or KRAS^G12V/C118D were injected subcutaneously into nude mice. Tumor growth was followed over time by measuring tumor volume every 3 days with a caliper, KRAS^G12V (n = 10), KRAS^G12V/C118S (n = 18) or KRAS^G12V/C118D (n = 16). (B) Tumor growth measure at endpoint in KRAS^G12V, KRAS^G12V/C118S, or KRAS^G12V/C118D sub‐cutis tumors (**P < 0.01, ***P < 0.005, one‐way ANOVA Kruskal–Wallis test). (C) Kaplan–Meier analysis of mice injected subcutaneously with KRas^lox KRAS^MUT MEFs expressing KRAS^G12V (n = 6, black line), KRAS^G12V/C118S (n = 8, turquoise line), or KRAS^G12V/C118D (n = 8, red line) (****P < 0.0001; log‐rank test [Mantel–Cox]).

Fig. 5

A model depicting the two reactive molecule pathways that regulate KRAS^G12V activity by reacting with C118. (A) Oxidative modification of cysteine with hydrogen peroxide (H₂O₂). (B) Oxidative modifications of cysteine and mimics of nonoxidizable and oxidized forms. Serine mimics a nonoxidizable cysteine, whereas aspartic acid mimics cysteine sulfinic acid, a doubly oxidized form of cysteine. (C) Mutant KRAS inhibiting redox‐pathway: in vitro and in vivo H₂O₂ selectively inhibits KRAS^G12V mutant driven oncogenesis by introducing cysteine 118 (C118) sulfinic acid modification without affecting wild‐type KRAS. The source of H₂O₂ can be paraquat, which produces superoxide that is converted by superoxide dismutases (SODs) to H₂O₂. The inhibitory effect of H₂O₂ oxidation on KRAS^G12V activity is permanently mimicked by the C118D substitution which resembles a cysteine sulfinic acid. Mutant activating redox‐pathway: Nitric oxide (NO) activates oncogenic KRAS^G12V mutant signaling introducing a C118 S‐nitrosylation. L‐NAME can inhibit the NO‐mediated activation of KRAS by inhibiting the NO production by nitric oxide synthases. Both the KRAS inhibiting and activating pathways are abolished by the C118S substitution, rendering KRAS redox‐insensitive. H₂O₂ and NO are depicted as acting on the same KRAS C118, although it is possible that the two reactive molecules act on different KRAS proteins and/or at different times during tumor development.