Glutamine antagonists may KEAP lung cancer in check

Abstract

The glutamine antagonist DRP-104 blocks purine synthesis and combines with checkpoint inhibitors to promote antitumor immunity in KEAP1/NRF2-mutant lung cancers.

Targeting metabolic pathways that use circulating nutrients, such as glucose and glutamine, has been a therapeutic focus in cancer for decades. Glutamine is a conditionally essential amino acid that provides carbon and nitrogen for pathways critical to cancer cell growth and proliferation (1). Glutamine feeds multiple metabolic pathways that contribute to cellular bioenergetics, biosynthesis, and redox balance (Fig. 1). Most tissues synthesize enough glutamine to maintain homeostasis, but rapid proliferation and other stresses increase glutamine demand. This is relevant to tumor biology, where oncogenes drive enhanced nutrient catabolism in support of malignant cell growth and reduce nutrient availability for other cells in the microenvironment (1).

Fig. 1. Effect of glutamine antagonists on cancer cell metabolism and the immune microenvironment.

(A) DRP-104 and CB-839 have distinct metabolic effects. DRP-104 inhibits amidotransferases that contribute to purines and other biosynthetic intermediates (dashed box). CB-839 suppresses flow into the TCA cycle. DRP-104’s blockade of purine synthesis is most responsible for tumor cell growth suppression. (B) In Keap1-mutant lung cancers, DRP-104 stimulates antitumor immunity by modulating several T cell subsets in the tumor microenvironment. Gln, glutamine; Glu, glutamate; α-KG, α-ketoglutarate; GSH, glutathione; GLS, glutaminase; PPAT, phosphoribosyl pyrophosphate amidotransferase; PFAS, phosphoribosylformylglycinamidine synthase; GMPS, guanosine 5'-monophosphate synthase; ASNS, asparagine synthetase; GFPT1, glutamine-fructose-6-phosphate transaminase-1; CTPS, cytidine 5′-triphosphate synthetase; Tregs, regulatory T cells. Credit: Ashley Mastin/Science Advances.

By driving expression of glutaminase (GLS), which initiates glutamine catabolism for the tricarboxylic acid (TCA) cycle, MYC was the first oncogene mechanistically linked to glutamine dependence (2, 3). Since that discovery, additional tumor genotypes have been identified as causes of glutamine dependence in preclinical models (1). Among these, KEAP1/NRF2-mutant non–small cell lung cancers (NSCLCs) have generated particular interest. Nuclear factor erythroid 2–related factor (NRF2) is a transcription factor that promotes expression of antioxidant genes, while Kelch-like ECH-associated protein 1 (KEAP1) negatively regulates NRF2 by targeting it for proteasomal degradation (4). Loss-of-function KEAP1 mutations and gain-of-function NRF2 mutations occur in 20 to 25% of human NSCLCs, and both result in increased NRF2-dependent gene expression (4–6). KEAP1/NRF2 mutations frequently co-occur with mutations in the oncogene Kirsten rat sarcoma virus (KRAS) and the tumor suppressor serine/threonine kinase 11 (STK11; which encodes the kinase LKB1) (4–7).

Tumors with mutations in these genes represent a therapeutically intractable group of NSCLCs known as the KLK subtype (KRAS, LKB1, KEAP1/NRF2). Such tumors tend to resist KRAS-targeting therapies and checkpoint inhibitors, and patients with these subtypes have a poor prognosis (4). At this time, no therapeutic options that selectively target this aggressive NSCLC subtype are available.

In cancer models, the KLK genotype confers glutamine dependence and sensitivity to GLS inhibition (4, 7). In patients with KEAP1/NRF2-mutated NSCLC, however, clinical trials combining the GLS inhibitor CB-839 with standard of care chemotherapy and immunotherapy demonstrated little-to-no benefit beyond chemotherapy and immunotherapy alone. Although these and other clinical trials with suboptimal outcomes have diminished enthusiasm for GLS blockade, they leave open the question of whether targeting multiple glutamine-dependent pathways simultaneously might provide benefit. Early clinical efforts using the broad-acting glutamine antagonist 6-diazo-5-oxo-l-norleucine (DON) revealed efficacy in some settings, but use of the drug was limited by a narrow therapeutic index (8). More recently, DON-like pro-drugs developed by Barbara Slusher and colleagues at Johns Hopkins University limit systemic exposure to DON. These compounds are activated by enzymes enriched within tumors, resulting in a large relative increase in DON concentration within tumors compared to the rest of the body (8). In mice, these compounds not only suppress cancer cell growth but enhance antitumor immunity by stimulating cytotoxic T cells, all without the toxicities that plagued earlier efforts with DON (9).

In this issue of Science Advances, Pillai et al. (10) tested DRP-104, a DON pro-drug developed for clinical trials, in mouse models of Keap1-mutant NSCLC. Keap1-mutant tumors were thought to be good candidates for DRP-104 because NRF2 drives expression of esterases that cleave the pro-drug to release DON. In tumors with mutant Kras with or without mutations in Stk11, loss of Keap1 conferred sensitivity to DRP-104 as a single agent in immunocompetent mice. Depleting cytotoxic T cells nearly eliminated DRP-104’s efficacy in these experiments, while combining DRP-104 with checkpoint inhibitors extended survival compared to DRP-104 alone.

These findings indicate that DRP-104 exerts its therapeutic effects in large part by inducing immunity against Keap1-mutant NSCLC. Additional molecular characterization of the tumors showed that DRP-104 markedly remodels the immune microenvironment, skewing it toward enhanced effector T cell function and away from T cell exhaustion. The data provide a preclinical rationale for advancing a combination of DRP-104 and immune checkpoint blockade into the clinic for KEAP1/NRF2-mutated NSCLC.

But which of DRP-104’s effects on glutamine metabolism are important for the therapeutic response (Fig. 1)? Among the many pathways blocked by DON, Pillai et al. show in vivo and in culture that the drug depletes nucleotides, with nearly complete inhibition of de novo purine synthesis in culture. Strikingly, proliferation of Keap1-mutant cells treated with DRP-104 was completely restored by hypoxanthine, a nucleobase that can be salvaged to produce purines independently of the de novo synthesis pathway. Therefore, although the drug can block many aspects of glutamine metabolism simultaneously, depletion of purines is most responsible for the DRP-104’s effect on cancer cell growth, at least in culture and in cells with Keap1 mutations.

The drug had minimal effects on glutamate synthesis or abundance, despite the fact that glutamine-dependent reactions in purine synthesis produce glutamate. This suggests that glutamate primarily arises from other reactions that are not inhibited by DRP-104. It is also notable that metabolites related to the TCA cycle failed to rescue growth in DRP-104–treated cells, indicating that DRP-104’s efficacy was unrelated to GLS blockade. Altogether, the findings imply that glutamine’s ability to provide nitrogen for biosynthetic pathways outweighs its role in central carbon metabolism in Keap1-mutant tumors. This might help explain the lack of efficacy of GLS inhibitors in tumors predicted to be glutamine-avid.

Many questions remain about how DON pro-drugs work, and how their clinical development can be accelerated. A key unresolved issue is the mechanistic link between glutamine blockade and T cell activation. Although purine depletion explains DRP-104’s effects on cancer cell growth, it is unknown if or how this pathway relates to T cell function. It may be that glutamine antagonism in cancer cells indirectly induces other changes in the tumor microenvironment, for example, improved pH and enhanced nutrient availability, that culminate in cytotoxic T cell stimulation (9). Better strategies to predict which tumors will respond to metabolic inhibitors would aid clinical development. Direct assessment of glutamine metabolism in patients could be helpful in this regard. Glutamine metabolism tracers developed for positron emission tomography, including ¹⁸F-(2S, 4R) 4-fluroglutamine (¹⁸F-Gln), (4S)-4-(3-[¹⁸F] fluoropropyl) l-glutamate (¹⁸F-FSPG), and ¹¹C-glutamine, could help identify tumors that take up glutamine in vivo and possibly help track therapeutic responses to drugs that block glutamine metabolism. Combining a better understanding of tumor metabolism with advanced tools to monitor glutamine utilization in vivo should help define the patients most likely to benefit from glutamine antagonists like DRP-104. Ultimately, this should increase the likelihood of successful clinical trials that improve outcomes in patients with cancer.