Small Molecule Inhibitor of NRF2 Selectively Intervenes Therapeutic Resistance in KEAP1-Deficient NSCLC Tumors

Abstract

Loss of function mutations in Kelch-like ECH Associated Protein 1 (KEAP1), or gain-of-function mutations in nuclear factor erythroid 2-related factor 2 (NRF2), are common in non-small cell lung cancer (NSCLC) and associated with therapeutic resistance. To discover novel NRF2 inhibitors for targeted therapy, we conducted a quantitative high-throughput screen using a diverse set of ∼400 000 small molecules (Molecular Libraries Small Molecule Repository Library, MLSMR) at the National Center for Advancing Translational Sciences. We identified ML385 as a probe molecule that binds to NRF2 and inhibits its downstream target gene expression. Specifically, ML385 binds to Neh1, the Cap 'N' Collar Basic Leucine Zipper (CNC-bZIP) domain of NRF2, and interferes with the binding of the V-Maf Avian Musculoaponeurotic Fibrosarcoma Oncogene Homologue G (MAFG)-NRF2 protein complex to regulatory DNA binding sequences. In clonogenic assays, when used in combination with platinum-based drugs, doxorubicin or taxol, ML385 substantially enhances cytotoxicity in NSCLC cells, as compared to single agents. ML385 shows specificity and selectivity for NSCLC cells with KEAP1 mutation, leading to gain of NRF2 function. In preclinical models of NSCLC with gain of NRF2 function, ML385 in combination with carboplatin showed significant antitumor activity. We demonstrate the discovery and validation of ML385 as a novel and specific NRF2 inhibitor and conclude that targeting NRF2 may represent a promising strategy for the treatment of advanced NSCLC.

Figure 1. Summary of the NRF2 qHTS screening results

(a) Pie chart displaying the breakdown of the screening hit classes; high-quality (HQ) inhibitors are compounds in curve class −1.1, −1.2, −2.1, and −2.2 with efficacy higher than 50%; low-quality (LQ) inhibitors are compounds in curve class −3 with single point activity or those with shallow or poor-fitting curves; toxic compounds are those showing prominent cell killing effects from the cytotoxicity readout; inactives are compounds with class 4 curves. Histogram displaying the potency distribution of the HQ-inhibitors. (b) Assay flowchart for the validation of small molecule inhibitors of NRF2. (c) Heat map showing the activity profile of 1712 qHTS hits in 3 ARE NRF2 assays (A549, H1437, and H838 cells), purified firefly luciferase biochemical counter assay and HEK293-CMV counter assay. Compound IDs are given at right and the assay types are listed at the bottom of the heat map. Hierarchical clustering analysis was done using Spotfire DecisionSite 8.2. (d) Concentration-response curves of ML385 in three ARE NRF2-luc assays (A549, H1437, and H838 cells), HEK293-CMV counter assay and glucocorticoid receptor (GR)-responsive beta-lactamase reporter cell-based assay.

Figure 2. ML385 inhibits DNA binding activity of NRF2-MAFG complex

(a) Chemical structure of qHTS hit 1 and subsequent active NRF2 inhibitor ML385 and inactive analog 3. (b) ML385 inhibits the binding of NRF2-MAFG protein complex to fluorescein-labeled ARE DNA. Fluorescence intensity was measured to get anisotropy value and IC₅₀ was calculated by fitting of sigmoid curve (R²>0.97).

Figure 3. ML385 directly interacts with purified NRF2 protein

(a) Chemical structure of active Biotin conjugated ML385 (AB-ML385) and inactive biotin conjugated analog (IB-ML385). (b) AB-ML385 (10 μM) attenuates NRF2 and GCLc mRNA expression, while IB-ML385 does not. ML385 was used at a concentration of 5 μM. Error bars represent ±S.D. ‘*’P<0.05 relative to vehicle. (c) Domain architecture of NRF2 protein. (d) AB-ML385 binds to purified histidine-tagged full length NRF2 protein in Ni⁺ pull-down assay and is competed away by ML385. (e) AB-ML385 selectively binds to Neh1 domain of NRF2. NRF2 full-length NRF2 protein; Neh1, Neh1 domain of NRF2, ΔNeh1, full-length NRF2 protein lacking Neh1 domain.

Figure 4. ML385 inhibits NRF2 signaling in lung cancer cells

(a) Dose-dependent inhibition of NRF2-mediated transcription by ML385. Error bars represent ±S.D. (b) Time-dependent reduction in NRF2 and its target genes in A549 cells after treatment with ML385 (5μM). Error bars represent ±S.D. ‘*’P<0.05 relative to vehicle. (c) Heat map showing expression of NRF2-dependent antioxidant and drug detoxification genes in A549 cells treated with ML385 for 48 h or 72 h. Inactive analog 3 (5 μM) was used as a control. Gene expression was measured by real time RT-PCR. (d) Immunoblot showing relative levels of NRF2 protein at 24 h and 48 h post-treatment with ML385. (e) Densitometric quantification of NRF2 immunoblot data. (f–h) Treatment with ML385 attenuates antioxidant enzyme activity resulting in lower NQO1 enzyme activity, total antioxidant capacity, and reduced glutathione levels in A549 cells. Error bars represent ±S.D. ‘*’P<0.05 relative to vehicle.

Figure 5. ML385 is selectively toxic to cells with KEAP1 mutations and potentiates the toxicity of chemotherapy drugs in NSCLC cells with KEAP1 mutations

(a) H460, a NSCLC line with a point mutation in KEAP1, is more sensitive to ML385 than H460-KEAP1 Knock-in H460 cells expressing WT KEAP1. ‘*’P<0.05 relative to H460-KEAP1 Knock-in cells. (b) Treatment with ML385 selectively inhibits the colony forming ability of lung cancer cells but has no effect on the growth of non-tumorigenic BEAS2B cells. (c–d) H460 and A549 cells were treated with different concentrations of paclitaxel, doxorubicin, or carboplatin alone or in combination with ML385 for 72 h. At the end of treatment, regular growth medium was added and cells were further incubated for 8–10 days and stained with crystal violet. (e) H460 cells treated with ML385 in combination with chemotherapy drug showed increased caspase 3/7 activity, a marker of apoptosis. Cells treated with chemotherapy drug alone or ML385 in combination with chemotherapy drug were incubated with luminogenic caspase substrate and change in luminescence was measured. Caspase activity was normalized to the number of viable cells using CellTiter-Blue assay. Error bars represent ±S.D. ‘*’P<0.05 relative to vehicle or ML385; ‘**’ P<0.05 relative to chemotherapy drug alone.

Figure 6. Therapeutic efficacy of ML385 as a single agent and in combination with carboplatin in subcutaneous lung tumor xenografts

(a) ML385 shows anti-tumor activity as a single agent in sensitized A549 xenograft tumors to carboplatin therapy. Values represent tumor volume ± S.E.M. for all groups. n = 7 mice/group; ‘*’P<0.05 relative to vehicle; ‘#’ P<0.05 relative to carboplatin. (b) Treatment with ML385 or ML385 in combination with carboplatin significantly reduced tumor weight as compared to vehicle group. Efficacy of ML385 alone was comparable to carboplatin. ‘*’P<0.05 relative to vehicle; ‘#’ P<0.05 relative to carboplatin. (c-d) ML385 sensitized H460 lung tumors to carboplatin treatment. Groups of H460 tumors treated with ML385 or ML385 in combination with carboplatin showed a significant reduction in tumor volume and weight as compared to the vehicle group. Efficacy of ML385 alone was comparable to carboplatin. n = 5–7 mice/group; ‘*’P<0.05 relative to vehicle; ‘$’ P<0.05 relative to ML385. (e) The proliferative index based on Ki-67 immuno-reactivity in A549 subcutaneous tumors. (f) Treatment with ML385 attenuated the expression of NRF2-dependent genes in H460 tumors. ‘*’P<0.05 relative to vehicle or carboplatin (g) Immunoblot showing reduction in NRF2 protein in H460 tumors treated with ML385. (h) Bar graph showing platinum levels in A549 and H460 tumors treated with carboplatin alone or ML385 in combination with carboplatin. ‘*’P<0.05, relative to carboplatin alone.

Figure 7. ML385 shows significant anti-tumor efficacy in combination with carboplatin in an orthotopic lung tumor model

(a) Volume rendered 3-D lung images demonstrate the effectiveness of combination therapy in A549 tumors. Representative images show the available lung volume that is not taken over by the lung tumor pre -and post-treatment with vehicle, ML385, carboplatin or ML385 in combination with carboplatin. (n=2–5 mice/group). ‘Tx’-treatment. (b) Bar graph showing average tumor free lung volume in different groups at 3 wks post-treatment suggesting that combination therapy is more effective in reducing tumor growth. Mice in the vehicle-treated group died after 3 wks. For each group, pretreatment available lung volume was defined as 100% and was compared with post-treatment lung volumes. Values represent lung volume ± S.E.M. ‘*’P<0.05 relative to vehicle; ‘#’ P<0.05 relative to carboplatin. (c) Combination therapy shows a significant growth reduction of H460 tumors. Representative images show the available lung volume that is not taken over by the lung tumor pre- and post-treatment with vehicle, ML385, carboplatin or ML385 in combination with carboplatin. Mice in the vehicle-treated group did not survive for 2 wks post treatment. (n=2–4 mice/group). (d) Bar graph showing H460 tumor-free lung volume in different groups at 2 wks post-treatment suggesting that combination therapy is more effective in reducing tumor growth compared to carboplatin monotherapy. Values represent lung volume ± S.E.M ‘*’P<0.05, relative to carboplatin.