Battles against aberrant KEAP1-NRF2 signaling in lung cancer: intertwined metabolic and immune networks

Abstract

The Kelch-like ECH-associated protein 1/nuclear factor erythroid-derived 2-like 2 (KEAP1/NRF2) pathway is well recognized as a key regulator of redox homeostasis, protecting cells from oxidative stress and xenobiotics under physiological circumstances. Cancer cells often hijack this pathway during initiation and progression, with aberrant KEAP1-NRF2 activity predominantly observed in non-small cell lung cancer (NSCLC), suggesting that cell/tissue-of-origin is likely to influence the genetic selection during malignant transformation. Hyperactivation of NRF2 confers a multi-faceted role, and recently, increasing evidence shows that a close interplay between metabolic reprogramming and tumor immunity remodelling contributes to its aggressiveness, treatment resistance (radio-/chemo-/immune-therapy) and susceptibility to metastases. Here, we discuss in detail the special metabolic and immune fitness enabled by KEAP1-NRF2 aberration in NSCLC. Furthermore, we summarize the similarities and differences in the dysregulated KEAP1-NRF2 pathway between two major histo-subtypes of NSCLC, provide mechanistic insights on the poor response to immunotherapy despite their high immunogenicity, and outline evolving strategies to treat this recalcitrant cancer subset. Finally, we integrate bioinformatic analysis of publicly available datasets to illustrate the new partners/effectors in NRF2-addicted cancer cells, which may provide new insights into context-directed treatment.

Keywords: KEAP1-NRF2 signaling; bioinformatics; metabolic reprogramming; non-small cell lung cancer; therapeutic vulnerabilities; tumor immune microenvironment.

Figure 1

Similarities and differences of KEAP1-NRF2 aberration in human non-small cell lung cancer. A, Left panel: schematic model showing the common location of lung squamous cell carcinoma (LUSC; central; associated with a smoking history) and adenocarcinoma (LUAD; peripheral). Middle panel: key components of KEAP1-NRF2 pathway. Right panel: genetic alterations of KEAP1, NFE2L2, CUL3 alone or in combination across The Cancer Genome Atlas (TCGA) LUSC and LUAD, respectively. Data were downloaded from the cBioPortal (https://www.cbioportal.org/). B, Oncoprint of genetic alterations of KEAP1, NFE2L2, CUL3 across TCGA LUSC and LUAD. The corresponding patient sex, tumor mutational burden (TMB) and three curated signatures calculating the hypoxia score were also shown, which displayed a different distribution between LUSC and LUAD. Data were downloaded from the cBioPortal (https://www.cbioportal.org/). C, Comparing the transcriptional molecular features between mutated (Mut) (KEAP1, NFE2L2, CUL3) and wildtype (WT) in patient samples of LUSC and LUAD. Nonnegative matrix factorization (NMF) clustering was performed to identify the molecular subtypes. Data were downloaded from the cBioPortal (https://www.cbioportal.org/).

Figure 2

Crosstalk between tumor metabolism and immunity in lung cancer with KEAP1-NRF2 aberration. KEAP1-NRF2 aberration plays a pleiotropic role in tumor metabolism, which has a close interplay with tumor immunity

Figure 3

Different hypoxia index between lung tumor with and without KEAP1-NRF2 aberration. A, Three different curated signatures were used to compare the hypoxia score between mutated (Mut) (KEAP1, NFE2L2, CUL3) and wildtype (WT) in patient samples of TCGA LUSC (lung squamous cell carcinoma) and LUAD (lung adenocarcinoma). Data were downloaded from the cBioPortal (https://www.cbioportal.org/). Of note, the “subtle and marginal significance” between the comparative subgroups might be due to: 1) the presence of high NRF2 activity in the KEAP1-WT lung tumors, as NRF2 activity can be regulated at multiple levels including genetic, transcriptional, and post-transcriptional. As such, NRF2 score , might better reflect the activity of NRF2 signaling. B, In LUAD, mutated KEAP1 is commonly co-occurring with KRAS or STK11 mutation. Different algorithms-based hypoxia scores were used to compare the difference between KRAS (upper panel) or STK11 (lower panel) wildtype (WT) and mutant (Mut) tumors under the context of KEAP1 mutation in LUAD. P-value was calculated by Wilcoxon Test.

Figure 4

scRNA-seq analysis of NRF2 expression across immune cell types within human NSCLC. A, UMAP plots showing the unsupervised clustering of tumor and immune cells from one non-small cell lung cancer (NSCLC) patient cohort. The middle and right panels showing the expression of NFE2L2 and G6PD (one key target gene of NFE2L2) across single cells. Single-cell RNA-seq (scRNA-seq) data were downloaded from the TISCH portal (Tumor Immune Single-cell Hub; http://tisch.comp-genomics.org/home/). B, Heatmap plots showing the distribution of NFE2L2 and G6PD expression across different cell types in NSCLC.

Figure 5

The association of tumor immunogenicity with the mutant status of the KEAP1-NRF2 pathway in lung cancer. Barplots showing the difference in tumor immunogenicity between KEAP1-mutant (mut) and -wildtype (WT) lung squamous cell carcinoma (LUSC) and adenocarcinoma (LUAD) samples of the TCGA cohort. Data were downloaded from the CAMOIP (Comprehensive Analysis on Multi-Omics of Immunotherapy in Pan-cancer) portal (https://www.camoip.net/).

Figure 6

Whole-genome screens reveal potential targeted dependency in KEAP1-mutant cancer cells. A, B, Top 10 essential genes whose mRNA level (upper panel) or genetic mutation (lower panel) is correlated with the dependence score of NFE2L2, based on CRISPR (A)- and RNAi (B)-based whole-genome screen (https://depmap.org/portal/). In the left panels of A and B, the random forest machine learning strategy was used to identify the most important molecular features related to NFE2L2 dependence. Right panels showing the Pearson correlation of the indicated genes with NFE2L2 dependence score. The negative correlation indicates that genes whose higher expression or mutation signifies higher sensitivity to knockout or knockdown of NFE2L2.

Figure 7

Therapeutic strategies for KEAP1-mutant cancer cells. Direct and indirect targets for the aberrant KEAP1-NRF2 pathway.